[Journal] Current Orthopaedics. Volume 20

Citation preview

Aims and Scope Current Orthopaedics presents a unique collection of international review articles summarizing the current state of knowledge and research in orthopaedics. Each issue focuses on a specific topic, discussed in depth in a mini-symposium; other articles cover the areas of basic science, medicine, children/adults, trauma, imaging and historical review. There is also an annotation, self-assessment questions and an exam section. In this way, the entire postgraduate syllabus will be covered in a 4-year cycle. The Journal is cited in: Cochrane Center, EMBASE/ Excerpta Medica, Infomed, Reference Update and UMI Microfilms.

Editor Professor R. A. Dickson MA, ChM, FRCS, DSc St James’s University Hospital Trust, Leeds, UK

Editorial Committee President of BOTA, M. A. Farquharson-Roberts (Gosport, UK), I. Leslie (Bristol, UK), D. Limb (Leeds, UK), M. Macnicol (Edinburgh, UK), I. McDermott (Ruislip, UK), J. Rankine (Leeds, UK)

Editorial Advisory Board L. de Almeida (Portugal) G. P. Songcharoen (Thailand) R. W. Bucholz (USA) J. W. Frymoyer (USA) R. W. Gaines (USA) S. L. Weinstein (USA) M. Bumbasirevic (former Yugoslavia)

A. K. Mukherjee (India) A. Kusakabe (Japan) A. Uchida (Japan) M.-S. Moon (Korea) R. Castelein (The Netherlands) R. K. Marti (The Netherlands) G. Hooper (New Zealand) A. Thurston (New Zealand) E. G. Pasion (Philippines)

D. C. Davidson (Australia) J. Harris (Australia) S. Nade (Australia) G. R. Velloso (Brazil) J. H. Wedge (Canada) S. Santavirta (Finland) P. N. Soucacos (Greece) M. Torrens (Greece) J. C. Y. Leong (Hong Kong)

Available online at www.sciencedirect.com

Amsterdam

K

Boston

K

Jena

K

London

K

New York

K

Oxford

K

Paris

K

Philadelphia

K

San Diego

K

St Louis

Volume 20 (2006) contents Mini-Symposium: Biomechanics for the FRCS Orth Exam (i) An introduction to basic mechanics R.K. Wilcox . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (ii) Deformation of materials S.M. Green . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (iii) Material properties of biological tissues related to joint surgery U. Hansen, S. Masouros and A.A. Amis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (iv) Basic biomechanics of human joints: Hips, knees and the spine T.D. Stewart and R.M. Hall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (v) Biotribology Z.M. Jin, M. Stone, E. Ingham and J. Fisher. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 9 16 23 32

Neuromuscular Conditions Poliomyelitis: Orthopaedic management A.A. Faraj. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

41

Radiology Functional imaging in orthopaedic infectionsFUpdate on immunoscintigraphy K.P. Iyengar, C.N. Ramesh and S. Vinjamuri . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

47

Basic Science The thermal properties of bone and the effects of surgical intervention S. Karmani. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

52

Trauma Ilizarov and trauma reconstruction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

59

CME Section CME Questions based on the Mini Symposium on Biomechanics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Answers to CME questions in Vol. 19, issue 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

72 75

Mini-Symposium: Soft Tissue Knee Problems (i) The crucial ligaments S. Bollen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (ii) Meniscal tears I.D. McDermott . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (iii) The dislocated knee A. Robertson and R.W. Nutton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (iv) Patellofemoral dysfunctionFExtensor mechanism malalignment S. Donell. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (v) Osteotomy in the management of knee osteoarthritis and of ligamentous instability A. Williams and N. Devic. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

77 85 95 103 112

Knee Alternative surgeries for patello-femoral disorders J. Sanchez-Ballester, P. McGraw, P.G. Turner and D.S. Johnson . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

121

Syndrome Morquio syndrome B.J.A. Lankester, M. Whitehouse and M.F. Gargan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

128

Spine Development in the management of tuberculosis of the spine M.-S. Moon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

132

VI

CURRENT ORTHOPAEDICS

Upper Limb Chronic wrist pain: Diagnosis and management S. Ankarath . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

141

CME Section CME Questions based on ‘‘Meniscal Tears’’. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Answers to CME questions in Vol. 19, issue 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Erratum to ‘CME section’. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

152 155 156

Mini-Symposium: Revision Hip Arthroplasty (i) International epidemiology of revision THR M. Skutek, R.B. Bourne and S.J. MacDonald . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

157

(ii) Current techniques and new developments in acetabular revision surgery A.W.R. Burns and R.W. McCalden. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

162

(iii) Cementless femoral revision: The role of monoblock versus modular stems S.M. Sporer and W.G. Paprosky. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

171

(iv) Periprosthetic fractures of the hip S. Patil, B.A. Masri and C.P. Duncan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

179

(v) Post-operative infection in total hip arthroplasty L. Mokete and D.D. Naudie . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

192

(vi) Economics of revision total hip arthroplasty A.W.R. Burns and R.B. Bourne. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

203

Wrist What’s new in the treatment of distal radius fractures? F. Lam, N. Jaysekera, S. Karmani and J.B. Jupiter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

208

Syndromes Down syndrome R. Amirfeyz, D. Aspros and M. Gargan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

212

Joint Replacement Obesity and total knee and hip replacement A.K. Amin, J.D. Sales and I.J. Brenkel. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

216

Adult Trauma Management of proximal humeral fractures D.J.C. Burton and A.T. Watters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

222

CME Section CME Questions based on the Mini Symposium on Revision Hip Arthroplasty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

234

Answers to CME questions in Vol. 20, issue 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

237

Book Reviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

238

Mini-Symposium: Children’s Orthopaedic Surgery (i) Indications for internal fixation of fractures in children I.H. Annan and M. Moran. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

241

(ii) Paediatric epiphyseal fractures around the knee M. Moran and M.F. Macnicol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

256

(iii) Congenital hand anomalies A.C. Watts and G. Hooper. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

266

(iv) Cervical spine problems in children J. Ross and L. Myles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

274

VOLUME 20 (2006) CONTENTS

VII

(v) The hip in cerebral palsy A.W. Murray and J.E. Robb. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

286

WRIST Carpal tunnel syndrome in men A.C. Watts and J. McEachan. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

294

TUMOURS Management of metastatic disease of the appendicular skeleton R.U. Ashford, S. Pendlebury and P.D. Stalley . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

299

CME SECTION CME questions based on the ‘‘Management of metastatic disease of the appendicular skeleton’’ . . . . . . . . . . . . . . . . . . . . . . . . . . . .

316

Answers to CME questions in vol. 20, issue 2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

319

BOOK REVIEWS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

320

Editorial Bullet and blast injuries R.A. Dickson . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

321

Mini-Symposium: Bullet and Blast Injuries (i) An overview of the pathophysiology of gunshot and blast injury with resuscitation guidelines S.A. Stapley and L.B. Cannon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (ii) Initial medical and surgical management P.J. Parker . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (iii) Military limb injuries/ballistic fractures D. Griffiths and J. Clasper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (iv) Anti-personnel mine injuries K. Trimble, S. Adams and M. Adams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

322 333 346 354

Children Fractures of the femoral neck in children C. Holton, P. Foster and P. Templeton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

361

The British Orthopaedic Trainees AssociationFpast, present and future M.R. Edwards and M. Freudman. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

367

Upper Limb Treatment of the painful biceps tendonFTenotomy or tenodesis? F. Lam and D. Mok . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

370

Knee Periprosthetic fractures above a total knee arthroplastyFA review of best practice G. Walsh, S. Ankarath and P.V. Giannoudis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

376

Radiology quiz C. Hammond and P. Robinson . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

386

CME Section CME questions based on the Mini-Symposium on Bullet and Blast Injuries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Answers to CME questions in vol. 20, issue 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

393 396

Mini-Symposium: Management of Fractures around the Knee Joint (i) Comminuted patellar fractures I. Mehling, A. Mehling and P.M. Rommens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (ii) The ‘‘floating knee’’ in adults and children B. Chalidis, S.S. Metha, E. Tsiridis and P.V. Giannoudis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (iii) Minimal invasive techniques in the management of tibial plateau fractures R. Venkatesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

397 405 411

VIII

CURRENT ORTHOPAEDICS

Syndromes Marfan syndrome A.R.T. McBride and M. Gargan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

418

Bone Physiology Bone morphogenetic proteins in orthopaedic surgery A. Cheung and A.M. Phillips . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

424

Spine Spinal muscular atrophy: Classification, aetiology, and treatment of spinal deformity in children and adolescents A.I. Tsirikos and A.D.L. Baker . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

430

Shoulder Modern management of calcifying tendinitis of the shoulder F. Lam, D. Bhatia, K. van Rooyen and J.F. de Beer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

446

Quiz Radiology quiz E. Hoey and P. Robinson . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

453

Children Fractures in the child’s hand A. Smit and G. Hooper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

461

CME Section CME Questions based on the Mini-Symposium on Management of Fractures around the Knee Joint . . . . . . . . . . . . . . . . . . . . . . . Answers to CME questions in Vol. 20, issue 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

467 470

Book Reviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

471

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

I

Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

III

Contents of Volume 20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

V

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 1–8

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: BIOMECHANICS FOR THE FRCS ORTH EXAM

(i) An introduction to basic mechanics Ruth K. Wilcox School of Mechanical Engineering, University of Leeds, Leeds LS2 9JT, UK

KEYWORDS Mechanics; Biomechanics; Statics; Dynamics

Summary The principles of solid mechanics have many important applications in orthopaedics. The aim of this paper is to cover the basic concepts of statics and dynamics, including both kinetics and kinematics. The equations most commonly used in biomechanical analysis are introduced and examples are given of their use in orthopaedic applications. & 2005 Elsevier Ltd. All rights reserved.

Introduction Mechanics is the name given to the branch of engineering science that deals with forces acting upon objects. It has many important applications in modern orthopaedics. Much of our modern understanding of mechanics is based on the work of Isaac Newton (1642–1729) and many of the principles discussed in this paper are derived from the three laws of motion published in his most famous work, Principia, in 1687. In more recent times, the concepts of mechanics have been applied to the study of the human body, its motion and the forces acting upon it. This subject is often referred to as ‘biomechanics’. Our increased understanding in this area has led to significant improvements in the design of many medical devices and interventions. In orthopaedics, for example, our ability to determine the method of force transmission between an implant and the surrounding bone has led to alterations in prosthetic design to prevent bone loss due to stress shielding. Our knowledge of the forces transmitted through the joints under different motions has also enabled the development of apparatus to simulate Tel.: +44 113 343 2214; fax: +44 113 242 4611.

E-mail address: [email protected]. 0268-0890/$ - see front matter & 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2005.09.003

these motions in the laboratory, which in turn has led to improvements in the design of joint replacements to enhance their wear characteristics. Mechanics is often divided into the study of solids and the study of fluids, although many of the same principles can equally be applied to both. In the clinical field, fluid mechanics is often associated with the study of the cardiovascular and respiratory systems, although the subject also has important applications in orthopaedics, particularly in the study of fluid-film lubrication at the synovial joints. This paper, however, will focus on solid mechanics, which in turn is divided into the fields of statics and dynamics. Statics is the study of equilibrium forces acting on objects in a state of rest; the principles of statics can be used in orthopaedics to estimate loads on the body and forces transmitted across joints when the body is stationary. Dynamics is concerned with the motion of objects and the forces acting upon them; the principles of dynamics can be used in orthopaedics to analyse the forces transmitted through the joints and the corresponding motions when the body is undertaking activity. The aim of this paper is to describe the basic principles of statics and dynamics in sufficient detail to cover many of the common applications in orthopaedics. Throughout the text, quantities are defined in terms of their SI units; the basic

ARTICLE IN PRESS 2

R.K. Wilcox

quantities of length, mass and time are measured in this system using the metre (m), the kilogramme (kg) and the second (s), respectively.

X a F

Statics (a)

Scalars and vectors W The physical quantities used in this article are of two types: scalars and vectors. A scalar quantity can be fully described by only its magnitude, and examples include mass, temperature and density. A vector quantity has both magnitude and direction and both are needed for the quantity to be fully described. Examples include velocity, force and moment. A vector is depicted in this text in bold type (e.g. F), and its magnitude in normal type (e.g. F).

b cosθ cos θ

X

b (b)

d

Force and moment One of the most important vector quantities in mechanics is force (F). It can be defined as the action between two bodies, such as a ‘push’ or ‘pull’. The SI unit of force is the Newton (N) which is equivalent to 1 kg m s
2. Since force is a vector, both the magnitude of the force and the direction in which it acts are important. The moment of a force (M) is defined as the product of the force and the perpendicular distance about which it rotates. Moment is also a vector quantity; the direction of a moment is described by the ‘right-hand rule’: if the fingers of the hand are curled to represent the direction of rotation then the thumb, if held perpendicular to the fingers, represents the direction of the moment (Fig. 1). The unit of moment is the Newton-metre (N m).

Scalar and vector addition Often in mechanics, we wish to add physical quantities together. With scalars, this is done simply by summing the magnitudes, but with vectors, the direction must also be taken into account. With two or more vectors in two dimensions, this can be done most easily using the parallelogram rule. Here, each vector quantity is depicted by an arrow in the appropriate direction, the length of which is proportional to the magnitude of the vector. The arrows form two sides of a parallelogram and the sum or resultant vector is the diagonal of the parallelogram (Fig. 2).

ϕ mg (c)

Figure 1 (a) Moment of force F about point X: magnitude ¼ Fa, direction ¼ out of page. (b) Moment of force W about point X: magnitude ¼ Fb cos y, direction ¼ into page. (c) Force of mass due to gravity ¼ mg, perpendicular distance from mass to shoulder ¼ d sin j, magnitude of moment on shoulder joint due to mass ¼ mgd sin j.

the recombination of the components. In three dimensions this enables mathematical methods to be used without the need to draw out complex parallelogram diagrams.

Equilibrium Newton’s first law states that an object will remain in a state of rest or of uniform motion (that is, travelling at a constant velocity) unless acted upon by a force. Therefore if an object is in a state of rest, then the sum of all of the forces and all of the moments acting upon it must be zero. It is also true that the sum of the components of all of the forces or moments in any given direction must be zero.

Components of a vector

Action and reaction

The same rule can also be used in reverse to determine the components of a vector in different directions. It is often useful to split a vector into two perpendicular directions (or three in three dimensions) as shown in Fig. 3a. The resolution of forces is particularly useful in biomechanics where there are often many loads being applied to a body in many different directions. Here, the forces can all be resolved into their component parts along the same perpendicular directions. In any of these directions, the components of all the forces can then be summed like scalars (Fig. 3b). The resultant force can be calculated by

Newton’s third law states that for every action there is an equal and opposite reaction. If a person stands on one leg, then all of the body weight of the person is transmitted through the foot to the ground. Since the ground does not accelerate away, it must exert an equal and opposite force on the foot. This force is often referred to as the ‘ground reaction force’.

Free body diagrams These laws are used in biomechanics to make estimates of the muscle and joint reaction forces acting in the body under

ARTICLE IN PRESS An introduction to basic mechanics

3 3. Sum of components of forces in x direction ¼ 0:

FR F1

F AM sin f
F JR sin y ¼ 0 ) F JR sin y ¼ F AM sin f.

F2

ð3Þ

4. From (2) and (3)

(a)

F JR sin y F AM sin f ¼ tan y ¼ . F JR cos y W þ F AM cos f

ϕ

(4)

θ FQT

FPT

Practical methods in biomechanics

FPT (b)

FQT FR

Figure 2 (a) The parallelogram rule is used to find the resultant of two forces F1 and F2. To differentiate forces from their resultants on diagrams, the resultant force is shown with a double arrowhead. (b) The quadriceps (FQT) and patella (FPT) tendon forces acting on the patella. The resultant force acts to pull the patella against the femur.

certain loading conditions. A ‘free body diagram’ is often drawn by isolating a region of interest and determining all of the external forces that act upon that region. An example of a free body diagram is shown in Fig. 4. Here, the region of interest is the hip joint. The forces acting on the joint from above are due to the body weight of the upper torso and the abductor muscle force, which is assumed to act along a single line of action. For this particular problem, the method by which the ground reaction force is transmitted through the leg is not of interest, but we know that this must cause a force to be transmitted into the region of interest via the femoral head. Therefore a third force is added to the diagram to represent this reaction force across the joint. Assuming that the body is known to be in equilibrium, then the sum of the forces acting upon it in any direction must be equal to zero. Likewise, the sum of the moments about any point must be equal to zero. This enables a series of equations to be written and solved to find the unknown abductor and joint reaction forces.

Although any number of equations could be written to state the equilibrium conditions, they would not all be independent of each other. In two dimensions, three independent equations can be written: two for the sums of the forces and one for the sum of the moments. In three dimensions, there are six independent equations. Mathematically, the number of unknowns that can be determined cannot be greater than the number of independent equations. In biomechanics, this means that simplifications often have to be made to limit the number of unknowns. This may involve grouping muscles together or ignoring smaller muscle groups completely. Data may also be obtained from EMG measurements to generate additional equations and our knowledge of how muscles and tendons behave may add further constraints. Computational methods using optimisation routines are also now used to estimate the most likely way muscles combine to generate a particular force.

Dynamics Introduction Many textbooks of mechanics introduce the principles of dynamics by considering a single particle, i.e., an object of negligible dimensions. This is usually followed by consideration of a rigid body, an object whose changes in shape are negligible compared to the size of the motions it undergoes. However, the human body is neither a particle nor a rigid body! Nonetheless, we can assume in many cases that the bony structure is dominant in the body’s behaviour and can be assumed to behave as a system of rigid bodies moving relative to each other. The subject of dynamics can be divided into kinematics and kinetics. Kinematics is the study of the relationship between the position, velocity and acceleration of objects while kinetics is concerned with both the motion of objects and the forces acting upon them.

1. Sum of moments about point P ¼ 0: F AM d 2
Wd 1 ¼ 0 Wd 1 ) F AM ¼ . d2

Kinematics ð1Þ

2. Sum of components of forces in y direction ¼ 0:
W
F AM cos f þ F JR cos y ¼ 0 ) F JR cos y ¼ W þ F AM cos f.

ð2Þ

Translation and rotation The plane motion of a rigid body may be divided into translation and rotation. Translation occurs when every line in the body remains parallel to its original position whilst pure rotation occurs when every point in the body moves in a circular path around the centre of rotation. Most motions of the human body are a mixture of rotation and translation.

ARTICLE IN PRESS 4

R.K. Wilcox F Fy = F sin θ

y x

θ

Fx = F cos θ

(a)

Resolving into

x y

y directions:

x FPT cos ϕ

ϕ

FPT sin ϕ

=

FPT =

- FPT cos φ - FPT sin φ

FPT FQT sin θ

θ FQT

=

FQT sin

ϕ

FQT =

FQT cos θ - FQT sin θ

F QT cos θ - FPT cos ϕ FQT + FPT =

FQT cos θ - FPT cos φ - (FQT sin θ + FQT sin φ )

(b)

= FQT sin φ + FPT sin ϕ

FR

Figure 3 (a) The force F resolved into its component parts in the x and y directions. (b) The quadriceps and patella tendon forces can each be split into their components in two perpendicular directions (x and y). The components in any direction can then be added as scalars with positive values taken in the direction of the x and y arrows shown. A vector format using [ ] is often used to show the x and y components. The components can be re-combined to give the same resultant force as in Fig. 2.

Velocity and acceleration The velocity of a body is defined as the rate of change of position of the body with respect to time, or more simply, it is the speed of the body in a given direction. So a car’s speedometer measures only the scalar quantity speed, while a GPS can also determine the direction in which the car is heading and therefore its velocity. Velocity is measured in m s
1. Acceleration is defined as the rate of change of velocity and is measured in m s
2. The rate of change of a quantity with respect to time is often depicted mathematically by a dot over the quantity. So if an object has a position represented by the vector x, € the velocity may be written as x_ and the acceleration as x. If the displacement of an object is plotted against time on a graph, then the velocity at any given time is equal to the gradient of the graph at that point. Likewise, the acceleration of the object is the gradient of the velocity–time graph. These relationships can be used to determine the velocity and acceleration of body segments by filming the position of markers fitted to the body during a particular activity. The area under the velocity–time graph is equal to the distance travelled and the area under the acceleration–time graph is equal to the velocity.

Angular velocity and angular acceleration Physical quantities to measure linear motion such as velocity and acceleration have equivalents in rotation and, although the equations may be less straightforward, the principles are the same as for translation. Thus angular velocity is the rate of turning or the change in angle per time and is measured in rad s
1 whilst angular acceleration is defined as the rate of change of angular velocity and is measured in rad s
2.

Instantaneous centre of rotation If a body is both translating and rotating in a single plane, the point about which it rotates moves with time. At any given time, the point is described as the instantaneous centre of rotation (ICR). This can be determined by taking two or more points on the object, determining the direction of travel of these points at a particular time and finding the intersection of lines drawn perpendicular to those directions as shown in Fig. 5. In three dimensions, instead of defining the rotation about a single point, rotation is considered about a line or axis. At any given point in time, a rigid body can be considered to be rotating about an axis and translating along the same axis.

ARTICLE IN PRESS An introduction to basic mechanics

5

d1

z

y

ϕ FAM

W P

d2 FJR θ

x y x

Figure 4 A free body diagram of a person standing on their left leg. The abductor muscle forces are assumed to act as a point force FAM at a known position and in a known direction. If the hip joint is assumed to be frictionless, then the moment caused by the weight of the upper body W must be balanced by the abductor muscle force. This can be calculated by taking moments about P. Once FAM is known, the joint reaction force FJR can also be calculated by taking the sum of the components of the forces in the x and y directions. This produces two equations that can be combined as shown to find the direction y and then the magnitude of FJR.

B B

A A ICR

Figure 5 The instantaneous centre of rotation can be found by determining the velocity vectors of two points on a rigid body. If the points A and B move to A0 and B0 over a small time interval, then the velocity vectors are as shown and the instantaneous centre of rotation is at the point where lines perpendicular to the velocity vectors cross.

Degrees of freedom In general, in two dimensions, three coordinates are required to fully define the position of a rigid body: two

Figure 6 In three dimensions, a rigid body such as the vertebra has six degrees of freedom. Relative to another object (such as the adjacent vertebra), it can rotate about and translate along any of the axes shown.

define the x and y coordinates of a specific datum point on the body and the third defines the orientation of the body relative to that point. The body may translate, which will alter the x and y coordinates of the datum point, or rotate, which will alter the orientation coordinate. Each coordinate is known as a degree of freedom. In three dimensions, there are six possible degrees of freedom: three translational along three perpendicular axes and three rotational about each of the three axes (Fig. 6). Relative motion at the joint surfaces In the diarthrodial joints, the relative motions of the bones are constrained by the geometry of the joint surfaces and action of the ligaments and muscles spanning the joint. When the two joint surfaces remain in contact, they may move relative to each other by rolling or sliding. Figure 7 shows the simple case of a circular wheel on a flat surface. Rolling occurs when there is no relative velocity, that is, no slip, between the two contacting points and the ICR is located at the point of contact. Sliding contact occurs when there is no resistive force between the two surfaces and the ICR is located at the centre of the wheel. The relative motion across the human joints is generally a combination of rolling and sliding. Both take place simultaneously in the knee joint, whereas in the hip and shoulder joints, sliding motion predominates. These concepts are important in the design of joint replacements, particularly if the natural relative motion is to be preserved.

Kinetics Linear momentum Linear momentum is defined as the product of mass and velocity and has units of kg m s
1. Since velocity is a vector, momentum is also a vector quantity. Newton’s second law states that the rate of change of momentum is equal to the applied force. Normally, the mass

ARTICLE IN PRESS 6

R.K. Wilcox

(a)

ICR

E Equal distance on ground and around wheel

No translation ICR

(b)

Figure 7 (a) Rolling contact occurs when there is no relative motion between the two surfaces. The point where the wheel touches the ground is the instantaneous centre of rotation since there is no motion at this point. (b) Sliding contact occurs when there is no relative translation of the two objects. The instantaneous centre of rotation is therefore located in the centre of the wheel.

of the object remains constant and thus: F ¼ rate of change of ðmvÞ, F ¼ mv_ , F ¼ ma. That is, the force applied to an object is equal to the mass of the object multiplied by its acceleration. An object of mass m falling under gravity will accelerate at 9.81 m s
2( ¼ g), so the gravimetric force acting upon the object is equal to mg. If the object is prevented from moving, for example, by the ground, then the ground must apply an equal and opposite force of mg to the object. If the object is a person, this force is known as body weight.

Angular momentum There is an equivalent rotational quantity to linear momentum called the moment of momentum or angular momentum. This is defined as the product of the linear momentum and perpendicular distance from a given point on the plane. Angular momentum therefore has units of kg m2 s
1. An equivalent form of Newton’s second law also exists for angular momentum; it states that the rate of change of angular momentum of a particle about any fixed point in the plane is equal to the moment of the force about that point. For convenience, a term called the moment of inertia (I) is used in calculations of angular momentum. The angular momentum is then given by Ix (where x is the angular velocity) and the principle of angular momentum may be stated as _ Moment of external forces ¼ Ix. This is exactly equivalent to the linear case with moment replacing force, moment of inertia replacing mass and angular acceleration replacing linear acceleration.

(a)

(b)

Figure 8 (a) A figure skater begins to spin with a high moment of inertia due to their outstretched arms and leg. (b) When the limbs are brought close to the body, the moment of inertia drops and the speed of rotation increases.

ARTICLE IN PRESS An introduction to basic mechanics

7

Moment of inertia For a particle of mass m rotating about an axis at radius r, the moment of inertia is defined as I ¼ mr2 . In biomechanics most objects of interest are not particles but solid bodies. Imagine the object is made up of many small particles; then the total moment of inertia of the object is equal to the sum of the moments of inertia of all of the individual particles. The r 2 term here is very important, the further the particle is from the axis of rotation, the greater its influence on the moment of inertia of the object. So if an object were rotated, the moment required to accelerate it would depend not just on the mass of the object, but also on how the mass was distributed relative to the axis of rotation. The principle of angular momentum is illustrated by an ice skater spinning on the spot (Fig. 8). The skater starts the spin with arms held outstretched and one leg extended outwards. Then, by moving the limbs closer to the axis of rotation, the moment of inertia of the skater is reduced. If the frictional effects of the ice are ignored, then no external moments or forces (other than gravity) are acting on the skater and therefore their angular momentum ( ¼ Ix) will remain constant. Thus if I decreases, the angular velocity x must increase and the skater will rotate faster. Work and energy Work is defined as the product of a force and the distance moved by the point of application of the force in the direction of the force, shown in Fig. 9. Energy is defined as the capacity to do work. Work and energy are measured in units of joules (J) where 1 J ¼ 1 N m. Potential energy and

F

d (a)

F θ

d

(b)

h

mg (c)

Figure 9 (a) Work done ¼ Fd. (b) Work done ¼ F cos y d. (c) Work done in lifting weight against gravity ¼ mgh.

kinetic energy are the most common forms of energy encountered in biomechanics. Potential energy relates to the position of an object relative to an initial point and equals the work done in moving it from the initial point to its present position, for example, lifting a weight against gravity. By lifting a mass m through a height h, the work done against the force of gravity is equal to the force acting on the object ( ¼ mg) multiplied by the distance moved along the line of action of the force: Potential energy ¼ mgh. Kinetic energy relates to the velocity of an object and is the work done in increasing the velocity of the object from zero to its present value. Kinetic energy ¼ 12mv 2 . This can be proved for the simple case of an object accelerating at a constant rate a from stationary to velocity v: initial velocity ¼ 0, final velocity ¼ v, time taken ¼ t, v acceleration ¼ , t vt ðthis can be calculated 2 from the area under the velocity
time graphÞ, force applied to body ¼ mass  acceleration mv ¼ , t work done ¼ force  distance mv vt ¼  t 2 1 ¼ 2 mv 2 . distance travelled ¼

Conservation of energy For most biomechanical problems, the effects of other forms of energy are negligible and it can be assumed that an object, or system of objects, contains only kinetic and potential energy. Such a system is said to be ‘conservative’ if there is no dissipation of energy, for example, through heat loss due to internal friction forces or cyclic loading. If no external work is done on a conservative system then the sum of the potential and kinetic energy will remain constant. Thus if there is an increase in kinetic energy, there must be a corresponding decrease in potential energy or vice versa. Energy equations in biomechanics The principle of the conservation of energy is useful in biomechanics to make estimates of the velocity of an object or body segment from a knowledge of its initial potential energy. For example, in the laboratory, fractures can be generated using a drop test rig in which a known mass m is dropped from a known height h onto a bone specimen, as shown in Fig. 10. The impact velocity of the mass can be calculated from the principle of conservation of energy by assuming that the initial potential energy of the mass is converted to kinetic energy at the point of impact.

ARTICLE IN PRESS 8

R.K. Wilcox

m

h

example, when opening a stiff door, we instinctively apply a force as far from the hinged edge as possible in order to create a larger moment. And we know far greater force is required to begin to move a trolley carrying a patient than an empty one. However, the laws of mechanics may sometimes have surprising results. It may seem counterintuitive that an ice skater will speed up as the arms are moved inwards or that a gyroscope can stay balanced in an apparently gravity-defying position. It is only by gaining an understanding of the underlying mechanical principles that these phenomena are explained. The aim of this paper has been to provide such an understanding of the key principles of solid mechanics and illustrate some of their many applications in orthopaedics. The emphasis has been on the concepts rather than the details of the mathematics and the suggestions given in the ‘‘Further reading’’ section should be referred to if more in-depth mathematical derivations are required.

Further reading

(a)

m

v

(b)

Figure 10 (a) When mass is held at hight h: potential energy of mass ¼ mgh, kinetic energy of mass ¼ 0 ) total energy ¼ mgh. (b) When mass impacts on specimen: potential energy ¼ 0 ) 2 kinetic p ffiffiffiffiffiffiffiffi energy ¼ total energy ¼ mgh ) 1/2mv ¼ mgh ) v ¼ 2gh.

Conclusion Many of the principles of mechanics may appear intuitive without any knowledge of the underpinning science. For

1. Dowson D. Basic mechanics. In: Dowson D, Wright V, editors. Introduction to the biomechanics of joint replacement. London: Mechanical Engineering Publications Ltd.; 1981. p. 11–20. 2. Meriam JL, Kraig LG. Engineering mechanics: dynamics, 5th ed., vol. 2, New York: Wiley; 2002. 3. Mow VC, Flatow EL, Ateshian GA. Biomechanics. In: Buckwalter JA, Einhorn TA, Simon SR, editors. Orthopaedic basic science. Biology and biomechanics of the musculoskeletal system. 2nd ed. Rosemont: American Academy of Orthopaedic Surgeons; 2000. p. 133–80. 4. Mow VC, Hayes WC. Basic orthopaedic biomechanics, 2nd ed. Philadelphia: Lippincott-Raven; 1997.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 9–15

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: BIOMECHANICS FOR THE FRCS ORTH EXAM

(ii) Deformation of materials S.M. Green Centre for Biomedical Engineering, School of Engineering, University of Durham, Durham DH1 3LE, UK

KEYWORDS Stress strain behaviour of materials; Mechanical properties of materials; Failure; Microstructure; Dislocations

Summary This article provides an overview of the mechanical properties of materials with particular emphasis placed upon orthopaedic biomaterials. Engineering concepts of stress, strain, shear, yielding and failure are introduced and through consideration of microstructural mechanisms, the behaviour of materials under loading is discussed. In providing a concise overview of the deformation of materials, the reader is introduced to fundamental concepts in engineering and materials science. & 2005 Elsevier Ltd. All rights reserved.

Introduction

Force and displacement

Understanding how and why materials deform is of fundamental importance in general engineering and bioengineering design. Such knowledge underpins appropriate material selection for devices and components ranging from hip prostheses to suture materials. This article outlines the underlying principles behind the material characteristics familiar to us, such as why alumina is brittle, stainless steel is tough and ductile and why polymers have low service temperatures. The starting point lies in consideration of the bonding prevalent at atomic and molecular level in the different classes of materials, from which the mechanical properties can be explained and understood. This article reviews these fundamental concepts and provides the reader with an overview of the topic that brings together elements of engineering, materials science and biomaterials.

According to Newton’s third law, every action has an equal and opposite reaction. Practically, when a solid is loaded by application of a force, it will respond to the applied load by displacement on the atomic scale. If this atomic displacement is significant, a macroscopic change in dimension will result. The magnitude of deformation realised is dependent both upon the applied load and the way in which the atoms that constitute the solid are held together, i.e. the atomic bonding. The ease of atomic displacement on loading is governed by the strength of the bonds making up the material—the higher the bond strength, the greater the load that must be applied to realise a given bond separation. Figure 1 shows the typical relationship between interatomic distance, x, and potential energy for two idealised atoms joined by an interatomic bond. Such bond-potential curves are qualitatively similar for all types of bonding, both primary (ionic, covalent and metallic) and secondary (hydrogen and van der Waals). Long-range attractive interactions grow stronger as the atoms/ions approach whereas short-range strong repulsive interactions keep matter from collapsing in upon itself.

Tel.: +44 191 334 2519; fax: +44 191 334 2550.

E-mail address: [email protected]. 0268-0890/$ - see front matter & 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2005.11.001

ARTICLE IN PRESS 10

S.M. Green

Superposition of attraction and repulsion produces a minimum in potential energy that specifies the equilibrium separation, or the bond length, x0. Pulling or pushing the atoms from x0, requires energy, and thus the potential energy of the atom-bond construct rises for values of x greater or less than x0. Mathematically, the 2nd derivative of the bond-potential curve, evaluated at the minimum, tells us how much force is necessary to stretch the bond by a small amount and this is a measure of the stiffness of the material. Therefore, the greater the curvature at the bondpotential curve minimum, the higher is the stiffness (or more correctly the elastic modulus), of the material. Beyond these small, elastic separations, the bond strength also influences the temperature at which materials melt. We find that strongly bonded solids are characterised by high melting points and high elastic moduli, whereas materials bonded via secondary bonds, for example many organic polymers, are characterised by low melting points and low elastic moduli (Table 1). Quantification of the strength and stiffness of materials requires knowledge of the force distribution under loading. Pulling forces are classified as tensile and positive, and cause elongation of a piece of material in the direction of the force. Pushing forces are compressive and negative, and cause contraction of a material in the direction of the force.

spread over a large enough area, but no if the force is very concentrated. The concept of force intensity, i.e. force per unit area, is needed when determining failure risk. This leads to the simplest definition of stress, being the force applied to a body, divided by its cross-sectional area. Figure 2 illustrates a solid cylinder of material of original cross-sectional area Ao. Upon application of a tensile force F, the normal stress (s) in the shaded plane is given by s¼

F . Ao

Stress is usually expressed using the unit of the Pascal (Pa), which is equivalent to 1 N/m2. The use of stress instead of force is of fundamental significance in engineering, since it relates directly to the ability of a material to withstand the loads to which it is subjected. It may simplistically be stated that all materials have their own characteristics which determine the maximum stress they can safely withstand. In order to describe these characteristics in more appropriate detail, it is necessary to introduce the concept of strain.

F

Stress Forces alone are by themselves not too helpful unless seen in the context of the material required to be subjected to such forces. Can a prosthetic ceramic femoral head withstand a force of 2 kN? The answer may be yes if the force is l

lo

Ao

F

Figure 1 Interatomic bond potential energy with respect to separation distance for an idealised atom pair.

Table 1

Figure 2 Axial loading of a solid cylinder of original length lo and cross-sectional area Ao.

Bonding-type comparisons.1

Bonding type

Bond energy

Example substance

Elastic modulus/GPa

Melting temp/1C

Covalent Ionic Metallic Van der Waals Hydrogen

High High Intermediate Low Low

C (diamond) Alumina ceramic Stainless Steel UHMWPE Polymer H2O

1000 380 200 1 n/a

3500 2500 1500 110 0

ARTICLE IN PRESS Deformation of materials

11

Strain Returning to the femoral head example, knowledge of the cross-section was needed in order to determine whether or not the 2 kN force could be carried. A similar concept applies when considering the deformation undergone by materials. A 0.5 mm deformation may not be cause for concern on a tibial bone plate that was 150 mm in length, but might cause complete failure in a PMMA cement mantle that was originally only 5 mm in thickness. Normal strain, e, is the response of the material to load, in the direction of loading, and will depend upon the applied force, crosssectional area, length and the type of material. Strain, e, is defined as the ratio of the change in length, l–lo, of the component compared to its original length, lo, l
lo e¼ . lo Since strain is defined as a length divided by a length, it has no units and is a dimensionless number. With the exception of elastomers and some polymers, for most materials and service conditions, strain will take a very small value (10
4). Figure 2 also illustrates the Poisson effect. Axial elongation of the cylinder is accompanied by radial contraction. Generally, any axial deformation in a material is accompanied by an associated lateral deformation of opposite sign. The Poisson’s ratio of a material, n, is given by elateral n¼
. eaxial

designed to resist more than one type of loading as in service, axial forces, bending moments and torques may all act simultaneously. A femoral stem experiences such combined loadings during gait. Such components can be analysed by superimposing the individual, constituent stresses due to each load. The combined, resultant stresses can then be calculated and compared with the known material strength. Such analyses are the remit of stress analysis and introductory texts on this subject are given at the end of this article.2,3

Stress–strain behaviour of materials Examination of the stress–strain behaviour of materials, for example via a simple uniaxial tension test, enables determination of a number of important material properties. As discussed earlier, solids will exhibit elastic deformation on loading, to a greater or lesser extent, as the macroscopic manifestation of changes in the mean interatomic separation distance, x. Plotted, there is often an initially linear relationship between stress and strain, the gradient of which is the elastic or Young’s modulus. The elastic modulus, E is given by E¼

Stress . Strain

In practice, for most materials nE0.3, with the exception of fully incompressible materials such as elastomers, for which n ¼ 0:5.

Unsurprisingly, the imposition of compressive, shear or torsional stresses also evokes elastic behaviour in solids. For low strains, the elastic modulus of most materials is independent of compressive or tensile loading. The shear modulus, G, is the constant of proportionality between shear stress, t, and shear strain, g:

Shear stress and strain

t G¼ . g

The discussion above has concentrated on forces applied perpendicularly to the section of interest, causing ‘normal stress’ s and ‘normal strain’ e. We also need to consider shear stresses which arise from shear forces. Figure 3 shows a cuboid ‘element’ of material subjected to shear forces, which cause a deformation of square faces into a parallelogram. The shear stress, t, is defined as the shear force, Q, divided by the shear area, A. Shear strain, g, is defined as the angle through which the square faces deform to become parallelograms. Most structural components have to be

For isotropic, linearly elastic materials of Poisson’s ratio n, the shear and elastic moduli are related to each other according to the expression E ¼ 2Gð1 þ nÞ. Figure 4 illustrates stress–strain plots typical of many ceramic, metal and polymer materials. The differing gradients of the linear elastic regions of these curves are apparent, reflecting the differences in elastic moduli and the different types of bonding found in these materials (Table 1).

δ

Q Shear force Q Angle γ

Shear force Q Shear area A (shaded)

Figure 3 Shear effects.

h Q

ARTICLE IN PRESS 12

S.M. Green

Ultimate tensile stress Yield stress Stress

High strength steel Ceramic

Yield Point E

Mild steel

Stress

Fracture Point

ElasticPlastic

Strain

Figure 5 Stress–strain behaviour of mild steel. Polymer Strain

Figure 4 Stress–strain behaviour of materials. 0.2% proof stress

Beyond the linear elastic region of the stress–strain response, which typically holds only for very small strains, most materials will either fail immediately in a brittle manner, due to simple cleavage of atomic planes within the structure, or exhibit permanent, plastic deformation, which is termed ductile behaviour. Ceramics and polymers such as PMMA fail in a brittle manner and the ultimate tensile stress of such materials is the stress at fracture. In contrast, metals such as stainless steels and titanium and polymers such as UHMWPE are ductile. For these ductile materials, which are characterised by appreciable strain prior to failure, the yield point marks the transition from elastic to plastic deformation. Figure 5 illustrates the elastic to plastic transition at the yield point of a mild steel sample, the stress at which is termed the yield stress. Also shown is the ultimate tensile stress and the fracture point, which gives the fracture stress. For materials that show appreciable plastic flow prior to failure, the fracture stress calculated using the simple definition of force/original cross-sectional area produces the anomalous result of apparent falling stress with increased strain, implying a reduction in the load-carrying capacity of the material. This erroneous result arises due to the diminishing instantaneous cross-sectional area of the sample once strained beyond the ultimate tensile stress, and is a phenomenon known as necking. Many ductile materials do not exhibit a clearly defined yield point, for example the high strength steel sample shown in Fig. 4, and for these materials the stress at a particular strain offset, for example 0.2% of strain, is used as a substitute for the yield stress in design calculations. This is also sometimes referred to as a proof stress (Fig. 6).

Work of fracture The area under the stress strain plot is the work to fracture, and represents the total energy expended in deforming the material to failure by fracture. This is a measure of both the elastic and plastic elements of deformation and quantifies the toughness of the material. A related property is the fracture toughness of the material, which can be assessed using pendulum type impact tests (Izod or Charpy methods. See, for example, Dieter, 1998).4

Stress

Yielding and failure

0.2

% Strain

Figure 6 Stress strain behaviour typical of a ductile material that does not exhibit a pronounced yield point.

Unlike the elastic modulus, which for a given base alloy is insensitive to composition and micro-structural form, the work to fracture of a material is strongly influenced by material micro-structure, and can thus be controlled by appropriate thermomechanical processing. Figure 7 illustrates the typical stress strain behaviours of cast and forged specimens of a hypothetical Ti-based alloy, exhibiting similar ultimate tensile strengths and elastic moduli but quite different ductilities, indicated by the different areas under the respective stress–strain plots. The cast alloy, although of similar ultimate strength, has a much lower toughness and fails in a catastrophic manner once overloaded. In comparison, the forged alloy has a more forgiving failure mechanism, exhibiting measurable yielding prior to fracture and in so doing providing an advanced warning of fracture failure. Other methods of measuring ductility commonly found in material design specifications are elongation to break and the reduction in area, both of which are based upon comparisons of specimen dimensions measured before and after tensile testing. For these methods, the greater the dimensional change, the higher the ductility.

Micro-structural mechanisms of failure The preceding discussion has highlighted generic differences between the different classes of materials based upon the bonding mechanism prevalent. Ceramics feature in

ARTICLE IN PRESS Deformation of materials

13

orthopaedics in a number of specific areas—Al2O3 and ZrO2 are used in some joint applications as hard wearing bearing surfaces whereas hydroxyapatite is often plasma sprayed onto metallic surfaces to enhance osseointegration. Metal alloys such as 316L stainless steel, CoCrMo and Ti6Al4V are used structurally to replace bone and polymers such as PMMA and UHMWPE find application in replacement joints as bone cement and low friction bearing surfaces, respectively. Based upon bond strengths alone (Table 1) we would perhaps expect ceramics to feature more prominently in higher strength structural in vivo applications, such as replacement joints. The reason why they do not, and metals such as stainless steels and forged Cr alloys do, is due to the differing ability of these materials to resist crack propagation. In metals, plastic deformation is facilitated by mobile lattice defects called dislocations which enable plastic flow to occur at stresses less than those needed for bulk plane cleavage.5,6 This incremental, localised defect motion has been likened to the movement of a caterpillar (Fig. 8) and enables less energy to be expended in achieving slip (plastic deformation) than if atomic planes moved over one another en masse. In real materials, surface and bulk defects, for example cracks and voids, are usually present. A flaw such as a crack tip acts as a stress raiser when the crack is pulled open under tension. In ductile materials, this stress raising effect is diminished by dislocation-facilitated plastic flow, which

blunts the crack tip. In ceramics, glasses and other brittle materials, although dislocations exist, they are of low mobility, partly as a consequence of high interatomic bond strengths. For ceramics loaded in tension, once a crack reaches a critical length such that the stress at the crack tip exceeds the cleavage strength, the crack propagates uncontrollably, causing failure. It is for this reason that brittle materials are usually designed to be loaded in crackclosing compression. The mobility of dislocations in metals can be controlled by thermomechanical processing to both alter the dislocation density within the metal and also to alter the microstructure. For example, plastic deformation at temperatures of less than 1/3 of an alloy’s melting temperature is termed cold work and produces work hardening via an increase in dislocation density. This increases the yield stress of the alloy as dislocation mobility falls with increasing dislocation density, due to interaction of the dislocation strain fields. Micro-structural features such as grain boundaries and second phase particles also impede the motion of dislocations and the prevalence and distribution of these features can be readily controlled during manufacture. We find that for most alloys, as the dislocation mobility falls, yield and tensile strengths rise but at a cost of reducing ductility (Fig. 9).

1400 Ultimate strength 1200

Work of fractureductile

0.2% Yield strength

800 600

80

400

Work of fracturebrittle

40

% Elongation

200 0 0 Strain

Figure 7 Stress strain plots of Ti alloys processed by casting, to produce a brittle sample and by forging, to produce a ductile sample.

20

40 % Cold Work

Unit step of slip Dislocation moves

Figure 8

0 80

Figure 9 Effect of cold work on the yield strength, ultimate tensile strength and elongation to failure of austenitic (type 316L) stainless steel.7

Caterpillar moves

Dislocation

60

Strain %

Stress/MPa

Stress

1000

Representation of the analogy between a caterpillar and dislocation motion.

ARTICLE IN PRESS 14

S.M. Green

For ceramics, employing mechanisms that impede dislocation motion produce no useful strengthening effect, as deformation in such materials is not facilitated by plastic flow.8 Engineered strength improvements instead rely upon reduction in flaw size by the use of manufacturing processes to minimise pre-existing defects such as cracks and pores. It is because of the high melting temperatures in these materials (Table 1) that powder processing routes, for example sintering, are often the only viable method of manufacture. Attaining a 100% dense, void free product from a sintered compact presents particular technological challenges, requiring hot isostatic pressing (HIP) during sintering and a fine grain sized starting powder. Given these conditions, high density ceramics can be produced with considerably improved toughness compared with their

Figure 10 Temperature dependence of the stress–strain behaviour of crystalline PMMA.

Table 2

conventionally processed counterparts. Typical grain sizes of HIPed alumina used for ceramic femoral heads are o2 mm which compares to the 30 mm typical grain size for CoCrMo alloy femoral components. For polymers, deformation generally occurs by slippage at the weaker van der Waals intermolecular bonds rather than by breakage of the stronger primary intramolecular bonds.9 There exists a strong temperature dependence for the stability of these intermolecular bonds, with loss of strength and tendency for plastic flow at temperatures much lower than those of most metals and ceramics (Fig. 10). Thermoset polymers such as vulcanised natural rubber and silicone elastomers are characterised by some primary bond crosslinks that connect the long chain molecules together. They are thus generally more stable with respect to temperature than thermoplastic polymers such as ultra high molecular weight polyethylene (UHMWPE) which have only secondary bonding between molecules. When an elastomer is loaded, the characteristic high strains are realised through molecular chain mobility, enabling uncoiling, stretching and alignment of the structure. On load removal, the cross-links provide a ‘return path’ for the chains to return to the preloaded disorganised structure. This strain-induced molecular alignment, or strain hardening, is a feature of many polymers and is a consequence of high chain mobility. A consequence of the low strength of the intermolecular bonds in polymers is that these materials are also characterised by time dependency of the stress strain behaviour. We find that for many polymers, the measured elastic modulus depends upon rate of loading during testing, with fast loading producing the highest moduli. Additionally, appreciable plastic flow can occur when polymers are

Mechanical properties of some orthopaedic biomaterials.

Material

Elastic modulus (GPa)

Yield strength (MPa)

Ultimate tensile strength (MPa)

Strain to failure (%)

193

170

480

40

Metals 316L stainless steel (annealed) 316L stainless steel (cold worked) CoCr Mo (as cast) CoCrMo (Hot Forged) Ti6Al4V

193

1200

1300

12

210 210 120

450 890 795

655 1400 860

8 28 10

Ceramics Alumina 499.5% Hydroxyapatiteb

380 50

n/a n/a

350a 400a

2 1

n/a 25

35 39

7 450

15 0.3

n/a n/a

150 15

3 6

Polymers PMMA bone cement UHMWPE Bone Cortical bone Cancellous bone Data from Black (1998).11 a Compressive strength. b Data from Kohn (1993).10

— —

ARTICLE IN PRESS Deformation of materials loaded at constant stress or equally, when held at constant strain, stress relaxation occurs with time. Time-dependent plastic flow is termed creep and the time dependency of the stress–strain behaviour is called viscoelasticity and is the product of both elastic and viscous (cold flow) responses of the material to loading. The viscoelastic response of a material to loading can be modelled theoretically by using combinations of elastic (spring) and viscous (dashpot) elements. The stress strain behaviours of metals and ceramics are practically independent of time and do not exhibit viscoelasticity whereas tissues such as tendons, cartilage and bone do.

Summary Representative mechanical properties of some of the common orthopaedic biomaterials, and for comparison the mechanical properties of cortical and cancellous bone, are listed in Table 2. The derivation and quantification of the mechanical properties included in the Table have been discussed in this article. The wide range in yield, UTS and ductility of the listed alloys with respect to material condition is evident and highlights the need to be aware of the thermomechanical history of the component being used in vivo. Similar considerations can apply to ceramics and polymers. In general, material specification on the basis of material composition alone, without knowledge of the manufacturing route, is insufficient for safe design. Postforming processes such as welding (produces locally an ascast micro-structure), machining (can introduce stress concentrating surface defects and sub-surface stresses) and cold working (raises the dislocation density and diminishes ductility) can all materially affect the strength

15 of a finished component and should be taken into consideration during design. In providing a concise overview of the deformation of materials, this article has given a brief introduction to a number of engineering disciplines. Introductory texts on specific topics are referenced for those readers wishing to gain a more detailed insight than that provided by the material presented here.

References 1. Callister WD. Materials science and engineering an introduction. New York: Wiley; 1997 [Chapter 2]. 2. Gere JM, Timoshenko SP. Mechanics of materials, 4th ed. Boston: PWS; 1997. 3. Timoshenko SP, Goodier JN. Theory of elasticity, 3rd ed. New York: McGraw-Hill Book Co.; 1990. 4. Dieter GE. Mechanical metallurgy. London: McGraw-Hill Book Company Ltd.; 1988. 5. Bacon DJ, Hull D. Introduction to dislocations, 4th ed. Woburn, UK: Butterworth Heinemann; 2001. 6. Cottrell AH. An introduction to metallurgy, 2nd ed. London: Edward Arnold Publishers Ltd.; 1975. 7. Park JB, Lakes RS. Biomaterials an introduction, 2nd ed. New York: Plenum Press; 1992. 8. Kingery WD, Bowen HK, Uhlmann DR. Introduction to ceramics, 2nd ed. New York: Wiley; 1976. 9. Young RJ, Lovell P. Introduction to polymers, 2nd ed. London: Chapman & Hall; 1991. 10. Kohn DH. Standard handbook of biomedical engineering and design. In: Kutz M, editor. New York: McGraw-Hill Inc.; 2003 [Chapter 13]. 11. Black J. Orthpaedic biomaterials in research and practice. New York: Churchill Livingstone; 1988.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 16–22

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI SYMPOSIUM: BIOMECHANICS FOR THE FRCS ORTH EXAM

(iii) Material properties of biological tissues related to joint surgery Ulrich Hansena, Spyridon Masourosb, Andrew A. Amisa, a

Biomechanics Group, Department of Mechanical Engineering, Imperial College London, Exhibition Road, London SW7 2AZ, UK b Department of Bioengineering, Imperial College London, Exhibition Road, London SW7 2AZ, UK

KEYWORDS Biomechanics; Bone; Cartilage; Menisci; Tendons; Ligaments

Summary The mechanical characteristics of biological materials underpin proper joint function. Changes to these structures through repair or replacement without appreciating the underlying structure–function relationships are likely to alter joint kinematics and predispose the joint to osteoarthritis and other degenerative changes. This paper will review the mechanical characteristics of bone, cartilage, the menisci, tendons and ligaments. In-depth knowledge of the mechanical behaviour of these structures is essential for replicating their mechanical function, but also provides a means to identify the perhaps still unappreciated function of many tissues. The paper aims to instil a general appreciation for all tissues of the human joints, hoping that future surgeons will proceed with caution and understanding when repairing or replacing tissue. & 2005 Elsevier Ltd. All rights reserved.

Introduction When performing an orthopaedic procedure to restore the function of a joint, it seems self-evident that a key requirement is to understand the function of the tissues that are being repaired or replaced. In turn, the functions of these tissues have resulted from the evolution of distinct structures that, from an engineering point of view, have Corresponding author. Tel.: +44 20 7594 7062;

fax: +44 20 7823 8845. E-mail addresses: [email protected] (U. Hansen), [email protected] (S. Masouros), [email protected] (A.A. Amis). 0268-0890/$ - see front matter & 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2005.12.001

been optimised in an impressive and noticeably hierarchical manner from the submicron level to the macroscopic structure of ligaments, menisci, etc. Not only is it important to be aware of the intricate nature of these tissues in order to replicate their mechanical function, but it can also be very informative in fully understanding their range of functions. In this paper the mechanical characteristics of bone, cartilage, menisci, tendon and ligament will be described. In each case, this will start with a description of the material constituents and their mechanical properties, progressing through the hierarchical organisation of the tissue while examining how this organisation affects the mechanical properties and function of the complete structure.

ARTICLE IN PRESS Material properties of biological tissues related to joint surgery

Biomechanics of bone Material and micro-structure The main material constituents of bone are 65% hydroxyapatite (calcium phosphate) and 35% collagen, cells and water. The collagen of bone (as in cartilage and tendons) is a tensile, load-bearing material and is made up of long chains of amino acid molecules arranged into a triple helix pattern. In bone, the crystal structure of the inorganic hydroxyapatite phase forms a bond with collagen at the level of this helical structure. The helical structures assemble to form microfibrils, which in turn are arranged into small sheet-like structures (lamellae). The orientation of the microfibrils within lamellae is relatively unidirectional, leading to higher lamellar strength in the direction of these string-like microstructures. Essentially, the ceramic-like hydroxyapatite gives this basic bone unit good compressive properties, the collagen fibres give it good tensile properties along the fibre direction, while the weak link of bone is its tensile properties perpendicular to the collagen fibre direction. The lamellae are further arranged into thick-walled tubelike structures called osteons (of diameter up to a few hundred microns) in which the walls are made up of several layers of lamellae. The lamellae have a predominant direction along the axis of the osteon, hence osteons are strongest along their length. The ‘hollow’ central part of these structures is known as the Haversian canal and it contains a blood vessel that supplies nutrients to the whole osteon through small canaliculae emanating centrifugally from the canal. Each osteon is covered by a boundary wall known as the cement line. Mature bone is mostly made up of many such subsystems, which collectively are known as Haversian bone. In some locations of the body, such as the endosteal or periosteal bone, the lamellae do not form concentric structures but are simply laid down as relatively big sheets of bone—this is known as lamellar bone. When bone is laid down relatively quickly, such as in immature bone or during fracture healing, the fibrillar structures are less organised. This type of bone is known as woven bone and is generally more flexible.

Cancellous bone and overall bone architecture Above the micro-scale bone can be divided into cortical (compact dense bone) and cancellous, or trabecular, bone (porous spongy bone). The peripheral ends of long bones are generally made up of cancellous bone covered by a thin outer layer of cortical bone. In the diaphyseal regions, the bone resembles a hollow tubular structure with thick walls of strong cortical bone. The trabecular structures of the low-density cancellous bone are arranged in such a way as to most effectively transfer the joint loads from the wide articulating surfaces into the strong cortices of the diaphyseal region. The great internal surface area of cancellous bone is exposed to bone regulating compounds and this results in this bone being very reactive. Therefore, the first and perhaps most significant bone changes caused by osteoporosis, or any other bone pathology are first seen in cancellous bone, hence the great interest in dual-energy X-ray scanning (DEXA) or ultrasonic measurements of cancellous-rich bones

17 when screening for osteoporosis. Even a cursory look at a radiograph shows marked regional variations in the cancellous bone density, so it is to be expected that there will be marked regional strength variations. Rice et al.1 found that cancellous bone modulus and strength both vary approximately with bone density squared, with the compressive strength usually in the range 1–15 MPa in human bone.2 The variation of strength with density is important for many reasons: knowledge of the link between local bone architecture and loads applied in normal use will show the directions of forces acting on the bone fragments that must be assembled and held together by fracture fixation devices. Knowledge of regional variations, for example, across the tibial plateau, will be very useful when deciding on the positions of fixation stems of joint replacements. The predominant load acting on the shafts of long bones is axial compression and intuitively it would therefore seem that compressive properties would be most relevant. However, most of this axial force is induced by the muscles. Typically, the muscle-induced loads are several times larger than those forces caused by body weight only. As the muscle forces are arranged alongside the bones and the bones are often curved, bones are subjected to both axial and bending forces. As the bone is subjected to a bending load, its central axis takes up a curvature. It follows that osteons on the outside of the bend are stretched, whereas those on the inside are compressed. It may also be concluded that between these two extremes there is an area where the bone is not stretched and therefore unloaded. It is this internal force distribution that resists the external bending moment applied to the bone. For the same magnitude of force, the stresses induced in the bone by bending loads are much greater than the stresses induced by axial forces. Furthermore, bending introduces tension in the bone, which is weaker in tension than in compression. Hence, the long bones have probably evolved primarily to resist bending. An engineering analysis of bending shows that the stress increases with the distance from the axis of the bone. A given quantity of bone will therefore resist bending more efficiently if it is distributed away from the axis of bending, so that the force resulting from the stress in a given fibre in tension or compression has as large a lever arm about the bending axis as possible. This is why steel girders are often of a deep I section, with their material concentrated into the flanges of the top and bottom, away from the centre line. The I section geometry is ideal for withstanding loads in one direction, but bones may have to resist bending loads from many directions, hence the circular section. Although not described in any detail here, the circular section is also the ideal shape to resist torsion. The reason for this is much the same as in bending; the torsion is resisted more efficiently if the stress in a given fibre has a large lever arm around the axis of torsion. This is also the explanation why periosteal callus formation is able to stabilise a fracture of the shaft of a long bone after the deposition of a relatively small quantity of immature bone tissue. Apart from evolving a geometry ideally suited to resist bending loads, the osteons, which act as reinforcing short fibres, are orientated along the axial direction of the bone. As a result, the tensile strength of diaphyseal bone is much higher in the axial direction than in the radial or circumferential directions. Typical tensile strengths for

ARTICLE IN PRESS 18

U. Hansen et al.

human bone are 133 MPa axially and 51 MPa circumferentially.3 While this material orientation is ideal for resisting bending loads, the resulting planes of weakness explain the occasional complication of longitudinal splitting of the femur during insertion of a hip stem into the diaphysis. The interference forces during insertion cause tensile circumferential stresses, which are occasionally sufficient to split the bone along the weak direction, perpendicular to the osteons. These stress and material distributions have other practical implications for the surgeon. For example, if the femur is subjected to an anteroposterior bending load, then there will be a tensile stress in the anterior cortex and compressive stress in the posterior cortex. It will be relatively safe to drill a hole on the lateral aspect, because this area will be relatively unloaded by this bending effect while it would be much less advisable to drill the anterior cortex. The stress distribution around a discontinuity, such as a hole, is complex and in the compressed posterior aspects would introduce tensile circumferential stresses. Although these tensile stresses are relatively low in comparison to the compressive stresses in this area, they are directed along the weak direction of the bone fibres and might result in splitting of the bone. Many orthopaedic surgical procedures involve saw cuts or the drilling of holes. It is a well-known principle in engineering practice that any sudden change in the crosssection or shape leads to localised high stresses. For example, a smooth round hole has a stress concentration factor of 3, which means that the stress in the vicinity of the hole may be raised by a factor three compared to the overall stresses. It is generally the case that the sharper the discontinuity, the greater the effect, sharp cracks being particularly harmful. If a hole must be made in the shaft of a bone, it is best kept to a rounded shape always avoiding sharp corners. The area involved should be enlarged axially rather than circumferentially, thereby minimising cutting of the load-bearing bone fibres.

Articular cartilage and meniscal tissue Material, micro-structure and mechanical properties The material of articular cartilage and meniscal tissue is relatively similar. Both can be described as porous perme-

Table 1

Intrinsic moduli. Compression (MPa)

Shear (MPa)

Permeability (m4/Ns)

3.24 (s) 1.01 (m) 0.32 (d)

0.79

0.68

4.7
1015

2.8

0.42

0.112

0.81
1015

Tension (MPa) Parallel Cartilage

Meniscus

able composites made up of an organic solid matrix saturated with water. The water phase constitutes 65–80% of the total weight of healthy tissue. The primary loadcarrying structural components of the solid matrix are collagen fibres and proteoglycans. In articular cartilage, collagen constitutes approximately 75% of the dry tissue weight, while the rest is mainly proteoglycans. In meniscal tissue, the proteoglycans account for less than 1% of the total weight and the concentration of collagen is slightly greater than in articular cartilage. Even so, proteoglycans play an important role, as they can entrain water 50 times their weight. Otherwise, the main constituent difference between articular cartilage and meniscal tissue is that the meniscus is composed almost exclusively of Type I collagen, while the collagenous fibres of articular cartilage are almost entirely Type II collagen. The collagen molecules of both articular cartilage and menisci assemble to form collagen fibres. However, the fibrillar structure of articular cartilage is generally much smaller in scale (0.1–10 mm) than the dominant fibre bundles (50–150 mm) of the menisci. Also the fibre distribution and orientation of articular cartilage and the menisci are characteristically different. On the surfaces of articular cartilage of the proximal tibia, the fibres lie densely packed and orientated in a plane parallel to the articular surface. In the bulk middle zone, the fibre orientation is more or less random, while in the deep regions near the subchondral bone, the fibres are oriented perpendicular to the bony interface. Although the articular cartilage of other bones may exhibit some differences from this description, it is more or less consistent. In contrast, the dominant fibre orientation of the menisci is circumferential, parallel to the periphery of the meniscus. The mechanical properties of the articular cartilage and meniscal tissue reflect the inhomogeneous distribution and directional orientation of the collagen fibre bundles as well as the porous water-saturated nature of these materials. The mechanical properties reported by Fithian et al.4 are shown in Table 1. It is worth pointing out that the above stiffness values are comparable or even lower than the reported joint pressures (20 MPa) the materials have to carry, suggesting that cartilage and menisci are very deformable materials. Also notable from Table 1, and clearly reflecting the ultrastructure of the materials, is that meniscal material is two orders of magnitude stiffer in the circumferential direction compared to a direction perpendicular to the

10.2 (s) 3.2 (m) 0.87 (d) 198.4

Perpendicular

s ¼ surface; m ¼ middle; d ¼ deep.  Parallel and perpendicular to dominant fibre direction.

ARTICLE IN PRESS Material properties of biological tissues related to joint surgery periphery. Similarly, meniscal material is 10 to 20 times stiffer than the superficial layers of cartilage, which in turn are an order of magnitude stiffer than the middle and deep layers of cartilage. In contrast, the meniscus is only half as stiff as cartilage in compression and only a sixth as stiff as cartilage in shear, both characteristics that render the meniscus highly deformable allowing it to conform to the surface geometries of the femur and tibia through a range of movements.

19 which resist the menisci being squeezed out of the joint. As a result of this arrangement, the menisci stay within the contact area carrying a large portion of the axial load and transferring it over a wide contact area to the tibia. Another important function of the meniscus, and to a lesser extent of the cartilage, is that of a shock absorber.7 For this function, the time-dependent nature of the material is critical. Essentially, the energy of the impact is dissipated by frictional drag forces of the fluid phase being squeezed through the matrix.

Structure–function relationships

Clinical relevance To understand the mechanical behaviour of these tissues, it is essential to appreciate their very low permeability and biphasic nature. These characteristics result not only in a functionally important viscoelastic response, which stems from the fluid phases slowly being squeezed out of the matrix, but also in very deformable materials that can carry loads far higher than one would expect based on a consideration of the individual constituents. Because of the low permeability of these materials, the fluid phase is ‘trapped’ within the matrix and pressure builds up within the fluid.5 Stolz and Ateshian6 reported that the fluid carries 95% of the load, thus significantly shielding the collagen– proteoglycan matrix. As a result, articular cartilage and menisci are able to carry very high joint loads while also being very deformable. An important function of cartilage relates to lubrication of the joint; fluid being extruded from the cartilage plays an important role in this function. The friction coefficient of the fluid pressurised cartilage surface is 20 times smaller than if the load was carried only by the collagen–proteoglycan matrix. It seems reasonable to think that the reason for the strong surface layer of articular cartilage—relative to the deeper layers—is to resist the frictional forces in the articulation, shielding the deeper weaker layers. Furthermore, the deeper layers are more permeable and richer in water-binding proteoglycan. Hence, damage of the surface layer will not only expose the weaker layers to higher forces but will also disrupt the pressure build-up in the fluid. This will break down the mechanism by which cartilage is able to carry high loads and will probably disturb lubrication of the articulation. One of the main functions of both cartilage and menisci is to transfer load from one bone to the next in a manner that spreads the load over a wide contact area, thus avoiding damage to the bone. As has already been pointed out, the biphasic nature of the tissue enables cartilage to carry out this function. Because of the characteristic orientation of the fibres, the menisci are even better suited for this function. As shown in Table 1, the shear and radial (compressive) stiffness of the menisci is much lower than articular cartilage and the menisci are therefore very deformable, enabling conformity with the joint geometries. As the menisci are wedge-shaped and occupy the periphery of the joint, axial loading of the joint will tend to squeeze the menisci out of the joint. The insertions at the anterior and posterior horns prevent this and as a result large circumferential (hoop) tensile stresses build up. This provides a structure–functional explanation for the circumferentially oriented strong fibre bundles of the menisci,

The first signs of osteoarthritic cartilage degeneration is a change in cartilage hydration. As indicated in the earlier sections, an imbalance in hydration levels is likely to disturb the load-carrying capacity of the tissue as well as the joint lubrication. Higher frictional forces and reduced loadcarrying capacity will lead to further damage of the cartilage. Damage to the cartilage, whether it is caused by osteoarthritis (OA)-related changes or trauma, will result in loss of fluid pressurisation and a continued and accelerated deterioration of the tissue. It is generally accepted that raised stress levels, due to injury or surgical procedures such as meniscectomy, predispose cartilage to degeneration. However, gross cartilage failure following for example a meniscectomy has never been reported. Hence, the progression of OA is clearly a continuous process and may occur even at non-injury stress levels. Early work by Weightman et al.8 showed that cartilage fatigue strength decreases with the number of load cycles. It was shown that these fatigue properties deteriorate with age to a level that may explain OA in later life. Despite the seemingly clear relevance of cartilage fatigue properties to OA, only one paper9 has followed up on Weightman’s early work. A related factor is the interplay between mechanically induced fatigue damage and biological factors. Vincent et al.10 showed that growth factors (bFGF) were released from the cartilage during cyclic loading. Growth factors are thought to play a role in cartilage maintenance but may also play a role in cartilage degeneration. Either way, understanding the association between cyclic loading, fatigue properties and biological factors seem essential to the understanding of the progression of OA. Surgical strategies to stimulate repair of damaged cartilage is an area of very active research. New techniques, based on tissue-engineering principles with cultured cells and scaffolds, are challenging established techniques based on generating a repair response from the bone marrow. The mechanism(s) by which chondrocytes convert physical stimuli, such as mechanically driven interstitial fluid flow, to intra-cellular signals, which in turn direct cell synthesis, will undoubtedly play an important role in the development of treatments. The ultrastructure of the menisci explains the characteristic circumferential bucket-handle tears of the meniscus as failures along weak planes in the material. This also provides the rationale for avoiding radial meniscectomies, as these will cut the important circumferential fibres. The clinical outcome of ignoring the function of the menisci was first

ARTICLE IN PRESS 20

Tendons and ligaments The material constituents of tendons and ligaments are relatively similar. The main material is collagen, with fibrils arranged in larger fibre bundles bound together in a mucopolysaccharide ground substance. The ratio of ground substance to collagen is higher in ligaments than in tendons. The main function of both tendons and ligaments is to transmit tensile forces and the orientation of the collagen fibre bundles governed primarily by the main tensile load direction. The fibre bundles of ligaments are not as unidirectional as those of tendons. The reason for this is that bones that are linked by a ligament often move greatly as the joint flexes and extends, the attachment sites rotating relative to each other. This means that the various fibre bundles within the ligament cannot all be tensed at the same time but rather different fibre bundles within the ligament will resist the load at different positions of joint rotation. Hence, ligaments often function as two or multiple structures, sliding or crossing each other, loosely connected by ground substance. One possible reason for the larger amount of ground substance in ligaments is to reduce the resistance between the fibre bundles when they slide and cross over each other. Tendons are usually longer than ligaments and the attachment to muscle much more flexible, allowing the fibre bundles to be equally tensed throughout joint movement. The differences in orientation of fibre ultrastructure between ligaments and tendons results in characteristically different failure patterns. Tendons tend to ‘snap’, failing completely, implying a ruptured tendon or at least a tendon, which has lost its load-carrying capacity. On the other hand, partial ligament rupture is a common occurrence. The reason is that those bundles of the ligament that are tensed at a particular joint flexion angle will rupture first. In some cases, this implies that certain parts of the ligament are more important than others and that the surgeon should put most efforts towards reconstructing these parts. In the ACL, for example, the anterior part of the ligament is tensed throughout flexion while the posterior parts are only tensed or functional at full extension.13 This suggestion is substantiated by superior mechanical properties of the anterior

part of the ACL, which might reflect the higher functional demands put on it in vivo. Both tendons and ligaments have structures that vary along their length, both microscopically and on a macroscale. This variation is optimised to maintain the great flexibility in the bulk part of the structures as well as allowing the transmission of forces from the soft tendon and ligament structures to the rigid bone at the attachment site. This is accomplished by means of a multilayer stack of materials of progressively increasing stiffness: the tendon or ligament blends into a fibrocartilaginous layer, which passes through a transition into a calcified form as it approaches the bone itself. Some collagenous fibres pass straight through these layers and blend directly into the collagenous matrix of the bone itself; these are known as Sharpey’s fibres. In some situations, such as the medial collateral ligament of the knee, the ligament approaches the bone surface tangentially and so the fibres do not pass through all the transitions into the bone structure. Instead, they blend into the periosteum over a wide area. The differences in insertion site morphology also lead to distinct failure patterns. Failure of a ligament or tendon that inserts into the bone via Sharpey fibres and a gradually stiffer and stronger fibrocartilagenous layer will often involve bone avulsion. In contrast, the tendons or ligaments that insert tangentially into the periosteum over a wide area fail gradually and without bone avulsion. Part of the explanation for this gradual or partial ligament failure is that the deep fibres of these ligaments are relatively short in relation to the superficial fibres. The effect of this is that when the whole ligament is stretched the deeper fibres will, relative to their length, be stretched more and ultimately these fibres will fail first.

Mechanical properties The primary function of both tendons and ligaments is to carry tensile loads and, therefore, it is the tensile mechanical properties that are most relevant. The force–extension graph of tendons or ligaments is similar to that shown in Fig. 1, in which the curve has been split into a number of different zones of behaviour.14 In zone 1, also called the toe-region, the structure is seen to elongate easily. However, as the elongation increases, so the gradient of the curve steepens, representing a stiffening of the

3

4

5 Load

established when Fairbank11 reported roentgenographic signs of post-meniscectomy knee joint degeneration. Since then many studies have reported on the importance of the menisci as shock absorbers and how total and/or partial tears increase the contact pressures on the articular cartilage. However, many issues remain concerning the mechanical effects of meniscal repair. Generally speaking, repairs in the vascular peripheral third of the menisci are reasonably promising, but repairs of the avascular inner parts of the menisci have not been very successful.12 In some cases, the meniscus is so damaged that replacement is indicated. Allogeneic or artificial replacements show some promise, but it is difficult to reproduce the complex biomechanical behaviour of the porous, permeable, biphasic and anisotropic menisci, and there is still much research and development to be done.

U. Hansen et al.

2

6

1 Extension

Figure 1 Load–extension curve for a ligament or tendon.

ARTICLE IN PRESS Material properties of biological tissues related to joint surgery tissue. This behaviour, which is common to collagenous structures, is explained by the crimped collagenous fibre micro-structure being stretched. Once the crimped structure has straightened out, further extension leads to elastic elongation of the fibres. This behaviour corresponds to zone 2, which shows the elongation response to increase linearly with load. The transition from zone 1 to zone 2 occurs at around 1% strain for tendons and 5% strain for ligaments. The reason for this relatively large difference is due to the very unidirectional orientation of the fibres of tendons as compared to ligaments. Because the fibre bundles of the ligament have different orientations and some may be slack initially, they are only gradually recruited to resist the elongation of the ligament, the effect being an extension of the toe-region. In zone 3 some of the length change represents damage to the tissue and is relatively permanent. In this region, the tissue has difficulty resisting further load and eventually sufficient damage has accumulated to cause sudden, catastrophic failure (point 4). The strength of ligaments and tendons generally depends on the density of collagen and a strength of 80 MPa has been measured for some tendons15 while ligaments may fail at 20 Mpa.16 The strength of the tissue is perhaps best put in perspective by comparing it with the strength of mild steel, which yields around 200 MPa. In Fig. 1 is also shown a zone 5 displaying a series of small load drops. This zone represents the mechanical behaviour exhibited by some ligaments. This behaviour is caused by the non-uniformity of the orientation and tension of the fibre bundles in certain ligaments. The non-uniformity results in some of the fibre bundles being tensioned more than others and the load drops are caused by sequential localised failures of particular fibre bundles reaching their ultimate strain before the rest of the structure. Finally, the graph shows a zone 6 of very low resistance to elongation after rupture has occurred. The resistance to deformation in this state stems from the load required to slide the fibres past each other. The appearance of the ligament in this state is relatively normal and if a ligament is arthroscopically assessed without also pulling it to assess its mechanical resistance, the erroneous conclusion might be that the tissue is intact.

Age and disuse effects Age and disuse dramatically change the mechanical properties of the bulk tissue of ligaments and tendons, as well as of the structures in and around the attachments. Noyes and Grood17 found that the strength of ACL specimens from people aged 16–26 years were 2.4 times stronger than those from people aged 48–83 years. Disuse may not dramatically alter the strength of the relatively avascular tendon, but the strength of the construct is also limited by the underlying bone. Laros et al.18 showed that 9 weeks in a plaster cast caused the medial collateral ligament of the canine stifle (‘knee’) joint to lose 39% of its strength, principally due to subperiosteal bone resorption. Although muscle weight returned to normal after only 12–18 weeks, bone mineralisation was not complete at 24 weeks and ligament

21 strength was still not normal at 30 weeks. This has obvious implications for postoperative rehabilitation.

Ligaments as joint stabilisers So far, we have considered the ligaments in isolation. However, their main function is to stabilise the joints and it seems relevant to consider their function as an integral part of the whole joint. Because ligaments are passive tensile structures, they can only have a role if the bone moves so as to tense them. A consideration of the positions of ligaments will to a large extent reveal their function in the joint. Ligaments aligned closely with the axis of rotation of a particular movement cannot prevent rotation around the axis but are more likely stabilisers of the relative translation between joint surfaces. In contrast, ligaments away from the axis of rotation benefit from the mechanical advantage of a large lever arm and will be much more effective stabilisers against joint rotations. Hence, the ACL does not play an important role in preventing torsion14 and its main function is not that of a primary constraint against varus or valgus opening of the joint—the collateral ligaments are much better placed for these functions. The anterior band of the ACL has been found to be tight throughout normal motion.13 Ligaments can only withstand relatively small cyclic strains (at the most a few millimetres of extension) so the anterior part of the ACL does not undergo any great length changes during knee motion. Therefore, the primary role of the anterior parts of the ACL is not that of a restraint to limit the range of motion, rather its function is that of controlling anterior subluxation during normal joint motion and, together with the posterior band, to prevent hyperextension, thus guiding the tibial kinematics as the knee flexes and extends.

Conclusion Some structures, for example, the anterior intermeniscal ligament, are only found in some patients and are possibly evolutionary remnants that can be ignored or removed when performing an orthopaedic procedure. However, the overall impression when studying the structure–function relationship of the tissues of the human body is that of an ingenious construction in which the initial assumption should be that all tissues perform an important role. Of course, the history of meniscal surgery and the consequence of ignoring its function is the most prominent example of this. However, the lesson is a general one and it is the hope of the authors that this paper will guide future surgeons toward a prudent respect for the importance of even seemingly inconsequential tissues.

References 1. Rice JC, Cowin SC, Bowman JA. On the dependence of the elasticity and strength of cancellous bone on apparent density. J Biomech 1988;21:155–68. 2. Cater DR, Hayes WC. The compressive behaviour of bone as a two-phase porous structure. J Bone Joint Surg (Am) 1977; 59A(7):954–62.

ARTICLE IN PRESS 22 3. Reilly DT, Burstein AH. The elatic and ultimate properties of compact bone tissue. J Biomech 1975;8:393–6. 4. Fithian DC, Kelly MA, Mow VC. Material properties and structure–function relationships in the menisci. Clin Orthop Relat Res 1990;252:19–31. 5. Mow VC, Holmes MH, Lai WM. Fluid transport and mechanical properties of articular cartilage: a review. J Biomech 1984;17: 377–94. 6. Stolz MA, Ateshian GA. Experimental verification and theoretical prediction of cartilage interstitial fluid pressurization at an impermeable contact interface in confined compression. J Biomech 1998;31:927–34. 7. Seedhom BB, Hargreaves DJ. Transmission of the load in the knee joint with special reference to the role of the menisci. Part II: experimental results, discussion and conclusions. Eng Med 1979;4:220–8. 8. Weightman BO. Tensile fatigue of human articular cartilage. J Biomech 1976;9:193–200. 9. Bellucci G, Seedhom BB. Tensile fatigue behaviour of human cartilage. Biorheology 2002;39:193–9. 10. Vincent TL, Hermansson MA, Hansen U, Amis AA, Saklatvala J. Basic fibroblast growth factor mediates transduction of mechanical signals when articular cartilage is loaded. Arthritis Rheum 2004;50(2):526–33.

U. Hansen et al. 11. Fairbank TJ. Knee joint changes after menisectomy. J Bone Joint Surg 1948;30B:664–70. 12. Messner K, Gao J. The menisci of the knee joint. Anatomical and functional characteristics, and a rational for clinical treatment. J Anat 1998;193:161–78. 13. Amis AA, Zavras TD. Isometricity and graft placement during ACL reconstruction. Knee 1995;2:5–17. 14. Amis AA. The Biomechanics of ligaments. In: Jenkins DHR, editor. Ligament injuries and their treatment. London: Chapman & Hall; 1985. p. 3–28. 15. Pring DJ, Amis AA, Coombes RRC. The mechanical properties of human flexor tendons in relation to artificial tendons. J Hand Surg (Br) 1985;10B:331–6. 16. Race A, Amis AA. The mechanical properties of the two bundles of the human posterior cruciate ligament. J Biomech 1994;27(1):13–24. 17. Noyes FR, Grood ES. The strength of the anterior cruciate ligaments in humans and rhesus monkeys: age-related and species-related changes. J Bone Joint Surg (Am) 1976;58A: 1074–82. 18. Laros GS, Tipton CM, Cooper RR. Influence of physical activity on ligament insertions in the knee of dogs. J Bone Joint Surg (Am) 1971;53A:275–86.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 23–31

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: BIOMECHANICS FOR THE FRCS ORTH EXAM

(iv) Basic biomechanics of human joints: Hips, knees and the spine T.D. Stewarta,, R.M. Halla,b a

Institute of Medical and Biological Engineering, c/o School of Mechanical Engineering, The University of Leeds, Leeds LS2 9JT, UK b Academic Unit of Orthopaedic Surgery, School of Medicine, The University of Leeds, Leeds LS2 9JT, UK

KEYWORDS Biomechanics; Hip; Knee; Spine

Summary The paper provides a basic introduction to the biomechanics of the hip, knee and spine with respect to the healthy joint and following joint replacement. The content is aimed specifically at persons with a medical background to introduce them to the concepts of forces and moments in application to the human body. & 2006 Elsevier Ltd. All rights reserved.

Basic principles of mechanics (forces and moments) Forces and moments (torque) can be described by referring to the child’s seesaw in Fig. 1. Standing stationary a child will exert a force onto the ground which is equal to the product of the child’s mass and the acceleration due to gravity (9.81 m/s2). Therefore, a child of 30 kg in mass would exert a downward force of 294.3 N to the ground (their weight). Note that if the child jumped onto the ground the acceleration term in the equation (force ¼ mass
acceleration) would increase, as would the force. A force acting on a body can produce rotation as well as translation. This rotation is caused by a rotational torque, also referred to as a moment. A moment is caused when a force acts at a particular distance from a point of reference. This point of reference may be a fixed axis of rotation as in a Corresponding author. Tel.: +44 0113 343 2133;

fax: .+44 0113 242 4611 E-mail address: [email protected] (T.D. Stewart). 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2005.12.004

ceramic-on-ceramic artificial hip replacement or variable as is the case in the functional spinal unit or knee. The magnitude of the moment is calculated by multiplying the force by the perpendicular distance from the line of action of the force vector to the reference point. The seesaw example below explains the basic principles. If Child A sits on the left end of the seesaw the bar will rotate downward, in the counterclockwise direction. In engineering terms, counterclockwise rotation is termed a positive moment, or positive torque. The magnitude of this torque would be equal to the product of the weight of Child A (294.3 N) multiplied by the distance (a) the child is sitting from the centre of rotation of the beam (2 m), in this case the magnitude of the moment would be equal to 588.6 Nm (294.3 N
2 m). The distance itself is also termed the moment or lever arm. A moment or torque is defined by a magnitude and a direction and is, therefore, a vector. If a heavier child (Child B 60 kg in mass) sits on the other side of the seesaw the beam will then rotate in the clockwise direction. For equilibrium to occur the sum of the moments acting on the bar must be equal to zero. Therefore, for the bar to be

ARTICLE IN PRESS 24

T.D. Stewart, R.M. Hall

balanced the product of the weight of Child A and distance a must be equal to the product of the weight of Child B and the distance b. As Child B is twice the weight of Child A this means that Child B must slide up the bar until his moment arm (distance b) is equal to one-half of the moment arm (distance a) of Child A for the seesaw to function effectively. The moments would then be said to be equal in magnitude but opposite in direction. To calculate the reaction force at the pivot of the seesaw you can add up all of the forces acting in the vertical direction. This reaction force would then be equal to the combined weight of the two children.

The hip The bony structures and ligaments of the natural hip create essentially a ball-in-socket joint. This structure limits anterior/posterior, and medial/lateral translation as well as subluxation (dislocation); however, it does not generally limit the range of motion of the hip during normal daily activities. The allowable range of motion is shown in Table 1 when compared to a selection of daily activities. The range of motion of the hip is far greater than what is required for normal activities, such as walking. This means that the surrounding bone and ligaments of the hip joint do not provide any rotational stability to the hip joint during the walking cycle and, therefore, this stability is provided almost entirely by the action of muscle forces. Principles of simple static mechanics can be used to analyse the loading applied within the body. Figure 2 shows a very simple twodimensional (2D) schematic of the leg at the heel-strike phase of gait. Contact with the ground produces a ground reaction force equal to the proportion of the person’s mass transferred to the ground multiplied by the acceleration of this mass (gravitational acceleration+linear acceleration). The ground reaction force at heel-strike can be measured experimentally using a simple force platform and was reported by Bassey et al.6 to be in the region of their body weight (BW). As the knee is fully extended at heel-strike the leg can be analysed in a similar manner to the seesaw as shown in Fig. 2 with the hip joint acting as the pivot. The ground reaction force acting at the foot (Fgr) produces a counterclockwise (positive) moment about the hip centre

Gr Hs

Fh Hamstrings Muscle Force

Child B

Child A

a

Figure 1

Table 1

equal to Fgr
Gr. The moment arm Gr is equal to the length of the leg (L) multiplied by the sine of the flexion angle (301); Gr is, therefore, equal to 0.5L. Thus, the ground reaction force (Fgr) produces a flexion moment of BW
0.5L. The flexion moment is balanced by the extensor muscles including the gluteus maximus and the hamstrings which act to stabilise the hip at heel-strike producing a counterclockwise negative moment. In this example, for simplicity, the hamstrings muscle alone has been considered. The moment produced by the hamstrings muscle is equal to the magnitude of the hamstrings muscle force (Fh) multiplied by the distance from the line of action of the muscle to the hip centre. Thus, the moment arm will vary from person to person; however, its value will be approximated in this example to be equal to 0.15L. Therefore, the moment produced by the hamstrings muscle group would be equal to Fh
0.15L. For the hip to be stable the ground reaction moment (BW
0.5L) must be approximately equal to the moment produced by the extensor muscles, in this case only the hamstrings. Therefore, the hamstrings muscle force Fh at gross approximation would be equal to 3.3 times BW. The vertical reaction force at the hip can be calculated by summing the forces acting in the vertical direction from Fig. 2. This includes the ground reaction force Fgr and the vertical component of the hamstrings muscle force vector Fh cos(30+y1), where y is the angle between the hamstrings muscle force vector and the line of action of the femur. In this case, y is assumed to be equal to 171. Therefore, for this simple example, the vertical reaction force at the hip is

Fgr Ground Reaction Force

b

Schematic of a seesaw.

Figure 2

Simplified 2D kinetics of heel-strike.

Range of motion (degrees) in the hip compared to daily activities.

Flexion/extension Internal/external rotation Abduction/adduction

Allowable1

Walking2,3

Tie Shoe4

Stairs5

140/30 90/90 90/30

30/15 4/9 7/5

129 18 abd. 13 ext.

40

ARTICLE IN PRESS Basic biomechanics of human joints: Hips, knees and the spine

25

3.00

2.50 Swing

Stance

Load (kN)

2.00

1.50

1.00

0.50

0.00 0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Time (s)

Figure 3 Vertical reaction force at the hip used for simulation of gait in wear simulators.

Body Weight

AB1 AB2

BW

Offset

Figure 4 Simplified schematic of standing demonstrating the concept of femoral offset.

equal to BW+3.3BW
0.68 for a total of 3.25BW. This 2D calculation is very crude; however, it demonstrates the general methods which can be utilised to analyse forces and moments in the body. It also demonstrates the inefficiency of our muscle forces due to the nature of our relatively long slender limbs and the resulting short muscular moment arms. The vertical hip joint reaction force during walking is shown in Fig. 3. The walking cycle is characterised by two peaks of load at heel-strike and at toe-off which generally range from 3–4 times our body weight7. Between these two loading peaks the body’s mass (head and torso) is moving smoothly and is not translated vertically a large amount. As such the reaction force at the hip between heel-strike and toe-off is relatively small and in the region of body weight. At toe-off the hip is extended 15 degrees. The quadriceps muscle acts to stabilise the knee whilst the gastrocnemius muscle produces plantarflexion at the ankle. These muscle forces combine to accelerate the body forward producing the second peak of load in the reaction force curve.

Hip replacement As the wear of replacement joints has improved over the past 20 years, correct positioning of the components during hip replacement is arguably the most important factor for the success of modern total hip prostheses. To restore normal function in the hip joint, an important biomechanical consideration in total hip replacement is the femoral offset. Normal function is in itself an arguable quantity as what one patient would consider normal may limit the activities of another. The femoral offset is the distance from the centre of rotation of the hip joint to the line of action of the femur, as outlined in Fig. 4.8 To understand the importance of offset consider the example shown below of a person standing on one leg whilst their body remains vertical. The BW of the person will create a clockwise (negative) moment or torque about the centre of the hip. This torque must be balanced by an equal but opposite counterclockwise (positive) moment produced by the abductor muscle force. The moment arm of the abductor muscle is directly related to the magnitude of

ARTICLE IN PRESS 26

T.D. Stewart, R.M. Hall

the femoral offset. As the offset increases in length the force required by the abductor muscles to balance the BW moment would reduce, thereby increasing the efficiency of the abductors. The reaction force at the hip would also decrease in this case since the sum of the forces acting in the vertical direction would reduce with the smaller abductor muscle force. Increasing the femoral offset may result in increased stress transferred to the femoral component and its fixation due to the larger bending moment produced by a longer neck length, however, as this is combined with a reduced joint load the overall effect on the stress distribution would be more complex with greater bending stress, but, reduced normal stress, a topic which is beyond the scope of this paper.

The knee The knee joint consists of two articulating joints the tibiofemoral joint and the patello-femoral joint. Unlike the ballin-socket geometry of the hip the femoral and tibial surfaces of the knee are not a close fit to one another. The variation

Table 2

in geometry allows a wide range of motion to occur which allows us to complete various daily activities. The allowable range of motion in the knee is shown in Table 2 when compared to a selection of daily activities. The corresponding reaction forces at the knee during walking are shown in Fig. 5. The reaction forces must be considered in parallel to the flexion–extension of the knee (Fig. 6) in order to fully understand their significance towards the characteristic three peaks of load which occur during walking. At heel-strike contact with the ground produces a flexion moment at the hip, and an extension moment at the knee, both of which are resisted by the hamstrings muscle. At this position, the knee is fully extended and there is a loading peak across the joint upon impact of 2–4 times BW, primarily due to muscle forces acting to stabilise the knee.9,12 When our heel hits the ground the knee is in its most stable position due to three factors.1,10 The medial/lateral spacing of the femoral condyles is the least when the knee is fully extended. At this position, the condyles tighten against the intercondylar notch (tibial spine) thereby providing bony stability. The radius of curvature of the femoral condyles is

Range of motion (degrees) in the knee compared to daily activities.

Flexion/extension Internal/external rotation Abduction/adduction Rollback (M/L)

Allowable1

Walking9,10

Sitting11

Stairs5

150/51 76/7301 0–101 5/15 mm

70/0 7101 01 8

100–1201

70–901

3.00

2.50

Load (kN)

2.00

1.50

1.00 Swing

Stance 0.50

0.00 0

20

40

60

80

100

Percentage of Gait Cycle

Figure 5

Vertical tibio-femoral knee reaction force used for simulation of gait in wear simulators.

ARTICLE IN PRESS Basic biomechanics of human joints: Hips, knees and the spine

27

60

50

Flexion Angle (degrees) Internal Rotation (degrees) Posterior Displacement (mm)

40

30 Swing

Stance 20

10

0 0

20

40

60

80

100

-10

-20

Figure 6

Percentage of Gait Cycle

Motion of the knee used for simulation of gait in wear simulators.

largest when the knee is fully extended. The tibial plateau is also sloped anteriorly and the combination of these two factors pull the collateral ligaments taught at extension. The result of the bony structures and taught ligaments creates a structure that is rotationally very stable at heel-strike. An additional factor to consider at heel-strike is the translational stability of the knee. When the foot contacts the ground a natural anterior translation of the femur with respect to the tibia occurs. Anterior translation of the femur is restricted by the posterior cruciate ligament which prevents forward dislocation of the bearing. Unlike the hip, stability in the knee at heel-strike is, therefore, provided by a combination of bony structures, ligaments and muscle forces. As walking progresses into mid-stance the knee begins to flex due to actions of the hamstrings muscles and the femoral condyles begin to roll on the tibial surface. This produces a natural posterior translation of the contact with the tibia along with an external rotation (51) of the knee since the lateral femoral condyle has a larger radius and rolls further. The knee flexes to 201 (hamstrings) and then extends (quadriceps) back to 01.1,10 At the change in direction from flexion to extension a second loading peak occurs in the knee as the muscle forces act to stabilise the joint. During mid-stance the knee is less stable since the medial/lateral spacing of the femoral condyles is larger and they no longer lock against the intercondylar notch. In addition to this, the radius of curvature of the femoral condyles during flexion is decreasing, and the natural rollback in the knee is pushing the contact down the posterior sloping tibial plateau which brings the femur closer to the tibia and increases the laxity in the ligaments. This laxity allows relative rotation between the femoral and tibial surfaces to occur. As walking continues the knee flexes once again. Rolling occurs up to 201 flexion until at which time the posterior

translation of the femur relative to the tibia that occurs during the rolling action is restricted by the anterior cruciate ligament which acts like an anchor preventing posterior translation. As flexion continues and the condyles can no longer roll backward on the tibia the motion between the femoral and tibial surfaces changes from rolling to sliding. This change occurs initially with the medial condyle and then the lateral condyle resulting in a natural external rotation of the knee. At toe-off (501 flexion) the quadriceps calf muscle (gastrocnemius) act to both stabilise the knee joint and produce plantar flexion of the ankle joint which accelerates the body forward resulting in a vertical joint reaction force of 2–4 times BW.9 A recent paper by Freeman and Pinskerova12 is recommended for further reading.

Knee replacement Much like the hip correct positioning of the components during knee replacement is also vital in its success. The general alignment issues of the hip all hold true for the knee with even greater importance since the knee has a less conforming geometry and high stresses can lead to accelerated wear and delamination.13 Among the important factors for knee replacement positioning include rollback, tibial position/size, varus/valgus positioning and lift-off. Rollback The posterior shift in the contact between the femur and the tibia during flexion (rollback) increases the moment arm and efficiency of the quadriceps muscle.14 This is outlined in the schematic of Fig. 7. As flexion occurs, such as during a squat, the efficiency of the quadriceps muscle force will be directly related to tension and moment arm (PT) of the

ARTICLE IN PRESS 28

T.D. Stewart, R.M. Hall

patella tendon. Clearly, without rollback the moment arm is short and a larger force is required by the muscle to produce the same action. While rollback occurs naturally in the healthy knee, replacement joints are sometimes designed with a cam arrangement to encourage natural motion. Knee biomechanics following total knee replacement has been shown to be highly variable in clinical studies, with great patient variability.14–17 Tibial positioning/size Incomplete support by the tibial cortex (cortical bone) for the tibial tray may lead to subsidence of the tibial tray if cancellous bone quality is poor.18 Generally, anterior medial and posterior lateral coverage is recommended when sizing components to prevent subsidence. An example of subsidence is shown in Fig. 8. The tibial insert in this case was lateralised as the patient had a valgus deformity and a paraQuadriceps

Femur

Femur “PT”

“PT”

patellar arthrotomy was conducted. This left very poor quality cancellous bone to support the medial side which subsided. Complete coverage is difficult due to the anatomical nature of the tibial plateau which extends further posteriorly on the medial side. Overhang of the tibial insert is equally as important as undersizing. In some areas an overhanging tray can lead to impingement with the ligaments and tendons surrounding the knee and cause discomfort, pain and the need for revision. Varus/valgus positioning and lift-off In the natural knee, it is generally considered that 60% of the load is transferred through the medial condyle while 40% is transferred through the lateral condyle. In a knee with valgus deformity, the biomechanics and of the knee have changed significantly so that the majority of the load is transferred through the medial condyle. In contrast in a knee with a varus deformity, the majority of the load is transferred through the lateral condyle. Following knee replacement the loading on the tibial insert should be restored to as normal as possible. Edge loading of the polyethylene tibial insert caused by a varus/valgus deformity or by condylar lift-off has been shown in in vitro studies to cause accelerated wear of the polyethylene and should, therefore, be avoided.19,20

Joint loading in the hip and knee Reaction Force

Reaction Force

Patella Tendon

Figure 7 Schematic of the effect of rollback on the quadriceps moment arm (PT).

There are two common methods for measuring the load acting on our joints. Inverse dynamics uses simple engineering mechanics as illustrated in Fig. 2 in the application of the body. The second method of measuring joint forces is to utilise instrumented prostheses.22 These are custom designed implants which contain complex instrumentation that

Anterior 0 330 320 310 300

340 350

30 40

30

50 60

20

290

Lateral

10 20

40

280

10

270

0

70 80 90

260

Medial

100

250

110

240 230 220 210

200 190

170 160

120 130 140 150

180 Posterior L3-91 L3-97

L3-92 L3-98

L3-93 L3-99

L3-94 L3-100

L3-95 Average

L3-96 Cad file 3

Figure 8 Medial tibial collapse (left) of a lateralised tibial component. Typical coverage of a tibial tray (dark line) and average patient data (right).

ARTICLE IN PRESS Basic biomechanics of human joints: Hips, knees and the spine

Table 3

Joint loading in the hip and knee.

Activity

Reference

Walking

Stair ascent/descent

Rising from a chair Rising from a squat

29

Hip load 12

Freeman and Pinskerova Paul7 Bergmann et al.22 (in vivo normal) Bergmann et al.22 (in vivo defective)

3–4 BW 2–3 BW 3–4 BW

Freeman and Pinskerova12 Costigan et al.23 Paul7 Bergmann et al.22 (in vivo normal) Bergmann et al.22 (in vivo defective)

2–3.5 BW 3.5–5 BW

Ellis et al.11 Bergmann et al.22 (in vivo normal)

1.75–2.25 BW

Dahlkvist et al.21 Paul7

directly records the load acting on the joints in-vivo. Ethical approval is required for the use of these devices. Typical loads measured within the body during various activities for the hip and knee are shown in Table 3. Kinesiology (inverse dynamics) has a tendency to overestimate joint forces due to assumptions related to the actions of muscles. In contrast, instrumented prostheses may produce more accurate results; however, there is a vast difference in the biomechanics from patient to patient. Therefore, to achieve an average value for a given population the sample must be very large to overcome the variability. The results from inverse dynamics and from instrumented prostheses are reasonably close and as long as you understand the limitations used in the analysis—either method is a very useful and valuable tool.

The spine The spine, comprising three joints at any level—a disc and two facet joints, is arguably the most complex and demanding of any joint systems within the human body. Harms and Tabasso24 noted the importance of restoring the normal biomechanical environment as far as is possible during a surgical intervention and have proposed four key biomechanical functions for the spine (Table 4). The key feature of this list is that the functions are listed in ascending order of importance and with the clinician give a set of principles by which an intervention should be approached. For example, the overriding concern is protection of the spinal cord which takes precedence over other considerations. If the functioning of the spinal cord can be assured then the stability of the spine should be the next important consideration as is the case in fusion surgery. This is achieved at the expense of the segment’s mobility.

Loads in the spine As previously noted the spine, particularly the lumbar segment, experiences arguably the most demanding biomechanical environment of any joint system in the human body.

Knee load 3.4 BW 3 BW

4.8/4.3 BW 3–6 BW 4.4/4.9 BW

3.2 BW 4.7–5.6 BW 4.2 BW

Table 4 1 2 3 4

Key biomechanical functions of the spine.24 Protection of the spinal cord Maintenance of trunk stability To provide mobility Aid movement in the upper and lower limbs

This is because of the high loads that must be sustained, and the complex neuromuscular control that is required to maintain a stable yet mobile unit. The limits of the compressive loads within the spine are defined, principally, by the axial strength of the individual vertebrae (Fig. 9) that rise from approximately 1300 N at C3 (the third cervical vertebra) to over 8000 N at L4 (the fourth lumbar vertebra).25 Whilst a considerable margin of safety is built into these failure strengths the loads observed in the spine are often several times BW. The only time the compressive load is less than BW is in mainly the prone position. The compressive forces arise largely from the muscle action that produces a counterbalancing moment to the weight of the upper torso and/or head that acts forward of the spine (Fig. 10) in a manner similar to the example given in Fig. 1. The posterior muscles have a relatively small moment or lever arm (b is typically 5 cm or less) and, therefore, have to produce a considerably larger force than the weight of the upper torso to produce a counterbalancing moment. The compressive load on the spine at that level is just the addition of the weight of the upper torso and the force generated by the posterior muscles. The axes of rotation of a given functional spinal unit, which is the disc and the adjacent two vertebrae, are generally located in or just below the disc, but the exact position will vary according to the type of motion being undertaken, the position of the spine and the nature of the individual as well as the functional spinal unit being observed. This simple figure can be amended to include a person carrying a weight in front of them. In this case, the posterior muscles will have to counterbalance an additional torque and therefore produce an even greater force to resist the forward flexing moment.

ARTICLE IN PRESS 30

T.D. Stewart, R.M. Hall 350 Force (% Bodyweight)

9000 8000

6000

250 200 150 100 50 0 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95100

5000

Stride (%)

Force (% Bodyweight)

3000 2000 1000 0

C3

T1

T5

T11

L1

Anterior

20

4000

15 10 5 0

Posterior

Compressive strength (N)

7000

300

-5

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95100

-10 -15 Stride (%)

L4 15

Weight of the Upper Torso

a b

Figure 10 A counterbalancing moment produced by the muscles posterior to the spine acts to prevent the body flexing forward in response to the weight of the upper body.

During level walking, the peak compressive forces developed across L4–L5 fluctuate between 1.5 and 3 times BW,26 and they do so at a frequency twice that of the gait cycle (Fig. 11). Impressively high transient loads (5–6 times BW or greater27–34) occur for more extreme forms of activity or at the extremes of motion, with trunk flexion and/or high external loading being especially demanding. However, substantial loads are sustained even during ‘inactively’ standing or sitting (1.5 and 2.0 times BW, respectively35). Also of importance are the observations of the shear forces in the lower lumbar spine. These loads are a source of much debate, in particular, on whether total disc replacements (TDRs) should be constrained or not, together with the effects these design features have on the adjacent structures, in particular, the facet joints. The important antero-posterior shear loads, which have been inferred from indirect measurement, can exceed 140 N during normal walking to more than 1000 N in more extreme activities.26,27,32–34 When considering any surgical intervention the clinician must also ensure that the forces and moments are distributed physiologically between the different structures within the spine with special reference to the facet joints. Typically, in the neutral position, 80% of the compressive load passes through the vertebral bodies and 20% through the facets in the lumbar spine, whilst in the cervical spine a greater proportion passes through the anterior column.

10 5 0

Right

Action of the Muscles Posterior to the Spine

Left

Strengths of vertebral bodies in different spinal Force (% Bodyweight)

Figure 9 regions.

-5

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95100

-10 -15 Stride (%)

Figure 11 Typical loads observed in L4–L5 functional spinal unit. The uppermost diagram outlines the axial compressive load, the middle the antero-posterior shear load and the lowest the lateral shear.26 Note the double peaks in both the compressive and antero-posterior shear loads that follow heelstrike of both the left and right foot during one gait cycle.

Motions in the spine Spinal motion is difficult to measure due to the segmented nature of the spine and the exceedingly high number of degrees of freedom together with elevated degree of redundancy that this allows. This must be coupled with issues of accessibility that may preclude routine non-clinical radiographic imaging of the spine with which to observe the behaviour of individual functional spinal units. Segmental motion assessment is more common in which surface measurement type devices are used. However, these have a number of shortcomings in that they are prone to the usual uncertainties found to be common with the use of surface markers but also the fact that the translational motion is often excluded from the analysis, which suggests that such motion is unimportant within the spine. Observations show that the overall range of motion in flexion/extension varies enormously throughout the spine and varies within anatomical segments as well as between them. Combined flexion–extension is relatively high in both the cervical and lumbar regions exceeding 101 whilst a minimum is observed in the upper and mid-thoracic regions. The range of motion for lateral flexion is more even throughout the spine with values typically between 51 and 101. The lower thoracic and lumbar spines are characterised by a relatively small range of motion in axial rotation which is typically of the order of 31, which results from the

ARTICLE IN PRESS Basic biomechanics of human joints: Hips, knees and the spine orientation of the facets which hinder this type of motion.25 During gait, the peak-to-peak flexion/extension motion in the lumbar spine increases with strenuousness of cadence, although the range utilised (typically 3–41) remains only a modest fraction of that fully available. As noted previously, further complexity arises from the fact that the motions of the lower lumbar spine are kinematically coupled: most significantly, flexion/extension results in translational motion, thus causing the instantaneous axis of rotation to migrate. These coupled translations are of the order of 1–2 mm for L4–L5 and 0.5–1 mm for L5–S1,36,37 but they are highly variable due to the differences in functional spinal unit recruitment patterns occurring in forward versus backward trunk bending between individuals.26,38–40

References 1. Kapandji IA. The physiology of the joints. In: Lower limb, vol. 2. Churchill Livingstone; 1987, ISBN 0 443 03618 7. 2. Johnson RC, Smidt GL. Measurement of hip-joint motion during walking. J Bone Joint Surg 1969;51-A(6):1083–94. 3. Heller MO, Bergmann G, Deuretzbacher G, Durselen L, Pohl M, Claes L, et al. Musculoskeletal loading conditions at the hip during walking and stair climbing. J Biomech 2001;34:883–93. 4. D’Lima DD, Urquhart AG, Buehler KO, Walker RH, Colwell CW. The effect of the orientation of the acetabular and femoral components on the range of motion of the hip at different headneck ratios. J Bone Joint Surg 2000;82A:315–21. 5. Andriacchi TP, Anderson GBJ, Fermier RW, Stern D, Galante JO. A study of lower limb mechanics during stair climbing. J Bone Joint Surg 1980;62A:749–57. 6. Bassey EJ, Littlewood JJ, Taylor SJG. Relations between compressive axial forces in an instrumented massive femoral implant, ground reaction forces, and integrated electromyographs from vastus lateralis during various ‘‘osteogenic’’ exercises. J Biomech 1997;30(3):213–33. 7. Paul JP. Forces transmitted by joints in the human body. Proc Inst Mech Eng 1966;181:8–15. 8. Charles MN, Bourne RB, Davey R, Greenwald AS, Morrey BF, Rorabeck CH. Soft-tissue balancing of the hip—the role of femoral offset restoration. J Bone Joint Surg 2004;86-A:1078–88. 9. Seedhom BB, Longton EB, Dowson D, Wright V. Designing a total knee prostheses. Eng Med 1972;1(2):28–32. 10. Palastanga N, Field D, Soames R. Anatomy and human movement—structure and function. Butterworth Heinemann; 1989. 11. Ellis MI, Seedhom BB, Amis AA, Dowson D, Wright V. Forces in the knee joint whilst rising from normal and motorised chairs. J Eng Med 1979;8(1):33–40. 12. Freeman MAR, Pinskerova V. The movement of the normal tibiofemoral joint. J Biomech 2005;38:197–208. 13. Fisher J, McEwen HMJ, Barnett PI, Bell CJ, Stewart TD, Stone MH, et al. Wear of polyethylene in artificial knee joints. Curr Orthop 2001;15:399–405. 14. Most E, Zayontz S, Li G, Otterberg E, Sabbag K, Rubash HE. Femoral rollback after cruciate retaining and stabilizing total knee arthroplasty. Clin Orthop Relat Res 2003;410:101–13. 15. D’Lima DD, Trice M, Uuquhart AG, Colwell CW. Comparison between the kinematics of fixed and rotating bearing knee prostheses. Clin Orthop Relat Res 2000:151–7. 16. Dennis DD, Komistek RD, Colwell CE, Ranawat CR, Thornhimm TS, Lapp MA. In vivo anterior posterior femorotibial translation of total knee arthroplasty. Clin Orthop Relat Res 1998; 356:47–57. 17. D’Lima DD, Poole C, Chadha H, Hermida JC, Mahar A, Colwell CW. Quadriceps moment arm and quadriceps forces after total knee arthroplasty. Clin Orthop Relat Res 2001;392:213–20.

31 18. Castle TH, Noyes FR, Grood ES. Posterior tibial subluxation of the posterior-cruciate deficient knee. Clin Orthop Relat Res 1990:193–202. 19. Jennings LM, Bell CB, Ingham E, Komistek R, Stome MH, Fisher J. The influence of femoral condylar lift-off on the wear of fixed bearing artificial knee joints. Proc Eur Soc Biomater 2004. 20. Insall JN, Scuderi GR, Komistek RD, Math K, Dennis DA, Anderson DT. Correlation between condylar lift-off and femoral component alignment. Clin Orthop Relat Res 2002;403: 143–52. 21. Dahlkvist NJ, Mayo P, Seedhom BB. Forces during squatting and rising from a deep squat. J Eng Med 1982;11:69–76. 22. Bergmann G, Deuretzbacher G, Heller MO, Graichen F, Rohlmann A, Strauss J, et al. Hip contact forces and gait patterns from routine activities. J Biomech 2001;34:859–71. 23. Costigan PA, Deluzio KJ, Wyss UP. Knee and hip kinetics during normal stair climbing. Gait Posture 2002;16:31–7. 24. Harms J, Tabasso G. Instrumented spinal surgery: principles and technique. Georg. Thieme Verlag; 1999. 25. White AA, Panjabi M. Clinical biomechanics of the spine, 2nd ed. Philadelphia: Lippincott Williams and Wilkins; 1990. 26. Callaghan JP, Patla AE, McGill SM. Low back three-dimensional joint forces, kinematics, and kinetics during walking. Clin Biomech 1999;14(3):203–16. 27. Ferguson SA, Gaudes-MacLaren LL, Marras WS, Waters TR, Davis KG. Spinal loading when lifting from industrial storage bins. Ergonomics 2002;45(6):399–414. 28. Rohlmannt A, Claes LE, Bergmannt G, Graichen F, Neef P, Wilke HJ. Comparison of intradiscal pressures and spinal fixator loads for different body positions and exercises. Ergonomics 2001; 44(8):781–94. 29. Rohlmann A, Arntz U, Graichen F, Bergmann G. Loads on an internal spinal fixation device during sitting. J Biomech 2001; 34(8):989–93. 30. Rohlmann A, Graichen F, Weber U, Bergmann G. 2000 Volvo Award winner in biomechanical studies: monitoring in vivo implant loads with a telemeterized internal spinal fixation device. Spine 2000;25(23):2981–6. 31. Rohlmann A, Bergmann G, Graichen F. Loads on internal spinal fixators measured in different body positions. Eur Spine J 1999; 8(5):354–9. 32. Kumar S, Moro L, Narayan Y. A biomechanical analysis of loads on X-ray technologists: a field study. Ergonomics 2003;46(5):502–17. 33. Fathallah FA, Marras WS, Parnianpour M. An assessment of complex spinal loads during dynamic lifting tasks. Spine 1998; 23(6):706–16. 34. Dennis GJ, Barrett RS. Spinal loads during individual and team lifting. Ergonomics 2002;45(10):671–81. 35. Callaghan JP, McGill SM. Low back joint loading and kinematics during standing and unsupported sitting. Ergonomics 2001; 44(3):280–94. 36. Frobin W, Brinckmann P, Leivseth G, Biggemann M, Relkeras O. Precision measurement of segmental motion from flexion–extension radiographs of the lumbar spine. Clin Biomech 1996;11(8):457–65. 37. McGregor AH, Anderton L, Gedroyc WM, Johnson J, Hughes SP. The use of interventional open MRI to assess the kinematics of the lumbar spine in patients with spondylolisthesis. Spine 2002; 27(14):1582–6. 38. Harada M, Abumi K, Ito M, Kaneda K. Cineradiographic motion analysis of normal lumbar spine during forward and backward flexion. Spine 2000;25(15):1932–7. 39. Gatton ML, Pearcy MJ. Kinematics and movement sequencing during flexion of the lumbar spine. Clin Biomech 1999;14(6): 376–83. 40. Pearcy MJ, Bogduk N. Instantaneous axes of rotation of the lumbar intervertebral joints. Spine 1988;13(9):1033–41.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 32–40

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: BIOMECHANICS FOR THE FRCS ORTH EXAM

(v) Biotribology Z.M. Jina,, M. Stoneb, E. Inghamc, J. Fishera a

Institute of Medical and Biological Engineering, School of Mechanical Engineering, University of Leeds, UK Institute of Medical and Biological Engineering, School of Biochemistry and Microbiology, University of Leeds, UK c Leeds Teaching Hospitals Trust b

KEYWORDS Tribology; Surfaces; Contact mechanics; Friction; Lubrication; Wear; Biotribology; Artificial hip joint; UHMWPE (ultra-high molecular weight polyethylene); Metal-on-metal; Ceramic-on-ceramic

Summary Basic principles of engineering tribology are briefly reviewed, in terms of surface metrology, contact mechanics, friction, lubrication and wear. In each of these topics, applications to artificial hip joints are discussed in detail. Various artificial hip joints with different bearing material combinations are considered, including ultra-high molecular weight polyethylene against metal or ceramic, metal-on-metal and ceramic-on-ceramic. & 2005 Published by Elsevier Ltd.

Introduction According to the Concise Oxford English Dictionary, ‘Tribo-’is derived from the Greek word ‘Tribos’, meaning rubbing and friction, and Tribology is ‘the study of friction, wear and lubrication, and design of bearings, science of interacting surfaces in relative motion’. Tribology was introduced in 1966 in the Jost Report (Lubrication (tribology) Education and Research, Department of Education and Science, HMSO, 1966) and was formally defined as ‘‘The science and technology of interacting surfaces in relative motion and Corresponding author. Tel.: +44 113 34 37471;

fax: +44 113 242 4611. E-mail address: [email protected] (Z.M. Jin). 0268-0890/$ - see front matter & 2005 Published by Elsevier Ltd. doi:10.1016/j.cuor.2005.09.005

the practices related thereto’’. In short, tribology deals with lubrication, friction and wear, which can be involved with a number of basic engineering subjects such as solid mechanics, fluid mechanics, lubricant chemistry, material science, heat transfer, etc. The importance of tribology in engineering is self-evident since virtually all engineering components and systems are involved with relative motion. Typical examples include ball bearings, gears, tyres, etc. The importance of tribology in biological systems is also clear. The term bio-tribology was introduced by Dowson and Wright1 in 1973 to cover ‘‘y all aspects of tribology related to biological systems’’. The best-known example of the subject is the numerous studies of natural synovial joint lubrication and the design, manufacture and performance of various forms of total joint replacements. Wear of bearing surfaces in humans and

ARTICLE IN PRESS Biotribology animals can result in pain and restricted movement. The consequences of excessive wear of the bearing material (articular cartilage) in synovial joints are well known. Typical examples of tribology applied to biology include:


Wear of dentures2,3
Friction of skin and garments, affecting the comfort of








clothes, socks and shoes,4,5 and slipperiness6,7 Tribology of contact lenses and ocular tribology8 Tribology at micro-levels—inside cells, vessels and capillaries such as lubrication by plasma of red blood cells in narrow capillaries9 The wear of replacement heart valves10 The lubrication of the pump in total artificial hearts11 The wear of screws and plates in bone fracture repair12 Lubrication in pericardium and pleural surfaces13 Tribology of natural synovial joints and artificial replacements.14,15

This review will focus on the tribology of artificial joints, in particular to review studies of surface, friction, lubrication and wear. It should be pointed out that tribological studies of bearing surfaces should ideally be considered in conjunction with biological studies of wear debris. This is particularly important in artificial joint replacements, since it is generally accepted that the major long-term factor limiting clinical outcome is loosening, caused by osteolysis and adverse tissue reactions to wear particles (see Ingham and Fisher16 for a recent review). Therefore, it is important to eliminate or, if that is impossible, to minimize wear. However, since biological reactions mainly depend upon the number and the size of wear particles, the wear volume as well as the size distribution of wear particles are important. For example, in hip implants using metal-onmetal bearings, the wear volume is generally greatly reduced compared with polyethylene-on-metal. However, the size of metallic wear debris is generally much smaller.

Surfaces All real surfaces are rough on microscopic scales. For example, typical atomic diameters are between 1 and 10 Å (1 Å ¼ 1010 m). The smoothest surface is achieved on mica, with irregularities in the order of 20 Å. The irregularities on quartz crystal are of the order of 100 Å. The smoothest bearing surface for artificial joints is usually found on ceramics, with irregularities in the region of 0.005 mm, while for metallic bearing surfaces, these are generally in the region of 0.01 mm. The most common parameter used to characterise the roughness of a surface is the arithmetical mean deviation (or average roughness or centre line average), which is denoted by (Ra ). The definition of other surface roughness parameters and application to artificial hip joints can be found elsewhere.17 Roughness parameters are usually measured with a contacting Talysurf profilometer or non-contacting white light interferometer. Typical (Ra ) values for surfaces produced by different engineering production processes relevant to orthopaedic implants are given in Table 1 (taken from Dowson and Wright18).

33

Table 1 Typical (Ra ) values for surfaces produced by different engineering production processes relevant to orthopaedic implants (taken from Dowson and Wright18). Production process

Ra (mm)

Sand casting Sawing Forging Drilling Turning Die casting Grinding (coarse) Grinding (fine) Polishing Super-polishing Super-finishing

12.5–25 3.2–25 3.2–12.5 1.6–6.3 0.4–6.3 0.8–1.6 0.4–1.6 0.1–0.4 0.05–0.4 0.025–0.2 0.005–0.01

Table 2 Comparison of (Ra ) values for surfaces between engineering and bioengineering applications (taken from Dowson and Wright18). Components

Ra (mm)

Plain bearings in turbines Rolling bearings in gear boxes Gears in engines Articular cartilage Endoprostheses (metal) Endoprostheses (plastic)

0.12–1.2 0.05–0.3 0.25–1 1–6 0.005–0.025 0.1–2.5

A comparison of the (Ra ) between engineering and bioengineering components is shown in Table 2. Typical (Ra ) values for various bearing surfaces used in current artificial hip joints are summarised in Table 3, as well as the composite surface roughness defined as qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Ra ¼ ðRa_Head Þ2 þ ðRa_Cup Þ2 . (1)

Friction Friction is loosely defined as the resistance to motion. Friction was first studied by Leonardo da Vinci (AD 1452–1519), (http://www.tribology-abc.com/abc/history.htm). There are three laws of dry friction: I. The force of friction (F) is directly proportional to the applied load (W) II. The force of friction (F) is independent of the apparent area of contact III. The kinetic force of friction (F) is independent of the sliding speed (V). The first two laws are attributed to Amontons who, in 1699, provided an explanation of friction as the work done to lift one surface over the roughness of the other, or from

ARTICLE IN PRESS 34

Z.M. Jin et al.

Table 3

Typical (Ra ) values for various bearing surfaces used in current artificial hip joints and their composite (Ra ) values.

Bearings

Femoral

Ra (mm)

Acetabular

Ra (mm)

Composite Ra (mm)

UHMWPE-on-metal Metal-on-metal Ceramic-on-ceramic

Cobalt chrome Cobalt chrome Alumina

0.01–0.025 0.005–0.025 0.005–0.01

UHMWPE Cobalt chrome Alumina

0.1–2.5 0.005–0.025 0.005–0.01

0.1–2.5 0.0071–0.035 0.0071–0.014

Table 4 Typical coefficients of friction for clean materials in dry contact in the presence of air (taken from Dowson and Wright18). Material combination

Coefficient of friction

Steel on steel Polyethylene on steel Polyethylene on polyethylene PTFE on PTFE PTFE on steel

0.6–0.8 0.3 0.2–0.4 0.04–0.2 0.04–0.2

R2 R1

Figure 1 Schematic diagram of a hip implant with radii of the femoral head (R1 ) and the outside of the cup (R2 ).

the deforming or the wearing of the other surface, although both were discovered by da Vinci. A non-dimensional ratio, known as the coefficient of friction and denoted by (m), is defined from the first law of friction m¼

F or F ¼ mW. W

(2)

The third law of friction is generally attributed to Coulomb in 1785. The kinetic coefficient of friction is generally less than or equal to the static coefficient of friction. Typical static coefficients of friction for clean materials in dry contact in the presence of air are given in Table 4. It should be pointed out that the coefficient of friction depends significantly on the nature of the bearing surfaces, particularly in the presence of biological lubricants. The coefficient of friction for the material combinations used for artificial joints in reality can be quite different from those listed in Table 4, as discussed later on in this section. Consideration of friction between bearing surfaces has played an important role in the development of artificial hip joints. Early hip replacements designed by Sir John Charnley utilised materials with optimal frictional properties as shown in Table 4 (e.g. Teflon (PTFE), with a low coefficient of friction). However, these failed very quickly due to wear. The majority of early hip implants using a metal-on-metal articulation also failed rapidly, largely due to the equatorial contact and resultant high friction and frictional torque generated. Low friction is preferential between the bearing surfaces to reduce the stresses transmitted to the fixation interface. This can readily be demonstrated through a simple analysis (Fig. 1). From the force/moment equilibrium, the frictional force (S) at the fixation interface between the outside of the acetabular cup and cement (or bone in case of cementless) is S¼

mWR1 , R2

(3)

where (R1 ) is the radius of the femoral head and (R2 ) the outside radius of the acetabular cup. It is clear that to reduce the likelihood of the interface failure, it is important to minimise the stress transmitted at the interface by not only reducing the coefficient of friction at the articulating surfaces, but also to reduce the femoral head radius. These are the basic principles for low friction (torque) arthroplasty (LFA) proposed by the late Sir John Charnley. Friction testing is a useful method to compare implants of various designs, materials and conditions. The measurement of friction may also be used as an indirect method to imply the lubrication of a bearing combination. The friction within an artificial joint can be measured using a pendulum simulator. One example is a single-station servo-hydraulic machine, controlled by a personal computer via a graphic user interface to provide a simplified gait cycle consisting of a dynamic vertical load and a horizontal flexion and extension motion as shown in Fig. 2. The measured frictional torque (T) is usually used to calculate a dimensionless parameter, called the friction factor, f¼

T . R1 W

(4)

This friction factor can be used to compare the effect of different variables, such as the material combination, implant size and design, lubricant, and load and motion profiles. Some of these parameters can be conveniently combined to form a dimensionless Sommerfeld number (z) defined below: z¼

ZuR1 , W

(5)

where (Z) is the viscosity of the lubricant, and (u) is the entraining velocity of the bearing surfaces. Typical friction factors associated with different lubrication regimens, to be discussed in the next section are given in Table 5. Furthermore, the variation in the friction factor against the Sommerfeld number can further indicate the mode of lubrication as schematically illustrated in Fig. 3. This is known as a Stribeck plot, in recognition of the great contribution to studies of journal bearing lubrication

ARTICLE IN PRESS Biotribology

35

Figure 2 A friction simulator at the Institute of Medical and Biological Engineering, Leeds University.

pointed out that wear cannot be eliminated completely, even with fluid film lubrication, since there are occasions when fluid-film lubrication can break down. Typically, this occurs during start-up and stopping. Wear can also occur in the absence of direct surface-to-surface contact, due to factors such as erosion and fatigue, as discussed in the section ‘Wear’. Furthermore, it should be pointed out that the lack of lubrication in some Judet and McKee-Farrar prostheses resulted in high friction and audible squeaking, which prompted the late Sir John Charnley to search for a LFA, as discussed in the section ‘Friction’. It is important to understand and to determine the lubrication mechanism in artificial joints, since such an understanding may help to optimise the bearing surfaces to minimise friction and wear. Conventional engineering methods of assessing lubrication regimens can be applied to artificial hip joints. These can be broadly classified into two categories, experimental measurements and theoretical predictions. The experimental methods involve either friction measurements, related to the Stribeck curve (see the Section ‘Friction’), or the detection of separation between the two bearing surfaces using a simple resistivity technique. The latter technique is particularly useful for conducting metal-on-metal bearings. For non-conducting ultra-high molecular weight polyethylene (UHMWPE) and ceramic bearing surfaces, a conductive coating is required. Theoretical predictions are based on the l ratio defined as follows15 l¼

Table 5 Typical friction factors for various bearings for artificial hip joints in the presence of bovine serum. Lubrication regimes

Friction factor

Boundary lubrication Mixed lubrication Fluid-film lubrication

0.1–0.7 0.01–0.1 0.001–0.01

by this German engineer in the 1920s. A constant friction factor with increasing Sommerfeld number indicates boundary lubrication. A reducing friction factor with increasing Sommerfeld number is indicative of a mixed lubrication and increasing friction factor with increasing Sommerfeld number indicates fluid film lubrication. Typical friction factors in various hip joints are summarised in Table 6.

Lubrication

hmin ¼ h Ra

Ra_Head

2

hmin  2 i1=2 . þ Ra_Cup

(6)

The key to the theoretical assessment of the potential for fluid-film lubrication is the accurate measurement of surface roughness (Ra ), discussed in the Section ‘Surfaces’, and the accurate prediction of a representative film thickness for the bearing (hmin). One of the film thickness equations developed in engineering can be used to estimate this as a first attempt.19  Zu 0:65  W 0:21 hmin ¼ 2:8 0 . (7) R ER E 0 R2 The equivalent radius (R) in Eq. (7) depends on the femoral head diameter (d) and the diametral clearance between the head and the cup (cd ), and can be calculated from the following equation:   dðd þ cd Þ d d R¼ ¼ . (8) 1þ 2cd 2 cd The entraining velocity (u) can be calculated from the angular velocity of the femoral head (o) od . (9) 4 Finally, the equivalent elastic modulus (E 0 ) is given by # ," 1  u2Head 1  u2Cup 0 E ¼2 , (10) þ E Head E Cup u¼

Lubrication refers to adding a lubricant between two bearing surfaces in order to control friction and wear. As indicated above, three distinct lubrication regimens exist in engineering: fluid-film lubrication, boundary lubrication and mixed lubrication. The characteristics of each lubrication regimen in terms of friction and wear are summarised in Fig. 4. It is clear from Fig. 4 that in order to minimise wear, the ideal lubrication regimen is fluid film. However, it should be

where (E) and (u) denote elastic modulus and Poisson’s ratio. A typical artificial hip joint with an UHMWPE cup against a metallic femoral head is used to illustrate how to calculate the (l) ratio and to determine the lubrication regimen. The

ARTICLE IN PRESS 36

Z.M. Jin et al.

0.1

Boundary lubrication

Friction factor

Mixed lubrication

Fluid film lubrication 0.01

z Figure 3

Typical friction factors and associated lubrication regimens.

Table 6 Typical friction factors for various bearings for artificial hip joints in the presence of bovine serum. Bearings

Friction factor

UHMWPE-on-metal UHMWPE-on-ceramic Metal-on-metal Ceramic-on-ceramic Ceramic-on-metal

0.06–0.08 0.06–0.08 0.22–0.27 0.002–0.07 0.002–0.07

femoral head is made of a cobalt–chromium alloy with an elastic modulus of 210 GPa and a Poisson’s ratio of 0.3. The elastic modulus of conventional UHMWPE is assumed to be 1400 MPa and the Poisson’s ratio is 0.3. Typical diameter of the femoral head (d) and the diametral clearance (cd ) are chosen as 28 and 0.2 mm. Typical load in the vertical direction and angular velocity representing the flexion/ extension in the human hip joint can be chosen as 2500 N and 1.5 rad/s. A typical viscosity for peri-prosthetic synovial fluid is assumed to be 0.0025 Pas. The equivalent radius, entraining velocity and equivalent elastic modulus can be calculated from Eqs. (8)–(10) as 1.974 m, 0.0105 m/s and 3309 MPa, respectively. The minimum film thickness can thus be determined from Eq. (7) as hmin ¼ 0:062 mm. A typical composite roughness for UHMWPE-on-metal material combination is 0.1–2.5 mm, as shown in Table 3. Therefore, the (l) ratio calculated is less than 1 (0.025–0.62) and this indicates a boundary lubrication regimen. Such an understanding is quite important when the wear is considered as will be discussed in the next section. Lubrication regimens in other types of artificial hip joint can be analysed readily using the above procedure. Predictions for typical hip implants with metal-on-metal and ceramic-on-ceramic bearings, respectively, are shown in Tables 7 and 8. It is clear from Tables 7 and 8 that ceramic-on-ceramic hip joints may achieve more favourable lubrication due to the smoothness of the bearing surfaces, although more

recently developed manufacturing techniques for metallic bearing surfaces are also capable of achieving a similar standard. The importance of design parameters, such as the femoral head diameter (d) and the diametral clearance (cd ), can be further explored for metal-on-metal bearings. It is clear from Eq. (7) that in order to promote fluid-film lubrication, it is necessary to increase the femoral head diameter and to reduce the diametral clearance so that the equivalent radius (R) is increased. The increase in the femoral head diameter also increases the entraining velocity. The importance of large diameter is manifest in the metal-on-metal hip resurfacing prosthesis. The estimated lubricant film thicknesses between a 28 mm total hip implant and a 50 mm diameter hip resurfacing prosthesis, both using a metal-onmetal bearing is compared in Table 9. However, it should be pointed out that the diametral clearance also plays an equally important role in the large diameter metal-on-metal hip resurfacing prostheses. An increase in the diametral clearance can lead to a decrease in the equivalent radius and consequently the predicted lubricant film thickness is reduced. This is particularly important for large diameter bearings. If the reduction in the lubricant film thickness moves the lubrication regimen towards boundary lubrication, the adverse effect of increased sliding distance associated with the large femoral head diameter must be considered. Such a comparison is shown in Table 10. These simple theoretical analyses have recently been confirmed with the experimental simulator studies.20,21 However, it should also be pointed that metal-on-metal bearings depend on protection from the boundary layers,22 and the effect of proteins can have a significant effect on the friction and wear.

Wear Wear is defined as progressive loss of substance from the operating surface of a body occurring as a result of relative motion at the surface. The importance of wear is related not only to the decreased function and replacement cost of a

ARTICLE IN PRESS Biotribology

37 Solid Surfaces

Boundary film Boundary lubrication: significant asperity contacts; Significant wear and friction; Protected by boundary films (physical and chemical properties)

Mixed lubrication: mixture of characteristics between boundary and fluid-film lubrication regimens

Fluid-film lubrication: Complete separation between two surfaces; Importance of lubricant viscosity; Minimum wear and low friction

Figure 4

Schematic lubrication regimens and associated characteristics.

Table 7 Calculation of (l) ratio and determination of lubrication in a typical metal-on-metal hip implant.

Table 8 Calculation of (l) ratio and determination of lubrication in a typical ceramic-on-ceramic hip implant.

Input parameters

Input parameters

Femoral head diameter Diametral clearance Elastic modulus (Co–Cr) Poisson’s ratio (Co–Cr) Load Angular velocity Viscosity Composite Ra

28 mm 0.06 mm 210 GPa 0.3 2.5 kN 1.5 rad/s 0.0025 Pas 0.014 mm

Femoral head diameter Diametral clearance Elastic modulus (Alumina) Poisson’s ratio (Alumina) Load Angular velocity Viscosity Composite Ra

28 mm 0.08 mm 380 GPa 0.3 2.5 kN 1.5 rad/s 0.0025 Pas 0.007 mm

Calculation Equivalent radius Entraining velocity Equivalent elastic modulus Minimum film thickness l ratio Lubrication regime

6.55 M 0.0105 m/s 230 GPa 0.024 mm 1.7 Mixed lubrication regimen

Calculation Equivalent radius Entraining velocity Equivalent elastic modulus Minimum film thickness l ratio Lubrication regime

4.914 M 0.0105 m/s 418 GPa 0.015 mm 2.12 Mixed lubrication regimen

component, but also the adverse effects of wear particles. For example, wear particles liberated from artificial joints have been shown to cause adverse tissue reactions, osteolysis and loosening. Similarly, wear particles can also have a detrimental effect on the quality of magnetic recording systems. The focus of wear studies should therefore be on both volume and particles. Fives types of wear are described: (a) Abrasive

The displacement of materials by hard particles

(b) Adhesive

(c) Fatigue (d) Erosive

The transference of material from one surface to another during relative motion by the process of solid-phase welding The removal of materials as a result of cyclic stress variations The loss of material from a solid surface due to relative motion in contact with a fluid which contains solid particles. This is often subdivided into impingement erosion and abrasive erosion. If no solid particles are present, erosion can still

ARTICLE IN PRESS 38

Z.M. Jin et al.

Table 9 Comparison of predicted lubricant film thickness between a total hip implant and a hip resurfacing prosthesis using a similar metal-on-metal bearing. Parameters

Total hip implant

Hip resurfacing prosthesis

Diameter (mm) Dia. clearance (mm) Load (N) Angular vel. (rad/s) Viscosity (Pas) Equivalent diameter (m) Entraining vel. (mm/s) Film thickness (mm)

28 60 2500 1.5 0.0025 13.1 10.5 0.024

50 100 2500 1.5 0.0025 25.1 (100%m) 18.75 (80%m) 0.058 (142%m)

Table 10

Effect of clearance on the predicted lubricating film thickness in metal-on-metal hip resurfacing prostheses.

Parameters

Hip resurfacing prosthesis

Hip resurfacing prosthesis

Diameter (mm) Dia. clearance (mm) Load (N) Angular vel. (rad/s) Viscosity (Pas) Equivalent diameter (m) Entraining vel. (mm/s) Film thickness (mm)

50 100 2500 1.5 0.0025 25.1 18.75 0.058

50 300 2500 1.5 0.0025 8.38 (70%k) 18.75 (0%) 0.025 (57%k)

(e) Corrosive

take place, such as rain erosion and cavitation A process in which chemical or electrochemical reactions with the environment dominates, such as oxidative wear

It should be pointed out that mechanical actions are involved in the above wear types (a–d) while wear type (e) is due to chemical action. Furthermore, it should be pointed out that the above wear types may occur simultaneously or sequentially. For example, wear particles, which may be produced as a result of adhesive wear, can then act as third bodies causing abrasive wear. In polymeric bearing surfaces, adhesive, abrasive and fatigue wear can all contribute to the overall wear. Some wear terms often described for artificial joints can all be related to the above mechanisms. For example, pitting, scratching, burnishing and delamination have all been described for retrieved total condylar knee joint replacements 23. Pitting and delamination are specific forms of fatigue wear, while burnishing and scratching are different degrees of abrasive wear. Understanding the wear mechanism is also important to design appropriate strategies to reduce wear. For example, abrasive wear can be minimised using hard smooth bearing surfaces such as ceramics. Effective cleaning during surgery and possible sealing of the whole joint can also prevent hard particles from entering the articulating surfaces. Fatigue wear mainly depends on the prosthesis design and materials used. It is important to minimise the contact stresses in order to avoid short-term fatigue failure. Effective lubrication, in terms of both boundary lubrication and fluid-film

lubrication, is the key to minimise adhesive wear. Corrosive wear mainly depends on the choice of the materials and for this reason generally similar metallic materials are used as metal-on-metal bearings. There are three laws of wear as listed below: I. Wear volume (V) increases as the normal load (W) increases II. Wear (V) increases as the sliding distance (x) increases III. Wear (V) decreases as the hardness (H) of the softer sliding component increases which can be expressed mathematically as follows: V/

Wx . H

(11)

A dimensionless constant can be introduced in the above equation, generally known as the wear coefficient (K 1 ) defined below: K1 ¼

VH . Wx

(12)

It is generally difficult to define the hardness of visco-elastic polymeric materials. Consequently, a dimensional wear factor is often used K¼

V . Wx

(13)

The common unit for the wear factor is mm3/(N m). A wide range of laboratory equipment, test methods and measuring systems has been employed to measure wear and to study wear mechanisms in total replacement hip joints.

ARTICLE IN PRESS Biotribology

39

Figure 5 Simple screening machines (a, b) and full simulator (c) for wear study of hip implants.

The three major forms of equipment listed below: (a) Pin-on-disc machines (b) Pin-on-plate machines (c) Joint simulators are schematically shown in Fig. 5. The pin-on-disc machine has been widely used in tribology and is particularly useful in the evaluation of the nature of wear and friction of material pairs under well controlled, steady-state conditions of load, sliding speed and environment. The pin-on-plate machine sacrifices the steady sliding speed between specimens, but partially simulates the reciprocating action broadly associated with the hip joint. More recently, it has become increasingly necessary to add an additional motion in either pin-on-disc or pin-on-plate machines to create a resultant multi-direction motion, particularly for UHMWPE materials. However, all these relatively simple wear machines are only useful as screening devices. A comparative performance evaluation of hip joints of different designs and material combinations requires joint simulators for laboratory studies. These simulators generate to a greater or lesser extent the three-dimensional loading and motion patterns experienced by hip joints, while providing a lubricating environment deemed to be physically and chemically similar to synovial fluid. Typical wear coefficients for various materials combinations are shown in Table 11. It is clear that the range of wear coefficients in Table 11 is much larger than that of the coefficient of friction in Table 4. It is also interesting to note that low friction is not synonymous with low wear, since the coefficient of friction between PTFE and steel is lowest among the materials combinations shown in Table 4 and yet the corresponding wear coefficient is much higher than that between

Table 11 Representative wear coefficients, K1, for various material combinations.18 Material combination

Wear coefficient

PTFE-on-steel High-density, high molecular weight polyethylene on steel

104–105 107–108

Table 12 Representative wear factors, K, for various material combinations tested in pin-on-plate machines. Material combination

Wear factor (mm3/(N m))

UHMWPE on metal Metal-on-metal Ceramic-on-ceramic

107 107 108

polyethylene and steel. Furthermore, small values of K 1 indicate that a very small fraction of asperity interactions are involved in the formation of wear particles. Typical wear factors for various biomaterials used for artificial hip joints are summarised in Table 12. It should be pointed out that the wear factors determined from the pin-on-plate machines may be quite different from those measured in joint simulators. It is often useful to compare wear rates between different bearings as shown in Table 13. Understanding the complex tribological mechanisms in terms of friction, wear and lubrication in artificial hip joints is important to optimise the bearing design in terms of

ARTICLE IN PRESS 40

Z.M. Jin et al.

Table 13 Representative volumetric wear rates for various artificial hip joints tested in simulators. Material combination

Volumetric wear (mm3/million cycles)

UHMWPE-on-metal UHMWPE-on-ceramic Cross-linked UHMWPE-on-metal Metal-on-metal Ceramic-on-ceramic

40 25 5–10 1.0 0.1

2. 3.

4.

5. 6.

clearance and head diameter to minimise wear. In UHMWPEon-metal material combinations, the lubrication regimen is predominantly boundary, and the change in the clearance and head diameter is unlikely to improve the lubrication significantly due to the relatively rough polymeric bearing surface. Therefore, a small head diameter should be chosen to minimise the sliding distance, but not to cause too much creep and linear penetration. Therefore, 28 mm diameter appears to be an optimum compromise in terms of wear. This is larger than the 22 mm diameter in the original Charnley hip prosthesis, which was intended to reduce the frictional torque. Other factors against the use of the 22 mm diameter include subluxation and possible dislocation if surgical technique is not precise. However, a mixed lubrication regimen is dominant in metal-on-metal bearings, which can either be close to boundary or fluid-film lubrication regimens. An increase in the femoral head diameter, particularly above 28 mm, can increase both the sliding speed and the effective radius, moving the lubrication regimen towards the fluid-film region. Consequently, wear can be significantly reduced. This has been best illustrated in large diameter metal-on-metal hip resurfacing prostheses introduced extensively recently. However, it should be pointed out that such an advantage of using a large diameter may be lost if the clearance between the head and the cup is excessive, such that the resultant lubrication regimen is close to the boundary lubrication region and the adverse effect of increasing sliding distance becomes dominant. On the other hand, too low a clearance may cause equatorial contact and requires tight manufacturing tolerances.

Summary Tribological studies of bearing surfaces for artificial hip joints play an important role in minimising wear particle generation and prolonging the lifetime of the implant. Integrated studies of friction, wear and lubrication are essential to gain a better understanding of the complex tribological mechanism in artificial joints. Application of tribological principles has contributed to the development of artificial hip joints. However, all these depend on the most important factor: the surgeon.

References 1. Dowson D, Wright V. Bio-tribology. In: Proceedings of the conference on the rheology of lubrication. The Institute of

7. 8. 9.

10.

11.

12.

13.

14.

15.

16. 17.

18.

19.

20.

21.

22.

23.

Petroleum, The Institution of Mechanical Engineers and the British Society of Rheology; 1973. p. 81–8. Litonjua LA, et al. Tooth wear: attrition, erosion, and abrasion. Quintessence Int 2003;34(6):435. Turssi CP, et al. Wear of dental resin composites: insights into underlying processes and assessment methods—a review. J Biomed Mater Res 2003;65B(2):280. Dowson D. Tribology and the skin structure. In: Berardesca E, et al., editors. Bioengineering of skin: methods and instrumentation, Part III: General aspects. Boca Raton, FL: CRC Press; 1995. p. 159. Sivamani RK, et al. Coefficient of friction: tribological studies in man—an overview. Skin Res Technol 2003;9(3):227. Gronqvist R, et al. Measurement of slipperiness: fundamental concepts and definitions. Ergonomics 2001;44(13):1102. Maynard WS. Tribology: preventing slips and falls in the workplace. Occup Health Saf 2002;71(9):134. Holly FJ, Holly TF. Advances in ocular tribology. Adv Exp Med Biol 1994;350:275. Secomb TW, Hsu R, Pries AR. Blood flow and red blood cell deformation in nonuniform capillaries: effects of the endothelial surface layer. Microcirculation 2002;9(3):189. Reul H, et al. In-vitro assessment of the wear development mechanism and stabilization of wear in the Edwards MIRA/Sorin Bicarbon mechanical heart valve orifice ring. J Heart Valve Dis 2002;11(3):409. Walowit JA, et al. The analysis, design, and testing of a blood lubricated hydrodynamic journal bearing. ASAIO J 1997;43(5): M556. Shahgaldi BF, Compson J. Wear and corrosion of sliding counterparts of stainless-steel hip screw-plates. Injury 2000; 31(2):85. Gouldstone A, et al. Elastohydrodynamic separation of pleural surfaces during breathing. Respir Physiol Neurobiol 2003; 137(1):97. Mow VC, Ateshian GA, Spilker RL. Biomechanics of diarthrodial joints: a review of twenty years of progress. J Biomech Eng 1993;115(4B):460. Dowson D. New joints for the Millennium: wear control in total replacement hip joints. Proc Inst Mech Eng J Eng Med 2001; 215(4):335. Ingham E, Fisher J. The role of macrophages in osteolysis of total joint replacement. Biomaterials 2005;26(11):1271–86. Hall RM, Siney P, Unsworth A, Wroblewski BM. The effect of surface topography of retrieved femoral heads on the wear of UHMWPE sockets. Med Eng Phys 1997;19(8):711–9. Dowson D, Wright V. Introduction to the biomechancis of joints and joint replacements. London: Mechanical Engineering Publications Ltd; 1981. Jin ZM, Dowson D, Fisher J. Analysis of fluid film lubrication in artificial hip joint replacements with surfaces of high elastic modulus. Proc Inst Mech Eng J Eng Med 1997;211: 247–56. Dowson D, Hardaker C, Flett M, Isaac GH. A hip joint simulator study of the performance of metal-on-metal joints: Part II: Design. J Arthroplasty 2004;19(8, Suppl 3):24–30. Rieker CB, Scho ¨n R, Konrad R, Liebentritt G, Gnepf P, Shen M, et al. Influence of the clearance on in-vitro tribology of large diameter metal-on-metal articulations pertaining to resurfacing implants. Orthop Clin N Am 2005;36:135–42. Buscher R, Tager G, Dudzinski W, Gleising B, Wimmer MA, Fischer A. Subsurface microstructure of metal-on-metal hip joints and its relationship to wear particle generation. J Biomed Mater Res B Appl Biomater 2005;15;72(1):206–14. Hood RW, Wright TM, Burstein AH. Retrieval analysis of total knee prostheses: a method and its application to 48 total condylar prostheses. J Biomed Mater Res 1978;17: 829–42.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 41–46

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

NEUROMUSCULAR CONDITIONS

Poliomyelitis: Orthopaedic management Adnan A. Faraj Department of Orthopaedics and Trauma, Airedale General Hospital, Skipton Road Steeton, Keighley, Bradford BD20 6TD, UK

KEYWORDS Post-polio syndrome; Paralysis; Orthopaedics; Orthosis; Surgery

Summary Although poliomyelitis has been eradicated in most parts of the world, orthopaedic surgeons occasionally encounter residual deformities in patients who suffered the disease in childhood. An understanding of the causative factors and the available treatment options are essential before surgical intervention is contemplated. It is also well recongnised that post-polio syndrome occurs in those who suffered poliomyelitis 20–40 years ago. It is important to note that inadequate or improper surgical intervention can potentially lead to more disability; a well-planned approach to the particular part of the body affected by polio, after considering the patient as a whole and understanding the principles involved, is the best option. & 2005 Elsevier Ltd. All rights reserved.

History of poliomyelitis Poliomyelitis is said to have first occurred nearly 6000 years ago in the time of the Ancient Egyptians. The evidence for this is in the withered and deformed limbs of certain Egyptian mummies. The following are the more important dates in the history of polio (Fig. 1): Ancient Egypt 3700 BC: An Egyptian mummy with probable polio. If this was polio, cases almost certainly occurred before then. 1580–1350 BC: The Priest Ruma with a withered leg and equinus foot—shown on a plaque and probably poliomyelitis. 1209 BC: Mummy Giptah with an equinus foot.

Tel.: +44 1535 652511; fax: +44 1535 292098.

E-mail address: [email protected]. 0268-0890/$ - see front matter & 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2005.10.005

(Middle Ages) 1559: A painting by Pieter Bruegel shows a crippled beggar. This is not definitely polio, although it probably did occur during this period in England. (Eighteenth century) 1789: First known description of poliomyelitis by Underwood. (Nineteenth century) 1834: First epidemic of poliomyelitis in the island of St. Helena. 1855: First description by Duchenne of the pathological process in poliomyelitis involving the anterior horn cells of the spinal cord. (Twentieth century) 1908: Transmission of poliomyelitis to a monkey by Landsteiner. 1909: Passage of the virus through a monkey by Flexner. 1949: Growth of the virus on tissue culture. 1951: Three types of polio virus isolated and identified. 1954: First large-scale trial of Salk (dead vaccine) by injection. 1958: First general use of Sabin (live attenuated vaccine) by mouth.

ARTICLE IN PRESS 42

A.A. Faraj developed countries, on the other hand, residual poliomyelitis is still occasionally seen in the elderly and immigrants.

Polio through the Ages Eygyptian Mummy with equinus foot 4000 B.C. Possible Polio

EGYPT Priest Ruma of Egypt stone carving

1500 B.C.

Probably Polio

Oil painting by Peter Bruegel of begger in 1559

1500 A.D. EUROPE

2000 A.D.

Possible Polio

THE WORLD

Polio as a Child

Residual paralysis as an adult

World wide immunization by 2000 A.D.

Acute poliomyelitis Poliovirus is primarily spread by faecal–hand–oral transmission from one host to another. The virus is shed in oral secretions for several weeks and in the faeces for several months. It destroys the anterior horn cells in the spinal cord. Poliovirus infections can be divided into minor and major forms: The minor illnesses occur 1–3 days before the onset of paralysis, with gastrointestinal complaints such as nausea and vomiting, abdominal cramps and pain, and diarrhoea. There are also systemic manifestations such as sore throat, fever, malaise, and headache. This stage lasts usually for 2–3 weeks but, may extend for up to 2 months; the presence of any tenderness in the muscles is evidence that the acute stage is not over. The major illness includes all forms of central nervous system (CNS) disease caused by poliovirus, including aseptic meningitis (or non-paralytic polio), polio encephalitis, bulbar polio, and paralytic poliomyelitis, alone or in combination. The clinical findings associated with an attack of polio are as follows:

No more new cases in children

Figure 1

Epidemiology Since the World Health Assembly resolved in May 1988 to eradicate poliomyelitis, the estimated global incidence of polio has decreased by more than 99% and three World Health Organization (WHO) regions (Americas, Western Pacific, and European) have been certified as polio-free. Since 1994, when the countries of the WHO South-East Asia Region (SEAR) began accelerating polio-eradication activities, substantial progress towards that goal has been made. By 2001, poliovirus circulation in India had been limited primarily to the two northern states of Uttar Pradesh and Bihar, with 268 cases reported nationwide. However, a major resurgence of polio occurred during 2002, with 1600 cases detected in India, of which 1363 (85%) were in Uttar Pradesh and Bihar.1 Problems remain due to the difficulties involved in extending immunization coverage to some regions (especially Africa), integrating new vaccines into routine immunization schedules and securing sufficient funding for programs. Injection safety is also a major problem that should be resolved by utilisation and proper disposal of single-use auto-disabling syringes. The forthcoming availability of new vaccines and the action of the Global Alliance for Vaccines and Immunization hold reasonable hope for the future. Other problems remain, such as new conditions resembling polio paralysis caused by viral infection other than by poliovirus2 and post-polio syndrome (PPS).3 It is safe to assume that acute and residual poliomyelitis is still unfortunately encountered in the developing world. In

1. There is fever, stiffness of the neck (nuchal rigidity), and a plecocytosis in cerebrospinal fluid. 2. Profound asymmetrical muscle weakness develops. 3. The initial phase is typically followed by some recovery of muscle strength, but permanent weakness results from necrosis of anterior horn cells. 4. Rarely, a transverse myelitis with paraparesis, urinary retention, sensory symptoms and signs along with autonomic dysfunction (including hyperhidrosis or hypohidrosis), and decreased limb temperature may occur. In this stage, the treatment is mainly medical involving the paediatric physicians. General supportive treatment for the pyrexia and irritation, the prevention of secondary respiratory infection and the treatment of any respiratory paralysis are the main aspects of the treatment. The paralysed legs are supported by plaster splints or pillows and sandbags to keep the hip joints in 51 of flexion and in neutral rotation. The knee joint is held at 51 of flexion and the foot is supported in a 901 position. Splinting relieves pain and spasm and prevents the development of deformities.

Recovery stage In this stage, also known as the convalescent stage, the acute symptoms and muscle tenderness disappear and the paralysed muscles begin to recover. This stage lasts for up to 2 years after the onset of the disease. During this whole period, there is gradual recovery of the muscles; the recovery is rapid in the first 6 months but is slower during the subsequent months.

ARTICLE IN PRESS Poliomyelitis: Orthopaedic management Treatment in this stage is mainly in the orthopaedic department involving physiotherapy and splinting. The aims of treatment are:

43 Ankle Dorsiflexors Plantar Flexors

(a) To assist in the recovery of paralysed muscles by remedial exercises. (b) To prevent deformities by the use of orthotic appliances. An assessment is first made of the extent of muscle paralysis by charting the power of various groups of muscles and grading them according to the international nomenclature (Medical Research Council grading) as follows: 0 —Complete paralysis 1 —Slight flicker of contraction present 2 —Muscle can move a joint only when gravity is eliminated. 3 —Muscle can move a joint against gravity 4 —Muscle can move a joint against gravity and resistance 5 —Full normal power. Total functional assessment of the limbs is made before planning treatment. This will include: (a) (b) (c) (d)

Charting the muscle power grades. Extent of contractures and deformities. Method of ambulation. Shortening of the limb.

Efficient physiotherapy is the mainstay of the management of this stage of poliomyelitis. A good physiotherapy department with facilities for exercise therapy, hydrotherapy and electrical stimulation of muscles is essential in the management of paralytic polio.

Orthotic management Appropriate orthotic appliances are prescribed to prevent deformities due to muscle imbalance, as shown in the following chart. The new international nomenclature for orthotic appliances describes the joints that are stabilised by the appliance e.g. the name ‘Ankle Foot Orthosis’ replaces the old term ‘below knee appliances’. When the power of muscles controlling the hip and knee are normal and the weakness is only in the dorsiflexors or plantar flexors of the ankle or invertors or evertors of the foot, the patient is prescribed an Ankle Foot Orthosis (Below Knee orthosis or Caliper). Indications for orthotic appliances Muscle paralysed

Deformity

Appliances

Foot Evertors

Inversion

Ankle Foot Orthosis (AFO) with inside bar and Outside T strap Ankle Foot Orthosis (AFO) with outside bar and Inside T strap

Invertors

Eversion of foot

Foot drop Calcanecus

AFO with 901 foot drop stop AFO with reverse 901 stop

Ankle and Foot All muscles

Flail foot

AFO with limited motion ankle joint

Knee Extensor: quadriceps

Genu Recurvatum

Knee Ankle Foot Orthosis (KAFO) with back knee support

Gluteus medius gait

Hip Knee Ankle Foot Orthosis (HKAFO), full caliper with pelvic band

Hip Abductors: gluteus medius, minimus and tensor fascia

When the quadriceps power is 2 or below, the knee has to be stabilized and hence a Knee Ankle Foot Orthosis (Full or Above Knee Calliper) is prescribed. If the hip abduction power is poor i.e. less than, 2, the appliance will include a pelvic band with a hip joint (Hip Ankle Foot Orthosis) to prevent the lurching gluteus medius gait. In the recovery stage, a child who starts with a full appliance with pelvic band may be able to gradually manage with shorter appliance and ultimately discard it when the muscles fully recover due to intensive physiotherapy treatment.

Residual paralysis stage The period beyond 2 years after the onset of the disease is called the residual paralysis stage. No recovery of muscle power occurs in this stage. Deformities are liable to occur due to imbalance of muscle power and poor posture. There is also disuse atrophy of muscles and shortening of the leg due to interference with growth. In neglected cases, gross fixed deformities of the hip, knee and foot occur with severe wasting of muscles. Children with extensive paralysis and gross deformities have to crawl on all fours to move from place to place.

Post-polio syndrome PPS is the term used for the newly occurring late manifestations of poliomyelitis that develop in patients 30 to 40 years after the occurrence of the acute illness. It has been estimated that 25–60% of the patients who had acute polio may experience these late effects of the disease.4 The specific cause of post polio syndrome is unknown; the aetiology has been attributed to pathophysiologic and functional causes. Pathophysiologic causes include chronic poliovirus infection, death of the remaining motor neurons with ageing, premature ageing, damage to the remaining motor neurons caused by increased demands or secondary insults, and immune-mediated syndromes. Functional

ARTICLE IN PRESS 44 aetiologies for PPS include greater energy expenditure as a result of weight gain and muscle weakness caused by overuse or disuse. PPS has been recognized for over 100 years, but is more common at the present time because of the large epidemics of poliomyelitis that occurred in the 1940s and 1950s. PPS is characterised by neurological, musculoskeletal, and general manifestations. Musculoskeletal manifestations include muscle pain, joint pain, spinal spondylosis and scoliosis, and secondary root and peripheral nerve compression. General manifestations include generalized fatigue and cold intolerance. The slowly progressive muscle weakness occurs in those muscle groups already involved such as the quadriceps and calf muscles.

Diagnostic criteria for post-polio syndrome 1. A prior episode of paralytic poliomyelitis with residual motor neuron loss (which can be confirmed through a typical patient history, a neurological examination, and, if needed, an electrodiagnostic exam). 2. A period of neurological recovery followed by an interval (usually 15 years or more) of neurological and functional stability. 3. A gradual or abrupt onset of new weakness or abnormal muscle fatigue (decreased endurance), muscle atrophy, or generalized fatigue. 4. Exclusion of medical, orthopaedic, and neurological conditions that may be causing the symptoms mentioned in 3.

A.A. Faraj decelerate the tibia is lost and, therefore, flexion of the knee will persist throughout the stance phase. In order to prevent this the patient may attempt to compensate with increased quadriceps activity for a longer portion of the stance phase. In the case of a weak quadriceps and hamstrings the occurrence of an equinus contracture, or a hinged AFO with a dorsiflexion block will both prevent excessive knee flexion and excessive ankle dorsiflexion during the stance phase. The pitfall of lengthening of the Achilles tendon should be avoided in these patients. These patients may require an ischial bearing, double upright locked knee orthosis, which helps prevent the knee from buckling during gait. Common foot and ankle deformities seen are pes cavovarus (hindfoot cavus) due to evertor paralysis (peroneus brevis and longus) and pronated everted foot due to invertor paralysis (tibialis anterior and posterior). Foot intrinsics are typically spared in polio. Claw toes result from relative overactivity of the long toe flexors and extensors (to compensate for weakness of the triceps

Mortality Excluding polio patients with respiratory failure, long-term mortality following polio appears to increase 20 years after recovery from the acute illness. Contracting severe paralytic poliomyelitis at a young age seems to increase long-term mortality.6

Management of post-polio syndrome

Treatment

Many patients require revision of orthotic devices such as braces, canes, and crutches or to use new lighter orthotic devices to treat new symptoms. Common issues include genu recurvatum, knee pain, back pain, degenerative arthritis, or arthralgia. Surgery for scoliosis or fractures may also be necessary to treat new conditions.4

It is more economic to prevent 100 polio cases than to treat one hopelessly crippled child. It is often quicker to straighten 100 deformed limbs by simple subcutaneous surgery than to treat a single patient with complicated procedures. The final aim should be that patients return to their own village or town, accepted and integrated into their own communities, and earning their own living among their own friends. Since overuse weakness is frequently present in these patients, the role of slowly progressive non-fatiguing exercise in their rehabilitation is emphasised. New muscle weakness of a mild-to-moderate degree responds well to a non-fatiguing exercise program and pacing of activity with rest periods to avoid muscle overuse. Generalised fatigue may be treated with energy conservation and weight loss programs and lower extremity orthoses.3,4 An orthosis is a device which externally supports an existing body part with the objective of supporting, correcting or compensating for skeletal deformity or weakness. The current range of available orthoses is many and varied and with the advent of new materials such as carbon fibre, advanced manufacturing techniques and the range of devices available to the prescriber is ever increasing. Orthoses are available for all parts of body and aid in conservative and definitive treatment of many deformities. The thermoplastic leaf spring AFO, or drop foot splint, is one good example of an orthosis commonly used. It assists dorsiflexion and uses 3-point pressure to stabilise the ankle joint.

Pattern of muscle weakness and deformities Upper limb involvement Late functional deterioration is common in long-term poliomyelitis patients. While upper-limb pain in individual functional regions is common, its overall prevalence and pattern in long-term poliomyelitis is poorly documented. There are data in support of ‘overuse’ due to greater mobility aid dependence as a cause of increasing upper limb pain in long-term poliomyelitis especially among severely paralysed polio patients.5

Lower limb involvement Typical osseous or soft-tissue abnormalities about the knees affected by poliomyelitis include external rotation of the tibia, excessive valgus alignment, ligamentous laxity, and genu recurvatum. With localised wasting, the quadriceps can help compensate for a weak calf. With hamstring weakness the ability to

ARTICLE IN PRESS Poliomyelitis: Orthopaedic management

45

Orthopaedic operations in patients with residual poliomyelitis7

This technique can still be helpful in skeletal immature patients.

Hip and knee contractures of over 301

Arthrodesis

In general these will all require surgery, unless one or both arms are weak in addition to bilateral lower limb paralysis when the use of crutches will be difficult or impossible. In a young child with fairly recent contractures the most important single factor responsible for the deformity is a tight tensor fascia lata and ilio-tibial band. In the older child or adult, however, other ligamentous and tendinous structures play an important part and must be divided as well. The subcutaneous method of division is very satisfactory for less severe contractures, provided it is done correctly and as extensively as necessary. Care must be taken to avoid damaging the femoral and popliteal-arteries and the common peroneal nerve. The biceps, however, should always be divided under direct vision because of the risk of damaging the adjacent lateral popliteal nerve.

Arthrodesis is used to correct deformity, relieve pain in arthritic joints and reduce the number of joints a weak muscle is acting across orthroclesis is more popular than tenodesis. In skeletal immature patients, extra-articular arthrodeses can be performed, allowing continued growth of the skeleton.8

Tendon transfer to re-establish muscle power In selecting a tendon to transfer, the muscle should be sufficiently strong to supplement the power of a paralysed muscle. The nerve and blood supply of the transferred muscle should be preserved in order to avoid iatrogenic weakness. For efficiency the transferred tendon should be securely attached (with tension) close to the insertion of a paralysed tendon, and should be routed in a direct line between its origin and the new insertion. The transferred tendon should also be retained in its own sheath, avoiding tunnels in fascia, bone, or an interosseous membrane to avoid adhesions. The joint across which the muscle acts must be in a satisfactory position; all contracted structures must be released before the tendon transfer. When possible an agonist muscle, with the same range of excursion of its tendon, should be chosen.

Muscle transplantation to replace a paralysed muscle In these procedures, unlike tendon transfer, both the origin and the insertion of a muscle are detached along with its neurovascular pedicle. This procedure is not as popular as tendon transfer, because of the difficulty in finding a normal muscle to transplant, donor side morbidity, the technical difficulty and the shortage of microvascular surgeons in the third world countries where residual polio is still seen.

Stabilization of relaxed or flail joint Tenodesis, fixation of ligaments, and construction of artificial check ligaments are used to restrict the range of movement or to eliminate abnormal motion of a joint. With few exceptions, these procedures have been discarded, as deformity in the opposite direction may occur, and the tendon or artificial check ligaments may stretch with time.

Limb lengthening Often poliomyelitis is unilateral causing limb length inequality, which occasionally requires limb lengthening. In leg lengthening for patients with poliomyelitis, callus maturation is slow and patients tend to develop contractures despite physiotherapy, bracing or joint fixation. Concomitant and secondary surgery is frequently required to treat associated problems or residual deformities. Lengthening along an intramedullary locked nail can significantly shorten the treatment time with relatively few complications.9

Joint replacement surgery In patients with post-polio residual deformities joint replacement can be indicated. In one study, pain and knee scores improved following total knee arthroplasty in patients with a history of poliomyelitis and antigravity quadriceps strength, but there was less pain relief in patients with less than antigravity quadriceps strength. Recurrence of instability and progressive functional deterioration is possible in all knees affected by poliomyelitis that have undergone total knee replacement, but they appear to occur more commonly in more severely affected knees.9

Ilizarov techniques There are many drawbacks to using conventional approaches for the treatment of complex foot deformities, like the increased risk of neurovascular injury, soft-tissue injury, and shortening of the foot. An alternative approach, that can eliminate these problems is the Ilizarov method. Pin-tract problems, contractures, residual and recurrence of deformity can complicate the Ilizarov method.10

Conclusion The surgeon managing the residual weakness of poliomyelitis and post-polio syndrome must possesses an understanding of the pathological process in poliomyelitis as well as the variations in the pattern of the disease in different parts of the body. Poliomyelitis causes a lower motor neuron disease unlike other types of neuromuscular paralysis. The neurological problems, and the pattern of paralysis following poliomyelitis is different from upper motor neuron paralysis or indeed lower motor neuron paralysis caused by other diseases.

ARTICLE IN PRESS 46 Knowledge of patho-anatomy before embarking on surgery is necessary. The mainstay of management remains physiotherapy and orthotic appliances.

References 1. Progress toward poliomyelitis eradication—India, 2003. Centers for Disease Control and Prevention. 2. Al-Shekhlee A, Katirji B. Electrodiagnostic features of acute paralytic poliomyelitis associated with West Nile virus infection. Muscle Nerve 2004;29(3):376–80. 3. Jubelt B. Post-polio syndrome. Curr Treat Options Neurol 2004;6(2):87–93. 4. Frustace SJ. Poliomyelitis: late and unusual sequelae. Am J Phys Med 1987;66(6):328–37.

A.A. Faraj 5. Koh ES, Williams AJ. Povlsen BUpper-limb pain in long-term poliomyelitis. QJM 2002;95(6):389–95. 6. Nielsen NM, Rostgaard K, Juel K, Askgaard D, Aaby P. Long-term mortality after poliomyelitis. Epidemiology 2003; 14(3):355–60. 7. Ingram AJ. Paralytic disorders in Campbell’s Operative Orthopaedics by A.H. Crenshaw. St. Luis, Washington, DC, Toronto: The C.V. Mosby Company; 1987. p. 2925–3060. 8. Nicholas JG, David GL. Total knee arthroplasty in limbs affected by poliomyelitis. Bone Jt Surg (American) 2002; 84:1157–61. 9. Moran MC. Functional loss after total knee arthroplasty for poliomyelitis. Clin Orthop 1996; (323):243–6. 10. Kocaoglu M, Eralp L, Atalar AC, Bilen FE. Correction of complex foot deformities using the Ilizarov external fixator. J Foot Ankle Surg 2002;41(1):30–9.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 47–51

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

RADIOLOGY

Functional imaging in orthopaedic infections—Update on immunoscintigraphy Karthikeyan P. Iyengar, Chandakacharla N. Ramesh, Sobhan Vinjamuri Department of Nuclear Medicine, Royal Liverpool University Hospital, Prescot Street, Liverpool L7 8XP, UK

KEYWORDS Osteo-articular infection; Infection imaging; Nuclear medicine; White cell labelling; 99m Tc-Sulesomab

Summary This article highlights the role of radionuclide imaging including isotope bone scanning, radiolabelled white cell scanning and newer imaging techniques such as monoclonal antibodies in orthopaedic infection. & 2005 Elsevier Ltd. All rights reserved.

Introduction The diagnosis of orthopaedic infection is based on clinical examination, laboratory investigations, musculoskeletal imaging and tissue culture. Nuclear medicine plays an important role in the evaluation of patients with suspected musculoskeletal infection (osteomyelitis and septic arthritis). This is especially so in joint prostheses or patients with metallic implants, where complementary imaging with Ultrasonography (USG), Computed Tomography (CT) or Magnetic Resonance Imaging (MRI) may not be helpful.

Imaging in orthopaedic infection Plain radiographs Plain radiographs are usually the first investigations in the diagnosis of orthopaedic infections. Although crucial, they Corresponding author. Tel.: +44 151 7064462;

fax: +44 151 7065844. E-mail address: [email protected] (S. Vinjamuri). 0268-0890/$ - see front matter & 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2005.09.008

are often inconclusive, non-specific and sometimes misleading.1 Changes do not occur until 1 to several weeks after the onset of infectious disease allowing bone or joint destruction if diagnosis is delayed.2

USG, CT and MRI When infection is established, USG, CT and MRI all have individual value in diagnosis including guiding biopsy and in therapy. USG offers a non-invasive, operator-dependant evaluation of musculoskeletal infection especially in septic arthritis and is free of radiation. However, the uses of ultrasound in the diagnosis of osteomyelitis are limited to the detection of the soft tissue abnormalities around the bone because the sonic beam does not cross the bone cortex and identify a bone marrow discontinuity. CT scanning provides an excellent assessment of bone and soft tissue structures. Cross-sectional images are created with the benefit of high density, high contrast and spatial resolution. The procedure involves significant radiation exposure.

ARTICLE IN PRESS 48

K.P. Iyengar et al.

MRI gives a greater contrast resolution and higher anatomical detail than plain radiographs and CT, and does not involve radiation exposure. It is a useful procedure for the detection and determination of the extent of infection. However, because of imaging interference, MRI and CT scans are often not helpful in patients with metallic implants, aneurysm clips, pacemakers or prosthetic joints.3

Nuclear medicine in infection imaging Nuclear medicine is a powerful diagnostic modality in the assessment of musculoskeletal abnormalities. Nuclear medicine applications in infection are most useful when complementary imaging is not helpful. Scintigraphic detection is based on physiochemical changes and hence provides a functional evaluation of bone pathology. Basis Radionuclide imaging uses applied tracer physiology in which the tracers are radionuclide compounds (radiopharmaceuticals). When the radiopharmaceutical is administered to the patient and it localises in the sites of abnormality, it emits gamma-rays, which are detected by scintillation crystals in a gamma camera. The crystal emits tiny scintillations of light, which reproduces the pattern of isotope distribution in the patient.4 Dealing with patients with presumed or established infective disorders, nuclear medicine techniques can answer the common clinical questions. Mechanism of action (1) Localisation in sites of enhanced vascular permeability. (2) Localisation in sites of leucocytes accumulated in infective foci by diapedesis and chemotaxis.

and will be seen on the bone scan as a ‘hot spot’. The introduction of 99mTc-labelled phosphates for skeletal imaging in 1971/72 has provided a non-invasive technique to detect and localise sites of osseous infection and became a reliable method for evaluating patients with suspected inflammatory bone disease and metastatic bone disease.5 A three-phase bone scan is the basic examination which detects sites of increased bone turnover with high sensitivity of more than 90%.6 Typically the scintigraphic appearance of osteomyelitis is associated with increase in uptake in all three phases (blood flow/angiographic, blood pool and delayed phases), especially the third. A normal 99mTcdiphosphonate bone scan excludes chronic osteomyelitis with a very high certainty of more than 90%.7 However, if any other cause of bone remodelling, complicating the diagnosis of infection is present, the sensitivity remains high but the specificity reduces markedly.8 Because of the low specificity of the isotope bone scan, more specific scintigraphic techniques are required when the bone scan is abnormal (Table 1). 67 Gallium citrate increases specificity for imaging infections, but has an accuracy rate of less than 70%.6,9 67Gallium citrate has been replaced by white blood cell scans labelled with either 111indium oxime or 99mTc hexamethyl propylene amine oxime. The accuracy of diagnosing bone infection is increased by 111indium-WBC imaging and is still recommended as a gold standard for skeletal infection imaging by many authors.2,9,10 However, both leukocyte procedures require in vitro granulocyte isolation and labelling, which are time consuming and involve biohazard risk to medical personnel and potential for misadministration of blood products to the wrong patient. The search is still on for the ideal (Table 2) and newer infection imaging radiopharmaceuticals (Table 3).

Immunoscintigraphy (antibody imaging) Scientific basis

Traditional bone scintigraphy in orthopaedic infection Bone scans rely on the property of orthopaedic lesions to excite a local osteoblastic response and increase in vascularity. This results in an accumulation of radiotracer Table 1

Immunoscintigraphy is an imaging procedure using antibodies labelled with radiopharmaceuticals. The antigen reacting Fab’ fragment of an immunoglobulin can be tagged with a radiopharmaceutical and thus be available for infection imaging. Antibody imaging can be both

Features of traditional agents commonly used for infection imaging.

Radio pharmaceutical

Advantages

Disadvantages

99m

Readily available Inexpensive High sensitivity

Low specificity

67

Easy to prepare Low toxicity Detects low grade infection

Time consuming, delayed imaging High radiation dose Low specificity

High target to background ratio

Time-consuming preparation Complex and expensive radiolabelling In vitro labelling required Poor availability

Tc-MDP

Gallium

111

In-WBC

ARTICLE IN PRESS Functional imaging in orthopaedic infections—Update on immunoscintigraphy 99m

Table 2 Criteria for an ideal infection imaging radiopharmaceutical.

 Should be easy to prepare and have low cost with wide availability

 Give rapid delineation of foci and extent of infectious lesion

 Should not have significant residual accumulation in normal body tissue

49

Tc-Sulesomab scintigraphy

99m

Tc-Sulesomab is 99mTc-labelled Fab’ fragment of IMMUMN3, an immunoglobulin G1 murine monoclonal antibody produced from a hybridoma developed by fusion of murine myeloma (SP2/0) cells with spleen lymphocytes obtained from a mouse immunised with cacinoembryonic antigen. The antibody reacts strongly with non-specific cross-reacting antigen (NCA-90) present on human granulocytes.11

 Have rapid wash out from background but good Mechanism

retention in infective foci

 Give discrimination between infection and non99m

microbial inflammation

 Have low toxicity and absent immune response  High specificity

Tc-Sulesomab uptake at the site of infection is explained by the migration of circulating antibody-labelled granulocytes to the site of infection.

Advantages of

 Table 3 Newer radiopharmaceuticals for orthopaedic infection imaging according to mechanism of action. Non-specific

Specific

Polyclonal human immunoglobulin G, e.g. 99m Tc-HIG

(1) Antigen binding antigranulocyte monoclonal e.g. 99mTc/In111AGABs (2) Antigen binding antigranulocyte monoclonal antibody fragment, e.g. 99m Tc Sulesomab (3) Receptor interacting chemotactic peptides e.g. 99m Tc colloids/nanocolloids (4) Bacterial binding, e.g. Ciprofloxacin derivatives (99mTc Infecton) (5) Uptake in metabolically active cells e.g. 18Fdeoxyglucose

non-specific, e.g polyclonal human immunoglobulin G (99mTc-HIG) and specific, e.g antigen binding anti-granulocyte monoclonal antibody/antibody fragments. Since immunoscintigraphy involves use of cloned murine antibodies, allergic reactions such as the human anti-mouse antibody (HAMA) response needs to be monitored.

Advantages of immunoscintigraphy

 Unlike autologous leukocyte techniques in the imaging of   

infection, immunoscintigraphy does not require isolation of white blood cells ex vivo for tagging. In-vivo tagging avoids chances of misadministration and contamination of healthcare professionals. Available in a readymade kit, easier to prepare, simpler to use. The antibodies can be tagged with either with 111indium or 99mTc, however, because of low radiation dose, better quality images and availability, 99mTc is preferred.

99m

Tc-Sulesomab scintigraphy

99m

Tc-Sulesomab is labelled with 99mtechnetium. Technetium is less expensive, readily available and gives superior quality images compared to 111indium. 99m Tc-Sulesomab preparation is a simple 5 min, one-step process, thus eliminating the 2 h delay, technically demanding cell isolation procedure and its concomitant biohazard risks to medical personnel and blood product misadministration risk to patients. 99m Tc-Sulesomab images are obtained the same day (within hours) in contrast to the 18–24 h delay with 111 indium WBC scan. With 99mTc-Sulesomab, the granulocytes can be specifically tagged in situ by the circulating radiolabelled antigranulocyte antibody. In contrast, WBC labelling is an invitro process labelling lymphocytes and granulocytes present in 30–50 ml of withdrawn blood. Thus 99mTcSulesomab is capable of tagging considerably more granulocytes than 111indium WBC scan. Chances of allergic reaction, especially the HAMA response are reduced with use of monoclonal antibody fragments instead of whole antibodies. Rapid clearance of radiotracer, since 99mTc-Sulesomab imaging involves antibody Fab’ fragments rather than whole antibody. 99m



 

 

Present clinical experience Hakki et al., in a total of 74 evaluable patients with a heterogeneous group of bone and joint infections including long bones, diabetic feet and prosthetic joints found that 99m Tc-Sulesomab had a sensitivity of 93%, a specificity of 91%, an accuracy of 92%, a positive predictive value of 86% and a negative predictive value of 96% compared to a sensitivity of 85%, a specificity of 82%, an accuracy of 83%, a positive predictive value of 73% and a negative predictive value of 90% for WBC scans done concurrently in the same patients1 (Figs. 1 and 2). In the other trial with 99mTc-Sulesomab, Becker et al. found a sensitivity of 90%, a specificity of 84.6% and a diagnostic accuracy of 87.9%, again in a heterogeneous group of patients with suspected musculoskeletal infections.12

ARTICLE IN PRESS 50

K.P. Iyengar et al.

Figure 1 A 30-year-old male patient had a closed fracture of tibia treated with intra-medullary nailing. Since he was complaining of pain at the fracture 6 months following surgery, he underwent nuclear imaging. 99mTc bone scan showed increased uptake at the fracture site, suggestive of infection. However, 99mTc-Sulesomab imaging revealed no increase or abnormal uptake. The patient’s fracture healed well without any further problems. This represents a true negative scan.

infection rather than confirming it in patients with suspected infection of prosthetic joints.14 This needs further evaluation.15

Practice points

 Nuclear medicine imaging plays an important role in   



early and accurate diagnosis of orthopaedic infection WBC imaging presently is considered gold standard Immunoscintigraphy has an emerging role 99m Tc-Sulesomab imaging is rapid, simpler and safer to use with negligible HAMA response rate and accuracy comparable to WBC scanning but with better quality images Immunoscintigraphy has the potential to replace WBC scans for infection imaging

Research directions Figure 2 A 45-year-old female patient had pain in the left knee 2 years after a left total knee replacement. The 99mTcSulesomab scan shows increased tracer uptake around the prosthesis, suggestive of infection. The patient responded well to antibiotics. This represents a true positive scan.

 Continuing and further evaluation of clinical safety

Experience at our institution reveals a sensitivity of 96.7%, specificity of 85.1% and an accuracy of 89.5%, along with a significant negative predictive value of 98.2% including those of long bones, joints and diabetic feet.13 This emphasises the emerging role of 99mTc-Sulesomab imaging in the evaluation of variety of orthopaedic infections. The accuracy of 99mTc-Sulesomab in the evaluation of suspected prosthetic joint infections was 81.5% with a sensitivity of 91%. With a high negative predictive value of 96% 99mTc-Sulesomab seems to be useful in excluding





especially regarding HAMA response and implications to long-term prognosis after undergoing the test Cost effectiveness in the context of manpower and ease of use Collaboration across centres would help the immunoscintigraphy results to be evaluated in a wider perspective (maybe with multi-centre trials)

References 1. Hakki S, Harwood SJ, Morrissey MA, Camblin JG, Laven DL, Webster WB. Comparative study of monoclonal antibody scan in diagnosing orthopaedic infection. Clin Orthop Relat Res 1996; 335:275–85.

ARTICLE IN PRESS Functional imaging in orthopaedic infections—Update on immunoscintigraphy 2. Propst-Proctor SL, Dillingham MF, McDougall IR, Goodwin D. The white blood cell scan in orthopaedics. Clin Orthop Relat Res 1982;168:157–65. 3. Modic MT, Pflanze W, Fieglin DH, Belhobeck G. Magnetic resonance imaging of musculoskeletal infection. Radiol Clin N Am 1986;24:247–58. 4. Calleja M, Alam A, Wilson D, Bradley K. Basic science: nuclear medicine in skeletal imaging. Curr Orthop 2005;19:34–9. 5. Turpin S, Lambert R. Role of scintigraphy in musculoskeletal and spinal infections. Radiol Clin N Am 2001;39(2):169–89. 6. Hans Van der Wall. Assessment of infection. Murray IPC, Ell PJ, editors. Nuclear medicine in clinical diagnosis and treatment, vol. 2. New York: Churchill Livingstone; 1994. p. 963–81. 7. Al-Sheikh W, Sfakianakis GN. Subacute and chronic bone infections: diagnosis using In-111, Ga-67 and Tc-99mMDP bone scintigraphy and radiology. Radiology 1985;155:501–6. 8. Schauwecker DS, et al. Evaluation of complicating osteomyelitis with 99mTc-MDP, 111-In granulocytes and 67Ga citrate. J Nucl Med 1984;25:849–53.

51

9. Merkel KD, Brown ML, Dewanjee MK, Fitzgerald RH. Comparison of In WBC imaging with sequential TcGallium scanning in diagnosis of low grade musculoskeletal sepsis. J Bone Jt Surg 1985;67A:465–76. 10. Gupta N, Prezio JA. Radionuclide imaging in osteomyelitis. Semin Nucl Med 1988;XVIII:287–99. 11. Becker W, Goldenberg DM, Wolf F. The use of monoclonal antibodies and antibody fragment in imaging infectious lesions. Semin Nucl Med 1994;24:142–53. 12. Becker W, Palestro CJ, et al. Rapid imaging of infections with a monoclonal antibody fragment. Clin Orthop Relat Res 1996; 329:263–72. 13. Iyengar KP, Nadkarni J, Vinjamuri S. Role of 99mTc-Sulesomab in the diagnosis of bone and joint infection. World J Nucl Med 2005;4:92–8. 14. Iyengar KP, Vinjamuri S. Role of 99mTc Sulesomab in the diagnosis of prosthetic joint infections. Nucl Med Commun 2005;26(6):489–96. 15. Ryan PJ. Leukoscan for orthopaedic imaging in clinical practice. Nucl Med Nucl Commun 2002;23:707–14.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 52–58

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

BASIC SCIENCE

The thermal properties of bone and the effects of surgical intervention S. Karmani Orthopaedic Department, Royal Surrey County Hospital, Egerton Road, Guildford, Surrey GU2 7XX, UK

KEYWORDS Bone; Drilling; Thermal effects

Summary Much is written about the mechanical properties of bone; we are aware of how bone responds to different modes and degrees of stress and can use this knowledge to influence the mechanical environment around bones and joints for therapeutic effect. The thermal properties of bone are less well documented, yet all orthopaedic interventions produce significant thermal effects. This paper explains basic material thermodynamics and how this relates to bone and impacts on an orthopaedic surgeon’s practice. & 2005 Elsevier Ltd. All rights reserved.

Introduction Bone is a complex biological tissue, with organic and mineral phases. The interaction of the different phases of bone account for its unique, complex mechanical properties. Bone is equally complex from a thermological perspective. This behaviour is difficult to study, sensitive to testing conditions, specimen preparation and is anisotropic.1 Heat is thermal energy transferred between a system and its surroundings. The transfer can take place by three different mechanisms:

In physics, materials can be defined thermologically by certain parameters: Specific heat: The energy required to raise the temperature of a system by 1 1C. More specifically this is defined per unit volume or unit mass. It is effectively a measure of how easily a material ‘‘heats up’’. Thermal conductivity: The thermal energy transfer in steady state per unit area and unit time between two infinite parallel planes, the distance and temperature gradient between which are 1 unit length and 11, respectively. It represents the ability of a material to transport heat. Various researchers have calculated these parameters for bone2 (Table 1).


Conduction—Thermal energy is transferred through the substance of the system.


Convection—Thermal energy is transferred by relative


motion of components of the system. Radiation—Thermal energy is transferred directly between separated parts of the system by electromagnetic radiation.

0268-0890/$ - see front matter & 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2005.09.011

The biological effects of excessive temperature on bone The importance of heat generation during bone drilling has long been recognised. In his theory and practice of

ARTICLE IN PRESS The thermal properties of bone and the effects of surgical intervention

Table 1

The thermophysical properties of bone.

Animal species

Man Elephant Ox Dog

53

Specific heat (cal/g 1C)

Thermal conductivity (cal/cm s 1C)

Fresh bone

Dry bone

0.3070.01 0.30

0.3070.01

0.2770.01 0.3070.01

medicine, written about 500 BC, Hippocrates recommended to drill slowly, to remove the tool frequently and plunge it in cold water to prevent the bone from heating when trephining discs of bone from the skull.3 In 1941, Gillies4 noticed necrosis around pins inserted into bone, which he attributed to heat from the drill. Anderson and Finlayson5 1943, proposed the term aseptic necrosis for local cauterisation, which they observed after pin insertion. The effect of heat on bone tissue depends on the temperature and duration of exposure.6 Direct comparisons between various investigations analysing the thresholds for thermal tissue injury to bone are difficult to make as studies have employed various exposure times, different criteria for tissue injury and observation at disparate points. The mechanisms underlying the tissue effects of thermal injury are not completely understood, but the main cellular factors are considered to be changes in protoplasmic proteins, with inactivation of enzymes and metabolic processes, and alterations in protoplasmic lipids. At higher temperatures phase changes may occur, with water being transformed into a vaporous state, resulting in dehydration, dessication, shrinkage, membrane rupture and finally carbonisation.7 Great differences exist in the thermal stability of different proteins. The enzyme lysyl-tRNA-synthetase, which is vital for cell function, is highly temperature sensitive. It has a half life of less than 5 min at 45 1C, with some denaturation occurring at 40 1C.8 Collagen is much more resistant, with a reported denaturation temperature of 60 1C.9 Bone alkaline phosphatase denatures in vitro at temperatures above 561 C for 10 s.10 Lundskog2 studied the diaphorase activity of bone tissue subjected to point thermal insults at various temperatures and exposure times. Diaphorase activity is a function of the enzymes NADH2 and NADPH which are essential to cellular aerobic and anaerobic function, respectively. In this study the vitality of a bone specimen was defined by the activity of these enzymes in the sample. Lundskog2 first studied exposure time, exposing bone specimens to 80 1C for 5, 10, 15, 25 and 30 s, respectively. He reported on the radius of cellular deactivation as determined by the absence of diaphorase activity in the osteocytes from the point heat source. There was a definite effect on the diaphorase activity of osteocytes at all exposure times, the radius of deactivation increasing with time. Exposure temperature was similarly assessed with a constant exposure time of 30 s. Again a definite dose response relationship was demonstrated, diaphorase deactivation being initiated at 50 1C.

0.2870.02 0.2870.01 0.2670.01

8.50  10 5.45  10 10.0  10 11.7  10

At the histological level, Erikson and Albrektsson11 noted cortical necrosis in living rabbit bone heated to 47 1C for 1 min. Cortical necrosis and a delay in the healing of surgical defects was reported for dog femora that were heated to between 43.3 and 68.7 1C by ultrasound after surgery.12 Rouiller and Majno13 described necrosis of osteocytes in the long bones of rabbits when they were exposed to a temperature of 55 1C for 1 min. Bonfield and Li14 reported that irreversible changes to the mechanical properties of bone occur when dog femora are heated to 50 1C in vivo. They attributed this to a reorientation of collagen molecules secondary to a weakening of bonds between collagen and hydroxyapatite. The effect of heat trauma has been most extensively studied in relation to skin. Moritz and Henriques15 concluded that the threshold temperature for epidermal necrosis at an exposure time of 30 s was 52–55 1C. Dermal blood vessels were more sensitive to thermal trauma than epithelial cells, with a threshold temperature for necrosis of 41–45 1 Leach et al.16 concluded that temperatures of 47 1C appear to kill all cell types. Between 42 and 47 1C death of all cells types can be produced, provided the time of heating is long enough. From the above discussion, it can be seen that the conditions for thermal bone tissue necrosis may be variable. Therefore, there is no agreement in the literature as to the threshold for thermal necrosis in bone and the ideal conditions to study this. A review of the literature reveals that many authors17–19 have taken 50 1C as the point at which bone necrosis occurs. This conclusion is drawn from the work of Lundskog2 and an average of the temperatures reported in the literature as being capable of causing necrosis.

Heat generation during drilling The drilling process The heat generated during drilling of bone comes from, firstly, the drilling process. Shear of the surface layer of a material by a drill bit breaks intermolecular bonds releasing energy. Secondly, friction from the non-cutting surfaces of a twist drill, such as the flank, flutes and shaft are another source of heat. The heat generated is partially dissipated by the presence of blood and tissue fluid, and part of the heat being carried away by the chips formed. However, bone is a poor conductor of heat and the temperature rise can be significant.20

ARTICLE IN PRESS 54

S. Karmani

The temperature developed within a section of bone tissue depends on specific heat capacity (how readily a material heats up) and on the thermal conductivity (how readily a material transfers heat from a source). When a heat source is applied to a specimen, a thermal gradient exists from the hotter heat source to the cooler specimen. Heat is transferred down this gradient. The higher the gradient the greater the rate of transfer. Heat continues to transfer as long as the gradient is maintained. The ultimate temperature a specimen develops depends on the balance between heat gain and heat loss. When these two phenomena are equal, the specimen is in a state of equilibrium. In the equilibrium state the temperature reached by a specimen is a function of the exposure temperature and thermal conductivity. Temperature equilibrium is, however, not normally reached when biological tissues are exposed to heat in a clinical situation such as drilling, as the exposure times are usually too short. The temperature reached in the non-equilibrium state of short exposures is a function of the exposure temperature and time, the specific heat capacity being constant.2 Different time-temperature relationships can produce similar effects in a specimen. Lundskog2 was able to measure the density of isotherms (lines connecting points with the same temperature) around a point heat source applied to a bone sample at various temperatures and exposures. He showed that the width of the 501 isotherm was approximately the same for such different thermal exposures as 901 for 10 s, 801 for 20 s and 751 for 30 s. As the exposure time shortens, so the distribution of isotherms becomes more compact. The reason for this is that heat requires a finite time to penetrate the tissue by conduction. The shorter the exposure time the more localised the heating process. The drilling process generates heat within the drill and in the bone. The rate of heat generation depends on various parameters of the cutting process, and these in turn influence the temperature of the tool, bone and bone chips, depending on their relative thermal properties.

Drill temperature The general formula correlating tool temperature Ti is given by21 T i ¼ C0 K s v 2n An ;

W 2n h1;2n

where Ks is the specific cutting energy, v the cutting velocity, A the chip cross-sectional area, W the thermal conductivity of work material, and h the thermal capacity (density  specific heat) of work material. C0 and n are constants and can be obtained experimentally. It is important to observe from the equation that the temperature Ti is directly proportional to the specific cutting energy Ks, which in turn is dependent on the dynamic shear strength of the material, since cutting is a dynamic shear failure process. The specific cutting energy increases with increase in the dynamic shear strength for a brittle material. As bone is a brittle material this can be assumed to apply.

Bone temperature The thermal capacity and conductivity of bone constitutes the other important factors affecting the temperature rise. These factors have an inverse relationship with the temperature. The lower the values, the higher the temperatures generated at the drill-bone interface. The thermal conductivity of the tool material has little effect on the temperature. When drilling metals the metal chips carry away nearly 85% or more of the heat generated. Unlike drilling metals, the temperature of the bone will rise more during the process due to its poor thermal capacity and conductivity, chips carrying away a smaller percentage of the heat. It should be pointed out that compared to the situation with metals, the total heat generated under identical heating conditions will be much less during bone drilling due to its lower K value. This is an important consideration for bone drill design22 (Table 2).

Factors effecting the temperatures generated when drilling bone The effect of drill force The literature is not clear regarding the influence of drill force on bone temperature. Some authors have found that the temperature rise or specific energy (energy required to drill a specific mass of material) decreases with force or feed rate (rate of drill advancement per unit time or per revolution of drill). Mathews and Hirsch,18 for example, applied loads of upto 118 N while drilling at various speeds in human cadaveric cortical bone and found that low temperatures are associated with high forces. They suggested that a high force causes the drill to cut more rapidly. Hence the drilling time is shorter and the temperature rise is smaller. Wiggins34 and Wiggins and Malkin35 found that specific energy is lower at higher feed rates; temperatures were not measured, but they would likely have been lower because less energy was consumed. This was confirmed in a later study by Krause et al.19 who drilled in bovine femoral cortical bone and showed that the temperature decreases as the feed rate increases. Other studies have found that temperature increases with force. Eichler and Berg,26 for example drilled in cortical sections of bovine femurs and recorded temperatures that increased with force. Henschel36 and Peyton37 investigated drilling in teeth and found that the temperature rises with force. However, in these investigations, the forces used were less than 30 N, much lower than the maximum force of 118 N applied by Mathews and Hirsch.18 In contrast to these results, Krause19 found that the specific cutting energy initially increases with feed rate and then decreases. His study, however, pertains to orthogonal bone cutting (the tool moves in a straight path parallel to the surface of the specimen) and not to drilling. Krause19 hypothesised that the decrease in specific cutting energy at high velocities is the result of changes in the material’s fracture properties. Abouzgia and James6 studied the temperature rise in bovine femurs using a surgical drill operating at 49,000 rpm with forces ranging from 1.5–9 N. They demonstrated that

Max temp (1C)

65.5 65 23.5 140 o50 95 38 300 76 89 74 50 460 55 130 41 57 96 185 33.8

Thompson23 Pallan24 Rafel25 Mathews and Hirsch18 Mathews and Hirsch18 Eichler and Berg26 Jacobs and Ray27 Tetsch28 Tetsch28 Lavelle and Wedgewood29 Lavelle and Wedgewood29 Lavelle and Wedgewood29 Pal and Saha30 Krause et al.19 Krause et al.19 Eriksson et al.31 Eriksson et al.31 Eriksson et al.31 Mathews et al.32 Eriksson and Adel33 2.5 2 3 0.5 0.5 0.5 2 1 1 0.5 0.5 0.5 1 Near Near 0.5 0.5 0.5 0.5 0.5

Distance from drill periphery

Temperatures recorded in bone during drilling.

