Clinical Ophthalmic Oncology : Uveal Tumors [3 ed.] 9783030178796, 303017879X

Citation preview

Clinical Ophthalmic Oncology Uveal Tumors Bertil E. Damato Arun D. Singh Editors Third Edition

123

Clinical Ophthalmic Oncology

Bertil E. Damato  •  Arun D. Singh Editors

Clinical Ophthalmic Oncology Uveal Tumors Third Edition

Editors Bertil E. Damato University of Oxford Oxford UK

Arun D. Singh Cole Eye Institute Cleveland Clinic Cleveland, OH USA

ISBN 978-3-030-17878-9    ISBN 978-3-030-17879-6 (eBook) https://doi.org/10.1007/978-3-030-17879-6 © Springer Nature Switzerland AG 2019 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Preface

Ophthalmic tumors are rare and diverse so that their diagnosis can be quite complex. Treatment usually requires special expertise and equipment and, in many instances, is controversial. The field is advancing rapidly, because of accelerating progress in tumor biology, pharmacology, and instrumentation. Increasingly, the care for patients with an ocular or adnexal tumor is provided by a multidisciplinary team, consisting of ocular oncologists, general oncologists, radiotherapists, pathologists, psychologists, and other specialists. For all these reasons, we felt that there was a need for the new edition of the textbook providing a balanced view of current clinical practice. Although each section of Clinical Ophthalmic Oncology now represents a stand-alone volume, each chapter has a similar layout with boxes that highlight the key features, tables that provide comparison, and flow diagrams that outline therapeutic approaches. The enormous task of editing a multiauthor, multivolume textbook could not have been possible without the support and guidance by the staff at Springer: Caitlin Prim, Melanie Zerah, ArulRonika Pathinathan, and Karthik Rajasekar. Michael D.  Sova kept the pressure on to meet the production deadlines. It is our sincere hope that our efforts will meet the high expectation of the readers. Oxford, UK Cleveland, OH, USA

Bertil E. Damato Arun D. Singh

v

Acknowledgments

To my family, Frankanne, Erika, Stephen, and Anna (BED) To my parents, who educated me beyond their means, and my wife, Annapurna, and my children, Nakul and Rahul, who make all my efforts worthwhile (ADS)

vii

Contents

1 Uveal Tumors: Examination Techniques ��������������������������������������   1 Bertil E. Damato and Iwona Rospond-Kubiak 2 Classification of Uveal Tumors ������������������������������������������������������  11 Bertil E. Damato and Sarah E. Coupland 3 Benign Melanocytic Tumors of the Uvea ��������������������������������������  17 Miguel Materin and Arun D. Singh 4 Uveal Melanoma: Epidemiologic Aspects��������������������������������������  53 Nakul Singh, Stefan Seregard, and Arun D. Singh 5 Uveal Melanoma: Clinical Features ����������������������������������������������  71 Leonidas Zografos and Ann Schalenbourg 6 Uveal Melanoma: Differential Diagnosis ��������������������������������������  85 Bertil E. Damato and Armin R. Afshar 7 Uveal Melanoma: Histopathologic Features���������������������������������� 109 Tero T. Kivelä 8 Uveal Melanoma: Molecular Pathology���������������������������������������� 121 Sarah E. Coupland, Helen Kalirai, Sophie Thornton, and Bertil E. Damato 9 Animal Models in Uveal Melanoma ���������������������������������������������� 135 Julia V. Burnier, Christina Mastromonaco, Jade Marie Lasiste, and Miguel N. Burnier Jr. 10 Iris Melanoma���������������������������������������������������������������������������������� 155 Arun D. Singh and Bertil E. Damato 11 Management of Patients with Posterior Uveal Melanoma���������� 185 Bertil E. Damato 12 Uveal Melanoma: Brachytherapy�������������������������������������������������� 201 Gustav Stålhammar, Stefan Seregard, and Bertil E. Damato 13 Uveal Melanoma: Proton Beam Radiation Therapy�������������������� 219 Anne Marie Lane, Ivana K. Kim, and Evangelos S. Gragoudas

ix

x

14 Uveal Melanoma: Stereotactic Radiation Therapy���������������������� 233 Karin Dieckmann, Roman Dunavoelgyi, Gerald Viktor Langmann, Roy Ma, Richard Pötter, Michael Sommer, Lisa Vajda, Werner Wackernagel, and Martin Zehetmayer 15 Uveal Melanoma: Phototherapy���������������������������������������������������� 241 Heinrich Heimann, Michael I. Seider, and Bertil E. Damato 16 Uveal Melanoma: Resection Techniques���������������������������������������� 249 Bertil E. Damato and Carl Groenewald 17 Uveal Melanoma: The Collaborative Ocular Melanoma Study ���������������������������������������������������������������� 261 Ezekiel Weis, Tero T. Kivelä, and Arun D. Singh 18 Uveal Melanoma: Prognostic Factors�������������������������������������������� 273 Robert Folberg and Jacob Pe’er 19 Uveal Melanoma: Prognostication Methods���������������������������������� 279 Sarah E. Coupland, Azzam Taktak, Antonio Eleuteri, Helen Kalirai, Sophie Thornton, and Bertil E. Damato 20 Uveal Melanoma: Mortality������������������������������������������������������������ 295 Bertil E. Damato and Azzam Taktak 21 Uveal Melanoma: Adjuvant Therapy�������������������������������������������� 305 Jessica Yang, Elaine Binkley, Arun D. Singh, and Richard D. Carvajal 22 Uveal Melanoma: Metastases���������������������������������������������������������� 317 Lucy T. Xu, Pauline Funchain, Ahmad A. Tarhini, and Arun D. Singh 23 Uveal Vascular Tumors�������������������������������������������������������������������� 331 Masood Naseripour, Jørgen Krohn, Pukhraj Rishi, and Arun D. Singh 24 Uveal Neural Tumors ���������������������������������������������������������������������� 353 Victoria M. L. Cohen, Bertil E. Damato, and Arun D. Singh 25 Intraocular Manifestations of Hematopoietic Disorders�������������� 365 Bercin Tarlan and Hayyam Kiratli 26 Uveal Myogenic, Fibro-histiocytic, and Histiocytic Tumors�������� 379 Paul A. Rundle and Hardeep Singh Mudhar 27 Uveal Lymphoproliferative Tumors ���������������������������������������������� 391 Sarah E. Coupland and Arun D. Singh 28 Uveal Metastatic Tumors���������������������������������������������������������������� 403 Norbert Bornfeld and Arun D. Singh

Contents

Contents

xi

29 Uveal Osseous Tumors �������������������������������������������������������������������� 423 Noel Horgan and Arun D. Singh 30 Bilateral Diffuse Uveal Melanocytic Proliferation������������������������ 441 Elaine Binkley, Ilya Leskov, and Arun D. Singh Index���������������������������������������������������������������������������������������������������������� 447

Contributors

Armin R. Afshar, MD, MBA, MAS  Ocular Oncology Service, Department of Ophthalmology, University of California, San Francisco, CA, USA Elaine  Binkley, MD Department of Ophthalmic Oncology, Cole Eye Institute, Cleveland Clinic, Cleveland, OH, USA Norbert Bornfeld, PhD  Klinik fuer Augenheilkunde, Universitaetsklinikum Essen, Essen, Germany Julia  V.  Burnier, PhD Department of Oncology and Pathology, McGill University, MUHC-RI, Montreal, QC, Canada Miguel N. Burnier Jr., MD, PhD, FRCSC  Department of Ophthalmology and Pathology, The MUHC-McGill University Ocular Pathology & Translational Research Laboratory, Montreal, QC, Canada Richard  D.  Carvajal, MD Division of Hematology/Oncology, Columbia University Medical Center, New York, NY, USA Victoria M. L. Cohen, MA, MB, BChir (Cantab), FRCOphth Department of Ocular Oncology, Moorfields Eye Hospital, London, UK Sarah E. Coupland, MBBS, PhD, FRCPath  Department of Molecular and Clinical Cancer Medicine, Institute of Translational Medicine, University of Liverpool, Liverpool, Merseyside, UK Bertil E. Damato, MD, PhD, FRCOphth  Nuffield Department of Clinical Neurosciences, University of Oxford, John Radcliffe Hospital, Oxford, UK Karin Dieckmann  Department of Radiation Oncology, Medical University of Vienna/General Hospital of Vienna, Comprehensive Cancer Center, Vienna, Austria Roman Dunavoelgyi, MD  Department of Ophthalmology and Optometry, Medical University of Vienna, Vienna, Austria Antonio  Eleuteri, PhD Department of Medical Physics and Clinical Engineering, Royal Liverpool University Hospital, Liverpool, UK Robert  Folberg, MD Foundational Medical Studies, Pathology, and Ophthalmology, Oakland University William Beaumont School of Medicine, Rochester, MI, USA

xiii

xiv

Pauline Funchain, MD  Taussig Cancer Center, Cleveland Clinic Foundation, Cleveland, OH, USA Evangelos  S.  Gragoudas, MD Ocular Melanoma Center, Department of Ophthalmology, Massachusetts Eye and Ear Infirmary/Harvard Medical School, Boston, MA, USA Carl Groenewald, MD  Ocular Oncology Service, St Paul’s Eye Unit, Royal Liverpool University Hospital, Liverpool, UK Heinrich  Heimann, MD Liverpool Ocular Oncology Service, Royal Liverpool University Hospital, Liverpool, Merseyside, UK Noel  Horgan, MD, FRCSI, MRCOphth Royal Victoria Eye and Ear Hospital, Dublin, Ireland St. Vincent’s University Hospital, Dublin, Department of Ocular Oncology and Medical Retina, Dublin, Ireland Helen  Kalirai, BSc, PhD Department of Molecular and Clinical Cancer Medicine, Institute of Translational Medicine, University of Liverpool, Liverpool, Merseyside, UK Ivana K. Kim, MD  Ocular Melanoma Center, Department of Ophthalmology, Massachusetts Eye and Ear Infirmary/Harvard Medical School, Boston, MA, USA Hayyam  Kiratli, MD  Ocular Oncology Service, Department of Ophthalmology, Hacettepe University School of Medicine, Ankara, Turkey Tero T. Kivelä, MD  Ocular Oncology and Ophthalmic Pathology, Helsinki University Hospital, Helsinki, Finland Department of Ophthalmology, University of Helsinki, Helsinki, Finland Jørgen  Krohn, MD Department of Clinical Medicine, Section of Ophthalmology, University of Bergen, Bergen, Norway Department of Ophthalmology, Haukeland University Hospital, Bergen, Norway Anne  Marie  Lane, MPH Ocular Melanoma Center, Department of Ophthalmology, Massachusetts Eye and Ear Infirmary/Harvard Medical School, Boston, MA, USA Gerald  Viktor  Langmann, MD Comprehensive Cancer Center Graz (CCC), Medical University Graz, Graz, Austria Jade  Marie  Lasiste, MD, MSc Department of Ophthalmology and Pathology, The MUHC-McGill University Ocular Pathology & Translational Research Laboratory, Montreal, QC, Canada Ilya Leskov, MD  Department of Ophthalmic Oncology, Cole Eye Institute, Cleveland Clinic Foundation, Cleveland, OH, USA Roy  Ma, MD, FRCPC Department of Radiation Oncology, British Columbia Cancer Agency, Vancouver, BC, Canada

Contributors

Contributors

xv

Christina  Mastromonaco, MSc, PhD Candidate Department of Ophthalmology and Pathology, The MUHC-McGill University Ocular Pathology & Translational Research Laboratory, Montreal, QC, Canada Miguel Materin, MD  Ophthalmic Oncology, Duke University Eye Center, Duke Cancer Institute, Durham, NC, USA Hardeep Singh Mudhar, PhD, MBBChir (Cantab), FRCPath Department of Histopathology, National Specialist Ophthalmic Pathology Service (NSOPS), Royal Hallamshire Hospital, Sheffield, South Yorkshire, UK Masood Naseripour, MD  Department of Ophthalmology/Ocular Oncology Service, Rassoul Akram Hospital/Iran University of Medical Sciences, Tehran, Iran Jacob  Pe’er, MD Department of Ophthalmology, Hadassah  – Hebrew University Medical Center, Jerusalem, Israel Richard Pötter  Department of Radiation Oncology, Medical University of Vienna/General Hospital of Vienna, Comprehensive Cancer Center, Vienna, Austria Pukhraj Rishi, MS, FRCS, FRCSEd  Shri Bhagwan Mahavir Vitreoretinal Services, Sankara Nethralaya, Chennai, Tamilnadu, India Iwona  Rospond-Kubiak, MD Ocular Oncology Service, Department of Ophthalmology, Poznań University of Medical Sciences, Poznań, Poland Paul A. Rundle, MBBS, FRCOphth  Ocular Oncology Clinic, Department of Ophthalmology, Royal Hallamshire Hospital, Sheffield, South Yorkshire, UK Ann Schalenbourg, MD  Department of Ophthalmology, Jules-Gonin Eye Hospital, University of Lausanne, FAA, Lausanne, Switzerland Michael  I.  Seider, MD Department of Ophthalmology, The Permanente Medical Group, San Francisco, CA, USA University of California – San Francisco, San Francisco, CA, USA Stefan Seregard, MD, PhD  Ophthalmic Pathology and Oncology Service and Department of Clinical Neuroscience, St. Erik Eye Hospital and Karolinska Institutet, Stockholm, Sweden Arun  D.  Singh, MD Department of Ophthalmic Oncology, Cole Eye Institute, Cleveland Clinic, Cleveland, OH, USA Nakul Singh, MS  School of Medicine, Case Western University, Cleveland, OH, USA Michael Sommer, MD  Department of Ophthalmology, Medical University of Graz, Graz, Austria Gustav  Stålhammar, MD, PhD Ophthalmic Pathology and Oncology Service and Department of Clinical Neuroscience, St. Erik Eye Hospital and Karolinska Institutet, Stockholm, Sweden

xvi

Azzam Taktak, B.Eng (Hons)  Department of Medical Physics and Clinical Engineering, Royal Liverpool University Hospital, Liverpool, UK Ahmad  A.  Tarhini, MD Taussig Cancer Center, Cleveland Clinic Foundation, Cleveland, OH, USA Bercin  Tarlan, MD  Ocular Oncology Service, Department of Ophthalmology, Hacettepe University School of Medicine, Ankara, Turkey Sophie Thornton, BSc, PhD  Department of Molecular and Clinical Cancer Medicine, Institute of Translational Medicine, University of Liverpool, Liverpool, Merseyside, UK Lisa Vajda, MD, PhD  Department of Ophthalmology, Medical University of Graz, Graz, Austria Werner  Wackernagel, MD Department of Ophthalmology, Medical University of Graz, Graz, Austria Ezekiel  Weis, MD, MPH University of Alberta, Department of Ophthalmology, Edmonton, AB, Canada University of Calgary, Department of Surgery, Calgary, AB, Canada Lucy T. Xu, MD  Department of Ophthalmic Oncology, Cole Eye Institute, Cleveland Clinic Foundation, Cleveland, OH, USA Jessica Yang, MD  Division of Hematology/Oncology, Columbia University Medical Center, New York, NY, USA Martin  Zehetmayer, MD, PhD Department of Ophthamology, Medical University, General Hospital of Vienna, Vienna, Austria Leonidas  Zografos, MD Ophthalmology, University of Lausanne, Lausanne, Switzerland

Contributors

1

Uveal Tumors: Examination Techniques Bertil E. Damato and Iwona Rospond-Kubiak

Introduction This chapter highlights procedures that are specific to the assessment of a patient with a uveal tumor. It is assumed that a full ophthalmic and systemic history is routinely obtained in all patients in addition to complete examination of both eyes and appropriate systemic assessment, consisting of clinically relevant ancillary investigations.

distinguish between a metastasis and other types of tumor, such as melanoma and hemangioma, because dual pathology is not uncommon. The history provides an understanding of the patient’s visual needs, which may help in the selection of the most appropriate form of treatment. The duration of the visual loss can have prognostic significance, for example, in patients with choroidal hemangioma in whom visual loss is irreversible if long-standing.

History Taking

Follow-Up

Initial Assessment

Routine use of a questionnaire ensures that at every follow-up visit, each patient is asked all ­relevant questions about general health, visual symptoms, ocular discomfort, and concerns about possible ocular complications and survival.

The past medical history can sometimes provide diagnostic clues, for example, if the patient has been a heavy smoker for many years or if a previous mastectomy has been performed. While such information might suggest the source of an intraocular metastasis, it should not be relied upon to

B. E. Damato (*) Nuffield Department of Clinical Neurosciences, University of Oxford, John Radcliffe Hospital, Oxford, UK e-mail: [email protected] I. Rospond-Kubiak Ocular Oncology Service, Department of Ophthalmology, Poznań University of Medical Sciences, Poznań, Poland

Visual Acuity If possible, the visual acuity should be measured using a LogMAR chart, which overcomes the limitations of the Snellen test, also facilitating statistical analysis of vision in any outcomes analysis [1]. If central vision is lost, the eccentric visual acuity should be measured using the optotype and, if necessary, finger counting before checking for hand movement vision.

© Springer Nature Switzerland AG 2019 B. E. Damato, A. D. Singh (eds.), Clinical Ophthalmic Oncology, https://doi.org/10.1007/978-3-030-17879-6_1

1

2

Slit-Lamp Examination It is necessary to define the primary tumor, recognizing any secondary effects, predisposing factors, and concurrent disease (Box 1.1): (i) Site of origin (iris, ciliary body, choroid). (ii) Location (superior, supero-nasal, nasal, etc.). (iii) Circumferential spread, ideally in clock minutes in a clockwise direction (e.g., 5–30 or 55–5). This is easier than using degrees and more precise than clock hours. For iris tumors, there may be a scope for recording circumferential spread at pupil margin as well as the midperipheral and peripheral iris. (iv) Posterior margin (pars plicata, iris surface). (v) Anterior margin (iris surface, angle, cornea). (vi) Consistency (solid, cystic, multicystic). (vii) Shape (flat, dome, multinodular). (viii) Margins (diffuse, discrete). (ix) Color (pink, white, yellow, red, orange, tan, brown, black, etc.). (x) Vascularity (present or absent). (xi) Seeding (across iris, into angle, vitreous). (xii) Angle involvement (in clock minutes). The first author (BD) finds it useful to describe the angle in each clock hour as normal, scanty pigment dusting, dense pigment dusting, flat confluent pigment, bulky confluent pigment, tumor, and uncertain (because of closed angle or hyphema) . (xiii) Extraocular spread (absent, nodular, diffuse). (xiv) Longitudinal and transverse basal dimensions, using the measure on the slit lamp. (xv) Secondary effects (dilated episcleral vessels, band keratopathy, glaucoma, hyphema, ectropion uveae and pupillary peaking, iris cyst formation, and lens abnormality such as cataract, deformity, and subluxation). (xvi) Predisposing factors (ocular or oculodermal melanocytosis, Sturge-Weber syndrome, and other vascular malformations).

B. E. Damato and I. Rospond-Kubiak

Box 1.1. Examination Techniques for Tumors • History taking, slit-lamp examination, and ophthalmoscopy. • Drawings complement photography, especially with peripheral tumors. • Tumor dimensions can be estimated using charts and ophthalmoscopically. • Three-mirror examination is useful in selected cases. • Transillumination gives an approximate indication of tumor extent.

The author (BD) has designed a diagram to facilitate documentation of slit-lamp findings (Fig. 1.1). Although intended primarily for conjunctival tumors, it is useful also with iris and ciliary body tumors.

Fig. 1.1  Template for documenting slit-lamp findings

1  Uveal Tumors: Examination Techniques

Indirect Ophthalmoscopy It is essential to examine the entire fundus, with indentation if necessary, to identify any other pathology. Both eyes should be examined, ideally with mydriasis. The first author (BD) has devised the mnemonic, MELANOMA, to alert the clinician to the presence of an intraocular tumor in situations where the pupils are not routinely dilated (Box 1.2). Box 1.2. Symptoms and Signs Indicating Presence of an Intraocular Tumor • Melanoma or other tumor visible externally in iris or episclera • Eccentric visual phenomena, such as photopsia, floaters, and field loss • Lens abnormalities, such as cataract, astigmatism, and coloboma • Afferent pupillary defect, mostly caused by secondary retinal detachment • No optical correction with spectacles because of blurring or metamorphopsia • Ocular hypertension, especially if asymmetrical • Melanocytosis, predisposing to melanoma • Asymmetrical episcleral vessels, indicating a ciliary body tumor

3

(ix) Circumferential involvement of disc, ciliary body, and perhaps choroid (in clock minutes) (x) Internal spread (subretinal space, retina, vitreous) (xi) Secondary effects (RPE changes such as drusen and orange pigment over the tumor, RPE changes adjacent to the tumor, exudative retinal detachment, and hemorrhage) (xii) Predisposing factors (ocular melanocytosis, melanocytoma, diffuse choroidal hemangioma)

Fundus Drawing Fundus drawings complement photography in several ways, for example, allowing important features to be highlighted by means of notes and markers. The technique has been described in detail elsewhere (Fig. 1.2) [2].

A. Ask the patient to lie supine on a couch or in a reclining chair. B. Stand at the head end of the patient and place the retinal chart on a tray next to the patient’s head. The top of the chart should be facing toward the patient’s feet. You should be able to move around so as to position yourself directly opposite the retinal quadrant being examined. C. Hold the indirect lens in the nondominant hand and a pencil in your dominant hand. D. Draw symbols for the optic disc and fovea, and then look at the fundus and rotate the drawing It is necessary to describe the primary tumor, pad so that the optic disc and fovea are aligned any secondary effects, and any predisposing facin the same way as the fundus image. tors as follows: E. Identify the meridians, in clock hours, of the (i) Tissue of origin (choroid, retina, retinal two lateral margins of the tumor, in relation to pigment epithelium) disc or fovea, and draw these lines on the chart. (ii) Quadrant (superotemporal, superior, F. Estimate the distance between posterior superonasal, etc.) tumor margin and disc or fovea and mark that (iii) Shape (flat, dome, collar stud) point on the chart. (iv) Margins (discrete, diffuse) G. Estimate the location of the anterior tumor (v) Color (pink, white, yellow, red, orange, margin in relation to equator or ora serrata tan, brown, black, etc.) and mark that point on the chart. (vi) Vascularity (vascular, avascular) H. Draw the profile of the tumor, using the marks (vii) Posterior extent, including distances to optic already on the chart as guides. disc margin and fovea (disc diameters or mm) I. Starting at the tumor and working backward (viii) Anterior extent (post-equatorial, pre-­ toward the optic disc, draw the major retinal equatorial, pars plana, pars plicata, etc.) blood vessels, placing conspicuous bifurca-

4

B. E. Damato and I. Rospond-Kubiak

b

a

c

13’

25’

d

13’

13’

25’

25’ 13’

e

25’ Fovea

f

Orange pigment

Disc

13’ Diffuse margins

Veins

Retinal detachment

Pale area 25’ Arteries

Fig. 1.2  Drawing a choroidal tumor. Color photograph (a). Locating tumor margins with respect to disc and fovea (b), delineating tumor margins (c), drawing vascular

details (d), adding tumor features (e), and annotations (f). (Reprinted from Damato [2]. With permission from Elsevier)

1  Uveal Tumors: Examination Techniques

tions and crossings in their correct positions in relation to tumor margins. J. Fill in details, such as texture, tumor vessels, RPE changes, hemorrhages, exudates, and retinal detachment. Fig. 1.3  Chart showing outer ocular surface, designed to help plan insertion of radioactive plaque or tantalum markers for the treatment of posterior segment tumors (a). The fovea is represented by a large ring, its distance from the limbus varying with axial length. Chart designed by the first author with the fundus demarcated circumferentially in clock minutes and radially in millimeters (b). The use of this chart as a template on an electronic tablet allows digital drawings to be prepared accurately and ergonomically, also facilitating transmission by the Internet and uploading onto electronic records

a

b

5

K. Ensure that the patient’s name and hospital number, the date of the examination, and your signature have all been documented. (Fig. 1.3a, b) shows an example of how fundus lesions are documented.

6

Estimation of Intraocular Tumor Basal Dimensions Schematic diagrams have been prepared to facilitate estimation of ocular dimensions on clinical examination [2]. We have developed a system for drawing fundus diagrams on an iPad using a template we developed to enhance accuracy (Fig. 1.3b). The preparation of digital drawings with inexpensive programs is ergonomic, especially in a dark room, and allows easy modification and uploading onto electronic records. Basal dimensions of intraocular tumors can also be measured by indirect ophthalmoscopy. The chord length tumor basal diameters (anteroposterior or longitudinal and circumferential or latitudinal) are estimated while performing indirect ophthalmoscopy by assessing the proportion of a specific condensing lens field that is filled by the tumor’s image. During this assessment, a 20D lens is considered to have a field diameter of approximately 12  mm, whereas a 28D lens is regarded to have a field diameter of 13 mm. For example, a tumor which fills one-half of the 20D lens field would be judged to have a diameter of approximately 6 mm, while a tumor filling twothirds of a 28D lens field would be considered to be about 8.5 mm in diameter. Tumor dimensions can also be measured by ultrasonography, as described elsewhere.

Three-Mirror Examination

B. E. Damato and I. Rospond-Kubiak

Transillumination Transillumination can be used to detect or locate tumor margins. In general, pigmented tumors and intraocular hemorrhage block transmission of light. It must be realized that not all pigmented tumors are melanoma and, conversely, not all melanomas are pigmented. Various techniques of transillumination include (Fig. 1.4): A. Trans-pupillary, placing the illuminator on the cornea. Care must be taken not to overestimate posterior extension because of an oblique shadow cast by a thick tumor. B. Trans-ocular, with a right-angled transilluminator on the globe directly opposite to the tumor. This is less convenient than trans-­ pupillary transillumination but more accurate. This is the first author’s preferred technique when identifying the lateral and posterior tumor margins intraoperatively, using a 20-gauge vitrectomy illuminator, which is bent 90°. C. Trans-scleral, with the light source on the sclera over the tumor. This only determines whether or not the tumor transmits light. Transillumination is also useful for identifying scleral necrosis and iris atrophy.

Color Photography

Color photography is useful for documenting the appearances of the tumor and other parts of the eye. The tumor color can vary greatly between A. Indentify the cause of raised intraocular cameras. For example, with the Optos camera, pressure. amelanotic metastases and hemangiomas may B. Determine whether a lesion behind the iris is falsely appear pigmented (Figs. 1.5 and 1.6) [3, solid or cystic. 4]. Tumor diagnosis should therefore not be C. Find a small, retinal angioma. based on color photographs. D. Determine the anterior extent of a pre-­ Photography also defines the tumor extent in equatorial tumor. relation to adjacent landmarks, such as retinal E. Measure the circumferential extent of ciliary blood vessels and pupil margin. Sequential phobody or angle involvement by a tumor, align- tography is usually helpful in detecting tumor ing in turn each lateral tumor margin with the growth over time (Fig. 1.7). False impressions of center of the mirror. growth can arise, however, as a result of inconsisThree-mirror examination is necessary to:

7

1  Uveal Tumors: Examination Techniques Fig. 1.4  Techniques of transillumination. Tumor extension assessed by trans-pupillary transillumination (a), trans-ocular transillumination (b), and trans-scleral transillumination (c). Note the exaggeration of posterior tumor extension with trans-pupillary transillumination [Reprinted from Damato [2]. With permission from Elsevier]

a

a

b

c

b

Fig. 1.5  Color photos of the metastasis from a lung adenocarcinoma in a 47-year-old woman. The tumor appears pigmented with the Optos camera (a) but not with the Topcon (b)

8

B. E. Damato and I. Rospond-Kubiak

a

b

c

Fig. 1.6  Color photos of a choroidal hemangioma photographed with the Panoret (a), Optos (b), and Topcon (c)

a

Fig. 1.7  Sequential color photographs of the right fundus of a 60-year-old man showing growth of a melanocytic tumor over a period of 13  years. Tumor appearances in

b

2000 (a) and in 2013 (b), with rupture of Bruch’s membrane and development of a collar-stud shape, with minimal lateral extension

1  Uveal Tumors: Examination Techniques

9

tent magnification or light exposure. Pseudo-­ and to ensure that the results are interpreted growth can also occur because of retinal properly. flattening, scleral flattening, and extension of secondary changes in the overlying retinal pigReferences ment epithelium [5].

Ancillary Tests Ancillary investigations such as color photography, autofluorescence imaging, optical coherence tomography, angiography, and ultrasonography are discussed in detail elsewhere in this series [6].

Conclusion The different examination techniques described in this chapter need to be used selectively. Each requires special expertise in performing the test

1. Bailey IL, Lovie JE. New design principles for visual acuity letter charts. Am J Optom Physiol Optic. 1976;53(11):740–5. 2. Damato B. Ocular tumours: diagnosis and treatment. Oxford: Butterworth-Heinemann; 2000. 3. Kernt M, Schaller UC, Stumpf C, et al. Choroidal pigmented lesions imaged by ultra-wide-field scanning laser ophthalmoscopy with two laser wavelengths (Optomap). Clin Ophthalmol. 2010;4:829–36. 4. Heimann H, Jmor F, Damato B.  Imaging of retinal and choroidal vascular tumours. Eye (Lond). 2013;27(2):208–16. 5. Schalenbourg A, Zografos L.  Pitfalls in colour photography of choroidal tumours. Eye (Lond). 2013;27(2):224–9. 6. Singh AD, Damato B, editors. Clinical ophthalmic oncology. Basic principles and diagnostic techniques. 3rd ed. Berlin/Heidelberg: Springer; 2019.

2

Classification of Uveal Tumors Bertil E. Damato and Sarah E. Coupland

Introduction Tumors of the same class should share a unique combination of features, which distinguishes them from all other classes [1, 2]. Classification, therefore, is the process of defining different tumor entities and correlating these with each other in a hierarchical manner based on the putative cell of origin. Classification also involves “grading” of tumor cell morphological features and consideration of their immunohistochemical and genetic profiles. It does not include tumor “staging,” which determines the extent of tumor involvement and hence the type of treatment required (i.e., local versus systemic treatment). Tumor classification has several uses. It can help the clinician to consider all relevant conditions when preparing a differential diagnosis. It can improve prognostication, by predicting how the tumor is likely to behave. This, in turn, can enhance treatment planning and enables proper evaluation of the results of treatment. Tumor classification also improves communication by B. E. Damato (*) Nuffield Department of Clinical Neurosciences, University of Oxford, John Radcliffe Hospital, Oxford, UK e-mail: [email protected] S. E. Coupland Department of Molecular and Clinical Cancer Medicine, Institute of Translational Medicine, University of Liverpool, Liverpool, Merseyside, UK e-mail: [email protected]

allowing standardization of disease categorization in multicenter studies. Classification is also valuable in research, contributing to investigations in tumor biology (Box 2.1).

Box 2.1. Classification of Uveal Tumors • Classification of tumors contributes to diagnosis, prognostication, treatment planning, evaluation of treatment results, communication between treatment centers, and oncological research. • Tumors can be classified according to location, etiology, histopathology, histogenesis, and other methods. • Each classification has its advantages and disadvantages in any particular situation so that different classification methods complement each other. • Tumor classification must be distinguished from staging.

Classification of Uveal Tumors Uveal tumors can be classified according to location, etiology, histopathology, histogenesis, genotype, and various other ontological methods. Each approach has its advantages and limitations.

© Springer Nature Switzerland AG 2019 B. E. Damato, A. D. Singh (eds.), Clinical Ophthalmic Oncology, https://doi.org/10.1007/978-3-030-17879-6_2

11

12

A classification based on tumor location within the uvea would need to mention some tumors more than once if these can arise at different sites, and, furthermore, it can be impossible to locate the origin of an extensive tumor. A classification that is superior in one situation may not be useful in other circumstances. For example, a histopathologic classification is helpful in a pathology laboratory but of limited value in an ophthalmic clinic when the patient is first seen, that is, before the tissue has been examined histologically. This chapter considers some conditions that are not strictly uveal, because these might be mistaken for a uveal tumor. For example, adenomas, adenocarcinomas, congenital hypertrophy of the retinal pigment epithelium, and iris cysts are all epithelial but need to be included in the differential diagnosis of several uveal tumors. For the same reasons, conditions, such as varix of the vortex vein ampulla, are mentioned, even if they are not neoplasms at all. Strictly speaking, the lists provided are not classifications, because they include tumors that are not biologically, clinically, and histologically related. In any case, it is hoped that this review will make it easier for clinicians to recall all relevant conditions in the differential diagnosis when the need arises, a mental feat that is facilitated by categorizing the different tumors and “pseudotumors” into meaningful groups.

Etiologic Classification This system categorizes uveal tumors according to their underlying causes, that is, as congenital, traumatic, inflammatory, neoplastic, degenerative, and idiopathic (Table  2.1). This is by no means an exhaustive list. First, the anatomical listing is only approximate, because tumors can arise in atypical locations. Second, the etiology of some tumors is not known (e.g., choroidal osteomas) (Fig. 2.1). It is important to appreciate that terminology may also influence clinical management inappropriately. For example, the term “suspicious nevus” may encourage passive clinical management, whereas if the same lesion is called

B. E. Damato and S. E. Coupland

“suspicious melanoma,” it is perhaps more likely to be treated. It may therefore be preferable to refer to an equivocal melanocytic tumor as an “indeterminate melanocytic tumor” or “melanocytic tumor of unknown malignancy.” Whether such a tumor is nevus, melanoma, or indeterminate is, of course, subjective if diagnosis is based on ophthalmoscopy alone. Finally, some tumors such as neurilemmoma, neurofibroma, and leiomyoma are classified separately despite being clinically indistinguishable.

Histopathologic Classification Histopathologic classification categorizes uveal neoplasms according to their cellular morphology, immunohistochemical, and genetic profiles. The World Health Organization (WHO) Classification of Tumours of the Eye has classified uveal tumors according to (a) anatomical site, (b) histologic type, and (c) genetic profiles [3]. This classification was developed by pathologists and ophthalmologists, specifically for diagnosing tumors based on histopathologic material. It provides differential diagnoses for each entity, including those that are nonneoplastic, such as inflammation or degenerative conditions. A summary of the WHO uveal tumor classification according to cell of origin is provided in Table 2.2.

Histogenetic Classification Histogenetic classification groups tumors hierarchically according to their embryonic lineage, with tumors being subclassified according to whether they originate from ectodermal, endodermal, or mesodermal cells [2]. Tumors are further subclassified according to whether they arise from primitive cells (e.g., totipotential cells forming teratomas) or differentiated cells (e.g., melanocytes). Proponents of this biological system of classification argue that it is simple, comprehensive, and capable of developing as molecular biology improves knowledge about different tumor types. Another advantage of this classification is that tumors with the same lineage

13

2  Classification of Uveal Tumors Table 2.1  Uveal tumors classified according to pathogenesis and location Category Congenital

Subtype Hamartoma Lisch nodules Choristoma Lacrimal gland choristoma Osseous choristoma Stromal iris cyst Inflammatory Infectious Granuloma (e.g., Tuberculosis) Noninfectious Sarcoidosis Juvenile xanthogranuloma Scleritis Uveal effusion Neoplastic/ Melanocytes Melanocytic nevus Hyperplastic Melanocytosis Melanocytoma Bilat. diffuse uveal melanocytic hyperplasia Blood vessels Circumscribed hemangioma Diffuse hemangioma Hemangiopericytoma Racemose angioma Fibroblasts Neurofibroma Neural tissue Neurilemmoma Muscle Leiomyoma Mesectodermal leiomyoma Lymphocytes Reactive lymphoid hyperplasia or chronic inflammation Melanoma Muscle Rhabdomyosarcoma Secondary Melanoma/carcinoma Hemopoietic Lymphoma Leukemia Metastatic Carcinoma/sarcoma Traumatic Foreign body Suprachoroidal hematoma Degenerative Disciform lesion (from choroidal neovascularization) Sclerochoroidal calcification Idiopathic Varicose vortex vein

a

Location Iris Ciliary body Choroid Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y

b

Fig. 2.1  Choroidal osteomas in the right eye (a) and left eye (b) of a 15-year-old woman. The etiology is not known as these tumors can be bilateral and have been reported in siblings

B. E. Damato and S. E. Coupland

14 Table 2.2  Pathologic classification of uveal tumors Subtype Benign Categories Melanocytes

Blood vessels Nerves Smooth muscle Striated muscle Fibroblasts Histiocytes Lymphocytes Ectopic tissue

Melanocytosis Melanocytoma Melanocytic nevus Diffuse melanocytic hyperplasia Hemangioma Hemangiopericytoma Schwannoma Leiomyoma Mesectodermal leiomyoma

Malignant Primary Melanoma

Secondary Conjunctival melanoma

Rhabdomyosarcoma Neurofibroma Juvenile xanthogranuloma Lymphocytic proliferation Lacrimal gland choristoma Osteoma

tend to share behavioral similarities. Histogenetic classification is not widely used for uveal tumors as the majority of uveal tumors are neuroectodermal or mesodermal in origin.

Genotypic “Classification” Tumors of the same class that seem identical on histology can behave very differently from each other. With developments in molecular biology, it has become possible to correlate tumor behavior with genetic aberrations and other novel characteristics. These advances have raised hopes of improving tumor classification. Molecular studies and microarray technology reveal marked differences between uveal melanomas, which have differing metastatic potential [4, 7, 8]. For example, metastatic disease from uveal melanoma occurs almost exclusively in patients whose tumor shows chromosome 3 loss, BAP1 mutations, and/or a class 2 gene expression profile [5–7]. Purists would argue that such categorizations or subgroupings actually constitute “discriminant analysis” rather than a “classification” [1]. This is because it appears that uveal melanomas may in rare cases undergo high-grade transformation from disomy 3 to monosomy 3 type, possibly as a result of a clonal

Lymphoma

Lymphoma Metastatic carcinoma Metastatic sarcoma

evolution and dominance [9]. A fundamental principle of tumor classification is that one class of tumor cannot transform into another class [1]. Categorization of uveal melanomas into different “classes” according to their gene expression ­profile [10] is, strictly speaking, incorrect, and perhaps should be considered as molecularly distinct subsets, as proposed by The Cancer Genome Atlas (TCGA) [7]. Different cancers can have the same genetic aberrations, so that they respond similarly to the same therapeutic agents. There is scope for classifying or subclassifying cancers according to their genetic makeup, as was put forward by the TCGA [7].

TNM Staging The TNM (Tumor, Node, Metastasis) staging system of the American Joint Committee on Cancer (AJCC) categorizes tumors anatomically according to the extent of the primary tumor and the presence or absence of lymph node involvement and metastasis [11, 12]. Clinical T staging of ciliary body and choroidal melanomas first groups tumors into size categories according to tumor thickness and basal diameter. The size categories are then integrated with ciliary body involvement and

15

2  Classification of Uveal Tumors

extraocular spread to group tumors into seven prognostic stages (I, IIA–B, IIIA–C, and IV) according to 10-year survival rates [13]. The AJCC TNM staging system has recently been revised so as to incorporate nonanatomical factors, such as mitotic count, to enhance prognostication. Staging is therefore “clinical” or “pathologic” or both. The TNM staging manual recommends collecting data on key prognostic factors, such as chromosomal alterations, even though these are not included in staging algorithms. The AJCC manual describes “Stage Classifications,” such as “Clinical,” “Pathologic,” “Post-therapy,” “Re-treatment,” and “Autopsy.” The manual therefore refers to “Clinical Classification,” “Pathologic Classification,” and so on. These different classifications enable patients to be grouped prognostically at different stages in their care pathway.

Conclusions Tumor classification is constantly evolving as advancing knowledge changes our perception about what makes a class of tumors unique and how that class relates to other groups in the hierarchy. Different classifications can complement each other, as seen with histologic and clinical groupings. Nevertheless, measures should be taken to ensure that alternative classifications match each other as closely as possible so that ambiguity and misunderstanding are avoided. There is no escaping the fact that terminology and semantics are important. It is essential to distinguish between factors such as class, grade, and stage of tumor, if the benefits of tumor groupings are to be maximized.

References 1. Berman JJ.  Tumor classification: molecular analysis meets Aristotle. BMC Cancer. 2004;4:10. 2. Berman J. Modern classification of neoplasms: reconciling differences between morphologic and molecular approaches. BMC Cancer. 2005;5:100. 3. Grossniklaus HE, Eberhart CG, Kivelä TT. WHO classification of tumours of the eye. 2018;12:4th ed. Lyon: WHO. 4. Coupland SE, Lake SL, Zeschnigk M, et  al. Molecular pathology of uveal melanoma. Eye (Lond). 2013;27(2):230–42. 5. Damato B, Dopierala JA, Coupland SE.  Genotypic profiling of 452 choroidal melanomas with multiplex ligation-dependent probe amplification. Clin Cancer Res. 2010;16(24):6083–92. 6. Onken MD, Worley LA, Char DH, et al. Collaborative Ocular Oncology Group report number 1: prospective validation of a multi-gene prognostic assay in uveal melanoma. Ophthalmology. 2012;119:1596–603. 7. Cancer Genome Atlas Research Network. The Cancer Genome Atlas Pan-Cancer analysis project. Nat Genet. 2013;45(10):1113–20. 8. Robertson AG, Shih J, Yau C, et al. Integrative analysis identifies four molecular and clinical subsets in uveal melanoma. Cancer Cell. 2017;32(2):204–20. 9. Callejo SA, Dopierala J, Coupland SE, et al. Sudden growth of a choroidal melanoma and multiplex ligation-­dependent probe amplification findings suggesting late transformation to monosomy 3 type. Arch Ophthalmol. 2011;129(7):958–60. 10. Onken MD, Worley LA, Ehlers JP, et al. Gene expression profiling in uveal melanoma reveals two molecular classes and predicts metastatic death. Cancer Res. 2004;64(20):7205–9. 11. Gospodarowicz MK, Miller D, Groome PA, et al. The process for continuous improvement of the TNM classification. Cancer. 2004;100(1):1–5. 12. American Joint Committee on Cancer. AJCC cancer staging manual. 7th ed. New York: Springer; 2010. 13. Kujala E, Damato B, Coupland SE, et al. Staging of ciliary body and choroidal melanomas based on anatomic extent. J Clin Oncol. 2013;31(22):2825–31.

3

Benign Melanocytic Tumors of the Uvea Miguel Materin and Arun D. Singh

Iris Nevus

Etiology and Pathology

Introduction

Most circumscribed iris nevi are composed of intra-stromal, bland, spindle cells [7]. In some cases, the lesion breaks through to the iris surface, to form a plaque or nodule. The diffuse iris nevus represents hyperplasia and hypertrophy of stromal melanocytes without mass effect. Other variants include epithelioid cell nevus, which is composed of bland, hypopigmented, epithelioid cells with a clear vacuolated cytoplasm, [8, 9] and balloon cell nevus, composed of cells with a clear cytoplasm [10].

Iris nevus is a stromal lesion and therefore quite different from an iris freckle, in which melanocytes collect only superficially, without stromal involvement. Iris freckles can been in up to 60% of population, whereas nevi are less common (4–6%) [1]. Both iris freckles and nevi are more frequent in light-colored irises [1]. Patients with dysplastic nevus syndrome may have a tendency to develop iris nevi [2, 3]. Although an association between iris nevi and uveal melanomas has been reported, [4] it is doubtful whether these two conditions are genuinely associated [1]. Recent studies have shown a potential role of a number of pigmented iris lesions as biomarkers for sun exposure and risk for cutaneous malignancies with higher risk in younger patients (30

White White

30

14+∗ 11∗ 1.9# (3.8)^ 3.1# (6.2)^ 30 7.9$

Not stated White

18–41 11–84

4.2 4.6

White Japanese Chinese Indian Non-­ Indigenous 2018

>49 28–86 40–101 30–100 >50

6.5∗(8.6)^ 0.34 2.9#= 0.15# = X 2.1

40–92

0.68

1965 USA 1970 Germany 1972 USA

Ganley

1973 USA

65

Gass Albert

1977 USA 1980 USA

250 1126

Lang Rodriguez-­ Sains Sumich Yoshikawa Jonas Nangia Keel

1982 Germany 3119 1986 USA 108

Clinic Older Population Chemical workers + controls Clinic Army Clinic Controls

1998 2004 2008 2012 2018

Population Clinic Population Population Population

3583 3676 4439 4711 3098

2018 Australia 3098 ∗ = Indirect ophthalmoscope not used +  = Including ciliary body nevi # = Only posterior to the equator ^ = Corrected for entire fundus $ = Calculated for white participants X = Calculated per eye

Prevalence % 1.1∗ 0.2∗

Race Not stated Race?

Hale Naumann Smith

Australia Japan Chinese India Australia

252 187 842

Age (years) Not stated 18–38

Population Consecutive cases Surgical trauma cases Autopsy Consecutive cases Autopsy Unselected cases Population Survey

Survey Normal volunteers Survey Survey Survey

Population Survey

95% White >18 Not stated All White (64%) >13

22

Terminology In contrast to a choroidal nevus, a choroidal freckle is composed of an increased density of normal melanocytes that do not disturb the normal architecture so that it is always flat, often with visible, normal choroidal vessels passing undisturbed through the lesion (Fig. 3.6).

Etiology and Pathology Choroidal nevi involve the full thickness of the choroid with sparing of the choriocapillaris [34]. They are composed of bland-looking nevus cells that are plumper than the normal uveal melanocytes [34]. The nevus cells vary in pigmentation and their shape may be polyhedral, fusiform, dendritic, spindle, some being of the clear, balloon type [34]. In the modified Callender’s classification of uveal melanoma, tumors composed entirely of spindle A cell are regarded as nevi [35]. Like uveal melanoma, choroidal nevus is associated with race, with an age-adjusted prevalence of 0.6% among Blacks, 2.7% among Hispanics, and 2.1% among others [36]. Also similar to uveal melanoma, choroidal nevus prevalence increases with age [36]. Interclass correlation coefficient for twins with choroidal nevi was

M. Materin and A. D. Singh

0.38 for monozygotic twins versus 0.02 in dizygotic twins suggesting genetic influence on naevogenesis [37].

Clinical Features Symptoms Choroidal nevi do not usually cause any symptoms and are mostly diagnosed on routine ophthalmoscopy. A macular nevus can cause visual loss RPE detachment, RPE atrophy, chronic foveal edema, and persistent subretinal fluid inducing photoreceptor atrophy (Fig.  3.7) [38]. Subretinal fluid in association with nevus may induce symptoms of metamorphopsia or photopsia (Fig. 3.8).

a

b

Fig. 3.6  Choroidal nevus and a choroidal freckle (arrow). Note normal choroidal vessels passing undisturbed through the choroidal freckle

Fig. 3.7  Retinal atrophy secondary to long-standing choroidal nevus. Fundus photograph (a) and spectral domain OCT (b)

23

3  Benign Melanocytic Tumors of the Uvea

a

b

c

d

Fig. 3.8 Choroidal nevus associated with subretinal fluid. Fundus photograph (a) and early (b) and late (c) fluorescein showing pinpoint RPE defects and leakage.

The presence of subretinal fluid is easily confirmed by spectral domain OCT (b)

Signs Choroidal nevus appears as a slate gray to brown lesion with minimal thickness. The margins are usually ill defined. Drusen and orange pigmentation (lipofuscin) are commonly associated features (Fig.  3.9a, b). Retinal pigment epithelium (RPE) atrophy/proliferation might be present as well as RPE detachment (Fig. 3.9c, d) and choroidal neovascularization as signs of chronicity (Fig.  3.10). Retinal pigment epithelial changes (prior resolved subretinal fluid) are recently described as long-term features (Fig. 3.11) [39].

Clinical Variants In addition to typical choroidal nevi, some larger choroidal lesions, and lesions with certain clinical features described below, have been variously categorized as suspicious nevi, intermediate lesions, indeterminate lesions, and even small melanomas [40]. Such lesions can be included as variants of a typical choroidal nevus. Amelanotic Nevus In some instances, the nevus may be completely amelanotic (Fig. 3.12), making it difficult to dif-

24

M. Materin and A. D. Singh

a

b

c

d1

d2

Fig. 3.9  Choroidal nevus with associated drusen (a) and orange pigmentation (lipofuscin) (b), retinal pigment epithelium atrophy (c), and RPE detachment (d)

25

3  Benign Melanocytic Tumors of the Uvea

a

b

c

d

Fig. 3.10  Choroidal nevus with associated choroidal neovascularization. Fundus photograph (a) and early (b) and late (c) fluorescein showing lacy hyperfluorescence

a

Fig. 3.11  A 50-year-old female presented with a choroidal mass measuring 8.1 mm × 7.3 mm in basal dimension. Note overlying retinal pigment epithelial proliferation and fibrous metaplasia and evidence of resolved subretinal fluid along the inferotemporal margin (a). The tumor was dome shaped (b) with a height of 3.0  mm and medium

and leakage. The presence of choroidal neovascular membrane is also suggested by spectral domain OCT (d)

b

reflectivity on A scan ultrasonography. The options of transpupillary thermotherapy, plaque brachytherapy, and enucleation were discussed, but the patient opted for close observation because of concerns of visual loss. On follow­up, most recently 3.8 years after initial presentation, the tumor continues to remain stable

26

M. Materin and A. D. Singh

a

b

c

d

Fig. 3.12  Amelanotic choroidal nevus. Amelanotic choroidal mass, 8 × 8 mm in basal dimension, in the inferior temporal quadrant of the right eye (a). Overlying drusen were present (arrows). Subretinal fluid and orange pigmentation were absent. Normal choroidal vasculature could be identified within the lesion (arrow head). Ultrasonographically the lesion measured about 1.2 mm in thickness with medium to high reflectivity (b). ICG was

also performed and confirmed the presence of normal choroidal vasculature within the lesion (c). Spectral domain OCT showed irregularly thickened Bruch’s membrane and outer retinal atrophy (arrow) suggestive of chronicity of the lesion (d). Details of the choroidal structures were not visible. Retinal edema, subretinal fluid, and dispersed lipofuscin associated with growing lesions were conspicuously absent

ferentiate them from choroidal metastases, choroidal hemangioma or inflammatory granuloma. Detailed past history (prior cancer), social history (tobacco use), and review of symptoms can provide supportive evidence. Intrinsic features of the lesion such as color (brown, yellow, orange, white), margins (sharp or ill defined), and visibility of intrinsic vessels (absent, normal, or abnormal) are important for differentiation. Secondary changes such as vitreous cells (inflammatory granuloma), drusen (chronicity), orange pigment (recent growth), SRF (acute), subretinal hemorrhage (indicative of secondary choroidal neovas-

cular membrane), lipid exudation (vascular incompetence), retinal pigment epithelial (RPE) atrophy (chronic), and choroidal folds (subchoroidal location) provide subtle but important clues regarding the nature of the choroidal mass (Table 3.2). Halo Nevus A choroidal nevus that has a distinct central pigmented region surrounded by a yellow halo is termed halo nevus (Fig.  3.13). Halo nevus may be more frequent in individuals with a prior history of cutaneous melanoma or vitiligo [41, 42].

Yellow Absent

Yellow Absent

Granuloma

Scleritis

Absent

Absent

Present

Present/ absent Absent

Drusen Present

Present/ absent Present/ absent

Present

Present/ absent Absent

Present/ absent Present/ absent

Present/ absent

Present

SRF Absent

Present

Orange pigment Absent

High/ medium High/ medium

High

Medium

USG-A scan High/ medium Low

Hypofl

Hyperfl (early) Hypofl (late) Hypofl

Hypofl

Variable

ICG Variable

Variable

Variable

OCT Retinal atrophy SRF/retinal edema SRF/retinal edema Variable

Choroidal folds

Systemic evaluation

ICG

Ophthalmoscopic features and USG A-scan Multiple/bilateral

Most diagnostic feature Ophthalmoscopic features

SRF subretinal fluid, USG A-scan ultrasonographic A-scan internal reflectivity, ICG indocyanine green angiography, Hypofl hypofluorescence, Hyperfl hyperfluorescence

Orange Absent

Intrinsic vessels Color visible Yellow Present/Normal vessels Yellow Present/abnormal vessels Yellow Absent

Hemangioma

Amelanotic melanoma Metastasis

Entity Amelanotic nevus

Table 3.2  Diagnostic features of amelanotic choroidal nevus. Clinical features and ancillary studies

3  Benign Melanocytic Tumors of the Uvea 27

28

M. Materin and A. D. Singh

Choroidal Melanocytosis Diffuse dark choroidal hyperpigmentation indicative of choroidal melanocytosis in otherwise normal fundus can be an isolated feature or occur in the setting of ocular dermal melanocytosis (Nevus of Ota) (Fig. 3.14) [43]. Isolated choroidal melanocytosis may be a risk factor for choroidal melanoma (Fig. 3.15).

Fig. 3.13  Halo choroidal nevus. Choroidal nevus that has a distinct central pigmented region surrounded by a yellow halo is termed halo nevus

Indeterminate Choroidal Melanocytic Lesions Choroidal melanocytic lesions that are larger in size than a typical choroidal nevus, i.e., those exceeding basal diameter of 5 mm or a thickness of 1  mm, have been labeled as small choroidal melanoma according to COMS criteria [44]. In the COMS small melanoma prospective study of 188

a

b

c

d

Fig. 3.14  Ocular dermal melanocytosis (Nevus of Ota). Note unilateral cutaneous hyperpigmentation in trigeminal (V1 and V2) distribution with varying hair color (a). Note dark iris and

episcleral pigmentation (b). Enucleated globe containing a large amelanotic melanoma (c). There was also widespread choroidal thickening due to excessive melanocytes (d)

29

3  Benign Melanocytic Tumors of the Uvea

a

b

c

d

Fig. 3.15  Isolated choroidal melanocytosis. A 40-year-­ old Indian asymptomatic woman. There was no significant prior medical or ocular history. Visual acuity with correction was 20/20 bilaterally. There was absence of eyelid or episcleral pigmentation, and her irides were blue in color. Anterior segment examination including gonioscopy was unremarkable. Fundus examination revealed large area of flat choroidal hyperpigmentation with indis-

tinct margins involving most of the posterior poles in each eye (a, b). The foveal reflexes were preserved. An occasional small drusen was present, but subretinal fluid and orange pigment were absent. Fluorescein angiography showed no evidence of blockage or leakage (c, d). Ultrasound-B scan showed absence of choroidal thickening. Optical coherent tomography showed normal retina

tumors that were initially observed, the Kaplan– Meier estimates of the probability of growth were 11%, 21%, and 31% at 1, 2, and 5 years of enrolment, respectively. It is important to realize that 63% of small tumors classified as melanoma did not grow during 5 years (Chap. 17).

chronicity such as drusen and RPE alterations suggest a diagnosis of nevus rather than melanoma (Fig. 3.16) [45].

Large or Giant Nevus Some nevi can reach larger proportions similar to that of a medium-size melanoma based upon size criteria devised for the COMS (5.1– 16.0  mm in largest basal diameter and 2.6 to 10.0 mm in height). Such cases are challenging to diagnose. However, presence of features of

 low Growth and Progressive Surface S Changes Choroidal nevi, being benign in nature, are expected to grow slowly, if at all. Such slow growth was observed in 4 of 75 adult Chinese subjects participating in The Beijing Eye Study [46]. Over a period of 5 years, largest diameter of the nevus increased by 0.25 mm (mean) in association with increase in drusen. None of the nevi regressed in size or showed clinical signs indicat-

30

a

Fig. 3.16  An 82-year-old man (current age) presented 19.5 years ago with a pigmented choroidal mass measuring 11.0 mm × 11.0 mm in basal dimension and height of 2.8  mm located posterior to the equator of his right eye

ing a malignant transformation. Similar slow growth of choroidal nevi, particularly in younger individuals ( 0.5 mm Observed lesions Retrospective Choroidal nevus without clinical evidence of malignant transformation

5 or more 3.3 (mean) 7 or more

Note varying inclusion criteria RPE retinal pigment epithelium, COMS Collaborative Ocular Melanoma Study a Data approximated for comparison b Inactive = minimal symptoms, good vision, and absence of subretinal fluid Table 3.4  Statistically significant clinical features predictive of subsequent growth of a choroidal nevus Risk factor (relative risk) Size (mm) Secondary Effects Basal Orange diameter pigment Drusen Thickness (present) (present) Author Gass S S S S Augsburger S NS S NS Butler NS NS S (2.7) S (1.8) Shields NS NS S (1.5) S (5.2) COMS S (5.2) S (0.2) S(6.4) S (17.7) Singh NS NS S (9.6) S (8.2)

Subretinal fluid NS

Adjacent RPE Changes S

Juxtapapillary location NA

Symptoms (present) NS

S

NS

S

S

S (3.0)

NS

NS

S (3.3)

S(1.4)

NA

S (2.6)

S (1.8)

NS

S (0.2)

NS

NA

NS

NS

S (6.3)

S (4.9)

S significant, NS not significant, NA not assessed, RPE retinal pigment epithelium, COMS Collaborative Ocular Melanoma Study

32

M. Materin and A. D. Singh

a

b

Fig. 3.17 Choroidal nevus followed for 10  years. Overlying drusen present with lack of risk factors for growth (a). Same tumor showing rupture Bruch’s mem-

brane, surrounded by hemorrhage, which has now transformed into a mushroom-shape choroidal melanoma (b)

neovascularization are uncommon secondary effects of choroidal nevi, which do not indicate malignancy; actually, these clinical features indicate chronicity [56]. Choroidal melanocytic lesions with chronic features need to continue to be monitored, since malignant transformation is still possible after longer follow-up (Fig.  3.17) [45]. An online referral guide that is based upon scoring system generated by input of relevant clinical features has been validated to allow appropriate follow-up in the community by stratification of individual risk [57]. In a large cohort of patients, the risk of growth can be estimated according to the presence or absence of associated clinical features (Table 3.5) [52, 58]. Nevertheless, in an individual patient, the distinction between small choroidal melanoma and choroidal nevus is not absolute [28] [59]. Supplemental analysis of data from the small melanoma prospective study component of the COMS revealed that tumor thickness (2.0–3.0 mm), larger diameter (12.1–16.0 mm), or presence of orange pigmentation was associated with 5-year growth rate of more than 50% suggesting higher probability of such an indeterminate lesion of being a small melanoma than a large nevus.

Table 3.5  Risk factors predictive of growth of a choroidal nevus

Diagnostic Evaluation Fundus photography, fluorescein angiography, autofluorescence imaging, and ultrasonography

Risk factors Thickness > 2.0 mm Posterior margin touching the optic disc Presence of visual symptoms Presence of orange pigment Presence of sub retinal fluid

Combination of Risk Factors None present Any one present

Risk 5% 36%

Any two present

45%

Any three present Any four present All present

50% 51% 56%

Based on data from Ref. [58]

can document the size, appearance, and secondary effects of a choroidal nevus. The fluorescein angiographic [28, 54] and ultrasonographic features [60] have been investigated as risk factors for growth, with mixed results. The role of fine-­needle aspiration biopsy remains controversial [61]. In the past, optical coherence tomography of a choroidal nevus was useful only for imaging the overlying retina for secondary changes, [62, 63] and OCT findings complement clinical examination by verifying and documenting secondary retinal and RPE changes associated with choroidal melanocytic lesions (Fig. 3.18) [63, 64]. With the development of enhanced depth imaging optical coherence tomography, a deeper and more accurate visualization of small choroidal melanocytic lesions that are undetectable by ultrasonography is achievable [65, 66]. Although EDI SD-OCT identified the tumor distinctly from

33

3  Benign Melanocytic Tumors of the Uvea

a

b

c

d

1,000 µm

Fig. 3.18  Fundus photograph of an indeterminate choroidal melanocytic lesion (small choroidal melanoma) with orange pigment and subretinal fluid (a). On FAF, orange pigment appears as a focal hyperautofluorescent spot (b). In addition, there is diffuse dispersion of orange

pigment within the SRF imparting diffuse hyperautofluorescence to the SRF. Time domain OCT only reveals shallow SRF (c) which is more clearly observed with SD-OCT (d). Note dispersed lipofuscin or photoreceptor within the subretinal space

the hyperreflective Bruch’s membrane/RPE layer (anterior) and normal choroid (lateral), the interphase between the posterior limit of the tumor and the inner sclera is only visible with tumors that are flat. Amelanotic nevi appear homogenous with a medium reflective band and visible choroidal vessels within the tumor (Fig.  3.19). Melanocytic nevi have a highly reflective band

within the choriocapillaris layer with posterior shadowing (Fig.  3.20). Choroidal melanomas lack visibility of posterior margins because of their pigment content and increased thickness. Further improvements in OCT technology are needed for better visualization of choroidal tumors. Choroidal imaging has also been improved by innovations in technology that use a

34

M. Materin and A. D. Singh

a

a

a

b

c

c

d

Fig. 3.19  Amelanotic nevus. Color fundus photograph shows amelanotic peripapillary nevus (a). B-scan ultrasound shows a mildly elevated choroidal lesion with low internal reflectivity and a basal diameter of 5.0 mm (b). SD-OCT conventional acquisition mode. A medium reflective choroidal band and some vessels under retinal pigment epithelium layer are seen in the tumor area (c). Enhanced depth imaging SD-OCT technique (inverted). A more detailed choroidal image is seen (d). A fusiform medium to low reflective band and intrinsic choroidal vessels are noticed. The inner sclera boundary is also visible. (Reprinted from Torres et al. [65]. With permission from Elsevier)

d

Fig. 3.20  Melanotic nevus. Color fundus photograph shows a flat and well-demarcated melanocytic nevus (a). B-scan ultrasound is unable to detect the lesion (b). SD-OCT (Topcon-1000): The lesion appears as a sharply highly reflective band at the Bruch’s/retinal pigment epithelium/choriocapillaris layer (c). SD-OCT enhanced depth imaging technique (inverted): The lesion is well distinguished from surrounding normal choroid as highly reflective band with posterior shadowing (d). A thin hyporeflective line separates the RPE and the anterior tumor surface. (Reprinted from Torres et al. [65]. With permission from Elsevier)

35

3  Benign Melanocytic Tumors of the Uvea

laser source across a range of wavelengths [67]. The swept laser light source sequentially emits various frequencies increasing the signal quality from deep tissues such as choroid [68]. Characteristics of choroidal nevi imaged by SS OCT include visualization of intralesional details such as vessels, cavities, and granularity [69]. Nevertheless, similar to EDI OCT, image quality is superior for choroidal lesions that are small, closer to the fovea, with lighter choroidal pigmentation [69, 70]. Further improvements in OCT technology are needed for better visualization of choroidal tumors.

Within indeterminate melanocytic tumors, indocyanine angiography may demonstrate an intrinsic vasculature, which would suggest malignancy [59, 71]. Indocyanine green angiography is more suitable for the evaluation of the choroidal vascular pattern than fluorescein angiography [72] and has been used with confocal scanning laser ophthalmoscopy for this purpose [59, 71]. In preliminary studies, a statistically significant association between complex microcirculatory patterns (parallel with cross linking, arcs with branching, loop and/or network) and tumor growth has been observed (Fig. 3.21) [59, 71].

a

b

c

d

Fig. 3.21  Indocyanine green angiographic vascular patterns of small choroidal melanocytic lesions. In the parallel vascular pattern, linear blood vessels seem to extend from normal adjacent choroid to the region of the tumor and could be traced undisturbed to go across the tumor margin into the surrounding choroid (a arrow). Tortuous

vessels were essentially similar to parallel vessels except minimal tortuosity of the vessels (b arrow). In a branching vascular pattern, the parallel vessels showed branching (c arrow). The loops vascular pattern comprised of circular vascular loops (d arrow). (Reprinted from Singh et  al. [40]. With permission from Elsevier)

M. Materin and A. D. Singh

36

Optical coherence tomography angiography (OCT-A) has shown different flow rates between choroidal nevus and choroidal melanoma. OCT-A on choroidal nevus demonstrated normal choroidal vasculature with similar flow rate compared to normal eye. Differently, choroidal melanoma presented with alteration on Bruch’s membrane-­ RPE complex and on the outer retinal layer. Choriocapillaris flow rate is higher on choroidal nevus compared to choroidal melanoma [73]. Other studies showed lack of blood flow in outer retina and normal choriocapillaris in choroidal nevus and irregular vascularity and choriocapillaris in choroidal melanoma [74].

Ciliary Body Nevus Clinical Features

Reactive retinal pigment epithelial hyperplasia develops after inflammation or trauma and is usually deeply pigmented, with discrete, irregular edges. Congenital hypertrophy of the retinal pigment epithelium (CHRPE) is dark brown or black, with sharply demarcated margins [75]. CHRPE may have central or peripheral clear areas (lacunae) but drusen, orange pigment, subretinal fluid and RPE fibrosis are absent. There can be slow growth over several years [75].

Nevi of the ciliary body have been rarely reported in the literature, but they are suspected to occur more frequently [78, 79]. A ciliary body nevus usually appears as a dome-shaped mass with a smooth surface. Intrinsic vascularity is usually not present. Unexplained sentinel vessels, sectoral cataract, or localized shallowing of the anterior chamber should prompt evaluation of the ciliary body. Gonioscopic evaluation provides a satisfactory view of the extent of the lesion. Ultrasound biomicroscopy provides useful information about the size, extent, and internal consistency (Fig.  3.22); however, no specific features are diagnostic of ciliary body nevus. Monitoring of these lesions with ophthalmic exam and UBM, when they are small, looking for documented growth, represents at this time the best way to diagnose a benign versus a malignant tumor. The differential diagnosis of a pigmented ciliary body nevus includes melanocytoma, melanoma, and adenoma or adenocarcinoma of the pigmented ciliary epithelium. These are described in the relevant sections of this text. In practice, the diagnosis is usually made histologically, with biopsy or after enucleation.

Treatment

Complications

At present, only periodic observation is recommended for choroidal nevi. Associated subretinal fluid has been treated with surface and surrounding laser photocoagulation [76]. Photodynamic therapy may be effective in rare cases when associated choroidal neovascularization is present [77]. However, treatment of subretinal fluid and/or neovascularization on the surface of a choroidal nevus is controversial since these treatments may alter the natural evolution of underlying lesion with a delay in the treatment of a possible small choroidal melanoma.

Many ciliary body nevi would remain asymptomatic until they reach a critical size when they may induce secondary cataract or glaucoma as a result of pigment shedding.

Differential Diagnosis

Treatment There is no consensus as to whether ciliary body tumors should be observed or excised. Before undertaking complex excisional surgery, there may be scope for incisional or fine-needle aspiration biopsy [80]. However, detection of benign

37

3  Benign Melanocytic Tumors of the Uvea

a

b

c

Fig. 3.22  Ultrasound biomicroscopy of an iridociliary mass. Slit lamp photograph (a). Ultrasound biomicroscopy revealed a ciliary body mass with anterior extension into the iris root (b). Note close correlation between histo-

pathological appearance (after iridocyclectomy) and biomicroscopic findings (c). (Reprinted from Bakri et  al. [143]. With permission from Wolters Kluwer Health, Inc.)

cells does not entirely exclude malignancy, because of the risk of sampling error [81]. Excision under a lamellar scleral flap is the treatment of choice for small, circumscribed ciliary body tumors (involving no more than three clock hours) and without extrascleral extension [81]. En bloc excision with simultaneous full-­thickness corneoscleral resection is indicated when extraocular extension is present [81, 82]. Another approach is to treat ciliary body tumors with plaque or proton beam radiotherapy, perhaps excising any extraocular tumor nodule if necessary.

Melanocytosis Introduction Ocular melanocytosis is a congenital condition characterized by hyperpigmentation of the episclera and uvea (Fig. 3.14) [83, 84]. Associated cutaneous hyperpigmentation in the distribution of the trigeminal nerve is called oculodermal melanocytosis (Nevus of Ota) (Chap. 4) [85]. The orbit and meninges can also be involved. Ota also described patients with a combination of vascular and melanocytic nevi in a Japanese

M. Materin and A. D. Singh

38

population (phakomatosis pigmentovascularis [PPV]) [86]. Five types of PPV are known, but there has been an attempt in the dermatology literature to reclassify this order into three subtypes: 1) phacomatosis cesioflammea (one or more blue spot with one or more port-­wine stain), 2) phacomatosis spilorosea (speckled lentiginous nevus of macular type with telangiectatic nevus), and 3) phacomatosis cesiomarmorata (blue spot with cutis marmorata telangiectatica congenita) [87]. Systemic associations with Sturge-Weber syndrome or Klippel-­ Trenaunay-­ Weber syndrome can occur. Recent work has shown that this is

likely caused by activating mutations in GNA11 and GNAQ [88]. Importantly, patients with this condition may also develop choroidal melanoma and should be monitored carefully for this [89]. Ocular involvement can also include congenital glaucoma, iris mammillations, and oculodermal melanocytosis (Fig. 3.23) [90].

Association with Uveal Melanoma In the Caucasian population, several observations support an association between oculo(dermal)

a

b

c

d

Fig. 3.23 Phakomatosis pigmentovascularis. Bilateral cutaneous hemangioma of Klippel-Trenaunay-Weber syndrome (a) with bilateral scleral hyperpigmentation (b, c ocular melanocytosis) and choroidal mass in the right eye causing exudative retinal detachment (d).

Ultrasonography B-scan (e, dome-shaped mass) and A-scan (f, low internal reflectivity) and ICG (g, hypofluorescence with intrinsic vessels) were suggestive of choroidal melanoma rather than hemangioma. Uveal melanoma was confirmed upon enucleation

3  Benign Melanocytic Tumors of the Uvea

e

39

f

0µs

10µs

20µs

30µs

40µs

g

Fig. 3.23 (continued)

melanocytosis and uveal melanoma (Fig.  3.14) [50, 91, 92]. These include the occurrence of uveal melanoma ipsilateral to ocular melanocytosis, [92] and the development of uveal melanoma in the sector of the eye affected with melanocytosis. Although, ocular and oculodermal melanocytosis is common in Orientals, the occurrence of uveal melanoma in this population is rare [93]. A few cases of uveal melanoma in Hispanics [94] and Blacks [95] with oculo(dermal) melanocytosis has also been observed. Uveal melanoma in patients with ocular melanocytosis might carry a higher risk for metastases compared to patients without this condition [96].

Biological Basis The biologic basis for a susceptibility to the development of uveal melanoma in oculo (der-

mal) melanocytosis is not known. Excessive melanocytes in the uveal tract of patients with oculo (dermal) melanocytosis may be the reason for susceptibility [84]. Recent genetic studies have identified GNAQ as a link between nevus of Ota and uveal melanoma offering an explanation for increased risk of developing uveal melanoma in patients with nevus of Ota [97, 98].

Risk Estimate The age at diagnosis of uveal melanoma in association with oculo (dermal) melanocytosis is no different to that of sporadic uveal melanoma (Fig. 3.24) [84]. Overall, it is estimated that the lifetime risk of developing uveal melanoma in a Caucasian with oculo (dermal) melanocytosis is about 1 in 400 [84].

40

M. Materin and A. D. Singh

100

ocular melanocytosis (melanosis oculi) [26, 106]. An alternative term of magnocellular nevus emphasizes its neural crest origin as a variant of a nevus [107]. Other descriptive terms such as benign melanoma of the optic nerve head are no longer used [108].

90

Percent

80 70 60 50 40 30

Etiology

20 10 0 10

20

30 40 50 60 70 80 Age at Diagnosis (years) Sporadic

90 100

ODM

Fig. 3.24  Comparison of age at diagnosis of uveal melanoma in patients with oculo(dermal) melanocytosis (ODM) and 100 randomly selected sporadic uveal melanoma patients. (Modified from Singh et al. [84]. With permission from Elsevier)

Clinical Variants In addition to ipsilateral unilateral, unifocal uveal melanoma, rare cases of bilateral [99] and multifocal uveal melanomas [100] tend to occur in the presence of oculo (dermal) melanocytosis. In cases of primary melanoma of the orbit [101] and central nervous system, the presence of ocular melanocytosis should be excluded [102].

Treatment It is generally recommended that patients with oculo (dermal) melanocytosis be monitored annually.

Optic Disc Melanocytoma Introduction Melanocytoma is a benign pigmented ocular tumor, which predominantly involves the optic disc and uvea [103]. Other rare sites include the sclera [104] and conjunctiva [105]. The term melanocytoma was proposed by Zimmerman and Garron because they observed a resemblance between melanocytoma cells and those seen in

Optic disc and uveal melanocytoma are considered to be congenital hamartomas, [26] arising from dendritic uveal melanocytes scattered throughout the uvea [109].

Pathology Optic disc melanocytoma is a dark pigmented mass, often extending from the optic disc to the surrounding choroid and retina [26, 106]. The cells are deeply pigmented because of numerous macromelanosomes, and cellular details are not evident histologically until bleaching is performed. Most cells are plump, round, or polyhedral. A smaller population of lightly pigmented spindle-shaped melanocytoma cells can also be present [110]. The nuclei are bland, uniformly small, and normochromic. Mitotic figures are usually absent.