Study

Table 2

Dog mandible Dog mandible Human mandible Human femur Human femur Human femur Rat radius Cat mandible Cat mandible Human femur Human femur Human femur Bovine long bones Bovine femur Bovine femur Rabbit femur Dog femur Human femur Human femur Human mandible

Bone type

In In In In In In in In In In In In In In In In In In In In

vivo vivo vitro vitro vitro vitro vivo vivo vivo vitro vitro vitro vitro vitro vitro vivo vivo vivo vitro vivo

Type of study

No No No No Yes No Yes No Yes No Ext irrigation Int irrigation No No No Yes Yes Yes No Yes

Cooling

125–2000 125–2000 35,000 345,885 2900 700 2500 20,000 20,000 350 350 350 65–2800 20,000 100,000 20,000 20,000 20,000 60–700 1500–2000

Free running drill speed rpm

20, 59 N 10, 20 and 30 N Not indicated Not indicated Not indicated 59 N 59 N 59 N 0.128 mm/rev 1.8–6.36 mm/s 1.8–6.36 mm/s Not indicated Not indicated Not indicated 60–120 N Low and intermittent

Not indicated Not indicated Not indicated

Force or feed rate

ARTICLE IN PRESS

The thermal properties of bone and the effects of surgical intervention 55

ARTICLE IN PRESS 56 the temperature rises with force to a certain point and then falls with greater force. The rise and fall of temperature with force is the result of competing factors. The total heat generated is the product of the rate of heat generation and the duration of drilling. The rate of heat increases with load, while the duration of drilling decreases. According to his data, when the force is high the duration is the dominant factor in temperature determination, while when the force is low the rate of heat generation is more important in determining peak temperatures. Conflicting results about the dependence of temperature rise on force may be the result, in part, of the wide variation in forces applied in different studies. A similar variation has been found in clinical studies. Hobkirk and Rusiniak38 measured the force exerted by 20 dental surgeons during drilling in bovine mandibular bone in vitro; they recorded maximum forces between 6 and 24 N. Much higher forces were recorded by Mathews et al.39 Dental burs cut in a different manner to orthopaedic drills, however, so the two processes may not be comparable.

S. Karmani signs of blunting of the cutting edges. The worn drills showed greater maximum temperature elevations and longer durations of temperature elevation. Blockage of the flutes of a drill also reduces its drilling efficiency. Natali et al.40 compared standard orthopaedic drills with blocked flutes with normal drills. The peak temperatures and the duration of drilling was significantly greater for blocked drills. Flutes are essential to take material away from the cutting zone.20 This removes heated debri and reduces friction. The flutes of a twist drill often tend to clog when the depth of the hole being drilled becomes appreciable compared to it diameter. Once clogged, friction increases excessively and over heating or even charring of the organic matrix of bone may result.30 Wiggins and Malkin35 using a general purpose twist drill have shown that the cutting torque and the specific cutting energy increases with increasing hole depth while drilling in the axial direction. The torque varies linearly and the energy exponentially. The torque may increase further with clogging.

The effect of coolants The effect of drill speed The amount of heat generated by a cutting bur is dependent upon the frictional forces and shearing forces at the cutting edges. Therefore the amount of heat generated is related to the number of revolutions in the cutting material and the number of cutting edges on the bur. Mathews et al.39 studied the temperature of a variety of skeletal pins inserted under different rotational speeds and feed rates. They found that an increase in rotational speed from 300 to 700 rpm increased the total cutting bone-bur contact. This supposedly increased the amount of friction and heat generated; however, this increase in temperature was only observed 0.5 mm from the drill tip. Krause et al.19 studied various burs at different speeds and found a decrease in temperature with increased operating speed for certain burs only. He concluded that increases in feed rate had greater effects on temperature response than rotational speed. Research in the dental literature has shown an increase in rotational speed reduced heat generation. However, dental burs operate at speeds of 3600–7500 rpm compared to orthopaedic drills at 60–800 rpm. The forces used are also different, dental 6–24 N and orthopaedic 60–120 N,9 so comparing the two situations is not straightforward. Many orthopaedic researchers have failed to show a relationship between drill speed and temperature elevation.18 A possible factor in the varied relationship between temperature and rotational speed is that the rotational speed of an electric drill depends on the applied force. Abouzgia and James6 measured the operating speeds of various drills and found them to be at times as low as 50% of the operating speed depending on the force applied. Therefore apparent rotational speeds may not be the actual speeds.

The effect of drill condition The condition of the drill is important. Mathews and Hirsch18 compared the temperatures generated by new drill bits to drills which had drilled more than 200 holes and showed

The use of coolants can minimise temperature elevations during bone drilling. The mean temperature of a high powered burr can be kept at or below 20 1C using cold Ringer’s irrigation. The temperatures achieved without irrigation for the same situation can be greater than 80 1C.41 In surgical practice both cortices of a bone are drilled to insert a screw. Irrigation can effectively cool the near cortex but the far cortex is difficult to cool. Mathews and Hirsch18 examined methods of cooling. Manual irrigation by an assistant is effective in cooling if a drill guide is not being used. Peak temperatures were reduced from 65 to 45 1C. When a drill guide was used, manual irrigation was only effective when the coolant was injected through the guide at a rate of 500 ml/min or greater, only then was it possible to keep temperature elevations below 50 1C. They recommended routine irrigation. When a drill guide was being used, they recommended that the drill guide be removed or back up the shaft of the drill to allow irrigation. If this was not possible, the drill guide should be modified to allow irrigation through the guide.

The effect of bone on drilling temperatures Bone is thermally anisotropic. In the study of Abouzgia et al.6 the temperatures recorded in the longitudinal direction were consistently higher than those recorded in the transverse direction. Zelenov1, investigating the thermophysical properties of human cadaveric femoral cortical bone, demonstrated a similar result. Lundskog’s2 work using infrared thermography to investigate temperature distribution in heterogenous specimens and in specimens perforated with drill holes found that the shape of isotherms was essentially circular and thus his bone specimens did not demonstrate thermal anisotropy. Lundskog’s2 work was performed on dry unspecified sections of elephant bone. The thermal conductivity of wet bone is approximately four times greater than dry bone and anisotropic according to Abouzgia.6 Cortical bone is denser than cancellous bone and drilling the former results in the generation of higher temperatures.

ARTICLE IN PRESS The thermal properties of bone and the effects of surgical intervention Cortical thickness is a major determinant of the peak temperatures generated when drilling. Erikson11 measured peak temperatures during drilling dog femur 56 1C; rabbit femur 40 1C and human femur 89 1C, he attributed the differences to cortical thickness. The mean cortical thickness of human femur being 6.5 mm, dog 3.5 mm and rabbit 1.5 mm, respectively. Cortical bloodflow in vivo may dissipate some heat produced by drilling during operative procedures. It is felt, however, that this cooling effect is unlikely to be significant. Cortical bloodflow is very low normally (2–3 ml/100 gm)42 and during drilling coagulation and occlusion of the small vessels probably occurs rapidly. Human red cells in vitro are totally haemolysed at 65 1C for 1.5 min.43 Mathews and Hirsch18 compared temperatures generated when drilling human femora in vivo and in vitro and found them to be equivalent. Prolonged ischaemia, as can occur when using a tourniquet, however, raises the threshold for thermal injury in bone.2,18 Interstitial fluid, particularly in cancellous bone, may also have a cooling effect. This can be mimicked by using wet bone specimens. Soft tissues surrounding bone are another insulating source or heatsink that can protect bone from excessive temperatures. This protective effect is lost in laboratory based studies. Predrilling the hole with a small diameter drill prior to enlarging the hole to the desired diameter is an effective way of reducing temperature rises. Peak temperatures can be reduced from 108 1C (3.2 mm drill) to 46 1C when the hole is predrilled with a 2.2 mm drill first.18,39 Smaller diameter drills require less energy to penetrate bone and thus generate less heat.21

Conclusion Bone is complex thermal tissue, the understanding of which is improving with research. All orthopaedic interventions produce significant thermal effects, the consequences of which can be far-reaching. Being more aware of the thermal effects of surgery allows one to control the possible detrimental effects, minimising the impact on our results.

References 1. Zelenov ES. Thermophysical properties of compact bone. Mech Compos Mater 1985;21(6):1092–5. 2. Lundskog J. Heat and bone tissue. An experimental investigation of the thermal properties of bone and threshold levels from thermal injury. Scand J Plast Reconstr Surg 1972;6(Suppl):5–75. 3. Phillips ED. Greek medicine. London: Camelot Press Ltd; 1973. 4. Gillies HD. The replacement and control of maxillofacial fractures. Br Dent Jt 1941;71:351–9. 5. Anderson R, Finlayson BL. Sequelae of transfixation of bone. Surgery 1943;13:46–54. 6. Abouzgia MB, James F. Temperature rise during drilling through bone. Int J Oral Maxfac Implants 1997;12(3):342–52. 7. Kuhns JG, Hayes J, Stein M, Helwig EB. Laser injury in skin. Lab Invest 1967;17:1. 8. Rymo L, Lagerkvist U, Wonacott A. Crystalisation of lysly transfer ribonucleic acid synthetase from yeast. J Biol Chem 1970;245:4308.

57

9. Viidik A. Functional properties of collagenous tissues. International review of connective tissue research, vol. 4. New York: Academic Press; 1972. 10. Eriksson RA. Heat induced bone tissue injury. Thesis, University of Gothenberg, Sweden. 11. Eriksson RA, Albrektsson T. Temperature threshold levels for heat induced bone tissue injury. A vital microscopic study in rabbit. J Prosth Dent 1983;50:101–7. 12. Ardan NI, Janes JM, Herrick JF. Ultrasonic energy and surgically produced defects in bone. J Bone Joint Surg 1957;39A:394–402. 13. Rouiller C, Majno G. Morphologische und Chemische Untersuchungen an knochen nach hitzecinwirkung. Beitr Z Path 1953; 113:100–20. 14. Bonfield W, Li CH. The temperature dependence of the deformation of bone. J Biomech 1968;1:323–9. 15. Mortiz AR, Henriques Jr. FC. Studies on thermal tissue injury III. The pathology and pathogenesis of cutaneous burns. An experimental study. Am J Path 1947;23:915. 16. Leach EH, Peters RA, Rossiter RJ. Experimental thermal burns, ecspecially the moderate temperature burn. Q J Exp Physiol 1943;32:67. 17. Bachus KN, Rondina MT, Hutchinson DT. The effects of drilling force on cortical temperatures and their duration. An in vitro study. Med Eng Phys 2000;22:685–91. 18. Mathews LS, Hirsch C. Temperature measured in human cortical bone when drilling. J Bone Joint Surg 1972;54(A):297–308. 19. Krause WR, Bradbury DW, Kelly JE, Lunceford EM. Temperature elevations during orthopaedic cutting operations. J Biomech 1982;15:267–75. 20. Saha S, Pal S, Albright JA. Surgical drilling: design and performance of an improved drill. J Biomec Eng 1982;104:245–52. 21. Kronenberg M. Machining science and application—theory and practice for operations and development of machining process, 1st ed. Oxford: Pergamon Press, p. 53. 22. Albright JA, Johnson TR, Saha S. In: Ghista DN, Raof, editors. Principles of internal fixation in orthopaedic mechanics; procedures and devices. New York: Academic Press; 1978. p. 124–229. 23. Thompson HC. Effect of drilling into bone. J Oral Surg 1958;16: 22–30. 24. Pallen FG. Histologic changes in bone after insertion of skeletal fixation pins. J Oral Surg Aneasth Hosp Dent Serv 1960;18: 400–8. 25. Rafel SS. Temperature changes during high speed drilling on bone. J Oral Surg Aneasth Hosp Dent Serv 1962;20:475–7. 26. Eichler J, Berg R. Temperatureinwirkung auf die Kompakta beim Bohren. Gewindeschneiden und Eindrehen von Schrauben. Z Orthop 1972;110:909–13. 27. Jacobs RL, Ray RD. The effect of heat on bone healing. A disadvantage in the use of power tools. Arch Surg 1972;104: 687–91. 28. Tetsch P. Development of raised temperature after osteotomies. J Maxillofac Surg 1974;2:141–5. 29. Lavelle C, Wedgewood D. Effect of internal irrigation on frictional heat generated from bone drilling. J Oral Surg 1980;38:499–503. 30. Pal S, Saha S. Effect of cutting speeds on temperature during drilling of bone. Proc Am Coll Eng Med Biol 1981;23:289. 31. Eriksson RA, Albrektsson T, Albrektsson B. Heat caused by cortical one drilling. Temperatures measured in vivo inpatients and animals. Acta Othop Scand 1984;55:629–31. 32. Mathews LS, Green CA, Goldstine SA. The thermal effects of skeletal fixation pin insertion in bone. J Bone Joint Surg 1984;66(A):1077–83. 33. Eriksson RA, Adell R. Temperature during drilling for placement of implants using osseointegration technique. J Oral Maxillofac Surg 1986;44:4–7. 34. Wiggins KL. Machining of bone. Thesis, Austin, TX: University Texas; 1974.

ARTICLE IN PRESS 58 35. Wiggins KL, Malkin S. Orthogonalmachining of bone. J Biomech Eng 1978;100:122–30. 36. Henschel CJ. Heat impact of revolving instruments on vital dentin tubules. J Dent Res 1943;22:323–33. 37. Peyton FA. Temperature rise and cutting efficiency of rotating instruments. NY Dent J 1952;18:439–50. 38. Hobkirk JA, Rusiniak K. Investigation of variable factors in drilling bone. J Oral Surg 1977;35:968–73. 39. Mathews LS, Green CA, Goldstine SA. The thermal effects of skeletal fixation pin insertion in bone. J Bone Joint Surg 1984;66(A):1077–83.

S. Karmani 40. Natali C, Ingle I, Dowell J. Orthopaedic bone drills—can they be improved? J Bone Joint Surg 1996;78B:357–62. 41. Kondo S, Yoshikazu O, Iseki H, Hori T, Takakura K, Kobayashi A, Napata H. Thermological study of drilling bone tissue with a high speed drill. Neusosurgery 2000;46(5): 1162–7. 42. Folkow B, Neil E. Circulation. New York: Oxford University Press; 1971. 43. Moritz AR. Studies of thermal injury III. The pathology and pathogenesis of cutaneous burns. An experimental study. Am J Path 1947;23:695.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 59–71

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

TRAUMA

Ilizarov and trauma reconstruction N. Shortt, G.F. Keenan New Royal Infirmary of Edinburgh, Dalkeith Rd, Edinburgh, UK

KEYWORDS Ilizarov; Non-union; Infected non-union; Malunion

Summary The Ilizarov method can be used to reform missing bone and soft tissue, whether bone loss has occurred because of trauma, or if bone has to be resected because of infection or nonunion. Any deformity of bone such as shortening, angulation or rotation may be addressed, often simultaneously, with correction of the deformity either in isolation or while treating non-union. The Ilizarov technique offers the opportunity to cure conditions not readily manageable by conventional internal fixation techniques. However, it does require specialist training and can be associated with its own set of problems and complications. & 2005 Elsevier Ltd. All rights reserved.

Introduction The Ilizarov technique has now been used in the UK and the US for 15–20 years. It offers an effective and reliable treatment for some of the most challenging conditions in orthopaedics, such as infected non-union of long bones1,2 and malunion.3 The Ilizarov method uses fine wires, inserted percutaneously into bone, which are attached to a circular frame and tensioned to provide a strong stable frame construct.4 The Ilizarov external fixator is a highly flexible apparatus which, due to the modularity of its components, can be constructed to correct any deformity or to address any mechanical problems. Gavril Ilizarov5 developed this method out of necessity, because of poor resources and a large population, in Kurgan in Russia. He first treated patients with this method in 1950s but only became well known in Russia after treating the

Corresponding author.

0268-0890/$ - see front matter & 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2005.09.002

Russian high jump gold medallist Valery Brummel. The athlete had sustained an open tibial fracture and subsequently developed an infected non-union that Moscow surgeons had failed to treat successfully. After treatment by the Ilizarov method Brummel returned to compete in the Olympic Games.

Basic science Ilizarov performed a large variety of experiments from the early 1950s and developed the principle of the tension-stress effect. An osteotomy performed through bone followed by slow controlled tension applied through soft tissues stimulated increased proliferative and metabolic activity in all tissue types. Ilizarov observed that the collagen produced was orientated along the plane of stress applied through the tissues. Subsequent formation of bone regenerate then consolidated proximally and distally, furthest away from the central area of regenerate bone, which remains highly active in regenerate production

ARTICLE IN PRESS 60

N. Shortt, G.F. Keenan

Table 1

      

Key factors in the Ilizarov method.

Stable frame construct Low energy corticotomy Preservation of periosteum Preservation of marrow Latent period Low rate of distraction (0.5–1 mm daily) Frequency of distraction

Table 2

Spectrum of non-unions.

Favourable factors

     

Hypertrophic Stiff Non infected Good soft tissue cover No mal-alignment Healthy host

Table 3

        

produces histological changes. Nerves and blood vessels show temporary features of degeneration that will resolve after distraction is completed.6 The newly formed tissues were noted by Ilizarov to be similar in structure to embryonic and foetal equivalents. Ilizarov found several factors to be important for the optimal production of new tissues. The configuration of the circular frame should provide as much stability to the fracture or osteotomised bone as possible. Excess movement between the distracted bone ends secondary to inadequate frame constructs can cause bleeding into the regenerate with resultant formation of cystic areas or cartilaginous islands, leading to non-union of the regenerate tissue.6 There should also be as little disturbance to the surrounding

Unfavourable factors

      

Atrophic Mobile Infected Poor soft tissue Deformed Multiply operated Debilitated host

Spectrum of hosts.

Young Healthy Elderly Steroid treatment Radiotherapy NSAIDs Smoking Alcohol intake Poor nutritional status

(intramembranous ossification). The active central area contains type 1 collagen and osteoid-producing osteoblasts. The new bone is laid down along the collagen bundles and consolidates into columns of bone lying parallel to the distraction force.6 In addition to new bone formation, the tension-stress effect promotes intense angiogenesis.7 Capillaries form between the new columns of bone, again orientated parallel to the distraction force. This phenomenon is responsible for increased regional perfusion of the limb being treated when compared to the control limb. The increased blood flow persists for up to 4 months after the corticotomy has been performed.6 Distraction also results in the growth of new muscle, fascial, nerve and skin tissues.7 Muscle tissues tolerate lengthening well, but increasing limb length by 30% or more

Figure 1 Case 1: Radiographs of first intramedullary nailing of fracture. Note absence of callus.

ARTICLE IN PRESS Ilizarov and trauma reconstruction

61 to form bone and remodel quicker than diaphyseal sites, but, overall, 1 cm of limb lengthening requires an average of 3–4 months of treatment using the Ilizarov method, including distraction and consolidation phases (Table 1).

Technique The basic technique involves fixation of two or more segments of bone, and an osteotomy performed percutaneously. At operation, fine wires of 1.8 mm diameter are pushed through the soft tissues to the bone surface, then driven through the bone at low speed to avoid thermal damage to the bone. If bone were burnt by excessive heat from drilling then this would promote infection of pin sites during treatment and the possibility of ring sequestra formation. The wires are placed in such a position as to penetrate as little soft tissue as possible to avoid tethering during the corrective stage of treatment. Wires that have an olive present are used to direct the bone during treatment by effectively pushing or pulling bone segments into place. All

Figure 2 Case 1: Deformity occurring in cast brace.

soft tissue envelope as possible, in order to preserve the periosteal and endosteal blood supplies of the bone. Finally the rate and frequency of distraction should be such that the tension-stress effect is maintained without causing damage to tissues by over-aggressive traction, which can result in abnormal muscle function,8,9 abnormal nerve structure10 and conduction.6,11 Additionally, too rapid distraction induces significant pain for the patient. From Ilizarov’s research, the optimal rate of distraction is 1 mm per day in four equal increments. Slower rates will produce premature bone consolidation, while faster rates are associated with abnormal structure of the soft tissue envelope and production of fibrous tissue at the regenerate site, i.e. the development of a fibrous non-union. Plain radiographs allow monitoring of the distraction and consolidation phases of treatment. Numerous studies have shown that the distraction gap mineralises from the bone ends toward the central area of the regenerate bone. The central radiolucent zone should be maintained between 4 and 6 mm to avoid premature consolidation, a risk if the central gap measures 2 mm or less, or non-union if the regenerate gap is 8 mm or more.6 Metaphyseal sites tend

Figure 3 Case 1: At presentation—hypertrophic non-union and varus deformity.

ARTICLE IN PRESS 62

N. Shortt, G.F. Keenan

Figure 4 Diagram showing the effect of hinge placement on compression, distraction and lengthening. (i) Closing wedge with compression opposite hinge. (ii) Simultaneous compression and distraction with no lengthening. (iii) Distraction on concave aspect of deformity with no increase in length. (iv) Opening wedge with distraction. (v) Increased length during correction of deformity.

Figure 6 Case 1: Final radiographs.

Figure 5 Case 1: Radiographs showing deformity corrected and non-union converting to union.

the wires are attached onto a circular frame that is built to correspond to the wire positions. They are secured by clamping them onto the frame with slotted and cannulated

bolts. One end is secured as tightly as possible to the frame, then the opposite end tensioned before being secured onto the frame. When the appropriate site for the osteotomy is decided, a small incision is made to allow multiple drill holes to be made across the bone. Again, this is done at low speed to avoid thermal damage to the periosteal tissues. An Ilizarov osteotome is used to connect the drill holes and complete the osteotomy. This surgery is followed by a latent period of 7–10 days to allow the bone to start healing prior to correction. Thereafter distraction of the osteotomy begins and regenerate bone is formed, the technique being termed ‘‘distraction osteogenesis’’. During this phase of treatment,

ARTICLE IN PRESS Ilizarov and trauma reconstruction

63

close follow-up is required with repeated radiographs to ensure the correction is proceeding as planned, and to ensure the patient understands their prescription for correction of the deformity or distraction. There follows a consolidation phase once the new bone has been made, and using these techniques 1 cm of mature bone can be ‘regenerated’ in 40 days. By this method, segmental defects of bone can be reconstructed.12

Figure 7

Patient assessment The typical patient with a non-union will probably have lost their job, and be suffering chronic pain. They are also likely to be adversely affected psychologically, many describing depressive symptoms and occasionally marital discord or even divorce. They may have had two to three unsuccessful operations in an attempt to get their fracture to heal before they present at a tertiary

Case 2: Mobile non-union (pseudarthrosis).

ARTICLE IN PRESS 64 reconstruction clinic. Some delayed unions or non-unions are easier to deal with than others and, like the fractures that produced them, they can have different personalities (see Tables 2 and 3). Non-unions are often classified into hypertrophic and atrophic morphology.13 Ilizarov classified non-unions as stiff or mobile. Stiff correlates with a hypertrophic non-union that has fibrous tissue present, mobile indicates an atrophic non-union with very little fibrous tissue present. The hypertrophic non-unions can be stimulated to form bone by altering the mechanical environment, either by distracting the non-union or compressing it. Atrophic non-unions tend to need surgery at the non-union site to promote biological activity, and radical debridement of dead bone, particularly in the infected cases, is required. We present a series of cases treated in our unit to illustrate the techniques most frequently employed when using the Ilizarov method for reconstruction following trauma.

N. Shortt, G.F. Keenan

Techniques Case 1: Hypertrophic non-union This case illustrates how the Ilizarov technique can be used to treat a hypertrophic tibial shaft non-union. This 25 year old patient injured his leg playing football, sustaining a midshaft tibial fracture. The initial treatment was with an intramedullary nail (Fig. 1) performed at another institution, but the fracture site was opened at the time due to difficulties passing the guide wire. The initial nail was exchanged at 3 months, but infection was present necessitating nail removal and intramedullary reaming. The fracture at this point had not healed, and the patient was put in a cast brace (Fig. 2), which resulted in tibial angulation. He presented to the reconstruction clinic with shortening and a hypertrophic non-union with a varus deformity in the tibia (Fig. 3). Erythrocyte sedimentation rate and C-reactive protein blood tests were near normal on presentation. Biopsies were taken from the fracture site for culture and sensitivity, although at the time of frame application infection was not clinically apparent. The position of hinges on the Ilizarov frame can be used to produce a variety of effects on the position of the limb (Fig. 4). In this case, the patient’s tibia had a varus position as a result of the non-union. Placing a hinge directly over the site of the non-union would compress the convex aspect of the deformity while distracting the concave aspect. Overall there is no increase in limb length. By placing the hinge at the apex of the convexity, an effective opening wedge is created, distraction of the non-union occurs in the concave aspect of the varus deformity. This allows correction of the deformity with the stimulus of distraction promoting conversion of the fibrous tissue in the non-union into mature bone with good reconstruction of alignment (Figs. 5 and 6). It is possible to increase limb length and correct deformity simultaneously by placing the hinge at a distance away from the apex of the concavity. As deformity correction proceeds, there is distraction across the whole of the non-union or corticotomy site resulting in limb lengthening.

Case 2: Atrophic non-union with true pseudarthrosis

Figure 8 Case 2: Preoperative elbow movement. (i) Flexion. (ii) Extension.

This 45 year old patient sustained multiple injuries in a road traffic accident, including femoral, pelvic and upper limb injuries. Despite several operations for non-union (Fig. 7i–iii) of the left humerus, complicated by infection, the patient was left with a useless arm. The shoulder and elbow (Fig. 8i–iii) were stiff because of movement occurring at the non-union. Abnormal movement at the atrophic non-union site would make plate fixation fail before union occurred. At operation a true pseudarthrosis with a false capsule was seen, requiring extensive debridement before the bone ends were approximated in the Ilizarov frame (Fig. 9). The basic surgical principle of debridement of infected or dead soft tissue and bone has been applied. The Ilizarov apparatus in this case was used to provide a stable mechanical environment for the fracture and to enable the healthy bone ends

ARTICLE IN PRESS Ilizarov and trauma reconstruction

65

Figure 9 Case 2: Humeral non-union compressed in an Ilizarov frame.

Figure 10 in situ.

Case 2: Function of upper limb with Ilizarov frame Figure 11

Case 2: Final radiographs showing humeral union.

ARTICLE IN PRESS 66

N. Shortt, G.F. Keenan

to be compressed and allow the bone to heal. The excision of bone in this case did alter the overall length of the limb, but not enough to cause functional or cosmetic difficulties for the patient. The patient described immediate benefit from the frame. He had stability with useful function in the limb compared to the flail limb at initial presentation. The stability of the frame allowed intensive physiotherapy to mobilise the shoulder and elbow. (Fig. 10). In this case, debridement and freshening the bone ends followed by strong compression in the Ilizarov frame was sufficient to obtain union of the humerus (Fig. 11). Alternatively, this non-union could have been plated, but plaster protection of the fixation would be essential to avoid fixation failure. The physiotherapy regimen would also be delayed and there would be no facility to stimulate union by increasing compression over the course of the treatment.

Case 3: Infected non-union This case illustrates the Ilizarov technique when dealing with an infected non-union of the tibia. The 36 year old patient was involved in an RTA with multiple injuries, including an open fracture of the distal tibia. This was initially treated with wound debridement and nailing. Six weeks following nailing, an abscess appeared over the distal tibia, which was debrided, washed out and treated with antibiotics. Further episodes of infection occurred necessitating an exchange nailing procedure, performed 6 months following the fracture (Fig. 12). At this time, the question of amputation was discussed and the patient referred for an opinion to the reconstruction clinic. Clinically, the limb was hot, swollen and the patient’s sleep pattern was disturbed because of pain. The patient was advised that treatment in the Ilizarov frame would take between 8 and 12 months, depending on the amount of nonviable bone that would require resection and thus regeneration at a different site. This case required bifocal treatment of the infected nonunion. At operation, thorough debridement of the non-union site was performed, 5 cm of infected bone were resected and the healthy, bleeding bone ends were approximated and compressed by two distal rings. Dead and infected bone is easily distinguished by its hard marble-like quality causing the bone to splinter when resected with bone nibblers. Dead bone does not bleed. Normal bone and callus feels less brittle on resection with nibblers, and capillary bleeding known as the ‘‘paprika sign’’ is seen at the interface of healthy and dead bone. It is important to remember that most resections are performed under tourniquet control and further confirmation of bleeding bone is obtained when the tourniquet is deflated. At the same operation, a corticotomy was performed in the proximal tibia to allow lengthening (Fig. 13). The proximal site was lengthened 34 mm a day. The distal site meantime was healing under the influence of compression and weight bearing in the frame. The biological effect of distraction of the corticotomy, which is known to produce sustained increases in blood flow distal to the limb14 also contributes to the union. The regenerate

Figure 12 Case 3: Radiographs of intramedullary nailing of fracture. Note bone lysis at distal cross screws.

Figure 13 Diagram illustrating the compression distraction technique used for treating an infected non-union. (i) Infected non-union with an intramedullary nail. Note bone lysis distally. (ii) Bifocal treatment with proximal corticotomy and distal nonunion debridement and compression. (iii) Distal fracture union with restoration of length by corticotomy distraction.

ARTICLE IN PRESS Ilizarov and trauma reconstruction

67

Figure 14

Case 3: Final radiographs.

consolidated proximally and union of the distal site was seen (Fig. 14i and ii). The frame was removed and he was placed in plaster to protect the regenerate in the final stage of consolidation. If the debridement necessitates the resection of more than 5–6 cm of bone then acute shortening and compression, as described above, is not possible. The technique of bone transport is then used. In this situation, limb length is maintained by the Ilizarov frame. A corticotomy is performed and a segment of remaining bone is transported into the segmental defect. New bone is created by distraction osteogenesis in the corticotomy gap. Ultimately the transported segment of bone will contact distal bone at the docking site. This area is then compressed using the Ilizarov frame to promote bone union while the regenerate

bone consolidates to restore the bony integrity of the limb (Fig. 15).

Case 4: Malunion—severe tibial varus deformity This 40 year patient presented with a varus tibia and medial joint line pain. He had sustained an open fracture of the proximal tibia 10 years prior to this, treated with plastic surgery and a standard external fixator. The treatment was complicated by recalcitrant pin tract infections, requiring early removal of the fixator. The fracture had not completely healed at this time and the patient developed a progressive deformity in the immature callus, and he was left with 201 of tibial angulation (Fig. 16), giving him medial

ARTICLE IN PRESS 68

N. Shortt, G.F. Keenan

Figure 15 Diagram illustrating technique of bone transport in segmental bone loss. (i) Segmental bone loss. (ii) Proximal corticotomy with secure fixation distally to maintain limb length. (iii) Distal transport of proximal fragment to unite distal fracture by compression. Regenerate bone produced proximally to restore limb length.

Figure 17 Case 4: Application of Ilizarov frame with proximal corticotomy.

joint line pain and altering the mechanical axis of the lower limb. A frame was applied with a single ring to the proximal tibia and two rings placed distally. A percutaneous proximal corticotomy was performed (Fig. 17). The frame was left in situ for 7 days before beginning distraction. The axis of the deformity was identified and the centre of rotation and angulation (CORA) identified. The CORA is identified by the intersection of the mechanical axes of the fracture fragments. The position of the hinge at this point is then able to correct the translation and angulation of the malunion simultaneously. The hinges could be placed along the line of the rotational axis in order to produce distraction or compression as required (Figs. 18 and 19). In this case, effectively an opening wedge callotasis was performed, restoring good alignment. The patient walked with the frame fully weight bearing during treatment and noted that the medial joint line pain disappeared during the treatment as the mechanical axis was corrected (Fig. 20).

Case 5: Rotational deformity correction with Taylor Spatial Frame This 68 year old man sustained a low energy open fracture of the tibia (Fig. 21). The injury was treated by debridement and washout, and was stabilised using a standard external fixator. It had been planned to nail the tibia on the night of presentation but because there was a technical problem with the image intensifier the definitive surgery was deferred. The patient had pre-existing chest pathology,

Figure 16 Case 4: Long leg radiograph showing tibial alignment deformity.

ARTICLE IN PRESS Ilizarov and trauma reconstruction

69

Figure 18 Case 4: Diagram showing the effect of hinge placement on malunion with angulation and translation. Centre of rotation and angulation (CORA) lies above area of deformity.

Figure 20 Case 4: Correction of mechanical axis and translational deformity.

too much callus to reduce back to an anatomical position acutely. The main deformity on application of the Taylor Spatial Frame was rotation. It is important to note the 901 rotational deformity between proximal and distal fragments. This device is capable of simultaneous threedimensional deformity corrections. Using computer software and measuring deformity parameters from antero-posterior and lateral radiographs a schedule for reduction of the deformity is obtained which can produce very accurate reductions. The frame was removed 6 weeks after gradual reduction.

Complications of treatment using Ilizarov method

Figure 19 Case 4: Hinge placed at level of the CORA.

which precluded his early return to theatre and he continued with treatment using the external fixator. The fracture displaced and rotated in the fixator but had formed

The Ilizarov method is associated with its own set of complications (Table 4) that require regular follow-up and review in order to prevent or to detect and treat early and effectively. Studies have shown that the incidence of complications reduces with experience and becomes low after a surgeon has performed 70–80 cases. Paley has weighted the adverse events15 during treatment into early and late problems. Most commonly, superficial pin site infections and pain related to soft tissue tension during distraction are reported. The majority of superficial pin site infections can be treated with oral antibiotics although occasionally inpatient treatment with intravenous antibiotics or wire removal is necessary. Pain associated with

ARTICLE IN PRESS 70

N. Shortt, G.F. Keenan

Figure 21 Case 5: (a) Lateral view with external fixator in situ. (b) AP view with external fixator in situ. (c) AP view with Taylor Spatial Frame in situ with rotational deformity evident. (d) Lateral view with Taylor Spatial Frame in situ with deformity evident. (e) AP view with Taylor Spatial Frame in situ with rotational deformity corrected. (f) Laterial view with Taylor Spatial Frame in situ with rotational deformity corrected.

distraction is usually related to soft tissue tension and a period of reduced or no distraction will allow the soft tissues to recover before further distraction continues.

Conclusions The Ilizarov method offers the orthopaedic surgical team a powerful tool with which to undertake some of the most

challenging and complex cases seen in orthopaedic trauma practise today. It offers a realistic and successful treatment option for patients whose only other options may include amputation or functional bracing of an effectively useless limb. Although treatment times for the patients are long, averaging between 8 and 12 months, cost analysis has shown that limb reconstruction using the Ilizarov method is more cost effective than amputation and its associated prosthetics costs.16

ARTICLE IN PRESS Ilizarov and trauma reconstruction

Table 4

    

Complications of the Ilizarov method.

Pin site infection Pain with distraction Poor regenerate formation Delayed union at docking site Regenerate fracture or deformity after frame removal

The techniques do require a period of specialist training and a degree of experience in treating these complex cases is essential. The Ilizarov method has a high success rate due to the trophic effect of distraction on bone growth.17 Ilizarov frames allow continual and progressive manipulation of the mechanical environment to give a predictable treatment course, with resultant bone of good quality. Furthermore, the mechanical strength of the fixators allows the removal of all metalwork to ensure the eradication of pre-existing infection. The techniques also have their unique set of complications, many of which are avoidable through the use of sound surgical technique and regular patient review in the outpatient department.

References 1. Ilizarov GA, Kaplunov AG, Degtiarev VE, Lediaev VI. Treatment of pseudarthroses and ununited fractures complicated by purulent infection, by the method of compression–distraction osteosynthesis. Ortop Travmatol Protez 1972;33(11):10–4. 2. Ring D, Jupiter JB, Bing SG, Israeli R, Yaremchuk MJ. Infected nonunion of the Tibia. Clin Orthop 1999;369:302–11. 3. Paley D, Chaudry M, Pirone AM, Lentz P, Kautz D. Treatment of malunions and mal-nonunions of the femur and Tibia by detailed preoperative planning and the Ilizarov techniques. Orthop Clin North Am 1990;21(4):667–91.

71 4. Fleming B, Paley D, Kristiansen T, Pope M. A biomechanical analysis of the Ilizarov external fixator. Clin Orthop 1989;241:95–105. 5. Ilizarov GA. The principles of the Ilizarov method. Bull Hosp Joint Dis Orthop Inst 1988;48(1):1–11. 6. Aronson J. Current concepts review–limb lengthening, skeletal reconstruction, and bone transport with the Ilizarov method. J Bone Joint Surg (Br) 1997;79-A:1243–58. 7. Ilizarov GA. The tension-stress effect on the genesis and growth of tissues. Part 1. The influence of stability of fixation and softtissue preservation. Clin Orthop 1989;238:249–81. 8. Ilizarov GA. The tension-stress effect on the genesis and growth of tissues. Part 2. The influence of rate and frequency of distraction. Clin Orthop 1989;239:263–85. 9. Simpson AHRW, Williams PE, Kyberd P, Goldspink G, Kenwright J. The response of muscle to leg lengthening. J Bone Joint Surg (Br) 1995;77-B:630–6. 10. Galardi G, Comi G, Lozza L, Marchettini P, et al. Peripheral nerve damage during leg lengthening. J Bone Joint Surg (Br) 1990;72-B:121–4. 11. Ippolito E, Peretti G, Bellocci M, Farsetti P, et al. Histology and ultrastructure of arteries, veins, and peripheral nerves during limb lengthening. Clin Orthop 1994;308:54–62. 12. Ilizarov GA, Ledyaev VI. The replacement of long tubular bone defects by lengthening a distraction osteotomy of one of the fragments. 1969. Clin Orthop 1992(280):7–10. 13. Webber BG, Cech O. Pseudoarthrosis. Pathophysiolgy, biomechanics, therapy, results. Bern: Huber; 1976. 14. Sveshnikov AA, Zhirov IuA, Saks RG. Radionuclide study of blood circulation in extremities during treatment of fractures by the GA Ilizarov method. Med Radiol (Mosk) 1984;29(8):52–6. 15. Paley D. Problems, obstacles and complications of limb lengthening using the Ilizarov technique. Clin Orthop 1990; 250:81–104. 16. Williams MO. Long-term cost comparison of major limb salvage using the Ilizarov method versus amputation. Clin Orthop 1994;304:156–8. 17. Sveshnikov AA, Shved SI, Mingazova NB, Karagodin EG, et al. Radionuclide research on reparative bone formation during treatment of spiral bone fractures of the leg by GA Ilizarov method. Med Radiol (Mosk) 1985;30(8):41–7.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 72–74

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

CME SECTION Three CME points available The following series of questions are based on the Mini Symposium on Biomechanics. Please read the articles in the mini symposium carefully and then complete the self assessment questionnaire by filling in the square corresponding to your response to each multiple-choice question. For true or false questions, please fill in one square only. After completing the questionnaire, either post or fax the answer page back to the Current Orthopaedics Editorial Office at the address at the bottom of the RESPONSE sheet. Please photocopy this page if you wish to keep your copy of Current Orthopaedics intact. Replies received before the next issue of Current Orthopaedics is published will be marked and those reaching an adequate standard will qualify for three external CME points. You will be notified of your marks and a CME certificate will be despatched for your records

Questions 1. Which of the following is a vector quantity? A. B. C. D. E.

Density Mass Moment Temperature Time

2. Forces of 3 and 5 N act at 901 to each other. What is the magnitude of the resultant force? A. B. C. D. E.

1N 2N 3N 4N 5N

3. How many rotational degrees of freedom are there for a three dimensional object?

C. Stress D. Strain E. Work 5. Which of the following has the greatest Yield Strength? A. B. C. D. E.

Annealed 316 L stainless steel Cold worked 316 L stainless steel Cast CoCrMo Hot forged CoCrMo Ti6Al4V

6. What is responsible for most of the axial force acting on the long bones of the leg during activity? A. B. C. D. E.

The mass of the body above the S2 vertebra Total body weight Externally applied loads The ground reaction force Muscle contraction 7. Which of the following is the stiffest?

A. B. C. D. E.

None 1 2 3 6

4. Which of the following relates directly to the ability of a material to withstand the loads to which it is subjected? A. Elasticity B. Force 0268-0890/$ - see front matter doi:10.1016/j.cuor.2006.01.001

A. Articular cartilage from the surface layer, measured parallel to the joint surface B. Articular cartilage from the middle layer, measured in any direction C. Articular cartilage from the deep layer, measured perpendicular to the subchondral plate D. Meniscal cartilage, measured parallel to the peripheral attachment E. Meniscal cartilage, measured perpendicular to the free edge

ARTICLE IN PRESS CME SECTION 8. Approximately what proportion of the applied load is carried by the fluid phase of articular cartilage? A. B. C. D. E.

10% 50% 75% 90% 95%

9. In the load extension graph for a tendon, what does the toe-in region represent? A. Straightening out of the crimped collagen molecules B. Redistribution of fluid within the ground substance of the tendon C. Loading of different collagen bundles that are not oriented along the line of application of force D. Stretching of the collagen molecule along its length E. The first sign of collagen failure that will leave slight permanent damage to the tendon 10. What sort of wear is occurring when material is lost from a solid surface due to its relative motion in contact with a fluid which contains solid particles? A. Abrasive B. Adhesive

73 C. Fatigue D. Erosive E. Corrosive 11. With a femoral head made of stainless steel and a cup of UHMWPE, which of the following is not a consequence of selecting a 22 mm diameter head instead of a 28 mm head? A. B. C. D. E.

The frictional torque will be less There will be less linear penetration There will be a greater susceptibility to creep Volumetric wear will be less Reduced surface roughness will give a greater chance of fluid film lubrication occurring

12. Which of the following best describes the most likely location of the axes of rotation of a functional spinal unit? A. B. C. D. E.

In the centre of the inferior vertebral body In the lower half of the superior vertebral body In or just below the disc The facet joints The spinal canal

ARTICLE IN PRESS 74

CME SECTION

Please fill in your answers to the CME questionnaire above in the response section provided below. A return address and fax number is given at the bottom of the page. ...............................................................................................

Responses Please shade in the square for the correct answer. 1 2 3 4 5 6 7 8 9 10 11 12

A A A A A A A A A A A A

& & & & & & & & & & & &

B B B B B B B B B B B B

& & & & & & & & & & & &

C C C C C C C C C C C C

& & & & & & & & & & & &

D D D D D D D D D D D D

& & & & & & & & & & & &

E E E E E E E E E E E E

& & & & & & & & & & & &

Your details (Print clearly) NAME. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ADDRESS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . FAX NO. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . EMAIL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

RETURN THE COMPLETED RESPONSE FORM by fax to +44-113-206-6791, or by post to CME, Current Orthopaedics, Orthopaedic Surgery, Clinical Sciences Building, St. James’s University Hospital, Beckett Street, Leeds LS9 7TF, UK.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 75

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

CME SECTION

Answers to CME questions in Vol. 19, issue 5 Please find below the answers to the Current Orthopaedics CME questions from vol. 19, issue 5 which were based on the article—‘Management of open pelvic fractures’ by E. Katsoulis, E. Drakoulakis and P.V. Giannoudis. 1

A&

B&

C’

D&

E&

2

A&

B&

C&

D’

E&

3

A&

B&

C&

D’

4

A&

B’

C&

D&

E&

5

A’

B&

C&

D&

E&

6

A&

B&

C&

D&

E’

7

A&

B&

C&

D’

E&

8

A&

B&

C&

D’

E&

9

A’

B&

C&

D&

E&

10 A &

B’

C&

D&

E&

11 A &

B’

C&

D&

E&

12 A &

B&

C&

D’

E&

0268-0890/$ - see front matter doi:10.1016/j.cuor.2005.10.002

Aims and Scope Current Orthopaedics presents a unique collection of international review articles summarizing the current state of knowledge and research in orthopaedics. Each issue focuses on a specific topic, discussed in depth in a mini-symposium; other articles cover the areas of basic science, medicine, children/adults, trauma, imaging and historical review. There is also an annotation, self-assessment questions and an exam section. In this way, the entire postgraduate syllabus will be covered in a 4-year cycle. The Journal is cited in: Cochrane Center, EMBASE/ Excerpta Medica, Infomed, Reference Update and UMI Microfilms.

Editor Professor R. A. Dickson MA, ChM, FRCS, DSc St James’s University Hospital Trust, Leeds, UK

Editorial Committee President of BOTA, M. A. Farquharson-Roberts (Gosport, UK), I. Leslie (Bristol, UK), D. Limb (Leeds, UK), M. Macnicol (Edinburgh, UK), I. McDermott (Ruislip, UK), J. Rankine (Leeds, UK)

Editorial Advisory Board L. de Almeida (Portugal) G. P. Songcharoen (Thailand) R. W. Bucholz (USA) J. W. Frymoyer (USA) R. W. Gaines (USA) S. L. Weinstein (USA) M. Bumbasirevic (former Yugoslavia)

A. K. Mukherjee (India) A. Kusakabe (Japan) A. Uchida (Japan) M.-S. Moon (Korea) R. Castelein (The Netherlands) R. K. Marti (The Netherlands) G. Hooper (New Zealand) A. Thurston (New Zealand) E. G. Pasion (Philippines)

D. C. Davidson (Australia) J. Harris (Australia) S. Nade (Australia) G. R. Velloso (Brazil) J. H. Wedge (Canada) S. Santavirta (Finland) P. N. Soucacos (Greece) M. Torrens (Greece) J. C. Y. Leong (Hong Kong)

Available online at www.sciencedirect.com

Amsterdam

K

Boston

K

Jena

K

London

K

New York

K

Oxford

K

Paris

K

Philadelphia

K

San Diego

K

St Louis

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 77–84

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI SYMPOSIUM: SOFT TISSUE KNEE PROBLEMS

(i) The crucial ligaments Steve Bollen Department of Orthopaedic Surgery, Bradford Royal Infirmary, Duckworth Lane, Bradford BD9 6RJ, UK

KEYWORDS Anterior cruciate ligament; Posterior cruciate ligament; Surgical treatment and outcomes

Summary Anterior cruciate ligament (ACL) damage is a high-risk sports injury which, if not treated surgically, leads to significant medium- to long-term disability and degenerative change. ACL reconstruction should be carried out using autogenous tendon grafting a month to 6 weeks after injury. With appropriate rehabilitation, return to sports can occur at around 6 months. Posterior cruciate ligament (PCL) injury is much less common and occurs by attenuation rather than midsubstance rupture and thus conservative treatment is an option. Surgery is not an option for isolated PCL injury, but is as part of chronic complex injury reconstruction. & 2006 Elsevier Ltd. All rights reserved.

The anterior cruciate ligament Anterior cruciate injury is common and has been shown to have an incidence of about 30 per 100,000 population per year.1 It can have a devastating effect on an individual’s sporting aspirations or daily life, and for a professional sportsman or woman is usually career threatening. In addition, we now live in an age when patients have high expectations and many wish to return to normal, recreational sporting activities following this injury, irrespective of age or occupation.

History and diagnosis The vast majority of these injuries occur during sport. Highrisk sports include soccer, rugby and skiing, which account for over 90% of the injuries in the author’s practice. Typically, there is a non-contact twisting injury or a valgus/external rotation strain. The individual often feels or hears a pop or snap, and is unable to carry on playing. Swelling usually occurs within 4 h. A study in the UK in 1995 showed that this was followed by a visit to casualty, a 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.02.012

normal X-ray and a discharge with an elasticated support bandage in 87% of cases!2 This study was repeated recently and things have not improved! The swelling generally takes 2–3 weeks to settle and thereafter the patients have few problems during straight line activities but find that a sudden change of direction causes their knee to give way. When they present to an orthopaedic clinic the signs of a positive Lachman and a pivot shift are always present, but can occasionally be difficult to elicit (especially in the acute setting). Clinical diagnosis in the hands of an experienced knee clinician is 99% sensitive and specific.3 A plain X-ray is important to identify anterior cruciate ligament (ACL) avulsion (which completely alters management); MRI is rarely indicated and rarely alters decision making.4 It is important to identify associated injuries, particularly of the posterolateral or posteromedial corners, as failure to recognise these and address them at the time of surgery, compromises the results and usually leads to early failure. Never mistake a posterior cruciate ligament injury for an anterior cruciate injury, as can occur if reliance is placed on an anterior draw test, failing to recognise that there is a posterior sag to start with (Fig. 1).

ARTICLE IN PRESS 78

S. Bollen injury.5,6 Noyes et al.7 found that 31% had disability for walking alone and there is evidence that by 8 years postinjury 70% of patients have radiographic osteoarthritis. What is unknown is whether ACL reconstruction can slow down or prevent the degenerative process. Some recent evidence8 does seem to show that there is no acceleration of osteoarthritis following reconstruction despite continuing participation in sport, provided that the menisci and articular cartilage are intact at the time of surgery (Fig. 2).

Biomechanics

Figure 1 This rugby playing lawyer presented 4 years after ACL reconstruction with constant pain and swelling in the knee. Posterior sag was obvious. When the old ACL reconstruction was taken out the original ACL was still present!

The knee is a complex joint with six degrees of freedom. For its stability it relies on the arrangement and interaction of ligaments, muscles and capsule. The ACL is the primary restraint to anterior translation of the tibia on femur. It also has a secondary role as a restraint to varus and valgus torques and internal and external rotation9,10. Its primary function however is in controlling rotation of the knee. ACL deficient patients do no complain of their tibias sliding forward but of giving way of the knee when trying to change direction. The ligament is made up of fibres arranged in such a way that some part of the ligament is taught throughout the full range of movement. It can be roughly divided into an anteromedial bundle, taught in flexion, and a posterolateral bundle, taught in extension.11 The ACL inserts on the tibia approximately in the centre, between the tibial spines, an ideal location to ensure normal physiological rotation during flexion and extension of the joint. When the ACL is ruptured, this normal control is absent and when a valgus torque is applied to the knee the centre of the axis of rotation shifts to the medial structures and an abnormal rotation occurs (Fig. 3)12. This is the basis of the pivot shift test. It is interesting that this test was socalled because the problem occurred when the patient was pivoting,13 but unknowingly the authors correctly described what happens biomechanically—the pivot shifts! Whilst the pivot shift occurs in all ACL deficient knees, the degree to which it occurs varies from one individual to another. Matsumoto found that in some knees little happens until a sudden shift occurs and in others, the abnormal

Figure 2 Weight bearing film of a patient who presented 13 years after his BPTB reconstruction, having ruptured the ACL in the other knee playing football. This shows no obvious signs of osteoarthritis.

Natural history Many believe that patients can manage well without an anterior cruciate. This may be true in the short term but by as little as 5 years post-injury 75% of patients have stopped playing sport and the majority have had a significant further

Figure 3 A schematic diagram illustrating what happens during the pivot shift. It is not an exaggerated physiological rotation, as occurs with a collateral ligament injury, but an abnormal rotation that occurs when the ACL is non-functional and a valgus torque shifts the axis of rotation to the medial structures.

ARTICLE IN PRESS The crucial ligaments

79

rotation occurs smoothly and gradually. This may be why some individuals are better able to cope with insufficiency and is probably related to the anatomy of the lateral compartment and the relative convexities of the lateral tibial plateau and lateral femoral condyle.

Treatment options Indications for reconstruction remain a matter of debate. Reconstruction (when performed satisfactorily) now has a low morbidity and a high degree of success—consequently the indications for surgical intervention have shifted. An active lifestyle and a refusal to change sporting activities to accommodate knee injury (particularly in a professional who’s livelihood depends on knee function), or recurrent giving-way episodes, are indications for reconstruction. Associated injuries, particularly reparable meniscal tears (which do not heal well in the presence of ACL deficiency), may prompt early reconstruction. Patients with low demand lifestyles and who have rare giving-way episodes may cope quite well with rehabilitation aimed at improving muscle strength and proprioception.14 ACL injury in skeletally immature adolescents presents particular problems. If an ACL avulsion has occurred the energy of the injury is dissipated through the fracture line and the ligament usually remains intact. Early anatomical fixation produces good results. Adolescents who suffer a mid-substance rupture, however, are an entirely different group. It has been shown that they are at high risk15 and if left untreated suffer significant morbidity in terms of reinjury, articular cartilage damage and meniscal tears. This has to be balanced against the potential for growth plate injury when drilling tunnels for an intra-articular ACL reconstruction. This risk, however, seems to be small compared to the risk of significant further injury if treated conservatively. Hamstring reconstruction fixed above and below the growth plate can allow return to competitive sport and seems to grow with the patient (Figs. 4a and b), though quite how this occurs is unknown. Partial ACL tears do occur and generally have a good prognosis if there is no clinical laxity at the time of presentation, irrespective of the amount of tearing seen during arthroscopy.16 If ACL injury occurs associated with a valgus strain, the pattern is often that of an avulsion from the femoral attachment. There is an anatomical synovial fold that runs from the ACL to the lateral wall of the notch and if this is not disrupted it holds the torn ACL in a roughly anatomical position, allowing healing to a slightly anteroinferior attachment site (Figs. 5a and b). This results in a set of consistent clinical signs: a larger Lachman than the normal limb but with a hard end stop, a large anterior draw with no end stop and a pivot glide or 1+ pivot.17

Surgical technique and outcome ACL reconstruction is not a difficult operation to do, but it is a very easy operation to get wrong. Biomechanical studies18 have shown that a misplacement of only 2 mm anteriorly from the ideal femoral tunnel position results in a restriction in range of movement. The most common reason (80%) for failure of an ACL reconstruction is technical error19 (Fig. 6).

Figure 4 (a) and (b) Radiograph and MRI 3 years after Hamstring ACL reconstruction in a 12 year old boy. He has subsequently grown 22 cm, participates in rugby and skiing, and has no leg length discrepency. MRI seems to show the graft has grown with him but how this occurs is unknown.

ARTICLE IN PRESS 80

S. Bollen

Figure 6 Grossly anterior femoral tunnel and anterior placement of the tibial tunnel in a referred patient who’s knee ‘‘has never been right’’ from day one after reconstruction.

Figure 5 (a) The normal synovial fold can be seen running from a normal ACL to the lateral wall of the notch. (b) An antero-inferior reattachment of the ACL showing laxity at 901 of flexion.

Unlike many orthopaedic procedures, a badly done ACL reconstruction does not work from day one and this cannot be corrected by prolonged physiotherapy! The vogue for reconstruction with synthetic materials has largely passed. Some of the initial reports were quite promising but this has not been sustained with time. Some quite serious biological reactions have occurred (Fig. 7) and artificial ligaments are often very difficult to revise when they fail. The principal grafts that are now used are the middle third patellar tendon (bone–patellar tendon–bone–BPTB) and four strand hamstring (doubled gracilis and semitendinosus tendons). They have stood the test of time and have a high success rate. The picture that seems to be emerging from the literature is that they can both provide satisfactory results and there is little to choose between them, other than that there is a higher incidence of kneeling pain following BPTB grafts.20–22 Successful reconstruction has even been shown to improve symptomatology in knees with significant degenerative change.23,24

Figure 7 Gross osteolysis in a femoral condyle following goretex ligament reconstruction.

ARTICLE IN PRESS The crucial ligaments The timing of intervention is important. The long-term outcome really depends on the state of the joint at the time of surgery and studies have shown that more than 6 months post-injury the likelihood of meniscal injury or articular cartilage damage increases with time.25–27 Immediate intervention can also be problematic, with a very high incidence of arthrofibrosis occurring if surgery is performed before the initial swelling has settled and a full range of motion has been regained.28,29 Generally, this will take somewhere between 4 and 6 weeks. There is an enormous literature on different types of graft fixation, their loads to failure, their ability to biodegrade, etc., but there is no hard evidence that any particular fixation method produces superior clinical results. Indeed two of the weakest forms of fixation—interference screws for hamstring grafts and sutures tied over a button for BPTB grafts—have been shown to produce clinical results at 2 and 5 years that are as good as any other study that has been published.30,31 The essentials for technical success are to place the graft so that it is isometric during flexion and extension and does not impinge in full extension (in most individuals this means a degree of hyperextension). In a right knee the femoral tunnel is placed at 10.30/11 O’clock with the whole tunnel in the posterior third of Blumenstedt’s line (seen on a lateral X-ray). The tibial tunnel is centred 2/3–3/4 along a line from the anterior horn of the lateral meniscus to the medial tibial spine. More detail can be obtained from the British Orthopaedic Association web site—good practice guidelines.

Rehabilitation Changes in rehabilitation have revolutionised the outcome and morbidity from ACL reconstruction. A patient 15–20 years ago may have been in plaster for 6 weeks followed by a cast brace for 6 weeks and it took nearly a year of rehabilitation to get back to anything like normal. You had to be a pretty good athlete to survive the surgery! It was when Don Shelbourne critically reviewed his patients and realised that the patients who ignored the protocols and just ‘did their own thing’ did better than the patients who followed the instructions to protect weight bearing, restrict range of motion, etc., that the concept of accelerated rehabilitation was born. Follow up has shown no deleterious effects of regaining full extension in the first couple of weeks and then progressing to activities such as jogging by 6–12 weeks post op and returning to sport by around 6 months.30,32

81 need to be seen, but this is currently the Holy Grail of knee reconstruction surgeons!

The posterior cruciate ligament Isolated posterior cruciate ligament (PCL) injury is much less common than ACL injury. In Myasakas’ study1 it accounted for less than 5% of ligament injuries. The true incidence is probably unknown for reasons that will be outlined below, and our management of this sometimes difficult problem lags significantly behind that of the ACL.

History and diagnosis Isolated PCL injury occurs usually with a fall onto a flexed knee or, more rarely, with a direct blow to the anterior aspect of the proximal tibia. It is most often a low violence injury—higher violence injuries such as road traffic accidents frequently produce combined injuries of the PCL and other structures. It is common in the sport of rugby league where tackling is done by two players, one round the legs and one round the shoulders, resulting in a fall onto flexed knees with an additional 120 kg across the shoulders. In the rugby league squads that the author looks after, approximately 30% of the players have a ruptured PCL in one knee, some in both! Typically there is pain in the popliteal fossa on injury, but the individual may be able to carry on playing. The injury is often dismissed as ‘‘minor’’ by the player. There is generally some swelling, but this may be a small amount. By a couple of weeks things are beginning to settle, but as the player tries to resume running, they are comfortable at a moderate pace but get pain or discomfort in the popliteal fossa when trying to sprint. This settles over a few weeks. If seen acutely there is a posterior sag (Fig. 8), increasing pain in the popliteal fossa on flexing the knee and a change in tibial step off (felt anteromedially at 901 of flexion, the anterior edge of the medial tibial condyle is usually anterior to the medial femoral condyle—compare to the normal knee).

The future The last 10 years have seen no dramatic changes in the treatment of ACL injury but rather a steady refinement of technique and better recognition and management of associated injuries. The race is on to develop synthetic collagen ligaments that can be seeded with the patients’ own fibroblasts and implanted 4–8 weeks post-injury, further reducing the morbidity, accelerating the maturation of the graft and allowing earlier return to normal activity. We will still need to place them correctly and the cost implications

Figure 8 Large posterior sag in a knee with a PCL+posterolateral corner injury.

ARTICLE IN PRESS 82 It is important to exclude associated injuries, as these have a major bearing on management. Pick the patients legs up by the toes—varus-recurvatum indicates damage to the posterolateral structures. A posterior draw of over 1 cm should also alert the examiner to the possibility of combined ligament injury. With the patient prone, an increased dial test at 301 and 901 of flexion may indicate associated posterolateral corner damage (it is important not to confuse anteromedial rotatory subluxation with posterolateral rotatory subluxation—you need to repeat the dial at 901 with the patient supine and carefully look at the tibial condyles to see if the medial tibial condyle is rotating forward or the lateral tibial condyle is going backwards). X-rays may show a PCL avulsion (although this is rare) but usually confirm posterior displacement of the tibia relative to the femur. MRI may be useful in acute, combined injuries but is generally unhelpful in isolated PCL injuries. In chronic injuries it is even less helpful, only picking up the diagnosis in 57% of cases.33

Natural history Patients who have sustained an isolated PCL injury usually return to sport and normal activities without too many problems, especially if quadriceps strength is built up.34–36 The fact that 30% of professional rugby league players, playing at an elite level, have a ruptured PCL, indicates that in the short term there is little functional disability. Following recovery from the initial injury, patients undergo a period of functional adaptation and then a long period of functional tolerance which may be 15–25 years before symptoms start to deteriorate and arthritic change becomes symptomatic.37

Biomechanics

S. Bollen returned to sport somewhere between 8 and 12 weeks post-injury. Surgery is rarely indicated in the acute situation except for PCL avulsions and combined injuries. There is some debate about reconstruction in grade 3 injuries, but the exact role for surgical intervention in isolated injuries remains to be defined.

Surgical treatment and outcome Techniques have undergone a gradual evolution over the last 10–15 years and so the long-term results from reconstruction are unknown. Certainly PCL reconstruction as part of a complex reconstruction in chronic complex injuries can provide satisfactory results in this group of patients, who usually have fairly low expectations. In 14 years of practice as a knee reconstruction surgeon the author has yet to perform a reconstruction for an isolated PCL injury. Unlike the ACL, where closed kinetic chain exercises put very little strain on the graft, it is difficult to minimise the tension on the PCL graft as soon as the knee is flexed. It is difficult to find a study that claims to have restored posterior laxity to normal after PCL reconstruction. The current discussion revolves around whether it is better to perform a one- or two-bundle reconstruction. Two bundle reconstruction offers the potential advantage of providing a more anatomical reconstruction but clinical studies have so far failed to show any significant improvement over a single bundle (effectively reconstructing the anterolateral bundle) technique. Technically, PCL reconstruction is more difficult to perform than ACL reconstruction, with the acute angle as the graft exits the back of the tibia and turns forward to the femoral tunnel, potentially causing difficulties in graft

The PCL may be twice the size of the ACL but it is not twice as strong.38 Like the ACL it has 2 major functional bundles but arranged the other way round—a posteromedial bundle taught in extension and an anterolateral taught in flexion. The PCL provides the primary restraint to posterior translation of the tibia on the femur but, unlike the ACL, does not have a significant role in controlling rotation. Experimental studies have shown that tension increases in the ligament in flexion and is reduced in extension.39,40 Very importantly, even in PCL deficiency the knee reduces anatomically in extension as long as the collateral ligaments are intact.41

Treatment options Unlike the ACL, which normally ruptures mid-substance, the PCL tends to attenuate, explaining the difficulty in identifying the injury on MRI in chronic cases. This also allows the possibility of conservative treatment in acute cases. We know the knee reduces in extension if the collateral ligaments remain intact (see above) and therefore bracing in extension for 4 weeks allows the possibility of the injured PCL to scar and shorten, reducing the residual laxity. Following this, the patient can be rehabilitated, concentrating on quadriceps strength, and generally can be

Figure 9 Image intensifier picture during drilling of the tibial tunnel in a complex reconstruction. The popliteal artery lies about 1 cm from the drill tip—it is important not to advance it too far!

ARTICLE IN PRESS The crucial ligaments passage. The exit position for the tibial tunnel also puts the posterior vessels at risk and it is advisable to carry out this part of the operation under image intensifier control (Fig. 9). BPTB grafts are difficult to fix and difficult to pass, although can be used with a tibial onlay technique—this does however mean turning the patient during the operation to use a posterior approach to allow access for fixation of the graft to the PCL fossa. Hamstring grafts are much easier to pass and allow the possibility for a double bundle technique but there are concerns about whether they are strong enough in many patients. Fresh frozen allografts (especially tendo-Achilleis) are useful in PCL reconstruction, being easy to pass from femur to tibia and are of sufficient size and strength to reproduce normal PCL characteristics.

83

15.

16. 17.

18.

19. 20.

The future 21.

The biomechanics of the PCL have been worked out, but much remains to be improved and consolidated with regard to reconstruction techniques and rehabilitation. It will probably be another 10 years before a lot of today’s practice will have been evaluated and validated. As with the ACL the possibilities of biological grafts and manipulation of the healing process offer the potential of reducing morbidity and improving outcome.

22.

23.

24.

References 25. 1. Myasaka KC, Daniel DM, Stone M. The incidence of knee ligament injuries in the general population. Am J Knee Surg 1991;4:43–8. 2. Bollen SR, Scott BW. Rupture of the anterior cruciate ligament—a quiet epidemic? Injury 1996;27:407–9. 3. O’Shea KJ, Murphy KP, Heekin D, Herzwurm H. The diagnostic accuracy of history, physical examination, and radiographs in the evaluation of traumatic knee disorders. Am J Sports Med 1996;24:164–7. 4. Gelb HJ, Glasgow SG, Sapega AA, Torg JS. Magnetic resonance imaging of knee disorders clinical value and cost effectiveness in a sports medicine practice. Am J Sports Med 1996;24:99–103. 5. Kannus P, Jarvinnen M. Conservatively treated tears of the ACL. JBJS 1987:1007–12. 6. Satku K, Kumar VP, Ngoi SS. ACL injuries—to counsel or operate? JBJS 1986;68B:458–61. 7. Noyes FR, Mathews DS, Mooar PA, Grood ES. The symptomatic ACL deficient knee. Parts 1&2 JBJS 1983;65A:154–74. 8. Hart AJ, Buscombe J, Malone A, Dowd GSE. Assessment of osteoarthritis after reconstruction of the anterior cruciate ligament. JBJS 2005;87B(11):1483–7. 9. Marshall JL, Wang JB, Furman W, Girgis FG, Warren R. The anterior drawer sign—what is it? Am J Sports Med 1975;3: 152–8. 10. Nielsen S, Ovesen J, Ramussen O. The ACL: an experimental study of its importance in rotatory knee instability. Arch Orthop Trauma Surg 1984;103:170–4. 11. Furman W, Marshall JL, Girgis FG. The ACL: a functional analysis based on post mortem studies. JBJS 1976;58A:179–85. 12. Matsumoto H. Mechanism of the pivot shift. JBJS 1990; 72B:816–21. 13. Galway HR, MacKintosh DL. The lateral pivot shift: a symptom and sign of ACL insufficiency. Clin Orthop 1980;147:45–50. 14. Buss DD, Min R, Skyhar M, Galinat B, Warren RF, Wickiewicz TL. Nonoperative treatment of acute anterior cruciate ligament

26.

27.

28.

29.

30.

31.

32. 33.

34.

35.

injuries in a selected group of patients. Am J Sports Med 1995; 23:160–5. Aichroth PM, Patel DV, Zorilla P. The natural history and treatment of rupture of the anterior cruciate ligament in children and adolescents: a prospective review. J Bone and Joint Surg 2002;84B:38–41. Noyes FR, Mooar LA, Moorman CL, McGiniss GH. Partial tears of the ACL. JBJS 1989;71B:825–33. Bollen SR. Antero-inferior reattachment of the anterior cruciate ligament—an unrecognosed pattern of injury. JBJS 2004; 86B(Suppl. IV):47. Grood ES, Hefzy MS, Lindfield TN. Factors affecting the most isometric femoral attachments: PCL and ACL. Am J Sports Med 1989;17:197–216. Ng A, Bollen SR. Failure of ACL reconstruction—the U.K. experience. Proc J Bone Joint Surg 1999(Suppl. III):278. Anderson AF, Snyder RB, Lipscomb AB. Anterior Cruciate Ligament Reconstruction—a prospective randomised study of three surgical methods. Am J Sports Med 2001;29:272–9. Shaieb MD, Kan DM, Chang SK, et al. A prospective randomised comparison of patellar tendon versus semitendinosus and gracilis tendon autografts fo anterior cruciate ligament reconstruction. Am J Sports Med 2002;30:214–20. Aglietti P, Giron F, Buzzi R, et al. ACL reconstruction BPTB compared to doubled ST&G tendon grafts. JBJS 2004;86A: 2143–55. Shelbourne KD, Wilckens JH. Intra-articular anterior cruciate ligament reconstruction in the symptomatic arthritic knee. Am J Sports Med 1993;21:685–9. Noyes FR, Barber-Westin SD. ACL reconstruction with autogenous patellar tendon graft in patients with articular cartilage damage. Am J Sports Med 1997;25:626–34. Shelbourne KD, Gray T. Results of ACL reconstruction based on meniscus and articular cartilage status at the time of surgery 515 year evaluation. Am J Sports Med 2000;28:446–52. Hogervorst T, Rijcken TH, Rucker D, et al. Changes in bone Scans after ACL Ligament reconstruction. Am J Sports Med 2002;30: 823–33. Karlsson J, Kartus J, Magnusson L, Brandsson S, Eriksson BI. Subacute versus delayed reconstruction of the anterior cruciate ligament in the competitive athlete. Knee Surg Traumatol Arthrosc 1999;7:146–51. Shelbourne KD, Wilckens JH, Mollabashy A, Decarlo M. Arthrofibrosis in acute ACL reconstruction The effect of timing on reconstruction and rehabilitation. Am J Sports Med 1991;19: 139–47. Waselewski SA, Covall DJ, Cohen S. Effect of surgical timing on recovery and associated injuries after anterior cruciate ligament reconstruction. Am J Sports Med 1993;21:338–47. Shelbourne KD, Gray T. Anterior cruciate ligament reconstruction with autogenous patellar tendon graft followed by accelerated rehabilitation—2–9 year follow up. Am J Sports Med 1997;25:786–95. Corry IS, Webb JM, Clingeleffer AJ, Pinczewski LA. Arthroscopic reconstruction of the anterior cruciate ligament A comparison of patellar tendon autograft and four strand hamstring tendon. Am J Sports Med 1999;27:444–54. Bollen SR. Rehabilitation after ACL Reconstruction. Knee 2001;8:75–7. Servant CTJ, Ramos JP, Thomas NP. The accuracy of magnetic resonance imaging in diagnosing chronic posterior cruciate ligament injury. Knee 2004;11:265–70. Dandy D, Pusey R. The long term results of unrepaired tears of the posterior cruciate ligament. J Bone Joint Surg 1982;64: 92–4. Cross MJ, Powell JF. Long term follow up of posterior cruciate ligament rupture : a study of 116 cases. Am J Sports Med 1984;12:292–7.

ARTICLE IN PRESS 84 36. Shelbourne KD, Davis TJ, Patel DV. The natural history of acute, isolated, no-operatively treated posterior cruciate ligament injuries: a prospective study. Am J Sports Med 1999;27: 276–83. 37. Dejour H, Walch G, Peurot J, Eberhard P. The natural history of rupture of the posterior cruciate ligament. Orthop Trans 1987;11:146–51. 38. Harner CD, Xerogeanes JW, Livesay GA. The human posterior cruciate ligament complex: an interdisciplinary study. Ligament morphology and biomechanical evaluation. Am J Sports Med 1995;23:736–45.

S. Bollen 39. Ogata K, McCarthy JA. Measurements of length and tension patterns during reconstruction of the posterior cruciate ligament. Am J Sports Med 1992;20:351–5. 40. Harner CD, Janauschek MA, Ma B, Kanamori A, Vogrin TM, Woo S. The effect of knee flexion angle and application of anterior tibial load at the time of graft fixation on the biomechanics of a posterior cruciate ligament reconstructed knee. Am J Sports Med 2000;28:460–5. 41. Ogata K, McCarthy JA, Dunlap J, et al. Pathomechanics of posterior sag of the tibia in posterior cruciate deficient knees An experimental study. Am J Sports Med 1988;16:630–6.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 85–94

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: SOFT TISSUE KNEE PROBLEMS

(ii) Meniscal tears Ian Douglas McDermotta,b, a

Ealing Hospital, London, UK Brunel University School of Sport & Education, Middlesex, UK

b

KEYWORDS Meniscus; Meniscal; Function; Tear; Meniscectomy; Repair; Allograft

Summary Meniscal tears are the most common injuries to the knee. The menisci have important functions within the knee, and loss of a meniscus from surgical meniscectomy significantly increases the risk of subsequent development of degenerative changes within the knee. In younger patients, some meniscal tears may be repairable, and there is now a wide variety of different techniques available for meniscal repair, with highly encouraging published clinical results. In patients with a painful knee where there has already been a previous meniscectomy, meniscal replacement by meniscal allograft transplantation may be a viable option. However, in the future, solutions may well lie within the field of tissue engineering. & 2006 Published by Elsevier Ltd.

Introduction

Incidence and aetiology of meniscal tears

The menisci of the knee are two crescentic wedges of fibrocartilage, positioned between the tibia and the femur in the medial and lateral compartments (Fig. 1). They used to be considered to be nothing more than the vestigial remnants of a muscle within the knee. However, it is now well recognised that they have various important functions within the knee. They act as load sharers,1 and shock absorbers2 within the joint, and are also secondary stabilisers,3 particularly in the ACL-deficient knee.4 Further roles of the menisci have been postulated with respect to joint lubrication,5 nutrition of the articular cartilage6 and proprioception.7

Meniscal tears are the most common injury of the knee, with a reported annual incidence of meniscal injury resulting in meniscectomy of 61 per 100,000 population.8 Medial meniscal tears are generally seen more frequently than tears of the lateral meniscus, to a ratio of approximately 2:1.9 Meniscal tears may occur in acute knee injuries in younger patients or as part of a degenerative process in older individuals. The acute tears frequently result from sports injuries where there is a twisting motion on the partially flexed, weight-bearing knee. Acute meniscal tears may also occur as part of a more major, combined injury to the knee. In particular, the triad of a rupture of the medial collateral ligament, rupture of the anterior cruciate ligament and a tear of the medial meniscus as a result of forced valgus angulation plus external rotation on the weightbearing knee from, for example, a rugby tackle or a highspeed skiing injury, is a well-recognised entity. In a study looking at 1236 patients aged between 18 and 60 years with arthroscopically proven meniscal injuries, 32%

Corresponding author at. 30 Park Way, Ruislip, Middlesex, HA4 8NU, UK. Tel.: +44 1895 678109; fax: +44 1923 678333; mobile: 07966 439688. E-mail address: [email protected].

0268-0890/$ - see front matter & 2006 Published by Elsevier Ltd. doi:10.1016/j.cuor.2006.02.010

ARTICLE IN PRESS 86

I.D. McDermott

Figure 1 Cadaveric specimen of the tibial plateaus, showing the menisci of the knee.

arose as sports injuries, 38% were from non-sporting injuries and the remaining 28% reported no specific history of injury.10 In the group with non-sporting injuries, approximately 50% of the patients reported that their injury occurred on rising from a squatting position. The mean age of the patients in the sports injury group was 33 years, in the non-sporting injury group was 41 years, and in the no injury group was 43 years. A report by Smillie in 1968 stated that out of 3000 cases of meniscectomy, 36% were due to acute tears (described as longitudinal tears), and 50% were due to degeneration.11 The degenerative tears were described as ‘horizontal’ tears, and occurred in an older age group (mean age 43, compared to a mean age of 31 in those with longitudinal tears). The degenerative tears were described as lesions of middle age, occurring in abnormal fibrocartilage. It was observed that this pattern of tear was most frequently seen in the posterior horn of the medial meniscus. Meniscal tears are frequently observed in conjunction with anterior cruciate ligament injuries at the time of anterior cruciate ligament reconstruction. In an arthroscopic study, Nikolic found that as many as 72% of knees with a recent ACL tear had a concomitant tear of the lateral meniscus, with a preponderance of circumferential tears of the posterior horn.12 In the chronically anterior cruciate ligament deficient knee, the incidence of meniscal tears has been reported to be as high as 98%.13 The medial meniscus, in particular, has been shown to significantly contribute to anterior stability of the tibia in the anterior cruciate ligament deficient knee.4 With chronic anterior cruciate ligament deficiency, repeated anterior translation of the tibia causes an increased load on the posterior horn of the medial meniscus.14 Furthermore, as noted by Vedi,15 the posterior horn of the medial meniscus is the least mobile portion out of both of the menisci. These factors seem to explain the frequency with which posterior medial meniscal tears, in particular, are found within the chronically anterior cruciate deficient knee.14 The relationship between the delay in reconstructive surgery for an ACL rupture, and the subsequent develop-

Figure 2 Meniscal tears: (A) radial tear in body; (B) vertical tear in posterior horn; (C) horizontal cleavage tear; (D) complex degenerative tear.

ment of meniscal tears as a direct result was demonstrated very elegantly in a study by de Roeck and Lang-Stevenson, where 10% of patients with ACL ruptures were found to have developed new meniscal tears between the time of their

ARTICLE IN PRESS Meniscal tears first arthroscopy (where the ACL rupture was confirmed) and the time of their subsequent ACL reconstruction.16 Acute meniscal tears may be radial, vertical-circumferential or horizontal-cleavage in orientation (Figs. 2A–C). The central portion of a circumferential tear may be unstable and can displace inwards within the relevant compartment of the knee. Such a tear is referred to as a bucket handle tear, and frequently causes mechanical locking of the knee. Degenerative tears tend to be complex in morphology (Fig. 2D).

Symptoms, signs and investigations The classic symptoms of a meniscal tear around the time of injury are pain around the affected side of the joint, possible locking of the joint, and swelling. The swelling tends to develop gradually over the first 24-h, as opposed to the knee with an acute ACL tear, where the swelling tends to develop rapidly, within the first hour. Ongoing symptoms of a meniscal tear include pain around the joint line, clicking (which may or may not be painful), giving way and locking. It is also common for patients to complain that they are either unable or else find it painful to squat fully (a position that heavily stresses the posterior horns of both menisci). The eponymous test that is commonly performed as part of the routine knee examination in order to test for the presence of a meniscal tear is McMurray’s test, which was described by McMurray in 1942 in the British Journal of Surgery.17 There are many differing written descriptions of how McMurray’s test is actually performed, and even more variations are witnessed in clinical practice. However, McMurray’s original description is as follows: ‘‘In carrying out the manipulation with the patient lying flat, the knee is first fully flexed until the heel approaches the buttock; the foot is then held by grasping the heel and using the forearm as a lever. The knee being now steadied by the surgeon’s other hand, the leg is rotated on the thigh with the knee still in full flexion. During this movement the posterior section of the cartilage is rotated with the head of the tibia, and if the whole cartilage, or any fragment of the posterior section, is loose, this movement produces an appreciable snap in the joint. By external rotation of the leg the internal cartilage is tested, and by internal rotation any abnormality of the posterior part of the external cartilage can be appreciated. By altering the position of flexion of the joint the whole of the posterior segment of the cartilages can be examined from the middle to their posterior attachments. Thus, if the leg is rotated with the knee at right angles the cartilages in their mid-section come under some pressure, but, anterior to this point, the pressure exerted on the cartilage is so diminished that accurate examination is impossible. When a loose segment of the cartilage is caught between the bones during rotation, the sliding of the femur over the loose fragment is accompanied by a thud or click, which can sometimes be heard but can always be felt, and the size of the detached portion can be judged by the rocking of the tibia, and usually by the severity of the sound produced. This method of examination is not easy to master; the rotation requires a considerable amount of practice, and the whole procedure must be carried out systematically if

87 success is to be attained. Probably the simplest routine is to bring the leg from its position of acute flexion to a right angle, whilst the foot is retained first in full internal rotation, and then in full external rotation. Any abnormality in the cartilage structure in the area under examination will be discovered during the straightening of the joint.’’ The role of MRI in the diagnosis of meniscal injuries can be the cause of some debate. In an ideal world, with no financial restrictions or limits on capacity, one might suggest that all patients with significant symptoms suggestive of soft tissue derangement within the knee should undergo MRI scanning. However, certainly within the NHS in the UK, the reality is such that waiting times for scans can be considerable in some centres. This may well influence decision making in some perhaps borderline cases. MRI is a fairly accurate investigation, with reported sensitivity of 80%, specificity of 71%, positive predictive value of 84% and negative predictive value of 71%.18 However, as can be seen from these values, MRI is far from infallible in the diagnosis of meniscal tears, and there is, therefore, good argument for proceeding directly to arthroscopy in those patients where there is a high index of clinical suspicion in situations where the requesting of an MRI scan is likely to cause undue delay.

Treatment Meniscectomy In the past open total meniscectomy was the appropriate treatment for tears of the menisci. Indeed, although the popular eponymous test for meniscal integrity carries his name, McMurray himself actually stated that ‘‘A far too common error is shown in the incomplete removal of the injured meniscus’’, and actually went as far as to suggest that if the knee were opened up on the clinical suspicion that there was a meniscal tear, but the meniscus was found to be intact, it should be excised anyway.17 However, reviewing the literature, McNeil Love recognised as far back as 1923 that the prognosis after meniscectomy should be guarded, and he observed that ‘‘the occasional pain y was usually associated with changes in the weather, suggesting its association with secondary osteo-arthritis’’.19 After King’s pivotal paper in 1936,20 there was a sea change in the literature, and as awareness of the true functional importance of the menisci grew, the consequences of meniscectomy became more fully appreciated. Loss of the meniscus leads to a decrease in intra-articular contact areas of approximately 75% and an increase in the peak local contact pressures of approximately 235%.21 Fukuda used mini-pressure transducers in cadaver knees and demonstrated significant increases in the stresses in subchondral bone after meniscectomy.22 A study by Roos23 reporting the long-term follow-up at 21 years of patients after meniscectomy compared to matched controls has shown a relative risk of 14 for the development of radiographic signs of osteoarthritis after meniscectomy. It has been shown that the chance of developing osteoarthritis after a lateral meniscectomy is greater than that after a medial meniscectomy24–26. This finding is most

ARTICLE IN PRESS 88

Figure 3 Diagrammatic representation of different geometries of medial and lateral compartments in the sagittal plane without menisci, leading to differences in joint congruity.

probably secondary to the fact that on the lateral side the meniscus carries 70% of the compartment load, compared with the medial meniscus, which only carries 50% of the medial compartment load.1 Furthermore, in the sagittal plane, on the medial side the convexity of the femoral condyle and the concavity of the medial tibial plateau give some degree of congruity, even in the absence of the medial meniscus. On the lateral side, however, the convexity of the femoral condyle is mirrored by convexity of the lateral tibial plateau. Thus, in the absence of the meniscus, on the lateral side there will be a greater tendency towards point loading and increased peak contact pressures (Fig. 3). Studies have shown that after meniscectomy, results can be affected by the quality and frequency of athletic activities.27 Jorgensen26 reviewed 147 athletes after meniscectomy for isolated meniscal injuries and found that radiographic deterioration started after 4.5 years. At 14.5 years follow-up, 89% of athletes had radiographic evidence of degeneration, and 46% had given up or reduced their sporting activity. These values are far higher than would be expected for the general population. However, the full importance of exercise and activity levels after meniscectomy have not yet been thoroughly studied.28 Given the clearly demonstrated and now universally acknowledged adverse effects of total meniscectomy, much interest has focused on the potential benefits of partial rather than total meniscectomy. In a biomechanical study of partial versus total medial meniscectomy, Burke29 showed that there was a linear correlation between increase in peak stress on the tibial joint surfaces and the amount of meniscal tissue removed. Similar findings were confirmed by Ihn.30 The clinical relevance of these observations was confirmed in a review of patients undergoing either partial or total meniscectomy, where Hede31 found that the function of the knee was inversely related to the amount of meniscal tissue excised. However, there was still a significant number of complaints from patients after partial meniscectomy.

I.D. McDermott ‘‘was completely separated from its anterior attachment, and was displaced backwards’’. The torn meniscus was repaired with three chromic catgut sutures, and the patient was ‘‘dismissed cured’’. Animal studies of the response of the menisci to injury have shown that at its periphery, meniscal tissue is capable of producing a reparative response.33–36 Cabaud35 performed transverse medial meniscal lacerations and repair with a single Dexon suture in 20 canine and 12 rhesus knee joints. By four months, only 6% of the menisci had failed to heal. The scar tissue that was present at the site of healing was shown to be composed of unorganised collagen without common ground substance components. Arnoczky34 performed a complete midportion transection of the medial meniscus in 15 canine knees. He showed that the response originated from the peripheral synovial tissues, and that the menisci had completely healed by fibrovascular scar by 10 weeks. Longitudinal incisions in the inner, avascular portion of the meniscus failed to heal. Using rabbit knees, Heatley36 performed meniscal incisions of varying magnitude, with resection of the peripheral meniscal rim, and with or without suturing of the resultant gap. He showed that healing occurred circumferentially, with cells proliferating at the synovial margin and migrating along the cut meniscal edge. The newly formed scar tissue was largely avascular. Heatley noted that synovial cells proliferate readily whilst chondrocytes have only a limited capacity for mitosis. He concluded that healing most probably depended not on the presence of a vascular supply, but on the presence of synovial cells invading the tissue from the periphery. The synovial cells were noted to form tissue that was initially very cellular, but which later became more fibrous. This fibrous tissue was also occasionally seen to transform into fibrocartilage, possibly as a response to compression of the tissue. Suturing the menisci facilitated the healing process by providing stability, and possibly by supplying bridges for synovial cells to migrate onto the meniscus. Despite Heatley’s observations, meniscal tears are still often classified according to the location of the tear relative to the blood supply of the meniscus (Fig. 4). With a ‘red–red’ tear, both the peripheral and inner margins of

Meniscal repair Given the above, over the past years there has been great interest in and effort towards avoidance of meniscectomy wherever possible, and meniscal repair has grown in popularity. The first case of meniscal repair was performed in 1885 by Thomas Annandale, and was reported in the British Medical Journal.32 Annandale described the case of a 30-year old miner who had felt something give within his knee whilst kneeling, followed by an effusion with subsequent pain and locking. An arthrotomy showed that the medial meniscus

Figure 4 The blood supply to the meniscus from the periphery, giving the vascular ‘red zone’ and the avascular ‘white zone’.

ARTICLE IN PRESS Meniscal tears the tear have a functional blood supply, and these peripheral tears reportedly have the best prognosis for healing.37 The ‘red–white’ tear has vascularised tissue on the peripheral side and avascular tissue on the inner side. The ‘white–white’ tear is completely in the avascular zone and is least likely to heal.34,37 Various techniques have been described in an attempt to facilitate healing of tears in the inner, avascular portion of the meniscus, including the creation of vascular access channels,34 trephination,38 rasping of the parameniscal synovium,39,40 and use of exogenous fibrin clot41–43 or free synovial autografts,44–46 or even laser welding.47 Techniques of open,48,49 inside-out,50–52 outside-in53,54 and all-inside arthroscopic repair55,56 have been described, and each has its merits.57 A number of biomechanical studies have investigated the properties of meniscal repairs using various different techniques of suturing,58–61 and all have confirmed that the vertical loop suture is the strongest, exhibiting the greatest load to failure when compared to horizontal or mulberry-knot sutures. Furthermore, numerous meniscal repair devices, such as bioabsorbable arrows, fasteners, and ‘T’-bar ended sutures (Fig. 5), are now available that may offer potential benefits compared to the traditional method of meniscal repair by suturing.62–69 Having tried many of the meniscal suturing devices available, my own personal preference is towards the use of the FasT-Fix device, from Smith & Nephew (Fig. 6). This is an all-inside suture repair system comprising two 5 mm polymer suture bar anchors, with a pre-tied, self-sliding knot of #0 non-absorbable polyester suture. It allows easy and rapid insertion of strong, tight horizontal or vertical loop sutures, which, biomechanically, remain the goldstandard.61 However, at the same time the FasT-Fix avoids some of the potential complications that have been observed with some of the bioabsorbable arrow or dart-like devices available, such as foreign body reactions in the soft tissues due to migrating broken devices, or severe chondral

89 damage from broken or protruding implants within the knee.70,71 The results reported with the use of the FasT-Fix for meniscal repair have been highly encouraging. Kotsovolos reported on the outcome of 58 meniscal repairs at an average follow-up of 18 months, and observed that just over 90% of repairs were clinically successful, with absence of joint-line tenderness, locking, or swelling, and a negative McMurray test.72 The potential long-term benefits of meniscal repair must, however, be weighed on a patient-by-patient basis against the significant differences in post-operative rehabilitation that are required compared to recovery after a simple partial meniscectomy. Patients are often told that there are no specific functional restrictions necessary after simple partial meniscectomy, and are frequently advised to return to full activities and work within a couple of weeks after surgery. However, although advice does vary, my personal post-operative recommendations to patients after meniscal repair is that they wear a hinged knee brace, locked between 01 and 601 for 6 weeks, and remain partial weight bearing during this period. At the end of this period they are then referred for physiotherapy to help them regain their range of motion and strength, and are told to refrain from running until 3 months after the repair. Such restrictions can be extremely imposing on, for example, a self-employed builder, who may well prefer the option of partial meniscectomy. Detailed, informed patient consent is therefore vital prior to any arthroscopic procedure where meniscal repair could potentially be considered, and treatment should be tailored to the needs and wishes of each individual patient.

Meniscal replacement Despite recent advances, a large proportion of meniscal tears remain irreparable, and partial, subtotal or even total meniscectomy may still unavoidably be indicated. In the

Figure 5 Meniscal repair devices. (A) Meniscal arrow (Atlantech, Harrowgate, UK). (B) Meniscal fastener (Mitek, Ethicon Inc., Waltham, MA, USA). (C) T-Fix (Smith & Nephew, Mansfield, MA, USA).

ARTICLE IN PRESS 90

I.D. McDermott

Figure 6 FasT-Fix meniscal repair system (Smith & Nephew).

past, a number of different tissues or materials have been used in an attempt to replace excised meniscal tissue. These include the use of silastic,73 carbon fibre,74 Dacron,75–78 and Teflon75,79 prostheses, patellar, Achilles or semitendinosus tendon autograft,80–82 fat pad autograft,83 and autologous rib perichondrial grafts.84 However, all have met with poor results. As an alternative, the concept of meniscal allograft transplantation was developed, and the first reported animal study describing meniscal allograft transplantation in dogs was by Canham in 1986.85 Since this study, meniscal transplantation has been described in sheep,86 rabbits, 87 mice,88 rats,89 goats,90 and monkeys.91 The first report of human meniscal allograft transplantation was published by Milachowski in 1987,92 and again in 1989.86 It has been estimated that over 4000 meniscal transplantations have now been performed in the USA (personal communication, Kevin Stone, San Francisco, USA), and this technique is also gaining in popularity in various countries across Europe. van Arkel93 reported more recently a larger and longerterm study of 63 consecutive meniscal transplantations with a mean follow-up of 60 months. He found a cumulative survival rate of 76% for lateral meniscal allografts, 50% for medial allografts, and 67% for medial and lateral allografts transplanted into the same knee. Verdonk94 reported his findings after implantation of fresh meniscal allografts into 31 patients. The medial meniscus was transplanted in 46% of cases, the lateral meniscus in 40%, and in the remaining 14%, grafts were inserted bilaterally. Follow-up arthroscopy or magnetic resonance imaging confirmed that all but 5 of the grafts were intact. The clinical outcome was assessed using the Hospital for Special Surgery (HSS) knee rating system,95 and showed that

initially the results were good, with 62% of patients achieving a score of 175 or above. However, after approximately the first 3 years, the clinical results were seen to deteriorate, and after 7 years, only 25% of patients reached an HSS score of 175. The difficulty with interpreting the clinical results of these studies to-date lies in the heterogeneity of the patients reported. The meniscal grafts were frequently performed in conjunction with various other major procedures, such as reconstruction of the anterior cruciate ligament,86,92,94,96,97 valgus or varus high tibial osteotomy or varus distal femoral osteotomy,94,98 or articular osteochondral allografting.96,99 Furthermore, a variety of different graft types were implanted, including fresh,94,96 lyophilised,86,92, fresh-frozen non-irradiated97, fresh-frozen gamma-irradiated98 and cryopreserved allografts.100–102 In addition, the grafts themselves were implanted using various different surgical techniques, ranging from suturing alone,94,96,103 to peripheral suturing plus attachment of the anterior and posterior horns by bone plugs into tunnels in the proximal tibia.97,102,104 What is still needed is a longterm, prospective, controlled study, to compare knees with an absent meniscus that receive a meniscal allograft, with those that do not. Until the results of such a trial are reported, human meniscal allograft transplantation should be considered an experimental, salvage procedure.105

Tissue engineering and the future The process of tissue engineering, as applied to meniscal regeneration, is the production of newly synthesizing meniscal tissue, in part or in whole, by the addition of exogenous cells, matrix scaffold, specific stimuli (growth

ARTICLE IN PRESS Meniscal tears factors, mechanical stress), or any combination thereof, in an in vitro or in vivo environment.106 Given the poor results following prosthetic meniscal replacement and the highly variable results of autografting using alternative tissues such as tendons, much interest and study is currently being directed towards the field of tissue engineering. In terms of cell sources, three basic cell types have been identified as potential sources for the regeneration of meniscal tissue: the meniscal fibrochondrocyte, the mesenchymal stem cell and the pluripotential fibroblast.106 The ideal matrix would allow cell proliferation, free diffusion of nutrients, access to cytokines, be mechanically durable and yet resorbable as the tissues own extracellular matrix develops. These are all characteristics shown by the body’s own regenerative scaffold, the fibrin clot. At present, much work is being undertaken in the development of collagen scaffolds. Collagen scaffolds obtained from enzymatically purified bovine Achilles tendon have been used.107 The collagen is highly crystalline and can be covalently cross-linked by dehydration, dramatically diminishing its susceptibility to collagenase degradation, without altering its natural triple-helix structure. The collagen can be further cross-linked by the use of aldehyde. Thus, the potential resorption rate of the scaffold can be controlled to suit the in vivo environment. The use of such collagen scaffolds has been described in canine studies where they were used to replace an 80% resection of the medial meniscus.107,108 The knees with the implanted scaffolds showed significantly less articular cartilage erosion and osteophyte formation compared to knees with meniscectomy alone, and there was also significant meniscal regeneration into the prosthetic scaffolds. Histologically, well-differentiated meniscal fibrochondrocytes were identified within the scaffolds, and by 12 months these cells were shown to be producing more proteoglycan than was seen in cells around the edges of resected menisci alone. Trials of collagen meniscal scaffolds have progressed to human studies, and Stone109 published the results of ten patients receiving purified, treated, gamma radiation sterilised, moulded bovine Achilles tendon collagen implants after meniscectomy to varying degrees. The patients were followed up by questionnaires, physical examination, X-rays, bone scans, serum analysis and magnetic resonance imaging. At thirty-six months postoperatively, the patients’ activity scores were similar to those that would be expected after simple meniscectomy. There were no significant changes demonstrated on X-ray. MRI showed that the interface between the remaining host meniscal rim and the implant regenerated tissue complex became less distinct as time progressed. Histology of biopsies of the implant regenerated tissue complex demonstrated progressive invasion and replacement by new collagen and cells typical of meniscal fibrochondrocytes. Immunological evaluation revealed no apparent immune responses to the implants. A further study by Rodkey,110 reported the results of the use of collagen scaffolds for reconstruction of irreparably damaged medial menisci. Eight patients received a collagen meniscus implant (Fig. 7), and they were followed up for a minimum of 24 months. All patients improved clinically, and radiographically there was no evidence of any progression of degenerative joint disease. Relook arthroscopy showed apparent meniscal tissue regeneration, and preservation of

91

Figure 7 Collagen meniscal implant (Sulzer Orthopaedics Inc., Austin, TX).

the articular surfaces of the joint. Histological analysis from biopsies at the site of implant insertion confirmed new fibrocartilage matrix formation. These studies confirm that human meniscal cartilage-like tissue does grow into resorbable collagen scaffolds, and initial experience suggests that their use is safe. Further studies examining the efficacy of such implants, compared to meniscectomy, repair, the use of allografts, or no treatment are currently underway.109 Other synthetic engineered bioabsorbable scaffolds have also been described for meniscal regeneration, including polyglycolic acid discs seeded with fibrochondrocytes,111 and 50/50 copolymers of lactide/caprolactone.112 Animal studies have shown that bovine fibrochondrocytes seeded onto polyglycolic acid scaffolds implanted subcutaneously into mice give rise to new meniscus-like tissue.111 Scaffolds of co-polymers of lactide and caprolactone have been used to replace sections of excised meniscus in canine knees, and it has been shown that there is ingrowth of fibrous tissue, formation of fibrocartilagenous tissue and gradual degradation of the polymer.112 Furthermore, only scaffolds manufactured with a high compressive modulus, close to that of normal meniscal tissue, showed formation of fibrocartilagenous tissue, whereas in implants with a lower compressive modulus, only fibrous tissue growth was seen. It was also observed that good adhesion of the periphery of the implant was not a prerequisite for fibrocartilagenous tissue formation. It was therefore suggested that the new fibrocartilage tissue was formed by metaplasia of fibrous tissue, and not by ingrowth from adjacent meniscal tissue, and that this only occurred in appropriate mechanical environments, as provided by scaffolds with compressive moduli comparable to normal meniscal tissue. Work using cultured mesenchymal stem cells implanted into collagen sponge scaffolds to fill meniscal defects in rabbits has shown that the repair process is augmented.113 Although degenerative changes within the rabbit knees were not prevented, seeding with the mesenchymal stem cells was shown to lead to the formation of fibrocartilage histologically similar to normal meniscal tissue, whereas

ARTICLE IN PRESS 92 use of collagen sponge alone, without cells, to fill meniscal defects, led to healing by fibrous tissue only. Further work has recently been directed towards the role of gene therapy for meniscal injury. In vivo work has shown that meniscal cells can be readily transduced by retroviral vectors carrying TGFß1 growth factor genes.114 The cells in monolayer culture then show greatly increased matrix synthesis. The discipline of tissue engineering is in its relative infancy, but technological advances are enabling the application of new techniques at a rapidly increasing rate, and instead of merely being in the realms of science fiction, the prospect of creating tailor-made replacement tissues by order now seems ever more likely to be a reality.

References 1. Seedhom BB, Dowson D, Wright V. Proceedings: functions of the menisci. A preliminary study. J Bone Joint Surg Br 1974;56B:381–2. 2. Voloshin AS, Wosk J. Shock absorption of meniscectomized and painful knees: a comparative in vivo study. J Biomed Eng 1983;5(2):157–61. 3. Markolf KL, Mensch JS, Amstutz HC. Stiffness and laxity of the knee—the contributions of the supporting structures. A quantitative in vitro study. J Bone Joint Surg Am 1976;58(5):583–94. 4. Allen CR, Wong EK, Livesay GA, Sakane M, Fu FH, Woo SL. Importance of the medial meniscus in the anterior cruciate ligament-deficient knee. J Orthop Res 2000;18(1):109–15. 5. MacConaill MA. The movements of bones and joints. III: the synovial fluid and its assistants. J Bone Joint Surg Br 1950;32:244–52. 6. Renstrom P, Johnson RJ. Anatomy and biomechanics of the menisci. Clin Sports Med 1990;9(3):523–38. 7. Zimny ML, Albright DJ, Dabezies E. Mechanoreceptors in the human medial meniscus. Acta Anat Basel 1988;133(1):35–40. 8. Baker PE, Peckham AC, Pupparo F, Sanborn JC. Review of meniscal injury and associated sports. Am J Sports Med 1985;13(1):1–4. 9. Campbell SE, Sanders TG, Morrison WB. MR imaging of meniscal cysts: incidence, location, and clinical significance. Am J Roentgenol 2001;177(2):409–13. 10. Drosos GI, Pozo JL. The causes and mechanisms of meniscal injuries in the sporting and non-sporting environment in an unselected population. Knee 2004;11(2):143–9. 11. Smillie IS. The current pattern of the pathology of meniscus tears. Proc R Soc Med 1968;61(1):44–5. 12. Nikolic DK. Lateral meniscal tears and their evolution in acute injuries of the anterior cruciate ligament of the knee. Arthroscopic analysis. Knee Surg Sports Traumatol Arthrosc 1998;6(1):26–30. 13. Warren RF, Marshall J. Injuries of the anterior cruciate and medial collateral ligaments of the knee. Clin Orthop 1978;136:191–7. 14. Wickiewicz TL. Meniscal injuries in the cruciate-deficient knee. Clin Sports Med 1990;9(3):681–94. 15. Vedi V, Williams A, Tennant SJ, Spouse E, Hunt DM, Gedroyc WM. Meniscal movement. An in-vivo study using dynamic MRI. J Bone Joint Surg Br 1999;81(1):37–41. 16. de Roeck NJ, Lang-Stevenson A. Meniscal tears sustained awaiting anterior cruciate ligament reconstruction. Injury 2003;34(5):343–5. 17. McMurray TP. The Semilunar Cartilages. Br J Surg 1942;29:407–14.

I.D. McDermott 18. Ryan PJ, Reddy K, Fleetcroft J. A prospective comparison of clinical examination, MRI, bone SPECT, and arthroscopy to detect meniscal tears. Clin Nucl Med 1998;23(12):803–6. 19. McNeill Love R. Prognosis after removal of semilunar cartilages. Br Med J 1923;ii:324–6. 20. King D. The function of the semilunar cartilages. J Bone Joint Surg Br 1936;18:1069–76. 21. Baratz ME, Fu FH, Mengato R. Meniscal tears: the effect of meniscectomy and of repair on intraarticular contact areas and stress in the human knee. Am J Sports Med 1986;14(4):270–4. 22. Fukuda Y, Takai S, Yoshino N, Murase K, Tsutsumi S, Ikeuchi K, et al. Impact load transmission of the knee joint-influence of leg alignment and the role of meniscus and articular cartilage. Clin Biomech 2000;15(7):516–21. 23. Roos H, Lauren M, Adalberth T, Roos EM, Jonsson K, Lohmander LS. Knee osteoarthritis after meniscectomy: prevalence of radiographic changes after twenty-one years, compared with matched controls. Arthritis Rheum 1998;41:687–93. 24. Allen PR, Denham RA, Swan AV. Late degenerative changes after meniscectomy. Factors affecting the knee after operation. J Bone Joint Surg Br 1984;66(5):666–71. 25. Johnson RJ, Kettelkamp DB, Clark W, Leaverton P. Factors effecting late results after meniscectomy. J Bone Joint Surg Am 1974;56(4):719–29. 26. Jorgensen U, Sonne HS, Lauridsen F, Rosenklint A. Long-term follow-up of meniscectomy in athletes. A prospective longitudinal study. J Bone Joint Surg Br 1987;69(1):80–3. 27. Hoshikawa Y, Kurosawa H, Fukubayashi T, Nakajima H, Watarai K. The prognosis of meniscectomy in athletes. The simple meniscus lesions without ligamentous instabilities. Am J Sports Med 1983;11(1):8–13. 28. Aagaard H, Verdonk R. Function of the normal meniscus and consequences of meniscal resection. Scand J Med Sci Sports 1999;9:134–40. 29. Burke DL, Ahmed AM, Miller J. A biomechanical study of partial and total medial meniscectomy of the knee. Trans Orthop Res Soc 1978;3:91. 30. Ihn JC, Kim SJ, Park IH. In vitro study of contact area and pressure distribution in the human knee after partial and total meniscectomy. Int Orthop 1993;17(4):214–8. 31. Hede A, Larsen E, Sandberg H. The long term outcome of open total and partial meniscectomy related to the quantity and site of the meniscus removed. Int Orthop 1992;16(2):122–5. 32. Annandale T. An operation for displaced semilunar cartilage. 1885 [classical article]. Clin Orthop 1990;260:3–5. 33. King D. The healing of the semilunar cartilages. J Bone Joint Surg 1936;18:333–42. 34. Arnoczky SP, Warren RF. The microvasculature of the meniscus and its response to injury. An experimental study in the dog. Am J Sports Med 1983;11(3):131–41. 35. Cabaud HE, Rodkey WG, Fitzwater JE. Medial meniscus repairs. An experimental and morphologic study. Am J Sports Med 1981;9(3):129–34. 36. Heatley FW. The meniscus—can it be repaired? An experimental investigation in rabbits. J Bone Joint Surg Br 1980;62(3):397–402. 37. Dehaven KE, Arnoczky SP. Meniscus repair: basic science, indications for repair, and open repair. J Bone Joint Surg Am 1994;76(1):140–52. 38. Zhang Z, Arnold JA, Williams T, McCann B. Repairs by trephination and suturing of longitudinal injuries in the avascular area of the meniscus in goats. Am J Sports Med 1995;23(1):35–41. 39. Okuda K, Ochi M, Shu N, Uchio Y. Meniscal rasping for repair of meniscal tear in the avascular zone. Arthroscopy 1999;15(3): 281–6.

ARTICLE IN PRESS Meniscal tears 40. Henning CE, Lynch MA, Clark JR. Vascularity for healing of meniscus repairs. Arthroscopy 1987;3(1):13–8. 41. Arnoczky SP, Warren RF, Spivak JM. Meniscal repair using an exogenous fibrin clot. An experimental study in dogs. J Bone Joint Surg Am 1988;70(8):1209–17. 42. Port J, Jackson DW, Lee TQ, Simon TM. Meniscal repair supplemented with exogenous fibrin clot and autogenous cultured marrow cells in the goat model. Am J Sports Med 1996;24(4):547–55. 43. van-Trommel MF, Simonian PT, Potter HG, Wickiewicz TL. Arthroscopic meniscal repair with fibrin clot of complete radial tears of the lateral meniscus in the avascular zone. Arthroscopy 1998;14(4):360–5. 44. Ochi M, Mochizuki Y, Deie M, Ikuta Y. Augmented meniscal healing with free synovial autografts: an organ culture model. Arch Orthop Trauma Surg 1996;115(3–4):123–6. 45. Jitsuiki J, Ochi M, Ikuta Y. Meniscal repair enhanced by an interpositional free synovial autograft: an experimental study in rabbits. Arthroscopy 1994;10(6):659–66. 46. Shirakura K, Niijima M, Kobuna Y, Kizuki S. Free synovium promotes meniscal healing. Synovium, muscle and synthetic mesh compared in dogs. Acta Orthop Scand 1997;68(1):51–4. 47. Vangsness Jr CT. Use of lasers for meniscal repair. Clin Sports Med 1996;15(3):531–5. 48. Dehaven KE, Black KP, Griffiths HJ. Open meniscus repair. Technique and two to nine year results. Am J Sports Med 1989;17(6):788–95. 49. Rockborn P, Gillquist J. Results of open meniscus repair. Longterm follow-up study with a matched uninjured control group. J Bone Joint Surg Br 2000;82(4):494–8. 50. Barber FA, Stone RG. Meniscal repair. An arthroscopic technique. J Bone Joint Surg Br 1985;67(1):39–41. 51. Jakob RP, Staubli HU, Zuber K, Esser M. The arthroscopic meniscal repair. Techniques and clinical experience. Am J Sports Med 1988;16(2):137–42. 52. Stone RG, Miller GA. A technique of arthroscopic suture of torn menisci. Arthroscopy 1985;1(4):226–32. 53. Morgan CD, Casscells SW. Arthroscopic meniscus repair: a safe approach to the posterior horns. Arthroscopy 1986;2(1):3–12. 54. Warren RF. Arthroscopic meniscus repair. Arthroscopy 1985;1(3):170–2. 55. Maruyama M. The all-inside meniscal suture technique using new instruments. Arthroscopy 1996;12(2):256–8. 56. Morgan CD. The ‘‘all-inside’’ meniscus repair. Arthroscopy 1991;7(1):120–5. 57. Rispoli DM, Miller MD. Options in meniscal repair. Clin Sports Med 1999;18(1):77–91. 58. Asik M, Sener N, Akpinar S, Durmaz H, Goksan A. Strength of different meniscus suturing techniques. Knee Surg Sports Traumatol Arthrosc 1997;5(2):80–3. 59. Kohn D, Siebert W. Meniscus suture techniques: a comparative biomechanical cadaver study. Arthroscopy 1989;5(4):324–7. 60. Post WR, Akers SR, Kish V. Load to failure of common meniscal repair techniques: effects of suture technique and suture material. Arthroscopy 1997;13(6):731–6. 61. Rimmer MG, Nawana NS, Keene GC, Pearcy MJ. Failure strengths of different meniscal suturing techniques. Arthroscopy 1995;11(2):146–50. 62. Albrecht OP, Kristensen G, Burgaard P, Joergensen U, Toerholm C. The arrow versus horizontal suture in arthroscopic meniscus repair. A prospective randomized study with arthroscopic evaluation. Knee Surg Sports Traumatol Arthrosc 1999;7(5):268–73. 63. Barber FA, Herbert MA. Meniscal repair devices. Arthroscopy 2000;16(6):613–8. 64. Boenisch UW, Faber KJ, Ciarelli M, Steadman JR, Arnoczky SP. Pull-out strength and stiffness of meniscal repair using

93

65.

66.

67.

68.

69.

70. 71. 72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

84.

absorbable arrows or Ti-Cron vertical and horizontal loop sutures. Am J Sports Med 1999;27(5):626–31. Dervin GF, Downing KJ, Keene GC, McBride DG. Failure strengths of suture versus biodegradable arrow for meniscal repair: an in vitro study. Arthroscopy 1997;13(3):296–300. Song EK, Lee KB. Biomechanical test comparing the load to failure of the biodegradable meniscus arrow versus meniscal suture. Arthroscopy 1999;15(7):726–32. Barrett GR, Treacy SH, Ruff CG. Preliminary results of the T-fix endoscopic meniscus repair technique in an anterior cruciate ligament reconstruction population. Arthroscopy 1997;13(2): 218–23. Hurel C, Mertens F, Verdonk R. Biofix resorbable meniscus arrow for meniscal ruptures: results of a 1-year follow-up. Knee Surg Sports Traumatol Arthrosc 2000;8(1):46–52. Koukoubis TD, Glisson RR, Feagin Jr JA, Seaber AV, Schenkman D, Korompilias AV, et al. Meniscal fixation with an absorbable staple. An experimental study in dogs. Knee Surg Sports Traumatol Arthrosc 1997;5(1):22–30. Oliverson TJ, Lintner DM. Biofix arrow appearing as a subcutaneous foreign body. Arthroscopy 2000;16(6):652–5. Calder SJ, Myers PT. Broken arrow: a complication of meniscal repair. Arthroscopy 1999;15(6):651–2. Kotsovolos ES, Hantes ME, Mastrokalos DS, Lorbach O, Paessler HH. Results of all-inside meniscal repair with the FasT-Fix meniscal repair system. Arthroscopy 2006;22(1):3–9. Kenny C, Krackow KA, McCarthy EF. An evaluation of the effects of a silastic meniscus prothesis on post-meniscectomy osteoarthrosis. Trans Orthop Res Soc 1983;8:335. Veth RP, den-Heeten GJ, Jansen HW, Nielsen HK. An experimental study of reconstructive procedures in lesions of the meniscus. Use of synovial flaps and carbon fiber implants for artificially made lesions in the meniscus of the rabbit. Clin Orthop 1983; (181): 250–4. Sommerlath K, Gallino M, Gillquist J. Biomechanical characterisitcs of different artificial substitutues for the rabbit medial meniscus and the effect of prosthesis size on cartilage. Trans Orthop Res Soc 1991;16(2):375. Sommerlath KG, Gillquist J. The effect of anterior cruciate ligament resection and immediate or delayed implantation of a meniscus prosthesis on knee joint biomechanics and cartilage. An experimental study in rabbits. Clin Orthop 1993; (289): 267–75. Sommerlath K, Gillquist J. The effects of an artificial meniscus substitute in a knee joint with a resected anterior cruciate ligament. An experimental study in rabbits. Clin Orthop 1993; (289): 276–84. Sommerlath K, Gillquist J. The effect of a meniscal prosthesis on knee biomechanics and cartilage. An experimental study in rabbits. Am J Sports Med 1992;20(1):73–81. Messner K. Meniscal substitution with a Teflon-periosteal composite graft: a rabbit experiment. Biomaterials 1994;15(3):223–30. Johnson LL, Feagin -JAJ. Autogenous tendon graft substitution for absent knee joint meniscus: a pilot study. Arthroscopy 2000;16(2):191–6. Kohn D. Autograft meniscus replacement: experimental and clinical results. Knee Surg Sports Traumatol Arthrosc 1993;1(2):123–5. Kohn D, Wirth CJ, Reiss G, Plitz W, Maschek H, Erhardt W, et al. Medial meniscus replacement by a tendon autograft. Experiments in sheep. J Bone Joint Surg Br 1992;74(6):910–7. Kohn D, Rudert M, Wirth CJ, Plitz W, Reiss G, Maschek H. Medial meniscus replacement by a fat pad autograft. An experimental study in sheep. Int Orthop 1997;21(4):232–8. Bruns J, Kahrs J, Kampen J, Behrens P, Plitz W. Autologous perichondral tissue for meniscal replacement. J Bone Joint Surg Br 1998;80(5):918–23.

ARTICLE IN PRESS 94 85. Canham W, Stanish W. A study of the biological behavior of the meniscus as a transplant in the medial compartment of a dog’s knee. Am J Sports Med 1986;14(5):376–9. 86. Milachowski KA, Weismeier K, Wirth CJ. Homologous meniscus transplantation. Experimental and clinical results. Int Orthop 1989;13(1):1–11. 87. Zukor DJ, Rubins IM, Daigle MR, Rudan JF, Roy I, Duval P, et al. Allotransplantation of frozen irradiated menisci in rabbits. Trans Orthop Res Soc 1990;15(1):219. 88. Ochi M, Ishida O, Daisaku H, Ikuta Y, Akiyama M. Immune response to fresh meniscal allografts in mice. J Surg Res 1995;58(5):478–84. 89. Wada Y. Meniscal transplantation using fresh and cryopreserved allografts—an experimental study in the genetically defined rat. Nippon Seikeigeka Gakkai Zasshi 1993;67(7):677–83. 90. Fabbriciani C, Lucania L, Milano G, Schiavone PA, Evangelisti M. Meniscal allografts: cryopreservation vs deep-frozen technique. An experimental study in goats. Knee Surg Sports Traumatol Arthrosc 1997;5(2):124–34. 91. Nagahara K, Wada Y. Allotransplantation of cryopreserved menisci in monkey: histological change. Nippon Seikeigeka Gakkai Zasshi 1994;68(9):784–9. 92. Milachowski KA, Weismeier K, Wirth CJ, Kohn D. Meniscus transplantation—experimental study and first clinical report. Am J Sports Med 1987;15:626. 93. van-Arkel ER, de-Boer HH. Survival analysis of human meniscal transplantations. J Bone Joint Surg 2002;84-B(2):227–31. 94. Verdonk R, Van-Daele P, Claus B, Van-den-Abbeele K, Desmet P, Hurel C, et al. Viable meniscal transplantation. Acta Orthop Belg 1998;64(Suppl.):270–5. 95. Insall JN, Dorr LD, Scott RD, Scott WN. Rationale of the Knee Society clinical rating system. Clin Orthop 1989; (248): 13–4. 96. Garrett JC, Stevensen R. Meniscal transplantation in the human knee: a preliminary report. Arthroscopy 1991;7(1):57–62. 97. Rodeo SA, Seneviratne A, Suzuki K, Felker K, Wickiewicz TL, Warren RF. Histological Analysis of Human Meniscal Allografts. J Bone Joint Surg 2000;82-A(8):1071–82. 98. Cameron JC, Saha S. Meniscal allograft transplantation for unicompartmental arthritis of the knee. Clin Orthop 1997;337:164–71. 99. Zukor DJ, Cameron JC, Brooks PJ, Oakeshott RD, Gross AE. The fate of human meniscal allografts. Orthop Trans 1988;12:658.

I.D. McDermott 100. van-Arkel ER, de-Boer HH. Human meniscal transplantation. Preliminary results at 2 to 5-year follow-up. J Bone Joint Surg Br 1995;77(4):589–95. 101. de-Boer HH, Koudstaal J. Failed meniscus transplantation. A report of three cases. Clin Orthop 1994;306:155–62. 102. Rath E, Richmond JC, Yassir W, Albright JD, Gundogan F. Meniscal allograft transplantation. Two- to eight-year results. Am J Sports Med 2001;29(4):410–4. 103. van-Arkel ER, de-Boer HH. Human meniscus transplantation. Agents Actions 1993;39(Suppl.):243–6. 104. Stollsteimer GT, Shelton WR, Dukes A, Bomboy AL. Meniscal allograft transplantation: a 1- to 5-year follow-up of 22 patients. Arthroscopy 2000;16(4):343–7. 105. McGinty JB. Meniscal allografts [editorial]. Am J Knee Surg 1996;9(1):1. 106. Arnoczky SP. Building a meniscus. Biologic considerations. Clin Orthop 1999;367(Suppl.):S244–53. 107. Stone KR, Rodkey WG, Webber RJ, McKinney L, Steadman JR. Future directions. Collagen-based prostheses for meniscal regeneration. Clin Orthop 1990;252:129–35. 108. Stone KR, Rodkey WG, Webber R, McKinney L, Steadman JR. Meniscal regeneration with copolymeric collagen scaffolds. In vitro and in vivo studies evaluated clinically, histologically, and biochemically. Am J Sports Med 1992;20(2):104–11. 109. Stone KR, Steadman JR, Rodkey WG, Li ST. Regeneration of meniscal cartilage with use of a collagen scaffold. Analysis of preliminary data. J Bone Joint Surg Am 1997;79(12):1770–7. 110. Rodkey WG, Steadman JR, Li ST. A clinical study of collagen meniscus implants to restore the injured meniscus. Clin Orthop 1999;367(Suppl.):S281–92. 111. Ibarra C, Jannetta C, Vacanti CA, Cao Y, Kim TH, Upton J, et al. Tissue engineered meniscus: a potential new alternative to allogeneic meniscus transplantation. Transplant Proc 1997;29(1–2):986–8. 112. de-Groot JH, Zijlstra FM, Kuipers HW, Pennings AJ, Klompmaker J, Veth RP, et al. Meniscal tissue regeneration in porous 50/50 copoly(L-lactide/epsilon-caprolactone) implants. Biomaterials 1997;18(8):613–22. 113. Walsh CJ, Goodman D, Caplan AI, Goldberg VM. Meniscus regeneration in a rabbit partial meniscectomy model. Tissue Eng 1999;5(4):327–37. 114. Goto H, Shuler FD, Niyibizi C, Fu FH, Robbins PD, Evans CH. Gene therapy for meniscal injury: enhanced synthesis of proteoglycan and collagen by meniscal cells transduced with a TGFbeta(1)gene. Osteoarthritis Cartilage 2000;8(4):266–71.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 95–102

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: SOFT TISSUE KNEE PROBLEMS

(iii) The dislocated knee A. Robertsona, R.W. Nuttonb, a

The Royal Infirmary of Edinburgh, Scotland, UK Department of Orthopaedics, Royal Infirmary, Little France, Old Dalkeith Road, Edinburgh EH16 4SU, Scotland, UK

b

KEYWORDS Knee joint; Knee injuries; Knee dislocation; Dislocations

Summary Knee dislocation is a rare injury usually resulting from high-energy trauma. Immediate complications can arise from injury to the popliteal artery or the common peroneal nerve. Early surgical reconstruction gives better results than delayed reconstruction, but this is a challenging procedure. The prognosis for knee function should be guarded, as there is a significant risk of long-term secondary degenerative change. & 2006 Elsevier Ltd. All rights reserved.

Introduction Knee dislocation is a serious injury with potential significant long-term implications for return to physical employment and recreational activity (Fig. 1a and b). The clinical presentation of knee dislocation may be either acute— under 3 weeks—or chronic—after 3 weeks. Late presentation with multiple knee ligament injuries is not uncommon, as spontaneous relocation of the dislocated knee may lead to an initial underestimation of the severity of the ligamentous disruption. This is particularly likely when the knee dislocation was part of multiple trauma and initial treatment priorities were focused on long bone fractures, etc. rather than on the stability of the knee.

Epidemiology Knee dislocation is uncommon representing less than 0.2% of all orthopaedic injuries.1 It predominantly occurs in younger Corresponding author. Tel.: +44 131 242 3466; fax: +44 131 242 3493. E-mail addresses: [email protected] (A. Robertson), [email protected] (R.W. Nutton).

0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.03.001

patients, with a male to female ratio of 4:1. Knee dislocation is reported to be a component of multiple trauma in between 14% and 44% of cases.2–7 Half are the result of motor vehicle accidents, sports injuries account for around one-third of cases and simple falls for approximately 10%. Bilateral dislocation is rare occurring in only 5% of patients2–7 (Table 1).

Classification In 1963 Kennedy proposed an anatomical classification based on the direction of tibial dislocation in relation to the femur.8 Five types of knee dislocation are described:

    

anterior, posterior, lateral, medial, rotatory.

Rotatory is further subdivided into anteromedial, posteromedial, anterolateral and posterolateral types.

ARTICLE IN PRESS 96

A. Robertson, R.W. Nutton

Figure 1 (a,b) Radiograph of postero-medial knee dislocation.

Table 1 Study

The epidemiology of knee dislocation. No. of patients Mean age M:F ratio MVA (%) Sports (%) Falls (%) Multiple trauma (%) Bilateral (%)

Harner et al. (2004)5 31 21 Liow et al. (2003) 4 Twaddle et al. (2003)3 60 Werier et al. (1998)2 36 Wascher et al. (1997)6 47 31 Almekinders, (1992)7 Pooled estimate 226

28 28 28 32 28 32 29

— 16:5 48:12 28:8 36:11 26:5 4:1

Although attractive in terms of simplicity, this directionbased classification system is an unreliable guide to specific patterns of ligament injury. An alternative classification system, which addresses the specific pattern of ligament disruption and the presence or absence of associated intra-articular fractures was described by Schenk9 and modified by others (Table 2). This gives a clearer guide to patterns of knee ligament injury and can thus be used to plan treatment.

26 43 57 69 79 52 56

23 43 38 11 21 — 32

6 14 5 17 6 20 10

Excluded 14 — 16 44 — 29

— 5 5 6 6 0 5

associated with tibial plateau fractures (Fig. 2). A combination of force vectors results in rotatory dislocations. Under the Schenk classification, the most common injury pattern is a bi-cruciate disruption with an associated medial (MCL) or lateral (LCL) collateral tear depending on the direction of the deforming force2,4–6 (Table 3).

The acutely dislocated knee Clinical assessment

Patterns of ligament injury Using the Kennedy classification, hyperextension of the knee resulting in anterior dislocation of the tibia on the femur is the commonest mechanism of injury accounting for approximately 40% of reported cases. Posterior dislocation (33%) usually occurs as a result of an anteroposterior force, as in the ‘dashboard’ type of injury. Varus or valgus loads may produce medial (4%) or lateral (18%) dislocations, which are

After initial patient assessment and management under ATLS protocols, (particularly if multiply injured), the vascular and neurological status of the injured limb must be carefully and repeatedly assessed for evidence of injury to the popliteal artery and common peroneal nerve. The clinical findings must be clearly recorded. The clinical diagnosis should be confirmed by X-ray and the dislocation should be reduced under sedation as soon as possible and congruent reduction confirmed radiologically.

ARTICLE IN PRESS The dislocated knee

97

Table 2 The classification of knee dislocation based on the extent of ligament injury (adapted from Schenk 1994). Classification

Type

KD-I KD-II KD-III

Dislocation Single cruciate involved Bicruciate disruption only Bicruciate+posteromedial or posterolateral disruption Bicruciate+posteromedial and posterolateral disruption

KD-IV

KD-V KD-V1 KD-V2 KD-V3M KD-V3L KD-V4

Associated ligament injury

Single cruciate involved Fracture Bicruciate disruption only Dislocation Bicruciate+posteromedial disruption Bicruciate+posterolateral disruption Posteromedial and posterolateral disruption

knee flexion, muscle spasm and other associated injuries. Recurvatum on passive limb elevation (Fig. 3) and gross laxity on varus or valgus testing with the knee in full extension (Fig. 4) indicate a major cruciate or capsular disruption as a result of knee dislocation. Plain AP and lateral radiographs should be obtained in all cases of suspected dislocation given the high incidence of associated fractures and avulsions.

Vascular injury and angiography The reported incidence of vascular injury in knee dislocation ranges from 4.8% in low-velocity injuries to 65% in higher energy trauma.10,11 Hence, a patient must be carefully assessed for signs of impaired circulation including ischaemic colour change, diminished or absent pulses with colour or temperature changes below the level of the knee. The ankle-brachial index (ABI) can supplement physical examination and may assist in deciding whether angiography is indicated. However, while very sensitive, the ABI has low specificity.12 Using a threshold of o0.90; however, it is possible to achieve a 100% positive predictive value for the presence of a vascular injury requiring surgical intervention.13

Associated injuries

Figure 2 Lateral tibial plateau fracture associated with knee dislocation.

While patients may present without clinical and radiological signs of a dislocation if the joint has reduced spontaneously, the clinical signs should alert the clinician to the extent of the soft tissue injury to the knee. For example, an uncontained haemarthrosis with extensive bruising and swelling on the medial or lateral side of the knee suggests major disruption of the joint capsule, which should alert the examiner to the possibility of a dislocation. However, acutely a thorough clinical assessment of specific knee ligaments for injury may be difficult due to pain, limited

In association with high-energy trauma, fractures of the distal femur or tibial plateau and damage to the common peroneal nerve can occur. Fractures of either the distal femur or proximal tibia are reported to occur in between 4.5% and 34% of cases of knee dislocation.2–4,6 Bony avulsion injuries of the PCL are not unusual and marginal lateral tibial plateau avulsion fractures (Segond fractures) or fibular head fractures are seen, indicative of significant capsular, collateral and cruciate disruption. Anteromedial tibial plateau fractures in particular are associated with disruption of the posterior cruciate ligament (PCL) and posterolateral ligament complex (PLC). Common peroneal nerve injury has an overall incidence of approximately 20%,2,3,6,7 but in dislocations with disruptions of the PCL and PLC the incidence may be as high as 45% due to the varus deforming force. Actual discontinuity of the nerve is present in 28% of cases but even if it is in continuity there may be extensive intraneural damage.18 When the nerve remains in continuity, spontaneous recovery occurs in 20% of cases with a more favourable prognosis if the nerve is damaged over short distances.4 Nerve grafting or tendon transfers may be considered as late reconstructive procedures if there is no recovery of nerve function.18

Radiology Plain X-rays of the knee demonstrate peri-articular fractures or avulsion fractures. Marginal fractures around the tibial plateau reflect significant capsule avulsions (Segond fracture) and anterior tibial plateau fractures are seen in hyperextension injuries with extensive disruption of the posterior capsule. Avulsion fractures occur at the tibial insertion of the PCL and at the tip of the fibula when the LCL has been avulsed. These reflect serious soft tissue

ARTICLE IN PRESS 98

A. Robertson, R.W. Nutton

Table 3

Patterns of knee ligament disruption associated with knee dislocation.

Study

No.

Harner et al. (2004)5 Liow et al. (2003)4 Werier et al. (1998)2 Wascher et al. (1997)6 Pooled estimate

Figure 3

ACL/PCL/MCL/ PLC (%)

ACL/PCL/MCL (%)

ACL/PCL/PLC (%)

ACL/PCL (%)

ACL or PCL+other/ # (%)

19

0

53

37

11

0

22

0

23

32

9

45

38

21

37

37

3

26

50

12

48

16

12

10

129

11

41

28

7

12

Recurvatum following a knee dislocation.

disruption. The extent of this is best shown by MRI scanning which should always be performed before undertaking surgical reconstruction. The necessity for routine angiography has been contentious, some authors considering it mandatory4,14,15 while others believe that it is only necessary in selected cases.16,17 Overall the published evidence suggests that routine angiography in knee dislocation is not necessary, only being performed if there is doubt about the circulatory status of the limb. Magnetic resonance angiography (MRA) is an alternative to standard angiography.

Early management and surgical reconstruction Following a thorough neurovascular assessment of the limb the immediate priority is to reduce the knee and provide temporary stabilisation of the knee with an external splint. Following reduction the vascular status of the limb should be reassessed. If there is any doubt about limb perfusion a vascular surgical opinion should be sought and urgent angiography performed. Surgical reconstruction of the popliteal artery

Figure 4

Varus instability following knee dislocation.

should take priority as delay in restoring circulation is associated with an increased risk of amputation. Under the same anaesthetic as the vascular repair, the collateral ligaments and capsule injury should be explored. The aim is to achieve sufficient stability of the knee to protect the vascular repair Primary repair of the capsular and collateral structures may be accomplished by ligament reattachment, augmentation or a combination of techniques and associated fractures of the tibial plateau or distal femur should be stabilised. If the knee remains unstable further temporary stabilisation may be obtained by using a bridging external fixator. Once initial stability and limb viability has been achieved, delayed reconstruction of the cruciate ligaments can be carried out as a staged procedure without a tourniquet at 6–12 weeks.

ARTICLE IN PRESS The dislocated knee If there is no vascular injury the surgical approach will be determined by the pattern of ligament injury. Immediate reconstruction of all injured ligaments is a major surgical challenge requiring expertise in knee ligament reconstructive surgery and access to essential resources including allograft tendons. Where expertise and such resources are not immediately available, a limited capsular and collateral ligament reconstruction should be performed. This may be achieved by reattaching avulsed structures using suture anchors, ligament staples or screws and ligament washers. Mid substance tears of the collateral ligaments may be repaired directly and augmented with locally available tissue from the medial hamstrings for the MCL and biceps femoris for the LCL. If cruciate ligament reconstruction is being considered in the acute phase the degree of capsular injury must be considered. Attempts to perform arthroscopic surgery before capsule healing increase the risk of fluid extravasation and may cause a compartment syndrome. If arthroscopic repair is undertaken, a pump should not be used. Cruciate ligament reconstruction (especially PCL reconstruction) in this situation is technically difficult and should not be attempted unless the surgeon has considerable experience in knee ligament reconstruction surgery. In addition access to tendon allografts is an advantage in order to reduce the demand for autogenous tissue from the damaged knee. It is possible to consider harvest from the opposite leg but this is not always practical and the patient may object to his ‘normal’ leg being used for this purpose. Arthroscopically assisted anterior cruciate ligament (ACL) and PCL reconstruction can be performed when the capsule has healed and the patient has started to regain range of movement. Under these circumstances the patient can be referred to a centre, which specialises in knee ligament reconstruction for the second stage of the reconstruction.

Chronic knee instability following dislocation Chronic knee instability may result from an episode of dislocation. The multiple ligament injuries caused by the dislocation causes complex patterns of instability, which can be very disabling. Such need careful objective assessment to accurately delineate the pattern of ligament damage. A management plan must address whether the patient will benefit from ligament reconstruction surgery, taking into account the complexity of the surgery and the prolonged rehabilitation. Additionally, if secondary changes of altered mechanical alignment and secondary degenerative changes have occurred, then corrective osteotomy may be required before considering ligament reconstruction. This complex scenario could require two operations, firstly to correct malalignment and secondly to perform a multiple ligament reconstruction. Patients must be aware of the implications of this complex and protracted course of treatment.

Primary and secondary ligament restraint The concept of primary and secondary restraint is central to understanding complex knee instability. A primary restraint is the main resistance to a deforming force on the knee e.g. the medial collateral is the primary restraint to valgus

99 force on the knee. A secondary restraint has a smaller but significant role in resisting a force, e.g. the ACL also resists a valgus force applied to the knee. If a secondary restraint is ruptured when the primary restraint is intact, there will be no discernable instability when a force is applied to the knee. However, if the secondary restraint is ruptured as well as the primary restraint, the magnitude of the instability is greater when the same force is applied, e.g. the degree of valgus instability is greater when the MCL and ACL are torn than when the MCL is torn in isolation. Other important primary and secondary restraints are the PCL and popliteofibular ligament (PFL) which restrain external rotation of the tibia and the ACL and PLC in controlling the dynamic instability demonstrated by the pivot shift test. The LCL is the primary restraint to varus force, whereas the PLC, particularly the PFL is the primary restraint to tibial external rotation.

Clinical examination The patient is first observed standing and walking to assess limb alignment at the knee. Increased varus and a lateral thrust on walking suggest varus malalignment. This can be confirmed by taking weight bearing long leg alignment radiographs. The normal knee should always be examined for comparison. The range of knee movement must be recorded. In particular abnormal hyperextension compared to the uninjured knee may indicate disruption of the posterior capsule. The varus recurvatum test assesses the passive extension of the knee as well as the tendency for the knee to fall into varus when held in an extended position suggesting damage to the LCL and PLC. The examination should progress in a logical stepwise fashion addressing each major ligament restraints to each of the main deforming forces, varus, valgus and anteroposterior in turn before assessing dynamic instability using the pivot and reverse pivot tests. The collateral ligaments should be examined with the knee in full extension and with the posterior capsule relaxed with the knee flexed to 201. The degree of joint line gapping and the quality of the endpoint should be noted. The cruciate ligaments are assessed with the knee flexed to 901. The position of the anterior margin of the tibia in relation to the femoral condyles to assess posterior sag is noted, remembering that when both cruciate ligaments are torn, the anteroposterior translation is increased such that it is difficult to determine the neutral anteroposterior position of the tibia. Abnormal posterolateral or posteromedial rotation of the tibia is then assessed. The Lachman test, so useful for detecting isolated ACL tears, is of less value in multiple ligament injuries. Increased tibial external rotation is an important feature of more severe types of knee instability, as the lateral ligament complex, particularly the popliteofibular ligament (PFL), is the primary restraint to external tibial rotation with the PCL a secondary restraint. Tibial external rotation is best examined with the dial test performed with the patient prone (Fig. 5a and b). At 301 of flexion the PFL is the primary restraint, therefore if the tibia (indicated by the medial border of the foot) externally

ARTICLE IN PRESS 100

A. Robertson, R.W. Nutton

Figure 5 (a,b) The dial test performed at 301 and 901 of knee flexion.

rotates by more than 101 compared to the uninjured knee, the PFL is damaged. At 901 of knee flexion however, the PCL will act as a secondary restraint and will limit external tibial rotation even if the PFL is damaged. If the dial test is positive at 901 of knee flexion, then both the PFL and PCL have been torn. The final part of the examination assesses the dynamic stability of the knee by performing the pivot and reverse pivot shift tests. The pivot shift is positive when the ACL is torn but the degree of instability is increased with the tibia tending to ‘‘lock out’’ when the PLC is damaged along with the ACL. The reverse pivot shift also assesses the PLC but can be positive in normal knees if the patient has a degree of joint hypermobility. A positive reverse pivot shift reflects injuries to the PCL and PLC resulting in a posterolateral rotatory instability (PLRI). Clinical examination should define the pattern of ligament injury, but MRI scanning is helpful in confirming clinical findings. Correctly advising the patient as to the most appropriate management will depend on many factors including the patient’s age, the physical demands they place on the knee, the degree of disability experienced by the patient and their ability to cooperate with the extended programme of rehabilitation required after surgery. If malalignment of the knee is present, ligament reconstruction surgery will fail and osteotomy should be considered with or without ligament reconstruction.

Surgical reconstruction Detailed description of the surgical techniques used for ligament reconstruction in complex knee instability is beyond the scope of this review but the basic principles

will be described. The principle behind all reconstructive procedures used for complex knee instability is to define all components of the instability and to reconstruct the primary restraints as anatomically and isometrically as possible. In patients with bi-cruciate injuries or if the PCL alone is ruptured the first objective is to reduce the posterior subluxation of the tibia by reconstructing the PCL. This reconstruction can be performed using autograft from the injured knee or from the opposite leg. Semitendinosus and gracilis, patella tendon and quadriceps tendon have all used for this purpose. In the multiply injured knee in which more than one ligament will be reconstructed it is not advisable to take more than one autograft from the affected leg where the soft tissues are already compromised as a result of injury. In these cases it is advisable to have access to allograft tendons which can be used to reconstruct one or both cruciate ligaments.19 An Achilles tendon allograft is suitable for reconstruction of the PCL and patella tendon allograft for the ACL reconstruction. Techniques for reconstruction of collateral ligaments depend on the precise nature of the injury. If the MCL has been avulsed from the medial epicondyle (Pellegrini Stieda type of injury), the femoral attachment can be recessed into the medial condyle of the femur, which should maintain the isometric position of the MCL. This is only effective if there has not been structural lengthening of the superficial band of the MCL, in which case, the superficial band can be augmented using gracilis or semitendinosus. These are left attached to the proximal tibia and fixed in an isometric position to the medial femoral condyle using a screw and ligament washer. Care should be taken not to over tighten the MCL reconstruction, which can ‘capture’ the knee preventing full extension.

ARTICLE IN PRESS The dislocated knee As the knee may be unstable in varus or in external rotation, or both, reconstruction of the lateral side of the knee must address either or both components of the instability if present. Advancement techniques as used on the medial side of the knee are less effective for deficiency of the lateral structures and will only address mild degrees of lateral ligament deficiency. Most reconstructive procedures require augmentation of either the LCL or popliteus/ PFL using autografts or allografts. The Larson technique using a semitendinosus autograft has been widely used and modified to address deficiencies of the LCL and PFL.20

Rehabilitation Although accelerated post operative rehabilitation programmes have transformed the recovery process for patients undergoing isolated ACL reconstruction, these regimes are not suitable for patients who have undergone multiple ligament repairs or reconstruction following a knee dislocation.4,5,21–25 In these patients the rehabilitation should be tailored to the particular pattern of injury and reconstruction. The principles of rehabilitation remain the same in the acutely injured knee and following delayed reconstruction. Close liaison with a physiotherapist familiar with this type of rehabilitation programme is essential and a team approach undoubtedly benefits the patient. In most cases the PCL has been reconstructed and posterior tibial subluxation reduced. Not only is it is essential to avoid exercises which will stress the PCL reconstruction, in particular open chain hamstrings contractions, but gravity is also acting on the reconstruction causing the tibia to fall back especially when the knee is flexed relaxing the posterior capsule. Thus if the PCL has been reconstructed as part of the multiple ligament repair it is advisable to splint the knee in extension (but avoiding hyperextension) for the first 6 weeks. Total immobilisation however is not beneficial for capsule repairs or augmentation procedures, which will benefit from controlled movement of the knee during the initial healing period. Many different approaches have been described to address this conflict between splinting the knee straight and starting controlled motion. In the senior author’s practice the knee is splinted in a hinged brace at 101 of flexion to avoid hyperextension. The brace is unlocked regularly to allow passive knee flexion and active knee extension, initially from 10–601 for the first 3 weeks, increasing the range to 10–601 from 3 to 6 weeks post-operatively. During this time the patient is allowed to ‘touch’ weightbear on their toes rather than remaining non-weightbearing. After 6 weeks the brace is unlocked to allow unrestricted knee flexion and is discarded after 8 weeks. From 6 to 10 weeks the patient is advised to increase weight bearing on the affected leg, usually progressing to full weight bearing by 10 weeks. Muscle strengthening progresses gradually after removal of the brace, but it is advisable to avoid open chain hamstring exercises for up to 6 months. Closed chain exercises and active knee extensions are encouraged. The process of recovery is gradual and individual to a particular patient however it is unusual for patients to return to manual work in less than 6 months and sporting activities by 12 months.

101

Outcomes Knee dislocation inevitably results in severe soft tissue disruption. Earlier studies suggested that knee dislocation could be treated non operatively with good results.5,27 However, while improvements in surgical techniques allied to a better understanding of the injuries have resulted in an increasing body of evidence demonstrating that early reconstruction, where feasible, will result in better outcomes,2,4,21–23 return to normal function is uncommon. In recent studies where patients were evaluated using the IKDC26 score, no patients were rated as normal at the time of review and overall, 38% were nearly normal, 40% abnormal and 21% severely abnormal.4–6 A recent meta-analysis of studies on the management of knee dislocation concluded that knee ligament reconstruction was associated with better overall results, with a significantly lower risk of residual stiffness and better Lysholm scores.28 Although, ligament reconstruction has improved the range of motion and functional outcome scores some residual impairment of knee function can be expected.28 Common complications after surgical reconstruction are joint stiffness and failure of some component of the reconstruction. Post-traumatic osteoarthritis may be expected in up to 50% of patients in the long term.2

Practice points

 Remain alert to the possibility of knee dislocation in high-energy trauma

 Careful evaluation and close monitoring of the neurovascular status of the limb is essential

 Early stabilisation with primary repair of capsular

 

structures and avulsed collateral ligaments prior to formal cruciate reconstruction may be performed if the facilities and expertise are not available to perform early definitive reconstruction The main objective in delayed reconstruction is to identify all components of the knee instability and reconstruct the primary restraints A team approach to rehabilitation is beneficial to the patient’s recovery

Research directions

 Epidemiological studies of factors determining outcomes in multiple ligament injuries

 Bioengineering studies on optimum techniques for ligament reconstruction

 Tissue culture research to engineer ligaments for implantation

 Rehabilitation techniques to optimise functional recovery following multiple ligament reconstruction

References 1. Rihn JA, Groff YJ, Harner CD, Cha PS. The acutely dislocated knee: evaluation and management. J Am Acad Orthop Surg 2004;12(5):334–46.

ARTICLE IN PRESS 102 2. Werier J, Keating JF, Meek RN. Complete dislocation of the knee—the long-term results of ligamentous reconstruction. Knee 1998;5:255–66. 3. Twaddle BC, Bidwell TA, Chapman JR. Knee dislocations: where are the lesions? A prospective evaluation of surgical findings in 63 cases. J Orthop Trauma 2003;17(3):198–202. 4. Liow RY, McNicholas MJ, Keating JF, Nutton RW. Ligament repair and reconstruction in traumatic dislocation of the knee. J Bone Jt Surg Br 2003;85(6):845–51. 5. Harner CD, Waltrip RL, Bennett CH, Francis KA, Cole B, Irrgang JJ. Surgical management of knee dislocations. J Bone Jt Surg Am 2004;86-A(2):262–73. 6. Wascher DC, Dvirnak PC, DeCoster TA. Knee dislocation: initial assessment and implications for treatment. J Orthop Trauma 1997;11(7):525–9. 7. Almekinders LC, Logan TC. Results following treatment of traumatic dislocations of the knee joint. Clin Orthop Rel Res 1992(284):203–7. 8. Kennedy JC. Complete dislocation of the knee joint. J Bone Jt Surg Am 1963;45:889–904. 9. Schenk RC. The dislocated knee. Instr Course Lect 1994;43: 127–36. 10. Meyers MH, Harvey Jr JP. Traumatic dislocation of the knee joint. A study of eighteen cases. J Bone Jt Surg Am 1971;53(1): 16–29. 11. McCoy GF, Hannon DG, Barr RJ, Templeton J. Vascular injury associated with low-velocity dislocations of the knee. J Bone Jt Surg Br 1987;69(2):285–7. 12. Applebaum R, Yellin AE, Weaver FA, Oberg J, Pentecost M. Role of routine arteriography in blunt lower extremity trauma. Am J Surg 1990;160:221–5. 13. Mills WJ, Barei DP, McNair P. The value of the ankle-brachial index for diagnosing arterial injury after knee dislocation: a prospective study. J Trauma 2004;56(6):1261–5. 14. Green NE, Allen BL. Vascular injuries associated with dislocation of the knee. J Bone Jt Surg Am 1977;59(2):236–9. 15. McCutchan JD, Gillham NR. Injury to the popliteal artery associated with dislocation of the knee: palpable distal pulses do not negate the requirement for arteriography. Injury 1989;20(5):307–10.

A. Robertson, R.W. Nutton 16. Klineberg EO, Crites BM, Flinn WR, Archibald JD, Moorman III CT. The role of arteriography in assessing popliteal artery injury in knee dislocations. J Trauma 2004;56(4):786–90. 17. Fanelli GC, Orcutt DR, Edson CJ. The multiple-ligament injured knee: evaluation, treatment and results. Arthroscopy 2005; 21(4):471–86. 18. Niall DM, Nutton RW, Keating JF. Palsy of the common peroneal nerve after traumatic dislocation of the knee. J Bone Jt Surg Br 2005;87(5):664–7. 19. Nutton RW, McLean I, Melville E. Tendon allografts in knee ligament surgery. J R Coll Surg Edinberg 1999;44(4):236–40. 20. Larson RL. Combined instabilities of the knee. Clin Orthop Rel Res 1980(147):68–75. 21. Wong CH, Tan JL, Chang HC, Khin LW, Low CO. Knee dislocations—a retrospective study comparing operative versus closed immobilization treatment outcomes. Knee Surg Sports Traumatol Arthrosc 2004;12(6):540–4. 22. Rios A, Villa A, Fahandezh H, de JC, Vaquero J. Results after treatment of traumatic knee dislocations: a report of 26 cases. J Trauma 2003;55(3):489–94. 23. Almekinders LC, Dedmond BT. Outcomes of the operatively treated knee dislocation. Clin Sports Med 2000;19(3):503–18. 24. Wascher DC, Becker JR, Dexter JG, Blevins FT. Reconstruction of the anterior and posterior cruciate ligaments after knee dislocation: results using fresh-frozen nonirradiated allografts. Am J Sports Med 1999;27(2):189–96. 25. Noyes FR, Barber-Westin SD. Reconstruction of the anterior and posterior cruciate ligaments after knee dislocation: use of early protected postoperative motion to decrease arthrofibrosis. Am J Sports Med 1997;25(6):769–78. 26. International Knee Documentation Committee. Knee ligament injury and reconstruction evaluation. In: Aichroth PM, Cannon WDJ, editors. Knee surgery: current practice. London: Martin Dunitz Ltd.; 1992. p. 759–60. 27. Taylor AR, Arden GP, Rainey HA. Traumatic dislocation of the knee: a report of forty-three cases with special reference to conservative treatment. J Bone Jt Surg Br 1972;54(1):96–102. 28. Dedmond BT, Almekinders LC. Operative versus nonoperative treatment of knee dislocations: a meta-analysis. Am J Knee Surg 2001;14(1):33–8.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 103–111

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: SOFT TISSUE KNEE PROBLEMS

(iv) Patellofemoral dysfunction—Extensor mechanism malalignment Simon Donell Institute of Health, University of East Anglia, Norwich NR4 7TJ, UK

KEYWORDS Patellofemoral joint; Patellar instability; Malalignment syndromes; Trochleoplasty

Summary Anterior knee pain has many causes. When it arises from the knee extensor mechanism, those with normal alignment must be distinguished from those with malalignment. Patellar dislocation is at one extreme of a spectrum of disorders that are due to malalignment of the extensor mechanism of the knee. This article discusses important aspects of extensor mechanism malalignment and a logical approach to operative treatment for patients with failed conservative therapy. & 2006 Elsevier Ltd. All rights reserved.

Introduction The patella is a sesamoid bone that acts as a marker for the extensor mechanism of the knee. Patellofemoral dysfunction comprises a number of disorders of the extensor mechanism of the knee. There are two main presentations: anterior knee pain, and abnormal patellar movements, but both may be present at the same time. The disorders presenting as anterior knee pain are listed in Table 1. Abnormalities of patellar movements range from permanent dislocation through subluxation to a tight lateral retinaculum.

Anatomy The patellofemoral joint includes the entire extensor mechanism of the knee, i.e. the quadriceps tendon, patella Tel.: +44 1603286286; fax: +44 1603287498.

E-mail address: [email protected]. 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.02.016

and patellar ligament. The trochlear groove and an arch of articular cartilage around the intercondylar notch make up the femoral side of the joint. It is important to remember that the tibial articular surface comes into contact with a different part of the femur to the patella, except at the tibial spines. It is also important to remember that intercondylar notch osteophytes usually arise from patellofemoral disease. The movements of the patellofemoral joint are complex.1 In full extension only the distal part of the patellar articular surface is in contact with the femoral groove. As flexion proceeds, the contact area on the patella sweeps proximally until at 901 flexion the proximal part is in contact with the distal groove. From 901 flexion the odd facet (the most medial) articulates with the lateral edge of the medial femoral condyle, and the lateral facet articulates with the medial edge of the lateral femoral condyle. The medial facet lies in contact with the synovium overlying the anterior cruciate ligament. There are synovial folds that fill in any space between the non-contacting articular surfaces of the patella and femur.2

ARTICLE IN PRESS 104

Table 1

S. Donell

The causes of anterior knee pain.

Adolescent anterior knee pain (Painful patella syndrome) Patellar malalignment syndromes including dislocation Patellofemoral pathology  Patellofemoral arthritis  Infections  Tumours  Complex regional pain syndromes Extensor mechanism pathology  Patellar tendonitis (Sindig–Larsen, Osgood–Schlatter’s)  Quadriceps tendonitis Tibiofemoral pathology  Medial meniscal tear  Lateral meniscal tear  Osteochondritis dissecans Hip pathology  Osteoid osteoma  Osteosarcoma/Ewing’s Postural abnormalities  Tight quadriceps/hamstrings  Hyperlordotic lumbar spine  Pronated feet Psychological problems Highlighted conditions covered in article.

In full extension the patella rests on the synovium of the supracondylar fat pad. This has a leading edge that moves 2–3 mm distally in the first 201 of knee flexion. As the contact area on the patella moves proximally, progressively more of the supracondylar fat pad lies in contact with the quadriceps tendon. In the inferior part of the patellofemoral joint the synovial folds are more complex. As the knee flexes, the inferior articular surface of the patella becomes more progressively covered by the alar folds of the infrapatellar fat pad. Initially the alar folds face inwards, but beyond 901 flexion the alar folds move away like opening curtains to face away from each other (the way they are typically depicted in anatomy textbooks). The movement of the synovial folds sweeps the articular cartilage and may be important for joint lubrication and nutrition.

Clinical features Pain is typically felt at the front of the knee around the patella. It is worse with exercise, on squatting or getting out of a chair, and on going upstairs. Increased pain on going downstairs is more typical of medial tibiofemoral pathology (medial meniscal tear or arthritis). Patients may complain of the knee locking, despite there being no mechanical block to extension. This pseudo-locking occurs when the patellar and trochlear lesions come into contact causing pain which inhibits knee movement. Feelings of instability and giving way are also typical, either because the quadriceps is weak from pain, or due to extensor mechanism malalignment.

Table 2 The signs that should be noted in a patient with patellofemoral dysfunction. Lower limb alignment including hip rotation and tibial torsion Quadriceps bulk, and the presence of the vastus medialis obliquus (VMO) The presence of an effusion The patellar tilt, and excursion, and tightness of the lateral retinaculum Patellar apprehension and areas of tenderness The range of knee movement and the presence of crepitus The tracking of the patella The tibiofemoral joint The hypermobility index

Clinically if the patient can squat easily, and duck waddle (move about in the squatting position), there is unlikely to be serious pathology within the knee. The clinical signs that should be noted are listed in Table 2.

Patellofemoral syndromes Adolescent anterior knee pain Adolescent anterior knee pain (AAKP) has many synonyms. It is common; in a case series of 446 children, 30% both boys and girls experienced it. Of these only 10% of the boys, and 30% of girls presented to doctors. There was no association with any anthropometric measure (although benign joint laxity syndrome was not recognised at that time), but it was associated with sporting activity.3 In a mean of 16 years follow-up of girls 71% were symptomatically improved, 88% took no painkillers, 90% played regular sports, and 25% had intermittent significant symptoms after 20 years.4 Chondromalacia patellae is a commonly used synonym. However, it is an arthroscopic diagnosis and can be inferred from an MRI scan. It should only be used when the macroscopic state of the articular cartilage of the patella is softened. However, it can be present secondary to chronic trauma, notably in trochlear dysplasia (see below), which is not AAKP. My view is that AAKP is result of chronic overload of the extensor mechanism in the growing sporty child. It occurs because the femur and tibia are growing faster than the soft tissues of the extensor mechanism, and therefore causes a relative overload during sports. The typical age of onset is around 12 years old. The pain is worse with exercise, and eases with rest. It usually resolves over 4–5 years. It is not a disease, and should not be treated operatively, but it is important to exclude other causes of anterior knee pain, most notably extensor mechanism malalignment with trochlear dysplasia.

Patellar malalignment syndromes The literature on the management of patellar instability is confusing. There is no standardisation of diagnoses, e.g.

ARTICLE IN PRESS Patellofemoral dysfunction

105

patients with patella alta are often lumped with patients with lateral subluxation in extension. Many reports are of the results of a single operative technique used for a variety of conditions. Consequently operations for patellar instability have a relatively poor record. What is clear is that while the patella is stabilised, operated patients have a higher rate of osteoarthritis than non-operated patients.

Classification of patellar instability The traditional classification of instability includes:

    

congenital traumatic habitual obligatory subluxation and dislocation.

Figure 1

Lateral patellofemoral angle on skyline X-ray.

Figure 2

Sulcus and congruence angles on skyline X-ray.

Dejour5 described a system based on clinical symptoms and on the radiological abnormalities found (Table 3). The radiological abnormalities are important to note. If they are present in a patient with anterior knee pain, then the cause is an extensor mechanism malalignment, and not classical AAKP. The latter may have a surgical option in the management, the former does not.

Imaging the extensor mechanism Although the patella has three degrees of freedom, in practice the important malalignments occur in the sagittal and frontal planes, and longitudinal axis. These are assessed on a lateral X-ray, a CT scan, and a tangential patella (or skyline) X-ray, respectively. These images give a static twodimensional analysis of the problem. Three-dimensional ‘moving’ images using MRI are possible, although this is not routinely available. Routine plain X-rays should include a strict lateral X-ray with the posterior condyles of the femur overlapping precisely. A skyline view (tangential patella) is essential. A sulcus angle of greater than 1401 is strongly associated with other abnormalities that lead to patellar dislocation. It may also show a medial ‘‘ossicle’’ which is pathognomic of a previous dislocation. See Figs. 1 and 2 for the typically measured angles on skyline views, although these are not for deciding on operative management as they change with knee flexion. Standardising images for knee flexion is very difficult within a department. Table 3 Dejour’s classification of extensor mechanism malalignment. Major patellar instability Objective patellar instability Potential patellar instability

More than one documented dislocation One dislocation with associated anatomical abnormalities Patellar pain with associated radiological abnormalities

Figure 3 The patellar height: Caton–Deschamp method on lateral plain X-ray. Arrow shows posterior condyles strictly overlapping.

On the lateral X-ray the patella height is measured. My preferred method is the Caton–Deschamps modification of the Blackburne–Peel method (TA/PA on the X-ray: 1.0 range 0.8–1.2 is normal) (see Fig. 3).5 Trochlear dysplasia and

ARTICLE IN PRESS 106

S. Donell

trochlear boss height can also be measured from the plain lateral X-ray (Figs. 4–6).6 CT scanning, made up of a series of transverse cuts with the knee in extension, allows the patellar tilt angle to be measured using the posterior condyles as the reference line

and the long axis of the patella as the measurement line. Normal is less than 201 (Fig. 7). Also on CT scan the tibial tubercle–trochlear groove (TTTG) distance can be measured. This measures the offset of the tibial tubercle relative to the true trochlear groove and is more accurate than the ‘Q’-angle. The normal is 10 mm with greater than 20 mm as abnormal, allowing for errors in the measurement (Fig. 7).

Surgical procedures for patellar instability Most patients with an acute patellar dislocation settle after a period of rest, and physiotherapy. The 10% of patients who have failed conservative therapy, and have significant disability from an unstable extensor mechanism can be considered for surgical treatment. However, the evidence base for operative intervention is very poor. Traditionally two types of procedure were described: proximal realignment and distal realignment (see Table 4). Proximal operations are soft tissue procedures intended to make the patella track more medially. Distal operations are related to the patellar ligament insertion and tibial tubercle. More recently operations on the groove, trochleoplasty, have come to the fore. Many surgeons take a traditional approach and perform a standard operation for all abnormalities, e.g. lateral release and medial reefing proximally, and a Roux–Goldthwaite distally. There is no doubt that this will stop dislocation in most cases, but may precipitate anterior knee pain (since any trochlear dysplasia has not been addressed), and not have a satisfactory functional outcome.

Proximal operations

Figure 4 The normal groove on: (a) plain lateral X-ray; (b) skyline X-ray.

Figure 5

Lateral release Lateral release may be open or closed (arthroscopic). Traditionally lateral release has been performed on all patients with instability, often as the only operation. The current view is that this is usually not necessary if the medial side is tightened with a medial patellofemoral

The Dejour classification of trochlear dysplasia on plain lateral X-ray.

ARTICLE IN PRESS Patellofemoral dysfunction

107

Figure 6 The Dejour classification of trochlear dysplasia on CT scan: (A) Shallow trochlear. (B) Flat trochlear. (C) Asymmetry of the trochlear facets: hyperplastic lateral, hypoplastic medial condyles. (D) Asymmetry of the trochlear facets: vertical join between the facets (cliff pattern).

ligament (MPFL) construction (see below) and may increase the risk of medial subluxation. Concerns about increasing the joint reaction force are overcome if a deepening trochleoplasty is included. There is also evidence that a lateral release significantly increases the loads on the patellar ligament.

described that tighten the medial side by a double-breasted reefing. This does not correct maltracking in extension, and itself stretches with time. The vastus medialis obliquus muscle is usually absent in these patients. Efforts to move the existing vastus medialis distally to compensate for this, fail because the muscle becomes functionless.

Medial reefing Medial reefing is used to correct excessive patellar tilt that develops as the medial structures stretch as part of the evolution of instability. Various techniques have been

Repair of the medial patellofemoral ligament The evidence base for the approach to managing an acute dislocation is sparse. Immediate repair of an acute rupture or avulsion of the insertion of the medial patellofemoral

ARTICLE IN PRESS 108

S. Donell Construction of the medial patellofemoral ligament Recently a number of operations have been devised that act as a check rein between the patella and medial femoral condyle, constructing the deficient MPFL.7 The isometric points are in the superior half of the medial border of the patella and just proximal to the medial epicondyle of the femur where a ridge is palpable.8 I use a semitendinosus free graft, passing it through a tunnel in the patella and fixing it to the femur with an absorbable screw as on the femoral side of an ACL reconstruction (see Fig. 8). This stops lateral tracking in extension, and guides the patella into the femoral groove during flexion. I rarely combine this with a lateral release, and now rarely perform a distal realignment since the patella tracks in the groove. Semitendinosus tendon is ten times stronger than the normal MPFL and so can overcome any resulting lateral displacement force. If combined with a trochleoplasty (see below), then the groove is moved more laterally. This decreases the TTTG and also means that tibial tubercle osteotomy is unnecessary. Dynamic semitendinosus transfer This procedure may be necessary in pre-physeal closure or severe adult dislocation. It aims to create a dynamic force somewhat akin to the effect of the vastus medialis obliquus. I use it in children as a temporising measure until they have stopped growing. They then undergo a definitive procedure depending on the underlying abnormalities.

Distal operations Figure 7 The patellar tilt angle and tibial tubercle–trochlear groove distance on CT scan. Table 4 Proximal

Distal

Operations for patellar instability. Lateral release Medial reefing Medial patellofemoral ligament Dynamic semitendinosus transfer Roux–Goldthwaite Tibial tubercle osteotomy

Repair Reconstruction

Roux–Goldthwaite This is a very popular procedure where the lateral half of the insertion of the patellar ligament is passed under the medial half and inserted into the medial plateau. Its effect is therefore medialisation of the insertion of the extensor mechanism. I do not favour using it for three reasons. There is no imaging technique that can measure its effect. Intraoperatively it is difficult to know precisely where to place it and how tightly it should be fixed. Finally, it usually requires immobilisation. In principle joints should not be splinted. It is however a popular operation used by many surgeons.

Elmslie (medialisation) Fulkerson (anteromedial) Maquet (anterior) Distalisation

Trochleoplasty Dejour Bereiter Albee

ligament as occurs with patella dislocation, has been advocated. My view is that most acute dislocations settle with an orthosis and early physiotherapy, without the need for an acute repair.

Figure 8 Lateral post-operative X-ray of a medial patellofemoral ligament construction; white arrow—patella tunnel, black arrow—femoral pit.

ARTICLE IN PRESS Patellofemoral dysfunction

Figure 9 Intra-operative photograph of medialisation of the tibial tubercle (Elmslie). The tubercle is held in the new position by an awl.

109 Tibial tubercle osteotomy Tibial tubercle osteotomy can be imaged and measured. Intra-operative displacement can be predicted, and immediate mobilisation of the knee allowed. Moving the tibial tubercle clearly affects the action of the extensor mechanism. The criticism for any tibial tubercle osteotomy is that it risks non-union, and stress fractures of the tibia. This is certainly true for distalisation (aimed at correcting patella alta). This is unusual in the Elmslie (see Fig. 9) and Fulkerson, which are based on a distal osteo-periosteal flap. The latter brings the extensor mechanism forwards, and therefore reduces the patellofemoral joint reaction force, which is useful if there is a mild trochlear dysplasia. The Maquet has the same effect, but was designed for anterior knee pain. It is not in vogue now, mainly because of its cosmetic effect. The Hauser procedure (not listed in Table 4), repositioned the tubercle posteromedially, but is well documented as leading to early patellofemoral osteoarthritis.

Trochleoplasty The earliest trochleoplasty was described by Albee.9 This was an elevating osteotomy of the lateral trochlear

Figure 10 Before and after the Dejour trochleoplasty: (a) per-operative photograph; (b) CT scans.

ARTICLE IN PRESS 110

S. Donell

facet. In the presence of the hyperplastic type of dysplasia (by far the commonest) this raises the joint reaction force and precipitates patellofemoral osteoarthritis. Despite this, it is useful in a rare condition of lateral condylar hypoplasia, although MPFL construction is my preferred management. There are currently two types of deepening trochleoplasty. Both remove subchondral bone approached from above. The Dejour10 (see Fig. 10) has a thicker resulting osteochondral flap compared to the Bereiter11 which creates a thin flexible flap. The Dejour uses screws or staples to anchor the flaps. The Bereiter uses a vicryl tape to create the new groove. The latter is only suitable with intact articular cartilage. It is an elegant operation suitable for the milder dysplasias. These operations are still in their infancy, and it is not yet known whether that truly improves the functional outcome or reduces the risk of later osteoarthritis.

county level but are struggling because of the pain. Sometimes the parents are basking in the reflected glory of the child, but the latter does not want to continue the sport! There is a small group of patients where the presentation is secondary to severe abuse (physical or sexual). The typical presentation is a girl with bilateral knee pain where both parents attend the clinic. The history is that the pain is severe, they are unable to run, and are taking time off school. On examination there is diffuse pain to light touch. All routine imaging is normal. If they are put on a waiting list (nowadays for an MRI scan), within a few days the family doctor rings asking that this is performed urgently as the child has deteriorated to being in a wheelchair and unable to go to school. Involvement of a paediatrician is mandatory. A psychiatrist will be needed to treat the child if suspicions are confirmed.

Summary of surgical options The advent of construction of the MPFL construction means that, for those surgeons who have taken it up, it is the most common operation that they do. Its strength and mechanism of action means that all other proximal and distal operations appear to be unnecessary for the vast majority of patients. The necessity for lateral release is contentious. Trochleoplasty is even more contentious, even though logical. Until the results of robust clinical trials are available to inform surgeons as to best practice, the majority of general orthopaedic surgeons (particularly in the USA) are likely to continue with lateral release, medial reefing, and some with Roux–Goldthwaite, Elmslie or Fulkerson.

Conclusion

Postural abnormalities

References

Anterior knee pain may be the result of postural abnormalities, most commonly rotational abnormalities of the lower limb either persistent femoral anteversion or external tibial torsion. Tightness of the lower limb muscles (quadriceps, hamstrings or gastrocnemius) may present with anterior knee pain. Additionally, there are two postural patterns that may be seen, especially in adolescents. Firstly, the hyperlordotic lumbar spine with squinting patellae where the symptoms are bilateral. Secondly, standing in the pelvic slouch position, with weight-bearing on one leg. The bent resting knee is the symptomatic one. The weight-bearing leg is being loaded through the iliotibial band. It may be that the iliotibial band is tight in these cases. This area needs further research.

1. Goodfellow J, Hungerford DS, Zindel M. Patello-femoral joint mechanics and pathology. Functional anatomy of the patellofemoral joint. J Bone Joint Surg [Br] 1976;58B:287–90. 2. Donell ST. The synovial folds of the patellofemoral joint: a dynamic study. Clin Anat 1992;5:107–12. 3. Fairbank JCT, Pynsent PB, Poortvilet JA, Phillips H. Mechanical factors in the Incidence of knee pain in adolescents and young adults. J Bone Joint Surg [Br] 1984;66-B:685–93. 4. Nimon G, Murray D, Sandow M, Goodfellow J. Natural history of anterior knee pain: a 14- to 20-year follow-up of nonoperative management. J Pediatr Orthop 1998;18:118–22. 5. Dejour H, Neyret P, Walch G. Factors in patellar instability. In: Aichroth PM, Dilworth Cannon W, editors. Knee surgery current practice. Martin Dunitz Ltd; 1992. p. 403–12. 6. Dejour H, Walch G, Neyret Ph, Adeleine P. Dysplasia of the femoral trochlea. Rev Chir Orthop 1990;76:45–54. 7. Cossey AJ, Paterson R. A new technique for reconstructing the medial patellofemoral ligament. Knee 2005;12:93–8. 8. Amis AA, Firer P, Mountney J, Senavongse W, Thomas NP. Anatomy and biomechanics of the medial patellofemoral ligament. Knee 2003;10:215–20. 9. Albee FH. The bone graft wedge in the treatment for habitual dislocation of the patella. Med Rec 1915;88:257–9. 10. Donell ST, Joseph G, Hing CB, Marshall TJ. Modified Dejour trochleoplasty for severe dysplasia: operative technique and early clinical results. Knee, in press. 11. Bereiter H, Gautier E. The trochleoplasty as a surgical therapy of recurrent dislocation of the patella in dysplastic trochlea of the femur. Arthroskopie 1994;7:281–6.

Psychological problems Finally, it is important to realise that many patients present with anterior knee pain (AKP) secondary to psychological problems. In children they fit in with a pattern of headaches and abdominal pain. Nowadays operations are avoided as MRI scans confirm that the knee is normal (having checked the hips and spine). Many adolescents present with AKP who are athletic. Beware the parents who say that their child is at national or

Patellofemoral dysfunction is common and has many causes. It presents as anterior knee pain, or patellar instability. Surgery aims to correct extensor mechanism malalignment in patients if conservative therapy fails. This requires proper imaging to define the abnormalities, and operations targeted at correcting them. The two new operations that are being popularised for this are MPFL construction and trochleoplasty. It has yet to be confirmed that these have improved functional outcomes over traditional operations.

ARTICLE IN PRESS Patellofemoral dysfunction

Further reading 1. Aichroth PM, Duri ZA. Dislocation and subluxation of the patella an overview. In: Aichroth PM, Dilworth Cannon W, editors. Knee surgery current practice. Martin Dunitz Ltd; 1992. p. 354–79. 2. Dandy DJ. Instructional Course Lecture on ‘Chronic Patellofemoral Instability’. J Bone Joint Surg [Br] 1996;78-B:328–35.

111 3. Remy F, Chantelot C, Fontaine C, Demondion X, Migaud H, Gougeon F. Inter and intraobserver reproducibility in radiographic diagnosis and classification of femoral trochlear dysplasia. Surgical and radiological anatomy. J Clin Anat 1998;20: 285–9. 4. Smirk C, Morris H. The anatomy and reconstruction of the medial patellofemoral ligament. Knee 2003;10:221–7.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 112–120

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: SOFT TISSUE KNEE PROBLEMS

(v) Osteotomy in the management of knee osteoarthritis and of ligamentous instability Andy Williamsa,, Natasa Devicb a

Department of Orthopaedic Surgery, Chelsea and Westminster Hospital, 369 Fulham Road, London SW10 9TR, UK Department of Materials, Imperial College London, South Kensington Campus, Exhibition Road, London SW7 2AZ, UK

b

KEYWORDS High tibial osteotomy; Anterior cruciate ligament; Medial compartment osteoarthritis; Varus mal-alignment

Summary Traditionally osteotomy has been used for redistribution of articular surface load in osteoarthritic knees. It was becoming a ‘lost art’ in orthopaedics due to the success of arthroplasty, and the perceived unpredictable outcomes of osteotomy. With newer techniques and improved predictability in results, osteotomy is rightly regaining popularity. In osteoarthritis the aim is to deliberately create deformity to unload diseased joint surface and load healthy articular surface. In ligament instability osteotomy is employed so that alignment is fine-tuned in favour of stability. A truly three-dimensional consideration of knee alignment is needed. Coronal alignment affects not only relative loading of medial and lateral tibio-femoral compartments, but stress on the collateral ligament complexes. The degree of tibial slope affects loading of the cruciate ligaments. Applying this knowledge allows enhanced results from complex ligament reconstruction of the knee. & 2006 Elsevier Ltd. All rights reserved.

Introduction Given the centuries over which anatomy has been studied and the slow pace of evolution, the knowledge of lower limb alignment should be certain. Sadly this is not the case. Nevertheless certain assumptions need to be made. In an average adult for the purposes of planning osteotomy it is assumed that the weight-bearing line (WBL) joins the centres of the hip, knee, and ankle of the lower limb Corresponding author. Tel./fax: +44 208 429 8411.

E-mail addresses: [email protected] (A. Williams), [email protected] (N. Devic). 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.04.002

concerned. Unfortunately at the knee it almost certainly passes through the medial compartment. In the coronal plane both tibial surfaces are slightly concave which is accentuated by the menisci. In the sagittal plane medially the posterior half of the tibial surface is flat and the anterior half slopes down from anterior to posterior. This overall incline accounts for the term ‘posterior tibial slope’ that is in common usage. Laterally the anterior and posterior tibia slopes down away from a central flat surface. Therefore the ‘posterior tibial slope’ that is referred to is really only relevant to the medial tibia. Since the lateral compartment effectively produces loading of two convex surfaces (in the sagittal plane) and medial loading (especially when one considers the role of

ARTICLE IN PRESS Osteotomy in knee osteoarthritis

113

Figure 1 (a–c) Bilateral varus is seen. It is well tolerated in the ligament intact knee but has led to progressive varus in the knee deficient in both cruciate ligaments and lateral ligament complex. The patient had a varus thrust on stance during walking. The knee was treated by CWHTO plus reconstructions of ACL and PCL and the posterolateral structures. Note division of the fibular neck to prevent de-tensioning of posterolateral corner structures.

the relatively fixed posterior medial meniscus) is between a convex femoral condyle and concavity, it is appropriate that the WBL passes medially where there is more inherent stability. It is generally accepted that the axes of the shafts of tibia and femur intersect at an average 5–71 (i.e. the ‘tibiofemoral angle’). The WBL passes 31 to the vertical in bipedal stance (i.e. ankles closer together than hips). To provide a horizontal joint line in this posture the joint line (and tibial surface) is in 31 of varus to the vertical/longitudinal axis of the tibia. Of course in many individuals, especially athletes, the alignment favours running where there is more varus in the proximal tibia. In the ligament-intact limb this overall varus does not lead to any dynamic deformity (i.e. lateral/varus thrust) during the stance phase of gait because the intact proprioceptive feedback in the limb allows dynamic muscular control of deforming forces. However, when cruciate ligaments are torn, as well as losing their passive restraining effect, their proprioceptive function is lost (Figs. 1a–c). Sometimes this is enough to allow dynamic instability (i.e. a thrust) to occur even with an initially intact lateral ligament complex whilst walking. In time, chronic overload of the lateral ligament complex causes it to stretch out and a dynamic thrust will appear (see below). Overall sagittal alignment has not received as much study as the coronal. Nevertheless, when standing it is likely that

the WBL passes slightly anterior to the mid-point of the knee so tending to maintain hyperextension. This allows prolonged stance without the need for quadriceps activity, since knee extension is maintained passively and balanced by tension in the posterior structures including the capsule. The tensor fascia lata tightens the ilio-tibial band like a rein just to maintain slight hyperextension when needed.

Modern osteotomy Osteotomy for osteoarthritis (OA) of the knee has, until recently, been in decline. This is especially true for North America and Britain where increasingly arthroplasty has been preferred in treating OA even in the younger patient. Many orthopaedic trainees in the UK, until the recent resurgence of osteotomy, were completing training having never seen a tibial nor femoral osteotomy. ‘Old’ osteotomy was often undertaken too late in the disease process, carried out badly, and employed poor fixation methods therefore necessitating reliance on casting with its associated problems. It is perhaps not surprising that the results were often disappointing. More recently devices have been developed making osteotomy a more simple and reliable operation. The fixation devices have improved greatly meaning that with

ARTICLE IN PRESS 114

A. Williams, N. Devic

certain devices, even for opening wedge techniques, weight-bearing, albeit not full, is possible immediately after surgery. Furthermore, some systems allow percutaneous techniques. As a result of these innovations the usage of osteotomy, including closing wedge techniques, has increased dramatically. There are pros and cons for opening and closing wedge techniques which should mean that the most appropriate for a given situation is carried out having considered the options. Unfortunately surgeons tend to employ either one or the other in all cases. Other advances have been in understanding which patients are likely to benefit most, how much to realign the limb, and in which plane. The technical advances have greatly improved the rehabilitation that can be undertaken in the early post-operative period which has reduced the problems of stiffness, weakness, and infrapatellar contracture that were common after older techniques.

Preoperative evaluation The following help identify suitable patients with OA. They are not absolute guidelines:

 chondral damage should be localized either to medial or lateral compartments;

 minimum range of movement of 15–1001;  maximum deformity of 15–201;  inactive, obese, and older patients are unlikely to be suitable. Preoperative gait analysis has also been used with some success. For cases with medial OA, a low adductor moment at the knee correlates with better results. The adductor moment reflects the medial displacement of the WBL from the centre of the knee during the stance phase of gait. The larger it is the more medial compartment loading and tendency to varus. External rotation of the foot (‘toe out’) reduces it and may be an adaptive phenomenon. Having a low adductor moment prior to surgery correlates to better outcome after osteotomy.1 Some surgeons use a period of bracing prior to operation. If the brace helps symptoms then it is suggested that osteotomy is more likely to help. Unfortunately the situation is rarely as clear-cut as this. Assessment of lower limb mal-alignment should be carried out with the aid of long leg weight-bearing films. These have limitations especially if the limbs are rotated allowing the femoral bow or tibial bow to influence ‘apparent alignment’. Fixed flexion or gross hyperextension at the knee too can be problematic. Whilst CT scanning is attractive for accuracy of measurements the findings are irrelevant to dynamic loading as they are non-weight-bearing. The lateral radiographs will aid the assessment of the posterior tibial slope. Standard ‘standing’ AP films will allow the assessment of medial/lateral pathological joint space narrowing and opening. It is very important to differentiate the contribution to standing deformity from bone alignment, joint space narrowing/loss, i.e. chondral damage, and joint opening from ligament laxity. The most common scenario is the varus knee. The alignment seen on standing X-ray results from one of, or a combination of, proximal tibial varus, medial joint space loss, and lateral ligament laxity. It is the former two

Figure 2 This example shows the contribution to varus of insufficiency of the lateral ligament complex. This lateral opening of the joint must be taken into account when calculating the magnitude of correction of a planned osteotomy.

factors which determine the degree of correction at the osteotomy. If the amount of varus secondary to lateral ligament laxity is included then the calculation of correction at the proposed osteotomy can be grossly excessive. As soon as the WBL is shifted into the lateral compartment the lateral ligament will be de-tensioned and its contribution to varus deformity disappears. Therefore on preoperative Xrays the excess lateral joint opening must be taken into account (Fig. 2). A good way of preventing excessive correction in this situation is to apply the following. Each millimetre of lateral tibiofemoral joint opening causes approximately an additional 11 of varus deformity and this needs to be taken into account in order to avoid valgus overcorrection.2 Preoperatively, the use of an image intensifier aids the correct placement of the osteotomy in both the coronal and sagittal axis. The ability to correct the WBL at the time of surgery is of the utmost importance as this is under the control of the surgeon. Although the image intensifier helps this, its small field of view and problems with controlling limb rotation restrict accuracy. This accuracy of coronal alignment is helped by placing a skin marker overlying the femoral head under Xray control which can later be palpated through the drapes during surgery. Using this and applying a metal rod or diathermy lead from this point to the palpated centre of the ankle approximates to the WBL. An X-ray of the knee will then show where the WBL, as indicated by the rod or diathermy

ARTICLE IN PRESS Osteotomy in knee osteoarthritis lead, passes through the knee joint. The axial rotation of the limb needs to be carefully controlled and is problematic during this assessment. Of course, intraoperatively the situation is not truly weight-bearing even if the surgeon applies axial load up through the foot. Therefore the preoperative plan remains key. For varus knees the osteotomy is undertaken at the tibia since it is where the pathology lies, i.e. chondral loss or proximal tibia varum. Conversely in the valgus knee the problem is in the femur. Correction at the correct bone is essential to maintain a joint-line that is horizontal.

Techniques Closing wedge high tibial osteotomy (CWHTO) This is the traditional procedure with modification of the technique used by Coventry.3 The osteotomy passes through the metaphyseal cancellous bone above the tibial tuberosity. It became clear that osteotomies below the tibia tuberosity have an unacceptable non-union rate. The advantages of the technique over opening wedge osteotomy are provision of bone to bone contact with excellent union rates, and the potential for full early weight-bearing. The disadvantages are creation of an abnormal shape to the bone with the tibial shaft ending up medial to the centre of the metaphysis (which has implications for future arthroplasty), loss of height, albeit offset by the lengthening effect of correction of deformity, risk to the common peroneal nerve, and the need to disrupt either the fibula or proximal tibio-fibular joint to allow the tibial osteotomy to close. Furthermore, the operation is more demanding to surgeons who undertake only occasional osteotomy. Opening wedge osteotomy advocates often suggest that the difficulty is more problematic than it is. With good training to learn the procedure—especially safe exposure of the anterior and posterior surfaces of the proximal tibia, and the ability to cut two flat surfaces, it is possible to perform this procedure as swiftly as its opening wedge counterpart despite reports to the contrary.

Opening wedge high tibial osteotomy (OWHTO) The healing potential of the proximal tibial metaphysis is again harnessed. The achievement of adequate access often requires significant elevation of the medial soft tissues including superficial medial collateral ligament (MCL). These tissues heal well and since the deep MCL is intact medial instability is rare. In fact with opening of the osteotomy the medial tissues are put on tension. Once the medial tibia is exposed the osteotomy is cut superiorly and lateral starting about the level of the tibial tuberosity and heading towards a point just below the level of the superior tibio-fibular joint. A major advantage of this technique is that neither the fibula nor superior tibio-fibular joint are violated. The common peroneal nerve should thus be at less risk. However, care must be taken when drilling guidewires across the bone, before undertaking the osteotomy, to avoid drilling too far especially in soft bone. Once cut the osteotomy

115 is opened whilst ‘hinging’ an intact lateral tibial cortex, which can be weakened, if needs be, by use of osteotomes. The amount of desired correction can be made according to the preoperative plan and guided by X-ray and wedged blocks of known angle. It is said that the control of the correction is easier than with CWHTO. Adjustments are certainly easier but with the small field of view from an image intensifier the surgeon is wise to keep to the preoperative plan. A reason for the current popularity of this procedure is development of decent implants to hold the osteotomy. Some even allow early restricted weight bearing. However, the period of protected weight bearing is much longer than with CWHTO. For all but minor corrections the ostotomy gap is best filled with bone graft from the iliac crest and/or bone substitute. The need for bone graft is a distinct disadvantage as compared to CWHTO. An inevitable consequence of OWHTO where the whole osteotomy passes above the patellar tendon insertion is a lowering of the patella relative to the knee joint-line. By modifying the technique this can be avoided by making an oblique extension inferiorly from the main osteotomy emerging anteriorly inferior to the tibial tuberosity. Patella baja was a frequent association of ‘old’ CWHTO and one of the reasons for difficulty in exposure at subsequent knee replacement. The actual technique causes a relative elevation of the patella and so this phenomenon was simply related to the use of plaster casting and very limited quadriceps work. Now casts and even braces are rarely needed allowing early restoration of motion and strengthening. The tibial slope is of great importance regarding antero-posterior stability (see below). Osteotomy to deliberately alter the tibial slope alone is required infrequently, but HTO for coronal plane realignment frequently alters the tibial slope without intention. Because OWHTO requires elevation of the medial soft tissue from anterior to posterior there is a tendency to have to place the fixation device from a more anterior position. This and importantly the fact that the cross-section of the tibia is triangular means that medial OWHTO produces more anterior opening than posterior even during what is supposed to be purely coronal correction—hence increasing the tibial slope.4 The tendency with CWHTO is the opposite, i.e. to decrease the slope by an average 51.18 This has implications for anteroposterior stability (see below).

Other techniques of HTO External fixation has been used with considerable success in obtaining correction of deformity. The technique employing the ortho-fix is established. For complex problems the Ilizarov family of frames are used. The senior author has no experience with these techniques as modern osteotomy allows acute correction with stable ‘buried’ fixation and no restriction of joint motion. When concomitant ligament reconstructions are undertaken these techniques are inappropriate. In addition there is a significant risk of infection with longer term implications. The ‘Dome’ HTO: via a direct anterior approach the tibia is transected from anterior to posterior to produce an arcuate osteotomy or dome. The fibula is divided through a second

ARTICLE IN PRESS 116 incision. The distal tibia can then be rotated relative to the proximal at the ‘dome’ osteotomy site. This has the attraction of preserving the shape of the proximal tibia whilst maintaining bone to bone contact throughout. However, in the senior author’s experience the complication rate for neurovascular injury and non-union seems higher than other techniques.

Distal femoral osteotomy (DFO) The need for this is much less common since primary OA of the lateral compartment of the knee is only 5% of the total. Furthermore, most patients with significant medial ligament laxity do not have constitutional valgus thereby making combined DFO and MCL reconstruction an uncommon event. Traditionally a medially based closing wedge osteotomy held with a blade plate device was used. New devices especially locking plates make laterally based opening wedge DFO attractive. As with HTO there are pros and cons for OW and CW techniques and the good surgeon will chose the most appropriate for the patient concerned.

Correction of the WBL In the treatment of OA the WBL is deliberately moved significantly to load the healthy compartment. Coventry’s work3 suggested that, in his series of HTOs for medial OA, undercorrection was associated with a poor outcome. He aimed for 3–51 more tibio-femoral valgus than normal (a tibiofemoral angle of around 101 and therefore a fairly modest increase). This has been confirmed by others.5,6 Unfortunately, the message that undercorrection leads to failure has been remembered more than the fact that modest corrections yield good results. There is a tendency for surgeons to emphasize the concept that overcorrection is a good thing. Excessive correction can lead to a feeling of instability with the knee which is usually perched on one compartment joint surface and, on occasion, a lurching of the joint contact into the other diseased compartment, as well as leading to relatively rapid overload and failure of the healthy compartment joint surfaces. In addition cosmetically the appearance is poor and a source of medico-legal claims. A reasonable guide is that the WBL should pass though the lateral edge of the base of the lateral tibial spine. Dugdale et al.2 advocate placing the WBL through the ‘62% coordinate’, i.e. 62% of the way along a line joining the medial edge of the tibial plateau to the lateral limit. This means that there has been a trend to less aggressive correction at the osteotomy than was hitherto practised. In cases where there is no OA and an osteotomy is purely used to produce a limb alignment to favour knee stability overcorrection is inappropriate and will lead to premature chondral failure from overload. The WBL in the coronal plane should be neutral so that there is no excess tension in any ligament reconstruction. The usual situation is in injury of the lateral ligament complex (i.e. ‘posterolateral corner’) in a knee with proximal tibia varum. Even if a sound posterolateral corner reconstruction is undertaken due to the lower limb alignment in the coronal plane there will be tension in the reconstruction when the knee is loaded. In the uninjured knee the varus is tolerated due to

A. Williams, N. Devic intact proprioception. In the ligament injured knee proprioception is lost as well as the increase in laxity. As a result a chronic overload of the reconstruction will lead to a stretching out of the tissues and failure. The calculation of the desired angle of correction can be done in a number of ways. The simplest is to mark on the measurement X-rays the point on the knee joint-line where the WBL is desired to pass. The angle subtended by lines passing from this point to the centres of the femoral head and ankle will bring the WBL to pass through the desired point at the knee. Many surgeons undertake correction at the osteotomy applying the notion that 1 mm at the base of the wedge equates to 11. Whilst this is a reasonable approximation it is more accurate to calculate the exact size from preoperative measurement X-rays since size of bones vary from case to case.

Tibial slope is important The normal tibial slope is said to measure between 71 and 131. The issue of the sagittal anatomy of the joint surface of the tibia is detailed above. It really refers to the medial side of the tibia. Since the WBL passes into the medial compartment and the medial femoral condyle is fairly stationary in the antero-posterior direction during knee flexion this overall slope has a profound influence on antero-posterior shear between the femur and tibia. Increasing the tibial slope increases the tendency to anterior tibial translation when the limb is axially loaded (Figs. 3a–f). In contrast, decreasing the tibial slope tends to posterior tibial translation. In cases of tibial slope reduction or reversal profound posterior tibial translation can result.7 By deliberately influencing the tibial slope the osteotomy can be used to favour one or other cruciate ligament. In cases in which there is an abnormal tibial slope even with intact cruciate ligaments the anteroposterior tibial translation can be dramatic, giving the incorrect impression that on or other cruciate ligament needs reconstruction. These patients need osteotomy first and only ligament reconstruction if the tibial translation has caused attritional stretching of the ligament. If coronal plane adjustment is required as well, a biplanar osteotomy is required. Hence for a combined PCL and posterolateral corner disruption in the presence of significant varus an osteotomy producing correction of the varus plus an increase in tibial slope is ideal. This can be achieved by medial plus anterior opening wedge, or lateral and posterior closing wedge osteotomies. The former is the easier to achieve. When reducing the tibial slope CWHTO is usually easier. Rather than always doing an osteotomy employing opening or closing wedge techniques, regardless of patient factors, the operation should be determined by various goals and circumstances. Using the knowledge detailed above the exact type of osteotomy for a patient should be designed for their needs and the ideal technique undertaken.

Clinical scenarios Medial OA with normal ligaments The standard OW or CW HTO is appropriate.

ARTICLE IN PRESS Osteotomy in knee osteoarthritis

117

Figure 3 (a) Gross posterior tibial subluxation in a knee with reversal of the tibial slope in a case of premature anterior tibial physeal arrest. The patient had multiple injuries as an adolescent and spent a long period in an intensive care unit where his knees were hyperextended. In addition there was rupture of the PCL. (b) This is the same case as in (a). The major effect of the tibial slope is apparent when the posterior subluxation is corrected with restoration of the normal angle of slope. (c) The same case as in (a and b). This shows how subluxation in the sagittal plane leads to an apparent but spurious coronal plane deformity. Once the tibial slope deformity was dealt with the coronal alignment was excellent. (d) The same case as in (a–c). A subsequent PCL reconstruction was undertaken. (e and f) Same patient as in (a–d). This shows how subluxation in the saggital plane leads to apparent but spurious coronal plane deformity. Once the tibial slope deformity was dealt with, the coronal alignment was excellent.

Medial OA plus ACL-deficiency CWHTO is theoretically the best choice since, as well as offloading the medial compartment, this osteotomy tends to reduce the tibial slope and so lessens anterior directed

stress on the proximal tibia, i.e. it favours the ACLdeficiency or offloads concomitant ACL reconstruction. The controversy is as to whether or not simultaneous ACL reconstruction should be performed. With OW or CW HTO the ACL reconstruction can be easily undertaken at the same sitting,

ARTICLE IN PRESS 118 but whether or not it is useful is debated. The senior author tends to believe so. There are advocates for both points of view, but those in favour of a combined procedure in carefully selected patients prevail.8,9 The osteotomy is performed first and therefore care has to be taken, if using an arthroscopic technique, that irrigation fluid does not fill the calf due to leakage across the osteotomy via the tibial tunnel.10

Varus plus posterolateral corner insufficiency In acute cases of posterolateral corner disruption repair or reconstruction of all injured structures is possible within 2–3 weeks from injury and seems to provide a better outcome than waiting until the situation is chronic. There are some knees with marked varus who present with an acute multiligament injury including the posterolateral corner disruption. In this scenario the question of whether or not an early ‘protective’ osteotomy should be undertaken arises. It is our philosophy not to undertake osteotomy early for two reasons. Firstly the added insult of an osteotomy would not be well tolerated, in view of the major injury and major ligament surgery. Also, in the acute scenario, the emphasis is on repair of tissues (even cruciate ligaments can be repaired and be expected to heal in such treatment of dislocated knees). The repairs may need to be augmented with grafts, but there is much more potential for maintaining intact proprioception in the repaired tissues. Hence, there is more likelihood of avoiding a dynamic varus thrust. In the chronic situation, secondary problems develop such as stretching out of soft tissue restraints such as the capsule and lateral ligament complex. Therefore, the anatomical abnormalities have to be defined to customize the osteotomy and concomitant procedures to combat all abnormalities present. With this in mind the following concept is useful. Noyes11 was the first to describe the concept of primary, double, and triple varus knees in order to classify anatomical abnormalities of the knees concerned. These terms take into account underlying bony malalignment, abnormal knee movement, subluxations and ligamentous insufficiency. Chronic ACL insufficiency leads to increased anterior translation of the tibia and shearing forces. Initially the kinematic consequence is lateral.12 The medial compartment kinematics are unchanged if the medial meniscus is competent. The meniscus is, however, subject to more load and in time will fail. This leads to medial compartment OA and thence deformity: ‘primary varus’. If the medial meniscus is torn or absent, the chances of medial compartment OA increase even further.13 The narrowing of the medial compartment leads to a medial shift of the WBL which increases strain on the posterolateral structures which may become lax: ‘double varus’. The term alludes to the fact that there are two reasons for the double varus knee, i.e. bone/joint alignment plus lateral ligamentous opening and lateral condyle ‘lift-off’. With time, excessive lateral stresses lead to pathological hyperextension and external rotation as the posterolateral capsule and ligamentous structures stretch: ‘triple varus’. Triple varus can also occur in a varus ACL-deficient knee with an acute injury to the posterolateral complex. Hence, when assessing a varus knee clinically it is important to see if the deformity is bone/joint alone (‘primary varus’), or

A. Williams, N. Devic has associated lateral laxity (‘double varus’), and even hyperextension/external rotation (‘triple varus’).