Clinical Features Symptoms Although optic disc melanocytoma is a congenital entity, it is rarely detected in childhood. The mean age at diagnosis is 50  years [111]. Optic disc melanocytoma is more commonly seen in Blacks and darker races than in Whites [109]. Most patients are asymptomatic, [109, 111] and detected on routine ophthalmoscopy. Signs On ophthalmoscopy, optic disc melanocytoma is a dark-brown or black and flat or slightly elevated mass, usually located inferotemporally (Fig. 3.25). The choroid is involved in about 54%

41

3  Benign Melanocytic Tumors of the Uvea

significant visual symptoms, unlike an optic disc melanocytoma. In addition, a melanoma is brown, does not usually infiltrate the nerve fiber layer, and may reveal intrinsic vasculature on ophthalmoscopy or angiography. Adenoma of the retinal pigment epithelium may be difficult to differentiate from an optic disc melanocytoma on the basis of clinical findings alone [114]. Combined hamartoma of the retinal pigment epithelium and retina is generally seen in a younger age group and has prominent vascular and gliotic components.

Fig. 3.25  Optic disc melanocytoma. Fundus photograph of the left eye showing a partially pigmented superficial mass that extends onto the disc and obscures retinal vessels

of cases, and a retinal component is present in about 30% of cases [111]. The retinal involvement usually appears darker than the choroidal component, and has feathery margins due to extensions into the nerve fiber layer. Larger melanocytomas can completely obscure the optic disc and may lead to pigment dispersion into the vitreous cavity (Fig. 3.25). Prominent intrinsic vasculature, subretinal fluid, and retinal exudation are not usually present.

Associations Unilateral optic disc melanocytoma is not associated with other ocular or systemic anomalies; however, bilateral tumors have been reported in association with optic nerve hypoplasia and central nervous system abnormalities [112].

Differential Diagnosis Optic disc melanocytoma should be differentiated from optic disc melanoma, [113] adenoma of the juxtapapillary retinal pigment epithelium, [114] and combined hamartoma of the retinal pigment epithelium and retina. Optic disc melanoma usually arises from the juxtapapillary choroid and extends over the optic nerve causing

Diagnostic Evaluation The diagnosis of optic disc melanocytoma is usually suspected on ophthalmoscopic examination. Fundus photographs are used to document and monitor the lesion over prolonged periods of time. On fluorescein angiography, optic disc melanocytoma appears as an area of dense hypofluorescence, which persists through all phases of the angiogram, unlike choroidal melanoma where hyperfluorescence due to intrinsic vasculature is noted. Intrinsic vasculature is characteristically absent in optic disc melanocytoma (Fig.  3.26). Similar findings are noted on indocyanine angiography. Ultrasonography B-scan detects an acoustically solid optic disc mass with high initial spike. Optical coherence tomography shows a high reflectance signal anteriorly, which is continuous with the retinal nerve fiber layer, and there is optical shadowing posteriorly [115]. Nevertheless, OCT can help delineate the margins of the tumor [116]. Two categories of OCT findings in optic disc melanocytoma have been identified: type 1, hyperreflective lesion with disorganized overlying retina and a posterior hyporeflective shadow, and (b) the less common, type 2 lesion overlaid by a relatively well-organized retina without a posterior hyporeflective shadow [117]. Magnetic resonance imaging (T1-weighted image with fat suppression technique) may be used to detect enlargement of the optic nerve and demarcate the posterior extension. The melanocytoma appears hyperintense with respect to the vitreous due to paramagnetic properties of mela-

42

M. Materin and A. D. Singh

a

b

c

Fig. 3.26 Fundus photograph of the right eye showing a large melanocytoma completely obscuring the optic disc. Note its feathery margins (a). The tumor was associated with pigment dispersion in the vitreous cavity (b).

Fluorescein angiography shows area of dense hypofluorescence that persist throughout all the phases of the angiogram corresponding to the location of the tumor (c). (Reprinted from Singh [144]. With permission from Elsevier)

nin. Enhancement with gadolinium may also be present. With availability of 7-T MRI, definition the involvement of the optic nerve can be more precise (Fig. 3.27) [118].

Prognosis

Treatment Once documented, most optic disc melanocytomas are kept under periodic observation [111]. Any eye demonstrating a rapid increase in the size of an optic disc melanocytoma is usually enucleated because of concerns of malignant transformation.

A large majority of optic disc melanocytomas remains stable over many years [109, 111]. Subtle growth over several years is observed in about 10% of cases [111, 119]. An afferent pupillary defect and nerve fiber bundle visual field defects are seen even when the visual acuity is normal, suggesting asymptomatic optic nerve dysfunction [111]. Rapid deterioration of vision from infarction and swelling of an optic disc melanocytoma may manifest as papillitis, neuro-­retinitis, and central retinal artery or vein occlusion (Fig.  3.27). Rapid enlarge-

43

3  Benign Melanocytic Tumors of the Uvea

a

b

c

d

Fig. 3.27  A 36-year-old Arabic woman presented with severe vision loss. Wide-field fundus photograph of the left eye on initial presentation demonstrating a markedly edematous optic nerve with subretinal fluid in the macula, nerve fiber layer hemorrhages, and vascular tortuosity (a). At 1-month follow-up, the edema of the nerve and retina has improved, revealing a darkly pigmented peripapillary lesion involving multiple layers. T1-weighted MRI (3 T)

of the orbit demonstrating apparent intrinsic enhancement of the optic nerve extending 15 mm posterior to the globe (c, arrow). Magnetic resonance imaging (7 T) of the orbit, indicating actual involvement of optic nerve was approximately 2 mm posterior to the globe (d, arrow). The lesion can be identified as the hypointensity in the juxtapapillary globe. (Reprinted from Baartman et  al. [118]. with permission from Wolters Kluwer Health, Inc.)

ment, indicative of malignant transformation into melanoma, is observed in about 2% of cases [111, 120, 121].

guishable from uveal nevus and melanoma, and most are probably managed as such. As with optic disc melanocytoma, uveal melanocytoma can give rise to melanoma [122].

Uveal Melanocytoma Uveal melanocytomas are histologically similar to optic disc melanocytoma [109]. Such tumors have been seen in iris, ciliary body, and choroid. Uveal melanocytomas are clinically indistin-

Iris Melanocytoma Clinical Features Only about 60 cases of iris melanocytoma are published [123]. In a study of 189 pathologic

44

M. Materin and A. D. Singh

a

b

c

d

Fig. 3.28  A 77-year-old woman complaining of floaters in the right eye. The visual acuity was of 6/12 in the right eye and 6/9 in the left eye. The left eye was normal. In the right eye, external examination did not reveal sentinel vessels. Anterior segment examination showed a mild sectoral cataract superonasally. In addition, there was pigment deposition around the posterior capsule of the lens and dispersed pigment in the anterior vitreous (a). Fundus examination revealed a dome-shaped ciliary body

mass between 1 and 2 o’clock hours position. The tumor was better visualized by gonioscopy and was black in color with a lobulated surface (b). Anterior extension of the ciliary body mass into the angle and pigment deposition in the angle inferiorly were apparent (c). B-scan ultrasonography confirmed the presence of a 5  ×  4  ×  5  mm ciliary body mass with high internal reflectivity on A scan (d). (Reprinted from Belfort et al. [130]. With permission from BMJ Publishing Group Ltd.)

samples of iris and ciliary body tumors suspected to be melanoma, about 5% were confirmed to be melanocytomas [7]. The usual age at presentation is about 35 years [123] with rare presentation in childhood [124]. The tumors are invariably darkly pigmented and nodular in appearance (Fig.  3.28). They most commonly involve the inferior quadrants and have a predilection for the iris root Associated features that are helpful in suspecting the diagnosis of iris melanocytoma are the presence of stromal and angle pigment seeding and the absence of ectropion iridis, sectoral cataract, and visible intrinsic vascularization.

Complications In a series of 47 patients with iris melanocytoma, the main complications included pigment shedding (34%), progressive enlargement (23%), and secondary glaucoma (11%) [123]. Tumor necrosis with the sudden onset of melanocytomalytic glaucoma is a rare complication [125, 126]. Malignant transformation of iris melanocytoma is extremely uncommon [127], and slow growth of the tumor usually does not imply such a change [122, 123]. Treatment Periodic observation is generally recommended, unless there are concerns about malignancy, in

45

3  Benign Melanocytic Tumors of the Uvea

which case fine-needle aspiration biopsy or local resection is performed.

mimic extrascleral growth of a ciliary body melanoma. One feature suggestive of melanocytoma is a uniform, black appearance, which is rare in ciliary body melanoma.

Ciliary Body Melanocytoma

Complications In a review of 40 patients with ciliary body melanocytoma, the main complications included anterior chamber extension (85%), progressive enlargement, and secondary glaucoma (13%) [128]. Tumor necrosis, uveitis, and pigment dispersion in the anterior chamber or vitreous cavity may also occur (Fig.  3.29) [129, 130]. Malignant transformation of ciliary body melanocytoma is extremely uncommon, and slow growth of the tumor usually does not imply such a change [122, 128].

Clinical Features Only about 40 cases of ciliary body melanocytoma are published [128]. The usual age at presentation is about 47 years [128, 129] with rare presentation in childhood [128]. The tumors are invariably darkly pigmented and nodular in appearance. There is no predilection for any particular quadrant. Although benign, ciliary body melanocytomas tend to extend into the anterior chamber and extraocularly. Such extension can

a

b

c

d

Fig. 3.29  Choroidal melanocytoma. Left fundus shows a tumor with black pigmentation along the base (a). Enucleated globe shows a pigmented peripapillary mass (b). The tumor cells are heavily pigmented with large granules (c Hematoxylin and eosin; original magnification × 500). The cells appear fairly uniform with benign cytologic features (d Hematoxylin and eosin after bleach-

ing; original magnification × 500). Electron micrograph demonstrating cytoplasmic macromelanosomes within the tumor cells (e, original magnification × 14,000). Macromelanosomes are about 10 times larger compared with melanosomes within normal choroidal melanocytes (f). (Reprinted from Brownstein et al. [133]. With permission from Elsevier)

46

e

M. Materin and A. D. Singh

f

Fig. 3.29 (continued)

Treatment Periodic observation is generally not recommended because of difficulty in differentiating ciliary body melanocytoma from ciliary body melanoma. Fine-needle aspiration biopsy followed by ultrasound biomicroscopic observation may be applicable in cases not associated with necrosis and glaucoma [131]. In general, iridocyclectomy under a lamellar corneoscleral flap is the preferred treatment. If the sclera is involved, a full-thickness corneoscleral graft may be required [81]. Enucleation is limited for cases with uncontrolled glaucoma [132].

Choroidal Melanocytoma Clinical Features Only about 15 cases of choroidal melanocytoma are published [133]. The usual age at presentation is between 30 and 50 years [133] with rare presentation in childhood [133, 134]. The tumors are usually darkly pigmented and dome shaped in appearance [133]. A diffuse variant has also been described [135]. A choroidal melanocytoma can be indistinguishable from melanoma on ophthalmoscopy, ultrasonography, and angiography [136, 137]. In fact, several cases of choroidal melanocytoma have been treated with plaque radiotherapy in the Collaborative Ocular Melanoma Study with a presumptive clinical diagnosis of melanoma, the correct diagnosis

being made only retrospectively, when the eye was enucleated for radiation-related complications [137, 138].

Complications Necrosis and inflammation can occur within a melanocytoma leading to blind painful eye [139, 140]. Choroidal melanocytoma can rarely undergo malignant transformation into melanoma [122, 139, 141]. Treatment The best management of choroidal melanocytoma is controversial. If clinically suspected, smaller lesions may be observed. Fine-needle aspiration biopsy can be misleading as it may only reveal melanocytoma and may miss associated regions of melanoma [139, 141]. The correct diagnosis may be achieved only when the globe is examined histologically after enucleation [137, 139, 141].

References 1. Harbour JW, Brantley MA Jr, Hollingsworth H, et  al. Association between posterior uveal melanoma and iris freckles, iris naevi, and choroidal naevi. Br J Ophthalmol. 2004;88(1):36–8. PubMed PMID: 14693769. 2. Rodriguez-Sains RS. Ocular findings in patients with dysplastic nevus syndrome. An update. Dermatol Clin. 1991;9(4):723–8. PubMed PMID: 1934647. 3. Toth-Molnar E, Olah J, Dobozy A, et al. Ocular pigmented findings in patients with dysplastic naevus syndrome. Melanoma Res. 2004;14(1):43–7. PubMed PMID: 15091193.

3  Benign Melanocytic Tumors of the Uvea 4. Bataille V, Sasieni P, Cuzick J, et  al. Risk of ocular melanoma in relation to cutaneous and iris naevi. Int J Cancer. 1995;60(5):622–6. PubMed PMID: 7860135. 5. Schwab C, Mayer C, Zalaudek I, et  al. Iris freckles a potential biomarker for chronic sun damage. Invest Ophthalmol Vis Sci. 2017;58(6):BIO174-BIBIO9. Epub 2017/07/18. https://doi.org/10.1167/iovs.1721751. PubMed PMID: 28715582. 6. Laino AM, Berry EG, Jagirdar K, et  al. Iris pigmented lesions as a marker of cutaneous melanoma risk: an Australian case-control study. Br J Dermatol. 2018;178(5):1119–1127. Epub 2018/01/10. https:// doi.org/10.1111/bjd.16323. PubMed PMID: 29315480. 7. Jakobiec FA, Silbert G.  Are most iris ‘melanomas’ really nevi? A clinicopathologic study of 189 lesions. Arch Ophthalmol. 1981;99(12):2117–32. PubMed PMID: 7305709. 8. Jakobiec FA, Moorman LT, Jones IS.  Benign epithelioid cell nevi of the iris. Arch Ophthalmol. 1979;97(5):917–21. PubMed PMID: 444127. 9. Shields JA, Karan DS, Perry HD, et al. Epithelioid cell nevus of the iris. Arch Ophthalmol. 1985;103(2):235– 7. PubMed PMID: 3977695. 10. Margo CE, Groden L. Balloon cell nevus of the iris. Am J Ophthalmol. 1986;102(2):282–3. PubMed PMID: 3740194. 11. Ticho BH, Tso MO, Kishi S.  Diffuse iris nevus in oculodermal melanocytosis: a light and electron microscopic study. J Pediatr Ophthalmol Strabismus. 1989;26(5):244–50. PubMed PMID: 2795414. 12. Cogan DG, Reese AB.  A syndrome of iris nodules, ectopic Descemet’s membrane, and unilateral glaucoma. Doc Ophthalmol. 1969;26:424–33. PubMed PMID: 5359536. 13. Eagle RC Jr, Font RL, Yanoff M, et  al. Proliferative endotheliopathy with iris abnormalities. The iridocorneal endothelial syndrome. Arch Ophthalmol. 1979;97(11):2104–11. PubMed PMID: 508177. 14. Ragge NK, Acheson J, Murphree AL. Iris mammillations: significance and associations. Eye. 1996;10(Pt 1):86–91. PubMed PMID: 8763309. 15. Gunduz K, Shields CL, Shields JA, et  al. Iris mammillations as the only sign of ocular melanocytosis in a child with choroidal melanoma. Arch Ophthalmol. 2000;118(5):716–7. PubMed PMID: 10815167. 16. Ragge NK, Falk RE, Cohen WE, et al. Images of Lisch nodules across the spectrum. Eye. 1993;7(Pt 1):95– 101. PubMed PMID: 8325432. 17. Carlson DW, Alward WL, Folberg R.  Aggressive nevus of the iris with secondary glaucoma in a child. Am J Ophthalmol. 1995;119(3):367–8. PubMed PMID: 7872400. 18. Paridaens D, Lyons CJ, McCartney A, et al. Familial aggressive nevi of the iris in childhood. Arch Ophthalmol. 1991;109(11):1552–4. PubMed PMID: 1755736. 19. Conway RM, Chew T, Golchet P, et  al. Ultrasound biomicroscopy: role in diagnosis and management in 130 consecutive patients evaluated for anterior segment tumours. Br J Ophthalmol. 2005;89:950–5.

47 20. Radhakrishnan S, Rollins AM, Roth JE, et al. Real-time optical coherence tomography of the anterior segment at 1310 nm. Arch Ophthalmol. 2001;119(8):1179–85. PubMed PMID: 11483086. 21. Bianciotto C, Shields CL, Guzman JM, et  al. Assessment of anterior segment tumors with ultrasound biomicroscopy versus anterior segment optical coherence tomography in 200 cases. Ophthalmology. 2011;118(7):1297–1302. Epub 2011/03/08. https:// doi.org/10.1016/j.ophtha.2010.11.011. PubMed PMID: 21377736. 22. Harbour JW, Augsburger JJ, Eagle RC Jr. Initial management and follow-up of melanocytic iris tumors. Ophthalmology. 1995;102(12):1987–93. PubMed PMID: 9098306. 23. Shields CL, Shields JA, Materin M, et  al. Iris melanoma: risk factors for metastasis in 169 consecutive patients. Ophthalmology. 2001;108(1):172–8. PubMed PMID: 11150284. 24. Territo C, Shields CL, Shields JA, et al. Natural course of melanocytic tumors of the iris. Ophthalmology. 1988;95(9):1251–5. PubMed PMID: 3211503. 25. Shields CL, Kaliki S, Hutchinson A, et  al. Iris nevus growth into melanoma: analysis of 1611 consecutive eyes: the ABCDEF guide. Ophthalmology. 2013;120(4):766– 772. Epub 2013/01/08. https://doi.org/10.1016/j.ophtha.2012.09.042. PubMed PMID: 23290981. 26. Zimmerman LE. Melanocytes, melanocytic nevi, and melanocytomas. Investig Ophthalmol. 1965;34:11– 41. PubMed PMID: 14271285. 27. Group COMS.  COMS manual of procedures: accession no. PBS 179693. Springfield: National Technical Information Service; 1995. 28. Anonymous. Factors predictive of growth and treatment of small choroidal melanoma: COMS report no. 5. The Collaborative Ocular Melanoma Study Group. Arch Ophthalmol. 1997;115(12):1537–44. PubMed PMID: 9400787. 29. Singh AD, Kalyani P, Topham A. Estimating the risk of malignant transformation of a choroidal nevus. Ophthalmology. 2005;112(10):1784–9. PubMed PMID: 16154197. 30. Yoshikawa H, editor. The incidence of choroidal nevus in Japan. New Orleans: American Academy of Ophthalmology; 2004. 31. Jonas JB, You QS, Xu L, et al. Choroidal nevi in adult Chinese. Ophthalmology. 2008;115(6):1102- 11e1. Epub 2008/06/04. https://doi.org/10.1016/j.ophtha.2007.12.022. PubMed PMID: 18519073. 32. Nangia V, Jonas JB, Agarwal S, et al Choroidal nevi in adult Indians: the Central India Eye and Medical Study. Br J Ophthalmol. 2012;96(11):1443–1444. Epub 2012/08/09. https://doi.org/10.1136/bjophthalmol-2012-302164. PubMed PMID: 22872672. 33. Keel S, Xie J, Foreman J, et  al Prevalence and characteristics of choroidal nevi: the Australian National Eye Health Survey. Clin Exp Ophthalmol. 2018;46(7):777–782. Epub 2018/03/03. https://doi. org/10.1111/ceo.13188. PubMed PMID: 29498193. 34. Naumann G, Yanoff M, Zimmerman LE. Histogenesis of malignant melanomas of the uvea. I. Histopathologic

48 characteristics of nevi of the choroid and ciliary body. Arch Ophthalmol. 1966;76(6):784–96. PubMed PMID: 5924936. 35. McLean IW, Foster WD, Zimmerman LE, et  al. Modifications of Callender’s classification of uveal melanoma at the Armed Forces Institute of Pathology. Am J Ophthalmol. 1983;96(4):502–9. PubMed PMID: 6624832. 36. Qiu M, Shields CL. Choroidal nevus in the United States adult population: racial disparities and associated factors in the National Health and Nutrition Examination Survey. Ophthalmology. 2015;122(10):2071–2083. Epub 2015/08/10. https://doi.org/10.1016/j.ophtha.2015.06.008. PubMed PMID: 26255109. 37. Csoma RZ, Toth-Molnar E, Varga A, et al. Risk factors and relationship of cutaneous and uveal melanocytic lesions in monozygotic and dizygotic twin pairs. PLoS One. 2016;11(8):e0160146. Epub 2016/08/04. https://doi.org/10.1371/journal.pone.0160146. PubMed PMID: 27486750; PubMed Central PMCID: PMCPMC4972429. 38. Shields CL, Furuta M, Mashayekhi A, et  al. Visual acuity in 3422 consecutive eyes with choroidal nevus. Arch Ophthalmol. 2007;125(11):1501– 1507. Epub 2007/11/14. https://doi.org/10.1001/ archopht.125.11.1501. PubMed PMID: 17998511. 39. Mashayekhi A, Siu S, Shields CL, et al. Retinal pigment epithelial trough: a sign of chronicity of choroidal nevi. Eur J Ophthalmol. 2012;22(6):1019–1025. Epub 2012/02/22. https://doi.org/10.5301/ejo.5000120. PubMed PMID: 22344466. 40. Singh AD, Mokashi AA, Bena JF, et al. Small choroidal melanocytic lesions : features predictive of growth. Ophthalmology. 2006;113(6):1032–9. 41. Fournier GA, Albert DM, Wagoner MD.  Choroidal halo nevus occurring in a patient with vitiligo. Surv Ophthalmol. 1984;28(6):671–2. Epub 1984/05/01. PubMed PMID: 6740487. 42. Shields CL, Maktabi AM, Jahnle E, et al. Halo nevus of the choroid in 150 patients: the 2010 Henry van Dyke Lecture. Arch Ophthalmol. 2010;128(7):859– 64. Epub 2010/07/14. https://doi.org/10.1001/ archophthalmol.2010.132. PubMed PMID: 20625046. 43. Kovoor TA, Bahl D, Singh AD, et  al. Bilateral isolated choroidal melanocytosis. Br J Ophthalmol. 2008;92(7):892, 1008. Epub 2008/06/26. https:// doi.org/10.1136/bjo.2008.138495. PubMed PMID: 18577638. 44. COMS G.  COMS manual of procedures: accession no. PBS 179693. Springfield: National Technical Information Service; 1995. 45. Li HK, Shields CL, Mashayekhi A, et al. Giant choroidal nevus clinical features and natural course in 322 cases. Ophthalmology. 2010;117(2):324–33. Epub 2009/12/09. https://doi.org/10.1016/j.ophtha.2009.07.006. PubMed PMID: 19969359. 46. You QS, Xu L, Jonas JB, et  al. Change in choroidal nevi during a 5-year follow-up study: the Beijing Eye Study. Br J Ophthalmol. 2010;94(5):575–8. Epub

M. Materin and A. D. Singh 2009/10/14. https://doi.org/10.1136/bjo.2009.165720. PubMed PMID: 19822910. 47. Mashayekhi A, Siu S, Shields CL, et  al. Slow enlargement of choroidal nevi: a long-term followup study. Ophthalmology. 2011;118(2):382–8. ­ Epub 2010/08/31. https://doi.org/10.1016/j.ophtha.2010.06.006. PubMed PMID: 20801518. 48. Elner VM, Flint A, Vine AK.  Histopathology of documented growth in small melanocytic choroidal tumors. Arch Ophthalmol. 2004;122(12):1876– 8. Epub 2004/12/15. https://doi.org/10.1001/ archopht.122.12.1876. PubMed PMID: 15596595. 49. Harbour JW, Paez-Escamilla M, Cai L, et  al. Are risk factors for growth of choroidal Nevi associated with malignant transformation? Assessment with a validated genomic biomarker. Am J Ophthalmol. 2018;197:168. Epub 2018/09/10. https://doi. org/10.1016/j.ajo.2018.08.045. PubMed PMID: 30195895. 50. Yanoff M, Zimmerman LE. Histogenesis of malignant melanomas of the uvea. 3. The relationship of congenital ocular melanocytosis and neurofibromatosis in uveal melanomas. Arch Ophthalmol. 1967;77(3):331– 6. PubMed PMID: 6019552. 51. Sahel JA, Pesavento R, Frederick AR Jr, et  al. Melanoma arising de novo over a 16-month period. Arch Ophthalmol. 1988;106(3):381–5. PubMed PMID: 3278703. 52. Singh AD, Schachat AP, Diener-West M, et  al. Small choroidal melanoma. Ophthalmology. 2008;115(12):2319–e3. Epub 2008/12/02. https://doi. org/10.1016/j.ophtha.2008.07.009. PubMed PMID: 19041491. 53. Augsburger JJ, Schroeder RP, Territo C, et al. Clinical parameters predictive of enlargement of melanocytic choroidal lesions. Br J Ophthalmol. 1989;73(11):911– 7. PubMed PMID: 2605146. 54. Butler P, Char DH, Zarbin M, et  al. Natural history of indeterminate pigmented choroidal tumors. Ophthalmology. 1994;101(4):710–6; discussion 7. PubMed PMID: 8152767. 55. Shields CL, Shields JA, Kiratli H, et al. Risk factors for growth and metastasis of small choroidal melanocytic lesions. Ophthalmology. 1995;102(9):1351–61. PubMed PMID: 9097773. 56. Zografos L, Mantel I, Schalenbourg A.  Subretinal choroidal neovascularization associated with choroidal nevus. Eur J Ophthalmol. 2004;14(2):123–31. PubMed PMID: 15134109. 57. Hemmerdinger C, Beech M, Groenewald C, et  al. Validation of an online referral guide for melanocytic fundus lesions. Ophthalmic Physiol Opt. 2011;31(6):574–9. Epub 2011/03/31. https://doi. org/10.1111/j.1475-1313.2011.00830.x. PubMed PMID: 21446922. 58. Shields CL, Cater J, Shields JA, et  al. Combination of clinical factors predictive of growth of small choroidal melanocytic tumors. Arch Ophthalmol. 2000;118(3):360–4. PubMed PMID: 10721958.

3  Benign Melanocytic Tumors of the Uvea 59. Singh AD, Mokashi AA, Bena JF, et  al. Small choroidal melanocytic lesions : Indocyanine angiographic features predictive of growth: a pilot study. Ophthalmology. 2006;113(6):1032–9. 60. Coleman DJ, Silverman RH, Rondeau MJ, et  al. Noninvasive in  vivo detection of prognostic indicators for high-risk uveal melanoma: ultrasound parameter imaging. Ophthalmology. 2004;111(3):558–64. PubMed PMID: 15019336. 61. Augsburger JJ, Correa ZM, Schneider S, et  al. Diagnostic transvitreal fine-needle aspiration biopsy of small melanocytic choroidal tumors in nevus versus melanoma category. Trans Am Ophthalmol Soc. 2002;100:225–32; discussion 32-4. PubMed PMID: 12545696. 62. Shields CL, Mashayekhi A, Materin MA, et al. Optical coherence tomography of choroidal nevus in 120 patients. Retina. 2005;25(3):243–52. PubMed PMID: 15805899. 63. Singh AD, Belfort RN, Sayanagi K, et  al. Fourier domain optical coherence tomographic and auto-­ fluorescence findings in indeterminate choroidal melanocytic lesions. Br J Ophthalmol. 2010;94(4):474–8. Epub 2009/10/14. https://doi.org/10.1136/ bjo.2009.162636. PubMed PMID: 19822920. 64. Sayanagi K, Pelayes DE, Kaiser PK, et  al. 3D spectral domain optical coherence tomography findings in choroidal tumors. Eur J Ophthalmol. 2011;21(3):271– 5. Epub 2010/11/03. https://doi.org/10.5301/ EJO.2010.5848. PubMed PMID: 21038307. 65. Torres VL, Brugnoni N, Kaiser PK, et  al. Optical coherence tomography enhanced depth imaging of choroidal tumors. Am J Ophthalmol. 2011;151(4):586– 93.e2. Epub 2011/01/25. https://doi.org/10.1016/j. ajo.2010.09.028. PubMed PMID: 21257150. 66. Shah SU, Kaliki S, Shields CL, et al. Enhanced depth imaging optical coherence tomography of choroidal nevus in 104 cases. Ophthalmology. 2012;119(5):1066– 72. Epub 2012/02/03. https://doi.org/10.1016/j.ophtha.2011.11.001. PubMed PMID: 22297027. 67. Mrejen S, Spaide RF. Optical coherence tomography: imaging of the choroid and beyond. Surv Ophthalmol. 2013;58(5):387–429. https://doi.org/10.1016/j.survophthal.2012.12.001. PubMed PMID: 23916620. 68. Adhi M, Duker JS.  Optical coherence tomography  – current and future applications. Curr Opin Ophthalmol. 2013;24(3):213–21. https://doi.org/10.1097/ ICU.0b013e32835f8bf8. PubMed PMID: 23429598; PubMed Central PMCID: PMCPMC3758124. 69. Francis JH, Pang CE, Abramson DH, et  al. Swept-­ source optical coherence tomography features of choroidal nevi. Am J Ophthalmol. 2015;159(1):169– 76.e1. https://doi.org/10.1016/j.ajo.2014.10.011. PubMed PMID: 25448319. 70. Ung C, Lains I, Silverman RF, et  al. Evaluation of choroidal lesions with swept-source optical coherence tomography. Br J Ophthalmol. 2018;103:88. https:// doi.org/10.1136/bjophthalmol-2017-311586. PubMed PMID: 29605814.

49 71. Mueller AJ, Freeman WR, Schaller UC, et al. Complex microcirculation patterns detected by confocal indocyanine green angiography predict time to growth of small choroidal melanocytic tumors: MuSIC Report II.  Ophthalmology. 2002;109(12):2207–14. PubMed PMID: 12466160. 72. Flower RW.  Infrared absorption angiography of the choroid and some observations on the effects of high intraocular pressures. Am J Ophthalmol. 1972;74(4):600–14. PubMed PMID: 4627576. 73. Ghassemi F, Mirshahi R, Fadakar K, et  al. Optical coherence tomography angiography in choroidal melanoma and nevus. Clin Ophthalmol. 2018;12:207–14. Epub 2018/02/07. https://doi.org/10.2147/OPTH. S148897. PubMed PMID: 29403262; PubMed Central PMCID: PMCPMC5784746. 74. Cennamo G, Romano MR, Breve MA, et al. Evaluation of choroidal tumors with optical coherence tomography: enhanced depth imaging and OCT-angiography features. Eye. 2017;31:906. https://doi.org/10.1038/ eye.2017.14. 75. Shields CL, Mashayekhi A, Ho T, et al. Solitary congenital hypertrophy of the retinal pigment epithelium: clinical features and frequency of enlargement in 330 patients. Ophthalmology. 2003;110(10):1968–76. PubMed PMID: 14522773. 76. Mines JA, Freilich DB, Friedman AH, et al. Choroidal (subretinal) neovascularization secondary to choroidal nevus and successful treatment with argon laser photocoagulation. Case reports and review of literature. Ophthalmologica. 1985;190(4):210–8. PubMed PMID: 2409499. 77. Parodi MB, Boscia F, Piermarocchi S, et al. Variable outcome of photodynamic therapy for choroidal neovascularization associated with choroidal nevus. Retina. 2005;25(4):438–42. PubMed PMID: 15933589. 78. Cogan DG, Kuwabara T. Tumors of the ciliary body. Int Ophthalmol Clin. 1971;11(3):27–56. PubMed PMID: 5152657. 79. Gordon E.  Nevus of the choroid and pars plana. Enucleation for suspected malignant melanoma. Surv Ophthalmol. 1963;123:507–11. PubMed PMID: 14084895. 80. El-Harazi SM, Kellaway J, Font RL.  Melanocytoma of the ciliary body diagnosed by fine-needle aspiration biopsy. Diagn Cytopathol. 2000;22(6):394–7. PubMed PMID: 10820536. 81. Rummelt V, Naumann GO, Folberg R, et al. Surgical management of melanocytoma of the ciliary body with extrascleral extension. Am J Ophthalmol. 1994;117(2):169–76. PubMed PMID: 8116745. 82. Raichand M, Peyman GA, Juarez CP, et al. Resection of uveal melanocytoma: clinicopathological correlation. Br J Ophthalmol. 1983;67(4):236–43. PubMed PMID: 6830742. 83. Fitzpatrick TB, Kitamura H, Kukita A, et  al. Ocular and dermal melanocytosis. AMA Arch Ophthalmol. 1956;56(6):830–2. PubMed PMID: 13371988.

50 84. Singh AD, De Potter P, Fijal BA, et  al. Lifetime prevalence of uveal melanoma in white patients with oculo(dermal) melanocytosis. Ophthalmology. 1998;105(1):195–8. PubMed PMID: 9442799. 85. Ota M, Tanino H.  Nevus fusco-ceruleus ophthalmomaxillaris. Tokyo Med J. 1939;63:1243–5. 86. Ota M, Kawamura T, Ito N. Phakomatosis pigmentovascularis. Ota Jpn J Dermatol. 1947;52:1–3. 87. Happle R. Phacomatosis pigmentovascularis revisited and reclassified. Arch Dermatol. 2005;141(3):385–8. PubMed PMID: 15781681. 88. Thomas AC, Zeng Z, Riviere JB, et al. Mosaic activating mutations in GNA11 and GNAQ are associated with phakomatosis pigmentovascularis and extensive dermal melanocytosis. J Invest Dermatol. 2016;136(4):770– 8. Epub 2016/01/19. https://doi.org/10.1016/j. jid.2015.11.027. PubMed PMID: 26778290; PubMed Central PMCID: PMCPMC4803466. 89. Tran HV, Zografos L. Primary choroidal melanoma in phakomatosis pigmentovascularis IIa. Ophthalmology. 2005;112(7):1232–5. PubMed PMID: 15939474. 90. Singh K, Dangda S, Mutreja A, et  al. Bilateral phacomatosis pigmentovascularis in a young male with developmental glaucoma and varicose veins. J Curr Glaucoma Pract. 2018;12(2):94–8. Epub 2018/11/27. https://doi.org/10.5005/jp-journals-10008-1251. PubMed PMID: 30473605; PubMed Central PMCID: PMCPMC6236123. 91. Gonder JR, Shields JA, Albert DM, et al. Uveal malignant melanoma associated with ocular and oculodermal melanocytosis. Ophthalmology. 1982;89(8):953–60. PubMed PMID: 7133641. 92. Velazquez N, Jones IS. Ocular and oculodermal melanocytosis associated with uveal melanoma. Ophthalmology. 1983;90(12):1472–6. PubMed PMID: 6677847. 93. Biswas J, Shanmugam MP, Gopal L. Malignant melanoma of the choroid in association with oculodermal melanocytosis: a case report. Indian J Ophthalmol. 1995;43(3):140–1. PubMed PMID: 8822492. 94. Infante de German-Ribon R, Singh AD, Arevalo JF, et  al. Choroidal melanoma with oculodermal melanocytosis in Hispanic patients. Am J Ophthalmol. 1999;128(2):251–3. PubMed PMID: 10458194. 95. Nik NA, Glew WB, Zimmerman LE. Malignant melanoma of the choroid in the nevus of Ota of a black patient. Arch Ophthalmol. 1982;100(10):1641–3. PubMed PMID: 7138332. 96. Shields CL, Kaliki S, Livesey M, et al. Association of ocular and oculodermal melanocytosis with the rate of uveal melanoma metastasis: analysis of 7872 consecutive eyes. JAMA Ophthalmol. 2013;131(8):993–1003. Epub 2013/05/18. https://doi.org/10.1001/jamaophthalmol.2013.129. PubMed PMID: 23681424. 97. Van Raamsdonk CD, Bezrookove V, Green G, et  al. Frequent somatic mutations of GNAQ in uveal melanoma and blue naevi. Nature. 2009;457(7229):599– 602. Epub 2008/12/17. https://doi.org/10.1038/ nature07586. PubMed PMID: 19078957; PubMed Central PMCID: PMC2696133.

M. Materin and A. D. Singh 98. Gerami P, Pouryazdanparast P, Vemula S, et  al. Molecular analysis of a case of nevus of Ota showing progressive evolution to melanoma with intermediate stages resembling cellular blue nevus. Am J Dermatopathol. 2010;32(3):301– 5. Epub 2010/01/30. https://doi.org/10.1097/ DAD.0b013e3181b96db7. PubMed PMID: 20110797. 99. Singh AD, Shields CL, Shields JA, et  al. Bilateral primary uveal melanoma. Bad luck or bad genes? Ophthalmology. 1996;103(2):256–62. PubMed PMID: 8594511. 100. Pomeranz GA, Bunt AH, Kalina RE.  Multifocal choroidal melanoma in ocular melanocytosis. Arch Ophthalmol. 1981;99(5):857–63. PubMed PMID: 7236090. 101. Rice CD, Brown HH.  Primary orbital melanoma associated with orbital melanocytosis. Arch Ophthalmol. 1990;108(8):1130–4. PubMed PMID: 2383203. 102. Sang DN, Albert DM, Sober AJ, et  al. Nevus of Ota with contralateral cerebral melanoma. Arch Ophthalmol. 1977;95(10):1820–4. PubMed PMID: 911255. 103. Howard GM, Forrest AW. Incidence and location of melanocytomas. Arch Ophthalmol. 1967;77(1):61– 6. PubMed PMID: 6015721. 104. Lee JS, Smith RE, Minckler DS. Scleral melanocytoma. Ophthalmology. 1982;89(2):178–82. PubMed PMID: 7070788. 105. Verdaguer J, Valenzuela H, Strozzi L. Melanocytoma of the conjunctiva. Arch Ophthalmol. 1974;91(5):363–6. PubMed PMID: 4821790. 106. Zimmerman LE, Garron LK. Melanocytoma of the optic disc. Int Ophthalmol Clin. 1962;2:431–40. 107. Cogan DG.  Discussion- pigmented ocular tumors. In: Boniuk M, editor. Ocular and adnexal tumors; new and controversial aspects. St Louis: CV Mosby; 1964. p. 385. 108. Reese AB.  Congenital melanomas. Am J Ophthalmol. 1974;77(6):789–808. PubMed PMID: 4599354. 109. Reidy JJ, Apple DJ, Steinmetz RL, et  al. Melanocytoma: nomenclature, pathogenesis, natural history and treatment. Surv Ophthalmol. 1985;29(5):319–27. PubMed PMID: 3992470. 110. Juarez CP, Tso MO.  An ultrastructural study of melanocytomas (magnocellular nevi) of the optic disk and uvea. Am J Ophthalmol. 1980;90(1):48–62. PubMed PMID: 7395958. 111. Shields JA, Demirci H, Mashayekhi A, et  al. Melanocytoma of optic disc in 115 cases: the 2004 Samuel Johnson memorial lecture, part 1. Ophthalmology. 2004;111(9):1739–46. PubMed PMID: 12545696. 112. Brodsky MC, Phillips PH.  Optic nerve hypoplasia and congenital hypopituitarism. J Pediatr. 2000;136(6):850. PubMed PMID: 10839891. 113. Loeffler KU, Tecklenborg H.  Melanocytoma-­ like growth of a juxtapapillary malignant mela-

3  Benign Melanocytic Tumors of the Uvea noma. Retina. 1992;12(1):29–34. PubMed PMID: 1565868. 114. Shields JA, Eagle RC Jr, Shields CL, et al. Pigmented adenoma of the optic nerve head simulating a melanocytoma. Ophthalmology. 1992;99(11):1705–8. PubMed PMID: 1454346. 115. Antcliff RJ, ffytche TJ, Shilling JS, et  al. Optical coherence tomography of melanocytoma. Am J Ophthalmol. 2000;130(6):845–7. PubMed PMID: 11124315. 116. Sutter C, Jean-Charles A, Merle H.  Clinical and paraclinical course of melanocytoma of the optic disk. Contribution of spectral- and time-domain OCT in the study of 10 patients. J Fr Ophtalmol. 2011;34(9):615–23. https://doi.org/10.1016/j. jfo.2011.02.009. PubMed PMID: 21663999. 117. Apinyawasisuk S, McCannel T, Arnold AC. Clinical and spectral-domain optical coherence tomography appearance of optic disc melanocytoma: a new classification and differentiation from pigmented choroidal lesions. Ocul Oncol Pathol. 2017;3(2):142–8. Epub 2017/09/05. https://doi. org/10.1159/000453110. PubMed PMID: 28868286; PubMed Central PMCID: PMCPMC5566758. 118. Baartman BJ, Ahmad B, Srivastava S, et  al. Melanocytoma or Juxtapapillary Melanoma? Retin Cases Brief Rep. 2017;13:15. Epub 2017/01/07. https://doi.org/10.1097/ICB.0000000000000520. PubMed PMID: 28060135. 119. Mansour AM, Zimmerman L, La Piana FG, et  al. Clinicopathological findings in a growing optic nerve melanocytoma. Br J Ophthalmol. 1989;73(6):410–5. PubMed PMID: 2751972. 120. Apple DJ, Craythorn JM, Reidy JJ, et al. Malignant transformation of an optic nerve melanocytoma. Can J Ophthalmol. 1984;19(7):320–5. PubMed PMID: 6525581. 121. Meyer D, Ge J, Blinder KJ, et al. Malignant transformation of an optic disk melanocytoma. Am J Ophthalmol. 1999;127(6):710–4. PubMed PMID: 10372881. 122. Roth AM.  Malignant change in melanocytomas of the uveal tract. Surv Ophthalmol. 1978;22(6):404– 12. PubMed PMID: 675502. 123. Demirci H, Mashayekhi A, Shields CL, et  al. Iris melanocytoma: clinical features and natural course in 47 cases. Am J Ophthalmol. 2005;139(3):468–75. PubMed PMID: 15767055. 124. Shields JA, Eagle RC Jr, Shields CL, et al. Progressive growth of an iris melanocytoma in a child. Am J Ophthalmol. 2002;133(2):287–9. PubMed PMID: 11812446. 125. Fineman MS, Eagle RC Jr, Shields JA, et  al. Melanocytomalytic glaucoma in eyes with necrotic iris melanocytoma. Ophthalmology. 1998;105(3):492–6. PubMed PMID: 9499781. 126. Teichmann KD, Karcioglu ZA.  Melanocytoma of the iris with rapidly developing secondary glaucoma. Surv Ophthalmol. 1995;40(2):136–44. PubMed PMID: 8533102.

51 127. Cialdini AP, Sahel JA, Jalkh AE, et  al. Malignant transformation of an iris melanocytoma. A case report. Graefes Arch Clin Exp Ophthalmol. 1989;227(4):348–54. PubMed PMID: 2777104. 128. LoRusso FJ, Boniuk M, Font RL.  Melanocytoma (magnocellular nevus) of the ciliary body: report of 10 cases and review of the literature. Ophthalmology. 2000;107(4):795–800. PubMed PMID: 10768345. 129. Frangieh GT, el Baba F, Traboulsi EI, et  al. Melanocytoma of the ciliary body: presentation of four cases and review of nineteen reports. Surv Ophthalmol. 1985;29(5):328–34. PubMed PMID: 3992471. 130. Belfort RN, Schoenfield L, Singh AD.  Ciliary body mass with vitreous pigment dispersion. Br J Ophthalmol. 2010;94(9):1260–1, 70-1. Epub 2009/02/13. https://doi.org/10.1136/ bjo.2008.154872. PubMed PMID: 19211607. 131. Mohamed MD, Gupta M, Parsons A, et al. Ultrasound biomicroscopy in the management of melanocytoma of the ciliary body with extrascleral extension. Br J Ophthalmol. 2005;89(1):14–6. PubMed PMID: 15615738. 132. Biswas J, D'Souza C, Shanmugam MP. Diffuse melanotic lesion of the iris as a presenting feature of ciliary body melanocytoma: report of a case and review of the literature. Surv Ophthalmol. 1998;42(4):378– 82. PubMed PMID: 9493281. 133. Brownstein S, Dorey MW, Mathew B, et  al. Melanocytoma of the choroid: atypical presentation and review of the literature. Can J Ophthalmol. 2002;37(4):247–52. PubMed PMID: 12095099. 134. Lehman LJ, Hohberger GG, Buettner H, et  al. Necrotic melanocytoma of the choroid in a 2-year-­ old child. J Pediatr Ophthalmol Strabismus. 1997;34(1):40–3. PubMed PMID: 9027679. 135. Haas BD, Jakobiec FA, Iwamoto T, et  al. Diffuse choroidal melanocytoma in a child. A lesion extending the spectrum of melanocytic hamartomas. Ophthalmology. 1986;93(12):1632–8. PubMed PMID: 3808622. 136. Shields JA, Font RL. Melanocytoma of the choroid clinically simulating a malignant melanoma. Arch Ophthalmol. 1972;87(4):396–400. PubMed PMID: 5018243. 137. Robertson DM, Campbell RJ, Salomao DR.  Mushroom-shaped choroidal melanocytoma mimicking malignant melanoma. Arch Ophthalmol. 2002;120(1):82–5. PubMed PMID: 11786063. 138. Group COMS.  Accuracy of diagnosis of cho roidal melanomas in the Collaborative Ocular Melanoma Study. COMS report no. 1. Arch Ophthalmol. 1990;108(9):1268–73. PubMed PMID: 2205183. 139. Kurli M, Finger PT, Manor T, et al. Finding malignant change in a necrotic choroidal melanocytoma: a clinical challenge. Br J Ophthalmol. 2005;89(7):921–2. PubMed PMID: 15965181. 140. Binkley EM, Biscotti CV, Singh A, et al. Conservative management of ciliary body mass with associated

52 ocular inflammation. Retina. 2019. [epub ahead of print]. 141. Shetlar DJ, Folberg R, Gass JD. Choroidal malignant melanoma associated with a melanocytoma. Retina. 1999;19(4):346–9. PubMed PMID: 10458304. 142. Aponte EP, Stern RM, Hayden BC, et al. Iridocorneal Endothelial (ICE) Syndrome. A disease often misdiagnosed as an iris tumor. Adv Ocular Care. 2010;2010:31–2.

M. Materin and A. D. Singh 143. Bakri SJ, Sculley L, Singh AD.  Imaging techniques for uveal melanoma. Int Ophthalmol Clin. 2006;46:1–13. 144. Singh AD.  Optic disc melanocytoma. In: Huang D, Kaiser P, Lowder CY, Traboulsi EI, editors. Retinal Imaging. Philadelphia: Elsevier, Inc; 2006. p. 556–8.

4

Uveal Melanoma: Epidemiologic Aspects Nakul Singh, Stefan Seregard, and Arun D. Singh

Introduction Melanomas of the uvea and conjunctiva comprise approximately 5% of all melanomas [1]. The majority (95%) of ocular melanomas are uveal in origin, whereas primary conjunctival melanomas are very rare [1–3]. Uveal melanoma is the most common primary intraocular malignant tumor in adults [4, 5]. In this chapter, the incidence of uveal melanoma and various etiological factors are briefly reviewed (Box 4.1).

N. Singh School of Medicine, Case Western University, Cleveland, OH, USA S. Seregard Ophthalmic Pathology and Oncology Service and Department of Clinical Neuroscience, St. Erik Eye Hospital and Karolinska Institutet, Stockholm, Sweden A. D. Singh (*) Department of Ophthalmic Oncology, Cole Eye Institute, Cleveland Clinic, Cleveland, OH, USA e-mail: [email protected]

Box 4.1. Important Epidemiological Features of Uveal Melanoma • 5% of all melanoma arise from the ocular and adnexal structures. • 95% of ocular melanomas arise in uvea. • The incidence of uveal melanoma in the United States is 5.2 per million per year. • The incidence of uveal melanoma has remained stable for the last 50 years. • There is strong racial variation in the incidence, with the white population most commonly affected. • Clinical, epidemiological, physiological, and genetic data argue against a major role of UV light in the causation of uveal melanoma. • Oculodermal melanocytosis predisposes to uveal melanoma. • BAP1 cancer predisposition syndrome should be suspected in phenotypic variants.

Incidence The reported incidence of uveal melanoma ranges from 5.3 to 10.9 cases per million population per year because of variations in inclusion criteria, diagnostic criteria, and the methodology

© Springer Nature Switzerland AG 2019 B. E. Damato, A. D. Singh (eds.), Clinical Ophthalmic Oncology, https://doi.org/10.1007/978-3-030-17879-6_4

53

54

N. Singh et al.

used in calculating the incidence. Both crude and age-standardized incidences have been published, along with different stratum weights in the standard populations chosen; therefore, these rates are not entirely comparable. In some studies, uveal melanoma was included with melanoma of the conjunctiva and eyelids [6]. Some of the older studies have included only histopathologically confirmed cases. Because an increasing proportion of uveal melanomas are treated by brachytherapy, cases diagnosed only clinically should be included, as well as those confirmed histologically [2]. In a recent study from the United States, only diagnostic codes with uvea as the primary site (iris, ciliary body, and choroid) were considered and other ocular sites were excluded [2]. The mean age-adjusted incidence of uveal melanoma in the United States was 5.2 per million per year (5.0–5.4; 95% CI) (Fig.  4.1) [7]. Males had a significantly higher age-adjusted incidence of 6.0 per million population (5.7– 6.3; 95% CI) compared to females, where the average age-adjusted incidence was 4.5 (4.3– 4.7; 95% CI) [3]. The data were derived from the Surveillance and Epidemiology and End Result (SEER) program of the National Institutes of Health (Maryland, USA). The SEER program collects and provides reliable population-based incidence data for a wide variety of cancers, including uveal melanoma, in the United States population [8].

The SEER data are considered to be “gold standard” with a high degree of ascertainment, quality, and completeness [9]. However, SEER data may be hampered when describing rarer malignancies such as choroidal melanoma, which at the time may involve few incident cases per year [10]. The greatest likelihood of underreporting is from nonhospital sources, which may not be applicable to choroidal melanoma as almost all (98%) cases of uveal melanoma are reported from the hospital-based reporting sources [10]. A recent report based on data from the North American Association of Central Cancer Registries, which tabulated data on 62% of US population (as compared to SEER data that covers about 10% of the population), reveals comparable incidence rates, indicating the robustness of SEER data [10]. However, lack of histologic confirmation in cases that are treated with radiotherapy may contribute to underreporting [11].

Fig. 4.1 Age-adjusted incidence of uveal melanoma, 1973–2013. Number of persons per million population (y-axis) adjusted to the US 2000 population (uveal mela-

noma C69.2-C69.4 only). (Modified from Aronow et al. [7]. Copyright © 2017, S. Karger AG, Basel)

Global Incidence The incidence of uveal melanoma has been reported in several countries (Table 4.1) [2, 12]. The incidence of uveal melanoma in the United States [7], England [13], and various European countries is similar to that in Australia [14] and New Zealand [15], where the population is exposed to a higher intensity of ultraviolet light.

4  Uveal Melanoma: Epidemiologic Aspects

55

Table 4.1  Published reports on national incidence of uveal melanoma Incidence/million per year

Author United States Strickland

Period

Country

Definition

No. of cases Criteria

1950–1974

United States

Eye melanoma

–

–

Scotto Singh

1969–1971 1973–1997

United States United States

Eye melanoma Uveal melanoma

341 2493

Clinical Clinical

Singh

1973–2008

United States

Uveal melanoma

4070

Clinical

Aronow

1973–2013

United States

Uveal melanoma

4999

Clinical

Europe Mallone Jensen Lommatzsch Swerdlow

1995–2002 1943–1952 1961–1980 1962–1977

Europe Denmark East Germany England (UK)

Uveal melanoma Uveal melanoma Eye melanoma Ocular melanoma

4097 305 4284

Clinical Histologic Clinical Clinical

Keenan Raivio Teikari Vidal Gislason

1999–2010 1953–1973 1973–1980 1992 1955–1979

England (UK) Finland Finland France Iceland

Uveal melanoma Cbd + choroid Cbd + choroid Uveal melanoma Cbd + choroid

359 382 412 29

Clinical Histologic Clinical Clinical Histologic

Mork Bergman

1953–1960 1960–1998

Norway Sweden

Ocular melanoma Uveal melanoma

220 2997

Histologic Clinical

Australia Kricker

1996–1998

Australia

Choroidal melanoma

539

Clinical

11.0 (Male) 7.8 (Female)

Asia Iscovich Kaneko Park Cheng

1961–1989 1977–1979 1999–2011 1979–1996

Israel Japan South Korea Taiwan

Cbd + Choroid Uveal melanoma Uveal melanoma Ocular melanoma

502 82 326 128

Clinical Histologic Clinical Clinical

5.7 (Jews) 0.3 0.42 0.39

9.0 (Male) 8.0 (Female) 5.6 4.9 (Male) 3.7 (Female) 5.8 (Male) 4.4 (Female) 6.0 (Male) 4.5 (Female) 5.1 7.4 10 7.2 (Male) 5.7 (Female) 10 5.3 7.6 7 7.0 (Male) 5.0 (Female) 9 9.4 (Male) 8.8 (Female)

Uveal melanoma: Iris, ciliary body (Cbd), and choroidal melanoma Eye melanoma: Uveal and conjunctival melanoma Ocular melanoma: Uveal, conjunctival, and eyelid melanoma

Sex- and Age-Specific Incidence

Uveal Melanoma in Pregnancy

Uveal melanoma is more commonly seen in the older age group, with a progressively rising age-­ specific incidence rate, which peaks at the age of 70 years (24.5 per million in males and 17.8 per million in females) (Fig.  4.1) [2]. Similar data regarding the age distribution have been reported from Sweden, although the peak incidence in females (26.5 cases per million) occurs a decade earlier than the peak incidence in the male population (36.6 per million) [16].

Following a report of increased incidence of uveal melanoma during pregnancy [17], and following exposure to sex hormones (in women with intact ovaries, on menopausal estrogen therapy, or who have ever been pregnant) [18], subsequent studies demonstrated that exposure to sex hormones (estrogen and progesterone and other hormonal variables, such as bilateral oophorectomy, reproductive history, oral contraceptives use, and menopausal hormone replacement ther-

56

N. Singh et al.

apy) does not impact uveal melanoma risk [19– 22]. On the contrary, pregnancy has been shown to protect against uveal melanoma and mortality from metastasis [20, 23]. Studies have demonstrated that uveal melanomas lack estrogen receptors [17, 24] and progesterone receptors [24]. There is scope for testing the hypothesis that pregnancy promotes rapid

tumor growth by inducing melanocyte-­stimulating hormone to cause growth of uveal melanoma precursor cells and by altering immunity [25]. Several cases of uveal melanomas occurring during pregnancy have been reported in the literature [25, 26], with at least two documented cases of accelerated melanoma growth during pregnancy (Fig.  4.2) [17, 25, 27]. Cases of benign

a

b

c

d

0µs

10µs

20µs

30µs

40µs

50µs

e

Fig. 4.2  A 28-year-old Caucasian female (Gravida 3, Para 2) was referred at 30 weeks of gestation for evaluation of blurry vision in the left eye for three months. A pigmented choroidal mass was observed against the posterior lens temporally, with pigment dispersion in the vitreous (a). Fundoscopy revealed total exudative retinal detachment. B-scan ultrasonography revealed a large dome-shaped choroidal mass 20  ×  20  mm in base and 12.0 mm thickness as well as total retinal detachment (b). Diagnostic A-scan ultrasonography demonstrated low internal reflectivity, consistent with the clinical diagnosis

of uveal melanoma (c). Enucleation of the left eye was performed under fetal heart rate monitoring. Histopathology confirmed uveal melanoma (d) that was composed predominantly of spindle cells and numerous tumor-infiltrating lymphocytes (e). Gene expression profile of the tumor was of Class 1A molecular signature (lower risk of metastasis). Therefore, systemic staging for metastasis was delayed until the pregnancy was completed. She subsequently delivered at full term normal female child with birth weight of 8 pounds. There was no evidence of metastasis to the placenta and the baby

4  Uveal Melanoma: Epidemiologic Aspects

choroidal nevi transforming into active melanomas during pregnancy have also been reported [26]. In considering the prognosis of the mother, such reports suggest a propensity for accelerated melanoma growth during pregnancy, cautioning against observation and delay of treatment while highlighting the need for close follow-up of uveal melanoma during pregnancy. An important question is whether maternal uveal melanoma affects the fetus. It is estimated that only 25% of at-risk infants are affected [28]. While placental metastasis of cutaneous melanoma is rare, fetal involvement confers poor fetal prognosis. Therefore, thorough examination of the placenta is advised [29]. In a large study examining uveal melanoma during pregnancy (n = 16) placental or infant metastases were not observed in any cases [26]. However, a case of primary choroidal melanoma metastasizing to the liver, skeleton, and placenta during pregnancy (1 year postenucleation) has been reported [30]. Overall, current evidence does not support the claim that pregnancy impacts the progression and prognosis of maternal choroidal melanoma.

Temporal Stability Unlike global trends for rising incidence of cutaneous melanoma, the incidence of uveal melanoma has either remained stable or declined slightly over last several decades [13]. In the United States between 1973 and 2013 [2, 3, 7] and even for 25  years prior to that, stable age-­ adjusted annual incidences have been reported (Fig.  4.1) [4]. In Sweden, over the last four decades, the incidence has declined significantly for males (11.7 to 8.4 per million), but not statistically significantly for females (10.3 to 8.7 per million) [16]. Similarly, the incidence of uveal melanoma has remained stable from 1999 to 2010 in England [13].

57

fied several risk factors for development of uveal melanoma [32]. The significant host factors include atypical cutaneous nevi (OR 2.82, 95% CI 1.10–7.26), common cutaneous nevi (OR 1.74, 95% CI 1.27–2.39), propensity to sunburn (OR 1.64, 95% CI 1.29–2.09), iris nevi (OR 1.53, 95% CI 1.03–2.27), cutaneous freckles (OR 1.27, 95% CI 1.09–1.49), fair skin color (OR 1.80, 95% CI 1.31–2.47), and light iris color (OR 1.75, 95% CI 1.31–2.34). Reported environmental factors include welding (OR 2.05, 95% CI 1.20– 3.51) and occupational cooking (OR 1.81, 95% CI 1.33–2.46). Outdoor leisure activity, occupational sunlight exposure, latitude of birth, and hair color were not statistically significant [32].

Skin Color/Race Among the host factors, skin color/race seems to be the most significant, as uveal melanoma is much more common in Whites (98%) than in Blacks (0.6%) [7, 31, 33] as well as being less common in Asians (Fig.  4.3) [34]. Based upon SEER data (US), the relative risk of uveal ­melanoma is 1.2 for Asian and Pacific Islanders, 5.4 for Hispanics, and 19.2 for non-Hispanic white patients as compared with the black patients [33].

Iris Color Light iris color, a surrogate for race, is a well-­ known risk factor for uveal melanoma [32, 35– 37]. Iris color is coded by several pigment genes. Recently, single nucleotide polymorphisms (rs12913832, rs1129038, rs916977) that are known to be risk factors for cutaneous melanoma were also identified as risk factors for uveal melanoma [38]. Further studies are necessary to confirm these recent observations.

Host Factors

Cutaneous Nevi

In the large majority of cases, the etiology of uveal melanoma remains obscure [31]. A systemic review of published meta-analyses identi-

Common cutaneous nevi (OR 1.74, 95% CI 1.27–2.39) and cutaneous freckles (OR 1.27, 95% CI 1.09–1.49) have been identified as risk

58

N. Singh et al.

a

b

c

d

Fig. 4.3  Uveal melanoma in a 39-year-old Black male. Dilated fundus examination of the left eye revealed an elevated juxtapapillary lesion with a large inferior exudative retinal detachment (a). There was evidence of retinal invasion at the tumor apex with intraretinal hemorrhage. Ultrasonography revealed a mushroom-shaped lesion with largest basal diameter of 10  mm and thickness of

6.9 mm (b). The lesion had low internal echogenicity, and no extraocular extension was noted (c). Following discussion about various therapeutic options, the patient elected to undergo enucleation. Histopathologic analysis confirmed the diagnosis of choroidal melanoma with mixed cell morphology (d)

factors for uveal melanoma [32]. However, presence of atypical cutaneous nevi represent an even higher risk (OR 2.82, 95% CI 1.10–7.26) for uveal melanoma [32]. Atypical cutaneous nevus (also known as dysplastic nevus) denotes a specific clinicopathologic entity that is associated with an increased risk for the development of cutaneous melanoma [39]. The syndrome of autosomal dominant predisposition to cutaneous melanoma was originally described by Clark as the BK mole syndrome [40]. Familial atypical

mole and melanoma (FAM-M) syndrome is now the preferred terminology [41]. Because cutaneous and uveal melanocytes share similar embryologic, morphologic, and antigenic properties, it is plausible that uveal melanoma may sometimes occur within the spectrum of a FAM-M syndrome. An increased number of uveal nevi and the occurrence of uveal melanoma in patients with FAM-M syndrome and their families support an association between FAM-M syndrome and uveal melanoma [42, 43].

59

4  Uveal Melanoma: Epidemiologic Aspects

Moreover, CDKN2A mutations that account for 50% of cases with FAM-M syndrome [44] have not been observed in patients with uveal melanoma [45]. Recently, germline BAP1 mutations have been identified in patients with uveal and cutaneous melanomas [46, 47]. The cutaneous melanocytic proliferation has been characterized as atypical Spitz nevus/ melanoma [47, 48] or atypical nevoid melanoma-like melanocytic proliferations [46]. Some of the discrepancy in the different studies could be attributed to differences in case selection, variations in the study population, and a lack of uniform criteria used for the diagnosis of atypical nevi or FAM-M syndrome.

Uveal Nevi Associations between iris freckles, iris nevi, choroidal nevi, and posterior uveal melanoma have not been consistently observed [49–51]. However, clinical and histopathological evidence suggests that choroidal melanomas arise from preexisting choroidal nevi [52, 53]. It is also possible for choroidal melanoma to arise de novo [54]. The prevalence of choroidal nevi based on fundus photography in the white US population ranges from 4.6 to 7.9% [55]. A higher prevalence of nevi is reported when using a ultra-wide-field scanning laser ophthalmoscope [56]. Like uveal melanoma, choroidal nevus is associated with race, with an ageadjusted prevalence of 0.6% among Blacks, 2.7% among Hispanics, and 2.1% among others [57]. Also similar to uveal melanoma, choroidal nevus prevalence increases with age [57]. Interclass correlation coefficient for twins with choroidal nevi was 0.38 for monozygotic twins versus 0.02  in dizygotic twins, suggesting a genetic influence on naevogenesis [58]. Assuming that all melanomas arise from pre-­ existing nevi, the estimated annual risk of malignant transformation of a choroidal nevus ranges from 1 in 4300 to 8845 [59, 60]. The relationship between choroidal nevus and choroidal melanoma is further discussed elsewhere.

Table 4.2  Genetic subtypes of uveal melanoma Clinical feature Familial occurrence Occurrence at an earlier age Bilateral involvement Multiple primary tumors Phenotypic associations

Subtype Familial uveal melanoma Uveal melanoma in young individual Bilateral primary uveal melanoma Multifocal primary uveal melanoma Oculo(dermal) melanocytosis Phakomatosis pigmentovascularis Familial atypical mole and melanoma syndrome Neurofibromatosis type 1 Li-Fraumeni syndrome BAP1 Cancer predisposition syndrome Germline BRCA1 /BRCA2 mutation

Genetic Predisposition Uveal melanomas usually occur sporadically [31]. Rare occurrences of uveal melanoma manifesting features indicative of an inherited predisposition [61], such as familial uveal melanoma, uveal melanoma in young individuals, bilateral primary uveal melanoma, and multifocal primary uveal melanoma, have been reported [42]. Phenotypic associations of uveal melanoma include uveal melanoma in patients with oculo(dermal) melanocytosis (ODM), phakomatosis pigmentovascularis, familial atypical mole and melanoma (FAM-M) syndrome, neurofibromatosis type 1 (NF1), and Li-Fraumeni syndrome (Table 4.2).

 amilial Uveal Melanoma F Silcock first reported the occurrence of uveal melanoma in a mother and her two daughters [62]. Familial uveal melanoma is very rare, occurring in only 0.6% of all uveal melanoma patients [63]. A review of published kindreds with familial uveal melanoma reveals that involvement of many generations, typical of autosomal dominant inheritance, is uncommon [63]. In most reported families, only two relatives are affected (Fig.  4.4) [64]. Other features of a genetic predisposition, such as earlier age at

60

N. Singh et al. 88

86

68

45

19

53

42

40

14

9

32

7

30

14

Fig. 4.4  Familial uveal melanoma. The pedigree shows the affected son (proband, single arrow) and the father (double arrows). All other first-degree relatives of the proband had a normal eye examination. (Reprinted from Singh et al. [64]. With permission from Elsevier.)

diagnosis, bilateral involvement, multiple primary tumors, and phenotypic associations, are not present in familial uveal melanoma [63, 65]. Therefore, the possibility of two individuals in a given family developing uveal melanoma by chance alone (1  in 10 million) cannot be completely ignored [65]. Alternatively, BAP1 cancer predisposition syndrome with germline BAP1 mutation can be the cause of familial uveal melanoma [66]. Review of published studies suggests that about 22% of familial cases have detectable BAP1 mutations, suggesting that yet-undetected additional genes confer a risk for familial uveal melanoma [66].

 veal Melanoma in Young Individuals U Approximately 1% of all uveal melanomas occur in patients who are less than 20 years of age [65, 67]. Most of the so-called pediatric uveal melanomas occur around puberty and have a tendency to preferentially arise in the iris [65, 68]. Uveal melanomas can very rarely be present at birth [68, 69]. The Pediatric Choroidal and Ciliary Body Melanoma Study estimated the cumulative frequency of having a CCBM diagnosed increased by a mean of 0.8% per year of age

between ages 5 and 10 years and, after a 6-year transition from 11 to 16 years, by a mean of 8.8% per year between 17 and 24  years of age [70]. Unlike adult uveal melanoma, females are affected roughly twice as frequently as males. In general, the management of uveal melanoma in young adults is similar to that of uveal melanomas in adults, but the survival tends to be better in young adults than uveal melanoma in older adults. The Pediatric Choroidal and Ciliary Body Melanoma Study found that melanoma-related survival for children 2, “probable melanoma”. Rarely, however, melanomas are indistinguishable even from typical nevi if detected very early. Shields et  al. have devised the mnemonic “TFSOM-UHHD,” in which T stands for thickness over 2  mm; F, fluid; S, symptoms; O, orange pigment; M, margin within 3  mm of the optic disc; UH,

ultrasound hollow; H, halo absent; and D, drusen absent [4]. It is conventional practice to observe unidentified melanocytic tumors for months or years, delaying treatment until growth is documented. Such management should be undertaken only with the patient’s fully informed consent, which requires to the patient to know that the differential diagnosis includes melanoma and that any opportunities for preventing metastasis may be missed if treatment is delayed [5]. Biopsy of choroidal melanomas is useful for confirming the diagnosis and estimating the risk of metastasis, which is high in the presence of chromosome 3 loss (“monosomy 3”), a class 2 gene expression profile, and/or BAP1 loss (Fig. 6.10).

92

B. E. Damato and A. R. Afshar

Table 6.1  Clinical findings Risk factor Mushroom shape

Orange pigment

Large size

Enlargement

Subretinal fluid

Severity Absent Unsure Present Absent Unsure/trace Confluent Thickness 2 mm None Unsure Definite Absent Trace Definite

Score 0 1 2 0 1 2 And And And/or And/or

Diameter 4 DD

0 1 2 0 1 2 0 1 2

Total score

Metastasis

Fig. 6.10  Choroidal tumor biopsy. Fundus photograph showing an inferior, anterior melanocytic choroidal tumor in the right eye of an 84-year-old woman. B-scan ultrasonography showed the tumor to have a basal diameter of 9.7 mm and a thickness of 1.1 mm. Trans-vitreal biopsy by the second author (AA) showed the tumor to be a melanoma with chromosome 3 loss

Ocular metastases can develop in a patient with a known history of cancer, with or without evident metastases elsewhere, or they can be the presenting feature in a patient without any previous history of malignancy, as with lung cancer (Chap. 28) [6]. Choroidal metastases can be solitary or multiple and unilateral or bilateral and are usually located post-equatorially. Most are white or tan (Fig. 6.11a, b), but may appear pigmented with some fundus cameras (Fig.  6.11c, d). Tumor vessels are not usually visible. There may be hyper- and hypo-fluorescent stippling over part of the tumor (Fig. 6.11e, f). Metastases arising from cutaneous melanoma can be pigmented (Fig.  6.12). Carcinoid tumors and metastases from thyroid or kidney can be orange. Metastases tend to have a dome or plateau shape with indistinct margins. On OCT, choroidal metastases have a “lumpy-bumpy” surface (Fig.  6.13c). On ultrasonography, they have a medium-to-high internal acoustic reflectivity (Fig.  6.13d). Metastases only rarely have a mushroom shape, and when they do the internal acoustic reflectivity is equally high throughout

93

6  Uveal Melanoma: Differential Diagnosis

a

b

c

d

e

f

Fig. 6.11  Choroidal metastases in a 65-year-old woman with lung cancer. (a) Right fundus color photograph showing multiple amelanotic metastases; (b) left color fundus photograph, showing similar appearances; (c)

a

right Optos photograph and (d) left Optos photograph, suggesting that some of the lesions are pigmented; (e) right and (f) autofluorescence imaging showing hyperand hypo-autofluorescent stippling

b

Fig. 6.12  Pigmented choroidal metastases from cutaneous melanoma in a 51-year-old man. (a) Right fundus, (b) left fundus

94

the tumor (unlike melanoma). Serous retinal detachment is usually present. Biopsy with immunohistochemistry is useful for confirming the diagnosis and, in some cases,

B. E. Damato and A. R. Afshar

identifying the site of the primary malignancy. Some centers perform biopsy in the first instance, whereas others undertake this procedure only if systemic imaging is negative.

a

b

c

d

e

Fig. 6.13  Imaging of choroidal metastasis. (a) Color photograph showing an inferotemporal choroidal tumor in the right eye, (b) autofluorescence image hyper- and hypo-autofluorescent lesions, (c) OCT showing the tumor

to have a lumpy surface, (d) B-scan ultrasound and (e) A-scan ultrasound, showing the tumor to have medium internal reflectivity

6  Uveal Melanoma: Differential Diagnosis

Hemangioma Choroidal hemangiomas can be nodular (syn. “circumscribed”) or diffuse, the latter tending to occur in patients with the Sturge-Weber syndrome a

95

(Chap. 23) [7]. Hemangiomas have the same color as the normal choroid (Fig.  6.14a). Indocyanine green angiography shows early hyper-fluorescence and late “washout” of the dye (Fig. 6.14b). Autofluorescence imaging shows stippling over c

b

d

Fig. 6.14  Choroidal hemangioma. (a) Color photograph, showing an inferior choroidal hemangioma in the right eye; (b) early fluorescein and ICG angiograms showing hyperfluorescent stippling; (c) autofluorescence imaging

e

showing hyperfluorescence from retinal detachment; (d) OCT, showing dome-shaped lesion obscuring normal choroidal vasculature; (e) A-scan and (f) B-scan ultrasounds showing high internal acoustic reflectivity

96

B. E. Damato and A. R. Afshar

f

Uveal Effusion Uveal Effusion is usually multilobulated, because of tethering of the choroid to the sclera by vortex veins (Fig. 6.16a). These lesions have a smooth, featureless, gray, or brown surface and discrete margins, possibly with serous retinal detachment. On autofluorescence imaging they are dark (Fig.  6.16b). Ultrasonography shows the multilobulated and hollow nature of these tumors (Fig.  6.16c, d). On transillumination, these lesions transmit light. Uveal effusion can be associated with a thick sclera and/or nanophthalmos.

Fig. 6.14 (continued)

Suprachoroidal Hemorrhage Suprachoroidal hemorrhage can develop after injury or surgery, or spontaneously if the patient is on anticoagulant therapy. The lesion is dome-­ shaped and gray or brown, having a smooth surface, possibly with choroidal folds (Fig.  6.17). Spontaneous resorption tends to occur after several weeks.

Choroidal Osteoma

Fig. 6.15  Fibrosis over a choroidal hemangioma

the tumor with inferior hyperfluorescence from RPE changes caused by retinal detachment (Fig.  6.14c). On OCT they appear acoustically dark (Fig. 6.14d). On ultrasonography, hemangiomas have a high internal acoustic reflectivity (Fig. 6.14e, f). Serous retinal detachment is usually present and may become total. If neglected, these tumors can result in neovascular glaucoma. Choroidal hemangiomas may develop overlying fibrosis, which can be extensive (Fig. 6.15).