ACL deficiency and varus malalignment (no OA, i.e. tibia varum) These may, in time, fit into Noyes’ double varus group as the varus, in the presence of reduced proprioception allows stretching of the lateral ligament complex (see below). ACL reconstructions in the presence of uncorrected varus malalignment can lead to graft failure (Fig. 4). In addition, in a varus ACL-deficient knee there are high adduction moments which lead to increased loading of the medial compartment and place stresses on the lateral ligaments.14 A recent study recommended HTO together with ACL reconstruction in cases of ACL deficiency with varus malalignment.19 This combined procedure can stabilize the knee, reduce pain and stop premature OA.15 In patients with ACL deficiency the importance of the tibial slope must not be underestimated and it should not be increased. Decreasing the tibial slope, although of proven benefit, is difficult with an OWHTO. The aim of the surgery in these patients is to avoid varus thrust and to reduce the stresses acting on lateral ligaments on ambulation. The correction provided by the osteotomy is therefore limited to provide a WBL in neutral at the knee. The difficulty is to decide when to undertake an osteotomy as usually a mild proximal tibial varum in the presence of isolated ACL tear usually does well with ACL reconstruction alone. In other words: when is too much varus present? This is not known and therefore considerable reliance on experience based judgement of the surgeon is required. It is correct to say that whenever such a scenario presents the potential for osteotomy should be considered.

Figure 4 This is an example of a patient with an isolated ACL tear reconstructed alone without dealing with the varus alignment years previously. In time the lateral ligament complex stretched‘double varus’. Subsequently the patient dislocated the other knee requiring ACL plus PCL plus posterolateral corner reconstruction and HTO. Whilst the knee with only an isolated ACL tear developed dynamic thrust on walking, the more severely injured knee was well aligned and more stable.

ARTICLE IN PRESS Osteotomy in knee osteoarthritis

Multiligament injury/triple varus These knees have gross instability. If the posterolateral structures were injured acutely with one or both cruciate ligaments at the same time then the knee has often actually dislocated at the time of injury. When the injury has become chronic HTO is needed if there is unfavourable varus. If CWHTO is chosen it has the advantage of allowing access laterally for posterolateral ligament reconstruction. In addition common peroneal nerve injury is not uncommon and neurolysis is often needed. However, it must be remembered that, to prevent proximal migration of the fibular head which will slacken off the posterolateral structures, it is best to divide the fibular shaft rather than disrupt the proximal tibiofibular joint. This can be undertaken via a more distal lateral incision, or, perhaps safer, through the neck of the fibula, if the nerve is dissected off it first. An OWHTO will require a second lateral incision for access to the posterolateral corner of the knee for ligament reconstruction/neurolysis. During the osteotomy the tibial slope can be altered to favour whichever cruciate ligament is reconstructed at the same time. If both cruciate ligaments are reconstructed, as they frequently are, the tibial slope should be left unaltered. Once the bony alignment is optimal the cruciate ligament(s) is/are reconstructed. Usually the osteotomy and cruciate ligament reconstructions can be completed within a safe tourniquet time and the extraarticular posterolateral reconstruction can then be undertaken without tourniquet if necessary. Of course a simultaneous HTO plus reconstruction of the posterolateral corner and one or both cruciate ligaments is demanding and may be lengthy. In experienced hands it is reasonable to aim for a single operation, as the patient avoids another procedure, but it is wise to stage procedures if time demands. The HTO is undertaken first and the ligament reconstructions delayed for a few months. Preoperative physiotherapy can be employed to improve the strength of lower limb musculature. In addition, gait training can be applied to those patients with hyperextension deformity, in order to avoid the patients waking with the hyperextension gait postoperatively,16 which would place a strain on the reconstructed ligaments. This preoperative optimization can lead to a reduction in joint pain and improvement in gait.

119 the knee as a consequence of long periods of traction in childhood such as in the treatment of hip disorders (‘Frame Knee’). In these cases anterior OWHTO is the key to treatment. Unless there has been major disruption of the PCL reconstruction of that ligament is rarely needed.

Injuries including medial collateral ligament (MCL) disruption As an advantage for medial OWHTO it is often said that in cases of MCL deficiency, the distraction offered by an opening wedge osteotomy can tension the lax ligaments. The combination of varus alignment plus MCL insufficiency is rare and in any case the osteotomy will cause valgus alignment so further stressing the suspect MCL! Uncontrolled MCL laxity can be a very difficult problem to deal with. Arthroplasty surgeons appreciate this problem especially. The problem is that the stresses involved in walking tend to exacerbate the situation further. During stance phase when the foot tends to be externally rotated relative to the line of propulsion (i.e. the ‘progression angle’) the knee is subject to a combined valgus and external rotation stress. This tends to stretch the MCL. The whole situation is worsened if the subject’s natural alignment at the knee is valgus and there is significant foot pronation (which produces further valgus and external rotation forces at the knee). This knowledge can be applied for treating MCL injuries. After an acute MCL rupture (usually but not always treated in a brace) for 6 weeks the subject should consciously internally rotate the limb so that during the stance phase the foot points forward without rotation or slight internal rotation. This minimizes the forces described above. Usually chronic MCL laxity can be dealt with by suture plication of the lax by medial ligament complex. In cases where the soft tissues are very attenuated or revision surgery is needed then graft (hamstring autograft or whole patellar tendon allograft with patella and tibial tuberosity) can be fixed into the medial knee. The need for a varus DFO is needed if there is significant bone/joint valgus or the gait is so poor it drives progression of the deformity and further MCL stretching. The latter problem often accompanies a neurologic limb such as that in lower limb cerebral palsy (Figs. 5a and b).

Postoperative rehabilitation PCL deficiency When combining an HTO and PCL reconstruction, it is useful to increase the posterior tibial slope, usually by placing cortical bone graft, or a fixation device (some have metal spacers built in) more anteriorly. This will decrease the load on a newly reconstructed PCL.17 In a case of pre-existing reversed tibial slope, as most commonly occurs with premature anterior physeal arrest, marked posterior sag of the tibia can be present. In addition hyperextension occurs at the knee joint. This is made cosmetically more obvious by extension deformity (precurvatum) just below the joint in the proximal tibia where growth arrest has occurred. The injury to the anterior tibial physis nowadays is usually secondary to direct trauma but previously was not uncommonly associated with prolonged hyperextension applied to

This is complex and lengthy. Each patient needs a customized programme taking into account a number of factors. Weight-bearing status, range of motion restrictions, muscle strengthening, proprioceptive retraining, and soft tissue care (dealing with swelling and the avoidance of fat pad contracture) should be defined.

Outcomes The importance of realistic treatment goals cannot be overemphasized. Patients need to be aware of the likely outcome of the surgery, and the low probability of return to a high level of sporting activity. Return to sports is certainly feasible after an HTO combined with an ACLR alone. In one study half of the patients studied

ARTICLE IN PRESS 120

A. Williams, N. Devic

References

Figure 5 (a) A case of absent ACL plus chronic stretching of MCL in a patient with spina bifida and contralateral below knee amputation. Her gait forced the knee into gross valgus plus external rotation. (b) The same case as in Fig. 5(a). The problem was dealt with by combining, at a single operation, distal femoral osteotomy, and allograft ACL and MCL reconstructions. The bone blocks of the patellar tendon allograft were impacted into recesses made in the tibia and femur.

returned to leisure sports, and 91% of all patients were satisfied with the results of the surgery.11 Patients who have had dislocations followed by multiple ligament reconstruction, and even osteotomy, do get back to sporting activity. However, the proprioceptive deficit must not be overlooked. This means that with the extra stresses to the knee occurring in such activity which is not resisted by normal neuromuscular control, chondral damage is likely. The patient should be warned that it is often best to ‘quit whilst they are ahead’.

Conclusions Osteotomy is an invaluable tool to be used in treating complex knee ligament problems as well as OA.

1. Noyes FR, Schipplein OD, Anriacchi TP, et al. The anterior cruciate ligament-deficient knee with varus alignment. An analysis of gait adaptations and dynamic joint loadings. Am J Sports Med 1992;20:707–16. 2. Dugdale TW, Noyes FR, Styer D. Preoperative planning for high tibial osteotomy. Clin Orthop 1992;274:248–64. 3. Coventry MB. Upper tibial osteotomy for gonarthrosis: the evolution of the operation over the last 18 years and long term results. Orthop Clin North Am 1979;10:191. 4. Marti CB, Gautier E, Machtl SW, Jakob MD. Accuracy of frontal and sagittal plane correction in open-wedge high tibial osteotomy. Arthroscopy: J Arthroscopic Related Surg 2004; 20(4):366–72. 5. Wada M, Imura S, Nagatani K, Baba H, Shimada S, Sasaki S. Relationship between gait and clinical results after high tibial osteotomy. Clin Orthop Related Res 1998;354:180–8. 6. Hernigou PH, Medevielle D, Debeyre J, Goutallier D. Proximal tibial osteotomy for osteoarthritis with varus deformity. A ten to thirteen year follow up study. J Bone Jt Surg 1987;69A: 332–54. 7. Amendola A. The role of osteotomy in the multiple ligament injured knee. Instructional course. Arthroscopy: J Arthroscopic Related Surg 2003;19(10) (Dec, Suppl 1). 8. Bonin N, Ait Si Selmi T, Donnell ST, Dejour H, Neyret P. Anterior cruciate reconstruction combined with a valgus upper tibial osteotomy: 12 years follow-up. Knee 2004;11:431–7. 9. Boss A, Stutz G, Oursin C, Gachter A. Anterior cruciate ligament reconstruction combined with valgus tibial osteotomy (combined procedure). Knee Sports Traumatol Arthrosc 1995;3: 1987–91. 10. Marti CB, Jakob RP. Accumulation of irrigation fluid in the calf as a complication during high tibial osteotomy combined with simultaneous arthroscopic anterior cruciate ligament reconstruction. Arthroscopy 1999;15(8):864–6. 11. Noyes FR, Simon R. The role of high tibial osteotomy in the anterior cruciate ligament-deficient knee with varus alignment. Orthopaedic sports medicine. Principles and practice. Philadelphia: WB Saunders; 1994. p. 1401–43. 12. Logan M, Dunstan E, Robinson J, Williams A, Gedroyc W, Freeman MAR. Tibiofemoral kinematics of the Anterior Cruciate Ligament deficient, weightbearing, living knee employing vertical access open ‘interventional’ MRI. Am J Sports Med 2004;32:720–6. 13. Neyret P, Donnell ST, Dejour H. Results of partial meniscectomy related to the state of the anterior cruciate ligament. Review at 20 to 35 years. J Bone Jt Surg (Br) 1993;75-B:36–40. 14. Noyes FR, Barber-Westin SD, Simon R. High tibial osteotomy and ligament reconstruction in varus angulated, anterior cruciate ligament-deficient knees. A two to seven-year follow up study. Am J Sports Med 1993;21(1):2–12. 15. DeJour H, Neyret P, Boileau P, Donnell ST. Anterior cruciate reconstruction combined with valgus tibial osteotomy. Clin Orthop 1994;12:220–8. 16. Noyes FR, Cummings JF, Grood ES, et al. The diagnosis of knee motion limits, subuxations and ligament injury. Am J Sports Med 1991;19:163–71. 17. Dejour H, Neyret P, Bonnin M. Monopedal weight bearing radiography of the chronically unstable knee. In: The knee and cruciate ligaments. Berlin: Springer; 1990. p. 568–76. 18. Hohmann E, Bryant A, Imhoff AB. The effect of closed wedge high tibial osteotomy on tibial slope: a radiographic study. Knee Surg Sports Traumatol Arthrosc 2005, November 16. 19. Badhe NP, Forster IW. High tibial osteotomy in knee instability: the rationale of treatment and early results. Knee Surg Sports Traumatol Arthroscopy 2002;10(1):38–43.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 121–127

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

KNEE

Alternative surgeries for patello-femoral disorders J. Sanchez-Ballestera,, P. McGrawb, P.G. Turnera, D.S. Johnsona a

Trauma and Orthopaedic Surgery, Stepping Hill Hospital, Poplar Grove, Stockport, SK2 7JE, UK Trauma and Orthopaedic Surgery, Fairfield General Hospital, Rochdale Old Road, Bury, BL9 7TD, UK

b

KEYWORDS Patello-femoral disorders; Pathology; Management; Surgery

Summary The treatment of patello-femoral disorders despite over 80% of cases being successfully treated by non-operative methods is one of the most challenging problems for the orthopaedic surgeon. The understanding of the underlying pathology in each patient before embarking on a treatment plan is paramount. This review intends to clarify the patello-femoral pathophysiology, discusses management plan and describes different surgeries for patello-femoral disorders. & 2006 Elsevier Ltd. All rights reserved.

Introduction While more than 80% of cases of patello-femoral disorders are successfully treated by non-operative methods, they remain one of the most challenging problems for the orthopaedic surgeon. However, most published studies are flawed; cited clinical tests and radiological assessments and the outcome measures are inconsistent, and most studies are retrospective including patients with mixed symptoms and pathology, and of insufficient statistical power. Patello-femoral disorders cover a wide range of pathology (Table 1). The aetiologies of these disorders include soft tissue imbalance, injury, dysplasia and degeneration. These may be linked, with one leading to another. They can present with a variety of symptoms including pain, locking, instability and dislocation. Corresponding author. Tel.: +161 483 1010.

E-mail addresses: [email protected] (J. Sanchez-Ballester), [email protected] (P. McGraw), [email protected] (P.G. Turner), [email protected] (D.S. Johnson). 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2005.11.002

Recent advances in the understanding and identification of the causative factors have allowed development of aetiologically based treatments. This article aims to set out guidelines for the surgical management of each of the underlying conditions, described above, that patients may present with.

Pathology Dejour showed that in patients with a dislocating patella, independent of a traumatic event, there are four factors contributing to pathology of patellar instability and pain. Eighty-five per cent have trochlear dysplasia, 83% have patellar tilt, 56% have a tibial tuberosity lateralised by at least 20 mm with respect to the centre of the trochlea and 24% have patella alta.2 There are other factors such as dysfunction of passive or active soft tissue restraints and extrinsic abnormalities, e.g. femoral neck anteversion, medial tibial torsion and valgus hindfoot with hyperpronated forefoot, all of which influence patellar tracking and stability.3

ARTICLE IN PRESS 122

J. Sanchez-Ballester et al.

Table 1 Insall classification of patello-femoral disorders (Data from Insall, Scott. Surgery of the Knee. Third edition, chapter 46, p. 955).


Presence of cartilage damage J J J J

Chondromalacia Osteoarthritis Osteochondral fractures Osteochondritis dissecans


Variable cartilage damage J J

Malalignment syndromes Synovial plica


Usually normal cartilage J J J J

Peripatellar causes: bursitis, tendonitis Overuse syndromes Reflex sympathetic dystrophy Patellar abnormalities

Unfavourable tracking or mal-alignment results in increased shear and compressive stress in the joint leading to cartilage damage and subsequent degenerative change, but before this point is reached, pain may arise from overstrain of soft tissues. Acutely, chondral injury may follow patellar subluxation or dislocation, usually affecting the lateral facet of the trochlea on dislocation and medial facet of the patella on relocation. Recently, damage to the medial patello-femoral ligament (MPFL) has been shown to be an important aetiological factor in recurrent patellar instability (Fig. 1). The MPFL is a medial passive primary restraint of the patella ruptured during lateral dislocation.3 It is a band of retinacular tissue overlaid by the vastus medialis obliquus (VMO) which connects the medial femoral epicondyle and adductor tubercle to the medial edge of the patella. It is approximately 55 mm long and 3–30 mm wide contributing to an average of 50–60% of the passive restraining force against lateral patellar displacement while the knee is extended or in early flexion. While clinical reports have shown rupture in 75–87% of cases after dislocation, biomechanical studies have found that the ligamentous collagen structure fails at about 20–30% elongation, equivalent to only 12–18 mm lateral patellar subluxation. Undoubtedly, during patellar dislocation this distance is exceeded.4 Patello-femoral arthritis is found in 79% of cadavers, either in isolation or in association with femoro-tibial arthritis. Clearly not all patients are symptomatic. Loss of articular cartilage is most commonly found on the lateral articular facet of the patella, implying excessive comparative loading of this facet. This may or may not be the result of pre-existing malalignment.

Assessment Pain and instability are the two most common complaints in patello-femoral disorders. The pain may be associated with instability or arthritis and is usually felt in the anterior aspect of the knee joint and is characteristically exacer-

Figure 1 Anatomical description of the medial patello-femoral ligament: Horizontal structure from medial aspect of the patella to medial epicondyle.

bated by extended periods of inactivity (including sitting or driving), kneeling, ascending or descending stairs, and walking/running especially downhill. Patients may also complain that their knee has a tendency to ‘‘give way’’. Instability may present as frank dislocation or only subluxation;5 true instability may be used to describe a feeling of the patello-femoral joint subluxing and can relate to deficiency of bony contours and/or ligament/tendon imbalance. However patients may also use it to describe a feeling of instability of the patello-femoral or knee joint secondary to pain with or without secondary muscle weakness.

Examination Examination of the knee includes an assessment of the back, hips, lower limb alignment and evaluation of the gait. There are a variety of specific patello-femoral joint signs, which can assist the surgeon to assess the underlying pathology: The J sign is a sign of pathological tracking during initiation of flexion and refers to the inverted J course of the patella that begins lateral to the trochlea and moves medially to enter the trochlea.5 The Q angle is widely reported in the current literature but it has questionable reproducibility and relevance to underlying malalignment.1,5 Nonetheless, the suspicion is that, if the Q angle is wide, the larger the lateral moment arm, and consequently the greater the lateral instability (normal angle o201).

ARTICLE IN PRESS Alternative surgeries for patello-femoral disorders

123

Examination for lateral retinacular tightness is carried out keeping the knee relaxed in full extension and pushing the patella medially. Normal mobility allows the patella to be parallel to the examination couch or beyond (patella gliding test); if otherwise, consider tight lateral retinaculum or iliotibial band. Patellar stability is assessed by gently pushing the patella laterally while flexing the knee. If the patient is apprehensive this is a sign of poor engagement and possible instability. Finally, the position of the tibial tuberosity with respect to the mid patella position will give some indication of an excessive force pulling the patella laterally.

Imaging Radiological assessment of the patello-femoral joint should aim to demonstrate the patellar configuration, trochlear depth, lateral patellar displacement, lateral tilt (lateral patello-femoral angle), patella height and the presence of arthritis. To achieve this, the minimum recommended radiographic views are the antero-posterior weight bearing, lateral and axial (skyline) views. The AP weight bearing semiflexed view provides information about the tibiofemoral joint.6 The lateral view allows assessment of patellar height, patellar tilt and trochlear groove depth. The lateral radiograph is taken with the knee in at least 301 of flexion to place the patellar tendon under tension.1 A true lateral with the posterior borders of the femoral condyle overlapping is needed to assess the trochlear groove depth; normally 7–8 mm measured lcm from its upper limit. Less than 5 mm is considered dysplastic. Several methods have been described to assess the vertical position of the patella. The most widely used is the Insall–Salvati method, where the ratio of the patella tendon length to the patella height is recorded. A ratio of o0.8 is considered to show patella baja, whilst a ratio of 41.2 is indicative of patella alta. This index has some pitfalls as the tibial tendon insertion may vary, especially after tibial tuberosity transposition.7 Caton described a ratio that compares the length of the patella articular surface with the distance between its lower end and the nearest point on the tibia,7 which is usually close to one. Different shapes of the trochlea have been described by Dejour2 (Fig. 2). In the axial view, tracking and alignment of the patella varies with the amount of knee flexion. The patella engages on the trochlea at 15–301 of flexion. This view is useful, but is only routinely requested at present by 23.1% of UK surgeons.8 In this position various measurements of patello-femoral congruency and dysplasia can be made9 (Fig. 3). These measurements may not be a true reflection of the actual angles, as bony outlines do not always correlate with that of the articular cartilage seen on MR. On all types of image the degree of osteoarthritis present may be underestimated. Cross-sectional imaging may give a better understanding of the dynamics of the patello-femoral joint, and help to plan surgery. We use the method described by Aglietty and Insall where a CT is taken at three different ranges of flexion with relaxed extensor mechanism: 01, 151, 301. In addition to the measurements made using the axial radiograph, the tibial tuberosity-sulcus femoralis distance (TT-SF) described

Figure 2 Dejours’s trochlear morphology: (A) normal knee, the sulcus line (corresponding to the trochlear floor) does not cross the condyle anteriorly. (B) Type I dysplasia. The medial femoral condyle is deficient (flatter). The sulcus line joins the line of the medial condyle distally. (C) Type II dysplasia. The medial femoral condyle is not deficient. The crossing of the two condylar outlines of the trochlear floor is symmetrical but situated distally. (D) Type III dysplasia: the crossing of the two condylar outlines with asymmetry of the outline of the trochlear floor2

Figure 3 Measurements of patello-femoral relationships on the axial radiograph. (A) Sulcus angle ¼ a–b (normal range 126–1501), congruence angle (curved line) ¼ c (the sulcus angle is bisected to produce a reference line) –d (line joining the apex of the sulcus and the lower point in the patella). Normal range no more than 161 lateral to the bisected sulcus angle. (B) Lateral patellar displacement ¼ e f line–g (dropping a perpendicular to this at the level of the summit of the medial femoral condyle). The distance of the medial margin of the patella from this perpendicular is measured. In the normal knee the medial patellar margin should lie no more than 1 mm lateral to the perpendicular. The lateral patello-femoral angle (curved arrow) ¼ e f–h. The angle is taken to be normal when it opens laterally, and abnormal when it opens medially.33

ARTICLE IN PRESS 124 by Goutallier can be measured. On the same sequences lower limb rotational alignment can be assessed, including femoral anteversion. Finally a CT scanogram scout film can measure any angular deformities in the coronal or sagittal plane, although interpretation needs to be guarded as these films will be non-weight bearing. MR scanning may be of some use in assessing the soft tissue structures around the patella and excludes other internal pathology. Assessment of the depth of the joint surfaces can be difficult depending on the resolution of the scanner. There may be a role in the future for dynamic MR scanning of patello-femoral joint motion.

Arthroscopy Arthroscopy ought only to be used when combined with treatment and not for purely diagnostic purposes. It is helpful in assessing the status of the articular cartilage, but the dynamics of patello-femoral engagement even when using the supero-lateral portal is not particularly easy to assess.

Surgical treatment The majority of patello-femoral problems can be treated with conservative measures. Before embarking on surgery the degree of soft tissue imbalance, injury, dysplasia and degeneration must be evaluated and correlated with the symptoms as surgery undertaken without a well-defined underlying pathology gives poor results. It is must not be forgotten that (Dejour), several anatomical defects may be present in the same knee. Thus while discussed separately here, more than one procedure may be necessary in a given knee.

Soft tissue imbalance and injury Tight lateral structures with otherwise normal anatomy, as demonstrated by lateral subluxation/tiit of the patella, usually respond to lateral retinacular release with satisfactory results.10 However, this procedure has an overall complication rate of 10.2%. The most common is haemarthrosis and in the long term excessive loading of the medial facet can give as severe symptoms as the lateral side did pre-operatively. Early complications can be reduced by performing the procedure by arthroscopic means without tourniquet, using diathermy and avoiding the use of a postoperative suction drain.11

Distal re-alignment procedures Many distal soft tissue realignment techniques such as the Goldthwait-Roux have been described in the past. They give moderately satisfactory results and are advocated for skeletally immature patients with recurrent dislocation. However, there is a risk of patella tendon rupture following this procedure.12 In the adult with associated patella alta or a laterally sited tibial tuberosity, medial tibial tuberosity transfer has been used successfully. This procedure was described

J. Sanchez-Ballester et al. originally by Roux in 1888 and modified later on by Elmslie and Trillat in 1964. Medial transposition alone is inadequate in most cases. It is usually supplemented with a lateral release and possibly medial reinforcement, although the latter’s efficacy is less well proven. A recent study has shown that medialisation alone prevents recurring dislocation in 87% of cases, but may not affect the development of long term patello-femoral degeneration. However, if the procedure is performed before significant cartilage damage has occurred the long-term results appear to be better.13

First time dislocator The management of first time dislocation has to be considered separately. Treatment of an initial dislocation is still controversial. Conservative treatment, with a period of immobilisation has a redislocation rate of up to 63% with restriction in activity level at 6 months of 58%.14 Rupture of the MPFL is the essential lesion in traumatic dislocation, but the only randomised trial for acute repair of the MPFL showed no improvement in the risk of recurrent dislocation after surgery, compared to patients treated non-operatively treatment. However, not all the operative patients underwent the same procedure.15 Nonetheless MPFL reconstruction is emerging as a treatment for recurrent patella dislocation and subluxation, often in association with correction of other predisposing factors (Fig. 4). Semitendinosus tendon, quadriceps tendon or medial retinacular autograft have been used as a biomechanical study that has shown that simple sutured repairs are weaker than the normal MPFL; 18% of the natural strength. However, the repair with suture anchors or tunnel reconstruction shows strengths of 68% and 94% respectively.16 Autograft reconstruction is a more extensive procedure with possible donor site morbidity.16 A recent study of 15 patients at 5 years after MPFL reconstruction had no redislocations, 94% were pain free and 88% showed normal tracking, but only 50% had returned to normal activity.17 On balance non-operative treatment is still the mainstay for the first time dislocator. While MPFL is an important structure to maintain stability of the patella, further research is needed to define the best reconstruction technique.

Correction of trochlear dysplasia Lateral condylar elevation (Albee osteotomy) was described as a salvage operation to treat recurrent patellar dislocations. A biomechanical study of patello-femoral joint pressures showed a significant increase of mean and peak contact pressures which may lead to chondral damage and arthrosis.18 More recently there has been interest in trocheoplasty or deepening the trochlea. Initially described by Dejour and published by Reynaud, nearly all studies are published in French (Fig. 5), and there is little in the English literature. To our knowledge, there is only one study written in English.19 Thirteen knees with trochlear dysplasia underwent Dejour trocheopiastv for recurrent patella dislocation and or persistent retropatellar pain. Good to very good

ARTICLE IN PRESS Alternative surgeries for patello-femoral disorders subjective results were achieved in 77%. However, there were five cases of arthrofibrosis. Two patients complained of persistent retropatellar pain and three patients developed impingement of the fixation material.

Surgery for patello-femoral arthritis Surgery may be appropriate for patello-femoral arthritis if non-operative measures fail to control pain. In the past, many surgical procedures were advocated including those now considered of historical interest, e.g. patellectomy.20,21 While soft tissue procedures have a limited role in the treatment of established arthritis, lateral release may be of benefit if there is patellar tilt in absence of patellar instability.22 While some have suggested incision of the retinaculum may provide analgesia through denervation,23 we would not recommend this as a procedure in isolation.

125 Osteotomy of the tibial tubercle can be used to correct malalignment of the extensor mechanism. Performed correctly it off-loads diseased articular cartilage and transfers loading onto healthier cartilage. Thus a Trillat (straight medial) tibial tubercle transfer is indicated when lateral tracking results in a lateral articular facet lesion.24 Anteriorisation of the tibial tubercle (Maquet procedure) is indicated for the treatment of an articular lesion confined to the distal aspect of the patella, with normal alignment. Schmid found that 28 out of 35 patients were rated as very good or good following this procedure for patello-femoral arthritis. The outcome of the remaining seven cases was attributed to wrong indications to perform this procedure or technical error.25 Schepsis achieved similar results with an excellent or good outcome in 48 out of 56 knees following anterior tibial tubercle transposition for patello-femoral osteoarthritis. The most common presentation is a lesion affecting the lateral and distal portion of the articular surface of the patella, consequent on lateral patellar tracking. An oblique osteotomy to transpose the tibial tubercle anteriorly and medially may be considered, to off-load the damaged articular cartilage and improve patellar tracking, (Fig. 6). Originally described in 1983,26 this technique avoids the necessity for bone graft required for the Maquet procedure, but offers the same off-loading of the patello-femoral joint. Fulkerson demonstrated good or excellent results in 90% of 30 patients at 35 months follow-up.27 An antero-lateral tibial tubercle transfer can be utilised in patients presenting with articular lesions of the medial articular facet of the patella.

Autologous chondrocyte implantation

Figure 4 Medial patello-femoral reconstruction. There are different alternatives to reconstruct MPFL. The new ligament is passed through the patellar tunnel and is fixed looping into the semimembranosus. Another alternatives is to fix to the medial epicondyle by screw.

Autologous chondrocyte implantation is indicated for large contained full thickness cartilage injuries in the presence of a maintained radiographic joint space in young patients with isolated patello-femoral non-inflammatory arthritis. It is essential to simultaneously correct alignment and tracking of the patella to optimise the outcome. It may be also be used to improve the outcome of antero-medial tibial

Figure 5 Trochleoplasty as described by Dejour. Lateral parapatellar arthrotomy. The subchondral bone is shaved from proximal to distal. Longitudinal cartilage incision along the new trochlear groove. The cartilage is impacted and fixed with screws.

ARTICLE IN PRESS 126

J. Sanchez-Ballester et al. with the Avon patello-femoral prosthesis (Stryker Orthopaedics, Limerick, Ireland), which was introduced in 1996.32 The design of this prosthesis is such that the biomechanics replicate better the normal patello-femoral joint, with less constraint than its predecessors. Whichever system is used it must be anatomically placed and any other anatomical deformity be addressed, such as lateral placement of the tibial tuberosity causing instability after insertion of the prosthesis.

Figure 6 Fulkerson anteromedial tibial tubercle transfer. The obliquity of the osteotomy gives more anteriorisation or more medial transfer of the tubercle. (A1) Less oblique osteotomy giving more medialisation and less anteriorisation. (A2) Marked oblique osteotomy, giving more anterorisation and less medialisation.

tubercle osteotomy in patients with medial facet or proximal pole articular lesions as such patients are known to have a poorer outcome compared to those with inferior pole or lateral facet lesions.28 An arthroscopy is performed to assess suitability for the procedure and to obtain the cartilage sample (from which the chondrocytes are to cultured). It is harvested from the non-weight bearing portion of the superior intercondylar notch. Approximately 4 weeks later the graft is inserted as an open procedure, either cells in suspension under a sutured patch (ACI) or as a matrix implanted with the cells (MACI). Peterson studied 101 patients receiving 110 autologous chondrocyte transplants. Fifteen patients had multiple focal femoral/patellar defects with a mean size 4.6 cm2, nine patients had an excellent or good result at a mean follow up of 2.7 years. The outcome of treatment of isolated patellar lesions in previous studies was less favourable. The more radical debridement of chondromalacic tissue around the defect and attention to correcting patello-femoral maltracking resulted in good or excellent outcomes in eleven out of nineteen patients.29 At present ACI or MACI performed in the UK has to be performed as part of a clinical trial according to NICE guidelines.

Arthroplasty Patello-femoral or total knee replacement is now becoming increasingly popular for established patello-femoral osteoarthritis. Patello-femoral arthroplasty is indicated in patients with isolated patello-femoral arthritis with an intact tibio-femoral joint, and normal ligaments and menisci. It may be considered for the younger patient with severe trochlear dysplasia resulting in marked instability. Clearly for whichever indication it is used, the need for future revision needs to be weighed against the benefits especially in the younger patient. Lubinus reported the first total patello-femoral arthroplasty in 1979,30 but this procedure was found by Tauro et al. to have an 8 year failure rate of 50%.31 Causes of failure included wear, impingement, disease progression and persistent malalignment. Better results have been obtained

Summary and conclusion A better understanding of patello-femoral pathophysiology, should lead to an improvement in the management of patello-femoral disorders. Thus it is important to understand the underlying pathology in each patient before embarking on treatment.

References 1. Insall S, Surgery of the Knee, third ed., Churchill Livingstone 2000. 2. Dejour H, Walch G. Factors of patella instability: an anatomic radiologic study. Knee Surg Sport Traumatol Arthrosc 1994;2: 19–26. 3. Post WR, Teitge R, Amis AA. Patellofemoral malalignment beyond the view box. Clin Sports Med 2002;21:521–46. 4. Amis AA, Firer P, Mountney J. Anatomy and biomechanics of the medial patellofemoral ligament. Knee 2003;10:215–20. 5. Post WR. Clinical evaluation of patients with patellofemoral disorders. Arthroscopy 1999;5(8):841–51. 6. Buckland-Wright JC. Substantial superiority of semiflexed views in the knee OA. J Rheumatol 1999;26912:2664–74. 7. Caton J, Mironeau A. Adolescent idiopatic patella alta: a review of 61 casestreated surgically. French J Orthop Surg 1990;4: 196203. 8. Vince AS. What X-rays do we need? Knee 2000;7:101–4. 9. Laurin CA, Dussauk MD. The tangential X-ray investigation of the patellofemoral joint. Clin Orthop Relat Res 1979;N144:16–26. 10. Boden BP, Pearsall AW, Garret WE, et al. Patellofemoral instability. Evaluation and management. JAAOS 1997;5:47–57. 11. Small NC. An analysis of complications in lateral retinacular release procedures. Arthroscopy 1989;5:282–6. 12. Hughston JC, Deese M. Reconstruction of the extensor mechanism for subluxing patella. Am J Sports Med 1972;1:6. 13. Nakagawa K, Wad Y, Minamide M, et al. Deterioration of longterm clinical results after the Elmslie-Trillat procedure for dislocation of the patella. J Bone Joint Surg 2002;85-B(6). 14. Arendt EA, Fithian DC. Current concepts of lateral patella dislocation. Cinic Sports Med 2002;21:499–519. 15. Nikku R, Nietosvaara Y, et al. Operative versus operative treatment of primary dislocation of the patella. Acta Orthop Scan 1997;68:419–23. 16. Mountney J, Senavongse W, Amis AA, Thomas NT. J Bone Joint Surg 2005;87-B(1). 17. Ellera Gomes JL, et al. Medial patellofemoral ligament reconstruction with semitendinosus autograft for chronic patellar instability: A follow up study. J Arthroscopy R Surg 2004;20(2):147–51. 18. Kuroda R, Kambic H, et al. Distribution of patellofemoral joint pressures after femoral trochlear osteotomy. Knee Surg Sport Traumatol Arthros 2002;10:33–7. 19. Verdonk R, Jansegers E, et al. Trochleoplasty in dysplastic knee trochlea. Knee Surg Sport Traumatol Arthrosc 2005.

ARTICLE IN PRESS Alternative surgeries for patello-femoral disorders 20. Kelly MA, Insall JN. Patellectomy. Orthop Clin N Am 1986;17: 289–95. 21. Giinal I, Karatosun V. Patellectomy: an overview with reconstructive procedures. Clin Orthop Relat Res 2001;389:74–8. 22. Aderinto J, Cobb AG. Lateral release for patellofemoral arthritis. J Arthroscop Relat Surg 2002;18:399–403. 23. Fulkerson JP, Tennant R, Jaivin JS, Grunnet M. Histological evidence of retinacular nerve injury associated with patellofemoral malaiignment. Clin Orthop 1985;197:196–205. 24. Trillat A, Dejour H, Couette A. Diagnosis and treatment of recurrent dislocations of the patella. Rev Chir Orthop Reparat Appar Mot 1964;50:813–24 (French). 25. Schmid F. The Maquet procedure in the treatment of patellofemoral osteoarthritis—long term results. Clin Orthop 1993; 294:254–8. 26. Fulkerson JP. Anteromedialisation of the tibial tuberosity for patellofemoral malaiignment. Clin Orthop 1983;177:176–81.

127 27. Fulkerson JP, Becker GJ, Meaney JA, Miranda M, Folcik MA. Anteromedial tibial tubercle transfer without bone graft. Am J Sports Med 1990;18:490–7. 28. Pidoriano AJ, Weinstein RN, Buuck DA, Fulkerson JP. Correlation of patellar articular lesions with results from anteromedial tibial tubercle transfer. Am J Sports Med 1997;25:533–7. 29. Peterson L, Minas T, Brittberg M, Nilsson A, Sjogren-Jansson E, Lindahl A. Two- to 9- year outcome after autologous chondrocyte transplantation of the knee. Clin Orthop 2000;374:212–34. 30. Lubinus HH. Patella glide bearing total replacement. Orthopaedics 1979;236:162–7. 31. Tauro B, Ackroyd CE, Neuman JH, Shah NA. The Lubrinus patellofemoral arthroplasty. A 5–10 year prospective study. J Bone Joint Surg Br 2001;83:696–701. 32. Ackroyd CE, Chir B. Development and early results of a new patellofemoral arthroplasty. Clin Orthop Relat Res 2005;436: 7–13.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 128–131

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

SYNDROME

Morquio syndrome Benedict J.A. Lankestera,, Michael Whitehousea, Martin F. Garganb a

Bristol Royal Infirmary, Bristol, BS2 8HW, UK Bristol Royal Infirmary and Bristol Hospital for Sick Children, Bristol, BS2 8HW, UK

b

KEYWORDS Morquio syndrome; MPS IV; Mucopolysaccharidoses; Skeletal dysplasia

Summary Morquio syndrome (mucopolysaccharidosis (MPS) type IV) is a rare inherited cause (autosomal recessive) of short-trunk dwarfism. Skeletal manifestations of this spondyloepiphyseal dysplasia include severe growth retardation, odontoid hypoplasia, thoracolumbar kyphosis, hip dysplasia, genu valgum and marked skin and joint laxity. Mental function is normal, and the coarse facial features associated with the other MPS types are not present. Treatment is supportive only, with most affected individuals living until early adulthood. & 2006 Elsevier Ltd. All rights reserved.

Introduction In 1929, Luis Morquio (1867–1935), a paediatrician in Montevideo, Uruguay, described a ‘‘familial skeletal dystrophy’’ in four out of five children of a consanguineous family of Swedish origin.1 Simultaneously but independently, James Brailsford (1888–1961), a radiologist in Birmingham, UK, described the clinical and radiological features of a child with ‘‘chondro-osteo dystrophy’’.2,3 Both are thought to be descriptions of patients with Morquio syndrome (also known as Morquio-Brailsford syndrome), a cause of short-trunk dwarfism. It is one of a group of inherited metabolic disorders, the mucopolysaccharidoses (MPSs) (Table 1). Each is due to a specific lysosomal enzyme defect leading to incomplete breakdown of complex proteoglycans, causing accumulation Corresponding author. Tel.: +44 117 923 0000;

fax: +44 117 928 2659; mob: 07787 552289. E-mail addresses: [email protected] (B.J.A. Lankester), [email protected] (M. Whitehouse), [email protected] (M.F. Gargan). 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2005.12.003

of glycosaminoglycans (previously called mucopolysaccharides) that interfere with cell function. They are defined by the enzymatic defect, the type of glycosaminoglycan excreted in the urine and by the clinical features. Skeletal effects are particularly marked in MPS types I H (Hunter), I S (Scheie) and IV (Morquio).

Epidemiology and genetics Morquio syndrome (MPS IV) is rare with an overall incidence of 1 in 40,000.4 The syndrome is found in all ethnic groups, but with higher incidence in French Canadians. Two types have been described—type A (more severe form) with a deficiency in N-acetyl-galactosamine-6-sulphatase (chromosome 16q) and a rarer type B with a deficiency in b-Dgalactosidase (chromosome 3p).5 A variety of different gene mutations that result in defective enzymes have been identified.6 The inheritance pattern is autosomal recessive, with parents therefore unaffected by the disease. Laboratory diagnosis is possible with detectable enzyme deficiencies in fibroblasts and amniocytes, allowing pre-natal diagnosis in subsequent pregnancies.

ARTICLE IN PRESS Morquio syndrome

Table 1

129

Mucopolysaccharidoses.

MPS

Syndrome

Enzyme deficiency

GAG excreted

Inheritance

IH IS I HS II III IV A IV B V VI VII

Hurler Scheie Hurler-Scheie Hunter Sanfilippo A–D Morquio

a-L-Iduronidase

Dermatan sulphate, heparan sulphate

AR

Iduronate sulphatase Varied–4 types N-acetylgalactosamine-6-sulphatase b-D-galatosidase

Dermatan sulphate, heparan sulphate Heparan sulphate Keratan sulphate

X-linked AR AR

N-acetylgalactosamine-4-sulphatase b-glucuronidase

Dermatan sulphate Dermatan sulphate, heparan sulphate

AR AR

Previously Scheie Maroteaux-Lamy Sly

Clinical presentation Children with Morquio syndrome have skeletal manifestations that result from a unique spondylo-epiphyseal dysplasia with ligament laxity. Growth and development are normal in the first year or two of life and the diagnosis is usually made between 2 and 4 years of age, although the severity of the features varies and milder forms of the syndrome may go undiagnosed. They can be distinguished from other MPSs at an early age as they have normal mental function and do not have coarse facial features.7 There is severe growth retardation, usually with disproportionate short stature (short trunk), a large head and short neck, a prominent maxilla with wide-spaced teeth, thoraco-lumbar kyphosis with other spine and rib abnormalities (see ‘‘radiological features’’), hip dysplasia, genu valgum and marked joint and skin laxity. The combined abnormalities result in a ‘‘duck-waddling’’ gait. Corneal clouding, deafness, aortic incompetence and hepatomegaly are all typical. Keratan sulphaturia is present, but decreases with age. Lifespan is variable, depending on disease severity, but many die in early adulthood with cardiopulmonary disease or the sequelae of neurological deficits.

Radiological features Axial skeleton (Figs. 1–3)

 Hypoplastic/absent odontoid with atlanto-axial subluxation

Appendicular skeleton (Figs. 3–5)

 Acetabular dysplasia with coxa valga  Genu valgum  Epiphyseal and metaphyseal irregularity (advanced stages)

 Metacarpals small and irregular with pointed proximal ends

Treatment The management of Morquio syndrome is currently limited to supportive care. Affected children need regular range of movement exercises and night splintage to limit progressive loss of motion. If surgery is contemplated, careful anaesthetic planning is required to address potential cervical instability, silent aortic incompetence, and respiratory insufficiency caused by spinal and thoracic deformities.8–10

Spine Spinal surgery is often necessary, as odontoid hypoplasia and the consequent C1-2 instability, if untreated, gives rise to myelopathy, quadriplegia or even sudden death.11 Hence posterior C1-2 fusion is often performed prophylactically aged 9–1012 to regain stability and protect the spinal cord. The thoraco-lumbar spinal deformity can be braced if necessary, but may need anterior decompression and strut grafting and posterior instrumented fusion if severe.

 Thoracolumbar kyphosis with wedging of apical     

vertebrae Anterior vertebral beaking at other levels Platyspondyly (flattened vertebrae) Pectus carinatum Rib flaring Constricted iliac wings—‘‘wine-glass’’ pelvis

Lower limbs Hip dysplasia may require containment osteotomies of the acetabulum and proximal femur. Genu valgum can be addressed with medial stapled hemi-epiphysiodesis, but realignment osteotomies may also be required, usually of the proximal

ARTICLE IN PRESS 130

B.J.A. Lankester et al.

Figure 3 Constricted iliac wings, acetabular dysplasia and coxa valga.

Figure 1 Odontoid hypoplasia.

Figure 4 Lower limb malalignment with genu valga.

tibia. Contractures around the hip, knee and ankle are difficult to address, as soft-tissue releases rarely successful.

Other

Figure 2 Thoracolumbar kyphosis with vertebra plana.

Gene therapy research promises new future treatments. The deficient enzyme N-acetylgalactosamine-6-sulfatase can be transferred via a recombinant retroviral factor, leading to correction of the metabolic defect in Morquio syndrome fibroblasts in the laboratory,13 but there has been no success yet in vivo. Bone marrow transplantation and enzyme

ARTICLE IN PRESS Morquio syndrome

Figure 5 Small pointed metacarpals.

replacement therapy have been partially successful in the treatment of Hurler syndrome with increased lifespan, but the musculoskeletal condition is unchanged.

References 1. Morquio L. Sur une forme de dystrophie osseuse familiale. Archives de me´decine des enfants 1929;32:129–35. 2. Brailsford JF. Chondro-osteo-dystrophy. Roentgenopgraphic & clinical features of a child with dislocation of vertebrae. Am J Surg 1929;7:404–10.

131 3. Brailsford JF. Chondro-osteo-dystrophy. Roentgenopgraphic & clinical features of a child with dislocation of vertebrae. Br J Radiol 1931;4:83–9. 4. Nelson J, Crowhurst J, Carey B, Greed L. Incidence of the muccopolysaccharidoses in Western Australia. Am J Med Genet A 2003;123:310–3. 5. Chudley AE, Chakravorty C. Genetic landmarks through philately: Luı´s Morquio. Clin Genet 2002;62:438–9. 6. Terzioglu M, Tokatli A, Coskun T, Emre S. Molecular analysis of Turkish MPS IVA (Morquio A) patients: identification of the novel mutations in the GALNS gene. Hum Mutat 2002; 20:477–8. 7. Mikles M, Stanton RP. A review of Morquio syndrome. Am J Orthop 1997;26:533–40. 8. John RM, Hunter D, Swanton RH. Echocardiographic abnormalities in type IV MPS. Arch Dis Child 1990;65:746–9. 9. Morgan KA, Rehman MA, Schwartz RE. Morquio’s syndrome and its anaesthetic considerations. Paediatr Anaesth 2002;12: 641–4. 10. Tobias JD. Anaesthetic care for the child with Morquio syndrome: general versus regional anaesthesia. J Clin Anaesth 1999;11:242–6. 11. Kopits SE. Orthopaedic complications of dwarfism. Clin Orth 1976;114:113–79. 12. Svensson O, Aaro S. Cervical instability in skeletal dysplasia. Report of 6 surgically fused cases. Acta Orthop Scand 1988;59: 66–70. 13. Toietta G, Severini GM, Taversari C, Tomatsu S, Sukegawa K, Fukuda S, et al. Various cells retrovirally transduced with N-acetylgalactosoamine-6-sulphate sulfatase correct Morquio skin fibroblasts in vitro. Hum Gene Ther 2001;12: 2007–16.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 132–140

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

SPINE

Development in the management of tuberculosis of the spine$ Myung-Sang Moona,b, a

Spine Center, Sun General Hospital, Daejeon, Republic of Korea Moon Kim’s Institute of Orthopaedic Research, Seoul, Republic of Korea

b

KEYWORDS Tuberculosis; Management; Development; Spine

Summary The history of the management of spinal tuberculosis is closely linked with the development of civilization. The author describes the development of antituberculous medicine and surgery, based of the era of medical art and science. Also the author describes the bacteriological behaviour of tubercle bacilli and drug response, and the expected development of surgery of tuberculosis of the spine. & 2006 Elsevier Ltd. All rights reserved.

Introduction Goals in the management have changed significantly during the last two centuries as a result of rapid and remarkable advances in science, particularly during the second half of the 20th century.1–13 It was once thought that the demand for spinal deformity prevention and correction was a luxury. Even as late as 1970s the primary goal of the management was to save the patients’ life by curing the disease rather than prevention and/or correction of the deformity the surgeons till this time, had no solutions to meet the

$ Oration delivered by the author as Professor P.S. Ramani’s oration at Neurospine Conference—2005 in Pune, Maharashtra, India on 16th July 2005. Corresponding author. Moon-Kim’s Institute of Orthopaedic Research, Shi-Bum Apt. 14-105, Yoe-Ee-Do-Dong, Young-Dung-PoKu, Seoul 150-761, Republic of Korea. Tel.: +82 2 780 5387; fax: +82 2 785 6065, +82 42 603 7387. E-mail addresses: [email protected], [email protected].

0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2005.12.002

patients’ aesthetic demand which has grown gradually and become more intense since early 1980s. This period was also one of economical growth. Antituberculous drugs changed the patients’ fate. It was soon realized that the specific chemotherapeutic agents alone could cure not only active tuberculosis but it also helped in the recovery from paralysis.11,14,15 The operative treatment was reserved for: (1) failure of the drug therapy, (2) recrudescence of the disease, (3) Pott’s paraplegia that did not resolve after 4–6 weeks of chemotherapy, (4) involvement of spinal cord, or (5) other complications. Antituberculous drugs also made the surgery safer.

Historical background The history of the management of spinal tuberculosis is closely linked to the development of civilization and economic growth in each country and each continent. The change started in the 19th century when old fashioned artistic medical practice slowly changed into scientific medicine. In the 19th century most practitioners used

ARTICLE IN PRESS Development in the management of tuberculosis of the spine ‘home methods’ for treatment as scientific progress had not reached all corners of the tuberculous zone. Pseudoscientific theories were rampant and popular. The contribution of Lister (surgery), Pasteur (bacteriology), Morton (anesthesiology), and Roentgen (X-rays) helped us to enter the medical science era in the 20th century. However, there was no progress in the development of specific anti-tuberculous drugs and surgery could not make any steady progress. The revolution produced later by chemotherapy relegated all traditional methods of conservative treatment of spinal tuberculosis to the museum. Even today, specific drugs are a most important weapon in the management of tuberculosis.11,14 In the ancient Indian religious text [Rig veda and Atharva veda; (3500–1800 BC): Samhita of Charaka and Sushruta (1000 and 600 BC)], tuberculosis was described as ‘Yakshma’ and ‘Sipudru’ was prescribed for spine tuberculosis, abscess and local inflammation, while ‘Jangida’ was prescribed for ‘Yakshma’ and also for persistent cough, pleurisy, lumbago, and rheumatic pain. In the ancient Hindu religious text there are descriptions about the traction treatment for hunchback.15 In 1570, Delechamps clarified the relationship of spine caries and lower limb paralysis12,16,17 but he did not mention the solution to the limb paralysis. In 1672 Wiseman used the scalpel for abscess drainage which is the first record in the world literature.16,17 Abscess drainage was known to be practised by the barber’s surgeons or the folk medicine practitioners in many countries during the medieval era.

Table 1

133 Abscesses were drained by any sharp instrument including the sharp edge of a broken ceramic dish or vase.18

Principles of management of spinal tuberculosis It is felt that the historical review of developments in the management of the spinal tuberculosis is timely in the era of the resurgence of the spinal tuberculosis associated with the world-wide spread of the human immunodeficiency virus disease (HIVD). It is more reasonable to explain the development of the management of the spinal tuberculosis based on the advent of the modern antituberculous chemotherapeutic agents. I have classified this into two eras: pre-chemotherapy and chemotherapy eras (Tables 1–3). Measures for management of the disease have been urgently needed, primarily to save life, and secondly to arrest or cure the disease. The progression of disease causing destruction and deformity needed to be halted and, prevent and overcome the dreaded neurological complication of ‘paraplegia’. We surgeons have to be prepared for the possible recrudescence of the disease in later years. In a disease as treacherous as tuberculosis, it is better to avoid the term ‘cure’ when the patient with advanced disease is treated by chemotherapy alone, regular follow-up is mandatory to

Development in the management of tuberculosis of spine, based on development of science.

ARTICLE IN PRESS 134

Table 2

M.-S. Moon

Development in the management of tuberculosis of spine, based on development of science.

diagnose the relapse. Treatment should be prolonged and a cautious attitude maintained for many years.

Conventional conservative management This management is nowadays purely a supplementary one though once in the medieval times evil-spirit chasing rituals, royal touch for King’s evil (scrofula) and other folk remedies were practised.16,19,20 Until the mid 20th century patients and physicians depended on nature to heal tuberculosis.

Classification of the conservative management I have classified the conservative management as follows: 1. Spiritual and religious methods (evil chasing ritual, Royal touch for Kings’ evil): Those belonged truly to psychological treatment and not to physical or medical treatment. 2. Non-drug management: For many centuries patients and physicians depended on the natural healing through exposure to sunlight, fresh air, maintaining good hygiene, taking good food and rest, etc. This kind of management was systematically practised through sanatorium care from mid 19th century till the early 1980s.

Now those measures are not in reality used any more for care of spinal tuberculosis except the hyperalimentation measures. In the pre-chemotherapy era, the disease progressed and produced various complications leading to deterioration in the patient’s general condition and sometimes many cases were fatal. 3. Drug management: Any combination of all of the nonantituberculous drugs were used for pure supplementary purposes such as pain alleviation and nourishment. In the Western world the extract of the willow bark or leaves (salix) was used for pain and fever by Stone in 1763. In 1853 acetylsalicylic acid (ASA) was synthesized by Gerherdt. ASA was used for pain alleviation in the early days, but since 1987 it is used for early bone tuberculosis, not only to suppress the inflammatory process but also to prevent bone absorption through the inhibition of prostaglandin E2 synthesis. I use ASA for 1–3 months even in the era of chemotherapy.21 4. Local supportive measures: In the pre-chemotherapy era, external spinal support measures were introduced for spinal tuberculosis in the late 19th century and were used until the advent of chemotherapy. Immobilisation by bed rest and body support or posterior spinal fusion (Hibbs and Albee, 1911) provided the local stability conducive to healing of the lesion.1,2 The revolution produced by chemotherapy has relegated all traditional methods of

ARTICLE IN PRESS Development in the management of tuberculosis of the spine

Table 3

135

Treatment of spinal tuberculosis.

conservative treatment to the museum.11,14 However, those methods are still widely used in certain countries. Those doctors still have reluctance to abandon traditional methods, and have advocated a ‘middle path’ regimen, combining the chemotherapeutics with rest on a hard bed or plaster bed.17 5. Chemotherapy: With the advent of specific antituberculous chemotherapy (SM in 1944, PAS in 1946, INH in 1951, PRZA in 1952, RF in 1965 and ETB in 1961) the clinical course of tuberculosis has been changed so much that the patients no longer die but are cured. The period of infectivity was considerably reduced, relapses were avoided and the chronicity was reduced. In the pre-chemotherapy era surgery for active spinal tuberculosis failed in the majority of cases because surgery could not eradicate the infection.1,2,5,6 In the chemotherapy era all surgery for active spinal tuberculosis is done under the cover of chemotherapy and the results are successful in eradicating tuberculosis and preventing and/or correcting spinal deformities.8,11,15,21 Chemotherapeutic agents are clearly known to be the most potent weapons in curing the tuberculosis and surgery was definitely found to be the supplementary one.11,14,22 However, surgery has made constant advancement, solving the patients’ complaints by curing the disease and its complications. During the early stage of its development the aim of surgery was purely to cure the disease but was unsuccessful.5,6 At a later stage surgery aimed to treat the disease-

related complications and involved spinal decompression for the limb paralysis, stabilization for the affected unstable spine and correction surgery for deformity.8,12,20,22,23 It must be borne in mind that any surgery for the management of spinal tuberculosis could not succeed without chemotherapy. The question of the choice of treatment of uncomplicated spinal tuberculosis had been largely answered by a series of controlled clinical trials carried out by the British MRC working party since 1964 to 1970. During this period the responsible opinions on the best method of treatment varied widely. Since the advent of chemotherapy, there has been tremendous change in the treatment protocol. Until 1970, the classical ‘first line’ drugs (streptomycin, INAH, PAS), the original triple therapy, were used. Rifampicin, ethambutol and pyrazinamide were added later. PAS was gradually eliminated in drug combination in view of its side effects. Now three or four drug regimens include four second line drugs (INAH, rifampicin, ethambutol, pyrazinamide). Three drug regimen had been the principal formula but since 1990 four-drug regimen has been adopted to shorten the duration of drug administration and to eradicate the resistant strains. Even today there is no standardised drug regimen applicable to every patient, although an ‘acceptable regimen’ has been suggested. Duration of drug administration is still an issue: 6, 9, 12, 18 or 24 months, though in most instances 9–12 months chemotherapy has been adopted for spinal tuberculosis.23,24

ARTICLE IN PRESS 136 In 1994 Upadhyay reported that 9 months short course chemotherapy was successful in patients undergoing anterior radical surgery while others are still insisting on the 12 months chemotherapy, particularly in the non-surgically treated cases.14,25 Dormant bacilli: Treatment of the dormant mycobacterium in the tuberculous lesion is a problem, because it does not completely respond to drugs. This is the cause of a relapse or a reactivation. As a result of very slow oxygen deprivation the mycobacterium in the tuberculous lesion becomes dormant. My suggestion as the first step is to awake the dormant bacilli through hyperbaric oxygen therapy and the increase of local circulation. Frequent hyperbaric oxygen therapy and oral prostaglandin E2 administration (Opalmon 5 mg, 3 times daily) have been used by me. To label the patient as a non-responder, based only on the drug response after a fixed duration, is unfair and unconvincing. There have been the two different criteria for drug response assessment for non-paralytic and paralytic patients. For the former, drug response was assessed at 3 months while for the latter it was assessed at 3–4 weeks. At present, the known preventive measures for the drug resistant strains are: (1) Consistent use of two or more effective agents. (2) Drug resistance usually reflects failure of a treatment programme; failure of adherence to therapy, as increase of the dose for 3–4 weeks and/or change of chemotherapy formula. To combat multi-drug resistant tuberculosis, it is essential to develop a novel anti-tuberculous composition of the drugs. Where do we go from here? We have to consider the adjunctive immunotherapy (immunopotentiation) which may also have a role in successfully treating spinal tuberculosis especially in HIV and other immunocompromised patients. Anti-tuberculous drugs changed the surgical approach to spinal tuberculosis.11 Surgery is only an augmentation to chemotherapy, and is not the key treatment. Surgery in tuberculosis has contributed significantly to improving the quality of patients’ lives and in meeting the patients’ aesthetic demands. The Hong Kong surgeons claimed that radical surgery with chemotherapy gave excellent results in their hands and felt that it was the obvious answer to spinal tuberculosis. They also stressed that healing must be accompanied by fusion.8 At the other extreme, Konstam and others impressed on us the excellent results of outpatient ambulant chemotherapy with or without spontaneous fusion.26 My personal experience supports this latter view. There are two reports on spontaneous fusion.14,26 One report said that chemotherapy alone brought about the stability of the affected segment through spontaneous fusion within 5 years and the other report said that in the untreated children with spinal tuberculosis spontaneous bony ankylosis took a longer time, average 9 years, during which period more severe deformity developed with luxation.

M.-S. Moon

Development of surgical management In the era of medical art, Menard in 1895 decompressed an abscess surrounding the spinal cord and was delighted to find that the patient recovered neurologically.27 Also two surgical procedures (posterior fusion introduced by Hibbs and Albee in 1911, and anterior surgery by Ito et al. in 1934)1,2,5 could not bring about successful results, because the two operations were fraught with danger at every stage due to lack of specific antituberculous chemotherapeutic agents (Fig. 1). Mukopadhya in 1956 was probably the first orthopaedic surgeon to recognize the usefulness of chemotherapy for skeletal tuberculosis and this reduced the numbers of operations performed.11 Thanks to the advent of chemotherapeutics in the 1940s, the surgical procedures started becoming extremely successful.8,10,12 Many new surgical techniques were introduced not only to treat the disease, but also to treat the complications of the disease in a safer and more effective manner.8,13,24,28

Incisional drainage of abscess In 1672 Wiseman used a scalpel for abscess drainage which was the first record in the world medical literature.16 However, abscess drainage has been known to be practised by the barber surgeons or folk medicine practitioners in many countries since the pre-medieval era. Old dictum of where there is pus, let it out, is only acceptable when it is associated with pressure symptoms. So it is good surgical practice to undertake aseptic evacuation of an abscess. However, it must be kept in mind that incisional drainage of a tuberculous abscess results in continuous wetting of the dressing for a much longer period, much to the discomfort of the patient. According to a literary review, in 1570 Dalechamps clarified the relationship of spine caries and lower limb paralysis, but he did not mention the solution for the limb paralysis.16 In 1895 Menard decompressed an abscess surrounding the spinal cord and was delighted to find that the patient recovered neurologically.27 This led him and other surgeons to decompress the spinal cord through a variety of posterolateral and anterolateral approaches. At that time even if the patients had aseptic surgery under current anaesthesia, some patients still showed further destruction of bone, paralysis and spread of the disease.15 This happened in the absence of chemotherapeutic agents.

Posterior spinal fusion (extrafocal bone grafting alone) This procedure was introduced at the turn of the 20th century by Hibbs and Albee in 1911, drawing the surgeons’ attention. However, it was shown quite clearly in 1937 by McKee7 and in 1938 by Seddon8 that the operation was actually harmful if performed in patients in whom the disease was still active and that the surgery conferred no demonstrable benefit upon the patients in whom the disease had healed. In spite of this information the operation was

ARTICLE IN PRESS Development in the management of tuberculosis of the spine

137

Ito and Tsuchiya (1934) Incisional drainage (Wiseman, 1672)

Hodgson & Stock (1956) Costotransversectomy (Menard, 1895)

Insterspinous wiring + fusion or cementation (Moon, 1975) Passive posterior compression rod-hook system in children (Moon, 1978)

Posterior fusion (Hibbs, Albee, 1911)

Posterior circumduction fusion Harrington distraction rod + AIF (Moon, 1981)

Harrington rod (Moon, 1983)

Fig. 1 Surgical treatment of active spinal tuberculosis with mild to moderate non-rigid (flexible) kyphosis.

being performed in many parts of the world by surgeons whose faith in it was so great that they refused to submit the procedure to the test of a controlled trial until mid-1980s. Thus, posterior fusion is now a wholly discredited traditional operation and is no longer an acceptable method of treatment, particularly since it does not affect progression of the disease in many instances. It is now rarely used except to reinforce an anterior spine fusion at points of greatest stress, the cervicothoracic and thoracolumbar junction. In young children, longitudinal growth was little affected by fusion, because even after solid posterior fusion the posterior elements continue interstitial growth, while the anterior column does not grow if the growth plates of the vertebral bodies are already destroyed.28,29,30 In summary, thanks to the advent of chemotherapeutics in the 1940s, the controversies on this procedure had a natural death. Nowadays, there is no need for posterior spinal fusion for active spinal tuberculosis even though some die-hards still swear by it.

been widely taught that chemotherapeutics cannot satisfactorily enter a spinal tuberculous lesion from the blood stream.10 Stevenson in 1959 considered chemotherapeutics ineffective when their use was not supported by operation22 and Roaf et al. in 1959 considered that in the presence of a large abscess, chemotherapy was relatively ineffective.28 At the time, sadly, the effectiveness of the controlled chemotherapy was not appreciated by many surgeons until the MRC reports of 197314 and 1976 demonstrated the incorrectness of such opinions. Focal debridement alone is rarely indicated nowadays because evidence suggests it does not improve healing nor prevent the kyphosis from progressing. Extirpation of the infected focus is contraindicated in children as it may damage the remaining growth plates in children. Although the procedure is simple, it carries a small risk of complications such as dural tear, cord injury, and injury to nerve roots, autonomic and peripheral nerves.

Anterior surgery

Anterior radical surgery

Limited focal surgery (focal debridement) This procedure was conceived together with abscess evacuation to enhance the healing of the tuberculosis under chemotherapy. Florey in 1954 stated that the exclusion of chemotherapeutic agents from established tuberculous abscess prevents their influence in such cases and it has

In the late 1950s and early 1960s, the responsible opinions on the best method of treatment varied widely: it was a struggle between conservative versus surgical management of spinal tuberculosis. The concept of radical excision of the affected vertebral bodies and of their replacement by bone-grafts was

ARTICLE IN PRESS 138 practised by Ito et al. as early as 1934 in Japan following the development of anterior surgery by Burns in 1933.31 The procedure was not well accepted as a safe and effective procedure until 1956 when Hodgson and Stock in Hong Kong published the first account of this procedure.8 The Hong Kong surgeons claimed that radical surgery under the cover of chemotherapy gave excellent result in their hands and that surgery was the obvious answer. Anterior radical surgery (Hong Kong operation) was then advocated as the treatment of choice for spinal tuberculosis until mid 1970. It gradually became clear that the only advantage of anterior radical surgery was essentially deformity correction and to a certain extent prevention of deformity progression. The author’s experience suggests that because of graft failure, the procedure was not always successful in preventing kyphosis progressing and/or correcting. The procedure could not meet patients’ aesthetic demands. This drawback of the Hong Kong operation compelled some surgeons in the mid 1970s to develop a newer method to meet the patients’ aesthetic demands.13,21,24

Stabilizing spinal instrumentation surgery Andry in 1741 suggested the use of the splintage method to treat spinal deformity.32 But approximately 180 years passed before the development of the internal spinal fixator, the Harrington rod-hook system in 1949.33 Since early 1975 in Korea I and my associates tackled the clinical problem of spinal tuberculosis with kyphosis in order to solve the patients’ and their parents’ aesthetic demands.13

Posterior instrument stabilization As the first surgical attempt, posterior interspinous wiring and fusion method was used in children. This was followed by interspinous wiring and cementation to stabilise the involved segment and to arrest the posterior spinal element growth. Those two combined techniques failed because of the loosening of the wires and cement. The child’s spine was then fixed with passive posterior compression using a rod and hook system. Even this procedure did not succeed. Lastly the combined posterior instrumentation and anterior radical surgery was evolved by the present author during the later part of 1981.13 Harrington rod-hook system for adults and Rush nail or Steinmann pins for children’s spines were used.24 Since 1987 Zielke and MOSS systems were used to replace the Harrington system. After 1991 the Cotrel-Dubousset (CD) system was utilised.34,35 In patients with early spinal tuberculosis, posterior instrumentation alone could prevent the progression of kyphosis under the cover of chemotherapy.21

Combined posterior instrumentation and anterior radical surgery My first report in 1983 did not draw any surgeons’ attention. Rather it raised several controversies. Most were strongly against the use of metals in the infected spine. But I was confident of doing so based on the experience of Kim in

M.-S. Moon Korea who had carried out THR for tuberculosis of the hip since 1973 and reported the successful outcome in 1979.35 The fact that metal was harmless in the tuberculosis lesion was experimentally proven by Oga in 199336 in Japan and Ha et al. in 2004 in Korea.37 This was due to the characteristic behaviour of tuberculosis bacilli. They are usually present in planktonic form, have very slow division, non-production of adhesion molecule and slime, and less bacillary population in the bony lesion.

Anterior radical surgery and anterior instrumentation I performed this operation in a patient with a tuberculosis of L3 in 1987. Since Eysel et al.’s report in 1997 recording combined anterior radical surgery and anterior instrumentation for pyogenic spondylodiscitis, this procedure has been gradually gaining popularity. Without sound scientific basis I am very concerned about it.38 This procedure has been done endoscopically since 1998 with great success. The biocompatibility of the metal in the tuberculous lesion was clinically proven through THR surgery.38 However, the controversy still exists among the surgeons because the use of biomaterials in the infected lesions of the pyogenic origin are strongly opposed by most surgeons.34 Pyogenic bacteria have different bacterial behaviour from tuberculosis bacilli.

Deformity correction surgery Non-instrumented posterior fusion has been attempted to stabilise the affected segment and preventing progress of the kyphosis was anticipated. However, the procedure failed during the early 20th century.1,23 Hong Kong surgeons attempted the correction of mild and moderate kyphotic deformity by anterior radical surgery from 1957 but there were limitations.9 For severe kyphosis, O’Brien et al. in 1973 used the combined pre-surgical halo-pelvic traction and subsequent additional combined deformity correction and stabilization surgery with a good outcome.39 He utilised the halo system developed by Perry and Nickel40 and the halo-femoral system developed by Kane and Moe (1969).41 For the flexible moderate degree kyphosis in an active stage, the combined two stage operation of posterior instrumentation and anterior radical surgery was introduced by Moon et al. in 1983 with a good outcome.13,21,24 This procedure is now well accepted and has been used extensively all over the world. Anterior radical surgery and anterior instrumentation has also been practised since 1987. In comparing the two methods I prefer the former method because it provides more deformity correction than the latter. Transpedicular decancellation osteotomy and posterior closed wedge osteotomy of the spine are being practised for the healed tuberculous kyphosis. In future I am expecting to see the segmental vertebral column shortening deformity correction osteotomy. The total en bloc spondylectomy procedure and Ilizarov’s callotasis procedure after spinal osteotomy will be introduced if the safety of the procedure is proven.

ARTICLE IN PRESS 139

19 34 )

Development in the management of tuberculosis of the spine

Ito

&

Ts uc hi ya (

Passive posterior compression rod-hook system in children (1978)

Anterior Radical surgery Hodgson & Stock (1956)

Decancellation osteotomy (1990) Combined Pre-surgical halo-pelvic traction , + anterior fusion (O Brien, 1971)

Vertical column shortening ostectomy (2004)

Fig. 2

Correction of severe rigid tuberculosis kyphosis.

All the corrective procedures for the severe rigid kyphosis are rather challenging and dangerous with high complication rates. The patient and their families should be cautioned about the possible complications of paraplegia before opting for the procedure and informed consent should be obtained after the detailed explanation about the surgical procedure (Fig. 2).

Pott’s paraplegia The first surgical procedure for Pott’s paraplegia was done by Menard in 1895. He decompressed an abscess surrounding the spinal cord, and was delighted to find that the patient recovered neurologically. This led him and other surgeons to decompress the spinal cord through a variety of posterolateral and anterolateral approaches. However, it was unable to survive until the advent of chemotherapy, because it could not contribute to curing the disease and more frequently worsened the clinical course. Chemotherapy changed the outcome of the decompression surgery completely.42,43 In addition to the direct decompression surgery, it has been found recently that posterior instrumented stabilisation surgery under the cover of chemotherapy can indirectly decompress the cord and hasten the neurological recovery.

Conclusion In the end, may I impress upon you that spinal tuberculosis is essentially a medical condition. When the disease advances and causes complications/residual sequelae, surgical intervention does have a role. To be a good surgeon, you must be a good knowledgeable chemotherapist first. The problems not yet solved in management shall be our future task and

not the invention of new implants. To alleviate human suffering shall remain our priority. I sincerely hope that the messages that I have tried to convey to you may turn into food for thought in your daily practice.

References 1. Albee FH. Transplantation of a portion of the tibia into the spine for Pott’s disease. A preliminary report. J Am Med Assoc 1911;57:885–6. 2. Hibbs RA. An operation for Pott’s disease of the spine. J Am Med Assoc 1911;59:433–6. 3. Hibbs RA. Treatment of vertebral tuberculosis by fusion operation. J Am Med Assoc 1918;71:1372. 4. Albee FH. The bone–graft operation for tuberculosis of the spine. J Am Med Assoc 1930;94:1467–71. 5. Ito H, Asami T. A new radical operation for Pott’s disease. Report of ten cases. J Bone Jt Surg 1934;15:499. 6. Ito H, Tsuchiya T, et al. New radical operation for Pott’s disease. J Bone Jt Surg 1934;16:31. 7. McKee GK. A comparison of the results of spinal fixation operation and non-operative treatment in Pott’s disease in adults. Br J Surg 1937;24:456–68. 8. Seddon HJ. Pott’s paraplegia, prognosis and treatment. Br J Surg 1938;22:769. 9. Hodgson AR, Stock FE. Anterior spinal fusion. A preliminary communication on radical treatment of Pott’s disease and Pott’s paraplegia. Br J Surg 1956;44:266–75. 10. Florey HW. Chemotherapy of tuberculosis (Listerian oration). Can Med Assoc J 1954;71:47–421. 11. Mukopadhya B. The role of excisional surgery in the treatment of bone and joint tuberculosis. Ann Roy Coll Surgeons Engl 1956;18:288–313. 12. Chalke HD. The impact of tuberculosis on history. Literature and art. Med Hist 1962;6:301.

ARTICLE IN PRESS 140 13. Moon MS. Treatment of spinal infections. Commemorative issue. Third spinal section congress, WPOA J Western Pacific Orthop Assoc 1983; p. 7–11. 14. Medical Research Council Working Party on Tuberculosis of Spine. A controlled trial of ambulant outpatient and inpatient rest in bed in the management of tuberculosis of spine in young Korean patients on standard chemotherapy. J Bone Jt Surg 1973; 55B:675–97. 15. Keswani NH. Ancient Hindu orthopedic surgery. Indian J Orthop 1967;1:81. 16. Shanmugasundaram TK. Bone and joint tuberculosis. Madras: Kothandaram and Co.; 1983. 17. Shanmugasundaram TK. Current concepts in bone and joint tuberculosis. Documentation Center, Internat Bone and Joint Club; 1985. 18. Allison MJ, Mendoza J, Pezzia A. Documentation of a case of tuberculosis in pre-Columbian America. Am Rev Resp Dis 1973; 107:985–91. 19. Morse D. Prehistoric tuberculosis in America. Am Rev Resp Dis 1961;83:489–504. 20. Morse D. Tuberculosis. In: Brothwell DR, Sandison AT, Thomas CC, editors. Disease in antiquity. Springfield, IL, 1967. 21. Moon MS, et al. Harrington rods in treatment of active spinal tuberculosis with kyphosis. J Western Pacif Orthop Assoc 1986;23:53. 22. Stevenson FH. The chemotherapy of orthopedic tuberculosis. J Bone Jt Surg 1954;36B:5–22. 23. Moon MS. Combined posterior instrumentation and anterior interbody fusion for active tuberculosis kyphosis of the thoracolumbar spine. Curr Orthop 1991;5:177. 24. Moon MS, Woo YK, Lee KS, et al. Posterior instrumentation and anterior interbody fusion for tuberculosis kyphosis of dorsal and lumbar spine. Spine 1995;20:1910–6. 25. Upadhyay SS, Saji MJ, Sell P, et al. Longitudinal changes in spinal deformity after anterior spinal surgery for tuberculosis of the spine in adults. Spine 1994;19:542–9. 26. Konstam PG, Blesovsky A. The ambulant treatment of spinal tuberculosis. Br J Surg 1962;50:26–38. 27. Menard V. Etude practique sur de Mal de Pott. Paris: Masson et Cie; 1900. 28. Roaf R, Kirkaldy-Willis WH, Cathro AJM. Surgical treatment of bone and joint tuberculosis. Edinburgh and London: Livingstone; 1959.

M.-S. Moon 29. Schulitz KP, Kothe R, Leong JCY, et al. Growth changes of solidly fused kyphotic bloc after surgery for tuberculosis. Spine 1997;22:1150–5. 30. Kirkaldy-Willis WH, Thomas T. Anterior approaches in the diagnosis and treatment of infections of the vertebral bodies. J Bone Jt Surg 1965;47A:87. 31. Burns BH. An operation for spondylolisthesis. Lancet 1933;1: 1233. 32. Andry N. L’Orthopedic ou L’Art De Prevenir et De Corriger Dans Les Enfans Les Difformites Du Corps, 1741. 33. Harrington PR. Treatment of scoliosis: correction and internal fixation by spine instrumentation. J Bone Jt Surg 1962;44A: 591–610. 34. Moon MS. Managing tuberculosis of the spine. Med Progr 2004;31:593–602. 35. Kim YY, et al. Replacement arthroplasty using the charnley prosthesis in old tuberculosis of the hip. Int Orthop (SICOT) 1979;3:81. 36. Oga M, Arizono T, Takasita M, et al. Evaluation of the risk of instrumentation as a foreign body in spinal tuberculosis. Clinical and biological study. Spine 1993;18:1890–4. 37. Ha KY, Chung YG, Ryoo SJ. Adherence and biofilm formation of Staphylococcus epidermis and Mycobacterium tuberculosis on various spinal implants. Spine 2004;29(24):1–6. 38. Eysel P, Hopf C, Vagel T, et al. Primary stable anterior instrumentation or dorsoventral spondylodesis in spondylodiscitis? Eur Spine J 1997;6:152–7. 39. O’Brien JP, Hodgson AR, Smith TK, Yau ACMC. Halo–pelvic traction. A preliminary report of external skeletal fixation for correcting deformities and maintaining fixation of the spine. J Bone Jt Surg 1971;83B:217–9. 40. Perry J, Nickel VL. The halo in spinal abnormalities. Practical factors and avoidance of complication. Orthop Clin North Am 1972;3:69. 41. Kane J, Moe JH. Methods and techniques of evaluating idiopathic scoliosis. AAOS symposium on the spine. St. Louis: C.V. Morby; 1969. p. 196–240. 42. Roaf R. Anterolateral decompression for Pott’s paraplegia in India (1952–1953). In: Pott’s paraplegia. Griffths DL, Seddon HL and Roaf R 1956. p. 84–8 [Chapter 8]. 43. Moon MS, Moon JL, Moon YW, et al. Pott’s paraplegia in patient with severely deformed dorsal and dorsolumbar spine: treatment and progrosis. Spinal Cord 2003;41:164–71.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 141–151

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

UPPER LIMB

Chronic wrist pain: Diagnosis and management S. Ankarath Department of Orthopaedics and Hand Surgery, Huddersfield Royal Infirmary, Acre Street, Lindley, Huddersfield HD3 3EA, UK

KEYWORDS Wrist; Chronic pain; Carpus

Summary Management of chronic wrist pain can be a challenging problem. A sound understanding of the anatomy is required to make a diagnosis. Here the commonest causes of chronic wrist pain are discussed with an overview on the principles of management. & 2006 Elsevier Ltd. All rights reserved.

Introduction Chronic wrist pain has been described as the ‘bad back’ of hand surgery. The management of chronic wrist pain can only be helped by having a good understanding of the basic anatomy and mechanics of the wrist joint as well as the effect that any ligament injury might have upon the intercarpal relationships. The wrist is a complex joint with two rows of carpal bones articulating with each other, which in turn articulate with the radius and ulna. In addition to this, there is a complex relationship between the distal radius and ulna. Stability of the wrist joint is provided by ligaments which interconnect these bones. The ligaments of the wrist can be grouped as extrinsic and intrinsic. The extrinsic ligaments of the wrist connect the forearm bones to the carpal bones on the dorsal and the palmar sides. The palmar ligaments were thought to be arranged as two rows in an inverted V-shape, the proximal one connecting the radius and ulna to the lunate and the distal one linking the radius to the capitate.1 However, more recently, anatomical studies and the introduction of arthroscopy have revealed that there are two layers of ligaments Tel.: +44 148 342345; fax: +44 1484 342888.

E-mail address: [email protected]. 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.01.004

on the palmar side.2,3 The superficial palmar ligaments are the radioscaphocapitate (RSC), the long radiolunate and the ulnocapitate. The deep ligaments are the short radiolunate, the ulnolunate, the ulnotriquetral and the radioscapholunate. On the dorsal side there is mainly one ligamentous complex, which is the radiolunotriquetral ligament. The intrinsic ligaments of the wrist originate from and insert into the carpal bones in the same row (interosseous ligaments) and also link the proximal and distal carpal rows by crossing the midcarpal joint (midcarpal ligaments). The two main interosseous ligaments are the scapholunate and the lunotriquetral ligaments, which hold these respective bones together.

Causes of chronic wrist pain For the ease of understanding the pathology and making a clinical diagnosis, the various causes of wrist pain may be classified as radial, central/dorsal or ulnar (Table 1). This is only a broad classification as some pathology, such as scapholunate dissociation (SLD) or scapholunate advanced collapse (SLAC), may present as radial and/or dorsal pain. Radial-sided wrist pain Tenosynovitis. Stenosing tenosynovitis of the first dorsal compartment tendons (DeQuervain’s disease) is one of the

ARTICLE IN PRESS 142

Table 1

S. Ankarath

Main differential diagnosis of wrist pain.

Radial-sided wrist pain Tenosynovitis OA 1st CMCJ STT OA Scaphoid non-union Ganglion Dorsal/central wrist pain Kienbock’s disease Scapholunate dissociation Scapholunate advanced collapse (SLAC wrist) Intra-osseous ganglion Ganglion Ulnar-sided wrist pain Ulnar abutment syndrome Ulnar impaction syndrome Distal radioulnar joint degenerative arthritis/ instability Ulnar head chondromalacia Triangular fibrocartilage complex (TFCC) tear ECU tendonitis/subluxation Lunotriquetral instability Piso-triquetral joint pathology Midcarpal instability

commonest causes of radial-sided wrist pain. Fritz de Quervain has been credited with the description in 1895 of an entity involving the abductor pollicis longus (APL) and extensor pollicis brevis (EPB) tendons.4 Seen more frequently in women, pain is localised to the radial side of the wrist and is aggravated by thumb movements. Finkelstein’s test (pain on ulnar deviation of wrist with the thumb clasped in the palm) is diagnostic. It often responds to conservative treatment with non-steroidal anti-inflammatory medication (NSAIDs), a short period of immobilisation in a splint which supports both wrist and thumb, or a steroid injection. Surgical decompression should be limited to patients who fail to respond to at least 6–8 weeks of non-operative measures. The compartment of EPB may in some cases be separate from that of APL, and care should be taken to decompress both APL and EPB tendons. Also, the APL may have two or more tendon slips, all of which have to be released for full benefit. Superficial branches of the radial sensory nerves should be identified and protected by blunt dissection at the incision site to reduce the risk of neuroma formation and avoid any disabling symptoms of altered sensation over the dorso-radial aspect of the hand. Flexor carpi radialis tendonitis is an uncommon cause of radial–volar wrist pain. The cause may be a primary inflammation as a result of overuse or it may be secondary to the soft tissue and bony abnormality adjacent to the tendon. Scaphotrapeziotrapezoid (STT) joint arthritis may be an underlying cause as the tendon runs adjacent to this joint. Immobilisation in a splint, NSAIDs and steroid injections are often successful in resolution of symptoms. In refractory cases, surgical decompression of the tendon including the synovial tunnel through which it passes before

it dips dorsally to attach to the base of the index finger metacarpal, bordered by the scaphoid tuberosity, trapezial ridge and the transverse carpal ligament, may be required. Intersection syndrome. DeQuervain’s disease must be differentiated from Intersection syndrome. Although this is generally thought to be due to friction between the APL and EPB muscle bellies and the radial wrist extensors,5 it was later demonstrated to be tenosynovitis of the second dorsal compartment.6 Pain and swelling about 4 cm proximal to the wrist joint is characteristic of this problem. Non-operative treatment, which includes modification of activities, the use of a thermoplastic wrist splint in 151 extension, and steroid injection into the second dorsal compartment should be tried first. Should these measures fail, surgical release of the second dorsal compartment should be considered.6 Trapeziometacarpal osteoarthritis of the thumb. Exactly why osteoarthritis affects the basal joint of the thumb is poorly understood. Stability of the saddle-shaped joint surfaces is provided by small ligaments. Instability of this joint leading to excessive movements may result in osteoarthritis. Eaton and co-workers7 have staged basal joint arthritis as follows, based on radiological appearance: Stage 1: Articular contours normal. Slight widening of joint space due to effusion or ligamentous laxity. Stage 2: Slight narrowing of the trapeziometacarpal (TMC) joint with minimal sclerosis of the subchondral bone. STT joint is unaffected. Joint debris less than 2 mm. Stage 3: Marked narrowing of the TMC joint. STT joint not affected. Joint debris more than 2 mm. Stage 4: Identical to Stage 3 but with involvement of the STT joint. Patents complain of radial-sided wrist pain which may be diffuse and poorly localised. There may be a visible prominence on the volar side of the base of thumb at the TMC joint level. The Swanson grind test may be used to confirm the diagnosis (pain on circular movements of the thumb metacarpal with axial compression). It should be differentiated from DeQuervain’s tenosynovitis, where the pain and tenderness are located more proximally over the radial styloid. Radiographic assessment should include standard PA, lateral and oblique views of the thumb (Fig. 1). Stage 1 and 2 disease should be managed initially with conservative measures such as anti-inflammatory medications and a well moulded thumb spica splint. Stages 3 or 4 also may respond at least partially to the non-operative measures before surgical intervention is considered. Various surgical procedures have been described in the literature. These can essentially be categorised into those that remove a portion of the metacarpal,8 those that remove a portion of the trapezium9 and those that remove whole of the trapezium.10 Reconstruction following removal of the whole of the trapezium may leave an empty space,11 or involve use of a spacer which can be biological,12 silicone13 or metal (Fig. 2). Davies et al. looked at comparative results and found no difference following trapeziectomy alone or with tendon interposition or ligament reconstruction.14 Total joint arthroplasty of the TMC joint is still in its infancy and has a high reported failure rate of the trapezial component.15

ARTICLE IN PRESS Chronic wrist pain: Diagnosis and management

Figure 1 Plain radiograph showing early trapeziometacarpal joint arthritis.

143

Figure 3 tis.

Plain radiographs showing isolated STT joint arthri-

Scaphotrapeziotrapezoidal (STT) joint osteoarthritis. Involvement of the STT joint is usually secondary to TMC joint osteoarthritis. Isolated involvement of the STT joint can be secondary to early rotatory subluxation of the scaphoid (RSS) (Fig. 3). Arthrodesis of the STT joint is generally accepted as the procedure of choice for the management of this problem.16

Figure 2 Silicone interpositional arthroplasty following trapeziectomy.

Scaphoid non-union. The scaphoid bone serves as a link between the proximal and distal carpal rows. It flexes on palmarflexion or radial deviation of the wrist and extends on dorsiflexion or ulnar deviation of wrist joint. Any loss in integrity of the scaphoid affects wrist motion and carpal alignment. Two main groups of blood vessels supply the scaphoid. The volar vessels, which are branches of the radial artery entering the distal tubercle, supply the distal 20–30% of the bone. The proximal 70–80% is supplied by branches of the radial artery entering through the foramina along the dorsal ridge.17 These vessels run from distal to proximal. Various anatomical studies have consistently demonstrated poor blood supply to the proximal pole. An interruption to the dorsal blood vessels may lead to ischaemia of the proximal end of the bone. An ulnar-directed force of sufficient magnitude on the radial side of the wrist with the wrist in 95–1001 of extension has consistently been shown to produce fractures of the scaphoid.18 Intrinsic forces in the wrist cause the proximal portion of the scaphoid to extend along with the lunate while the distal portion flexes, leading to the ‘humpback’ deformity (Fig. 4). This increases the risk of non-union or malunion. This

ARTICLE IN PRESS 144

Figure 4 Scaphoid non-union with collapse leading to ‘humpback’ deformity visualised in a CT scan.

extension of the lunate causes an adaptive carpal instability, which if left uncorrected may become static. Tenderness in the anatomical snuffbox and over the tuberosity are sensitive but not specific tests of scaphoid fracture. Plain radiographs are helpful in not only confirming a non-union, but also in assessing the other carpal bones for instability. Other imaging modalities such as computed tomography or magnetic resonance imaging can help in providing further assessment of the degree of collapse and deformity in the scaphoid. In most instances, non-unions of scaphoid fractures are treated by surgical means. If the non-union is stable with no radiological evidence of collapse, it may remain asymptomatic.19 In these instances, it is often an incidental finding when radiographs are obtained for an unrelated problem. However, Vender et al. in a retrospective review have demonstrated predictable progressive degenerative changes involving the radioscaphoid and midcarpal joints following untreated symptomatic fractures of the scaphoid20 (Fig. 5). Some have suggested operative stabilisation and grafting of all displaced scaphoid fractures before the onset of osteoarthritis.21 The goal of treatmentin patients presenting with degenerative arthritis of the wrist as a sequela of scaphoid nonunions—scaphoid non-union advanced collapse (SNAC wrist)— is to achieve a pain-free wrist. The options available include wrist denervation, limited arthrodesis by scaphoid excision and four-corner fusion, or full wrist arthrodesis. The first two have the advantage that they retain the movement of the wrist, but the symptomatic improvement may not be longstanding, necessitating conversion to wrist fusion later. Proximal row carpectomy (PRC) may also be considered, but the lunate fossa of the radius must be free of any involvement as the capitate will articulate with the lunate fosssa after removal of the proximal row of carpal bones. Scapholunate dissociation (SLD). The ligaments of the wrist fail in tension. The position of the wrist at the time of an injury is critical in determining which ligaments might fail. Most wrist injuries occur with the wrist in extension. When the wrist moves from radial to ulnar deviation, the scaphoid

S. Ankarath

Figure 5 Scaphoid non-union advanced collapse (SNAC) type of degenerative arthritis following untreated non-united scaphoid fracture.

extends. This puts the scapholunate ligament under tension. Loading of the wrist in this position can lead to rupture of the scapholunate ligament. In radial deviation, the triquetral–hamate ligament is stretched. An extension injury in this position can lead to injury to the lunotriquetral ligament. Mayfield and Johnson, in their cadaveric study, have demonstrated a predictable pattern of distribution of force across the wrist on loading a wrist in extension.22 Injury to the scapholunate ligament leads to dissociation of this joint. Although the natural tendency of the scaphoid is to fall into flexion and pronation, in early stages the alignment of the scaphoid would be maintained by the intact ligaments connecting its distal pole to the distal carpal row (STT ligaments). With time, however, these distal ligaments weaken resulting in flexion and pronation of the scaphoid. This is termed RSS and appears radiologically as widening of the scapholunate gap (Fig. 6). It is important to appreciate that the gap seen on the plain radiographs is not due to a pure radial movement of the scaphoid, but is a result of rotational displacement. Hence any surgical procedure to address this problem should aim to lift the distal pole of the scaphoid. Initially, SLD can be dynamic and may only be identified on PA radiographs by asking the patient to clench the fist (clenched fist view). In long-standing cases the instability becomes static. With time the capitate migrates proximally into the scapholunate gap. SLD leads to abnormal loading between the proximal pole of the scaphoid and the dorsal rim of the radius. This results in the development of degenerative changes in the radioscaphoid joint, which may progress to involve the midcarpal level. The symptoms of SLD can vary depending on the severity and the time after the injury. Pain on the radial side of the wrist, weakness of grip and swelling localised to the scapholunate area are often the main complaints. Pain may be associated with a clicking sensation. Patients often have tenderness localised to the scapholunate area dorsally or over the anatomical snuffbox. A positive shift test, described by Watson23 is diagnostic of SLD. This is performed with the examiner and patient facing each other, as for arm

ARTICLE IN PRESS Chronic wrist pain: Diagnosis and management

145 Arthroscopy is most accurate and is useful in directly visualising the ligament. Scapholunate advanced collapse (SLAC). The scaphoid acts as a link between the proximal and the distal carpal rows. If this link is broken either by fracture of the scaphoid or rupture of the scapholunate ligament, then the lunate and the triquetrum collapse into a more stable position of extension (dorsal intercalated segmental instability [DISI]). This position is not physiological, which in turn increases the force transferred through the wrist cartilage leading to early degenerative changes of the SLAC (wrist) type. SLAC wrists can be graded depending on the extent of the degenerative changes: Stage 1a: Degenerative changes between the scaphoid and the radial styloid. Stage 1b: Degenerative changes involving the whole of the radioscaphoid joint. Stage 2: Degenerative arthritis involving the scaphocapitate (midcarpal) joint.

Figure 6 PA radiograph showing increased scapholunate gap after injury to the scapholunate ligament (‘Terry Thomas’ sign).

wrestling. The examiner’s fingers are placed dorsally on the distal radius, while the thumb is placed on the palmar distal tuberosity of the scaphoid. The examiner’s other hand holds the metacarpals. Firm pressure is applied to the palmar tuberosity of the scaphoid while the wrist is moved from ulnar to radial deviation. This moves the scaphoid from a position of extension to flexion. The examiner’s thumb placed over the distal tuberosity prevents the scaphoid from flexing. In cases of scapholunate ligament tears or in patients with a lax wrist, this resistance to flexion of the distal pole will cause the scaphoid to move dorsally under the posterior margin of the radius, which can be felt by the examiner’s index finger, normally inducing pain. Sometimes this test may only be painful, without any perception of dorsal scaphoid displacement. When pressure over the distal tuberosity of the scaphoid is removed the scaphoid may go back into position with a clunk. Even though the Watson’s test is the best known for scaphoid subluxation, its sensitivity and specificity are low. Increased scapholunate joint space of more than 5 mm on plain radiographs is considered to be diagnostic of SLD (the Terry Thomas sign). With extreme flexion of the scaphoid, in a PA view of the wrist, the scaphoid tubercle may project in the form of a dense circle or ring over the distal two-thirds of scaphoid (the signet ring sign). Cine-radiography is often helpful in demonstrating abnormal movements between the scaphoid and the adjacent carpal bones. Arthrography or MR-arthrograms may help in further defining tears of the ligament; however, their use in partial tears is often limited.

The most common cause of a SLAC wrist is RSS. Other causes include Preiser’s disease, Kienbock’s disease, and midcarpal instability, or it can be post-traumatic, following injuries involving the radioscaphoid or capitolunate joints. Diagnosis can be confirmed on a plain radiograph (Fig. 7(a) and (b)). For chronic SLDs associated with RSS that are reducible, a dorsal capsulodesis, popularised by Blatt, is recommended.24 This has been modified by various others. The most popular technique is that of Brunelli, using a strip of the flexor carpi radialis tendon.25 The underlying principle of both is to lift the distal end of the scaphoid, thereby correcting the flexion and pronation deformity. In irreducible scapholunate dissociations, conservative treatment with NSAIDs, appropriate splints and steroid injection should be tried first before considering operative intervention. Chronic irreducible SLDs without secondary degenerative arthritis of the radiocarpal or midcarpal joints often need some form of bony procedure. Selective fusion of only the involved joints rather than the whole wrist will help to retain reasonable function. STT arthrodesis, also known as triscaphoid fusion, holds the scaphoid distally and helps to maintain its correct alignment. Other procedures such as scaphocapitate fusion or scapholunocapitate fusion (Fig. 8) have been shown to lead to more reduction of wrist movements. In cases with significant degenerative arthritis at the radiocarpal level, scaphoid excision and fusion of the capitate, hamate, triquetrum and lunate (four-corner fusion) may be the best option26 (Fig. 9). This retains some movement of the wrist at the radiocarpal level and helps to improve the discomfort. For this to be a success the articular cartilage of the radiolunate joint must be intact. For those individuals with involvement of the radiolunate joint, total wrist fusion would be the procedure of choice (Fig. 10). Dorsal/central wrist pain Kienbock’s disease. Peste, in 1843, first described collapse of the lunate which he though was as a result of acute trauma . Kienbock later postulated that this was the result of repeated minor injuries resulting in reduced blood supply to the lunate.

ARTICLE IN PRESS 146

S. Ankarath

Figure 8 Scaphocapitate fusion following scapholunate dissociation.

Figure 7 (a) and (b) PA and lateral view showing scapholunate advanced collapse (SLAC) type of degenerative arthritis following long-standing scapholunate dissociation. Note the dorsal tilt of lunate (dorsal intercalated segmental instability) in the lateral view suggesting secondary adaptive carpal instability.

The exact aetiology of this condition remains unknown. Many authors have suggested trauma as a cause. It is likely that repeated trauma is more often the cause than a single one. Uneven loading of the radiocarpal joint has been suggested as another likely cause. Palmer and Werner noted that in neutral ulnar variance 82% of the load though the wrist joint is taken by the radiocarpal joint. This increases to 96% when there is a shortening of the ulna by 2.5 mm (negative ulnar variance).27 However, Kienbock’s disease can also occur in ulnar positive as well as ulnar neutral variance. D’Hoore found no difference in ulnar variance between a group of patients with Kienbock’s disease and an age and sex matched control group.28 Staging is based on the radiological extent of the disease, as follows: Stage I: No changes in plain radiographs, but MRI will show deceased signal within the lunate.

Stage II: Increased density of the lunate in plain radiographs. Height of lunate maintained with no collapse. Stage III: There is collapse of lunate with loss of carpal height but the scapholunate relationship is maintained. Stage IIIB: Collapse of the lunate and loss of carpal height is associated with RSS. Stage IV: Generalised degenerative arthritis of the carpus asscociated with fixed RSS. Patients may complain only of dorsal and central wrist pain in the early stages. Collapse of lunate architecture is usually associated with restriction of wrist movements. For very early cases simple immobilisation of the wrist in a splint or POP cast may be of help in improving symptoms. Unloading of the lunate can be achieved by joint levelling procedures, such as shortening the radius or lengthening the ulna in those with negative ulnar variance or by limited intercarpal fusion of the STT or scaphocapitate joints. Although excision and replacement of the lunate has been reported, the long-term results of replacement arthroplasty are not encouraging and it is generally not recommended. Hori et al. suggested revascularisation of the lunate by direct implantation of the 2nd or 3rd metacarpal artery and vein into the lunate.29 More recently Moran et al have reported good results using the 4th–5th extensor compartmental artery.30 Wrist denervation may be helpful in controlling the symptoms. Other salvage procedures in advanced cases include proximal row carpectomy or wrist arthrodesis. Ganglion cysts. Ganglions are the most common soft tissue swellings of the wrist region. They are usually outpouchings of the capsule from the carpal joints, but may also arise in relation to the tendons. The commonest site of origin is from

ARTICLE IN PRESS Chronic wrist pain: Diagnosis and management

147

Figure 9 (a) and (b) Post-operative radiographs after scaphoid excision and four-corner fusion for scapholunate advanced collapse using a circular plate.

pain. The ulnar head, triangular fibrocartilage complex (TFCC), extensor carpi ulnaris (ECU) tendon or the intercarpal joints can all be possible causes (Table 1).

Figure 10

Total wrist fusion using a contoured titanium plate.

the scapholunate joint. There is almost always a connection to the underlying joint or tendon sheath which makes it difficult for these swellings to be removed completely on surgical excision. Patients often present with a painless swelling but in some cases they may complain only of wrist pain, with no visible swelling. This is particularly likely in early stages when the ganglion is small and not visible. An arthroscopy at this stage may help to identify a small ganglion as the cause of discomfort and an arthroscopic excision can be performed. As the swelling enlarges and becomes more clinically obvious, the pain may improve. Management options include reassurance, aspiration or surgical excision. Diaz and Buch, looking at their results of palmar ganglions, have found no difference in symptoms at 2- and 5-year follow-up regardless of whether the ganglions were excised, aspirated or left alone.31

Ulnar-sided wrist pain The ulnar side of the wrist joint is more complex than the radial side. This anatomical complexity often makes pain on this side more difficult to diagnose than a radial-sided wrist

Ulnar impaction syndrome (ulnocarpal abutment). With the forearm in supination, the radius lies alongside and parallel to the ulna. On pronation of the forearm, the distal radius moves around the ulna with the radial sigmoid fossa articulating with the ulnar head. This results in a relative lengthening of ulna. In people with a positive ulnar variance, this over lengthening of the ulna can cause abutment on the lunate, resulting in pain and discomfort. This happens more often with the wrist in slight ulnar deviation. The TFCC can get caught between the lunate and ulna. Over a period of time this may result in a central perforation of the TFCC. This degenerative tear of the TFCC is different to the traumatic variety32 (Table 2). Patients often complain of a catching sensation on forearm rotation with the wrist in ulnar deviation. Plain radiographs may show a positive ulnar variance but ulnar impaction can be associated with neutral ulnar variance as well. It is important to obtain radiographs of the wrist with the forearm in the neutral position and the shoulder and elbow at right angles, as the relative length of the ulna can vary depending on forearm rotation. MRI may show evidence of bone oedema on the ulnar aspect of the lunate. Arthroscopy will help confirm a TFCC tear and also can be helpful in debriding the tear. Treatment depends on various factors such as ulnar variance, TFCC integrity, lunotriquetral ligament status, DRUJ congruency and stability, and patient symptomatology. Patients with neutral or only mildly positive ulnar variance may respond to conservative measures such as rest, NSAIDs, splinting and possibly steroid injections. Surgical intervention will be needed in those with excessive positive ulnar variance. Individuals with TFCC tears may need debridement, which can be done either by open or arthroscopic

ARTICLE IN PRESS 148

Table 2

S. Ankarath

Triangular fibrocartilage complex classification.

Class 1: Traumatic A. Central perforation B. Ulnar avulsion With distal ulnar fracture Without distal ulnar fracture C. Distal avulsion D. Radial avulsion With sigmoid notch fracture Without sigmoid notch fracture Class 2: Degenerative (ulnocarpal abutment syndrome) A. TFCC wear B. TFCC wear+lunate and/or ulnar chondromalacia C. TFCC perforation+lunate and/or ulnar chondromalacia D. TFCC perforation+lunate and/or ulnar chondromalacia+LT ligament perforation E. TFCC perforation+lunate and/or ulnar chondromalacia+LT ligament perforation+ulnocarpal arthritis

methods.33 In ulnar neutral or negative variance, this is all that may be needed. However, in cases of positive ulnar variance, ulnar length should be corrected as otherwise the symptoms may recur. This can be achieved either by an extra-articular ulnar shortening osteotomy or by a limited resection of the articular surface of ulnar head at the point of impaction with the carpus (Wafer procedure).34 The extra-articular shortening of the ulna remains the gold standard but Constantine et al. compared the results of the two procedures and noted equally good improvement of impaction symptoms in both groups, although the group having open ulnar shortening osteotomy needed more secondary procedures for hardware removal and nonunion.35 Ulnar impingement syndrome. Ulnar impingement has to be differentiated from an abutment. This occurs due to impingement of the distal ulna on the radius causing a painful pseudarthrosis.36 It is often iatrogenic following excision of the distal end of the ulna (Darrach’s procedure). On making a fist the unstable distal end of ulna converges on the radius causing pain and discomfort (Fig. 11(a) and (b)). It can also occur with growth arrests of the distal radius or ulna epiphysis. Patients complain of weakness of grip, pain over the DRUJ on forearm rotation and tenderness over the distal ulna. Radiographs may demonstrate a scalloping on the ulnar side of the distal radius (Fig. 11(b)). This problem can only be addressed by restoring the length of the ulna to provide stability to the TFCC and the ulnar sling mechanism. Watson and Brown have reported satisfactory outcome with an ulnar lengthening osteotomy in conjunction with a ‘matched’ resection of the advanced ulna.37 Ulnar head replacement is another satisfactory method of regaining ulnar length and achieving stability of the distal radioulnar joint38 (Fig. 12(a) and (b)) Distal radioulnar joint pathology. Stability of the DRUJ is provided by various structures such as the TFCC,39 ECU subsheath and pronator quadratus. Of these the TFCC is probably the most significant. Instabilty of the DRUJ can be post-traumatic, following an acute injury to these structures or it can be as a result of inflammatory arthritis. Ballotment

of the prominent ulnar head confirms the diagnosis (the piano key sign). Acute tears of the TFCC with DRUJ instability can initially be treated with a period of immobilisation for about 4–6 weeks. Persistent instability will often need operative stabilisation. Arthritis of the DRUJ may be post-traumatic (Fig. 13) or as a result of systemic disorders such as rheumatoid arthritis. Patients complain of weakness of grip and decreased range of motion, often associated with pain and clicking. Plain radiographs may show narrowing and irregularity of the joint and osteophytes. For early stages, conservative treatment by NSAIDs, steroid injection and activity modification may be tried. Failure of conservative measures may necessitate operative intervention to restore the congruity of the joint. Excision of the distal end of the ulna, which was first described in 1913 by Darrach40 and involves complete resection of the distal end of ulna, has a very poor longterm result in non-rheumatoid patients. Most individuals have weak grip strength and develop ulnar impingement syndrome. Watson described the technique of resection of the distal ulna to match the shape of the ulnar border of the distal radius.41 This has the advantage of retaining the length of the ulna, thus avoiding the problem of instability of the ulnar stump. Pisotriquetral osteoarthritis. Instability of the pisiform leads to eventual degenerative arthritis of the pisotriquetral joint with pain over the ulnar aspect of the hand. Provocative tests such as the pisotriquetral grind test may be positive. Associated ulnar neuropathy may be seen in up to a third of patients.42 Diagnosis can be confirmed on radiographs from a 301 supination view. If initial nonoperative treatment with splinting, NSAIDs and steroid injections fail to improve the symptoms, excision of the pisiform may have to be considered. Extensor carpi ulnaris (ECU) subluxation. Anatomical studies have shown that the ECU tendon, which runs in the 6th extensor compartment, is held in the ulnar groove by a subsheath which is distinct from the extensor retinaculum of the wrist.43 In the event of a tear of this subsheath, the ECU tendon is likely to become unstable and sublux out of the

ARTICLE IN PRESS Chronic wrist pain: Diagnosis and management

Figure 11 (a) and (b) Ulnar impingement syndrome.(a) Excision of distal end of ulna for DRUJ instability. (b) On radial deviation of wrist, the ulnar stump impinges on the radius (note the scalloping on the ulnar border of the radius).

149

Figure 12 (a) and (b) Ulnar head replacement for painful ulnar impingement. (a) Preoperative radiograph following resection of distal end of ulna for DRUJ instability following distal radius fracture. (b) Post-operative radiograph showing ulnar head replacement in situ, restoring DRUJ stability.

ARTICLE IN PRESS 150

S. Ankarath step-off between the lunate and triquetrum in the PA view. MR arthrography may confirm communication between the radiocarpal and midcarpal joints. Arthroscopy is the most sensitive investigation in confirming the diagnosis. Acute injuries may respond to immobilisation in a cast for 6 weeks. Chronic tears may need stabilisation by repair or reconstruction of the ligament or by lunotriquetral fusion.

References

Figure 13 Early degenerative arthritis of the distal radioulnar joint. Note associated fracture of the ulnar styloid.

ulnar groove. ECU subsheath tears may occur in forced supination of the forearm with the wrist in palmar flexion and ulnar deviation. Patients complain of painful snapping on the ulnar border of the wrist. Acute injuries may be treated in a long arm cast with the wrist in dorsiflexion and some radial deviation, with the foream pronated. In chronic cases repair of the subsheath and surgical stabilisation would be required. Lunotriquetral ligament instability. Tears of the lunotriquetral ligament can occur as a result of an acute injury or as a consequence of ulnar abutment (degenerative). Hyperextension and radial deviation of the wrist results in tightening of the ulnocarpal ligaments and a fall onto the wrist in this position can disrupt the stretched lunotriquetral ligament. This results in dissociative carpal instability. However, when there is disruption of both the intrinsic (lunotriquetral) and extrinsic (radiotriquetral) supporting ligaments, the scaphoid and lunate as a unit fall into flexion resulting in a volar intercalated segmental instability (VISI).44 Initially this will manifest only in dynamic radiographs, but with time it may become static. Patients may complain of a painful clunk on ulnar devation of the wrist. A positive ballottement test, as described by Reagan et al.,45 where the lunate is held firmly by the thumb and index finger of one hand while the pisotriquetral complex is moved dorsally and palmarly with the other, is diagnostic. Plain radiographs may be normal in partial tears. Evidence of static instability in radiographs suggests involvement of both the extrinsic and intrinsic ligaments. There may be a

1. Taleisnik J. The ligaments of the wrist. J Hand Surg [Am] 1976;1:110–8. 2. North ER, Thomas S. An anatomic guide for arthroscopic visualisation of the wrist capsular ligaments. J Hand Surg [Am] 1988;13:815–22. 3. Cooney WP, Dobyns JH, Linschield RL. Arthroscopy of the wrist: anatomy and classification of carpal instability. J Arthrosc Rel Surg 1990;6:133–40. 4. Leao L. DeQuervain’s disease: a clinical and anatomical study. J Bone Jt Surg 1958;40A:1063–70. 5. Wood MB, Linschield RL. Abductor pollicis longus bursitis. Clin Orthop 1973;93:293–6. 6. Grundberg AB, Reagan DS. Pathologic anatomy of the forearm: intersection syndrome. J Hand Surg 1985;10A:299–302. 7. Eaton RG, Glickel SZ. Trapeziometacarpal osteoarthritis: staging as a rationale for treatment. Hand Clin 1987;3:455–69. 8. Kessler I. Silicone arthroplasty of the trapeziometacarpal joint. J Bone Jt Surg [Br] 1973;55:285–91. 9. Eaton RG, Glickel SZ, Littler JW. Tendon interposition arthroplasty for degenerative arthritis of the trapeziometacarpal joint of the thumb. J Hand Surg [Am] 1985;10:645–54. 10. Burton RI. Basal joint arthrosis of the thumb. Orthop Clin North Am 1973;4:347–8. 11. Dell PC, Bushart TM, Smith RJ. Treatment of trapeziometacarpal arthritis: resection of resection arthroplasty. J Hand Surg [Am] 1978;3:243–9. 12. Carroll RE. Fascial arthroplasty for carpometacarpal joint of the thumb. Orthop Trans 1977;1:15–9. 13. Swanson AB. Disabling arthritis of the base of the thumb: treatment by resection of the trapezium and flexible (silicone) implant arthroplasty. J Bone Jt Surg [Am] 1972;54:456–71. 14. Davis TR, Brady O, Barton NJ, Lunn PG, Burke FD. Trapeziectomy alone, with tendon interposition or with ligament reconstruction? J Hand Surg [Br] 1997;22(6):689–94. 15. Wachtl SW, Guggenheim PR, Sennwald GR. Cemented and noncemented replacements of the trapeziometacarpal joint. J Bone Jt Surg [Br] 1998;80(1):121–5. 16. Watson HK, Ryu J, DiBella A. An approach to Kienbock’s disease: triscaphe arthrodesis. J Hand Surg [Am] 1985;10:179–87. 17. Gelberman RH, Menon J. The vascularity of the scaphoid bone. J Hand Surg 1980;5:508–13. 18. Weber ER, Chao EY. An experimental approach to the mechanism of scaphoid waist fractures. J Hand Surg [Am] 1978;3:142–8. 19. Mazet Jr R, Hohl M. Conservative treatment of old fractures of the carpal scaphoid. J Trauma 1961;1:115–27. 20. Vender MI, Watson HK, Wiener BD, Black DMJ. Degenerative change in symptomatic scaphoid nonunion. Hand Surg [Am] 1987;12(4):514–9. 21. Mack GR, Bosse MJ, Gelberman RH, Yu E. The natural history of scaphoid non-union. J Bone Jt Surg 1984;66A:504–9. 22. Mayfield JK, Johnson RP, Kilcoyne RK. Carpal dislocations: pathomechanics and progressive periulnar instability. J Hand Surg [Am] 1980;5:226–41. 23. Watson HK, Ashmead D 4th, Makhlouf MV. Examination of scaphoid. J Hand Surg [AM] 1988;13(5):657–660.

ARTICLE IN PRESS Chronic wrist pain: Diagnosis and management 24. Blatt G. Dorsal capsulodesis for rotatory subluxations of the scaphoid. In: Gelberman RH, editor. Master techniques in orthopaedic surgery: the wrist. New York: Raven Press; 1994. p. 147–66. 25. Brunelli GA, Brunelli GR. Rotatory subluxation of the scaphoidcorrection using the flexor carpi radialis. In: Watson HK, Weinzweig J, editors. The wrist. Philadelphia: Lippincott Williams & Wilkins; 2001. p. 879–84. 26. Watson HK, Belniak R, Garcia-Elias M. Treatment of scapholunate dissociation: preferred treatment. STT fusion vs other methods. Orthopaedics 1991;14:365–70. 27. Palmer AK, Werner FW. Biomechanics of the distal radioulnar joint. Clin Orthop 1984;187:26–35. 28. D’Hoore K, De Smet L, Verellen K, et al. Negative ulnar variance is not a risk factor for Kienbock’s disease. J Hand Surg [Am] 1994;19:229–31. 29. Hori Y, Tamai S, Okuda H, et al. Blood vessel transplantation to bone. J Hand Surg [Am] 1979;4:23–33. 30. Moran SL, Cooney WP, Berger RA, Bishop AT, Shin AY. J Hand Surg [Am] 2005;30(1):50–8. 31. Dias J, Buch K. Palmar wrist ganglion: does intervention improve outcome. J Hand Surg [Br] 2003;28B(2): 172–6. 32. Palmer A. Triangular fibrocartilage complex lesions: a classification. J Hand Surg [Am] 1989;14:594–606. 33. Bednar JM, Osterman AL. The role of arthroscopy in the treatment of triangular fibrocartilage injuries. Hand Clin 1994;10(4):605–14. 34. Feldon P, Terrono A, Belsky M. The ‘‘wafer’’ procedure. Clin Orthop 1992;275:124–9.

151 35. Constantine KJ, Tomaino MM, Herndon JH, Sotereanos DG. Comparison of ulnar shortening osteotomy and the wafer resection procedure as treatment for ulnar impaction syndrome. J Hand Surg [Am] 2000;25(1):55–60. 36. Bell MJ, Hill RJ, McMurtry RY. Ulnar impingement syndrome. J Bone Jt Surg [Br] 1985;67:126–9. 37. Watson HK, Brown RE. Ulnar impingement syndrome after Darrach procedure: treatment by advancement lengthening osteotomy of the ulna. J Hand Surg [Am] 1989;14(2.1):302–6. 38. van Schoonhoven J, Fernandez DL, Bowers WH, Herbert TJ. Salvage of failed resection arthroplasties of the distal radioulnar joint using a new ulnar head prosthesis. J Hand Surg [Am] 2000;25(3):438–46. 39. Palmer AK, Werner FW. Biomechanics of the distal radioulnar joint. Clin Orthop 1984;187:26–35. 40. Darrach W. Partial excision of the lower shaft of ulna for deformity following Colle’s fracture. Ann Surg 1913;57:764–5. 41. Watson HK, Ryu J, Burgess RC. Matched distal ulnar resection. J Hand Surg [Am] 1986;11:812–7. 42. Carroll RE, Coyle MP. Dysfunction of the pisotriquetral joint. Treatment by excision of the pisiform. J Hand Surg [Am] 1985;10:703–7. 43. Palmar AK, Skahen JR, Werner FW, et al. The extensor retinaculum of the wrist. An anatomical and biomechanical study. J Hand Surg [Br] 1985;10:11–6. 44. Hoii E, Garcia-Elias M, An KN, Bishop AT, Cooney WP, Linscheid RL, et al. A kinematic study of luno-triquetral dissociations. J Hand Surg 1991;16A:355–62. 45. Reagan DS, Linscheid RL, Dobyns JH. Lunotriquetral sprains. J Hand Surg [Am] 1984;9A:502–14.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 152–154

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

CME SECTION

Three CME points available The following series of questions are based on the CME designated article for this issue—’’Meniscal Tears’’ by I. McDermott. Please read the article carefully and then complete the self-assessment questionnaire by filling in the square corresponding to your response to each multiple-choice question. For true or false questions, please fill in one square only. After completing the questionnaire, either post or fax the answer page back to the Current Orthopaedics Editorial Office at the address at the bottom of the RESPONSE sheet. Please photocopy this page if you wish to keep your copy of Current Orthopaedics intact. Replies received before the next issue of Current Orthopaedics is published will be marked and those reaching an adequate standard will qualify for three external CME points. You will be notified of your marks and a CME certificate will be despatched for your records.

Questions 1. Approximately what proportion of arthroscopically proven meniscal tears arise with no history of injury? A. B. C. D. E.

10% 30% 50% 70% 90%

2. In an arthroscopic study, what proportion of patients undergoing ACL reconstruction were found to have associated lateral meniscal tears at the time of reconstruction? A. B. C. D. E.

10% 30% 50% 70% 90%

identified an anterior cruciate ligament rupture went on to develop a meniscal tear by the time they were admitted for anterior cruciate ligament reconstruction? A. B. C. D. E.

2% 5% 10% 20% 33%

5. If 100 patients with meniscal tears undergo an MRI scan of the affected knee, how may scans will be positive for meniscal tear? A. B. C. D. E.

80 85 90 95 100

6. What increase in peak contact pressure is observed in the knee joint after meniscectomy?

3. What is the least mobile segment of the menisci? A. B. C. D. E.

Posterior horn of medial meniscus Middle 1/3 of medial meniscus Anterior horn of medial meniscus Posterior horn of lateral meniscus Anterior horn of lateral meniscus

4. In a study by de Roeck, what proportion of patients who had no meniscal tear at an initial arthroscopy that

doi:10.1016/j.cuor.2006.02.005

A. B. C. D. E.

Up to 49% 50–99% 100–149% 150–199% 200–249%

7. What did Roos find to be the relative risk for the development of the radiological signs of osteoarthritis after meniscectomy ?

ARTICLE IN PRESS CME SECTION A. B. C. D. E.

4 7 14 21 28

8. What proportion of the compartment load is borne by the meniscus on the lateral side of the knee ? A. B. C. D. E.

10% 30% 50% 70% 90%

9. In a study of athletes 14.5 years after meniscectomy, what did Jorgensen find concerning activity levels? A. Most were still active participants in sport, though 46% had developed radiological signs of steoarthritis B. 89% were still active participants in sport, all having radiological signs of osteoarthritis C. 46% still participated in sport, 89% of these having radiological signs of arthritis D. 46% had reduced their level of participation and of these, 89% had radiological signs of arthritis E. 89% had radiological signs of arthritis, 46% had reduced their level of participation.

153 10. What is the reported cumulative survival for medial meniscal allografts at 5 years? A. B. C. D. E.

10% 30% 50% 70% 90%

11. Which of the following is not a technique that has been used to facilitate meniscal healing for tears of the inner, avascular portion of the meniscus ? A. B. C. D. E.

Trephination Radiofrequency shrinkage Rasping of parameniscal synovium Free synovial grafts Laser welding

12. What has been shown to be the strongest repair technique for the meniscus? A. B. C. D. E.

Vertical loop suture Horizontal sutures Mason Allen suture Mulberry suture Laser weld

ARTICLE IN PRESS 154

CME SECTION

Please fill in your answers to the CME questionnaire above in the response section provided below. A return address and fax number is given at the bottom of the page. ...............................................................................................

Responses Please shade in the square for the correct answer. 1 2 3 4 5 6 7 8 9 10 11 12

A A A A A A A A A A A A

& & & & & & & & & & & &

B B B B B B B B B B B B

& & & & & & & & & & & &

C C C C C C C C C C C C

& & & & & & & & & & & &

D D D D D D D D D D D D

& & & & & & & & & & & &

Your details (Print clearly) NAME. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ADDRESS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . FAX NO. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . EMAIL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . RETURN THE COMPLETED RESPONSE FORM by fax to +44-113-206-6791, or by post to CME, Current Orthopaedics, Orthopaedic Surgery, Clinical Sciences Building, St. James’s University Hospital, Beckett Street, Leeds LS9 7TF, UK.

E E E E E E E E E E E E

& & & & & & & & & & & &

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 155

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

CME SECTION

Answers to CME questions in Vol. 19, issue 6 Please find below the answers to the Current Orthopaedics CME questions from Vol. 19, issue 6 which were based on the article—‘Novel treatments for early osteoarthritis of the knee’ by S.P. Krishnan and J.A. Skinner. 1

A&

B&

C’

D&

E&

2

A&

B&

C&

D’

E&

3

A&

B&

C&

D’

E&

4

A’

B&

C&

D&

E&

5

A&

B&

C’

D&

E&

6

A&

B&

C&

D’

E&

7

A&

B&

C’

D&

E&

8

A’

B&

C&

D&

E&

9

A&

B’

C&

D&

E&

10 A &

B&

C&

D&

E’

11 A &

B&

C&

D’

E&

12 A &

B’

C&

D&

E&

0268-0890/$ - see front matter doi:10.1016/j.cuor.2005.10.007

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 156

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

ERRATUM

Erratum to ‘‘CME section’’ [Current Orthopaedics (2006) 20, 72–74] The publisher regrets that there was an error in Question 2 of the CME section. It should read ‘‘Forces of 3 N and 4 N’’ not ‘‘Forces of 3 N and 5 N’’. The entire question is correctly printed below. We apologise for any inconvenience caused. 2. Forces of 3 N and 4 N act at 901 to each other. What is the magnitude of the resultant force? A. 1 N B. 2 N C. 3 N D. 4 N E. 5 N

DOI of original article: 10.1016/j.cuor.2006.01.001 0268-0890/$ - see front matter & 2006 Published by Elsevier Ltd. doi:10.1016/j.cuor.2006.04.001

Aims and Scope Current Orthopaedics presents a unique collection of international review articles summarizing the current state of knowledge and research in orthopaedics. Each issue focuses on a specific topic, discussed in depth in a mini-symposium; other articles cover the areas of basic science, medicine, children/adults, trauma, imaging and historical review. There is also an annotation, self-assessment questions and an exam section. In this way, the entire postgraduate syllabus will be covered in a 4-year cycle. The Journal is cited in: Cochrane Center, EMBASE/ Excerpta Medica, Infomed, Reference Update and UMI Microfilms.

Editor Professor R. A. Dickson MA, ChM, FRCS, DSc St James’s University Hospital Trust, Leeds, UK

Editorial Committee President of BOTA, M. A. Farquharson-Roberts (Gosport, UK), I. Leslie (Bristol, UK), D. Limb (Leeds, UK), M. Macnicol (Edinburgh, UK), I. McDermott (Ruislip, UK), J. Rankine (Leeds, UK)

Editorial Advisory Board

Amsterdam

K

Boston

K

Jena

L. de Almeida (Portugal) G. P. Songcharoen (Thailand) R. W. Bucholz (USA) J. W. Frymoyer (USA) R. W. Gaines (USA) S. L. Weinstein (USA) M. Bumbasirevic (former Yugoslavia)

A. K. Mukherjee (India) A. Kusakabe (Japan) A. Uchida (Japan) M.-S. Moon (Korea) R. Castelein (The Netherlands) R. K. Marti (The Netherlands) G. Hooper (New Zealand) A. Thurston (New Zealand) E. G. Pasion (Philippines)

D. C. Davidson (Australia) J. Harris (Australia) S. Nade (Australia) G. R. Velloso (Brazil) J. H. Wedge (Canada) S. Santavirta (Finland) P. N. Soucacos (Greece) M. Torrens (Greece) J. C. Y. Leong (Hong Kong)

K

London

K

New York

K

Oxford

K

Paris

K

Philadelphia

K

San Diego

K

St Louis

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 157–161

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: REVISION HIP ARTHROPLASTY

(i) International epidemiology of revision THR Michael Skutek, Robert B. Bourne, Steven J. MacDonald London Health Sciences Centre, 339 Windermere Road¸ London, Ont., Canada N6A 5A5

KEYWORDS Revision hip arthroplasty epidemiology

Summary The worldwide number of primary total hip replacements has increased over the last decade, and as the number of primary total hip arthroplasties (THAs) increases each year, the number of revision THAs can also be expected to increase. Revision THAs often lead to suboptimal results, more complications and high costs. The rapid growth of new surgical techniques and new hip implants warrants a continuous and objective monitoring of the results to lower revision rates and improve outcomes. From an international standpoint, the revision burden is one of the possible key figures that allow crude comparisons between different countries and health systems. Evidence based information and revision rates can be derived from national joint registry data. In countries without registries, data can be collected from large data sets used for billing, national hospital discharge surveys and quality surveys. In addition single- and multicentre data regarding revision THA in combination with randomised control trials (RCTs) can be used. However, certain strengths and weaknesses apply to non-registry data sources which will be discussed. It was shown that the overall revision burden which is as high as 17.5% in some countries, could be improved in countries with national joint registries. One way to lower revision rates in the future is to use evidence-based primary implants. Registry data have the power to identify such implants, surgical techniques and processes. Thus it can guide surgeons in an evidence-based fashion to the benefit of patient outcome. & 2006 Elsevier Ltd. All rights reserved.

Introduction Total joint arthroplasty (THA) surgery is widely recognised as one of the most cost-effective interventions in medicine today. The number of both primary and revision surgeries increased steadily over the last decade, and as the number of Corresponding author. Tel.: +519 663 3512; fax: +519 663 3789.

E-mail addresses: [email protected] (M. Skutek), [email protected] (R.B. Bourne), [email protected] (S.J. MacDonald). 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.06.012

primary arthroplasties increases each year, the number of revisions can also be expected to increase. For the US a rise of 137% over the next 25 years is estimated for revision THAs.1 Complications and multiple revisions often lead to suboptimal results and high costs. Costs are as high as $15,000 to $30,000 per revision. From an international standpoint, the revision burden is one of the possible key figures that will enable basic comparison between different countries and health systems. Despite continuous improvements in implant technology, the revision burden—defined as the ratio of revision to the total number of arthroplasties— has remained relatively constant in the US.1

ARTICLE IN PRESS 158 Worldwide there are different data sources available that are used to calculate crude annual revision rates. Nonregistry data sources include large data sets used for billing, national hospital discharge surveys (NHDSs) and quality surveys. In addition, single and multicentre data combined with randomised control trials (RCTs) can provide data regarding revision THA. In general different inclusion and exclusion criteria influence the power of these data sets. Also, many national reports regarding the results of total hip replacements are based on series from centres with a high annual volume of such procedures. Katz demonstrated, that outcome for low volume centres, however were worse than at high volume centres with higher revision rates.2,3 More comprehensive information regarding patient characteristics, prosthesis type and features, method of prosthesis fixation and surgical technique, however can be derived from national joint registries. Although there might be differences in the inclusion criteria such as age, end point definition or inclusion of late complications, registries provide high quality, evidence-based data. Many countries have used joint registries for several years. Other countries recently have started or are in the planning phase.4 The validity of these data is persistently improving. Computer network technologies enable doctors and medical staff to enter data on line and to make adjustments e.g. for death as it is done in Sweden. These prospectively collected data are more accurate compared to previous reports which were collected retrospectively on large billing data sets requiring the use of assumptions and estimates. The improved failure definitions and accuracy in the epidemiological data will also facilitate comparisons and benchmarking among different national registries. However, there is a time gap until valid data regarding survival and revision surgery (type, percentage, etc.) can be gathered from new registries. The rapid growth of both new surgical techniques (i.e. minimally invasive hip arthroplasty) and new hip implant technology highlight the need for a continuous and objective monitoring of the results (post-market surveillance) plus means to distribute these data to key stakeholders. Despite rigorous testing prior to introduction of new devices, examples of early failures exist and therefore systems which allow for identification of premature failure benefit physicians, health systems and patients. Since the expenditure for total joint replacements is increasing in many countries over the last years, also from a financial standpoint it is more and more important to gather data regarding surgical outcome, indications for joint replacements, types of prostheses and techniques. The purpose of this manuscript is to discuss strengths and weaknesses from data regarding the epidemiology of revision THAs derived from different international data sources.

Data collection—sources of data Worldwide there are different existing data sources that can be used to calculate or estimate crude annual revision rates. The ratio of revision to the total number of arthroplasties (revision burden) gathered longitudinally is a general indicator of expected longevity of total hip replacements. If more specific information regarding type of fixation,

M. Skutek et al. implant design etc. is added to the data, more accurate information (e.g. survivorship of certain implants) can be derived. In addition, there are often different inclusion criteria such as age, end point definition or inclusion of late complications. These might increase or decrease the meaning of the results. In the following the strengths and weaknesses of different data collection methods such as large data sets used for billing, single centre data, RCT and national registries are listed.

Large data sets From countries where national registries are not yet implemented, data can be derived from other sources (i.e. billing data sets, NHDSs, national quality surveys), but this information has many problems (i.e. poor documentation of primary versus revisions, side, implant type). In a study performed by Kurtz, data was derived from the NHDs in the United States. It is an annual survey conducted by the National Center for Health Statistics. Started in 1965, this survey programme has continuously compiled a representative sample of hospitalisations at non-federal and nonmilitary short-stay community hospitals throughout the United States. Information collected by the NHDs includes patient demographics (e.g. age and gender), disease diagnosis, type of procedure performed, institutional characteristics, and resource utilisation. On the basis of the information collected by the survey and with use of the provided sampling weights, national and regional estimates of characteristics of patients and of surgical and nonsurgical procedures in hospitals with various numbers of beds and types of ownership can be estimated.5 In Germany, a national hospital quality survey has been carried out over the last years. It is thought to identify hospitals with unusually high complication rates and to discuss reasons for that. Also crude revision burden can be derived from this database. However two-stage revisions, late complications or implant related problems after the initial hospital stay are not or only partially recorded.

Single centre data Most reports of the results of THAs are based on series from centres with a high annual volume of such procedures, whereas the outcomes in low volume centres have received little study. Both greater hospital volume and greater surgeon volume have been associated with lower rates of mortality and/or complications following several surgical procedures. It has been shown by Katz, that a higher volume surgeon was significantly associated with a lower rate of dislocation (P value for trend ¼ 0.0001 and, slightly less strongly with a lower rate of deep infection P ¼ 0:03). Patients who had primary THAs in hospitals where more than 100 of these procedures were performed per year had a lower rate of mortality than those who had primary replacement in hospitals in which 10 or 20 procedures were performed per year (mortality rate 0.7% compared with 1.3%). These analyses of Medicare claims are limited by a lack of key clinical information such as preoperative comorbidity, functional status and operative details.2 Nevertheless, it is likely that single centre data would reveal more

ARTICLE IN PRESS International epidemiology of revision THR biased results in terms of stating lower complication and revision rates as opposed to low volume community centres.

Randomised control trials (RCTs) In general randomised blinded control trials provide excellent data compared to retrospective studies with historical controls. Potential advantages of RCTs are the unbiased distribution of confounders and the statistical design of these studies (Level 1 evidence). However, these studies are expensive in terms of time consumption and money and ethically problems do sometimes arise. Since studies are likely to be carried out in high volume centres with experienced surgeons who have relatively low revision rates, the results do not necessarily reflect the nationwide surgical practise and are limited by nature to a small number of different fixation techniques and implant designs.

National registries National joint registries define the epidemiology of primary and revision surgery. The primary reason to document failures is to improve and redefine the primary indication, surgical technique and implant choice. It is thought that feedback of data stimulates the participating clinics to reflect and improve their health care accordingly (i.e. The Quality Cycle). Traditionally, revision THA has been used as an end point definition. In the future, however, patient based outcome measures and radiographic results are thought to improve sensitivity.4 In conjunction with individual, subjective, patient data and radiography, joint replacement registries contribute to the development of evidence-based THA. Many countries have utilised national joint replacement registries for several years (Australia, Canada, Denmark, Finland, Hungary, Norway, New Zealand, Sweden and Romania). Other countries have recently started or are in the planning phase of developing national joint replacement registries (Austria, Czech Republic, England, France, Germany, Moldova, Slovakia, Turkey, USA and Wales).4 In contrast, the Swedish hip register contains information on primary hip arthroplasties performed in Sweden since 1979. This information source has led to a stepwise reduction in annual crude revision rates, improvements in surgical centre performance and early removal of poorly functioning implants from use. In the 2003 report from the Swedish total hip joint register, the results were based on data for each primary procedure, captured since 1992, and adjustment for death was made on line, which was considered to be a major improvement compared with previous reports where part of the statistics were based on assumptions and estimates. The improved failure definitions and accuracy in the epidemiological data will also facilitate comparisons and benchmarking among different national registries.

Registry data So far a true comparison of crude revision rates from different countries is not possible without considerable bias. Even countries with established joint registries do have

159 different types of data collection and inclusion criteria (i.e. some are age-standardised, others not; some include population o20 years, others do not). The following data, are derived from national joint registries, alternate equivalent sources and from different years to give an overview (Table 1).4,6–12 In general the revision burden is considered as the proportion of hip replacements that were revisions. In Sweden, results are based on more than 90% all cemented THA. The revision burden for cemented implants is considerably lower than for uncemented implants (28.1% less 1992–2003) and even hybrid combinations (10.8% less 1992–2003). This is probably related to inferior acetabular fixation and polyethylene quality of the uncemented implants used in Sweden. As outlined before, data regarding the annual revision burden in the United States was calculated from the NHDs. Data were stratified by age and gender. Overall, when all data that were collected over the 13-year period were considered, a mean revision burden for THA of 17.5% (annual range 15.2–20.5%) was found for the period from 1990 to 2002. All patients regardless of age and gender were included in these data.13

Reasons for revisions Calculated from the Swedish registry, revision was the dominant subsequent procedure, accounting for 86% of all re-operations. The most common reason for revision surgery in this data pool was aseptic loosening (73.9%), followed by deep infection (7.9%), dislocation (7.5%) and periprosthetic fractures (5.7%).4 These data were similar in comparison to other countries (Australia, England/Wales).8,11,14 Although the true impact of joint registries is not exactly known, the rate of overall implant survival (including poor designs) has improved in Sweden from 89.4% to 92.5% between the two periods 1979–1991 and 1992–2003.4 Patients with index diagnoses of rheumatoid joint disease and sequel to childhood hip disease are overrepresented in the group of multiple revisions as are those revised due to deep infection, periprosthetic fractures and dislocation.4

Aseptic loosening trends In general the quality has improved in terms of fewer revisions because of aseptic loosening in Sweden. For Table 1

International revision burden.

Country (year)

Revisions (%)

Australia (2003–2004) Canada (2003–2004) Denmark (1998) England (2004) Germany (2004) Norway (2001) New Zealand (2004) Sweden (1992–2003) United States (1990–2002)

13.4 9.0 14.5 9.2 11.4 13.0 13.0 9.9 17.5

ARTICLE IN PRESS 160 cemented implants the results for the stem are generally better than the cup, with the flanged Charnley cup as the sole exception. In uncemented and hybrid implants, the stem results are generally good, whereas the cups show poorer results.4 This is in accordance with data from England, where cup revisions were more common than single stem revisions.

Infection trends One major contributor to revision surgery is prosthetic infection with an incidence of 0.2–4%. Resistance to antibiotics is a problem, too. It could be shown from the Norwegian arthroplasty register that an operating time of more than 150 min was associated with an increased risk of revision due to infection.15 In case of infection, it has been observed that the rate of recurrent infection is higher after one stage revision as opposed to two stage revision. In an investigation, performed in Sweden and the Netherlands, 864 revisions were performed for deep infection. A total of 245 were one-stage revisions, 455 were two-stage revisions and 164 ended up in permanent extractions. The average time in between stages was 115 days in Sweden. There were initially 93% cemented (27% with antibiotics), 2% uncemented and 5% hybrid THAs. There was a significant difference between one-stage versus two-stage revisions. Re-revision had to be performed in 29 (11.8%) of one-stage patients and in 33 (7.3%) of two-stage patients (w2: Po0.03). Thus, two stage revisions had a higher success rate in terms of eradication of infection. The risk for infection also was increased without the use of antibiotic bone cement and in male patients.16

Dislocation trends In Sweden, there is currently an increase of revisions due to dislocation and or technical reasons. For patients with 5 years follow up the cumulative revision rate is five to six times higher for the group operated on in 1998 compared to those operated on in 1984. This might be related to an increase in primary THA for displaced cervical femoral fractures in the elderly, which is in contrast to a long tradition of using percutaneous techniques with screws or pins as the primary intervention. Another explanation is the equally long tradition in Sweden with use of small head sizes (22 or 28 mm). The overall dislocation rate after revision THA in Germany is relatively low (2.72%, range 0.0–19.0%). However, in these data only early dislocations are recorded. Late complications after the initial hospital stay are not or only partially recorded.12 The implementation of a national joint registry collecting long-term data is planned, however, but not before the year 2008 (pers. communication Dr. Boy).12

Volume effects on outcomes of revision THA In the United States, the number of primary total hip replacements increased from 119.000 in 1990 to 193.000 in 2002, calculated from NHDS data. For this time period, a mean revision burden of 17.5% (range 15.2–20.5%) has been

M. Skutek et al. reported.1,17,18 Patients treated at hospitals and by surgeons with higher annual caseloads of primary and revision THA had lower rates of mortality and of selected complications.3,19 It is estimated that the cost of a routine revision THA was approximately $18,000. For a complex revision, the cost goes up to as high as $30,000. If the 2002 hip revision burden of 18.1% was reduced by 1% (a decrease of approximately 2844 hip revision procedures), the potential cost savings could range from $42.5 million to $ 112.6 million.1

Components utilised trends Over the last years, in some countries (e.g. Australia, Denmark), there is a trend to uncemented implants.8,9 Other countries predominantly use cemented fixation such England and Sweden, where 95% of the hip arthroplasties are cemented. In England, there is also an increasing trend to resurfacing in the group of patients o55 years. The high percentage of cemented implants in Sweden as opposed to other countries may be one reason for the comparatively lower revision rate, since a higher revision rate has been reported with first generation uncemented designs. Another reason for a lower revision rate in Sweden might be a positive effect on surgical quality, promoted many active years’ use of the national hip register.4,9

Epidemiology of components revised It is interesting to identify which components are being revised in revision THA. In England for example, more than 50% of revisions were of both femoral and acetabular components. However in those cases where only the stem or only the cup was revised, isolated cup revisions were done more often then single stem revisions (Table 2). The same trend applies for Australia, whereas the total number of stem and cup revisions is lower. Registry data demonstrate that the survivorship of implants in revision THA is less than in primary THA. In Norway, data (1987–2003) demonstrated a 26% failure rate at 10 years in 4762 revision THAs when infection was excluded as a cause for revision. Table 2 Percentage of revision hip surgery procedures in England and Australia. Procedure

Femoral head alone Femoral stem alone Acetabular cup or shell7liner7femoral head Acetabular cup or shell7liner and femoral stem7femoral head None/other

Australia (%)

England (%)

21.1 37.4

2.1 15.9 23.6

37.6

53.7

4.0

4.7

ARTICLE IN PRESS International epidemiology of revision THR

161

Conclusions The overall aims of a national joint replacement register is to pool data, recognise trends, improve the quality of THA and promote evidence-based surgical practise. The register can generate hypotheses suitable for either specific studies based on registry data or carefully planned prospective clinical studies. The failure end-point definition used in registries has traditionally been revision. The overall rate of complications and the need for revision total hip arthroplasty is influenced by the volume of cases performed at hospitals and by surgeons.2,20 It is important to clearly define and internationally agree on which key features should be presented in national registers in order to make comparisons unbiased. Different countries often have different perspectives (i.e. whether to use cemented or non-cemented stems) making comparisons difficult. An international Register Society could facilitate this process and there are ongoing efforts to initiate this.4

3.

4.

5.

6.

7.

8.

Perspectives/trends


Data derived from national joint registries is advanta












geous as opposed to data derived from large data sets and or single centre data. Registry data can contribute to lower revision rates. In Sweden, the proportion revised for the most common complication (aseptic loosening) has decreased to one third. The rate of implant survival might improve through the use of registry data. For example, the rate of implant survival has improved from 89.4% to 92.5% between the two periods 1979–1991 and 1992–2003 in Sweden. The rate of mortality and selected complications in high volume centres is lower as compared to low volume centres in the US. It is important to lower revision rates nationwide for the number of joint replacements is steadily increasing, so is the revision rate. As calculated for the United States, the number of THAs will increase 137% from 2005 to 2030. To meet future needs nationwide usage of joint registries and data evaluation might be a strong factor to identify potential sources for failure and to further lower revision rates.

9. 10. 11.

12. 13.

14.

15.

16. 17. 18. 19.

References 1. Kurtz S, Lau E, Zhao K, Mowat F, Ong K, Halpern M. The future burden of hip and knee revisions: US projections from 2005 to 2030. 2006. Scientific Exhibit No. SE53, AAOS, Chicago, Exponent. 2. Katz JN, Losina E, Barrett J, et al. Association between hospital and surgeon procedure volume and outcomes of total hip

20.

replacement in the United States medicare population. J Bone Joint Surg Am 2001;83-A:1622–9. Katz JN, Phillips CB, Baron JA, et al. Association of hospital and surgeon volume of total hip replacement with functional status and satisfaction three years following surgery. Arthritis Rheum 2003;48:560–8. Malchau H, Garelick G, Herberts P. The evidence from the Swedish hip register. In: Breusch, Malchau, editors. The wellcemented total hip arthroplasty, theory and practice. Berlin: Springer; 2006. p. 290–7. Kurtz S, Mowat F, Ong K, Chan N, Lau E, Halpern M. Prevalence of primary and revision total hip and knee arthroplasty in the United States from 1990 through 2002. J Bone Joint Surg Am 2005;87:1487–97. Furnes O, Havelin LI, Espehaug B, Engesaeter LB, Lie SA, Vollset SE. [The Norwegian registry of joint prostheses—15 beneficial years for both the patients and the health care]. Tidsskr Nor Laegeforen 2003;123:1367–9. Cyrus D, Gopinath S. Canadian Joint Replacement Registry Analytic Bulletin, June 2004. Revisions of Hip and Knee Replacements in Canada. 2004. Toronto, CIHI 2004. Graves S, Davidson D, Ingerson L. Australian orthopaedic association national joint replacement registry. Annual report. Adelaide: AOA; 2004. Lucht U. The Danish hip arthroplasty register. Acta Orthop Scand 2000;71:433–9. Rothwell A, Hobbs T. NZ National Joint Registry. Christchurch; 2005. van der Meulen J, Lewsey J, Hardoon S. National Joint Registry for England and Wales. 2nd Annual Report. Harwell: NJR Centre; 2005. Boy O, Dabisch I. Kapitel 11: Hu ¨ft-Totalendoprothesen-Wechsel; 2005. Kurtz S, Mowat F, Ong K, Chan N, Lau E, Halpern M. Prevalence of primary and revision total hip and knee arthroplasty in the United States from 1990 through 2002. J Bone Joint Surg Am 2005;87:1487–97. van der MEulen J, Lewsey J, Hardoon S. National Joint Registry for England and Wales. 2nd Annual Report. Harwell: NJR Centre; 2005. Smabrekke A, Espehaug B, Havelin LI, Furnes O. Operating time and survival of primary total hip replacements: an analysis of 31,745 primary cemented and uncemented total hip replacements from local hospitals reported to the Norwegian Arthroplasty Register 1987–2001. Acta Orthop Scand 2004;75:524–32. Malchau H. Revision THR: Lessons learned from national hip registries; 2004 Weinstein JN, Birkmeyer JD. The Dartmouth atlas of musculoskeletal health care. Chicago: AHA Press; 2000. Maloney WJ. National joint registries: has the time come? J Bone Joint Surg Am 2001;83:1582–5. Espehaug B, Havelin LI, Engesaeter LB, Vollset SE. The effect of hospital-type and operating volume on the survival of hip replacements. A review of 39,505 primary total hip replacements reported to the Norwegian Arthroplasty Register, 1988–1996. Acta Orthop Scand 1999;70:12–8. Smabrekke A, Espehaug B, Havelin LI, Furnes O. Operating time and survival of primary total hip replacements: an analysis of 31,745 primary cemented and uncemented total hip replacements from local hospitals reported to the Norwegian Arthroplasty Register 1987–2001. Acta Orthop Scand 2004;75:524–32.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 162–170

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: REVISION HIP ARTHROPLASTY

(ii) Current techniques and new developments in acetabular revision surgery Alexander W.R. Burns, Richard W. McCalden London Health Sciences Centre, University of Western Ontario, 339 Windermere Road, London, Ontario, Canada N6A 5A5

KEYWORDS Revision total hip arthroplasty; Acetabular; Reconstruction; Trabecular metal; Tantalum

Summary Revision of a failed acetabular component is one of the most challenging aspects of revision hip arthroplasty. The revision hip surgeon must have a systematic approach to preoperative, operative and post-operative management. The majority of acetabular revisions can be performed using uncemented ‘‘jumbo’’ components, however severe bone loss can make reconstruction difficult. An advanced skill set including practical knowledge of extensile exposures, special techniques in removal of components, management of bone defects, and reimplantation of revision components, is essential. The various surgical options available to the revision surgeon are discussed in this article with particular focus on new techniques, instruments, materials and prostheses which may make this challenging area less complex to manage and may improve outcomes for patients in the future. & 2006 Published by Elsevier Ltd.

Introduction Acetabular revision is perhaps the most challenging facet of revision hip reconstruction for the arthroplasty surgeon. Exposure may be difficult due to obliteration of the normal tissue planes from previous surgery and distortion of the anatomy. In addition acetabular bone stock may be grossly deficient leading to difficulties both with obtaining acetabular fixation with sufficient host bone contact and reproducing the hip centre, leg length and joint stability. Fractures of the acetabulum may be present or occur Corresponding author. Tel.: +1 519 663 5128; fax: +1 519 663 3780. E-mail addresses: [email protected] (A.W.R. Burns), [email protected] (R.W. McCalden).

0268-0890/$ - see front matter & 2006 Published by Elsevier Ltd. doi:10.1016/j.cuor.2006.06.008

intraoperatively and neurovascular structures lie within close proximity. It is therefore necessary for the revision arthroplasty surgeon to be mindful of possible intraoperative eventualities and to be armed with the necessary techniques, skills and inventory to manage them. Meticulous preoperative planning is required to ensure a satisfactory outcome.

Preoperative planning The surgeon must know the details of the implanted components including prosthetic sizes, femoral trunion size, acetabular locking mechanism and the available liner options. Bone defects should be estimated and bone graft (morsellised or structural) should be available. In this regard, applying a classification system of acetabular bone

ARTICLE IN PRESS Current techniques and new developments defects may aid to guide management. Inventory should include instruments for extraction of the components and prostheses to re-implant. Intra-pelvic migration of cement or components may warrant a preoperative angiogram to diagnose involvement of vascular structures and consultation from a general or vascular surgical colleague for assistance with operative exposure may prove beneficial. The exclusion of infection, by use of blood tests and the judicious use of joint aspiration, is mandatory in all revision surgery. While no absolute guidelines for interpretation of inflammatory markers exist, however in the absence of another cause an ESR greater than 30 mm/h or a CRP greater than 10 warrant careful consideration of preoperative hip joint aspiration. The operative procedure can then be divided into 4 phases; exposure of the acetabulum, extraction of implanted components, assessment and management of acetabular bone defects and re-implantation of revision components.

Exposure Surgical approach is dependant on various factors such as the surgeons preference and experience, previous surgical approach and anticipated reconstructive challenges. For instance, if a pelvic discontinuity exists, the posterior column may require plating and then a posterior approach to the hip is preferable. In most cases, the approach with which the surgeon is most adept is best, although it is imperative that the chosen approach is extensile. Some form of trochanteric osteotomy (standard, sliding or extended) is often required to enhance either femoral or acetabular exposure. If revision of the femoral component is also required and a trochanteric osteotomy or variation is used, this may also improve access to the acetabulum.

163

Figure 1 Explant acetabular removal system. Short followed by long blades are used to break the bone–prothesis interface with minimal bone resection. Any residual soft tissue or membrane should then be debrided to allow visualisation of the rim of the acetabulum circumferentially such that estimation of bone defect can be made.

Assessing and managing acetabular bone defects Preoperatively plain X-ray, CT scanning and MRI scanning can be used to attempt to assess acetabular bone defects. In general the defects seen on imaging will underestimate the intraoperative status. The AAOS and Paprosky1 classifications of acetabular deficiencies are useful in guiding further surgical management. The remainder of this article will outline the options currently available for revision surgery of the acetabular component with a focus on new developments.

Uncemented acetabular revision Extraction of implanted components The surgeon must have revision instruments including osteotomes, gouges, drills and high-speed burrs. The interface between the cement and bone should be exposed circumferentially. For removal of a cemented acetabular component, the prosthesis should initially be disrupted from the cement mantle and then the cement should be carefully removed using gouges and osteotomes to disrupt the cement bone interface in a piecemeal fashion with the minimum of bone loss. For removal of an uncemented acetabular component, the polyethylene liner should be disengaged from the shell and any screws removed. A 6.5 mm screw inserted through a 4.5 mm drill hole in the polyethylene will disengaged the liner from the cup in most instances where a proprietary liner removal tool is not available. The bone prosthesis interface can be broken with a series of curved gouges or the use of more specialised instruments such as the Explant (Zimmer, Warsaw, Indiana). Use of the Explant device requires re-insertion of the polyethylene liner or suitable trial liner after acetabular screw removal has been performed (Fig. 1).

In our practice, the majority of acetabular revisions are performed using uncemented revision acetabular components when greater than 50% host bone contact is available for implantation. The use of these so called ‘‘jumbo’’ cups has become the workhorse of acetabular revision surgery. Excellent outcomes have been reported for the use of uncemented acetabular revision with 12–15-year survivorship of 81–96%.2–4 The complexity of acetabular revision is determined by the amount and location of acetabular bone defects present. It is not uncommon for patients to present with several years of relatively minimal discomfort with large osteolytic defects and or migration of the acetabular components. The Paprosky classification1 of acetabular deficiency describes the amount of host bone contact for implantation of the prosthesis and takes into account the amount of rim, dome and anterior and posterior columns available. In Paprosky type 1 defects, bone loss is minimal and can usually be treated using an uncemented acetabular component supported by multiple screws and morsellised bone

ARTICLE IN PRESS 164 graft. Type 2 defects have intact anterior and posterior columns but loss of medial wall, dome or rim. Type 2A defects are considered contained with an intact acetabular rim but loss of acetabular dome and can be treated with either morsellised graft or femoral head allograft. Type 2B defects are uncontained defects with loss of the superolateral rim and may require bulk allograft such as a number 7 graft. Type 2C defects have an absent medial wall and require either particulate graft or sliced femoral head graft. Type 3 defects describe greater than 2 cm of superior bone loss and superolateral rim loss. Type 3A defects have an intact teardrop and inferomedial wall whereas type 3B defects posterior column deficiency is present with destruction of the inferomedial wall and teardrop. As the defect in the acetabulum is usually greater superoinferiorly rather than antero-posteriorly the AP diameter of the acetabulum will need to be widened to accommodate a hemispheric component. The posterior column is the most important structure for implant stability and thus after superior bone defects have been addressed, the anterior wall should be preferentially reamed until adequate component stability is gained. Uncemented acetabular components can also provide good long-term fixation in the presence of a pelvic discontinuity once the posterior column has been stabilised by plating. In principle, the amount of acetabular reaming should be kept to the minimum required to achieve adequate host bone contact and stability. The position of uncemented acetabular components should ideally be 451 of lateral opening with 15–251 of anteversion. This may be difficult in the absence of usual anatomical landmarks. In particular, the component will often be uncovered superolaterally although this is preferable to a component placed too vertically. In addition, efforts to restore correct anteversion will often result in the cup being uncovered posteriorly. As a general rule, somewhere between 50% and 70% of contact with host bone is required for long-term implant stability. Uncemented acetabular components should then be augmented with multiple screws into the ilium. The superior quadrant, described by a line between the ASIS and ischium and its bisector, is the safe area for placement of acetabular screws. Screws may also be required posteriorly into the ischium and anteriorly into the pubis, however this carries the increased risk of neurovascular damage. It has been our post-operative protocol to keep patients touch weight bearing for the first 6 weeks post-operatively, then gradually increase weight bearing by the 3-month mark.

Isolated polyethylene liner exchange and bone grafting Osteolysis behind a well-fixed uncemented acetabular component is a relatively common radiographic finding at our institution. Patients are often asymptomatic as the fixation is solid, but the process of osteolysis may continue without surgical intervention and may eventually lead to failure. It is important to note that the fixation of any revision acetabular component will initially be worse than that of a stable ingrown component even in the presence of

A.W.R. Burns, R.W. McCalden retro-acetabular osteolysis. Therefore, one option for this clinical problem is to perform an isolated polyethylene liner exchange, often combined with bone grafting of the defect, to halt the production of particle debris. Indeed, this is one of the prime rationales for the production and widespread adoption of modular acetabular components. The acetabular component suitable for this technique must be satisfactorily positioned and stable at the time of surgery, with a well designed locking mechanism which allows for exchange of liner. It is preferable to have multiple options for liner and head exchange including a range of head sizes and neck lengths, and lateralised, lipped or anteverted liners. Osteolytic defects may then be addressed by cancellous bone grafting. Bone grafting may be performed in several ways. Dome covers or screws may be removed and cancellous graft packed through these portals into defects. Specialised curettes with various angled necks have been designed specifically for working through dome and screw holes allowing improved debridement of granulation and osteolytic tissue (Wright Medical, Mississauga Canada.) It is possible to impact large quantities of cancellous graft through screw holes. Circumferential clearing of soft tissue from the bone prosthesis interface will often demonstrate defects around the cup which can also be used to debride and impact bone graft. Fenestrations in the superolateral ilium through which access may be gained may be made using a burr or osteotome. Bony septa are often present from the pelvis to the uncemented shell and are sometimes visible through rim defects or windows. These septa are critical for the continued stability of the implant and should be carefully preserved during debridement and grafting. High dislocation rates following isolated liner exchange have been reported, particularly with the posterior approach, and thus adequate intraoperative stability must be confirmed prior to closure.5 At our institution, using a direct lateral approach, dislocation has not been a complication of this procedure.6 If there is any question that stability cannot be maintained intraoperatively, or that the acetabular component is not well fixed, then complete acetabular revision should be performed.

Cementing a polyethylene liner into a well-fixed uncemented shell Another option for consideration in acetabular revision is the technique of cementing a liner into a well fixed cup. This option is applicable to a stable well-fixed shell with a nonmodular liner that can be removed, or a modular liner with a poor or damaged locking mechanism, and where the cup to be retained has proven good long term survivorship. The retained acetabular component should have an internal diameter large enough to accommodate the outer diameter of a polyethylene liner plus a 2 mm cement mantle. Patients with a history of dislocation may be treated by cementing a captured liner and good outcomes have been reported with use of this technique.7,8 For cementation of all polyethylene liners into an uncemented shell we use a high-speed metal cutting burr to make radial and circumferential grooves in the metal of the socket (2–3 mm deep) prior to cementing, and size the

ARTICLE IN PRESS Current techniques and new developments liner aiming for a 2 mm cement mantle. In addition, where a modular liner is being used, the backside should be roughened or grooved to improved cement fixation. Generally, we aim to use the largest bearing size available and use an anterolateral approach to minimise the risk of dislocation.

Use of structural bone grafts The majority of acetabular revision surgery can be accomplished using morsellised allograft to deal with small to large cavitatory bone defects. Structural bone grafts become necessary when there is insufficient bone stock available to provide adequate support for revision components. In the use of structural bone grafts size and complexity can range from using a femoral head to reinforce medial wall or superolateral rim defects to the use of total acetabular allografts in the case of massive defects. Precise indications for use of structural allograft have not been defined, and the use at our institution (which is a tertiary referral centre for revision arthroplasty) is approximately 5% of revision cases. Structural grafts are commonly used to manage superolateral acetabular defects in the presence of intact anterior and posterior columns. A femoral head or distal femur may be fashioned to make a ‘‘number 7’’ graft. The anteromedial quadrant of a femoral head is removed and the cancellous right angled bone surface so formed is placed over the superolateral acetabular rim such that the remainder of the femoral head is within the acetabulum and the neck is seated against the lateral ilium. The graft is then held in position using 6.5 mm screws with washers through the femoral neck portion of the allograft into the ilium such that they are well away from the acetabulum. The graft within the acetabulum is reamed to a hemispheric surface (Fig. 2).

165 If an uncemented revision socket is to be utilised, there must be greater than 50% host bone coverage as allograft bone will not reliably grow onto an uncemented socket.9 If less than 50% host bone coverage is present then a reconstruction ring should be chosen and fixed to the ischium and ilium with multiple screws at which point an all polyethylene cup is cemented into the construct.9 Postoperatively patients should be touch weight bearing for the first 6 weeks, with a gradual increase to 100% weight bearing by the 3-month mark.

Antiprotrusio rings and reconstruction cages When inadequate bone stock precludes the use of an uncemented acetabular revision component, two types of implants are available for use. An antiprotrusio ring fits within in the acetabulum supported by the intact acetabular rim whereas a reconstruction cage has flanges which can be attached to the ilium superiorly and into the ischium inferiorly, thus spanning the acetabulum. Both of these devices can provide a stable construct into which a polyethylene cup is cemented whilst providing protection and support for morsellised or structural bone graft (Fig. 3). The antiprotrusio ring is used in smaller defects and has to some extent been supplanted by the use of large uncemented acetabular components. The reconstruction cage is used when defects are much larger, particularly when bone stock is deficient around the rim of the acetabulum. Both types of implants are attached to the acetabulum or pelvis with screws allowing a cemented cup to be inserted in the optimal position. As rings and cages do not incorporate to the host bone they may fail in the longer term. However, this may be an acceptable solution as the bone graft beneath them has usually remodelled by this stage allowing revision into an improved bed compared to the status prior to the previous revision. Rings or cages must be supported either

Figure 2 Number 7 graft fashioned from femoral allograft (inset left) is attached to the superolateral acetabular defect with screws (left side) and reamed to a hemispheric socket (right). A reconstruction cage is usually required to protect this construct when less than 50% host bone contact exists.

ARTICLE IN PRESS 166

Figure 3 Antiprotrusio ring (left) and reconstruction cage (right).

by host bone ideally or allograft to limit the chance of fatigue failure. Reconstruction cages are sized with an external diameter equal to the last reamer used on the acetabulum. All bone grafting should be completed prior to insertion of the cage. The inferior flange of the cage can either be laid on the surface of the ischium (in close proximity to the sciatic nerve), or impacted into a groove cut into the ischium (more remote from the nerve).9 The superior flanges are then slid onto and contoured to the ilium and fixed with multiple screws. Screws may also be passed into the ischium through the inferior part of the cage. This results in a construct which is fixed solidly both superiorly and inferiorly and supported on either host bone or allograft throughout. Trial acetabular liners can be used at this time to provide the surgeon an idea of the leg length, orientation and stability of the hip. An all polyethylene cup is then cemented into the cage in an orientation independent of the cage but aiming for approximately 451 of lateral opening and 15–251 of anteversion. To improve hip stability larger size heads can be used (36 mm) or a captured liner may be required for stability particularly in the presence of weak or absent abductors. Review articles would suggest that there is approximately 75% survivorship at 5–10 years in acetabular revision using reconstruction rings or cages.9

Impaction bone grafting with cemented all polyethylene cups Impaction bone grafting is a technique which has now been used for more than 25 years with good clinical results.10 The technique involves packing a cavitatory defect with cancellous bone graft to create tight solid form which will then support a cemented all polyethylene liner. The technique can also be used for larger segmental defects by contouring malleable metal mesh to reconstruct a rim or wall and thus convert segmental defects into cavitatory defects. Primary implants can usually then be used. The technique allows bone stock to be restored which then remodels under load. Excellent incorporation of these grafts has been demonstrated histologically from biopsy studies thus confirming

A.W.R. Burns, R.W. McCalden restoration of bone stock.11 This technique is particularly applicable in young people where restoration of bone stock for future surgeries is desirable. The reconstitution of bone stock may also allow subsequent revision to an uncemented acetabular component in future should revision be required. The technique is not suitable for use in people who have undergone pelvic irradiation as the potential for revascularisation and reincorporation is reduced. After wide exposure of the acetabulum and removal of any soft tissue or membrane from the bony surface, any areas of eburnated bone are drilled to encourage host bone bleeding. Estimation of the extent and location of bone deficiency is made with a trial acetabular liner. Stainless steel mesh (Acetabular X-change mesh, Stryker-Howmedica, Newbury, England) is then cut and contoured to the desired size and position and held in place with screws. Cancellous bone chips preferably 7–10 mm in diameter are then implanted into this area and impacted vigorously in layers using a series of specialised hemispheric punches which are sized to allow for approximately 3 mm of cement mantle. Cement is then pressurised into this bone surface and a cemented cup is implanted in the appropriate orientation. Postoperatively, the patients are kept in touch weight bearing for the first 6 weeks then graduating up to full weight bearing by the three month mark with X-rays at both appointments. Survivorship of 78–93% at 20 years has been reported using this acetabular reconstruction technique.10

Bilobed uncemented acetabular components Isolated superolateral acetabular rim defects with intact anterior and posterior walls can also be treated with a bilobed cup. A bilobed cup is a hemispheric cup with a partial hemispheric extension. A bilobed cup can better match the superolateral acetabular defect and can theoretically conserve the bone that reaming of the anterior and posterior walls to insert a jumbo uncemented acetabular shell would otherwise sacrifice. Similarly, these prostheses may be applicable in primary arthroplasty for arthritis secondary to high dislocation in developmental dysplasia of the hip. The hip centre can then placed in the anatomical position avoiding a high hip centre, in which the superolateral bone defect is filled with the superior lobe of the implant, providing further bone contact. In comparison with a jumbo hemispheric cup, one disadvantage may be that it can be technically more difficult to orientate the version of a bilobed cup thus making stability more difficult to obtain. Nevertheless for selected indications acceptable mid-term results have been demonstrated12,13 (Fig. 4).

Custom-made triflange components The occurrence of massive pelvic osteolytic defects or pelvic dissociation is relatively rare in revision hip surgery. However, in the face of massive bone loss with or without pelvic discontinuity, a custom-made triflange reconstructive cage is an option for management of these difficult clinical problems.14 In comparison with a standard reconstruction cage, the custom triflange cage has the advantages of fitting the defect exactly, and, due to its specifically manufactured shape, the need for malleable flanges (which are weaker) is

ARTICLE IN PRESS Current techniques and new developments

167

Figure 4 Bilobed or oblong acetabular component. (DePuy, Warsaw, IN, USA).

removed thus providing a much more rigid and mechanically stronger construct than a standard cage. In addition it has the potential for osseous integration. The cost of such a customised component is significantly greater (upward of UK£3000) and its use is only indicated where a standard reconstruction cage would be inadequate.14 If pelvic X-rays (including Judet views) suggest a massive defect or pelvic dissociation, a three-dimensional CT reconstruction of the hemi-pelvis and acetabular defect is performed. Using stereolithography, a hemipelvis matching this is produced from which a clay model component can be made. In consultation with the medical engineer the size and shape of the flanges, the height, abduction angle and version of the component are decided upon by the surgeon and a plastic trial component is manufactured. The flanges are made large enough to allow multiple screw fixation options into both ilium and ischium. This trial can be sterilised and used intraoperatively. The definitive component is then manufactured usually from titanium and can be coated it with hydroxy-apatite to improve the osseous integration potential. After extensile exposure and acetabular preparation, including debridement and bone grafting, the ilial flanges are slid up underneath the abductor muscles and then the component is fixed with screws into first the ilium and then the ischium. A trochanteric osteotomy may be valuable to assist exposure, component placement and protect the abductor and superior gluteal neurovascular bundle. As the ischial area is usually the one most commonly associated with loss of fixation, multiple screw holes should be available in this part of the component. Short-term results reported with the use of these components have been encouraging with 89% survival at 4.5 years in this complicated patient population.14

Trabecular metal components Trabecular metal is a relatively new material made of the element tantalum from which acetabular components can be made. Trabecular metal consists of interconnecting pores resulting in a structural biomaterial that is 80% porous, allowing approximately 2–3 times greater bone ingrowth

Figure 5 Electron micrograph demonstrating trabecular metal structure (right half) and similarity with cancellous bone (left half). (Zimmer, Warsaw, IN, USA.)

compared to conventional porous coatings. This material has been shown to have a substantially higher coefficient of friction on cancellous bone, when compared to other commonly used prosthetic materials, and a higher bone interface shear strength compared to other fixation surfaces.15 This metal, due to its reduced stiffness, may also provide a more favourable environment for the remodelling of morsellised or structural bone graft in comparison with other materials (Fig. 5). Trabecular metal acetabular components come both with modular liners or as revision shells into which a cemented cup is implanted. Our own clinical experience with the trabecular metal at revision surgery is of excellent initial bony stability and scratch fit, even in clinical situations with very limited or poor quality bone stock. Traditionally, the use of an uncemented cup has required greater than 50% of host bone contact. However, some authors have described the use of trabecular metal with less than 50% bone contact and have found that its use has decreased the requirement for the use of reconstruction cages.15 One recent paper reported the use of a monoblock trabecular shell without screws for 55 of 60 acetabular revisions which would seem perilous with other implants.16 The porous nature of this material also allows additional screw holes to be burred through a revision component in whatever position is required to gain fixation (Fig. 6). The trabecular metal revision shell has been used for the technique of the cup-cage construct. In this construct morsellised or structural bone graft can be implanted into

ARTICLE IN PRESS 168 acetabular defects and then a trabecular shell fixed into the defect with multiple screws. To protect the trabecular metal shell until the bone graft has incorporated a reconstruction cage is then inserted into the cup and fixed through the holes in the flanges to the pelvis superiorly and inferiorly. Screws can also be drilled from the cage into the cup itself providing improved strength of construct. A liner is then cemented into the reconstruction cage. In this

Figure 6 Trabecular metal revision shell with augment. (Zimmer, Warsaw, IN, USA.)

A.W.R. Burns, R.W. McCalden construct, the reconstruction cage functions to protect the trabecular cup in the short term until cup ingrowth and bone graft incorporation has occurred after which the trabecular cup provides added support for the cage in the longer term, holding promise for greater survivorship.

Cup cage construct for massive pelvic defects Another recent advance is the development of trabecular metal augments used primarily for superior bone defects where a hemispheric cup will gain less than 50% host bone contact.17 Early results of this technique appear promising. Trabecular metal augments are available in various sizes and shapes to fill required bone defects. These can be screwed into position independent of the trabecular metal cup and then connected to the trabecular hemispheric cup with a small amount of bone cement at the time of cup impaction (Fig. 7). Recently, a series of transverse acetabular fractures leading to pelvic discontinuity were reported with the use of trabecular metal shells presenting at an average of 8-month post-revision THA.18 The authors stated that the acetabular components were all well fixed and thus surmised that this may represent an insufficiency fracture where the trabecular shells gain fixation in acetabuli with

Figure 7 Cup cage construct. Massive acetabular defect (left) treated with cancellous bone graft, a trabecular revision shell with a reconstruction cage and cemented polyethylene cup.

ARTICLE IN PRESS Current techniques and new developments less bone stock than previously possible, thus predisposing the acetabulum to later stress fracture. While no long-term results have been published in the literature, the short term results for trabecular metal technology are encouraging. In the future, the possibility of combining trabecular metal into reconstruction cages may provide both excellent biological fixation with the advantages of a reconstruction cage.

Osteo-inductive materials in acetabular revision Recent research interest has concentrated on the use of osteo-inductive materials such as osteogenic proteins (OP) or bone morphogenic proteins (BMP) to aid in the healing of bone. Advances in cloning and genetic expression have allowed the production of recombinant human OP-1 in large quantities. OP-1 (BMP-7) has been shown in preclinical canine models, either alone or in combination with auto graft or allograft bone, to improve healing in models of both morsellised graft, cortical strut allografts, and acetabular bone defects.19 Autologous platelet concentrate, which contains multiple growth factors (PDGF, TGF-ß, IGF) has also been shown to be efficacious in reconstructive surgery.20 It is therefore possible that osteo-inductive proteins could improve the amount of bone ongrowth to components as well as improving the restoration of bone stock. Currently, these substances are very expensive, and theoretical complications exist with respect to the formation of heterotopic ossification. In addition, some patients may develop antibodies after implantation of OP, and the effect of these antibodies on future bone induction and healing is unknown.19 Until proper clinical evaluation is performed, these substances should be used with caution. However, they definitely hold promise for the future of reconstructive arthroplasty surgery.

169 tions exists in the inhibition and treatment of early particle induced osteolysis as well as a prosthetic coating.

Summary and conclusions Acetabular component revision is a challenging part of revision THA surgery which requires a systematic approach with meticulous pre-operative planning, advanced reconstructive surgical techniques and careful post-operative supervision. Bone defects can be large and often underestimated and thus the revision surgeon must be skilled in bone grafting techniques. The majority of acetabular revisions can be accomplished with uncemented ‘‘jumbo’’ components. Massive defects may require structural allografts, protected by reconstruction cages where less than 50% host bone contact exists. In the future, the use of trabecular metal technology, modular reconstruction rings and cages, coupled with the use of bone induction and osteoclast inhibitor substances, may improve the outcomes and simplify complex acetabular reconstruction problems.

Practice points

 Meticulous pre-operative planning and imaging  Thorough knowledge of extensile exposure to the hip joint and pelvis

 Inventory to deal with implant removal, manage  

bone defects, prosthesis re-implantation and other eventualities Development of skill with advanced bone grafting techniques Careful post-operative care and radiographic evaluation

Future directions

 Use of trabecular metal in reconstruction rings and cages

Role of bisphosphonates in managing osteolysis Particle induced osteolysis, leading to aseptic loosening, is a major problem affecting the long-term survivorship of total joint replacements. Osteoclasts have been implicated in the pathophysiology of particle induced osteolysis. Bisphosphonates have well-described effects on the inhibition of osteoclastic action. In particular the third-generation substances (alendronate) appear to have a more specific action on osteoclasts with limited effects on osteoblasts.21 Alendronate has been shown in animal models to inhibit the osteolysis caused by polyethylene wear debris.22 An increase in bone volume and the reversal of particle induced bone loss has been shown following the administration of alendronate.23 Similarly, a recent study has shown an improvement in the mechanical stability of bisphosphonate-coated implants inserted into osteoporotic rats.24 Currently, there are no clinical trials to support this intervention; however, the widespread use of bisphosphonates for the prevention of osteoporosis has demonstrated the safety and efficacy of these medications with few complications. Therefore, a future role for these medica-

 Modularity in reconstruction rings and cages  Locking screw technology in shells of revision acetabular components

 Bone induction substances and HA coating to improve ingrowth

 Bisphosphonate use to limit osteolysis progression Reference 1. Paprosky WG, Perona PG, Lawrence JM. Acetabular defect classification and surgical reconstruction in revision arthroplasty. A 6-year follow-up evaluation. J Arthroplasty 1994;9(1): 33–44. 2. Della Valle CJ, Berger RA, Rosenberg AG, Galante JO. Cementless acetabular reconstruction in revision total hip arthroplasty. Clin Orthop Relat Res 2004(420):96–100. 3. Jones CP, Lachiewicz PF. Factors influencing the longer-term survival of uncemented acetabular components used in total hip revisions. J Bone Joint Surg Am 2004;86-A(2):342–7. 4. Whaley AL, Berry DJ, Harmsen WS. Extra-large uncemented hemispherical acetabular components for revision total hip arthroplasty. J Bone Joint Surg Am 2001;83-A(9):1352–7.

ARTICLE IN PRESS 170 5. Boucher HR, Lynch C, Young AM, Engh Jr. CA, Engh Sr. C. Dislocation after polyethylene liner exchange in total hip arthroplasty. J Arthroplasty 2003;18(5):654–7. 6. O’Brien JJ, Burnett RS, McCalden RW, MacDonald SJ, Bourne RB, Rorabeck CH. Isolated liner exchange in revision total hip arthroplasty: clinical results using the direct lateral surgical approach. J Arthroplasty 2004;19(4):414–23. 7. Beaule PE, Ebramzadeh E, LeDuff M, Prasad R, Amstutz HC. Cementing a liner into a stable cementless acetabular shell: the double-socket technique. J Bone Joint Surg Am 2004;86-A(5): 929–34. 8. Callaghan JJ, Parvizi J, Novak CC, Bremner B, Shrader W, Lewallen DG, et al. A constrained liner cemented into a secure cementless acetabular shell. J Bone Joint Surg Am 2004;86A(10):2206–11. 9. Gross AE, Goodman S. The current role of structural grafts and cages in revision arthroplasty of the hip. Clin Orthop Relat Res 2004(429):193–200. 10. Schreurs BW, Busch VJ, Welten ML, Verdonschot N, Slooff TJ, Gardeniers JW. Acetabular reconstruction with impaction bonegrafting and a cemented cup in patients younger than fifty years old. J Bone Joint Surg Am 2004;86-A(11):2385–92. 11. Schreurs BW, Slooff TJ, Buma P, Verdonschot N. Basic science of bone impaction grafting. Instr Course Lect 2001;50: 211–20. 12. Berry DJ, Sutherland CJ, Trousdale RT, Colwell Jr. CW, Chandler HP, Ayres D, et al. Bilobed oblong porous coated acetabular components in revision total hip arthroplasty. Clin Orthop Relat Res 2000(371):154–60. 13. Chen WM, Engh Jr. CA, Hopper Jr. RH, McAuley JP, Engh CA. Acetabular revision with use of a bilobed component inserted without cement in patients who have acetabular bone-stock deficiency. J Bone Joint Surg Am 2000;82(2):197–206. 14. Holt GE, Dennis DA. Use of custom triflanged acetabular components in revision total hip arthroplasty. Clin Orthop Relat Res 2004(429):209–14.

A.W.R. Burns, R.W. McCalden 15. Bobyn JD, Stackpool GJ, Hacking SA, Tanzer M, Krygier JJ. Characteristics of bone ingrowth and interface mechanics of a new porous tantalum biomaterial. J Bone Joint Surg Br 1999; 81(5):907–14. 16. Gross AE, Goodman SB. Rebuilding the skeleton: the intraoperative use of trabecular metal in revision total hip arthroplasty. J Arthroplasty 2005;20(4 Suppl 2):91–3. 17. Unger AS, Lewis RJ, Gruen T. Evaluation of a porous tantalum uncemented acetabular cup in revision total hip arthroplasty: clinical and radiological results of 60 hips. J Arthroplasty 2005; 20(8):1002–9. 18. Nehme A, Lewallen DG, Hanssen AD. Modular porous metal augments for treatment of severe acetabular bone loss during revision hip arthroplasty. Clin Orthop Relat Res 2004(429): 201–8. 19. Springer BD, Berry DJ, Cabanela ME, Hanssen AD, Lewallen DG. Early postoperative transverse pelvic fracture: a new complication related to revision arthroplasty with an uncemented cup. J Bone Joint Surg Am 2005;87(12):2626–31. 20. Cook SD, Barrack RL, Patron LP, Salkeld SL. Osteoinductive agents in reconstructive hip surgery: a look forward. Clin Orthop Relat Res 2003(417):195–202. 21. Franchini M, Dupplicato P, Ferro I, De Gironcoli M, Aldegheri R. Efficacy of platelet gel in reconstructive bone surgery. Orthopedics 2005;28(2):161–3. 22. Shanbhag AS, Hasselman CT, Rubash HE. The John Charnley Award. Inhibition of wear debris mediated osteolysis in a canine total hip arthroplasty model. Clin Orthop Relat Res 1997(344): 33–43. 23. Shanbhag AS, Hasselman CT, Rubash HE. The John Charnley Award. Inhibition of wear debris mediated osteolysis in a canine total hip arthroplasty model. Clin Orthop Relat Res 1997(344): 33–43. 24. Millett PJ, Allen MJ, Bostrom MP. Effects of alendronate on particle-induced osteolysis in a rat model. J Bone Joint Surg Am 2002;84-A(2):236–49.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 171–178

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: REVISION HIP ARTHROPLASTY

(iii) Cementless femoral revision: The role of monoblock versus modular stems Scott M. Sporera,, Wayne G. Paproskyb a

Department of Orthopaedic Surgery, Central Dupage Hospital, Rush University Medical Center, Chicago, 25 N. Winfield Road, Winfield, IL 60190, USA b Department of Orthopaedic Surgery, Rush Presbyterian St. Luke’s Medical Center, Chicago, Central Dupage Hospital, Winfield, IL, USA

KEYWORDS Hip revision; Femoral revision; Modular; Extensively coated stem

Summary A monoblock, extensively porous-coated stem can provide reliable long-term fixation for the majority of femoral revisions. However, stable long-term fixation may not be obtained in patients with moderate/severe bone loss or in patients with torsional femoral remodelling. If severe bone loss is present (Paprosky Type IV bone), a very large canal diameter is encountered (greater than 19 mm) or if torsion remodelling of the proximal femur has occurred, alternative methods of fixation, such as a distally based modular stem, should be considered. & 2006 Elsevier Ltd. All rights reserved.

Introduction

Preoperative planning

Total hip arthroplasty remains one of the most successful and reliable surgical procedures to relieve pain and improve function. Despite the overwhelming success of this operation, there are situations, such as aseptic loosening, septic loosening, recurrent dislocation and periprosthetic fracture, where a femoral revision is required. The number of revision hip surgeries is expected to increase as the indications for total hip replacement broaden and the average life expectancy increases.

The choice of implant used during the femoral reconstruction will be based largely upon the amount of femoral bone loss encountered at the time of revision surgery. The senior author has previously described a femoral classification system that can assist the surgeon with preoperative planning and predict the extent of bone loss.1 The Paprosky femoral defect classification places the remaining host bone into one of four defect types. Type I defects have minimal damage to the proximal metaphysis and resemble a primary hip. Type II defects demonstrate mild meta-diaphyseal bone damage with an intact diaphysis. Type III defects have significant meta-diaphyseal damage, Type IIIA allowing greater than 5 cm and Type IIIB allowing less than 5 cm of ‘‘scratch fit’’ at the isthmus. Type IV defects are

Corresponding author. Tel.: +1 630 339 2225;

fax: +1 630 682 8946. E-mail address: [email protected] (S.M. Sporer). 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.02.006

ARTICLE IN PRESS 172 characterised by extensive meta-diaphyseal damage with thin cortices and a widened femoral canal.

Intraoperative assessment The intraoperative assessment of the remaining femoral bone stock, once the femoral component has been removed, is important when considering various methods of femoral reconstruction. The three most important factors to consider when evaluating the remaining host bone are:

(1) is the remaining metaphysis supportive? (2) what is the remaining length of intact isthmus? and (3) what is the quality of the remaining cortex.

Principles of reconstruction A successful femoral reconstruction requires a femoral component that establishes initial axial and rotational stability. Bone ingrowth is less likely to occur if micromotion cannot be reduced to less than 50 mm. The ability to achieve initial implant stability in patients with compromised bone stock is challenging. Deficiencies of the isthmus may result in less surface area available to obtain a ‘‘scratch-fit’’ with an extensively coated device. Four to five centimeters of scratch fit is recommended in order to obtain adequate component stability. Post-operative and intra-operative hip stability is another consideration during femoral revision. The incidence of instability increases in patients with multiple procedures. The aetiology is likely to be multifactorial, including decreased muscle tone, leg length inequality, and inadequate restoration of offset/ anteversion.

S.M. Sporer, W.G. Paprosky

Monoblock extensively coated femoral components Femoral reconstruction with the use of a monoblock, extensively coated stem remains the ‘‘gold standard’’ for most femoral revisions2 (Fig. 1). This method of reconstruction uses an implant which relies upon distal fixation within the isthmus and bypasses any bone deficiency within the metaphysis. An implant should be chosen that will allow 4–5 cm of ‘‘scratch-fit’’.3 A standard six inch extensively coated stem can be used in the majority of femoral revisions (Paprosky Type I, II). The endosteal canal is reamed with a cylindrical reamer until the endosteal cortex is engaged. An implant 0.5 mm larger than the last reamer is chosen. Longer, extensively coated stems (7 and 8 in stems) are available for patients with more severe bone loss (Paprosky Type IIIA and IIIB). When using a longer extensively coated stem, care must be taken to avoid anterior femoral perforation during femoral preparation and component insertion due to the anterior bow of the femur. This is especially common in patients with short stature or in reconstructions that have not required an extended trochanteric osteotomy. Bowed implants are available in these situations and should minimise this complication (Fig. 2). Torsional remodelling of the proximal femur presents the surgeon with additional challenges when using a monoblock prosthesis. Unlike a straight monoblock stem, a curved monoblock stem will not allow the independent adjustment of femoral anteversion. The patient’s femoral bow will dictate the ultimate position of a bowed femoral stem. Inadequate femoral anteversion may result in patients that have a significant amount of proximal femoral torsion remodelling and can result in an increased risk of postoperative instability. The senior author has reported his average 14-year results of revision femoral surgery with the use of either a 7-in

Figure 1 Seventy-two-year-old female with increasing thigh pain 3 years post-operatively: (a) Preoperative radiograph demonstrating a fracture of the greater trochanter and a loose cemented femoral component. Patient has greater than 5 cm of remaining isthmic bone available (Paprosky Type IIIA defect). (b) Post-operative radiograph demonstrating reconstruction with the use of an 8 in extensively coated femoral component. Initial axial and rotational stability was obtained.

ARTICLE IN PRESS Role of monoblock versus modular stems

173 calcar or 8-in fully porous coated stem. The results from this study showed that reliable femoral fixation can be expected in more than 95% of patients.2 Several other authors have observed similar excellent results with the use of an extensively coated stem4,5 (Table 1). The results of a monoblock extensively coated stem deteriorate when this implant is used for more severe bone defects. The senior author has reviewed the results of patients treated with either a 9-in calcar or 10-in monoblock stem. The mechanical failure rate, defined as revision for aseptic loosening or radiographic evidence of unstable fibrous fixation, was 0% in patients with Type III B defects and femoral canals less than 19 mm, 18% in patients with Type IIIB defects and femoral canals greater than 19 mm, and 37.5% in patients with Type IV defects.6 Consequently, the use of extensively coated stems in the senior author’s practice is limited to patients with a Paprosky Type IIIA or less severe femoral defect or in patients with a Type IIIB defect with an endosteal canal less than 19 mm. The proximal metaphysis is also assessed and alternative methods of reconstruction are chosen if adequate anteversion cannot be obtained.

Modular femoral stems Figure 2 A bowed femoral component minimises the chance of anterior cortical perforation when a long extensively coated implant is chosen.

Modular femoral stems provide several potential advantages over a monoblock prosthesis. Modular devices allow independent preparation of the proximal and distal bone in

Table 1 Author(s)

Hips

Follow-up

Extensively coated revision femoral components Weeden, 170 hips 14.2 years (11–16 Paprosky years)

Sporer and Paprosky6

51 hips

6 years (2–11)

Engh et al.5

26 hips

13.3 years

Moreland and Moreno17

137 hips

9.3 years (5–16 years)

134 patients

Surgical variables

Outcome

Conclusion

3.5% revision for aseptic Extensively coated femoral loosening stems should be used during revision femoral surgery 4.1% mechanical failure rate Type IIIB or IV 0% failure in IIIB with Consider alternative defects treated canal o19 mm methods of fixation in with 9 or 10 in Type IIIB defects with a stem canal 419 mm or a Type IV defect 18% failure in IIIB with canal 419 mm 37.5% failure in IV Bone loss410 cm 89% Survival at 10 years. Extensively coated below lesser For aseptic femoral stems can be used to trochanter revision bypass severe proximal bone loss 15% aseptic femoral loosening rate 4% revision rate of Extensively coated aseptic femoral stems provide durable loosening fixation 83% radiographic bone ingrowth

ARTICLE IN PRESS 174

S.M. Sporer, W.G. Paprosky

Table 1 (continued ) Author(s)

Hips

Follow-up

Krishnamurthy et al.1

297 hips

8.3 years (5–14 years)

Surgical variables

297 patients Moreland and Bernstein4

175 patients

Lawrence et al.18 83 hips

9 years (5–13 years)

81 patients

Mulroy and Harris20

43 hips

15.1 years (14.2–17.5)

Calcar replacing component

Second generation cement technique

41 patients Katz et al.21

79 hips

47 hips with minimum 10year follow-up

Second generation cement technique

73 patients

Kavanagh and Fitzgerald22

Pellicci et al.23

45 hips

45 patients 99 hips

Kavanagh et al.24 166 hips

41 months

First generation cement technique

8.1 years (5–12.5 First generation years) cement technique 4.5 years First generation cement technique

162 patients Callaghan et al.14

139 hips

3.6 years (2–5 years)

Conclusion

1.7% femoral revision for aseptic loosening

Diaphyseal fixation should be used in femoral revisions to avoid bone loss in the proximal metaphysis

2.4% mechanical failure rate 96% component survival Extensively coated stems provide durable fixation 83% bone ingrown 10% femoral re-revision Extensively coated stems can provide reliable fixation during femoral revision 11% mechanical failure of femoral component

5 years

Cemented revision femoral components McLaughlin and 35 hips 10.8 years Harris19 (5.8–16.6 years) 38 patients

Outcome

First generation cement technique

18% rate of aseptic femoral revision 32% rate of mechanical failure 20% rate of aseptic femoral revision

26% rate of aseptic femoral loosening 9.5% rate of femoral revision for aseptic loosening at 10 years

26.1% rate of radiographic femoral failure at 10 years 16% rate of femoral revision. 28% rate of radiographic loosening

29% rate of mechanical failure 44% radiographic aseptic loosening 9% rate of femoral revision 15.8% definite mechanical failure

Second generation cement technique decreases the prevalence of aseptic femoral loosening

Second generation cement technique improves clinical and radiographic results. Cemented femoral revision demonstrates high rate of mechanical failure

Cemented re-revision demonstrates poor clinical and radiographic results Increased failure rate with longer follow-up of cemented stems The high incidence of radiographic signs of loosening is of a concern

Mechanical failure and progressive radiolucencies were

ARTICLE IN PRESS Role of monoblock versus modular stems

175

Table 1 (continued ) Author(s)

Hips

Follow-up

Surgical variables

Outcome

Conclusion

associated with poor quality of bone 8.6% femoral revision due to aseptic loosening Proximally coated revision femoral components Mulliken et al.25 52 hips 4.6 years

Monoblock titanium stem

51 patients Malkani et al.16

69 hips

Woolson and Delaney27

375 hips

25 hips

Monoblock stem

4.7 years

29% mechanical failure 82% 5 year survival Monoblock stems 58% survivorship at 8 years for aseptic femoral revision

5.5 years

Proximal modular femoral components Cameron28 320 hips 7 years (2–12 years)

Christie et al.29

163 hips

Chandler et al.30 52 hips

(4–7 years)

91 hips

Monoblock titanium stem

Proximally porouscoated stems in revisions with femoral bone loss show inadequate fixation

Proximally coated stems do not provide reliable fixation during revision surgery

Stable fixation with proximally coated femoral components cannot be reliably achieved

20% survivorship for aseptic femoral loosening 20% rate of femoral re- Proximally porous revision coated femoral components should not be used in revision surgery 45% mechanical failure

S-ROM stem

No primary stem revisions for aseptic loosening

109 primary stems, 211 revision stems Type II and III defects. S-ROM stem

1.4% revision for aseptic loosening with revision stem 2.9% rate of aseptic A proximal ingrowth, failure proximally modular stem can be used in revision hip surgery Mechanical loosening A proximal ingrowth, occurred in 5 hips proximally modular stem is versatile can be used in difficult revision hip surgery

3 years

S-ROM stem

(2–6 years)

22 pts. required structural allograft S-ROM stem

48 patients

Cameron31

40% intra-operative fracture 8.7% femoral revision

3 years

69 patients Berry et al.26

10% femoral revision for aseptic loosening; additional 14% radiographically loose

86% excellent/good results with primary stem—1 revision

A proximal ingrowth, proximally modular stem can be used in revision hip surgery

Modularity is an option for femoral revision

ARTICLE IN PRESS 176

S.M. Sporer, W.G. Paprosky

Table 1 (continued ) Author(s)

Hips

Surgical variables

Outcome

29 primary stems, 62 revision stems

80% excellent/good results with revision stem—10 revisions

Modular tapered revision femoral components Wirtz et al.11 142 hips 2.3 years

MRP-titan stem

95.8% component survival

Kwong et al.10

Link MP stem

97.2% component survival

143 hips

Follow-up

40 months (2–6 years)

Conclusion

The midstem junction provide a distinct advantage in its ability to independently adjust anteversion to maximise hip stability The link MP hip stem allows successful revision THA reconstruction of the proximally compromised femur

2.1 mm average subsidence Non-modular tapered revision femoral components Isacson et al.9 43 hips 25 months

Wagner SL stem

Grunig32

Wagner SL stem

40 hips

47 months (3–7 years)

order to maximise fill. Additionally, these implants can allow the independent adjustment of femoral anteversion, offset and leg length. Modular femoral stems may be classified into those that obtain fixation proximally and those that obtain fixation distally. Uncemented modular stems that obtain proximal fixation utilise a porous coated sleeve which is placed into the remaining host metaphyseal bone. A femoral stem is then secured to the proximal sleeve through the use of a Morse taper. The femoral stem is frequently slotted to aid in obtaining rotational stability. Unfortunately, most femoral revisions where a modular stem is considered have an unsupportive metaphysis. Consequently, initial stability is difficult to obtain without the use of an allograft. Bowed stems can also be used with a proximally based modular system to improve distal fill and bypass bone defects. Similar to a monoblock stem, the ultimate component anteversion of a proximally based bowed stem will be dictated by the anterior femoral bow. Uncemented modular stems that obtain distal fixation utilise either a porous or tapered distal segment which can be attached to a variety of proximal body segments. Modular systems which rely upon distal fixation allow the surgeon to

5 pts. with subsidence 420 mm.9 recurrent dislocations requiring 8 revisions 3 hips revised for aseptic femoral loosening

High rate of component subsidence and dislocation We recommend the Wagner SL femoral revision stem, not as a routine procedure to treat loosening, but for patients with severe femoral bone resorption after THA

obtain stable distal fixation prior to reconstructing the proximal femur. These systems allow the use of an extended trochanteric osteotomy which is frequently required due to varus femoral remodelling, or to facilitate component extraction. Extensively coated distal segments can be used in patients with adequate remaining isthmic bone (45 cm). As with a monoblock stem, sufficient bone must be available in order to obtain initial axial and rotational stability. A tapered stem geometry is an alternative to an extensively coated distal segment. Rotational stability in a tapered stem is obtained from the peripheral splines while axial stability is obtained via a wedge fit between the host bone and implant. This type of implant is useful in patients with minimal remaining isthmus (Paprosky Type IIIB and IV) (Fig. 3). Distally based systems allow the proximal body segment to be rotated independently of the distal segment in order to adjust the amount of femoral anteversion. There are potential disadvantages in using a modular femoral component. The junction between the proximal and distal segments introduces the potential for fretting and corrosion particles.7 Distally based modular femoral stems have also shown a small incidence of component fracture

ARTICLE IN PRESS Role of monoblock versus modular stems

177

Figure 3 Eighty-four-year-old female 2 years post-operatively with a 3 months history of increasing thigh and groin pain: (a) Preoperative radiograph demonstrating subsidence of a proximally coated long stem femoral component. (b) Preoperative aspiration demonstrated deep infection with S. epidermidis. Patient with a Type IIIB femoral defect. A temporary antibiotic spacer is in place. (c) Post-operative second stage reconstruction with the use of a distally based modular tapered stem.

when these implants are used in patients with unsupportive proximal bone. In these situations, an implant with a large proximal taper should be chosen to minimise this risk. The original fluted, tapered grit-blasted stem was the Wagner self-locking stem (Sulzer Medica, Baar, Switzerland). Several authors have reported their results with this stem during revision surgery with component survival rates of greater than 92% at 10 years.8 However, one of the problems seen with the monoblock tapered stem was early subsidence.9 This prompted manufactures to design implants with modular components to optimise the fit of the distal tapered stem with the isthmus and the proximal body against the remaining metaphyseal bone. Unfortunately, long term studies using modular tapered stems are lacking. Kwong et al. evaluated 143 patients at an average of 40

months with a 97% component survival rate using the Link MP prosthesis(Link America, Denville, NJ).10 Wirtz et al. reported on the results of 142 revision hips using the MRPTitan (Peter Brehm, Weisenorf, Germany) stem. The survival rate was 96% at an average 2.3-year follow-up. The authors noted the advantage of independent adjustment of the proximal body to minimise the risk of dislocation.11 Other authors have reported promising early clinical and radiographic results.6,12

Summary and conclusions Total hip arthroplasty remains an effective procedure to relieve pain and improve function in patients with arthritic

ARTICLE IN PRESS 178 conditions affecting the hip. Despite the excellent longterm survival of current generation implants, revision hip arthroplasty now accounts for 17% of hip procedures done on Medicare patients within the United States.13 Long-term follow-up studies of femoral revision have shown poor clinical results with cemented components.14 The high rate of mechanical failure is considered to be due to the decrease in shear strength at the bone–cement interface.15 Similar poor results have been observed in femoral revisions using a monoblock proximally coated implant.16 As a result, most surgeons will now use an extensively porous coated implant during routine femoral revision. Reliable fixation of a cementless stem requires immediate axial and rotational stability. The majority of femoral revisions that an orthopaedic surgeon encounters can be treated with either a 6-, 7- or 8-in extensively porous coated stem. This type of implant will provide reliable initial fixation with a high propensity for bone ingrowth. However, patients with significant femoral bone loss (Paprosky Type III B and Type IV femoral defects) require an alternative method of fixation. The senior author recommends that an extensively coated stem be used only in patients with a Paprosky Type I, II, or IIIA defect. A distally based modular tapered stem is recommended for patients with proximal torsional remodelling or patients with a Type IIIB or Type IV defect.