These tumors can be unilateral or bilateral (Chap. 29) [8]. They tend to be posterior, often involving the optic disc. “Active” tumors tend to be dome- or plateau-shaped and light brown, pink, yellow, or white, depending on whether there is any overlying RPE (Fig.  6.18b). The margins are discrete and characteristically “wavy.” After a period of growth, choroidal osteomas tend to undergo spontaneous regression, with RPE atrophy and loss of vision if the fovea is involved, resulting in a flat scar with pigment stippling (Fig. 6.18a). The RPE changes are also demonstrated by autofluorescence imaging (Fig. 6.18c, d), and fluorescein angiography (Fig.  6.18e, f), which can sometimes reveal a neovascular membrane. OCT is useful

97

6  Uveal Melanoma: Differential Diagnosis

a

b

c

d

Fig. 6.16  Uveal effusion. (a) Color photograph showing lobulated lesion, (b) autofluorescence image showing the lesion to be hypo-autofluorescent, (c) B-scan and (d) ultrasound biomicroscopy showing hollow nature of lesion

for showing areas of RPE atrophy and demonstrates characteristic horizontal lines in the tumor (Fig.  6.18g, h). On ultrasonography, choroidal osteomas have a highly reflective anterior surface, with orbital shadowing (Fig. 6.18i, j).

Sclero-choroidal Calcification

Fig. 6.17  Choroidal hemorrhage, with folds

These lesions can be unilateral or bilateral and tend to arise in the region of the oblique muscle insertions [9]. They are white or yellow with discrete, irregular margins, with a linear shape (Fig. 6.19). Ultrasonographic features are similar to those of choroidal osteoma.

B. E. Damato and A. R. Afshar

98

a

b

c

d

e

f

Fig. 6.18 Bilateral choroidal osteomas, which are regressed in the right eye and growing in the left eye. (a) Right color fundus image, showing flat area of RPE atrophy; (b) left color fundus image, showing pink mass; (c) right autofluorescence image showing extensive RPE atrophy around disc; (d) left autofluorescence image, showing preserved RPE and inferior RPE hyperfluores-

cence from retinal detachment; (e) right fluorescein angiogram, showing RPE atrophy; (f) left fluorescein angiogram, showing preserved RPE; (g) right OCT, showing flat atrophic area; (h) left OCT, showing persistent tumor with characteristic internal layering; (i) right B-scan and (j) left B-scan showing highly reflective lesions with orbital shadowing

6  Uveal Melanoma: Differential Diagnosis

99

g

h

i

j

Fig. 6.18 (continued)

a

b

Fig. 6.19  Sclero-choroidal calcification. (a) Right eye, (b) left eye

B. E. Damato and A. R. Afshar

100

a

b

Fig. 6.20  Vortex vein ampulla. (a) Engorged with blood, (b) collapsed

Vortex Vein Ampulla

Neurilemmoma

Prominent vortex vein ampulla is usually pink and located equatorially in any quadrant (Fig. 6.20) [10]. It tends to appear and disappear spontaneously every 30 seconds or so. It can also be made to disappear by pressing on the eye while performing ophthalmoscopy.

These tumors are usually located in the ciliary body, where they can appear pigmented if the overlying epithelium is intact (Fig. 6.22a) (Chap. 24) [12]. They can induce sentinel episcleral vessels. They transilluminate brightly. In the choroid, they are yellow/white, possibly with adjacent hard exudates (Fig. 6.22b). There can be a dilated feeder vessel.

Uveal Lymphoma This tumor is rare. It is usually unilateral (Chap. 27) [11]. On ophthalmoscopy, it is a similar color to the normal choroid, but normal choroidal vessels are obscured (Fig. 6.21a–d). Autofluorescence is normal (Fig. 6.21e, f). Indocyanine green angiography demonstrates obscuration of choroidal vessels (Fig. 6.21g, h). Optical coherence tomography shows the lumpy anterior tumor surface and obscuration of the choroidal vessels (Fig. 6.21i, j). On ultrasonography, uveal lymphoma has a low internal acoustic reflectivity and often extends extraocularly (Fig. 6.21k). Uveal lymphoma can involve ciliary body and can extend extraocularly to form a pink subconjunctival mass.

 ilateral Diffuse Uveal Melanocytic B Proliferation This is a rare, paraneoplastic phenomenon, characterized by bilateral uveal thickening, multiple pigmented tumors, hyper-autofluorescence, rapid, severe vision loss, serous retinal detachment, and cataract (Chap. 30). This condition can occur with an occult primary malignancy. The melanocytic proliferation can regress if the underlying malignancy is removed, but without improvement in vision [13].

6  Uveal Melanoma: Differential Diagnosis

101

a

b

c

d

e

f

Fig. 6.21  Uveal lymphoma in the left eye. (a) Right color image showing normal fundus; (b) left color image, showing obscuration of choroidal vasculature; (c) right Optos image showing normal choroidal vessels; (d) left Optos image, showing obscured choroidal vessels; (e) normal right autofluorescence image, (f) normal left autofluorescence image, (g) normal right fluorescein angio-

gram and ICG, (h) normal left fluorescein angiogram with partial obscuration of choroidal vasculature on ICG, (i) normal right OCT, (j) left OCT, showing lumpy surface and obscuration of choroidal blood vessels; (k) left B-scan, showing diffuse tumor with low internal reflectivity and small extraocular extension

B. E. Damato and A. R. Afshar

102

g

h

i

k

Fig. 6.21 (continued)

j

6  Uveal Melanoma: Differential Diagnosis

a

103

b

Fig. 6.22  Neurilemmoma. (a) Ciliary body neurilemmoma, which appears brown because of overlying RPE, (b) amelanotic choroidal neurilemmoma superior to the left optic disc, with inferior retinal detachment

a

b

Fig. 6.23  Eccentric disciforms. (a) Right eye, (b) left eye

Peripheral Exudative Chorioretinopathy (“Eccentric Disciform”)

Iris Tumors

These lesions arise in elderly patients and are usually located in the temporal periphery. They are pink or brown and multilobulated, with discrete margins, often associated with hemorrhages and exudates (Fig.  6.23). When regressed, they become gray or white and have sharp, irregular margins with adjacent RPE changes.

Iris nevi are usually small (Fig. 6.24). They can be nodular, multinodular (“tapioca”), diffuse, pigmented, or amelanotic. Although they can scatter pigment, they do not show growth around the angle. They can cause ectropion uveae, which is not necessarily a sign of malignancy.

Nevus

104

B. E. Damato and A. R. Afshar

a

b

c

d

Fig. 6.24  Iris nevus. (a) Slit-lamp color photograph, showing an inferotemporal nevus at the pupil margin; (b) gonio photograph, showing the nevus; (c) longitudinal

Melanoma Like nevi, iris melanomas can be nodular, “tapioca,” diffuse, dark brown, tan, yellow, white, and, rarely, red or pink. Nodular iris melanomas tend to be larger than nevi (i.e., more than 1 mm thick and more than 3 mm in basal diameter) (Fig. 6.25a–d) (Chap. 10). If diffuse, they can show annular spread around the angle, causing secondary glaucoma (Fig. 6.26a, b). There can also be pigment scatter. Peripheral iris tumors need to be distinguished from ciliary body melanomas invading the anterior chamber. Extraocular spread can also occur.

B-scan ultrasound and (d) transverse B-scan ultrasound, showing the nodule

Iris melanomas must be distinguished from ciliary body melanomas that have invaded the anterior chamber (Fig.  6.9). High-frequency ultrasonography is useful for this purpose (Fig. 6.9c, d).

Melanocytoma Iris melanocytomas tend to be deeply pigmented with a fluffy, granular surface, often with pigment scatter, which can cause glaucoma (Fig. 6.27). Necrosis and malignant transformation can occur.

105

6  Uveal Melanoma: Differential Diagnosis

a

b

c

d

Fig. 6.25  Iris melanoma. (a) Color slit-lamp photograph, showing an inferonasal tumor; (b) gonio photograph, showing the tumor thickness; (c) longitudinal B-scan and (d) transverse B-scan, showing the localized nature of the tumor

a

b

Fig. 6.26  Diffuse iris melanoma. (a) Slit-lamp photograph showing the melanoma with diffuse margins and (b) gonio photograph showing circumferential spread in the angle

106

B. E. Damato and A. R. Afshar

a

b

c

d

Fig. 6.27  Iris melanocytoma. (a) Slit-lamp photograph showing a superior pigmented tumor, (b) gonio photograph showing extension to the angle, (c) gonio photo-

graph showing pigment deposits in the inferior angle, (d) slit-lamp appearance after iridocyclectomy, which was successfully performed to prevent pigmentary glaucoma

Metastasis Iris metastases can be solitary or multiple, and yellow, pink or white, possibly associated with hyphema or pseudohypopyon (Fig.  6.28) (Chap. 29).

Cysts Iris pigment epithelial cysts are brown with a smooth surface and can be mistaken for melanoma, especially if large (Fig.  6.29). Although not uveal, they are mentioned here because they can be mistaken for uveal melanoma.

Fig. 6.28 Iris metastasis. Color photograph showing multilobulated, amelanotic tumor with a hyphema

107

6  Uveal Melanoma: Differential Diagnosis

a

b

c

d

Fig. 6.29  Irido-ciliary cyst. (a) Slit-lamp appearance, showing the iris pushed forward by the underlying cyst; (b) gonio photograph, showing narrowing of the angle; (c)

longitudinal B-scan and (d) transverse B-scan, showing the cystic nature of the lesion

Juvenile Xanthogranuloma These iris tumors tend to occur in young patients [14]. They are vascular and can be pink or light brown (Fig. 6.30). There can be recurrent spontaneous hyphema.

Iris Freckle Iris freckles are thin, lacy, pigmented lesions that are located on the iris surface and which do not disturb the normal architecture (Fig. 6.31). Fig. 6.30  Juvenile xanthogranuloma. Slit-lamp photograph showing a pink/orange iris tumor

108

B. E. Damato and A. R. Afshar

been a highly lethal monosomy 3/class 2 melanoma – especially they were not informed of the risk and/or if not all the appropriate tests were performed. Informed consent for nontreatment should be documented as meticulously as informed consent for treatment.

References 1. Shields JA, Shields CL, Eagle RC Jr. Melanocytoma (hyperpigmented magnocellular nevus) of the uveal tract: the 34th G.  Victor Simpson lecture. Retina. 2007;27(6):730–9. 2. Damato BE, Foulds WS.  Tumour-associated retinal pigment epitheliopathy. Eye. 1990;4(Pt 2):382–7. Fig. 6.31  Iris freckles. Slit-lamp photograph showing 3. Damato BE.  Tumour fluorescence and tumour-­ pigmented lesions, without disturbance of the iris associated fluorescence of choroidal melanomas. Eye. architecture 1992;6(Pt 6):587–93. 4. Shields CL, Kels JG, Shields JA.  Melanoma of the eye: revealing hidden secrets, one at a time. Clin Dermatol. 2015;33(2):183–96. Conclusions 5. Callejo SA, Dopierala J, Coupland SE, et al. Sudden growth of a choroidal melanoma and multiplex A well-informed clinician should be able to diagligation-­dependent probe amplification findings sugnose almost all uveal tumors by slit-lamp examigesting late transformation to monosomy 3 type. Arch Ophthalmol. 2011;129(7):958–60. nation or ophthalmoscopy, with the help of imaging such as OCT, autofluorescence, and 6. Konstantinidis L, Damato B.  Intraocular metastases  – a review. Asia Pac J Ophthalmol (Phila). ultrasonography, resorting to biopsy in a minority 2017;6(2):208–14. of cases. 7. Singh AD, Kaiser PK, Sears JE. Choroidal hemangioma. Ophthalmol Clin N Am. 2005;18(1):151–61, ix. The author has developed an online atlas that organizes intraocular tumors according to their 8. Shields CL, Shields JA, Augsburger JJ.  Choroidal osteoma. Surv Ophthalmol. 1988;33(1):17–27. location and color (www.oculonco.com); how- 9. Shields CL, Hasanreisoglu M, Saktanasate J, ever, care must be taken when interpreting color et  al. Sclerochoroidal calcification: clinical features, outcomes, and relationship with hypercalcephotographs because color of the tumor depends mia and parathyroid adenoma in 179 eyes. Retina. greatly on the camera that is used. 2015;35(3):547–54. If the diagnostic process is inconclusive, pos- 10. Gunduz K, Shields CL, Shields JA. Varix of the vortex sibly because biopsy is not possible, then the vein ampulla simulating choroidal melanoma: report of four cases. Retina. 1998;18(4):343–7. authors advise on further management according to what is likely to be the “least worst outcome” 11. Aronow ME, Portell CA, Sweetenham JW, et  al. Uveal lymphoma: clinical features, diagnostic studies, if the selected course of action turns out to be treatment selection, and outcomes. Ophthalmology. wrong. For example, when deciding between 2014;121(1):334–41. observation and treatment of a pigmented choroi- 12. Damato B, Damato EM, Konstantinidis L, et  al. Choroidal schwannoma: a case series of five patients. dal tumor of uncertain malignancy, the small but Br J Ophthalmol. 2014;98(8):1096–100. serious risk of metastatic disease needs to be 13. Sen J, Clewes AR, Quah SA, et  al. Presymptomatic weighed against a high risk of visual loss. diagnosis of bronchogenic carcinoma associated with bilateral diffuse uveal melanocytic proliferation. Clin Patients should be given enough knowledge to Exp Ophthalmol. 2006;34(2):156–8. be able to participate rationally in the decision-­ 14. Samara WA, Khoo CT, Say EA, et al. Juvenile xanmaking process. Quite understandably, they tend thogranuloma involving the eye and ocular adnexa: to get rather upset when, after years of observatumor control, visual outcomes, and globe salvage in 30 patients. Ophthalmology. 2015;122(10):2130–8. tion, their “suspicious nevus” turns out to have

7

Uveal Melanoma: Histopathologic Features Tero T. Kivelä

Introduction Uveal melanomas develop from melanocytes that reside within the stroma of the choroid, ciliary body, and iris. No basement membrane needs to be breached when the tumor develops [1]. The key histopathological features of uveal melanomas detailed in the present chapter are summarized in Box 7.1. Box 7.1. Histopathological Features of Uveal Melanoma

• Tumors show nodular, collar-stud, diffuse, and/or retinoinvasive growth. • Extraocular spread tends to occur through preexisting channels. • Uveal melanomas cause many secondary effects, such as glaucoma. • Melanoma cells can be spindle or epithelioid.

T. T. Kivelä (*) Ocular Oncology and Ophthalmic Pathology, Helsinki University Hospital, Helsinki, Finland Department of Ophthalmology, University of Helsinki, Helsinki, Finland e-mail: [email protected]

• Extravascular matrix patterns and microvascular density are important. • Infiltrating macrophages and lymphocytes are prognostically significant.

Growth Patterns Dome Shape The growing tumor is first flat but then develops a discoid, or dome, shape. Typically, it assumes dimensions in which its height is approximately half of its diameter (Fig. 7.1a). Most have a circular or oval contour, with irregular shape or multinodularity suggesting clonal evolution and variable growth rates.

Collar-Stud Shape A choroidal melanoma distends the overlying Bruch’s membrane, which eventually ruptures so that part of the tumor squeezes through the break to assume a characteristic mushroom, or collar-­ stud, shape (Fig. 7.1b). The relative sizes of the base and the collar-stud part vary widely, depending on how soon and to what extent Bruch’s membrane breaks. The collar stud is rich in dilated, sinusoidal capillaries because of a tourniquet effect of Bruch’s membrane, which strangu-

© Springer Nature Switzerland AG 2019 B. E. Damato, A. D. Singh (eds.), Clinical Ophthalmic Oncology, https://doi.org/10.1007/978-3-030-17879-6_7

109

110

a

T. T. Kivelä

b

c

Fig. 7.1  Uveal melanomas are dome-shaped (a) until they break through Bruch’s membrane and develop a collar-­button shape (b). Eosinophilic proteinaceous mate-

rial around the tumor is subretinal fluid. Diffuse uveal melanomas remain flat and grow laterally rather than vertically (c). Note extrascleral extension of the diffuse tumor

lates the vessels. The tumor can also erode through the retina, which leads to vitreous hemorrhage. Alternatively, the tumor detaches Bruch’s membrane along the rim of the optic disk or ora serrata and bulges in front of the disk or behind the lens, respectively.

than 7  mm in thickness and without breaking Bruch’s membrane (Fig.  7.1c). Diffuse tumors have a particular tendency to grow transsclerally.

Diffuse Growth Pattern

Ring Melanoma One subtype of the diffuse growth pattern is ring melanoma, which grows circumferentially around the ciliary body or chamber angle, often without extending into the choroid [3–5].

A third growth pattern is diffuse [2], in which the tumor, by definition, involves more than a quadrant of the fundus without growing to be more

Retinoinvasive Melanoma A variant of diffuse melanoma, retinoinvasive melanoma, disseminates to the vitreous and the

7  Uveal Melanoma: Histopathologic Features

retinal surface, eventually invading nonadjacent retina and the optic nerve [6, 7].

Local Invasion Uveal melanomas have indistinct borders. Tumor cells are often seen adjacent to visible tumor margin where the choroid is not thickened. This is particularly typical of the diffuse tumor type.

111

Optic Nerve Juxtapapillary melanomas can invade the optic disk, but the retrobulbar nerve is usually not invaded unless it has sustained prior glaucomatous damage and the intraocular pressure is high [14]. The rare, retinoinvasive uveal melanomas also invade the optic nerve [6].

Secondary Intraocular Effects

Sclera

Retinal Pigment Epitheliopathy

The sclera is resistant to invasion, and uveal melanomas gain access to it mainly along trans-­ scleral channels for ciliary nerves, ciliary arteries, vortex veins, and aqueous veins. Consequently, extrascleral extension typically occurs posteriorly adjacent to the optic nerve, equatorially adjacent to a vortex vein, and anteriorly adjacent to the limbus [8, 9]. Typically, the intrascleral part of the tumor is narrow, making the tumor dumbbell-shaped, and much narrower than the extrascleral part, which usually can be resected and the remaining tumor irradiated without full-­ thickness scleral resection [10]. Despite preoperative ultrasonographic assessment, extrascleral extension of uveal melanoma may go unnoticed and is detected only on histopathologic evaluation [11]. Extrascleral extension is not infrequent after thermotherapy [12] (Chap. 15) and can exceptionally occur even after brachytherapy [13].

Uveal melanomas induce proliferation and metaplasia of retinal pigment epithelial cells and atrophy or cystic degeneration of the retina over their surface, giving rise to subretinal membranes that consist of multilayered basement membrane material.

Retina The retina and optic nerve are likewise resistant to invasion by uveal melanoma. The retina overlying a tumor that has broken through Bruch’s membrane is typically eroded without invasion of the adjacent retina. Melanomas that have broken through the retina can seed tumor cells to the vitreous. These cells are mostly necrotic and accompanied by abundant melanophages and other macrophages.

Retinal Detachment Uveal melanomas are moderately to richly vascularized tumors. Retinal detachment develops as a result of leakage from tumor vessels, together with compromised function of the retinal pigment epithelial pump overlying the tumor [15]. On light microscopy, eosinophilic proteinaceous subretinal fluid is seen in most eyes, either adjacent to the tumor or displaced to the retinal periphery (Fig. 7.1a, b).

Secondary Glaucoma Uveal melanoma can induce a secondary glaucoma by several mechanisms. Closed-angle glaucoma can be caused by (1) compression or direct infiltration of the trabecular meshwork by a ciliary body tumor or melanoma of the anterior chamber angle [3, 4], (2) vitreous hemorrhage, or (3) a sudden increase in exudative retinal detachment after irradiation. Neovascular glaucoma can develop in an untreated eye with a large tumor and retinal

112

detachment but is more common after irradiation, especially if the tumor is located in the ciliary body or if ischemic radiation optic neuropathy has occurred. Melanomalytic glaucoma is caused by blockage of the trabecular meshwork by disseminated necrotic tumor cells and melanophages [3, 16].

Histopathologic Survival Predictors Primary Uveal Melanoma Cell Type Cell type was the first histopathologic feature of uveal melanoma to be associated with survival (Chap. 18) (Fig. 7.2). Originally, six histopathologic types of uveal melanoma were described, but the categories were later simplified into three (Table 7.1) [17–19]. It also correlates with the most later prognostic factors, including monosomy 3 and nuclear BAP1 staining loss (Fig. 7.2) [20]. Spindle cell melanoma is composed of fusiform cells orientated in bundles and whorls

Fig. 7.2 Associations between cell type and other major histopathologic and immunohistochemical features of uveal melanoma. The thicker the connecting line, the stronger the association. MLN mean of the ten largest nucleoli, MVD microvascular density, and LBD largest basal tumor diameter

T. T. Kivelä

(Fig. 7.3a). Spindle cells have variable amounts of fibrillar cytoplasm, and their borders are difficult to distinguish because the cells adhere to each other. Originally, spindle cell melanomas were divided into spindle A and B types. The former have narrow, oval nuclei and inconspicuous nucleoli, and the latter contain larger, round nuclei and more conspicuous nucleoli (Fig. 7.3a). Most spindle A cell tumors are currently classified as spindle cell nevi (Table 7.1) [18]. Epithelioid cell melanoma is composed of polyhedral cells, which are usually but not always large and which morphologically resemble epithelial cells (Fig.  7.3b). Their abundant cytoplasm is eosinophilic, and they characteristically crack apart from their neighbors during tissue processing, resulting in a non-cohesive appearance. The nucleoli are large and prominent. Mixed cell type melanoma contains variable proportions of spindle and epithelioid cells. Opinion is divided as to what proportion of epithelioid cells distinguishes spindle from mixed and mixed from epithelioid cell melanomas [18]. Increasingly, even a single well-defined epithelioid cell precludes classification as a spindle cell

113

7  Uveal Melanoma: Histopathologic Features Table 7.1  Classification of uveal melanoma based upon the cell type Histopathologic characteristics Original Callender [17] Spindle A melanomaa Spindle B melanoma Mixed cell melanoma Epithelioid cell melanoma Fascicular melanoma Necrotic melanoma

AFIP [18, 19] Spindle cell nevusa Spindle cell melanomaa Mixed cell melanoma Epithelioid cell melanoma

Cell shape and appearance Fusiform cohesive cells

Cell borders Indistinct

Fusiform cohesive cells

Indistinct

Nucleus Narrow, oval Plump

Nucleolus Inconspicuous Conspicuous

Mixed population of spindle and epithelioid cells. Fusiform cohesive cells mixed with (at least single) non-cohesive epithelioid cells Large polygonal cells, abundant Distinct Large, Large, eosinophilic cytoplasm round prominent Spindle cells arranged in fascicles Too extensive tumor necrosis to allow classification into other groups

AFIP Armed Forces Institute of Pathology, Washington, DC, United States a The majority of spindle A melanomas of Callender’s classification were reclassified as spindle cell nevi and a minority to spindle cell melanoma in the AFIP classification

a

b

Fig. 7.3  Main cell types of uveal melanoma. A spindle cell melanoma with mostly B type cells with round nucleoli intermixed with occasional slender A type cells with oval nuclei. Both are fusiform, with indistinct borders and arranged in bundles (a). A mixed cell melanoma with a population of polyhedral, non-cohesive, eosinophilic epithelioid cells, which have prominent nucleoli (b)

melanoma because the tumor is likely to harbor additional epithelioid cells elsewhere. Significant tumor necrosis is uncommon, but rarely, the tumor is too necrotic to be classified by cell type. These necrotic melanomas have a prognosis comparable to that of tumors with epithelioid cells. Widespread necrosis will cause a secondary inflammatory reaction.

Pigmentation Pigmentation of uveal melanomas ranges from heavy to amelanotic, and many tumors show regional variations in pigmentation. The grade of pigmentation can be semiquantitatively assessed by comparing unstained sections on white paper. A high grade of pigmentation is associated with a high risk of metastasis. Nucleolar Size Nucleolar size of uveal melanoma cells is typically large, so that nucleoli are conspicuous in hematoxylin–eosin-stained sections although better appreciated with special stains, especially the silver stain (Fig.  7.4) [21]. Melanin is first bleached with potassium permanganate and oxalic acid. Large nucleoli are associated with a high risk of metastasis [22–24]. The recommended method is to calculate the mean of the longest diameters of the ten largest nucleoli found along a 5-mm central strip of the tumor. A

114

a

b

Fig. 7.4  A mixed cell type uveal melanoma with prominent nucleoli (a), which are most distinct in a silver-­ stained section (b)

filar micrometer or digital photography can be used for the measuring. The exact measurements will depend on the staining and equipment used [24], and each laboratory must establish its own reference values.

Mitotic Figures Most uveal melanomas are slow growing, and mitotic figures are consequently usually few in number. The recommended method is to count the number of mitotic figures in 40 high-power fields. Higher numbers are associated with a higher chance of metastasis.  xtravascular Matrix Patterns E The stroma of uveal melanomas is scanty. The extravascular matrix can be highlighted with several stains, of which periodic acid–Schiff stain without counterstain is most popular (Fig.  7.5) [25]. The stroma of uveal melanomas is otherwise scanty. Melanin is first bleached with potas-

T. T. Kivelä

sium permanganate and oxalic acid, and, after staining, the slides are evaluated under a ­dark-­green filter to enhance the visibility of the matrix, which is purple (Fig. 7.5). Nine matrix patterns are distinguished, which often occur in combination in any given tumor (Table 7.2) [25–27]. The most widely evaluated patterns are closed loops and networks, the latter consisting of at least three closed loops that are linked back to back (Fig. 7.5). These two patterns are grouped with arcs and arcs with branching into a family of curved patterns. A second family of straight patterns consists of straight, parallel, and parallel with cross-linking. The remaining two patterns, silent and normal, may be seen in uveal melanomas but are more typical of uveal nevi [28]. Several extravascular matrix patterns are associated with a higher than average chance of metastasis. The association is strongest for loops, and in particular for networks [23, 25, 29]. In evaluating these patterns, loops of any size are accepted. They typically consist of thin matrix strands that separate nests of tumor cells, which range in number from fewer than 10 to several hundreds per nest (Fig. 7.5).

Cytological Diagnosis When diagnostic fine needle aspiration biopsy is considered, a stepwise diagnostic approach is useful [30]. Cytologists should first look for spindle cell morphology, observed in most uveal melanomas (Fig.  7.6a) [31], because the vast majority of uveal metastases are carcinomas [32] consisting of cohesive clusters of cells with epithelial cytomorphology [33]. Cytoplasmic melanin pigment is the second most important diagnostic feature, observed in 80% of uveal melanomas [31]. Melanin has a finely granular cytoplasmic distribution that is often focal and can be inconspicuous (Fig.  7.6b). Immuno­ histochemistry for markers such as HMB45 and MelanA is positive in melanoma (Fig. 7.6c) [30, 34]. Immunohistochemical Prognostic Indicators Immunohistochemical indicators of survival prognosis have been evaluated in uveal melanoma, sometimes with conflicting results. With

7  Uveal Melanoma: Histopathologic Features

115

Fig. 7.5  Extravascular matrix patterns are purple in specimens stained with periodic acid–Schiff stain (left column) and appear dark in red-free photographs (right column). The arcs with branching (top row), loops (middle

row), and network patterns (bottom row) all belong to the family of curved patterns. Note that the thickness of the patterns varies, and some are pencil-thin (middle row)

the exception of cytokeratins, the prognostic significance of each immunohistochemical marker discussed in this section has been confirmed by independent studies and by multivariate analyses, adjusted for the most commonly known clinical and histopathologic prognostic factors.

Immunohistochemical studies of uveal melanoma are complicated by the presence of melanin. This can be overcome by bleaching the sections with hydrogen peroxide and sodium dihydrogen phosphate following immunostaining [35]. This sequence excludes the possibility that antigenicity would be modified by the bleaching.

116

T. T. Kivelä

Table 7.2  Classification of extravascular matrix patterns [25–27] Pattern Normal

Histopathologic characteristics Tumor grows around normal choroidal vessels without compressing them in a portion of the tumor beneath Bruch’s membrane Silent No matrix patterns are demonstrated Straight Straight matrix strands, arranged in random orientation without dichotomous branching Parallel Straight matrix strands, arranged in parallel without dichotomous branching Parallel with cross-link Straight matrix strands, arranged in parallel and cross-linked to each other in a fashion reminiscent of rail tracks Arcs Curved matrix strands not attached to others forming incomplete loops Arcs with branching Curved matrix strands with dichotomous branching forming incomplete loops Loops At least one completely closed loop of matrix encircling a nest of tumor cells Networks At least three back-to-back closed loops

a

Group Nevus

Linear

Curved

b

c

Fig. 7.6  Presence of spindle shaped cells (a) and melanin pigment (b) are helpful cytologic diagnostic features of uveal melanoma. Immunohistochemical stains for melanoma markers, such as HMB45 (c), help to confirm the

identity of amelanotic melanoma cells if necessary. (Reprinted from Biscotti and Singh [30]. Copyright © 2011, Karger Publishers.)

117

7  Uveal Melanoma: Histopathologic Features

BAP1 is the gene that encodes BRCA1-­ Associated Protein 1. Pathogenic truncating and splice site variants in this gene lead to loss of nuclear immunostaining (Fig.  7.7) and have ­consistently been associated with a high risk of metastasis from uveal melanoma (Chap. 18). Uveal melanomas that are not uniformly labeled with antibodies to BAP1 protein are likewise associated with a shorter survival than melanomas showing uniform immunopositivity for this protein [36–38]. Although some studies have suggested nearly perfect correlation between the presence of BAP1 mutation and the absence of BAP1 staining [36], the staining pattern is sometimes difficult to interpret unequivocally, particularly in archival paraffin-embedded tissue, and pathogenic missense variants can abolish the function of BAP1 protein without altering its nuclear localization. Mutational analysis of BAP1 gene may then be necessary [37]. Microvascular density in uveal melanomas is best assessed using markers of vascular endothelium, which highlight the blood vessels (Fig. 7.8). The most commonly used antigen is the CD34 epitope, but comparable results can be obtained using antibodies to the CD31 epitope and factor VIII-related antigen and some lectins, such as the Ulex europaeus I agglutinin [39–41]. The recom-

a

a

b

Fig. 7.8  A mixed cell type uveal melanoma (a) with a high microvascular density (100 immunopositive elements/0.3 mm2) as revealed with antibodies to the CD34 epitope of endothelial cells (b)

b

Fig. 7.7  An epithelioid cell type uveal melanoma (a) with retained nuclear immunostaining for BAP1 protein in most of the tumor cells, suggesting a low risk of metastasis; with nuclear hematoxylin counterstain (b)

118

mended method is to count immunopositive elements under 40× magnification from a 0.3-mm2 area of tumor that is most densely vascularized when screened with 10× magnification. A high number of immunopositive elements per tumor cross-sectional area—called microvascular density (MVD)—are associated with an increased risk of metastasis [39–41]. Tumor-infiltrating macrophages vary in number. The macrophages can be identified by immunohistochemistry, especially using antibodies to the CD68 epitope (Fig.  7.9). Morphologically, the immunopositive cells vary in shape from dendritic to round [42]. Because dendritic cells are particularly difficult to count in histopathologic sections, the number of macrophages can be semiquantitatively assessed against standard photographs [42]. A high number of immunopositive cells is associated with an increased risk of

a

T. T. Kivelä

metastasis [42]. A similar observation has been made as regards infiltrating lymphocytes in uveal melanomas [43–45]. Cell proliferation antigens in cycling tumor cells can be identified with several antibodies, of which those recognizing proliferating cell nuclear antigen (PCNA) and Ki-67 antigen have been most widely used to evaluate uveal melanoma [47]. Larger numbers of cells with immunopositive nuclei are associated with a higher risk of metastasis [23, 45, 46]. Human leukocyte antigens (HLA) show decreased expression with progression in many cancers. This is thought to allow escape from immune surveillance. Low HLA class I expression in uveal melanoma is paradoxically associated with better survival than normal HLA class I expression [47]. This suggests that natural killer (NK) instead of T cells may be germane to immune surveillance of patients with uveal melanoma. Cytokeratins are intermediate filaments that are present in tumor cell populations of many uveal melanomas. Immunohistochemistry shows positivity for simple epithelial cytokeratins 8 and 18 and vimentin [48]. Tumor cells that coexpress cytokeratins and vimentin are more invasive in culture than tumor cells expressing only vimentin [49]. It has not been conclusively reported that the presence of cytokeratins significantly alters the risk of metastasis, but their presence has differential diagnostic implications.

Irradiated Uveal Melanoma b

Fig. 7.9  Tumor-infiltrating melanophages (a) in a spindle cell uveal melanoma can be identified in a hematoxylin–eosin-stained specimen, and antibodies to the CD68 epitope reveal additional macrophages (b)

Rather than being sampled systematically, irradiated uveal melanomas have been examined histologically in eyes enucleated because of tumor regrowth or complications resulting in a blind, painful eye or if the eye was removed at autopsy. Necrosis and microvascular density are different after secondary enucleation than after primary enucleation. In general, after secondary enucleation, melanomas show more necrosis and a lower microvascular density [50]. By matched-­ pair analysis, extravascular matrix loops and networks also tend to be less frequent after irradiation. The number of macrophages in non-­ necrotic tumor areas is not consistently higher in irradiated than in nonirradiated tumors [50].

7  Uveal Melanoma: Histopathologic Features

Melanoma cell type is more likely to include epithelioid cells in eyes enucleated after irradiation than eyes treated by primary ­ ­enucleation, which is surprising because epithelioid cells are believed to be more radiosensitive than spindle cells. Viable melanoma cells are present in many irradiated tumors. These cells are morphologically indistinguishable from those of untreated uveal melanomas, but seem to be dormant and nonproliferating [51]. Vascular and degenerative changes are also typical of irradiated tumors. These include obliteration of blood vessels and balloon cells. Local recurrence pattern after plaque brachytherapy is divided in marginal/horizontal, vertical/diffuse (increase in thickness with or without marginal recurrence), nonadjacent, and extrascleral that correlate with clinical patterns of recurrence [52, 53].

Metastatic Uveal Melanoma Metastases from uveal melanoma usually (but not always) reflect the cell type of the primary tumor. The metastases are less pigmented and have more epithelioid cells and a higher microvascular density than the primary tumor [27, 54]. The microvascular density of metastases may be associated with subsequent mortality (Chap. 18).

References 1. Folberg R. Tumor progression in ocular melanomas. J Investig Dermatol. 1993;100(3):326S–31S. 2. Reese AB, Howard GM. Flat uveal melanomas. Am J Ophthalmol. 1967;64(6):1021–8. 3. Stadigh A, Puska P, Vesti E, et al. Ring melanoma of the anterior chamber angle as a mimicker of pigmentary glaucoma. Surv Ophthalmol. 2017;62(5):670–6. 4. Demirci H, Shields CL, Shields JA, et al. Ring melanoma of the ciliary body. Retina. 2002;22(6):698–706. 5. Aziz HA, Modi YS, Plesec TP, et al. Amelanotic irido-­ ciliary ring melanoma: a clinicopathological correlation. Ocul Oncol Pathol. 2016;2(3):153–5. 6. Kivelä T, Summanen P.  Retinoinvasive malig nant melanoma of the uvea. Br J Ophthalmol. 1997;81(8):691–7. 7. Gorham JP, Szalai E, Wells JR, et al. Retinoinvasive uveal melanoma: report of 2 cases and review of the literature. Ocul Oncol Pathol. 2017;3(4):292–5.

119 8. Coupland SE, Campbell I, Damato B. Routes of extraocular extension of uveal melanoma. Ophthalmology. 2008;115(10):1778–85. 9. van Beek JGM, Koopmans AE, Vaarwater J, et  al. The prognostic value of extraocular extension in relation to monosomy 3 and gain of chromosome 8q in uveal melanoma. Invest Opthalmol Vis Sci. 2014;55(3):1284. 10. Muen WJ, Damato BE.  Uveal malignant melanoma with extrascleral extension, treated with plaque radiotherapy. Eye. 2006;21(2):307–8. 11. Burris CKH, Papastefanou VP, Thaung C, et  al. Detection of extrascleral extension in uveal melanoma with histopathological correlation. Orbit. 2018;37(4):287–92. 12. Singh AD, Eagle RC Jr, Shields CL, et  al. Clinicopathologic reports, case reports, and small case series: enucleation following transpupillary thermotherapy of choroidal melanoma: clinicopathologic correlations. Arch Ophthalmol. 2003;121(3):397–400. 13. Ayres B, Elner V, Demirci H.  Massive extraocular extension of choroidal melanoma after brachytherapy. Ophthalmology. 2017;124(10):1503. 14. Spencer WH.  Optic nerve extension of intraocular neoplasms. Am J Ophthalmol. 1975;80(3):465–71. 15. Kivelä T, Eskelin S, Mäkitie T, et al. Exudative retinal detachment from malignant uveal melanoma: predictors and prognostic significance. Invest Ophthalmol Vis Sci. 2001;42(9):2085–93. 16. McMenamin PG, Lee WR.  Ultrastructural pathol ogy of melanomalytic glaucoma. Br J Ophthalmol. 1986;70(12):895–906. 17. Callender GR. Malignant melanocytic tumors of the eye. A study of histologic types in 111 cases. Trans Am Acad Ophthalmol Otolaryngol. 1931;36:131–40. 18. McLean IW, Zimmerman LE, Evans RM. Reappraisal of Callender’s spindle a type of malignant melanoma of choroid and ciliary body. Am J Ophthalmol. 1978;86(4):557–64. 19. McLean IW, Foster WD, Zimmerman LE, et  al. Modifications of Callender’s classification of uveal melanoma at the Armed Forces Institute of Pathology. Am J Ophthalmol. 1983;96(4):502–9. 20. Shah AA, Bourne TD, Murali R.  BAP1 protein loss by immunohistochemistry: a potentially useful tool for prognostic prediction in patients with uveal melanoma. Pathology. 2013;45(7):651–6. 21. Moshari A, McLean IW.  Uveal melanoma: mean of the longest nucleoli measured on silver-stained sections. Invest Ophthalmol Vis Sci. 2001;42(6):1160–3. 22. McLean IW, Keefe KS, Burnier MN.  Uveal melanoma. Comparison of the prognostic value of fibrovascular loops, mean of the ten largest nucleoli, cell type, and tumor size. Ophthalmology. 1997;104(5):777–80. 23. Seregard S, Spångberg B, Juul C, et  al. Prognostic accuracy of the mean of the largest nucleoli, vascular patterns, and PC-10 in posterior uveal melanoma. Ophthalmology. 1998;105(3):485–91. 24. Al-Jamal RT, Mäkitie T, Kivelä T. Nucleolar diameter and microvascular factors as independent ­predictors of mortality from malignant melanoma of the cho-

120 roid and ciliary body. Invest Opthalmol Vis Sci. 2381;44(6):2003. 25. Folberg R, Rummelt V, Parys-Van Ginderdeuren R, et al. The prognostic value of tumor blood vessel morphology in primary uveal melanoma. Ophthalmology. 1993;100(9):1389–98. 26. Folberg R, Pe’er J, Gruman LM, et  al. The morphologic characteristics of tumor blood vessels as a marker of tumor progression in primary human uveal melanoma: a matched case-control study. Hum Pathol. 1992;23(11):1298–305. 27. Toivonen P, Mäkitie T, Kujala E, et al. Microcirculation and tumor-infiltrating macrophages in choroidal and ciliary body melanoma and corresponding metastases. Invest Opthalmol Vis Sci. 2004;45(1):1. 28. Rummelt V, Folberg R, Rummelt C, et  al. Microcirculation architecture of melanocytic nevi and malignant melanomas of the ciliary body and choroid. Ophthalmology. 1994;101(4):718–27. 29. Mäkitie T, Summanen P, Tarkkanen A, et  al. Microvascular loops and networks as prognostic indicators in choroidal and ciliary body melanomas. JNCI J Natl Cancer Inst. 1999;91(4):359–67. 30. Biscotti CV, Singh AD. Uveal melanoma: diagnostic features. Monogr Clin Cytol. 2012;21:44–54. 31. Medina CA, Biscotti CV, Singh N, et al. Diagnostic cytologic features of uveal melanoma. Ophthalmology. 2015;122(8):1580–4. 32. Shields CL, Shields JA, Gross NE, et  al. Survey of 520 eyes with uveal metastases. Ophthalmology. 1997;104(8):1265–76. 33. Bellerive C, Biscotti CV, Singh AD. Fine needle aspiration biopsy for suspected uveal metastases. Can J Ophthalmol. 2019. [epub ahead of print]. 34. Faulkner-Jones BE, Foster WJ, Harbour JW, et  al. Fine needle aspiration biopsy with adjunct immunohistochemistry in intraocular tumor management. Acta Cytol. 2005;49(3):297–308. 35. Kivelä T. Immunohistochemical staining followed by bleaching of melanin: a practical method for ophthalmic pathology. Br J Biomed Sci. 1995;52(4):325–6. 36. Koopmans AE, Verdijk RM, Brouwer RWW, et  al. Clinical significance of immunohistochemistry for detection of BAP1 mutations in uveal melanoma. Mod Pathol. 2014;27(10):1321–30. 37. van de Nes JA, Nelles J, Kreis S, et  al. Comparing the prognostic value of BAP1 mutation pattern, chromosome 3 status, and BAP1 immunohistochemistry in uveal melanoma. Am J Surg Pathol. 2016;40(6):796–805. 38. Herwig-Carl MC, Sharma A, Moulin A, et al. BAP1 immunostaining in uveal melanoma: potentials and pitfalls. Ocul Oncol Pathol. 2018;4(5):297. 39. Foss AJ, Alexander RA, Jefferies LW, et  al. Microvessel count predicts survival in uveal melanoma. Cancer Res. 1996;56(13):2900–3. 40. Mäkitie T, Summanen P, Tarkkanen A, et  al. Microvascular density in predicting survival of

T. T. Kivelä patients with choroidal and ciliary body melanoma. Invest Ophthalmol Vis Sci. 1999;40(11):2471–80. 41. Chen X, Maniotis AJ, Majumdar D, et  al. Uveal melanoma cell staining for CD34 and assessment of tumor vascularity. Invest Ophthalmol Vis Sci. 2002;43(8):2533–9. 42. Mäkitie T, Summanen P, Tarkkanen A, Kivelä T. Tumor-infiltrating macrophages (CD68+ cells) and prognosis in malignant uveal melanoma. Invest Ophthalmol Vis Sci. 2001;42(7):1414–21. 43. De Cruz POL, Specht CS, McLean IW. Lymphocytic infiltration in uveal malignant melanoma. Cancer. 1990;65(1):112–5. 44. Bronkhorst IH, Vu TH, Jordanova ES, et al. Different subsets of tumor-infiltrating lymphocytes correlate with macrophage influx and monosomy 3  in uveal melanoma. Invest Ophthalmol Vis Sci. 2012;53(9):5370–8. 45. Staibano S, Mascolo M, Tranfa F, et al. Tumor infiltrating lymphocytes in uveal melanoma: a link with clinical behavior? Int J Immunopathol Pharmacol. 2006;19(1):171–9. 46. Al-Jamal RT, Kivelä T.  KI-67 Immunopositivity in choroidal and ciliary body melanoma with respect to nucleolar diameter and other prognostic factors. Curr Eye Res. 2006;31(1):57–67. 47. Blom DJ, Luyten GP, Mooy C, et al. Human leukocyte antigen class I expression. Marker of poor prognosis in uveal melanoma. Invest Ophthalmol Vis Sci. 1997;38(9):1865–72. 48. Fuchs U, Kivelä T, Summanen P, et al. An immunohistochemical and prognostic analysis of cytokeratin expression in malignant uveal melanoma. Am J Pathol. 1992;141(1):169–81. 49. Hendrix MJ, Seftor EA, Seftor RE, et  al. Biologic determinants of uveal melanoma metastatic phenotype: role of intermediate filaments as predictive markers. Lab Investig. 1998;78(2):153–63. 50. Fuchs U, Kivelä T, Tarkkanen A, et al. Histopathology of enucleated intraocular melanomas irradiated with cobalt and ruthenium plaques. Acta Ophthalmol. 2009;66(3):255–66. 51. Saornil MA.  Histopathology of proton beam-­ irradiated vs enucleated uveal melanomas. Arch Ophthalmol. 1992;110(8):1112. 52. Echegaray JJ, Plesec TT, Bellerive C, et  al. Histopathologic patterns of recurrent choroidal melanoma following I-125 plaque brachytherapy. Ocul Oncol Pathol. 2019. [epub ahead of print]. 53. Harbour JW, Char DH, Kroll S, et al. Metastatic risk for distinct patterns of postirradiation local recurrence of posterior uveal melanoma. Ophthalmology. 1997;104(11):1785–92. 54. Griewank KG, van de Nes J, Schilling B, et  al. Genetic and clinico-pathologic analysis of metastatic uveal melanoma. Mod Pathol. 2013;27:175.

8

Uveal Melanoma: Molecular Pathology Sarah E. Coupland, Helen Kalirai, Sophie Thornton, and Bertil E. Damato

Introduction Intense efforts have been made in the last few decades to understand the molecular genetics involved in the development and progression of uveal melanomas (UM). These efforts have been undertaken in order to better predict which UM are likely to metastasize, for patient management purposes, and to identify signaling pathways and possible “druggable” molecules, which can be targeted using systemic therapies to improve the prognosis of patients with disseminated disease. Similar to other cancers, UM are characterized by an uncontrolled, cellular proliferation, occurring as a result of numerous genetic and epigenetic aberrations. They are characterized, however, by distinct genomic alterations, which distinguish them from other tumors, particularly from cutaneous melanomas. There are certain underlying risk factors unique to UM, such as congenital ocular melanocytosis, melanocytoma, and neurofibromatosis. Rare reports of families with UM have suggested a hereditary predisposiS. E. Coupland (*) · H. Kalirai · S. Thornton Department of Molecular and Clinical Cancer Medicine, Institute of Translational Medicine, University of Liverpool, Liverpool, Merseyside, UK e-mail: [email protected] B. E. Damato Nuffield Department of Clinical Neurosciences, University of Oxford, John Radcliffe Hospital, Oxford, UK

tion to this disease in some cases, with these patients also being at a higher risk of developing second primary cancers [1–5]. Most recently, germline mutations associated with BRCA1-­ associated protein 1 (BAP1) have been associated with a BAP1-tumor predisposition syndrome that predisposes patients to a range of cancers including UM, malignant mesothelioma, and renal cell carcinoma [6–10]. Signaling pathways altered in UM are most commonly associated with known initiating oncogenic drivers such as mutated GNAQ/11, resulting in the activation of multiple downstream factors such as MAPK, PI3K/AKT, Rho GTPase, and ARF6. In addition, several other pathways not directly associated with these oncogenic drivers are also critical in UM development, for example the retinoblastoma pathway, probably as a result of cyclin D1 overexpression and p53 signaling, possibly as a consequence of MDM2 overexpression. Characteristic chromosomal abnormalities are common and include 6p gain, associated with a good prognosis, as well as 1p loss, 3 loss, and 8q gain, which correlate with high mortality. Mutations in the BAP1 gene are most commonly associated with an increased risk of metastasis [11, 12]. Other genes known to influence the likelihood of metastatic spread (and located on some of these chromosomes) include SF3B1 [13] and EIF1AX [14] as well as LZTS1 [15], DDEF [16], PTP4A3 [17], TCEB1, and

© Springer Nature Switzerland AG 2019 B. E. Damato, A. D. Singh (eds.), Clinical Ophthalmic Oncology, https://doi.org/10.1007/978-3-030-17879-6_8

121

122

CNKSR3 [18]. This chapter provides a summary of the current insights into the molecular mechanisms underlying UM pathogenesis and briefly highlights potential therapeutic strategies (Box 8.1). Box 8.1. Molecular Pathology of Uveal Melanoma

• Objectives include diagnosis, prognosis, and treatment planning. • Important chromosomal abnormalities involve 1p, 3, 6p, and 8q. • Aberrations involve the RB, PI3K, and MAPK pathways. • Significant alterations involve mutations in GNAQ/11, CYSTLR2, PLCB4 BAP1, SF3B1, EIF1AX, and amplifications of CNKSR3. • BRAF mutations are rare in uveal melanomas.

Uveal Melanoma and the “Hallmarks of Cancer” Uveal melanomas are considered to arise from uveal tract melanocytes. The precursors of the differentiated melanocytes are nonpigmented immature melanoblasts derived from the neural crest, which migrate during embryogenesis. It is thought that the melanoblasts mature into melanocytes within the uvea and/or give rise to melanocytic stem cells, maintaining the ocular melanocytic “system” within the choroid, ciliary body, and iris. It is known that cutaneous melanocyte stem cells reside in the hair bulge in the skin [19]; the possible location of the ocular counterpart is not known at present. Various genetic and epigenetic alterations are thought to occur along the “melanoblast–melanocyte–nevus–uveal melanoma” pathway, which accumulate and ultimately cause the malignant transformation of the melanocytes, allowing for their self-autonomy and propensity to spread. It

S. E. Coupland et al.

is unclear whether melanoma “cancer stem-like” cells, demonstrated recently to be present in UM cell lines, are derived directly from ocular melanoblasts or from dedifferentiated melanocytes [20]. Recent functional analysis of gene expression data of UM would suggest that dedifferentiation does occur during UM development, but this is an area for future work [21, 22]. It also remains unclear whether the nevus stage is an absolute prerequisite during UM development or whether it can be “skipped” with melanomas directly arising de novo from either melanoblasts or transformed melanocytes [23]. It has been estimated that less than one in 8000 choroidal nevi undergo malignant transformation to form UM [24]. Confirming this observation, it is exceptionally rare to histologically observe residual nevi within the vicinity of a choroidal melanoma. This is in distinct contrast to other cancers, such as colon carcinoma, which is often associated with dysplastic mucosal polyps and with skin melanomas with their associated melanoma in situ. The six hallmarks of cancer are prerequisites for the survival and proliferation of a neoplastic cell at the primary site and for its ultimate ability to invade and metastasize [25, 26]. Despite this gap in our knowledge with respect to early UM pathogenesis, it has been demonstrated by several groups that the “hallmarks of cancer” can be applied when unraveling UM development. These (and their relationship to UM) include: Insensitivity to antigrowth signals [21, 27–30], Self-sufficiency in growth signals [21, 31–34], Avoiding apoptosis [21, 27, 29, 35–37], Limitless replicative potential [38, 39], Sustained angiogenesis [22, 40–42], and Tissue invasion and metastasis [43–45]. (Table 8.1). An increasing body of research suggests that two “emerging” hallmarks of cancer and two “enabling characteristics” are involved in cancer evolution [26]. These provisional capabilities (and their relationship to UM) include deregulating cellular energetics [22, 46], avoiding immune destruction [47–49], tumor-promoting inflammation [50–52], and genome instability and mutation (Tables 8.1, 8.2 and 8.3) [53–63].

123

8  Uveal Melanoma: Molecular Pathology Table 8.1  The classic and emerging “hallmarks of cancer” and their application to UM Hallmarks of cancer Example of gene/mechanism affected Classical hallmarks 1. Insensitivity to Loss of cell cycle inhibitor, such as retinoblastoma (Rb) suppressor anti-growth signals

2. Self-­ sufficiency in growth signals

Gain of cell cycle stimulator – activation of pathways.

3. Avoid apoptosis

p53 pathway alterations BCL-2 PTEN downregulation

4. Limitless replicative potential 5. Sustained angiogenesis

Produce insulin-like growth factor (IGF-1) survival factors Turn on telomerase

Production of vascular endothelial growth factor (VEGF) inducer either by the tumor cells or by accompanying inflammatory cells

6. Tissue invasion Activation of E-cadherin and metastasis

Emerging hallmarks Upregulating glucose transporters, 1. Deregulating e.g., GLUT1, resulting in substantial cellular increases in glucose import into energetics cytoplasm Hypoxia response of tumors acts by upregulating glucose transporters and multiple enzymes of the glycolytic pathway 2. Avoid immune Reduced tumor cell immunogenicity destruction

Mechanism(s) in UM The retinoblastoma tumor suppressor pathway is disrupted in most UM either through hyperphosphorylation of Rb, elevated expression of cyclin D1, or methylation and inactivation of the INK4A gene The PI3K–AKT prosurvival pathway is constitutively activated in UM. LOH of the PTEN locus occurs in 76% of UM The RAF/MEK/ERK pathway is constitutively activated: activating mutations in GNAQ or GNA11 occur in >80% of UM and can activate the RAF/MEK/ ERK pathway The p53 pathway is functionally blocked by its inhibitor MDM2 Defects in the Bcl2 pathway contribute to apoptosis resistance The PI3K–AKT prosurvival pathway is constitutively activated in UM to avoid apoptosis IGF1R is often upregulated and can activate the PI3K–AKT pathway Upregulated telomerase activity

Upregulated expression of VEGF; association with macrophage densities Increased expression of IGF-1 and IGF-1R Raised levels of hypoxia-inducible factor 1 alpha Upregulation of E-cadherin and Wnt/beta-catenin signaling pathways Downregulation of the helix-loop-helix inhibitor ID2 Increased expression of matrix metalloproteinases (MMPs) and downregulation of their tissue inhibitors (TIMPs) ALCAM expression NOTCH pathway activation Biallelic methylation of EFS Indirect evidence through the increased levels of the HIF1α and HIF2α transcription factors

Expression of PD-L1 by UM regulates T-cell function by suppressing IL-2 production and impairing T-cell function Downregulation of HLA class I and II Downregulation of HLA class I expression on UM cells expression T-cell exhaustion (continued)

S. E. Coupland et al.

124 Table 8.1 (continued) Hallmarks of cancer Example of gene/mechanism affected Enabling characteristics Lymphocytes 1. Tumor-­ Macrophages promoting Dendritic cells inflammation 2. Genome instability and mutation

Mechanism(s) in UM Varying densities of tumor infiltrating lymphocytes and macrophages; both associated with worse prognoses Tables 8.2 and 8.3

Table 8.2  Known genetic and epigenetic alterations described in some UM Alteration Proto-oncogenes NRAS BRAF NSB1 MYC DDEF1 (ASAP1) GNAQ/GNA11 CCND1 MDM2 BCL-2 Tumor suppressor genes LZTS1 CDKN2A-sporadic CDKN2A-familial PTEN BAP1 Other SF3B1 CNKSR3 Epigenetic alterations CDKN2A RASSF1 hTERT MicroRNA alterations let-7b miR18a miR-199a miR495 miR549

Mechanism

Chromosome

Frequency (%)

Mutation Mutation Amplification Amplification Amplification Mutation Amplification Amplification Amplification

1p13 7q34 8q21 8q24 8q24 9p21 11q13 12q15 18q21

∗ ∗ 50 43 50 >80 65 65 >95

Deletion Deletion, mutation Deletion, mutation Deletion, mutation Inactivating mutation

1p13 9p21 9p21 10q23 3p21

– ∗ ∗ 15 0–84%

Mutation Amplification

2q33.1 6q25.2

Hypermethylation Hypermethylation Hypermethylation

9p21 3p21.3

4–33% 13–70% 52%

Overexpression Overexpression Overexpression Overexpression Overexpression

n/a n/a n/a n/a n/a

−∗ −∗ −∗ −∗ −∗

Legend: ∗ = rare; −∗ = not documented; n/a = not applicable

Similar to other malignancies, it is thought that the genetic alterations outlined in Tables 8.1, 8.2, and 8.3 occur at differing stages of UM development. A critical bifurcation has been proposed to occur early in UM evolution whereby the resulting UM “subgroups” progress along either one of two genetic pathways

with very distinct genetic signatures (either 6p gain or class I molecular signature or a monosomy 3 or class II molecular signature) and differing metastatic propensities [46, 64–68]. Later genetic events are suggested to occur along the respective pathways, such as increasing aneuploidy.

8  Uveal Melanoma: Molecular Pathology Table 8.3  Most common chromosomal aberrations in UM

125

cancer types, producing inappropriate and autonomous proliferation of the tumor cells [77]. Most Chromosome Frequencya UM demonstrate constitutive activation of the 1p loss 28–34% MAPK pathway: [21, 32, 78] this is not a result of 1q gain 24% mutations of the “usual suspects,” such as muta3 loss 50–61% tions in KIT and the three RAS family members 6p gain 28–54% [32, 79–82], which can activate the MAPK path6q loss 35–37% way, or of RAF family members (ARAF, BRAF, 8p loss 17–28% 8q gain 36–63% CRAF) [32, 33, 78, 81, 83, 84]. Instead, MAPK/ a ERK pathway activation is caused in part by Combined data from Ref. [64, 69] mutations in guanine nucleotide-­binding protein Recent evidence suggests that this dichoto- G(q) subunit (GNAQ) in almost half of UM [33, mous model may be too simple, and that varying 34, 85]. GNAQ is one of a subfamily of genes, clones of malignant melanocytes occur within a comprising GNAQ, GNA11, GNA14, and single UM, with some of these clones having the GNA15/16. The GNAQ mutation is somatically potential to override or dominate others [69–73]. acquired occurring almost exclusively in exon 5 at This has been observed in longitudinal studies codon 209, resulting in substitution of the original using next-generation sequencing in paired pri- glutamine at this point. There are at least five mary and metastatic carcinomas [74] and war- known variants, with GNAQQ209L or GNAQQ209P rants further investigation using such techniques being the most frequent [34]. Mutations of codon 209 have also been recently found in GNA11, and in UM. Whole exome studies of UM have demon- both GNAQ and GNA11 can also have mutations strated that UM can be stratified into distinct in exon 4 affecting codon 183 [85]. Over 80% of molecular subgroups based on their genetic alter- UM were found to have GNAQ or GNA11 mutaations. An examination of 33 UM by deep-­ tions affecting either Q209 or R183 in a mutually coverage whole-genome sequencing identified exclusive pattern [85]. In the recent TCGA analyfour major subgroups associated with mutational sis, 94% of UM samples had mutually exclusive and metastatic status as a result of unsupervised GNAQ/GNA11 mutations with a single case dishierarchical clustering of copy number variation playing an R183 mutation in both genes [76]. Two [75]. Additionally, recent data from The Cancer further genes affecting the pathway downstream Genome Atlas (TCGA) analysis of 80 UM identi- of GNAQ/GNA11 have also been identified as fied four distinct molecular subgroups based on mutated in UM. These are CYSLTR2 and PLCB4, chromosomal copy number variations, muta- which were mutated in 4% and 1% of UM cases tions, and DNA methylation profiles. It would previously reported as wild type for GNAQ/ appear that although these molecular subgroups GNA11 [76, 86, 87]. GNAQ and GNA11 mutations are also found are distinct, they represent increasing metastatic risk from the lowest risk in group 1 through to the in uveal nevi and in most UM regardless of their size, chromosomal aberrations, or clinical feahighest risk in group 4 [76]. tures [88]. These mutations, which appear to be present at all stages of UM development and progression, are therefore necessary but not suffiMolecular Pathway Defects cient for malignant transformation to melanoma in Primary Uveal Melanoma [89]. Thus, GNAQ and GNA11 mutations are The mitogen-activated protein kinase/extracellu- considered to be initiating events in the molecular signal-regulated kinase (MAPK/ERK) path- lar evolution of UM. The PI3K/AKT pathway is constitutively way is integral for coordinating controlled cell cycle progression. Mutations in this pathway “switched on” in most UM (Table 8.1) [90]. Using result in it being constantly activated in numerous immunohistochemistry, it has been demonstrated

126

S. E. Coupland et al.

that phosphorylated AKT is associated with a worse prognosis in UM [91]. Loss of heterozygosity (LOH) of the PTEN locus has been shown in 76% of UM, with mutations in the PTEN coding region being seen in 0–11% of tumors [76, 92]. PTEN inactivation was also found to be associated with increased aneuploidy and decreased survival in UM [92, 93]. Taken together, these findings would suggest a role for PTEN in UM progression. In most UM, both the retinoblastoma (Rb) and p53 pathways are functionally inhibited, although actual mutations in either the RB1 or TP53 genes are rare [21, 35, 94–97]. The Rb protein is permanently hyper-phosphorylated and, therefore, functionally inactivated in most UM: this is probably due to cyclin D1 overexpression and/or CDKN2A promoter hypermethylation, which occurs in about two-thirds and one-third of cases, respectively [27, 29, 30, 98]. Increased cyclin D1 protein expression is associated with increased tumor basal diameter, epithelioid cell melanomas, and a poor prognosis [29]. Finally, inhibition of the p53 pathway downstream to p53  in UM is possibly a result of MDM2 overexpression [96], which is also observed in UM and associated with a poor outcome [28, 29]. The timing of the MAPK pathway alterations during UM development has been clarified partially while the temporal sequences of those affecting the Rb, p53, and PI3K/AKT.

situ hybridization (FISH) [102–105], comparative genomic hybridization (CGH) and array CGH [93, 106–109], spectral karyotyping [110, 111], microsatellite analysis (MSA) [112–117], multiplex ligation-dependent probe amplification (MLPA) [69, 118, 119], and single-nucleotide polymorphisms (SNPs) [120–125]. The above-mentioned chromosomal alterations in primary UM are clinically relevant because of their correlation with the risk of metastasis development and with the subsequent clinical course of UM patients (Chap. 20). Chromosome 3 loss is associated with a reduction of the 5-year survival probability from approximately 100% to less than 50% [126, 127]. Similarly, chromosome 8q gains and loss of chromosome 1p significantly correlate with reduced survival [69, 102, 104, 122]. Both chromosome 3 loss and polysomy 8q are also associated with other poor prognostic factors, which include increased tumor basal diameter, ciliary body involvement, the presence of epithelioid cells, high mitotic count, and closed connective tissue loops [69]. On the other hand, gains in chromosome 6p are associated with a good prognosis, suggestive of a protective effect initiated through these changes [128]. These chromosomal alterations correlate also with the gene expression profile UM subgroups: i.e., class I UM often has gains in chromosome 6p, and class II UM is associated with chromosome 3 loss [129].

Chromosomal Alterations in Primary Uveal Melanoma

 olecular Alterations Influencing M Uveal Melanoma Metastasis

Close to 20 years ago, chromosomal alterations specific to UM were first described: these are quite distinct from melanomas of other sites, particularly those of the skin (Table 8.3). The most striking abnormality in UM is the complete or partial loss of chromosome 3. Other common genetic abnormalities seen in UM include loss on 1p, 6q, 8p, and 9p and gain on 1q, 6p, and 8q (Table 8.3). Initially these alterations were identified by standard karyotypic analyses and chromosome banding [99–101] and then later using differing technologies, including fluorescence in

A number of genetic aberrations have been suggested to influence UM dissemination: [130] these are LZTS1 (located on chromosome 8p22) [15]; DDEF1 (“development- and differentiation-­ enhancing factor 1”, also known as ASAP1; chromosome 8q24.21) [16]; PTP4A3 (“protein tyrosine phosphatase type IV A member 3”; chromosome 8q24.3) [17]; TCEB1 (chromosome 8q21.11) [131]; BAP1 (chromosome 3p21.31­p21.2) [11]; SF3B1 (chromosome 2q33.1) [13]; EIF1AX (chromosome Xp22.12) and CNKSR3 (chromosome 6q25.2) [18].

8  Uveal Melanoma: Molecular Pathology

Of particular note are inactivating mutations of the BAP1 gene, which encodes for a deubiquitinating enzyme that binds to BRCA1 and BARD1 to form a tumor suppressor h­ eterodimeric complex. It is mapped to chromosome 3p21.31-p21.2, a region first noted by Trolet and coworkers to be of significance in UM [122]. Inactivating BAP1 somatic mutations have been described with varying frequencies in high metastatic-­ risk UM: Harbour and coworkers suggest that they occur in up to 84% of class II UM [11], although this has not been substantiated by other investigators. BAP1 somatic mutations have been implicated in other cancers including mesothelioma and lung, breast, and renal cell carcinomas [132]. Interestingly, germline BAP1 mutations have been described in families with high risk for hereditary cancer and a novel “BAP1 cancer predisposition syndrome” including UM has since been described by several groups [6, 7, 133–135]. Recent work also suggests that there may be genes which decelerate UM metastasis: these are EIF1AX [14], and CNKSR318 located on chromosomes X, 2 and 6, respectively. Although SF3B1 mutations were originally reported in low metastatic risk D3 UM, they have since been reported at low frequency in M3 UM. In both groups they are thought to be associated with an intermediate risk of developing metastatic disease, although this will be influenced by the presence of other molecular alterations. Other mutations which occur at a low frequency in UM may also contribute to metastatic progression. Royer-Bertrand et  al. identified additional 5 mutations in more than 1 of the 33 UM samples examined in their study, KTN1, DLK2, CSMD1, TTC28, and TP53BP1; however, the impact of these mutations has yet to be established [75]. Additionally, mutations in the splicing factor SRSF2 resulting in a change-offunction have also been identified in two separate studies in the same region that has been reported to disrupt splicing in myelodysplastic syndrome [67, 76, 136]. A mutation in FBXW7 in metastases was identified by Luscan et al., and this was also observed by Martin et al.; however, its involvement in tumorigenesis remains unknown [14, 137].

127

More recently, studies have shown that the cancer testis antigen gene, PRAME (“preferentially expressed antigen in melanoma”), has been associated with metastasis in disomy 3 or class 1 UM and UM with mutations in SF3B1 [138, 139]. The latter study demonstrated an association between T-cell reactivity and PRAME expression, suggesting a potential role for PRAME directed T-cell immunotherapy for metastatic UM.

 olecular Alterations in Uveal M Melanoma Metastases There are limited data available on cytogenetic or molecular genetic analyses of UM metastases. Trolet et al. examined 63 liver metastases using array CGH supported by a SNP array and could subdivide both the primary and secondary UM into distinct subgroups, groups 1 and 2, according to the presence or absence of chromosome 3. Following this, five subgroups (1a, 1b, 2a, 2b, and 2c) could be defined on the basis of imbalances of a few chromosome regions, mainly gains of 6p and 8q and losses of 1p, 8p, and 16q [122]. Most metastases were characterized by gain of 8q with a proximal breakpoint and losses of 3, 8p, and 16q, with the Group 2 UM dominating at 82% [122]. This is confirmed in the study by McCarthy et  al., who examined 12 hepatic UM metastases by aSNP and demonstrated that most were characterized by chromosome 3 loss and 8q gain [140]. Griewank et  al. examined the frequency of GNAQ, GNA11, SF3B1, and BAP1 mutations in 30 metastatic UM from different metastatic sites [141]. In this study, GNA11 mutations occurred more frequently than GNAQ mutations. Loss of BAP1 protein expression was also frequently seen, whilst mutations in SF3B1 were rare. Luscan et al. examined five hepatic UM using a targeted next-generation sequencing (NGS) approach and confirmed the presence of mutations in GNAQ, GNA11, BAP1, and SF3B1 [137]. As outlined by Hanahan and Weinberg [26], the metastatic process is multistepped and a ­complicated one and is dependent on numerous

128

supportive elements within the tumor’s microenvironment. The latter is also of significance at the metastatic site and is often the ­determinant for the ultimate location of successful tumor colonization. This concept of metastatic tropism of uveal tract “sarcomas” to the liver was first described by Ernst Fuchs [142]. It remains unexplained but may be accounted by the chemokine receptor– ligand axis (e.g., CXCR4 and CXCL12) [143], interactions between FAS and FAS ligand [144], IGF1 and IGF1-R, as well as C-Met and hepatocyte growth factor/scatter factor (HGF) [43, 145].

Strategies for Targeted Therapy in Uveal Melanoma Metastases It is beyond the scope of this chapter to review the numerous emerging chemotherapeutic regimens being considered in metastatic melanoma (Chap. 21) [146, 147]. However, it is worth briefly mentioning that potential targets for UM therapy include those affecting: the MAPK/MEK signaling pathway, inhibition of the PI3K/AKT pathway at the level of AKT, mTOR inhibitors, mTOR blockade combined with an IGF-1R antibody, tyrosine kinase inhibitors, c-Met pathway inhibitors, CXCR4 small molecule inhibitors, as well as histone deacetylase inhibitors (HDACi) [147, 148]. These are being considered and applied in both the adjuvant and advanced clinical settings. It can be anticipated that the rapidly developing field of molecular genetics will shed further light on key genes and signaling pathways involved in UM oncogenesis and progression, opening the way for target-based therapies. These could potentially be combined with emerging immunotherapies [147, 148] and epigenetic agents [149, 150], which are proving promising in other cancer types.

References 1. Canning CR, Hungerford J.  Familial uveal melanoma. Br J Ophthalmol. 1988;72(4):241–3. 2. Turner BJ, Siatkowski RM, Augsburger JJ, et  al. Other cancers in uveal melanoma patients and their families. Am J Ophthalmol. 1989;107(6):601–8.

S. E. Coupland et al. 3. Singh AD.  Familial uveal melanoma. Arch Ophthalmol. 1996;114(4):392. 4. COMS. Second primary cancers after enrollment in the COMS trials for treatment of choroidal melanoma. Archives of Ophthalmology. 2005;123(5):601. 5. Abdel-Rahman MH, Pilarski R, Ezzat S, et  al. Cancer family history characterization in an unselected cohort of 121 patients with uveal melanoma. Familial Cancer. 2010;9(3):431–8. 6. Abdel-Rahman MH, Pilarski R, Cebulla CM, et al. Germline BAP1 mutation predisposes to uveal melanoma, lung adenocarcinoma, meningioma, and other cancers. J Med Genet. 2011;48(12):856–9. 7. Carbone M, Ferris LK, Baumann F, et al. BAP1 cancer syndrome: malignant mesothelioma, uveal and cutaneous melanoma, and MBAITs. J Transl Med. 2012;10(1):179. 8. Murali R, Wiesner T, Scolyer RA.  Tumours associated with BAP1 mutations. Pathology. 2013;45(2):116–26. 9. Rai K, Pilarski R, Boru G, et  al. GermlineBAP1alterations in familial uveal melanoma. Genes Chromosom Cancer. 2016;56(2):168–74. 10. Masoomian B, Shields JA, Shields CL.  Overview of BAP1 cancer predisposition syndrome and the relationship to uveal melanoma. Journal of Current Ophthalmology. 2018;30(2):102–9. 11. Harbour JW, Onken MD, Roberson EDO, et  al. Frequent mutation of BAP1  in metastasizing uveal melanomas. Science. 2010;330(6009):1410–3. 12. Koopmans AE, Verdijk RM, Brouwer RWW, et  al. Clinical significance of immunohistochemistry for detection of BAP1 mutations in uveal melanoma. Mod Pathol. 2014;27(10):1321–30. 13. Harbour JW, Roberson EDO, Anbunathan H, et  al. Recurrent mutations at codon 625 of the splicing factor SF3B1  in uveal melanoma. Nat Genet. 2013;45(2):133–5. 14. Martin M, Maßhöfer L, Temming P, et  al. Exome sequencing identifies recurrent somatic mutations in EIF1AX and SF3B1 in uveal melanoma with disomy 3. Nat Genet. 2013;45(8):933–6. 15. Onken MD, Worley LA, Harbour JW. A metastasis modifier locus on human chromosome 8p in uveal melanoma identified by integrative genomic analysis. Clin Cancer Res. 2008;14(12):3737–45. 16. Ehlers JP. DDEF1 is located in an amplified region of chromosome 8q and is overexpressed in uveal melanoma. Clin Cancer Res. 2005;11(10):3609–13. 17. Laurent C, Valet F, Planque N, et al. High PTP4A3 phosphatase expression correlates with metastatic risk in uveal melanoma patients. Cancer Res. 2010;71(3):666–74. 18. Lake SL, Damato BE, Kalirai H, et  al. Single nucleotide polymorphism Array analysis of uveal melanomas reveals that amplification of CNKSR3 is correlated with improved patient survival. Am J Pathol. 2013;182(3):678–87.