References 1. Krishnamurthy AB, MacDonald SJ, Paprosky WG. 5- to 13-year follow-up study on cementless femoral components in revision surgery. J Arthroplasty 1997;12:839–47. 2. Weeden SH, Paprosky WG. Minimal 11-year follow-up of extensively porous-coated stems in femoral revision total hip arthroplasty. J Arthroplasty 2002;17:134–7. 3. Aribindi R, Barba M, Solomon MI, et al. Bypass fixation. Orthop Clin North Am 1998;29:319–29. 4. Moreland JR, Bernstein ML. Femoral revision hip arthroplasty with uncemented, porous-coated stems. Clin Orthop 1995:141–50. 5. Engh Jr CA, Ellis TJ, Koralewicz LM, et al. Extensively porouscoated femoral revision for severe femoral bone loss: minimum 10-year follow-up. J Arthroplasty 2002;17:955–60. 6. Sporer SM, Paprosky WG. Revision total hip arthroplasty: the limits of fully coated stems. Clin Orthop Relat Res 2003:203–9. 7. Cook SD, Barrack RL, Baffes GC, et al. Wear and corrosion of modular interfaces in total hip replacements. Clin Orthop 1994:80–8. 8. Bohm P, Bischel O. The use of tapered stems for femoral revision surgery. Clin Orthop Relat Res 2004:148–59. 9. Isacson J, Stark A, Wallensten R. The Wagner revision prosthesis consistently restores femoral bone structure. Int Orthop 2000;24:139–42. 10. Kwong LM, Miller AJ, Lubinus P. A modular distal fixation option for proximal bone loss in revision total hip arthroplasty: a 2-to 6-year follow-up study. J Arthroplasty 2003;18:94–7. 11. Wirtz DC, Heller KD, Holzwarth U, et al. A modular femoral implant for uncemented stem revision in THR. Int Orthop 2000;24:134–8.

S.M. Sporer, W.G. Paprosky 12. Whiteside LA. Major femoral bone loss in revision total hip arthroplasty treated with tapered, porous-coated stems. Clin Orthop Relat Res 2004:222–6. 13. Weinstein J. Dartmouth Atlas of musculoskeletal healthcare. American Hospital Association, 2001. 14. Callaghan JJ, Salvati EA, Pellicci PM, et al. Results of revision for mechanical failure after cemented total hip replacement, 1979 to 1982. A two to five-year follow-up. J Bone Jt Surg Am 1985;67:1074–85. 15. Dohmae Y, Bechtold JE, Sherman RE, et al. Reduction in cement-bone interface shear strength between primary and revision arthroplasty. Clin Orthop 1988:214–20. 16. Malkani AL, Lewallen DG, Cabanela ME, et al. Femoral component revision using an uncemented, proximally coated, long-stem prosthesis. J Arthroplasty 1996;11:411–8. 17. Moreland JR, Moreno MA. Cementless femoral revision arthroplasty of the hip: minimum 5 years followup. Clin Orthop Relat Res 2001:194–201. 18. Lawrence JM, Engh CA, Macalino GE, et al. Outcome of revision hip arthroplasty done without cement. J Bone Jt Surg Am 1994;76:965–73. 19. McLaughlin JR, Harris WH. Revision of the femoral component of a total hip arthroplasty with the calcar-replacement femoral component. Results after a mean of 10.8 years postoperatively. J Bone Jt Surg Am 1996;78:331–9. 20. Mulroy WF, Harris WH. Revision total hip arthroplasty with use of so-called second-generation cementing techniques for aseptic loosening of the femoral component. A fifteen-year-average follow-up study. J Bone Jt Surg Am 1996;78:325–30. 21. Katz RP, Callaghan JJ, Sullivan PM, et al. Results of cemented femoral revision total hip arthroplasty using improved cementing techniques. Clin Orthop Relat Res 1995:178–83. 22. Kavanagh BF, Fitzgerald Jr RH. Multiple revisions for failed total hip arthroplasty not associated with infection. J Bone Jt Surg Am 1987;69:1144–9. 23. Pellicci PM, Wilson Jr PD, Sledge CB, et al. Long-term results of revision total hip replacement A follow-up report. J Bone Jt Surg Am 1985;67:513–6. 24. Kavanagh BF, Ilstrup DM, Fitzgerald Jr RH. Revision total hip arthroplasty. J Bone Jt Surg Am 1985;67:517–26. 25. Mulliken BD, Rorabeck CH, Bourne RB. Uncemented revision total hip arthroplasty: a 4- to-6-year review. Clin Orthop Relat Res 1996:156–62. 26. Berry DJ, Harmsen WS, Ilstrup D, et al. Survivorship of uncemented proximally porous-coated femoral components. Clin Orthop Relat Res 1995:168–77. 27. Woolson ST, Delaney TJ. Failure of a proximally porous-coated femoral prosthesis in revision total hip arthroplasty. J Arthroplasty 1995;10 Suppl:S22–8. 28. Cameron HU. The long-term success of modular proximal fixation stems in revision total hip arthroplasty. J Arthroplasty 2002;17:138–41. 29. Christie MJ, DeBoer DK, Tingstad EM, et al. Clinical experience with a modular noncemented femoral component in revision total hip arthroplasty: 4- to 7-year results. J Arthroplasty 2000;15:840–8. 30. Chandler HP, Ayres DK, Tan RC, et al. Revision total hip replacement using the S-ROM femoral component. Clin Orthop 1995:130–40. 31. Cameron HU. The two- to six-year results with a proximally modular noncemented total hip replacement used in hip revisions. Clin Orthop Relat Res 1994:47–53. 32. Grunig R, Morscher E, Ochsner PE. Three- to 7-year results with the uncemented SL femoral revision prosthesis. Arch Orthop Trauma Surg 1997;116:187–97.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 179–191

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: REVISION HIP ARTHROPLASTY

(iv) Periprosthetic fractures of the hip Sanjeev Patil, Bassam A. Masri, Clive P. Duncan University of British Columbia, Department of Orthopaedics, Division of Adult Reconstruction & Oncology, Room 3114, 910 West 10th Avenue, Vancouver, BC, Canada V5Z 4E3

KEYWORDS Postoperative complications; Femoral fractures; Fractures; Hip prosthesis; Intraoperative complications

Summary Periprosthetic fractures are increasing in number and complexity. Appropriate precautions should be taken to prevent these fractures. A systematic approach is needed in the form of detailed assessment of the fracture, stability of the implant and the available bone stock for planning an appropriate treatment. The treatment options described in this article can be used as a framework for making the right surgical decision regarding appropriate method of reconstruction to ensure optimum result. & 2006 Elsevier Ltd. All rights reserved.

Introduction As the number of primary and revision hip joint replacements increases every year, so the number of ageing patients with joint replacements in place has been increasing steadily over the past few decades. They are frequently associated with polyethylene wear and osteolysis which predispose them to subsequent periprosthetic fractures. being seen quite frequently. Management of these fractures is difficult, complex, expensive and may be associated with complications and pre-existing co-morbidities commonly seen in older patients. Fractures can involve either the acetabulum or femur and can occur during primary or revision arthroplasty, or present at any time thereafter. Intraoperative fractures can be prevented by precise preoperative planning, i.e. by recogni-

Corresponding author. Tel.: +1 604 875 4192;

fax: +1 604 875 4677. E-mail address: [email protected] (B.A. Masri). 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.02.014

tion and treatment of predisposing factors such as significant osteolysis, and by careful surgical technique.

Periprosthetic fractures of the acetabulum There is surprisingly little reported in the literature about periprosthetic fractures of the acetabulum. They can be broadly grouped as perioperative or postoperative fractures.

Perioperative fractures The overall incidence is not known. McElfresh and Coventry1 reported a 0.2% incidence among 5400 cemented hip arthroplasties at the Mayo Clinic. Nor is there a specific classification system. Callaghan2 in an in vitro study described various fracture patterns including anterior wall, posterior wall, inferior lip, transverse and isolated column fractures. Intraoperative fractures of the acetabulum are commonly due to excessive undersizing of the reamed cavity and the subsequent need to use excessive force to seat an uncemented acetabular shell. Callaghan et al.,2 in an

ARTICLE IN PRESS 180

S. Patil et al.

in vitro study reported a greater than 50% fracture rate when under-reaming was more than 4 mm and a 12% fracture rate when it was limited to 2 mm. Intraoperative fractures may also occur due to a combination of excessive reaming and shell impaction force in the presence of a very weakened or deficient acetabulum due to osteolysis or osteoporosis. Most commonly this is through the floor of the acetabulum. Fractures recognised during surgery should be evaluated further by meticulous exposure of the fracture site. Any involvement of the columns should be identified as it affects the management. Routine use of intraoperative radiographs is not essential for this evaluation.

Postoperative fractures

Prevention During insertion of press-fit acetabular components, underreaming should be kept to 2 mm or less particularly for smaller sizes or for patients with bone loss due to osteoporosis or osteolysis. Appropriate hemisphere reaming of the acetabulum should be performed to avoid intraoperative fractures. In osteopenic bone one should consider under-reaming by 1 mm or even not at all and the routine use of screws. In revision cases careful reaming should be performed to avoid decreasing the bone stock further. If bone stock is to be sacrificed in order to prepare a hemispherical cavity, the posterior column and the dome should be preferentially protected at the expense of the medial wall and anterior column. Ideally as little host bone should be sacrificed as possible. In the presence of decreased bone stock, under-reaming by more than 2 mm can potentially result in pelvic discontinuity during the insertion of press-fit acetabular component and great care should be taken to avoid this potentially devastating complication. With the increasing use of acetabular shells that have an elliptical shape (such as trabecular metal shells and surface replacement components) under-reaming is unwise as the nominal size of the component is commonly 1–2 mm less than the true rim diameter. This is particularly important to understand in the revision scenario where the surgeon is dealing with weakened bone.



Treatment If a fracture is obvious during operation, careful evaluation of the fracture and implant stability should be made. Fractures with no involvement of the posterior column and a stable acetabular component can be treated by augmentation with screws. If the acetabular component is unstable and is associated with a posterior column fracture, buttress plating of the posterior column using a 3.5 mm pelvic reconstruction plate should be performed, in addition to augmentation of acetabular component fixation with screws. Many of these fractures are not obvious at the time of operation. If the fracture and the implant are judged to be stable, management can be expectant, with protected weight bearing. Otherwise, they should be managed as a postoperative fracture, see below.

The figures relating to the true incidence of these fractures are not available as very few reports exist in the literature. Berry3 reported a 0.9% prevalence (31of 3105) of postoperative fractures with pelvic discontinuity. They can result from:

 Trauma: Traumatic postoperative periacetabular fractures are rare.

 Secondary to osteolysis which is a major predisposing factor: Periodic follow-up of arthroplasty patients is helpful to detect any progressive osteolysis. Early intervention in these cases helps to prevent their progression to periacetabular fractures. Stress fractures: Stress fractures of the weakened medial wall of the acetabulum can be seen after cementless revisions.4 Patients with arthritis can have preoperative osteopenia due to inactivity. These patients can have stress fractures of the pubis due to increased activity levels after hip replacement. Peterson and Lewallen5 classified them as follows:

 Type I: Fractures associated with clinically and radiologically stable acetabular component.

 Type II: Fractures associated with unstable acetabular component.

Diagnosis Postoperative fractures should be included in the differential diagnosis of patients presenting with acute onset of groin pain. There may not be a history of trauma especially in patients with osteolysis. Osteopenia is not uncommon in elderly patients with arthritis due to immobility. Postoperative fractures in such patients due to increased activity level after surgery should be kept in mind. One should look for pelvic discontinuity in these patients. In most cases the discontinuity represents a transverse acetabular fracture non-union (Fig. 1). Pelvic discontinuities are rare but make for very difficult reconstructions. When postoperative fractures are detected on standard radiographs, they should be assessed further by Judet views and CT scans. The CT scans not only facilitate in fracture evaluation but also help in evaluation of the bone loss due to osteolysis and thereby assist in planning appropriate treatment. Juxta-articular insufficiency fractures can involve sacrum, ileum or rami. If they are not obviously seen on standard radiographs, isotope bone scans may be useful in alerting the surgeon. Treatment Fractures associated with a stable acetabular component (Type I fractures) detected during the early postoperative period can be treated conservatively with protected weight bearing for 6–8 weeks. Similar management can be done for late presenting type 1 fractures. This will allow the fracture to heal and may simplify the subsequent acetabular revision which is needed for the majority of these patients.5

ARTICLE IN PRESS Periprosthetic fractures of the hip

181

Figure 1 Pelvic discontinuity: (a) the black arrow points towards pelvic discontinuity; (b) postoperative radiograph showing reconstruction using pelvic reconstruction plate.

Figure 2 Intraoperative photograph showing reconstruction of posterior column fracture by plate.

These fractures are treated conservatively with activity modification until complete healing is noted, unless the injury has been so severe as to de-stabilise the acetabular component. In that instance early revision is required unless, in the case of a loosened but stable implant, it would seem the best judgment to await fracture union followed by revision. Fractures associated with an unstable acetabular component (Type II fractures) need revision surgery. The preoperative planning should include a detailed assessment of fracture pattern and remaining bone stock.

The goals of treatment are to provide structural integrity of the columns and to restore acetabular bone stock to provide a stable bed for the acetabular component. Isolated posterior column fractures should be fixed with a 3.5 mm pelvic reconstruction plate spanning from the ilium to the ischium (Fig. 2). Isolated anterior column fractures do not require any treatment unless they are so superior that a substantial portion of the acetabular roof is lost. In that case, treatment should be directed at restoring dome integrity using allograft or trabecular metal augments (Zimmer, Warsaw, IN). Isolated medial wall fractures or perforations do not need fixation. They can be treated by wafers of allograft combined with impacted morcellised allograft bone. Once the fracture is dealt with, a detailed assessment of acetabular bone stock should be performed. If there is loss of less than 50% acetabular host bone, porous coated hemispherical cups can be used for reconstruction. If there is more than 50% host bone loss, one should consider the use of either the combination of trabecular metal augments and trabecular metal revision cups (Zimmer, Warsaw, IN) or reconstruction cages.6 The cage should span from ilium to ischium and a polyethylene cup should be cemented within it. The cage should rest on the bone inferiorly, posteriorly and superiorly to prevent its future failure; hence, this type of reconstruction should be accompanied by restoration of bone stock either by bulk or morcellised allograft. The cage not only stabilises the fracture but in addition assists in graft incorporation. Every effort should be made to plate the posterior column. If this is not possible, the inferior flange of the cage should be fixed to the surface of the ischium with screws. Bone cement

ARTICLE IN PRESS 182

S. Patil et al.

Table 1 Risk factors for intraoperative and postoperative periprosthetic femoral fractures. General factors Female sex Osteoporosis: primary, secondary Osteopenia Rheumatoid arthritis Osteomalacia Paget’s disease Figure 3 Safe removal of the acetabular component through ileoinguinal approach. Arrows represent common iliac vessels.

augmentation of screw fixation may be required in these cases of much weakened bone. Postoperative fractures due to acute trauma should be assessed and treated individually. In cases where there is intrapelvic migration of the acetabular component, one should carry out a preoperative vascular assessment and strongly consider extracting the cup from a separate ilioinguinal approach (Fig. 3). Berry et al.7 reported on the treatment of patients with pelvic discontinuity by using reconstruction cages, uncemented components and cemented components without cages. Satisfactory outcome was seen in 77% of patients in the cage group and 56% of the patients in the uncemented group. None of the patients undergoing reconstruction using cemented components without a cage had a satisfactory outcome.

Periprosthetic fractures of the femur Periprosthetic fractures of the femur can be broadly grouped into perioperative and postoperative fractures.

Perioperative fractures These fractures can occur during primary or revision hip replacement but they tend to be more common in the revision setting. They are also more commonly associated with the use of uncemented femoral components. Berry3 reported a 1% prevalence (238 of 23,980) of these fractures during primary joint arthroplasty. About 71% of these fractures were associated with uncemented implants. The prevalence of these fractures was 7.8% (497 of 6349) during revision hip joint replacements. Sixty-five percent of these fractures during revision procedures were associated with the use of uncemented femoral components. Table 1 summarises the risk factors predisposing to periprosthetic femoral fractures. Several classifications of periprosthetic femoral fractures have been described. Most are based upon the level of fracture in relation to prosthesis. Some consider the fracture pattern, type of implant (cementless versus cemented) or the timing of the fracture (intraoperative versus postoperative fractures). An ideal classification system should assist in planning the treatment and

Local factors Stress risers within cortex Loose prosthesis Localised osteolysis Cortical perforation Revision arthroplasty Cementless femoral component

Osteogenesis imperfecta Thalassemia Clinical conditions resulting in loss of balance/falls Parkinsonism Neuropathic arthropathy Poliomyelitis Myasthenia gravis Cerebral palsy Seizures Ataxia

comparison of outcomes between different centres. The Vancouver classification for both intraoperative and postoperative fractures has been widely accepted in this respect.

Intraoperative fractures The Vancouver classification for intraoperative fractures was described by the senior authors Masri et al.8 in 2004. This is based upon the level of fracture and assists in planning the treatment. Type A fractures are proximal metaphyseal in location and do not extend into the diaphysis. Type B fractures are diaphyseal but these fractures do not extend into the distal diaphysis and hence can be bypassed by a long stem revision prosthesis. Type C fractures are distal diaphyseal or metaphyseal in location and the distal extent of these fractures precludes bypassing with even the longest revision stem. The fractures are then subdivided based upon fracture configuration and stability. Subtype 1 refers to a simple cortical perforation. These are common in the diaphyseal region (B1). These injuries tend to occur mainly during revision surgery at the time of cement removal or femoral canal preparation. Subtype 2 refers to an undisplaced linear crack. These are common in the proximal metaphysis (A2) and in the diaphysis (B2). They tend to occur during insertion of a rasp or at the time of cementless component insertion. Displaced or an unstable fracture pattern is designated as subtype 3. A3 fractures tend to occur during insertion of a rasp or cementless femoral component. They can also be as a result of inadvertent removal of the femoral component without clearing overhanging trochanter during revision hip surgery. B3 fractures commonly are seen during aggressive

ARTICLE IN PRESS Periprosthetic fractures of the hip

183

reaming or forceful torque applied to the lower limb either during dislocation or relocation of the hip. C3 fractures are rare but can occur during attempted forceful dislocation of the hip joint like in protrusio acetabuli.

upon the stability of the femoral component. Meticulous exposure of the fracture is vital, and the Vancouver classification system described earlier in this article aids in the management of these fractures.

Prevention

Type A intraoperative fractures (proximal metaphyseal) Type A1: cortical perforation. These injuries are unlikely to compromise the fixation of the prosthesis or increase the risk of postoperative fractures. They can be managed by application of locally harvested autografts, such as from acetabular reaming or can even be ignored. Type A2: undisplaced linear crack. These fractures are best treated by cerclage wires, performed as soon as the crack is noted and before the insertion of the final femoral component. If a proximally coated stem is used, its stability should be assessed carefully and if the fracture has compromised implant stability, it should be replaced with a fully porous-coated stem. If a cemented stem is used, any cement extruded within the crack should be removed. These fractures, once stabilised do not compromise the final outcome. Type A3: displaced fractures of the proximal femur. These fractures tend to compromise the fixation of standard femoral components, especially if they are proximally coated stems. These fractures need to be exposed and secured appropriately with cerclage wires. Diaphyseal fitting cementless stems should be used in the treatment of these fractures. Displaced fractures of the greater trochanter should be secured using cerclage wires or other trochanteric fixation devices. Displaced fractures of the greater trochanter can be associated with ETO employed during revision arthroplasty setting. Such fractures typically occur when the osteotomised fragment is being reapproximated with a cerclage wire. As a diaphyseal fitting stem is commonly used in this setting, the fracture fragments need to be just approximated, overlapped by a cortical allograft strut and secured with cerclage wires onto the prosthesis. In addition, the trochanteric fragment needs to be stabilised by using one of the methods described above. However, this fracture can be prevented using a cortical strut graft placed under the cerclage wire to dissipate the focal forces.

The incidence of periprosthetic fractures of the femur can be minimised by identifying the risk factors and addressing them appropriately. Detailed preoperative planning is an essential step in this process, which should include choice of surgical approach and of implant and back-up plans if the primary plan fails. Detailed clinical and radiographic assessment is essential. It is useful to have up-to-date radiographs as many risk factors like osteolysis and varus remodelling of the femur in revision arthroplasty setting are progressive. Intraoperative fractures in revision surgery commonly tend to occur during dislocation, cement removal, femoral canal preparation, trial reduction and insertion of the femoral component. It is important to use finesse rather than force as undue retraction and torsional forces on the bone that has been weakened by osteolysis can lead to intraoperative fracture. Dislocation of the hip should ideally use minimal force after adequate soft tissue release has been performed. It is wise to leave previous implants like plates and screws until the dislocation has been done to prevent any stress risers. Trochanteric overhang, if present, should be corrected by adequate clearance of overhanging bone before attempting to extract the femoral stem to avoid intraoperative trochanteric fracture. During revision of a femoral stem, an extended trochanteric osteotomy (ETO) facilitates the exposure. In addition, it assists in easy cement removal and reduces the risk of cortical perforation. Use of a guide wire and intraoperative radiographs help to avoid cortical perforation during revision surgery. Under-reaming can be dangerous especially in osteopenic bone as it can result in fracture while using long uncemented stems. We recommend under-reaming by 0.5 mm for diaphyseal fitting stems unless a bowed stem is to be used, in which case, so-called line-to-line reaming is preferred. If the bone is weak, prophylactic cerclage wires can be used to prevent fractures. Although use of uncemented stems in the revision setting has been shown to be associated with an increased incidence of fractures, no difference has been noted regardless of whether a straight or a bowed stem is used. Furthermore, the presence of an intraoperative fracture during revision total hip arthroplasty with a fully porous-coated cementless stem has not been shown to affect the final functional outcome.10 Periodic clinical and radiological review of patients after joint replacement helps to identify progressive osteolysis. Early intervention in these patients can prevent postoperative femoral fractures.

Treatment of intraoperative fractures The aim is to achieve near-anatomic fracture fixation and stable fixation of the prosthesis. The treatment depends not only upon the fracture location and configuration, but also

Type B intraoperative fractures (diaphyseal) Type B1: diaphyseal cortical perforation. These fractures act as stress risers and should be treated aggressively. They can be managed by bypassing the perforation with a longer stem by at least two cortical diameters. Cerclage wire applied at or just distal to the perforation prior to the insertion of the stem will prevent crack propagation. If the perforation is at the tip of the longest cementless stem it should be treated by application of an allograft strut. Type B2: undisplaced linear fracture. These fractures should be fixed by cerclage wires and bypassed with a longer cementless stem. If this is not possible, one should use cortical onlay graft with cerclage wires to secure the fixation. If these fractures are detected postoperatively, one should consider protected weight bearing for 6–12 weeks to allow the fracture to heal. Type B3: displaced fracture of the mid femur. These fractures should be stabilised by meticulous open reduction

ARTICLE IN PRESS 184 and internal fixation. Long oblique or spiral fractures can be treated by cerclage wires. However, comminuted fractures need to be reinforced with one or two cortical struts. The fracture should be bypassed with a long cementless stem by at least two cortical diameters. When these fractures occur distal to a well-fixed cementless femoral component and the stem is difficult to extract without compromising the bone stock, they can be managed by application of cortical onlay grafts and cerclage wires. Type C intraoperative fractures (distal diaphyseal/ metaphyseal) By definition, this fracture cannot be bypassed by the longest revision stem. C1 (cortical perforation) and C2 (linear crack) fractures tend to act as significant stress risers and hence need treatment in the form of cortical struts and cerclage wires. It is important to overlap the tip of the prosthesis with cortical struts to avoid any stress riser. The displaced (C3) fractures can be treated by the same technique or by using plate and strut fixation.

Fractures diagnosed in the immediate postoperative period Complex fractures are rarely diagnosed during the immediate postoperative period as they are almost always detected at the time of surgery. If seen on the postoperative radiograph, they should be treated by reoperation and the application of the same principles as described earlier. It is more common to see undisplaced fractures during the immediate postoperative period. These fractures should be imaged carefully to define the whole extent of the fracture. Non-operative treatment in the form of protected weight bearing for 6–12 weeks suffices in most cases without any compromise of the final outcome.

S. Patil et al.

Postoperative fractures The prevalence of postoperative fractures in the literature ranges from 0.1% to 2.1%.8 The largest series from the Mayo Clinic joint registry reports a prevalence of 1.1% (262 of 23,980) in primary arthroplasty and 4% (252 of 6349) in the revision setting. The most common cause of these fractures is often an episode of minor trauma. Beals and Tower reported minor trauma as a cause of 84% of the fractures (72 of 86). Only 8% of the fractures in their series were due to major trauma.9 Incidence of these fractures is expected to increase due to the increasing numbers of patients with hip replacements, and a growing number of them with compromised bone around the component due to osteolysis or revision procedures. Additionally, with ageing, these patients are at increased risk of falling. Classification Duncan and Masri11 published the Vancouver classification for postoperative fractures in 1995 based upon three important factors in the management; fracture location, stability of the implant and the quality of the femoral bone stock. This classification is widely accepted and has been shown to be both reliable and valid.12 It has three main categories A, B and C based upon the fracture location. Type A fractures occur in the peritrochanteric region and are subdivided based upon the involvement of either the greater trochanter (AG) or lesser trochanter (type AL). Type B fractures occur around the stem or just distal to the tip of the prosthesis. Type B fractures are subdivided into three categories based upon stability of the prosthesis and quality of the bone stock. Subtype B1 represents fractures associated with well fixed femoral components (Fig. 4). Fractures with loose femoral components are

Figure 4 B1 fracture: (a) ununited B1 fracture, treated earlier by combination of strut graft and cable grips; (b) follow-up radiograph showing successful treatment by using cable-plate fixation.

ARTICLE IN PRESS Periprosthetic fractures of the hip

185

Figure 5 B2 fracture: (a and b) B2 fracture around an uncemented femoral component; (c) treatment by using diaphyseal fitting stem and cerclage wires.

classed as subtype B2 (Fig. 5). B3 subtype includes fractures with severe femoral bone stock deficiency due to osteolysis or severe comminution or generalised osteopenia (Fig. 6). B3 fractures are commonly associated with a loose femoral component. At times the distinction between B2 and B3 fractures can only be done at the time of surgery. Type C fractures are located well distal to the tip of the prosthesis. The original article reported 75 patients with periprosthetic femoral fractures, wherein 4% of the fractures were type A, 86.7% were type B, and 9.3% were type C. Type B included 18.5% B1 fractures, 44.6% B2 fractures and 36.9% fractures.

Treatment of postoperative fractures The possibility of infection should always be considered while treating periprosthetic fractures of the femur. The presence of a fracture makes measurement of serological markers such as the ESR and C-reactive protein potentially unreliable. Thus if suspicion of a pre-existing covert infection is high based on the wound history from the last procedure, constitutional symptoms, local physical findings or radiographic parameters, diagnosis is made by performing an aspiration biopsy of the periprosthetic space. If this confirms infection, one should consider two-stage exchange method using an intermediate articulated antibiotic loaded spacer like PROSTALAC (DePuy, Warsaw, IN) to control the infection and to stabilise the fracture. The options in the management of postoperative periprosthetic fractures of the femur include non-operative methods, closed reduction and internal fixation, open reduction and internal fixation, standard revision hip arthroplasty with or without bone graft, and complex reconstruction sometimes including segmental replacement of the proximal femur.

Non-operative treatment We no longer advocate prolonged recumbence with traction or cast immobilisation due to associated complications such as atelectasis, pneumonia, pressure sores, thrombo-embolic disease, joint stiffness and disuse osteoporosis. Additionally, non-surgical treatment is commonly associated with malunion and non-union, making for increased technical difficulty during the next procedure. We recommend nonsurgical treatment only for undisplaced fractures with stable prostheses. Such patients can be treated with protected weight bearing and/or functional bracing as required. Patients with stable AG postoperative fractures are ideal for non-operative care. We would only rarely propose nonoperative management for a C type fracture.

Principles of operative treatment The Vancouver classification facilitates preoperative planning. Routine antibiotic and deep vein thrombosis prophylaxis is essential. The treating surgeon should be well versed in extensile approaches to the hip joint. Adequate exposure of the fracture is essential. However, intraoperative tissue specimens should be obtained for culture and antibiotics should be withheld until this step has been completed. Separation of the fracture fragments can aid in removal of cement debris from the canal and also canal preparation required for revision stem insertion. Judicious use of intraoperative radiographs may be helpful. Methods of operative treatment There are three broad generic approaches to treatment

 fracture fixation alone  fixation combined with revision replacement  complex reconstruction such as modified impaction grafting or proximal femoral replacement.

ARTICLE IN PRESS 186

S. Patil et al.

Figure 6 B3 fracture: (a) B3 fracture with evidence of gross bone stock deficiency of proximal femur; (b and c) treatment by using long uncemented femoral component and cortical strut graft along with cerclage wires.

There is no single approach that is applicable to all cases. We will discuss each treatment option in detail followed by a treatment algorithim based upon the Vancouver classification. Cerclage wire fixation Monofilament wire or braided cable used as cerclage fixation can be considered for the fixation of long oblique or spiral fractures around a well-fixed prosthesis. However on its own, this is biomechanically weak and does not provide the torsional or bending rigidity needed to reliably prevent fracture displacement. The exception is the management of a linear crack in the proximal femur occuring during

insertion of the femoral component. The technique can also be used to reinforce the femur to prevent initiation or propagation of such cracks during canal preparation. Cerclage fixation is usually combined with a plate or a strut graft to provide a more rigid construct. Plating Internal fixation with plates is indicated in diaphyseal fractures that occur around well-fixed implants (Vancouver B1 fractures and Vancouver type C fractures (distal diaphyseal–metaphyseal). As varus alignment of the stem is a contraindication for plate fixation,13 it is prudent to ensure that the prosthesis is in neutral alignment. The plate

ARTICLE IN PRESS Periprosthetic fractures of the hip should always overlap the intramedullary stem to prevent a stress riser between the two. It is also desirable to have fixation of a minimum eight cortices screw fixation on either side of the fracture to obtain a rigid reconstruction. It is often difficult in B1 fractures to achieve this with ease due to the presence of a canal filling cementless stem or the fear of violating the cement mantle in cemented implants. Options might include directing the screws posteriorly into the thicker bone underneath the linea aspera or using unicortical screws. To address the difficulties encountered in the proximal fixation, various cable plate designs (Ogden plate, Dall-Miles cable-plate system, etc.) have been introduced. These use cerclage fixation for the portion of the plate that spans the stem. The wires can either be locked into the plate itself or by a crimpling sleeve after adequate tensioning using a special tensioning device. Holes in plates allow additional screw placement of adequate length proximally as allowed by the prosthesis and bicortical screws distal to the tip of the stem. Dennis et al.14 compared five different options for plate fixation about femoral stems: plate with cable only, plate with proximal cable and distal bicortical screws, plate with proximal cable plus unicortical screws and distal bicortical screws, plate with proximal unicortical screws and distal bicortical screws and two allograft struts fixed with cables. These specimens were tested in torsion, lateral bending and axial compression. The constructs with proximal unicortical screws (with or without cerclage wires) were more stable than other constructs.

187 Locking screw plates have been recently introduced for the fixation of periprosthetic femoral fractures (Fig. 7).15 These plates act as internal–external splints and are especially useful in the presence of osteporotic bone. The fact that unicortical screws can provide adequate stability in this construct helps in fixation proximal to the fracture around the intramedullary stem. The forces are distributed over the whole plate increasing the pull-out strength of each screw. As they do not rely on the compression between the plate and the underlying bone, the periosteal vascular supply is least disturbed. They act as fixed-angle constructs and thereby improve the angular stability of the fracture. The modern designs with combination holes allow either fixed-angle locking screws or standard unicortical or bicortical compression screws, an important feature proximally where a large intramedullary stem can be avoided by angling the standard unicortical or bicortical screws. MIPO: Minimally invasive plate osteosysthesis This involves the least biological disturbance to the fracture site.16 In traditional open reduction, wide exposure of the fracture site results in evacuation of the fracture haematoma and stripping of the periosteum and soft tissue attachments to the fracture fragments which disturbs the local vascular supply to the fracture fragments and can have a negative influence on fracture healing. MIPO involves indirect reduction of the fracture by using a fracture table and an image intensifier. A standard or a locking plate is inserted by placing a small incision distal or proximal to the fracture and advanced submuscularly to span the fracture

Figure 7 B1 fracture: (a) B1 postoperative fracture around an uncemented femoral component; (b) fixation using Locking plate (Combi plate).

ARTICLE IN PRESS 188 site. The screws are commonly inserted through stab wounds with image intensifier guidance. This technique can be used in fractures with large areas of comminution or long oblique fractures or spiral fractures where the bending forces through the fracture tend to be over a large area.16 Ricci et al.,17 reviewed 41 Vancouver B1 fractures treated by open reduction using indirect reduction techniques and internal fixation with single lateral plate fixation at an average follow up of 24 months. They did not use cortical strut grafts to supplement the fixation. All the patients had satisfactory fracture union at an average period of 12 weeks without any evidence of implant loosening or malalignment. They concluded that care in preserving the soft tissue envelope around the fracture by the minimally invasive techniques led to consistent healing times.

Cortical allograft struts Cortical struts are generally obtained from freeze-dried femurs or tibias. They have the ability to augment the strength and stability of the construct and can be appropriately prepared to custom-fit any femur. As the modulus of elasticity of cortical struts is similar to host bone there is less stress shielding of the host bone in comparison to rigid fixation with devices like plates.18 They can be used alone as a biological bone plate, used to augment fixation with a metal plate or can be used to augment the junction between host bone and a massive structural allograft. Attention to some of the technical details is important to attain adequate reconstruction with cortical struts. We routinely use two allograft struts if no plate is used. However, a single allograft strut is used when augmenting fracture fixation with a plate. We ream the medullary canal of the allograft (femur or tibia) before sectioning it, until it is large enough to gain snug fixation over the host femur. We fashion each strut to about 16 cm in length and one third the circumference of the host femur. The two struts can be placed parallel to each other over medial and lateral surfaces of the femur or they can be placed perpendicular to each other over anterior and lateral surfaces of the host femur. We prefer the latter method as it entails less soft tissue stripping in comparison to the parallel technique. It is important to avoid edge contact between the two struts to prevent dissipation of the compressive force generated by the cerclage cables between the struts and thereby decreasing the stability of the construct. Three or four evenly spaced 2 mm multifilament cables should be passed around the host femur on either side of the fracture prior to positioning the struts. Morcellised allograft is generally packed within the medullary cavity of the strut before applying it on to the host femur. Central cables should be tightened first followed by peripheral cables to avoid displacement of the struts. Excessive tensioning should be avoided as it may fracture the allograft. Adequate soft tissue cover of the construct should be obtained with vastus lateralis muscle. The strength of the construct by using cortical struts and cables has been investigated by various authors. In a biomechanical study from our centre, Haddad et al.19 reported the effect of number of cables, cable tension, cable versus wires, strut configuration and strut length over

S. Patil et al. the rigidity of the construct. They concluded that at least three or more cables should be used on both sides of the fracture, the cable tension should be high, cables instead of wires should be used and that two struts should be used. They also reported that the strut length was inversely proportional to the rigidity of the construct. The advantages of cortical struts include their ability to heal to the host bone and remodel over time thereby increasing bone stock. The disadvantages include the prolonged time required for graft incorporation during which they are prone to stress fracture or loosening. They can also be a potential source of infection. Maximal weakness of these struts occurs between 4 and 6 months. Emerson et al.20 analysed 63 struts that had sufficient serial radiographs. They reported 96.6% union rate at 8.4 months. Cortical allografts have been used in combination with plates. A biomechanical study21 from our unit investigated different methods of fixation of a B1 fracture model using, lateral plate and anterior allograft, lateral plate fixation alone and lateral and anterior allograft struts. The results showed that the rotational and translational stability was highest in strut-plate combination constructs. In a multicentre study, Haddard et al.18 reported a 98% fracture union rate among 40 B1 postoperative periprosthetic femoral fractures patients treated by cortical strut grafts alone (19 patients) or combination of cortical strut graft and plate (21 patients). They concluded that the cortical strut grafts should be used routinely to augment fixation and healing of a periprosthetic femoral fracture.

Revision hip arthroplasty Revision hip arthroplasty is the technique of choice when a fracture is associated with a loose femoral component. Long cementless stems with distal fixation to bypass the fracture by at least two cortical diameters with or without cortical strut grafting is our method of choice. Adequate exposure of the fracture site to access the femoral canal for canal debridement and preparation minimises soft tissue stripping. The use of a long cemented stem may be appropriate in the frail elderly patient with limited functional demands and lifespan, with poor bone stock and a simple fracture pattern. Every effort should be made to limit the ingress of cement into the fracture interface, and morcellised bone should be packed around the fracture site. Modular cementless stems or extensively porous-coated cylindrical implants are used frequently to obtain distal fixation. The remaining proximal bone is usually overlapped onto the prosthesis and augmented with cortical struts. Distally fixed tapered cementless prostheses have been successfully used in Vancouver type B2 and B3 fractures. Berry22 in his series of seven B3 fractures treated by modular, tapered and distally fluted stems reported that, at short-term follow-up of 1.5 years all the patients had fracture union and stable implants.

Proximal femoral replacement This technique using a segmental allograft femur or a tumortype proximal femoral replacement is reserved for post-

ARTICLE IN PRESS Periprosthetic fractures of the hip

189

Treatment algorithm for postoperative periprosthetic femoral fractures This algorithm is based upon the Vancouver classification system for postoperative fractures (Fig. 9).

Figure 8 fracture.

Tumor prosthesis in the treatment of complex B3

operative Vancouver B3 fractures associated with severe, segmental bone loss. Elderly patients with limited life expectancy and associated poor bone stock can be treated by a tumor-type proximal femoral replacement prosthesis (Fig. 8). We no longer resect the proximal femur but bi-valve it (usually in the sagittal plane) and wrap it around the femoral component during closure. This facilitates reattachment of the greater trochanter and abductors as well as quadriceps function. This allows early mobilisation which is of paramount importance in this group of patients. However, in physiologically younger patients allograftprosthesis composite is desirable for obvious reasons. Wong and Gross23 have reported the use of this technique in 19 patients with B3 fractures. Thirteen patients of the 15 available for review at a mean follow-up of 5 years had a good result. It is notable that the use of segmental proximal femoral placements has declined in North America since the introduction of more sophisticated femoral components with multimodularity and improved fatigue resistance.

Postoperative management Postoperative management should be individualised, based upon the method of reconstruction used. Protected weight bearing is often advised for a period of 6–12 weeks until clinical and radiological union at the fracture site is obtained.

Type A postoperative fractures Type AG fractures are generally stable and they can be treated by protected weight bearing for 6–12 weeks. Active abduction is also avoided during this period. If the displacement of the fracture is greater than 2.5 cm, one should consider internal fixation. We prefer claw grip and cables over isolated monofilament wire fixation. If the fracture is secondary to severe osteolysis, the source of the osteolysis is addressed by liner exchange or cup revision in addition to managing the fracture. It is often impossible to fix these fractures associated with severe osteolysis. During operation, attempts should be made to preserve the soft tissue envelope of the abductors and the vasti as a single sleeve of tissue because it is a valuable fracture stabiliser in these cases24 and augments blood supply at the fracture site. Use of the trochanteric slide for exposure of the hip should be considered.25 Small cortical windows aid in bone grafting of these defects. Type AL fractures are usually minor and do not need surgical intervention. If they are major and involving the calcar femorale, they can affect implant stability due to lack of medial support, hence one should consider fixation using cerclage wires. Type B1 postoperative fractures If possible, we currently prefer fixation using minimally invasive plate osteosynthesis techniques to preserve local vascularity. If closed reduction of the fracture is not possible, we consider open reduction and internal fixation using a lateral plate and often anterior cortical strut allograft or two cortical strut grafts placed perpendicular to each other on anterior and lateral surfaces. We do not revise the stem unless it is in varus malalignment as this might result in treatment failure. Type B2 postoperative fractures We prefer to use long cementless femoral stems to bypass the fracture by at least two cortical diameters. Long cemented stems can also be used in older patients with due care to avoid extrusion of cement into the fracture site. We prefer to augment the fixation with a cortical allograft strut in the presence of unstable transverse fractures. This is often not necessary while treating long oblique or spiral fracture configurations as cerclage wires can provide adequate rotational stability. Type B3 postoperative fractures If the bone deficiency is not severe and there is some remaining supportive proximal femur, we prefer to use modular cementless tapered femoral components along with cortical struts. However if the deficiency is severe, segmental and not reconstructable, we prefer to use a segmental allograft/prosthesis composite in young patients and a tumor-type proximal femoral replacement prosthesis in low demand elderly patients.

ARTICLE IN PRESS 190

S. Patil et al. LOCATION OF THE FRACTURE

Type A (TROCHANTERIC FRACTURE) AG: Greater trochanter, symptomatic treatment with crutches and limited abduction Intervention only if displaced to avoid pain, weakness, limp or instability AL: Lesser trochanter, symptomatic treatment only even if displaced. Intervene if substantial segment medial cortex attached

Type B

Type C

(FRACTURE AT OR NEAR STEM TIP)

(FRACTURE WELL BELOW STEM) Ignore implant, fix fracture, if necessary address implant after fracture healed

FIXATION OF STEM

LOOSE STEM (B2)

WELL FIXED STEM (B1) Open reduction and internal fixation of fracture using any one or combination of techniques, or closed reduction and plate fixation (MIPO)

IS THERE GOOD BONE STOCK ?

YES (B2) Revision with long stem

Figure 9

NO (B2) Loose stem and poor bone stock. Revision and augmentation of bone stock with allograft in physiologically young or tumor prosthesis in elderly patient

Treatment algorithim based upon the Vancouver classification of postoperative periprosthetic femur fractures.

Type C postoperative fractures These fractures are treated with standard fracture fixation techniques employed in the treatment of distal femoral fractures. Minimally invasive plate osteosynthesis or bridge plating can be employed in the treatment of these fractures. It is important not to leave a short segment of bone between the plate and the tip of the femoral component above as it can act as a significant stress riser and predispose to later fracture.

Conclusion Periprosthetic fractures are increasing in number and complexity. Appropriate precautions should be taken to prevent these fractures. A systematic approach is needed in the form of detailed assessment of the fracture, stability of the implant and the available bone stock for planning an appropriate treatment. The treatment options described in this article can be used as a framework for making the right surgical decision regarding appropriate method of reconstruction to ensure optimum result.

References 1. McElfresh EC, Coventry MB. Femoral and pelvic fractures after total hip arthroplasty. J Bone Joint Surg 1974;56A:483–92.

2. Callaghan JJ. Periprosthetic fractures of the acetabulum during and following total hip arthroplasty. J Bone Joint Surg 1997; 79B:1416–21. 3. Berry DJ. Epidemiology of periprosthetic fractures after major joint replacement: hip and knee. Orthop Clin North Am 1999;30:183–90. 4. Andrews P, Barrack RL, Harris WH. Stress fracture of the medial wall of the acetabulum adjacent to a cementless acetabular component. J Arthroplasty 2000;17:117–20. 5. Peterson CA, Lewallen DG. Periprosthetic fracture of the acetabulum after total hip arthroplasty. J Bone Joint Surg 1996;78A:1206–13. 6. Goodman S, Saastamoinen H, Shasha N, Gross A. Complications of Iliooschial reconstruction rings in revision total hip arthroplasty. J Arthroplasty 2004;19(4):436–46. 7. Berry DJ, Lewallen DG, Hanssen AD, et al. Pelvic discontinuity in revision total hip arthroplasty. J Bone Joint Surg 1999;81A: 1692–702. 8. Masri BA, Meek RM, Duncan CP. Periprosthetic fractures evaluation and treatment. Clin Orthop Rel Res 2004(420):80–95. 9. Beals RK, Tower SS. Periprosthetic fractures of the femur: an analysis of 93 fractures. Clin Orthop 1996;327:238–46. 10. Meek RM, Garbuz DS, Masri BA, Greidanus NV, Duncan CP. Intraoperative fracture of the femur in revision total hip arthroplasty with a diaphyseal fitting stem. J Bone Joint Surg Am 2004;86-A(3):480–5. 11. Duncan CP, Masri BA. Fractures of the femur after hip replacement. Instr Course Lect 1995;45:293–304. 12. Brady OH, Garbuz DS, Masri BA, et al. The reliability and validity of the Vancouver classification of femoral fractures after hip replacement. J Arthroplasty 2000;15:59–62.

ARTICLE IN PRESS Periprosthetic fractures of the hip 13. Tadross TS, Nanu AM, Buchanan MJ, Checketts RG. Dall-Miles plating for periprosthetic B1 fractures of the femur. J Arthroplasty 2000;15(1):47–51. 14. Dennis MG, Simon JA, Kummer FJ, Koval KJ, DiCesare PE. Fixation of periprosthetic femoral shaft fractures occurring at the tip of the stem: a biomechanical study of 5 techniques. J Arthroplasty 2000;15(4):523–8. 15. Kolb W, Guhlmann H, Friedel R, Nestmann H. Fixation of periprosthetic femur fractures with the less invasive stabilization system (LISS)—a new minimally invasive treatment with locked fixed-angle screws. Zentralbl Chir 2003;128(1): 53–9. 16. Farouk O, Krettek C, Miclau T, Schandelmaier P, Guy P, Tscherne H. Minimally invasive plate osteosynthesis: does percutaneous plating disrupt femoral blood supply less than the traditional technique? J Orthop Trauma 1999;13(6):401–6. 17. Ricci WM, Bolhofner BR, Loftus T, Cox C, Mitchell S, Borrelli Jr J. Indirect reduction and plate fixation, without grafting, for periprosthetic femoral shaft fractures about a stable intramedullary implant. J Bone Joint Surg Am 2005;87(10): 2240–5. 18. Haddad FS, Duncan CP, Berry DJ, et al. Periprosthetic femoral fractures around well-fixed implants: use of cortical onlay allografts with or without a plate. J Bone Joint Surg 2002; 84A:945–50.

191 19. Haddad FS, Dehaan MN, Brady O, Masri BA, Garbuz DS, Goertzen DJ, Oxland TO, Duncan CP. A biomechanical evaluation of cortical onlay allograft struts in the treatment of periprosthetic femoral fractures. Hip Int 2003;13(3):148–58. 20. Emerson Jr RH, Malinin TI, Cuellar AD, Head WC, Peters PC. Cortical strut allografts in the reconstruction of the femur in revision total hip arthroplasty. A basic science and clinical study. Clin Orthop Rel Res 1992(285):35–44. 21. Howell J, Masri BA, Garbuz DG, Greidanus NV, Duncan CP. Cable Plates and Onlay Allografts in Perioprosthetic Femoral Fractures after Hip Replacement: Laboratory and Clinical Observations. Instr Course Lect 2004;53:99–110. 22. Berry DJ. Treatment of Vancouver B3 periprosthetic femur fractures with a fluted tapered stem. Clin Orthop Rel Res 2003(417):224–31. 23. Wong P, Gross AE. The use of structural allograft for treating periprosthetic fractures about the hip and knee. Orthop Clin North Am 1999;30:259–64. 24. Berry DJ. Periprosthetic fractures associated with osteolysis: a problem on the rise. J Arthroplasty 2003;18(3, suppl. 1): 107–11. 25. Jando VT, Greidanus NV, Masri BA, Garbuz DS, Duncan CP. Trochanteric osteotomies in revision total hip arthroplasty: contemporary techniques and results. Instr Course Lect 2005;54:143–5.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 192–202

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: REVISION HIP ARTHROPLASTY

(v) Post-operative infection in total hip arthroplasty Lipalo Moketea, Douglas D. Naudieb, a

Arthroplasty Unit, University of Western Ontario, London Health Sciences Center, University Campus, 339 Windermere Road, London, Ont., Canada N6A 5A5 b Orthopaedic Surgery, University of Western Ontario, London Health Sciences Center, University Campus, 339 Windermere Road, London, Ont., Canada N6A 5A5

KEYWORDS Hip; Arthroplasty; Infection; Revision

Summary The diagnosis of infection starts with a comprehensive history and thorough physical examination. First line investigations should include C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR). Elevation of one or both of these tests suggests infection and hip aspiration should be undertaken to confirm the diagnosis. Hip aspiration also allows identification of the infecting microorganism. Radionucleotide tests presently have a limited role but may be useful as first line investigations in patients with active inflammatory conditions. If there is still doubt regarding the diagnosis following these investigations then one should use clinical judgment and possibly proceed to intraoperative frozen section. Acute infections can sometimes be successfully treated with retention of the prosthesis but the gold standard treatment for chronic infections is two-stage reimplantation. & 2006 Published by Elsevier Ltd.

Introduction Total hip arthroplasty is one of the most successful surgical interventions with very high patient satisfaction rates. Deep infection is a catastrophic complication to what should otherwise be a predictably good outcome. The treatment of deep infection is associated with high patient morbidity and inordinately high hospital and physician resource utilisation. Making the diagnosis is often not straightforward and consequences of missed infection are dire to both the patient and physician. Indiscriminate use of the battery of Corresponding author. Tel.: +519 663 3407; fax: +519 663 3420.

E-mail addresses: [email protected] (L. Mokete), [email protected] (D.D. Naudie). 0268-0890/$ - see front matter & 2006 Published by Elsevier Ltd. doi:10.1016/j.cuor.2006.06.010

tests available to help make the diagnosis can easily confuse the issue. Possible treatment options range from conservative measures in the form of antibiotic therapy alone to staged procedures and in the extreme case, disarticulation (Table 1). Fortunately, evolution of practice has helped address some past management controversies but decisionmaking remains a challenge when one is dealing with an infected hip arthroplasty.

Pathogenesis Both aerobic and anaerobic bacteria have been implicated in hip arthroplasty infections. Rarely, fungi, mycobacterium and brucella may be the source of infection. Staphylococcus epidermidis which is part of normal skin flora and

ARTICLE IN PRESS Post-operative infection in total hip arthroplasty

Table 1 Illustrating the various treatment options for the infected total hip arthroplasty. Treatment options for the infected total hip arthroplasty Antibiotic therapy only Long-term antibiotic suppression Debridement and retention of prosthesis Single stage exchange arthroplasty Staged exchange arthroplasty Arthrodesis Girdlestone’s arthroplasty Disarticulation

Staphylococcus aureus are the most common infecting organisms.1 Streptococcus species, Coliforms, Pseudomonas, Anaerobes and other bacteria may be infective bacteria, their relative importance varying from hospital to hospital. The microorganisms are introduced onto the surface of the hip prosthesis either intraoperatively, secondary to superficial surgical site hematoma or infection or from hematogenous spread. Aside from existing in free-floating forms (planktonic bacteria) Staphylococcus aureus, Staphylococcus epidermidis and Pseudomonas aeruginosa have the ability to produce a hydrated matrix of extracellular polysaccharides (glycocalyx) and protein, forming a biofilm once adherent to the implant surface. The biofilm acts as a protective barrier against antimicrobial agents and host defences. Within the biofilm complex communities composed of one or more species of microorganisms can exist. From time to time bacteria resident within the biofilm may be shed into the blood stream resulting in systemic upset. Biofilm bacteria that are firmly established onto the prosthesis surface can in most instances only be eradicated by removing the colonised implant. Nishimura et al. have demonstrated that biofilm can form within 24 h of bacterial adherence to the surface of the prosthesis.2 Certain strains of S. aureus use a different mechanism to adhere onto the prosthesis surface. These strains elaborate binding factors known as adhesins which bind to host proteins, particularly collagen, that cover the implant surface.3 Infection by virulent strains of bacteria and polymicrobial infections have become more commonplace. Methicillin resistant S. aureus (MRSA) is increasingly being isolated from infected arthroplasties and is currently the commonest multiple drug resistant infective organism.4 S. aureus strains with reduced susceptibility to vancomycin have been isolated and will probably become increasingly important in the future.5 S. epidermidis and certain gram-negative bacilli infections are also showing multiple-drug-resistance.

193 noted. Multiple previous operations on the same hip are associated with increased infection risk. A history of concurrent infection such as urinary tract infection, infection of another implanted device, dental abscess or recent oral surgery is important. Hip pain is a consistent symptom in the infected arthroplasty. The duration of symptoms, presence of fever, rigors and the interval between previous surgery and symptom onset should be established. General examination should determine whether the patient is obese or has signs of malnutrition, as both factors are associated with increased risk of infection. Local examination of the wound will establish tenderness, presence of local inflammation, abscess, wound discharge, sinus and placement and relationship of surgical scars. Review of charts may reveal pre-morbid functional status in the form of hip scores, which can be compared to the patient’s present status. Any breach in operating room protocol may have been noted, including unusually high traffic, presence of a large number of non-essential personnel, or problems with airflow systems on the day. Other factors like poor choice or non-use of prophylactic antibiotics may have been recorded. The surgery may have been of particularly long duration, which is a factor that independently increases the risk of post-operative infection. Specimens sent for microbiological analysis may indicate the offending organisms and antibiotic sensitivities. Any antibiotics that the patient has had should be noted. Special investigations are indicated where the diagnosis of deep infection is in doubt following history taking and examination and in instances where the infecting organism has not been identified and antibiotic sensitivities are unknown.

Radiological investigations Radiographs are neither sensitive nor specific for diagnosing deep periprosthetic infection. Subtle radiographic signs such as localised osteopenia, periostitis and endosteal scalloping may suggest the presence of active infection. Significant osteolysis early on in the post-operative period also favors a diagnosis of infection provided the implanted materials were not sub-standard or defective. A review of serial radiographs will yield more useful information than a single radiograph. However, in the vast majority of patients, radiographic findings are not helpful in differentiating between aseptic and septic loosening. Ultrasonography has been used to locate abscesses and facilitate fluid aspiration for culture. The use of magnetic resonance imaging is presently limited to identifying sinus tracts and pockets of pus and fluid collection. Scatter artifact in the immediate area around implanted prostheses distorts the image quality and limits the usefulness of this imaging modality.

Diagnosis Radionucleotide scans History taking in the patient with a total hip arthroplasty that may be infected should be comprehensive. Systemic risk factors for infection should be elicited. Conditions resulting in immunocompromise including diabetes mellitus, rheumatoid arthritis, steroid therapy, psoriasis, renal failure, organ transplantation and advanced age should be

Radionucleotide scans, in the form of technetium scans, came into widespread use in the diagnosis of periprosthetic infections in the 1980s, following early favourable reports. 99m Tc MDP uptake is dependant on blood flow and reflects metabolic activity and bone turnover in skeletal tissue.

ARTICLE IN PRESS 194 Uptake is increased in instances of septic and mechanical loosening of implants, rendering the test of little use in distinguishing the two modes of loosening. Subsequent reports demonstrating the non-specific nature of this investigation in diagnosing periprosthetic infection have resulted in diminished enthusiasm for its routine use. Levitsky et al. reported a sensitivity of 33%, specificity of 86%, a positive predictive value of 30%, and a negative predictive value of 88%.6 Gallium (67Ga) scanning has been in use over the past four decades. The 67Ga tracer is produced by a cyclotron and following injection into the body it is initially bound to transferrin. The Gallium bound to transferrin accumulates at inflammatory foci and also binds to lactoferrin, which is present in high concentrations in these foci. Some of the circulating gallium is thought to be taken up directly by bacteria at the sight of infection. The distribution of the tracer within the body is picked up by a g-camera. Results have been disappointing with the use of this test in diagnosing periprosthetic infections due to low sensitivity and accuracy.7 In theory, the area surrounding an infected prosthesis should be teeming with inflammatory cells, including recruited white blood cells and immunoglobulins. This is the basis for use of radioactive labelled scans. In 111Ilabelled white blood cell scans, whole blood is harvested from the patient and erythrocytes are allowed to separate by a process of sedimentation. The remaining white blood cell-platelet mixture is centrifuged to separate the white blood cells and these are then labelled with radioactive Indium. Once reintroduced into the body, the white cells should preferentially cluster around the infected implant and be picked up by a g-camera as an area of increased activity. Neutrophils are the predominant type of white blood cell that take up the tracer and a minimum number of 2000 white cells/mm3 is required for satisfactory images. Despite the labor and expense entailed in this procedure, the test cannot be reliably used to diagnose the infected hip prosthesis. Reported sensitivities are variable ranging from 38% to 100%, with specificities of 15–100% and accuracy of 60–96%. In a bid to improve on the diagnostic accuracy of labelled white blood cell scans, 111I-labelled immunoglobin G (IgG) scans have been developed. When used in the diagnosis of infected hip prostheses the test has a reported sensitivity of 77.8% and specificity of 95.5%.8 Because of the questionable value of radionucleotide scans as first line diagnostic tests in infected total hip arthroplasties, combinations of tests were investigated to improve diagnostic accuracy. Kraemer et al.9 investigated technetium/gallium scanning in diagnosing infection in total hip arthroplasty. They reported a sensitivity of 38%, specificity of 100% and positive predictive value of 100%, negative predictive value of 79% and accuracy of 81%. In the same study, the accuracy of the combined test was compared to that of hip aspiration alone and combined technetium/gallium/aspiration. The combined accuracy of all three tests of 85% was just marginally higher than that of hip aspiration alone, at 84%. The authors concluded that technetium/gallium imaging is not an effective method for investigating painful hip protheses for sepsis and offered no additional advantage over hip aspiration. Joseph et al.10 looked at combined 111I-labelled WBC/99mTc sulfur and

L. Mokete, D.D. Naudie found 100% specificity, 46% sensitivity, 100% positive predictive value, 84% negative predictive value and 88% accuracy. Inclusion of blood pooling and flow phase data increased the sensitivity to 66%, negative predictive value to 89% and accuracy to 90%. However, the specificity dropped to 98% and the positive predictive value dropped to 91%. They stated that routine use of these combined radionuclide tests could not be supported. One new tracer that has been investigated in this field includes 18F-FDG. Activated inflammatory cells express increased numbers of inward glucose transporters. These same transporters carry 18F-FDG, which accumulates inside the cells. A high resolution imaging technique, positron emission tomography (PET), is used to determine the distribution of the tracer. Early reports suggested that this investigation might be useful in diagnosing infection but presently its use in differentiating septic from aseptic loosening in periprosthetic infections is limited as interpretation of results and diagnostic criteria are yet to be established.11 Quoted figures for the diagnostic accuracy of radionucleotide scans in the diagnosis of hip arthroplasty infections are variable, but on the whole there is little evidence to support their routine use as first line tests. The sensitivity and specificity of these tests are at best comparable to that of erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP). However, they may have a role in diagnosing infection in patients with active inflammatory conditions e.g. rheumatoid arthritis, accounting in part for raised ESR and CRP.

Blood investigations Useful blood investigations include inflammatory markers, such as the ESR and CRP. The white cell count is rarely elevated in chronic hip arthroplasty infections12 and is in general not useful in guiding treatment. Both ESR and CRP are non-specific markers that are influenced by a host of inflammatory conditions. When these conditions are excluded, the sensitivity, specificity, positive predictive value and negative predictive value of ESR in diagnosing total hip arthroplasty infection has been reported as 82%, 85%, 58% and 95%, respectively. The corresponding values for CRP are 96%, 92%, 74% and 99%, respectively.13 As these markers are not always elevated together, Spangehl et al.13 suggested that the combination should be used and they found it reliable in predicting periprosthetic infections. They defined elevated ESR as a level greater than 30 mm/h and elevated CRP as greater than 10 mg/l. Elevated levels of the cytokine Interleukin-6, a factor that is produced by monocytes and macrophages, have recently been shown to be correlated positively with the presence of chronic periprosthetic infection. With a normal serum IL-6, defined as o10 pg/ml, the test has a sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of 100%, 95%, 89%, 100% and 97%.14 Unfortunately, the levels of IL-6 may be elevated in patients with chronic inflammatory disease such as rheumatoid arthritis. Other conditions associated with elevated IL-6 levels include previous antibiotic treatment, acquired

ARTICLE IN PRESS Post-operative infection in total hip arthroplasty immunodeficiency syndrome, multiple sclerosis, and Paget’s disease of bone.

Aspiration Hip aspiration is a minimally invasive procedure with low morbidity. It can be undertaken in the operating room with the help of image guidance or in the radiology suite under local anaesthetic. The procedure should be done under sterile conditions and care should be taken not to introduce microorganisms into the hip. The aspirate is sent for microscopic analysis, cell count, gram staining and culture and antibiotic sensitivities. A synovial fluid cell count of more than 50,000 leukocytes per ml with predominantly polymorphonuclear cells (80%) is suggestive of bacterial infection.13 The glucose level in the fluid should also be compared to circulating blood glucose, with a lower level in the aspirate suggesting the presence of bacterial infection. Positive cultures other than contaminants indicate the presence of infection and allow antibiotic sensitivities to be determined. There is no clear definition of what should be considered contamination, with authors applying different criteria in published series. Williams et al.15 reported the sensitivity and specificity of aspiration to be 80% and 94%, respectively. Lachiewicz et al.16 reported a sensitivity of 92%, specificity of 97% and accuracy of 96% in a study performed on 193 revision hip arthroplasties. A review of the literature involving 1915 aspirates in 18 studies showed preoperative aspiration was generally deemed to be a useful test.17 The sensitivity of hip aspiration is affected by prior antibiotic administration. If the patient is already on antibiotic therapy these should be stopped for at least 2 weeks prior to aspiration. Other limitations of the procedure include the inability to sample biofilm microorganisms as the organisms within the aspirate consist of planktonic bacteria. The technique of lavaging the joint with normal saline and aspirating the fluid in the event of a dry aspirate remains controversial. More invasive preoperative tissue drill biopsies have no advantage over aspiration in terms of bacterial accuracy and result in more false positives.15 Based on a prospective analysis of 202 revision hip arthroplasties with 17% confirmed infection rate, Spagehl et al.13 concluded that a combination of normal ESR and CRP level is reliable for predicting the absence of infection. They suggest aspiration is indicated when the ESR or CRP is elevated or when a clinical suspicion of infection remains. Lachiewicz et al.16 recommended aspiration if the prosthesis had been in place for less than five years and the ESR was abnormal.

Intraoperative investigations Frozen section Biopsy specimens for frozen section analysis can be taken at the time of open procedure to the hip. Once a decision is made to go ahead with re-implantion on the basis of normal inflammatory markers and sterile aspirate, frozen section analysis can be used to help confirm absence of infection. Tissue specimens are taken from the most inflamed tissue in

195 the acetabulum, femur and surrounding areas. They are immediately sent for analysis by the pathologist. Not all hospitals are geared towards frozen section analysis and preoperative discussion with a pathologist with an interest in periprosthetic infections is often helpful. Lonner et al.,18 in a prospective study on 175 revision total joint arthroplasties (142 hip and 33 knee), reported a sensitivity of 84%, specificity of 96%, positive predictive value of 70% and negative predictive value of 98% in identifying infections when a positive result was defined as at least five polymorphonuclear cells per high power-field at microscopy. When an index of at least 10 polymorphonuclear cells per high-power field was used the sensitivity and negative predictive value remained the same but the specificity increased to 99% and the positive predictive value increased significantly to 89%. On the basis of the results of the study, a positive frozen section is generally considered to be more than 10 polymorphonuclear cells per high-power field. Spangehl et al.13 resort to intraoperative frozen section when the diagnosis of infection remains in doubt following ESR, CRP and hip aspirate or when inflammatory marker levels remain elevated on the basis of some other inflammatory condition. The value of intraoperative frozen section has been questioned of late. In a prospective study on 121 revision arthroplasties Banit et al.19 found that it lacked the positive predictive value and sensitivity for accurate determination of infection of hip arthroplasties (45% sensitivity, 92% specificity, 55% positive predictive value, and 88% negative predictive value).

Gram stain Intraoperative gram stain is not a reliable investigation for determining the presence of infection in revision hip arthroplasty. It is insensitive, has a poor positive predictive value and is of little use in determining suitability for reimplantation.20

Microbiologic tissue and/or fluid culture The gold standard for determining the presence or absence of periprosthetic infection remains the results of microbiologic culture of tissue and/or fluid obtained during revision arthroplasty. However, tissue and fluid cultures can still yield false-negative and false-positive results. For example, some authors have described cases in which, despite the presence of acute inflammation in the periprosthetic membrane and a clinical postoperative course consistent with infection, the intraoperative cultures remained negative.21 On the other hand, Padgett et al.22 reported that 30% of 142 hips treated with revision arthroplasty had at least one positive intraoperative culture. A clinically important infection later developed in only one case in their series, however, suggesting a high frequency of false-positive cultures probably caused by contamination of the tissue sample. In a prospective study involving revision arthroplasty in 297 patients with a total of 41 infections, Atkins et al.23 noted that only 65% of all samples obtained from the infected joints were culturepositive. They recommended obtaining five or six culture

ARTICLE IN PRESS 196

L. Mokete, D.D. Naudie

specimens from each patient and suggested that the cut-off for a definite diagnosis of infection be growth of the identical organism on culture of three or more specimens. In general, it is recommended that surgeons take special precautions to minimise tissue contamination, such as obtaining multiple samples from deep tissues, using clean instruments for tissue retrieval, transferring tissue to the culture bottle without allowing contact with the operative field or gloves, and transferring of the culture samples to the laboratory for processing as quickly as possible. Falsenegative cultures are likely when the patient received preoperative or intraoperative antibiotics, when the offending organism cannot be isolated by the routine laboratory protocols, or when the submitted tissue samples were extensively cauterised. To minimise the incidence of falsenegative cultures, representative samples should be obtained with sharp dissection, administration of antibiotics should be discontinued at least 2 weeks prior to the surgery, and intraoperative antibiotics should be withheld until the tissue samples are retrieved.24

Polymerase chain reaction Polymerase chain reaction (PCR) techniques have been investigated in the quest to detect subclinical joint infections. The ability to amplify minute levels of bacterial DNA to detectable levels should in theory improve our diagnositic accuracy of periprosthetic infections. However, present techniques are insensitive due to problems of contamination and result in a high rate of false positives.25 With further refinement these techniques may be applicable in the future.

Classification Early classification systems of periprosthetic hip infection did little to guide treatment. Fitzgerald et al.26 classified infection following total hip arthroplasty into three distinct groups: Stage I: Acute postoperative infections that included the classic fulminant postoperative infection, the infected hematoma and surgical site infection that progresses to deep infection. Stage II: Infections that become apparent 6–24 months postoperatively due to indolent infection. Stage III: Infections that develop two or more years postoperatively in a previously asymptomatic arthroplasty.

These were presumed to be due to hematogenous spread of infection. In a landmark study, Tsukuyama et al.27 divided hip arthroplasty infections into different types based on the above classification but with addition of a fourth type for positive intraoperative cultures taken at the time of direct exchange arthroplasty of presumed non-infected joints. Based on their experience of 106 patients treated according to a defined protocol, they were able to show success rates for eradication of infection in each of the types (Table 2). Although the strategies employed yielded modest success rates, the guidelines suggested formed a basis for a classification system that was useful in guiding treatment. The drawbacks of the classification were lack of consideration of the general health of the patient and of any operative factors at a local level. McPherson et al.28 have attempted to address this in a staging system that they proposed. They found significant positive correlation of systemic host grade and local extremity grade to a number of parameters in a sample of 50 patients with chronic infection. Further studies are required to validate the usefulness of this grading system for early infections and positive intraoperative cultures. For the purposes of this discussion acute infections are defined as postoperative infections within 2 weeks of surgery and hematogenous infections with symptoms for less than 2 weeks. Chronic infections include all infections recognised 2 weeks or later after surgery and hematogenous infections symptomatic for a period longer than 2 weeks.

Treatment The main aims of treatment should be eradication of infection and to leave the patient with a functional prosthesis. Treatment approaches to these patients must be multi-disciplinary with continuous consultation with, amongst others, infectious disease experts, nutritionists, physiotherapists and occupational therapists.

Antibiotic therapy alone This treatment strategy is indicated in patients with positive intraoperative cultures taken during revision for presumed aseptic loosening. Preoperatively, every effort should have been made to exclude infection. The distinction between mere contamination and potential and actual infection is not always clear-cut and this strategy may lead to

Table 2 Table showing the infection cure rate of patients treated by Tsukayama et al. according to defined treatment protocols. Infection type

Treatment protocol

Outcome

Type 1 (n ¼ 35)

Surgical debridement, exchange of polyethylene insert of acetabulum and intravenous antibiotics Removal of components, surgical debridement, antibiotic beads insertion and delayed reimplantation Surgical debridement, prosthesis retention and intravenous antibiotics Intravenous antibiotics

71% cure

Type 2 (n ¼ 34) Type 3 (n ¼ 6) Type 4 (n ¼ 31)

85% cure 50% cure 90% cure

ARTICLE IN PRESS Post-operative infection in total hip arthroplasty over-treatment. It is common practice to swab bone allografts before use in revision cases and these may on occasion present positive cultures. Antibiotic treatment may also be beneficial in these patients.

Long-term antibiotic suppression This treatment option is reserved for patients who have a stable prosthesis, harbor a sensitive microorganism and are not candidates for two-stage re-implantation either because of poor general health contra-indicating further surgery or refusal to consent to surgery. The oral antibiotic chosen for this purpose should be well tolerated by a compliant patient with minimal toxicity, but resistance remains a concern. The addition of rifampicin to treatment regimens for implant related staphylococcal infections has been shown to result in improved rates of eradication of infection and reduced rates of antibiotic resistance.29 Rifampicin containing regimens have been demonstrated to have activity against staphylococcal biofilms.30

Debridement and retention of prostheses Acute infections are often amenable to surgical debridement provided the prosthesis is stable. When dealing with modular implants, exchanging the bearings allows physical removal of microorganisms that may be interposed between the prosthesis and the bearing and removes biofilms that are resident in otherwise inaccessible areas. Copious amounts of saline lavage (at least 9 L) should be used during the procedure. Addition of antibiotic to the lavage may be beneficial. Patients are maintained on parenteral antibiotics for a minimum of 6 weeks post-op. Antibiotic choices are based on pre-operative cultures and they are modified depending on results of intraoperative cultures. Crockarell et al.31 found that this treatment strategy was unsuccessful if undertaken more than 2 weeks after onset of symptoms in acute infections.

197 concentrations of antibiotics locally. Local therapy is supplemented by parenteral antibiotics which may be converted to oral antibiotics depending on response to therapy. By convention the second stage is undertaken no earlier than 6 weeks following the first stage. A new prosthesis is re-implanted and antibiotic therapy is continued post-operatively for a minimum of 6 weeks.

Local antibiotics Delivery of high concentrations of local antibiotics can be achieved through use of antibiotic impregnated cement beads (Fig. 1) or antibiotic loaded cement spacers. Antibiotic loaded spacers can either be of the static variety or articulating. Static spacers are antibiotic loaded cement dowels loosely inserted into the femoral shaft and a corresponding spacer ball loosely inserted into the acetabulum (Fig. 2). Although some movement is possible at the hip joint it is restricted and there is associated discomfort favoring scar formation and soft tissue contracture over the interval between stages. Articulating spacers consist of femoral and acetabular components that are covered by antibiotic loaded cement with conventional articulating bearings (PROSTALAC), or more simply, a one-piece femoral component with a large head made of antibiotic loaded cement (Fig. 3). Articulating spacers allow maintenance of

Single stage exchange arthroplasty This treatment strategy is appropriate in a minority of cases of chronic infection. The prosthesis is removed, sinuses are excised, infected and devitalised tissue is debrided, and the wound is lavaged. Once a clean uninfected bed is achieved, then one proceeds to immediate re-implantation with a new prosthesis during the same procedure. The obvious advantages of this approach over delayed re-implantation are reduced patient morbidity of a single surgical procedure and attendant cost savings. Should there be any doubt regarding cleanliness of the wound following debridement then it is prudent to opt for staged surgery.

Staged exchange arthroplasty In the first stage of surgery, the prosthesis and all foreign material are removed, together with infected and devitalised tissue. Copious saline lavage (9 L) with one added antibiotic is used to irrigate tissues. The patient is then left with a resection arthroplasty and a device to deliver high

Figure 1 Radiograph of the right hip taken following first stage treatment of an infected hip arthroplasty using antibiotic impregnated cement beads.

ARTICLE IN PRESS 198

Figure 2 Radiograph of the left hip showing a static spacer consisting of antibiotic impregnated cement dowel shaped to fit the femoral cavity and an antibiotic impregnated ball placed in the acetabulum.

Figure 3 Radiograph of the pelvis showing an articulating spacer in the left hip fashioned to fit the femoral cavity and articulate with the acetabulum in a manner similar to a hemiarthroplasty prosthesis.

range of motion in the hip during the interval period, and ultimately an easier reimplantation procedure. Off-theshelf pre-molded spacers are also available but these are generally fabricated with low levels of antibiotic within the

L. Mokete, D.D. Naudie

Figure 4 Radiograph showing an example of a pre-molded offthe-shelf antibiotic loaded cement spacer.

cement (Fig. 4). Novel approaches include sterilisation of the removed prosthesis, coating with antibiotic loaded cement and use as a spacer. Hsieh et al.32 compared use of cement beads with a cement spacer prosthesis in 128 patients followed-up for an average of 4.9 years. The use of the spacer prosthesis was associated with a higher hip score, a shorter hospital stay, and better walking capacity in the interim period; a decreased operative time, less blood loss and a lower transfusion requirement at the time of re-implantation. The infection-free rates were similar in both groups. The choice of antibiotic is predicated on pre-operative cultures and sensitivities. Antibiotics loaded onto the cement should be thermostable and they should not adversely affect the properties of the cement. In vitro studies have demonstrated that combining two antibiotics in the cement improves elution of both, suggesting biantibiotic spacers may be better than mono-antibiotic spacers.33 Koo et al.34 recommend use of a combination of vancomycin, gentamicin and cefotaxime impregnated cement in two-stage revisions when the causative organism is not identified in preoperative cultures.

Decision making regarding proceeding to reimplantation The decision to proceed to re-implantation in the staged procedure is based on the general condition of the patient,

ARTICLE IN PRESS Post-operative infection in total hip arthroplasty the condition of the wound and special investigations. The patient’s condition should be optimised with respect to addressing immuno-incompetence, nutritional deficiencies and fitness to undergo further surgery. The wound should be healed, clean and without discharge or active sinuses.

Blood investigations A trend indicating a fall in both ESR and CRP during the interval period of antibiotic treatment is a reassuring sign that infection is being eradicated. Both these markers should be within normal range before proceeding to re-implantation. IL-6 may prove to be invaluable in guiding re-implantation as levels rapidly return to normal after total joint surgery in the absence of infection.

Aspiration Preoperative hip aspiration is useful in excluding ongoing infection. Continuing infection necessitates further surgical debridement or modification of antibiotic therapy. Aspiration is a particularly useful investigation in patients with active inflammatory disorders where the ESR and CRP would otherwise remain elevated despite eradication of infection.

Frozen section This investigation is appropriate in cases where infection eradication remains doubtful because of equivocal ESR, CRP and hip aspiration results. Della Valle et al.35 demonstrated that a negative finding on intraoperative analysis has a high predictive value with regard to ruling out the presence of infection. However, the sensitivity of the test for detection of persistent infection was poor (sensitivity of 25%, specificity of 98%, positive predictive value of 50%, negative predictive value of 95% and accuracy of 94%).

Salvage procedures Although arthrodesis, definitive excision arthroplasty and amputation remain in our armamentarium, these procedures are appropriate only in a select few patients in modern orthopaedics.

Arthrodesis Arthrodesis is indicated as a salvage procedure in young patients. Infection should be completely eradicated at the time of arthrodesis and the technique for fusion should ideally spare the abductors to enable reconversion to hip arthroplasty in later years.36 Long-term complications of hip arthrodesis include back pain, which dramatically improves on conversion to hip arthroplasty. Arthrodesis is contraindicated in patients with contralateral hip pathology, lumbar spine pathology and ipsilateral knee instability. Achieving hip fusion in the correct plane and sparing the abductors is a challenge in the arthroplasty patient because of pre-existing bone loss.

199

Girdlestone’s arthroplasty Planned definitive excision arthroplasty in the form of a Girdlestone’s arthroplasty is a reliable treatment option for eradicating infection (98%, 43 of 44 patients) and pain at the cost of functional compromise.37 The combination of a Trendelenberg gait and limb shortening makes for difficult walking and patients tire easily. Bourne et al.38 reported satisfaction rates of 79% in 33 patients followed for 6.9 years. They noted that results improved with time. Where cost constraints preclude further reconstructive surgery, Girdlestone’s arthroplasty is a reasonable salvage procedure in treating the chronically infected hip arthroplasty. The excision arthroplasty can be converted to hip arthroplasty when circumstances change, provided infection is eradicated. Occasionally, planned two-stage reimplantion surgery proceeds no further than the first stage due to patient infirmity leaving the patient with effectively an excision arthroplasty.

Hip disarticulation Disarticulation is associated with high morbidity and a very poor functional outcome with most patients confined to a wheelchair. The procedure may rarely be indicated in cases of overwhelming life threatening hip arthroplasty sepsis, as a last resort. It may also be considered in patients with intractable infection together with either major vascular insufficiency or massive bone loss, who are not deemed candidates for reconstructive surgery.

Single-stage vs. two-stage reimplantation The obvious advantages of single-stage reimplantation over two-stage surgery are the cost savings and reduced patient morbidity of a single procedure as opposed to two major procedures. In addition, single-stage reimplantation is a technically easier procedure. Proponents of the single-stage approach point to a higher incidence of mechanical complications including fracture, dislocation and limb length discrepancy with staged treatment.39 The disadvantages include less flexibility and lower success rate for curing infection. There is less flexibility in choice of high dose local antibiotic. There is only one opportunity, for instance, to choose the correct antibiotic to be impregnated into the cement at the time of reimplantation. Staged surgery allows revision of the antibiotic impregnated spacer should intraoperative cultures reveal infection by organisms resistant to the chosen antibiotic. There is also a documented lower success rate for curing infection in single-stage reimplantation procedures. Consistently high success rates have been reported for eradication of infection in two-stage reimplantations, making this treatment strategy currently the standard of care. Jackson et al.40 performed a literature review to determine when direct exchange was most likely to be successful. They found an 83% success rate at average 4.8 years follow-up of 1299 hips. The procedure was likely to be successful if the following criteria were met: patients who were generally of good health, absence of wound complications following the initial hip replacement, and if the

ARTICLE IN PRESS 200 infecting organisms are of low virulence and sensitive to antibiotic mixed into the bone cement used in the reimplantion. Polymicrobial infections, gram negative infections and infections by methicillin resistant organisms were associated with failure. The authors concluded that indications for direct exchange were limited as methicillin resistant organisms are common nowadays and current revision techniques use cementless implants.

Interval between stages Hansen and Rand41 combined results of multiple studies in order to investigate the effect of delay in reimplantation on the success rate of eradication of infection. They found that the timing of reconstruction and the use of antibiotic impregnated cement were closely linked variables. The results of two-stage reimplantation using cement without antibiotic were superior to those of single stage revision using cement without antibiotic and similar to those of single stage revision using antibiotic laden cement. An interval period between stages by itself seems to be beneficial. The ideal length of interval between stages has not been determined but by convention we wait a minimum of 6 weeks. Ketterl et al.42 found that reimplantation carried out early o4 weeks (average 2.1 weeks) was better than reimplantation 44 weeks (average 12.7 weeks), with respect to function, mortality and cure of infection.

L. Mokete, D.D. Naudie similar experience: Haddad et al.46 (50 patients) reported a 92% success rate at 5.8 years follow-up; Kraay et al.47 (33 patients) reported a 93% success rate. Reported success rates for two-stage cemented reimplantations range from 90% to 95%.

Bone grafting It is not unusual for the infected hip arthroplasty to present with significant bone loss. During debridement additional bone that is deemed to be non-viable is sacrificed, compounding bone stock deficiency (Fig. 5). In the past, there were concerns about the use of bone graft with fears that it may encourage persistence of infection or even introduce new infection. Patients with significant bone loss were treated with definitive resection arthroplasty, rendering them severely functionally compromised. The safety of bone grafts, including structural allogafts (Fig. 6) in secondstage reimplantation has now been reported in a number of studies with a reasonable length of follow-up.48,49

The future

Cementless vs. cemented implants

A single investigation with 100% sensitivity and specificity is required to diagnose infected hip arthroplasties. Different generations of radionucleotide tests have been touted to fulfil this purpose, only to be found inadequate on further clinical testing. However, with further refinement radionuclide tests may still come close to 100% sensitivity and specificity. Molecular biology techniques like PCR continue

The advantage of using cement lies in the ability to add antibiotic, which elutes over time, keeping infection in check. Indeed, early successful reimplantations involved use of antibiotic laden cemented prostheses. Buccholz et al.43 introduced the concept of antibiotic impregnated cement and popularised its use in Europe for single stage revision surgery, reporting infection cure rates of 77% in his series. Raut et al.44 reported a success rate of 84% using the same treatment strategy. A review of published series comparing use of plain bone cement and antibiotic impregnated cement in single-stage revisions found cumulative success rates of 60% and 83%, respectively, strongly suggesting there was no role for single-stage cementless revisions.41 The superiority of cementless hip revisions over cemented implants in aseptic loosening led investigators to consider their use in two-stage reimplantions. The limited timedependant ability of antibiotic to elute from bone cement in sufficient quantities to overcome infective bacteria coupled with the negative long-term effects of cement on the function of polymorphonuclear cells supported the move towards cementless two-stage reimplantations. Fehring et al.45 demonstrated a success rate of 92% (25 patients) at average 41 months follow-up using a two-stage cementless reimplantation approach. The authors highlighted the importance of using implants of appropriate design to ensure bone ingrowth of the femoral component. They used proximally milled cementless implants in cases where proximal femoral bone stock was adequate and extensively coated stems to obtain diaphyseal stability where there was proximal structural bone loss. Other authors have reported

Figure 5 Radiograph of the right hip illustrating severe femoral bone stock deficiency in a young patient following first-stage surgery for an infected revision hip replacement.

ARTICLE IN PRESS Post-operative infection in total hip arthroplasty

Figure 6 Radiograph of the same patient in Fig. 5, following second-stage treatment. The proximal femur has been reconstructed using a massive proximal femoral allograft in the form of an allograft-prosthesis composite.

to hold promise, but again need refinement. A technique for covalently tethering antibiotic to titanium surfaces has been developed.50 Facilitating high dose local antibiotic delivery may revolutionise treatment of chronic infections. This would make it possible to use cementless antibiotic coated revision implants and acceptable to dispense with staged treatment. The change in the microbiology of infecting organisms towards increased virulence is discouraging the move towards shorter antibiotic treatment times and use of more convenient oral antibiotics. Studies on biofilm organisms will hopefully lead to development of effective agents that can be taken orally. One of the biggest impediments in advancing management of the infected hip arthroplasty seems to be the falling incidence of infections. Studies with adequate numbers of patients are only possible in large referral centres that specialise in this particular complication, or combined multiple centres. There is certainly a need for these centres to rigorously evaluate new investigative and treatment modalities.

References 1. Salvati EA, Della Valle AG, Masri BA, Duncan CP. The infected total hip arthroplasty. AAOS Intruction Course Lecture 2003; 52:223–45. 2. Nishimura S, Toshiyuki T, Adachi K, Shindo H. Antimicrobial susceptibility of Staphylococcus aureus and Staphylococcus epidermidis biofilms isolated from infected total hip arthroplasty cases. J Orthop Sci 2006;11:46–50. 3. Montanaro L, Arciola CR, Baldassarri L, Borsetti E. Presence and expression of collagen adhesion gene (can) and slime production in Staphylococcus aureus strains from orthopaedic prosthesis infections. Biomaterials 1999;20:1945–9.

201 4. Ip D, Yam SK, Chen CK. Implications of the changing pattern of bacterial infections following total joint replacements. J Orthop Surg 2005;13(2):125–30. 5. Cosgrove SE, Carroll KC, Perl TM. Staphylococcus aureus with reduced susceptibility to vancomycin. Clin Infect Dis 2004; 39(4):539–45. 6. Levitsky KA, Hozack WJ, Balderston RA, et al. Evaluation of the painful prosthetic joint. Relative value of bone scan, sedimentation rate, and joint aspiration. J Arthrop 1991;6(3):237–44. 7. Love C, Tomas MB, Marwin SE, Pugliese PV, Palestro CJ. Role of nuclear medicine in diagnosis of the infected joint replacement. Radiographs 2001;21:1229–38. 8. Oyen WJ, Claessens RA, van der Meer JW, Corstens FH. Detection of subacute infectious foci with indium-111-labelled human autologous leukocytes and indium-111-labelled human nonspecific immunoglobulin G: a prospective comparative study. J Nucl Med 1991;32:1854–60. 9. Kraemer WJ, Saplys R, Waddell JP, Morton J. Bone scan, gallium scan, and hip aspitation in the diagnosis of infected total hip arthroplasty. J Arthrop 1993;8(6):611–5. 10. Joseph TN, Mujtaba M, Chen AL, et al. Efficacy of combined technitium-99m sulfur colloid/indium-111 leukocyte scans to detect infected total hip and knee arthroplasties. J Arthrop 2001;16(6):753–8. 11. Reinartz P, Mumme T, Hermanns B, et al. Radionuclide imaging of the painful hip arthroplasty. J Bone Joint Surg [Br] 2006;87B:465–70. 12. Canner GC, Steinberg ME, Heppenstall RB, Balderston R. The infected hip after total hip arthroplasty. J Bone Joint Surg Am 1984;66:1393–9. 13. Spangehl MJ, Masri BA, O’Connell JX, Duncan CP. Prospectiv analysis of preoperative and intraoperative investigations for the diagnosis of infection at the sites of two hundred and two revision total hip arthroplasties. J Bone Joint Surg [Am] 1999;81:672–83. 14. Di Cesare PE, Chang E, Preston CF, Liu C-J. Serum interleukin-6 as a marker of periprosthetic infection following total hip and knee arthroplasty. J Bone Joint Surg [Am] 2005;87:1921–7. 15. Williams JM, Norman P, Stockley I. The value of hip aspiration versus tissue biopsy in diagnosing infection before exchange hip arthroplasty surgery. J Arthrop 2004;19(5):582–6. 16. Lachiewicz PF, Rogers GD, Thomson HC. Aspiration of the hip joint before revision total hip arthroplasty: clinical and laboratory factors inlencing attainment of a positive culture. J Bone Joint Surg [Am] 1996;78:749–54. 17. Somme D, Ziza J-M, Desplaces N, et al. Contribution o routine joint aspiration to the diagnosis of infection before hip revision surgery. Joint Bone Spine 2003;70:489–95. 18. Lonner JH, Desai P, Di Cesare PE, Steiner G, Zuckerman JD. The reliability of analysis on intraoperative frozen sections for identifying active infection during revision hip or knee arthroplasty. J Bone Joint Surg [Am] 1996;78:1553–8. 19. Banit DM, Kaufer H, Hartford JM. Intraoperative frozen section analysis in revision total joint arthroplasty. Clin Orthop 2002;401:230–8. 20. Spangehl MJ, Masterson E, Masri BA, O’Connell JX, Duncan CP. The role of intraoperative gram stain in the diagnosis of infection during revision total hip arthroplasty. J Arthrop 1999;14(8):952–6. 21. Fehring TK, McAlister Jr JA. Frozen histologic section as a guide to sepsis in revision joint arthroplasty. Clin Orthop Relat Res 1994;304:229–37. 22. Padgett DE, Silverman A, Sachjowicz F, Simpson RB, Rosenberg AG, Galante JO. Efficacy of intraoperative cultures obtained during revision total hip arthroplasty. J Arthroplasty 1995;10: 420–6. 23. Atkins BL, Athanasou N, Deeks JJ, Crook DW, Simpson H, Peto TE, et al. Prospective evaluation of criteria for microbiological

ARTICLE IN PRESS 202

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

L. Mokete, D.D. Naudie diagnosis of prosthetic-joint infection at revision arthroplasty. The OSIRIS Collaborative Study Group. J Clin Microbiol 1998; 36:2932–9. Bauer TW, Parvizi J, Kobayashi N, Krebs V. Diagnosis of periprosthetic infection. J Bone Joint Surg Am 2006;88(4): 869–82. Clarke MT, Roberts CP, Lee PTH, Gray J, Keene GS, Rushton N. Polymerase chain reaction can detect bacterial DNA in aseptical loose total hip arthroplasties. Clin Orthop 2004;427:132–7. Fitzgerald Jr RH, Nolan DR, Ilstrup DM, Van Scoy RE, Washington II JA, Coventry MB. Deep wound sepsis following total hip arthroplasty. J Bone Joint Surg [Am] 1977;59:847–55. Tsukayama D, Estrada R, Gustillo RB. Infection after total hip arthroplasty. A study of the treatment of one hundred and six infections. J Bone Joint Surg [Am] 1996;78:512–23. McPherson EJ, Woodson C, Holtom P, Roidis N, Shufelt C, Patzakis M. Periprosthetic total hip infection. Outcomes using a staging system. Clin Orthop 2002;403:8–15. Zimmerll W, Widmer AF, Blatter M, Frei R, Ochsner PE. Role of rifampicin for treatment of orthopedic implant-related Staphylococcal infections: A randomized controlled trial. JAMA 1998;279(19):1537–41. Saginur R, St Denis M, Ferris W, et al. Multiple combination bactericidal testing of staphylococcal biofilms from implantassociated infections. Antimicrob Agents Chemother 2006; 50(1):55–61. Crockarell JR, Hanssen AD, Osmon DR, Morrey BF. Treatment of infection with debridement and retention of the components following hip arthroplasty. J Bone Joint Surg [Am] 1998;80: 1306–13. Hsieh P-H, Shih C-H, Chang Y-H, Lee MS, Shih H-N, Yang W-E. Two-stage revision hip arthroplasty for infection: comparison between the interim use of antibiotic-loaded cement beads and a spacer prosthesis. J Bone Joint Surg [Am] 2004;86:1989–97. Penner MJ, Masri BA, Duncan CP. Elution characteristics of vancomycin and tobramycin combined in acrylic bone-cement. J Arthrop 1996;11(8):939–44. Koo K-H, Yang J-W, Cho S-H, et al. Impregnation of vancomycin, gentamicin, and cefotaxime in a cement spacer for two-stage cementless reconstruction in infected total hip arthroplasty. J Arthrop 2001;16(7):882–92. Della Valle CJ, Bogner E, Desai P, et al. Analysis of frozen section of intraoperative specimens obtained at the time of reoperation after hip or knee resection arthroplasty for the treatment of infection. J Bone Joint Surg [Am] 1999;81:684–9. Beaule PE, Matta JM, Mast JW. Hip arthrodesis: current indications and techniques. J Am Acad Orthop Surg 2002;10(4): 249–58. Ballard WT, Lowry DA, Brand RA. Resection arthroplasty of the hip. J Arthrop 1995;10(6):772.

38. Bourne RB, Hunter GA, Rorabeck CH, Macnab JJ. A six-year follow-up of infected total hip replacements managed by girdlestone’s arthroplasty. J Bone Joint Surg [Br] 1984;66B:340–3. 39. Langlais F. Annotation: can we improve the results of revision arthroplasty for infected total hip replacement? J Bone Joint Surg [Br] 2003;85B:637–40. 40. Jackson WO, Schmalzried TP. Limited role of direct exchange arthroplasty in the treatment of infected total hip replacements. Clin Orthop 2000;381:101–5. 41. Hanssen AD, Rand JA. Evaluation and treatment of infection at the site of a total hip or knee arthroplasty. Instruct Course Lecture 1999;48:111–22. 42. Ketterl R, Henly MB, Strubinger B, et al. Analysis of three operative techniques for infected total hip replacements. Orthop Trans 1998;12:715. 43. Buchholz HW, Elson RA, Engelbrecht E, Lodenkamper H, Rottger J, Siegel A. Management of deep infection of total hip replacement. J Bone Joint Surg [Br] 1981;63-B:342–53. 44. Raut VV, Siney PD, Wroblewski BM. One-stage revision of total hip arthroplasty for deep infection. Long-term followup. Clin Orthop 1995;321:202–7. 45. Fehring TK, Calton TF, Griffin WL. Cementless fixation in 2-stage reimplantation for periprosthetic sepsis. J Anthroplasty 1999; 14(2):175–81. 46. Haddad FS, Muirhead-Allwood SK, Manktelow ARJ, BacareseHamilton I. Two-stage uncemented revision hip arthroplasty for infection. J Bone Joint Surg [Br] 2000;82-B:689–94. 47. Kraay MJ, Goldberg VM, Fitzgerald SJ, Salata MJ. Cementless two-staged total hip arthroplasty for deep periprosthetic infection. Clin Orthop 2005;441:243–9. 48. Haddad FS, Masri BA, Garbuz D, Duncan CP. The treatment of the infected hip replacement: the complex case. Clin Orthop 1999;369:144–56. 49. Alexeeff M, Mohamed N, Morsi E, Garbuz D, Gross A. Structural allografts in two-stage revisions for failed septic hip arthroplasty. J Bone Joint Surg [Br] 1996;78-B:213–6. 50. Parvizi J, Wickstrom E, Zeiger AR, et al. Titanium surface with biologic activity against infection: Frank Stinchfield Award. Clin Orthop 2004;429:33–8.

Suggested Further Reading 1. Bhandari M, Montori VM, Swiontkowski MF, Guyatt GH. User’s guide to the surgical literature: how to use an article about a diagnostic test. J Bone Joint Surg [Am] 2003;85:1133–40.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 203–207

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: REVISION HIP ARTHROPLASTY

(vi) Economics of revision total hip arthroplasty Alexander W.R. Burns, Robert B. Bourne London Health Sciences Centre, University of Western Ontario, 339 Windermere Road, London, Ont., Canada N6A 5A5

KEYWORDS Total hip arthroplasty; Total hip replacement; Cost effectiveness; Economics; THA; THR

Summary Total hip arthroplasty (THA) is a procedure which relieves pain, restores function and improves quality of life for patients with severe hip arthritis. The economics of total joint arthroplasty has become an increasingly important issue in the past decade, and will be more so in the future. If the number of patients treated with THA in the community increases as anticipated, so too will the requirement for revision of failed THA. Orthopaedic surgeons, medical administrators, and policy makers must work together to control costs while still maintaining high standards and quality outcomes for patients. This article examines issues relating to the economics of THA, including projections for the future burden of revision THA, the cost-effectiveness of revision THA, measures to improve survivorship of arthroplasties and the role of national arthroplasty registers in guiding decision-making based on evidence-based practice. & 2006 Elsevier Ltd. All rights reserved.

Introduction Total hip arthroplasty (THA) has revolutionized the care of patients with end-stage arthritic conditions of the hip, leading to marked improvements in health-related quality of life.1 The number of total hip replacements for arthritis is increasing as is the requirement for total hip replacement and hemiarthroplasty in hip fracture treatment.2 With the increased utilization of total hip replacement surgery, it is expected that the number of patients requiring revision total hip replacement will also grow. The increasing wave of individuals with degenerative joint disease and the projected requirement for THA in the

Corresponding author. Tel.: +519 663 3512; fax: +519 663 3780.

E-mail addresses: [email protected] (A.W.R. Burns), Robert. [email protected] (R.B. Bourne). 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.02.007

future are startling.2 Both healthcare providers and health administrators must co-operate to meet this challenge to provide THA to disabled patients, thus alleviating suffering, restoring function and improving quality of life. As a part of this initiative, health systems worldwide must ensure that any measures adopted to minimize arthroplasty expenditure are not to the detriment of excellent outcomes. The aim of this article is to explore the economics surrounding revision THA particularly with regard to the financial burden which it will place upon health systems in the near future. Ultimately the ideal solution would be the avoidance of the need for revision total hip replacement by the implantation of durable primary THA’s and the avoidence of postoperative complications. A critical analysis of all facets of arthroplasty is necessary to improve outcomes and both clinical and economic research should be aimed at addressing this issue. The use of national arthroplasty registers will be important with respect to guiding future directions based upon evidence based decision-making.

ARTICLE IN PRESS 204

A.W.R. Burns, R.B. Bourne

Burden of hip arthroplasties Recent data published by the American Academy of Orthopaedic Surgeons (AAOS) demonstrate the scope of the problem being faced with regard to THA.2 The number of primary total hip arthroplasties performed in the USA has increased from 134,000 in 1995 to 220,000 primary and 36,000 revision hip replacements in 2003, an appropriate revision rate of 14%. The AAOS projects that by the year 2030 the requirement for primary THA will increase to 274,000 and at the current rate of revision arthroplasties, the revision burden will be in the order of 45,700. Another potential source of revision surgery will be the number of hemiarthroplasties performed for the treatment of hip fracture, calculated to be 108,000 hemiarthoplasties in 2003 (Fig. 1).

Economic cost of primary and revision THA now and in the future

with greater technical difficulties and inferior outcomes. The crude revision THA rates (proportion of hip replacement procedures that are revisions) for various countries are demonstrated in Table 1. In 2003, the cost of primary THA in the USA was approximately $6.77 billion and the cost of revision THA was $1.66 billion, resulting in a total expenditure of $8.43 billion.2 Extrapolating from the current revision rate generates a revision burden of 45,700 revision THAs for 2030. The projected cost (at 2.5% inflation rate) of primary THA would be US$18.7 billion and a further US$4.01 billion for revision THAs. Thus the USA will be dealing with a US$22.7 billion dollar annual expenditure for primary and revision hip arthroplasty in 2030 (Fig. 2). Crude revision rates for THR vary around the world. Sweden started in 1979 with a crude revision rate similar to what other countries report today, but through evidencebased surgical practice, now has the lowest rate worldwide with 7%.9 In the USA the revision rate for 2003 was 14%, decreased from a rate of 18% in 2002. Obviously the financial benefits of reducing this rate of revision are considerable.

Revision total hip replacements are recognized as being both more expensive than primary procedures and associated

25 US $ 22.7 billion

Revision THA Primary THA

300.000

4.01

20

2003 250.000

Billions US $

2030 200.000

150.000

15

10

US $ 8.43 billion 18.7 1.66

100.000

5 6.77

50.000

0 2003 0 Primary THA

Revision THA

Figure 1 Projected numbers of primary and revision THA. Comparison of 2003 versus 2030.

Table 1

2030 Year

Figure 2 Cost of primary, revision and total cost of arthroplasty in the USA for 2003 and projected cost for 2030 (in US$ billions).

Crude THA revision rates between countries.

Country

Crude revision rate (%)

Year

Source

Australia Canada England/Wales New Zealand Norway Scotland Sweden USA

13 13 9 13 12 14 7 14

2004 2003 2004 2004 2003 2004 2003 2003

National joint replacement registry3 Canadian Joint replacement registry4 National joint registry5 NZ National joint register6 Norwegian arthroplasty register7 Scottish arthroplasty project8 Swedish national hip arthroplasty register9 American academy orthopedic surgeons2

Crude revision rate % ¼ number of revisions/total number of primary and revision hip arthroplasties.

ARTICLE IN PRESS Economics of revision total hip arthroplasty

Table 2

205

Cost of 10-point WOMAC improvement in primary and revision THA.

Funding allocation (US$) Increase in WOMAC score Cost (US$) per 10-point WOMAC improvement

The AAOS has estimated the average cost of a revision THA in 2003 was $45,000 and a decrease in the revision rate from 14% to the current Swedish rate of 7% would therefore produce a yearly saving of $810 million in 2003 or a projected saving of $2 billion extrapolated for the year 2030.

Recent reductions in the cost of revision THA Many changes in the management of arthroplasty patients have occurred over the past few decades and improved procedures have led to decreases in overall costs.10 Arthroplasty patients are often admitted on the day of surgery and now the average length of stay for primary THA is 4.5 days and for revision THA 5.4 days in the USA.2 Length of stay may decrease further for primary THA if inserted by minimally invasive techniques with improved anaesthesia, home support and accelerated rehabilitation.11,12 One recent study comparing outpatient THA (admission, surgery and discharge within 12 h) versus inpatient THA demonstrated $4000 less average hospital cost and $2500 reduction in the outpatient group per procedure.12 Surgeon remuneration has also decreased over the last decade as a percentage of the overall expenditure but implant costs still contribute substantially to the total.13 Currently the cost of alternate bearing surfaces (ceramic on ceramic, cross-linked poly-ethylenes or metal-on-metal articulations) are greater than a traditional metal on polyethylene bearing. This increased initial expenditure may produce potential long-term cost reductions if the low in vitro wear rates translate to improved in vivo survivorship.14 Many hospitals and health departments have a tendering process to negotiate contracts with the prosthesis suppliers in an attempt to control the cost of implants, leading to marked reductions in expenditures.15

Cost effectiveness of revision THA Health economics and cost-effectiveness is becoming one of the most important issues in orthopaedics today. There is an increasing volume of literature about the economics of both hip and knee arthroplasty.16 This is demonstrated in the findings of one recent study which showed that in the 10 years from 1993 to 2002, there were 70 economic evaluations of THA published in the English literature compared to only 10 in the preceding decade.16 Health economists have categorized healthcare interventions into extremely cost effective (cost-to-utility o$20,000), moderately cost-effective (cost-to-utility of $20,000 to $100,000), and effective but costly (cost-toutility 4$100,000).17 Rorabeck et al.18 and Chang et al.19

Primary THA

Revision THA

$7331 53 $1383

$8850 36 $2458

demonstrated that the cost-to-utility gain for primary THA was superior to treating moderate hypertension, coronary artery bypass surgery or renal dialysis. By these criteria a revision THA costing even twice that of a primary replacement would still be classified as excellent in terms of costeffectiveness. Unpublished data from our hospital supports the costeffectiveness of revision THA. We derived a dollar value for each 10-point WOMAC improvement comparing the pre- and postoperative status of 1264 primary and 133 revision hip replacement patients, using the local health department funding allocation.20 In our institution, US$1383 were spent for every ten-point increase in the WOMAC score for patients undergoing a primary THA as compared to US$2458 for each ten-point increase in patients undergoing a revision total hip replacement (Table 2). The ratio of cost per 10-point improvement in WOMAC scores comparing revision to primary arthroplasty was 1.78. Thus revision hip arthroplasty is approximately 56% as costeffective as a primary procedure at our hospital but is still rated as good or excellent. It should also be remembered that revision THA allows the majority of patients to maintain community independence, sparing some of the cost to society of permanent residential care for those individuals.

Importance of national arthroplasty registers National arthroplasty registers are now effective in Australia, Canada, Denmark, England/Wales, Hungary, New Zealand, Norway, Romania, Scotland, and Sweden generally funded by government sources. National joint registries have the power to provide accurate and timely data about the types and number of specific prostheses being implanted, and which prostheses are being removed. In a national arthroplasty register the endpoint taken for failure is revision surgery where revision is defined as exchange of one or both components or removal of prosthesis. It is desirable to identify the early failure of newer implants so that they can be promptly withdrawn from use. The continual collection of this data by joint registries allows more rapid recognition of problems than would otherwise occur through the conventional orthopaedic literature, and ensures that adverse outcomes come to light without signs. The Swedish Hip Registry9 was established in 1979 with the aim that individual hospitals would be able to compare their own arthroplasty outcomes to those of the rest of the nation. Those with poorer outcomes would then be able to identify improved practices and implement them for the benefit of their own patients. There is no question that the Swedish Registry has been a resounding success.

ARTICLE IN PRESS 206 The effect of the Registry in improving the quality of hip replacement in Sweden has been well documented.21 In the 1980s the Registry identified the benefits of modern cementing techniques in hip arthroplasty. The dissemination of this information led to a change in practice. As a result, the use of a distal femoral cement restrictor increased from 30% in 1982 to more than 95% in 1987. The crude rate of revision hip replacements in Sweden has halved over the last 20 years and is currently 7%. The implementation of techniques gained from the feedback of yearly outcome data has decreased the revision rates of their cemented implants.21 The Swedish Hip Registry has prompted a reduction in the number of implant types used, with 78% of cemented operations utilizing one of five designs all of which have documented long-term success. In this way the Registry has influenced both surgical technique and prosthesis type, leading to improvement of outcomes nationally. Arthroplasty registers are powerful tools that provide timely data to guide surgeons toward evidence-based surgical practice for the benefit of their patients. They should reduce the burden of revision THA surgery to society.

Funding allocation (instruments) and true cost of revision THA In the 1980s, health policy changed in many countries to a prospective system of payments based on the International Classification of Disease (ICD-9) and diagnostic-related groups (DRGs) in an attempt to reduce expenditure. The code for revision THA is represented by ICD-9 81.53 or DRG 545. Any excess between the true cost of the procedure and the DRG allocation is suffered as a loss by the hospital. Hospitals were initially able to cover lost revenue from individual cases by cost-shifting payments from third party providers. The increase in Health Maintenance Organizations (HMO) soon made this impossible. As a result, hospitals have adopted many of the measures previously discussed, streamlining procedures and reducing costs. Several studies have demonstrated that the average cost of a complicated revision THA surgery now lies between $30,000 and $52,000.2,22 Often these procedures require significant bone grafting as well as more costly modular revision implants. Crowe23 demonstrated that despite improvements in length of stay, use of clinical pathways, and negotiated discounts on implants, the hospital loss at their institution for each revision hip replacement procedure averaged $5402. Concerns have been raised that DRG reimbursements are now so low as to prevent hip arthroplasty.24 Other workers have questioned whether the surgical fee is so low as to make revision total hip replacement economically non-viable for surgeons.25 The Medicare reimbursement to surgeons for reconstructive hip surgery decreased 35% from 1991 to 2004 while the reimbursement to hospitals increased by 16%.26 It has even been suggested that tertiary referral hospitals may need to start limiting referrals of patients with failed THA who are in need of revision surgery.27 Grouping all revision THA under one DRG 545 code is unrealistic. An expansion of the basic DRGs, known as the All-Patient Refined, Diagnostic-Related Groups Severity of Illness (APR-DRG SOI), is a weighted index reflecting patient

A.W.R. Burns, R.B. Bourne baseline medical health ranging from one (lowest severity of illness) to 4 (highest severity of illness). Adoption of this refined DRG or a similar instrument allows categorization of the complexity of revision surgery to be taken into account.

Methods of decreasing revision THA rates The ultimate goal is to avoid the requirement for revision THA. The implantation of a durable primary THA with the minimal of complications is the main aim. Katz demonstrated that 12% of all primary THAs and 49% of revision THAs in the US Medicare population are performed at centres where ten or fewer of these procedures were undertaken each year.28 Significantly, 52% of primary and 77% of revision THA were performed by surgeons who carry out ten or fewer of these operations annually. Their study demonstrated lower complication rates and mortality for high-volume surgeons (those performing greater than 50 primary THAs or 10 revision THAs per year). Because certain complications such as deep infection or recurrent dislocation often necessitate early revision surgery, the reduction in these complications will have a beneficial effect upon the requirement for revision. Further work needs to be undertaken to elucidate exactly how the higher volume hospitals and surgeons achieve these results as the reasons are likely to be multifactorial. Adoption of similar practices may lead to improvements in outcomes globally at smaller centres. Functional outcomes 3 years after THA are not better when undertaken in busier units by surgeons with a greater turnover of procedures, despite the fact that early complication rates are lower.29 Technological improvements in materials also have the potential to decrease the revision rate in future if the theoretical advantages of contemporary bearing surfaces with very low wear rates translate to improved prosthetic longevity.14 Surgeons should base their choice of implant on best current data and be critical before adopting a new hip prosthesis. Ideally, the introduction of different materials and implants should be on a prospective, controlled basis beginning with a selected few centres to confirm safety before widespread acceptance. Evidence-based decisionmaking should be the primary guide, if we are to avoid failures from the implementation of unproven, if novel, implants.

Summary Revision THA is a cost-effective procedure which has been shown to improve quality of life, reduce pain and improve function for patients with failed primary arthroplasties. The increasingly aged population requires more primary arthroplasty than ever before and the need for revision THA is likely to increase as a result, leading to a greater financial burden for society. Health systems and surgeons must work to minimize this expenditure while ensuring that excellent outcomes are maintained. Reducing the rate of revision THAs by the implantation of durable primaries and the reduction in complications will be the most important longterm solution. The use of arthroplasty registers is a crucial part of this process. Evidence-based decision-making and

ARTICLE IN PRESS Economics of revision total hip arthroplasty the careful introduction of new technology in a prospective controlled fashion will assist toward better and changed practice.

Practice points

 Revision THA is a cost-effective procedure by health economic standards

 Hip arthroplasty cost the USA $8 billion in 2003, with projection for $23 billion for 2030

 Halving revision THA rates would currently save the USA almost US$ 1 billion per year

 Evidence-based medicine has the potential to reduce revision THA rates markedly

 National arthroplasty registers have great potential in this regard

Research directions

 Comprehensive healthcare cost analyses in the literature

 Epidemiologically-based estimation of future arthroplasty requirements

 Technological improvements (improved fixation, reduced wear)

 Biological improvements (antimicrobial surface 

coatings, bone stimulating proteins, bisphosphonates to reduce stress shielding or osteolysis) Worldwide adoption of national arthroplasty registers

References 1. Ethgen O, Bruyere O, Richy F, Dardennes C, Reginster JY. Health-related quality of life in total hip and total knee arthroplasty. A qualitative and systematic review of the literature. J Bone Jt Surg Am 2004;86-A(5):963–74. 2. American Academy of Orthopedic Surgeons Website. Data from 2003 analyzed from the National Center for Health Statistics. Accessed 20 November 2005; 2005. 3. Australian Orthopaedic Association National Joint Replacement Registry 2004 Annual Report. DMAC, University of Adelaide, pp. 6–7. 4. Canadian Joint Replacement Registry 2004 Report. Total hip and total knee replacements in Canada. Canadian Institute for Health Information; 2004. p. 43. 5. National Joint Registry for England and Wales. First Annual Report 2004. NJR Centre, Harwell, Didcot, Oxfordshire OX11 0QJ England; 2004. p. 4. 6. New Zealand National Joint Replacement Register. New Zealand Orthopaedic Association, Canterbury District Health Board; 2004. 7. Norwegian Arthroplasty Register. Department of Orthopaedic Surgery, Haukeland University Hospital; 2004. 8. Scottish Arthroplasty Register. NHS Scotland; 2004. 9. Swedish National Hip Arthroplasty Register. Department of Orthopaedics, Sahlgrenska University Hospital, Go ¨teborg, Sweden; 2003.

207 10. Scranton Jr. PE. The cost effectiveness of streamlined care pathways and product standardization in total knee arthroplasty. J Arthroplasty 1999;14(2):182–6. 11. Berger RA, Jacobs JJ, Meneghini RM, Della VC, Paprosky W, Rosenberg AG. Rapid rehabilitation and recovery with minimally invasive total hip arthroplasty. Clin Orthop Relat Res 2004(429):239–47. 12. Bertin KC. Minimally invasive outpatient total hip arthroplasty: a financial analysis. Clin Orthop Relat Res 2005(435): 154–63. 13. Lavernia CJ, Drakeford MK, Tsao AK, Gittelsohn A, Krackow KA, Hungerford DS. Revision and primary hip and knee arthroplasty. A cost analysis. Clin Orthop Relat Res 1995(311):136–41. 14. Hannouche D, Hamadouche M, Nizard R, Bizot P, Meunier A, Sedel L. Ceramics in total hip replacement. Clin Orthop Relat Res 2005(430):62–71. 15. Healy WL, Iorio R, Lemos MJ, Patch DA, Pfeifer BA, Smiley PM, et al. Single price/case price purchasing in orthopaedic surgery: experience at the Lahey Clinic. J Bone Jt Surg Am 2000;82(5):607–12. 16. Bozic KJ, Saleh KJ, Rosenberg AG, Rubash HE. Economic evaluation in total hip arthroplasty: analysis and review of the literature. J Arthroplasty 2004;19(2):180–9. 17. Laupacis A, Feeny D, Detsky AS, Tugwell PX. How attractive does a new technology have to be to warrant adoption and utilization? Tentative guidelines for using clinical and economic evaluations. CMAJ 1992;146(4):473–81. 18. Chang RW, Pellisier JM, Hazen GB. A cost-effectiveness analysis of total hip arthroplasty for osteoarthritis of the hip. J Am Med Assoc 1996;275(11):858–65. 19. Rorabeck CH, Bourne RB, Laupacis A, Feeny D, Wong C, Tugwell P, et al. A double-blind study of 250 cases comparing cemented with cementless total hip arthroplasty. Cost-effectiveness and its impact on health-related quality of life. Clin Orthop Relat Res 1994(298):156–64. 20. Ontario Ministry of Health and Long-Term Care. Funding allocations 2004/2005. Ontario Ministry of Health and Longterm Care; 2004. 21. Herberts P, Malchau H. Long-term registration has improved the quality of hip replacement: a review of the Swedish THR Register comparing 160,000 cases. Acta Orthop Scand 2000;71(2):111–21. 22. Bozic KJ, Katz P, Cisternas M, Ono L, Ries MD, Showstack J. Hospital resource utilization for primary and revision total hip arthroplasty. J Bone Jt Surg Am 2005;87(3):570–6. 23. Crowe JF, Sculco TP, Kahn B. Revision total hip arthroplasty: hospital cost and reimbursement analysis. Clin Orthop Relat Res 2003(413):175–82. 24. Boardman DL, Lieberman JR, Thomas BJ. Impact of declining reimbursement and rising hospital costs on the feasibility of total hip arthroplasty. J Arthroplasty 1997;12(5): 526–34. 25. Ritter MA, Carr KD, Keating EM, Faris PN, Bankoff DL, Ireland PM. Revision total joint arthroplasty: does medicare reimbursement justify time spent? Orthopedics 1996;19(2):137–9. 26. Mendenhall S. Hip and knee review. Orthop News Network 2004;15:1–16. 27. Sculco TP. The economic impact of infected total joint arthroplasty. Instr Course Lect 1993;42:349–51. 28. Katz JN, Losina E, Barrett J, Phillips CB, Mahomed NN, Lew RA, et al. Association between hospital and surgeon procedure volume and outcomes of total hip replacement in the United States medicare population. J Bone Jt Surg Am 2001; 83-A(11):1622–9. 29. Katz JN, Phillips CB, Baron JA, Fossel AH, Mahomed NN, Barrett J, et al. Association of hospital and surgeon volume of total hip replacement with functional status and satisfaction three years following surgery. Arthritis Rheum 2003;48(2):560–8.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 208–211

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

WRIST

What’s new in the treatment of distal radius fractures? F. Lama,, N. Jaysekerab, S. Karmanib, J.B. Jupiterc a

Department of Orthopedics, Eastbourne District General Hospital, Kings Drive, Eastbourne, East Sussex BN21 24D, UK Department of Orthopedics, Mayday University Hospital, 530 London Road, Thornton Heath, Croydon CR7 7YE, UK c Orthopedic Associates, 55 Fruit Street, YAW 2162, Boston, MA 02114-2696, USA b

KEYWORDS Distal radius fractures; Plates; Fixation

Summary The distal radius is the commonest site of fracture in the human skeleton and its treatment continues to evolve with advancing technology. Recently, there have been further refinements in the design of fixed-angle devices, pegs, locking plates and screws as well as arthroscopic instruments to assist management. In this article, we shall discuss the use of fixed-angle fixation devices, internal distraction plating, arthroscopically assisted reduction, fragment-specific fixation and the use of bone graft substitutes in the treatment of distal radius fractures. & 2006 Elsevier Ltd. All rights reserved.

Evolution of distal radius plates Since the introduction of the small fragment T plate by Mathys in 1973, there have been further refinements in the design of low-profile anatomically contoured plates, fixedangle devices and pegs. Since most distal radius fractures are dorsally displaced, it seems rational for the fixation plate to be applied dorsally. The Pi plate (Synthes, Paoli, Pennsylvania) was one of the earlier designs of anatomically contoured plates specifically intended for dorsal fixation of the radius. However, despite its low contact design, there were problems with tendon irritation limiting its use. One study compared the results of Pi plate fixation with two quarter tubular plate fixation and found that the incidence Corresponding author.

E-mail address: [email protected] (F. Lam). 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.02.008

of extensor tenosynovitis was higher with the Pi plate and the range of wrist motion was also less compared with the quarter tubular plating group.1 After this initial disappointment with dorsal plating, there was much interest in the use of volarly applied fixed-angle fixation devices for both dorsal and volar displaced fractures. These fixed-angle devices include fixed-angle blades, e.g. SCS plate (Subchondral Support plate, Avanta, San Diego, CA) which support the metaphyseal fragments and locking screws/pegs, e.g. DVR plate (Distal Volar Radius Plate, Hand Innovations, Miami, Florida). The DVR plate has two rows of pegs, a proximal row supporting the dorsal aspect of the subchondral bone and a distal row supporting the central and volar aspect of the subchondral bone. The plate also extends over the volar margin of the lunate fossa providing a buttressing effect. Biomechanically, a volarly placed fixed-angle device may provide more stability than a dorsal one since the articular

ARTICLE IN PRESS Treatment of distal radius fractures margin of the distal radius normally has a volar inclination which displaces the joint reaction force closer to the volar plate, and thus decreasing the bending moment.2 Recently, there have been further refinements in the design of dorsal fixation using 2.4 mm fragment-specific plates (AO/ASIF, Synthes, USA).3 There exists a radial column plate for fixation of the radial column and a T- or L-shaped plate for dorsal fixation of the intermediate column. Several modifications were made to reduce tendon irritation; these include polished surfaces with smooth tapered edges, precontouring of the plate thus reducing the need for cutting and bending and low-profile 2.4 mm screws.

Internal distraction plating Internal distraction plating of the distal radius involves the application of a dynamic compression plate from the radius to the middle-finger metacarpal bypassing the distal radius fracture site (Fig. 1). This technique is most useful in highly comminuted osteoporotic distal radial fractures involving the metaphyseal–diaphyseal junction. It may serve as an alternative to a bridging external fixator especially in elderly patients, avoiding the complications associated with external pins. The surgical technique involves two and sometimes three incisions.4 The first incision is centred over the middle-finger metacarpal shaft. The second incision is placed over the dorsal aspect of the radial shaft at least 4 cm from the

Figure 1 The distraction plate is applied to the middle-finger metacarpal and radial shaft with the wrist in neutral rotation.

209 fracture site. Sometimes, a third incision is required over Lister’s tubercle to facilitate passage of the plate under the extensor tendons and to retract the extensor pollicis longus tendon. In situations where there is extensive metaphyseal bone loss, supplementary bone graft can also be inserted into the defect through this incision. After regaining radial length by manual traction, the plate is fixed to the bone with the wrist in neutral position. A lag screw can also be inserted through the plate if the intra-articular fragment is large enough. In a series of 22 patients with a comminuted distal radial fracture treated with internal distraction plating, Ruch reported excellent results in 14 patients, good in six patients and fair in two.5 The average time to union was 110 days and the distraction plate was removed after fracture consolidation at an average of 124 days. Interestingly, there was no correlation between the duration of plate immobilisation and the range of wrist movements or functional scores at follow-up. The most common complication encountered was extensor lag of the middle finger (10%).

Fragment-specific fixation The concept of fragment-specific fixation evolved as it became clear that the use of a single-sized plate cannot be used for fixation for all fracture configurations. Since the introduction of the ‘3 column concept’ of the distal forearm by Rikli and Regazzone,6 a number of biomechanical studies have shown that rigid fixation of both the radial and intermediate columns is required for adequate stability, supporting radial inclination and radial tilt.7 The Trimed Wrist Fixation System (Trimed, Valencia, California) was introduced in 1996 by Medoff in Hawaii and consists of a highly versatile range of low-profile anatomically contoured plates, plate-supported pins (Fig. 2), wire forms and buttress wire forms (Fig. 3) to address each individual fracture fragment. The radial column is addressed with a radial plate, the dorsal intra-articular fragment fixed with a wire form, the dorsal ulnar fragment fixed with an ulnar pin plate and the volar articular fragment stabilised with an L plate. The surgical technique typically involves a combination of volar and dorsal approaches.8 The volar Henry approach between flexor carpi radialis and radial artery is used to expose the volar aspect of the distal radius after subperiosteal elevation of pronator quadratus. To obtain access to the radial column, the extensor tendons of the first compartment will also need to be elevated subperiosteally. Dorsally, a standard approach over the third extensor compartment is used and the extensor pollicis longus tendon is released and retracted radially. Fracture reduction should begin only after both of these approaches have been carried out to evaluate the fracture configuration adequately. To address a volarly displaced fragment, a volar wire form can be used. The wire form is first pre-contoured to fit the shape of the volar cortex and later impacted into the bone after drilling two holes into the volar cortex. The fixation is secured after application of two washers and screws (Fig. 3). The wire form can also be applied dorsally and in cases of a fracture involving the dorso-ulnar portion of the distal radius, an ulnar pin plate can be used which is ideal for small

ARTICLE IN PRESS 210

F. Lam et al. advantage of this fixation technique is related to its superior biomechanical stability since each major fragment is stabilised, thus allowing early mobilisation without the need for additional immobilisation. The main complication is tendon irritation and rupture, with up to 23% of patients (21/92) requiring hardware removal subsequently.

Arthroscopically assisted fixation

Figure 2 The radial pin plate can be used to stabilise the radial column by means of a combination of cortical screws proximally and K wires distally. The K wires are passed through the proximal pin holes of the plate into the distal fragment.

Figure 3 The volar pin plate is ideal for fixation of a displaced volar ulnar fragment. The volar wire form is prebent to fit the contour of the volar ulnar aspect of the radius and is secured with two screws and washers.

ulnar-sided corner fragments. The rationale of the pin plate is that it acts as a cortical buttress and also provides strong two-point K wire fixation of the small distal fragment (Fig. 2). The first point of fixation is achieved by the K wire penetrating through the far cortex of the proximal intact diaphysis and the second point of fixation is where the wire is supported by the plate. For the fixation of the radial column, there is a separate radial pin plate. There is also a buttress clip which is used for a central die punch fragment allowing elevation of the depressed fragment and supplementary bone graft if necessary. Several investigators have reported encouraging results of fragment-specific fixation with over 85% of AO-type C2 and C3 fractures achieving good or excellent results.9 The main

In 1986, Knirk and Jupiter demonstrated that patients with an articular step off at the radiocarpal joint of more than 2 mm were more likely to develop degenerative wrist disease.10 More recently, other investigators have found that the complication rate is significantly lowered if the articular incongruency is less than 1 mm.11 This level of accuracy is difficult to achieve using fluoroscopy alone and hence arthroscopy has been introduced as an adjunct to assist fracture reduction. This technique is particularly useful in 3-part and 4-part fractures with fragments greater than 1 cm. Furthermore, arthroscopy can also detect carpal ligamentous injuries, distal radioulnar joint instability as well as osteochondral flaps and loose bodies. In a series of 60 patients with an intra-articular fracture of the distal radius, Geissler found during arthoscopy that 49% had a tear of the triangular fibrocartilage complex, 32% had injury to the scapholunate ligament and 15% had injury to the lunotriquetral ligament.12 The surgical technique is well described in several papers.13,14 Essentially, the conventional 3–4 portal between extensor pollicis longus and extensor digitorum communis tendons is used for visualisation. Irrigation of the joint is achieved through the 6U portal and the 6R portal is the main working portal where a shaver can be inserted to debride the joint improving visualisation. The ideal fracture configurations for arthoscopically assisted reduction are radial styloid fractures, die punch fractures, and 3-part and 4-part fractures. Generally, the radial styloid fragment is reduced first and stabilised with either K wires or cannulated screws. Sometimes, an additional K wire can be passed into the radial styloid fragment acting as a joystick to aid manipulation. Following this, the depressed lunate facet fragment is then elevated to the radial styloid fragment which acts as a landmark. The reduction is then stabilised with multiple K wires and the accuracy checked arthroscopically. Finally, the fixation is completed after application of the plate. Several studies have demonstrated that arthroscopically assisted reduction and internal fixation yield better results with greater range of wrist motion than conventional open reduction and internal fixation.15–17 In addition, Ruch found that 10 of the 15 patients treated with arthroscopic reduction were found to have a tear of the triangular fibrocartilage complex, seven of which were peripheral and repaired acutely.17 This has important prognostic significance since treated acute tears do better than chronic ones. At final follow up, there were no cases of distal radio-ulnar joint instability in the arthroscopic group compared with 27% (4/12) in the closed reduction and external fixation group. In recent years, there has been an increased awareness of the significance of ulnar-sided wrist injuries which have previously been underestimated. It is estimated that 20% of

ARTICLE IN PRESS Treatment of distal radius fractures patients after sustaining a distal radius fracture complain of persistent ulnar-sided wrist pain. Common reasons for this include malunion and consequent shortening of the radius leading to a relatively long ulna causing abutment, tear of the triangular fibrocartilage complex, distal radio-ulnar joint instability, distal radio-ulnar joint incongruency and non-union of hypertrophic ulnar styloid.18

Bone graft substitutes In the past few years, there has been increasing interest in the use of bone graft substitutes in the treatment of distal radius fractures. Although autogenous bone graft has been traditionally regarded as the gold standard, its use is associated with significant donor site morbidity. Bone grafting is generally indicated in distal radius fractures: (1) for structural bone support, for example, if there is a significant bone defect particularly in osteoporotic bone, and (2) to augment healing in fracture non-union. The bone graft substitutes available include those based on naturally occurring materials such as demineralised allograft bone matrix, bovine collagen mineral composites, coralline hydroxyapatite and synthetic materials such as calcium sulphate pellets, bioactive glass, and calcium phosphate cement. Norian SRS is an injectable paste consisting of monosodium phosphate, tricalcium phosphate, calcium carbonate and sodium phosphate. After excellent results were reported by an initial study,19 the use of Norian SRS was further supported by a prospective randomised study showing that patients treated with Norian SRS with or without adjuvant external fixation had significantly earlier return to function than the group treated with casting or percutanous pin or external fixation alone.20 The use of carbonated hydroxyapatite in distal radius corrective osteotomies was reported by Luchetti who found a 100% union rate with radiographic evidence of complete graft integration into the bone tissue.21 There was improvement in range of wrist motion, forearm rotation as well as grip strength. With the increasing ageing of the population and a higher incidence of high-velocity trauma, the need for bone graft material is likely to expand. The future management of distal radius fractures is likely to be revolving around genetic engineering and refinements in internal fixation techniques.

References 1. Hahnloser D, Platz A. Internal fixation of distal radius fractures with dorsal dislocation: pi plate or two 1/4 tube plates? A prospective randomized study. J Trauma 1999;47:760–5. 2. Orbay J, Badia A, Khoury R, et al. Volar fixed angle fixation of distal radius fractures: the DVR plate. Tech Hand Upper Extremity Surg 2004;8(3):142–8.

211 3. Tavakolian J, Jupiter JB. Dorsal plating for distal radius fractures. Hand Clin 2005;21:341–6. 4. Papadonikolakis A, Ruch DS. Internal distraction plating of distal radius fractures. Tech Hand Upper Extremity Surg 2005;9(1): 2–6. 5. Ruch DS, Ginn TA, Yang CC, et al. Use of a distraction plate for distal radial fractures with metaphyseal and diaphyseal comminution. J Bone Jt Surg (Am) 2005;87(5):945–54. 6. Rikli D, Regazzoni P. Fractures of the distal end of the radius treated by internal fixation and early function: a preliminary report of 20 cases. J Bone Jt Surg 1996;78B(4):588–92. 7. Naidu SH, Capo JT, Moulton M, et al. Percutaneous pinning of distal radius fractures: a biomechanical study. J Hand Surg (Am) 1997;22:252–7. 8. Schumer ED, Leslie BM. Fragment specific fixation of distal radius fractures using the Trimed device. Tech Hand Upper Extremity Surg 2005;9(2):74–83. 9. Price JS, Korris M, Leslie B, et al. Initial outcome of distal radius fractures treated with the Trimed Wrist Fixation System. In: Presented at the 56th Annual Meeting of the American Society for Surgery of the Hand, Baltimore, MD, October 2001. 10. Knirk JL, Jupiter JB. Intrarticular fractures of the distal end of the radius in young adults. J Bone Jt Surg 1986;68A: 647–58. 11. Fernandez DL, Geissler WB. Treatment of displaced articular fractures of the radius. J Hand Surg (Am) 1991;16A: 375–84. 12. Geissler WB, Freeland AE, Savoie FH, et al. Intracarpal soft tissue lesions associated with an intrarticular fracture of the distal end of the radius. J Bone Jt Surg 1996;78A:357–65. 13. Guofen C, Doi K, Hattori Y, Kitajima I. Arthroscopically assisted reduction and immobilization of intrarticular fracture of the distal end of the radius: several options of reduction and immobilization. Tech Hand Upper Extremity Surg 2005;9(2): 84–90. 14. Geissler WB. Intra-articular distal radius fractures: the role of arthroscopy? Hand Clin 2005;21:407–16. 15. Stewart NJ, Berger RA. Comparison study of arthroscopic and open reduction of comminuted distal radius fractures. In: Presented at the 53rd Annual Meeting of the American Society for Surgery of the Hand [programs and abstracts]. January 11, 1998, Scotsdale, AZ. 16. Doi K, Hatturi T, Otsuka K, Abe T, Tamamoto H. Intraarticular fractures of the distal aspect of the radius arthroscopically assisted reduction compared with open reduction and internal fixation. J Bone Jt Surg 1999;81A:1093–110. 17. Ruch DS, Vallee J, Poehling GG, Smith BP, Kuzma GR. Arthroscopic reduction versus fluoroscopic reduction of intra-articular distal radius fractures. Arthroscopy 2004;20: 225–30. 18. Lindau T. Treatment of injuries to the ulnar side of the wrist occurring with distal radial fractures. Hand Clin 2005:417–25. 19. Jupiter JB, Winters S, Sigman S, et al. Repair of five dustal radius fractures with an investigational cancellous bone cement: a preliminary report. J Orthop Trauma 1997;11:110–6. 20. Ladd AL, Pliam NB. Use of bone graft substitute in distal radius fractures. J Am Acad Orthop Surg 1999;7:279–90. 21. Luchetti R. Corrective osteotomy of malunited distal radius fractures using carbonated hydroxyapatite as an alternative to autogenous bone grafting. J Hand Surg [Am] 2004;29(5): 825–34.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 212–215

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

SYNDROMES

Down syndrome Rouin Amirfeyza,, Demetris Asprosb, Martin Garganc a

Trauma and Orthopaedics, Yeovil District Hospital, Yeovil, UK Trauma and Orthopaedics, Southmead Hospital, Bristol, UK c Bristol Royal Hospital for Sick Children, Bristol, UK b

Introduction Dr. John Langdon Down, the English physician, first described this syndrome in 1866 in his article that some might consider it today as ‘politically incorrect’.1 It is one of the most common chromosomal disorders. The overall life expectancy of the sufferers has increased from 9 years in 1925 to over 40 years today. As a result the orthopaedic problems are now more commonly seen. Orthopaedic manifestations are mainly attributed to ligament laxity and include atlantoaxial instability (AAI), scoliosis, congenital hip dysplasia, hip dislocation (and secondary osteoarthritis), patellofemoral instability, pes planus, and metatarsus primus varus.2–4

Epidemiology and genetics The incidence of Down syndrome is approximately 1 in 700 live births increasing with maternal age. Trisomy of chromosome 21 is the most common genetic abnormality; however, cases of translocation and mosaicism were also reported.

Prenatal diagnosis and screening High free b chain of human chorionic gonadotrophin (b-HCG) and low pregnancy associated protein A (APAPP-A) measured in maternal blood between 8th and 14th week of gestation

Corresponding author.

0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.01.003

indicates a probable Down syndrome affected fetus. These blood tests can be combined with nuchal translucency measurement (NTM) to increase the accuracy of the diagnosis. NTM is ultrasound scanning (USS) detection of a thickened and oedematous flap of skin at the base of the neck. A measurement of more than 3 mm is considered to be abnormal (Fig. 1). NTM can detect up to 80% of Down syndrome affected fetuses during the first trimester of pregnancy.

Orthopaedics manifestations Atlantoaxial instability Approximately 15% of individuals with Down syndrome have AAI. However, most of these patients are asymptomatic, and only 1–2% have a serious neck problem that requires surgical intervention. Symptomatic AAI is either due to subluxation (which can be severe enough to damage the spinal cord) or dislocation. The manifestations of symptomatic AAI include fatiguability, difficulties in walking, abnormal gait, neck pain, limited neck mobility, torticollis, incoordination, sensory deficit, and hypereflexia. Signs and symptoms often remain stable for months or years, although sometimes patients can progress to paraplegia, quadriplegia, or death. The minimum required radiological investigations include an antero-posterior, flexion and extension lateral, and odontoid view radiographs. The atlantodens interval (ADI) is the measured distance between the posterior edge of the ring of C1 and the anterior edge of the odontoid peg (Fig. 2). It is normally less than 3.5 mm in adults.

ARTICLE IN PRESS Down syndrome

Figure 1

USS of fetus: note the measured nuchal translucency.

Treatment depends on history, physical examination, and the ADI. Recommended treatment for AAI (ADI ¼ atlantodens interval in mm) ADI 4–5 ADI 6–9

ADIX10

213

Restriction of high-risk activities MRI or CT is recommended to investigate the presence of any neurological compromise before planning potential treatment Posterior cervical fusion and wiring

Scoliosis Scoliosis is common in patients with Down syndrome. Treatment is the same as in other individuals, with bracing being the initial therapy. If necessary it is followed by surgical intervention i.e. arthrodesis of the spine.5,6

Figure 2 Atlantodens interval (ADI) is the distance between the posterior edge of the ring of atlas and the anterior edge of the odontoid peg.

The treatment for first-time dislocation begins with hip immobilisation in a cast. However, this is usually unsuccessful and frequently needs to be followed by surgical intervention. Surgical options include capsular plication, varus derotation osteotomy (as the proximal femur typically is anteverted and in valgus position) (Fig. 4), and periacetabular osteotomy.8–11 Slipped upper femoral epiphysis is more common in Down syndrome patients. This condition is often associated with obesity and hypothyroidism, both of which are common in teenagers with Down syndrome. Clinical presentation and treatment are the same as other individuals. Legg-Calve-Perthes disease is another disorder more commonly seen in children with Down syndrome. Usually it presents with a painless limp and a decrease in range of movement. Diagnosis is made radiologically. Treatment is the same as for other patients.

Hip abnormalities A total of 5–8% of children with Down syndrome will develop at least one form of the following hip pathologies:








dislocation, developmental dysplasia (Fig. 3), slipped upper femoral epiphysis, Legg-Calve-Perthes disease, avascular necrosis.

Dislocation of the hip is the most common hip abnormality, occurring in 4.5% of patients (Fig. 3). Down syndrome patients have increased external rotation of the hip. The acetabulum is deep, with a horizontal roof and reduced anteversion. The femoral head sits more deeply. Mechanically the hip joint is stable; however, due to ligament laxity and increased range of movement (especially in external rotation) the hip can dislocate. The proximal femur has a normal neck–shaft angle and a moderate increased anteversion.7 Interestingly, hip subluxation in children with Down syndrome is hardly ever found at birth but instead is most common between the ages of 3 and 13 years. The most common sign of dislocation is a limp, with or without pain. Such patients also have a delay in walking and abnormal mobility of the hip joint.

Patellofemoral instability Instability of the patellofemoral joint is seen in almost 20% of the patients with Down syndrome, but is rarely disabling. It is commonly missed and overlooked during clinical examination (Fig. 5). Dugdale and Renshaw classification of patellofemoral instability in Down syndrome12


Grade 1: stable patellofemoral joint
Grade 2: patella can be subluxated laterally to more than half of the patellar width


Grade 3: dislocatable
Grade 4: dislocated but reducible
Grade 5: dislocated irreducible

Fortunately, the majority of the patients are in grades 1 or 2. Patients with grades 4 or 5 are usually pain free and surprisingly patients with grade 5 PFI and good functional state were reported in the past.

ARTICLE IN PRESS 214

R. Amirfeyz et al.

Figure 3 Left: AP view of both hips showing severe bilateral DDH in a patient with Down syndrome. Right: severe osteoarthritis of right hip in a neglected case.

Figure 4 Bilateral varus derotation osteotomy (VDRO) to correct coxa valga and external rotation of proximal femur.

Treatment depends on the function, mobility, other deformities around the knee and the grading of instability. Surgical intervention without addressing other deformities will result in osteoarthritis of patellofemoral joint. Patients with grade 3 PFI are usually the candidates for realignment surgery.

Foot abnormalities Pes planus is seen in the vast majority of these patients. In mild cases, the heel is in a neutral position whereas in severe cases, the heel goes to valgus with the pronation and collapse of the midfoot. Flatfoot results in heavy calluses and bone spurs. Some cases respond to orthotics, but the majority are resistant to non-operative treatment and as a result surgical correction is preferred. Metatarsus primus varus is also commonly seen in people with Down syndrome. Mild or early cases may be treated with orthotics, but severe cases require surgical correction. Surgical techniques for correction of pes planus and hallux valgus are the same as for other patients.

Figure 5 CT scan of both patello-femoral joints in an individual with Down syndrome. Note bilateral subluxated patellae, more severe on the left side.

References 1. Down JLH. Observations on an ethnic classification of idiots. London Hosp Rep 1866;3:259–62. 2. Diamond LS, Lynne D, Sigman B. Orthopedic disorders in patients with Down’s syndrome. Orthop Clin North Am 1981;12(1):57–71. 3. Merrick J, Ezra E, Josef B, Hendel D, Steinberg DM, Wientroub S. Musculoskeletal problems in Down Syndrome European Paediatric Orthopaedic Society survey: the Israeli sample. J Pediatr Orthop B 2000;9(3):185–92. 4. Pueschel SM, Solga PM. Musculoskeletal disorders. In: Pueschel SM, Oueschell JK, editors. Biomedical concerns in persons with Down syndrome. Brookes Pub Co.; 1992. p. 147–57. 5. Roy M, Baxter M, Roy A. Atlantoaxial instability in Down syndrome—guidelines for screening and detection. J R Soc Med 1990;83(7):433–5. 6. Lerman JA, Emans JB, Hall JE, Karlin LI. Spinal arthrodesis for scoliosis in Down syndrome. J Pediatr Orthop 2003;23(2): 159–61.

ARTICLE IN PRESS Down syndrome 7. Shaw ED, Beals RK. The hip joint in Down’s syndrome. A study of its structure and associated disease. Clin Orthop 1992;278: 101–7. 8. Bennet GC, Rang M, Roye DP, Aprin H. Dislocation of the hip in trisomy 21. J Bone Joint Surg Br 1982;64:289–94. 9. Greene WB. Closed treatment of hip dislocation in Down syndrome. J Pediatr Orthop 1998;18:643–7.

215 10. Kioschos M, Shaw ED, Beals RK. Total hip arthroplasty in patients with Down’s syndrome. J Bone Joint Surg Br 1999;81:436–9. 11. Skoff HD, Keggi K. Total hip replacement in Down’s syndrome. Orthopedics 1987;10:485–9. 12. Dugdale TW, Renshaw TS. Instability of the patellofemoral joint in Down syndrome. J Bone Joint Surg Am 1986;68(3): 405–13.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 216–221

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

JOINT REPLACEMENT

Obesity and total knee and hip replacement Anish K. Amin, James D. Sales, Ivan J. Brenkel Department of Orthopaedics, Queen Margaret Hospital, Fife KY 12 0SU, UK

KEYWORDS Obesity; Total hip replacement; Total knee replacement

Summary As the prevalence of obesity rises, the number of hip and knee replacements performed on obese patients will rise. This article reviews the correlation between obesity and the development of osteoarthritis in the hip and knee, and outcomes. & 2006 Elsevier Ltd. All rights reserved.

Introduction The Body Mass Index (BMI) is an individual’s weight divided by height squared; this ratio correlates with the amount of total body fat.1 The normal range is 15–25 kg/m2, 25–30 kg/ m2 is overweight,430 kg/m2 is obese and 440 kg/m2 morbidly obese.2–4 It has long been recognised that BMI is a predictor of morbidity and mortality for several chronic diseases, including diabetes mellitus, coronary artery disease and stroke, with this health risk increases linearly. Recently, three primary care trusts in the United Kingdom have ruled that patients with a BMI of over 30 kg/m2 will not be entitled to hip and knee replacement surgery. This has sparked a debate by health professionals and politicians. This article reviews the published literature to see if this decision is justified in obese patients undergoing knee and hip replacements. The currently accepted definition of obesity1 is the BMI. Because earlier studies in the 1980s and early 1990s used relative body weight based on Metropolitan Life Insurance Tables5 to define obesity as 120% ‘ideal body weight’, they

Corresponding author. Tel.: +44 1283623623x3041.

E-mail address: [email protected] (I.J. Brenkel). 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.02.015

have been excluded to allow comparison of published results.

Prevalence of obesity in the United Kingdom The prevalence of obesity in males has increased from 6% in 1980 to 22% in 2002. Females have shown a similar increase (8–23%).6 Not only does the UK have some of the worst figures in Europe but it also has some of the worst trends. In the majority of European countries the prevalence of obesity has increased between 10% and 40% in the last 10 years, whereas in England it has more than doubled.6 In 1995, according to the World Health Organisation (WHO), there were an estimated 200 million obese adults worldwide and another 18 million children aged under five classified as overweight. However, by 2000, the number of obese adults had increased to over 300 million.7

Risk of obesity in the development of osteoarthritis of the hip and knee A positive association between obesity and BMI, and OA of the knee has been observed in cohort studies,8–12 a case– control study13 and cross-sectional studies.14–21 Weight loss has been shown to reduce the incidence of knee OA in

ARTICLE IN PRESS Obesity and total knee and hip replacement women in a cohort study.22 In the case–control studies,13 selection bias cannot be completely ruled out because they depend on symptoms for their case definition and patients with OA are more likely to have symptoms if they are obese.22 The relation with obesity and hip osteoarthritis is less clear. For example, several studies have suggested that obesity increases the risk of symptomatic23–31 but not radiographic hip osteoarthritis.32,33 Furthermore, because most studies have been cross-sectional23–30 it is not known if obesity preceded or followed hip pathology. Lievense et al.34 in a systematic review of the literature found associations between obesity and hip OA were stronger in studies in which the diagnosis of hip OA was based not only on radiological criteria but also on clinical symptoms. Overall, moderate evidence was found for a positive association between obesity and the occurrence of hip OA, with an odds ratio of approximately 2. Prospective cohort studies determine obesity status before the development of hip osteoarthritis. Karlson et al.,35 in a prospective cohort study of more than 120,000 women, found that only higher BMI and older age were associated with an increased risk of osteoarthritis requiring hip replacement surgery. In particular, women in the highest category of BMI had a twofold greater risk of hip arthroplasty, compared with those in the lowest category. A particularly intriguing finding by Karlson et al. was the relation of BMI at age 18 years to the risk of hip osteoarthritis. Risk estimates at age 18 years were significantly greater than those for ‘‘recent’’ BMI, which were reported closer to the date of surgery. For those who were obese at 18 there was a five fold increased risk of having a THR. Another study has estimated that if obesity were eliminated, the prevalence of hip osteoarthritis would decrease by 25%.36

217 undergoing total knee replacement. In one prospective study comparing 210 total knee replacements in non-obese patients with 160 total knee replacements in obese patients, there were no differences between the two groups for superficial wound infection, deep joint infection, deep vein thrombosis/pulmonary embolism or peri-operative mortality.44 In a retrospective matched study,42 the overall perioperative complication rate was reported as 3% in 78 total knee replacements performed in the obese group (a deep vein thrombosis and wound dehiscence in one patient and a foot drop in the other) and 0% in the non-obese group, but the difference was not statistically significant. In a crude analysis of ‘medical’ and ‘orthopaedic’ complications following total knee replacement, no difference was reported between patients sub-divided into various BMI categories.40 While these studies divided patients into obese and non-obese groups based on a BMI greater or less than 30 kg/m2 respectively, in one large prospective study of 1813 patients who underwent total knee replacement, the authors studied the peri-operative morbidity using a higher BMI value of 35 kg/m2 to divide the study sample into two groups.48 Patients with a BMI435 kg/m2 had a significantly higher deep joint infection rate compared to those with a BMIo35 kg/m2 (1.1% vs. 0.3%, respectively) although the rates of medical complications (cardiac, gastrointestinal, genitourinary and pulmonary events) and deep vein thrombosis was similar for both groups.48 In summary, although selection bias may be a significant confounding factor, reported results suggest the peri-operative complications in obese and non-obese patients following total knee replacement are similar, but infective complications may increase with BMI.

Clinical outcome scores

Total knee replacement in the obese patient Results of total knee replacement in the obese patient (BMI430 kg/m2) Several studies have compared the results of total knee replacement in obese patients, (BMI430 kg/m2) with nonobese patients, defined as BMIo30 kg/m2, with follow-up ranging from 1 to 15 years.37–44 Three main parameters must be considered when evaluating results of total knee replacement in obese and non-obese patients:


Short term the peri-operative complications.
Medium term, the clinical outcome scores.
Long-term survivorship of the prosthesis. Peri-operative complications While there are several reports regarding the influence of obesity on peri-operative complications in general, gynaecological and cardiac surgery,45–47 those following total knee replacement has been less well studied. Such studies as have been published have generally focused on wound complications, thromboembolic complications, medical events and peri-operative mortality. They can be criticised for selection bias, as high-risk patients are often discouraged from

Reported clinical outcome following total knee replacement is usually based on a composite score derived from joint specific and/or patient based formal outcome scoring systems. Studies comparing the 1-year clinical outcome using patient based outcome scoring systems have found no difference in the results between obese and non-obese groups.40,41 Studies with longer follow-up (Table 1) have used the Knee Society Score (KSS)49 to compare the clinical outcome following total knee replacement in obese and nonobese patients. The KSS is a joint-specific outcome scoring system and comprises a ‘knee score’ and ‘function score’. The knee score component of the KSS is derived by evaluating pain, range of movement and stability with deductions for flexion contracture, extensor lag and malalignment. The function score component is derived by evaluating walking distance and stair climbing ability with deductions for walking aids used. The knee and function scores are reported separately and both are scored out of a maximum of 100 points. Table 1 summarises the knee and function scores from studies comparing the results of total knee replacement in obese and non-obese patients with follow-up ranging from 5 to 15 years, but there are few large study samples beyond 7 years. It is clear that the clinical outcome based on the KSS does not show significant differences between obese and non-obese patients, except for one retrospective study42 in

ARTICLE IN PRESS 218

Table 1

A.K. Amin et al. Clinical outcome after TKR in obese (BMI 430 kg/m2) and non-obese (BMI o30 kg/m2) patients.

Study, follow-up

Groups

N

Mean KS

Mean FS

Summary

Level of evidencey

Amin et al.44 2006 5 years Spicer et al.37 2001 6 years Foran et al.42 2004 6.6 years Griffin et al.39 1998 10 years Foran et al.43 2004 15 years

Obese Non-obese Obese Non-obese Obese Non-obese Obese Non-obese Obese Non-obese

147 181 326 425 78 78 32 41 30 30

84 86 76 79 90 94 93 93 81 89

85 85 63 68 71 78 67 82 NR NR

No difference in scores

Level I, prospective study

No difference in scores

Level II, prospective study

Inferior KS in obese, FS similar

Level II, retrospective study

No difference in scores

Level II, retrospective study

No difference in KS, FS not reported

Level II, retrospective study

N—number of total knee replacements in each group, KS—knee score component of KSS, FS—function score component of KSS, NR—not reported.  Rounded to nearest decimal point. y Levels of evidence based on material published by the Centre for Evidence Based Medicine, Oxford, UK (www.cebm.net).

Table 2

Revision rates in obese patients compared to non-obese.

Study, follow-up

N obese vs. nonobese

Revision rate obese vs. nonobese

Radiographic analysis

Summary

Amin et al.44 2006 5 years Spicer et al.37 2001 6 years Foran et al.42 2004 6.6 years Mont et al.38 1996 7 years Griffin et al.39 1998 10 years

210 vs. 160 326 vs. 425 78 vs. 78 50 vs. 50 32 vs. 41

2.5% vs. 1.4% 4.9% vs. 3.1% 5% vs. 0% 8% vs. 4% 0% vs. 7.3%

No Yes Yes Yes Yes

No No No No No

statistical statistical statistical statistical statistical

difference difference difference difference difference

N—number of total knee replacements.  Calculated from numbers provided in study.

which the knee score component was found to be significantly inferior (P ¼ 0:04) in the obese group but not the function score component (P ¼ 0:05). In summary, the weight of the published evidence suggests that there is no difference in the clinical outcome in the mid-term between obese and non-obese patients assessed using patient based or joint specific outcome scoring systems. Larger, prospective studies are required to establish the long-term clinical outcome in obese patients following total knee replacement.

Prosthesis survivorship Survivorship of the implant is an important measure of longer term success of total knee replacement. In a survivorship analysis, it is assumed that all patients underwent the total knee replacement simultaneously, which allows analysis of data from patients with different lengths of follow-up with cases being ‘entered’ or ‘censored’ from the analysis at any stage and ‘failures’ identified. The survivorship for a particular length of time can be calculated either by constructing a life table (survivorship calculations based on failures per year of follow-up) or using Kaplan–

Meier methodology (survivorship recalculated each time a failure occurs).50 The literature comparing the survivorship of total knee replacement in obese and non-obese patients is difficult to analyse for two reasons. Firstly, a formal survivorship analysis has not been undertaken in the majority of the studies and most authors have reported revision rates with or without radiographic analysis (Table 2). Secondly, the two studies that have reported survivorship of the implant have been retrospective and involved relatively small numbers.42,43 Based on revision of the implant (for any reason) as the endpoint, none of the studies has demonstrated significant differences between obese and non-obese patients following total knee replacement (Table 2). Using revision, clinical failure and radiographic failure as endpoints however, one retrospective study found an inferior survivorship in obese patients (compared to non-obese patients) that became apparent after 60 months.42 At eighty months, survivorship of 78 total knee replacements in the obese group was 88%, compared to 99% in the same number of total knee replacements performed in a matched non-obese group.42 A further long-term study from the same centre comparing 30 total knee replacements in obese and non-obese groups

ARTICLE IN PRESS Obesity and total knee and hip replacement demonstrated a trend towards inferior survivorship in obese patients only after 14 years, but the differences were not statistically significant due to small numbers.43 In summary, present evidence suggests that mid-term survivorship is probably similar for obese and nonobese patients, but in the long-term, the survivorship may deteriorate in obese patients and requires further investigation.

Results of total knee replacement in morbidly obese patients (BMI440 kg/m2) Patients with a BMI440 kg/m2 are ‘morbidly obese’. If obesity were to have a negative influence on the results of total knee replacement, one would expect the inferior results to be most obvious in these patients with severe obesity. All published comparative studies using the BMI to divide patients into ‘obese’ (BMI430 kg/m2) and ‘non-obese’ (BMIo30 kg/m2) groups37–44 have included morbidly obese patients,37 within the ‘obese’ category. A separate analysis of results in the subgroup of obese patients who are morbidly obese was reported in one large study with an average follow up of about 6 years.38 A total of 326 knee replacements performed in the obese group were compared with 425 similar procedures performed in a matched group of non-obese patients.38 Of the total knee replacements performed in obese patients, 59 procedures were performed in the morbidly obese. The survivorship and clinical outcome scores were similar for obese and non-obese patients, but in the subgroup of obese patients who were morbidly obese, seven (12%) implants were revised or in need of revision by 6 years with a focal osteolysis rate five times higher than the non-obese control group. In a study with the primary aim of evaluating results of total knee replacement in the morbidly obese patient,51 50 primary total knee replacements performed in morbidly obese patients were compared with 1768 similar procedures performed in a control group of non-morbidly obese (BMIo40 kg/m2) patients by the same surgeon.51 The perioperative complication rate in morbidly obese patients was 26% compared to 2% in the control group. Additionally, at an average follow-up of about 5 years, the KSS was inferior in the morbidly obese with a 10% revision rate and a 10% deep joint infection rate (three revised, two re-operated with retention of implant). Although a high body weight results in increased stress across a total knee replacement and surrounding bone, it does not appear to produce high failure rates in obese patients (BMI430 kg/m2) who have total knee replacements. This is probably due to the lower activity levels in these patients compared to non-obese patients.52 It is possible however, that in the subgroup of obese patients who are morbidly obese (BMI440 kg/m2), lower activity levels may not compensate for the much higher stresses across the knee joint. The high rate of infective complications following total knee replacement in the morbidly obese patient appears to substantiate similar problems noted following gastric bypass surgery to treat morbid obesity.53

219

Total hip replacement in the obese patient There is little evidence describing the influence of BMI on the outcome of total hip replacements (THR).

Postoperative infection Our prospective study reported 800 consecutive hip replacements in 759 patients.54 Thirty-three per cent were obese. One of the strengths of our paper is the use of regression analysis to identify independently significant predictors. For example, in our initial univariate analysis, an association between BMI and infection was suspected. It would be erroneous to draw the conclusion that obesity is responsible for increased rates of infection as an increased BMI is associated with an altered incidence of other conditions, such as diabetes mellitus. Regression analysis allows us to separate out diabetes and BMI and test the effect of each on the incidence of infection. Other studies also found no increased infection.40,55 There was no correlation therefore for infection and obesity. The lack of multivariate analysis may explain why Namba et al.48 found a non-significant increase in infection rate in the obese hip patients.

Blood loss A paper comparing 41 obese and 125 non-obese patients55 noted increased blood loss in the obese group as did Bowditch and Villar56 in a series of 80 patients. Multiple regression analysis allows for the correction of other variables in assessing the individual influence of BMI. Even with the large numbers in our study, once other factors such as co-morbidity are taken into account, we did not find that BMI increased measured blood loss or transfusion requirement.54 It is likely that the comparison of obese and nonobese patients without factoring in confounding data oversimplifies the true state of affairs. In a separate study, in our unit, we performed a multivariate analysis on 1016 THR, looking at transfusion requirements. BMI did not affect the transfusion rate in these patients.57

Harris Hip and Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) Scores In our study BMI independently predicted a lower Harris Hip Score (HSS) at 6 and 18 months. Although its individual effect was significant statistically, the difference was small.54 Stickles et al.40 found no difference in the WOMAC scores between obese and non-obese patients.

Quality of life scores (QOL) Two studies used the Short Form 36 (SF36)40,54 and showed no difference in the outcomes between obese and non-obese patients. Chan and Villar58 used the Rosser Index Matrix to generate a QOL score on 176 prospective patients undergoing THR. Forty-nine percent were obese but was no difference in the QOL compared to the non-obese patients at 1 and 3 years post-operatively.

ARTICLE IN PRESS 220

Loosening and early failure In our study we saw no relationship between early failure of the THR and obesity54 confirmed by other studies.40,55

Summary Total joint replacement is established as a very successful operation in treating degenerative disease of the hip and knee when conservative measures fail to alleviate symptoms.59–62 As the prevalence of obesity is increasing in Europe and North America, the percentage of total joint replacements performed in obese patients is likely to increase dramatically,64,65 indeed published data suggests that over a third of all total hip and knee replacements are performed in obese patients.37,38,55,63 It is important to establish how results of total knee and hip replacement in such patients compare with the results obtained in nonobese patients. Present evidence suggests that the results of total knee replacement in obese (BMI430 kg/m2) patients are probably comparable to the results of the procedure in non-obese (BMIo30 kg/m2) patients, at least in the midterm, but larger, prospective studies are necessary to ascertain the influence of obesity on the long-term survivorship following total knee replacement. There is some evidence that in morbidly obese (BMI440 kg/m2) patients, the results of total knee replacement are consistently poor, but there is too little data on the effect of morbid obesity on the outcomes of THR. Finally, patients undergoing total hip or knee replacement who are obese often consider their disabling joint disease a cause of their increased weight. Two prospective studies showed that obese patients gained weight following total hip or knee replacement.66,67 Successful treatment of lower limb osteoarthritis does not lead to weight loss, and obesity is not just a result of inactivity.

References 1. Bray GA. Overweight is risking fate. Definition, classification, prevalence and risks. Ann NY Acad Sci 1987;9:14–29. 2. World Health Organisation. Obesity: preventing and managing the global epidemic. Report of a WHO consultation on obesity. World Health Organisation, Geneva, 1998. 3. National Institutes of Health, National Heart, Lung and Blood Institute. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults: the evidence report. Obes Res 1998;6:S51–210 4. Willett WC, Dietz WH, Colditz GA. Guidelines for healthy weight. NEJM 1999;341(6):427–34. 5. Metropolitan Life Insurance Company: Statistical Bulletin. New York: The Foundation, 1983. 6. www.publications.parliament.uk/pa/cm200304/cmselect/cmhealth/ 23/2302.htm 7. www.who/int/nut 8. Felson DT, Anderson JJ, Naimark A, Walker AM, Meenan RF. Obesity and knee osteoarthritis. The Framingham study. Ann Intern Med 1988;109:18–24. 9. Spector TD, Hart DJ, Doyle DV. Incidence and progression of osteoarthritis in women with unilateral knee disease in the general population: the effect of obesity. Ann Rheum Dis 1994;53:565–8.

A.K. Amin et al. 10. Manninen P, Riihima ¨ki H, Helio ¨vaara M, Ma ¨kela ¨ P. Overweight, gender and knee osteoarthritis. Int J Obes 1996;20:595–7. 11. Felson DT, Zhang Y, Hannan MT, Naimark A, Weissman B, Aliabadi P, et al. Risk factors for incident radiographic knee osteoarthritis in the elderly. Arthritis Rheum 1997;40:728–33. 12. Hart DJ, Doyle DV, Spector TD. Incidence and risk factors for radiographic knee osteoarthritis in middle-aged women. The Chingford Study. Arthritis Rheum 1999;42:17–24. 13. Oliveria SA, Felson DT, Cirillo PA, Reed JI, Walker AM. Body weight, body mass index, and incident symptomatic osteoarthritis of the hand, hip, and knee. Epidemiology 1999;10:161–6. 14. Davis NM, Ettinger WH, Neuhaus JM. The role of metabolic factors and blood pressure in the association of obesity with osteoarthritis of the knee. J Rheumatol 1988;15:1827–32. 15. Davis MA, Ettinger WH, Neuhaus JM. Obesity and osteoarthritis of the knee: evidence from the national health and nutrition examination survey (NHANES I). Semin Arthritis Rheum 1990; 20:34–41. 16. Bagge E, Bjelle A, Eden S, Svanborg A. Factors associated with radiographic osteoarthritis: results from the population study 70-year-old people in Go ¨teborg. J Rheumatol 1991;18:1218–22. 17. Hart DJ, Spector TD. The relationship of obesity, fat distribution and osteoarthritis in women in the general population: the Chingford study. J Rheumatol 1993;20:331–5. 18. Schouten JS, van den Ouweland FA, Valkenburg HA. A 12 year follow up study in the general population on prognostic factors of cartilage loss in osteoarthritis of the knee. Ann Rheum Dis 1992;51:932–7. 19. Sowers M, Hochberg M, Crabbe JP, Muhich A, Crutchfield M, Updike S. Association of bone mineral density and sex hormone levels with osteoarthritis of the hand and knee in premenopausal women. Am J Epidemiol 1996;143:38–47 [Ab]. 20. Huang C, Ross PD, Lydick E, Wasnich RD. Factors associated with joint pain among postmenopausal women. Int J Obes 1997; 21:349–54. 21. Cicuttini FM, Spector T, Baker J. Risk factors for osteoarthritis in the tibiofemoral and patellofemoral joints of the knee. J Rheumatol 1997;24:1164–7. 22. Felson DT, Zhang Y, Anthony JM, Naimark A, Anderson JJ. Weight loss reduces the risk for symptomatic knee osteoarthritis in women. Ann Intern Med 1992;116:535–9. 23. Hartz AJ, Fischer ME, Bril G, et al. The association of obesity with joint pain and osteoarthritis in the HANES data. J Chronic Dis 1986;39:311–9. 24. van Saase JL, Vandenbroucke JP, van Romunde LK, Valkenburg HA. Osteoarthritis and obesity in the general population. A relationship calling for an explanation. J Rheumatol 1988;15: 1152–8. 25. Vingard E. Overweight predisposes to coxarthrosis. Body-mass index studied in 239 males with hip arthroplasty. Acta Orthop Scand 1991;62:106–9. 26. Tepper S, Hochberg MC. Factors associated with hip osteoarthritis: data from the First National Health and Nutrition Examination Survey (NHANES-I). Am J Epidemiol 1993;137: 1081–8. 27. Heliovaara M, Makela M, Impivaara O, et al. Association of overweight, trauma and workload with coxarthrosis. A health survey of 7,217 persons. Acta Orthop Scand 1993;64:513–8. 28. Roach KE, Persky V, Miles T, Budiman-Mak E. Biomechanical aspects of occupation and osteoarthritis of the hip: a case– control study. J Rheumatol 1994;21:2334–40. 29. Vingard E, Alfredsson L, Malchau H. Lifestyle factors and hip arthrosis. A case referent study of body mass index, smoking and hormone therapy in 503 Swedish women. Acta Orthop Scand 1997;68:216–20. 30. Cooper C, Inskip H, Croft P, et al. Individual risk factors for hip osteoarthritis: obesity, hip injury, and physical activity. Am J Epidemiol 1998;147:516–22.

ARTICLE IN PRESS Obesity and total knee and hip replacement 31. Flugsrud GB, Nordsletten L, Espehaug B, et al. Risk factors for total hip replacement due to primary osteoarthritis: a cohort study in 50,034 persons. Arthritis Rheum 2002;46:675–82. 32. Kellgren JH. Osteoarthritis in patients and populations. Br Med J 1961;2:5243–8. 33. Goldin RH, McAdam L, Louie JS, et al. Clinical and radiological survey of the incidence of osteoarthrosis among obese patients. Ann Rheum Dis 1976;35:349–53. 34. Lievense M, Bierma-Zeinstra SMA, Verhagen AP, van Baa ME, Verhaa JAN, Koes BW. Influence of obesity on the development of osteoarthritis of the hip: a systematic review. Rheumatology 2002;41:1155–62. 35. Karlson EW, Mandl LA, Gideon N, Aweh GN, Sangha, Liang, et al. Total hip replacement due to osteoarthritis: the importance of age, obesity, and other modifiable risk factors. Am J Med 2003;114:93–8. 36. Felson T, Zhang Y. An update on the epidemiology of knee and hip osteoarthritis with a view to prevention. Arthritis Rheum 1998;41:1343–55.Morrison JB. The mechanics of the knee joint in relation to normal walking. J Biomech 1970;3:51–61. 37. Spicer DD, Pomeroy DL, Badenhausen WE, Schaper Jr LA, Curry J, Suthers KE, et al. Body mass index as a predictor of outcome in total knee replacement. Int Orthop 2001;25(4):246–9. 38. Mont MA, Mathur SK, Krackow KA, Lowey JW, Hungerford DS. Cementless total knee arthroplasty in obese patients: a comparison with a matched control group. J Arthroplasty 1996; 11(2):153–6. 39. Griffin FM, Scuderi G, Insall J, Wayne C. Total knee arthroplasty in patients who were obese with 10-year follow up. Clin Orthop 1998;1(356):28–33. 40. Stickles B, Phillips L, Brox WT, Owens B, Lanzer W. Defining the relationship between obesity and total joint arthroplasty. Obes Res 2001;9:219–23. 41. Deshmukh RG, Hayes JH, Pinder IM. Does body weight influence outcome after total knee arthroplasty? A 1-year analysis. J Arthroplasty 2002;17(3):315–9. 42. Foran JRH, Mont MA, Etienne G, Jones LC, Hungerford DS. The outcome of total knee arthroplasty in obese patients. J Bone Joint Surg [Am] 2004;86-A:1609–15. 43. Foran JRH, Mont MA, Rajadhyaksha AD, Jones LC, Etienne G, Hungerford DS. Total knee arthroplasty in obese patients: a comparison with a matched control group. J Arthroplasty 2004; 19(7):817–24. 44. Amin A, Patton JT, Cook RE, Brenkel IJ. Does obesity influence outcome following total knee replacement for osteoarthritis? A five-year analysis. J Bone Joint Surg [Br], 2006; in press. 45. Dindo D, Muller MK, Weber M, Clavien PA. Obesity in general elective surgery. Lancet 2003;361:2032–5. 46. Shapiro M, Munroz A, Tager IB, Schoenbaum SC, Polk BF. Risk factors for infection at the operative site after abdominal of vaginal hysterectomy. N Engl J Med 1982;307:1661–6. 47. Brandt M, Harder K, Walluscheck KP, Schottler J, Rahimi A, Moller F, et al. Severe obesity does not adversely affect perioperative mortality and morbidity in coronary artery bypass surgery. Eur J Cardiothoracic Surg 2001;19:662–6.

221 48. Namba RS, Paxton L, Fithian DC, Stone ML. Obesity and perioperative morbidity in total hip and total knee patients. J Arthroplasty 2005;20:46050. 49. Insall JN, Lawrence DD, Richard SD, Scott WN. Rationale of the Knee Society Clinical Rating System. Clin Orthop 1989;248: 13–4. 50. Murray DW, Carr AJ, Bulstrode C. Survival analysis of joint replacements. J Bone Joint Surg [Br] 1993;75-B:697–704. 51. Winiarsky R, Barth P, Lotke P. Total knee arthroplasty in morbidly obese patients. J Bone Joint Surg [Am] 1998;80A:1770–4. 52. Blair D, Buskirk ER. Habitual daily energy expenditure and activity levels of lean and adult-onset and child-onset obese women. Am J Clin Nutr 1987;45:540–50. 53. Colquitt J, Clegg A, Sidhu M, Royle P. Surgery for morbid obesity. Cochrane Database Syst Rev 2003;(2):CD003641. 54. Moran M, Wlamsley P, Gray A, Brenkel IJ. Does body mass index affect early outcomes of primary total hip arthroplasty. J Arthroplasty 2005;20:866–9. 55. Soballe K, Christensen F, Luxhoj T. Hip replacement in obese patients. Acta Orthop Scand 1987;58(3):223–5. 56. Bowditch MG, Villar RN. Do obese patients bleed more? A prospective study of blood loss at total hip replacement. Ann R Coll Surg Engl 1999;81:198–200. 57. Aderinto J, Brenkel IJ. Pre-operative predictors of the requirement for blood transfusion following total hip replacement. J Bone Joint Surg 2004;86-B:970–3. 58. Chan CLH, Villar RN. Obesity and quality of life after primary total hip arthroplasty. J Bone Joint Surg 1996;78-B:78–81. 59. Ritter MA, Campbell E, Faris PM, Keating EM. Long term survival analysis of the posterior cruciate condylar total knee arthroplasty: a 10-year evaluation. J Arthroplasty 1989;4:293–6. 60. Font-Rodriquez DE, Scuderi GR, Insall JN. Survivorship of cemented total knee arthroplasty. Clin Orthop 1997;345: 79–86. 61. Ranawat CS, Boachie-Adjei O. Survivorship analysis and results of total condylar knee arthroplasty: eight to eleven year followup period. Clin Orthop 1988;226:6–13. 62. Gill GS, Joshi AB. Long term results of kinematic condylar knee replacement: an analysis of 404 knees. J Bone Joint Surg [Br] 2001;83-B(3):355–8. 63. Callaghan JJ, Albright JC, Goetz DD, Olejniczak JP, Johnston RC. Charnley total hip arthroplsty with cement. Minimum twenty-five-year follow-up. J Bone Joint Surg 2000;82-A: 487–97. 64. Bostman OM. Prevalence of obesity among patients admitted for elective orthopaedic surgery. Int J Obes Relat Metab Disord 1994;18(10):709–13. 65. Bjorntorp P. Obesity. Lancet 1997;350:423–6. 66. Heisel C, Silva M, dela Rosa MA, Schmalzried TP. The effects of lower-extremity joint replacement for arthritis on obesity. Orthopaedics 2005;28:157–9. 67. Aderinto J, Brenkel IJ, Chan P. Weight change following total hip replacement: a comparison of obese and non-obese patients. Surgeon 2005;3:269–72.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 222–233

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

ADULT TRAUMA

Management of proximal humeral fractures D.J.C. Burton, A.T. Watters Deptartment of Orthopaedics and Trauma, Bradford Royal Infirmary, Duckworth Lane, Bradford, UK

KEYWORDS Shoulder fractures; Rehabilitation; Internal fixation

Summary All aspects of proximal humeral fractures produce much debate. From classsification systems to modes of treatment and rehabilitation, the influence of patient age and management of complications all may lead to confusion. Here we aim to give a balanced, contemporary overview of the subject and provide suggestions as to how to approach these injuries. & 2006 Published by Elsevier Ltd.

Introduction Proximal humeral fractures are common (approximately 4–5% of all fractures) and increasing in frequency, probably due to their association with osteoporosis in our increasingly aged population.1 They represent over 70% of humeral fractures occurring over the age 40. Overall this injury tends to follow a bimodal age distribution, though they are less common in younger people in whom they are usually the result of higher-energy trauma. The exception to this is 2 part fractures, which follow a unimodal distribution in the elderly. Fractures of the shaft and distal humerus on the other hand, are more common than proximal humeral fracture in the young population. Management decisions for the majority of these injuries are straightforward, though some cases lead to vigorous debate regarding classification and treatment decisions can range from the most conservative to the most aggressive operative intervention. The availability of new, specific implants spurs some on toward operative intervention, whilst some texts still recommend treatment techniques Corresponding author. Tel.: +44 1937 574609.

E-mail address: [email protected] (D.J.C. Burton). 0268-0890/$ - see front matter & 2006 Published by Elsevier Ltd. doi:10.1016/j.cuor.2006.02.017

of questionable value, making diagnosis and treatment confusing. In this article we will consider the common classification systems, investigations, relevant anatomy and the more commonly proposed treatment regimes—together with the preferred management options of the senior author (ATW).

Classification Fracture classification systems are most useful when they allow accurate anatomical description, guide treatment and allow an estimation of prognosis. The first classification that seemed to meet these requirements was the Neer classification of 1970 2, although this was based on the anatomical observations of Codman as early as 1934.3 Despite marked intra-and inter-observer error,4 this remains the most commonly used classification system today. Codman described the proximal humerus divided into 4 fragments, along the lines of physeal union (Fig. 1). Thus there are head, greater tuberosity, lesser tuberosity and shaft fragments, each with specific soft tissue attachments. Neer developed this scheme by describing the fragments as parts, when they are displaced from the other fragments by

ARTICLE IN PRESS Management of proximal humeral fractures

223

Figure 1 Codman’s fracture fragments: green—articular fragment, red—greater tuberosity, blue—lesser tuberosity. (a) anterior proximal humerus and (b) posterior proximal humerus.

10 mm or 451 of rotation. Hence displaced fractures, or fracture dislocations, may be 2, 3 or 4 parts. Additionally there are a group of humeral head fractures that are described as impression fractures, usually occurring with a dislocation, in which the articular surface and subchondral bone becomes indented by the glenoid rim. Impression fractures are divided according to the percentage of the articular surface affected, generally less than 20%, 20–45% and greater than 45%. Head splitting fractures are a further development of this mode of injury, as the humeral head is cleaved into at least 2 separate fragments. Jakob and the AO group developed a classification system that describes the degree of vascular isolation of the articular fragment.5 Type A fractures are extracapsular, of 2 anatomical (‘Codman’) fragments and lead to no vascular isolation of the humeral head. Type B fractures are intracapsular, of 3 fragments and lead to partial articular vascular isolation. Type C fractures are the most severe, leading to complete vascular isolation and are usually in 4 fragments. Like all AO classifications, each type is subdivided and, although this is useful for research, it leads to a rather unwieldy system less suited for general clinical applications. Siebenrock et al. considered the AO system, like the Neer system, to be of insufficient reproducibility to allow comparison of studies using these classifications.6 Despite the reservations expressed, it is still important to use a classification system. Otherwise it becomes impossible to compare different papers on the subject and reach any conclusions regarding the validity, effectiveness and success

rates of the various treatment modalities described. The authors prefer to use the Neer system.

Vascular anatomy of the humeral head Critical to the prognosis in proximal humeral fracture is the blood supply to the humeral head articular fragment. This receives its main blood supply from the ascending branch (also known as the anterolateral branch) of the anterior circumflex humeral artery (itself a branch of the third part of the axillary artery). This artery enters the bone at the intertubercular groove and supplies the tuberosities and the humeral head via the intraosseous arcuate artery, which lies within the head. To a lesser extent blood supply also arrives via the soft tissue attachments of the rotator cuff, mainly through the posterior circumflex humeral artery, which supplies a small part of the greater tuberosity and posteroinferior head. Gerber et al. defined the territories of these arteries in a series of cadaveric radiopaque injection experiments, proving that the vascularisation of the whole humeral head was only possible via the anterolateral branch of the anterior circumflex humeral artery.7 Injuries that disrupt the blood supply to the articular fragment tend to be those that detach both the tuberosities and damage the arcuate artery, i.e. Neer 4 part or AO C-type fractures and fracture dislocations. Head splitting fractures also tend to critically disrupt the blood supply to the articular fragment. Impaired vascularity increases the

ARTICLE IN PRESS 224

D.J.C. Burton, A.T. Watters

Figure 2 Positioning for shoulder trauma series radiographs: (a) anteroposterior, (b) scapular lateral, (c) axillary, and (d) modified axillary (Velpeau). (With thanks to J Henning, Senior Radiographer, Bradford Royal Infirmary.)

risk of Avascular Necrosis (AVN) and collapse of the humeral head.

Diagnosis Following standard history taking and examination of the patient with a possible shoulder injury, including neurovascular assessment, full radiological examination is required. The complete shoulder trauma radiological series consists of an anteroposterior view, scapula lateral view and an axillary view (Fig. 2) (a Velpeau view may be required if the shoulder cannot be abducted sufficiently for the axillary view). Without these 3 views the fractured proximal humerus cannot be completely assessed. The axillary view is particularly important to assess head splitting fractures, visualise posterior displacement of the greater tuberosity and to assess the relationship between the articular surfaces. Unfortunately this view is often considered

unnecessary by Accident and Emergency and Radiology departments, despite good evidence in the literature to the contrary.8 Computed tomography (CT) is useful to further visualise complex fractures and to assess the humeral and glenoid articular surfaces in impression, head splitting and glenoid rim fractures. It is also useful in assessing bone stock in cases of nonunion and anatomy in malunion (Fig. 3).

Treatment methods Treatment methods will be described by fracture pattern. Debate becomes fierce as the complexity and number of parts increase; there is also patient age, physiological state and comorbidities to take into account along with expected demands to be put on the limb.

ARTICLE IN PRESS Management of proximal humeral fractures

225

Figure 3 Computed tomography in proximal humeral fracture: (a) 2 part proximal humeral fracture dislocation unstable on closed reduction of the dislocation, (b) instability in (a) due to boney Bankart lesion of glenoid seen on CT and (c) 4 part fracture dislocation in a young patient, assessed prior to attempted open reduction and internal fixation.

Undisplaced and minimally displaced fractures There is general agreement that fractures that are not sufficiently displaced or angulated to represent a part by Neer’s description may be treated conservatively. This is done in a broad arm sling under clothes or with a body swathe. The ‘Polysling’ is an example of a rather more comfortable form of support made of synthetic material and is ideal. The shoulder is immobilised for 10–14 days, although elbow and wrist exercises may commence immediately to prevent stiffness. An interim radiograph is obtained to ensure no deterioration in position. At 10–14 days the patient may start gentle range of movement (ROM) passive exercise with the help of a physiotherapist. The exception to this situation arises when there is over 5 mm of displacement of the greater tuberosity, which should be treated in the same way as 2 part greater tuberosity fractures. These fractures are best seen on the axillary radiograph. The greater tuberosity fragment usually migrates proximally and posteriorly under the influence of supra- and infraspinatus. In cases of fracture dislocation the

closed reduction of the shoulder often leads to perfect reduction of the greater tuberosity, though this must be reimaged at 1 week to rule out late displacement. Any more than 5 mm of displacement may lead to impingement and the fragment should be reduced and fixed when fresh, as later osteotomy for malunion and impingement is fraught with difficulty (Fig. 4). The approach for fixation of greater tuberosity fragments utilises a deltoid split. With the arm draped free the humeral head may be rotated to visualise the fragment and its bed. If there is a single, large fragment then 3.5 mm cancellous screws may be used, carefully countersunk to avoid impingement. If the fragment is not amenable to screw fixation (due to comminution or poor quality bone), a suture technique is used, using a figure of eight pattern through the rotator cuff tendon and then through a drill hole in the shaft. The fragment is reattached and the surrounding cuff insertion repaired. A No. 2 Ethibond (Ethicon, USA) or Fiberwire (Arthrex, USA) is appropriate in this situation. Neer regime rehabilitation may start 2–3 days postoperatively.9

ARTICLE IN PRESS 226

D.J.C. Burton, A.T. Watters

Figure 4 (a) Scapular lateral radiograph in anterior dislocation of the shoulder with greater tuberosity fracture and (b) after reduction of the shoulder, significant posterior displacement of the fragment remains.

Two part fractures This is the commonest fracture pattern seen. Two-part fractures of the proximal humerus may be impacted, angulated or displaced. Where there is impaction the fracture will usually unite promptly, but with associated angulation into varus, valgus or posterior bow deformity. Significant functional impairment may result because of impingement or malposition of the articular surface. Displaced or ‘off-ended’ fractures, where the action of opposing muscle groups pull the distal fragment medially and anteriorly, may be slow to unite and in some cases may progress to nonunion. There are several schools of thought regarding the management of these injuries, ranging from ‘almost always conservative’ to ‘almost always operate’. Manipulative reduction under anaesthetic on its own has been described in the past but due to the high incidence of redisplacement has been largely discredited. Similarly ‘traction reduction’ using a hanging plaster-of-Paris cast has also been discredited for failure to achieve reduction and for the discomfort that it causes. The authors believe that neither of these treatment options should be offered. Minimal fixation techniques following manipulative reduction are well described.10,11 The technique most commonly used is to pass three or four 2.5 mm threaded Kirschner wires across the fracture site after reduction. Pin tract infection (particularly when the wires are not buried) and

wire migration are common complications. Where the wires pass through the shoulder girdle muscles they also cause pain and interfere with movement and rehabilitation until they are removed. There are also many reported cases of inadequate reduction, possibly due to tendon interposition, or loss of position despite wire fixation. Though there remain some well-known proponents of the closed percutaneous technique, for the reasons outlined above many surgeons find it unsatisfactory. Open reduction and internal fixation allows visualisation and accurate reduction of the fracture fragments. Many methods of maintaining the reduction have been described, including the use of sutures, sutures with eyeletted nails, tension band wiring, intramedullary nails and various types of plates and screws. More recently custom-designed proximal humeral systems have been introduced, of which the authors have experience of the Plant Tan (Medizentechnic, Germany) and AO Philos (Synthes, Switzerland) systems with fixed angle interlocking screws (Fig. 5). Court-Brown et al. have reported the results of treatment of a large series of patients, average age 72 years, and concluded that the functional results of operative management in this older group are statistically no better than those treated conservatively.12 On this basis they advocate conservative management, with its lower complication rate. However, it should be noted that the internal fixation system that they used (Enders nails and sutures), is not ideal for achieving secure fixation in osteoporotic bone and they

ARTICLE IN PRESS Management of proximal humeral fractures

Figure 5

227

(a) PlantTan proximal humeral fixator plate and (b) AO Philos plate and locking screw.

concede that technical difficulties with achieving and maintaining reduction may contribute to the overall poor results of surgery. Reports in the literature looking at the PlantTan proximal humeral fixator plate in 2 and 3 part fractures have shown good functional results in the younger age group, though the results in the older age group remain poor.13,14 The theoretical advantages of the interlocking, multidirectional screw system in the AO Philos plate (better hold in osteoporotic bone, low profile therefore less likely to cause impingement) has yet to be confirmed in clinical trials, but offers hope for better functional results in the elderly population (Fig. 6). It is the senior author’s current practice to advocate open reduction and internal fixation using the Philos system 15 in most displaced-pattern 2 part surgical neck fracture, except in the very elderly and infirm and also in the more deformed (440 1) angulated pattern fractures occurring in younger patients. Initial results are promising, but review and analysis are awaited.

Three part fractures In this injury one tuberosity usually remains attached to the articular fragment and the rotator cuff then acts as a displacement force on this fragment. In all cases reduction and fixation are advocated, allowing early mobilisation and rehabilitation (Fig. 7).

Resch (Austria) is perhaps the best known exponent of the percutaneous reduction of 3 and 4 part fractures, using elevators introduced through appropriately placed stab holes and subsequent fixation with cannulated screws.16 This avoids further damage to the vascularity of the humeral head during dissection. This technique is technically very demanding, particularly in 4 part fractures and is not common practice in the UK. Gerber (Switzerland) has employed similar techniques for reduction and fixation of some 3 part fractures, though the initial approach for other 3 part, and all 4 part fractures, was via a deltopectoral incision using a rasp introduced through the fracture lines for reduction.11 The senior author’s preferred method is similar to that for the internal fixation of the 2 part fracture, using the Philos plate and heavy sutures (Fibrewire for example) to reattach the tuberosity fragment either directly or by passing the suture through the rotator cuff tendon. The suture may then be passed through the holes in the plate or through drill holes in the humeral shaft. Care must be taken during dissection to not denude the proximal fragments of their soft tissue attachments as this increases the risk of AVN and nonunion with migration of the fracture fragments. In very osteoporotic bone and in cases of fracture dislocation in the elderly, with complete loss of soft tissue attachments, a primary shoulder hemiarthroplasty allows early mobilisation and produces a confortable shoulder.17 When done immediately, this is far less technically demanding and yields more satisfactory results than at an interval

ARTICLE IN PRESS 228

D.J.C. Burton, A.T. Watters

Figure 6

(a–c) Completely displaced 2 part fracture, treated by ORIF using the Philos system.

ARTICLE IN PRESS Management of proximal humeral fractures

229 screw fixation with an average Constant score of 87% for 4 part fractures. This was in a relatively young population (average age 54 years), using a technique not widely practiced in the UK. The treatment of choice in the elderly population is generally accepted to be primary hemiarthroplasty. Although hemiarthroplasty had been attempted before the 1950s, it was Charles Neer who revolutionised the procedure for the treatment of proximal humeral fractures.18 Arthroplasty performed after an interval is made more difficult by softening of the bone, malunion of the tuberosities, contracture and intrarticular adhesions. The treatment of middle-aged patients remains something of a grey area and decisions must be made on an individual basis after discussion with the patient. The decision to perform hemiarthroplasty in young, active patients is very difficult as it will almost inevitably lead, with the passage of time, to a requirement for revision surgery. The levels of discomfort and function from a nonweight-bearing joint makes the sequelae of AVN of the humeral head and femoral head noncomparable, making a trial of fixation of the 4 part fracture in young people a viable option in the first instance. Satisfactory levels of function and comfort have recently been reported in young patients sustaining complex proximal humeral fractures complicated by AVN, having had restoration of anatomy using a deltopectoral approach and a combination of plate, screw and transosseous suture.11 If humeral head replacement is performed the deltopectoral approach is preferred. The key to the deranged anatomy is identification of the bicipital groove and the

Figure 7 (a and b) Three part fracture proximal humerus.

following failed conservative or operative management (Fig. 8).

Four part fractures This pattern of injury is not amenable to closed reduction. There is also a high failure rate with open reduction and internal fixation due to avascular necrosis and nonunion. Four part fracture dislocation is associated with an even higher risk, the head fragment is usually found to be completely free and devoid of soft tissue at operation (Fig. 9). In the series from Austria alluded to above,16 there was an 11% AVN rate using percutaneous reduction and cannulated

Figure 8 Shoulder hemiarthroplasty for trauma.

ARTICLE IN PRESS 230

D.J.C. Burton, A.T. Watters amongst inexperienced surgeons to shorten the humerus. Finally, the tuberosities should lie below the level of the humeral head to ensure correct cuff tension and prevent impingement. The long head of biceps tendon should be replaced in the bicipital groove at the end of the procedure or tenodesed to the floor of the bicipital groove if it has become detached from the glenoid. The functional results of hemiarthroplasty are variable and the patient must be counselled that some stiffness and loss of range of movement are inevitable, although a comfortable shoulder is the primary goal. Functional result is influenced by age of the patient, correct positioning of the implant and the anatomical reduction of the tuberosities.20

Impression and head splitting fractures

Figure 9 Four part fracture dislocation in an elderly patient.

long head of biceps tendon, lying between the two tuberosities. It is usually unnecessary to divide the subscapularis tendon as the split that commonly occurs between the two tuberosities allows entry into the shoulder joint and retrieval of the humeral head fragment. Care must be taken when retrieving the humeral head in cases of fracture dislocation, particularly if there has been delay between injury and presentation, as it may lie very close to the brachial plexus and axillary sheath. A number of humeral head replacements are available and are broadly grouped into modular or monobloc. The original Neer implant of the 1950s was modified in the 1970s and has remained in its monobloc form ever since. The senior author currently uses the Bigliani-Flatow (Zimmer, USA) modular implant, which comes with interchangeable head sizes (both diameter and depth) and a range of lengths and diameters. Other implants have increased modularity (separate neck for example) and all vary in the number and distribution of holes in fins for reattachment of soft tissues. With increasing modularity comes increasing risk of dissociation 19 and theoretically increased risk of fret corrosion between the components. The key to a successful hemiarthroplasty for trauma is the avoidance of injury to the deltoid muscle, correct retroversion of the prosthesis (usually about 301) and secure reattachment of the tuberosities to each other, the humeral shaft and the prosthesis itself to ensure healing in the correct position.17 Correct soft tissue tension is vital and the height of the prosthesis must therefore replicate the height of the native humeral head closely, as there is a tendency

Impression fractures are sustained following dislocation of the shoulder. In general terms articular defects ofo20% may be treated conservatively in a shoulder immobiliser or sling after joint reduction. Fractures involving 20–40% of the articular surface may require a bone grafting procedure, either by transfer of a tuberosity into the defect (the McLaughlin procedure for posterior dislocation21), or by iliac crest grafting. Fractures of more than 40% of the head surface require hemiarthroplasty as they tend to be grossly unstable as the defect fails to engage the glenoid with any congruency. Head splitting fractures cleave the head into at least two avascular fragments. Primary hemiarthroplasty is the procedure of choice, however this is a difficult decision to make in the young patient. In this case an attempt at anatomical reduction and fixation may be made after forewarning the patient of the risks and benefits of each option.

Pathological fractures Pathological fractures involving the proximal humerus may be treated by intramedullary fixation if the prognosis for the malignant disease warrants stabilisation. The Polarus nail (Acumed, USA) range allows numerous possibilities for locking in the proximal fragment and may be augmented with PMMA cement for greater stability. Where there is extensive replacement of the humeral head by tumour then a long stem humeral hemiarthroplasty with proximal and distal locking options may be used. In many cases nonoperative treatment is most appropriate (Fig. 10).

Rehabilitation The rehabilitation of proximal humeral fractures is critical to optimise outcome. The sooner that rehabilitation can commence, the better. This is usually more successful under the supervision of a physiotherapist. The classical Neer regime involves 3 phases.22 Phase 1: Passive-assisted exercises, including initial pendulums in both side-to-side and small circle movements, with the patient bending over slightly to allow the limb to swing. This progresses to assisted forward elevation and assisted external rotation in the supine position. Pulleys attached to the ceiling may be used, with the opposite hand

ARTICLE IN PRESS Management of proximal humeral fractures

231

Figure 10 (a and b) Pathological fracture secondary to breast carcinoma metastasis. Stabilised with the Polarus intramedullary nail with PMMA cement augmentation.

as the motor to provide the assistance. The use of a ceiling pulley is not recommended in the presence of tuberosity fractures until they have healed, at approximately 6 weeks, as their use may lead to displacement. Exercises are repeated 4 times per day after initial warmup using a warm pack on the shoulder. These exercises are ideally commenced immediately after stable fixation of fractures, though they may be deferred 2–3 weeks if there is concern regarding the strength of fixation or in the case of potentially unstable fractures that are to be treated conservatively (greater tuberosity fractures for example). Phase 2: Involves active, resisted and stretching exercises, usually between 6 and 12 weeks post injury. Active exercises involve forward elevation against gravity and the use of rubber ‘TheraBands’ to resist internal and external rotation, one end held in the affected hand and the other attached to a fixed point. Stretching may be done in a doorway, using the door jamb and top of door frame as fixed points against which to stretch. The unaffected limb is also useful to aid in stretching exercises, especially in internal rotation and by clasping the affected hand to move it above and behind the head, leading to abduction and external rotation of the shoulder. Phase 3: Increases resistance with the use of weights, starting at 1 lb and progressing up to a maximum of 5 lb. If the patient develops pain then the weight should be reduced. Phase 3 starts at about 3 months post-injury and continues until functional gain reaches a plateau or preinjury levels. Stretching is continued throughout phase 3.