8  Uveal Melanoma: Molecular Pathology 19. Yu H, Fang D, Kumar SM, et  al. Isolation of a novel population of multipotent adult stem cells from human hair follicles. Am J Pathol. 2006;168(6):1879–88. 20. Bergeron MA, Champagne S, Gaudreault M, et  al. Repression of genes involved in melanocyte differentiation in uveal melanoma. Mol Vis. 2012;18:1813–22. 21. An J, Wan H, Zhou X, et  al. A comparative Transcriptomic analysis of uveal melanoma and normal uveal melanocyte. PLoS One. 2011;6(1):e16516. 22. Onken MD, Ehlers JP, Worley LA, et al. Functional gene expression analysis uncovers phenotypic switch in aggressive uveal melanomas. Cancer Res. 2006;66(9):4602–9. 23. Aleksidze N, Medina CA, Singh AD. De novo evolution of a small choroidal melanoma. Ocul Oncol Pathol. 2015;1(2):83–7. 24. Singh AD, Kalyani P, Topham A. Estimating the risk of malignant transformation of a choroidal nevus. Ophthalmology. 2005;112(10):1784–9. 25. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell. 2000;100(1):57–70. 26. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell. 2011;144(5):646–74. 27. Coupland SE, Bechrakis N, Schuler A, et  al. Expression patterns of cyclin D1 and related proteins regulating G1-S phase transition in uveal melanoma and retinoblastoma. Br J Ophthalmol. 1998;82(8):961–70. 28. Brantley MA Jr, Harbour JW. Deregulation of the Rb and p53 pathways in uveal melanoma. Am J Pathol. 2000;157(6):1795–801. 29. Coupland SE, Anastassiou G, Stang A, et  al. The prognostic value of cyclin D1, p53, and MDM2 protein expression in uveal melanoma. J Pathol. 2000;191(2):120–6. 30. van der Velden PA, Metzelaar-Blok JA, Bergman W, et al. Promoter hypermethylation: a common cause of reduced p16(INK4a) expression in uveal melanoma. Cancer Res. 2001;61(13):5303–6. 31. Naus NC, Zuidervaart W, Rayman N, et al. Mutation analysis of the PTEN gene in uveal melanoma cell lines. Int J Cancer. 2000;87(1):151–3. 32. Zuidervaart W, van Nieuwpoort F, Stark M, et  al. Activation of the MAPK pathway is a common event in uveal melanomas although it rarely occurs through mutation of BRAF or RAS.  Br J Cancer. 2005;92(11):2032–8. 33. Onken MD, Worley LA, Long MD, et al. Oncogenic mutations inGNAQOccur early in uveal melanoma. Invest Opthalmol Vis Sci. 2008;49(12):5230. 34. Shoushtari AN, Carvajal RD.  GNAQ and GNA11 mutations in uveal melanoma. Melanoma Res. 2014;24(6):525–34. 35. Brantley MA, Worley L, Harbour JW.  Altered expression of rb and p53  in uveal melanomas following plaque radiotherapy11The author has no proprietary interest in this study. Am J Ophthalmol. 2002;133(2):242–8.

129 36. All-Ericsson C, Girnita L, Müller-Brunotte A, et al. c-Kit–dependent growth of uveal melanoma cells: a potential therapeutic target? Invest Opthalmol Vis Sci. 2004;45(7):2075. 37. Economou MA, All-Ericsson C, Bykov V, et  al. Receptors for the liver synthesized growth factors IGF-1 and HGF/SF in uveal melanoma: intercorrelation and prognostic implications. Invest Opthalmol Vis Sci. 2005;46(12):4372. 38. Heine B. Telomerase expression in uveal melanoma. Br J Ophthalmol. 2000;84(2):217–23. 39. Rohrbach JM, Riedinger C, Wild M, et  al. Telomeraseaktivität in uvealen Melanomen. Ophthalmologe. 2000;97(5):359–63. 40. Abdel-Rahman MH, Craig EL, Davidorf FH, et al. Expression of vascular endothelial growth factor in uveal melanoma is independent of 6p21-region copy number. Clin Cancer Res. 2005;11(1):73–8. 41. Kirschmann DA, Seftor EA, Hardy KM, et  al. Molecular pathways: vasculogenic mimicry in tumor cells: diagnostic and therapeutic implications. Clin Cancer Res. 2012;18(10):2726–32. 42. Cools-Lartigue J, Marshall JC, Caissie AL, et  al. Secretion of hepatocyte growth factor and vascular endothelial growth factor during uveal melanoma-­ monocyte in  vitro interactions. Melanoma Res. 2005;15(3):141–5. 43. Wu X, Zhou J, Rogers AM, et  al. c-Met, epidermal growth factor receptor, and insulin-like growth factor-­1 receptor are important for growth in uveal melanoma and independently contribute to migration and metastatic potential. Melanoma Res. 2012;22(2):123–32. 44. Anastassiou G, Tschentscher F, Zeschnigk M, et al. Cadherin expression in uveal melanoma. Exp Eye Res. 2002;74(3):423–5. 45. Zuidervaart W, Pavey S, van Nieuwpoort FA, et al. Expression of Wnt5a and its downstream effector β-catenin in uveal melanoma. Melanoma Res. 2007;17(6):380–6. 46. Harbour JW.  The genetics of uveal melanoma: an emerging framework for targeted therapy. Pigment Cell Melanoma Res. 2012;25(2):171–81. 47. de Waard-Siebinga I, Hilders CGJM, Hansen BE, et al. HLA expression and tumor-infiltrating immune cells in uveal melanoma. Graefes Arch Clin Exp Ophthalmol. 1996;234(1):34–42. 48. Hurks HMH, Metzelaar-Blok JAW, Mulder A, et al. High frequency of allele-specific down-regulation of HLA class I expression in uveal melanoma cell lines. Int J Cancer. 2000;85(5):697–702. 49. Yang W, Chen PW, Li H, et  al. PD-L1: PD-1 interaction contributes to the functional suppression of T-cell responses to human uveal melanoma cells in  vitro. Invest Opthalmol Vis Sci. 2008;49(6):2518. 50. Whelchel JC, Farah SE, McLean IW, et  al. Immunohistochemistry of infiltrating lymphocytes in uveal malignant melanoma. Invest Ophthalmol Vis Sci. 1993;34(8):2603–6.

130 51. Makitie T, Summanen P, Tarkkanen A, et al. Tumor-­ infiltrating macrophages (CD68(+) cells) and prognosis in malignant uveal melanoma. Invest Ophthalmol Vis Sci. 2001;42(7):1414–21. 52. Pi T, Mäkitie T, Kujala E, et al. Microcirculation and tumor-infiltrating macrophages in choroidal and ciliary body melanoma and corresponding metastases. Invest Opthalmol Vis Sci. 2004;45(1):1. 53. Sharma A, Stei MM, Frohlich H, et al. Genetic and epigenetic insights into uveal melanoma. Clin Genet. 2018;93(5):952–61. 54. Helgadottir H, Hoiom V.  The genetics of uveal melanoma: current insights. Appl Clin Genet. 2016;9:147–55. 55. Werdich XQ, Jakobiec FA, Singh AD, et  al. A review of advanced genetic testing for clinical prognostication in uveal melanoma. Semin Ophthalmol. 2013;28(5–6):361–71. 56. Li Y, Jia R, Ge S. Role of epigenetics in uveal melanoma. Int J Biol Sci. 2017;13(4):426–33. 57. Russo A, Caltabiano R, Longo A, et al. Increased levels of miRNA-146a in serum and histologic samples of patients with uveal melanoma. Front Pharmacol. 2016;7:424. 58. Ragusa M, Barbagallo C, Statello L, et al. miRNA profiling in vitreous humor, vitreal exosomes and serum from uveal melanoma patients: pathological and diagnostic implications. Cancer Biol Ther. 2015;16(9):1387–96. 59. Zhou J, Jiang J, Wang S, et  al. Oncogenic role of microRNA20a in human uveal melanoma. Mol Med Rep. 2016;14(2):1560–6. 60. Li Y, Huang Q, Shi X, et  al. MicroRNA 145 may play an important role in uveal melanoma cell growth by potentially targeting insulin receptor substrate-1. Chin Med J. 2014;127(8):1410–6. 61. Chen X, Wang J, Shen H, et  al. Epigenetics, microRNAs, and carcinogenesis: functional role of microRNA-137  in uveal melanoma. Invest Ophthalmol Vis Sci. 2011;52(3):1193–9. 62. Yan D, Dong XD, Chen X, et  al. Role of microRNA-182 in posterior uveal melanoma: regulation of tumor development through MITF, BCL2 and cyclin D2. PLoS One. 2012;7(7):e40967. 63. Yan D, Zhou X, Chen X, et al. MicroRNA-34a inhibits uveal melanoma cell proliferation and migration through downregulation of c-Met. Invest Ophthalmol Vis Sci. 2009;50(4):1559–65. 64. Höglund M, Gisselsson D, Hansen GB, et  al. Dissecting karyotypic patterns in malignant melanomas: temporal clustering of losses and gains in melanoma karyotypic evolution. Int J Cancer. 2003;108(1):57–65. 65. Parrella P, Sidransky D, Merbs SL.  Allelotype of posterior uveal melanoma: implications for a bifurcated tumor progression pathway. Cancer Res. 1999;59(13):3032–7. 66. Tschentscher F, Husing J, Holter T, et  al. Tumor classification based on gene expression profil-

S. E. Coupland et al. ing shows that uveal melanomas with and without monosomy 3 represent two distinct entities. Cancer Res. 2003;63(10):2578–84. 67. Field MG, Durante MA, Anbunathan H, et  al. Punctuated evolution of canonical genomic aberrations in uveal melanoma. Nat Commun 2018;9(1). 68. Singh N, Singh AD, Hide W.  Inferring an evolutionary tree of uveal melanoma from genomic copy number aberrations. Invest Ophthalmol Vis Sci. 2015;56(11):6801–9. 69. Damato B, Dopierala JA, Coupland SE. Genotypic profiling of 452 Choroidal melanomas with multiplex ligation-dependent probe amplification. Clin Cancer Res. 2010;16(24):6083–92. 70. Mensink HW, Vaarwater J, Kilic E, et al. Chromosome 3 intratumor heterogeneity in uveal melanoma. Invest Opthalmol Vis Sci. 2009;50(2):500. 71. Dopierala J, Damato BE, Lake SL, et  al. Genetic heterogeneity in uveal melanoma assessed by multiplex ligation-dependent probe amplification. Invest Opthalmol Vis Sci. 2010;51(10):4898. 72. Bronkhorst IHG, Maat W, Jordanova ES, et al. Effect of heterogeneous distribution of monosomy 3 on prognosis in uveal melanoma. Arch Pathol Lab Med. 2011;135(8):1042–7. 73. Callejo SA.  Sudden growth of a choroidal melanoma and multiplex ligation-dependent probe amplification findings suggesting late transformation to monosomy 3 type. Arch Ophthalmol. 2011;129(7):958. 74. Swanton C.  Intratumor heterogeneity: evolution through space and time. Cancer Res. 2012;72(19):4875–82. 75. Royer-Bertrand B, Torsello M, Rimoldi D, et  al. Comprehensive genetic landscape of uveal melanoma by whole-genome sequencing. Am J Hum Genet. 2016;99(5):1190–8. 76. Robertson AG, Shih J, Yau C, et al. Integrative analysis identifies four molecular and clinical subsets in uveal melanoma. Cancer Cell. 2018;33(1):151. 77. Denhardt DT.  Signal-transducing protein phosphorylation cascades mediated by Ras/Rho proteins in the mammalian cell: the potential for multiplex signalling. Biochem J. 1996;318(3):729–47. 78. Weber A, Hengge UR, Urbanik D, et  al. Absence of mutations of the BRAF gene and constitutive activation of extracellular-regulated kinase in malignant melanomas of the uvea. Lab Investig. 2003;83(12):1771–6. 79. Mooy CM, Van der Helm MJ, Van der Kwast TH, et al. No N-ras mutations in human uveal melanoma: the role of ultraviolet light revisited. Br J Cancer. 1991;64(2):411–3. 80. Soparker CN, O'Brien JM, Albert DM. Investigation of the role of the ras protooncogene point mutation in human uveal melanomas. Invest Ophthalmol Vis Sci. 1993;34(7):2203–9.

8  Uveal Melanoma: Molecular Pathology

131

81. Cruz F 3rd, Rubin BP, Wilson D, et al. Absence of 97. Scholes AG, Liloglou T, Maloney P, et  al. Loss of heterozygosity on chromosomes 3, 9, 13, BRAF and NRAS mutations in uveal melanoma. and 17, including the retinoblastoma locus, in Cancer Res. 2003;63(18):5761–6. uveal melanoma. Invest Ophthalmol Vis Sci. 82. Pache M, Glatz K, Bosch D, et  al. Sequence 2001;42(11):2472–7. analysis and high-throughput immunhistochemical profiling of KIT (CD 117) expression in uveal 98. Brantley MA Jr, Harbour JW. Inactivation of retinoblastoma protein in uveal melanoma by phosphorymelanoma using tissue microarrays. Virchows Arch. lation of sites in the COOH-terminal region. Cancer 2003;443(6):741–4. Res. 2000;60(16):4320–3. 83. Cohen Y, Goldenberg-Cohen N, Parrella P, et  al. Lack of BRAF mutation in primary uveal mela- 99. Prescher G, Bornfeld N, Becher R. Nonrandom chromosomal abnormalities in primary uveal melanoma. noma. Invest Opthalmol Vis Sci. 2003;44(7):2876. JNCI J Natl Cancer Instit. 1990;82(22):1765–9. 84. Rimoldi D, Salvi S, Lienard D, et  al. Lack of BRAF mutations in uveal melanoma. Cancer Res. 100. Sisley K, Rennie IG, Cottam DW, et al. Cytogenetic findings in six posterior uveal melanomas: involve2003;63(18):5712–5. ment of chromosomes 3, 6, and 8. Genes Chromosom 85. Van Raamsdonk CD, Griewank KG, Crosby MB, Cancer. 1990;2(3):205–9. et  al. Mutations in GNA11in uveal melanoma. N 101. Horsman DE, Sroka H, Rootman J, et al. Monosomy Engl J Med. 2010;363(23):2191–9. 3 and isochromosome 8q in a uveal melanoma. 86. Johansson P, Aoude LG, Wadt K, et al. Deep sequencCancer Genet Cytogenet. 1990;45(2):249–53. ing of uveal melanoma identifies a recurrent muta 102. Sisley K, Rennie IG, Parsons MA, et  al. tion in PLCB4. Oncotarget. 2015;7(4):4624–31. Abnormalities of chromosomes 3 and 8 in posterior 87. Moore AR, Ceraudo E, Sher JJ, et  al. Recurrent uveal melanoma correlate with prognosis. Genes activating mutations of G-protein-coupled recepChromosom Cancer. 1997;19(1):22–8. tor CYSLTR2  in uveal melanoma. Nat Genet. 103. McNamara M, Felix C, Val Davison E, et  al. 2016;48(6):675–80. Assessment of chromosome 3 copy number in ocular 88. Bauer J, Kilic E, Vaarwater J, et  al. Oncogenic melanoma using fluorescence in situ hybridization. GNAQ mutations are not correlated with disease-­ Cancer Genet Cytogenet. 1997;98(1):4–8. free survival in uveal melanoma. Br J Cancer. 104. Damato B, Duke C, Coupland SE, et  al. 2009;101(5):813–5. Cytogenetics of uveal melanoma. Ophthalmology. 89. Van Raamsdonk CD, Bezrookove V, Green G, 2007;114(10):1925–31.e1. et  al. Frequent somatic mutations of GNAQ 105. Aronow M, Sun Y, Saunthararajah Y, et  al. in uveal melanoma and blue naevi. Nature. Monosomy 3 by FISH in uveal melanoma: vari2008;457(7229):599–602. ability in techniques and results. Surv Ophthalmol. 90. Babchia N, Calipel A, Mouriaux F, et al. The PI3K/ 2012;57(5):463–73. Akt and mTOR/P70S6K signaling pathways in human uveal melanoma cells: interaction with B-Raf/ 106. Hughes S, Damato BE, Giddings I, et al. Microarray comparative genomic hybridisation analysis of intraERK. Invest Opthalmol Vis Sci. 2010;51(1):421. ocular uveal melanomas identifies distinctive imbal 91. Saraiva VS, Caissie AL, Segal L, et  al. ances associated with loss of chromosome 3. Br J Immunohistochemical expression of phospho-­ Cancer. 2005;93(10):1191–6. Akt in uveal melanoma. Melanoma Res. 107. Ghazvini S, Char DH, Kroll S, et  al. Comparative 2005;15(4):245–50. genomic hybridization analysis of archival formalin-­ 92. Abdel-Rahman MH, Yang Y, Zhou X-P, et  al. fixed paraffin-embedded uveal melanomas. Cancer High frequency of submicroscopic hemizygous Genet Cytogenet. 1996;90(2):95–101. deletion is a major mechanism of loss of expression of PTEN in uveal melanoma. J Clin Oncol. 108. Kilic E, van Gils W, Lodder E, et  al. Clinical and cytogenetic analyses in uveal melanoma. Invest 2006;24(2):288–95. Opthalmol Vis Sci. 2006;47(9):3703. 93. Ehlers JP, Worley L, Onken MD, Harbour JW.  Integrative genomic analysis of aneu- 109. Petrausch U, Martus P, Tönnies H, et al. Significance of gene expression analysis in uveal melaploidy in uveal melanoma. Clin Cancer Res. noma in comparison to standard risk factors for 2008;14(1):115–22. risk assessment of subsequent metastases. Eye. 94. Bronkhorst IHG, Jager MJ.  Uveal melanoma: the 2007;22(8):997–1007. inflammatory microenvironment. J Innate Immun. 110. Naus NC, van Drunen E, de Klein A, et  al. 2012;4(5–6):454–62. Characterization of complex chromosomal abnor 95. Chana JS, Wilson GD, Cree IA, et  al. c-myc, p53, malities in uveal melanoma by fluorescence in situ and Bcl-2 expression and clinical outcome in uveal hybridization, spectral karyotyping, and comparamelanoma. Br J Ophthalmol. 1999;83(1):110–4. tive genomic hybridization. Genes Chromosom 96. Sun Y, Tran BN, Worley LA, et al. Functional analyCancer. 2001;30(3):267–73. sis of the p53 pathway in response to ionizing radiation in uveal melanoma. Invest Opthalmol Vis Sci. 111. White JS, Becker RL, McLean IW, et al. Molecular cytogenetic evaluation of 10 uveal melanoma cell 2005;46(5):1561. lines. Cancer Genet Cytogenet. 2006;168(1):11–21.

132 112. Tschentscher F, Prescher G, Zeschnigk M, et  al. Identification of chromosomes 3, 6, and 8 aberrations in uveal melanoma by microsatellite analysis in comparison to comparative genomic hybridization. Cancer Genet Cytogenet. 2000;122(1):13–7. 113. Scholes AGM, Damato BE, Nunn J, et al. Monosomy 3  in uveal melanoma: correlation with clinical and histologic predictors of survival. Invest Opthalmol Vis Sci. 2003;44(3):1008. 114. Thomas S, Pütter C, Weber S, et  al. Prognostic significance of chromosome 3 alterations determined by microsatellite analysis in uveal melanoma: a long-term follow-up study. Br J Cancer. 2012;106(6):1171–6. 115. Häusler T, Stang A, Anastassiou G, et al. Loss of heterozygosity of 1p in uveal melanomas with monosomy 3. Int J Cancer. 2005;116(6):909–13. 116. Shields CL, Ganguly A, Bianciotto CG, et  al. Prognosis of uveal melanoma in 500 cases using genetic testing of fine-needle aspiration biopsy specimens. Ophthalmology. 2011;118(2):396–401. 117. Shields CL. Chromosome 3 analysis of uveal melanoma using fine-needle aspiration biopsy at the time of plaque radiotherapy in 140 consecutive cases. Arch Ophthalmol. 2007;125(8):1017. 118. Damato B, Dopierala J, Klaasen A, et al. Multiplex ligation-dependent probe amplification of uveal melanoma: correlation with metastatic death. Invest Opthalmol Vis Sci. 2009;50(7):3048. 119. Damato B. Progress in the management of patients with uveal melanoma. The 2012 Ashton lecture. Eye. 2012;26(9):1157–72. 120. Onken MD, Worley LA, Person E, et al. Loss of heterozygosity of chromosome 3 detected with single nucleotide polymorphisms is superior to monosomy 3 for predicting metastasis in uveal melanoma. Clin Cancer Res. 2007;13(10):2923–7. 121. Singh AD, Aronow ME, Sun Y, et al. Chromosome 3 status in uveal melanoma: a comparison of fluorescence in situ hybridization and single-nucleotide polymorphism array. Invest Opthalmol Vis Sci. 2012;53(7):3331. 122. Trolet J, Hupé P, Huon I, et al. Genomic profiling and identification of high-risk uveal melanoma by array CGH analysis of primary tumors and liver metastases. Invest Opthalmol Vis Sci. 2009;50(6):2572. 123. Lake SL, Coupland SE, Taktak AFG, et al. Whole-­ genome microarray detects deletions and loss of heterozygosity of chromosome 3 occurring exclusively in metastasizing uveal melanoma. Invest Opthalmol Vis Sci. 2010;51(10):4884. 124. McCannel TA, Burgess BL, Nelson SF, et  al. Genomic identification of significant targets in ciliochoroidal melanoma. Invest Opthalmol Vis Sci. 2011;52(6):3018. 125. Abi-Ayad N, Kodjikian L, Couturier J.  Techniques d’analyse génomique du mélanome uvéal: une revue bibliographique. J Fr Ophtalmol. 2011;34(4):259–64. 126. Prescher G, Bornfeld N, Horsthemke B, et  al. Chromosomal aberrations defining uveal melanoma of poor prognosis. Lancet. 1992;339(8794):691–2.

S. E. Coupland et al. 127. Prescher G, Bornfeld N, Hirche H, et al. Prognostic implications of monosomy 3  in uveal melanoma. Lancet. 1996;347(9010):1222–5. 128. White VA, Chambers JD, Courtright PD, et  al. Correlation of cytogenetic abnormalities with the outcome of patients with uveal melanoma. Cancer. 1998;83(2):354–9. 129. Onken MD, Worley LA, Ehlers JP, et  al. Gene expression profiling in uveal melanoma reveals two molecular classes and predicts metastatic death. Cancer Res. 2004;64(20):7205–9. 130. Decatur CL, Ong E, Garg N, et  al. Driver mutations in uveal melanoma: associations with gene expression profile and patient outcomes. JAMA Ophthalmol. 2016;134(7):728–33. 131. Asnaghi L, Ebrahimi KB, Schreck KC, et al. Notch signaling promotes growth and invasion in uveal melanoma. Clin Cancer Res. 2012;18(3):654–65. 132. Jensen DE, Proctor M, Marquis ST, et  al. BAP1: a novel ubiquitin hydrolase which binds to the BRCA1 RING finger and enhances BRCA1-­ mediated cell growth suppression. Oncogene. 1998;16(9):1097–112. 133. Wiesner T, Obenauf AC, Murali R, et al. Germline mutations in BAP1 predispose to melanocytic tumors. Nat Genet. 2011;43(10):1018–21. 134. Testa JR, Cheung M, Pei J, et  al. Germline BAP1 mutations predispose to malignant mesothelioma. Nat Genet. 2011;43(10):1022–5. 135. Njauw C-NJ, Kim I, Piris A, et al. Germline BAP1 inactivation is preferentially associated with metastatic ocular melanoma and cutaneous-ocular melanoma families. PLoS One. 2012;7(4):e35295. 136. Komeno Y, Huang Y-J, Qiu J, et  al. SRSF2 is essential for hematopoiesis, and its myelodysplastic syndrome-related mutations dysregulate alternative pre-mRNA splicing. Mol Cell Biol. 2015;35(17):3071–82. 137. Luscan A, Just PA, Briand A, et al. Uveal melanoma hepatic metastases mutation spectrum analysis using targeted next-generation sequencing of 400 cancer genes. Br J Ophthalmol. 2014;99(4):437–9. 138. Field MG, Decatur CL, Kurtenbach S, et al. PRAME as an independent biomarker for metastasis in uveal melanoma. Clin Cancer Res. 2016;22(5):1234–42. 139. Gezgin G, Luk SJ, Cao J, et al. PRAME as a potential target for immunotherapy in metastatic uveal melanoma. JAMA Ophthalmol. 2017;135(6):541–9. 140. McCarthy C, Kalirai H, Lake SL, et  al. Insights into genetic alterations of liver metastases from uveal melanoma. Pigment Cell Melanoma Res. 2015;29(1):60–7. 141. Griewank KG, van de Nes J, Schilling B, et  al. Genetic and clinico-pathologic analysis of metastatic uveal melanoma. Mod Pathol. 2013;27(2):175–83. 142. Fuchs E.  Das Sarkom des Uvealtractus. Graefe’s Archiv für Ophthalmologie. 1882;12(2):233. 143. Scala S, Ieranò C, Ottaiano A, et al. CXC chemokine receptor 4 is expressed in uveal malignant melanoma and correlates with the epithelioid-­mixed cell type. Cancer Immunol Immunother. 2007;56(10):1589–95.

8  Uveal Melanoma: Molecular Pathology 144. Anastassiou G, Coupland SE, Stang A, et  al. Expression of Fas and Fas ligand in uveal melanoma: biological implication and prognostic value. J Pathol. 2001;194(4):466–72. 145. Bakalian S, Marshall JC, Logan P, et  al. Molecular pathways mediating liver metastasis in patients with uveal melanoma. Clin Cancer Res. 2008;14(4):951–6. 146. Triozzi PL, Eng C, Singh AD.  Targeted therapy for uveal melanoma. Cancer Treat Rev. 2008;34(3):247–58. 147. Yang J, Manson DK, Marr BP, et  al. Treatment of uveal melanoma: where are we now? Ther Adv Med Oncol. 2018;10:1758834018757175. 148. Triozzi PL, Singh AD.  Adjuvant therapy of uveal melanoma: current status. Ocul Oncol Pathol. 2014;1(1):54–62.

133 149. Bailey FP, Clarke K, Kalirai H, Kenyani J, Shahidipour H, Falciani F, Coulson JM, Sacco JJ, Coupland SE, Eyers PA. Kinome-wide transcriptional profiling of uveal melanoma reveals new vulnerabilities to targeted therapeutics. Pigment Cell Melanoma Res. 2018;31(2):253–66. https://doi. org/10.1111/pcmr.12650. Epub 2017 Oct 15. 150. Ambrosini G, Do C, Tycko B, Realubit RB, Karan C, Musi E, Carvajal RD, Chua V, Aplin AE, Schwartz GK. Inhibition of NF-κB-dependent signaling enhances sensitivity and overcomes resistance to BET inhibition in uveal melanoma. Cancer Res. 2019;79(9):2415–25. https://doi.org/10.1158/00085472.CAN-18-3177. Epub 2019 Mar 18.

9

Animal Models in Uveal Melanoma Julia V. Burnier, Christina Mastromonaco, Jade Marie Lasiste, and Miguel N. Burnier Jr.

Introduction Animal models play an important role in the study of malignant transformation and metastasis. They can be used to imitate and reflect human cancer and to help us better understand tumor growth and behavior. This equips us with the tools we need to study the underlying mechanisms of cancer development and metastasis and to develop treatments for various malignancies. Animal models also act as a bridge between in  vitro data and clinical application, with the ultimate goal of translating preclinical work into human clinical trials. While the rate of translation from animal studies to clinical cancer trials is low, animal studies have undoubtedly provided us with a means to advance cancer research and to develop new methods of treating patients. Animal models in oncology are becoming increasingly more sophisticated due to advances in technology, imaging, and molecular biology J. V. Burnier Department of Oncology and Pathology, McGill University, MUHC-RI, Montreal, QC, Canada e-mail: [email protected] C. Mastromonaco · J. M. Lasiste M. N. Burnier Jr. (*) Department of Ophthalmology and Pathology, The MUHC-McGill University Ocular Pathology & Translational Research Laboratory, Montreal, QC, Canada e-mail: [email protected]

tools and our understanding of tumor biology. As such, animal modeling is playing increasingly important roles in the study of the molecular mechanisms underlying tumors and in preclinical research, not only focusing on tumor growth and metastasis but also on defined steps during malignant transformation and progression, such as the role of inflammation and the tumor microenvironment, angiogenesis, cancer stem cells, tumor heterogeneity, response to treatment, and therapeutic resistance. By combining clinical and histopathological information, molecular tumor profiling, and increasingly sophisticated animal models, it is possible to uncover clinically useful prevention and treatment approaches that are personalized for patients with malignant diseases. In this chapter, the various types of animal models for UM will be discussed, with focus on the optimal applications of each model and their limitations.

 he Importance of Animal Models T in UM While control of primary uveal melanoma is highly successful, metastatic disease claims the lives of a large proportion of UM patients. The poor outcomes in UM are mainly a result of: (1) the difficulty in early diagnosis due to the asymptomatic nature of small tumors; (2) the

© Springer Nature Switzerland AG 2019 B. E. Damato, A. D. Singh (eds.), Clinical Ophthalmic Oncology, https://doi.org/10.1007/978-3-030-17879-6_9

135

J. V. Burnier et al.

136

absence of a sensitive tool to detect micrometastases; (3) the lack of biomarkers to track the molecular evolution of UM; and (4) few therapeutic options for metastatic disease. These gaps are rooted in the lack of understanding of the changes that occur between primary tumor induction, growth and evolution, and overt metastasis. Recently, increased understanding of the genetic, molecular, and environmental factors leading to UM has helped shed light into the pathways leading to the disease. However, our ability to study UM remains limited in that it largely relies on tissue sampling obtained through enucleation or biopsy. Intraocular biopsy carries the risk of rare but serious complications, such as retinal detachment and hemorrhage, and is therefore seldom performed in most centers. In addition, biopsy often results in inadequate sampling, and molecular profiling is not possible on necrotic tissue in patients who have previously undergone radiation [1]. But the most important disadvantage of a single assessment for studying molecular changes is that it neglects the dynamics of tumor evolution. Furthermore, establishing UM cell lines from patients has proved difficult and this has limited the number of distinct cells available to study the disease. Collectively, for these reasons, the use of animal models to study the biology of UM is essential. Many attempts have been made to create ocular melanoma models that mimic aspects of the human disease. An ideal model would reflect human UM in terms of genetics (monosomy 3, GNAQ/11, BAP1), inflammatory response, histopathology, tumor growth, and hematogenous dissemination to the liver [2]. While no perfect model currently exists, several different animals have been used to study different aspects of the disease. Currently, models differ by the type (if any) of melanoma cells, the species of animal, the manner in which the cells are introduced, and the method of tumorigenesis. There are several main types of animal models, with the three main types being spontaneous models, inoculation models and induced models, all of which exist for ocular melanoma. However, each model presents specific limitations, such as the use of cutaneous melanoma cells which possess marked genotypic and phenotypic differences to UM cells, the use of

immunosuppression in xenograft models, inoculation in non-ocular target organs, lack of metastatic manifestation, and differences in the anatomy of the animal versus human eye. The models also do not reflect the genetic and epigenetic heterogeneous nature of the disease and do not represent differences in the host genetic background and immune response. However, despite these limitations, animal models can be adapted to reflect certain aspects of the human disease. As such, the selection of the appropriate model is essential to yield clinically significant and biologically sound information. Ideally, the model should reflect the questions being asked and should mimic the behavior and pathological state of the corresponding phase in the human disease.

Types of Animal Models in UM Researchers have been searching for the ideal model to recapitulate the natural disease progression of UM.  Animal models have been refined throughout the years and many types of models have been produced. Herein, we will define these models and compare their different strategies and applications.

Spontaneous Models A spontaneous model is ideal to replicate the natural disease course and its progression, as no induced alteration or administration of any substance is required for UM tumor formation. In the case of UM, a number of species have been reported to spontaneously develop UM, including dogs, cats, horses, fish, sheep, and cattle [3, 4]. In dogs and cats, the most common form of uveal melanoma arises in the anterior chamber; however, canine UMs tend to be less metastatic than those of felines [5]. UMs in cats are predominantly diffuse iris melanomas, which are naturally more metastatic [6]. Histologically, the mitotic index in dogs is the most reliable feature predicting malignant behavior of the tumor [6], whereas in cats, neither mitotic index, cellular pleomorphism, or degree of pigmentation is a good indicator of malignancy [7]. Unfortunately, spontaneous UM development in mice is rare [8]; therefore, other types of models have been cre-

9  Animal Models in Uveal Melanoma

ated to develop UM in mice and will be discussed in the subsequent sections. The biggest advantage in using spontaneous models is the lack of manipulation required. The tumor environment has a dynamic immune system, allowing researchers to understand how the immune cells and surrounding factors promote tumor growth. However, the number of animals that develop UM spontaneously is small, as UM is a rare disease. Unpredictable growth and metastatic spread is not controlled, and may be more challenging in controlled experimental conditions.

Syngeneic Models Syngeneic models, otherwise known as allograft models, are models using cells injected into a host with the same genetic background. This model has been shown to produce less immunological rejections due to the similar nature of the cells towards its host, and does not require any immunodeficiency. The most frequently used syngeneic models were based on the mouse cutaneous melanoma model using B16 or the metastatic variant B16F10 cell line, microinjected into the eyes of the C57BL6 mice [9]. The B16LS9 cell line was further created through passage selection of higher metastatic potential cells targeting the liver, leading to the only model with an ocular tumor that specifically produces liver metastasis [9]. Recently, one study used a newer murine cutaneous cell line named HCmel12 in their syngeneic mice model, evaluating the histopathological and immunohistological features of UM in comparison to the human tumor; however, no liver metastasis occured [10]. Syngeneic models have also been utilized as preclinical evidence that certain immunotherapeutic drugs, such as the Toll-­like receptor 5 agonist, can reduce metastasis to the liver [11]. Other species, such as hamsters [12] and rabbits, have been deployed as syngeneic models, injecting cells into the anterior, posterior, or subchoroidal space [13]; however, these experiments were more complicated as tumors grew quicker, and metastasis was not reliable [14]. In this model, animals are immunocompetent, allowing researchers to study the immune

137

and angiogenic effects of the tumor [15, 16]. The major drawback of using cutaneous melanoma cell lines is the differences seen between the genetic makeup of cutaneous and uveal melanomas. B16 melanoma cell lines carry skin cancer mutations, such as the BRAF mutation, not typically seen in UM [17]. Furthermore, murine cell lines are not human cell lines, and thus do not reflect tumors produced in human patients.

Xenotransplantation Models Xenografts are defined as the implantation or transplantation of cells, tissues, or organs from one species into another. The first xenograft model in UM was established using hamster melanoma tissue implanted into the anterior chamber of rabbits [18]. Since then, human xenograft models, using human cutaneous and uveal melanoma cell lines, have been established in various species, including rabbits, rats, mice, zebrafish, and the chick embryo [19–21]. Rabbits have been frequently used due to their larger eye size, rendering them suitable for conducting ophthalmological examinations. Studies have continuously demonstrated the efficacy of using human cell lines for the establishment of animal models that replicate the primary and metastatic diseases, test possible therapeutic drugs and to further identify risk factors [22–27]. Consequently, rabbits need cyclosporine A immunosuppression for proper UM cell survival within the rabbit; therefore, conclusions on the role of the immune system are limited. Metastatic growth is seen in these models; however, metastasis was preferential to the lung instead of the liver. In an animal model using human cutaneous melanoma cells (WM-266-4), cells were inoculated into the suprachoroidal space of the eye of immunosuppressed rabbits [28]. Tumor growth was evaluated by fundoscopy. Interestingly, 66.7% of those animals had an anterior chamber tumor growth, which is not observed when human UM cells are inoculated. Necrosis was also a prominent feature of the tumors, which is also rare in UM. The authors speculated that cutaneous melanoma cells grew towards the conjunctiva, the only area of the eye that contains lymphatic vessels. These findings exemplify the importance of using UM cells in developing UM animal models (Fig. 9.1).

138

J. V. Burnier et al.

a

b

c

d

Fig. 9.1 Human cutaneous melanoma ocular model. 1 × 106 human CM cells (WM–266–4) were injected into the suprachoroidal space of the eye and fundoscopy was performed in week 3 (a). A large involvement of the anterior part of the eye, particularly the conjunctiva, can be seen 3  weeks after inoculation of the human cutaneous

melanoma cells (b). Gross picture of the corresponding enucleated eye shows that the globe is filled with an amelanotic tumor (c). The amelanotic intraocular tumor extends into the conjunctiva and external areas of the anterior aspect of the eye (d)

Mouse models using human cell lines have been used for histological characterization, identifying genetic pathways and assessing treatment efficacy [29–31]. Mice xerographs were used to investigate potential serum biomarkers, such as C-Met, for metastatic UM [32]. In a recent study, a novel model established good primary UM tumors using UMT2 cell lines injected into the choroid of BALB/c nude mouse [33]. Newer models have been established in recent years using zebrafish, enabling screening and preclinical drug testing in a fast and sensitive manner [34]. Low-cost models also include the chick embryo model, which requires no immunosuppression [35]. In that model, human UM cells

can simply be injected into the chick embryo between days 10–15, as the embryo has not yet acquired a functional immune system to reject the cells [36]. Patient-derived xerographs (PDXs) have gained increasing popularity, as it allows for fragments of the patient’s primary or metastatic tumor to be implanted into an animal model. The implanted tumor maintains its original characteristics [37], making it ideal for drug screening and analysis of combinational drugs for each tumor, overall taking a personalized medicine approach [38]. Studies using UM fragment transplants inserted into the subcutis or interscapular fat pad in immunosuppressed mice

9  Animal Models in Uveal Melanoma

demonstrated low engraftment rates [39]. Higher tumor engraftment rates are seen when the metastatic tumor, rather than the primary tumor, is implanted. Recently, the first orthotopic hepatic PDX model using two UM metastatic cell lines to recapitulate metastatic disease was found to be successful [40]. Primary cell-line orthotopic xenografts have been transplanted in the anterior or posterior chambers, ciliary body and/or choroid, to establish a primary tumor within the eye. Ectopic xenografts are made by implantation of human cell lines into the subcutaneous flank; however, this methodology does not result in metastasis unless cells are directly implanted into the liver [41]. Pseudo-­ metastatic xenotransplantation models have been created through injections into the tail vein, liver, or spleen [42, 43]. These types of models are ideal to analyze metastatic mechanisms and potential drug treatments against metastasis. An advantage of the xenograft model is the pronounced tumor stability, as genetic features of the original tumor are maintained. However, the disadvantage is that most models require immunosuppression. Furthermore, there are limited established human UM cell lines which exist and that are validated for research.