Complications Neurovascular injury The neurovascular injury rate has been reported to be as high as 4–6% in some series of displaced proximal humeral fractures.23 Damage to the axillary artery is a recognised, though uncommon, complication of proximal humeral fracture. It typically occurs in elderly, atherosclerotic vessels and after high-energy trauma. An expanding haematoma around the shoulder, with a pale, pulseless, paraesthetic forearm, raises suspicion of vascular injury and indicates angiography. The consequences of vascular injury are necrosis of the limb distally, distal embolisation and permanent damage to the brachial plexus secondary to compression by the haematoma. Any part of the brachial plexus may be injured. More commonly an axillary nerve palsy is seen following anterior fracture dislocation, due to the position of the nerve on the anterior capsule. After careful documentation of the deficit, progress may be monitored with electromyography. Discussion with a centre managing brachial plexus injury is advised at an early stage, as some surgeons prefer early exploration of the brachial plexus rather than an expectant policy.

Chest injury Pneumothorax and haemothorax are both possible association with proximal humeral fracture, usually in high-energy

ARTICLE IN PRESS 232

D.J.C. Burton, A.T. Watters block to movement must be ruled out before embarking upon arthroscopic or open capsular release, in the event of failed physiotherapy and stretches for this condition. Closed manipulation is hazardous and may lead to refracture.

Avascular necrosis Avascular necrosis may occur very occasionally in 2 part fractures of the proximal humerus, however it is following closed treatment (3–14%) of 3 part fractures and open treatment of 3 and 4 part fractures (13–34%) that it is more common.25 A combination of primary traumatic insult to the vascularity of the humeral head followed by the secondary insult of manipulation and/or surgical dissection leads to necrosis. If the fracture was reduced anatomically with a congruent joint and well positioned tuberosities then a satisfactory functional result may be achieved despite the avascular necrosis. However, a more common picture is that of periarticular soft tissue contracture, degenerative change, metalwork migration and a stiff, painful joint. Treatment is by hemiarthroplasty of the shoulder. A procedure that is more demanding in the chronic situation due to contracture, abnormal anatomy and soft bone (Fig. 11).

Fracture nonunion

Figure 11 (a and b) Avascular necrosis with metalwork migration following ORIF proximal humeral fracture.

This is an uncommon complication caused by excessive movement at the fracture site (over aggressive physiotherapy for example), extensive periosteal stripping in highenergy trauma, wide separation of fragments and soft tissue interposition. The humeral shaft fragment may form a pseudarthrosis with the undersurface of the head fragment, complete with synovial membrane formation, as a result of excessive movement of the fracture site that is bathed in synovial fluid from the shoulder joint above. Where possible, treatment is with open debridement, reduction and fixation with bone grafting. This is only possible when the humeral head fragment has sufficient bone stock to allow fixation and the articular surface is in good condition. A CT scan may be useful in planning. Humeral head replacement is the procedure of choice where the criteria above are not met, although painless pseudarthrosis may be very satisfactory in some elderly patients with no functional demands.

Fracture malunion trauma. Intrathoracic dislocation of the humeral head fragment is an interesting though rare complication ! 24

Frozen shoulder Post-traumatic frozen shoulder is largely preventable with adequate rehabilitation and physiotherapy starting as early as possible post injury. Failure to progress with range of movement (including passive range), with the cardinal loss of external rotation, suggests frozen shoulder. Impingement of fracture fragments and metalwork causing a mechanical

Greater tuberosity migration into a superior position produces subacromial impingement, whilst posterior migration produces impingement with the glenoid. Where there is gross malunion an osteotomy may be performed, through a superior approach. This also allows correction of the position of the rotator cuff, although this may have become shortened with time, making reduction difficult. This procedure is often associated with unrewarding results and is best prevented by early recognition of greater tuberosity migration and managed in the acute situation. This can only be achieved with correct trauma series radiographs to identify the lesion.

ARTICLE IN PRESS Management of proximal humeral fractures Angular malunion of the surgical neck may be associated with a surprisingly good range of movement, providing there is no frozen shoulder and tuberosity impingement. It may be better to perform capsular release and institute intensive physiotherapy before embarking on an angular osteotomy of the proximal humerus. There are some deformities that warrant corrective osteotomy, these are usually gross and are more commonly due to anterior angulation at the fracture site leading to reduced forward flexion.

Acknowledgement The authors’ thanks go to Jessica Henning, radiographer at Bradford Royal Infirmary, for the line drawings.

References 1. Bengner U, Johnell O. Changes in the incidence of fracture of the upper end of the humerus during a 30 year period. Clin Orthop 1988;231:179–82. 2. Neer II CS. Displaced proximal humeral fractures, 1: classification and evaluation. J Bone Jt Surg [Am] 1970;52A:1077–89. 3. Codman EA. Rupture of the supraspinatus tendon and other lesions in or about the subacromial bursa. In: The shoulder. Boston: Thomas Todd; 1934. p. 262–93. 4. Sidor ML, Zuckerman JD, Lyon T. The Neer classification system for proximal humeral fractures: an assessment of interobserver reliability and intraobserver reproducibility. J Bone Jt Surg [Am] 1993;75A:1745–50. 5. Jakob RP, Kristiansen T, Mayo K. Classification and aspects of treatment of fractures of the proximal humerus. In: Bateman JE, Welsh RP, editors. Surgery of the shoulder. Philadelphia: BC Decker; 1984. 6. Siebenrock KA, Gerber C. The reproducibility of classification of fractures of the proximal end of the humerus. J Bone Jt Surgery [Am] 1993;75A:1751–5. 7. Gerber C, Schneeberger AG, Vinh TS. The arterial vascularisation of the humeral head. An anatomical study. J Bone Jt Surg [Am] 1990;72A:1486–94. 8. Flatow EL, Cuomo F, Mady MG. Open reduction and internal fixation of two part displaced fractures of the greater tuberosity of the proximal part of the humerus. J Bone Jt Surg [Am] 1991;73A:1213–8. 9. Neer II CS. Displaced proximal humerus fractures, 2: treatment of three part and four part displacement. J Bone Jt Surg [Am] 1970;52A:1090–103.

233 10. Jaberg H, Warner JJ, Jakob RP. Percutaneous fixation of unstable fractures of the humerus. J Bone Jt Surg [Am] 1992;74A:508–15. 11. Gerber C, Werner CML, Vienne P. Internal fixation of complex fractures of the proximal humerus. J Bone Jt Surg [Br] 2004;86B:848–55. 12. Court-Brown CM, Garg A, McQueen MM. The translated two-part fracture of the proximal humerus: epidemiology and outcome in the older patient. J Bone Jt Surgery [Br] 2001;83B:799–804. 13. Sadowski C, Riand N, Stern R, Hoffmeyer P. Fixation of the proximal humerus with the PlantTan humerus fixator plate:early experience with a new implant. J Shoulder Elbow Surg 2003;12: 148–51. 14. Burton DJC, Wells G, Watters AT, Schilders E. Early experience with the PlantTan fixator plate for 2 and 3 part fractures of the proximal humerus. Injury 2005;36(10):1190–6. 15. Fankhauser F, Boldin C, Schippinger G, Haunschmid C, Szyszkowitz R. A new locking plate for unstable fractures of the proximal humerus. Clin Orthop 2005;430:176–81. 16. Resch H, Povacz P, Frohlich R, Wambacher M. Percutaneous fixation of 3 and 4 part fractures of the proximal humerus. J Bone and Jt Surgery [Br] 1997;79B:295–300. 17. Kralinger F, Schwaiger R, Wambacher M, Farrell E, Lajtai G, Menth-Chiari W, Hubner C, Resch H. Outcome after hemiarthroplasty for fracture of the head of the humerus. J Bone and Jt Surgery [Br] 2004;86B:217–9. 18. Neer CS, Brown TH, McLaughlin HL. Fracture of the neck of the humerus with dislocation of the head fragment. Am J Surg 1953;85:252–8. 19. Blevins FT, Warren RF, Deng X, Torzilli PA. Dissociation of humeral shoulder arthroplasty components. J Shoulder Elbow Surg 1988;3(suppl):575. 20. Boileau P, Coste JS, Ahrens PM, Staccini P. Prosthetic shoulder replacement for fracture:Results of the multicentre study. In: Walch G, Boileau P, Mole D, editors. 2000 shoulder prostheses: two to ten year follow up. Sauramps medical, 2002. p.561–73. 21. Hawkins RJ, Neer CS, Pianta RM, Mendoza FX. Locked posterior dislocation of the shoulder. J Bone Jt Surg [Am] 1987;69A:9–18. 22. Hughes M, Neer CS. Glenohumeral joint replacement and rehabilitation. Phys Ther 1975;55:850–8. 23. Stableforth PG. Four part fractures of the neck of the humerus. J Bone Jt Surg 1984 [Br];66B:104–8. 24. Hardcastle PH, Fisher TR. Intrathoracic displacement of the humeral head with fracture of the surgical neck. Injury 1981;12:313–5. 25. Hagg O, Lundberg B. Aspects of prognostic factors in comminuted and dislocated proximal humeral fractures. In: Bateman JE, Welsh RP, editors. Surgery of the shoulder. Philadelphia; BC Decker; 1984. p. 51–59.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 234–236

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

CME SECTION CME Questions based on the Mini Symposium on Revision Hip Arthroplasty The following series of questions are based on the Mini Symposium on Revision hip arthroplasty. Please read the articles in the mini symposium carefully and then complete the self assessment questionnaire by filling in the square corresponding to your response to each multiple-choice question. For true or false questions, please fill in one square only. After completing the questionnaire, either post or fax the answer page back to the Current Orthopaedics Editorial Office at the address at the bottom of the RESPONSE sheet. Please photocopy this page if you wish to keep your copy of Current Orthopaedics intact. Replies received before the next issue of Current Orthopaedics is published will be marked and those reaching an adequate standard will qualify for three external CME points. You will be notified of your marks and a CME certificate will be dispatched for your records.

Questions 1. What action is recommended during acetabular revision surgery if there is no proprietary tool available to remove the polyethylene liner from its metal shell: (a) remove the assembly as one piece by disrupting the bone cement interface, (b) chisel the polyethylene liner to split it into segments that can easily be removed, (c) drill a 4.5 mm drill hole in the polyethylene liner and tighten a 6.5 mm screw through the hole, (d) divide the polyethylene liner with a heated knife, (e) crack the metal shell using an osteotome allowing easy removal of the loosened liner. 2. What is the approximate 12–15 year survivorship reported for uncemented revision acetabular components: (a) (b) (c) (d) (e)

495%, 80–95%, 65–80%, 50–65%, o50%.

(d) superolateral rim loss, 42 cm superior bone loss but intact teardrop and inferomedial wall, (e) posterior column deficiency with destruction of the inferomedial wall. 4. What alignment of acetabular components should be achieved at revision: (a) (b) (c) (d) (e)

451 451 601 601 601

lateral lateral lateral lateral lateral

opening, zero anteversion, opening, 201 anteversion, opening 201 retroversion, opening, zero anteversion, opening 201 anteversion.

5. If an uncemented acetabular cup is to be inserted after reconstruction of an acetabular defect using a ‘number 7’ graft, what proportion of coverage must be achieved by host bone rather than graft if a reconstruction ring is not used: (a) (b) (c) (d) (e)

25%, 50%, 75%, 90%, complete.

3. What characterises Paprosky type 3A acetbular defects:

6. Which of the following is a contraindication to impaction grafting:

(a) intact acetabular rim but loss of the dome, (b) defect of the superolateral rim of the acetabulum, (c) intact rim but loss of the medial wall,

(a) unavailability of blood group matched allograft, (b) patient who is a smoker, (c) haemophiliac patient,

0268-0890/$ - see front matter doi:10.1016/j.cuor.2006.06.002

ARTICLE IN PRESS CME SECTION (d) previous irradiation of the pelvis, (e) patient aged over 70. 7. According to published survivorship figures, what can a patient be told is the chance of a cemented revision socket surviving 20 years: (a) (b) (c) (d) (e)

25%, 45%, 65%, 75%, 85%.

8. What unusual complication has been described as occurring several months after using trabecular tantalum shells to reconstruct difficult acetabular defects at revision hip replacement: (a) (b) (c) (d) (e)

rapid onset aseptic osteolysis, fracture of the cup liner, intrapelvic screw migration, heterotopic ossification, transverse acetabular fracture.

9. What action should be taken if a diaphyseal cortical perforation occurs whilst preparing the canal for a cemented revision femoral component: (a) ensure the stem bypasses the perforation by 2 cortical diameters and occlude the perforation whilst cementing such that the cement sets flush with the cortical surface, (b) change to an uncemented, hydroxyapatite coated stem that bypasses the perforation by 2 cortical diameters, (c) proceed as normal, occluding the perforation during cement hardening, then apply a cortical strut graft secured with cerclage wires, (d) select a stem that bypasses the perforation by 2 cortical diameters and apply a cerclage wire around the shaft

235 distal to the perforation prior to cementation and introduction of the stem, (e) select a stem that extends to the distal femur and can be secured distally with locking screws. 10. What has happened to the crude revision rate for total hip arthroplasty over the past 20 years, as documented by the Swedish Hip Arthroplasty Register: (a) (b) (c) (d) (e)

it has remained constant, the rate has halved to approximately 7%, the rate has halved to approximately 14%, the rate has increased by half to approximately 7%, the rate has increased by half to approximately 14%.

11. According to Koo, what antibiotic combination should be used in the cement when the organism has not been identified by preoperative culture during 2 stage revision: (a) (b) (c) (d) (e)

vancomycin, gentamicin and cefotaxime, vancomycin and gentamicin, gentamicin alone, vancomycin alone, cefotaxime and gentamicin.

12. What is the usual course of symptoms when a patient with a hip arthrodesis and chronic back pain undergoes conversion of the arthrodesis to total hip replacement: (a) the back pain remains unchanged, (b) the back pain increases for about 3 months, before returning to preoperative levels, (c) the back pain is likely to become associated with sciatica because of stretching of the nerve during surgery and restoring movement in the vicinity of the nerve’s proximal course, (d) the back pain briefly improves but returns to preoperative levels in about 3 months, (e) the back pain dramatically improves.

ARTICLE IN PRESS 236

CME SECTION

Please fill in your answers to the CME questionnaire above in the response section provided below. A return address and fax number is given at the bottom of the page. ...............................................................................................

Responses Please shade in the square for the correct answer. 1 2 3 4 5 6 7 8 9 10 11 12

A A A A A A A A A A A A

& & & & & & & & & & & &

B B B B B B B B B B B B

& & & & & & & & & & & &

C C C C C C C C C C C C

& & & & & & & & & & & &

D D D D D D D D D D D D

& & & & & & & & & & & &

Your details (Print clearly) NAME. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ADDRESS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . FAX NO. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . EMAIL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . RETURN THE COMPLETED RESPONSE FORM by fax to +44-113-206-6791, or by post to CME, Current Orthopaedics, Orthopaedic Surgery, Clinical Sciences Building, St. James’s University Hospital, Beckett Street, Leeds LS9 7TF, UK.

E E E E E E E E E E E E

& & & & & & & & & & & &

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 237

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

CME SECTION Answers to CME questions in Vol. 20, issue 1 Please find below the answers to the Current Orthopaedics CME questions from Vol. 20, issue 1 which were based on the articles in the Mini Symposium on Biomechanics. 1

A&

B&

C’

D&

E&

2

A&

B&

C&

D’

E&

3

A&

B&

C’

D&

E&

4

A&

B&

C’

D&

E&

5

A&

B’

C&

D&

E&

6

A&

B&

C&

D&

E’

7

A&

B&

C&

D’

E&

8

A&

B&

C&

D&

E’

9

A’

B&

C&

D&

E&

10

A&

B&

C&

D’

E&

11

A&

B’

C&

D&

E&

12

A&

B&

C’

D&

E&

0268-0890/$ - see front matter & 2006 Published by Elsevier Ltd. doi:10.1016/j.cuor.2006.01.002

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 238–239

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

BOOK REVIEWS Edward V. Craig, Beth E. Shubin Stein, An Atlas of Orthopaedic Surgery: A Guide to Management and Practice, Taylor & Francis Group, London, New York, ISBN 1842141-856, 2004 (136 pp., h 70). The authors indicate that this book is designed to be a useful reference guide to common orthopaedic disorders for clinicians and Orthopaedic trainees. The atlas covers the body by regions in eight chapters. Each chapter is prefaced with a short dialogue on common disorders relevant to that region of the body. This dialogue covers aspects of anatomy, history taking, examination and diagnostic studies. The remainder of each chapter is dedicated to drawings, X-rays, CT scans and clinical photographs illustrating these common Jisorders. In a book of this size the detail relating to each disorder is minimal and used for illustrative purposes only. However sufficient detail is present to allow those with minimal orthopaedic knowledge to establish a diagnosis of the common ortho-

paedic problems they might encounter. There is not enough detail relating to the conditions covered for this book to be considered a reference text for orthopaedic trainees. I do not think that the Orthopaedic trainee will find this a useful book. The details required to practice Orthopaedics can be found elsewhere and in greater detail, as seen from the bibliography. Where I do think this book will find a niche is in Accident & Emergency departments. It would be a good reference book for the A&E officer to refer to when examining patients with musculoskeletal problems. The clinical pictures are very good and I believe will help doctors who have limited Orthopaedic experience to examine, diagnose and treat patients with musculoskeletal problems. It is a book that the A&E officer can dip into. It will not, I believe, become a standard orthopaedic text.

Robert Dunsmuir

doi:10.1016/j.cuor.2006.02.003

David S. Bradford, Thomas A. Zdeblick (Eds.), The Spine: Master techniques in Orthopaedic Surgery, second Ed., Lippincot Williams & Wilkins, ISBN: 07817 4078 9 (480 pp., $199.00). Roby Thompson in his preface to this series of books aims to create a series of operative manuals which not only describe the approaches but also the operative details. The volumes in this series are also enhanced by allowing the authors to insert some personal hints to help the surgeon. He has certainly achieved his aims with this edition for spinal surgery. A number of well known spinal surgeons have contributed chapters which cover all aspects of spinal surgery. They have produced a comprehensive manual of operative techniques for the spinal surgeon. The volume is set out in 25 chapters which cover the spine from the atlas to the sacrum. The book is usefully divided into three parts covering the cervical, thoracolumbar and lumbar spine. Each part is divided into easily read chapters which cover anterior and posterior approaches to the spine. The narrative is easy to read and supported with beautifully clear radiography, drawings, and intra-opertive photographs. The ‘‘operation’’ in each chapter is planned doi:10.1016/j.cuor.2006.02.002

with detail given to indications, contra-indications, preoperative planning, operative technique, complications and after care. Each chapter is supported with a useful selection of references for further reading. The authors of each chapter enhance the description by making alternate suggestions for the procedure based on their own experience. The whole book gives the reader a chance to learn the operative technique, as any good operative manual should, but also teaches where established surgeons find difficulties and how they resolve them. I have personally found this book useful in my own practice. I would certainly recommend it to any trainee in Orthopaedics or Neurosurgery who is contemplating a career in spinal surgery. The prose is easily read and digested. The operative technique is easily followed and the illustrations enhance the learning experience. I would strongly recommend this book to the trainee spinal surgeon. In addition it would be a useful addition to any departmental library where practising spinal surgeons work.

Robert Dunsmuir

Aims and Scope Current Orthopaedics presents a unique collection of international review articles summarizing the current state of knowledge and research in orthopaedics. Each issue focuses on a specific topic, discussed in depth in a mini-symposium; other articles cover the areas of basic science, medicine, children/adults, trauma, imaging and historical review. There is also an annotation, self-assessment questions and an exam section. In this way, the entire postgraduate syllabus will be covered in a 4-year cycle. The Journal is cited in: Cochrane Center, EMBASE/ Excerpta Medica, Infomed, Reference Update and UMI Microfilms.

Editor Professor R. A. Dickson MA, ChM, FRCS, DSc St James’s University Hospital Trust, Leeds, UK

Editorial Committee President of BOTA, M. A. Farquharson-Roberts (Gosport, UK), I. Leslie (Bristol, UK), D. Limb (Leeds, UK), M. Macnicol (Edinburgh, UK), I. McDermott (Ruislip, UK), J. Rankine (Leeds, UK)

Editorial Advisory Board L. de Almeida (Portugal) G. P. Songcharoen (Thailand) R. W. Bucholz (USA) J. W. Frymoyer (USA) R. W. Gaines (USA) S. L. Weinstein (USA) M. Bumbasirevic (former Yugoslavia)

A. K. Mukherjee (India) A. Kusakabe (Japan) A. Uchida (Japan) M.-S. Moon (Korea) R. Castelein (The Netherlands) R. K. Marti (The Netherlands) G. Hooper (New Zealand) A. Thurston (New Zealand) E. G. Pasion (Philippines)

D. C. Davidson (Australia) J. Harris (Australia) S. Nade (Australia) G. R. Velloso (Brazil) J. H. Wedge (Canada) S. Santavirta (Finland) P. N. Soucacos (Greece) M. Torrens (Greece) J. C. Y. Leong (Hong Kong)

Available online at www.sciencedirect.com

Amsterdam

K

Boston

K

Jena

K

London

K

New York

K

Oxford

K

Paris

K

Philadelphia

K

San Diego

K

St Louis

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 241–255

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: CHILDREN’S ORTHOPAEDIC SURGERY

(i) Indications for internal fixation of fractures in children I.H. Annan, M. Moran Royal Hospital for Sick Children, Sciennes Road, Edinburgh, EH9 1LF, Scotland, UK

KEYWORDS Internal fixation; Fractures; Children; Indications

Summary The nature of most fractures in children makes them suitable for closed management. However, functional outcomes are not uniformly good and there is evidence that poorer outcomes may be under-reported in children. The role for internal fixation is becoming better-defined. Most Salter Harris type I and II fractures are treated by closed methods while Salter Harris type III and IV fractures almost always require internal fixation. Metaphyseal fractures can be managed as Salter Harris II fractures, but are more likely to require some form of internal fixation. The management of long bone diaphyseal fractures is more controversial. The specific management depends on the bone that is fractured and generalisations regarding treatment are difficult to make. While closed management remains appropriate in most situations, the use of intramedullary fixation has been shown to be effective and safe in a variety of fractures. Care is necessary to avoid rotational malalignment which will restrict range of movement. Both external fixation and plate osteosynthesis have a valuable role in the management of some children’s fractures. & 2006 Elsevier Ltd. All rights reserved.

Introduction The purpose of this article is to review the current indications for the internal fixation of pelvic and long-bone fractures in children with particular reference to recently published studies. The management of the more common fractures has been included but for comprehensive coverage of fractures in children the reader is referred to the standard texts. Whilst in adults the place for operative Corresponding author. Tel.: +44 131 536 0000;

fax: +44 131 536 0852. E-mail address: [email protected] (I.H. Annan). 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.07.001

treatment of fractures has become well established, the view that operative treatment in children should be avoided wherever possible has been more persistent. The study of fractures in children is hampered because of the relative infrequency of many injuries and the observation that the skeletal age of the child often has a profound influence over both the management and final outcome. Useful comparisons of treatment need to include practical classifications of fractures and accurate comparisons between children of similar skeletal age. There is a need for better-controlled multicentre prospective studies. Although the vast majority of fractures in children continue to be managed non-operatively, internal fixation is increasingly being advocated in a range of situations

ARTICLE IN PRESS 242

I.H. Annan, M. Moran

(Table 1). Factors which influence this include rising patient and parental expectations, a desire to rehabilitate the child back into society as soon as possible and the increasing popularity of less invasive techniques such as flexible intramedullary nailing.

Upper limb Proximal humerus fractures Paediatric fractures involving the proximal humerus follow characteristic patterns depending on the child’s age. Young children and infants tend to present with Salter Harris type I Table 1 Suggested relative indications for internal fixation of paediatric fractures. Relative indications for internal fixation Intra-articular fracture with 42 mm displacement Conservative management socially unacceptable (shoulder/hip spica) Multiple trauma Floating joints Open fractures Inadequate closed reduction Older child Inability to maintain reduction Surgeon’s expertise

Figure 1

injuries with metaphyseal fractures becoming more common in school age children up to the age of 10. After the age of 10, Salter Harris type II fractures predominate. The periosteum is tougher posteriorly and tends to tear anteriorly. This may explain the typical fracture pattern with anterior displacement of the humeral shaft. Neer and Horowitz1 classified proximal humeral fractures according to the degree of displacement. Fractures with o5 mm displacement were classified as grade I, grade II included displacement of up to one third shaft diameter and grade III up to two thirds of the shaft diameter. Fractures with displacement of greater than two thirds of the humeral shaft diameter, including completely displaced fractures, were classed as grade IV. Neer grade I and grade II fractures, which make up the majority of proximal humerus fractures in children, usually have little angulation, are relatively stable because of the intact soft tissues and can be treated conservatively. Neer grade III and IV fractures are associated with higher energy injury with more extensive soft tissue damage. The distal humerus may button-hole through the periosteum or the shoulder joint capsule into the deltoid muscle and interposition of soft tissues can occur. Both of these factors may make reduction difficult and increase the risk of redisplacement. Nevertheless, the majority of these fractures can usually be reduced by closed manipulation (Fig. 1). In children under 7 or 8 years of age with isolated injuries, treatment in a hanging cast is usually satisfactory. Some loss of position is acceptable since healing in as much as 701 angulation is consistent with a good outcome. Because of the capacity of the proximal humerus to remodel operative

Type IV proximal humerus fracture that has been manipulated to a satisfactory position.

ARTICLE IN PRESS Internal fixation of fractures in children treatment in older children, up to age 12 in girls and 14 in boys, is rarely indicated. The correct management of severely displaced fractures in older children remains controversial. The extent of soft tissue damage renders these fractures unstable and while the majority can be satisfactorily reduced, the reported incidence of loss of reduction is up to 50%. The attachment of the rotator cuff muscles proximal to the fracture tends to cause the humeral head to rotate into flexion, abduction and some external rotation. Maintenance of reduction may be better in a cast with the shoulder abducted to 901 (‘‘Stop the Bus’’ shoulder spica cast). However, the application of this type of cast is technically demanding and may be unacceptable to the parents and child. The remodelling potential in older children is less and therefore a loss of position is less tolerable in adolescents. The great majority of authors have recommended conservative management of all proximal humeral fractures2; however, there are relatively few functional outcome studies and those that exist suggest poorer results in older children. Over the age of 12, severe displacement, multiple trauma and a requirement for an open reduction are all relative indications for internal fixation. Hence grade III or grade IV fractures in adolescents should be considered for closed reduction and internal fixation with Kirschner wires or flexible intramedullary nailing.3,4

Humeral diaphyseal fractures Fractures of the humeral shaft constitute a small proportion of children’s fractures (approximately 2%). Most are relatively undisplaced without severe soft tissue damage and can be managed by non-operative means. Occasionally, children present with severely displaced fractures. This indicates substantial periosteal stripping and soft tissue tearing. In the absence of complicating factors most can still be treated by closed manipulation under general anaesthesia and application of a U slab or hanging cast. Since 80% of humeral growth occurs at the shoulder, the closer the fracture is to the shoulder and the younger the child, the greater the degree of angulation that can be accepted. Whilst angulation of 301 in a 10-year-old with a proximal humeral fracture will remodel well, the same angulation in the distal diaphysis of a 12-year-old girl is likely to result in a cosmetically unacceptable malunion. Although many humeral fractures will heal in some degree of internal rotation because of the position adopted by the arm when in a sling, this rarely appears to cause any functional problem. In common with other diaphyseal long-bone fractures, the indications for internal fixation include an inability to obtain or maintain a satisfactory reduction, multiple trauma, a floating elbow joint, open fractures and fractures associated with vascular injuries. The choice of fixation is primarily between flexible intramedullary nailing and plate osteosynthesis. This choice depends upon the indication for fixation, fracture configuration and local expertise. Plating remains a safe choice and has regained popularity over intramedullary nailing in the management of most adult diaphyseal fractures of the humerus. The use of flexible intramedullary nails is not a recent advance in humeral fractures. Brumback et al.5 described 63 adult fractures

243 managed with Rush pins or Enders nails in 1986. There is no evidence in the literature to guide the treating surgeon as to which technique is preferable in paediatric fractures. Whereas, radial nerve palsy associated with a closed humeral fracture (treated non-operatively) usually resolves spontaneously over a period of 8–12 weeks and can be observed, palsies associated with open fractures or presenting post-operatively require exploration.

Elbow: lateral condyle fractures Lateral mass fractures constitute up to one fifth of elbow fractures in children. The mechanism of injury is said to result from a fall on the outstretched hand resulting in either compression through the radio-capitellar joint or avulsion during a varus strain at the elbow. At surgical exploration of displaced fractures, extensive disruption of the lateral soft tissues is normally found. These fractures have a propensity for late displacement and delayed healing even when the initial displacement is slight. Rutherford6 classified this injury according to the degree of displacement: grade I including up to1 mm, grade II between 1and 2 mm and Grade III 42 mm of opening of the lateral cortex. This is a useful classification system as it guides treatment. Grade I fractures in children who have minimal swelling or bruising may be treated conservatively but will require weekly X-ray review until there is clear evidence of bone healing without displacement. Displaced fractures (Grade III) require an open reduction avoiding soft tissue stripping from the posterior aspect of the lateral condyle. Internal fixation can be achieved with two diverging K-wires or a screw. The management of minimally displaced fractures (Grade II) is controversial. Attempts to assess the risk of late displacement using further imaging such as MRI or contrast arthrography have been inconclusive. In our experience, the presence of an intact articular cartilage hinge does not prevent progressive lateral opening of these fractures, presumably caused by tension from the lateral wrist extensors. Closed reduction and percutaneous pinning of minimally displaced fractures has only a small risk of morbidity and avoids the problems associated with late treatment of displaced fractures which include delayed healing, non-union, cubitus valgus, tardy ulnar nerve palsy and loss of function. The accuracy of reduction can be confirmed by air or contrast arthrography.7

Elbow: medial epicondyle fractures Fractures of the medial epicondyle are generally avulsion injuries resulting from forced valgus applied to the extended elbow with the wrist dorsiflexed. As a result, they are associated with extensive soft tissue damage to the medial aspect of the elbow joint with resulting instability. They often become very swollen, and with prolonged immobilisation, recovery of function can be slow with persisting stiffness. The only absolute indication for fixation is incarceration of the fragment within the elbow joint, which may occur if there has been an associated dislocation of the elbow joint. Minimally displaced fractures (o5 mm) can be treated by cast immobilisation. It is debatable how much displacement is acceptable. If the epicondyle is displaced

ARTICLE IN PRESS 244 more than 5 mm, it is likely to heal with a pseudarthrosis. Whether this is clinically important is unknown. Farsetti et al.8 examined the long-term outcome of operative and non-operative treatment of fractures displaced 5–15 mm and found no difference in outcome. Our practice is to internally fix fractures displaced more than 5 mm using a single cannulated screw, to rehabilitate the patient more quickly and reduce the risk of elbow stiffness. There may also be a benefit in terms of grip strength.

Elbow: supracondylar fractures Extension supracondylar fractures account for nearly two thirds of elbow fractures seen in children. Over half are undisplaced, Gartland grade I. These are defined as having o1 mm of opening anteriorly with no comminution or deformity of either medial or lateral columns. These can be treated with confidence by simple immobilisation using a non-circumferential cast. While all Gartland grade III fractures require reduction and internal fixation, the literature is divided regarding the management of Gartland grade II fractures. These have a persisting posterior periosteal hinge but there may be significant comminution and displacement of one column, usually on the medial side. We have found that if the elbow is not grossly swollen, it is often possible to assess these fractures in the Accident and Emergency department. After simple analgesia the child is encouraged to extend the elbow sufficiently for the carrying angle to be assessed. If the carrying angle is similar to the normal side, the injury can be treated as for a grade I fracture. If there is clinical doubt or if the X-ray films suggest significant angulation and particularly medial comminution, then we prefer to perform an examination under anaesthesia. If the initial alignment is found to be satisfactory, the fracture is gently manipulated to achieve the best position and is subsequently treated as for a grade I fracture with a back-slab. If there is significant loss of the carrying angle or substantial hyperextension (centre of capitellum behind the anterior humeral line), we prefer to reduce and stabilise the fracture with two K-wires. There has been extensive review of the merits of laterally placed parallel wires compared with crossed lateral and medial K-wire insertion (Fig. 2). The use of crossed wires is associated with a small but significant incidence of iatrogenic ulnar nerve injury of up to 7%.9,10 Lee et al.11 found that crossed K-wires provided greater resistance to axial rotation. They also recommended that if two lateral Kwires are used, they should be divergent and not parallel. It is important to place the pins so that they cross the fracture as far apart as possible and equidistant from the respective medial and lateral extremities of the fracture. Rarely, there may be an intra-articular extension to the supracondylar fracture, which does not fit the Gartland classification easily. Abraham et al.12 has suggested that these fractures be classified separately as a new type IV and has recommended that fixation with one or two medial wires in addition to two parallel or divergent lateral wires is associated with improved stability and long-term functional outcome. Flexion supracondylar fractures are much less common. They are usually more difficult to manage conservatively,

I.H. Annan, M. Moran possibly because the intact soft tissues are anterior and cannot easily be used as a fulcrum to maintain reduction unless the arm is immobilised with the elbow fully extended. For this reason, we pin all angulated and displaced flexion supracondylar fractures.

Elbow: radial neck fractures Fractures of the radial neck make up approximately 10% of fractures of the elbow in children. They are caused by falls on the out-stretched arm with the forearm in supination and are the corollary of medial epicondylar avulsions. These are usually Salter Harris type I or II fractures in which the radial head is depressed and angulated in relation to the neck of the radius. There may have been a posterior dislocation of the elbow at the time of injury. Examination may demonstrate significant elbow joint instability, indicating possible medial soft tissue injury. These fractures are frequently associated with other injuries around the elbow such as fractures of the proximal ulna, and avulsions of the medial epicondyle. Restriction of forearm rotation is often apparent. Traditional teaching has been to accept up to 301 of angulation although it can be difficult to measure this accurately, particularly from standard X-ray films obtained in the emergency department.13 There is no agreement in the literature regarding the acceptable degree of displacement. Up to 501 of angulation may be expected to remodel in children under 5 years of age, but here is a risk of significant loss of forearm rotation in children over 10 years of age when the angulation is 301 or more. Various techniques for closed manipulation have been described in the literature. While open reduction has been associated with poor outcomes, this may reflect the severity of the original injury. More recently percutaneous pin reduction (with fixation only if the fracture remains unstable) has been popularised. We have found this technique to give good results in the majority of cases. However, open reduction is occasionally required and is preferred to significant residual displacement as judged by instability or restriction of forearm rotation during EUA. Care should be taken to avoid any additional damage to the periosteal attachment of the radial head and it is usually possible to steer the head back into position with minimal soft tissue surgery, where it is often remarkably stable. Internal fixation is achieved with a fine oblique K-wire. A protruding wire will prevent early mobilisation but localises surgical trauma to the site of injury. Alternatively a retrograde flexible intramedullary nail can be used to reduce and stabilise the fracture.14

Elbow: olecranon fractures The great majority of olecranon fractures in children have a largely intact periosteum and can be treated conservatively. Displacement of up to 2 mm is acceptable and associated with a good long-term outcome.15 In younger children with significant displacement, fixation with parallel longitudinal pins and tension band using a bio-absorbable suture such as PDS is satisfactory and allows easy removal of the metal without further surgery (Fig. 3). In older children the

ARTICLE IN PRESS Internal fixation of fractures in children

245

Figure 2 (a) Type III supracondylar fracture. (b) A provisional reduction was obtained in the Emergency department, as there was vascular compromise. (c,d) Following Brachial artery exploration, the fracture was fixed with two lateral K-Wires.

Figure 3

A displaced olecranon fracture with fracture of the radial head, treated by open reduction and pin fixation.

ARTICLE IN PRESS 246 fracture should be treated as for the adult equivalent with longitudinal K-wires and a wire tension band.

Forearm diaphyseal fractures The majority of forearm fractures in children present few therapeutic problems. The thick periosteum makes significant loss of rotation unlikely and the soft tissue hinge facilitates reduction and maintenance of alignment. Even completely displaced fractures in younger children may retain adequate soft tissues to allow accurate reduction and management in a cast with little risk of malunion or functional impairment. Careful radiological follow-up is needed to ensure acceptable alignment and rotation is maintained to union. There is little consensus in the literature regarding the degree of angulation or rotational malalignment that can be accepted. Children up to the age of 8 can be expected to correct up to 201 of angulation with growth but remodelling potential in children older than 10 years is unpredictable. Similarly, there is little agreement regarding the functional outcome in children with residual malalignment. Whereas some authors report satisfactory results despite residual angulation, it has been suggested that the true incidence of poor outcomes may be widely under-reported16 and other authors have stressed the risk of loss of rotation. The advice of Evans17 is that for complete fractures of the radius ‘‘there is nearly always a rotational deformity y and its correction is a dominant factor in the treatment’’. Tarr showed that for distal and middle third fractures loss of up to 18% of forearm rotation occurred with 101 of angular deformity. Additionally, rotational deformity of the radius produced a loss of rotation equal to the degree of the rotatory deformity.18 Since there exists little or no capacity for rotational remodelling, this suggests that even in young children accurate rotational alignment is important if normal function is to be maintained. There is therefore good evidence to support the recommendation that no more than 101 of angulation and minimal rotation in the diaphysis should be accepted, particularly in children over 10 years of age. Pronation of the forearm for the distal diaphyseal fracture, and supination of the proximal fracture helps to correct the underlying rotational deformity. Open reduction and internal fixation (ORIF) has been the mainstay of operative treatment for many years. More recently, intramedullary fixation of forearm diaphyseal fractures has been popularised.19 Lascombes recommended fixation of both bones using pre-bent nails. However, other authors have suggested that the use of a single nail may be sufficient if, following reduction of both bones and pinning of one, the second bone is stable.20 Yung et al.21 reported a technique of fixation of the radius using an intramedullary K-wire and then assessment of the stability of the ulna fracture. In his series, nearly 50% of ulna fractures were not nailed and this did not affect the outcome. While intramedullary fixation can provide reliable maintenance of alignment, it does not of itself guarantee correct rotational alignment. Luhmann et al.22 reported that 30% of children treated with intra-

I.H. Annan, M. Moran medullary K-wires or Rush rods had an average loss of forearm rotation of 131 of supination and 91 of pronation. Other complications associated with the use of intramedullary nails include skin irritation, pin site infection and injury to extensor tendons. Careful attention to surgical technique is necessary to avoid these problems.21 In situations where closed nailing cannot be achieved and where there is doubt regarding residual rotational malalignment, open reduction and plate fixation remains the treatment of choice. Our current practice is to carry out a closed manipulation and assessment in theatre. If a satisfactory stable reduction is obtained and the forearm can be rotated through a full range of movement an above elbow cast is applied. If there is doubt about the reduction or the fracture is unstable, surgical stabilisation is indicated (Figs. 4 and 5). In the rare cases where closed reduction is not possible, surgical exposure frequently reveals interposed soft tissues and mal-rotation. Monteggia23 described a subset of forearm fractures involving dislocation of the radial head. Bado’s classification24 is based on the direction of dislocation of the radial head and is helpful in determining treatment.

Type Type Type Type

I II III IV

anterior posterior lateral anterior+fracture of the radius

Many Monteggia fractures can be treated by closed manipulation provided reduction of the radial head is confirmed. Weekly X-ray review is required until the ulnar fracture is consolidated. The radial head may be felt to reduce and become stable, presumably becoming fully relocated within an intact annular ligament in which case some instability or residual deformity of the ulna can be accepted. Where there is any doubt about the stability of the reduction of the radial head, the ulna fracture should be stabilised either with a plate or intramedullary fixation. In type IV fractures, stabilisation of the radius either by ORIF or occasionally by intramedullary nail is generally recommended converting the injury to a type I configuration.

Distal radial fractures Fractures of the distal radius fall into two common groups. Salter Harris II fractures are almost always inherently stable once reduced and almost never require internal fixation. If an open reduction is required, then internal fixation is warranted. Fractures of the distal radial metaphysis may be unstable after reduction, particularly if the reduction is not complete. Up to 20% of completely displaced metaphyseal fractures will redisplace in cast. The use of a percutaneous K-wire in these children is effective in preventing re-displacement.25 Percutaneous K-wiring may also be indicated if there is extensive soft tissue damage, in the presence of an ipsilateral proximal fracture and following remanipulation.

ARTICLE IN PRESS Internal fixation of fractures in children

247

Figure 4 A displaced fracture of the radius and ulna. Concerns about rotation meant this fracture was treated with open reduction and plate fixation.

Figure 5 stable.

Galeazzi fracture. The radius was reduced anatomically and the distal radial-ulna joint reduced spontaneously and was

Lower limb Pelvic and acetabular fractures It requires a significant force to fracture the pelvis or acetabulum of a child. In pelvic ring disruptions there is a high incidence of injury to other body systems. Severe blood loss from a child’s pelvic fracture is less common than in the adult. This may be due to the more elastic nature of the

blood vessels in a child or the greater ability of arteries to vasoconstrict after injury. It is usually the injury to the other organs in the body that accounts for deaths due to paediatric pelvic fractures.26 The emergency stabilisation of the pelvis using an external fixator is therefore less often indicated in a child. Whilst the deforming forces in paediatric pelvic fractures are similar to adults, the resulting anatomy of the fracture may be quite different. These differences are due to the

ARTICLE IN PRESS 248

I.H. Annan, M. Moran

Table 2 Key and Conwell classification of pelvic fractures. Group

Examples

1. No break in pelvic ring

Avulsion fractures (of apophysis) Iliac wing fracture Fracture of pubis or ischium Fracture of sacrum or coccyx

2. Single break in pelvic ring

Fracture of two ipsilateral pubic rami Symphysis pubis disruption Sacroiliac joint disruption

3. Double break in pelvic ring

‘‘Straddle’’ fracture Complex fractures

elasticity of the paediatric pelvis and joints and the presence of apophyses. The ‘‘double break’’ rule does not apply in children. There are a variety of classification systems developed for paediatric fractures. They variably include extra-pelvic pathology (e.g. visceral injury) and pelvic and acetabular fractures. Anatomically, pelvic fractures may be divided into three broad groups (Table 2). The indications for internal fixation in paediatric pelvic fractures are poorly defined. The majority of fractures can be managed non-operatively. Apophyseal avulsion fractures typically occur in the older sporty child. Most do well with early rest and then intensive physiotherapy. Fixation of large, widely displaced fractures in an athletic child should be considered. Adult pattern fractures, in older children with closed triradiate cartilages, are probably best managed along the same lines as in an adult. Complex and unstable fractures are likely to benefit from internal fixation.27 It is difficult to give specific guidelines as the spectrum of injuries is wide but the numbers small. Tile type B and C fractures28 with greater than 1.1 cm of pelvic asymmetry have a better outcome if internally fixed.29 Referral to a specialist centre is required and an individualised approach necessary. Paediatric acetabular fractures are rare when compared to pelvic fractures. Most commonly these are transverse fractures or posterior wall fractures associated with posterior hip dislocation. Displaced fractures of weight-bearing areas require reduction and fixation. The rare central fracture dislocation carries a poor prognosis, which may not be greatly improved by internal fixation.

Paediatric hip fractures Paediatric hip fractures are rare (o1% all childhood fractures) and are frequently associated with high-energy injuries. The Delbet classification is most commonly used,

with Type 2 (transcervical) and Type 3 (basicervical) accounting for over 90% of fractures. Type 1 (transepiphyseal) and Type 4 (intertrochanteric) fractures are extremely rare. Paediatric hip fractures have a high incidence of complications. Avascular Necrosis (AVN) has been reported in up to 50% of Type 2 fractures and is the leading cause of a poor outcome.30,31 Other common complications include physeal growth arrest, non-union, coxa vara, chondrolysis and late degenerative changes. The rationale for the use of internal fixation is to minimise these complications. Completely undisplaced hip fractures are uncommon. They have been managed successfully by non-operative methods (traction and/or hip spica).30,32 The use of internal fixation for undisplaced fractures is unproven. Displaced hip fractures treated with closed reduction and spica cast immobilisation may be appropriate for the very young child. In older children the early literature reported poor results for the management of hip fractures with or without internal fixation.33 It was thought that the outcome was determined at the time of injury by the degree of fracture displacement and age of the child. Both of these factors are important determinants of the final outcome; however, recent studies have reported reduced complication rates with the use of internal fixation.32,34 The improved results from these small series are likely to be multifactorial. Modern internal fixation devices are stronger and less likely to fail. Anatomic reduction is strongly associated with a reduced risk of AVN, coxa vara and non-union.30 Decompression of the intraarticular haematoma may further reduce the risk of AVN.32 The timing of surgery has not been the subject of a separate study; however, there is indirect evidence that surgery within 24 h of injury is desirable.34 It is impossible to state with certainty which of these factors is the most important. Internal fixation is the most reliable means of ensuring maintenance of an anatomic reduction and therefore should be associated with a reduced risk of AVN, coxa vara and nonunion. For displaced fractures the treatment should consist of early anatomic reduction (closed or open), decompression of the hip joint and stable internal fixation (Fig. 6). A hip spica is commonly used to immobilise the joint for 4–6 weeks post-operatively. Open reduction may be achieved by a Watson–Jones approach. Internal fixation is frequently with cannulated screws, although smooth Steinmann pins may be used in the young child (age under 3). Type 4 fractures will require a dynamic hip screw device of appropriate size. If possible, penetration of the physis by metalwork should be avoided (except in Type 1 fractures) as it increases the probability of early physeal closure. However, fracture stability is the primary concern and if necessary the physis may be crossed.

Femoral diaphyseal fractures Femoral shaft fractures are common, with the majority of fractures occurring in the diaphysis. The choice of treatment depends on the age of the patient, fracture configuration and location in the bone, soft tissue damage, associated injuries and local expertise. In a child under the age of 6, most surgeons would opt for non-operative management in most patients. In children up

ARTICLE IN PRESS Internal fixation of fractures in children

Figure 6 Displaced intracapsular hip fracture in a 13-year-old multiply injured child. This fracture was detected 3 weeks postinjury and required open reduction and screw fixation. It has healed at 6 months.

to 24 months, gallows traction may be utilised, depending on the weight of the child. From the age of two, in-line skin traction with the use of a Thomas splint is the preferred method. After an initial period of traction, a spica cast can be applied to allow the child to return home. However in a young patient, when healing is rapid and the time in traction is short this is not as useful.

249 Between the ages of 6 and skeletal maturity, there are a number of management options, all of which have their own merits and problems. In this age group, there is little literature to guide the treating surgeon as to which method is best. As a child gets older, non-operative management becomes more problematic. The potential of the bone to remodel a malunion reduces and the risk of significant shortening is increased. Healing time increases and so the time needed in traction is longer. The effectiveness of traction in obtaining and maintaining reduction is also reduced. The use of a spica cast is less acceptable. A prolonged break from schooling is often necessary. Dissatisfaction with non-operative methods of managing femoral shaft fractures in this age group has resulted in a trend towards the use of operative methods, in particular flexible nailing (Fig. 7). In a non-randomised comparison of flexible nailing and skeletal traction/spcia cast management all fractures healed primarily and function scores at 1 year were similar. However, out of 35 fractures treated in traction/spica, there were three malunions and in a further two cases loss of reduction required re-operation. Children undergoing flexible nailing had reduced hospital stay and time off school. The most common complication in the flexible nailing group was irritation from the nails at the insertion site.35 Certain factors make flexible nailing a less attractive option. These are very proximal or distal fractures, highly comminuted or long spiral fractures. Heavier children may bend the non-rigid nails and therefore a supplementary cast may be required until callus formation has occurred. External fixation is an effective method of treating femoral shaft fractures. It is particularly useful in open fractures, very comminuted fractures and fractures in the distal femur, where bone healing is more rapid. The external fixator construct is rigid and there are concerns that it is associated with a longer time to union and less callus production. Dynamisation of the external fixator may not overcome this problem. The two major complications that occur with external fixation are pin track infections and refracture after frame removal. The risk of refracture is between 5% and 20%. A recent study published in the Lancet randomised children to External Fixation or Hip spica. At 2 years there were significantly more malunions in the hip spica group, although patient and parent satisfaction were similar as were functional outcomes.36 There is a single study comparing flexible nailing and external fixation. Recovery and return to school were quicker in the flexible nailing group. Additionally, parent satisfaction was greater and there were fewer complications.37 ORIF using a 4.5 mm plate is useful in proximal and distal femoral fractures. In addition, if a vascular repair is carried out, plate fixation is the treatment of choice. Pulmonary complications are rare in children with isolated femur fractures. However, the presence of a severe head injury and a femoral fracture increases the risk of pulmonary problems considerably. There is little guidance on the method of operative fixation that is preferable in the head-injured child. Children with a head injury or other multiple injuries can be safely managed with plate osteosynthesis (or external fixation) once they are adequately resuscitated. The timing of surgery may be less critical than in adults.38 Plate fixation is the only operative

ARTICLE IN PRESS 250

I.H. Annan, M. Moran

Figure 7 Transverse diaphyseal fracture of the femur in a 12-year-old boy. Treated with ESIN.

method that does not necessarily require image intensifier screening intraoperatively. The large incision, increased blood loss, fixation failure, plate removal and risk of refracture mean that plate fixation has largely been superseded by other methods, except in the circumstances mentioned above. Minimally invasive plating methods may expand the indications for plate fixation in children. Rigid intramedullary nailing has been used for the treatment of femoral shaft fractures in the skeletally immature patient. Most series report a small but significant risk of AVN of the femoral head, even with a greater trochanteric entry point. This complication is so severe that most surgeons have abandoned this technique, particularly as there are other effective treatment methods available. The use of rigid cephalo-medullary nailing is the treatment of choice in skeletally mature adolescents. In common with other joints, ‘‘floating’’ knee joint injuries are best treated operatively. As a minimum, the femur should be operatively stabilised. Non-operative management is associated with an increased risk of leg length discrepancy, malunion and secondary operative procedures.39

Proximal tibial metaphyseal fractures Fractures of the proximal tibial metaphysis typically occur in children aged between 2 and 8. If displaced, the tibia is usually in valgus. Displaced fractures are managed by closed

manipulation and a long leg cast. If a closed reduction cannot be achieved, there may be intervening periosteum or the tendons of the pes anserinus can become interposed in the fracture. In this circumstance an open reduction is necessary. When accurate reduction is obtained it is rare for these fractures to be sufficiently unstable to require internal fixation. Even following a perfect reduction, however, a valgus deformity of the tibia can occur with growth of the child. The rate of progression of this is maximal in the first year and then diminishes. The natural history of the valgus deformity is spontaneous correction by years 3–5, so early surgery is not indicated40 (Fig. 8). In the rare situation that the valgus does not correct sufficiently, it can be managed by hemiepiphysiodesis in early adolescence, or by proximal tibial osteotomy.

Tibial diaphyseal fractures Tibial shaft fractures are common in children, occurring most often in the distal tibial diaphysis and metaphysis. In contrast to tibial fractures in adults, paediatric tibial fractures are more commonly undisplaced, the fibula is often intact, open fractures are less common and so is the incidence of compartment syndrome. The vast majority of these fractures can be managed with closed manipulation and a long leg cast. Acceptable limits

ARTICLE IN PRESS Internal fixation of fractures in children

Figure 8

251

Proximal metaphyseal fracture. Significant valgus resulted which partially corrected by 2 years post-fracture.

for reduction are 101 of angulation in the coronal and sagittal planes, shortening of up to 10 mm and no malrotation. The indications for internal fixation are: failure to achieve and maintain a reduction within the above limits, open tibial fractures, polytrauma, compartment syndrome and a ‘‘floating knee’’. The techniques available include external fixation, flexible nailing, plate osteosynthesis, percutaneous pinning and rigid intramedullary nailing. Because fixation of tibial fractures is uncommon, there is a paucity of literature on the subject. The pros and cons of the various methods are similar to their use in femoral shaft fractures. External fixation is historically the most common technique used to stabilise tibial fractures. The most common reason to stabilise the tibia surgically is an open fracture and so most series include a large proportion of open fractures. External fixation is an effective method of stabilising tibial shaft fractures, particularly those with

extensive soft tissue damage. There is a relationship between the incidence of complications and the severity of the open wound. The main disadvantages of external fixation are delayed or non-union (approximately 10%), pin track infection (25%), refracture and tibial overgrowth. There is a single retrospective comparison of external fixation and flexible nailing in the management of tibial fractures. Approximately half of the fractures were open. External fixation was associated with a much longer time to union (18 versus 7 weeks). There were three non-unions and two delayed unions in the external fixator group (out of 15 patients). The non-unions required conversion to a ring fixator and healed eventually. Unlike femoral shaft fractures, the functional outcome at 3 years was poorer in the external fixation group.41 On the basis of this study there is evidence that flexible nailing is superior to external fixation. Further studies have shown flexible nailing to be a safe technique in the management of tibial shaft fractures.42

ARTICLE IN PRESS 252 Union is usually achieved within 3 months and the incidence of refracture, fixation failure and malunion is low. Flexible nailing has been used successfully in open fractures of the tibia, although its role in Type III open fractures is not clear. The very occasional use of rigid intramedullary nailing of the tibia in the skeletally immature has been reported. However, the proximal tibial growth plate is violated and it is therefore not recommended. Plate osteosynthesis is a viable option for the management of displaced, closed fractures of the tibial diaphysis. An attempt should be made to achieve maximal soft tissue cover for the plate, so that it is protected from superficial wound infections. The principles of treating an open tibial fracture in children are similar to those in adults. However, it has been suggested that the initial wound excision in children may be less extensive. In particular, apparently devascularised bone has been seen to survive and new bone may form from retained periosteum.43 Whilst primary closure of an adult open wound is inadvisable, a child’s tissues may tolerate this more readily.44,45

Fractures around the ankle Paediatric ankle fractures are common as the ligamentous structures in the child are stronger than the open physis. Attempts have been made to classify paediatric ankle fractures using anatomical schemes (Salter Harris) and also systems that relate the mechanism of injury (e.g. Dias–Tachdjian modification46 of the Lauge–Hansen classification). The Salter Harris system is easily remembered, prognostic and gives some guidance to treatment. Systems describing the mechanism of injury provide a guide to

I.H. Annan, M. Moran reduction (reversal of the deforming forces); however, they have poor interobserver agreement and can be confusing. Fibular fractures Isolated fractures of the fibula are common and are usually Salter Harris I or II. These rarely require anything other than immobilisation in plaster. Displaced isolated Salter Harris III or IV fibula fractures are rare and can usually be managed non-operatively. Occasionally, the degree of displacement warrants fixation. Fibular fractures associated with a distal tibial fracture usually reduce and are stable with reduction of the tibial fracture and do not often require fixation. Tibial fractures Undisplaced distal tibial fractures can be managed in a cast with a period of restricted weight bearing. Most displaced tibial fractures are Salter Harris I or II and can be treated with closed manipulation and a cast. If there is any suspicion of an incomplete reduction, for example with significant gapping medially, then the fracture site should be opened (Fig. 9). As with proximal fractures, soft tissue may become trapped in the fracture site and this needs to be removed to minimise the risk of late growth disturbance. Most fractures are stable once reduced; however, if the fracture is unstable then internal fixation is warranted. A large metaphyseal fragment in a Salter Harris II fracture can be secured with a transverse interfragmentary screw parallel to the joint. If the metaphyseal fragment is too small, or in a Salter Harris I fracture, K-wires can be inserted across the physis but need to be removed when the fracture has united. Salter Harris III fractures fall into two main groups. The Tillaux fracture is a Salter Harris III fracture of the anterolateral portion of the distal tibial epiphysis, caused

Figure 9 Following an attempted closed reduction, the fibula required ORIF. A gap persists medially, which required open reduction to remove interposed soft tissues from the fracture site.

ARTICLE IN PRESS Internal fixation of fractures in children by forced external rotation. It typically occurs in children aged 12–14, when the distal tibial physis is beginning to close. The physis closes over a period of 18 months, starting centrally, then medial and finally laterally. It may be possible to reduce a Tillaux fracture percutaneously, using a K-wire to manipulate the displaced fragment into position. Arthroscopically assisted reduction has also been described. The second common Salter Harris III fracture involves the medial malleolus. In common with other articular fractures, a maximum displacement of 2 mm can be accepted. CT or MR imaging is useful in the accurate assessment of articular disruption in distal tibial fractures and the images frequently suggest that the extent of articular damage may be greater than is apparent on plain X-ray films. If displacement is greater than 2 mm, an anatomic reduction (open or closed) and internal fixation is carried out. The triplane fracture is a Salter Harris IV fracture that can occur in two, three or four parts. A detailed description is beyond the remit of this paper. The key to management is correction of rotation and restoration of the articular surface and this may be aided by cross-sectional imaging. Biplanar interfragmentary fixation is usually required. The second variation of a Salter Harris IV fracture is a shearing of the medial malleolus and metaphysis. There is evidence that residual displacement of 42 mm is associated with a poorer long-term outcome, with pain, stiffness and degenerative change. These adduction injuries should be reduced anatomically and stabilised with a transverse, intraepiphyseal screw. The risk of growth disturbance following physeal ankle fractures is greatest with Salter Harris IV and V fractures. The energy of injury, Salter Harris grade and quality of reduction are predictive of late growth problems at the ankle.

Discussion The immature skeleton is characterised by tough, metabolically active periosteum, the open physis, relatively thick cartilage of the epiphysis and the rapidly remodelling metaphysis. These characteristics are most prominent in infants and progressively diminish throughout childhood and adolescence and are responsible for the typical morphology and behaviour seen in children’s fractures. As a result most children’s fractures occur with little disruption of the periosteum and are relatively stable. Satisfactory conservative fracture management depends upon the ability to obtain and maintain a sufficiently accurate reduction. Maintenance of a reduction is influenced by the quality of bone, configuration of fracture and the presence or absence of an intact soft tissue hinge that can be utilised in a cast or by traction. Fractures in children heal rapidly, generally having satisfactory outcomes. This is in part because they rarely involve the articular surface and because of their potential to remodel and their more rapid rehabilitation. The pattern of fractures in children is different and it would be erroneous to apply uncritically lessons learnt from adult fracture management. This difference is most obviously seen in growth plate injuries that can carry long-term sequelae due to interruption in bone growth.

253 There is a paucity of quality evidence on which to base recommendations for the management of children’s fractures. The bulk of the literature consists of case series and retrospective comparisons between techniques, which are both open to bias. Studies frequently include children of widely different ages when this may be the major factor determining outcome. Many of the controversial issues will need to be subjected to randomised trials to determine the most suitable method of management. Fractures in children can undergo a degree of correction after healing. This remodelling occurs according to two principles. Wolff’s law results in alterations in the form of bone in response to mechanical stresses upon it; after fracture healing resorption of bone which is mechanically redundant may allow some restoration of the original contour. According to the Heuter–Volkmann principle, asymmetrical growth may occur at the physeal plate, with compression of the physis inhibiting growth and tensile forces stimulating growth. Traditionally, reliance has been placed upon the potential for remodelling in the management of paediatric fractures. However, the ability of the child to adapt to deformity may disguise poor functional outcomes. Recent long-term follow-up studies have suggested that function may not always fully recover, despite apparent correction of deformity on X-ray.21 Even temporary loss of function during a child’s formative years may have significant long-term implications. In deciding on the indications for internal fixation, a number of factors need to be taken into consideration. The soft tissue envelope around a fracture is all important in the closed management of paediatric fractures. An intact soft tissue hinge can be used as a brace to stabilise the fracture, with the aid of a cast. The more displaced a fracture, the greater this soft tissue hinge is disrupted and the increasing likelihood of the fracture being unstable or irreducible by closed means. For internal fixation to be useful, it must carry a demonstrable improvement in the end result of treatment. Additionally, the fixation must carry acceptable risks and minimal morbidity. Rigid cephalo-medullary nailing of paediatric femoral fractures is an excellent method of fracture stabilisation; however, the small risk of AVN has such serious consequences that most surgeons would consider it to be an unacceptable form of treatment in a skeletally immature patient. Salter Harris I and II fractures are usually reduced closed and only occasionally require internal fixation. Salter Harris III and IV fractures frequently require open reduction and almost always require internal fixation. Metaphyseal fractures can be managed as Salter Harris II fractures, but are more likely to be unstable once reduced and hence may require some form of fixation (commonly with K-wires). The management of long-bone diaphyseal fractures can be simplified by asking a number of questions. Can the fracture be reasonably managed non-operatively? If not, is the fracture amenable to treatment with flexible intramedullary devices? If the fracture cannot be nailed, then the choice is between external fixation and plate osteosynthesis. This choice depends upon fracture morphology, soft tissue injury and surgical preference. Finally it is important to ask, can I manage this fracture? If not, then referral to a specialist centre is indicated.

ARTICLE IN PRESS 254

I.H. Annan, M. Moran

Practice points

 For most diaphyseal fractures the rule of 10 applies: over age 10 accept under 101 of angulation

 IM pinning reliably maintains axial alignment. If rotation is uncertain consider ORIF

 Mechanism of injury determines the pattern of soft 

tissue damage, the morphology of the fracture and the technique for reduction Need for internal fixation predicted from fracture and soft tissue injury morphology

References 1. Neer 2nd CS, Horowitz BS. Fractures of the proximal humeral epiphyseal plate. Clin Orthop Rel Res 1965;41:24–31. 2. Larsen CF, Kiaer T, Lindquist S. Fractures of the proximal humerus in children: 9 year follow up of 64 unoperated cases. Acta Orthop Scand 1990;61:255–7. 3. Chee Y, Agorastides I, Garg N, Bass A, Bruce C. Treatment of severely displaced proximal humeral fractures in children with elastic stable intramedullary nailing. J Pediatr Orthop B 2006;15:45–50. 4. Dobbs MB, Luhmann SL, Gordon JE, Strecker WB, Schoenecker PL. Severely displaced proximal humeral epiphyseal fractres. J Pediatr Orthop 2003;23:208–15. 5. Brumback RJ, Bosse MJ, Poka A, Burgess AR. Intramedullary stabilisation of humeral shaft fractures in patients with multiple trauma. J Bone Joint Surg Am 1986;68A:960–70. 6. Rutherford A. Fractures of the lateral humeral condyle in children. J Bone Joint Surg 1985;57:851–6. 7. Milbrandt TA, Copley LAB. Common elbow injuries in children: evaluation, treatment and clinical outcomes. Curr Opin Orthop 2004;15:286–94. 8. Farsetti P, Potenza V, Caterini R, Ippolito E. Long-term results of treatment of fractures of the medial humeral epicondyle in children. J Bone Joint Surg Am 2001;83A:1299–305. 9. Solak S, Aydin E. Comparison of two percutaneous pinning methods for the treatment of the pediatric type III supracondylar humerus fractures. J Pediatr Orthop B 2003;12:346–9. 10. Skaggs DL, Hale JM, Bassett J, Kaminsky C, Kay RM, Tolo VT. Operative treatment of supracondylar fractures of the humerus in children. The consequences of pin placement. J Bone Joint Surg Am 2001;83-A:735–40. 11. Lee SS, Mahar AT, Miessen D, Newton PO. Displaced pediatric supracondylar fractures: biomechanical analysis of percutaneous pinning techniques. J Pediatr Orthop 2002;22:440–3. 12. Abraham E, Gordon A, Abdul-Hadi O. Management of supracondylar fractures of humerus with condylar involvement in children. J Pediatr Orthop 2005;25:709–16. 13. Radomisli TE, Rosen AL. Controversies regarding radial neck fractures in children, in Symposium: controversies in skeletal trauma. Clin Orthop 1998;353:30–9. 14. Metaizeau JP, Lascombes P, Lemelle JL, Finlayson D, Prevot J. Reduction and fixation of displaced radial neck fractures by closed intramedullary pinning. J Pediatr Orthop 1993;13:355–60. 15. Caterini R, Farsetti P, D’Arrigo C, Ippolito E. Fractures of the olecranon in children. Long-term follow-up of 39 cases. J Pediatr Orthop B 2002;11:320–8. 16. Kay S, Smith C, Oppenheim WL. Both bone midshaft forearm fractures in children. J Pediatr Orthop 1986;6:306–10. 17. Evans EM. Fractures of the radius and ulna. J Bone Joint Surg Br 1951;33B:548–61.

18. Tarr RR, Garfinkel AI, Sarmiento A. The effects of angular and rotational deformities of both bones of the forearm. An in vitro study. J Bone Joint Surg Am 1984;66A:65–70. 19. Lascombes P, Prevot J, Ligier JN, Metaizeau JP. Elastic intramedullary nailing in forearm shaft fractures in children: 85 cases. J Pediatr Orthop 1990;10:167–71. 20. Pugh DMW, Galpin RD, Carey TP. Intramedullary Steinman Pin fixation of forearm fractures in children: long term results. Clin Orthop 2000;376:39–48. 21. Yung PSH, Lam CY, Ng BKW, Lam TP, Cheng JCY. Percutaneous transepiphyseal intramedullary Kirschner wire pinning: a safe and effective procedure for treatment of displaced diaphyseal forearm fracture in children. J Pediatr Orthop 2004; 24:7–12. 22. Luhmann SJ, Gordon JE, Schoenecker PL. Intramedullary fixation of unstable both-bone forearm fractures in children. J Pediatr Orthop 1998;18:451–6. 23. Monteggia GB. Instituzioni chirurgiche. Maspero: Milan; 1814. 24. Bado JL. The Monteggia lesion. Clin Orthop Rel Res 1967;50: 71–86. 25. McLaughlan GJ, Cowan B, Annan IH, Robb JE. Management of completely displaced metaphyseal fractures of the distal radius in children. A prospective, randomised controlled trial. J Bone Joint Surg Br 2002;84B:413–7. 26. Grisoni N, Connor S, Marsh E, Thompson GH, Cooperman DR, Blakemore LC. Pelvic fractures in a pediatric level 1 trauma center. J Orthop Trauma 2002;16:458–63. 27. Karunakar MA, Goulet JA, Mueller KL, Bedi A, Le TT. Operative treatment of unstable pediatric pelvis and acetabular fractures. J Pediatr Orthop 2005;23:34–8. 28. Tile M. Pelvic ring fractures. Should they be fixed? J Bone Joint Surg Br 1988;70:1–12. 29. Smith W, Shurnas P, Morgan S, Agudelo J, Luszko G, Knox EC, et al. Clinical outcomes of unstable pelvic fractures in skeletally immature patients. J Bone Joint Surg Am 2005;87A: 2423–31. 30. Morsy HA. Complications of fracture of the neck of the femur in children. A long-term follow-up study. Injury 2001;32:45–51. 31. Canale ST, Bourland WL. Fracture of the neck and intertrochanteric region of the femur in children. J Bone Joint Surg Am 1977;59:431–43. 32. Cheng JCY, Tang N. Decompression and stable internal fixation of femoral neck fractures in children can affect the outcome. J Pediatric Orthop 1999;19:338–43. 33. Ratliff AH. Fractures of the neck of the femur in children. J Bone Joint Surg Br 1962;44B:528–42. 34. Flynn JM, Wong KL, Yeh GL, Meyer JS, Davidson RS. Displaced fractures of the hip in children. Management by early operation and immobilisation in a hip spica cast. J Bone Joint Surg Br 2002;84:108–12. 35. Flynn JM, Luedtke LM, Ganley TJ, Dawson J, Davidson RS, Dormans JP, et al. Comparison of titanium elastic nails with traction and a spica cast to treat femoral fractures in children. J Bone Joint Surg Am 2004;86A:770–7. 36. Wright JG, Wang EE, Owen JL, Stephens D, Graham HK, Hanlon M, et al. Treatments for paediatric femoral fracture: a randomised trial. Lancet 2005;365:1153–8. 37. Bar-On E, Sagiv S, Porat S. External fixation or flexible intramedullary nailing for femoral shaft fractures in children. A prospective randomised study. J Bone Joint Surg Br 1997;79B: 975–8. 38. Hedequist D, Starr AJ, Wilson P, Walker J. Early versus delayed stabilisation of pediatric fractures: analysis of 387 patients. J Orthop Trauma 1999;13:490–3. 39. Yue JJ, Churchill RS, Cooperman DR, Yasko AW, Wilber JH, Thompson GH. The floating knee in the pediatric patient. Nonoperative versus operative stabilization. Clin Orthop Rel Res 2000;376:124–36.

ARTICLE IN PRESS Internal fixation of fractures in children 40. Skak SV. Valgus deformity following proximal tibial metaphyseal fracture in children. Acta Orthop Scand 1982;53:141–7. 41. Kubiak EN, Egol KA, Scher D, Wasserman B, Feldman D, Koval KJ. Operative treatment of tibial fractures in children: are elastic stable intramedullary nails an improvement over external fixation? J Bone Joint Surg Am 2005;87A: 1761–8. 42. O’Brien T, Weisman DS, Ronchetti P, Piller CP, Maloney M. Flexible titanium nailing for the treatment of the unstable pediatric tibial fracture. J Pediatr Orthop 2004;24(6):601–9.

255 43. Bartlett CS, Weiner LS, Yang EC. Treatment of type II and type III open tibia fractures in children. J Orthop Trauma 1997;11: 357–62. 44. Cullen MC, Roy DR, Crawford AH, Assenmacher J, Levy MS, Wen D. Open fracture of the tibia in children. J Bone Joint Surg Am 1996;78A:1039–47. 45. Hope PG, Cole WG. Open fractures of the tibia in children. J Bone Joint Surg Br 1992;74:546–53. 46. Dias LS, Tachdjian MO. Physeal injuries of the ankle in children: classification. Clin Orthop Rel Res 1978;136:230–3.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 256–265

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: CHILDREN’S ORTHOPAEDIC SURGERY

(ii) Paediatric epiphyseal fractures around the knee M. Moran, M.F. Macnicol Royal Hospital for Sick Children, Sciennes Road, Edinburgh, EH9 1LF, Scotland, UK

KEYWORDS Paediatric; Knee; Fracture; Osteochondral injury

Summary Physeal fractures around the knee are most common in children aged 9–14. The majority of these fractures will have a good outcome if they are adequately treated initially. There is a fairly clear consensus on how these fractures should be managed. Undisplaced fractures are treated with cast immobilisation and almost universally have a good outcome. Displaced Salter Harris I and II fractures of the proximal tibia and distal femur can usually be treated with closed reduction and fixation. Intraarticular fractures often require open reduction prior to internal fixation. Displaced fractures of the tibial spine, tibial tuberosity and patella are more difficult to reduce closed and an open reduction is frequently required. Most osteochondral fractures are simply excised arthroscopically. There are well recognised complications associated with paediatric knee fractures. Early complications include popliteal artery damage, ligament damage and compartment syndrome. Late complications are usually related to damage to the physis with leg length discrepancies and angular deformity. & 2006 Elsevier Ltd. All rights reserved.

Introduction As the child ages the incidence of physeal fractures around the knee increases with a corresponding decrease in the incidence of metaphyseal fractures. Isolated ligament injuries also become more common in the teenager (Fig. 1). With growth, the epiphysis becomes less cartilaginous and the secondary ossification centre increases in size. This relative rigidity of the epiphysis reduces its ability to absorb energy and there is a tendency for applied forces to be transmitted to the physis. This may account for the shift in fracture distribution. Most physeal knee fractures therefore

occur over the age of 10. The distal femoral and proximal tibial growth plates are responsible for the majority of longitudinal growth of the lower limb (Table 1). Fractures affecting these physes are associated with a risk of late angulation and limb shortening. It is fortunate that the majority of these fractures do occur in the older child with less remaining growth potential. The treating surgeon should be aware of the risk of injury to the soft tissues surrounding the knee. Injury to the popliteal vessels, compartment syndrome and ligament damage are all seen in association with paediatric knee fractures.

Radiological investigations Corresponding author. Tel.: +44 131 536 0000;

fax: +44 131 536 0852. E-mail address: [email protected] (M. Moran). 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.03.002

In many circumstances plain anteroposterior and lateral radiographs are all that are required. A Merchant or skyline1

ARTICLE IN PRESS Paediatric epiphyseal fractures around the knee

257 Harris Type II.2 Fractures of the distal femoral epiphysis account for just 1% of all paediatric long bone fractures.3

Classification The Salter Harris classification is used to classify distal femoral fractures. Type II fractures are the most common, accounting for over 80% of distal femoral physeal fractures.3 Displacement usually occurs in the direction of the metaphyseal fragment. In type III fractures the split in the epiphysis is usually in the midline. Type V fractures are rare and often diagnosed late (Fig. 2).

Mechanism of injury Figure 1

Graph of distribution of knee injuries by age.

Table 1 Approximate relative contribution of lower limb growth plates to leg length.

Proximal femur Distal femur Proximal tibia Distal tibia

Growth/ year (mm)

Percent contribution to length bone

Percent contribution to leg length

3

30

15

9 6

70 66

40 30

3

33

15

The mechanism of injury is either hyperextension or varus/ valgus stress. Historically, hyperextension injuries were most common (Fig. 3). These were often caused by a child’s leg becoming caught in the moving wheel of a horse-drawn cart. These injuries were associated with a high risk of vascular complications and hence amputation. Varus/valgus injuries are now much more common, usually sustained during sport or in road traffic accidents. The risk of vascular

view of the patella is helpful in the diagnosis of medial and lateral patellar avulsions. Tunnel views can be used to visualise the intercondylar area, for example when looking for loose osteochondral fragments. Further imaging is useful if there is doubt about the diagnosis or to delineate complex fracture patterns prior to surgery. Tomograms and stress radiographs have largely been replaced by the use of MRI and CT scanning. If stress radiographs are performed care should be taken not to further displace the fracture fragments. In general, when trying to delineate a complex fracture pattern, CT is the most useful investigation. MRI is helpful in tibial spine and patellar fractures with a very small bone fragment, to examine the integrity of the attached ligament. MRI is also used in the diagnosis of undisplaced fractures, particularly Salter Harris I fractures. Ultrasound is occasionally helpful in the assessment of the integrity of the patella tendon in sleeve fractures with a small bony avulsion.

Distal femoral epiphysial fractures Aetiology and anatomy The distal femoral epiphysis contributes 40% to total limb length, growing at a rate of approximately 1 cm per year (Table 1). Nearly 75% of all epiphyseal fractures are Salter

Figure 2 Asymmetric growth arrest following Salter Harris Type V distal femur fracture.

ARTICLE IN PRESS 258

M. Moran, M.F. Macnicol

Figure 4 Salter Harris II fracture of the distal femur treated with closed reduction and crossed Kirschner wire fixation. The large metaphyseal fragment could alternatively have been internally fixed with transverse cannulated screws.

Figure 3 femur.

Grossly displaced physeal fracture of the distal

complications in these injuries is lower. Type I injuries can occasionally occur in the neonate, during delivery.

Presentation In displaced fractures pain and deformity are the presenting features. The femoral epiphysis is displaced along with the tibia, so the femoral and tibial condyles are in normal alignment with each other. Arterial injury is rare (1%) in varus/valgus injuries.

Treatment Undisplaced fractures can be managed in a long leg cast for 4–6 weeks. Displaced type I and II fractures may be treated by closed reduction. An assessment of stability can be made at the time of reduction. The reduction can then be held with a long leg cast, with or without internal fixation. In

larger and heavier children it is more difficult to maintain a reduction with a plaster alone with a risk of redisplacement if the fracture is not internally fixed. In one series 3 out of 7 fractures redisplaced after cast immobilization.4 Internal fixation is usually carried out with crossed smooth Kirschner wires, inserted from distal to proximal (Fig. 4). Type II fractures with a large metaphyseal fragment can alternatively be managed by transverse percutaneous cannulated screw fixation of the metaphyseal fragment to the rest of the femur. Irreducible Type I and II fractures and all displaced type III and IV fractures require open reduction and internal fixation with cannulated screws or pins.

Outcome The final outcome is related to the initial displacement of the fracture.4 The irregular femoral epiphysis provides resistance to shear. Because of this irregularity, when a fracture does occur it is more likely to be associated with damage to the physis. Salter Harris type I and II fractures of the distal femoral epiphysis are more commonly associated with growth arrest and subsequent angulatory deformity than these patterns of fracture at other locations. In type II fractures the physis underlying the metaphyseal fragment is

ARTICLE IN PRESS Paediatric epiphyseal fractures around the knee spared and the fracture angulates away from the side of the metaphyseal piece i.e. a medial metaphyseal fragment results in a valgus deformity. As the force required to fracture the epiphysis in a younger child is greater and therefore the energy imparted to the growth plate is more than in an adolescent fracture, it can therefore be expected that the initial damage to the physis is greater in a younger child. In addition the younger child has more potential for future growth therefore any damage to the growth plate resulting in shortening and/or angulation will be exaggerated. Angulatory problems that develop may be managed with hemiepiphysiodesis or osteotomy. The degree of predicted Leg Length Discrepancy (LLD) will influence treatment of shortening. A significant LLD (42 cm or requiring contralateral epiphysiodesis) has been variably reported in 10–50% of patients.4,5 This is usually related to physeal bar formation.4 If children are followed to skeletal maturity, problems with LLD and angulation can be detected and treated at an early stage.

Proximal tibial epiphysial fractures Aetiology and anatomy Proximal tibial physeal fractures have similarities to distal femoral fractures in terms of classification, mechanisms of injury and treatment. They are approximately 3 times less

259 common than fractures of the distal femoral epiphysis. This is thought to be due to protection of the physis arising from the sites of insertion of the collateral ligaments. Both medial and lateral collateral ligaments attach to the femoral epiphysis. However, the medial collateral ligament attaches to the tibial metaphysis and the lateral ligament is inserted into the fibula (Fig. 5). When a deforming force is applied to the knee the collateral ligaments will therefore transmit force to the femoral epiphysis but spare the tibial epiphysis. At the level of the proximal tibia the popliteal artery is relatively fixed in position, due to its numerous branches. It is therefore at significant risk of injury in fractures of the proximal tibia.

Classification The Salter Harris classification is most widely used for proximal tibial physeal fractures. Type II and IV fractures are the most common. Type II fractures are usually due to valgus stress and the metaphyseal fragment is therefore usually lateral, with a corresponding valgus deformity. Type III fractures are often associated with avulsion of the tibial tuberosity (described below).

Mechanism of injury Hyperextension or varus/valgus strains are the most common mechanism of injury. Fractures may also occur at birth or during vaginal breach delivery. Periarticular knee fractures have been reported following manipulation of a child’s knee.6

Presentation A tense haemarthrosis is usually present. Because of the shape of the proximal tibial physis, the tibial metaphysis displaces posteriorly (Fig. 6). This is particularly so in hyperextension injuries. There is a significant risk of injury to the popliteal neurovascular structures. A careful examination of the neurovascular structures in the lower limb must be made.

Treatment Treatment options are similar to those used for fractures of the distal femoral epiphysis. Non-displaced Salter and Harris type I to IV fractures can be treated by cast immobilisation for 4–6 weeks. Displaced type I and II fractures may be reduced closed and then stabilised using smooth Steinmann pins. A large metaphyseal fracture may lend itself to cannulated screw fixation. Displaced type III and IV fractures will need open reduction and fixation, as will type I and II fractures that are not reducible by closed means.

Outcome

Figure 5

Collateral ligament insertion around knee.

These fractures usually occur in adolescents and the prognosis is generally good. Type IV fractures are at a higher risk of developing problems with angulation or limb length

ARTICLE IN PRESS 260

M. Moran, M.F. Macnicol

Figure 6 Proximal tibial epiphyseal fracture. The metaphyseal fragment is displaced posteriorly into the popliteal fossa.

discrepancy than other fracture patterns. Burkhart and Peterson described a series of open proximal physeal tibial fractures. These were all caused by lawnmowers and occurred in younger children. They had a universally poor outcome with a high incidence of problems with limb growth and angulation.7 Arterial injury is a well-recognised complication of proximal tibial physeal fractures, occurring in up to 10% of patients.

Mechanism of injury Fracture of the tibial tuberosity typically affects teenagers involved in sport. There is either forced flexion of the knee or a contraction of the quadriceps, which pulls off the tuberosity.9

Presentation

Tibial tuberosity avulsions

The tibial tuberosity develops from the anterior portion of the proximal tibial epiphysis. From the age of 8 years the secondary ossification centre of the tuberosity begins to appear, distinct from the main secondary ossification centre of the proximal tibia (which appears at 2 months). By age 17 the two ossification centres have merged.

There is a history of acute injury to the knee, which helps to distinguish this from Osgood–Schlatter’s disease. Osgood–Schlatter’s is a traction apophysitis affecting the superficial portion of the growth plate. Cases of fracture of the tuberosity have been reported in patients with pre-existing Osgood–Schlatter’s disease. It is uncertain as to whether pre-existing Osgood–Schlatter’s is a risk factor for tibial tuberosity fracture. In type I fractures it may still be possible to perform a straight leg raise. There is a risk of compartment syndrome in Type III fractures.

Classification

Treatment

The initial classification of these fractures was by Watson– Jones (Table 2). He recognised that the fracture starts at the distal tip of the physis of the tibial tuberosity. The extent to which the fracture propagates towards the knee joint surface determines the group into which the fracture falls. In type I fractures the fracture stops before the ossification centres of the tuberosity and proximal tibia meet. Type II fractures run up to the ossification centre of the proximal tibia (i.e. involve the whole tuberosity). Type III fractures are intra-articular. Ogden has subsequently modified this classification (Fig. 7).8

Undisplaced fractures may be treated in a long leg cast with the knee in extension. Displaced fractures are best treated by open reduction and internal fixation as it is important that intervening periosteum is removed from the fracture. In Type III fractures the menisci may become interposed in the fracture site or torn. Small bone fragments of the tibial tuberosity may need to be sutured back into place. Larger fragments can be held with screws or K-wires. Following removal of the cast physiotherapy is useful to regain movement and rebuild muscle strength. Typically sports may restart at 3–6 months post-injury.

Aetiology and anatomy

ARTICLE IN PRESS Paediatric epiphyseal fractures around the knee

Table 2

261

Watson–Jones and Ogden classification of tibial tuberosity fractures.

Watson–Jones

Ogden

I

Fracture occurs through tibial tuberosity ossification centre

Ia Ib

Undisplaced Displaced

II

Fracture occurs between ossification centres of tibial tuberosity and proximal tibia

IIa IIb

Undisplaced Comminuted and displaced

III

Fracture extends through proximal tibial ossification centre into knee joint

IIIa IIIb

Not comminuted Comminuted

Intercondylar eminence (tibial spine) fractures Aetiology and anatomy The intercondylar eminence is the central non-articulating part of the tibial plateau. The anterior tibial spine is most commonly avulsed, along with the insertion of the Anterior Cruciate Ligament (ACL). Rarely the posterior spine and insertion of the Posterior Cruciate Ligament (PCL) may be pulled off.

Classification Myers and McKeever have classified anterior tibial spine fractures.10 Type I fractures are minimally displaced. Type II fractures have an intact posterior hinge, however the anterior portion of the tibial spine is elevated. Type III fractures are completely displaced. Type III fractures may be further classified. IIIa fractures are displaced but not rotated and IIIb fractures are displaced and rotated. Most fractures are Type II or III. Radiographs often underestimate the size of the avulsed fragment, which is largely cartilaginous. The medial and lateral extents of the avulsed fragment have the appearance of ‘‘bat wings’’ and tissue can easily become interposed, blocking reduction.

Mechanism of injury Avulsions of the tibial spine are relatively common in children. The mechanism is similar to that of ACL injury in the adult. There is a valgus stress to the flexed knee or the knee is forcibly hyperextended. This usually occurs during sport or in a fall from a bicycle. Figure 7

Ogden’s tibial tuberosity classification.

Outcome The fracture typically occurs close to skeletal maturity and therefore the prognosis is good. In younger patients genu recurvatum is a possible complication. Retained metalwork may cause irritation as it often lies superficially. Patients with type III fractures should be closely monitored for compartment syndrome.

Presentation The knee is acutely swollen and it may not be possible to fully extend or flex it. Laxity is often evident, if the child allows clinical examination. Weight-bearing on the injured leg may not be possible.

Treatment Undisplaced fractures can be treated in a long leg cast with the knee in 201 of flexion (ACL maximally relaxed). Type II

ARTICLE IN PRESS 262

M. Moran, M.F. Macnicol

fractures should reduce when the knee is extended. They may therefore be managed in a long leg cast with the knee extended (not hyperextended). The cast is retained for 6 weeks. Type II fractures that do not reduce closed should be treated as type III injuries. Type IIIa injuries may reduce closed, but Type IIIb fractures will not. In practice Type III fractures are managed with arthroscopic reduction and fixation or with an open anteromedial arthrotomy, reduction and fixation (Fig. 8). Transosseous sutures, screws or direct suture may be used to secure fixation. Post-operatively repair is protected by

immobilisation of the knee in a long leg cast in slight flexion. A recent study has shown that it may only be necessary to perform an arthroscopic reduction (removing the transverse meniscal ligament from the fracture site). The reduction can then be maintained by casting the leg in hyperextension.11 There is very little information available on the management of PCL avulsions. If they are displaced the senior author’s practice is to perform an open reduction with internal fixation via a posterior approach (Fig. 9).

Figure 8 AP and lateral views of an anterior tibial spine fracture (8a) which has been internally fixed with a cannulated screw inserted into the epiphysis (8b). The physis has not been breached by the screw.

Figure 9

Posterior spine avulsion internally fixed via a posterior approach to the knee joint.

ARTICLE IN PRESS Paediatric epiphyseal fractures around the knee

263 A superior sleeve fracture has also been described. In addition to sleeve fractures, medial or lateral osteochondral avulsions may also occur.

Classification The patella is wholly cartilaginous until the age of 3. Ossification begins centrally. It is common for there to be more than one ossification centre, which can add to diagnostic problems in younger children. As there is no specific classification system the fractures are described by their appearance and anatomical location. Avulsion fractures may be inferior, superior, medial or lateral. The inferior ‘‘sleeve’’ fracture is the most well known variant.16 These fractures usually present acutely following a definite injury are often displaced and require fixation. A more chronic problem may present with no specific recollection of injury due to repetitive tension on the inferior pole (the Sinding–Larsen–Johansson lesion). Medial avulsions are also common. As the avulsion is largely cartilaginous they may not be diagnosed at the time of initial patella dislocation. It is only as the fragment ossifies that its full extent becomes evident. Superior avulsions are much less common. The extensor mechanism is not often compromised with superior avulsions. A chronic traction, similar to Sinding–Larsen– Johansson can also occur at the superior patella pole. Lateral avulsions are a chronic superolateral separation due to the tensile pull of vastus lateralis. These are equivalent to a bipartite patella. Figure 10

Patellar sleeve fracture with minimal displacement.

Mechanism of injury Outcome If the fracture is reduced adequately the prognosis is good. Proximal migration of the fracture can result in a block to extension. Following tibial spine avulsion, even after anatomic fixation, the ACL is often stretched.12 A positive anterior draw is common but the pivot shift test is almost always negative.13 In the long term this does not seem to be symptomatic, although the laxity does not recover.14 The children that are symptomatic following tibial spine fracture seem to suffer from activity-related pain rather than instability.15

Patellar fracture Aetiology and anatomy Adolescents approaching skeletal maturity may suffer the same spectrum of patella fractures as adults. Adult pattern fractures are much less common in the younger child, as the patella is largely cartilagenous. Children more frequently suffer from peripheral osteochondral fractures of the patella, of which the inferior ‘‘sleeve’’ fracture is the most common. The fracture occurs through the zone of osseochondrous transformation, where the patella tendon collagen fibres blend directly into cartilage. Radiographs often give the appearance of a small bony fragment, however as the bulk of the avulsed fragment is cartilagenous the true size of the fragment is generally much larger (Fig. 10).

Sleeve fractures occur when the quadriceps are contracted against a flexed knee, such as in jumping. Both a fixed flexion deformity and restricted flexion of the knee joint can predispose to patella fractures. Medial fractures are usually the result of an acute lateral patella dislocation. As many as 40% of lateral patellar dislocations may be complicated by a medial patella avulsion or osteochondral fracture (from the patella or lateral femoral condyle).17

Presentation Sleeve fractures present with a traumatic haemarthrosis. There may be a palpable gap in the infrapatella region. Full active extension and a straight leg raise are usually not possible. There may be patella alta. In medial and lateral fractures the extensor mechanism is intact. Medial fractures are often recognised following patellar dislocation.

Treatment Closed treatment in a long leg cast is appropriate for undisplaced fractures. Displaced fractures that involve the extensor mechanism (e.g. transverse and sleeve fractures) require open reduction and internal fixation. This is typically performed with a tension band technique. Medial avulsions are often best excised and the medial soft tissues repaired. If the fragment involves a significant amount of the joint surface it should be internally fixed. There is some evidence

ARTICLE IN PRESS 264

M. Moran, M.F. Macnicol

that if these medial avulsions are not treated operatively there is a risk of recurrent subluxation or dislocation.16 Prolonged physiotherapy may be required after any patella fracture.

Outcome Undisplaced and well-reduced fractures have a good prognosis. As might be expected displaced comminuted fractures do not have as good an outcome as undisplaced fractures. The presence of ipsilateral femoral or tibial fractures also predicts for a poorer outcome.18 Restoration of the pre-injury length of the extensor mechanism should minimise problems with restricted knee flexion, extensor lag and defunctioning of the quadriceps.

Mechanism of injury The three mechanisms of injury are a direct blow, compression of the tibia against the femur and patella dislocation. Patella dislocation may produce both a medial patella fracture (Fig. 11) (pull off) and a lateral femoral condyle fracture (shear).

Classification Beaty19 describes a classification of osteochondral fractures based on the work of Kennedy and Smillie. It is based on the anatomic location of the injury and the mechanism of injury (Table 3).

Presentation There may be a history of patellar dislocation. The knee may be locked or there may be a history of mechanical symptoms due to a loose body within the knee. The knee is frequently swollen and a typical aspirate yields blood and fat.

Osteochondral fractures Aetiology and anatomy Acute osteochondral fractures may arise from the patella or the medial or lateral femoral condyle. Even a large fragment may have only a small amount of ossified cartilage and their diagnosis can therefore be difficult.

Treatment The management of an osteochondral fracture depends on the size of the fragment and the location of the fracture. Large fragments (1 cm2 or greater) from weight bearing surfaces warrant reduction and internal fixation. Fixation may be obtained using Herbert screws, however the majority of fragments are satisfactorily treated by arthroscopic removal from the joint. After simple excision of a fragment weight bearing can start immediately. The postoperative regime following fixation of an osteochondral fragment is considerably longer. Typically the knee is immobilised in a long leg cast or brace for 6 weeks, during which the patient is non-weight bearing. Prolonged physiotherapy will be required after this.

Outcome Figure 11

Table 3 knee.

Medial patellar osteochondral avulsion.

Classification of Osteochondral fractures of the

Site

Mechanism of injury

Medial femoral condyle

Direct blow (fall) Compression and rotation (tibiofemoral)

Lateral femoral condyle

Direct blow (kick) Compression and rotation (tibiofemoral) Acute patella dislocation

Patella (medial margin)

Acute patella dislocation

As previously stated patellar dislocations associated with a medial patellar osteochondral fragment may be at a higher risk of future instability if the fracture is not treated operatively. There is little evidence available to calculate the risk of late osteoarthritic change with fragments involving weight-bearing surfaces.

Conclusion Epiphyseal fractures around the knee are uncommon. Vigilance is required with displaced fractures as complications can be limb threatening. The risk of late limb growth problems is related to the initial displacement of the fracture and the age of the child at the time of injury. Undisplaced fractures can be managed in a long leg cast. Displaced fractures frequently require open or closed reduction and internal fixation. If an accurate reduction is achieved the results are usually favourable. Despite adequate and prompt treatment a permanent deficit may occur. The ACL is frequently lax following tibial spine avulsion,

ARTICLE IN PRESS Paediatric epiphyseal fractures around the knee although this does not seem to be clinically important in the medium to long-term. The risk of osteoarthritis after excision of osteochondral fracture fragments and articular cartilage fragments is not known.

References 1. Merchant AC, Mercer RL, Jacobsen RH, Cool CR. Roentgenographic analysis of patellofemoral congruence. J Bone Jt Surg Am 1974;56A:1391–6. 2. Mizuta T, Benson WM, Foster BK, Paterson DC, Morris LL. Statistical analysis of the incidence of physeal injuries. J Pediatr Orthop 1987;7:518–23. 3. Mann DC, Rajmaira S. Distribution of physeal and nonphyseal fractures in 2,560 long-bone fractures in children aged 0–16 years. J Pediatr Orthop 1990;10:713–6. 4. Thomson JD, Stricker SJ, Williams MM. Fractures of the distal femoral plate. J Pediatr Orthop 1995;15:474–8. 5. Riseborough EJ, Barrett IR, Shapiro F. Growth disturbances following distal femoral physeal fracture-separations. J Bone Jt Surg Am 1983;65A:885–93. 6. Simonian PT, Staheli LT. Periarticular fractures after manipulation for knee contractures in children. J Pediatr Orthop 1995;15:288–91. 7. Burkhart SS, Peterson HA. Fractures of the proximal tibial epiphysis. J Bone Jt Surg Am 1979;61A:996–1002. 8. Ogden JA, Tross RB, Murphy MJ. Fractures of the tibial tuberosity in adolescents. J Bone Jt Surg Am 1980;62A:205–15.

265 9. Christie MJ, Dvonch VM. Tibial tuberosity avulsion fracture in adolescents. J Pediatr Orthop 1991;1:391–4. 10. Meyers MH, McKeever FM. Fracture of the intercondylar eminence of the tibia. J Bone Jt Surg Am 1970;52A: 1677–84. 11. Hallam PJ, Fazal MA, Ashwood N, Ware HE, Glasgow MM, Powell JM. An alternative to fixation of displaced fractures of the anterior intercondylar eminence in children. J Bone Jt Surg Br 2002;84B:579–82. 12. Janarv PM, Westblad P, Johansson C, Hirsch G. Long-term follow-up of anterior tibial spine fractures in children. J Pediatr Orthop 1995;15:63–8. 13. Baxter MP, Wiley JJ. Fractures of the tibial spine in children. An evaluation of knee stability. J Bone Jt Surg Br 1988;70B:228–30. 14. Willis RB, Blokker C, Stoll TM, Paterson DC, Galpin RD. Long-term follow-up of anterior tibial eminence fractures. J Pediatr Orthop 1993;13:361–4. 15. Smith JB. Knee instability after fractures of the intercondylar eminence of the tibia. J Pediatr Orthop 1984;4:462–4. 16. Grogan DP, Carey TP, Leffers D, Ogden JA. Avulsion fractures of the patella. J Pediatr Orthop 1990;10:721–30. 17. Nietosvaara Y, Aalto K, Kallio PE. Acute patellar dislocation in children: incidence and associated osteochondral fractures. J Pediatr Orthop 1994;14:513–5. 18. Maguire JK, Canale ST. Fractures of the patella in children and adolescents. J Pediatr Orthop 1993;13:567–71. 19. Rockwood CA, Wilkins KE, Beaty JH, editors. Fractures in children. 5th ed. New York, USA: LWW; 2001.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 266–273

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: CHILDREN’S ORTHOPAEDIC SURGERY

(iii) Congenital hand anomalies A.C. Watts, G. Hooper The Hand Unit, St John’s Hospital, Howden Road West, Livingston, West Lothian EH54 6PP, UK

KEYWORDS Congenital hand anomalies; Classification; Epidemiology; Management

Summary This is a review of congenital anomalies of the hand, taking into account the development of the upper limb, the known aetiological factors and classification of these disorders. The principles of their management are described. & 2006 Elsevier Ltd. All rights reserved.

Introduction Congenital upper limb anomalies are second only to congenital cardiac abnormalities in frequency and account for 10% of all malformations at birth. Congenital abnormalities may arise either from primary structural defects due to localized developmental failure or secondary changes after normal development has begun. A number of different classification systems have been proposed but the most widely known and that adopted in modified form by the international federation of societies for surgery of the hand (IFSSH) and the American society for surgery of the hand (ASSH) is the Swanson classification1 (Table 1). The aim of this classification is to use simple terms to describe the appearances, avoiding words based on Latin and Greek, or eponyms, which may be misunderstood or difficult to translate into other languages. This has not been entirely achieved, as will be apparent from the terms used in this article. It implies nothing about the causes of any anomaly. Most but not all anomalies can be classified using this system.2,3 It should be borne in mind that different types of

Corresponding author. Tel.: +44 1506 419666;

fax: +44 1506 460592. E-mail address: [email protected] (G. Hooper). 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.06.013

anomaly can occur in the same limb, making classification difficult.

Embryology The limb bud develops around 26 days after fertilization. At 52–53 days the embryo is 22–24 mm in length (crown-rump length) and the fingers are entirely separate. Eight weeks after fertilization all the limb structures are present. This period of embryogenesis when rapid development of the limb is occurring is when most congenital upper limb abnormalities occur. The fetal period that follows sees differentiation, maturation and enlargement of existing structures. The limb develops in a proximal to distal direction. Cells from the lateral plate mesoderm (which become bone, cartilage and tendon) and somatic mesoderm (which become muscle, nerve and vascular structures) migrate into the overlying ectoderm to form the limb bud. Three signalling centres control development: the apical ectodermal ridge (AER) regulates proximodistal development, the zone of polarizing activity (ZPA) radioulnar development and an ectodermal signalling centre regulates dorsoventral development (differentiation of dorsum of finger with nail, and ventral surface with pulp). There is increasing

ARTICLE IN PRESS Congenital hand anomalies

Table 1 lies.

267

IFSSH classification of congenital hand anoma-

Type Failure of formation Transverse Longitudinal Central Failure of differentiation (separation)

Examples

Congenital transverse amputation Radial absence Cleft hand Radioulnar synostosis Carpal coalition Syndactyly Polydactyly Macrodactyly Short bones

Duplication Overgrowth Undergrowth Constriction ring syndrome Generalized Skeletal dysplasias abnormalities and syndromes Miscellaneous, Five-fingered hand unclassifiable elsewhere

knowledge of how these signalling centres control cellular differentiation and positioning by growth factors and gene expression. Congenital anomalies may be isolated (confined to the upper limb, possibly bilateral) or part of a malformation syndrome, with several congenital abnormalities affecting different systems, for example the heart and renal systems. Malformation syndromes are presumably the result of some process affecting several developing systems at the same time.

Aetiology The precise aetiology of congenital upper limb anomalies is largely unknown. Some have a clear genetic pattern, which may be autosomal, recessive or sex-linked, but others are sporadic and non-hereditary. An apparently sporadic anomaly with no family history may be the result of new mutation. A clinical geneticist should assess all families in which there is a child with a congenital anomaly. Physical and chemical agents acting on the developing embryo may cause non-hereditary anomalies. Relatively few such factors have been identified, the most well-known being the drug thalidomide.4 The former diagnostic process of chorionic villus sampling was also associated with limb disorders.5

Epidemiology3,6,7 The precise epidemiology of congenital upper limb abnormalities is difficult to ascertain. Minor anomalies may go unreported, and consideration must be given whether to include stillbirths when calculating the population prevalence. It is possible that the quoted incidences may change

if antenatal diagnosis by ultrasound is followed by termination or better maternal health changes the risks of embryogenesis. These factors will only be apparent from long-term studies. Excluding stillbirths the overall prevalence is 11.46–19.77 per 10 000 live births. Little variation has been shown between different population groups in the prevalence of most anomalies but ring constriction syndrome was found to be 4–6 times more common in Japan than in Edinburgh6 and ulnar polydactyly is known to be more common amongst Afro-Americans.8 Male children are more commonly affected than female with a reported ratio of 3:2 for all categories of abnormality other than generalized skeletal abnormalities. There is a reported increase in prevalence with increasing maternal age: for mothers over the age of 40 years the prevalence ratio is twice that for mothers under 30 years. Between 5% and 20% of upper limb anomalies are associated with an identifiable syndrome. Bilateral anomalies occur in half of the cases and the majority are of the same type in each hand. Multiple upper limb anomalies occurring in the same individual are reported to occur in 17% of cases. Failures of differentiation and duplications are the most common anomalies. Polydactyly is the most common individual diagnosis. Failure of formation is the next most common category with both transverse and longitudinal defects occurring equally. Radial ray deficiency is the most common single diagnosis in this category. The rate of mortality is high. Eighteen per cent die before the age of 6 years due to other congenital disorders.

Failure of formation In transverse absence all proximal structures including nerves, arteries and tendons will be present. In longitudinal absence lateral structures are absent or deficient. The aetiology of transverse absence is multifactorial but experimentally removal of the AER during embryogenesis results in a truncated limb similar to that seen in congenital amputations. Intercalary defects, in which the proximal and distal parts of the limb are present but an intermediate segment is missing, probably do not exist.9

Transverse absence Transverse absence, incorrectly known as congenital amputation and also incorrectly attributed at times to constriction bands, occurs when the limb fails to form below a certain level. The most common level is the proximal forearm (more often on the left side), followed by transmetacarpal, distal forearm and transhumeral. The anomaly is usually unilateral and is classified according to the last remaining bone segment. Finger nubbins may be present at the distal end. Developmentally the child is usually normal but may not crawl. Prostheses are more likely to be accepted if introduced before the age of 2 years. Initially the child is given a passive hand, but may start to use an active device between the age of 2 and 3. Active devices may be cable operated or myoelectric. The latter may be perceived as superior but they are heavy and need a lot of maintenance. Despite the

ARTICLE IN PRESS 268

A.C. Watts, G. Hooper

total difference in control mechanism, most children can use both types, although sufficient muscle bulk to generate a control signal is necessary to use a myoelectric arm.

Longitudinal absence Radial hypo- and aplasia Radial deficiency includes a spectrum of anomalies affecting the radial side of the forearm and hand, including the thumb, which may be normal or hypoplastic. In radial absence (Fig. 1) the forearm is usually very short, the wrist unstable and the digits have reduced range of movement and power. It is an uncommon anomaly but is the most common longitudinal deficiency. There is a strong association with congenital syndromes, some of which may be lifethreatening, such as the vertebral anomalies, anal atresia, cardiac anomalies, tracheoesophageal fistula, renal anomalies, limb anomalies syndrome (VACTERL), Holt-Oram syndrome (cardiac septal defects and limb anomalies), thrombocytopenia, anaemia, absent radius (TAR), Fanconi anaemia and chromosomal anomalies such as trisomy 13 and 18. The condition is often bilateral. Function may be severely affected especially where the deformity is bilateral. The primary issues are thumb hypoplasia, wrist instability and inequality of the forearm bones. The aim of surgery is to obtain carpal centralization and stabilization on the end of the ulna and thumb reconstruction by pollicization may be required. External fixators are often used to stretch the soft tissue structures on the radial side of the wrist prior to centralisation. Recurrence of deformity and stiffness have been reported in long-term follow-up studies. Surgery should not be undertaken in the presence of severe major organ defects due to the risks of anaesthesia, and when active elbow flexion is absent. Ulnar deficiency This is much less common than radial deficiency with an incidence of one in 100 000 live births. It is most commonly unilateral and does not have the same systemic associations as radial deficiency. The entire upper limb may be hypoplastic with a malformed or fused elbow joint. The hand and carpus are always affected with absent ulnar digits in 90%, syndactyly in a third and thumb abnormalities in 70%.

Figure 1

Complete radial aplasia.

Severe bowing of the radius may result in a bizarre appearance and marked malposition of the hand. The aim of surgery is to enable the child to bring the hand to the mouth. Resection of the fibrocartilaginous anlage of the distal ulna, once considered mandatory, has been shown to be unnecessary in most cases,3 although it may still be required if the radius is extremely bowed and corrective radial osteotomy is necessary. Humeral osteotomy may be indicated if the arm is internally rotated, and closing wedge osteotomy can be employed to place the elbow in a more flexed position. Surgical correction of associated digital anomalies can improve hand function. Cleft hand True cleft hand (stigmatizing terms such as ‘‘lobster-claw hand’’ should be avoided) is manifest by a central V-shaped absence, which may be associated with absence of one or more digits. It is commonly inherited as an autosomal dominant trait, sometimes leading to local lusters of cases in affected families. It may also occur in a number of syndromes. The aetiology is thought to be a wedge-shaped defect of the apical ectoderm of the limb bud. It may be unilateral or bilateral and may involve the feet. Cleft hand produces a particular challenge because function may be good but the appearance may be socially unacceptable. Those with a family history are more likely to accept an observational approach. Early surgery should be aimed at preventing worsening of the deformity, for example by removing a transverse bone, which will widen the cleft as it grows (Fig. 2). Other surgery can be delayed until the child is older, with emphasis placed on reconstruction of the first web space and closure of the cleft. Symbrachydactyly This term is discussed here as it occurs in the literature and causes confusion. The use of the Greek-based term symbrachydactyly goes against the principle of simple language in the IFSSH classification but it perhaps illustrates one of the difficulties of that classification, at least for certain disorders. Symbrachydactyly has been used to describe a spectrum of distal hand deformities, usually occurring unilaterally, which are all thought to result from a sporadic failure of mesenchymal differentiation. The 4 types are seen as: (1) short stiff fingers and a normal thumb;

Figure 2 Cleft hand. Note the transverse bone, which will widen the cleft with growth.

ARTICLE IN PRESS Congenital hand anomalies (2) absence of the central 3 fingers and relatively normal thumb and little finger; (3) absence of all fingers with preservation of the thumb; and (4) transverse absence of all digits at the metacarpal level. However, types 2, 3 and 4 differ from true transverse absences in that there are digital nubbins that bear nail remnants. Using the IFSSH classification these abnormalities would be classified respectively as: (1) undergrowth; (2) terminal central defect; and (3) and (4) as terminal transverse defects. However, it should be remembered that the IFFSH classification is descriptive and does not attempt to ascribe a pathological mechanism to each anomaly, whereas the concept of ‘‘symbrachydactyly’’ is based on a proposed commonly pathological process. A more detailed overview of this topic and other aspects of classification is given in the article by McCarroll.3

Failure of differentiation Radioulnar synostosis This is not of course a congenital hand disorder but is mentioned here because of the effect that it may have on hand function and the fact that it is commonly seen by orthopaedic surgeons. The radius and ulna normally separate late in the first trimester of pregnancy. Failure of this process results in synostosis. This condition may be inherited in an autosomal dominant fashion and is bilateral in 60% of cases. The forearm will be fixed in pronation in most cases. Diagnosis is typically delayed, sometimes until adolescence. Radiographic evidence of a bony bar between radius and ulna confirms the diagnosis. It is usually an isolated anomaly but other musculoskeletal anomalies co-exist in a third of cases. If the deformity is unilateral surgery may not be required as the child can compensate with shoulder and elbow movement. Surgery is indicated if the forearm is pronated greater than 601. Derotation osteotomy, which should be combined with slight shortening by bone resection, aims to place the hand in neutral to 151 of pronation in unilateral cases or 10–201 of pronation of the dominant arm and neutral rotation of the non-dominant arm in bilateral cases. The patient must be observed closely post-operatively for compartment syndrome.10

Carpal coalition

269 fixed. Most cases are sporadic but it may be inherited in an autosomal dominant fashion with variable penetrance. The little finger is most commonly affected. The pathological mechanism has not been defined for all cases but some are associated with anomalous lumbrical insertions, which can be released. Generally speaking the results of surgical treatment are otherwise poor, particularly if the proximal interphalangeal joint is fixed. Conservative treatment with stretching, splints and serial casts can be employed with some success.

Clinodactyly In this condition there is angulation of the digit in the coronal plane distal to the metacarpophalangeal joint. It is associated with many syndromes. The underlying abnormality is in the alignment of the interphalangeal joints due to asymmetrical longitudinal growth. The overall incidence is difficult to ascertain as many cases never present. The most common type shows radial angulation which is often bilateral and is inherited in an autosomal dominant fashion with variable penetrance. Males are more likely to express the phenotype. The most severe form of growth insult results from the formation of a ‘‘delta’’ phalanx often associated with a Cshaped longitudinal bracketed epiphysis (Fig. 3). Hand function is rarely significantly altered by isolated clinodactyly but intervention may be sought for cosmesis. Indications for surgery are severe deformity with shortening or involvement of the thumb, and moderate deformity with functional impairment. Osteotomy is required to correct the underlying bony abnormality. Opening wedge osteotomy has the advantage of gaining or preserving length. Single stage correction is preferred. An alternative technique, which can only be used when the growth plate is open, is by epiphyseal bracket resection and fat grafting, allowing ‘catch-up’ growth on the side of the concavity.12

Symphalangism This is a curious condition in which an interphalangeal joint, usually the proximal interphalangeal joint in the small finger, fails to form by the usual mechanism of cavititation of the cartilage between the cartilaginous precursors of the

During the embryonic phase of limb development the carpal bones develop from a cartilaginous condensation. Failure of this cavitatory process results in coalition of the carpal bones. The most commonly fused bones are the lunate and the triquetrum in isolated carpal coalition but it can occur as part of a syndrome. The overall incidence is estimated to be 1 in 1000 of the population but there is increased frequency in females and in those of African descent. This condition is usually asymptomatic and an incidental finding but symptomatic cases have been reported.11

Camptodactyly Camptodactyly is a flexion deformity of the proximal interphalangeal joint that may be progressive and is often

Figure 3 Clinodactyly affecting the small finger. Note the longitudinal bracketed epiphysis.

ARTICLE IN PRESS 270 phalanges. It has an autosomal dominant form of inheritance and its interest lies in the fact that the condition has been traced in families for many generations, far longer than any other congenital condition.

Flexed Thumbs Congenital trigger thumb This is a misnomer. There is no evidence that the thumb is abnormal at birth and the majority of thumbs present with fixed flexion rather than triggering. Bilateral involvement occurs in about a quarter of children. A nodule lying just proximal to the A1 pulley is readily palpable on the tendon of flexor pollicis longus and prevents full excursion of this tendon. The thumb may be passively manipulated into extension, often with a ‘pop’, and may be splinted in extension. There is debate as to the natural history. It is currently accepted that in those diagnosed before 1 year of age the thumb may be observed with or without splintage up to 3 years of age. The indications for surgery are failure of resolution with conservative management, a child presenting over the age of 2–3 years and the presence of a rigid deformity. The treatment is by surgical release of the A1 pulley under general anaesthesia and with magnification. Care must be taken to avoid injury to the digital nerve. Recurrence is very rare. Congenital clasped thumb Congenital clasped thumb describes a spectrum of anomalies from minor deficiency of the extensor mechanism to severe abnormality of the thenar muscles, web space and soft tissues. A type I deformity is usually supple with absence or hypoplasia of the extensor mechanism. Type II is complex with additional joint contractures, collateral ligament abnormality, web space contracture or thenar muscle abnormality. A type III anomaly is associated with arthrogryposis or its related syndromes. The diagnosis may be delayed because the newborn typically holds the thumb clenched in the palm. The diagnosis is made by the appearance of a thumb that rests in flexion with extensor lag that is usually at the metacarpophalangeal joint. This indicates hypoplasia of the extensor pollicis brevis muscle. Extension lag at the interphalangeal joint and an adduction deformity of the metacarpal indicate deficiency of extensor pollicis longus and abductor pollicis longus, respectively. The treatment is by splintage in extension to prevent further hypoplasia and allow hypertrophy. The results are good with splintage retained for 2–6 months in cases identified at less than 1 year of age. Surgery is indicated if splintage fails or the child is over the age of 2 years. Mild deformity that is missed may not require treatment, as function may not be affected. Extensor indicis transfer is the operation of choice where this is present.

A.C. Watts, G. Hooper apoptosis occurs from distal to proximal between the digits to form the web-space. Syndactyly is commonly bilateral and a family history is reported in 10–40% of cases. The degree of digital fusion is variable, being described as complete or partial. The fusion may include soft tissue only (simple syndactlyly) (Fig. 4) or soft tissue and bony parts (complex syndactyly). Abnormalities of the nails, digital nerves and vessels, and tendons may be seen. Extra digits may be contained within a fused mass of digits in synpolydactyly. Syndactyly may be isolated, in which case the long/ring finger web space is most commonly affected, or involve several fingers. The tendon mechanisms may be normal, shared or deficient. Syndactyly results in cosmetic, functional and developmental problems. Tethering of longer digits by the adjacent shorter digit leads to flexion contracture and deviation of the longer digit towards the shorter. In all cases surgery should be considered unless the deformity is minor, other comorbidities prevent intervention or the deformity is so severe that intervention may have a deleterious affect on function. There is debate as to the ideal timing of surgery; the aim should be to complete any corrective procedures by school age. Better outcomes have been shown with surgery after 18 months of age, but some prefer to operate earlier to prevent developmental problems. Staged procedures are indicated when release is required on both sides of a digit to reduce the risk of vascular compromise. Reconstruction is performed primarily by ‘‘Z’’ plasty. A full description of the many different techniques can be found in larger texts.13

Syndactyly Syndactyly is a common congenital hand anomaly that occurs when the normal processes of digital separation and web space formation fail to occur. The digits form from condensations of mesoderm within the terminal paddle of the embryonic limb under the control of the AER. Regulated

Figure 4 (a,b) Simple syndactly between the ring and long fingers.

ARTICLE IN PRESS Congenital hand anomalies Skin grafts should be full thickness where possible and should be avoided in the commissure to prevent interdigital contractures and web creep. Outcomes are better for simple syndactyly. Syndactyly may be associated with other deformities, and be a feature of a syndrome, such as Apert syndrome (severe complex syndactyly with bicoronal craniosynostosis) or Poland syndrome (simple syndactyly, brachydactyly and absence of the sternocostal portion of pectoralis major), or present in some chromosomal abnormalities.

Duplication The terms radial and ulnar are now preferred to pre-axial and post-axial when describing polydactyly. Radial polydactyly Polydactyly of the thumb is more common amongst Caucasians and Native Americans and is usually sporadic although a family history is not uncommon. It is classified according to the system of Wassel14 into seven types depending on the degree of duplication and separation of the bony parts (Fig. 5). The most common of these (30–40%) is the type IV duplication, followed by type II. The condition is not a true duplication as neither digit is as robust as a normal thumb nor do they both contain the usual soft tissue attachments. The thumb should be examined for stability and mobility of joints. One of the duplications will tend to be larger and in surgery for type III and IV duplications the smaller digit may be ablated. Surgery is usually carried out between 6 and 12 months of

271 age before pinch grip develops. The aim is to create a stable functioning thumb and this may require using parts from both digits in reconstruction. Tendons should be centralized on the retained thumb and intrinsic muscles should be transferred from the ablated digit to the one that is retained. Any residual instability or malalignment may progress to the formation of a lateral ‘Z’ thumb deformity.15 Ulnar polydactyly Ulnar side polydactyly is frequently inherited in an autosomal dominant pattern but with variable penetrance. Its prevalence is highest amongst black Americans, and when seen in Caucasians it is suggestive of an underlying syndrome. The accessory digit can either be well-formed (type A) or a rudimentary pedunculated nubbin (type B). A well-formed digit must be surgically removed and this is probably safer than simple ligation for the type B digit. Central polydactyly Central polydactyly is the occurrence of an accessory digit within the hand but not on the ulnar or radial border. The ring finger is most commonly duplicated followed by the long and lastly the index finger. It has been suggested that central polydactyly is the result of the same mechanisms that produce typical cleft hand deformity, a view supported by the occurrence of typical cleft hand deformity and central polydactyly in identical twins.16 If the accessory digit is fully formed and functional it need not be removed, otherwise ray resection is the treatment of choice. Synpolydactyly is treated by separation of the digits and debulking of the accessory digit. Complete resection is often

Figure 5 Wassel classification of thumb duplication. Type (I) bifid terminal phalanx, Type (II) duplicate terminal phalanx, Type (III) bifid proximal phalanx, Type (IV) duplicate proximal phalanx, Type (V) bifid metacarpal, Type (VI) duplicate metacarpal and Type (VII) triphalangism.

ARTICLE IN PRESS 272

A.C. Watts, G. Hooper

impossible leading to somewhat disappointing results. If the duplication is complex no treatment may result in a better functional outcome than surgical intervention.

Overgrowth and undergrowth Macrodactyly The most common form of macrodactyly, a disproportionately enlarged digit, is an isolated deformity associated with lipomatosis of a proximal nerve. The anomaly may be progressive (disproportionate growth) or static (the enlarged digit maintains in the same proportion to the rest of the hand throughout growth). Multiple digital enlargement is more common than single digit, but is usually confined to an area supplied by either the median or ulnar nerves. The index finger and long finger are most commonly involved. Macrodactyly may be present as part of a more widespread anomaly. It is associated with neurofibromatosis, gigantism and other syndromes such as Ollier disease, Maffucci syndrome, Klippel–Trenaunay–Weber syndrome and Proteus syndrome. Treatment is extremely difficult and is aimed at limiting growth or reducing the size of the digit, for example by epiphysiodesis and debulking. Amputation may be necessary in extreme cases.

Thumb hypoplasia The hypoplastic thumb presents as a spectrum of anomaly from a slightly small thumb to complete absence. It is frequently part of a radial deficiency. The modified Blauth classification17 is used to guide treatment (Table 2). The status of the intrinsic and extrinsic muscles should be assessed, along with the stability of the MCP and CMC joints. The diagnostic challenge lies in correctly identifying the type III deformity and whether it is a type A or B anomaly. This may require serial examinations before treatment, but

Table 2 plasia.17

Modified Blauth classification of thumb hypo-

Type

Findings

Treatment

I

Minor generalised hypoplasia Absence of intrinsic thenar muscles First web space narrowing Ulnar collateral ligament insufficiency As above plus: Extrinsic muscle and tendon abnormalities Skeletal deficiency A: Stable CMC joint B: Unstable CMC joint Floating thumb Absence

Augmentation

II

III

IV V

may not be distinguishable until the child develops pinch and grasp in the hand. The stable type A thumb will be incorporated into the pinch grip but the unstable type B thumb will be bypassed with pinch grip performed between the index and middle finger. Type I deformity may require no surgical intervention as function may be good. Pollicization is the treatment of choice for type IIIB, IV and V anomalies. Theoretically pollicization should be done between 6 and 12 months of age, before development of pinch grip, but in practice later operation does not seem to affect the outcome. The functional results are generally good but pinch strength is reduced.18 In type II and IIIA deformity the first web space can be widened with Z-plasty and release of the interosseous fascia, and absence of functioning thenar muscles can be circumvented by opposition transfer using abductor digiti minimi or the flexor digitorum superficialis tendon from the ring finger. Extrinsic extensor absence in type IIIA anomaly can be reconstructed with extensor indicis transfer but adequate long flexor reconstruction remains a challenge.

Brachydactyly This term describes a short digit, or ray, in which all the skeletal elements are present but 1 or more are shortened. It may occur in isolation or as part of a syndrome. Similar appearances may be due to injuries resulting in physeal growth arrest (trauma, infection, frostbite). It is a common feature of autosomal dominant skeletal dysplasias such as hereditary multiple exostoses and multiple epiphyseal dysplasia, but is also seen in conditions such as pseudohypoparathyroidism. The condition is not functionally disabling and treatment is not necessary.

Constriction ring syndrome This is characterized by partial or complete circumferential constrictions around digits or limbs (Fig. 6). The aetiology is unknown and contentious but is currently thought to be related to endogenous mesenchymal degeneration within the developing limb rather than damage by extrinsic amniotic bands. The constriction band results in mechanical

Opponensplasty First web release UCL reconstruction A: Reconstruction B: Pollicization

Pollicization Pollicization Figure 6

Constriction ring syndrome.

ARTICLE IN PRESS Congenital hand anomalies strangulation of the distal appendage. This may be tense with incompressible oedema and occasional nerve palsies. It has been proposed that this condition may result in transverse absence characterized by normal proximal anatomy. The digits may be fused at their distal parts (acrosyndactyly). When the child is born with a congenital amputation due to constriction band syndrome, the condition may be confused with a true transverse deficit. However, in constriction band syndrome there is invariably evidence of skin grooving or other damage in the affected limb or other limbs. Treatment is aimed at functional and aesthetic improvement. Urgent surgery to salvage a threatened appendage or for nerve decompression may be required. Complete circumferential excision of the band with ‘‘Z’’ plasty is the treatment of choice.

273 anomalies should not be underestimated and must be addressed in each case. The surgeon should work as part of a team that must include an experienced children’s hand therapist. Detailed descriptions of surgical procedures are outside the scope of this article. The common procedures involve removal of extra digits, realignment of digits, separation of digits and restoration of a functional thumb by pollicization of a normal digit or toe-to-hand transfer if no other digit is available. It is preferable to complete surgery before normal school age. Advances in antenatal ultrasound are enabling earlier detection of congenital upper limb deformities. Relton McCarrol in his stimulating review article,3 has identified in-utero surgery and the use of chromosomal analysis and recombinant DNA technology as future methods for the treatment and prevention of congenital hand anomalies.

Generalized abnormalities and syndromes The hands may have characteristic shape in many common skeletal dysplasias. For example, in Marfan syndrome the hands are long and thin and in achondroplasia they are short and square. Examples of malformation syndromes have been mentioned elsewhere in the text.

Miscellaneous disorders This group contains those anomalies that cannot readily be classified elsewhere. A couple of examples will be given.

Madelung deformity Madelung deformity is excessive ulnar and palmar angulation of the distal radius. It is often bilateral and more commonly affects girls, usually presenting between 6 and 13 years of age with deformity. It is caused by growth arrest of the ulnar and palmar part of the distal radial growth plate. The ulna is relatively longer than the radius and the lunate is wedge shaped. The appearance is often seen in dyschondrosteosis, a common skeletal dysplasia. No treatment is required for painless deformity but physiolysis and ligament release can be done in skeletally immature patients to prevent worsening of the deformity. Symptomatic skeletally mature patients may require correction of the deformity with osteotomy and distal ulnar resection or a Sauve ´-Kapanji procedure to treat pain arising from the distal radio-ulnar joint.

Five-fingered hand The radial digit is finger-like, with three phalanges, and it lies in the same plane as the other digits. It can be treated by pollicization of the radial digit.

Principles of surgery Surgery is only a part of the management of children with congenital hand anomalies. The aim of surgical treatment is to achieve the maximum function possible for each patient. The psychological and social impact of congenital hand

References 1. Swanson AB, Swanson GD, Tada K. A classification for congenital limb malformation. J Hand Surg [Am] 1983;8:693–702. 2. Flatt AE. A test of a classification of congenital anomalies of the upper extremity. Surg Clin North Am 1970;50:509–16. 3. McCarroll HR. Congenital anomalies: a 25-year overview. J Hand Surg [Am] 2000;25A:1007–37. 4. Lenz W. Thalidomide and congenital anomalies. Lancet 1962;i:45. 5. Matroiacovo P, Botto LD, Calvacanti DP, et al. Limb anomalies following chorionic villus sampling. Am J Med Genet 1992;44: 855–64. 6. Lamb DW, Wynne-Davies R, Soto L. An estimate of the population frequency of congenital malformations of the upper limb. J Hand Surg [Am] 1982;7A:557–62. 7. Giele H, Giele C, Bower C, Allison M. The incidence and epidemiology of congenital upper limb anomalies: a total population study. J Hand Surg [Am] 2001;26A:628–34. 8. Woolf CM, Myrianthopoulos NC. Polydactyly in American negroes and whites. Am J Hum Genet 1973;25:397–404. 9. Tytherleigh-Strong G, Hooper G. The classification of phocomelia. J Hand Surg [Br] 2003;28B:215–7. 10. Simmons BP, Southmayd WW, Riseborough EJ. Congenital radioulnar synostosis. J Hand Surg [Am] 1983;8A:829–38. 11. Simmons BP, McKenzie WD. Symptomatic carpal coalition. J Hand Surg [Am] 1985;10A:190–3. 12. Vickers D. Clinodactyly of the little finger: a simple operative technique for reversal of the growth anomaly. J Hand Surg [Br] 1987;12B:335–42. 13. Kay S, McCombe D, Kozin S. Deformities of the hand and fingers. In: Green DP, Hotchkiss RN, Pederson WC, Wolfe SW, editors. Green’s operative hand surgery. Philadelphia: Elsevier, Churchill Livingstone; 2005. p. 1381–444. 14. Wassel HD. The results of surgery for polydactyly of the thumb. A review. Clin Orthop 1969;64:175–93. 15. Ogino T, Ishii S, Takahata S, Kato H. Long-term results of surgical treatment of thumb polydactyly. J Hand Surg [Am] 1996;21A: 478–86. 16. Manske PR. Cleft hand and central polydactyly in identical twins: a case report. J Hand Surg [Am] 1983;8A:906–8. 17. Kozin SH, Weiss AA, Webber JB, Betz RR, Clancy M, Steel HH. Index finger pollicization for congenital aplasia of hypoplasia of the thumb. J Hand Surg [Am] 1992;17:880–4. 18. Manske PR, Rotman MB, Dailey LA. Long-term functional results after pollicization for the congenitally deficient thumb. J Hand Surg [Am] 1992;17A:1064–72.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 274–285

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: CHILDREN’S ORTHOPAEDIC SURGERY

(iv) Cervical spine problems in children Jerard Ross, Lynn Myles Royal Hospital for Sick Children, Sciennes Road, Edinburgh, Scotland, EH9 1UW, UK

KEYWORDS Child; Cervical vertebrae; Spinal injuries; Spinal fusion; Bone neoplasm; Klippel–Feil; Osteochondrodysplasia; Mucolipidoses

Summary The paediatric spine is the site of a number of conditions, both congenital and acquired, which are unusual in adult practice if not unique to paediatrics. Some insight into the embryological development of the cervical spine, the normal paediatric anatomy and its development into the standard adult anatomy is crucial to understanding paediatric spinal pathology. Conditions which frequently come under the remit of orthopaedic surgeons such as trauma, bony tumours and congenital bony and ligamentous disorders are discussed. & 2006 Published by Elsevier Ltd.

Introduction Disorders of the bony paediatric spine are the responsibility of many clinicians including orthopaedic surgeons, neurosurgeons, paediatric surgeons, paediatricians and geneticists. Some understanding of the embryological development of the cervical spine, its growth patterns and normal anatomy, in both the child and adult, is necessary to appreciate the spectrum of disorders which may be encountered in paediatric practice. Disorders of the cervical spine in children are uncommon and therefore any one clinician, unless super-specialised, is unlikely to build up much personal experience and therefore the purpose of this review is to synthesise current opinions on a variety of the more common disorders encountered.

Corresponding author.

E-mail addresses: [email protected] (J. Ross), [email protected] (L. Myles). 0268-0890/$ - see front matter & 2006 Published by Elsevier Ltd. doi:10.1016/j.cuor.2006.06.009

Embryological development and normal anatomy Once the embryo undergoes the process of gastrulation (i.e., the formation of a trilaminar disc), the formation of the notochord induces the overlying ectoderm to form neuroectoderm, from which the brain and spinal cord are derived, and the para-axial mesoderm is induced to form somites. The craniocervical junction and the cervical spine develop from the four occipital and seven cervical somites. The ventromedial portions of the somites become the sclerotomes and one sclerotome contributes to two adjacent vertebral bodies while the neural arch arises solely form the caudal portion of the sclerotome. This process is driven by, at least in part, two families of regulatory genes: the Box and Pax families and is reviewed in detail by David and Crockard.1 Chondrification of these mesenchymal elements begins at 6 weeks of development in utero and ossification proceeds by endochondral ossification from 9 weeks. The subaxial vertebrae ossify from three primary centres, one in the vertebral body and one in each of the neural arches. The

ARTICLE IN PRESS Cervical spine problems in children synchondrosis between the two neural arches fuses by age three and the synchondroses between the arches and the body between the ages of 3 and 6 years. In addition, secondary ossification centres exist at the tips of both the transverse and spinous processes, and ring apophyses at the superior and inferior endplates which may remain unfused until early adulthood. The first two vertebrae are unique in their ossification. The atlas (C1) is composed of three centres of ossification with individual centres in the anterior arch and in each neural arch. The anterior arch is not ossified in most children at birth, generally becoming visible as an ossification centre by 1 year and fusing with the neural arches around 7 years. The two neural arches generally fuse posteriorly by 3 years. Therefore, developmentally normal non-fusion can appear pathological. The axis (C2) is more complex still, with four ossific centres at birth (body, both neural arches and one in the odontoid process which itself forms from two centres fusing in utero). Fusion of these synchondroses proceeds first with the neural arches posteriorly by the age of 3 years, while the odontoid process fuses with the body of C2 by 3–6 years; this subdental synchondrosis is frequently confused with a fracture (Fig. 1). In addition, a secondary centre of ossification appears in the tip of the dens at 3–6 years fusing with the rest of the dens by age 12.2 Radiographically synchondroses appear sclerotic while fractures do not.

Figure 1 Lateral radiograph in a child showing the subdental synchondrosis and pseudosubluxation of C2 upon C3 and C3 upon C4.

275

Normal radiographic variants There are a number of anatomical differences between the adult and the paediatric spine which can cause confusion when assessing for pathology. In flexion, the normal adult anterior atlanto-dental interval (ADI) is p3 mm while in children it may be normal up to 5 mm. In children it is normal to have some overlap of the lateral masses of C1 relative to the dens and up to 6 mm can be seen in children as old as 8 years. In adults, X8 mm usually indicates disruption of the transverse ligaments indicating an unstable injury. The variable fusion of the C1 synchondroses can further confuse the picture for the unwary. In addition, flexion can result in an increase in the distance between the tips of the spinous processes of C1 and C2 which is a normal finding in children. Pseudosubluxations of C2 on C3 and, to a lesser extent, C3 on C4 are probably the most commonly over-interpreted findings on paediatric cervical spine films (Fig. 1). Up to 40% of children less than 8 years old may demonstrate this phenomenon,3 with C2 moving 2–3 mm on C3. The spino-laminar line can be used to differentiate normal from abnormal movements (Fig. 2). The cervical lordosis seen in adults, the absence of which may be pathological, is commonly not seen in children. Prevertebral soft tissue swelling, a useful sign in adult cervical spine injury, is unreliable in children, being dramatically increased when crying Table 1.

Figure 2 Lines used in assessing the alignment on a lateral cervical spine radiograph; (1) anterior vertebral line; (2) posterior vertebral line; (3) spinolaminar line and (4) spinous process line.

ARTICLE IN PRESS 276

J. Ross, L. Myles

Table 1 injury.

Practice points for paediatric cervical spine

 All traumatised children should be suspected of having sustained spinal column or cord injury

 Maintenance of alignment of the cervical spine in children under 8 years requires elevation of the thorax

 Children unable to communicate due to age or injury    

are at high risk of having sustained a cervical spine injury Children without neurological deficit, neck pain, distracting injury or intoxication are unlikely to require imaging Lateral and AP plain views are useful in all traumatised children CT is useful to assess difficult to visualise areas i.e., craniocervical junction MRI is useful in non-assessable children or those with fixed or transient neurological deficit

Fractures are easy to misdiagnose in children especially those in the region of the occiput to C2 where the complex normal developmental anatomy can be confusing. Avellino et al.4 found that 19% of paediatric cervical spine injuries were misdiagnosed on initial assessment, especially in children under 8 years old, and the majority of these were due to misinterpretation of normal findings.

the facet joints which in the child are more horizontally oriented predisposing to subluxation and translational movement of the vertebral bodies. This translational movement is exacerbated further by the normal anterior wedging of the superior aspects of the vertebral bodies in children (Fig. 3). These anatomical differences result in a pattern of injury which differs from that seen in adults. The differential elasticity of the spinal column in comparison with the neural elements results in the syndrome of spinal cord injury without radiological abnormality (SCIWORA). In addition, children are twice as likely to injure the bony upper cervical segments (C1-4) rather than the lower segments although the frequency of lower cervical injury increases with age. While fractures are the most common bony injury regardless of age, dislocations occur more frequently in the upper segments and in the younger age group (o8 years).6 Injuries in the younger child are also more likely to result in an associated cord injury. The change from a paediatric pattern of injury to that of adults has traditionally been held to occur around the age of 8–9 years when injuries in the mid-cervical spine become more frequent. By this age all the synchondroses between the centres of ossification have fused. However, some authors maintain that the trend to an adult pattern of

Traumatic injury to the cervical spine Spinal injury in children is rare but can have catastrophic consequences. In the UK, 3.4% of paediatric trauma admissions have spinal column or cord injury and of these approximately 45% are in the cervical spine. Spinal cord injury presents in about 16.5% of all spinal injury admissions and 55% of all cord injuries were in the cervical spinal cord.5 Spinal column injuries are more frequent after road traffic accidents and falls (especially from heights in excess of 2 m) and should be particularly suspected in those with a depressed level of consciousness (from head injury) as well as in patients with chest/multiple injuries.

Biomechanical factors The anatomical differences between the adult and the paediatric spine result in a different injury profile between the two groups which lessens with development. Overall the elasticity of the spinal column results in a lower incidence of spinal column injuries when compared to adults; this elasticity does not, however, extend to the neural structures. Young children have proportionally larger heads and therefore a higher centre of gravity, resulting in relatively greater torque and acceleration forces applied to relatively underdeveloped neck musculature and relatively more lax and elastic ligaments and joint capsules. They are therefore more prone to flexion/extension injuries. In addition, the fulcrum of this force differs from the normal adult fulcrum of C5/6 and instead lies at C2/3. The stability of the adult cervical spine rests, in part, on the coronal orientation of

Figure 3 Normal lateral radiograph in a 1-year-old child showing anterior wedging of the vertebral bodies and horizontal orientation facet joints.

ARTICLE IN PRESS Cervical spine problems in children injury does not start until 10–11 years and does not equal that of the adult until late adolescence.7

Initial management All traumatised children should be suspected of having sustained a significant spinal column or cord injury and managed by standard principles namely appropriate airway management with control of the cervical spine, and assessment of the adequacy of ventilation and circulatory status.

Immobilisation Immobilisation is more complex in children than in adults where ‘sand-bags, cervical collar, tape and spinal board’ are the standard of care. To achieve a neutral alignment in the paediatric cervical spine, account must be taken of the relatively large head compared to the torso which tends to force the neck into a degree of flexion when lying on a flat surface,8 a position not prevented by application of a cervical collar. This applies to children under 8 years but particularly to those under 4 years. To counteract this angulation, the torso may need to be elevated 25 mm or the head placed in an occipital recess.

Imaging Imaging is, even to a greater degree than in the adult, a contentious issue. Particular causes of dissent are the imaging of asymptomatic children, of comatose children and those under 5 years of age. Laham et al.9 evaluated 268 children who they retrospectively divided into high risk (those complaining of neck pain and those unable to communicate because of age (o2y) or injury) and low risk for spinal injury after head injury. They used the standard trauma cervical spine series (AP, lateral and open mouth radiographs) and none of the low-risk group versus 7.5% of the high-risk group had injuries. A prospective study applying the NEXUS decision instrument to children indicated that the presence of specific risk factors (midline cervical tenderness, altered level of consciousness, intoxication, neurological abnormality or other injury such as a long bone fracture which might distract from cervical pain) rendered children at high risk of cervical spine injury, correctly identifying all children with cervical spine injury. However, none of the injured children studied were under 2 years of age and the authors were cautious about applying their study to infants and toddlers.10 Other studies have questioned the requirement for open mouth odontoid views with some authors indicating that the number of injuries missed by their omission is minimal, particularly in children under 5 years old.11,12 In addition, obtaining these views can be challenging in the young and frequently uncooperative child. Helical computed tomographic scanning (CT scanning) of the cervical spine of adult victims of trauma is gaining ground as an effective method of assessment of the bony anatomy. When reconstructed in three dimensions, the alignment is easily appreciated particularly in areas that are

277 difficult to assess by plain radiographs (e.g., C0–C2 and the cervicothoracic junction). CT is the imaging modality of choice in unconscious adults and adolescent patients. In young children, however, it is generally much easier to visualise the whole cervical spine and in addition they are particularly prone to ligamentous injury which CT will not reveal. It is reasonable to utilise CT in a similar way to its use in adults in older children who have spinal biomechanics and injury patterns similar to adults and to screen abnormal areas on radiographs to assess for fractures in younger children. Magnetic resonance imaging (MRI) has been used to diagnose the cervical spine of intubated/obtunded children and those children with equivocal plain radiographs. Dorman’s group demonstrated that MRI picked up a significant number of missed injuries which could not be evaluated on plain radiology (15 of 64 children had their management changed) and reduced costs. SCIWORA is another area of its particular diagnostic utility (vide infra).13,14 In summary, evidence supports lateral and antero-posterior cervical spine radiographs for injured children who cannot communicate because of age or injury, have a neurological deficit, neck pain, or distracting injury or who are intoxicated. Children without these risk factors are unlikely to require imaging. Open mouth views are worthwhile in children over 9 years. Fine cut (1 mm) CT scanning with 3D reconstruction is required to assess non-visualised (particularly C0–C3) or suspicious areas in younger children and may supplant plain radiographs in children at high risk of cervical spine injury. MRI is useful in non-assessable children and in children with fixed or transitory deficits.

Paediatric patterns of injury Some patterns of injury are unique in children and are particular to certain stages of development i.e., neonatal spinal cord injuries and odontoid epiphysiolysis. Other injuries that are most common in, but not unique to, children are: os odontoideum, SCIWORA, atlanto-occipital dislocation (AOD) and atlanto-axial rotatory fixation. Older children in addition present with injuries seen in the adult population and management of these pathologies is not discussed here.

Neonatal spinal cord injuries Spinal cord injury at the time of delivery is rare, occurring in around 1/60 000 live births. It occurs most frequently in the cervical or cervico-thoracic region, the former being most common and associated with rotational forceps deliveries of cephalic presentations, the latter in breech presentations. Clinically, they tend to present with apnoea and flaccid quadriplegia. Of the 22 neonates reported by MacKinnon et al.15 five of the 14 with cervical cord injury died before 3 months of age and six of the seven still alive at last followup were ventilator dependent. A smaller series from Oakes indicated that bedside diagnosis was often incorrect and in their series of five infants only one had an abnormal lateral radiograph.16

ARTICLE IN PRESS 278

Odontoid epiphysioloysis The subdental synchondrosis (neurocentral synchondrosis) is relatively vulnerable to injury in children until fused. The lateral radiograph may show abnormal angulation (usually anteriorly) or indeed frank dislocation of the dens. As it is a fracture through a physis it is very likely to heal and it can be successfully managed in a halo orthosis. If it is significantly dislocated, then reduction in the halo ring under fluoroscopic guidance is appropriate. Successful fusion rates of 80% have been described following 10–18 weeks of immobilization. Others have opted for operative management (anterior odontoid screw fixation or posterior C1–C2 fixation) although the published numbers are much smaller.17

Spinal cord injury without radiographic abnormality (SCIWORA) The SCIWORA lesion results in a sensory and/or motor deficit with no injury apparent on imaging.18 The name is somewhat misleading as in reality it refers to no evidence of bony injury on plain radiology or CT; with the advent of MRI it has become apparent that these children often have neural (haemorrhage/oedema in spinal cord) or extraneural pathological processes (traumatic disc protrusion, epidural haematomata, ligamentous injuries) which correlate with the clinical findings. SCIWORA is rare, accounting for around 4.5% of children with spinal injuries, and following serious trauma (motor vehicle accident, fall from height). It usually occurs in the cervical spine and less frequently in the thoracic spine. It presents with a complete or partial deficit although initial assessment is often clouded by haemodynamic instability. About a quarter of patients present in a delayed fashion.19 It was the incidence of delayed deterioration and of recurrent SCIWORA that has led to the suggestion that there exists in such children a state of ‘occult instability’, which renders them particularly at risk, of further deterioration from repeated stress.20 Children with abnormal or transiently abnormal symptoms should be immobilised as for an unstable injury. If there are any doubts about the normality of whole spine radiographs, then the patient should undergo high-resolution CT and 3D reconstruction, particularly of poorly visualised areas. If there is any evidence of a fracture or fracture/dislocation, then these children are managed as appropriate for that fracture. If there are no radiographic abnormalities, then MRI should be obtained with some urgency, looking for a compressive lesion (e.g., traumatic soft disc herniation, haematoma). In patients with neurological abnormalities, the MRI is almost always abnormal with respect to the neural and/or the extraneural tissues although these abnormalities may be subtle. Children who have had transient symptoms or ongoing neck pain/paravertebral spasm should be immobilised and flexion/extension views performed looking for overt instability when neck pain has settled.21 The majority of cases of SCIWORA with no compressive lesion on MRI are managed non-operatively with a short course of corticosteroids and rigid external immobilisation for 12 weeks, checking flexion/extension views after this

J. Ross, L. Myles period to rule out late instability. Outcome is ultimately related to the admission neurological status. Those admitted with complete lesions rarely improve; hence, younger children tend to do worse as they often have more severe and segmentally higher injuries. Patients with mild-tomoderate injuries or no cord changes on MRI will tend make a better recovery.

Atlanto-occipital dislocation (AOD) Traumatic AOD is a common cause of death.22 It is an injury that children are anatomically predisposed to because of the higher fulcrum and increased ligamentous laxity already described. Over the last decade, there have been around 100 reports of survivors of AOD of whom 41 have been in the paediatric age group.23,24 The stability of the craniovertebral junction in children is very reliant on the ligamentous structures as the reciprocal concavities of the atlas and convexities of the occipital condyles, a feature of the spine at maturity (48–10 years), are absent. The particular ligamentous structures of relevance are the tectorial membrane (a continuation of the posterior longitudinal ligament), and the cruciate, alar and apical ligaments. Traumatic AOD results from high-energy impacts causing rupture of the tectorial membrane and the alar ligaments, most frequently by hyperextension/rotation injuries. Patients often die at the scene or present after major trauma, often with an associated head injury, a quadriplegia and cardiopulmonary instability. There may in addition be lower cranial nerve palsies. Many will have been endotracheally intubated and ventilated making assessment of neurology very difficult. A small number of patients present without initial deficit. AOD may therefore be missed on the initial examination. These injuries have been classified into three specific types, namely anterior occipital displacement (Type I), longitudinal distraction (Type II) and posterior displacement (Type III).25 A number of methods have been devised to diagnose AOD from plain lateral radiographs; some rely upon absolute distance measured on lateral radiographs (Densbasion interval, Dublin measure), others on ratios of measurements (Powers ratio, Lee’s X-line). None of these techniques is ideal in children because visualisation of the opisthion is difficult and the odontoid process is incompletely ossified in young children. More recently, Sun et al.26 introduced the C1-2:C2-3 posterior interspinous ratio as a sensitive method of diagnosing tectorial membrane abnormalities on MRI. Cursory examination of the above assessments reveals that all of the measurements are taken from bony landmarks which do not themselves participate in the articulation of the occiput and C1. Therefore, one of the best methods of assessing the articulation is with fine cut (1–2 mm) CT through C0–C3 and coronal/sagittal reconstructions. Generally no condyle to C1 measurement is greater than 3 mm.21 MRI gives a good assessment of the joints and the ligamentous structures, as well as identifying spinal cord injury. MRI when combined with CT may allow diagnosis of partial AOD when there is bony dislocation but the tectorial membrane is intact. The tectorial membrane is probably the critical structure for ensuring stability.26

ARTICLE IN PRESS Cervical spine problems in children

279

Initial management comprises in line immobilisation followed by early application of a halo vest. Definitive management thereafter requires fusion across the whole occipito-atlantoaxial unit either with a contoured loop construct or a combined transarticular C1-2 screw fixation and suboccipital screw fixation. Both methods require bone grafting in addition.27

Atlanto-axial rotatory fixation (AARF) Fixed subluxation of the atlanto-axial complex, responsible for 40% of cervical rotation, is caused by a variety of conditions including otolaryngological infection/surgery (synonymous with Grisel’s syndrome) and is predisposed to by a variety of others (Down syndrome, juvenile chronic arthritis, the mucopolysaccharidoses (MPS)). Trauma, often insignificant, accounts for a substantial minority of cases. Patients present with neck discomfort, limited rotation and torticollis. They may assume what has been described as the ‘cock robin’ deformity with the head rotated to one side and laterally flexed to the other. Patients will be unable to rotate the head past the midline to the opposite side. Neurological deficit is rare (Fig. 4). On both plain films and static CT, it can be difficult to differentiate normal rotation from a fixed subluxation, although the value of both of these investigations is in defining associated congenital abnormalities or fracture. This problem has been studied systematically by Pang and Li.28,29 They utilised 3-mm axial CT from the clivus to the base of C3 firstly to, define the dynamic relationship between the occiput and C1 (which move as a single unit) as well as C1 and C2, in normal children. This allowed construction of a normal motion curve, which was highly reproducible between individuals. Around 01, C1 crosses C2 in either direction to start rotation, and for the first 231 of rotational movement from neutral C1 rotates without C2 moving. From 231 to 651, C2 starts to rotate progressively with C2 and by 65–901 of rotation C1 and C2 move in unison at a fixed angle of separation (mean 431 of separation). Pang and Li then went on to define the motion curves obtained in forty children presenting with painful ‘cock-robin’ deformities by imaging them in the presenting position, in a neutral rotational position and with the head forcefully turned to the opposite side as much as the patient could tolerate. They defined five groups ranging from a fixed dynamically irreducible angle between C1 and C2 (AARF I), to an abnormal but reducible C1-2 separation angle which never reached zero no matter how forceful the rotation (AARF II), to patients in whom C1 would cross C2 but only in extremes of rotation to the contralateral side (AARF III). Some of the patients had normal dynamics of motion despite the deformity and a small number fell into a diagnostic grey zone.29 The success of treatment modalities in AARF is related to the duration of the preceding period of subluxation. The quicker the intervention the more likely it is to succeed. In the paper from Pollack’s group in Pittsburgh a quarter of the children were initially treated in a collar with antiinflammatory drugs, with an 80% success rate. Of the other three-quarters treated with halter traction (approximately 2 kg) and immobilisation in a collar (for 8 weeks), successful

Figure 4 3D reconstruction of cervical spine CT in a child with atlanto-axial rotatory fixation. The left C1 articular facet is dislocated anteriorly over the C2 articular facet.

reduction was achieved in a mean of 4 days in 93% but a third of the patients had recurrent AARF and some of these children went on to require open reduction and fusion.30 The third part of the landmark series by Pang and Li assessed 50 children presenting with a painful torticollis. It subjected them to the three position CT described above; 29 were diagnosed with AARF fairly equally distributed between the three classes. All went on to halter or calliper traction on diagnosis and upon reduction were either immobilised in a cervicothoracic brace or halo jacket, those whose deformity was irreducible or recurred in a halo orthosis underwent fusion in the best possible alignment. There was a clear gradation in failure of reduction, degree of difficulty in obtaining and then maintaining reduction, and the need for operative fusion between Type I and III patients with Type II being intermediate. As in other studies, delays in diagnosis and treatment resulted in worse outcomes independent of degree.31 Patients with ‘cock-robin’ deformities should be rapidly assessed with dynamic CT, then typed and put in traction rapidly to reduce the deformity with acute cases being placed in cervicothoracic bracing. Early intervention is likely to result in improved outcome and avoidance of operative interventions. Chronic cases may require halo bracing when the deformity is reduced. Recurrent dislocation and failure of reduction are indications for operative intervention.

ARTICLE IN PRESS 280

Non-traumatic disorders of the paediatric cervical spine Primary tumours of the cervical spine Primary tumours of the cervical spine in children are extremely rare. It is therefore unlikely that any one surgeon will build up a large series of cases, and indeed such series are lacking from the literature. Fortunately, the majority of such neoplasms are benign and present in a fairly typical fashion with persistent neck pain, especially at night or when resting; neurological compromise is infrequent. Tumours may be primary tumours of bone, tumours of nonosseous bony elements or be related to diffuse neoplastic processes elsewhere (acute leukaemia and lymphoma). Plain X-rays may demonstrate lytic lesions or vertebral collapse but MRI is the most useful screening tool. Once an abnormality is demonstrated on MRI, then fine-cut computed tomography through the vertebra involved gives better definition of the anatomy.

J. Ross, L. Myles walled, blood-filled, cystic cavities and are often located in the metaphyses of long bones; they comprise about 15% of primary spine tumours with a predilection for the thoracolumbar area (Fig. 5). They tend to occur in the posterior elements extending into the pedicles and vertebral bodies. Individuals tend to present with nocturnal/recumbency pain or localised swelling and less frequently with neurological compromise. There are reported cases of rapid onset of neurological compromise after vertebral collapse. They are most frequent in the first and second decades of life (mean age at diagnosis of 14 years), with a slight female preponderance. Optimal therapy is controversial and depends upon the age of the patient, the size and accessibility of the lesion and the need to minimise intraoperative blood loss. Preoperative embolisation followed by wide excision and if necessary spinal instrumentation to maintain stability is probably the intervention of choice. Curettage alone is

Osteochondroma Osteochondromata (synonymous with exostoses) are the most common benign bone tumours although less than 3% occur in the spine. They may cause local pain but are often asymptomatic, presenting with a painless mass related to a spinous process or with symptoms of neural compression. A recent literature review identified 165 reported cases of spinal exostoses of which about half were solitary and onethird part of the multiple hereditary exostosis syndrome (diaphyseal aclasis).32 About half of the reported lesions were in the cervical spine and the majority arose from posterior elements of the vertebra giving rise to neurological symptoms. A significant proportion of cases were younger than 20 years. Complete excision is curative with a very low recurrence rate, although thought should be given to stabilisation in wide resections. Osteoid osteoma/osteoblastoma These tumours are histopathologically similar entities, both bone-producing tumours that arise in the long bones and in the posterior vertebral elements especially the neural arch. They tend to occur in adolescents and young adults. Osteoid osteomas are less than 2 cm in diameter and tend to present with localised pain, which classically, although infrequently, responds to salicylate analgesia. Cervical spine osteomas can present with torticollis or radicular pain. Radiographically both lesions have a lytic nidus surrounded by a sclerotic margin and appear ‘hot’ on bone scans. Histologically both tumours consist of osteoid trabecula containing a fibro-vascular stroma surrounded by dense cortical bone. Osteoblastomas are differentiated by size (osteoblastomas being greater than 2 cm in diameter) and have an increased predilection for the spine (30–50%) with a tendency to form expansile masses. Osteoid osteomas are curable by resection; similarly a proportion of osteoblastomas are cured by surgery although they have a significant recurrence rate. Frank malignant transformation is rare.33,34 Aneurysmal bone cyst Aneurysmal bone cysts are benign bone pseudo-tumours of unknown aetiology. They are expansile, containing thin-

Figure 5 (a) AP radiograph of destructive lesion in lateral portion of C6 vertebra, (b) Axial CT through C6 showing expansile lesion in the transverse process and posterior elements of C6. Histologically this was demonstrated to be an aneurysmal bone cyst.

ARTICLE IN PRESS Cervical spine problems in children likely to result in early progression of residual disease and the role of radiotherapy is controversial.35,36 Giant cell tumour Giant cell tumours are expansile, vascular lesions that account for less than 2% of all vertebral tumours. They are rapidly progressive, aggressive tumours and often present with pain and neurological compromise. They are most often found in the vertebral body (especially in the sacrum) and appear radiolucent and expansile. Aggressive surgical treatment, often requiring multiple surgical approaches, is the usual requirement with recurrence rates of 8–16% after complete excision and bone grafting. Radiotherapy may have a role for tumours in which complete resection is not possible (Fig. 6).37 Langerhans cell histiocytosis (LCH) and eosinophilic granulomata Eosinophilic granulomata arise from the reticuloendothelial system and are a localised form of LCH. LCH is a challenging disease which presents in a spectrum of disorders ranging from a spontaneously regressing solitary bone lesion to a multisystem and life-threatening disorder. The incidence of LCH is 2–5/million/annum and is probably underdiagnosed; the peak age is between 1 and 4 years with a slight male preponderance.

Figure 6 Midsagittal MRI showing expansile space occupying lesion involving the C4 spinous process. Histologically this was demonstrated to be a giant cell tumour.

281 Bone lesions are found in both localised and disseminated forms of the disease and are frequently apparent in the vertebra where they are typically osteolytic lesions which may progress to a collapsed vertebral body (vertebra plana). These lesions are frequently asymptomatic or can present with localised pain. Treatment depends upon the extent of disease, and for those with limited disease (i.e., bone disease without visceral involvement) the prognosis is excellent. Therefore, appropriate staging of the disease is vital and children should be referred to a paediatric oncologist and screened radiologically. Treatment for localised bone disease is primarily conservative with rest, analgesia and bracing; over time the vertebra often reconstitutes. Some recommend curettage of lesions and corticosteroid instillation.38

Congenital anomalies of the cervical spine There are a large number of congenital malformations of the cervical spine and craniocervical junction, the detail of which are beyond the scope of this review. However, some of the more common and illustrative have been selected for discussion.

Down syndrome Trisomy 21, or Down syndrome, is the commonest inherited chromosomal disorder occurring in 1 in 660 live births. Children and adults with Down syndrome are prone to cervical spine disease, including instability, although the natural history is unclear. The majority of pathology occurs at the craniocervical junction particularly at the atlantoaxial and atlanto-occipital joints. The reason for this increased incidence is two-fold: abnormal anatomy and the effects of ligamentous laxity. At maturity, the atlanto-occipital joint is stabilised by the reciprocal joint surfaces of the occipital condyles on the superior C1 articular surface, the associated joint capsules, atlanto-occipital membranes and the tectorial membrane. Down syndrome patients have variable joint morphology with flat or ‘rocker-bottom’ joints and lax ligaments. At the atlanto-axial joint where stability relies on the bony integrity of the dens and the transverse ligament, ligamentous laxity again predisposes to instability, further exacerbated by the significant incidence of bony abnormalities of the dens (os odontoideum, hypoplasia and abnormal ossification of the arch of C1). One of the major areas of controversy is the demarcation between instability i.e., pathological intersegmental motion which threatens neural integrity and hypermobility i.e., increased mobility not indicative of loss of structural integrity that is known to occur in trisomy 21. The standard measurements used to indicate instability may not apply in the patients with Down syndrome. Atlanto-occipital hypermobility is observed in more than 60% of individuals with Down syndrome although it is not usually associated with neurological risk.39 If hypermobility exists in the context of bony anomalies of C1 or the skull base then there is an increased risk of neurological compromise, which usually co-exists with atlantoaxial instability.40

ARTICLE IN PRESS 282

J. Ross, L. Myles

The majority of studies of this population have concentrated on atlanto-axial instability, the incidence of which is variously reported as between 10 and 30% depending on the study methodology. Pueschel and Scola41 found an incidence of 14.6% based upon measurements of the ADI in flexion and extension in 404 children with Down syndrome; of these children only six were symptomatic (1.5% of the cohort). The ADI (distance from the back of the arch of C1 to the front of the dens) is normally less than 4 mm in children and gives indirect information about the neural canal width. When the ADI is 10 mm or more it is likely that the cord is significantly compromised (Fig. 7). This method of screening has been criticised and is probably useful only as a guide as to which patients should go on to MR imaging looking for evidence of cord pathology.42 Children tend to present with gait abnormalities, diminished exercise tolerance and occasionally neck pain or with a spinal cord injury. Posterior cervical fusion either instrumented or with autologous bone graft is warranted after reducibility has been established. The occiput should be incorporated into the construct when there is, in addition, significant atlanto-occipital subluxation, a congenital osseous abnormality of the C1 ring following transoral odontoidectomy for ventral compressive pathology.43 Anomalies of the odontoid process/os odontoideum Congenital anomalies of the odontoid process include aplasia (its complete absence), hypoplasia (its partial absence) and os odontoideum, defined as an ossicle with smooth circumferential cortical margins lying cephalad to, but without osseous continuity with, the body of C2. It can be described as orthotopic when the ossicle moves with the ring of C1, and dystopic when functionally fused to the basion. It is debatable whether os odontoideum should be included in a section on congenital disorders of the spine. Although it is found with increased frequency in children with congenital disorders of the spine characterised by instability at the craniovertebral junction e.g., Down syndrome and Morquio syndrome, there is evidence of a

Figure 7

traumatic origin for the condition; there are, for example, well-documented cases of normal odontoid processes being replaced by an os on later examination. Crockard and Stevens44 have suggested that the instability secondary to the underlying congenital condition results in fracturing of the dens and the subsequent interposition of the lax transverse ligament results in an established non-union and an unstable atlanto-axial joint. Children tend to present asymptomatically after screening radiographs or symptomatic with neck pain, loss of neck movement or severe neurological injury after relatively minor trauma. Asymptomatic patients without symptoms or signs of neurological compromise may be managed conservatively and followed closely. Symptomatic individuals require posterior C1–C2 fixation and fusion with postoperative Halo fixation or alternatively transarticular screw fixation.45

Klippel–Feil syndrome (KFS) and disorders of segmentation KFS describes a heterogeneous cohort of patients unified only by the presence of a failure of normal segmentation of the cervical spine (cervical synostosis). The classic triad of short neck, low posterior hairline and limited neck movement is seen in fewer than 50% of people with a KF type anomaly. A wide variety of congenital anomalies are associated with the finding of failure of cervical segmentation including congenital scoliosis, rib abnormalities, deafness, cardiac and genitourinary abnormalities and Sprengel’s deformity (an abnormally elevated and hypoplastic scapula). In some children, the KF anomaly may be associated with another eponymous syndrome e.g., Goldenhar syndrome (synonymous with hemifacial microsmia and oculoauricular dysplasia) or Wildervanck syndrome (synonymous with cervico-oculoacoustic syndrome). A minority of cases of KFS have a genetic inheritance basis with both autosomal dominant and recessive variants.46 Excitingly KFS has been linked to Sprengel deformity and Chiari malformations as being primarily a defect in post-otic neural crest cell fate

Axial MRI demonstrating a large atlanto-dental interval in a child with Down syndrome.

ARTICLE IN PRESS Cervical spine problems in children

283

Figure 8 (a) Lateral radiograph showing asymptomatic fusion of C4 and C5. (b) Mid sagittal MRI demonstrating asymptomatic C2–C3 fusion.

choices i.e., ectopic ossification or connective tissue formation.47 KFS occurs in 1:40 000 live births with a slight female preponderance. Clinical features are determined by the pathoanatomy of the fusion with more rostral and more extensive fusions tending to present earlier with cosmetic complaints, neck pain or cervical radicular or myelopathic symptoms. Alternatively asymptomatic KFS may be noted during assessment of other congenital anomalies (Fig. 8). Three specific patterns of fusion are of particular concern as they connote high risk for symptomatic instability: C2–C3 fusion with occipitalisation of the atlas, extensive cervical fusion with an abnormal occipitocervical junction and two fused segments with an interposed open joint space. In each instance, congenital fusion of motion segments predisposes to altered biomechanics at mobile segments and predisposition to pathological instability and neurological compromise. Patients with symptomatic instability or neurological compromise are candidates for an appropriate surgical fusion.48

Osteochondral dysplasias Osteochondral dysplasias are defined as abnormalities of bone or cartilage growth and remodelling in development. They are a heterogeneous group of heritable disorders with more than 120 having been identified. Achondroplasia and spondyloepiphyseal dysplasia (SED) and its subtypes are the most frequently encountered. Achondroplasia is a common form of disproportionate dwarfism with autosomal dominant (20%) and sporadic (80%) inheritance with an incidence of 1 in 26 000 live births. Generalised spinal stenosis with spinal cord compression can occur which may require operative decompression in child-

hood; more frequently reported, however, is cervicomedullary compression. Ryken and Menezes49 described six achondroplastic children presenting with pain, ataxia, incontinence, recurrent apnoeic spells and respiratory arrest. Radiological evaluation revealed marked foramen magnum stenosis, ventrolateral cervicomedullary junction compression secondary to basilar invagination and dorsal cervicomedullary junction compression secondary to ligamentous hypertrophy and invagination of the posterior atlantal arch. All improved with dorsal decompression without fixation. In a prospectively evaluated series of 11 children, two had radiological evidence of cervicomedullary damage on MR1 and underwent immediate decompression; two underwent interval decompression after developing signs of compression. All 11 children were asymptomatic at a mean of 4.6 years.50 Simple posterior decompression without fusion may predispose to deformity and simultaneous fusion may be most appropriate. SED is a term for a group of conditions resulting in short trunk disproportionate dwarfism. SED congenita (SEDC) is complicated by ligamentous laxity often combined with odontoid hypoplasia or os odontoideum. This, as in Down syndrome, can result in atlanto-axial subluxation and myelopathy. The incidence of myelopathy may be as high as 35% in SEDC and symptomatic myelopathy requires posterior decompression and fusion.51

Mucopolysaccharidoses (MPS) The MPS are a heterogeneous group of lysosomal storage disorders each caused by a deficiency in an enzyme involved in degradation of glycosaminoglycans, resulting in cell, tissue and organ dysfunction. Children with both Morquio

ARTICLE IN PRESS 284

J. Ross, L. Myles

Figure 9 (a) Lateral radiograph demonstrating craniocervical instability and os odontoideum in Morquio syndrome. (b) Sagittal MRI demonstrating the thickened posterior longitudinal ligament found in Maroteaux–Lamy syndrome.

syndrome (MPS IV) and Maroteaux–Lamy syndrome (MPS VI) are at risk of cervical myelopathy secondary to craniovertebral junction abnormalities. In Morquio syndrome, children have an os odontoideum secondary to ligamentous laxity with repeated trauma retarding ossification; in addition there is a stenotic foramen magnum secondary to invagination of the posterior rim of C1 and a soft tissue mass of reactive tissue around the unossified dens (Fig. 9a). This results in ventral and posterior compression and a myelopathy.52,53 Craniovertebral anomalies in MPS VI are less common than in Morquio’s syndrome and atlanto-axial subluxation is rare. They too suffer from foramen magnum stenosis, in this syndrome secondary to marked thickening of the posterior longitudinal ligament (Fig. 9b).54

Acknowledgements

5. 6. 7.

8. 9.

10. 11.

Thanks to Dr. R Gibson of the Department of Radiology in the Western General Hospital and Messrs Cowie and Thorne of the Department of Neurosurgery in the Royal Manchester Children’s Hospital.

References 1. David KM, Crockard A. Early development of the craniovertebral junction and the cervical spine. In: Crockard A, Hayward R, Hoff JT, editors. Neurosurgery. The scientific basis of clinical practice. Oxford: Blackwell; 2000. p. 74–86. 2. Lustrin ES, Karakas SP, Orlando Ortiz AO, Cinnamon J, Castillo M, Vaheesan K, et al. Paediatric cervical spine: normal anatomy, variants and trauma. Radiographics 2003;23:539–60. 3. Catell HS, Filtzer DL. Pseudosubluxation and other normal variations in the cervical spine in children: a study of one hundred and sixty children. J Bone Joint Surg Am 1965;47: 1295–309. 4. Avellino AM, Mann FA, Grady MS, Chapman JR, Ellenbogen RG, Alden TR, et al. The misdiagnosis of acute cervical spine injuries

12.

13.

14.

15.

16. 17.

18.

and fractures in infants and children: the 12-year experience of a level 1 paediatric and adult trauma centre. Childs Nerv Syst 2005;21(2):122–7. Martin BW, Dykes E, Lecky FE. Patterns and risks in spinal trauma. Arch Dis Child 2004;89:860–5. Patel JC, Tepas JJ, Mollitt DL, Pieper P. Paediatric spine injuries: defining the disease. J Pediatr Surg 2001;36(2):373–6. Kokoska ER, Keller MS, Rallo MC, Weber TR. Characteristics of paediatric cervical spine injuries. J Pediatr Surg 2001;36(1): 100–5. Nypaver M, Treloar D. Neutral cervical spine positioning in children. Ann Emerg Med 1994;23:208–11. Laham JL, Cotcamp DH, Gibbons PA, Kahana MD, Crone KR. Isolated head injuries versus multiple trauma in paediatric patients: do the same indications apply? Pediatr Neurosurg 1994;21:221–6. Viccellio P, Simon H, Pressman BD, Shah MN, Mower WR, Hoffman JR, for the Nexus Group. Pediatrics 2001;108(2):E20. Buhs C, Cullen M, Klein M, Farmer D. The paediatric trauma cspine: is the odontoid view necessary? J Pediatr Surg 2000;35:994–7. Swischuk LE, John SD, Hendrick EP. Is the open mouth view necessary in children under 5 years? Pediatr Radiol 2000;30: 186–9. Flynn JM, Closkey RF, Mahboubi S, Dormans JP. Role of magnetic resonance imaging in the assessment of paediatric cervical spine injuries. J Pediatr Orthop 2002;22:573–7. Frank JB, Lim CK, Flynn JM, Dormans JP. The efficacy of magnetic resonance imaging in paediatric cervical spine clearance. Spine 2002;27(11):1176–9. MacKinnon JA, Perlman M, Kirpalani H, Rehan V, Suave R, Kovacs L. Spinal cord injury at birth: diagnostic and prognostic data in 22 patients. J Pediatr 1993;122:431–7. Rossitch Jr E, Oakes WJ. Perinatal spinal cord injury. Pediatr Neurosurg 1992;18:149–52. Mandabach M, Ruge JR, Hahn YS, McGlone DG. Pediatric axis fractures: early halo immobilisation, management and outcome. Pediatric Neurosurg 1993;19:225–32. Pang D, Wilberger JE. Spinal cord injury without radiographic abnormality in children. J Neurosurg 1982;57:114–29.

ARTICLE IN PRESS Cervical spine problems in children 19. Launay F, Leet AI, Sponseller PD. Paediatric spinal cord injury without radiographic abnormality: a meta-analysis. Clin Ortho Rel Res 2005;433:166–70. 20. Pang D, Pollack IF. Spinal cord injury without radiographic abnormality in children-the SCIWORA syndrome. J Trauma 1989;29:654–64. 21. Pang D, Sun PP. Pediatric vertebral column and spinal cord injuries. In: Winn RH, editor. Youmans neurological surgery. Philadelphia: Elsevier; 2004. p. 3515–57. 22. Bucholz RW, Burkhead WZ. The pathological anatomy of fatal atlanto-occipital dislocations. J Bone Joint Surg Am 1997;61: 248–50. 23. Hadley M. Diagnosis and management of traumatic atlantooccipital dislocation injuries. Neurosurgery 2002;50(Suppl. 3):S105–13. 24. Steinmetz MP, Lechner RM, Anderson JS. Atlanto-occipital dislocation in children: presentation, diagnosis and management. Neurosurg Focus 2003;14(2):1–7. 25. Traynelis VC, Marano GD, Dunker RO, Kaufman HH. Traumatic atlanto-occipital dislocation. Case report. J Neurosurg 1986;65(6):863–70. 26. Sun PP, Poffenbarger GJ, Durham S, Zimmerman RA. Spectrum of occipito-atlantoaxial injury in young children. J Neurosurg 2000;93(Spine 1):28–39. 27. Brockmeyer DL, Apfelbaum RI. A new occipitocervical fusion construct in pediatric patients with occipito-cervical instability. Technical note. J Neurosurg 1999;90(Spine 2):271–5. 28. Pang D, Li V. Atlantoaxial rotatory fixation: part 1—biomechanics of normal rotation at the atlantoaxial joint in children. Neurosurgery 2004;55:614–26. 29. Pang D, Li V. Atlantoaxial rotatory fixation: part 2—new diagnostic paradigm and a new classification based on motion analysis using computed tomographic imaging. Neurosurgery 2005;57:941–53. 30. Subach BR, McLaughlin MR, Albright AL, Pollack IF. Current management of atlantoaxial rotatory subluxation. Spine 1998;23(20):2174–9. 31. Pang D, Li V. Atlantoaxial rotatory fixation: part 3—a prospective study of the clinical manifestation, diagnosis, management, and outcome of children with fixed atlantoaxial rotatory fixation. Neurosurgery 2005;57:954–72. 32. Bess RS, Robbin MR, Bohlmann HH, Thompson GH. Spinal exostoses: analysis of twelve cases and review of the literature. Spine 2005;30(7):774–80. 33. Lucas DR, Unni KK, McLeod RA, O’Connor MI, Sim FH. Osteoblastomas: clinicopathological study of 306 cases. Hum Pathol 1994;25:117–34. 34. Zileli M, C - agli S, Basdemir G, Ersahin Y. Osteoid osteomas and osteoblastomas of the spine. Neurosurg Focus 2003;15(5): 1–6. 35. Liu JK, Brockmeyer DL, Dailey AT, Schmidt MH. Surgical management of aneurysmal bone cysts of the spine. Neurosurg Focus 2003;15(5):E4.

285 36. Garg SM, Mehta S, Dormans JP. Modern surgical treatment of primary aneurysmal bone cyst of the spine in children and adolescents. J Pediatr Ortho 2005;25(3):387–92. 37. Hart RA, Boriani S, Biagini R, Currier B, Weinstein J. A system for surgical staging and management of spine tumours: a clinical outcome study of giant cell tumours of the spine. Spine 1997;22(15):1773–82. 38. Egeler RM, D’Angio GJ. Medical progress Langerhans cell histiocytosis. J Pediatr 1995;127(1):1–11. 39. Tredwell SJ, Newman DE, Lockitch G. Instability if the upper cervical spine in Down syndrome. J Pediatr Orthop 1990;10(5): 602–6. 40. Brockmeyer D. Down syndrome and craniovertebral instability. Pediatr Neurosurg 1999;31:71–7. 41. Pueschel SM, Scola FM. Atlantoaxial instability in individuals with Down syndrome: epidemiologic, radiographic and clinical studies. Pediatrics 1987;80(4):555–60. 42. Pizzutillo PD, Herman MJ. Cervical spine issues in Down syndrome. J Pediatr Orthop 2005;25(2):253–9. 43. Nader-Sephai A, Casey ATH, Hayward R, Crockard HA, Thompson D. Symptomatic atlantoaxial instability in Down syndrome. J Neurosurg (Pediatrics) 2005;103:231–7. 44. Crockard HA, Stevens JM. Craniovertebral junction anomalies in inherited disorders: part of the syndrome or caused by the disorder? Eur J Pediatr 1995;154:504–12. 45. Hadley M. Os odontoideum. Neurosurgery 2002;50(Suppl. 3):S148–55. 46. Clarke RA, Catalan G, Diwan AD, Kearsley JH. Heterogeneity in Klippel–Feil syndrome: a new classification. Pediatr Radiol 1998;28:967–74. 47. Matsuoka T, Ahlberg PE, Kessaris N, lannarelli P, Dennehy U, Richardson WD, et al. Neural crest origins of the neck and shoulder. Nature 2005;436:347–55. 48. Tracy MR, Dormans JP, Kusumi K. Klippel–Feil syndrome: clinical features and current understanding of aetiology. Clin Ortho Rel Res 2004;424:183–90. 49. Ryken TC, Menezes AH. Cervicomedullary compression in achondroplasia. J Neurosurg 1994;81:43–8. 50. Keiper GC, Koch B, Crone KR. Achondroplasia and cervicomedullary compression: prospective evaluation and surgical treatment. Pedialr Neurosurg 1999;31:78–83. 51. Miyoshi K, Nakamura K, Haga N, Mikami Y. Surgical treatment for atlantoaxial subluxation with myelopathy in spondyloepiphyseal dysplasia congenita. Spine 2004;21:E488–91. 52. Ransford AO, Crockard HA, Stevens JM, Modaghegh S. Occipitoatlanto-axial fusion in Morquio–Brailsford syndrome a ten year experience. J Bone Joint Surg [Br] 1996;78-B:307–13. 53. Hughes DG, Chadderton RD, Cowie RA, Wraith JE, Jenkins JPR. MRI of the brain and craniocervical junction in Morquio’s disease. Neuroradiology 1997;73:381–5. 54. Thorne JA, Javadpour M, Hughes DG, Wraith E, Cowie RA. Craniovertebral abnormalities in Type VI mucopolysaccharidosis (Maroteaux–Lamy Syndrome). Neurosurgery 2001;48(4):849–52.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 286–293

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: CHILDREN’S ORTHOPAEDIC SURGERY

(v) The hip in cerebral palsy Alastair W. Murraya,, James E. Robbb a

Royal Hospital for Sick Children, Yorkhill NHS Trust, Dalnair Road, Glasgow, G3 8SJ, UK Royal Hospital for Sick Children, 9 Sciennes Road, Edinburgh, EH9 1LF, USA

b

KEYWORDS Hip; Cerebral palsy; Dislocation; Surveillance; Gait

Summary Deformity and displacement of the hip is the second most common orthopaedic problem to affect children with cerebral palsy. The severity and incidence of hip pathology increases with the severity of cerebral palsy and up to 70% of patients with total body involvement cerebral palsy suffer from hip displacement. Hip deformity can also be problematic in patients who walk as it results in rotational malalignment of the lower limb in gait. Awareness of the problems and appropriate screening permits timely intervention. This may involve management of muscle tone or soft tissue procedures. In more advanced cases of hip displacement and deformity, femoral and pelvic osteotomies are usually required. The goals of treatment range from improving the efficiency of gait to prevention of severe postural problems and pain in the most severely affected. & 2006 Elsevier Ltd. All rights reserved.

Introduction

Aetiology and incidence

Cerebral palsy is a non-progressive abnormality of the central nervous system resulting in an impairment of motor function. By definition, the insult to the central nervous system should occur before the age of two years but the manifestations of the condition can result in progressive musculoskeletal pathology throughout life. Cerebral palsy may also be associated with other disabilities such as speech disorders, visual disturbance, epilepsy and intellectual impairment. The orthopaedic surgeon should always bear in mind that surgery is only managing one aspect of the disorder and that objectives of treatment often differ with every patient.

Hip displacement is the second most common deformity after equinus. These children have normal hips at birth but the hip displaces under the influence of abnormal muscle balance and tone (Fig. 1). The problem is therefore quite distinct from developmental dysplasia of the hip (DDH) (Table 1). It is thought that increased tone in the hip flexors and adductors overcomes the relatively weaker extensors and abductors resulting in transfer of the centre of hip rotation from the centre of the femoral head to the lesser trochanter. The abnormal forces prevent the normal development of the acetabulum and proximal femoral geometry. In unaffected children the acetabulum normally deepens around a concentrically located femoral head during the first 6 years of life resulting in an acetabular index (AI) of less than 301. In cerebral palsy the AI is commonly found to be higher than this. Multi-planar

Corresponding author. Tel.: +44 141 201 0275.

E-mail addresses: [email protected] (A.W. Murray), [email protected] (J.E. Robb). 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.06.011

ARTICLE IN PRESS The hip in cerebral palsy CT 3-D reconstruction has also shown that both anterior and posterior acetabular cover is reduced in people with cerebral palsy.1 Chung et al.2 found that posterior defects were most common in subluxed hips whereas global defects predominated in dislocated hips. These acetabular deformities are thought to be due to eccentric pressure from the femoral head on the soft, cartilaginous acetabular anlage at the periphery of the developing acetabulum.

Figure 1 Progressive, bilateral hip subluxation in a patient with cerebral palsy.

Table 1 (DDH).

287 The effects of abnormal tone on the proximal femur are equally detrimental. The neonate usually has a neck-shaft angle (NSA) of 1401 that reduces to a mean of 1201 in the adult. At birth femoral neck anteversion is about 401 and gradually reduces to about 12–151 at skeletal maturity. These changes may not occur in cerebral palsy and a persistent femoral anteversion of greater than 301 and NSA of 140–1501 is often found. The mildest manifestation of persistent anteversion is internal rotation of the lower limb. This can also prove problematic for the patient who walks as it is one of the causes of an internally rotated foot progression angle in gait. Femoral anteversion can be estimated clinically using the method of Ruwe et al.3 This can be assessed with the patient either prone (walkers) as part of the assessment of the rotational profile4 or supine (non-walkers). With the knee flexed to 901, the limb and the hip is internally rotated until the greater trochanter is felt at its maximum prominence maximally over the lateral aspect of the upper thigh. The angle of the shank from the vertical corresponds to the degree of femoral anteversion (Fig. 2). A CT scan comparing the femoral anteversion with the transcondylar axis of the knee can also be used, but carries the disadvantage of exposure to radiation. There is, however, a relatively poor correlation between static

Figure 2 Clinical assessment of femoral neck anteversion of the right hip. The greater trochanter is at its maximum prominence on the lateral aspect of the thigh.

Summary of differences between hip displacement in cerebral palsy (CP) and developmental dysplasia of the hip

CP

DDH

Stable at birth Dislocation occurs years later Bracing ineffective Dislocation usually postero-superior Part of systemic condition

Unstable at birth Dislocation pre/perinatal Bracing mainstay of early treatment Dislocation usually antero-superior Usually isolated problem

ARTICLE IN PRESS 288

A.W. Murray, J.E. Robb

measurements and the dynamic situation during gait.5 Planning for surgical intervention in patients who walk can be greatly enhanced by preoperative 3-D gait analysis to identify the anatomical level(s) which is/are contributing to a pathological foot progression angle in gait. Excessive femoral anteversion in the walker can be treated by derotation femoral osteotomy. In the non-walker the effect of abnormal muscular forces on the immature skeleton may produce hip subluxation and dislocation. The great majority displace in a posterosuperior direction and a small minority anteriorly. A dislocated hip is often associated with difficulties with access to perineal hygiene and seating. The association between scoliosis, pelvic obliquity and hip dislocation is not clear and prevention of hip dislocation may not prevent development of scoliosis.6 Equally, hip dislocation also occurs in patients who have a level pelvis. The ‘‘windswept’’ deformity, where there is adduction and internal rotation of the hip on one side and abduction and external rotation on the other, may also make seating very difficult. Pain may be a problem and carers of patients may report that the patient is distressed when being dressed or undressed, seated or having a nappy changed. Although degenerate joint disease (DJD) does develop in displaced hips some investigators have found no correlation between the degree of DJD and pain7 while other have suggested there is always a link.8 From the literature it would appear that more than 50% of patients with a subluxed or dislocated hip complain of hip pain, which can be severe.9 Prevention of DJD in the longer term is also a reasonable additional indication to the presenting difficulties of pain, seating and perineal access. About 3–7% of patients who walk develop hip displacement. A recent study has shown a linear relationship between hip displacement and the Gross Motor Function Classification System (GMFCS) score10 (Table 2). The association is not surprising as the inability to walk is often

Table 2

associated with severe spasticity which is also highly undesirable for normal hip development. It would however, appear that walking itself and avoidance of prolonged sitting is protective for hip development (Table 3).

Surveillance for hip displacement Hip subluxation and eventual dislocation is a gradual process and dislocation does not often occur before the age of six, even in severely affected individuals. For the reasons mentioned above, patients with total body involvement cerebral palsy and non-walkers should be regularly screened clinically and radiologically. Hip abduction of less than 301 and hip flexion contracture are signs of a hip at risk but clinical examination alone is unreliable. Most clinicians would request an AP X-ray of the pelvis 12 monthly before the age of six to look for signs of subluxation. It has been recommended that the first X-ray is taken before the age of 18 months as hip displacement can begin before that age.11 It is important to obtain a true AP of the hip. Figure 3 appears to show excessive femoral neck valgus but much of this appearance is due to excessive femoral anteversion. A break of Shenton’s line, a progressive decrease of the centre-edge angle of Wiberg or an increasing AI are all used as signs of hip subluxation. Since the centre-edge angle is unreliable under the age of five and the AI is affected by pelvic tilt and rotation, many clinicians use the migration percentage (MP)12 which quantifies the lateral displacement of the hip on an AP film and expresses it as a percentage. An MP of greater than 25–30% (normal 10%) or an increase of more than 10% in one year have been used as criteria for intervention.11 Some studies have shown the MP to have a relatively good interobserver reliability but others have been less supportive. It certainly has the disadvantage that it does not assess anterior or posterior displacement and can also be affected by the patient’s position at the time of the

The gross motor function classification system (GMFCS).19

Level

Function

1 2 3 4 5

Can climb stairs, run and jump but balance and coordination impaired Can walk unassisted but uses railing to climb stairs and experiences difficulty on uneven surfaces Can walk on level surfaces with walking aids and manually propel a wheelchair Relies upon wheeled mobility at home but may walk short distances with a walker No independent mobility. Loss of antigravity head and trunk control

Table 3

Key points.

Hip displacement in cerebral palsy Affects 28% of people with cerebral palsy Risk of hip problems increases with increasing severity of cerebral palsy Up to 75% of patients with total body involvement have hip displacement Neonatal femoral anteversion and neck shaft angle do not usually decrease with time AP pelvis X-ray surveillance recommended for patients at risk of hip displacement Skeletal surgery usually requires pelvic and femoral osteotomy Retain realistic objectives for treatment, tailored to each patient

ARTICLE IN PRESS The hip in cerebral palsy

289 overcome fixed contractures. Some studies have shown an improvement in MP after SDR but this has tended to occur in less affected patients; worsening of the MP has been reported in patients with a higher GMFCS score.13 Its role in prevention of hip displacement therefore remains uncertain and the procedure is not often practiced in the UK.

Adductor and psoas lengthening

Figure 3 AP of a hip with increased anteversion giving the appearance of marked coxa valga. Note also the flattening of the medial aspect of the femoral head and the saucer shape of the acetabulum which have occurred as the femoral head has gradually displaced.

X-ray. Even with these limitations the MP is in widespread use and avoids the greater doses of radiation from CT scans. Walkers with spastic diplegia are often included in the higher risk group for surveillance particularly if their spasticity is severe. Patients with hemiplegia or predominantly athetoid or ataxic cerebral palsy can be considered low risk for hip displacement. Most hips displace before the age of ten and continuing screening for high-risk patients until this age seems advisable. A small number of hips displace during adolescence and carers should be made aware of this possibility.

Orthopaedic management of the hip in the nonwalker Tone management The underlying problem is increased muscle tone and there are several ways of reducing this. Botulinum toxin inhibits the release of acetyl choline at the neuromuscular junction. In very young patients botulinum injections to the hip adductors has been shown to result in a reduction of the MP. The main disadvantage of botulinum in the treatment of hip displacement is its transient effect so that it is unlikely to be successful as a sole treatment for this condition. Reduction of severe spasticity with intrathecal baclofen has been shown to prevent progression or even reduce the MP in some patients. Its more prolonged effect makes this a potentially more effective treatment than botulinum toxin. Its long term effect on hip displacement is yet to be evaluated. Selective dorsal rhizotomy (SDR) is a procedure where nerve rootlets between L2-S1 are sectioned if they demonstrate abnormal responses to electrical stimulation. This has been shown to reduce muscle tone but does not

An intramuscular tenotomy of adductor longus and myotomy of gracilis can be effective in protecting the hip from displacement in young non-walkers who have an adduction contracture, but no fixed hip flexion deformity.14 More extensive procedures to the adductors should be avoided as they may lead to a very disabling fixed abduction (Fig. 4). Psoas tenotomy may be included if there is a hip flexion contracture. Surgery should therefore be performed while the acetabulum retains the ability to remodel and before a deformity is visible on an X-ray.

Bony surgery Management of the displaced hip will usually entail pelvic and femoral osteotomies to correct abnormal femoral rotation, neck-shaft angle and the acetabular deformity. Femoral varus derotation shortening osteotomy Femoral derotation varus osteotomy, without a pelvic procedure, has been shown to give poorer long term results than when a pelvic procedure has been included.15 The aim of the osteotomy is to shorten the femur to facilitate reduction of the femoral head by lowering soft tissue tension, to derotate it to give physiological anteversion of about 151 and a neck shaft angle of about 1201. Excessive varus can produce excessive adduction and should be avoided. A closing wedge osteotomy is usually performed at the intertrochanteric or subtrochanteric level. The psoas tendon should be lengthened or divided. One advantage of performing the osteotomy at the intertrochanteric level is that the lesser trochanter is excised and the psoas tendon divided as a part of the procedure. It is also possible to include flexion or extension into the femoral osteotomy. The osteotomy and fixation with a blade plate will now be described.

Figure 4 Fixed abduction following excessive adductor release. Note that the patient has to be supported to sit.

ARTICLE IN PRESS 290 The patient is placed on standard operating table and a small rolled up towel placed under the buttock of the affected hip. A sandbag should not be used as this will obscure X-rays. A true AP X-ray of the hip is taken using an image intensifier. Usually this is obtained with the hip is maximally internally rotated. This will give the desired angle of varus correction; for example if the NSA is 1401 and a 1201 NSA is required, a 201 wedge should be excised. The proximal femur is then exposed subperiosteally. The greater trochanteric apophysis, if open, should be identified and safeguarded. We prefer to reflect the vastus lateralis

Figure 5 (a) Intraoperative radiograph showing the use of a swan neck chisel to create the pelvic osteotomy. The femoral osteotomy has already been performed. (b) Post operative radiograph showing the pelvic osteotomy held open after insertion of the bone graft.

A.W. Murray, J.E. Robb anteriorly rather than splitting it. If, for example, 201 of varus is required a guide wire is passed into the femoral neck under radiographic control at 201 to the transverse plane of femoral shaft on the AP view. It should be introduced proximal to the trochanteric apophysis to leave room for the seating chisel in the neck and positioned centrally in the neck on the lateral view. This is checked by taking a frog lateral view of the hip. A seating chisel is then used to cut a parallel track to the guide wire. The femur is then divided by ensuring that the saw blade is parallel to the seating chisel. It is essential to ensure that there is a sufficient bone between the chisel and the saw cut; otherwise there is risk of insufficient purchase of the blade within the femoral neck. The distance between the chisel and saw cut will depend on the size of blade plate selected but, as a guide, there should be a minimum of 1–1.5 cm of bone. A second transverse osteotomy is then performed at the level of the lesser trochanter. Shortening of the femur will

Figure 6 (a) Preoperative radiograph of a dislocated hip in a patient with cerebral palsy. (b) Follow up AP radiograph at 1 year after pelvic and femoral osteotomy showing the hip reduced, and well covered.

ARTICLE IN PRESS The hip in cerebral palsy occur as a wedge has been created, but more shortening can be added by making the transverse cut more distal. The blade of the blade plate is then introduced into the proximal femur and the shaft approximated to the plate and the two cut surfaces of the femur reduced. The amount of derotation can now be determined by relating the inclination of the guidewire, which should have been placed in the middle of the femoral neck on the lateral X-ray, to the transcondylar axis of the knee visually. As an approximate guide, the patella should be pointing vertically.

291 The lateral wall of the ilium is exposed using a SmithPetersen approach via a bikini incision. A radiolucent retractor is placed subperiosteally in the sciatic notch. The hip capsule can be opened the hip inspected and, if necessary, the elongated ligamentum teres excised and the transverse acetabular ligament incised. This is not always

Pelvic osteotomy While the acetabulum is dysplastic it is incorrect to consider its pathology as the same as in DDH (Table 1) as the hip in CP is usually normal at birth. In DDH there is commonly an antero-superior deficiency whereas in cerebral palsy the deficiency is more commonly postero-superior. For this reason, the uni-directional osteotomies, such as Salter’s osteotomy which enhances antero-superior cover, should be avoided as it risks uncovering the hip posteriorly which is a potential disadvantage for patients who will be life long sitters. While the Pemberton pericapsular osteotomy has some similarities it can provide better superior cover than the Salter and is used in cerebral palsy patients,16 particularly for the much less common anterior hip dislocation in CP. The Dega periacetabular osteotomy shares the same advantages and disadvantages as the Pemberton but has had equally good long term results reported.17 The Dega type of periacetabular osteotomy will now be described.

Figure 7 Lateral radiograph showing hetertopic ossification around the hip after femoral and pelvic osteotomy.

Figure 8 (a) AP radiograph of the pelvis showing longstanding, bilateral hip dislocation, severe pelvic obliquity and scoliosis. (b) AP radiograph of the pelvis in the same patient following bilateral proximal femoral resection. (c) Clinical photograph of the excised proximal femora showing gross dysplasia.

ARTICLE IN PRESS 292

A.W. Murray, J.E. Robb

Figure 9 Clinical photographs of the same patient with total body involvement before and after hip surgery. This resulted in comfortable seating in a wheelchair which was impossible before surgery.

necessary. The osteotomy of the lateral wall of the ilium passes from the anterior inferior iliac spine, where a 1 cm bicortical cut of the ilium is made, around the supraacetabular bone in a cephalad direction. It then passes either into the sciatic notch or down the posterior column approximately 1 cm in front of the sciatic notch according to the preference of the surgeon. Swan neck chisels or curved osteotomes are used to cut between the inner and outer pelvic cortices into the interval under X-ray control. It is important not to penetrate the hip joint. The osteotomy is levered open anteriorly and laterally and held open by the excised wedge of bone form the femoral osteotomy (Figs. 5a and b). The graft is inserted so as to give maximum coverage of the femoral head and fixation is not usually needed (see Fig. 6). After wound closure a double spica or hip orthosis is applied for 3–6 weeks depending on the age of the patient. Complications after surgery include heterotopic bone formation (Fig. 7), loss of fixation, implant failure and femoral fracture as the blade plate can act as a stress riser.

Salvage procedures In the case of a longstanding, painful dislocation with established degenerative changes, a proximal femoral resection18 has been shown to be useful salvage procedure. This is not the same as a Girdlestone resection and should be performed extraperiosteally at least 3 cm below the lesser trochanter to minimise the occurrence of heterotopic ossification and proximal migration of the femoral shaft (Figs. 8a–c). A valgus proximal osteotomy to position the femoral head lateral to the pelvis, hip arthrodesis and arthroplasty are other possibilities but there are no published reports of long term follow-up of these procedures in adequate numbers of patients.

Summary and conclusions The manifestations and severity of cerebral palsy are wide ranging and the objectives of management must be tailored to each patient. Reducing hip pain in the patient with total body involvement or assisting in keeping the non-walking patient seated comfortably in their wheelchair are no less worthwhile objectives than efforts to improve the gait of the patients who are able to walk (Fig. 9). An awareness of the potential for hip problems in cerebral palsy and their surveillance are essential.

References 1. Abel MF, Wenger DR, Mubarak SJ, Sutherland DH. Quantitative analysis of hip dysplasia in cerebral palsy: a study of radiographs and 3-D formatted images. J Pediatr Orthoped 1994;14(3): 283–9. 2. Chung CY, Park MS, Choi IH, et al. Morphometric analysis of acetabular dysplasia in cerebral palsy. J Bone Joint Surg [Br] 2006;88B(2):243–7. 3. Ruwe PA, Gage JR, Ozonoff MB, Deluca PA. Clinical determination of femoral anteversion. J Bone Joint Surg [Am] 1992;74A(6):820–30. 4. Staheli LT. Rotational problems in children. J Bone Joint Surg [Am] 1993;75A:939–49. 5. Kerr AM, Kirtley SJ, Hillman SJ, et al. The mid-point of passive hip rotation range is an indicator of hip rotation in gait in cerebral palsy. Gait Posture 2003;17(1):88–91. 6. Lonstein J, Beck K. Hip dislocation in spastic cerebral palsy. J Pediatr Orthoped 1986;6(5):521–6. 7. Noonan KJ, Jones J, Pierson J, Honkamp NJ, Leverson G. Hip function in adults with severe cerebral palsy. J Bone Joint Surg [Am] 2004;86A(12):2607–13. 8. Boldingh EJ, Jacobs-van der Bruggen MA, Bos CF, Lankhorst GJ, Bouter LM. Determinants of hip pain in adult patients with severe cerebral palsy. J Pediatr Orthop B 2005;14(2):120–5.

ARTICLE IN PRESS The hip in cerebral palsy 9. Bagg MR, Farber J, Miller F. Long term follow up of hip subluxation in cerebral palsy patients. J Pediatr Orthoped 1993;13(1):32–6. 10. Soo B, Howard JJ, Boyd RN, et al. Hip displacement in cerebral palsy. J Bone Joint Surg [Am] 2006;88A(1):121–9. 11. Dobson F, Boyd RN, Parrott J, et al. Hip surveillance in children with cerebral palsy: impact on the surgical management of spastic hip disease. J Bone Joint Surg [Br] 2002;84B(5):720–6. 12. Reimers J. The stability of the hip in children: a radiological study of the results of muscle surgery in cerebral palsy. Acta Orthop Scand 1980;184:1–100. 13. Hicdonmez T, Steinbok P, Beauchamp R, et al. Hip joint subluxation after selective dorsal rhizotomy for spastic cerebral palsy. J Neurosurg 2005;103(1 Suppl):10–6. 14. Miller F, Cardoso Dias R, Dabney KW, et al. J Pediatr Orthoped 1997;17(5):571–84.

293 15. Brunner R, Baumann JU. Long term effects of intertrochanteric varus derotation osteotomy on femur and acetabulum in spastic cerebral palsy. An 11–18 year follow up study. J Pediatr Orthoped 1997;17(5):585–91. 16. Shea KG, Coleman SS, Carroll K, et al. J Bone Joint Surg [Am] 1997;79A(9):1342–51. 17. Jozwiak M, Marciniak W, Piontek T, et al. Dega’s transiliac osteotomy in the treatment of spastic hip subluxation and dislocation in cerebral palsy. J Pediatr Orthop B 2000; 9(4):257–64. 18. Castle ME, Schneider C. Proximal femoral resection-interposition arthroplasty. J Bone Joint Surg [Am] 1978;60A(8):1051–4. 19. Palisano R, Rosenbaum P, Walter S, et al. Development and reliability of a system to classify gross motor function in children with cerebral palsy. Dev Med Child Neurol 1997; 39(4):214–23.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 294–298

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

WRIST

Carpal tunnel syndrome in men A.C. Watts, J. McEachan The Hand Surgery Service, Queen Margaret Hospital, Whitefield Road, Dunfermline, Fife KY12 0SU, Scotland

KEYWORDS Carpal tunnel syndrome; Female patients; Male patients; Treatment; Outcome

Summary Carpal tunnel syndrome is thought of as affecting female patients in their sixth or seventh decades. This article reviews the literature and shows that it has an overall prevalence in males nearly equivalent to women. Males present with milder symptoms but more severe electrophysiological changes. Male gender is not associated with poor outcome but many male patients are involved in heavy manual work activities which is a poor prognostic indicator. & 2006 Elsevier Ltd. All rights reserved.

Introduction

Epidemiology

Carpal tunnel syndrome (CTS) is a clinical diagnosis made on the basis of a group of symptoms and signs caused by compression of the median nerve within the carpal tunnel. These include pain, sensory disturbance, motor loss and electrophysiological change.1,2 Pain may be felt in the hand, wrist or forearm. Weakness if present may be observed in the abductor pollicis brevis and opponens pollicis. CTS is classified as an industrial disease by the department of Work and Pensions, for which disablement benefit may be claimed, and as it accounts for the majority of upper extremity disorders attributed to the working environment, it represents a high cost to society.3 Many clinicians feel that men with CTS respond less well to treatment. This review examines the particular issues surrounding male patients, and gender differences with respect to CTS.

There are little good data on the incidence of CTS in the general population. Most studies have looked at specific industry populations in whom rates of CTS are likely to be higher than the general population.4,5 One large Swedish population based study demonstrated that women have a significantly higher prevalence of pain, numbness and/or tingling in the median nerve distribution than men.6 This observation is supported by recent prospective studies demonstrating that women have significantly greater preoperative symptoms and disability than men using well-validated measures7,8 but that there are no differences in outcome. When only those with clinically and electrophysiologically confirmed CTS are considered this difference is diminished (Table 1). Beyond retirement age the prevalence in women is 4 times that in men in Atroshi et al.’s6 study. It is likely given the multifactorial nature of this syndrome that there is some variation between different populations in the prevalence of CTS.

Corresponding author. Orthopaedic SpR, South East Scotland Rotation, New Royal Infirmary, Little France, Old Dalkeith Road, Edinburgh EH16 4SU, UK. Tel.: +44 131 2426464. E-mail address: [email protected] (A.C. Watts).

Anatomy

0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.02.013

The carpal tunnel is defined on the volar aspect of the wrist by the flexor retinaculum, which spans between the hamate

ARTICLE IN PRESS Carpal tunnel syndrome in men

Table 1

295

Gender-specific prevalence in the general population6.

Gender

Male Female

Pain, numbness and/or tingling in median nerve distribution

Clinically certain CTS

Prevalence % (95% CI)

Prevalence % (95% CI)

Clinically and electrophysiologically confirmed CTS Prevalence % (95% CI)

10.4 (8.6–12.2) 17.3 (15.3–19.4)

2.8 (1.8–3.8) 4.6 (3.5–5.7)

2.1 (1.3–3.0) 3.0 (2.1–3.9)

and triquetrum on the ulnar side to the scaphoid and trapezium on the radial side. It contains the median nerve and the tendons of flexor pollicis longus, four of flexor digitorum superficialis and four of flexor digitorum profundus. The carpal tunnel is narrowest at a point 2 cm from its proximal edge corresponding to the level of the hook of the hamate.9 However, the anatomy of the median nerve in the carpal tunnel is variable; a detailed description is provided in the study by Lanz.10 There is little evidence in the literature to suggest significant gender-specific differences in the anatomy of the carpal tunnel. There are a greater number of case reports pertaining to a variety of muscular anomalies in relation to CTS in men11–24 than women25–28 which may imply that this is a more common aetiological factor in males; there is however no scientific evidence to support this.

fabrication workers, food processing and construction but the second largest group of claims came from clerical workers. Longitudinal studies in industry workers indicate that increasing age, female gender, obesity, work posture and vibration associated occupations were associated with an increased risk of CTS.5,38 However, in one study with 17year follow-up only greater relative weight and female gender were associated.39 A literature review investigating the causal relationship between work and CTS concluded that except in the case of work that involves very cold temperatures such as butchery, work is less likely than demographic and disease-related variables to cause CTS.4,39 At a histological level Pickering has shown there is no significant association between heavy occupational hand use and tenosynovial thickening in the carpal tunnel.40

Other causes

Aetiology and pathophysiology Normal tissue pressures in the upper limb are approximately 8 mmHg. In CTS the resting pressure in the carpal tunnel may be 30 mmHg rising to over 90 mmHg with wrist movement.29,30 This increase in pressure can either lead to ischaemia or with prolonged pressure to focal demyelination.31–33 Common conditions associated with CTS are diabetes, rheumatoid arthritis, osteoarthritis, hypothyroidism and wrist fractures.34 However, in most cases, the cause of the increased pressure is idiopathic. In Atroshi et al.’s study6 only obesity (BMI of greater than 25) was shown to be statistically more common in patients with proven CTS. In patients with a work-related cause, Manktelow reported an association with tendonitis or epicondylitis in up to 39% of patients.35

Relationship to work The relationship between CTS and work is contentious. The prevalence of CTS is higher in patient series where the cohort of patients is from a particular ‘‘high risk’’ industry when compared to population studies.5,36,37 However, the lack of consistency between studies in the definition of CTS makes it difficult to draw direct comparisons. It can be argued that men are more likely to be affected if there is a relationship between heavy repetitive work or work with vibrating tools and CTS. A review of industrial workers compensation claims for a population of three million in North America revealed that half of those claiming for CTS were men.35 The majority of claims came from assembly and

Anatomical variations that impinge on the carpal tunnel,12,18,41 metabolic conditions42 and tumours43 are all well-described causes of CTS.

Electrophysiological studies The need for nerve conduction studies remains controversial.44 As a diagnostic tool nerve conduction studies can be viewed as an adjunct to clinical history and examination. Nerve conduction studies showed a high correlation with a clinical questionnaire set by Kamath and Stothard45 who advocate its use as an alternative. There is no evidence to support the use of nerve conduction studies to measure outcome although they may be useful in deciding when to re-operate following failed carpal tunnel decompression46. In a multicentre study from Italy, Padua reported that while men complain less of discomfort due to CTS they tend to have more severe electrophysiological findings.47

Management Non-surgical—injections/splints/behaviour modification The use of surgical splints for the management of CTS should not be dismissed. In a randomised control trial of splints versus surgery performed in the Netherlands the success rate was greater in the surgery group, with 90% success at 18 months. However, over a third (37%) of those in the splint arm of the trial had a successful outcome at 18 months and

ARTICLE IN PRESS 296

A.C. Watts, J. McEachan

the benefit of surgery was tempered by an increase in complications related to treatment including a painful or hypertrophic scar (61%), wound haematoma or infection (11%), severe pillar pain (2%) and one case of reflex sympathetic dystrophy.48 Predictors of success of treatment with a splint included a shorter duration of symptoms (o1 year) and milder symptoms at the time of randomisation.49 Gender was not shown to independently alter outcome. Similarly, the use of steroid injection in conjunction with wrist splints for 3 weeks is adequate treatment in some patients (11% of a series of 50 patients at a minimum of 6 months follow-up) with greater efficacy in those with symptoms for less than 1 year and those with more mild symptoms.50 A successful initial response to local steroid injection is a predictor of success following surgical treatment in those whose symptoms recur.51 In Gorsche’s study, meat-packers behaviour modification to less intensive duties was shown to be effective in managing the symptoms of patients in whom treatment with splints failed.52

Surgical There is no consensus regarding endoscopic versus open carpal tunnel release. Randomised control trials have not demonstrated any improvement in success rate of endoscopic release over open carpal tunnel decompression.53 A Cochrane review concluded that there was no strong need to replace open carpal tunnel decompression with alternative procedures.54 Proponents of the endoscopic technique point to a more rapid return to work following this procedure55; however, this is at the cost of a higher complication rate, including iatrogenic nerve injury.56 For male patients of working age, while there is clearly a benefit in enabling them to return to work more rapidly, the authors are not aware of any cost benefit analysis comparing the benefit of more rapid return to work to the greater cost of the procedure, longer operating time and financial cost of the complications of endoscopic release. In our opinion the personal cost to the individual in the event of a complication tends to favour open carpal tunnel decompression.

Outcomes Carpal tunnel decompression generally produces excellent outcomes, with over 80% of patients satisfied.57–60 However, while patients may declare satisfaction with the outcome of surgery this does not mean that all symptoms are abolished (only 51%59 to 63%57 of patients in reported series have complete relief of symptoms). In Agee’s study,55 one in four patients still reported pain 6 months after surgery, one in

Table 2

eight reported tingling and one in six numbness. Weakness was reported by a third of patients. Important complications related to carpal tunnel decompression, open or endoscopic, are in Longstaff’s series59 scar tenderness 13%, pillar pain 10% and weakness 5%. A prospective study of 97 patients with CTS found no differences in outcome between male and female patients following carpal tunnel decompression.7 However, Manktelow’s large study concluded that female gender was one of the significant independent predictors of worse clinical outcome when adjusted for age.35 In the same study the better outcome scores were amongst those who underwent surgery. There is no correlation between preoperative electrophysiological parameters and outcome,59 nor is the outcome affected by preoperative psychological disturbance.58 Men with occupation-related vibration exposure have a poorer outcome with only 60% having relief of their symptoms after surgery.61 Thirty percent still complained of persistent night time paraesthesia, with persistence of symptoms related to degree of preoperative vibration exposure.62 An association has been demonstrated between physically strenuous work activities and poorer outcomes, all these cases were involved in workers compensation claims and correlated with their inability to return to work.63 This link between poor outcome and workers compensation claims has been demonstrated in other studies.64,65

Return to work The majority of men with CTS are of working age.6 The return to work interval is affected by several factors (social security insurance, workers compensation claims, manual occupation) and ranges from an average of 11 days for nonmanual independent sector workers to 72 days for state employed workers compensation claimants (Table 2).66 Between 56% found that 67% of workers are able to return to their original jobs.35,60,67 Compared to clerical workers assembly and fabrication workers, food processing workers and mechanics are reported to have poorer clinical and return to work outcomes.35 From the same study the presence of other symptoms of repetitive strain (tendonitis, epicondylitis) is a negative predictor of return to work. An ongoing workers compensation claim is reported to be a critical predictor of work absence.67

Conclusion The majority of men affected by CTS are of working age. Open surgical decompression remains the mainstay of treatment. There is no evidence that male patients will

Average return to work period after surgery in days66.

Self-employed Private sector wage earners State employed

Manual

Non-manual

Social security

Workers compensation

29 42 63

11 21 49

17 31 56

34 46 72

ARTICLE IN PRESS Carpal tunnel syndrome in men have a less successful outcome following surgery unless their symptoms are related to the use of vibrating tools or they have an ongoing compensation claim. A third of patients will need to consider modifying working practices or seeking alternative employment, especially if working in heavy manual industry.

References 1. Dawson DM. Entrapment neuropathies of the upper extremities. N Engl J Med 1993;329:2013–8. 2. Nora DB, Becker J, Ehlers JA, Gomes I. What symptoms are truly caused by median nerve compression in carpal tunnel syndrome? Clin Neurophysiol 2005;116:275–83. 3. Feuerstein M, Miller VL, Burrell LM, Berger R. Occupational upper extremity disorders in the federal workforce. Prevalence, health care expenditures, and patterns of work disability. J Occup Environ Med 1998;40:546–55. 4. Falkiner S, Myers S. When exactly can carpal tunnel syndrome be considered work-related? Anz J Surg 2002;72:204–9. 5. Werner RA, Franzblau A, Gell N, Hartigan AG, Ebersole M, Armstrong TJ. Incidence of carpal tunnel syndrome among automobile assembly workers and assessment of risk factors. J Occup Environ Med 2005;47:1044–50. 6. Atroshi I, Gummesson C, Johnsson R, Ornstein E, Ranstam J, Rosen I. Prevalence of carpal tunnel syndrome in a general population. JAMA 1999;282:153–8. 7. Hobby JL, Venkatesh R, Motkur P. The effect of age and gender upon symptoms and surgical outcomes in carpal tunnel syndrome. J Hand Surg [Br] 2005;30:599–604. 8. Greenslade JR, Mehta RL, Belward P, Warwick DJ. Dash and Boston questionnaire assessment of carpal tunnel syndrome outcome: what is the responsiveness of an outcome questionnaire? J Hand Surg [Br] 2004;29:159–64. 9. Cobb TK, Dalley BK, Posteraro RH, Lewis RC. Anatomy of the flexor retinaculum. J Hand Surg [Am] 1993;18:91–9. 10. Lanz U. Anatomical variations of the median nerve in the carpal tunnel. J Hand Surg [Am] 1977;2:44–53. 11. Cossey AJ, Stranks GJ. Intramuscular lipoma in an anomalous muscle belly of the middle finger lumbrical as a cause of carpal tunnel syndrome and trigger wrist. Orthopedics 2003;26:85–6. 12. Eriksen J. A case of carpal tunnel syndrome on the basis of an abnormally long lumbrical muscle. Acta Orthop Scand 1973;44:275–7. 13. Feingold MH, Hidvegi E, Horwitz SJ. Bilateral carpal tunnel syndrome in an adolescent. Am J Dis Child 1980;134:394–5. 14. Imran D, Bainbridge LC. Carpal tunnel syndrome after distal release of the flexor digitorum profundus and subsequent retraction of the lumbrical muscle into the carpal tunnel. J Hand Surg [Br] 1999;24:303–4. 15. Kono H. Acute carpal tunnel syndrome caused by anomalous muscle bellies: a case report. Hand Surg 2003;8:141–3. 16. Lange H. Carpal tunnel syndrome caused by the palmaris profundus muscle. Case report. Scand J Plast Reconstr Surg Hand Surg 1999;33:251–2. 17. Robinson D, Aghasi M, Halperin N. The treatment of carpal tunnel syndrome caused by hypertrophied lumbrical muscles. Case reports. Scand J Plast Reconstr Surg Hand Surg 1989;23:149–51. 18. Schon R, Kraus E, Boller O, Kampe A. Anomalous muscle belly of the flexor digitorum superficialis associated with carpal tunnel syndrome: case report. Neurosurgery 1992;31:969–70. 19. Schultz RJ, Endler PM, Huddleston HD. Anomalous median nerve and an anomalous muscle belly of the first lumbrical associated with carpal-tunnel syndrome. J Bone Jt Surg Am 1973; 55:1744–6.

297 20. Touborg-Jensen A. Carpal-tunnel syndrome caused by an abnormal distribution of the lumbrical muscles. Case report. Scand J Plast Reconstr Surg 1970;4:72–4. 21. De Smet L, Wouters C. Severe carpal tunnel syndrome in a patient with juvenile idiopathic arthritis due to proximal migration of hypertrophic lumbrical muscles. Clin Rheumatol 2004;23:552–4. 22. Boya H, Ozcan O, Arac S, Tandogan R. Incomplete scapholunate and trapeziotrapezoid coalitions with an accessory carpal bone. J Orthop Sci 2005;10:99–102. 23. Ragoowansi R, Adeniran A, Moss AL. Anomalous muscle of the wrist. Clin Anat 2002;15:363–5. 24. Guler MM, Celikoz B. Anomalous palmaris longus muscle causing carpal tunnel-like syndrome. Arch Orthop Trauma Surg 1998;117:296–7. 25. Pruzansky ME. Compression of the radial branch of the median nerve due to an anomalous muscle belly of the first lumbrical in a child. Mt Sinai J Med 2004;71:285–6. 26. Malloy MA, Finger DR. Clinical image: carpal tunnel syndrome from accessory lumbrical muscles. Arthritis Rheum 2000;43:707. 27. Depuydt KH, Schuurman AH, Kon M. Reversed palmaris longus muscle causing effort-related median nerve compression. J Hand Surg [Br] 1998;23:117–9. 28. Sanchez LJ, Canada M, Diaz L, Sarasua G. Compression of the median nerve by an anomalous palmaris longus tendon: a case report. J Hand Surg [Am] 1996;21:858–60. 29. Gelberman RH, Hergenroeder PT, Hargens AR, Lundborg GN, Akeson WH. The carpal tunnel syndrome. A study of carpal canal pressures. J Bone Jt Surg Am 1981;63:380–3. 30. Okutsu I, Ninomiya S, Hamanaka I, Kuroshima N, Inanami H. Measurement of pressure in the carpal canal before and after endoscopic management of carpal tunnel syndrome. J Bone Jt Surg Am 1989;71:679–83. 31. Fern R, Harrison PJ. The contribution of ischaemia and deformation to the conduction block generated by compression of the cat sciatic nerve. Exp Physiol 1994;79:583–92. 32. Gupta R, Steward O. Chronic nerve compression induces concurrent apoptosis and proliferation of Schwann cells. J Comp Neurol 2003;461:174–86. 33. Diao E, Shao F, Liebenberg E, Rempel D, Lotz JC. Carpal tunnel pressure alters median nerve function in a dose-dependent manner: a rabbit model for carpal tunnel syndrome. J Orthop Res 2005;23:218–23. 34. Geoghegan JM, Clark DI, Bainbridge LC, Smith C, Hubbard R. Risk factors in carpal tunnel syndrome. J Hand Surg [Br] 2004;29:315–20. 35. Manktelow RT, Binhammer P, Tomat LR, Bril V, Szalai JP. Carpal tunnel syndrome: cross-sectional and outcome study in Ontario workers. J Hand Surg [Am] 2004;29:307–17. 36. Falck B, Aarnio P. Left-sided carpal tunnel syndrome in butchers. Scand J Work Environ Health 1983;9:291–7. 37. Kim JY, Kim JI, Son JE, Yun SK. Prevalence of carpal tunnel syndrome in meat and fish processing plants. J Occup Health 2004;46:230–4. 38. Nathan PA, Meadows KD, Istvan JA. Predictors of carpal tunnel syndrome: an 11-year study of industrial workers. J Hand Surg [Am] 2002;27:644–51. 39. Nathan PA, Istvan JA, Meadows KD. A longitudinal study of predictors of research-defined carpal tunnel syndrome in industrial workers: findings at 17 years. J Hand Surg [Br] 2005. 40. Pickering SA, Stevens A, Davis TR. Work practices and histopathological changes in the tenosynovium in carpal tunnel syndrome in men. J Hand Surg [Br] 2004;29:325–8. 41. Kimura H, Ikuta Y, Ishida O. Carpal tunnel syndrome in radial dysplasia. J Hand Surg [Br] 2001;26:533–6. 42. Pai CH, Tseng CH. Acute carpal tunnel syndrome caused by tophaceous gout. J Hand Surg [Am] 1993;18:667–9.

ARTICLE IN PRESS 298 43. Nakamichi K, Tachibana S. Unilateral carpal tunnel syndrome and space-occupying lesions. J Hand Surg [Br] 1993;18:748–9. 44. Smith NJ. Nerve conduction studies for carpal tunnel syndrome: essential prelude to surgery or unnecessary luxury? J Hand Surg [Br] 2002;27:83–5. 45. Kamath V, Stothard J. A clinical questionnaire for the diagnosis of carpal tunnel syndrome. J Hand Surg [Br] 2003;28:455–9. 46. Cobb TK, Amadio PC, Leatherwood DF, Schleck CD, Ilstrup DM. Outcome of reoperation for carpal tunnel syndrome. J Hand Surg [Am] 1996;21:347–56. 47. Padua L, Padua R, Aprile I, Tonali P. Italian multicentre study of carpal tunnel syndrome. Differences in the clinical and neurophysiological features between male and female patients. J Hand Surg [Br] 1999;24:579–82. 48. Gerritsen AA, de Vet HC, Scholten RJ, Bertelsmann FW, de Krom MC, Bouter LM. Splinting vs surgery in the treatment of carpal tunnel syndrome: a randomized controlled trial. JAMA 2002;288:1245–51. 49. Gerritsen AA, Korthals-de Bos IB, Laboyrie PM, de Vet HC, Scholten RJ, Bouter LM. Splinting for carpal tunnel syndrome: prognostic indicators of success. J Neurol Neurosurg Psychiatry 2003;74:1342–4. 50. Gelberman RH, Aronson D, Weisman MH. Carpal-tunnel syndrome. Results of a prospective trial of steroid injection and splinting. J Bone Jt Surg Am 1980;62:1181–4. 51. Edgell SE, McCabe SJ, Breidenbach WC, LaJoie AS, Abell TD. Predicting the outcome of carpal tunnel release. J Hand Surg [Am] 2003;28:255–61. 52. Gorsche RG, Wiley JP, Brant R, Renger RF, Sasyniuk TM, Burke N. Comparison of outcomes of untreated carpal tunnel syndrome and asymptomatic controls in meat packers. Occup Med (Lond) 2002;52:491–6. 53. Wong KC, Hung LK, Ho PC, Wong JM. Carpal tunnel release. A prospective, randomised study of endoscopic versus limitedopen methods. J Bone Jt Surg Br 2003;85:863–8. 54. Scholten RJ, Gerritsen AA, Uitdehaag BM, van Geldere D, de Vet HC, Bouter LM. Surgical treatment options for carpal tunnel syndrome. Cochrane Database Syst Rev 2004; CD003905. 55. Agee JM, McCarroll Jr HR, Tortosa RD, Berry DA, Szabo RM, Peimer CA. Endoscopic release of the carpal tunnel: a randomized prospective multicenter study. J Hand Surg [Am] 1992;17:987–95.

A.C. Watts, J. McEachan 56. Chuang HL, Wong CW. Carpal tunnel syndrome induced by tophaceous deposits on the median nerve: case report. Neurosurgery 1994;34:919. 57. Weber RA, Rude MJ. Clinical outcomes of carpal tunnel release in patients 65 and older. J Hand Surg [Am] 2005;30:75–80. 58. Hobby JL, Venkatesh R, Motkur P. The effect of psychological disturbance on symptoms, self-reported disability and surgical outcome in carpal tunnel syndrome. J Bone Jt Surg Br 2005;87:196–200. 59. Longstaff L, Milner RH, O’Sullivan S, Fawcett P. Carpal tunnel syndrome: the correlation between outcome, symptoms and nerve conduction study findings. J Hand Surg [Br] 2001;26:475–80. 60. Adams ML, Franklin GM, Barnhart S. Outcome of carpal tunnel surgery in Washington State workers’ compensation. Am J Ind Med 1994;25:527–36. 61. Bostrom L, Gothe CJ, Hansson S, Lugnegard H, Nilsson BY. Surgical treatment of carpal tunnel syndrome in patients exposed to vibration from handheld tools. Scand J Plast Reconstr Surg Hand Surg 1994;28:147–9. 62. Hagberg M, Nystrom A, Zetterlund B. Recovery from symptoms after carpal tunnel syndrome surgery in males in relation to vibration exposure. J Hand Surg [Am] 1991;16:66–71. 63. al Qattan MM, Bowen V, Manktelow RT. Factors associated with poor outcome following primary carpal tunnel release in nondiabetic patients. J Hand Surg [Br] 1994;19:622–5. 64. Katz JN, Keller RB, Simmons BP, Rogers WD, Bessette L, Fossel AH, et al. Maine Carpal Tunnel Study: outcomes of operative and nonoperative therapy for carpal tunnel syndrome in a community-based cohort. J Hand Surg [Am] 1998;23:697–710. 65. Atroshi I, Johnsson R, Nouhan R, Crain G, McCabe SJ. Use of outcome instruments to compare workers’ compensation and non-workers’ compensation carpal tunnel syndrome. J Hand Surg [Am] 1997;22:882–8. 66. Chaise F, Bellemere P, Fril JP, Gaisne E, Poirier P, Menadi A. Return-to-work interval and surgery for carpal tunnel syndrome. Results of a prospective series of 233 patients. J Hand Surg [Br] 2004;29:568–70. 67. Katz JN, Lew RA, Bessette L, Punnett L, Fossel AH, Mooney N, et al. Prevalence and predictors of long-term work disability due to carpal tunnel syndrome. Am J Ind Med 1998;33:543–50.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 299–315

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

TUMOURS

Management of metastatic disease of the appendicular skeleton Robert U. Ashforda,, Susan Pendleburyb, Paul D. Stalleya a

Department of Orthopaedics, New South Wales Bone & Soft Tissue Sarcoma Service, Royal Prince Alfred Hospital, Missenden Road, Camperdown, NSW 2050, Australia b Radiation Oncology, New South Wales Bone & Soft Tissue Sarcoma Service, Royal Prince Alfred Hospital, Missenden Road, Camperdown, NSW 2050, Australia

KEYWORDS Metastases; Bone; Surgery

Summary Treatment advances have resulted in improved prognosis for patients with cancer metastatic to bone. The management of appendicular skeletal metastases requires a multi-disciplinary approach. The overall disease prognosis is important in determining the appropriate surgical treatment of metastases, with simple measures for those with the poorest prognosis and resection with reconstruction for those expected to survive more than 1 year. This article reviews management options available for patients with skeletal metastases affecting the appendicular skeleton and broaches the controversies of prophylactic fixation of impending fractures. & 2006 Elsevier Ltd. All rights reserved.

Introduction Advances in multi-modality management mean that the prognosis for patients with cancer continues to improve.1,2 Furthermore, advances in chemotherapy have conferred improvement on both local and systemic control of malignancies. Figures from the American Cancer Society (1995–2000) quote an age-adjusted 5-year survival of 88% for breast cancer and 99% for prostatic cancer. Bone is the third most common site of metastatic disease after lung and liver. Harrington3 demonstrated that, at postmortem, 84% of patients with breast or prostate carcinomas had skeletal metastases. Figures for thyroid, lung and renal Corresponding author.

E-mail address: [email protected] (R.U. Ashford). 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.03.005

carcinomas were 50%, 44% and 37%, respectively.4 Those cancers most likely to metastasise to bone are:

Prostate Skeletal metastases are very common in prostate cancer. At autopsy, 84% of those with prostatic adenocarcinoma have skeletal metastases, while in 1982, of 20,000 new cases of prostate cancer, 21.5% of patients presented with clinical stage D (metastatic) disease. Skeletal metastases are generally associated with a poor prognosis. Only 23% of patients survived 5 years from initial diagnosis, and the 10year survival rate is 10%. The bones most commonly involved are the vertebrae, sternum, pelvic bones, ribs, and femora. The most common sites for pathological fracture are the medial cortex of the proximal femur and the vertebral

ARTICLE IN PRESS 300 bodies, these two carrying heavy loads. In a prospective study of 112 patients with bone metastatases from prostate carcinoma, 10% developed pathological fractures.5 Eightyfour percent of prostatic skeletal metastases are osteoblastic, 12% are mixed osteolytic and osteoblastic. Pure osteolytic metastases are rare.6 The bony metastases are characterised by an excess of abnormally dense bone. This indicates that there is increased bone turnover. Both osteoblastic and osteoclastic activities increase, but the relative amount of osteoblastic activity exceeds that of the osteoclasts, and bone formation is the net result. Treatment of osteoblastic metastases is principally hormonal and radiotherapy. Predictors of skeletal morbidity include the extent of bone involvement, severity of pain, serum alkaline phosphatase level, and urinary deoxypyridinoline levels.7 Even in those cases in which fractures do occur, the rate of healing approaches that of normal bone. Normal healing, in conjunction with effective radiotherapy and hormonal manipulation, limits the need for surgical stabilisation to only about one-fourth of the patients who develop a pathologic fracture.

Breast Bone is the first site of metastasis in 26% of breast cancer patients. It is the most common site of metastatic disease, both at presentation and as the site of first recurrence.7 Up to 85% of breast cancer patients develop skeletal metastases at some stage in the disease process. Higher incidences are seen in those with steroid receptor positive lesions and welldifferentiated lesions.8 Up to one-third of these patients develop a major complication9 and approximately 17% require orthopaedic surgery,10 two-thirds of these because of femoral metastases. Du ¨rr8 reported a 20% surgical complication rate in these patients. Radical resection of solitary lesions in breast cancer does not appear to improve survival, and therefore intralesional surgery, methyl methacrylate augmentation and stabilisation is often most appropriate.8 Survival in breast cancer patients with skeletal metastases is increasing with newer adjuvant therapies. Patients with osteoblastic breast metastases or osteolytic metastases with no risk of fracture are usually treated with radiotherapy. The mean survival of patients after the first metastasis is detected was 34 months in 199211 and is likely to be longer now. The role of bisphosphonates in the prevention of skeletal events in metastatic disease and in the prevention of metastases is currently being explored.

Kidney Between 30% and 60% of patients with renal cell carcinoma have skeletal metastases at the time of primary tumour diagnosis.12 In between 1% and 3% of cases the metastatic lesion is a solitary skeletal metastasis. Lesions are often aggressively osteolytic, progress rapidly and are highly vascular. The skeletal pathology can be the presenting complaint in metastatic renal cell carcinoma. Fifty percent of renal metastases are synchronous13,14 and there appears to be a survival advantage for synchronous over metachronous skeletal metastases.13 Pre-operative embolisation of the vascular supply to the tumour is useful in reducing intra-

R.U. Ashford et al. operative bleeding.15–17 Typically, renal metastases respond poorly to radiotherapy in terms of healing but bone pain responds reasonably well.18,19 Chemotherapy and hormone therapy are also relatively ineffective.17 Isolated metastases with a prolonged latent period should be treated by en-bloc resection rather than stabilization.12,20,21 The 5-year survival of patients with a solitary bone metastasis from renal cell carcinoma has been quoted as high as 55%.21 The evidence is variable and Fuchs14 reported no survival advantage of wide excision over intralesional surgery. Thirty-five percent of patients with locally advanced renal tumours have skeletal metastases. Nephrectomy for patients with a solitary skeletal metastasis and a synchronous primary renal tumour appears to confer a survival advantage14 (88% versus 66% 1-year survival, 31% versus 0% 5-year survival).

Lung Skeletal metastases usually occur in small cell carcinoma. Up to 60% of patients with lung cancer develop skeletal metastases. Lung cancer metastasis most commonly affects the spine, ribs, pelvis, and proximal long bones. A unique feature of this lesion is its ability to spread to the bones of the hands and feet. Half of all metastases to the hand bones are from lung, as well as 15% of lesions in the feet. Lung cancer metastases normally (approximately 75%) appear purely lytic, with poor margination, no matrix and cortical destruction, but may also be blastic. The prognosis is usually poor22 and the bone disease aggressive. There are also a number of reports in the literature of metastases from lung carcinoma to skeletal muscle. Systemic symptoms may also occur, such as those due to hypercalcemia and hypertrophic pulmonary osteoarthropathy. Lung cancer with metastasis to bone is one of the most aggressive tumours and has a very unfavourable prognosis. The average survival after the diagnosis of a metastasis is about 6 months. There is virtually no role of curative surgery.

Thyroid Approximately 4% of thyroid malignancies metastasise to bone and are usually osteolytic. Of these around half are present at the time of initial diagnosis. It is typically adenocarcinomas that metastasise. Seventy percent of metastases occur in the axial skeleton.23 The 10-year survival of patients with thyroid carcinoma metastatic to bone is 13% from the time of diagnosis of the first skeletal metastasis.23 Treatment is usually surgical with adjuvant radiotherapy. Bone pain from thyroid metastases responds well to bisphosphonate therapy.24 If the lesion is solitary, aggressive wide local excision plus appropriate reconstruction as for a primary bone tumour may be the desirable treatment. Thyroid tumours can be highly vascular and pre-operative angiography and selective embolisation may be needed. Radioactive iodine 131I has been found to be beneficial in some cases and has been demonstrated to be a prognostic indicator.23

Gastro-intestinal Bone metastases are relatively rare and usually osetolytic. Colorectal tumours carry a better prognosis than gastric and

ARTICLE IN PRESS Management of metastatic disease of the appendicular skeleton pancreatic. Hatfield25 recommended routine bone scanning of pancreatic primary carcinomas based on a detection rate of 6%. Kanthan26 reported a 6.6% incidence of skeletal metastases from colorectal primary tumours. Eighty-three percent of these were in combination with lung, liver and brain metastases. Bone lesions are principally of the axial skeleton. Solitary skeletal metastasis from colonic carcinoma was rare, with an incidence of only 1.1%. Besbeas27 reported from Memorial Sloan-Kettering Cancer Center the incidences of osseous metastases were 8.9% from rectal carcinoma and 5.1% from colonic carcinoma. Of these, 14 (1.8%) had osseous metastases only. The mean period from manifestation of skeletal metastasis to death was 13.2 months. In gastric primaries, skeletal metastases are more common to the axial skeleton, principally affecting the thoraco-lumbar vertebrae.28 Osteoblastic metastases are not uncommon from gastric carcinomas.

Gynaecological The reported incidence is approximately 10% of ovarian, 2–5% of cervical and less than 2.5% of uterine carcinomas metastasising to bone, most frequently to the axial skeleton. Skeletal metastases are usually associated with more advanced local disease. Cervical cancer metastasises to the lungs more frequently than it does to bone. Blastic metastases are rare and axial metastases are much more common than appendicular. However, when investigated at post-mortem, in a study between 1948 and 1984, autopsies were performed on 305 patients with primary carcinomas of the cervix, endometrium, ovaries, fallopian tubes, vulva, and vagina. Skeletal metastases were detected pre-mortem and at autopsy in 49 cases (16.1%) of gynaecological primaries. These were 20 (40.8%) of cervical carcinoma, 17 (34.7%) endometrial, 7 (14.3%) ovarian, 4 (8.2%) vulval, and 1 (2%) fallopian tube. This suggests the true incidence may be higher.29 The skeletal metastases tend to behave aggressively and should usually be treated by simple osteosynthesis techniques. One-third of new cancers diagnosed are breast tumours in females and prostatic in males. These provide the largest burden of metastatic patients to the orthopaedic oncologist.

Principles of management Carcinomas are much more likely to metastasise to bone than sarcomas. The axial skeleton is seeded more than the appendicular skeleton, presumably due to the persistence of red bone marrow in the former. The ribs, pelvis and spine are usually first affected. Batson’s vertebral venous plexus allows cells to enter the vertebral circulation without first passing through the lungs. The sluggish blood flow in this plexus is more conducive to tumour survival, perhaps accounting for the high rate of prostate cancer metastasis to the spine. It is a valveless system that also allows retrograde tumour embolisation. Pain, pathological fractures and hypercalcaemia are the major sources of morbidity in patients with bone metastasis. The other common complaint is difficulty mobilising. Pain is the most common symptom found in 70% of patients with bone metastases. Pain has been attributed to stretching of

301

the periosteum by the tumour, nerve stimulation in the endosteum, tumour growth leading to local soft tissue inflammation or by pathological fracture. The pain is usually dull in character and constant. It becomes gradually progressive and night pain not relieved by rest should alert the surgeon to a potentially sinister cause. Pathological fractures are most common in breast cancer due to the lytic nature of the lesions. They are less common in lung cancer due to short survival and in prostate cancer, when the lesions tend to be osteoblastic. Lytic bone metastases must be greater than 1 cm across and have destroyed 30–50% of the bone density in order to be seen by X-ray. Up to 13 of patients with skeletal metastases may suffer pathological fractures.30 Even when pathologic fractures occur the degree of soft tissue damage is small and extensive haematoma is rare. They do not require internal fixation in the middle of the night and appropriate planning is desirable. These fractures can be stress fractures initially that progress to full-blown pathological fractures without expedient management. Hypercalcaemia is usually due to calcium release from malignant bone destruction. The serum calcium should be measured in all patients with bone pain and malignant disease and before undertaking surgery for skeletal metastases. Hypercalcaemia should be addressed by rehydration and bisphosphonates. Metastatic bone lesions can be described as osteolytic, osteoblastic and mixed. The osteolytic lesions are most common where the destructive processes outstrip the laying down of new bone. Osteoblastic lesions result from new bone growth that is stimulated by the tumour. The majority of skeletal metastases occur in the spine, pelvis and long bones. Virtually any malignant tumour can metastasise to any bone and cause a pathologic fracture. In the appendicular skeleton, approximately two-thirds of pathologic fractures occur in the femur and the majority of the remainder are in the humerus.31 Survival following pathologic fracture is related to the prognosis of the primary tumour, although for most primary tumours the long-term survival following pathological fracture has more than tripled in the past 25 years.20 Options for management of skeletal metastases include radiotherapy, surgery, chemotherapy, hormonal therapy and immunotherapy. Improved medical treatment of cancers has resulted in increased survival and this requires thought to be given as to the most appropriate surgical procedure. It is desirable that skeletal metastasis management be through a multi-disciplinary team. Patients with a predicted survival of less than 6 weeks require analgesia and radiotherapy. More than 6 weeks warrants consideration of fracture stabilisation and more than 6 months, endoprosthetic reconstruction. Put in simple terms, the recovery from the operation must be shorter than the anticipated survival. Damron1 identified four key principles in the surgical management of metastatic bone lesions. Firstly, patient selection is critical, in that the operation is appropriate for the predicted survival of the patient. Secondly, the construct must be stable enough to allow full weight-bearing immediately. The third principle is that all areas of the bone that are affected by tumour are addressed in any planned reconstruction and finally post-operative radiotherapy is utilised to minimise disease progression. Where the bone has been reamed, the entire bone should be irradiated.

ARTICLE IN PRESS 302

R.U. Ashford et al.

The aims of surgery for patients with skeletal metastases are no different to the aims in other aspects of orthopaedics–pain relief, preserving or restoring function and maximising the quality of life. Occasionally, for the solitary renal or thyroid metastasis, cure may also be a goal. It is important not to look for a short-term fix to what may well be a long-term problem.

and are superior at demonstrating marrow replacement, skip lesions, quantifying oedema and assessing neurovascular involvement. Bone scinitigraphy will show the intraosseous extent of lesions. It may also show multiple hot lesions (Fig. 4A), but beware of cold lesions (myeloma, renal cell carcinoma). It is also useful for demonstrating singularity versus multiplicity

Investigation and diagnosis A thorough history and examination is mandatory in all patients, including examination of the breasts in all female patients and the prostate in all males. Metastatic disease is most common in patients over 40 years of age, and lymphoma and myeloma are also common. Table 1 outlines the investigations that are often employed in the investigation of patients with suspected skeletal metastases. Radiographs should be taken in two planes including the joint above and below, so as to enable all lesions within the bone to be identified and stabilised. Plain radiographs offer good integration of the overall bone structure and alignment, though initially metastases may be subtle or invisible. Radiographic features of skeletal metastases are variable. Tumours can be:

 osteolytic (Fig. 1), osteoblastic (Fig. 2) or mixed (Fig. 3).    

Lytic metastases are often much more extensive than on initial radiologic impression; solitary or multiple; well or poorly defined; permeative and/or destructive; varied in their soft tissue involvement.

A Computerised Tomography (CT) scan or a Magnetic Resonance Imaging (MRI) scan may be appropriate, depending on the site of the lesion. CT scans are best for assessing bone quality, bone destruction, calcified tumour matrix and cortical erosions. MRI scans are highly sensitive and specific

Table 1

Figure 1 Osteolytic metastasis of the proximal femur in a 47 year old female with metastatic breast cancer.

Investigations in patients with metastatic bone disease. Known primary

Unknown primary

Serological

Full blood count Bone biochemistry Liver function tests Coagulation screen Group and save

Radiological

AP/lateral of whole bone CT/MRI bone Chest radiograph PET scan or isotope bone scan

Biopsy

Not necessary

Full blood count Bone biochemistry Liver function tests Coagulation screen Group and save Thyroid function tests Prostate specific antigen Tumour markers Myeloma screen AP/lateral of whole bone CT/MRI bone PET scan or isotope bone scan Mammography CT chest, abdomen and pelvis USS as appropriate Most easily accessible lesion, for histology and microbiology (including TB)

ARTICLE IN PRESS Management of metastatic disease of the appendicular skeleton

303

Figure 2 Osteoblastic metastasis of the proximal femur and pelvis from breast carcinoma.

Figure 4 (A) Isotope bone scan demonstrating multiple areas of increased uptake consistent with multiple metastases from lung primary. (B) PET scan in the same patient. (C) PET–CT scan of the same patient.

Figure 3 Mixed lytic and blastic pattern of metastasis of the proximal femur again from prostatic primary.

of lesions and for identifying asymptomatic lesions. Isotope scans often reveal lesions to be much more extensive than initial radiographs suggested. Positron Emission Tomography (PET) scanning is emerging as a useful technology and we

utilise it in the majority of our patients to determine the extent of tumour disease (Figs. 4B and C). F18-FDG (fluorine-18 fluorodeoxyglucose) is a glucose analogue taken up into tumour cells. This enables the most active and viable part of the tumour to be identified and this is the appropriate portion of the tumour to target on biopsy. A combination of these modalities will help to identify other sites (metastases), both skeletal and visceral, that may potentially require treatment. Angiography is not commonly indicated. Specific indications are large pelvic tumours and those tumours where preoperative embolisation is considered. Those patients with

ARTICLE IN PRESS 304 metastatic renal cell carcinoma should be considered for pre-operative embolisation, as torrential bleeding may be encountered at surgery. If ever there is a doubt, and especially if the lesion is solitary, a biopsy should be performed. The ‘‘whoops’’ operation (intramedullary nailing of a primary bone tumour) can have catastrophic consequences, disseminating tumour cells throughout the bone marrow, rendering the limb unsalvageable and having a negative impact on survival. An MRI scan (or PET scan) should be performed prior to the biopsy to allow for appropriate planning of the biopsy. The biopsy should be a core needle biopsy through a surgical stab incision with violation of only a single compartment. The biopsy can be performed on an outpatient basis,4,32 although we prefer a general anaesthetic with frozen section. The biopsy should be well planned in case surgical resection is subsequently needed. This can be performed at the time of the planned nailing procedure; a frozen section is performed and, assuming confirmation of metastasis, the nailing can proceed. It is worth remembering the general principles of biopsy: usually a longitudinal incision (the person doing the biopsy should be aware of the approach for definitive resection), single compartment violation, no neurovascular

Figure 5 PET scanning in the unknown primary. The scan demonstrates a glucose avid bone lesions in the ischium and a lesion in the oesophagus consistent with an oesophageal primary.

R.U. Ashford et al. violation, meticulous haemostasis and a drain, if warranted, exiting in the line of the incision. The isolated hip lesion is worthy of special consideration, as chondrosarcoma needs to be excluded. In 15% of cases the primary tumour is never (or only posthumously) identified.33 With the advent of PET scanning this figure will fall. Our staging investigations for the unknown primary at present consist of only a PET scan and CT chest. With the advent of PET–CT this may become the only investigation necessary for the unknown primary (Fig. 5).

Medical management Radiotherapy Radiotherapy is a well-established and effective local treatment for patients with painful bone metastases. Ninety percent of patients with painful osseous metastases experience some relief of pain and 54% eventually obtain complete relief, often in combination with other analgesia.34 The effect of the radiotherapy is thought to be modulated by humeral mediators. Three systematic reviews35–37 and a meta-analysis38 have demonstrated an equivalence of pain relief achieved by radiotherapy delivered as a single fraction and that delivered in multiple fractions. Single fraction doses of 4–18 Gy have all produced some response. Four gray produces overall response rates of approximately 45–59%. Two randomised trials39,40 have demonstrated 8 Gy as a single fraction to be superior to 4 Gy. Eight gray produces response rates of 70–85%. Results from 6 Gy treatments are only slightly lower. Doses above 8 Gy do not produce measurably superior response rates but produce toxicity in some situations. Detailed analysis of the largest study, the Dutch Bone Metastasis Study41 confirmed 8 Gy is equivalent to 24 Gy in six fractions, with response rates of 71%. In this study 14% of patients responded completely requiring no analgesia and experiencing no pain. A similar figure of 15% has been reported after 10 Gy. The mean time to response was 3 weeks and the mean duration of response was 18 weeks. This is typical of many studies. Mean response duration to 52 weeks has been described in some series. Seventy percent of patients achieve durable relief, maintained for the rest of their life.34 Most studies show better response rates in breast cancer, prostate cancer and myeloma metastases than in lung, kidney or colorectal cancer. The Dutch Bone Metastasis Trial showed response rates of 78% in breast cancer and 79% in prostate cancer. Other histologies do respond to radiation treatment; the lung cancer response rate was 58% and other histologies 60–65%. Some studies have not shown a difference between histologies. Re-treatment of patients with painful bone metastases is also effective.42 Regardless of their response to initial treatment, up to 90% respond to a re-treatment. Mithal43 reported an 84% response rate to a second treatment with no relationship to primary tumour type or site of metastasis. Jeremic44 reported a 46% response rate to re-treatment in initial non-responders and a 73% response rate in those who had responded previously. The Dutch Bone Metastasis Trial

ARTICLE IN PRESS Management of metastatic disease of the appendicular skeleton reported a 66% response to re-treatment. The best response to re-treatment is in breast cancer with an 89% response rate in those who initially responded and an 82% response rate in a group of non-responders to initial radiation therapy. Some evidence exists that multiple fraction courses may be more effective in patients with solitary bone metastases.45 Additionally, where normal tissue tolerances are approached, especially around the spinal cord, it may be necessary to deliver the radiotherapy in a series of fractionated doses.

Other radiation techniques Hemibody irradiation is an alternative to localised radiation where metastases are widespread. This is commonly seen in prostate cancer. Treatment is delivered using 6–8 Gy in a single dose to the upper, middle or lower part of the body. Response rates up to 77% are recorded.46 Nausea and vomiting and haematological suppression occur in up to 30% of patients. Radioisotopes. The advantages of this modality are that all osseous sites are addressed simultaneously. Because uptake is selectively into bones, irradiation of normal tissues is limited. Treatment is administered as a single outpatient injection. Their use is limited by the requirement for good renal function and bone marrow reserve, often not present in patients with advanced malignancy. The common isotopes are Strontium-89 (89Sr), and Phosphorus-32 (32P).47 Samarium-153 and Rhenium-186 are also reported. Strontium-89 decays by beta emission with a half-life of 50–60 days. It is absorbed into bone matrix. Elimination is through the kidneys, requiring careful disposal of urine for 7–10 days. There is little gamma emission, so patients are not a radiation hazard to their families. Most work on this modality has been in prostate cancer and response rates of 75% are reported. One study showed equivalence with external beam radiotherapy.48 There can be a transient flare of pain in 10% of patients at 1–2 weeks. The principal toxicity is haematological, with a drop of up to 50% in blood counts (white cells and platelets) that nadir at 4–8 weeks and can last up to 12 weeks. The use of iodine-131 (131I) specifically targets thyroid cancer metastases.

Systemic therapies Definitive palliative treatment of the primary malignancy produces improvement in bone pain in most patients. Hormonal manipulation has a response rate of about 50% in breast cancer-related bone metastases, and 70–80% with prostate cancer. Chemotherapy can be effective in breast cancer and other malignancies. The goals of hormonal manipulation and chemotherapy in skeletal metastatic disease are pain control, disease stabilisation and a reduction in the risk of skeletal morbidity. Bisphosphonates are a bone-targeted therapy. They act by inhibiting osteoclast-mediated bone resorption. Compared with placebo, they reduce the frequency of all skeletal-related events—pain, pathological fracture, hyper-

305

calcaemia and the requirement for radiotherapy. These effects have been most clearly seen with pamidronate and zoledronate in osteolytic bone disease due to breast cancer and multiple myeloma, although zoledronate in particular can be effective in other solid malignancies (including lung, kidney and colorectal).7,49 They are administered as oral or intravenous agents. The pharmacokinetics differ for the two routes of administration, with oral bisphosphonates having low and variable absorption rates. Intravenous bisphosphonates show a biphasic pharmacokinetic pattern, with rapid renal clearance of approximately 50% in the first 24 h and urinary elimination persisting for many months.7 Pamidronate has been a standard of care for treating complications of bone metastases and there is increasing evidence for the efficacy of zoledronate.

Prediction of fracture risk The prediction of fracture risk and prophylactic fixation of impending fractures is controversial. In general, osteolytic lesions are more likely to fracture than osteoblastic metastases. Permeative and focal areas of osteolysis are equally likely to fracture. If more than 50% of the bone diameter is involved, then there is increased risk of fracture. Similarly, an area subject to higher stress is more prone to fracture, examples being the femoral neck, intertrochanteric and subtrochanteric regions of the femur and supracondylar and diaphyseal regions of the humerus. Persisting pain from a metastatic lesion despite medical treatment has also been found to be statistically significant in predicting increased risk of pathological fracture. Cortical lesions extending for more than 2.5 cm are also at increased risk of fracture.6 Harrington3 proposed four risk factors that he felt warranted prophylactic stabilisation based on several studies. 1. 2. 3. 4.

Cortical bone destruction greater than 50%. A lesion of more than 2.5 cm in the proximal femur. A pathological avulsion fracture of the lesser trochanter. Persisting stress pain despite irradiation.

Mirels’ score is a weighted scoring system to analyse the risk of pathological fracture in long, weight-bearing bones. It combines four radiological and clinical risk factors. A score of 4–6 indicates a lesion has a low risk of fracture and can be irradiated safely, but a score of 8 or higher demands prophylactic internal fixation prior to irradiation (Table 2).50

Table 2 Mirels’ scoring system for risk of pathological fracture (1989). Variable

Site Pain XR appearance Size

Score 1

2

3

Upper limb Mild Blastic o1/3

Lower limb Moderate Mixed 1/3–2/3

Peritrochanteric Functional Lytic 42/3

ARTICLE IN PRESS 306

R.U. Ashford et al.

CT and MRI are more accurate at assessing bone destruction by tumour. The risk of a fracture developing depends on the load applied to the bone and the strength of that bone. Bone strength is itself proportional to activity level and lifestyle as well as extent of tumour disease. Hipp and colleagues51 proposed a system whereby an estimate of the load-bearing capacity of a bone can be calculated taking into account defect size, overall bone density and the range of load-bearing requirements. This has been validated for the spine, but not peripheral bones. Van der Linden52 followed 102 patients with 110 femoral metastases and the only risk factors they found to predict fracture were axial cortical involvement greater than 30 mm and circumferential cortical involvement of greater than 50%. They concluded that Mirel’s scoring system would result in unnecessary prophylactic fixation.52 They had, however, excluded patients with suspected impending or actual pathological fracture. All these different methods only serve as a guide. The simplest approach, and the one we use, is that if the lesion is in the weight-bearing skeleton and causes mechanical pain, especially on weight bearing, then it should be stabilised, whereas if it is not then it can be treated by radiotherapy. Wedin,53 in his review of 192 patients treated surgically for skeletal metastases, showed a crude failure rate of 9% for prophylactic fixation compared with 15% for fixation following pathological fracture. They did, however, have two peri-operative deaths directly related to intramedullary nailing of the femur.54

Surgical management Surgical management of appendicular metastases should never be considered in isolation. It should always be part of a multi-disciplinary team approach, including orthopaedic surgeon, medical and clinical oncologist, radiologist and pathologist plus appropriate paramedical and support staff. Immediacy of surgery is less important than pre-operative planning.30 Primary bone tumour should always be excluded, as nailing of a long bone lesion that turns out to be a primary

Table 3

tumour (the ‘‘whoops’’ procedure) is a disaster that often precludes limb-salvage surgery and compromises long-term survival. Surgery may be considered in those lesions that progress following radiotherapy or fail to respond to it. The aim of surgery for metastatic disease is usually the palliation of symptoms. This includes the relief of pain and the restoration of the ability to walk. It is essential to preserve stability and function. Palliative surgery is unlikely to result in an increased duration of survival. Pathological fractures are prone to slow union and any surgery should aim to improve fracture healing or render it unnecessary by endoprosthetic replacement. The introduction of cement into a fracture site, or the use of local radiotherapy, both slow fracture union. Conservative management should be reserved for the bedridden with a poor life expectancy. The surgical construct should be strong enough to allow immediate mobilisation. Capanna et al.6,55 grouped patients into four classes and subdivided the treatment accordingly (Table 3). Surgical management of spinal metastases is indicated for decompression of the spinal cord and/or nerve roots and to stabilise the spine. This is outside the scope of this review and has been covered in a recent edition of this journal.56 As we have outlined earlier, the goal of metastatic surgery is usually to improve mobility and quality of life for the patients remaining time. The surgeon must take into account the degree of functional impairment, the general condition of the patient and their life expectancy. Bauer and Wedin53 reported five positive criteria for survival: absence of visceral metastases, absence of pathological fracture, solitary skeletal metastasis, a primary tumour that is breast, prostate, lymphoma or myeloma, but not lung cancer. Katagiri57 has recently proposed and validated a prognostic scoring system for patients with skeletal metastases (Tables 4 and 5). This enables an estimate to be made as to the patient’s chance of survival and the appropriate type of surgical or non-surgical management to be chosen. With a score of 0–2, a nearly 90% chance of 1-year survival should lead the surgeon towards resection and reconstruction rather than simple osteosynthesis techniques. Pre-operative planning is essential. Hypervascular lesions (renal, thyroid and myeloma) may be subjected to

Surgical options for differing types of metastases (from Capanna 1999 and 2001).6,55 Metastasis

Surgery Aim for long-term cure Wide resection7endoprosthetic replacement (EPR)

II

Solitary metastatic lesion Primary with good prognosis (thyroid, prostate, sensitive breast, renal and colorectal) More than 3 years since primary detected Pathological fracture at any site

III

Impending fracture in weight-bearing bone

IV

Osteoblastic metastasis all sites Osteolytic metastasis non-structural Osteolytic metastasis no impending fracture

I

Proximal femur/humerus—resect and EPR Diaphyseal—a scoring system is used to predict survival and depending on score: simple osteosynthesis, reinforced osteosynthesis or EPR As II Conservative management

ARTICLE IN PRESS Management of metastatic disease of the appendicular skeleton

Table 4

307

Significant prognostic factors and score for each factor (from Katagiri, 2005).57

Prognostic factor

Score

Primary lesion

Rapid growth Slow growth Moderate growth

Hepatocellular, gastric, lung carcinoma Breast, prostate, thyroid carcinoma, multiple myeloma, lymphoma Other carcinoma and sarcoma

Visceral or cerebral metastases ECOG performance status 3 or 4 Previous chemotherapy Multiple skeletal metastases

3 0 2 2 1 1 1

Table 5 Prognostic scores and survival rate (from Katagiri, 2005).57

healing complications and infection. Care should also be paid to tension free soft tissue coverage and closure.

Prognostic score

Amputation

0–2 3–5 6–8

Survival rate (%) 6 months

12 months

24 months

98 71 31

89 49 11

75 28 2

embolisation.58 Median blood loss has been demonstrated to be lower in spinal,17,59 pelvic and appendicular metastases15 from renal cell primaries when embolisation is utilised on the preceding day or day of surgery. The whole bone must be assessed (MRI or PET or isotope bone scan) as an implant that ends at a more distal lesion will be a stress riser. It is also possible to perforate the cortex at a lesion. Polymethyl methacrylate (PMMA) bone cement augmentation of internal fixation is a useful adjunct60 that has little effect on bone healing and provides good symptomatic relief. Thought must also be given to the reconstruction utilised. In their series from our unit, Marsden et al.61 reviewed 69 patients with proximal femoral metastases. Of the 41 that underwent internal fixation, 8 (19.5%) required revision fixation at a mean time of 11 months. This compared with 1 of 28 (3.6%) of those undergoing total hip arthroplasty. Similar results were demonstrated by Wedin53 where a 14% revision rate (at a median time to failure of 8 months) was necessary for osteosynthesis techniques compared with a 2% rate for endoprosthetic replacement. The rate of failure was highest in renal metastases and distal femoral metastases. The early failures were due to poor implant choice (hip screws, sliding hip screw, femoral plate and an unlocked nail). Later failures were due to non-union, stress fractures or tumour progression. In Wedin’s10 series of breast cancer metastases managed throughout Stockholm, there was an 11% (14 of 129) re-operation rate. The most common site for failure was the proximal femur (9 of 60)—of these, seven were treated by osteosynthesis and two by endoprosthesis. Les62 again showed a higher failure rate for intralesional surgery than wide excision, and also a survival advantage for wide excision. It is also desirable to avoid placing incisions at sites of previous radiotherapy to reduce the incidence of wound

Amputation is always an option in metastatic disease. However, for patient psychology and acceptance it is infrequently performed. Indications include, local control of unremitting disease, painful non-union of pathological fractures (either non-treated or those where osteosynthesis techniques have failed), and involvement of skin, soft tissues and neurovascular structures. Severe lymphoedema may also warrant amputation, as may post-irradiation neuropathy and fibrosis. It may also be appropriate where reconstruction is impossible, the limb is functionless, catastrophic bleeding occurs at the tumour site, in the presence of overwhelming sepsis or for pain management. Forequarter amputation for involvement of the brachial plexus and axillary vessels in metastatic tumours is an effective pain relieving procedure.63

Femur Reconstruction of the pelvis and acetabulum is a complex issue outside the scope of this review. There are numerous techniques including cages, saddle prostheses and extracorporeal irradiation and reimplantation.64,65 The topic has been well covered in a review by Choong.2 Appendicular metastases are commonest around the proximal femur and the knee. The proximal femur is the most common site for pathological fracture and demands surgery in all except those with a life expectancy of less than 6 weeks and the bedridden. Algan and Horowitz66 have demonstrated good outcomes in terms of pain relief and function with low morbidity in 29 patients with hip lesions that were metastatic. A pre-operative PET scan, bone scan or MRI scan of the entire femur should be performed to assess dissemination locally. If the acetabulum is not involved, we still advocate a cemented total hip replacement rather than a bipolar prosthesis, although others2,6,20 preferred the inherent stability of the bipolar prosthesis. Calcar replacement prostheses may be necessary, as may tumour endoprostheses (Fig. 6). For a lesion confined to the femoral head, a conventional hip replacement may suffice. When using a tumour endoprosthetic replacement, we choose to

ARTICLE IN PRESS 308

R.U. Ashford et al.

Figure 6 Proximal femoral replacement following resection of the metastasis in Fig. 1.

use a double purse-string technique in the retained hip capsule to enhance stability. We also retain a sliver of greater trochanter in metastatic lesions, freezing it with liquid nitrogen prior to reattaching the trochanteric fragment to the prosthesis. This is of course not an option in primary bone tumours where margins are critical. Retaining the greater trochanter and the abductors also enhances stability. The length of the femoral stem should bypass the most distal lesion by two bone diameters. Tumour endoprostheses should be considered when the proximal bone is too weak to provide stable fixation.1 Pelvic (acetabular) metastases may require complex periacetabular reconstruction at the time of surgery, however, if the lesion is confined to the acetabulum then the management may be restricted to radiotherapy. Internal fixation of metastatic femoral neck fracture is unwise given the association with an unacceptably high risk of further subsequent pathologic fracture and the fact that they rarely unite.10 We avoid the sliding hip screw construct, as it results in stress on the device during ambulation. Tumour progression is highly likely to result in failure of such a fixation device. There is also a stress riser, following radiotherapy, at the distal end of the implant. We therefore prefer to use a cephallo-medullary nail, locked proximally and distally for maximum stability, which is biomechanically superior. In more proximal lesions a conventional arthroplasty or a tumour endoprosthesis is used. For the proximal femur, PMMA augmented internal fixation is no better in terms of failure than simple internal fixation. Endoprosthetic reconstruction remains a better option. Cemented implants are preferred if radiotherapy is to be given (and we advocate radiotherapy in everyone) as the radiotherapy may affect bone ingrowth or ongrowth in uncemented prostheses. A significant proportion of femoral metastases are subtrochanteric or diaphyseal (Fig. 7). Once a primary bone tumour has been excluded, these can be treated by antegrade intramedullary nailing using a cephallo-medullary nail such as the Reconstruction Nail (Smith and Nephew, Cambridge, UK ) or the Long Proximal Femoral Nail (Synthes, Welwyn, UK). These nails are biomechanically stronger than conventional nails. Cardiopulmonary complications are relatively common. Barwood67 reported an incidence of

Figure 7 Lytic metastasis of the femoral diaphysis from a nonsmall cell lung primary. Based on Mirel’s scoring this lesion would score 10 as the patient had functional pain and therefore the lesion was treated by cephallo-medullary nailing.

24% acute oxygen desaturation and hypotension, and of these 27% died. Peak intramedullary pressures during reaming have been shown to reach 450 mmHg, whereas only 100 mmHg can cause embolisation of marrow contents.68,69 In the absence of a fracture, higher pressures occur due to the absence of an ‘‘exhaust vent’’.70 Consideration should therefore be given to venting the femur (somewhere between the site of the proximal and distal interlocking screws) during the nailing procedure, as this reduces intramedullary pressure by more than 50%,71 or alternatively opening the metastasis prior to nailing.54 Venting can also be carried out both proximally and distally72 but does risk extra-skeletal spread of disease. Increases in intramedullary pressures can also be reduced by using only small, progressive increases in reamer diameter and slow advancement of the reamer.68 Cardiopulmonary complications can further be reduced by pulsatile lavage of the medullary canal.72 There are reported incidences of on-table and immediate post-operative deaths where bilateral intramedullary nailings have been performed either in one sitting or staged54,73 as well as single femoral nailings.69,73,74 In Kerr’s series from Bristol, the incidence of death was 100% for bilateral femoral nailings and 13% for unilateral.73 There is little difference in mortality irrespective of whether the nail is inserted reamed or unreamed. Other authors have reported no increase in mortality with bilateral sequential nailings.75 Where there is significant proximal bone stock loss, and stable fixation is unlikely to be achieved, consideration must be given to a proximal femoral replacement. Solitary diaphyseal lesions, particularly from a renal

ARTICLE IN PRESS Management of metastatic disease of the appendicular skeleton

309

or thyroid primary, may be resected and treated by intercalary reconstruction. Prophylactic stabilisation of a diaphyseal lesion is technically easier than fixation after a fracture has occurred and patients frequently make a rapid recovery from surgery. Capanna has introduced a scoring system for diaphyseal lesions based on survival, biomechanics, the defect size and response to adjuvant therapy. Dependent on the score, the treatment recommended was either osteosynthesis or implant (endoprosthetic reconstruction). Osteosynthesis was subdivided into minimal (Rush nails), simple (locked intramedullary nails) or reinforced (locked intramedullary

Figure 9 (A) Distal femoral metastasis from metastatic malignant melanoma treated by (B) distal femoral replacement.

nails with PMMA cement augmentation). This system is yet to be validated. Distal diaphyseal femoral lesions have been treated successfully by retrograde nailing with good pain relief and function achieved in most patients, although its use should not be over extended. Short nails are inappropriate as they will lead to stress risers and therefore a long nail should be inserted. Curettage and supplementary internal fixation is also an option in the distal femur (Fig. 8). Additional local measures can be utilised after curettage to sterilise the margins, including phenol, alcohol, and cryotherapy. Cryotherapy with liquid nitrogen has been shown to be the most effective. PMMA bone cement can also be added prior to internal fixation to stabilise the construct with successful results.60 It is also possible to add chemotherapeutic agents, appropriately targeted, to the PMMA cement without disturbing the mechanical properties of the cement76,77 or the effects of the chemotherapeutic agent.76 In the distal femur if more than 50% of the epiphysis or metaphysis is replaced by tumour then endoprosthetic replacement is our preferred treatment (Fig. 9). We have used a rotating hinge prosthesis with good effect.

Tibia

Figure 8 (A) Lytic lesion in the distal femur (renal cell metastasis). Initial treatment (B) was a locked cephallomedullary nail. Ongoing pain and tumour progression necessitated a further procedure and the lesion was curetted and fixation augmented with PMMA cementation (C).

Endoprosthetic replacement of the proximal tibia, whilst an option, is not as successful as in the distal femur, principally because of problems with the soft tissue cover. Where possible we try to retain the tibial tubercle and the attached patellar tendon. We also swing a medial head of gastrocnemius as a flap around the front of the prosthesis to cover the implant. This often necessitates a split skin graft to obtain skin cover. Internal fixation with a tibial nail must be sufficiently sound as to allow early weight bearing. Therefore, nails should be locked both proximally and distally. It may be necessary to augment fixation with PMMA, and the whole bone must be addressed. If less than half the epiphysis or distal tibial metaphysis is involved by tumour then satisfactory outcomes (pain relief, avoidance of fracture) can be obtained with curettage, cementation and plate fixation.

ARTICLE IN PRESS 310

R.U. Ashford et al.

Foot and ankle Metastases distal to the knee are far less common than those occurring more proximally. Usually radiotherapy suffices for most metastases but pain can be unremitting and resistant to radiotherapy. Options for treatment include local curettage with either bone grafting or cementation (Fig. 10) or below knee amputation. Pathologic fractures can usually be treated in a cast with supplemental radiotherapy. Metastatic lung cancer can affect the small bones of the foot (Fig. 11) and is often rapidly destructive. Radiotherapy is the first line treatment.

Humerus Endoprosthetic replacement of the proximal humerus (Fig. 12A and B), whilst sacrificing function, is effective at

Figure 11 Lung metastases affecting the distal phalanx of the great toe.

Figure 12 (A) Metastatic prostatic carcinoma of the proximal humerus treated by (B) proximal humeral replacement. Figure 10 (A) Isotope bone scan and (B) CT scan demonstrating large lytic lesion in the talus (metastatic breast cancer) treated by (C) curettage and cementing.

relieving pain. If the lesion is confined to the humeral head then a conventional shoulder hemiarthroplasty can be utilised. Infection and instability can be problems after proximal

ARTICLE IN PRESS Management of metastatic disease of the appendicular skeleton

311

Figure 13 (A) Destructive renal cell metastasis of the distal humerus (B) treated by PMMA cement augmentation and plating.

humeral replacement. The main indications for prosthetic replacement are inadequate bone to stabilise a fracture or excessive bone destruction. It is important to reattach the muscles of the rotator cuff. Again we use a double purse-string reconstruction of the retained capsule to enhance stability if there is sufficient capsule. Where there is insufficient capsule we use polypropylene mesh augmentation of the cuff and capsule. Excision of a metastasis without reconstruction may be appropriate in some patients and is cosmetically more acceptable than a forequarter amputation. Intramedullary nailing, either antegrade or retrograde can treat the majority of metastatic lesions in the humerus. The nail should provide sufficient stability for early mobilisation. Subsequently radiotherapy is utilised. If the patient has a very limited life expectancy then a humeral brace may be sufficient. The distal humerus is more complex. Endoprosthetic replacements are available but infrequently utilised. More

Figure 14 (A) Multiple myeloma bone deposits throughout the humerus treated by (B) flexible intramedullary nails.

ARTICLE IN PRESS 312

R.U. Ashford et al.

Figure 15 Destructive lesion in the distal radius from metastatic non-small cell carcinoma of the lung.

commonly plating augmented with PMMA bone cement is adequate (Fig. 13). Flexible intramedullary nailing is a technique usually reserved for children. However, in exceptional cases the technique may be used for pathological humeral fractures. If a metastatic lesion is very distal or very extensive, this may preclude both antegrade and retrograde conventional nailing. Stability can be provided by two flexible nails inserted retrograde through the epicondyles, enabling recovery of function78 (Fig. 14).

Radius, ulna, carpus and hand Metastases in the forearm bones are uncommon (Fig. 15). Small lesions can be treated by curettage, cementation and conventional plating. More extensive lesions can be resected. It may simply be necessary to biopsy the lesion to establish a diagnosis and manage the fracture nonoperatively in a plaster cast. Metastases affecting the small bones of the wrist and hand warranting surgery are unusual. Radiotherapy may, however, result in post-operative contractures and functional impairment.

Risks and complications of surgery Patients with skeletal metastases tend to have co-morbidities increasing the general risks of surgery and anaesthesia. There are other more specific risks also worthy of consideration: (1) Embolism: particularly in those patients with metastases affecting the long bones and on reaming or insertion of an intramedullary nail. Tumour emboli or fat emboli can result in cardiac arrest. Trans-oesophageal echocardiography can be diagnostic. (2) Infection: chemotherapy and radiotherapy especially pre-operative but also post-operative can make patients susceptible to infection. We use a 6 week course of antibiotics when we implant a tumour endoprosthesis. (3) Haemorrhage: the risk of haemorrhage can be reduced

Figure 16 (A) Radiotherapy-induced osteonecrosis resulting in pathological fracture of the femoral neck in a patient with metastatic oesophageal carcinoma. (B) MRI scan showing avascular necrosis but no active tumour. (C) Reconstruction with cemented total hip arthroplasty with roof reinforcing ring.

by judicious use of pre-operative embolisation of tumours likely to be highly vascular (e.g. renal cell metastases). We try to embolise either on the morning of surgery or the preceding day. (4) Dislocation: tumour endoprostheses have a reputation for dislocation. Often this is because of the degree of

ARTICLE IN PRESS Management of metastatic disease of the appendicular skeleton

Figure 17 CT scan demonstrating radiotherapy-induced osteonecrosis resulting in a complex pathological fracture through the sacrum.

soft tissue sacrifice necessary to obtain tumour clearance. Usually surgery for metastatic disease is palliative and therefore such radical margins are not necessary. We try to reduce our dislocation rate by a double pursestring of the capsule (in the hip and shoulder) as well as retaining the hip abductors by freezing the remnant of greater trochanter with liquid nitrogen. We also utilise polypropylene mesh to augment the capsule where necessary (particularly in the shoulder). (5) Radiotherapy-related complications: infection has already been mentioned but if the native bone remains, and is part of the radiation field, avascular necrosis can occur (Fig. 16). This can result in pathological fracture. There is a theoretical increase in the rate of stress fractures or nonunion but this complication developed in only 13/136 patients in the series by Wedin.53 Radiotherapy-induced osteonecrosis can be marked, verging on the untreatable, as illustrated in Fig. 17.

Conclusion The management of appendicular skeletal metastases requires a multi-disciplinary approach. It is essential to exclude a primary bone tumour prior to employing osteosynthesis techniques. Impending fractures are technically easier to stabilise than displaced fractures, but prophylactic stabilisation remains somewhat controversial. Prognosis is important in determining the appropriate surgical treatment with simple measures for those with the poorest prognosis and resection and reconstruction for those expected to survive more than 1 year.

References 1. Damron TA, Sim FH. Operative treatment for metastatic disease of the pelvis and proximal end of the femur. J Bone Joint Surg 2000;82A:113–26.

313

2. Choong PFM. Reconstructive surgery following primary and secondary tumours of the hip. J Orthop Surg 2000;8(2):83–94. 3. Harrington KD. Orthopaedic management of metastatic bone disease. St. Louis: Mosby; 1988. p. 7. 4. Aaron AD. The management of cancer metastatic to bone. J Am Med Assoc 1994;272(15):1206–9. 5. Berruti A, Dogliotti L, Bitossi R, et al. Incidence of skeletal complications in patients with bone metastatic prostate cancer and hormone refractory disease: predictive role of bone resorption and formation markers evaluated at baseline. J Urol 2000;164:1248–53. 6. Capanna R, Campanacci DA. The treatment of metastases in the appendicular skeleton. J Bone Joint Surg 2001;83B(4):471–81. 7. Blum RH, Novetsky D, Shasha D, Fleishman S. The multidisciplinary approach to bone metastases. Oncology 2003; 17(6):845–57. 8. Du ¨rr HR, Mu ¨ller PE, Lenz T, Baur A, Jansson V, Refior HJ. Surgical treatment of bone metastases in patients with breast cancer. Clin Orthop 2002;396:191–6. 9. Coleman RE, Rubens RD. The clinical course of bone metastases from breast cancer. Br J Cancer 1987;55:61–6. 10. Wedin R, Bauer HCF, Rutquist L-E. Surgical treatment for skeletal breast cancer metastases. Cancer 2001;92(2): 257–62. 11. Koenders PG, Beex LVAM, Kloppenborg PWC, Smals AGH, Benraad THJ. Human breast cancer: survival from first metastasis. Breast Cancer Res Treat 1992;21:173–80. 12. Baloch KG, Grimer RJ, Carter SR, Tillman RM. Radical surgery for the solitary metastasis from renal cell carcinoma. J Bone Joint Surg 2000;82B:62–7. 13. Jung ST, Ghert MA, Harrelson JM, Scully SP. Treatment of osseous metastases in patients with renal cell carcinoma. Clin Orthop Relat Res 2003;409:223–31. 14. Fuchs B, Trousdale RT, Rock MG. Solitary bony metastasis from renal cell carcinoma: significance of surgical treatment. Clin Orthop Relat Res 2005;431:187–92. 15. Rowe DM, Becker GJ, Rabe FE, et al. Osseous metastases from renal cell carcinoma: embolization and surgery for restoration of function. Radiology 1984;150:673–6. 16. Sun S, Lang EV. Bone metastases from renal cell carcinoma: preoperative embolization. J Vasc Interv Radiol 1998;9–2: 263–9. 17. Manke C, Bretschneider T, Lenhart M, et al. Spinal metastases from renal cell carcinoma: effect of preoperative particle embolization on intraoperative blood loss. Am J Neuroradiol 2001;22:997–1003. 18. Halperin EC, Harisiadis L. The role of radiation therapy in the management of metastatic renal cell carcinoma. Cancer 1983;51:614–7. 19. Lee J, Hodgson D, Chow E, et al. A phase II trial of palliative radiotherapy for metastatic renal cell carcinoma. Cancer 2005;104(9):1894–900. 20. Bo ¨hm P, Huber J. The surgical treatment of bony metastases of the spine and limbs. J Bone Joint Surg 2002;84B(4):521–9. 21. Althausen P, Althausen A, Jennings LC, Mankin HJ. Prognostic factors and surgical treatment of osseous metastases secondary to renal cell carcinoma. Cancer 1997;80(6):1103–9. 22. Hansen BH, Keller J, Laitinen M, et al. The Scandinavian Sarcoma Group skeletal metastasis register. Acta Orthop Scand 2004;75(Suppl 311):11–5. 23. Pittas AG, Adler M, Fazzari M, et al. Bone metastases from thyroid carcinoma: clinical characteristics and prognostic variables in one hundred forty-six patients. Thyroid 2000; 10(3):261–8. 24. Vitale G, Fonderico F, Martignetti A, et al. Pamidronate improves the quality of life and induces clinical remission of bone metastases in patients with thyroid cancer. Br J Cancer 2001;84(12):1586–90.

ARTICLE IN PRESS 314 25. Hatfield DR, DeLand FH, Maruyama Y. Skeletal metastases in pancreatic carcinoma: study by isotopic bone scanning. Oncology 1976;33(1):44–7. 26. Kanthan R, Loewy J, Kanthan SC. Skeletal metastases in colorectal carcinomas: a Saskatchewan profile. Dis Colon Rectum 1999;42(12):1592–7. 27. Besbeas S, Stearns Jr. MW. Osseous metastases from carcinomas of the colon and rectum. Dis Colon Rectum 1978;21(4):266–8. 28. Nakanishi H, Araki N, Kuratsu S, Narahara H, Ishikawa O, Yoshikawa H. Skeletal metastasis in patients with gastric cancer. Clin Orthop Relat Res 2004;(423):208–12. 29. Abdul-Karim FW, Kida M, Wentz WB, Carter JR, Sorensen K, Macfee M, et al. Bone metastasis from gynecologic carcinomas: a clinicopathologic study. Gynecol Oncol 1990;39(2):108–14. 30. Tillman RM. The role of the orthopaedic surgeon in metastatic disease of the appendicular skeleton. Working Party on Metastatic Bone Disease in Breast Cancer in the UK. J Bone Joint Surg 1999;81B(1):1–2. 31. Galasko CSB. Principles of management of metastatic disease. Curr Orthop 1988;2:158–67. 32. Berman AT, Hermantin FU, Horowitz SM. Metastatic disease of the hip: evaluation and treatment. J Am Acad Orthop Surg 1997;5(2):79–86. 33. Rougraf BT, Kniesl JS, Simon MA. Skeletal metastases of unknown origin. A prospective study of a diagnostic strategy. J Bone Joint Surg 1993;75A:1276–81. 34. Tong D, Gillick L, Henderson F. The palliation of symptomatic osseous metastases. Cancer 1982;50:898–9. 35. Roos D. Continuing reluctance to use single fractions of radiotherapy for metastasis to bone: an Australian and New Zealand practice survey and literature review. Radiother Oncol 2000;56:315–22. 36. Jeremic B. Single fraction external beam radiation therapy in the treatment of localized metastatic bone pain. A review. J Pain Symptom Manage 2001;22:1048–58. 37. McQuay H, Collins S, Carroll D. Radiotherapy for the palliation of painful bone metastases (Cochrane review). The Cochrane Library, UK; 2002. 38. Wu J, Wong L, Johnston M. Meta-analysis of dose-fractionation radiotherapy trials for the palliation of painful bone metastases. Int J Radiat Oncol Biol Phys 2003;55: 594–605. 39. Jeremic B, Shibamoto Y, Acimovic L, et al. A randomized trial of 3 single dose radiation therapy regimens in the treatment of metastatic bone pain. Int J Radiat Oncol Biol Phys 1998;42: 161–7. 40. Hoskin P, Price P, Easton D, et al. A prospective randomized trial of 4 Gy or 8 Gy single doses in the treatment of metastatic bone pain. Radiother Oncol 1992;23:74–8. 41. Steenland E, Leer J, Van Houwelingen H, et al. The effect of a single fraction compared to multiple fractions on painful bone metastases: a global analysis of the Dutch bone metastasis study. Radiother Oncol 1999;52:101–9. 42. Van der Linden Y, Lok J, Steenland E, et al. Single fraction radiotherapy is efficacious: a further analysis of the Dutch Bone Metastasis Study controlling for the influence of retreatment. Int J Radiat Oncol Biol Phys 2004;59:528–37. 43. Mithal N, Needham P, Hoskin P. Retreatment is radiotherapy for painful bone metastases. Int J Radiat Oncol Biol Phys 1994; 29:1011–4. 44. Jeremic B, Shibamoto Y, Igrutinovic I. Single 4Gy Re-irradiation for painful bone metastasis following single fraction radiotherapy. Radiother Oncol 1999;52:123–7. 45. Blitzer P. Reanalysis of the RTOG study of the palliation of symptomatec osseous metastases. Cancer 1985;55:1468–72. 46. Salazar O, Rubin P, Hendrikson F, et al. Single dose half-body irradiation for palliation of multiple metastases from solid tumours: RTOG report. Cancer 1986:58–9.

R.U. Ashford et al. 47. Ben-Josef E, Porter AT. Radioisotopes in the treatment of bone metastases. Ann Med 1997;29:31–5. 48. Quilty P, Kirk D, Bolger J, et al. A comparison of the palliative effects of Strontium-89 and external beam radiotherapy in metastatic prostate cancer. Radiother Oncol 1994; 31:33–40. 49. Clines GA, Guise TA. Mechanisms and treatment for bone metastases. Clin Adv Hemat Oncol 2004;2(5):295–302. 50. Mirels H. Metastatic disease in long bones. A proposed scoring system for diagnosing impending pathologic fractures. Clin Orthop Rel Res 1989;249:256–64. 51. Hipp JA, Springfield DS, Hayes WC. Predicting pathologic fracture risk in the management of metastatic bone defects. Clin Orthop Relat Res 1995;312:120–35. 52. Van der Linden YM, Dijkstra PDS, Kroon HM, et al. Comparative analysis of risk factors for pathological fracture with femoral metastases. J Bone Joint Surg 2004;86-B:566–73. 53. Wedin R, Bauer HC, Wersall P. Failures after operation for skeletal metastatic lesions of long bones. Clin Orthop Relat Res 1999;(358):128. 54. Persson EV, Bauer HCF. Sudden hypotension and profuse bleeding during intramedullary nailing of the femur in cancer patients. Acta Orthop Scand 1994;65:564–7. 55. Capanna R, Campanacci D, Martelli L, Michels S, Bacci G. The treatment of metastases in bone. European Instructional Course Lectures, Vol. 4. 1999. 56. Finkelstein JA, Ford MH. Diagnosis and management of pathological fractures of the spine. Curr Orthop 2004;18(5): 396–405. 57. Katagiri H, Takahashi M, Wakai K, Sugiura H, Kataoka T, Nakanishi K. Prognostic factors and a scoring system for patients with skeletal metastases. J Bone Joint Surg 2005;87B:698–703. 58. Barton PP, Waneck RE, Karnel FJ, Ritschl P, Kramer J, Lechner GL. Embolization of bone metastases. J Vasc Intervent Radiol 1996;7:81–8. 59. Roscoe MW, McBroom RJ, St Louis E, Grossman H, Perrin R. Preoperative embolization in the treatment of osseous metastases from renal cell carcinoma. Clin Orthop Relat Res 1989;238:302–7. 60. Harrington KD, Sim FH, Enis JE, Johnston JO, Diok HM, Gristina AG. Methylmethacrylate as an adjunct in internal fixation of pathological fractures: experience with three hundred and seventy-five cases. J Bone Joint Surg 1976;58A:1047–55. 61. Marsden FW, Hale DG, Sullivan JA, Findlay MPM. In: Langlais F, Tomeno B, editors. Treatment of metastatic disease involving the proximal femur and hip joint. Major reconstructions in oncologic and non-tumoral conditions. Berlin, Heidelberg: Springer; 1991. 62. Les KA, Nicholas RW, Rougraff B, Wurtz D, Vogelzang NJ, Simon MA, et al. Local progression after operative treatment of metastatic kidney cancer. Clin Orthop Relat Res 2001;390:206–11. 63. Wittig JC, Bickels J, Kollender Y, Kellar-Graney KL, Meller I, Malawer MM. Palliative forequarter amputation for metastatic carcinoma to the shoulder girdle region: indications, preoperative evaluation, surgical technique, and results. J Surg Oncol 2001;77(2):105–13. 64. Davidson AW, Hong A, McCarthy SW, Stalley PD. En bloc resection, extra-corporeal irradiation, and re-implantation in limb salvage for bony malignancies. J Bone Joint Surg 2005; 87B:851–7. 65. Hong A, Stevens G, Stalley P, et al. Extracorporeal irradiation for malignant bone tumours. Int J Radiat Oncol Biol Phys 2001;50:441–7. 66. Algan SM, Horowitz SM. Surgical treatment of pathological hip lesions in patients with metastatic disease. Clin Orthop Relat Res 1996;332:223–31. 67. Barwood SA, Wilson JL, Molnar RR, Choong PFM. The incidence of acute cardiorespiratory and vascular dysfunction following

ARTICLE IN PRESS Management of metastatic disease of the appendicular skeleton

68.

69.

70.

71.

72.

intramedullary nail fixation of femoral metastases. Acta Orthop Scand 2000;71(2):147–52. Peter RE. Influence of the reamer shape on intraosseous pressure during closed intramedullary nailing of the unbroken femur: a preliminary report. Injury 1993;24(Suppl 3):S48–55. Peter RE, Schopfer A, Le Coultre B, Hoffmeyer P. Fat embolism and death during prophylactic osteosynthesis of a metastatic femur using an unreamed femoral nail. J Orthop Trauma 1997;11(3):233–4. Assal M, Zanone X, Peter RE. Osteosynthesis of metastatic lesions of the proximal femur with a solid femoral nail and interlocking spiral blade inserted without reaming. J Orthop Trauma 2000;14(6):394–7. Roth SE, Rebello MM, Kreder H, Whyne CM. Pressurization of the metastatic femur during prophylactic intramedullary nail fixation. J Trauma 2004;57:333–9. Choong PFM. Cardiopulmonary complications of intramedullary fixation of long bone metastases. Clin Orthop Relat Res 2003;415S:S245–53.

315

73. Kerr PS, Jackson, Atkins RM. Cardiac arrest during intramedullary nailing for femoral metastases. J Bone Joint Surg 1993; 75(6):972–3. 74. Karachalios T, Geeurickx A, Newman JH. Fatal air embolism after prophylactic intramedullary nailing. J Bone Joint Surg 1992;74A(7):1101–2. 75. Cole AS, Hill GA, Theologis TN, Gibbons CLMH, Willett K. Femoral nailing for metastatic disease of the femur: a comparison of reamed and unreamed femoral nailing. Injury 2000;31(1):25–31. 76. Wang HM, Galasko CS, Crank S, Oliver G, Ward CA. Methotrexate loaded acrylic cement in the management of skeletal metastases. Biomechanical, biological and systemic effect. Clin Orthop Relat Res 1995;312:173–86. 77. Healey JH, Shannon F, Boland P, DiResta GR. PMMA to stabilize bone and deliver antineoplastic and antiresorptive agents. Clin Orthop Relat Res 2003;415S:S263–75. 78. Dalal S, Gibson RJ. Pathological fracture of the distal humerus treated with Nancy nails. Injury 2004;353:327–8.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 316–318

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

CME SECTION

Three CME points available The following series of questions are based on the CME designated article for this issue—‘‘Management of metastatic disease of the appendicular skeleton’’ by Ashford, Pendlebury and Stalley. Please read the article carefully and then complete the self-assessment questionnaire by filling in the square corresponding to your response to each multiple-choice question. For true or false questions, please fill in one square only. After completing the questionnaire, either post or fax the answer page back to the Current Orthopaedics Editorial Office at the address at the bottom of the RESPONSE sheet. Please photocopy this page if you wish to keep your copy of Current Orthopaedics intact. Replies received before the next issue of Current Orthopaedics is published will be marked and those reaching an adequate standard will qualify for three external CME points. You will be notified of your marks and a CME certificate will be despatched for your records.

Questions 1. Approximately, what percentage of patients who die with breast or prostate cancer are found to have skeletal metastases at post-mortem? a. b. c. d. e.

15% 25% 50% 75% 85%

2. Which of the following is not a predictor of skeletal morbidity in patients with prostatic cancer? a. b. c. d. e.

The extent of bone involvement Serum alkaline phosphatase level Serum acid phosphatase level Urinary deoxypyridinoline levels The severity of pain

3. What is the most likely reason for a patient with welldifferentiated breast cancer to require orthopaedic surgery as a consequence of their disease? a. b. c. d. e.

Treatment of pathological spinal fracture Relief of spinal cord compression Restoration of the hip damaged by acetabular disease Treatment of a femoral metastasis Treatment of humeral or clavicular metastases

doi:10.1016/j.cuor.2006.07.003

4. A patient is found to have a metastatic deposit in the second metacarpal. What is the most likely source? a. b. c. d. e.

Breast Lung Kidney Prostate Thyroid

5. After the femur, which bone is the commonest site for pathological fracture due to metastatic disease? a. b. c. d. e.

Clavicle Humerus Metacarpal Pubic ramus Tibia

6. An analogue of which compound is administered then detected when investigating patients for metastatic disease using Positron Emission Tomography a. b. c. d. e.

Glucose Technetium Appatite Hydroxyproline Gadolinium

7. What is the preferred technique for performing biopsy of a tumour in bone?

ARTICLE IN PRESS CME SECTION a. Core needle biopsy through a stab incision crossing one compartment only b. Core needle biopsy after opening the skin along an incision that can later be used for excision, approaching the bone along an intercompartmental plane c. Open biopsy of the margin of the tumour through a small incision crossing one compartment only d. Open biopsy of the margin of the tumour, approaching the tumour along an intercompartmental plane e. Open biopsy through an incision that can later be used for definitive surgery, sampling the core of the tumour

317 a. Lower limb blastic disease affecting 1/2 the bone diameter with mild pain b. Upper limb mixed lytic/blastic disease affecting o1/3 the bone diameter with moderate pain c. Peritrochanteric blastic lesion affecting o1/3 the bone diameter with mild pain d. Upper limb lytic disease affecting o1/3 the bone diameter with mild pain e. Lower limb mixed lytic/blastic disease affecting 1/3 the bone diameter with moderate pain

8. Which cancer shows the best response to retreatment with radiotherapy for bone pain after relapse following previous successful radiotherapy for the same indication?

11. Intramedullary pressures of about 100 mmHg result in bone marrow embolisation. What pressures are reached during standard unvented reaming using conventional techniques?

a. b. c. d. e.

a. b. c. d. e.

Thyroid Lung Kidney Breast Prostate

20 mmHg 50 mmHg 100 mmHg 200 mmHg 4250 mmHg

9. Which of the following is not one of the risk factors for pathological fracture identified by Harrington as warranting prophylactic stabilisation?

12. When should embolisation of renal metastases take place in relation to planned surgery?

a. Cortical bone destruction 450% b. Neurological symptoms without signs in a patient with early vertebral collapse c. A pathological avulsion fracture of the lesser trochanter d. A lesion of more that 2.5 cm in the proximal femur e. Persisting stress pain despite radiation

a. At least 6 weeks beforehand with confirmation of tumour shrinkage by rescan b. At least 6 weeks beforehand with no scan necessary to confirm shrinkage c. Two weeks beforehand d. On the morning of surgery or during the preceeding day e. Immediately prior to surgery

10. According to Mirel’s score, which of the following patients with metastatic disease and pain requires prophylactic fixation rather than radiotherapy alone?

ARTICLE IN PRESS 318

CME SECTION

Please fill in your answers to the CME questionnaire above in the response section provided below. A return address and fax number is given at the bottom of the page. ...............................................................................................

Responses Please shade in the square for the correct answer. 1 2 3 4 5 6 7 8 9 10 11 12

A A A A A A A A A A A A

& & & & & & & & & & & &

B B B B B B B B B B B B

& & & & & & & & & & & &

C C C C C C C C C C C C

& & & & & & & & & & & &

D D D D D D D D D D D D

& & & & & & & & & & & &

Your details (Print clearly) NAME. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ADDRESS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . FAX NO. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . EMAIL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . RETURN THE COMPLETED RESPONSE FORM by fax to +44-113-206-6791, or by post to CME, Current Orthopaedics, Orthopaedic Surgery, Clinical Sciences Building, St. James’s University Hospital, Beckett Street, Leeds LS9 7TF, UK.

E E E E E E E E E E E E

& & & & & & & & & & & &

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 319

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

CME SECTION Answers to CME questions in Vol. 20, issue 2 Please find below the answers to the Current Orthopaedics CME questions from Vol. 20, issue 2 which were based on the article—‘‘Meniscal Tears’’ by I. McDermott. 1

A&

B’

C&

D&

E&

2

A&

B&

C&

D’

E&

3

A’

B&

C&

D&

E&

4

A&

B&

C’

D&

E&

5

A’

B&

C&

D&

E&

6

A&

B&

C&

D&

E’

7

A&

B&

C’

D&

E&

8

A&

B&

C&

D’

E&

9

A&

B&

C&

D&

E’

10

A&

B&

C’

D&

E&

11

A&

B’

C&

D&

E&

12

A’

B&

C&

D&

E&

doi:10.1016/j.cuor.2006.02.004

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 320

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

BOOK REVIEWS James Bono, Richard Scott (Eds.), Total Knee Arthoplasty, Springer, Berlin, ISBN 0387223525, 2005 (285pp., £111.50). This book is a superbly presented glossy tome which aims to cover all aspects of revision knee arthroplasty from diagnosis and evaluation of the failed knee, to general principles of surgery and special situations. Most of the chapters are multi-authored but in general has been edited to have consistency of reading style. All contributors bar one are from the USA. Amongst them are some global names in knee surgery. The book is clearly laid out into three main sections, subdivided into chapters with clear subject headings, copious photographs and diagrams, many in colour. As a practising knee surgeon I found that there are numerous useful technical tips and a practical approach throughout the book. One gets the impression that the chapter content has been well thought out with practicality in mind. For example, a chapter on the management of the extensor mechanism during revision tries to address ques-

tions which we would all like to ask: such as, what releases to do, what to do with a patella remnant, how to manage the ruptured patellar tendon. All chapters are thoroughly referenced though as is the nature with multi authored books of long gestation, I could not find any references later than 2003. This is a minor and unavoidable issue. There is a useful index but I did find it to be deficient in some areas: for example, though there are excellent sections within the book on soft tissues and balancing the knee, it was not easy to look up. This overall is an excellent clear text, laid out attractively and logically. I was unable to think of any situation I have come across which was not addressed clearly within its pages. Consultants with an interest in revision knee arthroplasty, particularly newly appointed, would do well to consider acquiring a copy of this book and it will soon become well thumbed in the Post-Graduate Library. Highly recommended.

Stuart Calder

doi:10.1016/j.cuor.2006.06.004

‘Shoulder Arthroplasty’ by Louis Bigliani and Evan Flatow, Springer, Berlin, ISBN 0387223363, 2005 (214pp., £111.50). This short but helpful textbook is written primarily by a New York Faculty, though the late Ian Kelly and his SpR from Glasgow have contributed one of the eight chapters. Although each chapter is written by a different team of senior and junior specialists, the sequencing is logical. Preoperative evaluation and surgical approaches appear as chapter 1 whilst the final chapter deals with rehabilitation. There is no separate discussion of outcomes, as chapters on inflammatory and non-inflammatory arthropathy, trauma and cuff deficiency each deal with results relevant to the indication. The quality of illustrations is good in general, many being in colour. In places operative photographs could be clarified by the use of accompanying line diagrams but these are rarely used. The contributions are well written and clear, providing the level of information that is appropriate for a doi:10.1016/j.cuor.2006.07.002

surgeon leaving a training programme and for those with a general practice in orthopaedic and trauma surgery. The price that is paid for this being a short but concise text is that each author describes their favoured methods without detailed consideration of alternatives. This has some disadvantages for those working outside the United States and this is nowhere more apparent than in the chapter on rotator cuff deficiency, where scant mention is made of the so-called reverse shoulder arthroplasty. The chapter on rehabilitation is an excellent summary that would be useful for physiotherapists. However, for the reasons described above it appears as a prescriptive solution when the authors point out that in reality the rehabilitation has to be individualised and reactive. This book will be most useful for trainees considering a period of fellowship training in shoulder surgery and for those preparing for specialty examinations.

David Limb

Aims and Scope Current Orthopaedics presents a unique collection of international review articles summarizing the current state of knowledge and research in orthopaedics. Each issue focuses on a specific topic, discussed in depth in a mini-symposium; other articles cover the areas of basic science, medicine, children/adults, trauma, imaging and historical review. There is also an annotation, self-assessment questions and an exam section. In this way, the entire postgraduate syllabus will be covered in a 4-year cycle. The Journal is cited in: Cochrane Center, EMBASE/ Excerpta Medica, Infomed, Reference Update and UMI Microfilms.

Editor Professor R. A. Dickson MA, ChM, FRCS, DSc St James’s University Hospital Trust, Leeds, UK

Editorial Committee President of BOTA, M. A. Farquharson-Roberts (Gosport, UK), I. Leslie (Bristol, UK), D. Limb (Leeds, UK), M. Macnicol (Edinburgh, UK), I. McDermott (Ruislip, UK), J. Rankine (Leeds, UK)

Editorial Advisory Board

Amsterdam

K

Boston

K

Jena

L. de Almeida (Portugal) G. P. Songcharoen (Thailand) R. W. Bucholz (USA) J. W. Frymoyer (USA) R. W. Gaines (USA) S. L. Weinstein (USA) M. Bumbasirevic (former Yugoslavia)

A. K. Mukherjee (India) A. Kusakabe (Japan) A. Uchida (Japan) M.-S. Moon (Korea) R. Castelein (The Netherlands) R. K. Marti (The Netherlands) G. Hooper (New Zealand) A. Thurston (New Zealand) E. G. Pasion (Philippines)

D. C. Davidson (Australia) J. Harris (Australia) S. Nade (Australia) G. R. Velloso (Brazil) J. H. Wedge (Canada) S. Santavirta (Finland) P. N. Soucacos (Greece) M. Torrens (Greece) J. C. Y. Leong (Hong Kong)

K

London

K

New York

K

Oxford

K

Paris

K

Philadelphia

K

San Diego

K

St Louis

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 321

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

EDITORIAL

Bullet and blast injuries These are four excellent papers on this very important topic. Necessarily there is a fair amount of repetition but Michael Farquharson-Roberts, who commissioned this mini-symposium, and I, as the Editor, felt that a degree of repetition was indeed desirable and reinforced some of the many important messages to take on board from this mini-symposium. The

0268-0890/$ - see front matter & 2006 Published by Elsevier Ltd. doi:10.1016/j.cuor.2006.08.002

difference between civilian and military life seems to be getting ever narrower and so there is a lot for everyone to learn from these hands on experts.

Robert A. Dickson

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 322–332

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: BULLET AND BLAST INJURIES

(i) An overview of the pathophysiology of gunshot and blast injury with resuscitation guidelines S.A. Stapley, L.B. Cannon Ministry of Defence Hospital Unit, Portsmouth and Portsmouth NHS Trust, UK

KEYWORDS Gunshot injuries; Ballistic principles; Low energy wounds; High energy wounds; Blast injury; Blast physics

Summary Because of the blurring of the ‘front line’ the knowledge and skills once only the province of the military surgeon is now required by the civilian trauma team. The mechanisms of injury and basic resuscitative principles are set out. & 2006 Elsevier Ltd. All rights reserved.

Introduction Recent years have seen an increase in gunshot wounding and blast injury in the UK associated with criminal and terrorist activities. Injury patterns previously confined to the military environment are now seen in civilian practice. Thus any doctor treating trauma victims should have a basic understanding of the mechanisms involved in such injuries. This overview intends to provide a short explanation of the biophysics behind these injury mechanisms, the pathophysiology of the main organ systems affected and early resuscitation guidelines.

Gunshot injuries Basic ballistics Ballistics is the study of the motion of an object when it has been launched, shot, hurled, thrown, etc. or by other means Corresponding author.

0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.07.008

projected such as bullets. The wounding effects of projectiles are determined by:


the physical properties of the missile (mass, shape, composition),


the flight characteristics (velocity, stability and yaw or tumbling characteristics). In addition the reaction of the differing types of tissue the missile encounters affect wounding outcome. Modern cartridges (Fig. 1) have in one container the components necessary to propel a bullet from a gun. As the cartridge is fired the burning powder produces a gas. As the volume of this gas increases, the internal pressure within the barrel of the gun rises, forcing the bullet down the barrel. As it does so, it is engaged by the spiral grooving or rifling of the barrel causing it to spin. This stabilises the flight of the projectile, maintaining its trajectory, range and accuracy. Velocity is the single most important factor in creating a wound usually classified as either low-velocity (o600 m/s or o2000 ft/s) or high-velocity (4600 m/s or 42000 ft/s)

ARTICLE IN PRESS An overview of the pathophysiology of gunshot and blast injury

323 through the air is also affected by yaw. Yaw is produced by the inherent asymmetry of any bullet as it deviates from its longitudinal axis during flight. This is identified mostly during the initial and final phases of flight producing phases of instability and thus deceleration of the round. The rifling of the barrel reduces yaw to a minimum. The design and composition of a bullet varies. Usually they have a lead core to increase mass, which may or may not be jacketed (enveloped) with copper. A bullet is described as having a nose or point, a body and a base or heel. The nose may be jacketed or not and can be rounded, pointed, flat, hollow, semi-hollow or full. A rounded nose makes for non-expansion on impact; a pointed nose, most a military type, tends to penetrate tissues more deeply, whereas a flat nosed round tends to expand on impact creating a larger superficial wound. The body contains the core, usually lead, providing the necessary concentricity and balance for flight. The base or heel may be flat or boat tailed, the latter to reduce air drag to improve velocity and range. The combination of these properties further affect tissue damage. The energy transfer from the bullet to the tissues depends on six factors2:

Figure 1 Cartridges demonstrating the round at the tip, and the casing carrying the explosive powder.

injuries. The initial velocity is described as the velocity of the projectile at 5 m (15 ft) from the muzzle. Low-velocity wounds are more common in civilian practice and are usually less severe, whereas high-velocity wounds are more severe in nature, causing widespread tissue damage, and are more commonly seen in the military setting. The terms high and low velocity however can be misleading. For instance, a shotgun injury is technically a low-velocity injury but is frequently responsible for major soft tissue, nerve, vascular and joint injuries. More useful is perhaps the use of low-energy and high-energy wounds, which is indicative of the amount of tissue damage sustained suggesting the concept that energy transfer from the missile to the tissue is responsible for the severity of the wound.1 This relationship is where the kinetic energy of the bullet or other projectile: 2

kinetic energy ðKEÞ ¼ 0:5  mass  ðvelocityÞ . Thus as the mass of the bullet increases, there is a linear increase of its kinetic energy, but if the velocity is increased, KE and wounding potential increase exponentially. The velocity is dependent upon shape, weight and calibre (calibre is equivalent to the diameter of the bore of the weapon) known as the ballistic properties of the bullet and weapon. Differences in ballistic properties affect flight through the air, e.g. some slow more rapidly than others. The ballistic coefficient represents the ease with which the missile penetrates the atmosphere. Controlled missile flight is a factor of its ballistic coefficient and the spin imparted to it by the rifling of the barrel. The flight of the missile

1. The residual KE at impact. This is dependant on range, i.e. the further the target is away from the weapon the lower the KE. 2. The stability and entrance profile of the round. i.e. if the bullet has already started to tumble then the wound on impact will be ragged and more superficial. If flight has remained stable then impact wounding will be smaller and more circumscribed. 3. The calibre, construction and design of the bullet. 4. The penetration of the bullet, i.e. the distance it travels within the tissues. If it is penetrating, it is retained within the wound and thus its whole KE is delivered to the surrounding tissues. If it perforates, i.e. passes through the tissues and exits, the KE transfer is significantly less (Fig. 2). 5. The biological characteristics of the tissues the round impacts, i.e. tissue elasticity, cohesiveness and density. 6. The mechanism of tissue disruption, i.e. stretching, tearing, crushing. The impact velocity necessary to penetrate skin is approximately 40–50 m/s (150–170 ft/s).3 Most entry wounds are small, round to oval in appearance with clean well-defined edges. Tattooing surrounding the entry wound, caused from the powder within the cartridge, suggests that a close range injury has occurred. However, the wound may be insignificant and be easily missed or be surrounded by a ring of damaged skin with the appearance of an abrasion or bruise. A flat nosed bullet may cause a large superficial entry wound and appear to have caused far greater tissue damage. As the bullet traverses the tissues, a temporary cavity forms for approximately 10–30 ms (Fig. 3) due to stretching of the tissues as the bullet passes through them forming a cavity with a relative vacuum behind it. The volume of the cavity formed is proportional to the KE of the bullet and its

ARTICLE IN PRESS 324

S.A. Stapley, L.B. Cannon

Figure 2 Rounds within tissues. Pictures (a) and (b) show AP and lateral views of the cervical spine with an AK47 rifle round remaining within the soft tissue. The nose of the round remains pointed for deep penetration, although the body of the round has already been destorted by ricochet. Fortunately this round was at the end of its range and the patient survived. Pictures (c), (d), (e) and (f) show a round on the shoulder X-ray, the round after being removed and the patient from whom it was removed. The round had entered on the left side of the neck, passed through the soft tissue and exited on the right side only to re-enter the patient to be stopped by the distal clavicle. This demonstrates a typical rounded nose, fully jacketed with copper, flat-based round frequently used in the military setting. The initial extry and exit wounds are the same size.

Figure 3

Demonstrates the passage of a round as it penetrates a gelotine block with the formation of a cavity.

size. The maximum size of 10–40 times the diameter of the round is reached within 1–4 ms of impact. Internal pressures within the cavity of between 100 and 200 atm (10–20 MPa) have been measured.2 In high-velocity injuries this may create damage of an almost explosive nature. The subsequent vacuum sucks foreign material such as pieces of clothing, dirt, etc. into the wound causing further contamination. The temporary cavity collapses and reforms

repeatedly with diminishing amplitude, eventually leaving a permanent small cavity. The more the elastic nature of the surrounding tissues has been exceeded the larger the size of the permanent cavity. Surrounding this cavity therefore is a zone of contused soft tissue. In low-energy wounds, such as an airgun rifle wound, this may be only a few cells in depth. However, in high-energy transfer wounds such as those with a velocity of 4600 m/s (2000 ft/s) this volume of devitalised

ARTICLE IN PRESS An overview of the pathophysiology of gunshot and blast injury tissue becomes increasingly significant and may extend several centimetres away form the track of the projectile. Fascial planes can serve as channels along which energy dissipates leading to remote tissue damage. Therefore disruption of muscular capillary beds, fractures and rupture of gas containing viscera can occur remotely from the impact without direct injury. The densities of the tissues within the path of the projectile have an important effect on the overall wounding. The skin and lung, with relatively low density but large elasticity can be virtually spared from significant damage. Bone however, with a higher density but low elasticity, can be completely shattered. The liver, spleen and muscle are widely disrupted when the KE transferred exceeds the elasticity of the tissue. Larger arteries and nerve trunks are remarkably resistant to injury, and while neurapraxias do occur, nerves are rarely completely disrupted unless in the direct path of the missile. Exit wounds occur when the round has sufficient KE to perforate and travel through the body. Frequently, but not always, the exit wound is larger than the entry wound. As the round loses energy it begins to tumble within the tissue area thus creating a larger more irregular, less well-defined wound. In high-velocity injuries cavity formation may occur at the exit wound additionally sucking in substantial quantities of contaminated material (Fig. 4).

325 initially cleaned and dressed with a betadine soaked gauze dressing. In the civilian environment, patients are generally clean and present with low-energy transfer wounds (i.e. wounds from a handgun) and can often report the type of weapon used. This has lead to a conservative approach being adopted in the management of these wounds. Examples of low-energy wounds include airgun rifle injuries where the entrance wound can be as small as 0.5 cm in diameter and the round retained within the soft tissues, with minimal injury to the skin, subcutaneous tissue and muscle. Fractures, if present are of a stable configuration, can be treated conservatively not requiring operative intervention. Military-type wounds are the antithesis and must be managed more aggressively with early surgery. The presence of massive soft tissue damage with gross contamination, unstable fracture configurations with or without joint involvement, neurological deficit, vascular injury, gastrointestinal tract involvement and late presentation (wounds 48 h post-injury) require immediate surgical intervention with washout and formal wound excision. Other indications for surgery are tendon injury, superficial fragments in the palm of the hand or sole of the foot, certain cases with spinal involvement and injuries to the bony pelvis. A retained metallic fragment within a joint cavity is an absolute indication for intervention (as most rounds contain lead). Further management considerations are discussed later within this mini-symposium.

Bone A velocity of approximately 60 m/s (195–200 ft/s) is required to breach the cortex of bone.3 Fracture patterns are variable and may be complete or incomplete, and obviously fractures have concomitant soft tissue injury. Low-velocity projectiles tend to cause relatively minor, stable fracture configurations such as unicortical involvement, passing completely through the bone, or a chip to the bone. Highvelocity projectiles are more likely to cause unstable fracture configurations with butterfly fragments and large amounts of comminution. Bones are fractured either due to direct contact with the bullet or by a secondary energy transfer via the temporary cavitation. Bone fragments may themselves become secondary missiles, causing further damage to more distant structures, although retracting to the original site on dissipation of the KE. In the immature skeleton, physeal injuries are usually as a result of direct injury from the projectile passing close to the growth plate. This can be easily identified on initial Xrays. However, physeal arrest has been associated with remote injury.

Wound assessment and initial resuscitation As with any injury, assessment following ATLS guidelines should be undertaken at presentation with , if possible, an immediate full history and examination. Tetanus prophylaxis, an initial dose of broad spectrum IV antibiotics should be administered. Full assessment of the patient should include head to foot examination including a log roll and rectal examination. Identification of entrance and exit wounds should be made and documented before being

Blast injury Blast physics An explosive is a material capable of producing an explosion by its own energy. Explosives produce heat and gas. An explosion or blast follows a sudden release of energy4 from a chemical, gaseous, mechanical or even nuclear means dissipated by a blast wave, propelling fragments and surrounding material, and causing heat formation. The gas is the primary mechanism by which the explosive produces its effects. Thus following detonation of an explosive in air, a shock front (a wave travelling in excess of the speed of sound in that medium) travels away from the centre of the charge. The initial shock wave following an explosion is a special form of high-pressure stress wave, with an instantaneous wave front. The surrounding atmosphere is heated by the passage of this shock wave and then forced outwards by the expansion of gases formed within the explosion. After a short distance of travel of the shock front a further shock front is formed in the air (Fig. 5). This has a lower peak pressure and initial velocity than the detonation shock wave and has a zone of rarefied air immediately behind the highpressure area. This second shock front is called the blast wave. Within this blast wave are the products of the explosion, i.e. the gas and fragments of debris. The blast wave travels supersonically before decaying into an acoustic wave as it loses velocity and magnitude. The blast wave travels further than the detonation shock wave, exerting its effects further from the explosion centre. If the explosion is in the open, without being confined by buildings, etc. or is underwater, then a simple blast wave is produced (Fig. 6).

326

Figure 4 Entry and exit wounds. These pictures demonstrate the entry and exit wounds with the associated bony injury observed. In both cases the entrance wounds are small and the exit wound significantly greater in size. The top patient unfortunately had a complete disruption of the brachial plexus and axillary artery demonstrating the high-energy nature of the firearm that sustained this injury. This resulted in amputation. The lower patient was neurovascularly intact.

ARTICLE IN PRESS S.A. Stapley, L.B. Cannon

ARTICLE IN PRESS An overview of the pathophysiology of gunshot and blast injury

Figure 5 Simplified diagram of the components making up a blast wave.

Figure 6 Simplified waveform diagram of shock wave following an explosion in air.

This simple wave form has an almost instantaneous rise to peak overpressure, which then declines exponentially through ambient pressure to sub-atmospheric pressure, corresponding to the rarefied zone behind the blast front. The overpressure lasts for approximately 10 ms, with the sub-atmospheric pressure zone lasting for considerably longer. Confinement of the explosion within a building or underwater produces a complex blast wave pattern containing multiple overpressure peaks, due to reflections of the blast wave. The biological effects of the blast wave depend on the peak overpressure and its duration.

Classification of blast injuries Blast injuries fall into four main categories: (i) Primary blast injury relates to the interaction of the initial shock wave with the body. Gas containing structures such as the ear, lungs, and gastrointestinal tract are at particular risk. Solid organs including the skin are more resistant to the blast wave. Thus a patient with pure primary blast injury may display little external evidence of trauma. (ii) Secondary blast injury occurs as a result of the blast wave or wind and is caused by bomb fragments and other ‘secondary’ projectiles energised by the explo-

327 sion causing penetrating or non-penetrating wounds. Any part of the body may be affected. (iii) Tertiary blast injury occurs as the result of gross body displacement, i.e. the body being thrown through the air as a result of the blast wind. This leads to crush injuries. The combination of primary blast injury from the shock wave and tertiary injury due to body displacement leads to limb avulsion injuries. This occurs as the initial shock wave causes long bone fractures and the body displacement flailing of the limb with subsequent avulsion.5 (iv) Quaternary blast injury is a miscellaneous collection of all other mechanisms. These include thermal injury to exposed skin caused by the radiant and convective heat of the explosion, methaemoglobinaemia due to poisoning by dinitrobenzene or potassium perchlorate (from the explosive), and acute septicaemic meliodosis due to inhalation of soil particles contaminated with Pseudomonas pseudomallei. A high incidence of psychological sequelae in injured and uninjured survivors is also seen.6 Primary blast injuries result from the interaction of the shock wave with the body and are therefore a type of nonpenetrating injury. The interaction of the blast wave with the body wall generates two types of waves.7 These are known as stress waves and shear waves. Stress waves are longitudinal pressure waves with similar properties to sound waves. They travel at approximately the speed of sound, but differ from sound waves because of their high amplitude and velocity. The initial shock wave following an explosion is a special form of high-pressure stress wave, with an effectively instantaneous wave front, which travels through a medium at speeds greater than the speed of sound in that medium. The properties of this wave form explain the effects produced on tissues. Effects include high local forces produced with small rapid distortions, thus producing microvascular disruption, without gross lacerations. Organs with differing acoustic impedance are affected i.e. gas containing organs more readily. Tissue interfaces reflect and reinforce stress waves causing enlargement of wave pressures far from the site of body impact. The coupling of stress waves through the abdominal and thoracic walls is responsible for the primary blast injury to the gastro-intestinal tract.7 This occurs due to: (a) The development of pressure differentials across the delicate structures such as alveolar septa causing disruption. (b) As the stress wave passes from a solid into a gas filled tissue interface, a component of the compressive stress wave is reflected back as a tension wave. Most materials are weaker in tension than in compression and thus disruption and therefore damage at the tissue interface occurs. (c) When the stress wave compresses a gas containing structure such as an alveolus or bowel segment, the subsequent expansion causes damage to the wall of the structure. Shear waves are pressure waves transmitted perpendicular to stress waves and are of long duration and low

ARTICLE IN PRESS 328 velocity. These result from the deformation of the body wall and compression of the visceral structures. The tearing of structures from their attachments and shearing of solid organs is caused by the asynchronous movements of tissues with differing inertia. Shear waves are thus responsible for the primary blast injury of solid abdominal viscera, mesenteries and large bowel. The musculoskeletal system being solid is relatively resistant to the pressure waves although it has been demonstrated that initial shock waves of sufficient intensity can cause long bone fracture.5

S.A. Stapley, L.B. Cannon receptors located in the alveolar interstitial spaces close to pulmonary capillaries. An increase in pulmonary interstitial pressure or volume, due to pulmonary haemorrhage and oedema, could distort and therefore activate the pulmonary C-fibres. However, this triad is not shown in animals undergoing abdominal blast exposure.9 Therefore it is likely, that such cardiorespiratory responses are multifactorial, and that the additional conventional injuries associated with haemorrhage are likely to alter the physiological picture, but this has yet to be shown.10

Clinical features of blast lung Pathophysiology of primary blast lung injury Human post-mortem data of fatal primary blast injury are rare. Gross findings are of heavily consolidated haemorrhagic lungs. (Fig. 7) Haemorrhages may be multiple, subpleural lesions or those coalescing to involve the entire thickness of a lobe. These are often bilateral and multilobar. Animal experiments of primary blast injury demonstrate that the most consistent lesion was bilateral traumatic haemorrhage. Tension pneumothorax resulting from the rupture of subplueural cyst formed following alveolar septal tearing may occur and air embolism may occur. Haemorrhage into the alveoli and the resultant pulmonary oedema cause a ventilation perfusion mismatch with increased intrapulmonary shunt, reduced lung compliance, resulting in hypoxia and increased work of breathing.8 This response is similar to that seen in other non-penetrating lung injury. Initial clinical observations of the physiological responses to blast injury vary considerably, largely due to the varying times at which these observations are made post-injury and the secondary associated injuries sustained. Observations of victims dying immediately following blast exposure with little external evidence of injury led to the theory that the blast wave causes an acute cardiovascular and respiratory response. Experimental work on animals subjected to a thoracic blast has demonstrated a reflex triad of apnoea, bradycardia and hypotension.9 This triad of effects is thought to be mediated by pulmonary afferent C-fibre

Initial symptoms of blast lung include dyspnoea, cough (which may be dry or productive), haemoptysis, chest pain or discomfort (characteristically retrosternal). Signs include tachypnoea, cyanosis, reduced breath sounds with dullness to percussion, coarse crepitations and rhonchi are well described. Surgical emphysema and retinal artery emboli may also be evident. Pneumothorax/haemopneumothorax presenting with sudden shortness of breath, pain and deviated trachea require immediate attention.

Pathophysiology of intestinal primary blast injury Because of its many tissue/gaseous interfaces the intestine is highly susceptible to primary blast injury.11 Secondary and tertiary penetrating injuries must be managed in a conventional manner. The primary characteristic of intestinal primary blast injury is the intramural haematoma, although extreme overpressure shock waves will cause immediate gut laceration. Intramural haematomas may be minor, mucosal or submucosal haemorrhage only with oedema, through to complete disruption of the muscular layers and serosa, causing perforation. Individuals sustaining blast injury are observed to sustain injury mainly in the ileocaecal region and colon which are more likely to be gas filled.11 Prelaparotomy detection of these injuries is difficult. The natural history is uncertain and later perforation up to 14 days post-injury is reported. Thus Cripps et al.12 have proposed a histological classification for small and large bowel intramural haemorrhage which identifies contusions at high, intermediate and low risk of perforation. However, this requires direct visual examination to assess each injury and this may not always be possible or necessary. The overall consequences of intramural haemorrhage range from complete resolution, immediate perforation, delayed perforation and late stricture formation.

Clinical features of intestinal primary blast injury

Figure 7 Blast lung demonstrating multiple haemorrhages and septal disruption.

Abdominal pain, vomiting, haematemesis, distension, rectal pain with tenesmus and the presence of loose stools with fresh blood, or melena are indicative of such injury. Abdominal guarding with rebound tenderness and absent bowel sounds are indicative of intra-abdominal injury. Plain X-ray, ultrasonography and CT scanning may assist in the diagnosis of immediate perforation and fluid within the abdomen. Diagnostic peritoneal lavage has undisputed value in the diagnosis of intraperitoneal bleeding, but it is

ARTICLE IN PRESS An overview of the pathophysiology of gunshot and blast injury insensitive to the retroperitoneal and mesenteric haematoma, likely to occur in intramural haemorrhage of blast injury.13 In some patients the indications for laparotomy are obvious, in others primary blast injury represents a diagnostic challenge, injury remaining clinically silent until complications are manifest.

Pathophysiology of auditory injuries The ear is particularly susceptible to rapid pressure changes, but because these injuries are not life threatening, they can be easily overlooked and frequently under recorded. A small explosive causing an overpressure wave in a confined area can cause significant auditory damage.14 Injuries are categorised anatomically. The external ear is more likely to sustain injury from secondary or tertiary blast effects and should be managed accordingly. The shock wave on reaching the tympanic membrane (TM) causes displacement medially, creating a range of injury from intra-tympanic haemorrhage to radiation laceration. Perforation may disintegrate the TM completely or leave inverted or everted drum flaps. Small fragments of keratinising squamous epithelium may be distributed throughout the middle ear and mastoid system. If not removed they may subsequently form cholesteatoma. Disruption of the ossicular chain may be associated with TM rupture. The frequencies of such injuries are variable in the reported literature. Many blast survivors experience a profound, short-lived sensory neural hearing loss with tinnitus. The duration varies from a few hours to a permanent deficit. A few individuals experience vertiginous problems after blast injury. This may be caused by post-concussional states rather than labyrinthine damage. Otolaryngological examination and follow up is recommended in all blast injury patients suspected of auditory injury.

Pathophysiology of central nervous system primary blast injury Serious head injury is the most common cause of death in terrorist bombings.15 The enclosure of the brain within the skull makes survivable primary blast injury unusual. Brain and spinal cord injuries are usually caused by the secondary and tertiary effects of blast. Experimental studies on rabbits sustaining a large energy blast16 have demonstrated haemorrhage into the ventricles from the choroid plexus and haemorrhage into the pia, but none into the brain substance itself. Extradural spinal cord haemorrhages were also observed in these animals. Other animal studies in rats using histological techniques have demonstrated hypertrophy of microglial cells, a characteristic feature of neural degeneration, evident in the superficial zones of the cerebral and cerebellar cortices between day 1 and day 14 post-injury. These have reverted to normal levels by day 28.17 While clinical and experimental evidence exists for directly induced damage to the nervous system by the shock wave, more is known about the effects of air emboli in cerebral vessels. This is probably a more significant mechanism of injury. Air gains access to the circulation via alveolar-venous fistulae secondary to pulmonary blast injury

329 and has been considered to be one of the principal causes of immediate and early death due to primary blast injury. Thus symptoms and signs of air emboli should be sought early in patient assessment. These include headache, vertigo, ataxia, convulsions, altered levels of consciousness, weakness or sensory loss, facial or tongue blanching and retinal artery air emboli. The immediate action for this is the administration of oxygen. Definitive treatment requires hyperbaric oxygen, although this is not universally available. It reduces the volume of gas bubbles and improves blood flow to hypoperfused tissues.

Pathophysiology of orthopaedic blast injuries Most skeletal injuries, as already suggested, occur due to the effects of secondary and tertiary blast injury. Traumatic amputation occurs as the result of a combined primary and tertiary injury. In survivors of blast injury only 1.5% have traumatic amputations of limbs, with the exception of those injured by anti-personnel mines, discussed elsewhere in this symposium. However, traumatic amputation is relatively common in those patients dying early, reported as 19% by Mellor et al.18 in an analysis of servicemen killed or injured by blast in Northern Ireland between 1970 and 1984. It was considered that the blast wind caused rapid displacement as being the primary mechanism for this injury. However, it is now considered not to be the case. If traumatic avulsion were to occur through flailing alone, then amputation would occur through or close to joints as observed in fast jet pilots ejecting from aircraft and being exposed to slipstream wind speeds of 1100 km/h, which approximates to blast wind speeds.6 However, post-mortem data demonstrate that traumatic amputations following blast occur through the shafts of long bones and are not associated with joints (Fig. 8). Experiments undertaken on goats by Hull et al.5 have shown that the coupling of the initial shock waves into long bones generates stress waves which cause fracture of the long bone shaft. These occur most frequently in the upper third of the tibia and upper or lower third of the femur. The displacement of the body by the blast wind causes separation of the limb from the body through the fracture site. Large external, foreign fragments may also

Figure 8 Traumatic amputation through the lower tibia.

ARTICLE IN PRESS 330 cause traumatic amputation of limbs, but this is not the most usual mechanism.

Fragment wounds The current favoured weapon of the insurgent and terrorist bomber is the improvised explosive device (IED). These often crude devices are capable of inflicting massive injury to those in the immediate locality of the explosion. The casing of the bomb disintegrates into injurious fragments with nails and ball bearings sometimes added to maximise wounding. In a military environment the limbs are most frequently injured with relative sparing of the chest and abdomen due to the widespread use of personal body armour (Fig. 9). In those individuals closest to the blast epicentre the wounds sustained may be as a result

S.A. Stapley, L.B. Cannon of a combination of all four categories of blast injury (Fig. 10). In the case illustrated, it is clear that surgical wound excision is mandatory to remove all nonviable and foreign material. In situations were fragments have not penetrated vital structures and where the limb wounds are small a conservative approach can be taken. Subsequent infection rates are low.19 This has been the experience of the two authors during tours of duty in Iraq.

Initial resuscitation of blast injury Isolated primary blast injury is almost never seen, unless under experimental conditions. More frequently and therefore of more relevance to this review, is the combined primary blast injury associated with secondary and tertiary

Figure 9 Orthopaedic blast injuries. These pictures show the results of explosive devices which have gone off very close to the vehicles in which these patients were travelling. They are the result of two separate incidents. The X-rays demonstrate multiple pieces of retained fragments but also the severity of the bony injuries. Neither of these patients were propelled through the air following the explosion but were retained within their vehicles. Patient (a) was wearing body amour at the time of the explosion and did not develop any signs of primary blast injury to the lungs or intestine within the first 48 h. There is, however, no experimental evidence to suggest that body armour reduces transmission of the initial shock wave, but it could be considered that the bony injury results from a combination of primary blast injury and secondary fragment injury. This patient was neurovascularly intact on presentation. Patient (b) had a complete disruption of the tibial artery trifurcation and an asensate foot on presentation despite the soft tissue injury not appearing to be as significant on initial examination. This patient also sustained intra-abdominal pathology and multiple burns to his exposed skin.

ARTICLE IN PRESS An overview of the pathophysiology of gunshot and blast injury

331

Figure 10 This patient sustained injuries when an IED went off as the vehicle in which he was travelling went past. The explosive was positioned high up so that as this patient was standing in the back of the vehicle he sustained all four types of blast injury. Blast lung from the initial shock wave, secondary injury from the multiple fragments caused by the blast wind. He was not ejected from the vehicle but thrown against the sides. Some superficial skin burns due to the heat from the explosion. The pictures show his injury before and after debridement.

blast injury leading to combined penetrating and nonpenetrating injuries. First aid and resuscitation should be commenced immediately. The airway should be checked and cleared, high flow oxygen should be utilised and intravenous fluid resuscitation commenced. Currently, the regimen chosen for fluid resuscitation in this type of casualty remains a subject of some debate. Most recently, the NICE guidelines of 200420 have reported that hypotensive resuscitation should replace normotensive fluid resuscitation in trauma casualties. However, evidence supporting this is limited to those victims with penetrating injuries having short delays to definitive surgical management. Recent randomised trials in animal models investigating the physiology of resuscitation after blast, combined a primary blast injury with an associated controlled haemorrhage of 30% blood volume.21 Normotensive resuscitation was compared with hypotensive resuscitation for prolonged periods of up to 8 h post-injury. Normotensive resuscitation follows ATLS guidelines providing the equivalent of 2 l of 0.9% NaCl to a 70 kg man in order to regain a systolic blood pressure of 110 mmHg. Further bolus doses of 0.9% NaCl are given in order to maintain this level of systolic blood pressure. Hypotensive resuscitation follows BATLS guidelines and provides 3 ml/kg/min of 0.9% NaCl until a systolic pressure of 80 mmHg is obtained or a palpable radial pulse and maintained at this level with bolus doses of 0.9% NaCl. These studies demonstrate a considerable survival advantage in those animals receiving normotensive resuscitation and that prolonged hypotensive resuscitation was not compatible with survival after primary blast injury. Also identified within these experiments was that the blast injury exacerbated the fall in arterial blood pH when treated with hypotensive resuscitation which contributed to the more rapid demise of the blast injured model. The risk of re-bleeding associated with normotensive resuscitation remains high on the list of complications in such cases. However, each individual case must be considered separately. In the civilian setting fortunately in the UK such injuries remain relatively uncommon and time to

definitive surgical intervention is short. However, this has more relevance within the military setting where delay to evacuation and definitive surgery can be significantly prolonged.

References 1. Cooper GJ, Ryan JM. Interaction of penetrating missiles with tissues: some common misapprehensions and implications for wound management. Br J Surg 1990;77:606–10. 2. Bartlett CS, Helfet DL, Hausman MR, Strauss E. Ballistics and gunshot wounds: effects on musculoskeletal tissues. J Am Acad Orthop Surg 2000;8:21–36. 3. Ordog GJ, Wasserberger J, Balasubramaniam S. Shotgun wound ballistics. J Trauma 1988;28:624–31. 4. Cullis IG. Blast waves and how they interact with structures. J R Army Med Corps 2001;147:16–26. 5. Hull JB, Cooper GJ. Pattern and mechanism of traumatic amputation by explosive blast. J Trauma 1996;40:S198–205. 6. Horrocks CL. Blast injuries: biophysics, pathology and management principles. J R Army Med Corps 2001;147:28–40. 7. Cooper GJ, Taylor DEM. Biophysics of impact injury to the chest and abdomen. J R Army Med Corps 1989;135:58–67. 8. Cohn SM. Pulmonary contusion: a review of the clinical entity. J Trauma 1997;42:973–9. 9. Guy RJ, Kirkman E, Watkins PE, Cooper GJ. Physiological responses to primary blast. J Trauma 1998;45:983–7. 10. Guy RJ, Glover MA, Cripps NPJ. The pathophysiology of primary blast injury and its implication for treatment. Part 1: the thorax. J Roy Naval Med Service 1998;84:79–86. 11. Cripps NPJ, Glover MA, Guy RJ. The pathophysiology of primary blast injury ands its implications for treatment. Part II: the auditory structures and abdomen. J Roy Naval Med Service 1999;85:13–24. 12. Cripps NPJ, Cooper GJ. Risk of late perforation in intestinal contusions caused by explosive blast. Br J Surg 1997;84: 1298–303. 13. Gru ¨essner R, Mentges B, Du ¨ber CH, Ru ¨ckert K, Rothmund M. Sonography versus peritoneal lavage in blunt abdominal trauma. J Trauma 1989;29:242–4.

ARTICLE IN PRESS 332 14. Garth RJN. Blast injury of the ear; an overview and guide to management. Injury 1995;26:363–6. 15. Frykberg ER, Tepas JJ. Terrorist bombings. Lessons learned from Belfast to Beirut. Ann Surg 1988:569–76. 16. Guy RJ, Glover MA, Cripps NPJ. Primary blast injury: pathophysiology and implications for treatment. Part III: injury to the central nervous system and the limbs. J Roy Naval Med Service 2000;86:27–31. 17. Kaur C, Singh MK, Lim B, Ng BL, Yap EPH, Ling EA. The response of neurons and microglia to blast injury in the rat brain. Neuropathol Appl Neurobiol 1995;21:369–77.

S.A. Stapley, L.B. Cannon 18. Mellor SG, Cooper GJ. Analysis of 828 servicemen killed or injured by explosion in Northern Ireland 1970–84: the Hostile Action Casualty System. Br J Surg 1989;76:1006–10. 19. Bowyer GW. Management of small fragment wounds: experience from the Afghan border. J Trauma 1996;40(Suppl. 3): S170–2. 20. NICE. The clinical and cost effectiveness of prehospital intravenous fluid therapy in trauma. 2004. 21. Parry C, Garner J, Bird J, Watts S, Kirkman E. Reduced survival time with prolonged hypotensive versus normotensive resuscitation. Br J Surg 2005;92(Suppl. 1):112.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 333–345

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: BULLET AND BLAST INJURIES

(ii) Initial medical and surgical management Paul J. Parker MDHU(N), Friarage Hospital, Northallerton DL6 1JG, UK

KEYWORDS Military; British Army; Conflict; Mine; Gunshot; Blast; Ballistic; Trauma; Red cross; Limb injury; Amputation; Surgery

Summary This paper sets out the basic principles of the initial medical and surgical management of those affected by blast, mine and ballistic injury. The principles are unchanged since the American Civil War—and many come from pre-history; put simply: Resuscitate, Penicillin, Anti-Tetanus, Debride, Wash, Fasciotomise, Pack, stabilise, Leave—open and alone! & 2006 Elsevier Ltd. All rights reserved.

Introduction Limb wounds are common in all walks of life, accidents in the home, on the roads and in industry. In our western society, the well-nourished victim is transported with extreme rapidity by a dedicated ambulance service to a local hospital with a well-staffed operating theatre where a highly trained medical team will perform a familiar surgical operation. Thence they will go to a clean, plumbed, well-lit ward with dedicated and trained nursing staff where they will receive seamless holistic, rehabilitative care whilst the state cares for the needs of their family and home. Tel: +44 1609 764901; fax: +44 1609 764638.

E-mail addresses: [email protected], [email protected] (P.J. Parker). 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.07.006

There are those who would say that our medical management of blast, gunshot wounds (GSW) and mine injuries should not differ significantly from this secure civilian norm: that our normal protocols for wound closure, limb stabilisation and post-operative care may be safely applied, wherever in the world we are. Sadly, this perceptual error is repeated at the start of every new conflict and even during established conflicts when new medical staff arrive, keen to show that being a good doctor can overcome ‘old-fashioned’ wound care protocols. We must have a different mindset when we deal with these injuries. The most reliable civilian analogue of war and conflict wounds would be those that occur in slurry covered farmyards or sewage pits. Where civilian wounds are contaminated on average by a single bacterial species, wounds of conflict harbour between four and six. Thus, wounds greater than 6 h old need to be treated as infected

ARTICLE IN PRESS 334

P.J. Parker

rather than contaminated. The life should be saved first, then the limb. If all wounds were treated in such a fashion, then articles like this would not be necessary. Military surgeons, unfortunate civilian surgeons caught up in conflict zones, and those who now volunteer for overseas service with Non-Governmental organisations (NGOs) such as Me ´decins sans Frontieres and the Red Cross will find themselves living ‘above the shop’, in-danger, with limited resources and little or no back-up. They will be expected to operate on severely mangled limbs or wounds of unexpected severity. Small simple wounds will not heal and patients will sell their drugs to pay for food for their family. Malnutrition will be rife, clean water unavailable—and all this, often from the day that they arrive in that country.

History Wounding agents are largely unchanged since the Vietnam war. The multiple penetrating wounds, blast trauma and burns seen from today’s improvised explosive devices (IEDs), car bombs, suicide bombs and rocket propelled grenades (RPGs) are little changed from those encountered by the French in Dien Bien Phu, the British in Northern Ireland or the Coalition Forces in Iraq and Afghanistan.1 The anatomical distribution of these military wounds has remained essentially unaltered. Conflict

Penetrating limb wounds (%)

WW1 WW2 Vietnam Gulf War 1 (UK) Afghanistan (US)

70 75 74 71 61

The selective availability of munitions is a determinant of wounding pattern. In year one of the Soviet War in Afghanistan, 60% of injuries were GSW. Five years later, as the Mujahadeen captured or acquired explosives, mortars and mines, blast accounted for over 70% of all injuries. In general though, roughly 1/3 of all injuries presenting to a conflict hospital of any sort will be from blast injury (RPG/ Mine/IED) and 1/3 will be ballistic (GSW/Grenade) trauma. The remainder will be RTAs, burns and other injuries. We may think that we have made major steps in our care of the war-injured patient. This is untrue as despite substantial improvements in resuscitation (including novel haemostatic techniques—QuikClot, HemoCon, Factor V11a), ‘far-forward damage-control’ surgery, front-line intensive care and critical care evacuation, advances in body armour, vehicle armour and basic first aid—the battlefield surgical salvage rates from our current campaigns in Iraq and Afghanistan remain exactly the same as in WWII and Vietnam—3.5%. As armour has improved, so have sighting and offensive weapon capabilities. A current trick in Iraq is for a sniper to quietly shoot out the radiator of a vehicle. When the occupant gets out to check why the engine has overheated—he is shot in the exposed infra-axillary gap in the body armour when he lifts his hands and the bonnet above

his head (Fig. 1). The resulting trans-mediastinal wound is rapidly fatal. There is now newer body armour to counter this threat (Enhanced CBA), which has upper arm cuffs. Unfortunately, this item is heavy and cumbersome and thus disliked by troops. Mine injuries themselves are emotive, disabling events. Whilst they can be life threatening, most mines are not designed to kill. The PMA-3 mine, ubiquitous in Bosnia, contains just 35 g of Tetryl explosive. The PMN mine, favoured in Afghanistan contains 240 g of TNT. Anti-tank mines require much higher pressures to set the mine off. However, it is not uncommon for some to bury anti-tank mines with anti-personnel mines on top. In total, 3 kg pressure triggers the smaller mine whose explosion activates the larger device.2 It is estimated that there are 110 million landmines across 64 countries worldwide. Although there are well publicised mine clearance programs, mines are continually being laid. There are 600,000 victims of landmines in Afghanistan, one in 50 of the population. One third are female. There are still 10,000,000 mines in Afghanistan. Removal is said to occur ‘limb by limb’ with even the most diligent of the agencies only able to achieve 99.6% clearance.3 The military ethos of mine-laying is denial of ground. This may be to delineate border areas such as in Egypt. They also act against the morale of the fighting troops. In a section of eight men, if one is shot dead, the remaining seven members of the group are likely to fight back all the harder. If one man suffers traumatic amputation of a leg after a mine strike, at least two men will have to carry him (if not four) and one will have to render

Figure 1

Non-enhanced body armour vulnerability.

ARTICLE IN PRESS Initial medical and surgical management

335 target. Local drivers and car owners should be offered bonuses and incentives for bringing blast and bullet victims to medical care. This simple approach should also apply to western agencies in the resuscitation room. Pressure dressings and well-padded tourniquets are useful when there is traumatic amputation—a pneumatic theatre tourniquet being the best type. Antibiotics (Benzylpenicillin 2.4 g i.v., Flucloxacillin 2 g i.v. given as a loading dose with half these doses given qid for at least 2 days) may be given intravenously. Tetanus toxoid booster and immunoglobulin (Tetanus Immunoglobulin of Human Origin—HTIG, 500 units i.m.) for the nonvaccinated individual is also necessary. Crystalloid is administered intravenously to maintain a palpable radial pulse. The most important factor in this initial care is not to be distracted in any way by the catastrophic limb injury. Although limb bleeding can be troublesome, an unrecognised chest injury such as a pneumothorax will kill the patient just as surely as exsanguination. Similarly unrecognised perineal or rectal injury—common in lower limb mine strikes may also be ultimately fatal. This is covered in the resuscitation section of this mini-symposium.

first-aid. As soldiers generally do not fear death, merely disability, the fighting spirit of the two remaining soldiers not directly involved in the care of the casualty is not generally improved by the screams of the injured man. There is also the concept of ‘mine starvation’. Agricultural production falls by over a third in mined areas, due to land denial, the inability of the amputee farmer to till the soil and a general lack of investment. Mines may purposefully be sown on riverbanks to deny a population fresh water. Further deaths are then caused by the infantile diarrhoea, malnutrition and iodine deficiencies that result. Because vaccination teams do not visit the area, there are outbreaks of polio and measles amongst the villagers. Prosthetic fitting charities report two types of provision in the most heavily mined areas; callipers for the polio epidemic survivors and artificial limbs for the minestrike victims. Finally, mines do not disappear. Mines laid in Libya in 1942 still kill. The concept of mine marking is specious. Soil and sand erode. Snow melts. Those mines left on riverbanks are washed downstream to kill again. Cluster munitions are often brightly coloured and attractive to children who pick them up—so-called ‘butterfly bombs’. They continue to maim and blind long after the fighting troops have gone. For every two mine victims who reach hospital, there are three who died of shock, sepsis, blood loss and gangrene. There are now 100,000 minestrike amputees in Angola.

Initial surgical care

Initial medical management

Mine injury

Village first aid—Villagers should be trained in first aid. This should mean some provision of basic equipment with advice on bandaging/compressive wrapping, intramuscular antibiotic administration and mouth-to-mouth resuscitation. Simple written tourniquet protocols are appropriate for mine injuries, but these must be clear as pre-hospital transfer times often average 6–9 h. Splintage of the injured limb relieves pain for transport. Pentazocine (50 mg p.o.) or morphine (10 mg i.m.) are useful for pain control. Local barbers, shopkeepers and pharmacists are the best group to

If primary debridement is radical, there is no need for serial (piecemeal) debridement. This is the basic tenet of all surgery for mine injuries and is the core principle of the treatment protocols of the International Committee of the Red Cross (ICRC). Mine injuries are dirty and bacterially contaminated. The tissues are impregnated with organic matter, dirt and debris, often quite proximal to the injury up the intermuscular and interfascial tissue planes. By comparison, ballistic wounds

Figure 2 (a) Distracting Type I mine injury to left lower leg. (b) Post-debridement picture showing extensive damage to contralateral leg.

ARTICLE IN PRESS 336

P.J. Parker

are relatively clean—minestrike surgery is difficult and time consuming. Four patterns of wounding are recognised from mines: I. Mine stand (30%)—The victim stands on a mine producing a mangled foot (including the hindfoot) with evidence of proximal blast injury several centimetres up the leg (Fig. 2a). There may be blast injury to the contra-lateral leg (Fig. 2b). The anterior and lateral compartments fare badly but the posterior compartments of the leg normally flap back and are relatively spared. Transtibial amputation with a long posterior flap is the general rule. II. Fragmentation injury (50%)—a mine explodes at distance and secondary projectiles cause injury. In a Claymore type mine or grenade, this will be from multiple ball-bearing or preformed notched wire fragments (Fig. 3). III. Mine pick up (5%)—This usually causes significant hand and forearm injury. Blindness from globe penetration is common. Penetrating chest injury must be ruled out (Fig. 4). IV. No recognisable pattern (15%)—Often seen when smaller mines and anti-tank mines are used in conjunction, or where there has been an explosion in a confined space.

further tissue damage. If or when substantial bleeding is encountered, the tourniquet can be inflated in a matter of seconds and released again when the bleeding is controlled. Using a skin marker pen and ruler it is then determined what skin and bone are available—a preoperative X-ray may be useful in the stable patient (Fig. 5). In total, 12–15 cm of stable tibia and a further 12–15 cm of posterior skin flap beyond this ideal. Less than 5 cm of tibia is does not usually produce a functional result. Because the anterior and lateral compartments are tightly bound down to the tibia and fibula—they do not survive blast injury well. They are often severely contused and contaminated. They are best dealt with by resecting them transversely at the site of bony election. This makes a rapid, definitive and simple single surgical step as opposed to slow piecemeal distal to proximal nibbling.

General or Ketamine anaesthesia is used. For type I and III injuries, the majority of the contaminated distal limb is placed in an sterile impermeable stockinette bag and a crepe wrap applied. The remainder of the limb is prepped and draped. Use of an uninflated proximal pneumatic thigh tourniquet is very reassuring at this stage. Generally, it is best to operate without a tourniquet—this will prevent any Figure 4 Traumatic forearm amputation after mine pick-up. Forearm fasciotomies were also necessary.

Figure 3

Type II injury from fragmentation device.

Figure 5 X-ray image of Fig. 2.

Figure 6

Amputation sequence: (a,b) initial surgery, protection of soft tissues, (c) stump at 5 days, (d) long healthy posterior flap, (e) fashioning the final stump, (f) closure.

ARTICLE IN PRESS

Initial medical and surgical management 337

ARTICLE IN PRESS 338 By contrast, the posterior compartment with its overlying skin will have been able to flap back and absorb the shockwave better (the superficial posterior compartment with gastrocnemius especially as it is attached distally only to the flail foot and proximally only to the distal femoral condyles). The soleal mass is quite bulky and often needs filleting, but this is not so important at the time of the initial surgery. It is important to check up all tissue planes for grass, mud, pieces of clothing and shoe. Interseptal contamination will extend proximally for many centimetres above the obvious external injury. As each compartment is identified and dealt with, a simple proximal fasciotomy is then performed with curved round-ended scissors. The tibia should be cut transversely 12–15 cm below the joint line: use a ruler to formally measure this distance. A Gigli saw can be used but a tenon or bone saw is best. The surrounding tissues must be protected. The medullary cavity is curetted and washed out as debris may have been forced into it. The bone edges are bevelled down with a rasp, anteriorly the sharp tibial crest should be smoothed down for a few millimetres. The fibula should be resected some 2 cm proximal to the tibia. This can be done with bone cutters and again, any bony spicules should be removed. Disarticulation of the proximal tibio-fibular joint is not recommended. If access is difficult to the fibula, a small 3–4 cm ‘back-cut’ may be made on the lateral side of the flap for access (Fig. 6a–f). If possible, the plane between the gastrocnemius muscle and the skin should be left alone; this avoids damage to the perforating vessels and cutaneous blood supply. Although the muscle itself may have some embedded fragments, the soleus muscle acts as a bodyguard for the gastrocnemius and this plane is usually therefore clean. The three neurovascular bundles are identified. The main arteries are best double ligated with vicryl ties. The veins can be singly ligated. The nerve ends are best cut under gentle traction and allowed to retract proximally: This avoids the possibility of a later painful distal stump neuroma. Although asymmetric flaps can be created and used, the surgical reconstruction of most mine injuries relies on this persistent long posterior myocutaneous flap. The wound and muscles are then copiously irrigated with sterile saline. However, this is a bulky expensive item and use of potable or bottled water is the norm. The final wash can then be with a few 100 ml of sterile saline or water, but this is not strictly necessary. Fluffy gauze is loosely placed in the wound on top of the gastrocnemius and the skin is curved, without kinking, around the stump. Further fluffy gauze is placed around the stump with a layer of cotton wool and several crepe bandages on top. The last layer should include a Plaster of Paris (POP) backslab to splint the knee in full extension for if a significant flexion contracture develops in a stump, it becomes functionally useless. The ICRC teaching, in countries where over 25% of all surgical workload may be mine-related, is to leave all dressings in place for 5 days before a return to theatre for delayed primary closure (DPS). Although the dressing may smell and stain yellow–green, it is not the smell of putrefaction, it is the so-called ‘good–bad smell’ and the patient is otherwise well. Only if the patient deteriorates with pyrexia and the dressing becomes extremely offensive,

P.J. Parker the ‘bad–bad smell’, is the patient returned to theatre before 5 days. In low intensity conflict or single patient scenario where circumstances allow, a return to theatre at 48 h may be appropriate, especially in the first few cases performed. This can serve as a feedback as to the adequacy of the initial debridement. There will be an increased requirement for blood transfusion if this is done routinely. In a third world setting, a high-protein diet with mineral and vitamin supplementation is useful at this stage. Hookworm and helminth infestations can be treated. Children especially, may have been anaemic pre-operatively; transfusion to normocythaemia if logistically possible will aid wound healing. If the patient cannot eat, enteral feeding using locally available resources is a possibility.4 At day five, the patient is returned to the operating theatre. Any remaining non-viable areas, foreign matter and debris are removed. These areas should be minimal if the initial surgery was done correctly. Filleting of the bulky soleus may be necessary, sometimes a little of the gastrocnemius belly also until the flap sits nicely over the end of the bone. Two 3.2 mm holes are drilled in the end of the tibia with a hand or power-drill, two sutures are passed through the end of the flap picking up the muscle, which will have the beginnings of the Achilles tendon within it, but not the skin. This myodesis will aid propulsion by firmly attaching the muscle to the bone. The skin is then closed with interrupted 2 ‘0’ nylon mattress sutures over a Penrose drain (the cut-off finger of a glove is a useful substitute). Some attention to peripheral ‘dog-ears’ may be necessary. Fluffy gauze, cotton wool and crepe are used to dress the stump. Again, a backslab is useful to hold the stump in full extension for a day or so. Physiotherapy begins 24–48 h after wound closure, specifically range of motion and quadriceps exercises. Stump wrapping (Fig. 7) and ambulation on crutches are also commenced at this stage. The drain is pulled on the ward at 48 h. The patient should be ambulating on a prosthetic limb at 12 weeks after surgery (Fig. 8). Re-operation may be

Figure 7 Stump wrapping and physiotherapy, note the split skin grafts to the previously exposed areas on the contralateral leg.

ARTICLE IN PRESS Initial medical and surgical management

339 equal length skin flaps are constructed. In the hand, as much useful tissue is preserved as possible, the thumb in particular. Even a short stiff thumb acting as a post is better than no thumb at all. Surplus palmar and dorsal skin can be used to cover bony defects. Krukenberg amputation: This is used by some agencies where the victim has been blinded as well as losing part of one or both forearms. It provides a chopstick style sensate limb. The skin of the forearm is wrapped around the radius and ulna separately in the distal 2/3 of the forearm. This allows motion between the two bones and a degree of sensate manual dexterity.5

Guillotine amputation The ICRC does not recommend this technique and its associated postoperative care for the following reasons:

Figure 8 The patient (centre), 3 months later, ambulating independently.

necessary for prosthesis fitting, scar tethering, sinus or sequestrum formation, neuroma, protruding bone spikes and growing bone in children.

Above knee amputation Through-knee disarticulation is not recommended. Divide the femur 12–15 cm above the joint line of the knee (25 cm below the greater trochanter), bevel and smooth the bone end. The nerves and vessels are dealt with as for BKA. Fashion a 12–15 cm anterior and a 10 cm posterior skin flap such that the final scar will not overlay the point of the stump. This often equates to an anterior cut where the quadriceps tendon inserts into the patella and the posterior cut 3–5 cm proximally. Two or three 3.2 mm drill holes are made in the distal femur and the quadriceps tendon containing anterior muscle group attached to the bone. The skin is closed using 2 ‘0’ nylon over a Penrose drain. A bulky stump always remodels.

Above elbow amputation There is no optimal level of amputation. The bone should be divided as low as possible. This is usually about 2/3 of the way down the humeral shaft using 5 cm equal anterior and posterior skin flaps.

Below elbow amputation Above the wrist, the bones are cut cleanly across at equal length. The edges are filed smooth. The flexor and extensor musculature is sewn over the bone ends. In total, 3–4 cm

I. It shows a lack of understanding of the patho-physiology of mine injury; proximally injured dead muscle and infective debris in tissue planes well above the signs of any external damage will be missed. II. It is unsuitable in mid limb situations because of subsequent muscle swelling. III. It will require revision stump surgery if the patient survives. IV. It may result in an amputation higher than necessary. V. It produces a non-standard stump that is difficult to fit with a standard prosthesis. In a guillotine amputation, skin traction is required postoperatively to prevent skin retraction—this is cumbersome, difficult to do and makes transportation difficult. It is better to perform the long posterior flap technique as described above.

Ballistic injury and gun shot wounds (GSW) There is often much discussion (and confusion) about the following terms; high or low velocity and high or low energy transfer. Quite simply, velocity relates to the projectile: energy transfer to the wound. As discussed elsewhere in this mini-symposium, a high velocity projectile may pass through tissue (Fig. 9) without giving up any of its energy and produce a low energy transfer wound. Alternately it may be significantly retarded or come to rest in the tissue having expended all of its energy and produce a high energy transfer wound. Additionally, a heavy, yet low or intermediate velocity projectile designed to expand or fragment on impact (hollow point bullet) may produce a high energy transfer wound. Therefore, it is best to treat the wound and not the projectile. Again, the principles of treatment of ballistic injury are not difficult. The ICRC and British Military guidelines are simple and easy to understand: I. II. III. IV. V.

Complete wound excision, Delayed Primary Suture (DPS), antibiotics, antitetanus vaccine and immunoglobulin, splintage without internal bone fixation.

ARTICLE IN PRESS 340

P.J. Parker

Figure 9

Low energy transfer wound right calf.

Effective and early wound excision reduces the mortality and morbidity from these wounds. It reduces the chance of death from sepsis and gas gangrene and the number of operations required to remove remaining dead tissue and allows primary closure to succeed.

Wound classification Simple wound scoring systems are useful, for both documentation and clinical decision making purposes. The ICRC has a simple six parameter scoring system which is shown below: E ¼ entry wound size in cm X ¼ exit wound size in cm (X ¼ 0 if no exit) C ¼ can the cavity of the wound admit two of the surgeons fingers before surgery? C 0 ¼ no C 1 ¼ yes F ¼ fracture F 0 ¼ no fracture F 1 ¼ simple fracture, hole, insignificant comminution F 2 ¼ clinically significant comminution V ¼ Vital structure injury (dura, pleura, peritoneum, major vessel injury) V 0 ¼ no injury V 1 ¼ yes M ¼ metallic body M 0 ¼ no M 1 ¼ yes, one metallic body M 2 ¼ yes, multiple metallic bodies Thus, a small simple track without fracture and no cavity—a low energy transfer wound might be graded: E1 X1 C0 F0 V0 M0. A high energy transfer wound with a large exit wound, tissue cavitation and a comminuted femoral fracture might be graded: E1 X5 C1 F2 V0 M1. The greater the cavity and tissue damage, the greater and more complex the surgery. Wound excision is the technique best suited to these injuries. All the dead and damaged tissue grossly contaminated with bacteria and debris is completely cut away. This leaves behind healthy perfused tissue capable of combating infection as long as the wound is

not sutured shut. Thus, single or multiple superficial penetrating fragment wounds (E1 X0 C0 or E1 X1 C0) such as those caused by preformed notched wire fragments in modern grenades and explosive munitions (Fig. 3) do not usually require exploration. The ‘‘Death by a Thousand Cuts’’ or ‘‘Swiss Cheese’’ surgery—an attempt to excise all wounds and fragments through multiple incisions—is not necessary. In regard to the fragments themselves, unless they are intra-articular (where tissue acids such as hyaluronic acid may dissolve lead and other heavy metals causing a lead arthropathy), they need not be removed. For distal injuries, the placement of a proximal, uninflated pneumatic tourniquet is often reassuring. In more proximal or junctional injuries, a decision needs to be made whether formal proximal vascular control is necessary. The whole limb is scrubbed clean, then prepped and draped. All posterior wounds should be dealt with first. Different tissues respond in different ways to ballistic trauma. Skin is elastic, strong and remarkably resistant to damage. Only skin that is pulped or multiply lacerated e.g. in the typical stellate contaminated fashion, need be removed. It is wise to be conservative with skin—it will be needed later. Only a millimetre or so of the edge need be excised with a sharp knife. For E1 X1 C0 wounds i.e., simple tracks, this may be all that is necessary (Fig. 10). Skin excision should be followed by wound flossing (use of a single swab opened lengthways passed repeatedly through the wound to remove any small pieces of debris) and copious irrigation. Simple dressings are then applied. For E1 X4 C1 type larger wounds (Fig. 11a), after the wound edges have been debrided, generous longitudinal incisions (Fig. 11b), in the long axis of the limb but curved over flexion creases are made in the skin. The commonest error is to make these too short. Subcutaneous fat with its poor blood supply, generally high levels of contamination and poor resistance to infection and its associated shredded fascia is generously removed. Formal fasciotomy is often necessary and if you think about fasciotomy—you should perform it! All compartments should be released in the injured limb. There are four in the lower leg, three in the thigh, two in the arm and three in the forearm. For wound debridement though, the deep fascia is

Figure 10

Low energy transfer right thigh, wound flossing.

ARTICLE IN PRESS Initial medical and surgical management

341

Figure 11 (a) High energy transfer wound right thigh, posterior entrance, anterior exit. (b) Adequate longitudinal incisions for wound debridement.

Figure 12 High energy transfer wound left thigh, note extensive muscle damage and blue–black clot—the femoral vein had been disrupted.

incised along the length of the skin incision; this then allows the gentle gloved finger to become a diagnostic probe to estimate track size and tissue damage. If a large area of blue-black clot is encountered—this is a reliable sign that a major vessel has been damaged (Fig. 12). Before further exploration is undertaken, proximal vascular control should be considered. Attention can now be turned to the muscle: all dead muscle must be excised. Dead muscle, contaminated at time of wounding, becomes infected muscle six hours after wounding. The bacteria exit their latent phase and begin actively dividing and invading. If unrecognised clostridial infection then occurs, the limb threatening injury becomes a life threatening one. Muscle viability is assessed using the four Cs: Colour, Consistency, Contractility and Capillary bleeding. Healthy muscle is a succulent glistening brick-red colour. It will contract and ooze when cut. If the patient has been given a muscle relaxant (blocking the neuromuscular junction), the normal pinch reaction of muscle may be partly lost. Low intensity diathermy (so as not to char the tissues) can then be used as a diagnostic adjunct.

At its simplest level then, the basic technique is to pick up lumps of muscle with forceps, pinch them, and if they do not contract, excise them with scissors. The lumps should be no bigger than 2 cm2 to prevent damage to vital structures or healthy tissue. Small vessels should be carefully and individually coagulated: avoid the tendency to fulgurate large areas. Larger vessels are controlled with vicryl ties. Frayed tendons are trimmed; pieces of bone without attachment are removed. No attempt is made at tendon, nerve or internal bone repair at this stage. If major vessels are damaged primary repair, vein-patching or reversed saphenous vein grafting are usually necessary. In certain circumstances, it might be better for the inexperienced surgeon to ligate a distal artery and perform wide fasciotomies rather than spend several fruitless hours attempting an unfamiliar arterial repair or graft. This is especially true in multiple casualty situations. Figures from the Second World War indicate that gangrene is not always a foregone conclusion—even in ‘end-artery’ ligation.6 It is also important not to open up fresh planes in healthy tissue. Complete metallic fragment removal is unnecessary but all foreign debris, clothing wads, vegetation and dirt must be completely eliminated. At the end of this phase, the wounds are copiously irrigated with bottled or potable water (Fig. 13a). All residual debris and clot is removed. If available, a few 100 ml of sterile normal saline is used for the last wash. The wound is then packed with fluffed up dry gauze. No tight packs or antibiotic soaks (e.g. betadine) are used. The most memorable historical aphorism about this phase is; Wounds should be packed like a lady’s handbag, that is lightly and with delicate fluffy things. A gamgee or cotton wool layer is placed over this and several layers of crepe bandage steadily but not tightly applied. The aim is for inflammatory fluid to be drawn out of the wound and into the dressing. Therefore, Vaseline gauze and other occlusive dressings should not be used. As with mine injuries, the wound is left untouched for 5 days before DPS is undertaken. The previous notes about the ‘good/bad’ smell and the ‘bad/bad’ smell remain extant. The dressing is taken down in theatre under Ketamine/GA, any small remaining devitalised areas are removed and the wounds closed without tension using 2 ‘0’

ARTICLE IN PRESS 342

P.J. Parker

Figure 13 (a) High energy transfer wound right tibia. After nibbling of bone ends and general debridement, the wound is copiously irrigated. (b) The wound and fracture are stabilised with an external fixator.

nylon sutures. If 2 ‘0’ nylon will not close the wound, then it is either not ready for closure or some alternate means will need to be used to achieve skin cover such a split-skin graft or local flap.

Bony injury All dead and contaminated bone has to be removed. The prevention of bone sepsis is paramount. In the acute phase it may be life-threatening, in the chronic phase it will prevent formal reconstruction. Even with modern technology, the viability of bone fragments can be difficult to determine. Large cortical fragments with firm muscle attachments are usually fine. Small pieces (o1 cm) with only filamentous strands of tissue holding them in place are best removed. Dirty bone ends are cleaned best by judicious use of bone nibblers back to healthy white punctate bleeding bone. Medullary cavities need to be curetted and irrigated clean. Both ends of the fractured bone need to be delivered into the wound, thus adequate longitudinal incisions and a bone hook are useful. In neglected wounds—the bone may be yellow (‘bad teeth’) in colour and appearance—this needs to be aggressively resected back to healthy bone. If there is a suspicion that a fragment of bullet has penetrated or traversed a joint, then that joint must be washed out. X-rays are helpful as they may show air or debris in the joint. Retained metallic fragments, especially if they contain lead need to be removed because of the risk of lead arthropathy. Where there has been damage to the joint, the same rules apply as to bone. Attached osteochondral fragments are retained in place as best as is possible, but it is best not to fix them at the initial debridement stage. After copious irrigation the joint capsule should be closed, if possible, to further reduce the risk of infection. The skin and other structures can be closed at the time of delayed primary suture.

Fracture stabilisation In total, 35–50% of all limb wounds will have some bone involvement. Such injuries are compound and are frequently infected by the time they reach a surgical facility. Wound debridement should take place as described above, then the wound and fracture must be stabilised and if the limb seems in any way tense, fasciotomy is immediately performed. X-ray is a useful but not essential adjunct.

Realignment and stabilisation The benefits of realignment and stabilisation of fractures are:


Provides pain relief and reduces overall analgesic requirements.


Reduces secondary neurovascular and tissue damage.
Stabilises the ‘wound organ’ and reduces harmful cytokine release.


Reduces the chance of fat embolism.
May convert a stretcher case to a self-care walking casualty.


Allows for safe and prolonged transport.
Begins the rehabilitation and healing process. The stabilisation method used will vary from case to case and limb to limb, but POP remains the safest and commonest method. Although external fixation devices may be available (Fig. 13b), their presence is not an absolute indication for their use. Humeral, tibial and forearm injuries are all suitable for POP splintage. In the tibia, this can be augmented with proximal and distal Steinmann pins to hold the limb out to length when there is significant bone loss (Bo ¨hler technique). In all cases, the foot should be incorporated and placed at 901. This prevents any later equinus deformity.

ARTICLE IN PRESS Initial medical and surgical management A fractured femur will require some form of skeletal traction as the pull of the attached muscles causing femoral and tissue shortening may allow for untamponaded intracompartmental haemorrhage. For the femur, placement of a proximal tibial or distal femoral Steinmann or Denham (threaded) pin is optimal. This may be attached to a Thomas splint for transport (Fig. 14) and allows for easy conversion to beam, balanced or Obletz traction for treatment to union. Well-moulded POP splintage is suitable for tibial fractures; windows may be cut in the plaster for wound care (Fig. 15). External fixation, when available is an alternative and is best suited to the tibia. However, if a patient is likely to be repatriated to a western standard of care facility within a day or so, consideration should be given to avoidance of the technique in uncomplicated fractures as an early pin-site infection can contaminate the medulla and this may then preclude intra-medullary nailing. The ICRC currently uses the AO/ASIF tubular fixator. The US, NL and UK military use the Hoffmann II pattern fixator. The latter has the advantage of self-drilling, self-tapping pins. External fixation is particulary suitable for:


Unstable fractures with extensive bone loss.
Fractures with extensive soft tissue loss or burns.

343


Vascular injuries requiring repair.
Polytrauma.
Casualty transfer/definitive care. However, external fixation is not without its complications. Pin-site infections, thermal necrosis of bone at insertion, unicortical placement and poor biomechanical placement of the pins will all cause early loosening and failure. Even trained orthopaedic surgeons can have difficulty with fixator placement. It is best learnt as a formal practical skill in a workshop environment. Technical points to consider are:













General limb realignment before beginning. 41 cm stab incisions. Avoidance of bony thermal damage. Subcutaneous tibial border insertion. Bi-cortical placement. Avoidance of pin-placement within joint capsule or 3 cm of fracture. Two solid pins above, two solid pins below the fracture. Pins in the ‘near-far/near-far’ configuration. Stable pin/bar configuration. Daily pin-site care.

Fasciotomy The technique of fasciotomy is not difficult. It is therefore surprising that many surgeons agonise over the procedure. In the conflict scenario or times of high casualty flow, it may be several days before the surgeon sees the patient again. The message is therefore very simple. If you even think the word ‘fasciotomy?’—then just do it. Do not agonise over ‘pain on passive stretch’, ‘moderate but not uncontrollable pain’, ‘weak but palpable distal pulses’. Indications for Fasciotomy:

Figure 14 Thomas splintage for compound femoral fracture post-debridement.









Ballistic wounds directly involving the calf. Ipsilateral major venous injuries. Clinical signs or any suspicion of compartment syndrome. Warm ischaemic times in excess of 4–6 h. Crush or prolonged entrapment scenarios. Major arterial injury, ligation or repair.

Fasciotomy is best learnt as a technical skill. But it is not difficult to do. There is no place for single skin incision fasciotomy or decompression via fibulectomy (the fibula is then lost as a reconstructive aid). There is no place for limited incisions or so-called subcutaneous fasciotomy. Fasciotomy incisions must be generous, correctly placed and include at least a 7 cm skin bridge between incisions.

Technique

Figure 15 POP windows allow for limb splintage and wound care.

There are four compartments to release in the lower leg. The guide to the anterior and lateral compartments is the lateral intermuscular septum; this is found about 4–5 cm lateral to the lateral subcutaneous border of the tibia. It may be palpated through the skin. An incision is made which extends from just anterior and 2 cm below the head of the fibula; to 5 cm above and anterior to the lateral malleolus. The subcutaneous fat is

ARTICLE IN PRESS 344 also incised taking care at this stage not to violate the fascia. The lateral intermuscular septum is easily felt. A 2 cm transverse incision is made at mid incision level in the fascia centred on the septum. The sharp tense edge of the septum as it dives deep between the two compartments is unmistakeable. Long curved scissors used at the outer edges of this transverse cut are then used in a twopass technique to produce an H-shaped release. The first pass with the tips closed lifts the fascia off the muscle, the second with tips open divides the fascia completely, up to and slightly beyond the skin incision margin. The middle finger of the surgeon’s free hand should then be able to pass freely proximally and distally without encountering any remaining fascial leashes or bands. This technique is used proximally and distally in each compartment. The commonest technical errors are:


Inadequate skin incisions and failure to confirm the




compartment being released, the swollen lateral compartment is then mistaken for the anterior and two incisions are made in the lateral compartment instead. The anterior compartment dies and the leg or the patient is lost. Damage to the lateral peroneal nerve. It perforates the lateral septum two-thirds of the way down the leg. It is avoided by angling the tips of the scissors away from the initial incision. In the lateral compartment, the tips are aimed towards the head of the fibula proximally and the lateral malleolus distally. In the anterior compartment, the tips are aimed towards the tibial metaphyseal flare both proximally and distally.

The superficial and deep posterior compartments are released next. A ruler is used to determine a seven centimetre skin bridge between the two releasing incisions. For the posterior compartment, the incision is made one cm behind the medial tibial border from just below the proximal metaphyseal flare to 7 cm above the tip of the medial malleolus. The long saphenous vein and its accompanying nerve are gently retracted. A 2 cm transverse cut is made to identify the thin septum between the superficial and deep posterior compartments. The superficial compartmental fascia containing gastrocnemius and soleus is released proximally and distally using the curved roundended scissors. The deep compartment is released distally by dividing the filmy deep fascia over flexor digitorum longus and tibialis posterior. This dissection is continued proximally through the soleus bridge—a transverse leash of 2–3 veins where soleus begins its bony attachment to the tibia. Some mobilisation of the soleus off the tibia may be necessary to fully release the proximal fascia over the deep compartment. Again, a finger is used to confirm release. After haemostasis, the wounds are dressed and left open for later DPS. There is often considerable angst about the techniques for fasciotomy wound closure. The medial wound often contains exposed tibia and therefore should be closed first and directly. The lateral wound is then problematic, if there is residual swelling then a delay of a few more days is reasonable before closure. Skin is viscoelastic and retracts.

P.J. Parker Consideration, if facilities allow, should be given to skinstretching techniques. The easiest is to use a skin pulley stitch. An ‘2’ nylon is used in the near-far-far-near configuration, the skin edges opposed and the wound formally closed with 2 ‘0’ nylon. The pulley stitches are removed. However, it needs to be remembered that excessive tension must be avoided. The other common alternative is to place a meshed split-skin graft over the anterolateral defect. In the upper arm, the two compartments; anterior containing biceps and brachialis, posterior containing triceps, are released via a single lateral incision. In the forearm, the superficial and deep flexor compartments are released through a lazy ‘S’ incision from the transverse ante-cubital fold, over palmaris longus, down to and including the carpal tunnel. The extensor compartment is released in the pronated arm from the lateral wad down to the centre of the wrist. In the hand the transverse carpal ligament is released on the palmar side and two incisions used to release the interossei on the dorsal side. In the thigh, the three compartments (anterior/quadriceps, posterior/hamstring, medial/adductor) are released via a two incision technique. A single lateral incision with division of the fascia over the flexor and extensor muscles usually suffices. If there is a penetrating injury, a medial incision over adductor longus releases this side.

Special situations Suicide bomb incidents Standard multiple fragment injuries remain the norm. However, the biologic material produced in such explosions presents an added complication as patients will sustain injuries caused by the bone fragments of the bomber. These samples have tested positive for Hepatitis B virus.7 In the non-immune, initial medical treatment of these patients needs to include Hepatitis B vaccination and Hepatitis B immunoglobulin (HBIG 500 units). If there is any suspicion of HIV infection in the bomber, post-exposure prophylaxis (currently Zidovudine 300 mg p.o. b.d., Lamivudine 150 mg p.o. b.d., Nelfinavir 1.25 g p.o. b.d.) should be started immediately. If possible, the remains of the bomber should also be tested.

Junctional injuries These continue to pose a significant challenge to the orthopaedic surgeon. Proximal thigh and axillary injuries are special cases. The best advice is to have a vascular surgeon in theatre and to ‘double-team’ the case. If in solitary practice, then proximal arterial control is wise. In the thigh, the femoral artery can be controlled, with a circumferential sloop, in the groin crease, if the wound is in the groin, then extra-peritoneal iliac control is prudent. For complex proximal subclavian injuries, be prepared to undertake a rapid midline sternotomy to gently clamp the intrathoracic right subclavian or brachiocephalic arteries. During resuscitation, a Foley catheter can be placed in a neck or proximal wound and inflated, to tamponade bleeding until formal control is obtained.

ARTICLE IN PRESS Initial medical and surgical management

Training for blast and ballistic injury Whilst this mini-symposium is intended to provide a general overview of the pathophysiology, resuscitation, initial treatment and care of victims of ballistic and blast trauma, there is no doubt that practical training is also vital. The Royal College of Surgeons of London runs a twice-yearly ‘Definitive Surgical Trauma Skills’ (DSTS) course. This is a cadaver practical-based course with emphasis on great vessel access and control, amputation, fasciotomy, external fixation and damage control laparotomy for penetrating abdominal trauma. The ICRC runs a twice yearly 2–3 days ‘Surgery for Victims of War’ course. It also runs longer courses for those about to deploy overseas to one of their third world hospitals. The Professorial Department of Military Surgery of the UK Defence Medical Services runs yearly War Surgery courses and Mangled Extremity courses.

References 1. Rush RM, Stockmaster NR, Stinger HK, et al. , Supporting the global war on terror: a tale of two campaigns featuring the 250th Forward Surgical Team (Airborne). Am J Surg 2005;189:564–70.

345 2. /www.mineactioncanada.orgS. 3. /www.humanitarian-demining.orgS. 4. Frizzi JD, Ray PD, Raff JB. Enteral nutrition by a forward surgical team in Afghanistan. Southern Med J 2005;98(3):273–8. 5. Garst R. The Krukenberg Hand JBJS-B 1991;73(3):385–8. 6. Tudor-Edwards A. Limb Amputation Rates. Inter-Allied Conference on War Medicine. Paris, January 1945. p. 169. 7. Eshkol Z, Katz K. Injuries from biologic material of suicide bombers. Injury 2005;36(2):271–4.

Further reading 1. The British Military Surgery Pocket Book. UK: British Army Publication AC 12552; 2004. 2. Surgery for Victims of War. An ICRC Handbook. 3rd ed. 1998 /www.icrc.orgS. 3. Emergency War Surgery Handbook 3rd US Revision 2004, Department of Defense, USA. 4. Ballistic Trauma. A practical guide, 2nd ed. Berlin: Springer; 2004. 5. Definitive Surgical Trauma Skills Manual. Royal College of Surgeons of England, 2002.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 346–353

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: BULLET AND BLAST INJURIES

(iii) Military limb injuries/ballistic fractures Dylan Griffiths, Jon Clasper Trauma and Orthopaedic Surgery, Frimley Park Hospital, Surrey GU16 7UJ, UK

KEYWORDS Military; Open; Fracture; Contamination; Ballistic

Summary Military limb injuries and ballistic fractures occur most commonly in the lower limbs and the degree of tissue damage is related to the energy transfer from the missile. Most are multi-fragmentory. Contamination is a far greater problem in the military environment, and wound exploration, debridement and irrigation is mandatory for all military ballistic injuries. Stabilization of military ballistic fractures poses specific problems. & 2006 Elsevier Ltd. All rights reserved.

Introduction Seventy per cent of all wounds seen during armed conflict are wounds to the limb; caused by various mechanisms such as landmines, bomb blasts and bullets. The lower limb is predominantly affected. Most penetrating injuries are due to fragments following explosions, although in some conflicts, such as Northern Ireland in recent years, bullet wounds may predominate. Ballistic limb injuries show a wide spectrum of severity ranging from superficial low-energy wounds caused by shrapnel to high-energy bullet wounds. In the latter more than half are associated with fractures. While the majority of limb injuries are not life threatening, they are associated with significant morbidity, hence timely assessment and management are essential.

General principles of ballistic injuries The severity of a ballistic injury is directly related to the transfer of energy from the missile to the tissues. The Corresponding author.

E-mail addresses: [email protected] (D. Griffiths), [email protected] (J. Clasper). 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.07.007

damage caused by missiles can be either direct or indirect. In low-energy injuries such as from knives, tissue damage is caused directly by the missile and is confined to the wound track, either by cutting or crushing. Significant injury will only occur if a vital structure lies directly in that path. With high-energy missiles, energy is dissipated to the surrounding tissues and produces damage remote to the wound track. This indirect effect is most commonly seen in injuries from high velocity rifles used in military conflicts. This contributes to the increased mortality from high velocity (more correctly termed high-energy transfer) wounds. It is important to recognize however, that outcome is determined by the structures that are damaged and not the total energy that is transferred. Therefore, a knife passing through the heart is more likely to be fatal than a highenergy transfer wound to the leg! A number of factors determine the transfer of energy to tissues, including the velocity of the missile, its mass, its shape, its stability and type of tissue through which missile passes. The available kinetic energy of a missile is determined by the formula: kinetic energy ¼ 12
mass
velocity2 . From this, the importance of velocity in the kinetic energy of a missile is obvious and accounts for the severe tissue

ARTICLE IN PRESS Military limb injuries/ballistic fractures damage seen with high-velocity weapons. However, it is wrong to purely define wounds in terms of velocity. For example, a close range shotgun injury can cause significant tissue damage because of the large mass of the missile1 as shown in Fig. 1. Wounds should be defined in terms of energy transfer rather than the velocity of the weapon. The transfer of energy to the tissues depends on the interaction of the missile with the tissues. A key factor is the resistance provided by the tissues to the missile, which occurs either as a result of the surface area presented to the tissues by the missile or the intrinsic resistance of the tissues. With regard to the shape of the missile, the smaller the surface area of missile that is presented to the tissues, the smaller the resistance and the smaller the amount of energy transferred. A spherical object such as a ball bearing will cause less damage than an irregular, flattened piece of shrapnel2 of the same mass and velocity. Due to their shape and centre of gravity, bullets are very unstable and can yaw (tumble), thus presenting a larger surface area to the tissues and increasing energy transfer to the tissues.3 This often occurs within the deeper tissues such that there is a very small ‘bullet-sized’ entry wound (Fig. 2), but far greater tissue damage beneath this and a large ‘ragged’ exit wound.4 An additional factor that contributes to the energy transfer to the tissues is any deformation or breaking-up of the missile in the tissues. This produces more extensive wounding and is the basis of illegal, soft-nosed bullets or ‘dum-dum’ bullets that are more likely to fragment on entering the body. Even with legal bullets, fragmentation can occur, particularly if the bullet hits bone. The intrinsic resistance of tissue to a missile is related to the density and rigidity of the tissue, such that bone offers greater resistance than softer tissues such as lung or muscle. This results in greater energy transfer and therefore the missile inflicts greater damage than if it passed through soft tissues. Damage to tissues from energy transfer is thought to occur by 3 main mechanisms: 1. Cutting—direct laceration by the missile 2. Overpressure—waves of pressure radiating out from the missile as a result of energy loss due to the resistance of the tissues. The greater the resistance the greater the

347

Figure 2 Small bullet entry wound which if from a high-energy weapon which may mask significant tissue damage beneath the skin.

overpressure. There has been debate amongst authors as to the significance of this in producing tissue injury.5,6 3. Cavitation—this is the formation of a temporary cavity behind the missile and is thought to be the most significant factor in tissue injury from high-energy transfer wounds (Fig. 3). This occurs as a result of energy transfer from the missile, which accelerates tissues away from it causing a cavity, which persists beyond the time the missile is in contact. Such a mechanism causes injury by compressive/shearing forces, but the temporary cavitation also sucks material into the wound causing increased contamination. Once again, greater damage will occur with a missile passing through more resistant tissue such as bone and tissue contained within a nonelastic capsule such as liver.

Specifics of missile injury to bone

Figure 1 Close range shotgun wound, high-energy transfer wound from a low-velocity weapon.

The resistance of bone may promote high-energy transfer from the missile and the bone commonly fractures. Such fractures may be complete or incomplete; some bony continuity may remain. Complete fractures can be divided into simple (with only two main fragments—see Fig. 4) and multi-fragmentory (with multiple fragments present—see Fig. 5). Incomplete fractures can also be subdivided into ‘drill-hole type’ in which a tunnel is driven through the bone

ARTICLE IN PRESS 348

D. Griffiths, J. Clasper

Figure 3 Cavitation in a gelatin block from high-energy transfer.

Figure 5

Multifragmentary ballistic fracture.

in this instance, as opposed to the multi-fragmentory type from direct injury.2,9 The degree contamination of indirect fractures has also been found to be far less, in vitro, than for direct fractures.9

Injury specifics of landmines

Figure 4

Simple (2-part) ballistic fracture.

and ‘chip type’ in which part of the cortex was removed but there was no tunnel through the bone.7 In wounds from highenergy weapons, fractures are always complete and multifragmentory, but low energy weapons can cause simple or incomplete fractures.7 In vitro models have shown that for high-energy wounds the average length for the comminuted segment of bone was 42% of the total length,8 which accounts for the difficulty in achieving fracture stability. The fractures described so far are from direct missile injuries to bone, but with high-energy transfer wounds, indirect injuries to bone can occur. This is thought to be from the effects of cavitation, accelerating bone away from the missile track. The fracture pattern is commonly simple

Landmines can be broadly divided into two groups—antitank or antipersonnel devices. Both are designed to deny ground to the enemy whether on foot or in armoured vehicles such as tanks. The antipersonnel devices are smaller explosive devices that are designed to injure rather than kill a casualty, and the channeled explosive frequently causes devastating limb injuries including traumatic amputations. Retrospective analysis of 757 antipersonnel mine injuries, by the International Committee of the Red Cross,10 revealed predominantly 3 patterns of injury amongst the survivors. Pattern 1: Pattern 1 occurs when a buried mine is stepped on. Severe lower limb injuries occur including traumatic amputations as well as genital injuries. Pattern 2: Pattern 2 occurs when the device explodes near the victim. Lower limb injuries occur, but are less severe than in pattern 1, with traumatic amputations less likely. Pattern 3: Pattern 3 injuries occur when the device explodes whilst victim is handling it. Severe facial and upper limb injuries are common in this group.

ARTICLE IN PRESS Military limb injuries/ballistic fractures

349

Assessment and management of ballistic injuries Resuscitation should follow ATLS guidelines. While the history may be useful in determining the type of weapon used, it is important to ‘treat the wound and not the weapon’. If there is extensive soft tissue damage, plastic surgery opinion should be sought early if possible. As indicated earlier, it is important to be aware that a small skin wound sustained from a high-energy missile may conceal extensive soft tissue and bony damage beneath the skin. Figures 6 and 7 show a high-energy wound to the upper arm from several bullets: the skin wounds are small, but despite the external appearance, an extensive soft tissue and bony wound is present and the outcome was relatively poor. In contrast Figs. 8 and 9 illustrate that the apparently more severe shotgun wound shown in Fig. 1, resulted in only a simple fracture pattern beneath. Early internal fixation and flap coverage were possible with a good functional outcome. The principles of management of military limb injuries/ ballistic fractures are simple11:

Figure 8 Radiograph of the humerus in Fig. 8 demonstrating a simple complete fracture.

1. surgical debridement and adequate lavage, 2. stabilization of the limb, and

Figure 9 Patient shown in Figs. 8 and 9 showing early closure using a latissimus dorsi flap.

3. use of appropriate antibiotics (including at initial presentation).

Figure 6 Multiple gunshot wounds to the upper arm; despite the small wounds extensive soft tissue and bony damage is present.

Figure 7 Radiograph of humerus in Fig. 6. Extensive fracture has been stabilized by external fixation.

The principle of debridement is to remove all non-viable tissue. Debridement of additional skin beyond the wound edges can often be minimal, but may need to be more extensive in the case of degloving injuries. It is important, however, that the wound is extended to allow visualization of the full extent of the subcutaneous wound, which may be considerable in high-energy transfer wounds. Extension is performed along the long axis of the limb, apart from when crossing the flexor surface of a joint, when incisions should be oblique to prevent contracture. Fat should be excised, but this should not be over-generous to avoid additional areas of degloving. The deep fascia should be incised along the complete length of the wound, including any extensions. Full-length compartment fasciotomies should be carried out in most high-energy wounds (Fig. 10). The adequate debridement of muscle is essential and it may be necessary to excise a large amount of necrotic muscle. The aim is to remove all non-viable muscle leaving only pink, healthy-looking, contractile muscle. The 4 C’s of capillary bleeding, consistency, colour and contractility act as an objective way of deciding if muscle is viable, but experience is the most important factor. Bleeding from debridement of muscle may result in significant blood loss

ARTICLE IN PRESS 350

D. Griffiths, J. Clasper However, the decision is far more difficult in less severe injuries. The following criteria are helpful:

 is there a large bony defect?  are there extensive wounds that will require flap coverage?

 is there soft tissue injury (including vein injury) that will impair function?

 is there vascular injury that requires repair?  is there neurological injury, particularly involving the hand or sole of the foot?

 what resources are available locally? Figure 10 Fasciotomy of the forearm following a gunshot wound: wound continued into the palm to include the carpal ligament.

and both surgeon and anaesthetist must be prepared for this. A tourniquet will make recognition of non-viable tissue more difficult and is not recommended for most lower limb injuries, but may be indicated with upper limb injuries to avoid damage to vital structures. Nerves and patent blood vessels should be left along with tendons that are in continuity with muscle (although they may dessicate and need to be excised at a later date). Divided nerve ends should be marked with a non-absorbable monofilament suture. Washout with copious amounts of sterile fluid is essential; it has been recommended that 9 l be used for open fractures. With high-energy transfer wounds, contamination can spread along tissue planes and these must be thoroughly irrigated. In the military environment it may be impossible to get hold of such volumes of sterile fluids and in these circumstances drinking water can be used with a final irrigation with sterile fluid. The debridement of bone can often be quite difficult, particularly with small bony fragments. Bony fragments without any soft tissue attachment are avascular and should be removed. However, fragments often do retain periosteal/ soft tissue attachments and fragment viability can be difficult to determine. Again experience is the most important factor in determining the viability of bone fragments. Experimental work8 has shown that there is limited spread of contamination of bone beyond the fracture site. Although this should be approached with caution, if the fracture site itself is well visualized and washed out, exposure of intact bone beyond the fracture site is not necessary. In general, delayed primary closure of ballistic wounds should be carried out. This is commonly done after 5 days due to the heavy contamination of military wounds. Highenergy transfer wounds with comminution of bone should never be closed primarily and plastic surgical input may be required.

Primary amputation Primary amputation of a limb may be required as part of the initial debridement. This is a relatively easy decision in a limb that is ‘hanging off’ and completely non-viable.

Scoring systems have been developed in an attempt to determine objectively whether a limb requires amputation, such as the Mangled Extremity Severity Score (MESS). However, review of limb injury severity scores by Bosse et al. in 200112 cast significant doubt on their value in predicting the need for amputation and operator experience is thought to be the most important factor. Even in the military environment, a second opinion should be sought if possible, prior to carrying out an amputation, and these should be carried out at the lowest level possible, rather than creating formal flaps at the time of initial surgery.

Contamination of military wounds Most open fractures are contaminated at the time of injury but military wounds are likely to be more heavily contaminated with bacteria than civilian wounds. Much of the evidence for this comes from Lindberg et al. 195513 who studied the degree of contamination of medium and large wounds (41 cm) in the Korean War. All were found to be contaminated with Clostridia, with a mean of 2 different strains of Clostridium per wound. There was also a seasonal variation in the aerobic bacteria contaminating wounds. In summer 89% of wounds were found to be contaminated, with faecal organisms predominating. In winter 81% of the wounds were contaminated with aerobic bacteria, but these were most commonly staphylococci and streptococci. A further report from Korea by Strawitz in 195514 looked at serial biopsies of 11 penetrating wounds, isolating Clostridia in 9 of 11 wounds, as well as isolating Gram-positive cocci from all wounds and gram negative bacilli from 9 of 11 wounds (mean of 5.5 species of bacteria isolated per open wound). Notably, Strawitz found that wounds remained contaminated after debridement unlike civilian open fractures described by Robinson et al. 1989.15 Wound biopsies 5-6 days after debridement isolated a total of 11 species of gram negative bacteria and 10 of gram positive cocci (with a mean of 1.9 species of aerobic bacteria isolated per open wound). Looking at reports collectively, they confirm that most military wounds are heavily contaminated with 3–4 different species of bacteria and sometimes more,16 as compared with only 1 species found in most civilian wounds.13,17,18 Also of note was that although the species of aerobic bacteria appeared to be quite similar, anaerobic bacteria contaminated most military wounds but were rarely isolated from civilian wounds. Of particular relevance to the contamination of military wounds is the delay in treatment that often occurs. In a

ARTICLE IN PRESS Military limb injuries/ballistic fractures retrospective study of trauma centres in the United States,19 a mean delay of 37 min was reported between the 911 call and the patient arriving at the trauma centre. This is in stark contrast to the time it took from injury to arriving at a British Surgical Hospital during the first and second Gulf Wars. During the first Gulf War of 1991, the mean delay was 10.2 h,20 whilst during the recent second Gulf War with much shorter lines of communication and better casualty evacuation, the mean delay was still 6 h.21

Stabilization of fractures Stabilization of fractures provides pain relief and helps to prevent further bone and soft tissue injury. There are several methods for fracture stabilization in both the military and civilian environments, but there can be some additional problems within the military environment. Each method of splinting has advantages and disadvantages, but adequate local surgery is far more important than the method of stabilization. Plaster of Paris This is an inexpensive and easy to apply method of external splintage. The first use of plaster as a splint was by Antonius Mathlisen (1805–1878) to treat fractures on the battlefield.22 Its value within the military environment should not be underestimated because relatively little equipment is required and it can be performed relatively close to the front line. Plaster back-slabs should be used, which can be supplemented by lateral slabs when used in the lower limb, particularly with the knee and ankle joints. It is essential in the acute situation, or after initial debridement, that plaster does not encircle the limb; if cylinders are used then they must be split along their entire length down to skin. The disadvantages of plaster are in its inability to control movement at the fracture site, and shortening and malunion are common with multi-fragmentary fractures. This is clearly a problem with high-energy transfer wounds where there is often significant bone loss and there may have been vascular repair or soft tissue transfer, but plaster is often suitable for low-energy transfer wounds. Other problems include difficulty of access to wounds and other injuries such as burns and the polytrauma patient, where surgical stabilization is indicated. Infections are common in open tibial fractures treated with plaster, but have been shown to be comparable to other methods of treatment.23 Other external splints The Thomas splint was specifically designed for the evacuation of soldiers with ballistic fractures of the femur during the First World War. With the increasing use of intramedullary nailing for civilian femoral fractures, its use has been significantly reduced, but it remains a useful method of stabilizing fractures in the military environment, either alone or in combination with plaster. The disadvantages of the Thomas splint are related to the prolonged immobilization necessary and the difficulty of wound access.

351 Traction Before the advances in both external and internal fixation, traction was widely used to stabilize fractures that were difficult to manage in plaster, such as unstable or open fractures. It still does have a place in the management of fractures when limited resources are available and has been used extensively by the Red Cross. The problem in the military environment is that rapid, prolonged or repeated evacuation may be necessary, making this method of stabilization less than ideal. For good results to be obtained with traction requires experience of the technique, such that regular adjustments may be required to ensure good position, which once again may be more difficult to achieve within the military environment. However, within a suitable environment, traction can be used both in the initial and definitive treatment of ballistic fractures.

Intramedullary fixation Intramedullary (IM) fixation with a nail (Fig. 11) is currently considered to be the method of choice for the stabilization of open tibial or femoral fractures in the civilian environment.24 The advantages of IM nailing are high rates of healing for both wound and fracture and no requirement for additional splintage, allowing full access to the wound for inspection, dressings or plastic surgical procedures. In the USA IM nailing is used extensively in the treatment of ballistic long bone fractures, but these are normally low energy, minimally contaminated fractures that are usually treated within 6 h of wounding. Its main disadvantages are that it is technically demanding and requires a large amount of equipment, including image intensification. For this to be appropriate within the military environment, the surgical facility would need to be relatively well established, static, and not too close to the front line. Rich et al. 197125 reviewed the results of open fractures from Vietnam that required vascular repair but also discussed the method of fracture stabilization. They reported that 50% of those that underwent IM nailing required removal of the nail due to complications directly related to the implant. The most common complication was infection and they concluded that within the military

Figure 11 Ballistic fracture of the femur stabilized by internal fixation with an IM nail; the fragments of the bullet did not need to be removed.

ARTICLE IN PRESS 352 environment external splints with the use of transfixion pins was a safer option for stabilization of fractures with associated vascular injuries. Animal models have also confirmed the high infection rate after primary IM nailing of a heavily contaminated wound.26 This agrees with a civilian recommendation to avoid nailing certain open fractures when dirty water or agricultural contaminants are present,27 which is likely to apply to the military environment. In summary, primary IM nailing is not recommended in the military environment. It could be carried out after evacuation of the casualty, possibly after plaster immobilization, which would provide good initial stabilization. Internal fixation The US military have used delayed internal fixation for open fractures with some success,28 but it was dismissed as a technique by the British military in both the First and Second World Wars. The advantages of this type of fixation with plates and screws, are accurate reduction and rigid fixation of fractures as shown in Fig. 12, but it is technically demanding and requires very good post-operative management. This makes it an unsuitable technique close to the front line and its only role is as a secondary mode of treatment after initial stabilization with plaster. Although delayed internal fixation has been shown to have a lower rate of complications than acute plating, the complication rates of both infection and delayed union are still high. Therefore internal fixation probably has little place in the management of the military ballistic fracture. Little information is currently available on the use of percutaneous or locking plates and it is possible these may be indicated in the secondary management of military fractures. External fixation External fixation is one of the main methods of stabilizing military ballistic fractures. Indications include:

 extensive bone loss,  large soft tissue wounds,  vascular injuries that require repair,

Figure 12 Ballistic fracture of the humerus stabilized by internal fixation with a plate; the fragments of the bullet did not need to be removed.

D. Griffiths, J. Clasper

 fractures in association with burns,  multiple injuries and  to facilitate casualty evacuation. Although external fixators are often considered easy to apply, in the case of ballistic fractures they can be technically quite difficult because of the multifragmentory nature of most of these fractures29, which can lead to a high complication rate. Dubravko et al., 199430 reported that during the war in Croatia of 116 external fixations, complications occurred in 79 (68.1%); pin tract infection occurred in 35.3% and pin tract osteomyelitis in 7.8%. Additionally 8 (6.9%) of the fractures required re-operation for loss of reduction at the fracture site. A further retrospective review of the treatment of limb injuries in Croatia by Has et al. 1995,31 found that of the 1320 open fractures treated, external fixation was used in 215 (16.3%). Of these 215, 20 (9.3%) developed osteomyelitis and 21 (9.8%) developed non-union and then 9 of these 21 went on to subsequently develop osteomyelitis. Unfortunately information on the other forms of fixation used and their complication rates was not given and comparisons could not be made. Although external fixation has been accepted as the treatment of choice by many surgeons in the management of open fractures from missiles, this is in spite of the fact that there are very little follow-up data, and the data that are there shows quite high complication rates. The technique of external fixation. As with the civilian use of external fixators, it is essential to ensure safe placement of pins because damage to adjacent structures is one of the complications of external fixation. Hence, pins should preferably be inserted into the subcutaneous surface of a bone. If this is not possible, pins can be inserted through ‘‘stab’’ incisions, which should be at least 1 cm in length. Open insertion should be used for distal humeral and distal radial pins. External fixation pins must pass across both cortices and with many, pre-drilling is essential. This is not the case with the British military pattern fixators, which have been designed to be self-drilling and self-tapping. Ideally all pins are connected to the same bar, which is best achieved by inserting proximal and distal pins first, then the fracture is reduced as accurately as possible. These pins are connected

Figure 13 Ballistic fracture of the elbow stabilized by a bridging external fixator.

ARTICLE IN PRESS Military limb injuries/ballistic fractures to a single bar by pin-to-bar connectors and the bar then acts as a guide for further pin placements. A second bar should also be inserted in most cases to increase frame stability. Sometimes it is not possible to use just one bar and 2 or 3 bars are connected in a variety of arrangements, as shown by the bridging external fixator in Fig. 13.

References 1. Shephard GH. High energy, low velocity close-range shotgun wounds. J Trauma 1980;20:1064–7. 2. Liu Y, Chen X, Chen SLX, et al. Wounding effects of small fragments of different shapes at different velocities on soft tissues of dogs. J Trauma 1988;28:S95–8. 3. Kirby NG, Blackburn G. Field surgery pocket book. London: Her Majesty’s Stationary Office; 1981. 4. Janzon B, Hull JB, Ryan JM. Projectile material interactions: soft tissue and bone. In: Cooper GJ, Dudley HAF, Gann DS, Little RA, Maynard RL, editors. Scientific foundations of trauma. Oxford: Butterman-Heinemann; 1997. 5. Janzon B, Seeman T. Muscle devitalisation in high-energy missile wounds, and its dependence on energy transfer. J Trauma 1985;25:138–44. 6. Ryan JM, Rich NM, Ochsner MG. Biophysics and pathophysiology of penetrating injury. In: Ryan JM, Rich NM, Dale RF, Morgans BT, Cooper GJ, editors. Ballistic trauma. London: Arnold; 1997. 7. Rose SC, Fujisaki CK, Moore EE. Incomplete fracture associated with penetrating trauma: aetiology, appearance and natural history. J Trauma 1988;28:106–9. 8. Clasper JC, Bowley DWG. Comminution and contamination after high-energy ballistic fractures of the tibia. J Bone Jt Surg (Br) 2000;82-B(Suppl 1):55. 9. Hill PF. Intramedullary nailing of contaminated tibial fractures. M. Chir., University of Cambridge, 2000. 10. Coupland RM, Korver A. Injuries from antipersonnel mines: the experience of the international committee of the red cross. Br Med J 1991;303:1509–12. 11. Coupland RM. Technical aspects of war wound excision. Br J Surg 1989;76:663–7. 12. Bosse MJ, Mackenzie EJ, Kellam JF, et al. A prospective evaluation of the clinical utility of the lower extremity injury severity scores. J Bone Jt Surg (Am) 2001;83-A:3–14. 13. Lindberg RB, Wetzler TF, Marshall JD, Newton A, Strawitz JG, Howard JM. The bacterial flora of battle wounds at the time of primary debridement. Ann Surg 1955;141:369–74. 14. Strawitz JG, Wetzler TP, Marshall JD, Lindberg RB, Howard JM, Artz CP. The bacterial flora of healing wounds: a study of the Korean battle casualty. Surgery 1955;37:400–7.

353 15. Robinson D, On E, Hadas N, Halperin N, Hofman S, Boldur I. Microbiologic flora contaminating open fractures: its significance in the choice of primary antibiotic agents and the likelihood of deep wound infection. J Orthop Trauma 1989; 3:283–6. 16. Levaditi MC, Gerard-Moisonnier, Brechot MMH, Tournay R. Nouvelles recherches sur la flore microbienne des traumatisms de guerres. Acad Med 1939;7:371–81. 17. Gustilo RB, Anderson JT. Prevention of infection in the treatment of one thousand and twenty-five open fractures of long bones. J Bone Jt Surg (Am) 1976;58-A:453–8. 18. Lawrence RM, Hoeprich PD, Huston AC, Benson DR, Riggins RS. Quantitative microbiology of traumatic orthopaedic wounds. J Clin Microbiol 1978;8:673–5. 19. Cornwell EE. Current concepts of gunshot wound treatment: a trauma surgeon’s perspective. Clin Orthop 2003;408:58–63. 20. Spalding TJW, Stewart MPM, Tulloch DN, Stephens KM. Penetrating missile injuries in the Gulf war 1991. Br J Surg 1991; 78:1102–4. 21. Hinsley D, Rosell P, Henman P, Clasper J. Penetrating injuries of the 2003 Gulf conflict. Society of Military Orthopaedic Surgeons. December 15–20 2003: Paper No.6 22. Van Assen J, Meyerding HW. Antonius Mathijsen, the discoverer of the plaster bandage. J Bone Jt Surg (Am) 1948;30-A:1018–9. 23. D’Aubigne RM, Maurer P, Zucman J, Masse Y. Blind intramedullary nailing for tibial fractures. Clin Orthop 1974;105:267–75. 24. Clasper JC. The interaction of projectiles with tissues and the management of ballistic fractures. J R Army Med Corps 2001; 147:52–61. 25. Rich NM, Metz CW, Hutton JE, Baugh JH, Hughes CW. Internal versus external fixation of fractures with concomitant vascular injuries in Vietnam. J Trauma 1971;11:463–73. 26. Hill PF, Parker SJ, Clasper JC, Watkins PE. Contaminated fractures of the tibia: the results of immediate intramedullary nailing in an animal model. Proc Br Orthop Res Soc 1998;5–6:28. 27. Templeman DC, Gulli B, Tsukayama DT, Gustilo RB. Update on the management of open fractures of the tibial shaft. Clin Orthop 1998;350:18–25. 28. Hampton OP. Delayed internal fixation of compound battle fractures in the mediterranean theatre of operations. Ann Surg 1946;123:1–26. 29. Clasper JC, Phillips SL. Early failure of external fixation in the management of war injuries. J R Army Med Corps 2005;150. 30. Dubravko H, Zarko R, Tomislav T, Dragutin K, Vjenceslav N. External fixation in war trauma management of the extremities—experience from the war in Croatia. J Trauma 1994;37: 831–4. 31. Has B, Jovanovic S, Wertheimer B, Mikolasevic I, Grdic P. External fixation as a primary and definitive treatment of open limb fractures. Injury 1995;26:245–8.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 354–360

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: BULLET AND BLAST INJURIES

(iv) Anti-personnel mine injuries Karl Trimblea,, Scott Adamsb, Michael Adamsc a

Ministry of Defence Hospital Unit, Derriford Hospital, Plymouth PL6 8DH, UK Edinburgh Orthopaedic Trauma Unit, UK c Avon Orthopaedic Centre, Bristol, UK b

KEYWORDS Anti-personnel mine injuries; Amputation

Summary Anti-personnel landmines give rise to injuries years after conflicts. Many injured are in third world countries. The predominantly lower limb injuries often are to children and care of these crippling wounds must reflect available local after care. & 2006 Published by Elsevier Ltd.

Introduction An estimated 160 million anti-personnel mines (APMs) are currently laid worldwide, many in unmarked mine fields. While the treatment of APM injuries has traditionally been thought of as the preserve of military and non-government organisation (NGO) surgeons,1 landmines can be found within the borders of mainland Europe, following recent civil wars. Hence, even European residents may potentially be victims of an injury often thought only to occur in war zones in the developing world. The International Committee of the Red Cross (ICRC) registry of mine incidents between 1991 and 1996 showed almost 10,000 casualties of APM injury were registered at ICRC hospitals worldwide, with half of those affected being civilian. Even when a country is declared ‘safe’ the legacy of unmarked minefields can continue indefinitely.2 In the recent conflict in Kosovo, APMs were laid along 120 km of the border with Albania and between 1998 and 2003 over 240 people were killed or injured by them. A quarter of those were children and many were injured long Corresponding author. Tel.: +44 1752 777111;

fax: +44 1752 768976. E-mail address: [email protected] (K. Trimble). 0268-0890/$ - see front matter & 2006 Published by Elsevier Ltd. doi:10.1016/j.cuor.2006.07.005

after hostilities had ceased. The conflict in Bosnia-Herzegovina, which ended in 1995, left thousands of APMs in both rural and urban locations; by 2003 over 4800 people had been killed or injured. Croatia, a popular tourist destination not commonly thought of as a problem area for minefields, has had over 1000 deaths and injuries from landmines. The final cost to life and limb of the conflict in Chechnya will be vast; current estimates by the ICRC are that there are more than 7000 APM-related amputees in the republic.3 APM injuries have wide-ranging physical, social and economic effects on all those affected by them. While the injuries may be un-survivable, many APMs are designed not to kill but incapacitate and maim their victims.4 The sight of colleagues who have suffered severe but survivable multiple injuries and subsequent amputations is thought to have more of a psychological effect on the local population than if they had been killed instantaneously (Fig. 1). APMs can be bought for as little as US$3 and laid with ease, yet de-mining is painstakingly slow and dangerous (Fig. 2). The huge costs of clearance to a country with already limited resources, means minefields and the potential for APM injury remains long after conflict ceases. They are often located in rural or farming areas and are of particular threat not to the military, but to farmers, women and children who repopulate these areas.

ARTICLE IN PRESS Anti-personnel mine injuries

355

Figure 4 Figure 1

The ‘PFM’ APM is widely available and cheap.

Figure 2 De-mining in progress.

Line of stones crudely indicating APMs.

scarce healthcare resources.2,5 Landmine victims present a difficult logistical problem to medical services involved in their evacuation and treatment. The average length of inpatient stay after a landmine injury is 32.3 days, with each victim undergoing an average of 4 operations and receiving 3.2 units of blood.3 The costs of prosthetics and rehabilitation are considerable. The victim may also be unable to provide financially for the family and may be seen as a burden. Many families may lose several members to injury and severe disability and the overall cost to the local community and country of this incapacitated yet young population with a near-normal life expectancy is vast.6 In 1997, the Ottawa Convention prohibited the use, stockpiling, production and international trade of APMs. Since its inception, over 140 governments have fully committed to the ideals of the convention; this number includes most of the countries devastated by land mine usage.4 There are, however, some notable exceptions; many major APM manufacturing nations have refused to sign up to the treaty. The convention has helped reduce the number of circulating landmines; in 1998 estimates of global landmine stockpiles were greater than 250 million in 108 countries; by 2003 this figure had fallen to approximately 200 million in 67 countries. However, by the end of 2004, there were still 163 million landmines worldwide.7

Mechanism of injury and medical management APM design

Figure 3 Anti-tank mines by the roadside (marked).

Minefields are sometimes marked with warning signs (Fig. 3), but more often locals who find mines are only able to mark their presence crudely with stones (Fig. 4). The presence of minefields limits agricultural development and economic reconstruction and APM injuries overload already

Most APMs are small enough to be laid in large numbers over a vast area (Fig. 5) at any one time, usually from the back of a truck or by air. Those laid by air may cover vast expanses of land rendering it impassable on foot. Runways can also be mined from the air rendering them useless as a taking off or landing point. The mine is usually triggered by being stepped upon, though some are activated by trip-wire, or designed to be sensitive only to heavier objects such as vehicles. The trigger sets off an explosive mobilising the mine casing and casing content, which varies depending upon design.

ARTICLE IN PRESS 356

K. Trimble et al.

Figure 5 Minefield. Figure 6

Contents include wire, ball-bearings, flechettes (small arrows) and phosphorous.

Mechanism of injury An explosive changes its physical structure by combustion (a thermal reaction) or detonation (an electro-chemical reaction) into an expanding gaseous by-product of the explosion. During this process energy is released and propagates outwards. The expanding gaseous mass is the blast wave. The blast wave maybe subdivided into its component parts which include the shock front, the blast wind and the thermal front. The shock front is the leading edge of the blast wave and is an expanding front of compressed air and energy that travels at speeds greater than sound in air and is propagated quickly within the body. Trailing the shock front, the blast wind is the expanding mass of explosive products and debris. Behind the blast wind is the thermal front, containing the heat energy of the explosion. With regard to APM injury, when triggered the mine detonates and the shock front is propagated through the foot and up the limb as well as radiating out into the surrounding air. The high-energy content of the shock front disrupts fascial planes and induces micro-fractures within bone. The vast kinetic energy of the debris and fragmentcontaining blast wind follows microseconds later and when incident upon the limb leads to massive tissue disruption and also imposes a turning moment about the limb (Fig. 6). Research has shown that partial or complete amputation usually occurs through the site of shock-front induced microfractures.8 The extent and level of the energy depends upon the amount of explosive and fragments contained within the APM. At first inspection, though the injury may appear to be contained at a particular level, the shock front and blast wind induced fascial plane separation propagates far proximally towards the groin and often the other limb. This means that extensive soft tissue injury and contamination will have occurred throughout the limb often with neurovascular injury and compromise far above the obvious injury level.

APM injury requiring amputation of foot.

Immediate treatment The ‘ABCs’ are initiated like any high-energy trauma injury and ATLS or BATLS (Battlefield Advanced Trauma Life Support) guidelines should be followed. Fragmentation injury to the face neck and thorax may present airway problems and rapid assessment is vital. Reversible life threatening conditions such as tension pneumothorax, open chest wounds and haemorrhage should be identified and dealt with immediately. Direct pressure on obvious haemorrhagic sites is advised. The use of tourniquets remains controversial as misuse may induce secondary hypoxic tissue injury and necrosis and subsequent renal failure if used for considerable lengths of time. There is most definitely a role for tourniquets in the immediate ‘at the scene’ resuscitation phase, especially in the battlefield situation, but care must be taken to place them as distally as possible. Intravenous access remains a priority, as this may be increasingly difficult in the shocked patient. Fluid resuscitation should be given in accordance with local policy be it military or civilian. The content and amount of fluid given remains a moot point. Ongoing concerns have again been raised in the recent literature with regard to excessive fluid-resuscitation of the trauma casualty. The use of hypotensive resuscitation is not the remit of this article, but a familiarisation with the pros and cons is advised to all those involved in trauma management. Stabilisation of the limb with splints to reduce pain and haemorrhage and intravenous pain relief and antibiotics are advisable (Fig. 7).

Surgical treatment Pre-operative planning is essential in the management of APM casualties, and basic surgical principles should not be ignored. A thorough examination of all limbs, the perineum, abdomen, thorax (Fig. 8) and face is necessary to ascertain extent of injuries and neurovascular status and extent of tissue loss in the affected limb(s). Several scoring systems exist that can assist in the decision making process though caution is advised in attempting to apply these preoperatively. The MESS and ICRC classifications are discussed below.

ARTICLE IN PRESS Anti-personnel mine injuries

Figure 7 Traumatic amputation from APM.

357

Figure 10 Following thorough wound and wound-track excision.

Figure 8 Upper limb APM injury with penetrating thoracic trauma. Figure 11 Fragmentation injury to elbow.

Figure 9 APM fragmentation to gluteal region.

At surgery all dead and devitalised tissue should be excised, including devascularised bone. Wound edges should be excised back to clean vascularised margins (Figs. 9 and 10) and copious lavage is necessary to remove all debris and contamination (Figs. 11 and 12). Muscle must be assessed for

Figure 12 Fragments removed from elbow in Fig. 11 after wound debridement.

ARTICLE IN PRESS 358 contamination, colour, consistency, contractility and capillary bleeding. Debridement should be thorough even though the consequence will be an increased chance of a more proximal surgical amputation. Few APM injuries to the contact limb are salvageable and most result in amputation; it is merely the level that is in question, not the decision to perform one. The contralateral limb must also be assessed independently of the other limb. Again first principles should not be ignored in the vain hope of a favourable outcome. The question of limb salvage versus sacrifice should not put the patient’s life or rehabilitation at risk. In western societies the MESS score often assists our surgical decision making, but in the developing world the patient may well fare better with an early well performed amputation and prosthesis than a limb that is salvaged, requiring many operations and the outcome of which is unpredictable (Fig. 13). The level of amputation is dictated by the involvement of the soft tissues and thus the extent of debridement. While some in the published literature suggest a favourable outcome of limited amputations of the foot and hind-foot, in general a below knee amputation allows easier fitting of widely available prostheses than a limited foot/hind-foot amputation. Debate exists as to the benefits of performing initial ‘guillotine’ type amputations and then performing a definitive amputation as a second procedure. However, the ICRC does not recommend this, feeling that it may compromise surgical debridement. It is rarely if ever performed in ICRC hospitals, the only exception is as a rapid life-saving procedure in the multiply injured patient. Definitive through knee amputations should be avoided at all costs. Of vital importance is not to close wounds primarily. Once adequately debrided the wounds and stump should be dressed with dry gauze and then left for about 5 days. All APM injuries are contaminated, and primary closure may lead to local or systemic sepsis and is to be discouraged. Repeated wound inspections are discouraged as little information is obtained from interfering with the wounds; insufficient debridement will manifest itself with erythema proximal to the wound and persistent pyrexia. Several visits to theatre may be necessary before delayed closure is possible. The use of a pneumatic tourniquet is recommended prior to removal of any dressing. It does not

K. Trimble et al.

Figure 14 Severe (pattern 3) hand injury.

interfere with identification of devitalised tissue and will limit further blood loss. The tourniquet should be released prior to the end of the procedure and haemostasis achieved. The purpose of delayed closure is to permit swelling of the injured tissues and allow drainage of the wound thus preventing the formation of an anaerobic environment; the second or third ‘look’ facilitates reassessment of devitalised or contaminated tissue.9 The ‘first line antibiotic’ is surgery and though there are schools of thought that argue that an adequate excision removes the need for chemical antibiotics, the significant long-term problems of infection make early use of appropriate antibiotics sensible.10 It is worth re-iterating that at the time of initial debridement the bone injury is assessed and appropriate excision carried out. The type of fixation employed for fractures or bone defects depend on a number of factors: the environment, equipment and surgical follow-up available, the associated soft tissue defect and surgical experience. Techniques available include plaster of Paris, skin and skeletal traction and external fixation. Once the wound is ready for closure, the decision as to the method of definitively restoring skin integrity can be made. Direct suture and split skin grafting are commonly used, but if skin is closed directly the closure should be tension free. While more complex flap coverage including free flaps may allow limb salvage in severe injuries, it must be re-iterated that the majority of APM injuries require amputation. A salvaged but persistently discharging, ununited, deformed limb is a sub-optimal outcome when compared to a well healed, appropriately placed and rehabilitated amputation.

Level of amputation

Figure 13 Traumatic amputation of contact foot with severe injury to contralateral foot.

The level of amputation is dictated by the injury. There are two appropriate levels for below knee amputation. For nonsalvageable foot injuries, with little involvement of the leg, an amputation may be performed through the middle third of the tibia, covered by a soleus muscle flap. However, more commonly due to injury proximal to the foot an amputation through the proximal tibia is performed, 10–15 cm below the knee (approx. 10 cm below the tibial tuberosity) with the fibula cut 2 cm more proximal. The degree and distribution of soft tissue injury usually dictates where the skin flaps are based. Many surgeons use a long posterior

ARTICLE IN PRESS Anti-personnel mine injuries

359

musculo-cutaneous flap. Though easily performed in the elective situation, in trauma surgery this is often difficult as the distal part of the flap may be involved in the injury and thus not leaving enough soft tissue to fashion the lengthy flap. The ‘skew’ flap technique uses two shorter but identical flaps, one based antero-medial, the other postero-lateral. Due to muscle bulk of the stump, the ICRC recommends sacrificing excess muscle and using medial gastrocnemius alone as a myoplastic flap allowing tensionfree facial closure over the bone ends. Again the pattern of injury will dictate where the muscle flaps are based. Drill holes in the tibia can facilitate anchoring of the soft tissues with sutures. More proximal injury often necessitates an above knee amputation. The level depends on degree of involvement of the thigh in the injury, but ideally the femur is divided approximately 12 cm above the knee, again with bevelling of the bone end. Equal semi-circular-shaped anterior and posterior skin flaps are created with their longest points just above the knee. The ICRC recommends coverage with vastus medialis. The wound is then closed with the scar offset to avoid scar sensitivity and breakdown on weightbearing. In the upper limb there is no optimum level of amputation and the bone should be divided as low as practical without compromising debridement. This is particularly true in the hand, where every centimeter possible should be conserved (Fig. 14). The bone should be bevelled and then filed smooth; the remaining soft tissues are then sutured over the ends. As weight bearing is less of an issue, simple skin graft techniques have an increased place in preserving length.10

Wound classification Of the numerous scoring systems available to categorise traumatic injuries and therefore aid management decisions only a handful are robust enough to help in the initial management of APM causalities. The pattern of injury and mechanism have been classified by Coupland and Korver (Table 1).11 These categories alert the clinician to associated occult injuries such as compartment syndrome with pattern 1 and chest or ocular trauma with pattern 3. The dilemma between a complex reconstructive procedure and a definitive amputation is not easily resolved. Knowledge of patient expectations in a 1st world civilian environment or surgical follow up available in a wartime or developing region help in the decision making process. The

Table 1 Pattern 1

ICRC APM injury classification.

Pattern 2

Traumatic amputation of part of the lower limb, less severe injuries elsewhere Multiple fragment wounds

Pattern 3

Injury to hands and face

Stood on mine Exposed to blast Handling mine

Table 2

Mangled extremity severity score.

Skeletal/soft tissue injury

Score

Low energy Stab, simple fractures, hand gun injury

1

Medium energy Open multiple-level fractures, moderate crush

2

High energy Rifle, close range shotgun injury, high-speed RTA 3 Very high energy Massive crush, gross contamination

4

Shock Systolic BP always 490 mmHg Transient hypotension Persistent hypotension

0 1 2

Limb ischaemia (Score doubled if warm ischaemic time 46 h) Perfusion normal, pulse reduced or absent Pulseless, paraesthesia, decreased capillary refill Cool paralysed, insensitive, numb

3

Patient age o30 31–50 451

0 1 2

1 2

Mangled Extremity Severity Score (MESS),12 which grades significant clinical factors, is held by some as a useful guide. Amputation is indicated in a case with a score of greater than 7 (Table 2). Nerve Injury, Ischemia, Soft Tissue Injury, Skeletal Injury, Shock, and Age of Patient Score (NISSSA)13 initially developed to address perceived weaknesses in the MESS has similar limitations. Both are specific in predicting limbs, which should not undergo amputation, but neither is sensitive. This lack of sensitivity may lead to complications resulting from delay of amputation.14 Pattern 2 injuries usually result from multiple fragments rather than the traumatic amputation related to patterns 1 and 3. These penetrating wounds are managed with adequate primary debridement and subsequent delayed primary closure. The classification system of choice in penetrating trauma is The Red Cross Wound Classification (Table 3).15 This allows for assessment of individual wounds and the energy imparted to the individual rather than the weaponry involved. Injury classification systems such as the injury severity score (ISS) and revised trauma score (RTS) are useful in predicting mortality and may have some use in the mass casualty situation. They are, however, of limited value in the management of individual patients. The predictive salvage index (PIS), Howe 1987 has been show to be less sensitive and specific that either the MESS or NISSSA systems.16

ARTICLE IN PRESS 360

Table 3

K. Trimble et al.

Red Cross Wound Classification

E X C F V M

Maximum diameter of the entry wound Maximum diameter of the exit wound (no exit wound, score X ¼ 0) Can wound cavity (may be established after skin incision) take 2 fingers? (C ¼ 0 no, C ¼ 1 yes) Fracture (F ¼ 0 no fracture, F ¼ 1 simple fracture, F ¼ 2 clinically significant comminution) Vital structure, brain, viscera, major vessel (V ¼ 0 no, V ¼ 1 yes) Metallic foreign body on radiography (M ¼ 0 none, M ¼ 1 one, M ¼ 2 multiple)

This is an anatomical classification and does not include a physiological variable. Dimensions are in centimetres.

Acknowledgements Pictures reproduced with kind permission of: Lieutenant Colonel Paul Parker RAMC Squadron Leader Teresa Griffiths RAF Corporal Myatt RAF.

References 1. Mannion SJ, Chaloner E. ABC of conflict and disaster: principles of war surgery. BMJ 2005;330:1498–500. 2. Geneva International Centre for Humanitarian Demining (GICHD) brochure. GICHD publication. /http://www.gichd.ch 2003S. 3. Trimble K, Clasper J. Anti-personnel mine injury; mechanism and medical management. J R Army Med Corps 2001;147:73–9. 4. ICRC. Special report; mine action. 2003. ICRC publication. /http://www.icrc.orgS. 5. Coupland RM. Assistance for victims of antipersonnel mines: needs, constraints and strategy. ICRC publication. /http:// www.icrc.org 1997S. 6. Coupland RM. The effects of weapons; defining superfluous injury and unnecessary suffering. Med Global Survival 1996; 3:A1. 7. Bannon I. Landmine contamination: a development imperative. Social development notes; conflict prevention and reconstruction. No. 20, October 2004. /http://www.mineaction.orgS. 8. Hull JB, Cooper GJ. Pattern & mechanism of traumatic amputation by explosive blast. J Trauma 1996;40(3 Suppl): S198–205. 9. Gray R. War wounds: basic surgical management. ICRC publication. /http://www.icrc.org, 1994S. 10. Roberts P. The British military surgery pocket book. Crown Copyright; 2004. p. 236–42.

11. Farquharson-Roberts M, in: Ryan JM, Rich NM, Dale RF, Morgans BT, Cooper GJ. Ballistic trauma. London: Arnold; 1997. p. 129–31. 12. Coupland RM, Korver A. Injuries from antipersonnel mines: the experience of the International Committee of the Red Cross. BMJ 1991;303:1509–12. 13. Helfet DL, Howey T, Sanders R, Johansen K. Limb salvage versus amputation: preliminary results of the mangled extremity severity score. CORR 1990;256:80–6. 14. McNamara MG, Heckman JD, Corley EG. Severe open fracture of the lower extremity: a retrospective evaluation of the Mangled Extremity Severity Score. J Orthop Trauma 1994;8:81–7. 15. Bosse MJ, MacKenzie EJ, Kellam JF, et al. A prospective evaluation of the clinical utility of lower extremity injury severity scores. J Bone Jt Surg Am 2001;83:3–14. 16. Coupland RM. The Red Cross Wound Classification. ICRC publication. /http://www.icrc.org, 1997S.

Further Reading 1. The British military handbook is an extremely useful guide to treating battlefield injuries. 2. The ICRC produce several paperback publications that are essential reading for those involved in war surgery or severe extremity trauma. The website is /http://www.icrc.orgS. 3. Coupland R. Amputation for war wounds. ICRC publication; 1992. 4. Gray R. War wounds: basic surgical management. ICRC publication; 1994. 5. Rowley D. War wounds with fractures: a guide to surgical management. ICRC publication; 1996.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 361–366

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

CHILDREN

Fractures of the femoral neck in children C. Holton, P. Foster, P. Templeton Department of Orthopaedics and Trauma, Leeds General Infirmary, Great George Street, Leeds LS1 3EX, UK

KEYWORDS Fracture; Femoral neck; Child

Summary Femoral neck fractures account for less than 1% of all children’s fractures. They are much rarer than in adults. Delbet classified them from types 1 to 4. Treatment is aimed at reducing the fracture and maintaining the reduction long enough to allow union within 12–16 weeks. Stabilisation of the fracture is most commonly with cannulated screws. Complications include avascular necrosis, coxa vara, non-union, premature physeal arrest and occasionally infection. & 2006 Published by Elsevier Ltd.

Introduction Femoral neck fractures account for less than 1% of all children’s fractures.1 The femoral neck in children is tough, dense bone with a thick, strong periosteum that requires highenergy trauma to fracture2 compared to the commonly seen hip fractures in the elderly osteoporotic population. Treatment is aimed at reducing the fracture and maintaining the reduction long enough to allow union within 12–16 weeks.3 Fracture patterns and classification differ in children compared to adults in part due to the different anatomy. A more precarious vascular anastomosis between the femoral neck and head leads to a higher incidence of avascular necrosis (AVN) after displaced fractures within the paediatric population. Growth disturbance is an important complication of surgical fixation due to the presence of the physeal plate.

trochanteric physis and a medial subcapital physis. The lateral physis becomes evident at roughly 1 year of age. The subcapital epiphysis ossifies at 4–8 months of age, with the greater trochanter ossific nucleus appearing at about 4 years of age.4 Fusion of the proximal femoral physis occurs in both sexes at about age 18, while fusion of the trochanteric physis occurs earlier at 16–18 years of age.4 Canale points out that these paediatric anatomical differences are important for two reasons; firstly that the radiographs highlighting the bony elements often fail to demonstrate the cartilage model of the femoral head, neck and trochanter, and secondly that growth arrest is possible after almost any hip fracture due to the large amount of cartilage present in children’s hips.4 The vascular anatomy of the paediatric proximal femur has been studied in-depth due to the high complication rate of AVN occurring in fractures around this area. Chung5 highlighted several points including:

Anatomy

 Ligamentum teres contributes very little blood supply The proximal femur grows from birth with one proximal femoral physis, which later splits into a lateral greater Corresponding author.

0268-0890/$ - see front matter & 2006 Published by Elsevier Ltd. doi:10.1016/j.cuor.2006.06.001



to the head until the age of 8, and as an adult serves only 20%. Medial and lateral circumflex metaphyseal vessels that traverse the femoral neck predominately supply the head

ARTICLE IN PRESS 362



C. Holton et al.

at birth. These later become virtually non-existent by the age of 4 owing to the development of the cartilaginous physis that forms a barrier to these penetrating vessels. As the metaphyseal vessels diminish their supply to the femoral head the lateral epiphyseal vessels become the main blood supply as they bypass the physeal barrier. These vessels can be identified as the posteroinferior and posterosuperior branches of the medial circumflex artery which supply the femoral head throughout the rest of its life.

Mechanism of injury High-energy trauma accounts for most paediatric hip fractures; five large studies found this to be the main cause in approximately 80% of cases,1,6–9 usually attributable to road traffic accidents or falls from height.9 Hip fractures caused by low energy trauma should raise suspicion of a pathological fracture; causes include unicameral bone cyst, aneurysmal bone cyst or fibrous dysplasia. Non-accidental injury may rarely be responsible.

reduction, followed by secure fixation by way of smooth pins if the patient is less than 3 years of age and threaded screws if older.

Type 2 Internal fixation with threaded or cannulated hip screws was suggested by Canale4 as the treatment of choice for transcervical fractures even if they are undisplaced.

Type 3 Cervico-trochanteric fractures which are definitely undisplaced can be treated in an abduction hip spica cast4 However, if displacement is evident then internal fixation

Classification Delbet’s10 simple classification is the most widely used and accepted (approximate percentage of incidence).3 Type 1 Type 2 Type 3 Type 4

Transphyseal, with or without dislocation of the femoral head (3%) Transcervical, displaced or undisplaced (50%) Cervico-intertrochanteric (37%) Intertrochanteric (10%) Figure 1 AP pelvis radiograph showing transcervical fracture of the left proximal femur.

Treatment There is no recognised agreed method for treatment of each type of fracture. Ratliff recommended a plaster hip spica to be used to treat undisplaced fractures,1 although fixation in the older child is more appropriate.

Type 1 Canale4 advocates that trans-physeal fractures with or without dislocation should be treated with anatomical

Figure 2 Lateral left hip radiograph showing transcervical fracture of the left proximal femur.

ARTICLE IN PRESS Fractures of the femoral neck in children should be performed to reduce the risk of coxa vara and possible non-union.

Type 4 Intertrochanteric fractures can be treated by skin traction followed by an abduction spica in young children.4 If the fracture cannot be reduced or maintained by casting then internal fixation with a screw and side plate is necessary. Blockey3 advocates manipulative reduction, threaded screw fixation and postoperative immobilisation no longer than 4 months for all types of fractures. He states that if manipulative reduction is unsuccessful then open reduction is required, with the interposed soft tissue removed. The final option is a sub-trochanteric osteotomy. Ratliff1 felt that subtrochanteric osteotomy should be used routinely for displaced type 2 transcervical fractures. Canale4 concluded that all types 2, 3 and 4 fractures in older children should undergo open reduction and internal fixation whatever the degree of displacement, particularly in the presence of multiple injuries.

Complications The incidence of complications varies between 20% and 60%.4,6,11,12 It should be noted that the higher figure of 60% occurred in two studies with long-term follow-up of their cases.6,12 This suggests that these injuries may well not reveal their full impact until several years after the initial insult.

Figures 3 and 4 femur.

363

Avascular necrosis Morsy’s6 study of 53 children found that 40% developed AVN, similar to Ratliff’s 42% in his review of 73 cases in 1962.1 His study revealed no AVN in undisplaced fractures. Type 1 displaced fractures were found to have a 75% chance of developing AVN by Canale8 and Ng.13 This is in marked contrast to the low AVN frequency of less than 10% with type 4 displaced fractures.1,8,13 Studies6,11 of types 2 and 3 fractures have confirmed that the amount of initial fracture displacement is directly proportional to the risk of AVN due to the compromise of blood supply to the proximal femur at time of the injury. Swiontkowski7 proposed that this vascular compromise was likely to be due to rupture or kinking of the vessels by the displaced fracture. This lead Swiontkowski,7 Ng13 and Cheng14 to advocate early hip decompression in displaced types 2 and 3 fractures, to successfully reduce the incidence of AVN from 50% to 10%. The decompression was performed at the time of open reducte of the fracture by either aspiration or capsulotomy. Anatomical reduction of displaced fractures was also found to be a significant factor predicting the likelihood of developing AVN by Morsy.6 He found a decrease from 71% in the non-anatomical reduction group to only 17% in the anatomically reduced group, therefore advising anatomical reduction in every case to reduce the risk of AVN. Several authors conclude that it is these two factors, the initial severity of the trauma and degree of displacement, rather than the mode of treatment that most closely relates to the risk of AVN.1,4,15

AP and lateral radiographs 8 weeks postoperatively showing double cannulated screw fixation of the left proximal

ARTICLE IN PRESS 364

Coxa vara Lam’s16 most common complication was coxa vara (31%), which is consistent with Morsy’s (36%). Morsy’s affected patients included 11 cases of transcervical fractures and eight patients with basicervical fractures.6 Togrul9 found 8% of their cases developed coxa vara but this was confined to the conservatively treated cases. Several studies have found the incidence of coxa vara to be reduced by internal fixation of these injuries.1,4,6

Non-union The incidence of non-union varies from 1.6% in Togrul’s review13 to 36% in Morsy’s study.6 Morsy comments that this significant increase was probably due to infection in 23% and non-anatomical reduction in 38% of their cases. This would be consistent with Togrul’s series accounting for their low incidence of non-union being due to their high rate of anatomical repositioning.13

C. Holton et al. developmental delay and epilepsy. Her mobility was poor, only able to walk a few steps with the aid of one stick. Following the fall she was unable to weight bear and held her left hip and knee flexed. Radiographs showed generalised osteopenia with a transcervical fracture (Delbet type 2) of the left proximal femur (Figs. 1 and 2). She was transferred to our centre and was treated with reduction under image intensifier and fixation with two 6.5 mm cannulated screws (Figs. 3 and 4), a hip spica for 6 weeks and follow up for 3 years. She was then pain-free with a good range of movement of the hips, although overall mobility was still poor due to the underlying syndrome. There was no evidence of avascular necrosis on subsequent radiographs.

Case 2: cervicotrochanteric fracture (type 3) following a significant fall A 5-year-old boy with ADHD fell from a first floor window on to concrete. Sustaining a right cervicotrochanteric proximal femur fracture (Delbet type 3) and no other significant

Premature physeal closure The incidence of premature physeal closure varies greatly from 6.5% in Ratliff’s1 to 62% in Canale’s.8 Morsy found a correlation between AVN and premature physeal closure6. If AVN is linked to initial displacement of these fractures it seems reasonable to propose that Canale’s cases may well have undergone more severe initial displacement to cause their increased incidence of premature physeal closure. If an implant crosses the physis, an increase in the rate of premature physeal closure1,6,8 led Togrul13 to comment that this complication is mostly a result of an intra-operative error by an inexperienced surgeon. This means that premature physeal closure is linked to both the force of energy involved as well as the use of internal fixation rather than there being an increased incidence related to the fracture type.

Figure 5 AP pelvis radiograph showing cervicotro-chanteric fracture of the right proximal femur.

Infection Canale4 quotes infection as being uncommon. Morsy6 reported a high rate of infection, stating that 67% of these cases were treated with internal fixation and the rest underwent subtrochanteric osteotomy. He makes no comment in his methods as to whether prophylactic antibiotics were used in these patients.

Cases Three brief cases are shown which demonstrate different fracture types.

Case 1: Delbet Type 2 fracture in an 8-year-old child with Rett’s syndrome treated with 2 cannulated screws An 8-year-old girl presented to a local district general hospital having fallen from the edge of a bed from a height of 90 cm. She had a diagnosis of Rett’s syndrome with

Figure 6 Lateral radiograph of the right hip showing cervicotrochanteric fracture of the proximal femur.

ARTICLE IN PRESS Fractures of the femoral neck in children

Figures 7 and 8

365

AP and lateral radiographs of the right hip 6 weeks postoperatively following fixation with three cannulated screws.

Figure 9 AP pelvis showing a type 4 fracture of the left proximal femur.

injuries (Figs. 5 and 6). He was placed in a Thomas splint and transferred to our centre. The fracture was reduced under image intensifier and fixed with three 4.0 mm cannulated screws (Figs. 7 and 8) and protected in a spica cast for 6 weeks. At 16 months postoperatively he was noted to be fully weight bearing with a good range of movement in the hips. There was no evidence of avascular necrosis on subsequent radiographs.

Case 3: Intertrochanteric (type 4) fracture following RTA A 13-year-old boy was a passenger in a head-on vehicle collision sustaining an intertrochanteric fracture (Delbet

Figure 10 Lateral radiograph of the left hip showing a type 4 fracture of the proximal femur.

type 4) of the left proximal femur and no other significant injuries (Figs. 9 and 10). He was treated with an 80 mm dynamic hip screw and 1351 four-hole plate (Figs. 11 and 12). Within 3 weeks he was fully weight bearing without pain.

ARTICLE IN PRESS 366

C. Holton et al.

Figures 11 and 12

AP and lateral radiographs 6 months postoperatively with dynamic hip screw.

Summary Proximal femoral fractures in children are uncommon and are usually due to high energy trauma. The presence of growth plates and a precarious blood supply makes them different from adult fractures. Complication rates are high and often do not manifest until several years after the initial fracture management. These patients should therefore be followed up until they reach skeletal maturity. The high complication rates and rarity of these fractures suggest that ideally only surgical teams with a high degree of experience of paediatric trauma should manage them. In our experience we would suggest anatomical reduction and fixation with cannulated screws for types 1, 2 and 3 fracture patterns. For management of type 4 intertrochanteric fractures we suggest the use of a sliding screw and side plate.

References 1. Ratliff AHC. Fractures of the neck of the femur in children. J Bone Jt Surg 1962;44B:528–42. 2. Meyers MH. Fractures of the Hip. Chicago: Year Book Medical Publishers; 1985. p. 93–106. 3. Blockey NJ. Fractures around the hip. Bennet GC, editor. Paediatric Hip Disorders, vol. 9. Oxford: Blackwell Science; 1987. p. 188–96.

4. Canale ST, King RE. Fractures of the hip. In: Rockwood CA, Wilkins KE, King RE, editors. Fractures in children, vol. 3B, 3rd ed. Philadelphia: Lippincott JB; 1991. p. 1046–93. 5. Chung SMK. The arterial supply of the developing proximal end of the human femur. J Bone Joint Surg 1976;58A:961–70. 6. Morsy HA. Complications of fracture of the neck of the femur in children. A long-term follow-up study. Injury 2001;32:45–51. 7. Swiontkowski MF, Winquist RA. Displaced hip fractures in children and adolescents. J Trauma 1986;26:383–8. 8. Canale ST, Bourland WL. Fractures of the neck and intertrochanteric region of the femur in children. J Bone Surg 1977;59A:431–43. 9. Togrul E, et al. Fractures of the femoral neck in children: longterm follow-up in 62 hip fractures. Injury 2005;36:123–30. 10. Delbet P. Fracture of the neck of the femur in childhood. A report of six cases. Ann. Surg. 1928;88:902. 11. Morrissy R. Hip fractures in children. Clin Orthop 1980;152: 202–10. 12. Ovesen O, Arreskoc J, Bellstrom T. Hip fractures in children. A long-term follow-up of 17 cases. Orthopaedics 1989;12:361–7. 13. Ng GPK, Cole WG. Effect of early decompression on the frequency of avascular necrosis in children with fractures of the neck of the femur. Injury 1996;27:419–21. 14. Cheng JC, Tang N. Decompression and stable internal fixation of femoral neck fractures in children can affect the outcome. J Pediatr Orthop 1999;19:338–43. 15. Hughes LO, Beaty JH. Fractures of the head and neck of the femur in children: current concepts reviewed. J Bone Joint Surg 1994;76A:283–92. 16. Lam SF. Fractures of the neck of the femur in children. J Bone Joint Surg 1971;52A:1165–79.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 370–375

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

UPPER LIMB

Treatment of the painful biceps tendon—Tenotomy or tenodesis? F. Lam, D. Mok Shoulder Unit, Department of Orthopaedics, Epsom General Hospital, Dorking Road, Epsom, Surrey KT18 7EG, UK

KEYWORDS Biceps; Tenodesis; Tenotomy; Shoulder surgery

Summary The function of the long head of biceps tendon in the shoulder remains controversial. Pathology of the biceps tendon such as tenosynovitis, subluxation and pre-rupture are intimately associated with rotator cuff disease. Treatment therefore varies widely among surgeons and range from non-operative management to biceps tenotomy or tenodesis. The purpose of this article is to provide an up to date review on the indications and results of biceps tenotomy and tenodesis. & 2006 Elsevier Ltd. All rights reserved.

Anatomy The anatomical origin of the long head of biceps tendon is variable. It arises most commonly from the glenoid labrum (45%), less commonly from the supraglenoid tubercle (30%) and in the remaining it arises from both the glenoid labrum and the supraglenoid tubercle (25%). The tendon travels obliquely within the glenohumeral joint to exit beneath the transverse humeral ligament along the intertubercular sulcus or bicipital groove. In the glenohumeral joint the tendon is encased within a synovial sheath, which ends as a blind pouch at the end of the bicipital groove. As a result, the biceps tendon is intraarticular but extrasynovial. The average length of the tendon is 102 mm. It is interesting to note that the shape and cross-sectional area of the tendon changes as it runs from proximal to distal. At its proximal attachment near the glenoid, it has an average cross-section of 8.4  3.4 mm. This decreases to 4.5  2.1 mm as the tendon leaves the bicipital groove. Corresponding author.

0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.06.007

Therefore, rupture of the biceps tendon most commonly occurs proximally near the glenoid labrum and distally in the bicipital groove.

Function The biceps muscle-tendon unit is one of many structures in the human body to cross two joints. In the elbow, it serves primarily as a forearm supinator. Its secondary role is as an elbow flexor. Whilst its function at the elbow is clear, its role in the shoulder joint remains controversial. Cadaveric studies1–3 suggest that the long head of biceps acts as a humeral head depressor, anterior stabiliser, posterior stabiliser, limiter of external rotation, lifter of the glenoid labrum as well as a humeral head compressor. Warner4 studied the change in acromiohumeral interval on plain radiographs in patients with isolated loss of the proximal attachment of the long head of biceps. He found that there was 2–6 mm of superior translation of the humeral head in all patients in all positions of shoulder abduction except at zero degrees. He concluded that the long head of biceps acts

ARTICLE IN PRESS Treatment of the painful biceps tendon as a stabiliser of the humeral head in the glenoid during shoulder abduction in the scapular plane. The most recent biomechanical data come from Youm et al.5 who found that loading of the long head of biceps tendon significantly affects the glenohumeral joint range of movement, translations and kinematics. They concluded that the long head of biceps acts as a ligament at the extreme of motion to shift the humeral head to a position more centred on the glenoid. Electromyographic studies6–8 have produced conflicting results. There have been seven studies confirming that the long head of biceps acts as a shoulder flexor, three studies supporting its role as a shoulder abductor, two for internal rotator, one for external rotator and one as a shoulder extensor, one as anterior stabiliser of the shoulder and one study showing that it has abductor function only with resistance. A common limiting factor in all these studies is that the electromyographic activity from the motion of the elbow and forearm was not controlled during measurement of shoulder activity. Levy et al.6 have controlled this variable using a long arm brace, locking the elbow in extension and forearm in a neutral position. They concluded that the long head of biceps is not active in isolated shoulder motion when the elbow and forearm are controlled. They postulated that activity of the biceps tendon in the shoulder is achieved by either passive proprioception of the tendon or by active tension in association with the elbow and forearm activity. It seems evident therefore that the long head of biceps does not have a primary function in the shoulder but instead has multiple secondary roles. It is hardly surprising, therefore, to know that there is no single reliable clinical test to diagnose biceps pathology, as there is no primary function that can be isolated.

Pathophysiology Pathology of the biceps tendon can be broadly divided into three main types: inflammatory, instability and traumatic. Clearly, there is a huge overlap between these categories and in fact biceps pathology is very rarely a single entity (Fig. 1). The pathology most commonly seen is biceps tenosynovitis associated with a rotator cuff tear. This is related to its anatomical arrangement, since the biceps tendon sheath is continuous with the synovium of the glenohumeral joint and therefore any inflammatory process affecting the rotator cuff is likely to affect the long head of biceps as well. Hence, the detection of fluid in the biceps

Figure 1 There is often an overlap between different pathologies of the biceps tendon.

371 sheath on ultrasound is highly sensitive for rotator cuff disease.

Primary versus secondary biceps tendinitis Primary biceps tendonitis, in which there is isolated pathology affecting the tendon, is rare. One of the few studies that supported the existence of primary biceps tendonitis comes from Berlemann and Bayley9 who reported the long term results of 14 patients (15 shoulders) following keyhole biceps tenodesis. Fifty-three percent of patients had previously undergone a subacromial decompression but symptoms persisted until the biceps tenodesis was carried out. This would suggest that biceps tendinitis is a primary event. Other researchers, however, believe that biceps tendonitis is secondary to an ongoing subacromial impingement.10,11 Since the biceps tendon occupies a relatively antero-superior location within the impingement zone, it is prone to mechanical impingement. Neer10,11 believes that 95% of biceps tendonitis is secondary to impingement. Neviaser12 has also reported that there is a strong association between rotator cuff tear and biceps tendonitis. In a large series of 210 patients with impingement, Walch13 found that 70% had concomitant biceps pathology. This is supported by another large series of 200 patients by Murthi et al.;14 49% had evidence of biceps pathology and 40% required subsequent tenodesis.

Treatment The treatment of biceps tendonitis remains controversial. Spontaneous rupture of the long head of biceps is very common but is seldom associated with any significant long term functional deficits. Mariani et al.15 compared 30 patients with spontaneous rupture of the long head of biceps treated non-operatively with 26 patients who underwent early biceps tenodesis. They found that there was a loss of 21% of supination strength in the group treated non operatively compared to 8% in the tenodesis group. There was no difference in elbow flexion strength but the group treated non-operatively returned to work earlier. It seems therefore that spontaneous rupture of the long head of biceps tendon can be treated adequately without surgery (Fig. 2). In the context of rotator cuff disease, treatment of the degenerate biceps tendon is more controversial. Surgical options include benign neglect with treatment of concomitant rotator cuff disease only, inspection and synovectomy, repair of partial tear, tenotomy and tenodesis. Proponents for biceps tenotomy advocate that the procedure is simple to do, has limited surgical morbidity, bears no postoperative restriction, avoids implant complications such as hardware loosening, tendon pull-out, bicipital spasm etc.12,15 Moreover, most patients requiring this procedure are elderly with low functional demand. One of the largest studies of the results of biceps tenotomy comes from the data in Lyon by Walch et al.13 Between 1988 and 1999, he carried out 390 biceps tenotomies for full thickness rotator cuff tear. The cuff was not repaired but 35% of patients did have a subacromial decompression. After a mean follow up of 57 months, the mean Constant score improved from 48 preoperatively to 67

ARTICLE IN PRESS 372

F. Lam, D. Mok

Figure 2 Arthroscopic view of a ruptured long head of biceps tendon.

postoperatively. The patient satisfaction rate was 87%. Repair of the rotator cuff was required in only 1% of patients and further surgery for cuff tear arthropathy was required in 2%. The authors concluded that in selected patients in whom repair of the rotator cuff is neither possible nor desirable, good objective improvement can be expected with arthroscopic biceps tenotomy. They advocated that the rotator cuff should not be repaired in:

1. Patients aged over 65. 2. Patients with an irreparable rotator cuff tear, indicated by a reduced acromiohumeral interval of less than 7 mm and evidence of extensive fatty infiltration of the rotator cuff musculature on MRI. 3. Patients not willing to undergo post-op rehabilitation for rotator cuff repair.

Gill et al.16 have also reported favourable results of biceps tenotomy. In 30 patients with biceps tenosynovitis, dislocation or partial rupture treated with a simple arthroscopic tenotomy, they found that there was significant improvement in functional score and reduction in pain. The time taken for return to work averaged 1.9 weeks. There was generally a high patient satisfaction rate although one patient did require revision tenodesis due to cosmetic deformity. The overall complication rate was 13% and included loss of overhead function secondary to impingement, persistent pain and cosmetic deformity. The incidence of the Popeye sign (Fig. 3) caused by distal migration of the long head of biceps stump following biceps tenotomy is in fact far more common. In 54 patients treated with long head of biceps release carried out as an adjunctive procedure for a variety of conditions including rotator cuff tear, glenohumeral osteoarthritis and instability, the overall incidence of Popeye sign was 70% with 38% complaining of persistent biceps fatigue discomfort after resisted elbow flexion.17 It is interesting to note that there was a marked difference in the incidence of Popeye deformity between men and women, 83% and 37%, respectively.

Figure 3 Popeye sign following proximal rupture of long head of biceps.

The great majority of biceps pathology is encountered with degenerative rotator cuff disease, many of which may not be suitable for repair. In the treatment of a full thickness rotator cuff tear, Maynou et al.18 advocates biceps tenotomy as he believes that the biceps tendon is often the cause of part or all of the pain. His reasoning is that the long head of biceps occupies an antero-superior position in the shoulder, and as the shoulder forward flexes, the biceps can impinge against the vault of the acromion. In his series of 40 shoulders, he reported satisfactory mid-term results with an average of 341 gain in forward flexion of the shoulder. Although this was described as a simple procedure with limited functional consequences, he recommended this procedure for irreparable rotator cuff tears. Klinger et al.19 compared the results of arthroscopic debridement in massive irreparable rotator cuff tears with and without biceps tenotomy. Both groups had significant improvement in Constant scores although the difference between the two groups was not statistically significant. In the group treated with arthroscopic debridement and biceps tenotomy, he found that there was a longer duration of post operative pain relief although the final pain scores were similar. Overall, they found that the additional tenotomy did not significantly improve the outcome.

Biceps tenodesis Advocates for biceps tenodesis20 argue that tenodesis is required to:

 Re-establish the resting muscle length and thereby    

maintain the length-tension relationship. Prevent muscle atrophy. Avoid cramping pain. Maintain elbow flexion and supination strength. Avoid cosmetic deformity (Popeye).

ARTICLE IN PRESS Treatment of the painful biceps tendon Yamaguchi21 recommends biceps tenodesis when there is:

 More than 25% partial thickness biceps tear.  Chronic atrophic changes in the tendon.  Any subluxation of the biceps tendon from the bicipital groove.

 Altered anatomy of the bicipital groove which would  

make auto-tenodesis unlikely. For example, four part fracture of the proximal humerus. More than 25% reduction or atrophy of normal tendon width. Failed decompression in context of rotator cuff tendonitis [relative indication].

The most common indication for biceps tenodesis in Boileau’s series was massive, degenerative and irreparable rotator cuff tear, accounting for 79%.22 Less common is the case of isolated biceps pathology in the presence of an intact cuff (14%). Sometimes, biceps tenodesis is required as part of the arthroscopic cuff repair. For instance in the repair of subscapularis (both open and arthroscopic), Burkhart 23 advocates biceps tenodesis routinely as an adjunctive procedure when there is evidence of biceps instability. He found that attempts at trying to preserve the biceps by stabilising it in the bicipital groove commonly fail due to redislocation of the biceps causing persistent pain and subsequent disruption of the subscapularis repair. The biceps pathologies most commonly encountered include prerupture (35%), luxation (30%), subluxation (26%) and tenosynovitis (9%). The commonly used methods of biceps tenodesis include:

   

Subpectoral bone tunnel technique. Interference screw technique. Suture anchor technique. Keyhole technique.

A detailed description of the surgical technique of biceps tenodesis is beyond the scope of this article and is covered in detail in other sources.23–25 There have been several biomechanical studies comparing the fixation strengths of various methods of biceps tenodesis. Using fresh sheep shoulders, Ozalay et al.26 compared the fixation strength of four methods of biceps tenodesis. They found that the strongest construct was using the interference screw; this was followed by the tunnel technique, then the anchor technique and lastly the keyhole technique. Using fresh frozen cadavers dissected free of soft tissue, Mazzocca et al.27 compared the cyclic displacement and ultimate failure strength of four methods of biceps tenodesis: the open subpectoral bony tunnel technique, the arthroscopic suture anchor tenodesis, the open subpectoral interference screw fixation technique and the arthroscopic interference screw technique. Using a Materials Testing System, they found that the subpectoral bony tunnel technique was the weakest and showed significantly greater displacement than the other three methods. The difference between the other three methods was not statistically significant.

373 Kilicoglu et al.28 studied two intraosseous techniques (suture sling and tenodesis screw) and one extraosseous technique (2 suture anchors). Using sheep sacrificed at 3, 6 and 9 weeks, the time dependent changes in fixation strengths following biceps tenodesis were evaluated. They found that the load to failure was similar in all three techniques although the interference screw group had a significantly greater increase in fixation strength within the first 3 weeks. In a similar biomechanical study using fresh frozen cadavers, Richards and Burkhart29 compared the fixation strengths of the interference screw with the double suture anchor. They found that the group treated with an interference screw had a significantly greater resistance to pullout than the suture anchor group. In view of the greater fixation strength, the authors recommended the use of the interference screw in biceps tenodesis, thereby permitting early active elbow flexion. Using cadaveric specimens, Jayamoorthy et al.30 evaluated the initial fixation strength of keyhole tenodesis with two types of interference screw fixation: a cannulated metallic interference screw and a bioresorbable one. They found that the keyhole tenodesis technique was significantly stronger than the cannulated metallic interference screw but it was similar in strength to the bioresorbable interference screw. The mode of failure of both interference screws was by tendon slippage at the screw-tendon-bone interface. On the other hand, the keyhole fixation failed by tendon splitting and slippage out of the restraining key hole.

Tenotomy verus tenodesis In comparing the results of arthroscopic biceps tenotomy versus tenodesis, Osbahr et al.31 studied 160 patients with chronic refractive bicipital pain. Half the group was treated with tenotomy and the other half with tenodesis. They found there was no significant difference between the two treatment methods in terms of cosmetic deformity, muscle spasm in biceps and anterior shoulder pain. They concluded that biceps tenotomy may be a reasonable alternative to tenodesis in patients with chronic refractive bicipital pain. One biomechanical study examined the likelihood of distal migration of the biceps stump following simple tenotomy as compared to tenodesis under physiological loading conditions.32 The cyclical load through which the specimen was tested equates approximately to the force produced during gentle active range of motion without resistance. Wolfe et al. reported that 40% of the tenotomised specimens failed during physiological loading as compared to 0% in the tenodesed group. They therefore concluded that in patients who are concerned about the potential cosmetic deformity and associated dysfunction caused by distal migration of the long head of biceps stump, a tenodesis rather than a tenotomy should be carried out. However, it is important to remember that not all cases of ruptured long head of biceps lead to the Popeye sign since the biceps tendon does not always retract distally into the upper arm. There are three hypotheses for this observation.31 Autotenodesis phenomenon—As mentioned earlier, the biceps tendon has a variable cross-sectional area and is

ARTICLE IN PRESS 374

F. Lam, D. Mok

wider proximally near the glenoid than distally as it leaves the bicipital groove. Following an arthroscopic intraarticular biceps tenotomy, the proximal biceps stump becomes entrapped in the bicipital groove and in time, this becomes autotenodesed to the proximal humerus. Chinese finger trap—Being intraarticular and extrasynovial, the biceps tendon is surrounded by a synovial sheath, which consists of two layers. There is an outer parietal and an inner visceral layer. Distal to the level of the transverse humeral ligament, the synovial sheath reflects back on itself. Following biceps tenotomy, the tendon together with the visceral layer of the synovium slide distally but this sliding is soon stopped by the parietal layer akin to the Chinese finger trap. Mesotendon—The biceps tendon receives its blood supply from a mesotendon (vincula tendinum) containing the terminal branch of the anterior humeral circumflex artery. This mesotendon is attached to the biceps tendon through the visceral layer of the tendon sheath. The mesotendon is thought to act like a checkrein to prevent excessive distal migration of the tendon.

Summary and conclusion The functional role of the long head of biceps tendon in the shoulder remains unclear despite numerous cadaveric and electromyographic studies. It is therefore not surprising that the modern management of biceps pathology in the shoulder remains controversial despite major advances in arthroscopic techniques. An algorithm for treatment based on current evidence is proposed and presented in Fig. 4. Until more is known about the functional significance of the long head of biceps tendon in the shoulder, algorithms can only act as a guide to treatment and the final decision must rest with the operating surgeon intraoperatively.

1. Biceps subluxation 2. Tenosynovitis 3. Pre-rupture (> 25% PT tear, atrophic changes, > 25% reduction of tendon width) 4. Failed SAD with biceps +RC tendinitis

elderly >65 female low functional demand large irreparable RC tear

TENOTOMY

Figure 4

young male high athletic demand concerned with cosmesis absence of RC fatty infiltration

TENODESIS

Algorithm for treatment of chronic bicipital pain.

References 1. Itoi E, Kuechle DK, Newman SR, Morrey BF, An KN. Stabilising function of the long head of the biceps in stable and unstable shoulders. J Bone Joint Surg Br 1993;75:546–50. 2. Kumar VP, Satku K, Balasubramaniam P. The role of the long head of biceps brachii in the stabilization of the head of the humerus. Clin Orthop 1989;244:172–5. 3. Lucas DB. Biomechanics of the shoulder joint. Arch Surg 1973; 107:425–32. 4. Warner JP, McMahon PJ. The role of the long head of biceps brachii in superior stability of the glenohumeral joint. J Bone Joint Surg 1995;77A:336–72. 5. Youm T, ElAttrache N, Tibone J, McGarry MH, Lee TQ. Loading the biceps affects the glenohumeral range of motion, translation and kinematics. Presented at the Annual Meeting of the American Academy of Orthopaedic Surgeons, Chicago, March 2006, p.683 of abstract booklet. 6. Levy AS, Kelly BT, Lintner SA, Osbahr DC, Speer KP. Function of the long head of biceps at the shoulder: electromyographic analysis. J Shoulder Elbow Surg 2001;10(3):250–5. 7. Glousman R, Jobe F, Tibone J. Dynamic electromyographic analysis of the throwing shoulder with glenohumeral instability. J Bone Joint Surg Am 1988;70:220–6. 8. Habermeyer P, Kaiser E, Knappe M, Kreusser T, Wiedemann E. Anatomical and electrophysiological evaluation of the stabilising mechanism of the long head of biceps brachii. Unfallchirurg 1987;90:319–29. 9. Berlemann U, Bayley I. Tenodesis of the long head of biceps brachii in the painful shoulder; improving results in the long term. J Shoulder Elbow Surg 1995;4:429–35. 10. Neer CS. Impimgement lesions. Clin Orthop 1972;173:70–7. 11. Neer CS. Anterior acromioplasty for the chronic impingement in the shoulder. J Bone Joint Surg Am 1972;54:41–50. 12. Neviaser TJ, Neviaser RJ, Neviaser JS. The four in one arthroplasty for the painful arc syndrome. Clin Orthop 1982; 163:107–12. 13. Walch G, Edwards TB, Boulahia A, Nove-Josserand L, Neyton L, Szabo I. Arthroscopic tenotomy of the long head of biceps in the treatment of rotator cuff tears: clinical and radiographic results of 307 cases. J Shoulder Elbow Surg 2005;14:238–46. 14. Murthi AM, Vosburgh CL, Neviaser TJ. The incidence of pathologic changes of the long head of the biceps tendon. J Shoulder Elbow Surg 2000;9(5):382–5. 15. Mariani EM, Cofield RH, Askew LJ, Li G, Chaos EYS. Rupture of the tendon of the long head of the biceps brachii. Surgical versus non surgical treatment. Clin Orthop 1988;228:233–9. 16. Gill TJ, McIrvin E, Mair SD, Hawkins RJ. Results of biceps tenotomy for treatment of pathology of the long head of the biceps brachii. J Shoulder Elbow Surg 2001;10:247–9. 17. Kelly AM, Drakos MC, Fealy S, Taylor SA, O’Brien SJ. Arthroscopic release of the long head of the biceps tendon: functional outcome and clinical results. Am J Sports Med 2005;33(2): 208–13. 18. Maynou C, Mehdi N, Cassagnaud X, Audebert S, Mestdagh H. Clinical results of arthroscopic tenotomy of the long head of the biceps brachii in full thickness tears of the rotator cuff without repair: 40 cases. Rev Chir Orthop Reparatice Appar Mot 2005; 91:300–6. 19. Klinger HM, Spahn G, Baums MH, Steckel H. Arthroscopic debridement of irreparable massive rotator cuff tears—a comparison of debridement alone and combined procedure with biceps tenotomy. Acta Chir Belg 2005;105:297–301. 20. Ahmad CS, ElAttrache NS. Arthroscopic biceps tenodesis. Orthop Clin North Am 2003;34(4):499–506. 21. Sethi N, Wright R, Yamaguchi K. Disorders of the long head of biceps tendon. J Shoulder Elbow Surg 1999;8:644–54.

ARTICLE IN PRESS Treatment of the painful biceps tendon 22. Boileau P, Krishnan SG, Coste JS, Walch G. Arthroscopic biceps tenodesis: a new technique using bioabsorbable interference screw fixation. Arthroscopy 2002;18(9):1002–12. 23. Lo I, Burkhart S. Arthroscopic biceps tenodesis—indications and technique. Arthroscopy: J Arthorosco Relat Surg 2002;10(2): 105–12. 24. Klepps S, Hazrati Y, Flatow E. Arthroscopic biceps tenodesis. Arthroscopy: J Arthorosco Relat Surg 2002;18(9):1040–5. 25. Gartsman GM, Hammerman SM. Arthroscopic biceps tenodesis—operative technique. Arthroscopy: J Arthorosco Relat Surg 2000;16(5):550–2. 26. Ozalay M, Akpinar S, Karaeminogullari O, Balcil C, Tasci A, Tandogan RN, et al. Mechanical strength of four different biceps tenodesis techniques. Arthroscopy 2005;21(8):992–8. 27. Mazzocca AD, Bico J, Santangelo S, Romeo AA, Arciero RA. The biomechanical evaluation of four fixation techniques for proximal biceps tenodesis. Arthroscopy 2005;21(11):1296–306.

375 28. Kilicoglu O, Koyuncu O, Demirhan M, Esenyel CZ, Atalar AC, Ozsoy S, et al. Time dependent changes in failure loads of 3 biceps tenodesis techniques: in vivo study in a sheep model. Am J Sports Med 2005;33(10):1536–44. 29. Richards DP, Burkhart SS. A biomechanical analysis of two biceps tenodesis fixation techniques. Arthroscopy 2005;21(7): 861–6. 30. Jayamoorthy T, Field JR, Costi JJ, Martin DK, Stanley RM, Hearn TC. Biceps tenodesis: a biomechanical study of fixation methods. J Shoulder Elbow Surg 2004;13(2):160–4. 31. Osbahr DC, Diamond AB, Speer KP. The cosmetic deformity of the biceps muscle after long head tenotomy versus tenodesis. Arthroscopy: J Arthorosco Relat Surg 2002;18(5): 483–7. 32. Wolfe RS, Zheng N, Weichel D. Long head of biceps tenotomy versus tenodesis: a cadaveric biomechanical analysis. Arthroscopy 2005;21:182–5.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 376–385

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

KNEE

Periprosthetic fractures above a total knee arthroplasty—A review of best practice Graham Walsha,d,, Sudhindran Ankarathb, Peter V. Giannoudisc a

Department of Orthopaedics, Pinderfields Hospitals, Aberford Road, Wakefield WF1 4EE, UK Department of Orthopaedics, Huddersfield Royal Infirmary, Acre Street, Huddersfield, West Yorkshire HD3 3EA, UK c Department of Trauma and Orthopaedic Surgery, St. James’s University Hospital, Beckett Street, Leeds LS9 7TF, UK d The Stables, Church Lane, Southowram HX3 9TD, UK b

KEYWORDS Periprosthetic; Fracture; Total knee arthroplasty; Total knee replacement; Femoral

Summary Periprosthetic fractures around knee replacements remain one of the most challenging problems that face the modern day orthopaedic surgeon. The incidence is on the rise with both an increasing elderly population and the increased use of prosthetic implants. This review looks at the options available to treat these fractures and, based on published results, presents an algorithm as a guide for the management of periprosthetic fractures of the femur above total knee arthroplasties. & 2006 Elsevier Ltd. All rights reserved.

Introduction Periprosthetic fractures around knee replacements remain one of the most challenging problems that face the modern day orthopaedic surgeon. The incidence is on the rise with both an increasing elderly population and the increased use of prosthetic implants. Currently myriad methods are available to treat these fractures aiming to restore both the biological and mechanical environment to allow optimal fracture healing. This review looks at the available literature and offers an algorithm suggesting the current

Corresponding author. The Stables, Church Lane, Southowram

HX3 9TD, UK. E-mail addresses: [email protected] (G. Walsh), [email protected] (S. Ankarath), [email protected] (P.V. Giannoudis). 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.05.004

best treatment for periprosthetic fractures of the femur above total knee arthroplasties (TKAs).

Incidence and aetiology The elderly population is increasing. It has been estimated that within the next 50 years the population of those over the age of 65 years will increase five-fold.1,2 In tandem with this the indications for joint replacements are also broadening, inevitably leading to an increased number of patients with joint arthroplasties3–5 and periprosthetic fractures. The incidence of periprosthetic femoral fractures above TKAs has a reported incidence between 0.3% and 2.5%.6–10 One of the biggest published series to date following patients after primary TKAs was by Merkel and Johnson11 looking at 4539 patients. They showed a 0.6% incidence of periprosthetic fractures following primary total knee

ARTICLE IN PRESS Periprosthetic fractures above a total knee arthroplasty

Table 1

Risk factors for periprosthetic fractures.

Patient factors

Local factors

Rheumatoid arthritis Osteoporosis Steroid use Neurological disorders Smoking Immunosuppression Female sex Frequent falls

Osteolysis Anterior femoral notching Stress risers Loosening

arthroplasty. Within the same review 637 patients underwent revision surgery and an increased fracture incidence of 1.6%, nearly three times as much, was noted. Most periprosthetic fractures that occur in the femur do so as a result of only low energy trauma, be it a fall or spontaneously. In most series considered when compiling this review, most fractures occurred as a result of minor trauma, such as a fall. The risk factors for these fractures can be divided into two broad groups: those related to the patient and those related to the prosthesis (Table 1). Patient factors include disorders leading to osteopenic bone, such as rheumatoid arthritis,10,12,13 steroid use10 and osteoporosis.5,6,8–10,12 Other patient factors include neurological disorders, immunosuppression, smoking, female sex, frequent falls in the elderly6,8,10,12 and previous fracture.14 Factors local to the prosthesis include osteolysis associated with particulate debris from bearing surfaces or loosening,15 and stress risers from previous surgery.16 Historically, the most commonly reported local factor responsible for an increased risk of a periprosthetic fracture above a TKA is anterior femoral notching. The risk of fracture was initially thought to be due to a decrease in bending and torsional strength associated with notching.17,18 These results however, were based on mathematical and biomechanical studies but in clinical practice little evidence is available to support this theory. Ritter et al.19 reviewed a total of 670 TKAs, 180 of these having had some degree of femoral notching. They reported that only two of these developed a periprosthetic fracture. They concluded that anterior femoral notching was of minimal concern in fracture risk beyond the first 6 months postoperatively.

377

Table 2 tures.

Rorabeck classification of periprosthetic frac-

Type

Description

I II

Undisplaced fracture with an intact prosthesis Displaced fracture with an intact prosthesis (Fig. 1) Fractures associated with a loose or failing prosthesis (Fig. 2)

III

Figure 1

Type II Fracture associated with a stable prosthesis.

drawback of this classification system is that it does not differentiate between infected or aseptic loosening, which is of paramount importance in the management of these fractures.

Treatment of fractures above TKAs Classification In order for a fracture classification to be useful it must suggest treatment, estimate prognosis, give an indication of likely outcome12 and also allow comparison of results by different centres.20 For fractures around TKAs a number of systems have been described,6–9,11 However only that proposed by Lewis and Rorabeck21 appears to have achieved universal acceptance (Table 2). The Rorabeck classification (Fig. 1), as it is known, offers a good basis for treatment. Type I fractures are undisplaced fractures with an intact prosthesis, Type II are displaced fractures around an intact prosthesis and Type III are those associated with a loose or failing prosthesis (Fig. 2). One

The ideal outcome after fixation of these fractures, as with most fractures, would be union, ability to weight bear and restoration of the range of movement around the joint. Historically a good outcome for the patient is one which results in a minimum of 901 of knee flexion or restoration of preoperative motion, shortening of less than 2 cm, a varus/ valgus mal-alignment of less than 51 and flexion/extension mal-alignment of less than 101.21,22 For the purpose of this review results reported in the literature will be compared to these ‘standards’ and if they are achieved the outcome will be seen as a success. Fundamental to achieving a successful outcome is good preoperative planning. This must take into account the type

ARTICLE IN PRESS 378

Figure 2

G. Walsh et al.

Type III Fracture associated with a loose prosthesis.

of fracture, degree of displacement, type of TKA in-situ, presence of infection or loosening of the prosthesis and patient factors such as pre-fracture morbidity, mobility and bone quality.9,10,23,24 It has been demonstrated that fracture displacement, comminution and intra-condylar extension are all associated with the poorest postoperative outcomes.9 A good history and examination are paramount. It is important to elucidate the status of the implant prefracture. A careful history will often give clues as to whether the implant was initially loose or stable. If loosening is suspected then infection must be ruled out. If infection is present early, defined as being within 3 months of knee surgery, the diagnosis is often obvious with an acutely painful, erythematous joint with an effusion prior to fracture. The patient may also have had systemic signs of infection. Delayed infections occur within 3–24 months after surgery and are often hard to distinguish from aseptic loosening. The patient will have had chronic pain associated with a loose prosthesis. The late infected joints, diagnosed after 24 months, are a result of haematogenous seeding and often start insidiously. Diagnosis of late infection prior to surgery may be difficult but is essential for a successful outcome. Blood tests include C-reactive protein (CRP) and leukocyte count.25–27 Aspiration of the joint can be useful, and gram staining synovial fluid has been shown to have a high specificity, over 97%, although there is a low sensitivity of less than 26%.28 Culture of the synovial fluid has a much better sensitivity of between 65% and 95%,28 but requires a longer period and may delay surgery. Work carried out by the Nuffield Orthopaedic Centre in Oxford attempted to improve preoperative diagnosis of prosthetic joint infection.29 They prospectively looked at 297 patients undergoing revision surgery of which 41 were found to be infected. Samples were taken at the time of surgery. They found that only 65% of the samples taken from the infected patients were culture positive. They demon-

strated that isolation of an indistinguishable micro-organism from three or more specimens was highly predictive of infection, with a sensitivity of 65% and specificity of 99.6%. They recommended that five or six culture samples should be sent at the time of surgery, with a definitive diagnosis of infection being made if three or more specimens yielded an indistinguishable organism. They also showed poor results with gram staining, with a sensitivity of 12% and specificity of 98% and suggested its use should be abandoned. More recently, work has looked at molecular based techniques using the polymerase chain reaction (PCR) to diagnose infections around prosthetic joints. The main advantage of this technique over culture-based methods is that diagnosis can be made with only a very small amount of genetic material in a retrieved specimen. This method is currently in its infancy and the results are varied, with sensitivities ranging from 100% to 71% and specificity ranging from 100% to 49%.30 Work is ongoing assessing the validity of this technique in diagnosing infection. Good quality radiographs are essential to classify the fracture pattern, assess the quality of the bone stock and also determine loosening. Further imaging modalities are often un-necessary unless a pathological fracture is suspected. It is useful to obtain previous radiographs for comparison. The type of the original implant and insertion technique are also important factors which must be considered.

Conservative treatment The conservative approach to treating periprosthetic supracondylar knee fractures has fallen from favour in recent years due to the poor results reported in the literature. The conservative approach includes casting, skeletal traction and bracing.11,13,18,24,31,32 If the conservative route is taken, treatment involves a long period of bed rest-a practice which is associated with an increase in morbidity and mortality.33 It also restricts early mobilisation and movement, inevitably leading to a loss of motion and ambulatory capacity.9,11,18,34 The conservative approach was first described in the literature by Short et al.31 who treated four fractures with traction and plaster immobilisation. Two of the four patients suffered complications, with both requiring open reduction and internal fixation (ORIF) for non-union. They concluded that ORIF was necessary to allow early movement and mobilisation. Merkel and Johnson11 published the results of 26 fractures which were treated conservatively. They reported success in 17 (65%) with non-union being the main complication in the unsuccessful group. Their study also reviewed a further 10 fractures treated operatively and although they showed a better outcome in this group they found an unacceptable high complication rate of 100%. They therefore concluded that the conservative approach was more favourable than operative management. Cain et al.24 also reported success with conservative treatment. They looked at the treatment of 14 fractures, 10 of which were conservatively managed. They reported a 70% success rate in the conservative group, loss of function and non-union being the main cause of failure. A further two

ARTICLE IN PRESS Periprosthetic fractures above a total knee arthroplasty fractures treated conservatively were reported by Cordeiro et al.32 both of which went on to achieve a good result. Figgie et al.13 published the results of their series of 24 periprosthetic fractures. They treated 10 fractures conservatively and reported a good result in nine of these. In this study only un-displaced fractures with good bone stock were treated conservatively. The largest series to date was published by Culp et al.18 looking at 61 fractures, of which 30 were treated conservatively. They reported a successful outcome in 17 (57%) of those fractures treated conservatively. It was however noted that casting was associated with a significant average loss of motion of 261 and traction with a 121 loss of motion. They also showed patients treated conservatively had a significant increase in postoperative pain when compared to their operative treatment group. They concluded that ORIF was the best form of treatment. From the available literature a total of 82 fractures treated conservatively have been reported. Overall a successful outcome was deemed to have been achieved in 54 patients, a 66% success rate. The main reasons for failure were stiffness and non-union. However, it must be stressed that in a number of these reports accurate data on postoperative alignment and movement were not recorded. It is likely the real success of conservative management is much worse than that reported. The literature only recommends this form of treatment in fractures that were undisplaced and above an area of good bone stock.

Revision surgery Revision surgery should be considered in any periprosthetic fracture associated with a prosthesis that is loose and deemed to be failing prior to injury. Revision surgery should also be considered in those patients who have had failed attempts at fixation of their fracture. Revision surgery allows early mobilization and stable fixation.9,13,35 However, it is technically demanding with osteopenia, osteolysis, bony defects, marked comminution and possible infection all contributing to make revision surgery a challenge for even the most experienced knee surgeon. The use of revision knee surgery as a primary treatment for periprosthetic fractures around a TKA is poorly described in the literature. Cordeiro et al.32 reported on four fractures treated with a femoral long stem revision. Full weight bearing was achieved on average by 14 days. All patients went on to union and achieved either 901 or their preoperative degree of flexion with no varus or valgus deformities. No other complications were described. Srinivasan et al.36 used a long stem cemented revision (Fig. 3) to treat six patients with periprosthetic fractures. They reported union in all cases with two complications, one wound dehiscence requiring plastic surgical cover and one posterior knee dislocation requiring a cylinder cast. An average valgus deformity of 61 was reported. Mean flexion was 65% with only one patient achieving 901. They concluded that the use of the long stem revision should be considered when there is implant loosening, gross comminution and poor bone stock. The use of a structural allograft with the revision prosthesis has also been reported. The allograft allows the

379

Figure 3

The use of a long-stem revision.

comminuted often osteopenic distal part of the femur to be removed and allows a better platform to be introduced for the revision.6,37,38 They can be customised to fit the femur and because they mechanically resemble the host femur they are thought to reduce stress shielding.37 Wong and Gross38 used structural allografts on a series of five fractures. They cemented the allograft to the femoral implant and then inserted the construct into the host femur after splitting it longitudinally from anterior to posterior. One patient required further revision due to instability and one patient was described as having an asymptomatic nonunion. Success was achieved in three fractures. Recently Kassab et al.37 published data on 10 fractures that were treated with a distal femoral allograft. Again they used a long stem revision with the allograft cemented to it prior to insertion. Range of motion exercises were commenced 48 h postoperatively. Three patients required further surgery, with one requiring amputation for deep sepsis, one requiring revision and the other needing soft tissue stabilisation. The average degree of flexion achieved was 981, with three patients not achieving either 901 or their preoperative range of movement. Kray et al.39 used a revision prosthesis with structural allograft in seven acute fractures. Apart from one death unrelated to surgery, they achieved union in all fractures. An average range of motion of 961 was achieved. Freedman et al.40 used revision surgery without allograft on five patients. They reported four good results and one infection. An average range of motion of 991 of flexion was achieved. Overall the results from revision surgery are good. Its use is recommended in fractures associated with a loose prosthesis, and combined with the use of structural allograft it gives a good result in very distal fractures where

ARTICLE IN PRESS 380 significant bone loss would be expected. The revision technique however is technically demanding and requires the skills of a surgeon experienced in treating this type of fracture.

LISS plate The less invasive stabilisation system (LISS) is currently increasing in popularity as a method for stabilizing supracondylar femur fractures.41–45 The use of the LISS plate relies on the concepts of both indirect reduction and minimally invasive plate osteosynthesis (MIPO) in order to preserve bone biology.45 It acts as a so called ‘internal fixator’41,43,44 whereby the screws (pins) are locked into the plate (frame).46 Forces are then transferred across the screws, preventing compression between the plate and the bone, thus preserving the blood supply to the bone beneath the plate.47 The plate can also be inserted percutaneously, further minimising the disruption of the fracture’s soft tissue envelope.48 It is also suggested that the failure mechanics of the plate are changed. Traditionally plates fail one screw at a time. It is hypothesised that the LISS plate, if it is to fail, must do so as a unit with all the screws pulling out together. This is thought to add structural support, which is of particular importance in osteoporotic bone.23,49 The use of the LISS to treat fractures above an implant is an obvious choice. It has a theoretical advantage in osteoporotic bone and its minimally invasive insertion technique make it appear an ideal solution. Currently the experience of the LISS is at an early stage and data on the clinical results are still sparse. The biggest published series to date was by Kregor et al.50 A series of 13 periprosthetic fractures were treated with the LISS. The average operative time was 140 min. Postoperatively range of motion exercises were begun immediately, without the use of a brace, and progressive weight bearing occurred with full weight bearing achieved by 13 weeks. The review highlighted two complications. One fixation failed in a fracture associated with a loose femoral component; this patient required revision surgery. The other complication was a delayed union and this patient required bone grafting to promote union. No varus or valgus collapse occurred and range of movement averaged 2–901. Ultimately a 92% success rate was achieved. Althausen et al.51 reported on five periprosthetic fractures treated with the LISS. Their operative time averaged 135 min, with a blood loss of only 180 ml. Postoperatively, range of motion was begun immediately. By the end of follow up patients achieved an average range of movement of 0–911 with an average valgus deformity of 51, with no shortening. All fractures went on to union and all patients returned to their pre-injury mobility. In a further series, nine periprosthetic fractures were treated with the LISS by Wick et al.52 They reported very good results with only one patient reported to have a failed fixation, requiring revision for infection. From the small number of cases in the literature the LISS has proved to be a successful implant in the treatment of periprosthetic fractures around a total knee replacement. Good results were achieved in 93% of patients with a low operative morbidity. A word of caution must be offered

G. Walsh et al. however, as this is a new technique which has proved to be technically demanding, it is recommended that the use of the LISS is reserved for use by surgeons experienced in the technique.

Intramedullary nailing The use of an intramedullary device is an attractive option for the treatment of these fractures. Using the intramedullary route is advantageous as it is a minimally invasive technique with minimal soft tissue disruption.34,53–55 It also affords early knee mobilisation.22,54–57 Nailing is also attractive as gives a better fixation in osteoporotic bone and also provides good axial, angular and rotational stability.9,22,54,55,58 Many nailing options are available including the retrograde supracondylar nail, flexible nail and anterograde femoral nail (Fig. 4).

The retrograde nail The retrograde supracondylar nail is inserted in a retrograde fashion via a medial parapatellar incision (Fig. 5).54–57 Important to the success of nail insertion is the design of the intercondylar notch of the knee prosthesis. Most studies recommend reserving the retrograde nailing technique for fractures above cruciate retaining prostheses,7,22,54 these prostheses usually have notch diameters ranging from 11 to 20 mm10 and will therefore facilitate the passage of the nail. Cruciate substitution components often have much narrower notches10 or no notch and may therefore prevent the passage of the nail. In these situations an alternative fixation method is recommended.7,10,54 In those knees

Figure 4 A Type II fracture treated with an anterograde femoral nail.

ARTICLE IN PRESS Periprosthetic fractures above a total knee arthroplasty

381 fractures. Again results were all reported as good with union by 3 months, a range of movement of at least 0–901 and no varus/valgus deformities. Overall, from the literature available, the retrograde nail seems to be a very good fixation method for supracondylar fractures above TKAs. A total of 44 fractures have been reported with only two poor results described. This gives an overall success rate of 95%. The only hinderance to this method of fixation appears to be when the fracture is above a cruciate substitution prosthesis which will not allow the passage of the nail.

Flexible nailing

Figure 5 The use of a supracondylar nail.

where the box is too small, a technique of widening the notch with a high speed carbide drill has been described.54 With this technique it is possible that metal fragments may be retained within the knee joint from the drilling process and these fragments may stimulate osteolysis and implant failure. This may be both directly, due to macrophage activation, or indirectly by accelerating polyethelene debris generation by a third body wear mechanism. The use of the retrograde nail for periprosthetic fractures was first described by McLaren and colleagues.59 They published a series of seven fractures treated with a supracondylar rod. All patients achieved a good result and returned to a pre-fracture level of mobility by 3 months. In 1995 two further groups published similar results in the same journal. Murrell and Nunley60 described four fractures and Jabczenski and Crawford58 a further four fractures treated with the supracondylar nail. Both these studies reported good outcomes with no complications. Smith et al.56 described a single fracture treated with a retrograde nail but reported a poor result with a range of movement of only 7–75156 after 3 months. Althausen et al.51 also reported on a single case treated with a retrograde nail. Their results described a good postoperative result with a range of movement of 0–1101 with varus of 31. The retrograde nail was used by Bezwada et al.61 to treat a series of 18 fractures. A good outcome was reported with an average range of movement of 5–1001 achieved and an average valgus of 41. Surgical time was relatively speedy at a mean of 45 min. Only one complication occurred; a deep infection in a diabetic requiring amputation. All other fractures united by 10 weeks. The Huckstep intramedullary retrograde nail was used by Chen et al.54 in their series of two patients. Both patients had a good result with union by 3 months. Range of movement was 0–951 with no varus/valgus deformities. In this series patients were immobilized in plaster for 6 weeks postoperatively. Weber et al.57 reported on a further seven

The use of flexible nail has also been advocated in the treatment of these fractures as it is both minimally invasive and relatively inexpensive. The use of the Rush nail for periprosthetic fracture fixation was first described by Cain et al.24 They looked at 14 fractures, three of which were treated with Rush nails. Of these three fractures, two failed, with one non-union and one mal-union. In this report percutaneous flexible nailing was not recommended. Ritter et al.62 reviewed 22 periprosthetic fractures treated with Rush nails. Union occurred in all fractures by 4 months. All patients achieved flexion over 901, with an average of 1081. Valgus averaged 101 with two cases in 151 of valgus. The authors felt their results were good but on review the degree of valgus postoperatively does not support this conclusion. The Rush nail was also used in the series reported by Althausen et al.57 They treated four fractures. Their procedures had a quick mean operative time of 68 min and average blood loss of only 95 ml. Overall clinical results were poor with an average range of movement 0–681 postoperatively. All fractures collapsed into a valgus deformity with an average of 101, with two fractures exhibiting a flexion deformity and the other two an extension deformity. Overall the results were poor. The use of flexible intramedullary nails to treat periprosthetic fractures around TKAs does not appear to be a suitable option from the literature available. It is understandable why this method of fixation would be an attractive option as it is percutaneous with minimal soft tissue disruption and short operative times. However in practice this appears to be a biomechanically unstable method of internal fixation.

Open reduction and internal fixation The use of ORIF to treat supracondylar fractures above TKRs has been extensively described. ORIF allows visualisation of the fracture to allow a more accurate reduction and provides a good degree of coronal stability to prevent collapse. Many devices have been utilised including condylar plates, fork plates and conventional plates and screws. Fixed angle devices such as the condylar plate and the fork plate have proved popular. The use of these devices relies on adequate distal bone stock to allow hold in the distal fragment. Unfortunately, due to the osteoporotic nature of these fractures the distal bone stock is often poor. The condylar buttress plate was used by Zehntner and

ARTICLE IN PRESS 382 Ganz63 in six fractures. Union was achieved by 14 weeks in all fractures with an average valgus of 51 and average flexion of 971. The results were good in all patients. The fork plate was used by Ochsner and Pfister64 in a series of six fractures. It was suggested that this device would be superior to the condylar plate as it allows a more distal fixation in qualitatively poor bone, providing better anchorage.64 Immediately postoperatively patients were made to partial weight bear and active range of motion exercises were begun. One patient died at 9 weeks, all other fractures united by 4 months. No valgus/varus deformities were found. From these results it would appear the fork plate achieves a biologically stable result. Althausen et al.51 treated two fractures with a plate. Their average operative time was a relatively lengthy 160 min with a large blood loss of 1350 ml. The results were unsatisfactory with an average postoperative flexion of only 571, with both varus and valgus collapse described. Cain et al.24 also reported a poor result. In their only case treated with a plate, the fracture went on to non-union. Cordeiro et al.32 treated three fractures with ORIF. Two patients required revision surgery when the fractures collapsed into varus. Only one patient had a good outcome. Bezwada et al.61 used the periarticular distal femoral plate plus cement on 12 fractures. Operative time averaged 74 min with an average blood loss of 450 ml. All but one fracture went on to union, with an average valgus of 51. Four patients had a deformity greater than 51. Range of movement averaged 5–951. In this series only 58% of the results could be described as good. ORIF was used by Figgie et al.13 in 10 fractures. Their series showed poor results with only a 50% success rate, with union averaging 4 months. In a further series by Moran et al.65 15 fractures were treated with a variety of internal methods including condylar plates, blade plates and conventional plates. Of these fractures the results of 10 were described as good; a 66% success rate. The use of the semi-rigid carbon fiber plate has been described by Al-Shawi and colleagues66 who used the implant to fix five fractures above knee replacements. In their series one patient died postoperatively of a pulmonary embolus and all the other fractures went on to union, although this took up to 6 months, with an average of 16 weeks. A cast brace was used postoperatively. The resultant average range of movement was 0–851, with only two patients achieving 901 of movement. There was one malunion. The theoretical advantage of these plates is that the lower elastic modulus increases the stress sharing between plate and bone therefore enhancing callus formation and limiting osteopenia from stress shielding, a potential advantage in osteoporotic bone.66 In practice a good result was only achieved in two patients, with a long time to union. Overall 64 cases have been described in the literature treated with ORIF. A good result was only achieved in 56% of patients. There are disadvantages in the treatment of periprosthetic fractures with ORIF, one of which is the degree of soft tissue damage that is incurred by the exposure, compromising subsequent healing with further devitalisation of the fracture fragments. Another disadvantage is that often the prosthesis is exposed, leading to a theoretical risk of infection and subsequent loosening of the

G. Walsh et al. prosthesis. In this review however, no episodes of infection were documented, but follow-up was relatively short. The biomechanical stability of the fixation has also to be questioned. The majority of these fractures occur in osetopaenic bone with a small distal fragment. The use of plates and screws relies on a good cortical hold to offer stability. However in poor bone stock this cannot reliably be achieved, restricting early mobilisation.

Other methods The use of external fixators to treat these fractures has seldom been described in the literature. Figgie et al. 13 used a monolateral fixator on one fracture in their series. This resulted in failure with a pin site infection leading to deep sepsis, necessitating revision surgery. The Ilizarov fixator has also been used in one reported case. Simon and Brinker67 used this method of fixation with a successful result. The patient was able to fully weight bear on the first postoperative day with union occurring at 10 weeks, at which point the frame was removed. No complications were described. Their range of movement achieved was 0–1101, with no valgus/varus deformity.

Discussion The treatment of periprosthetic fractures above TKAs has never been as important as it has become in recent years, and will become in years to come. With increases in the elderly population and in the range and indications for joint replacement surgery the occurrence of periprosthetic fractures can only increase. It is therefore inevitable that the majority of orthopaedic surgeons will, at some point in their career, face the challenge of fixing these fractures. With this borne in mind it is important that a management plan is in place that reflects best practice to aid in the decision making process of treating these difficult fractures. When treating these fractures an ideal outcome is one which achieves early mobilisation and movement leading to early rehabilitation, reducing morbidity and mortality. The earliest reports in the literature recommended conservative treatment if the fracture was un-displaced. Results appeared promising, with an overall 66% success rate. However, these results should be viewed with a degree of scepticism, as the data recorded often gave little information on postoperative alignment and range of movement. Studies that appeared in the literature later, gave a more thorough description of postoperative results, and suggested poorer outcomes from conservative management. The conservative approach to treatment often requires a prolonged period of immobilisation. This has an adverse effect on patient recovery. The only indication for conservative treatment should be in patients whose medical comorbidities preclude them from surgery. Most fractures described in the literature are either type I or II and indeed these two groups could be grouped together as the treatment modalities are the same, both requiring fixation. Type III fractures are a much more challenging entity. Infection has to be ruled out as a cause of loosening, and this involves checking blood inflammatory markers and joint

ARTICLE IN PRESS Periprosthetic fractures above a total knee arthroplasty aspiration or possibly multiple biopsy and culture. If infection is present then it must be cleared prior to definitive treatment. This may involve removing the prosthesis and carrying out a 1 or 2 stage revision. Revision can be with either a mega-prosthesis or revision prosthesis with allograft. The decision on which method to use is dependent on assessment of the bone quality and amount of residual bone stock. In those fractures where infection has been ruled out as the cause of loosening then management becomes more straight-forward, but by no means easy. Revision surgery has a good success rate, 68% in this review, and should be carefully planned. If the bone stock is poor then either adjunctive allograft or a mega-prosthesis should be used for fixation. This method of surgery is technically demanding and should only be undertaken by a specialist. Fractures associated with a stable prosthesis, either type I or type II, require operative fixation for best results. The literature available demonstrates a better outcome with operative treatment, with a success rate of 72% compared to 66% with the conservative approach. Treatment success is very much dependent on the quality of the bone and the degree of exposure of the fracture. The use of ORIF, by whatever means, was associated with a high failure rate and a successful outcome in only 56% of cases. Failure was probably due to two main reasons, paramount being the quality of the bone stock fixation relied upon. Often in these fractures the bone is markedly osteopenic and therefore the screw hold within the bone is very poor and prone to cut out

383 and failure. The second reason for failure is related to rehabilitation. Inevitably ORIF requires a period of immobilisation to protect the fixation. However, in these fractures this can greatly affect recovery and lead to stiffness and reduced range of movement. Overall, from the data considered in this review the retrograde femoral nail seems to demonstrate the best results, with an overall success of 95%. The advantage of the nail is that it is relatively quick to insert with minimal blood loss, it allows early mobilisation and therefore is associated with a good postoperative range of movement. The disadvantages of the nail include difficulty with the entry point of the nail relating to the design of the original prosthesis. With a cruciate substituting knee prosthesis the notch may be too small to allow passage of the nail and is so then another method must be employed. Locking the nail can also be problematic and relates to the bone stock distally. For the technique to be successful fixation requires sufficient bone substrate distally in the femur to allow the nail to be locked. The more distal the fracture and the poorer the bone quality, the less feasible it may be to carry out distal locking and therefore another fixation technique may need to be sought. Infection is another possible disadvantage. Nail insertion requires opening up the knee joint. By doing so there is a theoretical risk of introducing infection around the knee prosthesis, leading to later failure. No reports of infection were reported in the literature. However, the follow-up times were relatively short. Another possible disadvantage

Figure 6 Algorithm for treatment of femoral periprosthetic fractures around a total knee arthroplasty.

ARTICLE IN PRESS 384 is related to the nail abutting the prosthesis, leading to metalosis and possible loosening of the prosthesis. Again, this was not reported in the literature but may occur much longer after fixation. Overall, from the results available in the literature the retrograde femoral nail appears to be one of the better methods available for the fixation for periprosthetic femoral fractures around a TKA. The LISS plate has also been shown to be a successful implant for fixation of periprosthetic fractures. If the knee prosthesis will not allow passage of a nail then the LISS is a very useful alternative to the retrograde nail. It is also has advantages in bone of poor quality where locking the retrograde nail would not be possible. The LISS does not rely on cortical bone strength and therefore failure is less likely in the osteoporotic bone. The disadvantage of the LISS is that it requires a certain degree of expertise and therefore any fracture treated with this system must be carried out by a surgeon competent in its use.

Conclusion Overall the treatment of periprosthetic fractures above a total knee arthroplasty is a complex challenge that has become more commonplace in recent years. With the forecast increased incidence of these fractures a strategy must be in place to manage these as effectively as possible. We have therefore come up with an algorithm to ensure best practice (Fig. 6).

References 1. Cooper C, Campion G, Melton J. Hip fractures in the elderly: a worldwide projection. Osteoporosis Int 1992;2:285–9. 2. WHO Study Group. Assessment of fracture risk and its application to screening for postmenapausal osteoporosis. WHO Technical Reports Series No. 843; 1994. 3. Garbuz DS, Marsi BA, Duncan CP. Periprosthetic fractures of the femur: principles of prevention and management. Instruct Course Lect 1998;47:237–42. 4. Lewallen DG, Berry DJ. Preiprosthetic fracture of the femur after total hip arthroplasty: Treatment and results to date. Instruct Course Lect 1998;47:243–9. 5. Duncan CP, Masri BA. Fractures of the femur after hip replacement. Instruct Course Lect 1995;44:293–304. 6. Engh GA, Ammeen DJ. Periprosthetic fractures adjacent to total knee implants: treatment and clinical results. Instruct Course Lect 1998;47:437–48. 7. Haidukewych GJ, Jacofsky DJ, Hanssen AD. Treatment of periprosthetic fractures around a total knee arthroplasty. J Knee Surg 2003;16:111–7. 8. Ayers DC, Dennis DA, Johanson NA, Pelligrini VD. Common complications of total knee arthroplasty. J Bone Joint Surg Am 1997;79:278–311. 9. Dennis DA. Periprosthetic fractures following total knee arthroplasty: the good, bad & ugly. Orthopaedics 1998;21: 1048–50. 10. Su ET, DeWal H, Di Cesare PE. Periprosthetic femoral fractures above total knee replacements. J Am Acad Orthop Surg 2004; 12:12–20. 11. Merkel KD, Johnson EW. Supracondylar fracture of the femur after total knee arthroplasty. J Bone Joint Surg 1986;68:29–43. 12. Schmidt AH, Kyle F. Periprosthetic fractures of the femur. Orthop Clin North Am 2002;33:143–52.

G. Walsh et al. 13. Figgie MP, Goldberg VN, Figgie HE, Sobel M. The results of treatment of supracondylar fracture above total knee arthroplasty. J Arthroplasty 1990;5:267–76. 14. Sarvilinna R, Huhtala HAS, Puolakka TJS, Nevalainen JK, Pajamaki KJ. Periprosthetic fractures in total hip arthroplasty: an epidemiological study. Int Orthop 2003;27: 359–61. 15. Berry DJ. Periprosthetic fractures associated with osteolysis: a problem on the rise. J Arthroplasty 2003;18:107–11. 16. Tsiridis E, Haddad FS, Gie GA. The management of periprosthetic fractures around hip replacements. Injury 2003;34: 95–105. 17. Lesh ML, Schneider DJ, Deol G, Davis B, Jacobs CR, Pellegrini VD. The consequences of anterior femoral notching in total knee arthroplasty: a biomechanical study. J Bone Joint Surg Am 2000;82:1096–101. 18. Culp RW, Schmidt RG, Hanks G. Supracondylar fracture of the femur following prosthetic knee arthroplasty. Clin Orthop 1987;222:212–22. 19. Ritter MA, Faris PM, Keating EM. Anterior femoral notching and ipsilateral supracondylar femur fracture in total knee arthroplasty. J Arthroplasty 1988;3:185–7. 20. Sung-Rak L, Bostrom MPG. Periprosthetic fractures of the femur after total hip arthroplasty. AAOS Instruct Course Lect 2004;53:111–8. 21. Lewis PL, Rorabeck CH. Periprosthetic fractures. In: Engh GA, Rorabeck CH, editors. Techniques of revision surgery. Philadelphia, PA: Williams & Wilkins; 1997. p. 275–95. 22. Rorabeck CH, Taylor JW. Periprosthetic fractures of the femur complicating total knee arthroplasty. Orthop Clin North Am 1999;30:265–77. 23. Bong MR, Egol KA, Koval KJ, Kummer FJ, Su ET, Iesaka K, et al. Comparison of the LISS and retrograde-inserted supracondylar intramedullary nail for fixation of a Periprosthetic distal femur fracture proximal to a total knee arthroplasty. J Arthroplasty 2002;17:876–81. 24. Cain PR, Rubash HE, Wissinger HA, McCain EJ. Periprosthetic femoral fractures following total knee arthroplasty. Clin Orthop 1986;208:205–14. 25. Cukler JM, Star AM, Alavi AM, Noto RB. Diagnosis and management of the infected total joint arthroplasty. Orthop Clin North Am 1991;22:523–30. 26. Evans BG, Cuckler JM. Evaluation of the painful total hip arthroplasty. Orthop Clin North Am 1992;23:303–11. 27. McAuley JP, Moreau G. Sepsis: etiology, prophylaxis and diagnosis. In: Callaghan JJ, Rosenberg AG, Rubash HE, editors. The Adult Hip. Lippincott-Raven; 1997. 28. Trampuz A, Steckelberg JM, Osman DR, Cockerill FR, Hanssen AD, Patel R. Advances in the laboratory diagnosis of prosthetic joint infection. Rev Med Microbiol 2003;14:1–14. 29. Atkins LA, Athanasou N, Deeks JJ, Crook DWM, Simpson H, Peto TEA, et al. Prospective evaluation of criteria for microbiological diagnosis of prosthetic-joint infection at revision arthroplasty. J Clin Microbiol 1998;36:2932–9. 30. Gallo J, Raska M, Dendis M, Florschutz AV, Kolar M. Molecular diagnosis of prosthetic joint infection: a review of evidence. Biomed Papers 2004;148:123–9. 31. Short WH, Hootnick DR, Murry DG. Ipsilateral supracondylar fracture femur fractures following knee arthropasty. Clin Orthop 1981;158:111. 32. Cordeiro EN, Costa RC, Carazzato JG, Silva JDS. Periprosthetic fractures in patients with total knee arthroplaties. Clin Orthop Rel Res 1988;252:182–9. 33. Hefley Jr FG, Nelson CL, Puskarich-May CL. Effect of delayed admission to the hospital on the preoperative prevalence of deep vein thrombosis associated with fracture around the hip. J Bone Joint Surg Am 1996;78:581–3.

ARTICLE IN PRESS Periprosthetic fractures above a total knee arthroplasty 34. Chen F, Mont MA, Bachner RS. Management of ipsilateral supracondylar femur fractures following total knee arthroplasty. J Arthroplasty 1994;9:521–6. 35. Booth RE. Revision total knee arthroplasty: management of periprosthetic fractures. Orthopeadics 1994;17:845–7. 36. Srinivasan K, Macdonald DA, Tzioupis CC, Giannoudis PV. Role of long stem revision knee prosthesis in periprosthetic and complex distal femoral fractures: a review of eight patients. Injury 2005;36:1094–102. 37. Kassab M, Zalzal P, Azores GMS, Pressman A, Liberman B, Gross AE. Management of Periprosthetic femoral fractures nafter total knee arthroplasty using a distal femoral allograft. J Arthroplasty 2004;19:361–8. 38. Wong P, Gross AE. The use of structural allografts for treating Periprosthetic fractures about the hip and knee. Orthop Clin North Am 1999;30:259–64. 39. Kraay MJ, Goldberg VM, Figgie MP, Figgie HE. Distal femoral replacement with allograft/prosthetic reconstruction for treatment of supracondylar fractures in patients with total knee arthroplasty. J Arthroplasty 1992;7:7–16. 40. Freedman EL, Hak DJ, Johnson EE, Eckardt JJ. Total knee arthroplasty including a modular disatl femoral component in elderly fracture or nonunion. J Orthop Trauma 1995; 9:231–7. 41. Kregor PJ, Stannard J, Zlowodzki M, Cole PA, Alonso J. Distal femoral fracture fixation utilizing the less invasive stabilization system (LISS): the technique and early results. Injury 2001; 32:32–47. 42. Ricci AR, Yue JJ, Taffet R, Catalano JB, Defalco RA, Wilkens KJ. Less invasive stabilization system for treatment of distal femur fractures. Am J Orthop 2004;33:250–5. 43. Schandelmaier P, Partenheimer A, Koenemann B, Grun AO, Krettek C. Distal femoral fractures and LISS stabilization. Injury 2001;32:55–63. 44. Schutz M, Muller M, Krettek C, Hontzsch D, Regazzoni P, Ganz R, et al. Minimally invasive fracture stabilization of distal femoral fractures with the LISS: a prospective multicenter study. Results of a clinical study with special emphasis on difficult cases. Injury 2001;32:48–54. 45. Syed AA, Agarwal M, Giannoudis PV, Matthews SJE, Smith RM. Distal femoral fractures: long-term outcome following stabilization with the LISS. Injury 2004;53:599–607. 46. Frigg R, Appenzeller A, Christensen R, Frenk A, Gilbert S, Schavan R. The development of the distal femur less invasive stabilization sysytem (LISS). Injury 2001;32:24–31. 47. Tepic S, Perren SM. The biomechanics of the PC-fix internal fixator. Injury 1995;26:5–10. 48. Borrelli J, Prickett W, Song E, Becker D, Ricci W. Extraosseous blood supply of the tibia and the effects of different plating techniques: a human cadeveric study. J Orthop Trauma 2002;16:691–5. 49. Marti A, Fankhauser C, Frenk A, Cordey J, Gasser B. Biomechanical evaluation of the less invasive stabilization system for the internal fixation of distal femur fractures. J Orthop Trauma 2001;15:482–7. 50. Kregor PJ, Stannard JA, Zlowodzki M, Cole PA. Treatment of distal femur fractures using the less invasive stabilization

385

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

system: surgical experience and early clinical results in 103 fractures. J Orthop Trauma 2004;18:509–20. Althausen PL, Lee MA, Finkemeier CG, Meehan JP, Rodrigo JJ. Operative stabilization of supracondylar femur fractures above total knee arthroplasty: a comparison of four treatment methods. J Arthr 2003;18:834–9. Wick M, Muller EJ, Kutscha-Lissberg F, Muhr G. Periprosthetic supracondylar femoral fractures: LISS or retrograde intramedullary nailing? Problems with the use of minimally invasive technique. Unfallchirurg 2004;107:181–8. Engh GA, Ammeen DJ. Periprosthetic fractures adjacent to total knee implants: treatment and clinical results. Instruct Course Lect 1998;47:437–48. Chen YH, Chen WM, Huang CK, Chen TH. Retrograde intramedullary Huckstep nailing for supracondylar fracture of femur after total knee arthroplasty. J Chin Med Assoc 2003;6:755–8. Sekel R, Newman AS. Supracondylar fractures above a total knee arthroplasty: a novel use of the Huckstepp nail. J Arthroplasty 1994;9:445–7. Smith JS, Martin SL, Mabrey JD. The use of a supracondylar nail for treatment of a supracondylar fracture of the femur following total knee arthroplasty. J Arm 1996;11:210–3. Weber D, Pomeroy DL, Schaper LA, Badenhausen WE, Curry JI, Smith MW, et al. Supracondylar nailing of distal Periprosthetic femoral fractures. Int Orthop 2000;24:33–5. Jabezenski FF, Crawford M. Retrograde intramedullary nailing of supracondylar femur fractures above total knee arthroplasty. J Arthroplasty 1995;10:95–101. McLaren AC, Dupont JA, Schroeber DC. Open reduction internal fixation of supracondylar fractures above total knee arthroplasties using the intramedullary supracondylar rod. Clin Orthop 1994;302:194–8. Murrell GA, Nunley JA. Interlocked supracondylar intramedullary nails for supracondylar fractures after total knee arthroplasty: a new treatment method. J Arthroplasty 1995;10: 37–42. Bezwada HP, Neubauer P, Baker J, Israelite CL, Johanson NA. Periprosthetic supracondylar femur fractures following total knee. Arthroplasty 2004;19:453–8. Ritter MA, Keating M, Faris PM, Meding JB. Rush rod fixation of supracondylar fractures above total knee arthroplasty. J Arthroplasty 1995;10:213–6. Zehntner MK, Ganz R. Internal fixation of supracondylar fractures after condylar total knee arthroplasty. Clin Orthop 1993;293:219–24. Ochsner PE, Pfister A. Use of the fork plate for internal fixation of periprosthetic fractures and osteotomies in connection with total knee replacements. Orthopaedics 1999;22:517–21. Moran MC, Brick GW, Sledge CB, Dysart SH, Chien EP. Supracondylar femoral fracture following total knee arthroplasty. Clin Orthop 1996;324:196–209. Al-Shawi AK, Smith SP, Anderson GH. The use of a carbon fiber plate for periprosthetic supracondylar femoral fractures. J Arthroplasty 2002;17:320–4. Simon RG, Brinker MR. Use of Ilizarov external fixation for a periprosthetic supracondylar femur fracture. J Arthoplasty 1999;14:118–21.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 386–392

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

Radiology quiz QUESTIONS Question 1 Figures 1a and b are axial magnetic resonance (MR) images through the L5/S1 disc space in a patient with recurrent sciatica following previous microdiscectomy.

Figure 1

What are the MR sequences? What are the findings? What is the diagnosis?

0268-0890/$ - see front matter doi:10.1016/j.cuor.2006.02.009

ARTICLE IN PRESS Radiology quiz

387

Question 2 Figures 2a–c are axial MR images through the L5/S1 disc space in another patient with recurrent sciatica following previous microdiscectomy. Figure 2d is a sagittal midline image of the same patient.

Figure 2

What are the MR sequences? What are the findings? What is the diagnosis?

ARTICLE IN PRESS 388

Radiology quiz

Question 3 Figures 3a and b are computed tomographic (CT) images of the pelvis in a patient with ongoing right sacro-iliac joint (SIJ) pain following insertion of a SIJ diastasis screw for pelvic trauma.

Figure 3

What are the findings? What is the diagnosis? What other investigations may be helpful?

ARTICLE IN PRESS Radiology quiz

389

Question 4 Figures 4a–c are axial MR images through the wrist at the level of the distal carpal row. Figure 1a is T1 weighted, 1b is a STIR image and 1c is a T1 fat-saturated image with gadolinium enhancement. Figure 4d is a T1 weighted long axis view through the wrist at the level of the trapezoid.

Figure 4

What are the findings? What is the diagnosis? What should happen to the patient next?

ARTICLE IN PRESS 390

Radiology quiz

Question 5 Figure 5a is a radiograph of a patient with wrist pain following a fall. Figures 5b and c are MR arthrogram images of the same wrist (5b a coronal section and 5c an axial section through the level of the proximal carpal row).

Figure 5

What are the radiographic findings? What are the findings on the MR arthrogram?

ARTICLE IN PRESS Radiology quiz

391

Question 6 Figure 6a (sagittal) and 6b (coronal) are MR images of the knee in a young man complaining of pain and swelling following a twisting injury.

Figure 6

What are the findings? What is the diagnosis?

ARTICLE IN PRESS 392

ANSWERS Answer 1. Epidural fibrosis Figure 1a is a T1 weighted MR image (the cerebrospinal fluid is dark). An area of low signal (arrow) can be seen surrounding the right S1 nerve root, replacing the normally high signal epidural fat. Post contrast (Fig. 1b) the area of low signal diffusely enhances consistent with epidural fibrosis. The enhancement of the scar tissue diminishes after about 2 years but some residual enhancement always persists.

Answer 2. Recurrent sequestrated disc Figure 2a is a T1-weighted MR image. A small mass, of intermediate signal intensity, can be seen obliterating the left lateral recess. The mass indents the left side of the thecal sac ( ) and the exiting nerve root is compressed and cannot be identified. Figure 2b is a T2-weighted MR image (the cerebrospinal fluid is bright) showing the thecal indentation clearly. Post contrast (Fig. 2c) the lesion (arrow) enhances peripherally suggesting an intervertebral disc prolapse. Figure 2d is a sagittal T2-weighted MR image confirming the origin of the disc prolapse (arrow) at the L4/ 5 level. The prolapse has lifted the posterior longitudinal ligament away from the back of the vertebral bodies. The ligament (arrowhead) can be seen as a thin dark line tented over the prolapsed disc. The L4/5 and L5/S1 level discs are dehydrated (compare their low T2 signal with the signal from other levels) consistent with degenerative disease. It is important to differentiate recurrent disc from epidural fibrosis as a cause of recurrent sciatic pain following spinal surgery. Fibrosis is managed conservatively but a recurrent disc prolapse may be removed surgically.

Answer 3. Post-operative SIJ infection Figure 3a shows irregular and eroded right SIJ surfaces, widened joint space and bony sclerosis in the surrounding bone of the sacrum and ilium. There has been some significant bony destruction posteriorly in the joint (arrow). The left SIJ is normal (compare the two sides). The infection also involves the metalwork: a dark ‘halo’ is seen surrounding the SIJ screw (Fig. 3b, black arrow). This indicates periprosthetic bone resorption. Whilst such resorption may sometimes be aseptic in origin it must be assumed to be infective in this patient given the other changes in the sacro-iliac joint. An MRI would give information about the extent of bone and soft tissue involvement surrounding the SIJ. Fluoroscopic or CT guided joint aspiration and biopsy would provide tissue for microbiological analysis.

Answer 4. Giant cell tumour of tendon sheath at wrist There is a mass lesion over the dorsum of the wrist (arrowheads) which is heterogenously low signal on all sequences. It shows some peripheral enhancement with gadolinium (Fig. 4c, arrowheads) but no central enhancement (Fig. 4c, long arrow). On the sagittal image, the lesion

Radiology quiz is seen to lie in close approximation to one of the extensor tendons of the back of the wrist (Fig. 4d, long arrow). These features are consistent with a synovial proliferative disorder. These usually arise in the synovium of joints where they are known as pigmented villonodular synovitis (PVNS). The same disease arising in the synovium of a tendon sheath is known as giant cell tumour (as in this example). The lesions consist of histiocytes, hyperplastic synovial elements and giant cells. Haemosiderin deposition occurs and accounts for the low signal seen on all MR sequences. They usually present as a monoarticular arthropathy or a gradually enlarging lump in a young adult. The patient should be referred to a soft tissue sarcoma service for resection. Although the lesion is benign, it has a tendency to recur both at the same site and in neighbouring tendon sheaths.

Answer 5. Scapholunate dissociation (Terry Thomas sign) The distance between the scaphoid and lunate on a plain radiograph should be less than 2 mm. A gap of greater than this (Fig. 5a, arrow; Fig. 5b) indicates scapholunate ligament disruption and scapholunate dissociation. The MR arthrogram images confirm the disruption. Gadolinium is injected at the side of the wrist at the level of the proximal carpal row. The contrast is normally constrained within the radio-ulnar-carpal joint by the intrinsic ligaments of the proximal carpal row. In this patient, the scapholunate ligament disruption has allowed contrast to flow abnormally into the midcarpal joints (Fig. 5b, small arrows). Small fragments of bone can be seen in the gap between scaphoid and lunate as low signal opacities (Fig. 5b, long arrow). The axial image (Fig. 5c) demonstrates the tear directly (long arrow). The residual damaged scapholunate ligament is seen as a ragged piece of tissue (short arrow) surrounded by gadolinium in the widened joint space between the scaphoid (S) and lunate (L). The scapholunate ligament is important for stabilising the wrist joint and damage to it, particularly its dorsal component, can result in dorsal intercalated segment instability (DISI).

Answer 6. Meniscal injury (bucket handle tear) Figure 6a shows a large knee joint effusion, most marked in the suprapatellar pouch (long arrow). The posterior horn of the medial meniscus is small (double arrowhead). It should be at least as big as the anterior horn (arrowhead). Extra meniscal tissue is seen lying anterior to the anterior horn of the medial meniscus (short arrow). Figure 6b shows extra meniscal tissue lying in the intercondylar notch (long arrow) and a tiny residual fragment of meniscal tissue in the medial compartment of the knee (short arrow). Compare with the normal lateral meniscus (arrowhead). These features are consistent with a ‘bucket handle’ tear of the medial meniscus. The ‘bucket handle’ has prolapsed forward and medially and now lies anterior to the anterior horn of the meniscus and in the intercondylar notch. A small remnant of posterior horn is left in the posterior joint. Christopher Hammond, Philip Robinson Department of Radiology, St. James’ University Hospital, Beckett Street, LEEDS LS7 7TF, UK

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 393–395

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

CME SECTION Three CME points available The following series of questions are based on the Mini Symposium on Bullet and Blast Injuries. Please read the articles in the mini symposium carefully and then complete the self-assessment questionnaire by filling in the square corresponding to your response to each multiple-choice question. For true or false questions, please fill in one square only. After completing the questionnaire, either post or fax the answer page back to the Current Orthopaedics Editorial Office at the address at the bottom of the RESPONSE sheet. Please photocopy this page if you wish to keep your copy of Current Orthopaedics intact. Replies received before the next issue of Current Orthopaedics is published will be marked and those reaching an adequate standard will qualify for three external CME points. You will be notified of your marks and a CME certificate will be despatched for your records.

Questions 1. What is the recognised pattern of mine injury that gives rise to a mangled hindfoot with relative preservation of the posterior compartments of the leg? (a) (b) (c) (d) (e)

Drive-over Mine stand Fragmentation injury Pick up No recognisable pattern

2. Which of the following statements concerning lower limb below-knee amputation for mine injury is most often correct? (a) Skew flap amputations are preferred, as there tends to be circumferential injury to the skin and soft tissues defining a level of amputation. (b) Long posterior flaps are usual—the gastrosoleus should be closed over bone before fashioning a posterior fasciocutaneous flap to obtain accurate muscle cover. (c) Long posterior flaps are most often used, the skin, fascia and gastrosoleus being kept together as one soft tissue flap, folded over resected bone ends. (d) Guillotine amputations are normally carried out above the level of obvious injury and left open until there is no evidence of infection. (e) Bone cuts are made 10 cm below the knee joint in all cases and the soft tissues are resected according to the level at which visible injury has occurred. Flaps are 0268-0890/$ - see front matter doi:10.1016/j.cuor.2006.05.002

created opportunistically to use available tissue to cover the bone stumps. 3. What is the correct course of action, according to the ICRC, if the dressings on a below knee amputation stump created 3 days previously begin to stain yellow-green and smell, the patient remaining systemically well? (a) Carry out a dressing change (b) Leave the dressings but change include cover for pseudomonas (c) Leave the dressings but change cover anaerobes (d) Return the patient to theatre debridement (e) Return the patient to theatre inspection and closure

the antibiotic regime to the antibiotic regime to immediately for repeat at five days for stump

4. If munitions leave multiple lead fragments in the body in the sites listed below, which one should be removed ? (a) (b) (c) (d) (e)

Adjacent to a major neurovascular bundle In contact with dura In the mediastinum In contact with the synovial cavity of a joint Within a fracture site

5. What additional prophylaxis should be considered in injuries caused by suicide bombers as distinct from injuries caused by most land mine explosions?

ARTICLE IN PRESS 394 (a) (b) (c) (d) (e)

CME SECTION Tetanus immunoglobulin Penicillin Metronidazole Zidovudine Oral Polio vaccination

6. What impact velocity is required for a missile to penetrate skin? (a) (b) (c) (d) (e)

10–20 m/s 25–35 m/s 40–50 m/s 75–100 m/s 150–250 m/s

7. If a round passes through the chest and mediastinum, which of the following tissues will be most susceptible to damage? (a) (b) (c) (d) (e)

Lung Aorta Cardiac muscle Bronchi Nerve tissue

8. Which of the following pathology is not associated with blast injury?

(a) (b) (c) (d) (e)

Intracerebral haemorrhage Extradural cord haemorrhage Cholesteatoma Intramural intestinal haemorrhage Retinal artery emboli

9. Which component of a land mine explosion is associated with the most significant energy exchange, causing massive soft tissue injury? (a) (b) (c) (d) (e)

The shock front The blast wind The thermal front Heat energy behind the thermal front Chemical residue

10. What is the optimum level of amputation in the upper limb for subsequent orthotic management? (a) (b) (c) (d) (e)

Mid forearm Through elbow Supracondylar elbow Mid Humeral As distal as possible without compromising debridement

ARTICLE IN PRESS CME SECTION

395

Please fill in your answers to the CME questionnaire above in the response section provided below. A return address and fax number is given at the bottom of the page. ...............................................................................................

Responses Please shade in the square for the correct answer. 1 2 3 4 5 6 7 8 9 10

A A A A A A A A A A

& & & & & & & & & &

B B B B B B B B B B

& & & & & & & & & &

C C C C C C C C C C

& & & & & & & & & &

D D D D D D D D D D

& & & & & & & & & &

Your details (Print clearly) NAME. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ADDRESS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . FAX NO. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . EMAIL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . RETURN THE COMPLETED RESPONSE FORM by fax to +44-113-206-6791, or by post to CME, Current Orthopaedics, Orthopaedic Surgery, Clinical Sciences Building, St. James’s University Hospital, Beckett Street, Leeds LS9 7TF, UK.

E E E E E E E E E E

& & & & & & & & & &

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 396

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

CME SECTION Answers to CME questions in Vol. 20, issue 4 Please find below the answers to the Current Orthopaedics CME questions from Vol. 20, issue 4 which were based on the Mini Symposium on Revision Hip Arthroplasty. 1

A&

B&

C’

D&

E&

2

A&

B’

C&

D&

E&

3

A&

B&

C&

D’

E&

4

A&

B’

C&

D&

E&

5

A&

B’

C&

D&

E&

6

A&

B&

C&

D’

E&

7

A&

B&

C&

D&

E’

8

A&

B&

C&

D&

E’

9

A&

B&

C&

D’

E&

10

A&

B’

C&

D&

E&

11

A’

B&

C&

D&

E&

12

A&

B&

C&

D&

E’

0268-0890/$ - see front matter & 2006 Published by Elsevier Ltd. doi:10.1016/j.cuor.2006.06.003

Aims and Scope Current Orthopaedics presents a unique collection of international review articles summarizing the current state of knowledge and research in orthopaedics. Each issue focuses on a specific topic, discussed in depth in a mini-symposium; other articles cover the areas of basic science, medicine, children/adults, trauma, imaging and historical review. There is also an annotation, self-assessment questions and an exam section. In this way, the entire postgraduate syllabus will be covered in a 4-year cycle. The Journal is cited in: Cochrane Center, EMBASE/ Excerpta Medica, Infomed, Reference Update and UMI Microfilms.

Editor Professor R. A. Dickson MA, ChM, FRCS, DSc St James’s University Hospital Trust, Leeds, UK

Editorial Committee President of BOTA, M. A. Farquharson-Roberts (Gosport, UK), I. Leslie (Bristol, UK), D. Limb (Leeds, UK), M. Macnicol (Edinburgh, UK), I. McDermott (Ruislip, UK), J. Rankine (Leeds, UK)

Editorial Advisory Board

Amsterdam

K

Boston

K

Jena

L. de Almeida (Portugal) G. P. Songcharoen (Thailand) R. W. Bucholz (USA) J. W. Frymoyer (USA) R. W. Gaines (USA) S. L. Weinstein (USA) M. Bumbasirevic (former Yugoslavia)

A. K. Mukherjee (India) A. Kusakabe (Japan) A. Uchida (Japan) M.-S. Moon (Korea) R. Castelein (The Netherlands) R. K. Marti (The Netherlands) G. Hooper (New Zealand) A. Thurston (New Zealand) E. G. Pasion (Philippines)

D. C. Davidson (Australia) J. Harris (Australia) S. Nade (Australia) G. R. Velloso (Brazil) J. H. Wedge (Canada) S. Santavirta (Finland) P. N. Soucacos (Greece) M. Torrens (Greece) J. C. Y. Leong (Hong Kong)

K

London

K

New York

K

Oxford

K

Paris

K

Philadelphia

K

San Diego

K

St Louis

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 397–404

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: MANAGEMENT OF FRACTURES AROUND THE KNEE JOINT

(i) Comminuted patellar fractures Isabella Mehling, Andreas Mehling, Pol M. Rommens Department of Trauma Surgery, University of Mainz, Germany

KEYWORDS Patellar fracture; Tension-band wiring; Cannulated screw fixation; Partial and total patellectomy

Summary Purpose of review: This article reviews current best practice for the diagnosis and treatment of comminuted patellar fractures. Recent findings: Patellar fractures make up about 1% of all fractures. As a rule, fractures of the patella are caused by direct trauma to the knee. A transverse fracture is the most common fracture type. Open reduction and internal fixation is the treatment of choice for the majority of displaced patellar fractures. Treatment must achieve anatomic reduction of the articular surface and reestablish the continuity of the extensor mechanism. Tension-band wiring, interfragmentary screw fixation and a combination of cerclage wiring and screw fixation are the most accepted techniques for stabilisation. Partial or total patellectomy is generally indicated when the patella is so severely comminuted that an accurate reduction and reconstruction of the retropatellar joint surface cannot be achieved. Summary: Different methods of stabilisation for patellar fractures are used, depending mainly on the fracture pattern and the amount of displacement. The aims of operative treatment are basically accurate reduction and stable fixation that allows early mobilisation. & 2006 Elsevier Ltd. All rights reserved.

Introduction The patella is the largest sesamoid bone of the human skeleton. It is integrated into the extensor apparatus and with its articular surface, it is also a component of the patellofemoral joint. The patella serves as the fulcrum for the extensor mechanism between the quadriceps tendon and the patellar tendon. Forces transmitted across the femoropatellar joint can reach up to three to seven times body weight.1 In 1985, Corresponding author.

0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.11.004

Kapandji2 reported that maximal forces within the quadriceps tendon could be as high as 3200 N, and up to 6000 N in the patellar tendon in young, physically fit men. Patellar fractures represent 0.5–1.5% of all skeletal injuries.3 Typically, patients are between 30 and 60 years old. As a rule, the mechanism for this injury is direct trauma, such as an impact onto the knee (‘‘dashboard injury’’). Indirect trauma mechanisms may produce bony avulsions of the adjacent tendons, and occur infrequently. Fractures can also occur as complications after total knee replacement surgery or after patellar tendon graft transplantation for ACL rupture.

ARTICLE IN PRESS 398

I. Mehling et al.

Classification systems There are different classification systems for patellar fractures: based on the trauma mechanism in direct or indirect injuries, on the grade of displacement in displaced or non-displaced fractures and on the basis of the configuration of the fracture lines in transverse, vertical, marginal, osteochondral or comminuted fractures. These systems are too imprecise for scientific use. The OTA classification, similarly to the AO classification,1,4 describes the different fracture types in extra-articular (A), partial articular (B) and complete articular fractures (C) (Fig. 1). Each fracture type has its own code, consisting of three elements, e.g., 45-C1.3. The first element, 45, identifies the bone: the patella. The second element describes the fracture type: (A) extra-articular, extensor mechanism disrupted, (B) partial articular, extensor mechanism intact, for example, often vertical fractures, (C) complete articular, disrupted extensor mechanism. The classification of Speck and Regazzoni5 also differentiates the fracture in three types (A, B, C), with three subgroups for each fracture type. The classification of Rogge et al.6 differentiates seven fractures types.

Diagnosis The medical history is the first step of the diagnostic workup. Information on the mechanism of trauma helps for estimation of the severity of injury and the fracture pattern. The activity level of the patient and his/her current medical problems are also important for decision making on further treatment.

The clinical examination should include an inspection of the whole extremity. Clinical signs of a patellar fracture are swelling and pain in the knee joint. Fracture blisters, skin lacerations, abrasions or contusions are signs of direct trauma and should be documented. Wounds should be checked to confirm whether the fracture is open or closed. In displaced patellar fractures, a defect zone between the fragments may be palpable. Often there is a hemarthrosis of the knee. Flexion and extension in the knee joint is limited and painful. Active extension and lifting of the leg is ususally impossible. However, the ability to extend the knee does not rule out a patellar fracture, because the medial and lateral retinacula may be still intact.1 The stability of the knee joint should be carefully examined. Of course a check of the peripheral pulses, the compartments of the leg, and a neurological examination should always be performed. Special interest should be paid to potential ipsilateral concomitant injuries, e.g. acetabular fractures, femoral fractures or tibial fractures, which are signs of serious trauma. For radiographic examinations, a standard X-ray of the knee in two planes as well as a 301 tangential view of the patella should be performed. In the anteroposterior (AP) view, the patella is normally centred on the medullary axis of the femur. In the lateral view, the patellar fracture is best visible: displacement, intra-articular involvement and degree of comminution can be assessed. Vertical patellar fractures are best seen on the axial view. With the Insall7 method of relating the greatest diagonal lengths of the patella and the patellar tendon, abnormal position of the patella, e.g., patella alta or patella baja or a rupture of the patellar ligament are recognisable. Evaluation of the true degree of damage of the patellar fracture in the conventional radiographic examinations is not always possible due to the cancellous bone structure of the patella. An additional CT-scan is seldom necessary.

Figure 1 OTA classification for patellar fractures (J Orthop Trauma, 1996).4

ARTICLE IN PRESS Comminuted patellar fractures

399

However, a CT-scan can be helpful for the evaluation of articular incongruity in cases of non-union, malunion and patellofemoral alignment disorders. Special attention should be paid to bipartite or tripartite patellae, which occur due to a lack of convergence of the bone during growth. Bipartite and tripartite patellae have characteristic signs on the radiographs, with rounded, sclerotic lines in contrast to the sharply edged lines of a fractured patella. For differentiation it can be helpful to compare the radiographs with the contralateral side. Magnet resonance imaging is only recommended in special cases, for example in stress fractures. Differential diagnoses for patellar fractures are contusion of the knee, tendon ruptures (rupture of the quadriceps tendon or the patellar ligament), injuries of the capsular ligament of the knee and patellar dislocation or growth abnormalities.

Non-operative versus operative treatment Operative management is the treatment of choice for the majority of patellar fractures, especially those with displacement and intra-articular involvement. The AO recommends the following treatment method, depending on the fracture type3 (see Table 1). Displacement of more than 3 mm or articular incongruity of more than 2 mm is considered an indication for surgical treatment. The grade of dislocation can be checked in the radiographic examination whereas stability can be proven by clinical examination. The extensor apparatus is intact if the patient can lift up the extended leg. A secondary loss of position is unlikely. Contraindications and relative contraindications for surgical treatment include non-displaced or minimally displaced stable fractures. These fractures can be treated nonoperatively. Contused or injured skin areas, which preclude a safe surgical approach to the fracture, active infection on the extremity or medical conditions of the patient, which do

Table 1

not permit safe surgical intervention, are contraindications. Open fractures of the patella are treated as other open joint injuries. Open fractures are surgical emergencies; other lesions are best treated as soon as possible, depending on the condition of the local soft tissues.

Therapy The aim of the therapy is an anatomic reduction and stable fixation of the fracture, which allows early mobilisation. Stable fractures without dislocation or minimally displaced stable fractures can be treated conservatively. We prefer immobilisation in a semicircular plaster cast, with the knee extended. In longitudinal fractures, full weight bearing is allowed. In transverse fractures, in the first 4 or 6 weeks, only half weight bearing is allowed because of the risk of secondary loss of reduction. Physiotherapy should be performed out of the plaster cast. Active and passive knee mobilisation are limited in the beginning up to 601 of flexion, later on up to 901 of flexion. After 6 weeks, increased weight bearing and free knee motion are allowed. The immobilisation in the plaster cast should be not longer than 6 weeks.

Surgical technique The patient is positioned supine on a radiolucent operation table. To avoid external rotation of the leg, a cushion is used under the ipsilateral hip. With cushioning below the lower leg, a knee flexion of 30–401 is achieved. This is the optimal position for Kirschner-wire drilling. A tourniquet can be placed high around the involved thigh and inflated to about 300 mmHg, depending on the size of the leg and on the patient’s blood pressure. Use of a tourniquet is not absolutely necessary. The surgeon should take into account the fact that the inflated tourniquet can complicate the reduction of the fracture, as under tourniquet pressure the patella can be displaced proximally due to contraction of

Recommended therapy from the AO1 for each fracture pattern.

Patella #4545-A

Therapy Extra-articular

Pole fractures

Extensor mechanism disrupted Operative:

 Lag srew+tension-band wire or cerclage

 Transosseous suture+cerclage 45–B

Partial articular

Vertical fractures

 Non-displaced  Displaced, simplemultifragmentary

 Non-operative  Transverse lag screw+cerclage  Circumferential cerclage+tension band

45-C

Complete articular

Transverse fractures

Disrupted extensor mechanism  K-wire+tension-band wire  Lag screw or K-wire+tension-band wire  + Third fragment  K-wires, screws+tension-band  Four or more fragments  Partial or total patellectomy  Comminuted fracture

ARTICLE IN PRESS 400 the quadriceps muscle. Therefore, the knee should be carefully flexed and the patella manually pulled distally before inflating the tourniquet. The incision can either be longitudinal or transverse. We prefer the longitudinal incision over the patella, because, if necessary, it can be enlarged distally and proximally and does not interfere in case of later revision. For the best cosmetic result, the transverse incision is preferable because it lies within the Langer’s lines. However, one should consider that this approach may injure the infrapatellar branch of the saphenous nerve. Care should also be taken of the vessels of the geniculate arteries, and an en-bloc preparation of the fasciocutaneous layers under the bursa prepatellaris should follow. Under direct vision, an anatomical reconstruction is performed with the aid of one or several bone reduction forceps. By reducing smaller fracture fragments to each other, we convert a complex fracture pattern into a simple one.8 Anatomical reduction of the articular surface is checked by digital palpation of the patellofemoral joint inside the knee. The most common osteosynthesis technique is tensionband wiring. The principle of this technique is to transform distraction forces into compression forces. Two different techniques of tension-band wiring are in use: the outside-in technique and the inside-out technique. In the outside-in technique, the fracture is first reduced and then fixed with two Kirschner wires (1.6 mm stainless-steel wire), which are drilled in the axial direction through the reduced fragments. In the inside-out technique, first the Kirschner wires are drilled into one of the unreduced fragments, and then reduction and completion of the fixation follows. We prefer to drill the Kirschner wires from the distal to the proximal pole because it is easier to find the optimal entry portal in the distal patellar pole for Kirschner-wire drilling. Following Kirschner-wire fixation, a 30 cm segment of a 1.25 mm wire should pass adjacent to and behind the Kirschner wires. Close approximation of the wire to the proximal and distal pole of the patella is recommended. The cerclage wire is placed in the form of a figure-zero or figure-of-eight fashion. The figure-zero fashion in comparison seems more resistant

I. Mehling et al. to torsion forces.1 Nevertheless, both the figure-zero and the figure-of-eight wire enhance total stability and compress the superficial parts of the fracture fragments. Especially the frontal wire changes distraction into compression during knee flexion. For symmetrical tensioning of the wire, a double-loop technique is recommended. (Figs. 2–5 show examples of tension-band wiring). Then, the ends of the two cerclage wires are hand-tightened by lifting them up on the clamp. The proximal pins of the two Kirschner wires are bent, shortened and turned towards the quadriceps tendon and put into the patella to avoid skin irritation and loosening. The distal pins are cut at short length so that they are not prominent within the patellar tendon. Finally, the quality of reduction should be checked again and the knee gently flexed to assess the stability of the fixation. Before wound closure, the tourniquet is deflated and haemostasis is achieved with electrocoagulation. A suction drain is placed into the knee joint and closure of the wound can be performed in layers. First, closure of the prepatellar bursa with 2–0 resorbable sutures is performed. Then the subcutaneous tissues are closured with simple inverted 2–0 resorbable sutures, and finally skin closure is performed. With the tension-band wiring technique even comminuted fractures can be reduced and stabilised.1 In these fractures, the first step of internal fixation after fragment reduction is the placement of the circumferential cerclage wire in order to avoid recurrent displacement of the fragments. An additional figure-of-eight cerclage wire must be combined with the tension-band wiring technique. Screw osteosynthesis is an alternative to the tension-band wiring technique. Depending on the thickness and bone quality of the patellar bone, 6.5 mm cancellous bone screws or 3.5 mm cortical screws can be used.9 Also, fractures of the superior or inferior pole of the patella or fractures with small fragments can be stabilised by lag screws (see Fig. 6). In vertical patellar fractures, transverse lag screw fixation in combination with a cerclage wire is a sufficient fixation technique (see Fig. 7).

Figure 2 Left patellar fracture in an 86-year-old female. The fracture has a transverse component with smaller additional fracture fragments: (a, b) preoperative X-rays ap and lateral and (c, d) postoperative X-rays of fracture healing ap and lateral. The fracture is stabilised with two K-wires and tension-band wiring with steel cable in double-loop technique.

ARTICLE IN PRESS Comminuted patellar fractures

401

Figure 3 Right transverse patellar fracture in a 56-year-old female stabilised with three K-wires and tension-band wiring with braided cable in figure-of-eight fashion: (a, b) preoperative X-rays AP and lateral and (c, d) postoperative X-rays of fracture healing AP and lateral.

Figure 4 A minimally displaced multi-fragment fracture of the right patella in a 64-year-old female: (a, b) preoperative X-rays AP and lateral, (c, d) postoperative X-rays AP and lateral showing slight distal migration of one of the K-wires and (e, f) X-rays AP and lateral after hardware removal 1 year after osteosynthesis.

Another variation of the tension band technique in combination with screws is the use of 4.0 mm cannulated screws with the tension-band wire passed through the

cannulated screws and tightened in a double-loop technique. Berg10 described equivalent clinical results for transverse patellar fractures for fixation with a tensioned anterior

ARTICLE IN PRESS 402

I. Mehling et al.

Figure 5 A comminuted patellar fracture in a 46-year-old female stabilised with three K-wires and double-loop tension-band wiring: (a, b) preoperative X-rays AP and lateral, (c, d) postoperative X-rays AP and lateral with two visible suction drains, (e, f) postoperative X-rays AP and lateral after 1 week showing a failure of the circular loop cable, (g, h) postoperative X-rays AP and lateral after 4 months after removal of the broken cable and (i, j) postoperative X-rays AP and lateral 8 months after stabilisation. Complete hardware removal was performed; the patellar fracture was completely healed.

figure-of-eight wire placed through parallel cannulated screws in comparison to reports of fixation with modified tension-band wiring. Burvant et al.11 and Carpenter et al.12 showed, in biomechanical studies, a significantly higher stability for osteosynthesis with screws only or for the combination of cannulated screws with a tension-band wire compared to a single tension-band wire. Scilaris et al.13 described, in a

biomechanical comparison in transverse non-comminuted patellar fractures, a better fixation with two Kirschner wires and a 1.0 mm braided cable tension loop as opposed to a monofilament wire tension loop. Fortis et al.14 demonstrated in an experimental investigation that tension-band wiring is highly effective for the fixation of the fractured patella but is improved by an additional circular wire.

ARTICLE IN PRESS Comminuted patellar fractures

403

Figure 6 Avulsion fracture of the distal pole of the patella in a 23-year-old male. Osteosynthesis was performed with two cancellous bone screws: (a, b) preoperative X-rays AP and lateral and (c, d) postoperative X-rays of fracture healing AP and lateral.

Osteochondral fragments can be fixed with biodegradable pins. There are implants of polyglycolic acid (PGA), polydioxanone (PDS) or polylactic acid (PLA). These implants are only recommended for adaptation of unloaded fragments and not in areas of high mechanical stress. Their advantage is that implant removal is not necessary.1,15 A new stabilisation technique for comminuted patellar fractures is described by Yammis et al.16 Instead of the tension-band technique, they used a circular external fixator, which is placed under arthroscopic control. The authors suggested that this treatment can provide enough stability to allow active knee motion in the early postoperative period. In addition, arthroscopic examination of the knee allows assessment of other intra-articular lesions. With a comminuted patellar fracture, a partial patellectomy may be necessary if adequate anatomic reduction of the displaced fragments is not possible. Whenever possible, partial patellectomy is preferred to total patellectomy because it keeps the fulcrum function of the patella intact. A comminuted upper or lower pole or a comminuted zone in the middle of the patella can be managed by removing the small fragments. Non-resorbable transosseous sutures have to be used after resection of the upper pole for fixation of the quadriceps tendon and after lower pole resection for fixation of the patellar tendon, securing the suture with a tension-band wire. A tilt of the remaining patella should be avoided in all cases. In cases of severe comminution and extended cartilage damage a patellectomy may be the only option. However, one should be aware that this always means a decrease in muscle strength of the quadriceps muscle. All bony fragments and the damaged tissue should be removed leaving as much extensor apparatus as possible. If the defect zone is more than 4 cm wide and a direct adaptation is not possible, an inversion of the quadriceps tendon in accordance to Miyakawa17 may be necessary.

Postoperative treatment The postoperative treatment depends on the fracture type and the stability of surgical fixation. In patients with stable fractures, knee motion exercises begin immediately. Knee

flexion is most important for transforming distraction forces into compression forces, and this supports bone healing. Continuous passive motion (CPM) is helpful in the early days after surgery. Drains are removed on the second postoperative day or depending on the amount of wound drainage. The patient begins with isometric exercises and out of bed mobilisation with the help of a physiotherapist. For protection of the osteosynthesis, we prefer mobilisation with a semicircular plaster cast. Normally, full weight bearing is allowed. Walking exercises begin with partial weight bearing of 15–20 kg or half body weight bearing and the help of two crutches for a 6-week duration. Actively assisted motion of the knee is allowed from full extension to 901 of flexion. Total weight bearing without the plaster cast and free motion are allowed after the seventh postoperative week, provided that the patient has a good clinical feeling and that X-rays show ongoing bony healing. Implant removal is possible after 1 year, on average.

Complications and prognosis Feared complications are disturbed wound healing and deep infection. Careful soft tissue debridement, spared resection and secondary wound closure are recommended to reduce wound healing disturbances. In cases of larger skin defects, split skin grafting or gastrocnemius rotation flaps covered with split skin grafts may be necessary. Infection should be treated aggressively with radical debridement with drainage and with antibiotics. Bowing of the Kirschner wires or failure of the figure-zero or figure-of-eight fashion wire may happen in patients who perform unlimited and aggressive early active mobilisation. When fracture displacement occurs in relation to the above, re-osteosynthesis is mandatory. Sometimes, rupture of the wire is discovered later on. There is no need for intervention when the patellar fracture is already healed. Another possible complication after surgical treatment of patellar fractures is loss of fixation and reduction. If there is a loss of fixation without loss of reduction, the leg should temporarily be immobilised. If there are signs of loss of fixation and reduction, a revision of internal fixation should be performed.

ARTICLE IN PRESS 404

I. Mehling et al. remaining steps or gaps. Nevertheless, an unsatisfactory outcome is also possible when the radiographic examination shows an anatomical healing. As with other articular lesions, there is evidence that optimal reduction will give the best long-time results. Steps and gaps in the articular surface will be responsible for knee complaints such as swelling, pain and diminution of movement.

References

Figure 7 Vertical patellar fracture in a 61-year-old male stabilised with two transversal lag screws: (a, b) preoperative X-rays AP and lateral and (c, d) postoperative X-rays of fracture healing AP and lateral after.

Delayed union and non-union are usually the result of a failure of fixation or inadequate initial reduction. This can be avoided by anatomical reduction and good fixation techniques during primary surgery as well as close postoperative follow-up. If there is a delayed union, it can be treated with repeated cerclage-wire techniques. Significant mal-unions may require revision osteosynthesis or patellectomy. Klassen and Trousdale18 showed in a retrospective study that patients with minimal symptomatic delayed union or non-union of the patella can be successfully treated nonoperatively with the knowledge that the fracture will not unite. Operative management of symptomatic patients can be expected to achieve union and increased function of the knee. Especially in comminuted fractures, arthrofibrosis and loss of knee function are relatively common complications. An aggressive and persistent physical therapy regimen is necessary to avoid these complications or to treat them. For arthrofibrosis, an arthroscopic debridement may be beneficial. Post-traumatic osteoarthrosis is the consequence of traumatic cartilage damage or suboptimal surgery with

1. Nerlich M, Weigel B. Patella. In: Riiedi TP, Murphy WM, editors. AO principles of fracture management. Stuttgart, New York: Thieme; 2000. p. 482–97. 2. Kapandji IA. Funktionelle Anatomie der Gelenke. Enke Verlag: Stuttgart; 1985. 3. Bostrom A. Fracture of the patella. A study of 422 patellar fractures. Acta Orthop Scand Suppl 1972;143:1–80. 4. Orthopaedic Trauma Association. Fracture and dislocation compendium. Orthopaedic Trauma Association Committee for Coding and Classification. J Orthop Trauma 1996;10(Suppl 1): 1–155. 5. Speck M, Regazzoni P. Klassifikation der Patellafrakturen. Z Unfallchir Versicherungsmed 1994:27–30. 6. Rogge D, Oestern HJ, Gosse ´ F. Die Patellafraktur. Orthopade 1985;14:266–80. 7. Insall JN. Anatomy of the knee. Surgery of the knee. ChurchillLivingstone: New York; 1984. p. 1–20. 8. Wilber JH. Patellar fractures: open reduction internal fixation. In: Wiss DA, editor. Fractures. Philadelphia: Lippincott-Raven Publisher; 1998. p. 335–46. 9. Galla M, Lobenhoffer P. Frakturen der Patella. Der Chirurg 2005; 76:978–99. 10. Berg EE. Open reduction internal fixation of displaced transverse patella fractures with figure-eight wiring through parallel cannulated compression screws. J Orthop Trauma 1997;11: 176–573. 11. Burvant JG, Thomas KA, Alexander R, Harris MB. Evaluation of methods of internal fixation of transverse patella fracture: a biomechanical study. J Orthop Trauma 1994;8:147–53. 12. Carpenter JE, Kasman RA, Patel N, Lee ML, Goldstein SA. Biomechanical evaluation of current patella fracture fixation techniques. J Orthop Trauma 1997;11:351–6. 13. Scilaris TA, Grantham JL, Prayson MJ, Marshall MP, Hamilton JJ, Williams JL. Biomechanical comparison of fixation methods in transverse patella fractures. J. Orthop Trauma 1998;12:356–9. 14. Fortis AP, Milis Z, Kostopoulos V, et al. Experimental investigation of the tension band in fractures of the patella. Injury 2002;33:489–93. 15. Hoffmann R, Weller A, Helling HJ, Krettek C, Rehm KE. Local foreign-body reactions to biodegradable implants. A classification system. Der Unfallchirurg 1997;100:658–66. 16. Yammis I, Oguz E, Atesalp AS, et al. Application of circular external fixator under arthroscopic control in comminuted patella fractures: technique and early results. J Trauma 2006; 60:659–63. 17. Baker CL, Hughston JC. Miyakawa patellectomy. J Bone Joint Surg Am 1988;70:1489–94. 18. Klassen JF, Trousdale RT. Treatment of delayed and nonunion of the patella. J Orthop Trauma 1997;11:188–94.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 405–410

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: MANAGEMENT OF FRACTURES AROUND THE KNEE JOINT

(ii) The ‘‘floating knee’’ in adults and children Byron Chalidis, Saurabh S. Metha, Eleftherios Tsiridis, Peter V. Giannoudis The Academic Unit of Orthopaedic Surgery, A Floor, Clarendon Wing, Leeds General Infirmary, Great George Street, Leeds LS1 3EX, UK

KEYWORDS Ipsilateral femoral and tibial fractures; Knee injuries; Polytrauma

Summary The treatment of simultaneous ipsilateral femoral and tibial fractures is a therapeutic challenge, often complicated by concomitant multisystem injury. This combination necessitates vigilance in identifying and promptly treating associated neurovascular compromise. In the context of multiple trauma, a spanning external fixator could be considered to allow restoration of haemodynamic stability and to minimise the immunoinflammatory response to injury. Definitive reconstruction can be performed at later stage. Intramedullary nailing appears to be the most common method of stabilisation. & 2006 Elsevier Ltd. All rights reserved.

Introduction ‘‘Floating knee’’ is the term applied to a flail knee joint segment resulting from fractures of the shaft or adjacent metaphysis of the ipsilateral femur and tibia.1 The injury, which was initially described by Blake and McBryde,1 is generally caused by high-energy trauma. Local trauma to the soft tissues is often extensive and life-threatening injuries to the head, chest or abdomen may co-exist.2,3 The initial evaluation of the extent of injuries is of critical importance and should be followed by an appropriate sequence of emergency diagnosis and therapeutic measures according to the ATLS guidelines. These injuries have received a lot of attention during the last few decades. The rates of infection, non-union, malunion and stiffness of the knee are relatively high, leading to functional impairment and frequently unsatisfacCorresponding author. Tel.: +44 (0) 113 39 22750;

fax: +44 (0) 113 39 23290. E-mail address: [email protected] (P.V. Giannoudis). 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.10.005

tory results.4 In the paediatric population this fracture combination is less common than adults but epiphyseal injury can adversely affect the open growth plates, predisposing to limb length discrepancy and angular deformities.5 The aim of this review is to address the current management options for ipsilateral femoral and tibial fractures, including a consideration of the prognostic factors that may be correlated with the final outcome.

Mechanisms of injury—associated injuries A traffic accident is the most common mechanism of trauma, reported in up to 97% of cases, followed by gunshot wounds and a fall from a height.4,6 There is a male preponderance, particularly in young adults and especially in the age range of 20–30 years.7 ‘‘Floating knee injuries’’ are often combined with extensive multisystem injuries and they are accompanied by significant morbidity.1,2,5 Ostrum8 reported that the average Injury Severity Score9 was higher in patients with

ARTICLE IN PRESS 406 ipsilateral femoral and tibial fractures than in cases of isolated femoral shaft fractures (19 and 16.6, respectively). Furthermore, Rios et al.4 found 42% of patients to have a head injury, 28% chest injuries and 16% abdominal trauma. The above figures indicate that these combined fractures are more severe than simply the sum of the skeletal injuries and must be included in a polytrauma-based assessment and treatment protocol. Vascular damage and nerve lesions are also quite common. Adamson et al.10 and Paul et al.11 reported arterial lacerations in 30%, mainly in the popliteal and posterior tibial arteries. The majority of vascular injuries were associated with limb ischaemia necessitating vascular reconstruction. The incidence of nerve dysfunction was approximately 10% and the most commonly affected nerve was the peroneal nerve, as a result of traction injury. Traction usually causes neurapraxia, and resolution of nerve palsies is anticipated in the majority of cases.10,11 The incidence of open fractures is very high, approaching 50–70% at one or both fracture sites.2,4,7,10,12 The most common combination is a closed femoral fracture with an open tibial fracture. In the 89 floating knee injuries described by Hee et al.,12 there were 55 open fractures, 32 open tibial fractures, 3 open femoral and 20 both femoral and tibial open fractures. Out of 55 open fractures, 29 were Grade III and 3 of them required knee amputation. Knee ligament injury occurring in association with ipsilateral femoral and tibial fractures is well documented, occurring in roughly 13 of patients.13,14 Anterolateral rotatory instability seems to be the most common instability pattern. Knee ligament injury is not always suspected and joint swelling due to haemarthrosis may be mistaken for a ‘‘sympathetic effusion’’. It seems apparent that the ipsilateral femoral and tibial shaft fracture and knee ligament injury is part of a continuum of combined injuries resulting from complex high-energy forces. There is no clear correlation between the side or type of the fracture and which of the ligaments are involved. On the basis of the experience of many authors, a high index of suspicion for knee ligament damage is advisable.13,14 In skeletally immature patients the injury is uncommon; few existing studies report on floating knee in children.5,15–17 Data from the studies that are available show that the results are comparable with adults in terms of the mechanism of fracture, the incidence of associated major injuries and the complexity of treatment.

B. Chalidis et al. IIb with fractures of distal femur and shaft of the tibia and IIc with distal femur and tibial plateau fractures (Fig. 1). In both classifications, type II fractures with intraarticular involvement have been linked with higher complication rates and poorer functional results than type I injuries.10 There is wide acceptance of the evaluation of functional outcome according to criteria established by Karlstrom and Olerud.19 Subjective symptoms from thigh or leg, subjective symptoms from knee or ankle joint, walking ability, participation in work and sports, angulation and/or rotational deformity, shortening and restricted joint mobility (hip, knee, or ankle) are recorded and the result is characterised as excellent, good, acceptable or poor (Table 1).

B. Children In children ‘‘floating knee’’ injuries are classified according to the Bohn–Durbin5 and the Letts et al.17 classification systems. The Bohn–Durbin5 classification describes the double-shaft pattern of fracture as Type I, the juxtaarticular pattern as Type II and the epiphyseal type as Type III. However, it does not take into consideration open fractures and cannot be used to predict complications and prognosis. The authors determined that femoral union in a position of greater than 301 anterior angulation, 151 valgus angulation, 51 posterior or varus angulation, or more than 2 cm of shortening, should not be accepted. Tibial malunion was defined as greater than 51 angulation in any plane or greater than 1 cm of shortening, whereas rotational malunion was defined as any internal rotation deformity exceeding that of the unaffected side or greater than 201 external rotation of the extremity as detected during gait or stance.5 Letts and Vincent17 designed a new classification system, recognising not only diaphyseal, metaphyseal or epiphyseal knee fractures (types A, B, C) but also open fractures (types D and E), (Fig. 2). The drawback of their classification system is that they do not indicate how to classify patients

Classification—evaluation A. Adults Blake and Mcbryde1 used the terms ‘‘true’’ or type I injury and ‘‘variant’’ or type II injury to classify the ‘‘floating knee’’ fracture pattern. In type I injury there is a pure diaphyseal fracture of the femur and tibia, whilst in variant type II injury the fracture extends into the knee, hip or ankle joint. Fraser et al.18 classified the injury in a similar way, according to knee involvement. Their type I is the same as Blake’s and McBryde’s ‘‘true’’ injury, with extra-articular fractures of both bones. Type II is subdivided into three groups: IIa, with femoral shaft and tibial plateau fractures,

Figure 1

Fraser’s classification of floating knee injuries.

ARTICLE IN PRESS Floating knee in adults and children

Table 1

407

Karlstrom’s criteria for assessment of functional outcome.

Criterion

Excellent

Good

Acceptable

Poor

Subjective complaints from thigh or leg

0

Intermittent slight symptoms

More severe symptoms impairing function

Subjective symptoms from knee or ankle joint Walking ability

0

Same as above

Same as above

Considerable functional impairment; pain at rest Same as above

Unimpaired

Same as above

Work and sports

Same as before accident

Use cane, crutch, or other support Permanent disability

Angulation, rotational deformity, or both Shortening Restricted joint mobility (hip, knee or ankle)

0

Given up sport; work same as before accident o101

Walking distance restricted Change to less strenuous work 10–201

4201

1–3 cm 10–201 at ankle; 20–401 at hip, knee or both

43 cm 4201 at ankle; 4401 at hip, knee or both

0 0

o1 cm o101 at ankle; o201 at hip, knee or both

with epiphyseal separation in the distal femur and tibia, or how to describe the location of any open fracture in the epiphysis, metaphysis or diaphysis. The subjective outcome of floating knee injuries can be evaluated using Yue et al.15 criteria as follows: (a) Excellent, no complaints or limitations secondary to the extremity injury. (b) Good, occasionally minor extremity pain or decreased ability in athletic competition. (c) Fair, intermittent moderate extremity pain but able to perform all activities of daily living and most recreational activities. (d) Poor, constant extremity pain and inability to perform activities of daily living secondary to their extremity injury.

Treatment Adults

Figure 2 Letts et al.’s classification of floating knee injuries in children.

Several approaches have been reported in the literature concerning the treatment of ipsilateral femoral and tibial fractures.1,2,5,8,11,19 Early reports favoured a non-operative approach to one or both fractures because of the unpredictable methods of internal fixation that were available, high infection rates and lack of knowledge of posttraumatic critical care.4 However, conservative methods entailed considerable risks of fracture shortening and angulation in combination with a delay in fracture union and time to return to normal activities.19 As a result, non-operative methods became unfavourable and gradually replaced by internal fixation techniques.4 Lately, advances in internal fixation methods have led orthopaedic surgeons to recommend more aggressive

ARTICLE IN PRESS 408 treatment, with early stabilisation of both fractures, integrated with a multisystem approach that emphasises early mobilisation of the patient to facilitate better care and quicker respiratory recovery.4,8,12,20 The use of intramedullary nails either side of the knee joint is presently the best treatment modality for fracture management.4,7,8 An antegrade technique is usually used when nailing fractures.The procedure is performed after adequate resuscitation and physiological stabilisation of the injured patient. This is especially important in cases with a high Injury Severity Score due to associated injuries. If immediate intramedullary nailing is not possible then a spanning external fixator can be applied to stabilise the entire limb. This can be replaced by internal fixation when the patient’s general condition is stable and/or neurovascular lesions and soft tissue coverage have been managed.7

B. Chalidis et al. However, the technique of antegrade nailing has the drawbacks due to difficulties in patient positioning, two incisions, prolonged anaesthetic and surgical time and the inability to perform other surgical procedures at the same time.8 In order to overcome the above concerns, retrograde intramedullary nailing of the femur with antegrade nailing of the tibia through the same incision has been proposed as an alternative4,8 (Fig. 3). Ostrum8 using the above technique through a 4 cm medial parapatellar tendon incision, reported 88% good or excellent results with a full range of knee motion in 19 out of 20 patients. The femoral shaft fracture is always addressed first. The reasons for this are twofold. First, stabilisation of the femur allows mobilisation of the patient without traction, should the patient suffer decompensation during the operation necessitating abandonment of the second procedure. The extremity with the tibial fracture can be placed in a splint, cast, or an external fixator can be applied quickly. Second, stabilisation of the femoral shaft fracture allows the surgeon to flex the knee sufficiently to obtain access to the proximal tibial starting point and gives a stable proximal limb for support.8 In cases where the fracture pattern involves the metaphyseal region of the tibia and/or femur, with or without intraarticular extension, stabilisation can be achieved using locking plates.21 In 21 cases with ‘‘floating knee injury’’, Hung et al.20 treated 16 femoral fractures and 15 tibial fractures with plates and screws. All the patients had a type II or ‘‘variant’’ injury. The authors concluded that in cases with knee involvement intramedullary nails are not recommended. Plate fixation can offer anatomic reduction of the articular surface allowing rigorous mobilisation and maximising the functional outcome. Early diagnosis of ligament injuries is essential to facilitate an appropriate rehabilitation programme. In cases of isolated injury to medial collateral ligament, conservative treatment is preferred. It is suggested that anterior or posterior cruciate ligament reconstruction is delayed if it is necessary until after union of the fractures. Posterolateral corner injuries and avulsion fractures of cruciate ligaments either from femur or tibia should be repaired at the time of the initial procedure or in the early postoperative period.13,14

Children

Figure 3 (a) AP radiograph illustrating a mid shaft femoral and tibial fracture. (b) AP and lateral radiographs showing stabilisation of the floating knee injury pattern with retrograde femoral and antegrade tibial intramedullary nailing via a single incision.

The treatment of ipsilateral femoral and tibial fractures in children is controversial, especially in patients under 10 years of age.5,15–17 Non-operative treatment consists of skeletal traction for the femoral fracture with closed reduction and casting or splinting of the tibial fracture. A hip spica cast is applied when sufficient femoral healing has occurred. Operative treatment options include flexible nails, plates or external fixators for diaphysis fractures and crossed K-wires in epiphyseal or metaphyseal fractures.15 A disappointingly high incidence of complications of fracture healing occurs in older children treated with conservative methods.5,17 Malunion, non-union, refracture and limb-length discrepancy are documented in 40–50% of closed treatment cases. Bone et al.5 highlighted a difference between the results in patients who were treated by closed methods and those who had operative stabilisation of

ARTICLE IN PRESS Floating knee in adults and children the femoral fracture. Most authors suggest operative treatment of at least the femoral fracture in patients older than 10 years of age.5,15,16 Letts et al.17 recommend that even in younger patients (less than 9 years old) at least one fracture must be rigidly fixed, and it is usually most appropriate for this to be the tibia, despite the fact that none of their patients were treated with operative fixation. Moreover, juxtaarticular fractures are difficult to stabilise with closed methods and assessment of knee instability cannot be evaluated without fixation of one or both of the fractures. Recently, Yue et al.15 in a comparative study on closed and operative methods, reported that operative treatment reduced complications, length of hospital stay and time to unsupported walking. Therefore, the above authors concluded that both femoral and tibial fractures should be treated operatively in all age groups. Similar results were recorded by Arslan et al.16 who stated that previous recommendations of closed treatment for children younger than 9 years old with a floating knee injury should be shifted to rigid stabilisation of both fractures.

Results and complications Healing time Healing of femoral and tibial fractures occurs in approximately 15–20 weeks after injury and can be increased to 40 weeks in type II fractures.10,20 As union time may be prolonged with external fixation or casting, rigid internal fixation appears to be the best method of treatment.11,20

Non-union–malunion Delayed union, non-union, malunion and stiffness of the knee are more prevalent in patients with this combination of fractures than in patients with an isolated femoral or tibial fractures.2 Adamson et al.10 reported a 12% incidence of limb shortening (range 2–7 cm), 9% of malrotation and 12% of angulation 4101. Similarly, Hee et al. described 67% delayed union, 31% non-union, 2% incidence of malrotation 4101 and 6% of shortening of up to 2 cm. In the same study factors such as older age, increased number of pack years smoked at the time of injury, high injury severity score, open and comminuted fractures adversely affected the bony union time and reoperation rate. In the paediatric population, open growth plates presented an additional dynamic factor associated with complications that are unique to children (overgrowth of bone after fracture or premature closure of the ipsilateral physis, genu valgum and physeal arrest). The mean overgrowth of femur and tibia is 1.4 and 1.1 cm, respectively.15,16 Patients younger than 9 years of age and those who were treated non-operatively were observed to have an increased incidence of leg length discrepancy. However, a subjective limp and lower extremity length discrepancy can occur regardless of the fracture type, extent of soft tissue injury, or treatment method. Patients and parents must be counseled and informed of these possible outcomes. Patients should be followed up until skeletal maturity to monitor for signs of these complications.15

409

Infection The incidence of infection and osteomyelitis is relatively high in ‘floating knee’ injuries, especially in open and type II fracture patterns. Rios et al.4 and Veith et al.2 in their series reported a 16% incidence of wound infection and 4.4% of osteomyelitis. In addition, Yokoyama et al.6 found that deep infection developed in 13.8% of patients and meticulous debridement, continuous suction/irrigation drainage, polymethylmethacrylate beads or sticks impregnated with antibiotics led to eradication of infection in the majority of cases.

Functional outcome By using the criteria of Karlstrom and Olerud, most published series describe excellent to good results in up to 65% of operatively treated cases,2,8,10 while after conservative methods the success rates drop to only 29% of patients.18 In children, good to excellent results have been reported either with conservative or operative methods.5,15–17 Fixation of one or two fractures in children up to 9 years of age offers superior results, minimising the incidence of long-term dysfunction of the extremity.15,17 Considering the intraarticular injuries (type II), good or excellent results are reported in only 24% of patients.10 The difficulty in obtaining satisfactory function in type II injuries may result from the possibility of morbidity by severe softtissue injuries or knee joint damage and the complexities of achieving sound reconstruction.6 Delayed rehabilitation and associated injuries around the knee result in limitation of knee movement.20 In two studies with intraarticular extension of the fracture into the knee joint, average knee flexion of 791 and 961 was reported.10,20 Half of the patients in the study of Fraser et al.18 considered that their activity, in relation to work or sports, was still moderately or markedly reduced as a result of their injuries; this was regardless of the treatment group to which they belonged (closed or operative). Using multivariate analysis Yokoyama et al.22 found that apart from knee involvement and open Grade III femoral fractures, significant contributory factors affecting the functional outcome of floating knee injuries are: (1) (2) (3) (4) (5)

involvement of knee joint, severity of soft tissue injury in the tibia, fixation time after injury in the tibia, AO fracture grade in femur and tibia, and fixation time after injury in the femur and severity of open femoral fractures.

Conclusions Ipsilateral femoral and tibial fractures present following high energy trauma and they are accompanied by an increased risk of morbidity and mortality. Isolated fractures in stable patients could be treated acutely while in unstable or ‘‘in extremis’’ patients temporary fracture stabilisation with external fixation is indicated according to the concepts of Damage Control Orthopaedics (Table 2). When the patient’s physiological state has been stabilized conversion to internal fixation is desirable. Involvement of knee joint,

ARTICLE IN PRESS 410

Table 2

B. Chalidis et al.

Treatment protocol of floating knee injuries.

severe soft tissue trauma, fracture comminution and open fractures are associated with higher complication rates resulting in poorer functional outcome. In children, when the physis is involved the possibility of leg length discrepancy must be always taken into consideration.

References 1. Blake R, McBryde Jr A. The floating knee: ipsilateral fractures of the tibia and femur. South Med J 1975;68:13–6. 2. Veith RG, Winquist RA, Hansen Jr ST. Ipsilateral fractures of the femur and tibia: a report of fifty-seven consecutive cases. J Bone Joint Surg Am 1984;66:991–1002. 3. Behr JT, Apel DM, Pinzur MS, Dobozi WR, Behr MJ. Flexible intramedullary nails for ipsilateral femoral and tibial fractures. J Trauma 1987;27:1354–7. 4. Rios JA, Ho-Fung V, Ramirez N, Hernandez RA. Floating knee injuries treated with single-incision technique versus traditional antegrade femur fixation: a comparative study. Am J Orthop 2004;33(9):468–72. 5. Bohn WW, Durbin RA. Ipsilateral fractures of the femur and tibia in children and adolescents. J Bone Joint Surg Am 1991;73A: 429–39. 6. Yokoyama K, Nakamura T, Shindo M, Tsukamoto T, Saita Y, Aoki S, et al. Contributing factors influencing the functional outcome of floating knee injuries. Am J Orthop 2000;29(9):721–9. 7. Dwyer AJ, Paul R, Mam MK, Kumar A, Gosselin RA. Floating knee injuries: long-term results of four treatment methods. Int Orthop 2005;29(5):314–8. 8. Ostrum RF. Treatment of floating knee injuries through a single percutaneous approach. Clin Orthop 2000;375:43–50. 9. Baker SP, O’Neill B, Haddon Jr W, Long WB. The injury severity score: a method for describing patients with multiple injuries and evaluating emergency care. J Trauma 1974;14:187–96. 10. Adamson GJ, Wiss DA, Lowery GL, Peters CL. Type II floating knee: ipsilateral femoral and tibial fractures with intraarticular extension into the knee joint. J Orthop Trauma 1992;6:333–9.

11. Paul GR, Sawka MW, Whitelaw GP. Fractures of the ipsilateral femur and tibia: emphasis on intra-articular and soft tissue injury. J Orthop Trauma 1990;4:309–14. 12. Hee HT, Wong HP, Low YP, Myers L. Predictors of outcome of floating knee injuries in adults: 89 patients followed for 2–12 years. Acta Orthop Scand 2001;72(4):385–94. 13. Van Raay JJ, Raaymakers EL, Dupree HW. Knee ligament injuries combined with ipsilateral tibial and femoral diaphyseal fractures: the ‘‘floating knee’’. Arch Orthop Trauma Surg 1991;110(2):75–7. 14. Szalay MJ, Hosking OR, Annear P. Injury of knee ligament associated with ipsilateral femoral shaft fractures and with ipsilateral femoral and tibial shaft fractures. Injury 1990;21:398–400. 15. Yue JJ, Churchill RS, Cooperman DR, Yasko AW, Wilber JH, Thompson GH. The floating knee in the pediatric patient: nonoperative versus operative stabilization. Clin Orthop 2000;376:124–36. 16. Arslan H, Kapukaya A, Kesemenli C, Subasi M, Kayikci C. Floating knee in children. J Pediatr Orthop 2003;23(4):458–63. 17. Letts M, Vincent N. The ‘‘floating knee’’ in children. J Bone Joint Surg Br 1986;68:442–6. 18. Fraser RD, Hunter GA, Waddell JP. Ipsilateral fracture of the femur and tibia. J Bone Joint Surg Br 1978;60:510–5. 19. Karlstro ¨m G, Olerud S. Ipsilateral fracture of the femur and tibia. J Bone Joint Surg Am 1977;59:240–3. 20. Hung SH, Chen TB, Cheng YM, Cheng NJ, Lin SY. Concomitant fractures of the ipsilateral femur and tibia with intraarticular extension into the knee joint. J Trauma 2000;48(3): 547–51. 21. Cole PA, Zlowodzki M, Kregor PJ. Less Invasive Stabilization System (LISS) for fractures of the proximal tibia: indications, surgical technique and preliminary results of the UMC Clinical Trial. Injury 2003;34(Suppl 1):A16–29. 22. Yokoyama K, Tsukamoto T, Aoki S, Wakita R, Uchino M, Noumi T, et al. Evaluation of functional outcome of the floating knee injury using multivariate analysis. Arch Orthop Trauma Surg 2002;122(8):432–5.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 411–417

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

MINI-SYMPOSIUM: MANAGEMENT OF FRACTURES AROUND THE KNEE JOINT

(iii) Minimal invasive techniques in the management of tibial plateau fractures Ramakrishnan Venkatesh Leeds Teaching Hospitals NHS Trust, Great George Street, Leeds LS1 3EX, UK

KEYWORDS Minimal invasive surgery; Arthroscopic reduction and internal fixation (ARIF); Schatzker classification

Summary Intra-articular fractures of the proximal tibia present a wide spectrum of injury patterns with associated soft tissue injury. The last two decades have seen the techniques of management evolve from extensive open reduction and rigid internal fixation to arthroscopy-assisted minimal invasive surgery (MIS) and biologically benign internal fixation. The ultimate aim is to prevent the occurrence of late degenerative arthritis. This could be achieved in selected patients using minimal invasive surgery, which offers the advantages of better visualisation and management of intra-articular soft tissue injuries, confirmation of fracture reduction viewed from the joint surface, faster rehabilitation and fewer wound complications. & 2006 Elsevier Ltd. All rights reserved.

Introduction Tibial plateau fractures present a wide spectrum of injuries with a range of fracture patterns involving varying degrees of joint surface depression and displacement. The tibial articular surface slopes approximately 101 anterior to posterior and the lateral tibial articular surface is higher and dome shaped when compared to the medial surface. The lateral meniscus is broader and covers a larger portion of the lateral tibial plateau than does the medial. The majority of the fractures affect the lateral tibial plateau (55–70%) whilst 10–23% of reported series have isolated medial tibial plateau fractures.1 There is a high incidence of associated intra-articular soft tissue injuries Tel.: +44 113 3925647; fax: +44 113 3924585.

E-mail address: [email protected]. 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.11.003

with both. With the increasing use of MRI scans and arthroscopy after acute tibial fractures, meniscal tears are now being reported in 35–91% of fractures and cruciate ligament tears are reported in 15–77% of tibial plateau fractures.2–6 In Schatzker type II fractures, the presence of depression greater than 6 mm and fracture displacement greater than 5 mm is associated with an 83% incidence of lateral meniscal tears.7 The adequacy of fracture reduction and presence of soft tissue injuries relate to the long-term functional outcome and incidence of late arthritis.8–10 Good long-term results have been achieved with open reduction11 but there is a higher risk of imperfect reduction,12 wound complications and stiffness.13 Arthroscopy-assisted minimal invasive surgery has been shown to achieve similar results with fewer wound complications and faster rehabilitation.6,14–17 Surgical techniques have been facilitated by the development of specific arthroscopic instruments and implants (Fig. 1).

ARTICLE IN PRESS 412

R. Venkatesh displacement, tibial spine or tuberosity avulsions, fibular fracture, meniscal tears and ligament injuries. 151 caudal view radiographs or fluoroscopy images are helpful. The axial slices from a CT scan are more helpful in assessing the size of the depressed fragment and planning the placement of screws. Routine MRI scans for Schatzker types I–III are not necessary, as long as the surgeon performs an arthroscopy on all these knees. Midsubstance ACL injuries, even if diagnosed prior to surgery, are treated with rehabilitation and appropriate delayed reconstruction.

Classification The most widely used system is the Schatzker classification.19

Figure 1 The development of specific instruments, such as the core reamer and angle tamps (top—courtesy of Arthrex) and absorbable interference screws (bottom), has facilitated the development of arthroscopic-assisted fracture reduction and fixation.

Evaluation History: Understanding the mechanism of injury and the fracture personality is important in planning the management of these fractures. Fractures of the tibial plateau occur with varus or valgus forces coupled with strong axial loading in the range of 1600–8000 lb.18 The majority of lateral tibial plateau fractures are secondary to low to medium energy injuries, whilst medial plateau and bicondylar fractures happen with high-energy injuries. Examination: It is important to assess the soft tissue status, neurovascular injuries and the extensor mechanism. It is important to look for collateral ligament tenderness and gently assess stability. A lot of information is also obtained by examination under anaesthesia prior to surgery and also after fracture stabilisation.

Investigations In addition to knee radiographs, CT scans and/or MRI scans can offer valuable information about the injuries. The scans provide information about the extent of displacement of any split, the size and location of any depression, rim

(1) Type 1—Lateral split fracture—the bone quality is usually good and these fractures are amenable to percutaneous fixation. (2) Type II—Lateral split depression fracture—intra-articular soft tissue injuries are common. (3) Type III—Lateral isolated depressed fracture—usually lower energy. (4) Type IV—Medial plateau fracture—high-energy injury. Commonly associated with ligamentous and neurovascular injuries. (5) Type V—Bicondylar fracture—high-energy. Avulsion fractures and neurovascular injuries must be assessed. (6) Type VI—Bicondylar fracture with metaphyseal diaphyseal dissociation. Soft tissue envelope could be compromised.

Goals of surgery and treatment planning The non-operative treatment of these fractures is limited to patients who are very poor surgical candidates and those with a very small area of central depression, with less than 5 mm depression, and an intact rim. The aims of surgical treatment are to obtain anatomical reduction, restore the mechanical axis, achieve stable internal fixation and regain early joint movements.

The role of minimal invasive surgery Minimal invasive surgery plays a key role in almost all tibial plateau fractures. Arthroscopy-assisted surgery is useful in the most common fracture types I–III. In the high velocity injuries, arthroscopic surgery poses significant risks of fluid extravasation, hence minimal invasive fixation alone with screws and external fixators is favoured. Arthroscopy—Arthroscopy improves the visualisation of intra-articular soft tissue injuries (Fig. 2). It is synergistic with the C arm in confirming fracture reduction and guiding the placement of screws. Arthroscopy also facilitates fracture reduction, as it is easier to visualise and clear any soft tissue interposed at the fracture gap. Using the tibial guide, a guide wire can be placed at the centre of the depression. It is also technically easier to repair posterior meniscal tears with arthroscopy than using open techniques (Fig. 2).

ARTICLE IN PRESS Tibial plateau fractures

413

Figure 2 Case 1—top shows axial CT that is most useful for determining the size and location of the depressed fragment. Bottom shows arthroscopic images before (bottom 2 images) and after (top 2 images) reduction.

Fracture reduction—Corkscrew tip guide pins or threaded guide wires through small stab incisions can be used to joystick fracture fragments. Large pointed bone clamps are also useful for fracture reduction. After placing a guide wire at the base of the depressed fragment, the fracture can be elevated with combined fluoroscopy and arthroscopy control. Fixation—Percutaneous raft screws, antiglide plates and LISS plates can be used to fix the fracture. A bio-cortical interference screw could also be placed in the metaphyseal core to avoid any delayed subsidence of the bone graft.

fracture split, is necessary. The displaced split lateral plateau also produces lateral malalignment. If the overlying soft tissue envelope is not compromised, surgery is best performed early (24–72 h) to facilitate easier indirect reduction. The degree of osteopenia and extent of depression will help decide the need for bone grafting or the use of bone graft substitutes. Fixation can usually be achieved by using multiple raft screws7antiglide plate and for the more complex injuries using the LISS plate.

Surgical technique Preoperative planning A three-dimensional understanding of the fracture, including the size and amount of depression and orientation of the

Examination under anaesthesia: This is performed before the leg holder is applied in the anaesthetic room and also after the fracture is reduced and fixed.

ARTICLE IN PRESS 414 Theatre set-up: The patient is supine and the affected limb is placed in an arthroscopy leg holder with tourniquet in place. The hip is slightly abducted with the patella facing about 201 externally. This is to enable the C arm to clear the opposite leg when it comes from the ipsilateral side and swings underneath for a true lateral view. If autogenous graft is required then the iliac crest needs to be prepared. The arthroscopy monitor is placed on the contralateral side. Arthroscopy: Other than the high velocity injuries with significant soft tissue damage, or when there is a severely displaced split, it is preferable to start with a diagnostic arthroscopy. The use of the powered 4.5 mm synovator blade to clear any clots or haemarthrosis increases the speed of the surgery. The author initially starts with the standard anterolateral and anteromedial portals then subsequently swaps the arthroscope to the anteromedial portal. A further inframeniscal lateral portal could be used to obtain better

R. Venkatesh visualisation. The goldfinger meniscal retractor set (Arthrex, Inc., Florida) may be used. Soft tissue management: Posterior meniscal tears are easier to repair using the arthroscope with all-inside techniques. The more common anterior third meniscal tears are repaired through mini-open outside–in technique. Anterior cruciate avulsion fracture can also be reduced and fixed arthroscopically using either cannulated screws or a trans-tibial suture technique. MCL injuries are managed in a hinged knee brace. Fracture reduction:

(a) Type I—Articular gap is reduced using a large pointed reduction forceps placed perpendicular to the fracture line. If there is a step then the fragment is manipulated using a threaded guide wire or cork–screw tip guide pin (Arthrex, Inc., Florida).

Figure 3 (a) Case 2—CT images demonstrating depressed segment. (b) Case 2—arthroscopic views before and after reduction. (c) Case 2—after fixation.

ARTICLE IN PRESS Tibial plateau fractures

415

Figure 3

(b) Type II—Using the ACL tibial guide, a guide wire is placed just beneath the centre of the depressed fragment. A core reamer is used to take an 8–10 mm core of metaphyseal bone up to 5 mm beneath the fragment. Then this core along with bone graft/bone graft substitutes are gently tapped up using angled bone tamps to elevate the depressed fragment. Both arthroscopy and fluoroscopy help in confirming the reduction. The split fragment is then elevated and reduced as mentioned for Type I (Fig. 3a–c). (c) Type III—The technique is very similar to that used to elevate the depressed fragment mentioned above. Fracture fixation: Guide pins for cannulated raft screws are inserted lateral to medial, through stab incisions, perpendicular to the plane of fracture at the subchondral level. The bone tamp is held in position until the guide wires are in place. Multiple cannulated screws are used to achieve fixation in most circumstances. Once bone graft substitute has been packed into the cylindrical tunnel, an angled biocortical interference screw (Arthrex, Inc., Florida) is placed flush on to the tibial cortex to support the graft. To further support the split fracture, an antiglide screw with washer or plate could be used. The LISS plate could be used in more severe split depressed fractures, bicondylar fractures or if there is poor bone quality.

Postoperative management Pain relief is achieved through a femoral nerve block and intra-articular infiltration of local anaesthetic agent. In

(Continued)

most situations, the author uses continuous passive mobilisation over the first 48 h and the patients mobilise nonweight bearing with a hinged knee brace for the first 6 weeks. If there was more posterior depression or comminution, flexion is restricted to 901 but it is important to regain full extension as soon as possible. After 6 weeks the patients commence progressive weight bearing and should aim to achieve full flexion.

Results ARIF There are no prospective randomised studies comparing open versus arthroscopic techniques for management of tibial plateau fractures. Fowble12 retrospectively compared the two techniques for local compression and split compression fractures. The results were superior in the ARIF group with better anatomical reduction. There was also a significant reduction in hospital stay (5.36 vs. 10.27 days). There are many reports of ARIF providing anatomical reduction without long-term loss of fixation and excellent results at follow-up.6,14–17 Young patients with sporting injuries have also been reported to have 92% good to excellent results with ARIF.16 Eighty-four per cent of these patients returned to full sporting activity. Cassard15 performed a clinical and radiological assessment of 19 tibial plateau fractures treated arthroscopy assisted with an average follow-up of 32.7 months. The average knee society score was 94.1 for the knee and 94.7 for function. There was

ARTICLE IN PRESS 416 no secondary bony depression. They also noticed 8 meniscal injuries at arthroscopy. The lesions at injury (cartilage, menisci and ligaments) and the late condition (articular incongruity, axial abnormality and instability of the knee) influenced the functional score and the importance of late arthritis.8 Instability is a major cause of unacceptable results in tibial plateau fractures.10 MRI scans have shown an 80% incidence of meniscal tears and 40% complete ligament tears even in undisplaced tibial plateau fractures3 and in displaced fractures 91% meniscal tears and 77% complete ligament disruptions.2 Similarly arthroscopy for tibial plateau fractures has shown 57% meniscal injuries and 25% ACL injuries in 98 knees.4 Bennett5 reported 20% meniscal injuries and 10% ACL tears and found these commonly associated with Types II and IV injuries. Van Glabbeek6 found 35% meniscal lesions in 20 arthroscopies.

MIS fixation Limited internal fixation with percutaneous screws has been found to be effective in the fixation of tibial plateau fractures, with no reports of late subsidence.6,20,21 Simpson and Keating21 compared 13 tibial plateau fractures treated with open reduction and bone grafting with a similar group of well matched fractures treated with minimal internal fixation and injectable calcium phosphate bone cement. They found more favourable anatomical results in the MIS group at 1-year follow-up. A biomechanical comparison of various fixation methods for Type II fractures shows that 3.5 mm raft screws with an antiglide plate provided superior longitudinal and depression stiffness than the conventional large fragment buttress plates.22 The LISS plating system provides a minimally invasive fixed angle construct and Cole23 reports 89 tibial plateau fractures treated using this system achieving 97% stable fixation. This allows minimal invasive management of complex bicondylar tibial plateau fractures allowing early range of knee motion with favourable clinical results.24

Bone graft substitutes Biomechanical comparative studies have shown that bone graft substitutes can prevent fracture subsidence better than autograft.21,25,26 The treatment of centrally depressed tibia plateau fractures with calcium phosphate cement provides equivalent or better stability than conventional open reduction and internal fixation in pure depression tibial plateau fractures and provides the option of faster rehabilitation and earlier weight bearing.25

Complications Very few complications have been reported that are specific to arthroscopy-assisted minimal invasive surgery. There is a risk of fluid extravasation and compartment syndrome during arthroscopy,27 hence it is important to use gravity inflow and good outflow. This is more of a risk in the high velocity injuries. Cassard15 reported one case of septic arthritis.

R. Venkatesh

Summary Arthroscopy is a useful aid in the management of lower energy tibial plateau fractures. Meniscal tears and ligament injuries are commonly associated with tibial plateau fractures and the use of MRI scans and arthroscopy has increased their diagnosis. Understanding the fracture anatomy is a prerequisite to minimal invasive techniques. Theatre set-up is important when using both arthroscopy and fluoroscopy during surgery. Gravity inflow with adequate outflow is important during arthroscopy. The placement of a guide wire under the depressed fragment is performed with arthroscopy assistance whilst elevation of the depressed fragment is more fluoroscopy assisted, the arthroscope confirming the reduction. Screw placement has to be perpendicular to the fracture plane and both axial images of CT scans and the arthroscope aid during this step. Multiple percutaneous screws offer good fixation and depression stiffness.

Learning points: (1) Examination includes assessment under anaesthesia before and after fracture fixation (2) Displacement of the split greater than 5 mm is highly associated with lateral meniscal tears (3) During arthroscopy, use gravity inflow only to avoid extravasation of fluid and use an outflow cannula (4) Preoperative planning using axial images essential for correct screw placement

References 1. Hohl M. Part I. Fractures of the proximal tibia and fibula. In: Rockwood C, Green D, Bucholz R, editors. Fractures in adults. 3rd ed. Philadelphia: JB Lippincott; 1991. 2. Gardner MJ, Yacoubian S, Geller D, Suk M, Mintz D, Potter H, et al. The incidence of soft tissue injury in operative tibial plateau fractures: a magnetic resonance imaging analysis of 103 patients. J Orthop Trauma 2005;19(2):79–84. 3. Shepherd L, Abdollahi K, Lee J, Vangsness Jr. CT. The prevalence of soft tissue injuries in nonoperative tibial plateau fractures as determined by magnetic resonance imaging. J Orthop Trauma 2002;16(9):628–31. 4. Abdel-Hamid MZ, Chang CH, Chan YS, Lo YP, Huang JW, Hsu KY, et al. Arthroscopic evaluation of soft tissue injuries in tibial plateau fractures: retrospective analysis of 98 cases. Arthroscopy 2006;22(6):669–75. 5. Bennett WF, Browner B. Tibial plateau fractures: a study of associated soft tissue injuries. J Orthop Trauma 1994; 8(3):183–8. 6. Glabbeek FV, Van Reit R, Jansen N, D’Anvers J, Nuyts R. Arthroscopically assisted reduction and internal fixation of tibial plateau fractures: report of twenty cases. Acta Orthop Belg 2000;68(3):258–64. 7. Gardner MJ, Yacoubian S, Geller D, Pode M, Mintz D, Helfet DL, et al. Prediction of soft-tissue injuries in Schatzker II tibial plateau fractures based on measurements of plain radiographs. J Trauma 2006;60(2):319–23.

ARTICLE IN PRESS Tibial plateau fractures 8. Kohut M, Leyvraz PF. Cartilaginous, meniscal and ligamentous lesions in the prognosis of tibial plateau fractures. Acta Orthop Belg 1994;60:81–8. 9. Blokker CP, Rorabeck CH, Bourne RB. Tibial plateau fractures. An analysis of the results of treatment in 60 patients. Clin Orthop 1984;182:193–9. 10. Delamarter RB, Hohl M, Hopp Jr. E. Ligament injuries associated with tibial plateau fractures. Clin Orthop 1990;250: 226–33. 11. Stevens DG, Beharry R, Mckee MD, Waddell JP, Achemitsch EH. The long-term functional outcome of operatively treated tibial plateau fractures. J Orthop Trauma 2001;15(5):312–20. 12. Fowble CD, Zimmer JW, Schepsis AA. The role of arthroscopy in the assessment and treatment of tibial plateau fractures. Arthroscopy 1993;9(5):584–90. 13. Young MJ, Barrack RL. Complications of internal fixation of tibial plateau fractures. Orthop Rev 1994;23(2): 149–54. 14. Kiefer H, Zivaljevic N, Imbriglia JE. Arthroscopic reduction and internal fixation (ARIF) of lateral tibial plateau fractures. Knee Surg Traumatol Arthrosc 2001;9(3):167–72. 15. Cassard X, Beaufils P, Blin JL, Hardy P. Osteosynthesis under arthroscopy control of separated tibial plateau fractures. 26 case reports. Rev Chir Orthop Reparatrice Appar Mot 1999;85(3):257–66. 16. Gill TJ, Moezzi DM, Oates KM, Sterett WI. Arthroscopic reduction and internal fixation of tibial plateau fractures in skiing. Clin Orthop 2001;383:243–9. 17. Hung SS, Chao EK, Chan YS, Yuan LJ, Chung PC, Chen CY, et al. Arthroscopically assisted osteosynthesis for tibial plateau fractures. J Trauma 2003;54(2):356–63. 18. Kennedy JC, Bailey WH. Experimental tibial plateau fractures. J Bone Jt Surg Am 1968;50:1522.

417 19. Schatzker J, McBroom R, Bruce D. The tibial plateau fracture: the Toronto experience 1968–1975. Clin Orthop 1979; 138:94–104. 20. Duwelius PJ, Rangitsch MR, Colville MR, Woll TS. Treatment of tibial plateau fractures by limited internal fixation. Clin Orthop 1997;339:47–57. 21. Simpson D, Keating JF. Outcome of tibial plateau fractures managed with calcium phosphate cement. Injury 2004;35(9): 913–8. 22. Karunakar MA, Egol KA, Peindl R, Harrow ME, Bosse MJ, Kellam JF. Split depression tibial plateau fractures: a biomechanical study. Journal of Orthopaedic Trauma 2002;16(3):172–7. 23. Cole PA, Zlowodzki M, Kregor PJ. Treatment of proximal tibia fractures using the less invasive stabilization system: surgical experience and early clinical results in 77 fractures. J Orthop Trauma 2004;18(8):528–35. 24. Egol KA, Su E, Tejwani NC, Sims SH, Kummer FJ, Koval KJ. Treatment of complex tibial plateau fractures using the less invasive stabilization system plate: clinical experience and a laboratory comparison with double plating. J Trauma 2004; 57(2):340–6. 25. Yetkinler DN, McClellan RT, Reindel ES, Carter D, Poser RD. Biomechanical comparison of conventional open reduction and internal fixation versus calcium phosphate cement fixation of a central depressed tibial plateau fracture. J Orthop Trauma 2001;15(3):197–206. 26. Welch RD, Zhang H, Bronson DG. Experimental tibial plateau fractures augmented with calcium phosphate cement or autologous bone graft. J Bone Jt Surg Am 2003;85A(2):222–31. 27. Belanger M, Fadale P. Compartment syndrome of the leg after arthroscopic examination of a tibial plateau fracture. Case report and review of the literature. Arthroscopy 1997; 13(5):646–51.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 418–423

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

SYNDROMES

Marfan syndrome Andrew R.T. McBridea, Martin Garganb, a

Department of Trauma & Orthopaedic Surgery, Weston General Hospital, Grange Road, Weston-super-Mare BS23 4TQ, UK Department of Trauma & Orthopaedic Surgery, Bristol Royal Hospital for Sick Children, Bristol BS2 8HW, UK

b

KEYWORDS Marfan syndrome; Connective tissue disease

Summary Marfan syndrome is an inherited connective tissue disorder with multi-organ system involvement caused by mutations in the gene encoding the glycoprotein fibrillin-1 (FBN1). The condition presents to the orthopaedic surgeon with an array of musculoskeletal problems. This article will review the pathogenesis, diagnosis, clinical manifestations and treatment options. & 2006 Elsevier Ltd. All rights reserved.

Introduction

Antoine Bernard-Jean Marfan

Marfan syndrome is an autosomal dominant inherited disorder of connective tissue. Clinical manifestations are varied in range and severity and involve many organs. The majority of anomalies are found in the cardiovascular, respiratory, ocular and skeletal systems. Advances in the medical and surgical treatment of life threatening cardiovascular complications of the condition now enable more individuals to expect an improved life expectancy.1 However this longevity is associated with an increased frequency of musculoskeletal problems. Specific musculoskeletal anomalies may be found in the spine, hip, knee, hand, and foot. Generalised joint laxity and an increased incidence of joint dislocation also feature. Arthralgia is a common complaint in all age groups.

Antoine Bernard-Jean Marfan was born in Castlenaudary, Southern France in 1858. He completed his medical training in Paris in 1886. A career in paediatrics followed, culminating in a professorship at the University of Paris School of Medicine. In 1896, Marfan presented the case of a 5-year-old girl to the Socie ´te´ Me´dicale des Ho ˆpitaux de Paris. He described her disproportionately long limbs and her fingers as ‘‘pattes d’araigne´e’’ (spider fingers). He called her condition ‘‘dolichoste´nome ´lie’’ (slender limbs). Retrospective analysis has shown this was probably a case of Beal’s syndrome. However Marfan’s description of over one hundred similar cases led to the discovery of the disorder named after him.2

Corresponding author. Tel.: +44 117 928 2121; fax: +44 117 928 2659. E-mail addresses: [email protected] (A.R.T. McBride), [email protected] (M. Gargan).

0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.09.007

Epidemiology and genetics The incidence of the disorder is approximately 1 in 5000, sporadic mutations making up a quarter of all cases. No particular racial or geographical factors are recognised features.2 The condition is inherited in an autosomal dominant pattern with a high degree of penetrance but variable

ARTICLE IN PRESS Marfan syndrome

Table 1

419

Revised (Ghent) diagnostic criteria.5

Major criteria Skeletal system  pectus carinatum  pectus excavatum requiring surgery  reduced upper to lower segment ratio or arm span to height ratio 41.05  positive wrist and thumb signs  scoliosis of 4201 or spondylolisthesis  reduced extension of the elbows (o1701)  medial displacement of the medial malleolus causing pes planus  protrusio acetabuli of any degree

Minor criteria

   

pectus excavatum of moderate severity joint hypermobility high arched palate facial appearance (dolichocephaly, malar hypoplasia, enophthalmos, retrognathia, down-slanting palpebral fissures)

At least two components of the major criteria or one major criterion and two minor criteria must be present for the skeletal system to be considered in the diagnosis. Ocular system  ectopia lentis

 abnormally flat cornea (as measured by keratometry)

 increased axial length of globe (as measured by ultrasound)

 hypoplastic iris or hypoplastic ciliary muscle causing decreased miosis At least two minor criteria must be present for the ocular system to be considered in the diagnosis. Cardiovascular system  dilatation of the ascending aorta with or without aortic regurgitation and involving at least the sinuses of Valsalva; or  dissection of the ascending aorta

 mitral valve prolapse with or without mitral valve regurgitation

 dilatation of the main pulmonary artery in the absence of valvular or peripheral pulmonic stenosis or any other obvious cause, below the age of 40 years  calcification of the mitral annulus below the age of 40 years; or  dilatation or dissection of the descending thoracic or abdominal aorta below the age of 50 years

One major criterion or one minor criterion must be present for the cardiovascular system to be considered in the diagnosis. Pulmonary system  none

 spontaneous pneumothorax; or  apical blebs (ascertained by chest radiography)

One minor criterion must be present for the pulmonary system to be considered in the diagnosis. Skin and integument  none

 striae atrophicae (stretch marks) not associated with marked weight changes, pregnancy or repetitive stress, or  recurrent or incisional herniae

One minor criterion must be present for the skin and integument to be considered in the diagnosis. Dura  lumbosacral dural ectasia (CT or MRI)

 none

The major criterion must be present for dura to be considered in the diagnosis.

ARTICLE IN PRESS 420

A.R.T. McBride, M. Gargan

Table 1 (continued ) Major criteria

Minor criteria

Family/genetic history  having a parent, child or sibling who meets these criteria independently  presence of a mutation in FBN1 known to cause the Marfan syndrome; or  presence of a haplotype around FBN1, inherited by descent, known to be associated with unequivocally diagnosed Marfan syndrome in the family

 none

One major criterion must be present for the family/genetic history to be considered in the diagnosis.

Figure 1 Arachnodactyly.

phenotypical expression. Mutations in the gene encoding the glycoprotein fibrillin-1 (FBN1) situated on chromosome 15q21 cause Marfan syndrome.3 Despite the discovery and genetic mapping of the FBN1 gene, its large size and a lack of mutational hot spots have prevented the development of a single and therefore clinically useful genetic test.1 FBN1 is an important component of extracellular microfibrils. The exact role of microfibrils is unclear but they are found in a wide range of connective tissues and are vital for normal elastogenesis, elasticity and homoeostasis of elastic fibres. The molecular pathogenesis of Marfan syndrome has yet to be accurately defined. Mutated FBN1 proteins may interfere with the correct assembly of mature microfibrils or the defective proteins may increase their susceptibility to proteolysis.4

unfortunately late diagnosis is not uncommon. Due to the lack of a molecular diagnostic test, diagnosis depends upon the revised (Ghent) clinical criteria5 (Table 1). The large variation in phenotypical expression of clinical features can make diagnosis difficult. Large variation can even be encountered between affected individuals of the same family. Overlap of clinical features with other Marfanlike disorders also requires care when diagnosing a new case of Marfan syndrome. The differential diagnosis includes homocystinuria, Beal’s syndrome (congenital contractual arachnodactyly), Ehlers–Danlos syndrome, Stickler syndrome (hereditary arthro-ophthalmopathy), Klinefelter syndrome, familial mitral valve prolapse syndrome, multiple endocrine adenomatosis and X-linked mental retardation with marfanoid habitus.

Skeletal features Clinical diagnosis The progressive and potentially fatal clinical features of Marfan syndrome make early diagnosis very important;

The typical body habitus in the Marfan syndrome comprises tall stature, with disproportionately long, slender limbs compared with the trunk (dolichostenomelia). At birth the

ARTICLE IN PRESS Marfan syndrome

421

Figure 2 ‘‘Thumb sign’’. Figure 4

Figure 3

‘‘Wrist sign’’.

Figure 5

Scoliosis.

Dural ectasia (MRI scan).

ARTICLE IN PRESS 422

A.R.T. McBride, M. Gargan

Marfan infant is longer than the normal infant; this difference remains through childhood into adult life. Growth rates compared to a normal population are increased in childhood. The puberty-associated growth spurt and skeletal maturation occur early.6 The upper body to lower body segment ratio is reduced (o0.85), this may be exaggerated by spinal deformity. Arachnodactyly is a non-specific feature (Fig. 1). However, the thumb sign (positive if the thumb nail, when the thumb is maximally opposed within a clenched fist, projects beyond the ulnar border of the hand) (Fig. 2), or the wrist sign (overlap of the thumb on the terminal phalanx of the little finger when held around the contralateral wrist) (Fig. 3), feature as major skeletal criteria in the revised (Ghent) diagnostic system. Pectus excavatum (depression) or carinatum (protrusion) deformities of the anterior chest wall due to longitudinal overgrowth of the ribs occur, often progressing and evolving with age. The facial skeleton is affected, including a high-arched palate, a narrow jaw (resulting in crowded dentition) and retrognathia leading to malocclusion and obstructive sleep apnoea. Abnormal FBN1 may affect calcium binding in bone; however the investigations of lowered bone mineral density by DEXA analysis have been inconclusive so far. Whether or not the improving life expectancy of the Marfan population will render individuals more susceptible to later onset osteoporosis remains to be seen.7

infancy tends to progress relentlessly. Around half of individuals will have a kyphosis of more than 501.8 The prevalence of a lumbosacral spondylolisthesis is only slightly higher in the young Marfan population (6%) compared to the general population (3%). However the degree of forward slip, when it is seen, is probably much greater (30%) than in the normal population (15%).8 The frequency of back pain in the Marfan population is greater than in an age-matched normal population (age range 35–45 years: 53% vs. 13%, p ¼ 0.001). Functional impairment due to pain is also significantly greater (p ¼ 0.001).8 Atlantoaxial rotatory instability has also been reported.9 Dural ectasia, a dilatation of the dural sac or nerve root sleeve, occurs in up to 90% of individuals, usually at the L5-S1 level (Fig. 5). The severity can be graded 1–3 with most patients having a grade 2 (moderate) dilatation.10 Dural ectasia occurs as a result of the gradual pulsatile effect of CSF in combination with gravity upon the abnormally composed dura mater of the dural sac. The expansion of the dural sac can cause erosion and scalloping of the vertebral bodies and the development of a meningocele. Mild cases are usually asymptomatic, but moderate and severe cases can be associated with, the symptoms of headache, low back pain, radicular pain radiating into the buttocks and legs and rarely urinary symptoms or motor weakness.11

The lower limb

Specific orthopaedic problems Spine Spinal deformity is one of the most common skeletal manifestations encountered. Between 50% and 60% of individuals have a scoliosis at one or more vertebral levels8 (Fig. 4). The prevalence between the sexes is equal. Approximately one-fifth of these will have a severe deformity. Deformity usually worsens during the rapid growth of early adolescence so that scoliosis detected in

Figure 6

Developmental dysplasia of the hip is probably increased but does not represent a common clinical problem.12 Protrusio acetabuli is common affecting up to 77% of individuals.13 Genu valgum can occur late in childhood. The foot is long and narrow with arachnodactyly. Joint laxity in the foot leads to a number of problems. Severe deformities left untreated can lead to significant discomfort and gait disturbances. Pes planus (with or without calcaneoplano valgus) affects 25% of individuals (Fig. 6), hallux valgus is also common.14

Pes planus.

ARTICLE IN PRESS Marfan syndrome

Other joint problems Joint laxity is frequently encountered, with hyperextensible fingers, wrists, elbows and knees. Some individuals have joint contractures, usually affecting the fingers and elbow. These can co-exist with lax joints in a proportion of patients. Recurrent dislocations can occur, most commonly affecting the first metacarpophalangeal joints and patellae. Ligamentous sprains are commonly encountered. Arthalgia is a common complaint in all age groups and the older Marfan patient commonly has arthritis in those previously lax joints.15

423

 If a mutation known to cause Marfan syndrome in others is detected, a major criterion in an organ system and involvement of another system must be present. For a relative of an index case:

 Presence of a major criterion from the family/genetic history and from one organ system with involvement of a second.

Orthopaedic management References A multi-disciplinary team approach is advised. Close cooperation with a geneticist, paediatrician and cardiologist is required because of the life threatening complications that can occur. Early diagnosis and treatment of deformity can slow progression and improve function. Spinal complications may require early surgical intervention including fusion and instrumentation in childhood or adolescence depending on the severity of deformity, symptoms and age of the patient. Bracing is generally unsuccessful because of the severity or rigidity of curves or the poor maintenance of correction obtained.16 Protrusio acetabuli in childhood should be closely followed with serial radiographs. Early surgical intervention is recommended with closure of the triradiate cartilage in patients with significant radiographic progression before waiting for the onset of pain or stiffness.13 Foot problems are common but can usually be treated with orthotics, a triple fusion maybe rarely be required for the pes planovalgus foot to improve pain relief, stability and function. Severe genu valgum in late childhood can be treated with hemiepiphyseal stapling.

Pre-operative considerations Cardiovascular assessment is mandatory before embarking on any major orthopaedic surgery because of the threat to life from mitral valve prolapse and aortic dilatation. Blood pressure is usually maintained below normal with antihypertensive agents. Antibiotic prophylaxis is indicated before the majority of surgical procedures. Anti-coagulation for prosthetic heart valves may have to be negotiated. Individuals are at risk from spontaneous pneumothoraces, usually arising from an apical bulla, which may complicate an already compromised pulmonary reserve from spinal or chest wall deformity. Spinal surgery requires careful planning and the judicious use of pre-operative CT and MRI scans to assess pedicle adequacy and the presence of dural ectasia. Large blood losses should be anticipated and staged procedures considered when possible.17 Requirements for diagnosis of Marfan syndrome For an index case:

 If the family/genetic history is not contributory, at least two different organ systems and involvement of a third system must be present.

1. Silverman DI, Burton KJ, Gray J, Bosner MS, Kouchoukos NT, Roman MJ, et al. Life expectancy in the Marfan syndrome. Am J Cardiol 1995;75:157–60. 2. Ho NCY, Tran JR, Bektas A. Marfan’s syndrome. Lancet 2005 [published online August 2, 2005]. 3. Dietz HC, Cutting GR, Pyeritz RE, Maslen CL, Sakai LY, Corsen GM, et al. Marfan syndrome caused by a recurrent de novo missense mutation in the fibrillin gene. Nature 1991;352: 337–9. 4. Robinson PN, Booms P. The molecular pathogenesis of the Marfan syndrome. Cell Mol Life Sci 2001;58:1698–707. 5. De Paepe A, Devereux RB, Dietz HC, Hennekam RCM, Pyeritz RE. Revised diagnostic criteria for the Marfan syndrome. Am J Hum Genet 1996;62:417–26. 6. Erkula G, Jones KB, Sponseller PD, Dietz HC, Pyeritz RE. Growth and maturation in Marfan syndrome. Am J Med Genet 2002;109:100–15. 7. Giampietro PF, Peterson M, Schneider R, Davis JG, Raggio C, Myers E, et al. Assessment of bone mineral density in adults and children with Marfan syndrome. Osteoporos Int 2003;14: 559–63. 8. Sponseller PD, Hobbs W, Riley LH, Pyeritz RE. The thoracolumbar spine in Marfan syndrome. J Bone Jt Surg 1995;77A: 867–76. 9. Herzaka A, Sponseller PD, Pyeritz RE. Atlantoaxial rotatory subluxation in patients with Marfan syndrome. Spine 2000;25: 524–6. 10. Fattori R, Nienaber CA, Descovich B, Ambrosetto P, Reggiani LB, Pepe G, et al. Importance of dural ectasia in phenotypic assessment of Marfan’s syndrome. Lancet 1999;354: 910–3. 11. Foran JRH, Pyeritz RE, Dietz HC, Sponseller PD. Characterization of the symptoms associated with dural ectasia in the Marfan patient. Am J Med Genet 2005;134A:58–65. 12. Sponseller PD, Tomek IM, Pyeritz RE. Developmental dysplasia of the hip in the Marfan syndrome. Br J Paediatr Orthoped 1997;6:255–9. 13. van de Velde S, Fillman R, Yandow S. Protrusio acetabuli in Marfan syndrome. Indication for surgery in skeletally immature Marfan patients. J Pediatr Orthoped 2005;25:603–6. 14. Lindsey JM, Michelson JD, MacWilliams BA. The foot in Marfan syndrome. Clinical findings and weight-distribution patterns. J Paediatr Orthoped 1998;18:755–9. 15. Grahame R, Pyeritz RE. Marfan syndrome: joint and skin manifestations are prevalent and correlated. Br J Rheumatol 1995;34:126–31. 16. Robins PR, Moe JH, Winter RB. Scoliosis in Marfan’s syndrome. J Bone Jt Surg 1975;57A:358–68. 17. Jones KB, Erkula G, Sponseller PD, Dormans JP. Spine deformity correction in Marfan syndrome. Spine 2002;27:2003–12.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 424–429

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

BONE PHYSIOLOGY

Bone morphogenetic proteins in orthopaedic surgery A. Cheung, A.M. Phillips Limb Reconstruction Unit, Department of Trauma and Orthopaedics, King’s College Hospital, Denmark Hill, London SE5 9RS, UK

KEYWORDS Bone morphogenetic protein; Bone healing; Orthopaedic; Surgery; Fracture healing; Nonunion

Summary Bone morphogenetic proteins (BMPs) are growth factors which induce new bone formation. They are an increasingly important adjunct in the treatment of certain musculoskeletal disorders. Their underlying basic science and role in bone healing is explained. Delivery systems, safety issues and current evidence-based clinical applications of BMPs in orthopaedic surgery are described. & 2006 Published by Elsevier Ltd.

Introduction In 1965 Urist reported that intramuscular injection of demineralised bone matrix elicited new bone formation.1 This led to research into a family of growth factors known as ‘‘bone morphogenetic proteins’’ (BMPs).2 These are proteins that act as cellular signaling agents and BMPs play a crucial role in cell signaling in bone growth and healing. When they bind to their mesenchymal cell surface receptors, they activate a signaling cascade to the cell nucleus. Genes are then expressed leading to synthesis of macromolecules involved in bone/cartilage repair, and the mesenchymal cell differentiates towards the chondrocyte or osteoblast phenotype.3 BMPs are of interest to the orthopaedic surgeon for many reasons. In the treatment of tibial nonunion, BMPs have been shown to be equally effective as allograft, possibly obviating the need to harvest allograft.4 BMPs are also

becoming an important adjunct in treatment of segmental defects,5,6 fresh fractures,7 cartilage8,9 and intervertebral disc repair,10 allograft incorporation,11,12 impaction bone grafting,13 spinal fusion,14 acceleration of regenerate ossification in callotasis,15 and enhancement of tendon incorporation in ligament reconstruction.16 Hence orthopaedic surgeons should understand the basic science to utilise their full therapeutic potential.

Osteoinduction, osteogenesis and osteoconduction

 Osteoinduction implies the recruitment of immature cells

 

Corresponding author. Tel.: +44 2077 374000;

fax: +44 2077 3463445. E-mail address: [email protected] (A.M. Phillips). 0268-0890/$ - see front matter & 2006 Published by Elsevier Ltd. doi:10.1016/j.cuor.2006.09.003

and the stimulation of these cells to develop into preosteoblasts,17 leading to de novo bone formation in—usually—mesenchymal tissue. Osteogenesis is new bone formation from osteocompetent cells in connective tissue or cartilage. Osteoconduction is the process of bone formation on a three-dimensional implant or graft through ingrowth of capillaries, mesenchymal tissue and osteoprogenitor cells from the recipient host.

ARTICLE IN PRESS Bone morphogenetic proteins in orthopaedic surgery

Basic science Osteoinduction was initially thought to be due to one protein in the bone matrix, i.e. ‘‘BMP’’. However more than 40 BMPs have been identified to date. Most are distinguished by their potential to induce new bone formation.18,19 Of these, recombinant human (rh) BMP-2 and rhBMP-7 (also known as Osteogenic Protein-1 or OP-1) appear to be the most effective, and are the only ones that have currently been developed for clinical use. BMPs are part of the transforming growth factor-b (TGF-(3)) superfamily, a group of growth factors that play an important role during embryogenesis and postnatal tissue repair. Within this group, BMPs appear to be the most selective for, and have the greatest effect upon, osteogenesis.18 They are dimeric molecules with two polypeptide chains of over 400 amino acids linked by a single disulphide bond and on X-ray crystallography exhibit a characteristic cysteine knot. Genes expressing BMPs 2, 7 and 14 are located on chromosome 20.3,20

BMP in the bone morphogenesis cascade Expression of BMPs is a complex intra- and extracellular signaling process which is variable throughout the bone morphogenesis cascade. BMPs 2 and 4 are expressed early by primitive mesenchymal stem cells, and throughout the cascade. BMP-2, 6, and 9 are involved at the early stage of differentiation of mesenchymal progenitor cells to preosteoblasts. BMP-7 is expressed by osteogenic cells from day 7, peaking at 2–4 weeks. BMPs 2, 4, 6, 7 and 9 increase osteocalcin expression and alkaline phosphatase expression in pre-osteoblasts, leading to mineralisation.21 BMP-3 has been shown to be osteoinductive, but may also be inhibitory in the presence of BMP-2 and 7, acting as a negative regulator of bone formation.22 Animal and human studies of BMPs demonstrate a typical sequence of events in bone morphogenesis; chemotaxis and mitosis of mesenchymal cells, differentiation of these cells into cartilage, and replacement of cartilage with bone. What initiates this cascade is unclear, but it has been hypothesised that plasma fibronectin bonds to implanted demineralised matrix, facilitating recruitment and proliferation of mesenchymal cells (maximal on day 3).23 The mesenchymal cells differentiate into chondrocytes (days 5–8) which then hypertrophy. Calcification of the cartilage matrix ensues (day 9). Angiogenesis and vascular invasion occurs simultaneously with differentiation of osteoblasts and bone formation (days 10–11). Finally the newly formed woven bone remodels into trabecular and cortical bone and bone marrow formation occurs. In a callus this remodelling is not homogeneous and depends on ‘zonal ossification’—broadly speaking the medullary and inter-cortical areas create ‘soft callus’ and enchondral ossification occurs, whilst the subperiosteal area and soft tissues surrounding the fracture form ‘hard callus’ and bone through intramembranous ossification.24

BMP receptors and Smads Only certain cell types are able to respond to BMPs. Their actions are mediated via specific BMP receptors that

425 activate intracellular messenger proteins called Smads.a,25 The activated Smads are then translocated to the cell nucleus, activating BMP responsive genes, resulting in transcription of macromolecules involved in bone and cartilage formation.

Clinical applications of BMPs The best evidence for the clinical use of BMPs is from randomised, controlled trials involving BMP-2 and BMP-7, which show that BMP-7 is at least as effective as autograft in treatment of nonunion. BMP-2 enhances healing of open tibial fractures, and BMP-2 is at least as effective as autograft used in anterior interbody spinal fusions in degenerative disc disease. Other BMPs have not been subject to this level of evidence testing, nor have other growth factors. Evidence for other clinical applications is taken from smaller uncontrolled observational trials, and animal models.26

Potential applications of BMPs in bone Treatment of non-union In 2001, a prospective randomised controlled multicentre trial compared rhBMP-7 (also known as OP-1) with fresh bone autograft in 122 patients with 124 tibial nonunions.4 At 9 months, 81% of the rhBMP-7 group and 85% of the autograft group achieved clinical union, and 75% of the rhBMP-7 group and 84% of the autograft group were deemed to have united radiologically. There was no statistical significance in outcome between the two groups at 2 years. The authors concluded that rhBMP-7 was equally effective to autograft in the treatment of tibial nonunion. However 20% of those treated with autograft experienced chronic donor site pain, and the incidence of postoperative osteomyelitis was significantly greater in the autograft treated group (21%) compared to the rhBMP-7 treated group (3%). On the basis of these data, the United States Food and Drug Administration (FDA) licensed rhBMP-7 for use as an alternative treatment to autograft in long bone nonunions. An observational, non-randomised study assessed clinical application of rhBMP-7 in 653 patients treated for a variety of conditions across the UK. Of these patients, 395 (60.5%) had a fracture nonunion. Mean time from initial injury to BMP-7 application in this nonunion group was 19.7 months, and mean number of procedures perfumed previously was 3.5. The authors showed an overall success rate of 82% clinically and radiologically using rhBMP-7 in a wide variety of clinical scenarios.27 An observational study of treatment of femoral nonunion using allograft and partially purified human BMP demonstrated that 24 out of 30 femoral nonunions healed at an average of 6 months (follow up 55 months). In 26 of these a

The term ‘Smad’ originated from identification of two genes involved in cell signaling. ‘Mad’ (mothers against decapentaplegic) is a gene involved in a signaling pathway in Drosophila, identified in 1997.25 A similar gene, ‘Sma’, was identified in Caenorhabditis elegans, and both genes are involved in signal transduction pathways activated by BMPs.

ARTICLE IN PRESS 426 femurs, a one stage lengthening technique was used to correct shortening deformity.28

Healing of fresh fractures In 2002, the BESST study group7 compared treatment of 450 patients with open tibial fractures using intramedullary nail fixation alone, or with 6 mg (0.75 mg/mL) or 12 mg (1.5 mg/ mL) of rhBMP-2 in a prospective, controlled, randomised trial. Randomisation was stratified according to the severity of the open fracture. The main outcome measure was whether a patient required secondary intervention within 12 months due to delayed union or nonunion. Ninety-four percent (n ¼ 421) of patients were followed up at 12 months. The group receiving the higher dose of rhBMP-2 (12 mg) had a significantly greater rate of healing, fewer infections and invasive interventions and lower rate of nonunion compared to the control group.

Autograft substitute in spinal surgery RhBMP-2 was licensed for use by the United States FDA for single-level interbody fusions of the lumbar spine in 2001. This decision was based upon data from a prospective, randomised, non-blinded multicentre study of 279 patients with degenerative lumbar disc disease undergoing anterior interbody fusion with use of two tapered threaded fusion cages.14 The control group received autograft (n ¼ 136) whilst the treatment group (n ¼ 143) received rhBMP-2 on a collagen sponge. Radiologically fusion rates were higher for the rhBMP-2 group compared to the autograft group at 24 months (94.5% vs. 88.7%) while clinically outcome was similar in both groups. More recently a prospective, randomised, multicentre study of 131 patients compared the use of rhBMP2 with allograft cortical struts (ACS) (n ¼ 79) against autograft (n ¼ 52) in anterior lumbar spinal arthrodesis.29 Radiologically the patients receiving rhBMP-2/ACS achieved significantly higher rates of fusion, and improved functional disability scores compared to the autograft group. The authors concluded that rhBMP-2/ACS is equally effective as autograft and avoids the morbidity associated with autograft harvest. Pilot clinical trials have all taken place comparing use of rhBMP-2 to autograft in posterior lumbar interbody fusion, anterior cervical discectomy and fusion and posterolateral lumbar fusion, and await further study.30

Healing of segmental/critical-sized defects (CSDs) A CSD is the smallest sized defect in a bone, which will not heal spontaneously. CSDs have been established in animal models using long bone, cranial and mandibular defects. Studies of healing of CSDs have been carried out using BMP-2 in the rat femur and BMP-7 in the canine ulna with promising results.5,6 However direct comparisons of humans with animal models should be made with caution.

A. Cheung, A.M. Phillips

Impaction grafting/ primary and revision arthroplasty Several observational studies in animal models have evaluated the effect of BMPs upon osteoconduction. In a canine model, uncemented total hip arthroplasties were performed on 15 dogs.31 A 2 mm deep artificial defect was created behind the acetabular component to mimic bone loss. Five dogs received rhBMP-2 in a carrier (aBSM), five dogs received aBSM alone, and five were controls. At 12 weeks the control group showed no bone formation in the defect, whereas the rhBMP-2 group showed bone formation and filling of the defect. In another study using a canine model (n ¼ 28), 3 mm defects were created between a titanium implant and the proximal humerus.13 These defects were treated with rhBMP-2 alone, rhTGF-(b2 alone, both rhBMP-2 and rhTGFb2, or autogenous grafting. The greatest implant strength was obtained using combined rhBMP-2 and rhTGF-(b2 and was equivalent to autogenous grafting alone.

Treatment of congenital pseudo-arthosis of the tibia Congenital pseudo-arthrosis of the tibia is a rare disorder which is difficult to treat and often has an unsatisfactory outcome. In 2006 an isolated case report described usage of BMP-7 in treatment of this disorder in a 13-year-old child. He had been treated unsuccessfully with nine previous operations. Intramedullary nailing was performed using autograft and 3.5 mg of BMP-7. Bony union was noted at 5 weeks, and full weight bearing allowed at 5 months. Fracture union did not occur until 28 months, and despite a 5 cm leg length discrepancy, he and his family were reportedly satisfied with the result.32 A case series of five children with congenital pseudoarthrosis of the tibia treated with rhBMP-7 (3.5 mg combined with allograft) was also reported in 2006. There was evidence of radiographic union in only one out of the five cases (follow up 12–18 months). The authors concluded that use of BMPs alone may not be enough to overcome the poor healing environment associated with congenital pseudoarthrosis of the tibia.33

Treatment of avascular necrosis of the femoral head A retrospective analysis was performed on 15 patients involving 17 hips with avascular necrosis of the femoral head.34 Fifteen of the hips were classified at Ficat Stage IIA. Treatment consisted of core decompression combined with allograft and 50 mg of partially purified BMP. Average follow up was 53 months. Fourteen of the 15 Ficat Stage IIA hips treated were judged clinically and radiologically to be a success.

Potential applications of BMPs in musculoskeletal tissue Cartilage repair In vitro and in vivo studies have demonstrated that BMPs are required for chondrocyte differentiation and regulation of

ARTICLE IN PRESS Bone morphogenetic proteins in orthopaedic surgery SOX proteins required for type II collagen expression during chondrogenesis.35 In a rat model, BMP-4 delivered locally through genetically engineered muscle-derived stem cells (MDSCs) has been shown to enhance chondrogenesis and improve repair of articular cartilage.8 In a canine model rhBMP-7-induced hyaline cartilage like repair of full-thickness osteochondral defects.9

Degenerative intervertebral disc repair A number of studies in different animal models have evaluated the role of BMPs, in particular BMP-7, in treatment of degenerative disc disease. Injection of BMP-7 into iatrogenically compressed rat intravertebral disc has been shown histologically to enhance discal extracellular matrix and clinically to inhibit pain-related behaviour.10 Disc degeneration was iatrogenically created in a rabbit model (n ¼ 90) by puncture of intervertebral discs with an 18 gauge needle. Radiographic evaluation showed induction of disc degeneration at 4 weeks. Following injection with BMP-7 into the disc, histological and radiographic evaluation showed significant improvement in disc height at 6 weeks compared to a control group, which remained until end of follow up at 24 weeks.36

BMP specific antagonists Over expression of BMP pathways has been implicated in heterotopic bone formation in fibrodyplasia ossificans progressiva (FOP), for example in the posterior longitudinal ligament of the spine.37 BMP specific antagonists exist (e.g. ‘noggin’, ‘gremlin’) which bind to BMPs themselves.38,39 In a mouse model, under-expression of noggin, the extracellular BMP antagonist, has been shown to result in heterotopic ossification in FOP in vivo.40 It has been hypothesised that BMP specific antagonists may be of therapeutic value in treatment of pathologies with excessive bone formation.

BMPs are pleiotropic BMPs also have actions upon development of nonmusculoskeletal tissue. Knockout of BMP genes in mice have resulted in developmental defects in the heart (BMP-2), eye and kidney (BMP-7).41,42

Delivery When the osteoinductivity of individual recombinant human BMPs (rhBMPs) was discovered, it was found that this follows a dose-response ratio43 and that the dose of BMPs must reach a threshold value before osteoinduction can occur. It followed that local concentration of BMPs was the most important factor, more than patient characteristics or total dose. As systemic clearance of BMPs is high, local application of BMPs is essential. From primate models, it has been estimated that the human therapeutic dosage is 0.88 mg/mL of sterile saline solution for rhBMP-7, and 1.50 mg/mL of sterile water for rhBMP-2.44 The concentrations of BMPs used in current clinical applications are several orders of magnitude greater

427 than BMPs occurring naturally in the body (e.g. 3.5 mg in a single dose of BMP-7 vs. several nanograms). The expense of BMPs may be a prohibitive factor for combination therapies of BMPs and means that development of efficient, costeffective delivery systems remains all important, until the complex cascade of fracture healing is better understood and can be replicated. Several methods of delivering BMPs have been tried, e.g. direct application of the recombinant protein to the regeneration site by injection, with or without a carrier such as bone matrix or formulation buffer, local vs. systemic application, and gene therapy, in which a protein is delivered indirectly by its encoding gene.

Direct application of rhBMPs BMP in a buffer solution may be lost to surrounding tissue, but by suspending BMP in a carrier it may possible to retain the agent in the required location for a longer time. Combinations of BMPs with carrier vehicles such as gelatin foam, collagen or calcium phosphate directly applied to a regeneration site give BMP retention rates of 15–55%, compared to o5% retention using a buffer delivery system alone.45 One disadvantage of solid carriers is that they require an open procedure whereas injectable carriers of BMPs are easier to administer and preferable where open surgery would not otherwise be necessary (e.g. closed treatment of fractures).

Gene therapy Gene therapy is still at an early stage, with promising results in animal models.46 Either genes expressing BMPs can be delivered to the regeneration site directly resulting in transfection of cells and protein expression (in vivo transduction), or through transfection of cultured cells (in vitro transduction) which are then implanted and express the required protein. Advantages such as percutaneous delivery, relatively low cost of manufacture and high local concentration must be weighed against immunogenicity, inflammatory response, and variation of gene expression.

Safety of BMPs The concentration of BMPs used in clinical application is often at least four to six orders of magnitude greater than the concentration of naturally occurring BMPs. Ectopic bone formation and overgrowth has been reported after the use of extremely high BMP doses.43 This appears to be regulated by the limited number of mesenchymal stem cells which express BMP receptors, as well as osteoclastic activity stimulated by high doses of BMPs.47 However, using current therapeutic doses, there appears to be a relatively small risk of excessive bone formation, and it is usually eventually remodelled to the natural contour of the bone.48 Systemic clearance of rhBMPs is rapid and human studies have not demonstrated any systemic toxicity so far. Collagen carrier materials have been shown to induce antibody formation in 5–20% of patients in one series,7 while antibodies against BMPs developed in 6–10% patients in another.4 Currently, use of BMP on more than one occasion in

ARTICLE IN PRESS 428 the same patient is contra-indicated. It is unknown whether immunogenic reactions will occur if higher doses of BMPs or collagen are used in initial applications, or when used again in a previously sensitised patient. Long-term genetic effects in humans are unknown. No studies have demonstrated any evidence of carcinogenicity. Usage of BMPs during pregnancy is contraindicated, and must be used with caution in children.44

Discussion BMPs show great clinical potential in the treatment of musculoskeletal disorders.49 However a particular criticism leveled at current usage of BMPs is that the method of using a huge concentration of a single growth factor is a crude (yet effective) tool. BMPs are just one type of growth factor in a complex healing cascade which is not fully understood. Nonetheless single BMPs seem to be able to initiate the normal fracture healing cascade in practice and there is good evidence for the clinical application of BMPs in orthopaedic surgery. In particular BMP-7 has been shown to be of benefit in the treatment of nonunion, whilst BMP-2 is effective in fresh fractures and interbody spinal fusion. BMPs are not an orthopaedic panacea, and further evidence is required to demonstrate other areas of therapeutic potential as, despite promising results for potential clinical applications in animal models, extrapolation of these results to the human model requires caution and further study. Combinations of BMPs may be more effective. Other growth factors such as fibroblast growth factor and growth hormone have also shown promise in animal models, particularly in tandem with BMPs.50 The main problems are to refine delivery methods, optimise delivery time and therapeutic dosages. Greater understanding of the potential applications of BMPs combined with meticulous surgical technique will enhance usage of this therapeutic tool. In future BMPs may also be used in combination with other growth factors, mesenchymal stem cells and gene therapy.51,52

A. Cheung, A.M. Phillips

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

References 1. Urist MR. Bone: formation by autoinduction. Science 1965;50(698):893–9. 2. Urist MR, Strates BS. Bone morphogenetic protein. J Dent Res 1971;50(6):1392–406. 3. Reddi AH. Bone morphogenetic proteins: from basic science to clinical applications. J Bone Jt Surg Am 2001;83-A(Suppl. 1, Part 1):S1–6. 4. Friedlaender GE, Perry CR, Cole JD, Cook SD, Cierny G, Muschler GF, et al. Osteogenic protein-1 (bone morphogenetic protein-7) in the treatment of tibial nonunions. J Bone Jt Surg Am 2001;83-A(Suppl. 1, Part 2):S151–8. 5. Vogelin E, Jones NF, Huang JI, Brekke JH, Lieberman JR. Healing of a critical-sized defect in the rat femur with use of a vascularized periosteal flap, a biodegradable matrix, and bone morphogenetic protein. J Bone Jt Surg Am 2005;87(6):1323–31. 6. Salkeld SL, Patron LP, Barrack RL, Cook SD. The effect of osteogenic protein-1 on the healing of segmental bone defects treated with autograft or allograft bone. J Bone Jt Surg Am 2001;83-A(6):803–16. 7. Govender S, Csimma C, Genant HK, Valentin-Opran A, Amit Y, Arbel R, et al. BMP-2 Evaluation in Surgery for Tibial Trauma

18.

19. 20.

21. 22. 23.

24.

(BESTT) Study Group. Recombinant human bone morphogenetic protein-2 for treatment of open tibial fractures: a prospective, controlled, randomized study of four hundred and fifty patients. J Bone Jt Surg Am 2002;84-A(12):2123–34. Kuroda R, Usas A, Kubo S, Corsi K, Peng H, Rose T, et al. Cartilage repair using bone morphogenetic protein 4 and muscle-derived stem cells. Arthritis Rheum 2006;54(2):433–42. Cook SD, Patron LP, Salkeld SL, Rueger DC. Repair of articular cartilage defects with osteogenic protein-1 (BMP-7) in dogs. J Bone Jt Surg Am 2003;85-A(Suppl. 3):116–23. Kawakami M, Matsumoto T, Hashizume H, Kuribayashi K, Chubinskaya S, Yoshida M. Osteogenic protein-1 (osteogenic protein-1/bone morphogenetic protein-7) inhibits degeneration and pain-related behavior induced by chronically compressed nucleus pulposus in the rat. Spine 2005;30(17):1933–9. Jones AL, Bucholz RW, Bosse MJ, Mirza SK, Lyon TR, Webb LX, et al. BMP-2 Evaluation in Surgery for Tibial Trauma-Allograft (BESTT-ALL) Study Group. Recombinant human BMP-2 and allograft compared with autogenous bone graft for reconstruction of diaphyseal tibial fractures with cortical defects. A randomized, controlled trial. J Bone Jt Surg Am 2006; 88(7):1431–41. Pradhan BB, Bae HW, Dawson EG, Patel VV, Delamarter RB. Graft resorption with the use of bone morphogenetic protein: lessons from anterior lumbar interbody fusion using femoral ring allografts and recombinant human bone morphogenetic protein-2. Spine 2006;31(10):E277–84. Sumner DR, Turner TM, Urban RM, Virdi AS, Inoue N. Additive enhancement of implant fixation following combined treatment with rhTGF-beta2 and rhBMP-2 in a canine model. J Bone Jt Surg Am 2006;88(4):806–17. Burkus JK, Gornet MF, Dickman CA, Zdeblick TA. Anterior lumbar interbody fusion using rhBMP-2 with tapered interbody cages. J Spinal Disord Tech 2002;15(5):337–49. Mizumoto Y, Moseley T, Drews M, Cooper III VN, Reddi AH. Acceleration of regenerate ossification during distraction osteogenesis with recombinant human bone morphogenetic protein-7. J Bone Jt Surg Am 2003;85-A(Suppl. 3):124–30. Martinek V, Latterman C, Usas A, Abramowitch S, Woo SL, Fu FH, et al. Enhancement of tendon-bone integration of anterior cruciate ligament grafts with bone morphogenetic protein-2 gene transfer: a histological and biomechanical study. J Bone Jt Surg Am 2002;84-A(7):1123–31. Albrektsson T, Johansson C. Osteoinduction, osteoconduction and osseointegration. Eur Spine J 2001;10(Suppl. 2):S96–S101. Cheng H, Jiang W, Phillips FM, Haydon RC, Peng Y, Zhou L, et al. Osteogenic activity of the fourteen types of human bone morphogenetic proteins (BMPs). J Bone Jt Surg Am 2003;85-A(8):1544–52 [Erratum in: J Bone Jt Surg Am 2004; 86-A(1):141]. Matthews SJ. Biological activity of bone morphogenetic proteins (BMPs). Injury 2005;36(Suppl. 3):S34–7. Wang EA, Rosen V, D’Alessandro JS, Bauduy M, Cordes P, Harada T, et al. Recombinant human bone morphogenetic protein induces bone formation. Proc Natl Acad Sci USA 1990;87(6): 2220–4. Phillips AM. Overview of the fracture healing cascade. Injury 2005;36(Suppl. 3):S5–7. Bahamonde ME, Lyons KM. BMP3: to be or not to be a BMP. J Bone Jt Surg Am 2001;83-A(Suppl. 1, Part 1):S56–62. Weiss RE, Reddi AH. Synthesis and localization of fibronectin during collagenous matrix-mesenchymal cell interaction and differentiation of cartilage and bone in vivo. Proc Natl Acad Sci USA 1980;77(4):2074–8. Gerstenfeld LC, Cullinane DM, Barnes GL, Graves DT, Emhorn TA. Fracture healing as a post-natal developmental process: molecular, spatial, and temporal aspects of its regulation. J Cell Biochem 2003;88(5):873–84.

ARTICLE IN PRESS Bone morphogenetic proteins in orthopaedic surgery 25. Heldin CH, Miyazone K, ten Dijke P. TGF-beta signalling from cell membrane to nucleus through SMAD proteins. Nature 1997;390(6659):465–71. 26. Simpson AH, Mills L, Noble B. The role of growth factors and related agents in accelerating fracture healing. J Bone Jt Surg Br 2006;88(6):701–5. 27. Giannoudis PV, Tzioupis C. Clinical applications of BMP-7: the UK perspective. Injury 2005;36(Suppl. 3):S47–50. 28. Johnson EE, Urist MR. Human bone morphogenetic protein allografting for reconstruction of femoral nonunion. Clin Orthop Relat Res 2000(371):61–74. 29. Burkus JK, Sandhu HS, Gornet MF, Longley MC. Use of rhBMP-2 in combination with structural cortical allografts: clinical and radiographic outcomes in anterior lumbar spinal surgery. J Bone Jt Surg Am 2005;87(6):1205–12. 30. Sandhu HS, Khan SN. Recombinant human bone morphogenetic protein-2: use in spinal fusion applications. J Bone Jt Surg Am 2003;85-A(Suppl. 3):89–95. 31. Bragdon CR, Doherty AM, Rubash HE, Jasty M, Li XJ, Seeherman H, et al. The efficacy of BMP-2 to induce bone ingrowth in a total hip replacement model. Clin Orthop Relat Res 2003(417):50–61. 32. Fabeck L, Ghafil D, Gerroudj M, Baillon R, Delince P. Bone morphogenetic protein 7 in the treatment of congenital pseudarthrosis of the tibia. J Bone Jt Surg Br 2006;88(1):116–8. 33. Lee FY, Sinicropi SM, Lee FS, Vitale MG, Roye Jr DP, Choi IH. Treatment of congenital pseudoarthrosis of the tibia with recombinant human bone morphogenetic protein-7 (rhBMP-7). A report of five cases. J Bone Jt Surg Am 2006;88(3):627–33. 34. Lieberman JR, Conduah A, Urist MR. Treatment of osteonecrosis of the femoral head with core decompression and human bone morphogenetic protein. Clin Orthop Relat Res 2004(429): 139–45. 35. Goldring MB. Are bone morphogenetic proteins effective inducers of cartilage repair? Ex vivo transduction of musclederived stem cells. Arthritis Rheum 2006;54(2):387–9. 36. Masuda K, Imai Y, Okuma M, Muehleman C, Nakagawa K, Akeda K, et al. An HSOsteogenic protein-1 injection into a degenerated disc induces the restoration of disc height and structural changes in the rabbit annular puncture model. Spine 2006; 31(7):742–54. 37. Hayashi K, Ishidou Y, Yonemori K, Nagamine T, Origuchi N, Maeda S, et al. Expression and localization of bone morphogenetic proteins (BMPs) and BMP receptors in ossification of the ligamentum flavum. Bone 1997;21(1):23–30. 38. Zimmerman LB, De Jesus-Escobar JM, Hariand RM. The Spemann organizer signal noggin binds and inactivates bone morphogenetic protein 4. Cell 1996;86(4):599–606. 39. Hsu DR, Economides AN, Wang X, Eimon PM, Hariand RM. The Xenopus dorsalizing factor Gremlin identifies a novel family of

429

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51. 52.

secreted proteins that antagonize BMP activities. Mol Cell 1998;1(5):673–83. Glaser DL, Economides AN, Wang L, Liu X, Kimble RD, Fandl JP, et al. In vivo somatic cell gene transfer of an engineered Noggin mutein prevents BMP4-induced heterotopic ossification. J Bone Jt Surg Am 2003;85-A(12):2332–42. Zhang H, Bradley A. Mice deficient for BMP2 are nonviable and have defects in amnion/chorion and cardiac development. Development 1996;122(10):2977–86. Dudley AT, Lyons KM, Robertson EJ. A requirement for bone morphogenetic protein-7 during development of the mammalian kidney and eye. Genes Dev 1995;9(22):2795–807. Valentin-Opran A, Wozney J, Csimma C, Lilly L, Riedel GE. Clinical evaluation of recombinant human bone morphogenetic protein-2. Clin Orthop Relat Res 2002(395): 110–20. Termaat MF, Den Boer FC, Bakker FC, Patka P, Haarman HJ. Bone morphogenetic proteins. Development and clinical efficacy in the treatment of fractures and bone defects. J Bone Jt Surg Am 2005;87(6):1367–78. Winn SR, Uludag H, Hollinger JO. Carrier systems for bone morphogenetic proteins. Clin Orthop Relat Res 1999(367 Suppl.):S95–S106. Betz OB, Betz VM, Nazarian A, Pilapil CG, Vrahas MS, Bouxsein ML, et al. Direct percutaneous gene delivery to enhance healing of segmental bone defects. J Bone Jt Surg Am 2006;88(2): 355–65. Itoh K, Udagawa N, Katagiri T, lemura S, Ueno N, Yasuda H, et al. Bone morphogenetic protein 2 stimulates osteoclast differentiation and survival supported by receptor activator of nuclear factor-kappaB ligand. Endocrinology 2001;142(8): 3656–62. Ripamonti U, Van Den Heever B, Sampath TK, Tucker MM, Rueger DC, Reddi AH. Complete regeneration of bone in the baboon by recombinant human osteogenic protein-1 (hOP-1, bone morphogenetic protein-7). Growth Factors 1996;13(3–4): 273–89 [color plates III–VIII, pre.bk]. Mont MA, Ragland PS, Biggins B, Friedlaender G, Patel T, Cook S, et al. Use of bone morphogenetic proteins for musculoskeletal applications. An overview. J Bone Jt Surg Am 2004;86-A(Suppl. 2):41–55. Lieberman JR, Daluiski A, Einhorn TA. The role of growth factors in the repair of bone. Biology and clinical applications. J Bone Jt Surg Am 2002;84-A(6):1032–44. De Biase P, Capanna R. Clinical applications of BMPs. Injury 2005;36(Suppl. 3):S43–6. Muschler GF, Nakamoto C, Griffith LG. Engineering principles of clinical cell-based tissue engineering. J Bone Jt Surg Am 2004;86-A(7):1541–58.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 430–445

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

SPINE

Spinal muscular atrophy: Classification, aetiology, and treatment of spinal deformity in children and adolescents$ Athanasios I. Tsirikos, Alexander D.L. Baker Scottish National Spine Deformity Centre, Royal Hospital for Sick Children, Sciennes Road, Edinburgh EH9 1LF, UK

KEYWORDS Spinal muscular atrophy; Spinal deformity; Neuromuscular scoliosis; Surgical treatment; Paralytic scoliosis

Summary Spinal muscular atrophy is a hereditary neurological condition, which presents with symmetrical limb and trunk weakness. Spine deformity is the most frequent orthopaedic manifestation of the disease in patients who survive beyond the first year of life. Scoliosis in this group of severely disabled children decreases their sitting tolerance, causes pain from impingement of the ribs against the pelvis, affects ambulatory ability, and creates further respiratory compromise accelerating their death. Spinal arthrodesis is the only treatment that has a well-documented positive impact in restoring trunk balance and preserving function. This is associated with significant technical challenges and a high rate of life-threatening complications. A comprehensive review of the condition and a strategy for treating spinal deformity are presented in this paper. & 2006 Elsevier Ltd. All rights reserved.

Introduction

$ The authors did not receive grants or outside funding in support of the preparation of this manuscript. They did not receive payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution or other charitable or nonprofit organization with which the authors are affiliated or associated. Corresponding author. Tel.: +44 131 536 0784; fax: +44 131 536 0924. E-mail address: [email protected] (A.I. Tsirikos).

0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.09.006

Spinal muscular atrophy (SMA) is an uncommon hereditary condition of autosomal recessive inheritance, which affects the anterior horn cells of the spinal cord and the neurons of the lower bulbar nuclei. It manifests clinically with symmetrical limb and trunk weakness affecting the proximal more than the distal trunk muscles and the lower more than the upper limbs. Sensation and mental function are not impaired.1 Fasciculation of the tongue and the deltoid are often seen, as well as a fine tremor affecting the hands.2,3 This tremor resolves during relaxation or sleep. The development of scoliosis, joint contractures and dislocation of the hip are the most common orthopaedic problems associated with the condition. The natural history of SMA is premature death,

ARTICLE IN PRESS Scoliosis in spinal muscular atrophy

431

which is caused by respiratory failure as the consequence of a gradually deteriorating respiratory muscular function. The purpose of this article is twofold. Firstly, to provide a comprehensive overview on SMA, which is helpful to orthopaedists involved in caring for these children, including recent advancements in classification, understanding of the aetiology, and principles of management of the condition. The focus of the paper is, however, to investigate the characteristics of spinal deformity that develops in this group of patients and to describe a strategy for treatment. We believe that it is important to increase awareness of the complexity of these deformities in regard to their specific surgical considerations, as well as the coexistence of a multitude of associated medical problems that occur in this particular patient population and can jeopardize an inherently challenging surgical procedure.

Classification There is no universally accepted classification for SMA. Currently, there is more than one classification systems in use with each one of these systems applying different terms to the various types of the condition (Table 1). A recent attempt at resolving this problem by the international SMA collaboration has led to the definition of specific diagnostic criteria and the development of a new classification system, which is based on the age at presentation of symptoms, functional abilities and life expectancy (Tables 1 and 2).4 The classification system most widely used by orthopaedic surgeons is largely historical, but it correlates well with disease severity. In 1893, Hoffmann5 and in 1894, Werdnig6 described SMA for the first time in the literature. Kugelberg and Welander7 described in 1956 a similar condition of later onset that was also less progressive in nature.

Table 1

According to this classification, type I, acute Werdnig– Hoffmann or acute infantile SMA is characterized by generalized muscular weakness and hypotonia that has its onset before the age of 6 months. This is the most severe form of SMA with the earliest onset of symptoms and most rapid progression. In 30% of cases the onset is prenatal.8 There may be a history of poor foetal movement and crying is of low volume and not sustained following delivery. These children are never able to walk or sit unaided. They have lack of head control and often bulbar paralysis. Voluntary movement of their extremities may be limited to their fingers and toes and there is often fasciculation of the tongue and a fine tremor affecting the hands.3 Spinal reflexes are globally absent. Facial expression has been described as bland with a striking facial resemblance among patients due to muscle weakness. The cranial nerves may be also involved. Nasogastric feeding may be required as these children may suffer from dysphagia. As the intercostal muscles are affected, respiration depends mainly on the diaphragm. Respiratory movements are paradoxic with severe collapse of the ribs and a bell-shaped lower thorax. Progressive respiratory insufficiency can lead to recurrent chest infections and atelectasis that ultimately result in early death from pulmonary failure. The prognosis is poor and life expectancy ranges from 6 months to 2 years. As a consequence, orthopaedic intervention is rarely required. Type II, chronic Werdnig–Hoffmann or chronic infantile SMA has its onset before the age of 18 months and is less severe than type I. Muscle weakness is not generalized at first and there is less disease involvement. After initial presentation, neuromuscular function in these patients may remain static for long periods before eventual progression. Most commonly the weakness begins in the proximal lower limb muscles (gluteal and quadriceps). The legs are affected earlier than the arms. The upper limbs are also involved but

Classification systems and clinical features of SMA.4–7,9,10

Comparison of the classifications and clinical features of SMA Historical Type I, Werdnig–Hoffmann, Acute Infantile Age of onset before 6 months Generalized muscle weakness Type II, Werdnig–Hoffmann, Chronic Infantile Age at onset 2 months to 2 years Lower limbs and proximal muscles affected first Unable to walk unaided Unable to run or climb stairs

Kugelberg–Wellander, Chronic Childhood, Juvenile Age at onset over 2 years May walk until adolescence May survive to fifth decade

SMA Collaboration Type I

Evans et al. Group I

Dubowitz Mild

Onset o6 months Never sit Do not live beyond 2 years Type II

Never able to sit

Could stand unaided at same stage in life

Group II

Intermediate

Can sit but are unable to walk

Could sit at some stage but was never able to stand unsupported

Onset o18 months Never stand live beyond 2 years Type III Onset 4 18 months Stand alone Live into adulthood

Group III Have limited walking ability Group IV

Severe

Walk, run and climb stairs

Was never able to sit unsupported

ARTICLE IN PRESS 432

Table 2

A.I. Tsirikos, A.D.L. Baker Criteria for the diagnosis of SMA.4

Criteria for diagnosis

Exclusion criteria

1. 1. Muscle weakness 2. Symmetrical 3. Proximal4distal Legs4arms Trunk involved 2. Denervation demonstrated By EMG 4. By muscle biopsy 5. Clinical fasciculations

CNS dysfunction Arthrogryposis Involvement of other neurologic systems or other organs (i.e., hearing, cardiac or vision) Sensory loss Eye muscle weakness 6. Marked facial weakness

this typically occurs later. A fine tremor of the hands and fasciculation of the tongue may be seen but it is less common than in type I. Fasciculation of the eyelids may be present. Respiratory problems develop after a variable initial asymptomatic period of time and thoracic deformity is less pronounced. Patients can typically sit if placed in position and may be able to stand with assistance. They have a waddling gait, increased lumbar lordosis, genu recurvatum, a protuberant abdomen and scoliosis. Prognosis in this form varies considerably and life expectancy ranges from 2 years to adulthood. Most patients live longer than 10 years and some may survive up to their fifth decade. The principal cause of death is respiratory compromise. Type III, Kugelberg–Welander, chronic childhood, or juvenile SMA has a later onset and better prognosis. The disease manifests clinically after the age of 18 months and usually before the age of 10 years. The onset of muscle weakness is gradual and typically follows a slowly progressive course. The affected children achieve normal or slightly delayed motor milestones. These patients are usually able to walk, run and climb stairs until late childhood and sometimes into adolescence. On clinical examination, they have atrophic proximal lower limb muscles with characteristic pseudohypertrophy of the gastrocnemius, which is spared. A positive Gower sign can also be elicited due to marked weakness in the pelvic girdle. Their ability to walk may remain for many years after the initial diagnosis and often continues into the third decade of life. Survivorship can extend to the fifth decade. More recently, Evans et al.9 described a functional classification system, which has been based on the maximum physical function achieved by the affected patients rather than the age of onset or age at diagnosis of the condition (Table 1). This classification only applies to the chronic Werdnig–Hoffmann, and the Kugelberg–Welander forms of SMA and divides the children into four groups: Group I. Patients never develop the ability to sit independently and have poor head control. Group II. Patients have head control and can sit but are not able to walk. Group III. Patients can pull themselves up and walk in a limited fashion frequently requiring orthoses. Group IV. Patients develop the ability to walk, climb stair and run before the onset of weakness. Dubowitz developed another classification system in 1978 according to the severity of the disease and incorporated

three types (Table 1).10 Mild when the child has been able to stand unaided at some stage in life. Intermediate when the child has been able to sit without support at some stage in life. Severe when the child has never been able to sit without support.

Epidemiology Pearn et al.11 reported an incidence of 1 in 25,000 live births for SMA type I in the United Kingdom, making it one of the most common causes of genetically determined neonatal death. This translates to a gene frequency of about 1 in 160 and a carrier frequency of 1 in 80.12 Epidemiological data from Canada, Finland, Hungary, and Norway have recorded an incidence of SMA type I ranging from 1 in 25,000 to 1 in 15,000 live births.13–16 The incidence in types II and III SMA account for similar numbers of patients.17,18 As anticipated for a genetically transmitted disease, the reported incidence of SMA in population groups where consanguineous marriages are a frequent practice is significantly higher. In Dammam Saudi Arabia, an incidence of the condition of up to 1 in 518 live births has been described.19 The disease is even more common among the Karaite community in Israel with an increased incidence of 1 in 400.20

Aetiology and pathogenesis The complete pathogenesis of SMA is still undetermined; however, the genetic cause of many cases has been attributed to two genes. The identification of the Survival of Motor Neuron 1 (SMN1 or telomeric SMNT) gene on the long arm of chromosome 5 (5q11.2–13.3) was first described by Brzustowicz et al. in 1990.21 The SMN1 gene lies within an inverted duplication which contains an almost identical copy of the gene (labelled SMN2 or centromeric SMNC).22 The majority of SMA patients (98.7%) with the autosomal recessive form of proximal SMA lack SMN1 but carry at least one copy of SMN2, which only partially functions and is unable to compensate completely for the lack of SMN1.23 Both SMN1 and SMN2 encode for the survival motor neuron (SMN) protein.24 SMA develops as a consequence of an insufficient amount of the SMN protein. The level of SMN protein expressed from SMN2 varies between patients, accounting for different disease phenotypes, and also shows a variable expression between different human tissues. High levels of SMN protein are expressed in the spinal cord, brain, liver and kidney.24 In SMA patients the tissue that shows the most significant change in SMN expression is the spinal cord. SMN expression is reduced 100-fold in the spinal cord of SMA type I patients when compared to non-SMA controls. The SMN protein is likely to have many functions but one of them is essential for cell survival. The loss of SMN protein has a catastrophic effect on the motor neuron cell, resulting in cell death. Exactly which metabolic processes are disrupted remains unknown. Examination of the spinal cord reveals loss of anterior horn cells (motor neurones) and secondary changes occur in the roots and nerves as a consequence of Wallerian degeneration.25 Denervation of the associated muscle groups leads to muscular atrophy and the clinical

ARTICLE IN PRESS Scoliosis in spinal muscular atrophy

433

presentation of SMA. Clinically evident sensory functional loss is rarely detected in these patients; however, microscopic changes in the dorsal spinal columns have been previously reported.26 Apart from the autosomal recessive form of SMA there are other genetically distinct forms that have been described including an autosomal dominant form,27–29 an X-linked type of SMA,30–32 distal SMA,33 and SMA with associated respiratory distress.34 SMA is considered by most investigators to be a progressive neuromuscular condition; however, the cause of this observed functional deterioration is not clear. In a prospective study using reproducible quantitative measurements, Russman et al.35,36 have examined muscle strength over a period ranging from 2 to 6 years in 73 patients with an age range of 4 to 57 years. Muscle strength was ‘an order of magnitude’ less in SMA patients (total muscle score 17.1 kg in patients less than 15 years of age; 21.1 kg in those aged more than 15 years) when compared to controls (total muscle score 101.7 and 229.5 kg respectively). However, with time no measurable loss in muscle strength was detected in individual patients with SMA. When comparing patients and controls over the age of 15 years to those aged less than 15 years, the authors found a 200% increase in total muscle score in the control group and only a 25% increase in the SMA group. It has been suggested that the functional loss observed in patients with SMA over the years may be due to secondary factors, such as the development of scoliosis and hip subluxation or may be related to medical complications rather than be attributed to a continued functional motor unit loss.4,35,36

and magnitude in direct relation to the severity and duration of the disease. Fibrillation potentials following denervation of muscle are the most accurate diagnostic features. Spontaneous motor unit activity occurs frequently, is present at all ages, and is specific for the condition. Conduction velocity in motor units is often normal. However, it can be decreased in severely affected children and increased in children with less extensive involvement.3,39,40 Evidence of a muscle tremor defect may be present on the electrocardiogram.35 Genetic testing for autosomal recessive SMA using a PCRbased DNA test is now possible.41 Liu et al.42 used magnetic resonance imaging to investigate the appearance of affected muscles in 17 patients with SMA (3 severe, 9 intermediate and 5 mild according to the Dubowitz classification). In the severe form, the authors found significant atrophy of all the affected muscles of the thigh and calf. In the intermediate form, ragged atrophy of the muscle bundles of the thigh and calf with selective sparing of the adductor longus muscle was observed. In the mild form, fatty infiltration of muscle bundles and increased intermuscular fat planes was detected. The differential diagnosis should include all conditions that could manifest as a hypotonic baby. SMA should be distinguished from congenital myopathies where the electromyographic studies show myopathic findings. Hypotonic cerebral palsy or mixed forms with predominant flaccid paralysis have to be excluded. Neuromuscular conditions characterized by generalized muscle weakness, such as congenital muscular dystrophy or Duchenne muscular dystrophy also have to be ruled out.

Diagnosis

Orthopaedic problems

The diagnosis of SMA is primarily based on the clinical features and can also be supported by a positive family history. The diagnostic criteria for the condition have been described by an international SMA collaboration (Table 2).4 The diagnostic studies include muscle and nerve biopsy, electromyography, and conduction velocity. Muscle biopsy can be used to assist diagnosis and should be performed on moderately affected muscles using an open technique. A careful, atraumatic surgical technique is necessary to obtain a muscle sample (preferably from the vastus lateralis) and maintain its length and orientation. Local anaesthetic can be utilized but must spare the muscle. Compromised areas of musculotendinous junctions, or sites of previous scarring due to injections, recent trauma or electromyography needle application should be avoided. Histological samples show evidence of loss and atrophy of muscle fibres within normal groups of fibre bundles and no evidence of a primary myopathy.37,38 There is proliferation of perimysial connective tissue, and groups of giant fibres (type I) can also be seen. In neonates, perimysial fibrosis and a non-specific variation in muscle fibre diameter are the only histological features present. In adults with SMA, the same histochemical pattern of involvement is preserved with the addition of a progressive denervation and necrosis of muscle fibres occurring within groups of hypertrophic fibre bundles. Electromyography may be used to support the diagnosis of SMA.38 Abnormal findings on this study increase in frequency

Orthopaedic manifestations of the condition that may require treatment can occur primarily in patients with SMA type II or III who have a longer life expectancy. Scoliosis and hip subluxation or dislocation are the two most common orthopaedic problems encountered. In the study by Evans et al.9 scoliosis affected all patients with moderate to severe disease. Progressive hip subluxation and dislocation occurred in up to 50% of patients with more severe disease often after the age of 5 years.9 Hip and knee contractures may develop in children with SMA as a consequence of immobility. This poor lower limb function may accelerate the progression of the spinal deformity.43 Walking may be promoted by the early provision of lower limb orthoses and the use of a walking frame. The regular use of a standing frame can also retain some degree of muscular function in non-ambulatory children without severe fixed hip and knee contractures. Patients with SMA type I are significantly hypotonic and have considerable respiratory and feeding difficulties leading to an early death. Orthopaedic treatment is rarely required and is mostly limited to range-of-motion exercises in order to prevent fixed joint contractures.43 In patients with SMA type II, subluxation and subsequent dislocation of the hip are caused by muscle imbalance and the inability to walk. The presence of muscular imbalance and the lack of mobility frequently produce a valgus deformity of the proximal femur, which progresses to

ARTICLE IN PRESS 434 gradual displacement of the femoral head from the acetabulum. Unilateral hip subluxation results in pelvic obliquity that can contribute to the development of scoliosis. Passive stretching exercises and soft tissue releases may be required to treat hip deformity. Shapiro and Specht43 have recommended performing a proximal femoral varus osteotomy to maintain hip alignment and restore a good sitting balance. Early attention should be paid to the hips in children where a long life expectancy is predicted, even though the incidence of re-operation is high in these patients.43 Other investigators have reported poor results following hip osteotomies in this patient population,44 or minimal clinical symptoms when no treatment has been offered.45 In patients with SMA type III who are able to walk and have developed acetabular dysplasia with valgus deformity of the proximal femur, surgical reconstruction of the hip on the femoral side may further affect their poor abductor muscle strength and lead to loss of walking function. Periacetabular osteotomy has been recommended as an alternative to proximal femoral osteotomy in these patients in an attempt to restore congruity and containment of the hip joint and prevent further weakening of the hip abductors.43

Spinal deformity Epidemiology: Spinal deformity in patients with SMA who survive early childhood constitutes a particularly common problem. There have been several previous studies to calculate the incidence of deformities of the spine in this group of patients.9,37,46–50 However, it is not likely that these studies have included the whole population of patients affected with SMA. A large number of children with SMA type I do not survive long enough for orthopaedic problems to require treatment and are not included in any of these reports. The development of spinal deformity should be anticipated in practically all patients who will have a life expectancy beyond the first few years of their life. The reported incidence of scoliosis varies according to the length of clinical follow-up and the type of SMA; it ranges from 60%51 to almost 100%9,46. In contrast, Benady52 recorded a relatively low incidence of scoliosis of 54% in a group of 46 patients with SMA. Seventeen of these 25 patients developed a curvature before the age of 4 years. Among the patients without scoliosis there were children with less severe SMA and a favourable expected survivorship; it is, therefore, likely that at least some of these patients will develop scoliosis in later life. Characteristics of the spinal deformity: The collapsing long C-shaped scoliosis, typically seen in neuromuscular conditions, represents the most frequent pattern of spinal decompensation in children with SMA and is due to the inability of the trunk muscles to support the spine in the upright position. The curvature may involve a large segment of the spine with the apex located often in the thoracolumbar junction, extending into the pelvis and leading to the development of significant trunk and pelvic imbalance. Merlini et al.53 described the pattern of scoliosis in 40 patients with SMA. A single curve developed in 33 of these patients, with the most common type being a right

A.I. Tsirikos, A.D.L. Baker thoracolumbar scoliosis (67%; 22 patients).53 An associated thoracic or thoracolumbar hyperkyphosis can occur in 30% of the patients.46 Primary thoracic hyperkyphosis and lumbar hyperlordosis have also been described with or without coronal spinal deformity.9,48 The patient age at onset of scoliosis ranges from birth to adolescence.47,48,54 The age of onset, severity and progression of the scoliosis are related to the type of SMA and the extent of neurological involvement. The more severely affected the child, the higher the possibility of developing scoliosis at an earlier age and the greater the anticipated curve progression. The presence of familial involvement may also affect the severity and progression of the curve. Riddick et al.54 reported on 36 patients with SMA whose ages ranged from 2 to 35 years and found an incidence of spinal deformity of 86%. Twenty-eight patients had a scoliosis or kyphoscoliosis, which was most commonly thoracolumbar (17 patients), and three patients had a thoracic hyperkyphosis. The average age at onset of the deformity was 6 years, ranging from 1 week to 12 years. Natural history: The evolution of the scoliosis in this group of patients is characterized by an early onset with rapid progression and poor response to orthotic management, especially during the adolescent growth spurt. Scoliotic curvatures demonstrate a particularly higher incidence in children with more severe disease, directly proportionate to the extent of neuromuscular impairment and inversely proportionate to the ambulatory function. The spinal deformity develops as a consequence of the generalized muscle weakness and the poor ambulatory function. Previous studies have reported an association between the poor level of motor function and the development of scoliosis in this group of children.9,35,37,50,53 The limited ability for these patients to mobilize is directly related to the extent of neurological disorder. Patients with SMA and mild scoliosis who lost ambulatory capacity and became wheelchair-bound showed rapid deterioration of their scoliosis.46 In addition, scoliosis developed as soon as a previously ambulatory patient became wheelchair dependent.46 Aprin et al.49 reported in his group of SMA children an intervening time between loss of walking ability and the development of scoliosis of 2 years. Marked curve progression is also associated with periods of increased growth velocity.55 Granata et al.46 studied the natural history of spinal deformity in 63 patients with mild and intermediate SMA. Twenty-five of 32 patients with mild and all but one patient with intermediate SMA had developed a scoliosis. The mean age at onset of scoliosis was 4.3 years for the intermediate group and 9.9 years for the group with mild SMA. A single curve developed in 74% of patients and this was most commonly thoracolumbar convex to the right. The authors found a linear correlation between the development of scoliosis and age in patients with the intermediate form and in those with the mild form of SMA who could not walk. Rodillo et al.50 reviewed 63 patients, 37 with intermediate and 26 with mild SMA whose ages ranged from 2 to 26 years. In the intermediate form, all patients developed a scoliosis, which had an early onset and rapid deterioration. In the mild form, only 30% of the patients had a scoliosis that was progressive during puberty in those patients who had lost ambulatory function.

ARTICLE IN PRESS Scoliosis in spinal muscular atrophy Hensinger et al.48 reported on 50 patients with SMA; 29 patients were found to have a significant scoliosis (58%) and two patients had an increased lumbar lordosis without scoliosis. The average age of onset of the spinal curvature was 7.6 years ranging from birth to 13 years. In three patients, the scoliosis developed after loss of walking ability. In four cases the scoliosis was considered the cause of decreased ambulatory function. Schwentker and Gibson37 reviewed the records of 130 patients with SMA and identified scoliosis as their major orthopaedic problem. Fifty patients were re-examined. Seven of 15 ambulatory patients had a scoliosis as opposed to 28 of 35 non-ambulators. The scoliosis in wheelchairbound patients was more severe leading the authors to correlate the inability to walk with the severity of the scoliosis. They suggested that loss of ambulatory function was related to muscular weakness due to the underlying neurological disease rather than related to the spinal deformity. The pattern of scoliosis was analysed in 30 patients. Twenty-six were single curves, most commonly right thoracolumbar, and four were double curves. Evans et al.9 reviewed 54 patients and reported an incidence of spinal deformity of 92%. Thirty-nine patients had a single and eight had a double scoliosis. There were three patients with primary sagittal plane deformity. In this study, the more severe the inherent muscle weakness, the earlier the onset of the scoliosis and the more progressive the deformity. In patients with mild neurological involvement, progression of the scoliosis coincided with loss of ambulatory function. Functional impairment: The development of a pathological spinal curve in addition to the underlying neurological disorder may significantly restrict the patient’s functional capacities and increase the need for nursing care. Ambulatory children may become confined to a wheelchair as the deformity progresses. Wheelchair-dependent patients gradually loose their ability to sit due to the severe trunk and pelvic decompensation and become hand-dependent sitters. The imbalance of the spine and the pelvis may lead to persistent back pain and pain caused by impingement of the ribs against the iliac crest on the concavity of the scoliosis especially when the child is in the sitting position. Asymmetric weight bearing on the pelvis can result in the development of pressure sores. As the scoliosis progresses, it has an adverse effect on the respiratory function which also deteriorates and increases the risk for life-threatening complications.10,50

Treatment of patients with SMA The treatment of SMA focuses on the prevention and management of medical complications, most commonly associated with poor respiratory function, and early interventions to correct musculoskeletal deformities in children with a favourable life expectancy. Despite the recent advances in our understanding of the genetic basis of SMA it remains difficult to predict which children that develop SMA at a young age will survive long enough to necessitate orthopaedic treatment. However, due to significant improvements primarily in respiratory support, prolonged survival should be anticipated for many patients with mild

435 to moderate disease. Bach56 has recently reported the benefit on lung development and prevention of chest wall deformities by the use of nocturnal positive pressure ventilation. The identification of the SMN1 and SMN2 genes and their protein product as part of the pathogenesis of SMA has resulted in the exploration of novel genetic treatments for the condition. As patients with SMA lack the SMN1 gene but maintain the SMN2, which produces at least some SMN protein, one possible therapeutic target is upregulating expression of the SMN2 gene. A number of compounds have been examined in an attempt to increase the expression of the SMN2 gene including valproic acid, sodiun butyrate, phenylbutyrate, aclarubicin, and the aminoglycosides tobramycin and amikacin with encouraging in vitro results.57–61 Valproic acid has been tested in patients with SMA and has shown to increase SMN production in fibroblasts.57

Treatment of spinal deformity Conservative management There is no non-operative measure that has a documented effect on preventing scoliosis progression or the final outcome of the spinal deformity in children with SMA. The goal of all conservative modalities is not to correct the deformity but instead to retain function as the curve continues to progress and delay surgery for a later stage. Seating support: Appropriate seating adaptations should be considered as the cornerstone of managing patients with scoliosis before surgical correction is indicated. These supports can be built on the patient’s wheelchair, which becomes the primary seating device for those children over time. They include offset chest lateral rests, shoulder harnesses and straps, and can significantly improve sitting balance and maintain an upright posture. Extended head rests are essential in order to provide adequate support in patients with poor head and neck control. Curve control-orthoses: Bracing has also proved to be ineffective in controlling the deformity and does not seem to change the natural history of the scoliosis. Moreover, the traditional rigid thoracolumbosacral orthosis (TLSO) applied in the treatment of patients with idiopathic scoliosis can be restrictive for children with SMA who have respiratory compromise, feedings disorders, and occasionally skin problems.46,49 Alternatively, a soft brace or a bi-valved brace could be used whenever the patient is not using his adapted wheelchair to improve sitting position and facilitate activities of daily living. Previous authors have suggested that bracing can be useful as a measure to delay curve progression and increase sitting tolerance in patients with SMA.37,46,54 Schwentker et al.37 reported on 23 patients with SMA treated with spinal orthoses. In many patients, bracing started before the spinal deformity developed with the aim to support the child in the sitting position. In these children, the initiation of brace therapy did not prevent the development of scoliosis. In 18 patients the scoliosis occurred or progressed after the brace was applied. In only

ARTICLE IN PRESS 436 three patients with moderate curves the deformity remained static during the period of orthotic treatment. Riddick et al.54 treated conservatively 20 patients with SMA by using Milwaukee braces, total-contact underarm or sitting support orthoses, and serial casting. Their results were mostly discouraging on preventing scoliosis progression but allowed for a lower rate of curve deterioration in certain patients. Aprin et al.49 presented 15 patients with SMA and scoliosis that were treated with a spinal brace; brace therapy was found to be uniformly ineffective in controlling the deformity. Orthotic treatment was discontinued in five patients because of respiratory complications. In the remaining 10 patients, the scoliosis progressed relentlessly and required spinal arthodesis. Evans et al.9 reported on the comparative use of four different types of orthoses in patients with SMA. The trunk and lower extremity sitting orthosis has been described as a moulded posterior plastic shell, which controls the patient from axilla to toes and is being used to support the young severely affected children. This orthosis can be extended behind the head and may have a removable anterior component. The authors found this orthosis to be well tolerated by children with SMA providing them with satisfactory trunk support and allowing them to be more interactive with their surroundings. In the same study, the Milwaukee brace could only be fitted in patients with the mildest forms of SMA. One of these children developed severe dental deformities as a consequence of the chin rest. A plastazote lined total-contact orthosis was used in 12 patients and was associated with rapid progression of the deformity. The thoracic suspension orthosis, described as a moulded body jacket suspended from a wheelchair was used in five patients with a good response in three of these children. The stage at which bracing should be attempted still remains debatable. Schwentker et al.37 applied a brace before any spinal deformity developed. It has been suggested that the success of brace treatment is more likely in patients who can stand and walk more than in those who are non-ambulators.9,62 However, compliance can be more challenging in children who still retain good ambulatory function. Shapiro et al.63 proposed that bracing should be considered when the scoliotic curvature reaches 15–201 in order to slow progression of the deformity. In our opinion, a bi-valved brace can be used in young patients with SMA and a mild scoliosis, which retains flexibility; the aim is to improve sitting balance when the patient is not using his adapted wheelchair, potentially delay deterioration of the deformity, allow further spinal growth, and postpone surgical treatment for a later age. Orthotic management in these children should be closely monitored and the underarm brace discontinued if respiratory compromise or chest wall deformities develop and if the patient’s compliance is poor. If the deformity progresses to more than approximately 501 or if it becomes rigid we believe that spinal arthrodesis should be considered.

Surgical management Spinal arthrodesis with the use of instrumentation is the only treatment that has a well-established positive effect in

A.I. Tsirikos, A.D.L. Baker children and adolescents with SMA. Rapid progression of the scoliosis and decompensation of the trunk leads to loss of function and cardiorespiratory impairment usually in an unexpectedly short period of time.9,12 Indications: Spinal fusion is recommended in the presence of documented scoliosis progression and a curve size of between 401 and 601 in children ideally 10 years of age or older, especially if there is a recorded deterioration in their functional skills or their pulmonary capacities.9,55,63 At this size of deformity, an isolated posterior spinal arthrodesis should be sufficient to achieve satisfactory correction of the deformity without the need for additional anterior surgery or preoperative spinal traction. Other investigators have suggested earlier fusion for curves of approximately 351 or greater.51 If the scoliosis is left untreated for too long a severe and fixed curvature develops and pulmonary function is likely to become further compromised. A combined anterior release and posterior instrumented spinal fusion will then be required to maximize flexibility of the curve and allow for increased correction; this is associated with significantly higher perioperative risks compared to the posterior-only procedure. Therefore, we believe that spine surgery should not be delayed even in the expense of growth if this delay will result in the necessity for an anterior release, which will cause substantial increase in the rate of perioperative lifethreatening complications. Principles of surgery: Restoration of coronal and sagittal trunk balance is considered equally if not more important than the percentage of scoliosis correction in patients with SMA. The spinal arthrodesis should be extended from the upper thoracic region, in order to prevent the development of recurrent proximal kyphosis, to the pelvis, especially if pelvic obliquity is present, if the curve involves the sacrum, or if the patient has poor sitting balance (Fig. 1). Progression of the deformity following short fusions both above and below the instrumented levels has been previously reported by Aprin and Dorr.49,64 Long fusions to include the lumbosacral articulation are indicated even in the ambulatory children with SMA who have developed significant pelvic obliquity. Deterioration of the scoliosis and pelvic imbalance in these patients will lead to a gradual decline in the ambulatory function. It has been the authors’ experience that extension of the spinal arthrodesis to the sacrum does not significantly affect walking ability in most children, as long as a good coronal and sagittal spinal alignment has been achieved and the patients are mobilized with intensive physical therapy in the immediate postoperative period. In the small cohort of patients with milder neurological involvement (usually type III SMA) and good functional skills where the scoliosis is associated with minimal or no pelvic imbalance we are performing spinal arthrodesis from the upper thoracic to the lower lumbar spine (T1–T2 to L4–L5 level) sparing the lumbosacral joint in order to maintain some degree of spinal flexibility. In these less severely affected patients, who most commonly have a wellpreserved walking ability, the iliac crest has good bone quality and can be used as bone graft.65 The use of autograft bone harvested from the iliac crest in these children increases the rate of fusion and prevents the development of a non-union. The absence of significant osteopenia in the

ARTICLE IN PRESS Scoliosis in spinal muscular atrophy

437

Figure 1 Preoperative anteroposterior (A) and lateral (B) radiographs of the spine of a boy aged 13 years and 8 months with a left thoracolumbar scoliosis measuring 791 with 391 of associated pelvic obliquity (black line) and increased thoracolumbar kyphosis. The patient underwent a posterior spinal arthrodesis using segmental instrumentation and allograft bone extending from T1 to the sacrum with pelvic fixation. This resulted in a very satisfactory correction of his scoliosis, which measured 221 at 2 years follow-up, levelling of the pelvis and restoration of a normal sagittal balance (C,D).

vertebral bodies also allows for the use of third generation instrumentation systems with segmental pedicle screw fixation as opposed to sublaminar wires to correct the spinal deformity (Fig. 2).

Children with SMA and severe neurological disease occasionally have considerable weakness in their neck muscles. This weakness can be aggravated following spinal fusion to T1 or T2 and the patients may develop significant

ARTICLE IN PRESS 438

A.I. Tsirikos, A.D.L. Baker

Figure 2 Preoperative anteroposterior (A) and lateral (B) radiographs of the spine of a boy aged 16 years and 10 months with mild SMA and a right thoracolumbar scoliosis measuring 601 with minimal associated pelvic obliquity (black line) and increased thoracolumbar kyphosis. Due to adequate bone quality, the patient underwent a posterior spinal arthrodesis using third generation instrumentation, autologous iliac crest and allograft bone extending from T1 to L5 with segmental pedicle screw fixation. This resulted in a very satisfactory correction of his scoliosis, which measured 171 at 2 years follow-up, levelling of the pelvis with a mobile lumbosacral joint, and restoration of a normal sagittal balance (C,D).

neck pain. The way to overcome these symptoms is to engage these children post-surgery on a regular physical therapy programme, which should focus on strengthening their neck muscles. Occasionally, the use of a supportive soft neck collar as a temporary measure can alleviate the neck discomfort. In extreme cases, where the child has a scoliosis and very poor head control, extension of the spinal arthrodesis proximally to the occiput can be considered. However, that increases the risk of pseudarthrosis and requires an instrumentation system with tapered rods and occipital plate fixation. Anterior Surgery: Anterior spinal release can be used to allow for better scoliosis correction in patients with rigid

deformities.9,39,43,63,66 However, the severe risk of consequent respiratory compromise following a thoracotomy or a thoracoabdominal approach to the spine suggests that an anterior spine arthrodesis should only be attempted in patients with mild disease and good preoperative pulmonary function. The need for postoperative tracheostomy in this group of children has been previously related to the use of an anterior spinal operation.49 As the severity of the scoliosis is directly associated with the degree of neurological involvement, it is unlikely that patients with significant curves, who would be candidates for an anterior release, will have sufficiently preserved respiratory reserves to survive this procedure.

ARTICLE IN PRESS Scoliosis in spinal muscular atrophy Anterior spinal instrumentation alone provides stabilization of a short segment of the spine and is, therefore, not indicated in patients with SMA. An isolated anterior instrumented fusion cannot correct pelvic obliquity and has a high risk of proximal and distal add-on or residual deformity. Anterior instrumentation can be also used after the multi-level discectomies as the first stage of a combined anterior and posterior spinal fusion; this is then followed by an instrumented posterior spinal arthrodesis. In the vast majority of our patients we do not use anterior instrumentation when we are performing the anterior release. We have found that in these severe deformities, a greater degree of correction can be achieved with the use of segmental posterior instrumentation following a circumferential spinal release, which includes complete anterior discectomies and annulectomies, as well as extensive posterior facetectomies and capsulectomies at every level from T1 to the sacrum. Preoperative assessment: Surgery to correct scoliosis in children with SMA is a significant undertaking and carries considerable risks. These patients may have a number of concomitant medical problems, which will need to be addressed before surgical management is considered. A multi-disciplinary approach will be required and the expert input from a wide group of specialists is indispensable to optimize the patients’ medical condition prior to surgery and minimize the potential for complications. The spine surgeon should not only limit his role to performing the operation; he needs to recognize that much of the success of the surgical procedure depends on a good coordination of the actions of the medical, nursing and therapist teams. In our experience, the involvement of dedicated anaesthetists, paediatric neurologists, cardiac and respiratory physicians, gastroenterologists, dieticians, physiotherapists, occupational therapists, and intensive nursing care in the perioperative management of these children is essential for a favourable outcome. A thorough preoperative evaluation of the patients is routinely performed, including monitoring of their respiratory capacity, cardiac function, immune system, coagulation mechanisms, nutritional status, urinary system and their overall level of functional deficiencies. Respiratory: The major limiting factor for scoliosis surgery in children with SMA is their respiratory function. Pulmonary function at the time of surgery may already be significantly impaired and this is an essential prognostic parameter for postoperative respiratory complications.52,67 Patients with a vital capacity of 50% or less of predicted value are at risk of developing pulmonary edema or atelectasis.67 In the same study, patients with a vital capacity of more than 60% did not have respiratory complications after scoliosis surgery.67 Patients with a vital capacity of less than 25% are at significant risk of developing life-threatening pulmonary complications.68 In this case, the risk of death involved in the procedure possibly outweighs the gains from correcting the deformity and these patients are not considered by most surgeons as adequate surgical candidates. Rodillo et al.50 reported on treating surgically patients with SMA and a vital capacity of less than 35% of predicted with a good outcome. Over the last 2 years we have performed posterior spinal fusion in a number of children

439 with SMA whose vital capacity was as low as 20% of predicted with encouraging results (Figs. 3 and 4). Our patients were kept intubated for 24–72 h following surgery and were then placed electively on assistive ventilation through a C-pap or bi-pap machine for a period varying from a few days to a few weeks. With the provision of positive pressure ventilation, aggressive respiratory physical therapy to clear secretions and prevent the development of chest infections, and patient mobilization on their adapted wheelchair soon after surgery all our patients survived the procedure and had a very satisfactory outcome. Most patients with a vital capacity of 25% or above of predicted did not require assistive ventilation long-term. We believe that scoliosis surgery should not be attempted in patients with respiratory function of less than 20% who are at the greatest risk of postoperative death.69 The preoperative application of a halo-femoral traction has been suggested as a measure to achieve improvement of the respiratory function (less than 10%) in children with SMA.69 However, this should be weighed against the associated morbidity and the poor patient compliance. We do not perform halo-femoral traction at our institution as we have not found it useful and young patients do not tolerate it well. Nutritional: Malnutrition is a common finding in children with SMA. This can be attributed to the combination of a poor diet and a high metabolic demand due to recurrent chest infections or other medical illnesses. In addition, the neurological condition may alter normal metabolic pathways.70 Muscle function is a critical nutritional reserve for protein, carbohydrate and mineral metabolism. Reduction of muscle mass may lead to loss of this nutritional reserve and limit the patient’s ability to adjust to simple nutritional changes such as overnight fasting. Fatty acid metabolism may be affected and patients with SMA may have a lower dietary fat tolerance; conversely a high carbohydrate diet (the result of a low fat diet) may place additional stress on the respiratory system by increasing carbon dioxide production. Poor nutrition predisposes to delayed wound healing and a poor immunological response to infection. Spine surgery should be postponed in the malnourished patient; supplemental nutrition in the preoperative period may be necessary to optimize a surgical candidate. Soon after the procedure, we place our patients on total parenteral alimentation or nutritional supplementation through a nasogastric tube until they can feed orally. Bleeding: A considerable amount of intraoperative blood loss should be anticipated during scoliosis correction in children with SMA. Bone quality is often inherently poor, especially in the non-ambulatory patients because of disuse osteopenia. Poor bone quality increases the risk of intraoperative blood loss and instrumentation failure during spinal fusion. Preoperative management of osteoporosis with the administration of intravenous bisphosphonates might be a consideration, especially in non-ambulators with marked osteoporosis, to maximize their bone quality before spine surgery. When these children are scheduled for spinal arthrodesis there must be at least one blood volume of blood typed and cross-matched. The use of a cell saver system during the procedure may limit the need for transfusion. Controlled

ARTICLE IN PRESS 440

A.I. Tsirikos, A.D.L. Baker

Figure 3 Preoperative anteroposterior (A) and lateral (B) radiographs of the spine of a boy aged 7 years and 6 months with a very severe right thoracolumbar scoliosis measuring 1251, 401 of associated pelvic obliquity (black line) and increased thoracolumbar kyphosis measuring 941. Due to his poor respiratory function (vital capacity of 23% of predicted) the patient underwent a posterior spinal arthrodesis using segmental instrumentation and allograft bone extending from T2 to the sacrum with pelvic fixation. This resulted in a satisfactory correction of his scoliosis, which measured 501 at 2 years follow-up, marked reduction of his pelvic obliquity (black line), which measured 151, and restoration of a normal sagittal balance (C,D). At discharge, the patient did not require assistive ventilation.

hypotensive anaesthesia, normovolemic haemodilution and meticulous haemostasis during tissue dissection are required to reduce blood loss. Early administration of fresh frozen plasma should be considered in the presence of a gradually increasing intraoperative blood loss, especially in patients with identified coagulopathies or those with severe neurological involvement. Recently, aprotinin has been reported to reduce blood loss during spinal surgery in this group of children.71 During surgery, transfusion of a significant volume of packed red cells and clotting factors may be required. There are many problems associated with large-volume blood transfusions, which include anaphylaxis and lesser febrile

reactions, hyperkalemia, hypocalcemia, and hypothermia. A consumptive coagulopathy (disseminated intravascular coagulation) may be the result of large volume transfusions and must be avoided with the appropriate administration of clotting products.

Surgical considerations Historical perspective: The introduction of Harrington instrumentation provided a tremendous evolution in the operative management of these complex deformities, but the incidence of pseudarthrosis and consequent curve

ARTICLE IN PRESS Scoliosis in spinal muscular atrophy

441

Figure 4 Preoperative anteroposterior (A) and lateral (B) radiographs of the spine of a girl aged 11 years and 7 months with a very severe right long thoracic scoliosis measuring 1151, 211 of associated pelvic obliquity (black line) and increased thoracic kyphosis measuring 100o. Due to her poor respiratory function (vital capacity of 24% of predicted) the patient underwent a posterior spinal arthrodesis using segmental instrumentation and allograft bone extending from T2 to the sacrum with pelvic fixation. This resulted in a satisfactory correction of her scoliosis, which measured 531 at 2 years follow-up, marked reduction of her pelvic obliquity (black line), which measured 101, and restoration of a normal sagittal balance (C,D). At discharge, the patient did not require assistive ventilation.

progression remained high.47,72 Harrington instrumentation was used in the surgical management of neuromuscular scoliosis until the late 1970s. The distraction techniques applied through the Harrington system failed to restore normal sagittal spinal balance and were often impractical in the soft bone of patients with SMA. In addition, postoperative immobilization of the spine in a cast was required and that could result in further compromise of the respiratory function. In 1977, Luque73 developed the concept of segmental spinal fixation with the application of translational correc-

tive forces through the use of multiple-levels sublaminar wires and two single rods. His technique achieved a wide distribution of forces over each vertebra, with an increased initial spinal stabilization and a low risk of instrumentation failure. This led to a greater degree of correction and decreased the risk of implant related complications. However, later reports indicated a high rate of pseudarthrosis associated with the Luque system of up to 10%, instrumentation-related complications up to 21%, and curve progression postoperatively in up to 30% of the cases.74,75

ARTICLE IN PRESS 442 The Galveston technique of intramedullary placement of the rod in the iliac bed was developed by Allen and Ferguson,76 and accomplished a secure pelvic fixation.74 The combination of the Luque wiring technique and the Galveston intrailiac rod fixation resulted in fewer instrumentation-related complications.47,50,51 However, it was soon recognized that the two unconnected soft rods introduced by Luque were moving independently, which was the reason why they failed to provide adequate stabilization. The use of postoperative immobilization did not prove to resolve the problem.77 The Unit rod, a further development of the Luque–Galveston technique, accomplished the requirement for rigidly connected rods and has been widely used in North America in the treatment of neuromuscular scoliosis. This instrumentation technique has provided improved correction of the spinal deformity in both the coronal and sagittal planes, as well as balancing of the pelvis. Newer techniques: More recently, newer instrumentation systems using primarily pedicle screws have been introduced in the management of neuromuscular scoliosis. These are based on the same principle of segmental fixation with pedicle screws or hooks instead of sublaminar wires and the alternative fixation of iliac bolts for lumbo-pelvic or sacroiliac plates for sacro-pelvic fixation. A common problem in the application of third generation instrumentation that uses pedicle hooks or screws in patients with SMA and marked associated osteopenia is poor vertebral fixation, which can limit the ability to perform corrective maneuvers and may increase significantly the risk of pseudarthrosis. It is our experience that in this group of children with considerably poor bone quality, the lamina provides the strongest point of fixation compared to the pedicle or the vertebral body and can withstand segmental translational forces applied through sublaminar wires that can achieve correction of the deformity and balancing of the spine in both the frontal and lateral planes. On the contrary, if repeat surgery is required, for example to address a non-union, it is technically considerably easier and safer to revise an instrumentation system that uses pedicle screws with or without hooks as opposed to sublaminar wires. As these modern techniques become common and widely utilized, it is essential to establish benchmarks for degree of deformity correction and complication rate related to the technical aspects of the individual procedure. As third generation spinal instrumentation is being widely utilized, there is also a significant increase in implant cost; this has to be balanced against possible benefits obtained by the use of these latest techniques.

Surgical technique used by the authors As previously mentioned, anterior release is performed in very few selected patients with severe scoliosis and increased curve rigidity who retain satisfactory pulmonary function. We perform the anterior procedure in the lateral decubitus position, using a thoracic or thoracolumbar retroperitoneal approach, through a removed rib, depending on the apex of the deformity and the levels of anterior release that are required. The anterior approach allows for an extensive release of the anterior longitudinal ligament,

A.I. Tsirikos, A.D.L. Baker complete annulectomy and discectomy, with the intention to provide angular and rotational mobility of the spinal segments, while at the same time enhancing anterior fusion in the excised disc spaces with the application of morselized rib graft. Anterior instrumentation is very rarely used. Once the anterior procedure is accomplished, the patients who are in a good general medical condition to tolerate the second stage procedure under the same anaesthetic are immediately rolled into the prone position and a posterior fusion with instrumentation is then performed. For most of our patients we select to stage the anterior and posterior procedures under separate anaesthetic sessions. In this second group, after the anterior stage is completed, the patients are taken to the intensive care unit (ICU), where hyperalimentation and aggressive pulmonary care are initiated until the posterior surgery 7–8 days later. The chest drain is removed when the drainage is less than 100 ml/day, usually on the third or fourth day after surgery. A posterior instrumentation system, which includes two independent rods and multi-level sublaminar wire fixation, is being used in all patients to apply translational corrective forces and provide segmental stability, with the spinal arthrodesis routinely extending from T1 or T2 to the sacrum with pelvic stabilization using the Galveston technique (Fig. 1). Distally, iliac screws are placed intramedullary in the iliac bed and these are attached to the bilateral rods using lateral connectors. Cross-connectors are bridging the two rods proximally and distally to create a rectangular frame and increase stability of the construct with the aim to prevent vertical movement between the two independent rods. Decortication of the transverse processes and lateral laminae with extensive excision of the facet joints and their capsule is performed at every level and the posterior instrumentation is always supplemented by abundant allograft bone mixed with autogenous graft harvested from the spinous processes. Iliac crest bone is not obtained in these patients due to various degrees of osteopenia. Wound drains are not routinely placed and the lumbosacral fascia is closed meticulously in order to obliterate dead space and aid wound healing. All patients receive prophylactic antibiotic treatment with the administration of a first-generation cephalosporin immediately before and for 24 h after surgery. Before anaesthesia induction, arterial and central venous lines are placed, and the central venous line is maintained until the second stage in patients who will require a two-stage procedure. A nasogastric tube is used to decompress the stomach and a Foley catheter to monitor urinary output. Cell-saver is used intraoperatively, and our patients receive homologous blood transfusions at the discretion of the anaesthetist. Spinal cord monitoring with the use of somatosensory evoked potentials is used during surgery in all patients. Motor-evoked potentials are recorded only in patients with ambulatory function. Postoperative care: The most important concern after scoliosis surgery is respiratory care. Coughing ability decreases following the operation and this along with the restriction of the patient’s mobility increases significantly the risk of atelectasis and pulmonary infection. Patients with severe neurological involvement and a decreased vital

ARTICLE IN PRESS Scoliosis in spinal muscular atrophy capacity often require ventilatory support (C-pap or bi-pap) or endotracheal intubation. Close monitoring by a respiratory physician is essential as patients may have to remain for several days or even weeks on non-invasive ventilation. If the previous measures prove ineffective, tracheostomy can be performed to provide pulmonary support for a longer time. Intensive respiratory physiotherapy and early mobilization of the patient are also fundamental to prevent pulmonary complications. Nutritional supplementation is initiated soon after surgery in the form of nasogastric feedings or total parenteral alimentation in order to provide adequate dietary coverage until the patients can feed orally. We do not use postoperative immobilization or external trunk support in our patients. The patients are mobilized early to an upright position in their wheelchair and are engaged in an intensive physical therapy programme. Their wheelchair is modified to adapt with the corrected seating posture. A reclining wheelchair has to be used for the initial mobilization following spinal instrumentation to the sacrum and the pelvis; an upright sitting position can be usually achieved gradually within the next few days. Postoperative complications: Scoliosis surgery in children with SMA is associated with complication rates as high as 45%, which are more prevalent in older patients and in those children with more severe deformities.48,72 Patients with extensive neurological deficits can be particularly sensitive to analgesic medications, such as opioids, that suppress respiratory function.77 Respiratory insufficiency, atelectasis, and pulmonary infections are the most common complications following surgery. Other reported complications include rupture of the diaphragm, pulmonary embolism, acute gastric volvulus, and narrowing of the diameter of the chest as a consequence of cast immobilization.49,54,78,79 Complications related to the technical aspects of the surgery include neurological damage to the spinal cord and the nerve roots, failure of the spinal instrumentation and subsequent pseudarthrosis, which can occur with an incidence ranging from 6% to 28% of the patients, wound infection, and postoperative urinary tract infection.47,48,72,79 Outcome after spinal fusion: Spinal arthrodesis in children with inherent neurological impairment constitutes a major physical insult and can be associated with an initial decline in functional skills in some of these patients.47 This initial deterioration in motor activity can be reversed with intensive physical therapy and should be weighed against the loss of function that would occur if the deformity was left untreated due to further progression. In previous reports, the negative impact of the scoliosis surgery on function has improved 5 years after the operation.47,80 Other investigators have reported improvement in function following spinal fusion.48,54,79 Scoliosis correction has been associated with an 86% patient satisfaction rate; this was mainly due to a better sitting balance, which allows the patients to use their upper limbs for functional tasks other than supporting their trunk in the upright position.49

Conclusions The development of spinal deformity constitutes the most common musculoskeletal problem in SMA and should be

443 anticipated in almost all the patients who survive beyond early childhood. The development of an abnormal spinal curvature in this group of severely disabled children decreases their sitting tolerance, creates pain from impingement of the ribs against the pelvis, and accelerates deterioration in respiratory function. Scoliosis surgery in paediatric patients with SMA who have severe neurological compromise and complex medical co-morbidities is associated with significant technical difficulties and an increased risk of life-threatening complications related primarily to their poor pulmonary capacity. However, there is a welldocumented positive impact on these children by correcting their spinal deformity in performing activities of daily living. With recent advancements in medical management and a multidisciplinary approach, life expectancy for this population of patients can be longer than previously reported. Operative procedures to restore the balance of the spine and the pelvis have a definitive effect in improving the patients‘ quality of life, preserving function, and prolonging their survivorship.

References 1. Von Gontard A, Zerres K, Backes M, et al. Intelligence and cognitive function in children and adolescents with spinal muscular atrophy. Neuromusc Dis 2002;12(2):130–6. 2. Dubowitz V. Infantile muscular atrophy: a prospective study with particular reference to a slowly progressive variety. Brain 1964;87:707–18. 3. Moosa A, Dubowitz V. Spinal muscular atrophy in childhood: two clues to clinical diagnosis. Arch Dis Child 1973;48: 386–8. 4. Munsat TL. Workshop report, International SMA collaboration. Neuromusc Dis 1991;1:81. 5. Hoffman J. Ueber chronischespinale Muskelatrophie im Kindersalter, auf familiarer Basis. Dutsch Z Nervenheulkd 1893;3:427. 6. Werdnig G. Die Fruhinfantile progressive spinale Amyotrophie. Arch Psychiatr Nervenkr 1894;26:706. 7. Kugelberg E, Welander L. Heredofamilial juvenile muscular atrophy simulating muscular dystrophy. Arch Neurol Psychiary 1956;75:500. 8. Pearn J. Classification of Spinal Muscular Atrophies. Lancet 1980;1(8174):912–22. 9. Evans GA, Drennan JC, Russman BS. Funtional classification and orthopaedic management of Spinal Muscular Atrophy. JBJ S (Br) 1981;63:516. 10. Dubowitz V. Muscle disorders in childhood. Philadelphia, PA: W.B. Saunders; 1978. 11. Pearn J. The gene frequency of acute Werdnig–Hoffmann disease (SMA Type I). A total population survey in North East England. J Med Genet 1973;10:260–5. 12. Munsat T. The spinal muscular atrophies. Current Neurol 1994;3(14):55–71. 13. Winsor EJ, Murphey EG, Thomson MW, et al. Genetics of childhood spinal muscular atrophy. J Med Genet 1971;8: 143–8. 14. Ignatius J, Donner M. Epidemiology of childhood spinal muscular atrophy (in Finnish). In: Lang H, editor. Lihastautien Kehittyva Tutkimus Ja Hoito. Turku: Kiasma; 1989. 15. Czeizel A, Hamula J. A Hungarian study on Werdnig–Hoffmann disease. J Med Genet 1989;26:761–3. 16. Tangsrud S-E, Halvorsen S. Child neuromuscular disease in Southern Norway. Prevalence, age, and distribution of diagnosis with special reference to non-Duchenne muscular dystrophy. Clin Genet 1988;34:145–52.

ARTICLE IN PRESS 444 17. Pearn J. Incidence, prevalence and gene frequency studies of chronic childhood spinal muscular atrophy. J Med Genet 1978;15:409–13. 18. Emery AEH. Population frequencies of inherited neuromuscular diseases—A world survey. Neuromusc Disorders 1991;1(1): 19–29. 19. Al-Rajeh S, et al. Werdnig–Hoffmann disease (spinal muscular atrophy type I): a clinical study of 25 Saudi Nationals in AlKhobar. Ann Saudi Med 1992;12:67–71. 20. Fried K, Mundel G. High incidence of spinal muscular atrophy type I (Werdnig-Hoffmann disease) in the Karaite community in Israel. Clin Genet 1977;12(4):250–1. 21. Brzustowicz LM, Lehner T, Caxtilla LH, et al. Genetic mapping of childhood-onset spinal muscular atrophy to chromosome 5q11.2–13.3. Nature 1990;344:540–1. 22. Lefevre S, Burglen L, Reboullet S, et al. Identification and characterization of a spinal muscular atrophy-determining gene. Cell 1995;80:155–65. 23. Monani UR. Spinal Muscular Atrophy: a deficiency in a ubiquitous protein; a motor neuron-specific disease. Neuron 2005;48: 885–96. 24. Coovert DD, Le TT, McAndrew PE, et al. The survival motor neuron protein in spinal muscular atrophy. Hum Mol Genet 1997;6(8):1205–14. 25. Thieffry A, Arthus M, Bargeton E. Quarante cas de maladie de Werdnig–Hoffman avec onje examens anatomiques. Rev Neurol 1955;93:621–44. 26. Carpenter S, Karpati G, Rothman S, et al. Pathological involvement of primary sensory beurones in Werdnig–Hoffmann disease. Acta Neuropath 1978;42:91–7. 27. Sambuughin N, Sivakumar K, Selenge B, et al. Autosomal dominant distal spinal muscular atropy V (dSMA-V) and CharcotMarie-Tooth disease type 2D (CMT2D) segregate within a single large kindred and map to a refined region on chromosome 7p15. J Neurol Sci 1998;161:23–8. 28. Zellweger H, Simpson J, McCormick WF, et al. Spinal Muscular Atrophy with autosomal dominant inheritance. Neurology 1972;22:957. 29. Van der Vleuten AJ, van Ravenswaaij-Arts CM, Frijns CJ, et al. Localisation of the gene for a dominant congenital spinal muscular atrophy predominantly affecting the lower limbs to chromosome 12q23–q24. Eur J Hum Genet 1998;6:376–82. 30. Greenberg F, Feuolia KR, Hetmancik JF, et al. X-linked infantile spinal muscular atrophy. Am J Dis Child 1988;142:217. 31. La Spada AR, Wilson EM, Lubahn DB, et al. Androgen receptor gene mutations in X-linked spinal and bulbar muscular atrophy. Nature 1991;352:77–9. 32. Kobayashi H, Baumbach L, Matise TC, et al. A gene for a severe lethal form of X-linked athrogryposis (X-linked infantile spinal muscular atrophy) maps to human chromosome Xp11.3–q11.2. Hum Mol Genet 1995;4:1213–6. 33. Viollet L, Barois A, Rebeiz JG, et al. Mapping of an autosomal recessive chronic distal spinal muscular atrophy to chromosome 11q13. Ann Neurol 2002;51:585–92. 34. Grohmann K, Schuelke M, Diers A, et al. Mutations in the gene encoding immunoglobulin mu-binding protein 2 casuses spinal muscular atrophy with respiratory distress type 1. Nat Genet 2001;29:75–7. 35. Russman BS, Iannacone ST, Buncher CR, et al. Spinal muscular atrophy: new thoughts on pathogenesis and classification schema. J Child Neurol 1992;7(4):347–53. 36. Iannaccone ST, Russman BS, Browne RH, et al. Prospective analysis of strength in spinal muscular atrophy. J Child Neurol 2000;15(2):97–101. 37. Schwentker EP, Gibson DA. The orthopaedic aspects of spinal muscular atrophy. JBJS (Am) 1976;58(1):32–7. 38. Buchthal F, Olsen PZ. Electromyography and muscle biopsy in infantile spinal muscular atrophy. Brain 1970;93:15–30.

A.I. Tsirikos, A.D.L. Baker 39. Drennan JC. Skeletal deformities in spinal muscular atrophy. Clin Orthop 1978;133(Abst):266–7. 40. Renault F, Raimbault J, Praud JP, et al. Etude electromyographique de 50 cas de maladie de Werdnig–Hoffman. Rev Electroencephalogr Neurophysiol Clin 1983;13:301–5. 41. Van Der Steege G, Grootscholten PM, Van Der Vlies P, et al. PCR-based DNA test to confirm clinical diagnosis of autosomal recessive spinal muscular atrophy. Lancet 1995;345: 985–6. 42. Liu GC, Jong YJ, Chiang CH, et al. Spinal muscular atrophy: MR evaluation. Pediatr Rad 1992;22(8):584–6. 43. Shapiro F, Specht L. Current concepts review. The diagnosis and orthopaedic treatment of childhood Spinal Muscular Atrophy, Peripheral Neuropathy, Friedreich Ataxia, and Athrogryposis. JBJS (Am) 1993;75(11):1699–714. 44. Thompson CE, Larsen LJ. Recurrent hip dislocation in intermediate spinal atrophy. J Ped Orthop 1990;10:638–41. 45. Sporer S, Smith B. Hip dislocation in patients with Spinal Muscular Atrophy. J Ped Orthop 2003;23(1):10–4. 46. Granata C, Merlini L, Magni E, et al. Spinal Muscular Atrophy: natural history and orthopaedic treatment of scoliosis. Spine 1988;14(7):760–2. 47. Brown JC, Zeller JL, Swank SM, et al. Surgical and functional results of spine fusion in Spinal Muscular Atrophy. Spine 1989;14(7):763–70. 48. Hensinger RN, MacEwan GD. Spinal deformity associated with hereditable neurological conditions: Spinal Muscular Atrophy, Freidreich’s Ataxia, familial dysautonomia, and Charcot-MarieTooth Disease. JBJS (Am) 1976;58(1):13–24. 49. Aprin H, Bowen R, MacEwen GD, et al. Spine fusion in patients with Spinal Muscular Atrophy. JBJS (Am) 1982;64(8):1179–87. 50. Rodillo E, Marini ML, Heckmatt JZ, et al. Scoliosis in Spinal Muscular Atrophy: review of 63 cases. Jr Child Neurol 1989;4(2):118–23. 51. Phillips DP, Roye DP, Farcy JC, et al. Surgical treatment of scoliosis in a Spinal Muscular Atrophy population. Spine 1990;15(9):942–5. 52. Benady SG. Spinal muscular atrophy in childhood: review of 50 cases. Dev Med Child Neurol 1978;20:746–57. 53. Merlini L, Granata C, Bonfiglioli S, Marini M, Cervellati S, Savini R. Scoliosis in Spinal Muscular Atrophy: Natural history and management. Dev Med Child Neurol 1989;31:501–8. 54. Riddick MF, Winter RB, Luter LD. Spinal deformities in patients with spinal muscular atrophy: a review of 36 patients. Spine 1982;7(5):476–83. 55. Bradford DS, Hu SS. Neuromuscular spinal deformity. Moe’s textbook of scoliosis and other spinal deformities, 3rd ed. Philadelphia: W.B. Saunders Company; 1978. 56. Bach JR, Bianchi C. Prevention of pectus excavatum for children with Spinal Muscular Atrophy Type1. Am J Physic Med Rehab 2003:815–9. 57. Britcha J, Hofmann Y, Hahnen E, et al. Valproic acid increases the SMN2 protein level: a well-known drug as a potential therapy for spinal muscular atrophy. Hum Mol Genet 2003;12(19):2481–9. 58. Chang JG, Hsieh-Li HM, Jong YJ, et al. Treatment of spinal muscular atrophy by sodium butyrate. Proc Natl Acad Sci USA 2001;98:9808–13. 59. Brahe C, Vitali T, Tiziano FD, et al. Phenyl butyrate increases the SMN gene expression in spinal muscular atrophy patients. Eur J Hum Genet 2005;13:256–9. 60. Wolstencroft EC, Mattis V, Bajer AA. A non-sequence-specific requirement for SMN protein activity: the role of aminogycosides in inducing elevated SMN protein levels. Hum Mol Genet; 14:1199–1210. 61. Andreassi C, Jarecki J, Zhoe J, et al. Aclarubicin treatment restores SMN levels to cells derived from type I spinal muscular atrophy patients. Hum Mol Genet 2001;10(24):2841–9.

ARTICLE IN PRESS Scoliosis in spinal muscular atrophy 62. Savini R, Cervellati S, Granata C, et al. La Scoliosi nelle atrofie muscolari prossimaii infantile. In: Grupo Italiano di studio della scoliosis, ed. Progressi in patalogia vertebrate. 1980 Bologna: Editore Aulo Gaggi. 63. Sharpio F, Bresnan MJ. Orthopaedic management of childhood neuromuscular disease, 1: Spinal Muscular Atrophy. JBJS (Am) 1982;64:785–9. 64. Dorr JR, Brown JC, Perry J. Results of posterior spinal fusion in patients with spinal muscular atrophy. A review of 25 cases. JBJS (Am) 1973;55:436–7. 65. Kinali M, Banks LM, Mercuri E. Bone mineral density in a paediatric Spinal Muscular Atrophy population. Neuropaediatrics 2004;35:325–8. 66. Freeman BL, Scoliosis and Kyphosis. Campbell’s Operative Orthopaedics. 10th ed., vol. 38. Mosby Philadelphia 2003, p. 1855–7. 67. Padman R, McNamara R. Postoperative pulmonary complications in children with neuromuscular scoliosis who underwent posterior spinal fusion. Delaware Med J 1990;62(5): 999–1003. 68. Frownfelter DL. Chest physical therapy and pulmonary rehabilitation. An interdisciplinary approach, 2nd ed. Chicago: Year Book; 1987. 69. Lonstein JE, Renshaw TS. Neuromuscular spine deformities. American Academy of Orthopaedic Surgeons: instructional course lectures, 36. Mosby St. Louis; 1987. 70. Leighton S. Nutritional issues associated with spinal muscular atrophy. Nutrition and Dietetics 2003;60:92–6.

445 71. Cole J, Murray DJ, Snider RJ, et al. Aprotinin reduces blood loss during spinal surgery in children. Spine 2003;21:2482–5. 72. Dayher YH, Lonstein JE, Winter RB, et al. Spinal deformities in patients with Freidreich ataxia: a review of 19 patients. J Ped Orthop 1985;5:553–7. 73. Luque ER. Paralytic scoliosis in growing children. Clin Orthop 1982;163:202–9. 74. Gau YL, Lonstein JE, Winter RB, et al. Luque–Galveston procedure for correction and stabilization of neuromuscular scoliosis and pelvic obliquity: a review of 68 patients. J Spinal Disord 1991;4(4):399–410. 75. Sussman MD, Little D, Maxwell AR, et al. Posterior instrumentation and fusion of the thoracolumbar spine for treatment of neuromuscular scoliosis. J Ped Orthop 1996;16(3):304–13. 76. Allen BL, Ferguson RL. The Galveston technique for L-rod instrumentation of the scoliotic spine. Spine 1982;7:276–84. 77. Kepes ER, Martinez LR, Andrews I, et al. Anaesthetic problems in hereditary muscular abnormalities. NY State J Med 1972;72: 1051. 78. Linston M, Bresnan M, Eraklis A, et al. Acute gastric volvulus following Harrington rod instrumentation in a patient with Werdnig–Hoffman disease. Spine 1981;6:522–3. 79. Piasecki JO, Mahinpour S, Levine DB. Long-term follow up of spinal fusion in spinal muscular atrophy. Clin Orthop 1986;207: 44–54. 80. Furumasu J, Swank SM, Brown JC, et al. Functional activities of spinal muscular atrophy patients after spinal fusion. Spine 1989;14:771–5.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 446–452

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

SHOULDER

Modern management of calcifying tendinitis of the shoulder F. Lam, D. Bhatia, K. van Rooyen, J.F. de Beer Cape Shoulder Institute, Medgroup Anlin House, Plattekloof, Panorma, Cape Town, South Africa

KEYWORDS Calcifying tendinitis; Shoulder; Extracorporeal shock wave therapy; Needle aspiration irrigation; Arthroscopic shoulder surgery

Summary Calcifying tendinitis of the rotator cuff is a common disorder and the underlying cause is still not fully understood. About 90% of patients can be treated non-operatively but some are resistant to conservative treatment and surgery is indicated. Non-operative treatments include non-steroidal anti-inflammatory drugs, subacromial injection of steroid, physiotherapy, extracorporeal shockwave therapy and needle aspiration irrigation. When conservative treatment fails, arthroscopic excision of calcium, sometimes combined with an acromioplasty and/or rotator cuff repair, reliably produces excellent results with high patient satisfaction. In this article, an up-to-date review of the published papers evaluating each treatment modality is presented. & 2006 Elsevier Ltd. All rights reserved.

Introduction The term ‘calcifying tendinitis’ was first coined by De Seze and Welfling1 and is preferred to calcific tendinitis as this reflects better the continually changing nature of the disease. It is a common disorder of the rotator cuff and accounts for approximately 10% of all consultations for painful shoulder. It affects women more often than men; its peak incidence is in the fifth decade. The prevalence among asymptomatic individuals was reported to be 2.7% by Bosworth, who studied 6061 volunteers from an insurance office.2 The histopathological findings of calcifying tendinitis have been extensively reported by Uhthoff, who described three distinctive stages through which the disease process Corresponding author.

E-mail address: [email protected] (F. Lam). 0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.09.005

progresses.3 The first stage is the precalcific stage, characterised by metaplasia of the tendinous tissue into fibrocartilage. This is followed by the calcific stage, which consists of a phase of formation and a phase of resorption. In the post-calcific stage, following resorption of the calcium deposit, tendon reconstitution occurs. For a detailed description of the pathological stages of calcifying tendonitis the reader is advised to refer to a previous issue of this journal, where it is well described.4

Radiological classification The radiographic findings of calcifying tendinitis were first described by Painter in 1907. Since then, several authors have proposed various classification systems based on the size of the deposit on radiographs,2 stage of the disease process5 and its morphological appearance.6,7 As calcifying tendinitis is a multifocal and polymorphic disease, with

ARTICLE IN PRESS Modern management of calcifying tendinitis of the shoulder

447

different parts of the tendon simultaneously undergoing varying stages of the evolutionary process, these classification systems serve as a useful guide to treatment and ensure that therapy is targeted according to the individual and to the stage of the disease (Figs. 1–3, Tables 1–4).

Non-operative treatment Conservative treatment is usually successful in up to 90% of patients. The main treatment modalities are:

    

non-steroidal anti-inflammatory drugs, subacromial injection of steroid, physiotherapy, extracorporeal shockwave therapy, needle aspiration and irrigation.

Non steroidal anti-inflammatory drugs are the mainstay of non-operative treatment. Although steroid injections are commonly used in the treatment of calcifying tendinitis, there is still no conclusive evidence that they promote resorption of the calcium deposit. Uhthoff and Sarkar8 believe that steroids actually impair the cell-mediated resorption of carbonated apatite crystals. Noel et al.9 found that steroid injections administered before needle aspiration had no effect on the clinical outcome. The efficacy of physiotherapy in the form of therapeutic ultrasound, in the treatment of calcifying tendinitis, remains uncertain. The Cochrane Musculoskeletal Database Review of twenty six trials found that both ultrasound and pulsed electromagnetic field therapy resulted in significant improvement in pain, compared to placebo, in calcific tendonitis.10 However, a further meta-analysis of 35 randomised controlled trials, of which 10 were suitable for inclusion, found that only 2 studies supported the use of therapeutic ultrasound over placebo. The remaining 8 showed that therapeutic ultrasound is no more effective than placebo.11

Figure 2 Heterogeneous well-defined calcific deposit.

Figure 3 Heterogeneous ill-defined calcific deposit with a punctate appearance.

Table 1 Bosworth’s classification based on the size of the calcium deposit on the radiograph.2 Small Medium Large

o0.5 cm 0.5–1.5 cm 41.5 cm

Extracorporeal shock wave therapy

Figure 1 Homogeneous well-defined calcific deposit.

Extracorporeal shock wave therapy utilises acoustic waves to induce fragmentation of the mechanically hard crystals. Its use as an alternative treatment for calcifying tendinitis

ARTICLE IN PRESS 448

Table 2

F. Lam et al. De Palma and Kruper classification.5

Type

Radiological appearance

Correspondence to Uhthoff’s pathological stages

I

Fluffy, fleecy with ill-defined periphery Homogeneous with clearly defined periphery

Resorptive phase

II

Formation phase

Table 3 French Arthroscopic Society classification based on appearance.6 Type A Type B

Type C

Type D

Homogeneous calcification with welldefined limits (Fig. 1) Heterogeneous and fragmented calcification with well-defined limits (Fig. 2) Heterogeneous calcification with poorly defined limits and sometimes with a punctate appearance (Fig. 3) Dystrophic calcification of the tendon insertion

Table 4 Patte and Goutallier classification based on morphology.7 I II III IV

Sharp and dense Blunt and dense Sharp and translucent Blunt and translucent

has gained increasing popularity in the last few years, especially in Europe. The efficacy of extracorporeal shock wave therapy has been confirmed in several prospective studies and favourable results have been reported in terms of patient satisfaction, improvement in functional scores and disappearance of calcific deposit confirmed radiographically.12–15 A recent single-blind, randomised controlled study of 90 patients with radiographically verified calcific tendinitis found that extracorporeal shock wave therapy led to complete disappearance of calcifications in 86.6% of the subjects in the treatment group and reduction in size of deposit in 13.4% of subjects.13 In the control group, only 8.8% of the subjects displayed partially reduced calcifications and none disappeared totally. There was significant reduction in pain and improvement of shoulder function after 4 weeks, with no adverse effects reported. The optimum energy level for extracorporeal shock wave therapy, for successful treatment of calcifying tendinitis, was evaluated by Peters et al.14 who compared extracorporeal shock wave therapy at two different energy levels (0.15

and 0.44 mJ/mm2) with placebo. Those treated with a lower energy level of 0.15 mJ/mm2 had significantly less pain during treatment but required more treatments and had a significantly higher recurrence of calcification at the 6 months follow-up. On the other hand, those treated with a higher energy level of 0.44 mJ/mm2 had no residual calcification or recurrence of pain. It seems that the effectiveness of extracorporeal shock wave therapy is directly related to the energy level. Overall, there were no major side effects with either treatment, except for a small number of haematomas. Most of the studies demonstrating the effectiveness of extracorporeal shockwave therapy in the treatment of calcific tendinitis have the common limitation of having only a short term follow up. Daecke et al.15 carried out a prospective long-term follow-up study and found that 4 years after the shockwave therapy, 20% of the study population had undergone surgery on the involved shoulder. Thus, it seems that the failure rate following extracorporeal shockwave therapy is higher than previously reported.

Needle aspiration and irrigation The aim is to drain a substantial portion of the calcium deposit, thereby stimulating cell-mediated progressive resorption. There are data to suggest that the outcome following this procedure is directly related to the amount of calcium that can be aspirated.16,17 The procedure was initially done under fluoroscopic guidance but the use of ultrasound has provided greater accuracy without the risk of irradiation.18 Needle aspiration has an advantage over arthroscopic treatment in that this can be readily done under local anaesthesia in the outpatient setting. The procedure is most easily accomplished with the patient in the lateral decubitus position. Local anaesthetic infiltration along the soft tissue planes leading to the calcium deposit is administered. Using a large bore needle (15 G), the calcium deposit is punctured under direct ultrasound guidance (Fig. 4). An attempt at aspiration should be made, and sometimes the creamy material can be withdrawn from the needle. A second needle is introduced

Figure 4 Ultrasound image of a needle traversing into the calcium deposit.

ARTICLE IN PRESS Modern management of calcifying tendinitis of the shoulder

449

anteriorly and saline is injected, thus creating an inflowoutflow irrigation system between the two needles. It is important to avoid multiple punctures of the deposit, as this causes leakage of the irrigation fluid, making the flushing less effective. On completion of the procedure a steroid injection is often given. The major drawback for this procedure is hyperalgesic rebound, which is most common in the first 2–3 days following needling. The best results are obtained in patients with an acutely painful shoulder, typically during the resorption stage in which the calcium is of toothpaste like consistency. It is contraindicated in patients with adhesive capsulitis and probably not suitable for small (o1.5 cm) ill-defined deposits, which are difficult to locate and deposits which are of hard chalky consistency, commonly found in the formation stage.9 Favourable results have been reported by Aina et al.19 who prospectively evaluated 30 consecutive patients with a follow-up of 1 month and found that there was a statistically significant improvement in pain (30.5%) and disability (23.9%) index scores. In a longer term follow up study of 1 year by Farin et al.,20 74% (45 of 61) of the calcifications had decreased in size, including 28% (17 of 61) which disappeared totally, and 26% (16 of 61) that were unchanged. Clinical results were excellent in 74% (45 of 61), moderate in 16% (10 of 61) and poor in 10% (6 of 61) of cases. One of the most recent reports comes from Galletti et al.,21 who found that after a period of 2 years there was complete regression of calcification in 67% of patients and more than 60% reduction in calcific deposits in another 28% of patients on X-ray. Considerable reduction in symptoms was also evident in 87% of patients within a few days of treatment. They concluded that ultrasound-guided needle aspiration is simple to execute, with low cost and is easily repeatable, with good relief of symptoms. One prospective randomised controlled trial compared the outcome of patients treated with ultrasound-guided needling followed by high-energy shock-wave therapy versus shock-wave therapy alone.22 They found that the group treated with ultrasound-guided needling followed by highenergy shock-wave therapy had significantly better clinical results, with a reduction in the need for surgery. Complete disappearance of the calcific deposit was also observed in 60% compared with 32.5% in the group treated with shockwave therapy alone.

calcium and acromioplasty.25 Proponents of the open approach argue that this is technically simpler to perform and the defect within the tendon can also be repaired easily and quickly.24 With advances in technology, these superior results have also been reproduced arthroscopically.27 The procedure involves a glenohumeral arthroscopy with special attention to the ‘critical zone’ of the rotator cuff. A cherry red spot is often visible on the articular side of the rotator cuff close to the footprint and represents an area of increased vascular proliferation (Fig. 5). This is a useful landmark for the location of the calcium deposit and some surgeons recommend marking out this lesion with a suture to aid subsequent identification of the deposit in the subacromial space. Next, a subacromial bursoscopy and bursectomy is carried out to adequately visualise the rotator cuff. When an acromioplasty is indicated, the coracoacromial ligament is released using electrocautery and a subacromial decompression carried out using a burr. The calcific deposit is usually self-evident and is most commonly found in the supraspinatus, 1.5–2 cm from its attachment to the greater tuberosity. In cases where it is difficult to identify the calcium deposit a spinal needle can be used to probe the rotator cuff. Intraoperative fluoroscopy is sometimes helpful in locating the deposit and confirms that the evacuation is complete at the end of the procedure. Once the calcium deposit is identified, the capsule is carefully incised with an arthroscopic knife in line with fibre orientation of the tendon (Fig. 6). To minimise tendon damage a blunt instrument, such as a curette, is then used to milk out the toothpaste-like contents (Fig. 7). When the calcium is of a hard chalky texture, the arthroscopic rotating blade can be used to decompress the deposit, creating a typical snowstorm appearance. At the end of the procedure, a thorough washout of both the glenohumeral joint and subacromial space is necessary to prevent leaving behind any calcium fragments and some surgeons also give an intraarticular steroid injection. Washout is thought to be important to prevent the development of secondary stiffness, which is relatively common following calcium deposit removal and has been reported in 9–15% of cases.6,28 It is thought to be caused by residual calcium fragments provoking an inflammatory reaction within the subacromial bursa, triggering the so called hyperalgesic crisis.28

Operative treatment Whilst there is still controversy regarding the optimal operative treatment, most would agree that in patients with severe disabling symptoms which have persisted for more than 6 months and are resistant to conservative treatment, surgery is indicated.23 The first case of operative removal of calcific deposit was carried out by Harrington and Codman in 1902. Since then, favourable results have been reported by numerous authors with a subjective improvement of 82% and 71% achieving excellent objective results following open excision of the calcium deposit via a deltoid split approach combined with an acromioplasty.24–26 Similarly, good results were reported by Rochwerger et al. who found that the Constant score improved from 52 to 89 after a mean follow-up of 23 months following open removal of

Figure 5 tion.

Cherry spot—area of increased vascular prolifera-

ARTICLE IN PRESS 450

Figure 6 Incision of the calcium deposit in line with the tendon fibre orientation.

Figure 7 Arthroscopic view of the toothpaste like contents from the calcific deposit.

How important it is to remove all of the calcium deposit remains uncertain. Jerosch29 has shown that the functional outcome following surgery is inversely related to the amount of calcification remaining. More recent studies, however, suggest that absolute eradication of the calcium deposit is probably not necessary, as cell-mediated resorption is already initiated by the surgery.30

Acromioplasty The question of whether acromioplasty should be performed or not remains controversial. Some surgeons, such as Neer,31 strongly believe that calcific tendinitis is independent of rotator cuff impingement. This is supported by Uhthoff’s3 histological findings that the subacromial bursa in calcifying tendinitis usually has minimal signs of inflammation. Other researchers believe that the vascular invasion and influx of

F. Lam et al. phagocytic cells during the acute resorptive phase lead to oedema of the rotator cuff and rise in the intratendinous pressure.30 This theoretically can lead to secondary impingement as the thickened and indurated calcified tendon bulges into the subacromial space. However, radiological studies have found little correlation between calcifying tendinitis and osseous subacromial impingement, with only 16% of patients with calcifying tendinitis having the socalled type III or hooked acromion on supraspinautus outlet view radiograph.32 Interestingly, Resch et al.33 reported that patients with diffusely spread small (o5 mm) calcium deposits often had fair or poor functional outcome following deposit removal alone and pain was only relieved later by performing an additional acromioplasty. They therefore advocated that patients with diffusely spread small calcium deposits without evidence of substantial surrounding inflammatory changes should also have an acromioplasty. Some authors even advocate performing an acromioplasty alone without excision of the calcium deposit.25,34 The study from Tillander et al.34 showed that by performing a subacromial decompression alone without interfering with the calcific deposit, 79% of patients had disappearance or decrease in the size of the calcific deposit after a mean period of 2 years from surgery. They question whether the calcific deposits disappear more quickly after an acromioplasty as a result of a reduction in the pressure within the subacromial space. Thus, summarising the currently available evidence, the most commonly accepted indications for performing an acromioplasty are:

1. Radiological evidence of mechanical impingement e.g. type III acromion, sclerosis of undersurface of acromion and greater tuberosity. 2. Intraoperative evidence of mechanical impingement e.g. kiss lesion—partial bursal sided rotator cuff tear with mirror changes on the undersurface of anterior acromion. 3. Type C calcium deposits with an ill-defined contour and heterogeneous appearance on X-ray.6 This is because the calcium deposit is diffusely infiltrated and even following surgery, some minute microscopic deposits of calcium will inevitably remain within the tendon.

Repair of the rotator cuff Traditionally, it was thought that calcifying tendinitis progresses through distinct stages, as described by Uhthoff,3 and the tendon always reconstitutes after calcium deposit removal. Neer35 recommended excision of the calcifying tendon as ‘a quarter orange’ without the need for complementary suturing. Recent evidence, however, suggests that spontaneous healing of the tendon does not always occur and the cyclical natural history can be interrupted at any stage of the disease.30,36 Seil et al.30 found that 65% of patients 2 years following surgery had persistent discrete flattening of the tendon on ultrasound. The incidence of persistent rotator cuff defects following surgery has been found to be 25% with 7% having persistent pain.36 This is more common following removal of large (42 cm) deposits. Some surgeons therefore recommend a

ARTICLE IN PRESS Modern management of calcifying tendinitis of the shoulder

451

primary side to side repair of the rotator cuff defect if the residual defect following excision is large.37

16. Normandin C, Ortiz-Bravo E. Ponctions de calcification de l’epaule. Concours Med 1987;109:2559–64. 17. Nizard J, Maugars Y, Horreard F, Castagne A, Prost A. Ponctionfragmentation-lavage de calcifications tendineuses de l’epaule: etude prospective de 40 cas. In: Pelisser J, Simon L, Rodineau J, editors. Pathologie de la Coiffe des Rotateurs de L’Epaule, Collection de Pathologie Locomotrice. Paris: Masson; 1993. p. 120–6. 18. Cooper G, Lutz GE, Adler RS. Ultrasound-guided aspiration of symptomatic rotator cuff calcific tendonitis. Am J Phys Med Rehabil 2005;84(1):81. 19. Aina R, Cardinal E, Bureau NJ, Aubin B, Brassard P. Calcific shoulder tendinitis: treatment with modified US-guided fineneedle method. Radiology 2001;221(2):455–61. 20. Farin PU, Rasanen H, Jaroma H, Harju A. Rotator cuff calcifications: treatment with ultrasound-guided percutaneous needle aspiration and lavage. Skeletal Radiol 1996;25(6): 551–4. 21. Galletti S, Magnani M, Rotini R, Mignani G, Affinito D, Pelotti P, et al. The echo-guided treatment of calcific. Chir Organi Mov 2004;89(4):319–23. 22. Krasny C, Enenkel M, Aigner N, Wlk M, Landsiedl F. Ultrasoundguided needling combined with shock-wave therapy for the treatment of calcifying tendonitis of the shoulder. J Bone Joint Surg Br 2005;87(4):501–7. 23. Rotini R, Bungaro P, Antonioli D, Katusic D, Marinelli A. Algorithm for the treatment of calcific tendinitis in the rotator cuff: indications for arthroscopy and results in our experience. Chir Organi Mov 2005;90(2):105–12. 24. Gazielly DF, Bruyere G, Gleyze PTT. Open acromioplasty with excision of calcium deposits and tendon suture. In: Gazielly DF, Gleyze PTT, editors. The cuff. Paris: Elsevier; 1997. p. 172–5. 25. Rochwerger A, Franceschi JP, Viton JM, Roux H, Mattei JP. Surgical management of calcific tendinitis of the shoulder: an analysis of 26 cases. Clin Rheumatol 1999;18(4):313–6. 26. Postel JM, Goutallier D, Lambotte JC, Duparc F. Treatment of chronic calcifying or post calcifying shoulder tendinitis by acromioplasty without excision of the calcification. In: Gazielly DF, Gleyze PTT, editors. The cuff. Paris: Elsevier; 1997. p. 159–63. 27. Costouros JG, Bassi O, Gerber C. Arthroscopic management of calcific tendonitis of the shoulder. Presented at the American academy of orthopaedic surgeons annual meeting, 2006. p. 695. 28. Kempf JF, Bonnomet F, Nerisson D, Gastaud F, Lacaze F, Geraud H. Arthroscopic isolated excision of rotator cuff calcium deposits. In: Gazielly DF, Gleyze PTT, editors. The cuff. Paris: Elsevier; 1997. p. 164–7. 29. Jerosch J, Strauss JM, Schmiel S. Arthroskopische Therapie der Tendinitis calcarea. Wie wichtig ist die Kalkentfernung? Arthroskopie 1996;9:241–5. 30. Seil R, Litzenburger H, Kohn D, Rupp S. Arthroscopic treatment of chronically painful calcifying tendinitis of the supraspinatus tendon. Arthroscopy: J Arthrosc Rel Surg 2006;22(5):521–7. 31. Neer CS. Anterior acromioplasty for the chronic impingement syndrome of the shoulder. J Bone Joint Surg (Am) 1972;54A: 41–50. 32. Loew M, Sabo D, Wehrle M, Mau H. Relationship between calcifying tendinitis and subacromial impingement: a prospective radiography and magnetic resonance imaging study. J Shoulder Elbow Surg 1996;5:314–9. 33. Resch H, Povacz P, Seykora P. Excision of calcium deposit and acromioplasty? In: Gazielly DF, Gleyze PTT, editors. The cuff. Paris: Elsevier; 1997. p. 169–71. 34. Tillander BM, Norlin RO. Change of calcifications after arthroscopic subacromial decompression. J Shoulder Elbow Surg 1998;7:213–7.

Summary and conclusions Calcifying tendinitis of the rotator cuff is a polymorphic disease characterised by multifocal deposition of calcium in non degenerative tissue. The majority of patients can be treated effectively with non-operative measures such as non steroidal anti-flammatory drugs, subacromial injection of steroid, physiotherapy, extracorporeal shockwave therapy and needle aspiration irrigation. Approximately 10% are resistant to conservative treatment and surgical removal of the calcium deposit is necessary. In selected patients, a concurrent acromioplasty and rotator cuff repair is also indicated. The key to successful management is to understand the natural history of the condition thereby devising the optimum treatment based on the pathology.

References 1. De Seze S, Welfling J. Tendinites calcifiantes. Rhumatologie 1970;22:5–14. 2. Bosworth BM. Calcium deposits in the shoulder and subacromial bursitis: a survey of 12,122 cases. J Am Med Assoc 1941;116: 2477–82. 3. Uhthoff HK. Anatomopathology of calcifying tendinitis of the cuff. In: Gazielly DF, Gleyze PTT, editors. The cuff. Paris: Elsevier; 1997. p. 144–6. 4. Hughes PJ, Bolton-Maggs B. Calcifying tendinitis. Curr Orthop 2002;16(5):389–94. 5. DePalma AF, Kruper KS. Long term study of shoulder joints afflicted with and treated for calcific tendinitis. Clin Orthop 1961;20:61–72. 6. Mole D, Kempf JF, Gleyze P, Rio B, Bonnomet F, Walch G. Resultat du traitement arthroscopique des tendinopathies non rompues, Il: les calcifications. Rev Chir Orthop 1993;79:532–41. 7. Patte CF, Goutallier D. Calcifications. Rev Chir Orthop 1988;74: 277–8. 8. Uhthoff HK, Sarkar K. Calcifying tendinitis. Rockwood Jr CR, Matsen III FA, editors. The shoulder, vol. 2. Philadelphia: WB Saunders; 1990. p. 774–90. 9. Noel E, Carillon Y, Gaillard T, Bouvier M. Needle aspiration irrigation in calcifying tendinitis of rotator cuff. In: Gazielly DF, Gleyze PTT, editors. The cuff. Paris: Elsevier; 1997. p. 152–7. 10. Green S, Buchbinder R, Hetrick S. Physiotherapy interventions for shoulder pain. Cochrane Database Syst Rev 2003(2): CD004258. 11. Robertson VJ, Baker KG. A review of therapeutic ultrasound: effectiveness studies. Phys Ther 2001;81(7):1339–50. 12. Moretti B, Garofalo R, Genco S, Patella V, Mouhsine E. Mediumenergy shock wave therapy in the treatment of rotator cuff calcifying tendonitis. Knee Surg Sports Traumatol Arthrosc 2005;13(5):405–10. 13. Cacchio A, Paoloni M, Barile A, Don R, de Paulis F, Calvisi V, et al. Effectiveness of radial shock-wave therapy for calcific tendinitis of the shoulder: single-blind, randomized clinical study. Phys Ther 2006;86(5):672–82. 14. Peters J, Luboldt W, Schwarz W, Jacobi V, Herzog C, Vogl TJ. Extracorporeal shock wave therapy in calcific tendinitis of the shoulder. Skeletal Radiol 2004;33(12):712–8 [Epub 2004 October 8]. 15. Daecke W, Kusnierczak D, Loew M. Long-term effects of extracorporeal shockwave therapy in chronic calcific tendinitis of the shoulder. J Shoulder Elbow Surg 2002;11(5):476–80.

ARTICLE IN PRESS 452 35. Neer CS, Marberrey TA. Calcium deposits. In: Neer CS, editor. Shoulder reconstruction. Philadelphia, London: WB Saunders Co; 1990. p. 774–89. 36. Lam F, Chidambaram R, Mok D. Will rotator cuff defects heal after arthroscopic excision of calcium and subacromial decompression? Abstract booklet of the British shoulder and elbow society annual scientific meeting, Edinburgh, 2006.

F. Lam et al. 37. Porcellini G, Paladini P, Campi F, Paganelli M. Arthroscopic treatment of calcifying tendinitis of the shoulder: clinical and ultrasonographic follow-up findings at two to five years. J Shoulder Elbow Surg 2004;13:503–8.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 453–460

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

Radiology quiz QUESTIONS Question 1 Figures 1a and b are antero–posterior and lateral lumbar spine radiographs of an 11-year-old patient. Figure 1c is a coronal T2-weighted MRI scan of another patient with the same condition. What are the abnormalities and what is the diagnosis?

Figure 1

0268-0890/$ - see front matter doi:10.1016/j.cuor.2006.09.004

ARTICLE IN PRESS 454

E. Hoey, P. Robinson

Question 2 Figure 2a shows a Technetium 99 m bone scan from a 60-year-old female patient who presented with low back pain, exacerbated by walking. Pelvic radiographs were unremarkable. What is the abnormality and what further imaging would you advise?

Figure 2

ARTICLE IN PRESS Radiology quiz

455

Question 3 Figures 3a and b are axial and coronal fat suppressed T2-weighted MR images through the left knee of a 20-year-old female patient who presented with knee pain and swelling. What is the diagnosis?

Figure 3

ARTICLE IN PRESS 456

E. Hoey, P. Robinson

Question 4 Figure 4a shows a lateral lumbar spine radiograph of an 18-year-old male patient with low back pain exacerbated by exercise. What is the abnormality? What further imaging could confirm the diagnosis?

Figure 4

Question 5 Figure 5a is a coronal T2-weighted fat suppressed MR image of the thighs in a patient who presented with pain and swelling of the left thigh. A firm lump was palpable on the medial aspect. The patient had a history of insulin dependant diabetes secondary to lipodystrophy. What is the diagnosis? What further imaging sequence could help confirm this?

Figure 5

ARTICLE IN PRESS Radiology quiz

457

Question 6 Figures 6a–c are T2-weighted fat suppressed MR images of the posterior abdominal wall (Fig. 6a—coronal), Ilium (Fig. 6b—axial) and hip joint (Fig. 6c—axial) in an intravenous drug abuser who presented with right hip pain and fever. What are the abnormalities?

Figure 6

ARTICLE IN PRESS 458

E. Hoey, P. Robinson

ANSWERS Answer 1 Figure 1a shows dysplastic right L2, L3 and L4 pedicles. Figure 1b shows widened exit foraminae (asterisks) and posterior vertebral body scalloping (arrows). Figure 1c shows dural ectasia (arrows). The posterior surface of the vertebral bodies constitute the anterior wall of the spinal canal and may be re-modelled by enlargement of the canal (dural ectasia) or longstanding raised pressure e.g. chronic hydrocephalus or slow growing tumour of the cord such as an ependymoma. Causes of dural ectasia include Neurofibromatosis, Marfans syndrome, Ehlers-Danlos syndrome and sacral peri-neural cysts (Tarlov cysts). This patient has Neurofibromatosis Type 1. The dural ectasia has also caused the exit foraminae to be widened and has resulted in dysplasia of the right sided pedicles.

Answer 2 Intense uptake of radio-isotope in the right sacro-iliac region. MRI pelvis. Figures 2b and c are coronal oblique MR images of the pelvis. A low signal vertical fracture line is seen in the right sacral ala paralleling the sacro-iliac joint on the T1-weighted image (Fig. 2b) with corresponding high signal due to bone marrow oedema on the T2-weighted image (Fig. 2c). These are the appearances of a sacral insufficiency fracture which is caused by normal stresses on weakened bone, most commonly due to post-menopausal osteoporosis. The classical description of sacral insufficiency fracture is of an H-shaped pattern (Honda sign) in which the vertical limbs lie within the sacral alae on each side and the transverse component extends across the sacral body. However asymmetric, incomplete patterns such as in this case are commonly seen in the early stages and may progress if not treated appropriately.

Figure 2 continued.

ARTICLE IN PRESS Radiology quiz

459

Answer 3 Transient lateral patellar dislocation. See Fig. 3a*. There is bony contusion (high-signal marrow oedema) of the lateral femoral condyle (asterisk), contusion of the medial facet of the patella (arrowhead), oedema of the medial retinaculum (arrow) and a joint effusion. This constellation of findings is typical for transient lateral patellar dislocation, and results from impaction of the medial aspect of the patella against the lateral femoral condyle as it dislocates. Most dislocations undergo rapid spontaneous reduction with the patient unaware of the exact nature of their injury. Clinical examination findings rarely pin-point the diagnosis. Standard projection plain films may just show a knee joint effusion or lipohaemarthrosis. Axial (Skyline) views sometimes demonstrate impaction fractures of the medial facet of the patella and lateral margin of the lateral femoral condyle (especially in cases of recurrent dislocation). However MRI is the imaging modality of choice because of its ability to demonstrate the bone and tissue oedema that occurs with this injury.

Answer 4 Spondylolysis of L5. CT scan of lumbar spine. The plain film shows a break in the pars-interarticularis of the L5 vertebra (spondylolysis) see arrows in Fig. 4a*, of note there is no anterior subluxation of L5 on SI (spondylolisthesis). The patient went on to have a CT scan (Fig. 4bc) which shows bilateral fractures of the pars interarticularis at this level. A sagittal reformat (Fig. 4c) clearly shows the right-sided pars defect. Stress fractures of the pars are typically seen in young athletes with the L5 vertebrae most commonly affected. Patients complain of low back pain exacerbated by exercise. In the early stages before any spondylolisthesis occurs the plain film findings may be unremarkable; oblique views can be more sensitive than standard lateral radiographs (looking for a break in the neck of the ‘Scottie Dog’) but CT is the imaging modality of choice for fracture detection. There is an increasing role for MR imaging in this condition as it is also sensitive to associated marrow oedema and does not involve a high-radiation dose.

Figure 4 continued.

ARTICLE IN PRESS 460

E. Hoey, P. Robinson

Answer 5 Myonecrosis. T1-weighted post-gadolinium sequence. The left Vastus Medialis muscle and adjacent subcutaneous tissues show marked high signal (oedema) on this T2-weighted fat suppressed MR image. Linear striations due to oedema between the muscle fibres can be seen. Within vastus medialis there is a central area which has irregular margins and appears less oedematous. Figures 5b and c are T1-weighted post gadolinium axial and coronal images which show vastus medialis enhancing with relative sparing of the central area. The linear striations can be seen even more clearly on these sequences due to contrast agent extravasation from damaged vessels between the muscle fibres. These are the MRI features of myonecrosis (muscle infarction)—the central area is the necrotic infarcted tissue. A similar lesion can be seen developing in the right thigh! Also of note in this patient are diffuse high signal bone marrow changes reflecting high marrow turnover seen in Lipodystrophy. Myonecrosis is an uncommon complication of poorly controlled diabetes. Its pathogenesis is poorly understood but it is thought to reflect severe small vessel atherosclerosis causing an arteriosclerosis obliterans. The muscular arteries supplying the quadriceps muscles seem to be particularly susceptible. Recognition of the MRI appearances can avert the need to perform a biopsy and the potential risk of delayed healing and infection.

Figure 5 continued. (b) Axial T1-weighted post-gadolinium MR image of the left thigh showing enhancement of vastus medialis (asterisk) with a central area of infarction (non-enhancement) (arrow). (c) Coronal T1-weighted post-gadolinium MR image of both thighs demonstrates the muscle infarction in the left thigh (arrow) with a similar lesion seen to be developing in the right thigh (arrowhead).

Answer 6 Right iliopsoas/gluteal infective myositis. Right hip and right sacro-iliac joint septic arthritis. See Figs. 6b* and c*. The right iliopsoas (arrow) and gluteal (arrowhead) muscles are swollen and of diffusely high signal on these T2-weighted fat suppressed images. High signal fluid can also be seen within the right hip (arrow) and sacro-iliac joints. Several discrete foci (asterisk) of very high signal are visible within the muscles indicating abscesses. These are the appearances of infective myositis with abscess formation which has spread to involve the hip and sacro-iliac joints as a septic arthritis. The patient had been injecting into the right groin which is the presumed site of infection origin. It has tracked up iliopsoas, involved the hip (which this muscle bursa communicates with) and right sacro-iliac joint as well as tracking down through the sciatic notch to involve the gluteal muscles. The commonest organism (as in this case) is staphylococcus aureus. Surgical debridement is needed.

E. Hoey, P. Robinson Department of Radiology, St. James’s University Hospital, Beckett Street, Leeds LS9 7TF, UK

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 461–466

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

CHILDREN

Fractures in the child’s hand Adriaan Smita, Geoffrey Hooperb, a

Wrightington Hospital, Hall Lane, Appley Bridge, WN6 9EP, UK St. John’s Hospital, Howden Road West, Livingston, West Lothian, EH54 6PP, UK

b

KEYWORDS Fracture; Child; Hand

Summary Most children’s hand fractures can be successfully treated by either simple protective splinting or closed reduction and a short period of immobilisation. However, failure to correct rotational deformities can produce long-term problems. Certain injuries that may require early operative treatment are frequently missed. These include fractures of the neck and condyles of the proximal phalanx and dislocation of the nail combined with an epiphyseal fracture of the distal phalanx. & 2006 Elsevier Ltd. All rights reserved.

Diagnosis

frightened and in pain, accompanied by an anxious parent or guardian. Other factors to take into account are inability to perform specific tests of hand function, unwillingness to have the other hand examined and a lack of bony detail on plain radiographs. Patience and minimal handling will help establish trust with the examiner. Pain management is very important. Obtaining useful radiographs may be impossible without sedation. The treating doctor will often need to accompany the child to the X-ray department to ensure correct positioning. Radiographs of the other hand may help in identifying subtle abnormalities. Posteroanterior and lateral films are often insufficient to establish the nature of the injury. Oblique films are often more reliable in identifying intra-articular fractures of the phalanges.

Difficulties may be encountered in both clinical and radiographic assessment due to an uncooperative patient who is

Ossification of the child’s hand

Corresponding author. Tel.: +44 1506 419666; fax: 44 1506 460592. E-mail addresses: [email protected] (A. Smit), [email protected] (G. Hooper).

Radiographs of the child’s hand and wrist are notoriously difficult to interpret due to the extent of nonossified cartilage in the immature hand. Carpal injuries are particularly easy to overlook. Knowing the order of

Introduction Injuries of the hand and wrist are among the most common injuries in the skeletally immature population.1,2 Hand fractures and dislocations are uncommon in the very young, but the prevalence of these injuries increases sharply after the eighth year and peaks around age 13.2–4 This coincides with the involvement of many boys in competitive contact sports and as children are introduced to such sports at a younger age, these injuries will inevitably increase.

Special considerations in the child

0268-0890/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.cuor.2006.09.008

ARTICLE IN PRESS 462 ossification of the carpal bones may assist in identifying injuries. The centre of ossification of the capitate is the first to appear radiologically at 3–4 months after birth, closely followed by the hamate, the triquetrum at 2–3 years, the lunate at 4 years, the scaphoid (where ossification begins distally) at 4–5 years, the trapezium at 5 years, the trapezoid at 6 years and the pisiform at 9 years. It will be noted that ossification centres appear in the carpus in anticlockwise fashion (when looking at the dorsum of the right wrist). The epiphyses of the phalanges and the thumb metacarpal are situated at their bases, whereas those of the other metacarpals lie distally. Occasionally there are anomalies such as an extra basal epiphysis of the index metacarpal.

Soft tissue anatomy and its relevance The strength of the child’s soft tissues to withstand tensile forces often exceeds that of the adjacent growth plate and epiphysis. This predisposes to growth plate injuries whereas ligamentous ruptures and tendon avulsions are seldom seen. The soft tissues around the growth plate provide relatively poor protection. At the metacarpophalangeal joint the growth plates of both the metacarpal and proximal phalanx are particularly exposed due to the fact that the attachments of the collateral ligaments are on the epiphyses. Not surprisingly, fractures of the proximal phalanx are the most common injuries in the child’s hand.3 At the interphalangeal joints the collateral ligaments originate in the collateral recess of the juxta-epiphyseal region of the proximal segment and insert into the metaphysis as well as the epiphysis of the respective middle and distal phalanges. The collateral ligaments therefore protect the single growth plate at each interphalangeal joint. The palmar plate and the weak extensors provide poor protection of the growth plate. The anatomy of the palmar plates of the metacarpophalangeal and interphalangeal joints is similar. They originate from the palmar aspect of the metaphysis of the proximal segment and insert into the epiphysis of the distal segment. Extensor tendons insert into the epiphyses of the middle and distal phalanges, making these growth plates vulnerable to avulsion injuries, especially at the distal phalanx. Both superficial and deep flexor tendons insert into the shafts of their respective phalanges. Transverse midshaft fractures of the middle phalanges can be difficult to reduce or to control after reduction, due to the pull of intact flexor tendons. These fractures are often more stable with the fingers in a flexed position. The periosteum around the long bones in the hand is a tough membrane. Its mechanical properties can be used to aid reduction of fractures and maintain them in reduction. However there is a tendency for displaced fractures to perforate the periosteum, which may then prevent reduction. There should be a relatively low threshold for operative exploration of an irreducible fracture, as not only periosteum but also other structures, such as tendons, may be interposed.

A. Smit, G. Hooper

The growing skeleton Fractures in children heal more quickly than in adults but the short healing time makes early diagnosis and prompt fracture reduction essential, particularly for injuries in the vicinity of the growth plate, where 5 days is the limit for safe reduction. Repeated, forceful and late attempts at reduction must be avoided to prevent iatrogenic injury where the growth plate is involved. Compression injuries of the growth plate in the upper limb are often the result of manipulation, while in the lower limb they usually result from the initial trauma. Future growth and fracture remodelling provide a safety margin in many, but not all, fractures. Uncorrected angular deformity has the best potential for remodelling and this occurs by asymmetrical growth, with the growth plate changing its orientation to align with the mechanical axis rather than with the adjacent metaphysis.5 Remodelling is most rapid after metaphyseal injuries because they are close to the growth plate. Deformity in the plane of joint motion can be compensated for to some extent by the multiaxial movements of the metacarpophalangeal joints. However, the uniplanar movements of flexion and extension at the interphalangeal joints mean that radial and ulnar angulations of the proximal and intermediate phalanges are poorly compensated. Malrotation does not correct by remodelling. Even when considerable epiphyseal development remains, there may be only limited adaptive changes in the shape of the joint surfaces. Viewed end-on with the fingers extended, the nails of the fingers form an arc. Malrotation of even 101 can be identified by loss of the arc. The effects of rotational malunion are more marked in fractures at the bases of the fingers. The resultant overlapping of fingers in flexion should be prevented by identifying and correcting rotation at the fracture site at an early stage. The hypertrophic zone of the growth plate is most likely to fail under load. The resting and proliferative zones are stronger due to their high collagen content.1,6 Simple Salter–Harris type 1 and 2 injuries are not associated with significant long-term problems and the child can be discharged on return to full function. Salter–Harris type 3 and 4 injuries require to be followed for at least 1 year. Late growth disturbances due to premature fusion are usually seen after non-accidental injuries, thermal injuries7 and high-energy injuries.8 Fractures caused by such mechanisms should be followed up for long enough to exclude premature fusion. Parents should always be warned of the possibility of a poor outcome after intra-articular fractures, even when they have been reduced anatomically.

Treatment Most fractures can be managed by protective splinting, after reduction if required. Young children have a remarkable ability to remove their limbs from dressings and plasters. Casts should in most cases be extended to above the elbow. On the rare occasions when internal fixation is required, it can be obtained using smooth Kirschner wires.9 Should it be necessary to cross the growth plate with a K-wire, it should

ARTICLE IN PRESS Fractures in the child’s hand be inserted a right angles to the plate and be centrally placed to avoid the peripheral perichondrial ring. Such pins should be removed around three weeks after fixation. Screw fixation is an alternative to wiring but is technically much more demanding. Plate fixation should be avoided in children’s hands to prevent adhesions, prominence of the fixation device and the invariable need for a second procedure to remove the device.

Fractures of the phalanges Distal phalanx Crush injuries, such as trapping a finger in a closing door, cause fractures of the body of the phalanx with or without nail bed injury. Hyperflexion forces cause epiphyseal injuries. A radiograph should be requested in all cases of nail bed injury to exclude a fracture, which is present in many cases. Management is by careful repair of the nail bed using 6–0 or 7–0 absorbable sutures. If the nail is relatively undamaged it should be sutured back into the nail fold to act as a splint; alternatively the foil from a suture pack can be used to keep the nail fold open, but it is less effective as a splint. The characteristic epiphyseal plate injury of the terminal phalanx is sometimes called the Seymour fracture.10 It occurs in the 12–14-year-old age group and is effectively a type 2 Salter–Harris injury. The significance of the injury is

Figure 1 (a) Dislocation of the nail plate from the fold due to a fracture of the terminal phalanx, (b) Salter–Harris type 2 open fracture of the terminal phalanx.

463 that it is an open, displaced fracture with displacement of the nail from its fold (Fig. 1). Unless the nail is replaced in its correct position the fracture cannot be reduced and the risk of infection is high. The open fracture should be irrigated thoroughly before reducing the fracture and repositioning the nail. This can be done by removing the nail and replacing it in the fold, or by opening the fold laterally to allow reinsertion. The fracture is usually stable after reduction in this way and K-wire stabilisation is rarely necessary. Salter–Harris type 3 and 4 growth-plate injuries can usually be reduced by extending the terminal phalanx and ‘‘massaging’’ the displaced fragment back into place. The joint is then held in slight extension using a splint on the flexor aspect of the finger. Dorsal splintage can cause pressure necrosis of the skin. Rarely a large fragment of the epiphysis may prove irreducible and open reduction and K-wire fixation may be indicated, but this is a procedure that can be difficult and associated with complications such a poor healing and infection. It should be borne in mind that there is considerable scope for remodelling of even widely displaced epiphyseal injuries in this area in children.

Fractures of the proximal and middle phalanges These fractures are divided into fractures of the shaft, neck and condyles and those involving the growth plate. As one would expect, those involving the growth plate are the most common. Growth plate injuries The extra-articular Salter–Harris type 2 injury at the base of the proximal phalanx of the small finger is extremely common.2 It is an abduction-type injury that can occur in a variety of childhood activities. Because of the characteristic posture of the hand it is often called the ‘‘extra-octave’’ fracture. A similar deformity may be the result of a buckle fracture in the metaphyseal region. There may be considerable rotation at the site of injury (Fig. 2). Reduction is straightforward if the finger is flexed at the metacarpophlanangeal joint, which tautens the collateral ligaments and stabilises the proximal fragment. Reduction over a pen placed between the fingers as a fulcrum is best avoided because of the possible risk of pressure damage to the digital nerves. The hand can be rested on a palmar slab with the metacarpophalangeal joints flexed for a few days and thereafter the little finger may be supported against the ring finger. Intra-articular Salter–Harris type 3 and 4 injuries are more common in older children. Open reduction and stabilisation with a K-wire may be indicated if the fragment accounts for 25% of the joint surface or is displaced by more than 1.5 mm (Fig. 3). It has been suggested that a haemarthrosis in the metacarpophalangeal joint may lead to avascular necrosis of the metacarpal head.11 This could occur after an intraarticular fracture at the base of the proximal phalanx or one involving the metacarpal head, a rare injury in childhood. There is no evidence that this is a common problem but there is no harm in aspirating the joint if this situation is suspected. An abduction injury of the thumb in childhood may damage the ulnar collateral ligament but frequently there is

ARTICLE IN PRESS 464

A. Smit, G. Hooper films due to overlying shadows. A lateral film of the injured digit is necessary. Although remodelling of angulated fractures does occur, dorsal angulation of the proximal phalanx should be corrected as it results in imbalance between the flexors and extensors of the fingers, giving a clawed appearance. Oblique fractures with unacceptable shortening are rare but are an indication for internal fixation using transverse wires.

Figure 2 (a) Fracture of the base of the proximal phalanx of the small finger. Note the rotation which must be corrected, (b) Salter–Harris type 2 fracture.

Fractures of the neck and condyles Such injuries account for only about 1% of phalangeal fractures in children.4 Their significance is often not appreciated by junior staff who first deal with the child. Fractures of the neck of the phalanx in young children are difficult to diagnose and difficult to treat. They typically occur in young children when a finger is trapped in a door. The bone is poorly ossified in the young child and the extent of the injury can only be inferred from radiographs (Fig. 4). There should be a readiness for early open exploration of such injuries as displaced fragments are impossible to reduce after a few days.12 Failure to reduce displaced fractures results in deformity and loss of movement. Callus formation can cause a mechanical block to flexion and it may be necessary to remove obstructing spikes later from the subcondylar fossa to improve the range of motion.13 Nevertheless, in late presentations (more than a few days after injury) the deformity should be accepted and corrected later, rather than risk further damage by interfering with the fracture in the healing phase, which might result in avascular necrosis of the head of the phalanx. In very young children the remodelling of an untreated displaced fracture of the neck of the phalanx can be surprisingly good14,15 but this cannot be relied on in the older child. Fractures involving the distal third of the proximal phalanx are approached by elevating the lateral bands through a curved dorsal skin incision, or by elevating a V-shaped slip of the extensor tendon based on the central slip, which allows excellent visualisation of the PIP joint. Fractures of the neck and those through the condyles are held by cross pinning. Unicondylar or avulsion fractures are held by wires or screws inserted parallel to the joint (Fig. 5).

Figure 3 Displaced type 3 fracture of the basal epiphysis of the proximal phalanx of the small finger. Open reduction and fixation is indicated with this degree of displacement but is not always necessary.

an epiphyseal injury or a buckle injury of the adjacent metaphysis of the proximal phalanx. A type 3 fragment that is widely displaced may require open reduction and fixation using a transverse K-wire placed in the epiphysis. Buckle fractures heal and remodel very quickly.

Fractures of the shaft These are relatively uncommon. Dorsal angulation of transverse midshaft fractures is easily overlooked on lateral

Figure 4 A grossly displaced fracture of the neck of the proximal phalanx with 901 dorsal angulation.

ARTICLE IN PRESS Fractures in the child’s hand

465

Figure 6 A type 2 epiphyseal injury at the base of the thumb metacarpal. Remodelling is usually rapid.

Figure 5 (a) Displaced unicondylar fracture of the proximal phalanx. (b) Anatomical reduction and fixation with a small fragment screw.

K-wires should not be placed through the lateral bands, as these must be allowed to move freely after fixation has been achieved. Neck and condylar fractures of the middle phalanx are managed in similar fashion but it is often possible to hold a displaced transverse fracture through the neck of the intermediate phalanx with a longitudinal K-wire. This is driven distally from the fracture site and used as a joystick to place the fragment in its correct position relative to the terminal phalanx, before driving the wire out through the terminal phalanx and back across the fracture and into the shaft of the intermediate phalanx.

Metacarpal fractures Metacarpal fractures are seldom problematic due to effective splinting of the shaft and protection of the growth plate by surrounding soft tissues. The remodelling potential is considerable after fractures of the metacarpal neck and shaft and there is seldom a poor outcome unless malrotation has been overlooked, although this is rarely present. In most cases a protective cast only is required until the acute discomfort subsides. Midshaft fractures with sufficient displacement to cause shortening are rare in childhood. Again the remodelling potential is significant, but very occasionally fixation using an intramedullary pre-bent K-wire might be considered, particularly if there are multiple fractures and soft-tissue damage.

Fractures of the metacarpal shaft of the thumb are common and again remodel quite quickly. They are often overtreated since there is concern that the function of the thumb will be affected if there is even slight angulation. In fact the multiplanar motion of the trapeziometacarpal joint readily compensates for any malunion and 301 of lateral angulation will rapidly disappear with remodelling. If reduction is required, it is readily achieved by pressure over the fracture and counter-pressure at the metacarpal head. Basal metacarpal fractures of the thumb are either growth plate injuries or buckle fractures of the metaphysis. The latter heal and remodel very rapidly. Salter–Harris type 2 injuries are very common and occur by various mechanisms of injury (Fig. 6). Since the injury usually occurs after forceful abduction of the thumb, the metaphyseal fragment attached to the growth plate lies on the ulnar side. Reduction may be incomplete because of inability to stabilise the proximal fragment but remodelling is usually rapid. The metaphyseal fragment lies on the palmar side in the much rarer type 2 injury caused by forced adduction. In this pattern of fracture the periosteum may become interposed between the bone fragments, making closed reduction impossible. Type 3 and 4 injuries are rare and a decision must be made about open reduction and stabilisation with K-wires if displacement is marked. Closed manipulation will seldom reduce these fractures when they are displaced and closed insertion of K-wires is not recommended. The wires may be removed after three weeks and free mobilisation is allowed.

Carpal fractures The carpal bones are resistant to injury in early childhood since they are largely cartilaginous and ligamentous injuries of the carpus are also rare. Not uncommonly in the older child a scapholunate injury may be suspected after a wrist radiograph has been taken following an injury. However the scapholunate gap appears to be wide in the normal child’s wrist because ossification is incomplete. A comparison film of the other wrist will clarify this situation.

ARTICLE IN PRESS 466

A. Smit, G. Hooper tree. As with adults in the same situation, there may be other, possibly life-threatening injuries that draw attention away from the wrist. Even if the wrist is examined radiologically, the injury may not be recognised due to inexperience in assessing radiographs of children’s wrists. Trainees should be taught the importance of Gilula’s lines19 when assessing wrist radiographs (Fig. 7). A clue to the presence of a more major carpal injury is a displaced scaphoid fracture. The capitate is the bone that is most frequently injured in association with the scaphoid and the proximal pole may rotate through 1801. Distraction films of the wrist may be helpful in identifying these complex injuries, which almost invariably require open reduction and stabilisation with K-wires.

Figure 7 Gilula’s lines drawn at the proximal border of the carpus and the margins of the midcarpal joint. They are parallel in the normal wrist. When there is a fracture-dislocation the lines become discontinuous and may overlap.

Scaphoid fractures The pattern of scaphoid fractures in children differs from those seen in adults. The typical fracture through the scaphoid waist due to a hyperextension injury is relatively uncommon until later childhood. In children 50% of fractures occur at the distal pole, most often on the dorsoradial aspect, as a result of ligamentous avulsions.16 Treatment of scaphoid fractures in children is no different from that in adults. A clinically suspect fracture without abnormal radiographic findings is treated in a thumb spica cast, usually extending above the elbow in younger children. Further treatment is based on clinical findings rather than special investigations as bone scans are not particularly reliable in children and an MR scan may require a general anaesthetic. Most fractures are minimally displaced, stable and suitable for closed treatment. They will heal after 6–8 weeks treatment in a cast. Open reduction and fixation is usually only necessary when the scaphoid fracture is part of a more extensive carpal injury. Nonunions are seldom seen in scaphoid fractures in childhood due to the rich blood supply of the bone, the predominantly distal situation of the fracture and the excellent healing potential after childhood fractures. Even in established nonunions of the waist where there are cystic changes, casting is the first line of treatment and union will often occur.16 The natural history of scaphoid nonunions in childhood has not been documented in the longer term, but is seems unlikely that they follow the same course as in adults, with the gradual onset of characteristic patterns of osteoarthritis. Only minimal symptoms have been reported at long–term follow-up of the so-called bipartite scaphoid, which is now regarded as being a long-standing scaphoid nonunion rather than a congenital variation.17,18

Other carpal injuries Carpal fracture-dislocations do occur in children, almost invariably after a high-energy injury such as a fall from a

References 1. Grad JB. Children’s skeletal injuries. Orthop Clin North Am 1986;3:437–49. 2. Hastings H, Simmons BP. Hand fractures in children. A statistical analysis. Clin Orthop Rel Res 1984;188:120–30. 3. Worlock PH, Stower MJ. The incidence and pattern of hand fractures in children. J Hand Surg [Br] 1986;11B:198–200. 4. Barton NJ. Fractures of the phalanges of the hand in children. Hand 1979;11:134–43. 5. Karaharju EO, Ryoppy SA, Makinen R. Remodelling by asymmetrical epiphyseal growth: an experimental study in dogs. J Bone Joint Surg Br 1976;58B:122–6. 6. Torre BA. Epiphyseal injuries in the small joints of the hand. Hand Clin 1988;4:113–20. 7. Nakazato T, Ogino T. Epiphyseal destruction of children’s hands after frostbite. J Hand Surg Am 1986;11A:289–92. 8. Vickers DW. Premature incomplete fusion of the growth plate: causes and treatment by resection (physiolysis) in 15 cases. Aust NZ J Surg 1980;.50:398–9. 9. Campbell RM. Operative treatment of fractures and dislocations of the hand and wrist region in children. Orthop Clin North Am 1990;21:217–43. 10. Seymour N. Juxta-epiphyseal fracture of the terminal phalanx of the finger. J Bone Joint Surg Br 1966;48B:347–9. 11. McElfresh EC, Dobyns JH. Intraarticular metacarpal head fractures. J Hand Surg Am 1983;8A:383–93. 12. Al-Qattan MM. Phalangeal neck fractures in children: classification and outcome in 66 cases. J Hand Surg [Br] 2001;26B: 112–21. 13. Simmons BP, Peters TT. Subcondylar fossa reconstruction for malunion of fractures of the proximal phalanx in children. J Hand Surg [Am] 1987;12A:1079–82. 14. Hennrikus WL, Cohen MR. Complete remodelling of displaced fractures of the neck of the phalanx. J Bone Joint Surg [Br] 2003;85B:273–4. 15. Cornwell R, Waters PM. Remodeling of phalangeal neck fracture malunions in children. J Hand Surg [Am] 2004;29A:458–61. 16. Vahvanen V, Westerlund M. Fractures of the carpal scaphoid in children. Acta Orthop Scand 1980;51:909–13. 17. Louis DS, Calhoun TP, Garn SM, et al. Congenital bipartite scaphoid—fact or fiction? J Bone Joint Surg Am 1976;58A: 1108–12. 18. Littlefield WG, Friedman RR, Urbaniak JR. Bilateral non-union of the carpal scaphoid in a child. J Bone Joint Surg Am 1995; 77A:124–5. 19. Gilula LA. Carpal injuries: analytic approach and case exercises. AJR 1979;133:503–17.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 467–469

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

CME SECTION Three CME points available The following series of questions are based on the Mini Symposium on Management of Fractures around the Knee Joint. Please read the articles in the Mini Symposium carefully and then complete the self-assessment questionnaire by filling in the square corresponding to your response to each multiple-choice question. For true or false questions, please fill in one square only. After completing the questionnaire, either post or fax the answer page back to the Current Orthopaedics Editorial Office at the address at the bottom of the RESPONSE sheet. Please photocopy this page if you wish to keep your copy of Current Orthopaedics intact. Replies received before the next issue of Current Orthopaedics is published will be marked and those reaching an adequate standard will qualify for three external CME points. You will be notified of your marks and a CME certificate will be dispatched for your records.

Questions

4. What proportion of patients suffering a floating knee injury will have associated knee ligament injury?

1. What degree of displacement in fractures of the patella is recommended as the threshold for surgical treatment?

(a) Less than 5%, as the occurrence of fractures ‘protects’ the knee from ligament injury. (b) About 1 in 10. (c) About 1 in 5. (d) About 1 in 3. (e) More than 1 in 2.

(a) Any articular incongruity whatsoever or fracture displacement of 2 mm. (b) Any articular incongruity whatsoever or fracture displacement of more than 5 mm. (c) Articular displacement of 2 mm or more or fracture displacement of 3 mm or more. (d) Articular displacement of 2 mm or more or fracture displacement of more than 5 mm. (e) Articular displacement or fracture displacement of 5 mm or more. 2. Which of the following injuries is never an essential element of a floating knee injury? (a) Midshaft fracture of the tibia. (b) Schatzker VI fracture of the tibial plateau. (c) Subtrochanteric femoral fracture. (d) Supracondylar Salter Harris grade II growth plate femoral injury. (e) Transverse displaced patellar fracture.

5. Which of the following associated injuries is best treated by reconstruction at the time of fracture fixation or as soon as possible thereafter? (a) Medial collateral ligament injury. (b) Anterior cruciate ligament rupture. (c) Posterior criuciate ligament rupture. (d) Combined anterior and posterior cruciate ligament rupture. (e) Posterolateral corner injury. 6. What is the approximate incidence of delayed union in either one of the fractures making up a floating knee injury? (a) About 1 in 20. (b) About 1 in 10. (c) About 1 in 5. (d) About 1 in 3. (e) About 2 in 3.

3. Which of the following most frequently complicates a floating knee injury?

7. Which of the following factors has least effect on the outcome after floating knee injury?

(a) Superficial femoral artery laceration. (b) Popliteal or posterior tibial artery laceration. (c) Peroneal nerve injury. (d) Posterior tibial nerve injury. (e) Sciatic nerve traction injury.

(a) AO fracture grade in femur and tibia. (b) Length of time between injury and fixation of tibia. (c) Method of fixation used. (d) Severity of soft tissue injury in tibia. (e) Involvement of the knee joint by fracture line.

0268-0890/$ - see front matter doi:10.1016/j.cuor.2006.12.002

ARTICLE IN PRESS 468 8. What is the approximate incidence of lateral meniscal tears in Type II lateral plateau fractures depressed by 6 mm? (a) 100% (b) 80% (c) 50% (d) 25% (e) 10% 9. Which of the following is most commonly found in the population of patients with tibial plateau fractures? (a) Anterior 1/3 tears of the lateral meniscus. (b) Posterior 1/3 tears of the lateral meniscus. (c) Anterior 1/3 tears of medial meniscus. (d) Posterior 1/3 tears of medial meniscus. (e) Anterior cruciate ligament avulsion. 10. Which of the following is the most important early rehabilitation goal after minimally invasive treatment of tibial plateau fractures? (a) To regain full extension of the knee. (b) To regain 901 of knee flexion. (c) To be able to bear 50% of the body weight through the knee. (d) To be able to fully weight bear on the operated side. (e) Restoration of normal proporioception.

CME SECTION 11. Which of the following associations is strongest in terms of the proportion of patients with the injury having the named associated lesion? (a) ACL tears in type II and type IV fractures. (b) Arthroscopically diagnosed ACL injury after tibial plateau fracture. (c) Arthroscopically diagnosed meniscal injury after tibial plateau fracture. (d) MRI diagnosed complete ligament ruptures in displaced tibial plateau fractures. (e) MRI diagnosed meniscal injury in undisplaced tibial plateau fractures. 12. Which of the following fixation techniques has been shown experimentally to provide superior longitudinal stiffness and resistance to depression when used to treat type II tibial plateau fractures? (a) 3.5 mm raft screws alone. (b) 3.5 mm raft screws with medial external fixator. (c) 3.5 mm raft screws with an antiglide plate. (d) Large fragment buttress plate. (e) 6.5 mm lag screws with large fragment buttress plate.

ARTICLE IN PRESS CME SECTION

469

Please fill in your answers to the CME questionnaire above in the response section provided below. A return address and fax number is given at the bottom of the page. ...............................................................................................

Responses Please shade in the square for the correct answer. 1 2 3 4 5 6 7 8 9 10 11 12

A A A A A A A A A A A A

& & & & & & & & & & & &

B B B B B B B B B B B B

& & & & & & & & & & & &

C C C C C C C C C C C C

& & & & & & & & & & & &

D D D D D D D D D D D D

& & & & & & & & & & & &

E E E E E E E E E E E E

& & & & & & & & & & & &

Your details (Print clearly) NAME. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ADDRESS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . FAX NO. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . EMAIL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

RETURN THE COMPLETED RESPONSE FORM by fax to +44-113-392-3290, or by post to CME, Current Orthopaedics, Academic Department of Orthopaedic Surgery, ‘‘A’’ Floor Clarendon Wing, Leeds General Infirmary, Great George Street, Leeds LS1 3EX, UK.

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 470

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

CME SECTION Answers to CME questions in Vol. 20, issue 5 Please find below the answers to the Current Orthopaedics CME questions from Vol. 20, issue 5 which were based on the article—‘‘Management of metastatic disease of the appendicular skeleton’’ by R.U. Ashford, S. Pendlebury and P.D. Stalley. 1

A&

B&

C&

D&

E’

2

A&

B&

C’

D&

E&

3

A&

B&

C&

D’

E&

4

A&

B’

C&

D&

E&

5

A&

B’

C&

D&

E&

6

A’

B&

C&

D&

E&

7

A’

B&

C&

D&

E&

8

A&

B&

C&

D’

E&

9

A&

B’

C&

D&

E&

10

A&

B&

C&

D&

E’

11

A&

B&

C&

D&

E’

12

A&

B&

C&

D’

E&

doi:10.1016/j.cuor.2006.07.004

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 471

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

BOOK REVIEWS Rockwood and Green’s Fractures in Adults, sixth ed. Robert W. Bucholz, James D. Heckman, Charles CoutrBrown (Eds.). Lippincott Williams & Wilkins, Philadelphia, PA (2006) (2400pp., £203.00, ISBN: 0-7817-21608). Thirty-one years since the first edition was published, Rockwood and Green’s series remains the definitive textbook on fracture care. It is not difficult to see why—the book has evolved to keep pace with a rapidly changing field yet provides both the foundations and structure for the advice it dispenses. A historical review of management precedes discussion of the treatment options before the ‘authors preferred treatment’ is presented. Over 30 years, there has been a significant accumulation of knowledge, and this has been dealt with sensibly. Instead of increasing the size of an already substantial work, the historical elements have been condensed to make room for new chapters and substantial revision of existing ones. Indeed, of the 93 contributing authors almost two-thirds are new to the work, with a distinct and welcome international expansion. The editors have done a very good job in maintaining the style that made the earlier volumes so popular and, subject to the threat to printed works posed by the technological revolution, I see no reason why this incarnation should not be equally successful. As before, there are four main sections: general principles, upper extremity, spine and lower extremity. The contents of each particular volume are no longer apparent from glancing at the spine and there is a rather unusual division between first and section volumes, which occurs part way through the upper limb section. I presume this was done to make the two volumes of equal size. There is limited but effective use of a single colour (blue) and text boxes have been increasingly used, this helping to condense information from previous

editions as well as offering highly visible summaries of important information and treatment tips. It is well illustrated and reproduction, apart from occasional radiographs with poor contrast, is good. New are two compact discs containing videos of surgical approaches performed on cadavers. I doubt that these will survive long in institutional copies but they are useful for trainees and certainly an improvement on text descriptions of the same with line diagrams. The video quality is good, though distinction of anatomical structures can be difficult, a problem with all cadaveric, and indeed operative, videos. The CDs are visibly sponsored and a link on the screen takes the reader through to residentscorner.com, a commercially sponsored website where the same material is available. A second value-added service, of perhaps much greater utility, is a unique code that permits one user to register at rockwoodsolutions.com. Once activated this allows web access to the full contents of ‘fractures in adults’. The site also permitted this reviewer full access to ‘fractures in children’ and an image bank consisting of all illustrations from both of the above. It also promised access to the ‘Journal of Orthopaedic Trauma’, though the link only led to a single article from the most recent edition. With the availability of computers in theatre this gets round the problem of taking the weighty tome into the operating room. This remains one of the standard texts in Trauma and Orthopaedic Surgery and this latest edition represents a significant development on the last, not only in terms of updating information but also in adding valuable services to make it competitive in the modern world. It would be an enduring and useful investment for anyone celebrating appointment to a training programme in our specialty.

David Limb

doi:10.1016/j.cuor.2006.08.001

Complications of Spine Surgery: Treatment and Prevention. Howard S. An, Louis G. Jenis (Eds.), Lippincott Williams & Wilkins, Philadelphia, PA (2005) (195pp., $139.00, ISBN: 0781757916). This book is a useful addition to any spine surgeon’s library but will be of particular benefit to those clinicians embarking on a career in spine surgery or to those who are likely to encounter post-surgical spinal patients whether on an occasional or regular basis. At only 195 pages this book is not too intimidating. The 19 chapters are arranged within three sections, cervical spine, thoracolumbar/lumbosacral spine and miscellaneous disorders. Although many of the points in each chapter seem to be repetitive and at times stating the obvious this does emphasise the importance of pre-operative planning, appropriate patient positioning, careful surgical technique and post-operative doi:10.1016/j.cuor.2006.09.001

care. There are no bullet point lists of ‘‘do’s and don’t’s’’ but the text is clearly written and even the casual reader would have no difficulty constructing their own list. The text is appropriately illustrated. Chapters 11 and 12, complications related to thoracic and lumbar pedicle screw instrumentation could have been condensed into one chapter. Chapter 14 on lumbar pseudarthrosis has some helpful case illustrations and it is perhaps a weakness of the book that this approach was not adopted in more of the chapters. The operative management of pseudomeningoceles is mentioned but the technique is then left to the reader’s imagination. Overall, I would recommend this book to any spinal surgery registrar or resident as one of those texts to read very early on in their careers.

Crispin Wigfield

ARTICLE IN PRESS Current Orthopaedics (2006) 20, 472

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/cuor

BOOK REVIEWS Orthopaedic Knowledge Update—Trauma 3. Baumgaertner Michael R., Tornetta III Paul (Eds.), American Academy of Orthopaedic Surgeons (2005) (400pp., £94.95, ISBN: 0892033479). The first two editions of OKU trauma appeared in 1996 and 2000. After a slightly longer interval, the third edition takes up more shelf space than previous editions, but is more than simply an update on previous versions. The structure of the OKU series is familiar. After a general introductory section, the book is split up along anatomical lines and specific problems are discussed. For each topic, an expert in the field reviews what is new or controversial and comes to a position statement that is carefully offered as just that. A foreword emphasises that alternative methods of treatment may be valid. As such this is not a comprehensive textbook on trauma, though it was never intended to be so. The very nature of the book makes it ideal for trainees wishing to be fully up to date before sitting exams and those in practice wishing to update. Changes to the first section of this book further enhance this function, with particularly

useful sections on basic sciences. Indeed, most of the expansion that seems to have occurred since the last issue affects this part of the book. Instead of a general section followed by upper extremity, lower extremity and spine sections there is now a section on new measurements and new technology followed by one on the polytrauma patient and fracture healing. The next sections are on upper extremity, the axial skeleton (including spine and pelvis) and lower extremity. Children’s trauma now sits in a new section at the end with geriatric trauma, a pairing familiar to those up to date with ATLS. This is a book to be recommended, provided your learning style suits the OKU format. There is some crosstalk with the OKU 8 Home Study Syllabus but not as much as one might imagine. Even where the same topics are discussed a different author has provided a new perspective. For many this will be an excellent book for revision and update, but complements rather than replaces the traditional textbooks.

David Limb

doi:10.1016/j.cuor.2006.09.002

Vertebroplasty and Kyphoplasty. D.K. Resnick, S.R. Garfin (Eds.), American Association of Neurological Surgeons, Thieme, Stuttgart (2005) (138pp., $99.95, ISBN 3-13-1353414). Vertebroplasty and kyphoplasty are now well-established techniques for the treatment of pain from vertebral compression fractures (VCFs). The techniques are becoming important options in the treatment of VCFs for the spine surgeon and radiologist alike. This book contains 129 pages and 16 short chapters written solely by the American authors. The book guides the reader through the mechanisms of pain relief, patient selection, techniques, complications and compares vertebroplasty to kyphoplasty with the advantages and disadvantages of each technique. Chapters doi:10.1016/j.cuor.2006.07.009

begin with a brief statement of the objectives that are to be achieved by the reader. The book is reasonably well illustrated and well laid out making it quite readable. For the novice or trainee, Chapter 1 provides an excellent overview and summary of the technique and is highly recommended if one wants to grasp the concept in a short amount of time. The following chapters discuss in more detail the techniques, rationale and evidence and are essential reading for anyone who will be performing either technique. In summary, this is a comprehensive wellreferenced book that flows well about a relatively new technique, which will only expand in clinical practice.

Jake Timothy

Author Index A ADAMS M, 354 ADAMS S, 354 AMIN AK, 216 AMIRFEYZ R, 212 AMIS AA,16 ANKARATH S,141, 376 ANNAN IH, 241 ASHFORD RU, 299 ASPROS D, 212

H HALL RM, 23 HAMMOND C, 386 HANSEN U,16 HOEY E, 453 HOLTON C, 361 HOOPER G, 266, 461 I

B

INGHAM E, 32 IYENGAR KP, 47 J

BAKER ADL, 430 BHATIA D, 446 BOLLEN S, 77 BOURNE RB,157, 203 BRENKEL IJ, 216 BURNS AWR,162, 203 BURTON DJC, 222

JAYSEKERA N, 208 JIN ZM, 32 JOHNSON DS,121 JUPITER JB, 208 K C KARMANI S, 52, 208 KEENAN GF, 59

CALDER S, 320 CANNON LB, 322 CHALIDIS B, 405 CHEUNG A, 424 CLASPER J, 346

L

D

M

DE BEER JF, 446 DEVIC N,112 DICKSON RA, 321 DONELL S,103 DUNCAN CP,179 E EDWARDS MR, 367 F FARAJ AA, 41 FISHER J, 32 FOSTER P, 361 FREUDMAN M, 367 G GARGAN M, 212, 418 GARGAN MF,128 GIANNOUDIS PV, 376, 405 GREEN SM, 9 GRIFFITHS D, 346

LAM F, 208, 370, 446 LANKESTER BJA,128 LIMB D, 238, 320, 471, 472

MACDONALD SJ,157 MACNICOL MF, 256 MASOUROS S,16 MASRI BA,179 MCBRIDE ART, 418 MCCALDEN RW,162 MCDERMOTT ID, 85 MCEACHAN J, 294 MCGRAW P,121 MEHLING A, 397 MEHLING I, 397 METHA SS, 405 MOK D, 370 MOKETE L,192 MOON M-S,132 MORAN M, 241, 256 MURRAYAW, 286 MYLES L, 274 N NAUDIE DD,192 NUTTON RW, 95

II

CURRENT ORTHOPAEDICS P

PAPROSKY WG,171 PARKER PJ, 333 PATIL S,179 PENDLEBURY S, 299 PHILLIPS AM, 424

STEWART TD, 23 STONE M, 32 T

R RAMESH CN, 47 ROBB JE, 286 ROBERTSON A, 95 ROBINSON P, 386, 453 ROMMENS PM, 397 ROSS J, 274

TEMPLETON P, 361 TEMPLETON PA, 238 TIMOTHY J, 472 TRIMBLE K, 354 TSIRIDIS E, 405 TSIRIKOS AI, 430 TURNER PG,121 V VAN ROOYEN K, 446 VENKATESH R, 411 VINJAMURI S, 47

S W SALES JD, 216 SANCHEZ-BALLESTER J,121 SHORTT N, 59 SKUTEK M,157 SMITA, 461 SPORER SM,171 STALLEY PD, 299 STAPLEY SA, 322

WALSH G, 376 WATTERS AT, 222 WATTS AC, 266, 294 WHITEHOUSE M,128 WIGFIELD C, 471 WILCOX RK,1 WILLIAMS A,112

Subject Index 99mTc-Sulesomab, 47 E A Acetabular,162 Allograft, 85 Amputation, 333, 354 Anterior cruciate ligament, 77,112 Anti-personnel mine injuries, 354 Arthroplasty,192 Arthroscopic reduction and internal ¢xation (ARIF), 411 Arthroscopic shoulder surgery, 446 Arti¢cial hip joint, 32 B Ballistic principles, 322 Ballistic, 333, 346 Biceps, 370 Biomechanics,1,16, 23 Biotribology, 32 Blast, 333 Blast injury, 322 Blast physics, 322 Bone,16, 52, 299 Bone healing, 424 Bone morphogenetic protein, 424 Bone neoplasm, 274 British Army, 333 British OrthopaedicTrainees Association (BOTA), 367 C Calcifying tendinitis, 446 Cannulated screw ¢xation, 397 Carpal tunnel syndrome, 294 Carpus,141 Cartilage,16 Ceramic-on-ceramic, 32 Cerebral palsy, 286 Cervical vertebrae, 274 Child, 274, 361, 461 Children, 241 Chronic pain,141 Classi¢cation, 266 Con£ict, 333 Congenital hand anomalies, 266 Connective tissue disease, 418 Contact mechanics, 32 Contamination, 346 Cost e¡ectiveness, 204

Economics, 204 Epidemiology, 266 Extensively coated stem,171 Extracorporeal shock wave therapy, 446 F Failure, 9 Female patients, 294 Femoral, 376 Femoral fractures,179 Femoral neck, 361 Femoral revision,171 Fixation, 209 Fracture healing, 424 Fractures,179, 241, 256, 346, 361, 376, 461 Friction, 32 Function, 85 G Gait, 286 Gunshot, 333 Gunshot injuries, 322 H Hand, 461 High energy wounds, 322 High tibial osteotomy,112 Hip, 23,192, 286 Hip prosthesis,179 Hip revision,171 I Ilizarov, 59 Indications, 241 Infected non-union, 59 Infection,192 Infection imaging, 47 Internal ¢xation, 223, 241 Intraoperative complications,179 Ipsilateral femoral and tibial fractures, 405 K Klippel ^Feil, 274 Knee, 23, 256 Knee dislocation, 95 Knee injuries, 95, 405 Knee joint, 95

D Development,132 Dislocations, 9, 95, 286 Distal radius fractures, 209 Drilling, 52 Dynamics,1

L Ligaments,16 Limb injury, 333 Low energy wounds, 322 Lubrication, 32

IV

CURRENT ORTHOPAEDICS

M Malalignment syndromes,103 Male patients, 294 Malunion, 59 Management,121,132, 266 Marfan syndrome, 418 Mechanical properties of materials, 9 Mechanics,1 Medial compartment osteoarthritis,112 Meniscal, 85 Meniscectomy, 85 Menisci,16 Meniscus, 85 Metal-on-metal, 32 Metastases, 299 Microstructure, 9 Military, 333, 346 Mine, 333 Minimal invasive surgery, 411 Modular,171 Morquio syndrome,128 MPS IV,128 Mucolipidoses, 274 Mucopolysaccharidoses,128

N Needle aspiration irrigation, 446 Neuromuscular scoliosis, 430 Nonunion, 59, 424 Nuclear medicine, 47

R Reconstruction,162 Red cross, 333 Rehabilitation, 223 Repair, 85 Revision,192 Revision hip arthroplasty epidemiology,157 Revision total hip arthroplasty,162 S Schatzker classi¢cation, 411 Shoulder, 446 Shoulder fractures, 223 Shoulder surgery, 370 Skeletal dysplasia,128 Spinal deformity, 430 Spinal fusion, 274 Spinal injuries, 274 Spinal muscular atrophy, 430 Spine, 23,132 Statics,1 Stress strain behaviour of materials, 9 Surfaces, 32 Surgery, 41,121, 299, 333, 424 Surgical treatment, 430 Surgical treatment and outcomes, 77 Surveillance, 286 T

O Obesity, 217 Open, 346 Orthopaedics, 41, 424 Orthosis, 41 Osteo-articular infection, 47 Osteochondral injury, 256 Osteochondrodysplasia, 274 Outcome, 294

P Paediatric, 256 Paralysis, 41 Paralytic scoliosis, 430 Partial and total patellectomy, 397 Patellar fracture, 397 Patellar instability,103 Patello-femoral disorders,121 Patellofemoral joint,103 Pathology,121 Periprosthetic, 376 Plates, 209 Polytrauma, 405 Posterior cruciate ligament, 77 Postoperative complications,179 Post-polio syndrome, 41

Tantalum,162 Tear, 85 Tendons,16 Tenodesis, 370 Tenotomy, 370 Tension-band wiring, 397 THA, 204 Thermal e¡ects, 52 THR, 204 Total hip arthroplasty, 204 Total hip replacement, 204, 217 Total knee arthroplasty, 376 Total knee replacement, 217, 376 Trabecular metal,162 Trauma, 333 Treatment, 294 Tribology, 32 Trochleoplasty,103 Tuberculosis,132 U UHMWPE (ultra-high molecular weight polyethylene), 32 V Varus mal-alignment,112 W Wear, 32 White cell labelling, 47 Wrist,141