Transgenic Models Transgenic models allow for specific modifications of the genome, replicating aspects of the human disease. The ideal transgenic animal model would carry the same human mutation within the endogenous locus [44]. Early metallothionein-I (MT)-ret transgenic mice models, overexpressing the RET oncogene (primarily used in cutaneous melanoma experiments), were crossed with AAD mice expressing the chimeric major histocompatibility complex (MHC) molecule. The RET. AAD models exhibited hyperplastic lesions with melanocyte-­containing tissue [45]. RET mutations, however, have not been found in human melanomas, even though they activate the MAP kinase pathway, simulating melanomas [46].

139

However, in some transgenic models in mice, the intraocular tumor appears to be of retinal pigment epithelium (RPE) cell origin [47]. To address this issue, the Tyr-RAS+ Ink4a/Arf −/− transgenic male was developed, and has been demonstrated to spontaneously form UM and CM tumors that are not of RPE origin [48]. In this model the H-ras gene is integrated into the Y chromosome, which becomes activated by its tyrosine promoter driving melanocyte gene expression. The Ink4a/Art deficient gene promotes its overexpressive drive, producing a UM tumor with both epithelioid or spindle shaped characteristics, though no metastasis was detected. Other markers of UM are crucial to produce models with similar features to the human disease. A model using the glutamate receptor 1 (Grm1) transgene has been used as Grm expression is increased in UM [49]. The TG(Grm1) transgenic mice model spontaneously develop UM with features that included high expression of Ki-67+ proliferating cells in the choroid, and positive MelanA and S100B cells; all characteristics of a human UM tumor [50]. A recent mouse model, showed to be the first of its kind, reproduced the GNAQ oncogenic mutation found in the majority of human UM tumors [51]. G-protein mutations, GNAQ or GNA11, are the most common mutation in early UM tumorigenesis, as their alterations are present in 80% of patients. GNAQ and GNA11 play a role in activating G protein signalling cascade via I3P, DAG, and cAMP, which activates MAP kinase/AKT signalling [52]. While 15–20% of patients do not show a mutation in GNAQ or GNA11, many have a mutation in CYSLTR2, which codes for a G protein-coupled receptor that leads to constitutive activation of GNAQ signalling, or PLCB4, which activates GNAQ downstream signalling [53]. A mutually exclusive mutation in GNAQ, GNA11, CYSLTR2, and PLCB4 is found in almost all cases of UM and thus may act as initiating mutations [54]. The GNAQ(Q209L) oncogene is under the control of the Rosa26 promoter, overexpressing Yap protein in the eye, leading to blood vessels with melanocytic invasion [55]. Results showed neoplastic proliferation of the choroid

140

J. V. Burnier et al.

Table 9.1  Model types of uveal melanoma disease Model Types Spontaneous

Syngeneic

Xeno-­ transplantation

Transgenic

Induced

Advantages No manipulation is required No substrate administration Replicate natural disease Analyze immune system interactions Immunocompetent model Ability to study the immune system and the interaction of angiogenic factors High efficacy when using human UM cell lines and replicating human UM Good for preclinical drug testing and finding biomarkers Metastasis does form Available lower cost models (Zebrafish and chick) PXD models ideal for a personalized medicinal approach Models exist that can replicate GNAQ mutation seen in human UM Good to study early stage tumor progression Genetic pathways can be discovered Good to be used in combination with other model types

Disadvantages UM is rare Unpredicted growth or metastasis Challenging for controlled experimental conditions Cutaneous melanoma cells have different genetic makeup than UM Not using human cells Some models need immunosuppression (rabbits) Lack of human UM cell lines

Some models can have RPE originating cells Some models have metastasis development

Artificially inducing UM Not reproducible and uncontrolled

The advantages and disadvantages of the main types of animal model in UM are shown

and other melanocytic neoplasia, with metastasis only to the lung. Transgenic mice models using GNAQ(Q209L) oncogene have been used to test inhibitors for UM treatments [56]. Zebrafish transgenic models have further been established, also inducing the GNAQ(Q209P) oncogene in melanocytes in combination with the inactivation of the tumor suppressor gene p53 [57]. These tumors progressed at an early onset and showed increased hyperplasia of the melanocytes when compared to just GNAQ overexpression on its own. Transgenic models have good potential for their use in studying early stage tumor progression, as most of these models do not form metastasis. They permit the study of molecular pathways and genetic alterations, and the analysis of specific therapeutic targets.

administration. These stresses cause mutations within the genome, leading to hyperplasia and eventually tumor formation. Older models have used oncogenic viruses to induce neoplasia [58], and chemical induction in rabbits [59]; however, results are not reproducible. The disadvantage of this model is that the conditions are uncontrolled, and there is uncertainty of tumor development for all groups. Future models may be a combination of a transgenic model with an inducible source to target specific mutations that can generate tumors. Applications of these models are endless. As researchers replenish, refine, and create new models, we can one day produce an ideal model that replicates human UM pathogenesis more completely (Table 9.1).

Induced Models

 pplications of Human Uveal A Melanoma Models

Induced models are artificially created through the exposure of a carcinogenic agent such as radiation and chemical, physical, or biological

As we have shown in the previous section, many UM animal models exist, and while none perfectly mimics the natural history of disease in

141

9  Animal Models in Uveal Melanoma

humans, several have been developed that are tailored for particular research questions. These models have been used to study the pathophysiology of the disease, to elucidate potential prognostic markers and therapeutic targets, and to evaluate the safety and efficacy of treatment strategies. In this section, we review the animal models that have been reported in the literature to represent the different stages of UM: primary ocular disease, tumor dormancy, and metastasis. We also consider the models that have been utilized for specific objectives, such as for the study of tumor angiogenesis and for preclinical investigations of disease interventions.

Primary Ocular Disease Many models to induce UM in mice, hamsters, rats, and rabbits have been reported. Attempts to model UM in mice have concentrated on using

the B16 cutaneous melanoma cell line; although successful in generating tumors, it is questionable whether the biology of these tumors reflects that of human UM [8]. Human UM cells lines have been used more frequently in rats, although the injection of tumor spheroids have been shown to have a higher success rate in generating intraocular tumors, perhaps due to decreased leakage of tumor cells [60]. However, the albino rabbit is widely considered to be the best experimental organism for UM because its eyes approximate the size and resembles anatomically that of the human eye. The immunosuppressant Cyclosporine A (CsA) has to be administered to rabbits to increase the success rate of developing ocular tumors after orthotopic implantation, although different concentrations of CsA do not affect the incidence of metastasis [23] (Figs. 9.1, 9.2, 9.3, and 9.4). Three methods of orthotopic implantation of UM have been described: transscleral choroidal

a

b

c

d

Fig. 9.2  Experimental model of UM in albino rabbits. In this 12-week experiment, 1 × 106 human UM (92.1) cells are inoculated in the suprachoroidal space of the rabbit

eye (a). By week 3, intraocular tumors can be seen by fundoscopy (b and c). Blood is obtained throughout the experiment to detect circulating malignant cells (d)

142

J. V. Burnier et al.

a

b

c

d

Fig. 9.3  Experimental model of UM in albino rabbits. Dilated fundus examination reveals a lightly-pigmented choroidal tumor by week 5 (a). The corresponding enucleated eye in that rabbit reveals the mildly-pigmented

diffuse choroidal tumor (b). In another animal, a moderately pigmented choroidal tumor can be seen by funduscopic examination and in the enucleated eye by week 7 (c and d)

injection; suprachoroidal injection; and implantation through a surgically-induced cyclodialyis cleft. Both human UM cell lines in a tumor suspension and fragments of UM xenotransplanted from mouse flanks have been used; suprachoroidal implantation of tumor fragments resulted in a higher success rate (70%) of developing UM [61]. Of the many available UM cell lines used, it was found that the suprachoroidal injection of the more aggressive 92.1 and SP6.5 were most efficient in generating intraocular tumors, both with a success rate of 100%, as confirmed by

histological analysis post-mortem [24]. It is this model that is most widely used at present.

Natural History of Uveal Melanoma A rabbit model, wherein suprachoroidal injection of 92.1 cells was performed to orthotopically induce ocular UM, was used to study the natural growth and progression of UM over 10  weeks. With this technique, intraocular tumors formed in 89% of the experimental rabbits. Tumors were

143

9  Animal Models in Uveal Melanoma

a

b

Fig. 9.4  Experimental model of UM in albino rabbits. Gross pathology image (a) and corresponding histopathological section (b) of an enucleated rabbit on week

9 showing a mushroom-shaped choroidal melanoma with a large extraocular extension, indicated by the arrows

noted beginning with 1 week post-injection and progressed initially as a thickening of the choroid before evolving into a dome-shaped mass by 4  weeks. Extraocular and extrascleral extension were also observed in 90% of the animals after 4 weeks. By 8–9 weeks post-injection, the vitreous cavity was filled with the tumor and, at the end of the experiment, the tumor filled the globe in 30% of animals. The tumors were characterized histopathologically to be of a mixed cell type, with predominantly epithelioid cells, with marked lymphocytic infiltration. Immunohistochemistry markers were consistent with UM [22]. This model was also used to study other aspects of UM disease, such as metastasis and the presence of circulating malignant cells. Micrometastatic lesions in the lung were noted at 4 weeks after injection, and larger nodules were seen at 7 weeks; by 10 weeks, 100% of the rabbits had pulmonary metastasis. Meanwhile, hepatic micrometastases were observed at 10  weeks post-injection and in only 18% of the animals. Circulating malignant UM cells (CMCs) were detected at 6 weeks after injection and in 26% of the experimental rabbits [22]. Ideally, spontaneous or transgenic models of UM would be preferred over induced animal models, especially when a research endeavor aims to study disease progression. However,

spontaneous tumors are limited by their unpredictability in occurrence and spread. Different transgenic mice have also been developed, but the intraocular tumors that grow in these mice have a component that originates from the retinal pigment epithelium (RPE) [8]. In addition, these models do not develop metastases unless tumor cells are transplanted intracamerally [62]. Tg (Grm1) transgenic mice that develop nodular melanomas on hairless skin areas also have a thickened choroid and numerous Ki-67 positive cells, suggesting a high proliferative capacity and potential tumor growth [50]. As mentioned previously, the first genetically modified mouse that bears the driver mutation GNAQ (Q209L) has been created using a Rosa26-­ floxed stop-GNAQQ209L allele via a loxP-flanked stop cassette that prevents transcription. Mice bearing this allele were crossed with Mitf-cre transgenic mice, which express Cre recombinase under the control of the melanocyte-specific promoter of the Microphthalmia gene. Mice progeny developed all the three tumors that have been associated with the GNAQ mutation: blue nevi, uveal melanoma, and melanocytomas of the central nervous system (CNS). On histopathology, there was note of localized invasion of blood vessels and multiple lung tumors at 3 months of age, thereby suggesting that GNAQQ209L mutations could also contribute to disease progression. While there

J. V. Burnier et al.

144

was no hepatic metastasis observed in these mice, this transgenic model is a promising tool to study the biology of UM and to screen for potential prognostic markers and therapeutic targets [51].

Cancer Angiogenesis and Immunology Rabbits have been used to model microcirculation in UM, with tumors generated by implantation of tumor fragments or injection of malignant cells into the choroid, followed by analysis and visualization with confocal indocyanine green (ICG) angiography [61]. But smaller animals, such as mice and rat, have been much utilized in studying UM angiogenesis, hemodynamics, and immunology. With the research focus on the ­circulation instead of the eye, smaller animals are more advantageous to use because of ease of handling and maintenance. Posterior chamber inoculation of the mouse B16 cell line into the C57BL/6 mouse results in an invasive tumor that requires enucleation within 2 weeks and which produces metastases in a month [8]. This model has been used to correlate the blood volume within the tumor, as detected by high-frequency contrast-enhanced ultrasound, with histologic tumor vascularity [63]. It has also been used to determine that constitutively overexpressing pigment epithelium-derived growth factor (PEDF), an angiogenesis suppressor, via lentiviral transduction decreased both the growth of the intraocular tumor and hepatic micrometastasis [64]. Another model used nude athymic rats to study the hemodynamics of choroidal melanoma. With this technique, small fragments of subcutaneously-­ grown OCM-1 xenografts from donor rats were implanted into the choroidal space. After 6 to 8  weeks, rhodamine-labeled liposomes and red blood cells (RBCs) were injected intravenously and monitored via epifluorescence microscopy. This experiment demonstrated increased RBC flux and hematocrit (viscosity) in medium to large tumor vessels, implying increased vascular resistance and decreased blood flow. Such findings are important considerations in terms of efficacy of drug delivery and in the understanding of angiogenesis [20].

Tumor Dormancy Tumor dormancy is defined as the interval between a disease-free state, after successful treatment of the primary tumor, and the development of overt metastasis. In UM, tumor dormancy can last more than 25  years. Understanding the phenomenon of tumor dormancy can improve insights into the mechanisms underlying UM metastasis, which is the leading cause of death in patients. A bulb-c nu/nu mouse model has been reported to investigate tumor dormancy in UM. Cells of the UM cell line 92.1 were transfected with green fluorescent protein (GFP) and injected into the tail vein. Unlike other models that utilized the portal vein, this model simulated the entry of UM cells into systemic circulation [65] (Fig. 9.5). The cells were tracked using an epifluorescence microscope, and it was observed that the UM cells were present in the lung, kidney, spleen, and liver immediately after injection. After two weeks, however, the UM cells were detected only in the liver. Consistent with the fact that the most common site of UM metastasis is the liver, these findings show that the microenvironment plays a role in the homing of cancer cells to target organs. There were no changes in cell number and size throughout the 6 weeks of observation, confirming that the cells were quiescent. In this model, however, no micrometastatic lesions were noted, and so the transition from the dormant to active phase is yet to be studied [65] (Fig. 9.6).

Metastasis UM metastasis is associated with a very poor prognosis: the mortality rate is 90% within 2 years after the diagnosis of metastasis, with a median survival time of less than a year. Research geared towards improving treatment options for metastatic disease is hampered by the sparse number of metastatic cell lines and animal models. Most experimental models of UM metastasis utilize either ocular, hepatic, splenic, or intravascular injection of tumor cells to generate metastatic lesions, although some utilize ectopic sites such as the skin.

145

9  Animal Models in Uveal Melanoma

a

b

Fig. 9.5  Experimental model of GFP-transfected human UM cells in nude mice. The cells were injected in the tail vein of nude mice (a and b)

Metastatic cell lines developed from UM hepatic lesions have been implanted subcutaneously on SCID immunocompromised mice and have been used to study the therapeutic potential of pharmacologic agents [66]. Subcutaneous tumors, however, do not mirror the hepatic microenvironment. Another model has used liver-­selected cells from the cutaneous mouse melanoma B16 cell line for retroorbital injection. Although metastatic lesions in the liver develop, the genetic makeup of the tumor in murine cutaneous melanoma is markedly different from human UM [11]. Orthotopic liver models of metastasis are preferred in UM because these allow the study of tumor-organ interactions in the microenvironment. Direct injection of the metastatic UM MUM2B cells into the liver parenchyma in SCID mice has resulted in the development of small liver nodules. This method, found to exhibit vasculogenic mimicry patterns of human laminin identical with primary and metastatic UM [67], has been used to screen for potential serum biomarkers of early hepatic metastatic disease. An orthotopic liver tumor xenograft model that was developed for UM utilized a novel surgical implantation technique wherein a pocket in the liver parenchyma was created to house the tumors. The implantation site was closed not with sutures, which require refined techniques, but with absorbable hemodynamic materials; this

rendered the surgery easier and reduced the risk of hemorrhage. The tumors implanted came from three samples: suspensions of human UM hepatic metastatic cell line tumors; tumors that developed from the initial xenograft; and ­patient-­ derived xenografts. Analysis revealed that the xenografts resembled the histopathology and preserved the genetic mutations and copy number variations present in the original hepatic metastasis of the patient [40]. This model is limited by the number of serial transplantations that can be done on recipient mice, as too many passages may cause genetic drifts. Also, given that human tissue is implanted into murine liver and that the interactions between the microenvironment and the tumor are species-­ specific, a chimeric mouse will be more useful in studying molecular interactions [40]. In addition, with this model, the natural progression of tumor initiation to the development of metastasis is ignored. As such, this model is best suited for the study of overt metastatic disease with regards to pathology, prognostic markers, and therapeutic targets. Intrasplenic injection of UM cells can also result in hepatic metastasis. This method is fairly simple to perform, is highly reproducible and has a low mortality rate. The cells are injected into the upper pole of the spleen and are allowed to enter the portal circulation after 1 minute, after which a splenectomy is completed. Hepatic tumors with

146

a

J. V. Burnier et al.

b

c

Fig. 9.6  Experimental model of GFP-transfected human UM cells in nude mice. Live-animal epifluorescence microscopy reveals GFP-labelled UM cells in the liver of nude mice at week 8 (a). Fluorescence confocal micros-

copy shows interaction between cultured GFP-transfected human UM cells and red-fluorescence protein (RFP)transfected hepatocytes (b and c)

vascularization form within 6  weeks [68]. This model has been used to study the mechanisms of metastasis in UM, such as the role of the chemokine receptors CXCR4 and CCR7 on the directional migration of tumor cells to the liver [69]. Metastatic lesions can also be developed through the orthotopic injection of melanoma cells into the eye. The advantage of such models is that they more closely resemble the natural history of the disease, as seen in the clinics. Transscleral injection into the posterior chamber of C7B5L6 mice results in pulmonary and hepatic metastases in 90% and 80% of mice,

respectively, after 4  weeks of enucleation [70]. Micrometastases have also been detected 4–6  weeks after posterior chamber inoculation of the UM cell line Mel290 transduced by lentiviral enhanced green fluorescent protein (EGFP) in athymic nude mice [42]. Suprachoroidal injection of the aggressive 92.1 UM cell line in immunosuppressed rabbits also resulted in overt pulmonary metastasis and micrometastasis in the liver [22] (Fig. 9.7). Finally, to mimic hematogenous dissemination of UM, cells can be introduced via intracardiac injection. Hamster melanoma cells have been

9  Animal Models in Uveal Melanoma

a

147

b

c

d

e

f

Fig. 9.7  Primary and metastatic human UM in an albino rabbit model. Funduscopic examination in week 10 reveals a large diffuse pigmented choroidal tumor (a). The corresponding gross pathology of the enucleated eye

shows the large and thick heavily pigmented choroidal tumor (b). Autopsy of the same animal reveals large metastatic pulmonary (c and d) and hepatic (e [arrow] and f) lesions by H&E (Arrow shows a hepatic metastasis)

injected into the left ventricle of albino rabbits, which resulted in the development of metastatic lesions in the liver, lung, and the kidney [71]. In another experiment, luciferase-positive OCM-1

cells were injected into the left ventricle and were monitored using whole-body bioluminescent reporter imaging. This technique successfully resulted in the development of metastasis in the

148

maxillofacial and femoral bones, as well as the lung and mediastinum [72]. These methods have been helpful in deepening our understanding of metastasis and in testing for potential therapeutic agents. Still, due to the complex biology of UM, care must be taken when interpreting data obtained from these experimental models [73].

Liquid Biopsy UM is one of the few cancers that relies on clinical examination and non-invasive imaging for diagnosis, without a tissue-based diagnosis. Fine needle aspiration biopsy (FNAB) is done to aid in prognostication through cytogenetic analysis. In recent years, cytogenetic analysis of UM has been used for the prognostication of tumors through classification into Class 1 or Class 2, which suggest low- and high-risk, respectively, for developing metastasis [1]. Chromosome 3 monosomy remains the single strongest cytogenic factor to predict UM metastasis. However, the use of tumor biomarkers and gene profiling is still limited in that it relies on tissue sampling by enucleation or biopsy. Liquid biopsy has emerged as a minimally invasive approach to detect and monitor disease progression, recurrence, and response to treatment [74] and to potentially replace the need for biopsy. Detection of circulating tumor cells (CTCs), circulating tumor DNA (ctDNA) and exosomes may contribute to early diagnosis and treatment [75]. It is also a valuable tool for longitudinal assessment of the genetic heterogeneity of cancer: recurrent or metastatic lesions may exhibit a different mutational profile than the primary lesion, and these changes may dictate the most effective treatment. Liquid biopsy can thus inform on the changing mutational status of the disease, and help to guide therapy [76]. CTCs have been found to provide prognostic information and monitor the response to therapy in patients with various metastatic carcinomas. In UM, CTCs have been isolated using different technical approaches [77–79]. However, the prognostic value of CTCs in UM remains unclear, and comparison between studies is difficult due to varying methodologies. Recent

J. V. Burnier et al.

studies have also focused on blood exosomes, which are small vesicles (30–100  nm) secreted by cells that contain functional biomolecules that reflect their cell of origin [80]. A previous study reported the ­exosomal miRNA profile in patients with UM, using vitreous humor and serum as samples [81]. In another study using plasma from liver perfusates, metastatic UM patients had a higher concentration of exosomes compared to healthy controls [82]. ctDNA are small fragments of DNA that are released by tumor cells into the circulation. Because blood, irrespective of health state, contains circulating free DNA (cfDNA) derived from the physiologic apoptosis of hematopoietic cells, evaluation of ctDNA must be done via detection of a tumor-specific mutation that identifies cfDNA that is of tumor origin. As such, malignancies with defined mutations, such as GNAQ/11 in UM, are good candidates for assessing the value of ctDNA to monitor disease. The quantity of ctDNA found in the blood has been correlated to tumor burden and cell turnover [83]. In one prospective study in UM, CTCs were found in 12 of 40 patients, and ctDNA (GNAQ/GNA11 mutations) in 22 of 26 patients with known mutations. CTC and ctDNA levels were associated with the presence and volume of metastasis as well as progression-free survival and overall survival [84]. Investigations into the applications of liquid biopsy in UM have mostly been done using samples acquired from patients. As research on liquid biopsy in UM advances, it is expected that animal and clinical studies will be done either sequentially or concurrently. Any model thus far discussed may serve suitably depending on the objectives set forth, the stage of disease that is of interest, and the type and amount of sample needed. Larger amounts of blood and aqueous samples, for instance, will require larger animals. One study used rabbits, in which 92.1 cells were injected into the suprachoroidal space, to confirm the presence of circulating malignant cells (see Fig. 9.2) [22]. In a follow-up analysis, CMCs were subjected to transcriptional profiling alongside the intraocular and metastatic tumors generated. Comparisons of these profiles identified significant changes in gene expression and potential key proteins regulating metastasis [27].

149

9  Animal Models in Uveal Melanoma

Treatment Strategies Several animal models have been used to assess the effect of different therapies in UM, and strategies to address metastatic disease have been described in a previous section. Historically, the first prominent animal model of UM used to evaluate the effectiveness of a treatment was in the Syrian golden hamster, as developed by Greene. The tumor was from a spontaneous, highly progressive melanotic melanoma that over serial passages transformed into an amelanotic melanoma and was eventually successfully grown in the eye. Subsequent subcutaneous implantation of the amelanotic melanoma tissue on the abdomen of a hamster resulted in tumor growth. It was on such a model that the effectiveness of transpupillary thermotherapy (TTT), as evidenced by the necrosis a day after treatment with infrared irradiation [85]. This work eventually led to the clinical use of TTT as an adjunct to treat UM.  Likewise, a modification of this model involving covering the tumor with donor scleral tissue enabled the evaluation of transscleral laser thermotherapy

[86]. Rabbit UM models, as described above, and using hamster and murine cutaneous melanoma cell lines, have been used to test the efficacy of adjuvant therapy such as neodymium-yttrium and argon endolasers and transscleral conductive heating [87, 88]. These, however, have yet to be validated in clinical studies. Recently, the light-activated drug AU-011 has been approved for clinical testing in UM patients. AU-011 is a small molecule conjugated to amines on the surface of a viral nanoparticle and, once injected into the vitreous, binds to the heparan sulfate proteoglycans (HSPG) on the UM cells. Irradiation of the molecule with infrared light results in rapid cellular necrosis. Prior to the commencement of clinical trials, this drug was first tested on two animal models: (1) in mice, with tail vein injection of 92.1 UM cells to simulate hematogenous dissemination; (2) and in rabbits, with 92.1 UM cells injected into the suprachoroidal space. Tumor uptake of AU-011 was noted in both models, with the drug exhibiting potency while sparing the retina and adjacent ocular structures [89] (Table 9.2).

Table 9.2  Summary of animal models used to study the disease stages in uveal melanoma Primary Ocular Disease Animal Cell Line Mouse B16LS9 (CM) B16LS9 (CM)

Inoculation Posterior chamber Posterior chamber (transcorneal) Rats C918, OCM-1 spheroids Suprachoroidal space Rabbits OCM-1 tumor fragments from Transscleral choroidal, suprachoroidal injection and xenotransplants, OCM-1 implantation through a suspension cyclodialysis cleft MKT-BR, OCM-1, 92-1, SP Suprachoroidal space 6.5 Natural History of Uveal Melanoma Animal Cell Line Inoculation Rabbit 92.1 Suprachoroidal space None; thickened, highly Mouse None; transgenic mouse Tg(Grm1) proliferative choroid Mouse Genetically modified, bears None; UM spontaneously GNAQ (Q209L) mutation developed Cancer Angiogenesis and Immunology Animal Cell Line Inoculation Rabbit OCM-1 tumor fragments from Transscleral choroidal, suprachoroidal injection and xenotransplants, OCM-1 implantation through a suspension cyclodialysis cleft

Metastasis Lung [90] Lung, liver, lymph node [70] ND [60] ND [61]

Lung [24]

Metastasis Lung, liver micrometastasis [22] ND [50] None [51]

Research Objective Microcirculation patterns using indocyanine green [61]

(continued)

150

J. V. Burnier et al.

Table 9.2 (continued) Mouse

B16LS9 (CM)

Rats

Suprachoroidal space

Choroidal implantation

Tumor blood volume using high-­ frequency contrast-enhanced ultrasound [63], role of PEDF [64], interferon-alpha [91], and natural killer T cells [92] RBC flux and hematocrit [20]

OCM-1 subcutaneous xenotransplanted tumor fragments Tumor Dormancy Animal Cell Line Mouse 92.1 (transfected with GFP)

Inoculation Tail vein injection

Metastasis None; liver homing of UM cells observed [65]

Metastasis Animal Cell Line Mouse 92.1, OMM1.3

Inoculation Subcutaneous

Metastasis OMM1.3 is a cell line derived from UM metastasis to the liver [66]. Lung, liver [11]

B16LS9 (CM) MUM2B, OCM1, M619 TJU-UM001,TJU-UM004 and PDX (tumor suspensions and fragments) B16 (CM) Mel290 OCM-1 Rabbit Hamster melanoma cells Liquid Biopsy Animal Cell Line Rabbit 92.1 Treatment Strategies Animal Cell Line Hamster Greene (hamster, amelanotic) Mouse 92.1 Rabbit

92.1

Posterior chamber (transcorneal technique) Liver parenchyma Liver parenchyma (liver pocket method)

Microscopic and expansile nodules in the liver, lung [67] Liver [40]

Intrasplenic injection Posterior chamber Left ventricle (hematogenous) Left ventricle (hematogenous)

Liver Liver micrometastasis [42] Bone, lung, mediastinum [72] Liver, lung, kidney [71]

Inoculation Suprachoroidal space

Observations Circulating malignant cells noted

Inoculation Subcutaneous Tail vein injection (hematogenous) Suprachoroidal injection

Treatment Modality Tested Transpupillary therapy [85] AU-011 [89] AU-011 [89]

CM cutaneous melanoma, ND not determined, PDX patient-derived xenografts

Conclusion and Future Directions UM is a complex disease, for which the underlying mechanisms driving tumor development, treatment response, and metastasis remain largely unknown. In order to advance our understanding of the biology of UM, animal models are needed. While no single ideal model exists, each approach has advantages and limitations, and understanding the limitations of each model and applying them appropriately to address specific scientific questions is essential. While syngeneic models may be best suited for immunologic and tumor biology aspects, xenografts

can help investigate treatment approaches and new therapeutics. In addition, there have been recent efforts to establish transgenic mouse models of spontaneous uveal melanoma that mimic human metastatic UM. Moving forward, future models should continuously aim to better imitate the clinical reality, with tumors that show similar genetics, histopathology, growth characteristics, hematogenous dissemination, and hepatic metastasis. Models in which animals can also be treated similarly to the clinic, by plaque radiation and enucleation, may also help us to better understand the natural history of the disease and the outcomes from the current treatment paradigm.

9  Animal Models in Uveal Melanoma

References 1. Harbour JW, Chen R.  The DecisionDx-UM gene expression profile test provides risk stratification and individualized patient care in uveal melanoma. PLoS Curr. 2013;5. 2. Stei MM, Loeffler KU, Holz FG, Herwig MC. Animal models of uveal melanoma: methods, applicability, and limitations. Biomed Res Int. 2016;2016:4521807. 3. Malho P, Dunn K, Donaldson D, Dubielzig RR, Birand Z, Starkey M.  Investigation of prognostic indicators for human uveal melanoma as biomarkers of canine uveal melanoma metastasis. J Small Anim Pract. 2013;54(11):584–93. 4. Planellas M, Pastor J, Torres MD, Pena T, Leiva M. Unusual presentation of a metastatic uveal melanoma in a cat. Vet Ophthalmol. 2010;13(6):391–4. 5. Giuliano EA, Chappell R, Fischer B, Dubielzig RR. A matched observational study of canine survival with primary intraocular melanocytic neoplasia. Vet Ophthalmol. 1999;2(3):185–90. 6. Grahn BH, Peiffer RL, Cullen CL, Haines DM.  Classification of feline intraocular neoplasms based on morphology, histochemical staining, and immunohistochemical labeling. Vet Ophthalmol. 2006;9(6):395–403. 7. Day MJ, Lucke VM. Melanocytic neoplasia in the cat. J Small Anim Pract. 1995;36(5):207–13. 8. Cao J, Jager MJ. Animal eye models for uveal melanoma. Ocul Oncol Pathol. 2015;1(3):141–50. 9. Grossniklaus HE, Barron BC, Wilson MW.  Murine model of anterior and posterior ocular melanoma. Curr Eye Res. 1995;14(5):399–404. 10. Kilian MM, Loeffler KU, Pfarrer C, Holz FG, Kurts C, Herwig MC.  Intravitreally injected HCmel12 melanoma cells serve as a mouse model of tumor biology of intraocular melanoma. Curr Eye Res. 2016;41(1):121–8. 11. Yang H, Brackett CM, Morales-Tirado VM, et  al. The Toll-like receptor 5 agonist entolimod suppresses hepatic metastases in a murine model of ocular melanoma via an NK cell-dependent mechanism. Oncotarget. 2016;7(3):2936–50. 12. Romanowska-Dixon B, Urbanska K, Elas M, Pajak S, Zygulska-Mach H, Miodonski A.  Angiomorphology of the pigmented Bomirski melanoma growing in hamster eye. Ann Anat. 2001;183(6):559–65. 13. Lambrou FH, Chilbert M, Mieler WF, Williams GA, Olsen K. A new technique for subchoroidal implantation of experimental malignant melanoma. Invest Ophthalmol Vis Sci. 1988;29(6):995–8. 14. Krause M, Kwong KK, Xiong J, Gragoudas ES, Young LH.  MRI of blood volume and cellular uptake of superparamagnetic iron in an animal model of choroidal melanoma. Ophthalmic Res. 2002;34(4):241–50. 15. Fabian ID, Rosner M, Fabian I, et  al. Low thyroid hormone levels improve survival in murine model for ocular melanoma. Oncotarget. 2015;6(13):11038–46.

151 16. Xue S, Yang H, Qiao J, et  al. Protein MRI contrast agent with unprecedented metal selectivity and sensitivity for liver cancer imaging. Proc Natl Acad Sci U S A. 2015;112(21):6607–12. 17. Rimoldi D, Salvi S, Lienard D, et  al. Lack of BRAF mutations in uveal melanoma. Cancer Res. 2003;63(18):5712–5. 18. Greene HS, Harvey EK. The growth and metastasis of amelanotic melanomas in heterologous hosts. Cancer Res. 1966;26(4):706–14. 19. Mueller AJ, Maniotis AJ, Freeman WR, et  al. An orthotopic model for human uveal melanoma in SCID mice. Microvasc Res. 2002;64(2):207–13. 20. Braun RD, Abbas A, Bukhari SO, Wilson W 3rd. Hemodynamic parameters in blood vessels in choroidal melanoma xenografts and rat choroid. Invest Ophthalmol Vis Sci. 2002;43(9):3045–52. 21. van der Ent W, Burrello C, Teunisse AF, et  al. Modeling of human uveal melanoma in zebrafish xenograft embryos. Invest Ophthalmol Vis Sci. 2014;55(10):6612–22. 22. Blanco PL, Marshall JC, Antecka E, et  al. Characterization of ocular and metastatic uveal melanoma in an animal model. Invest Ophthalmol Vis Sci. 2005;46(12):4376–82. 23. Blanco G, Saornil AM, Domingo E, et  al. Uveal melanoma model with metastasis in rabbits: effects of different doses of cyclosporine A.  Curr Eye Res. 2000;21(3):740–7. 24. Lopez-Velasco R, Morilla-Grasa A, Saornil-Alvarez MA, et  al. Efficacy of five human melanocytic cell lines in experimental rabbit choroidal melanoma. Melanoma Res. 2005;15(1):29–37. 25. Di Cesare S, Maloney S, Fernandes BF, et  al. The effect of blue light exposure in an ocular melanoma animal model. J Exp Clin Cancer Res. 2009;28:48. 26. Marshall JC, Fernandes BF, Di Cesare S, et  al. The use of a cyclooxygenase-2 inhibitor (Nepafenac) in an ocular and metastatic animal model of uveal melanoma. Carcinogenesis. 2007;28(9):2053–8. 27. Marshall JC, Nantel A, Blanco P, Ash J, Cruess SR, Burnier MN Jr. Transcriptional profiling of human uveal melanoma from cell lines to intraocular tumors to metastasis. Clin Exp Metastasis. 2007;24(5):353–62. 28. Odashiro AN, Pereira RP, Marshall J-C, Godeiro K, Burnier MN Jr. Skin cancer models. Drug Discov Today Dis Model. 2005;2(1):5. 29. Xu XL, Hu DN, Iacob C, et  al. Effects of zea xanthin on growth and invasion of human uveal melanoma in nude mouse model. J Ophthalmol. 2015;2015:392305. 30. Hu K, Babapoor-Farrokhran S, Rodrigues M, et  al. Hypoxia-inducible factor 1 upregulation of both VEGF and ANGPTL4 is required to promote the angiogenic phenotype in uveal melanoma. Oncotarget. 2016;7(7):7816–28. 31. Yu FX, Luo J, Mo JS, et  al. Mutant Gq/11 pro mote uveal melanoma tumorigenesis by activating YAP. Cancer Cell. 2014;25(6):822–30.

152 32. Barisione G, Fabbi M, Gino A, et  al. Potential role of soluble c-Met as a new candidate biomarker of metastatic uveal melanoma. JAMA Ophthalmol. 2015;133(9):1013–21. 33. Susskind D, Hurst J, Rohrbach JM, Schnichels S.  Novel mouse model for primary uveal melanoma: a pilot study. Clin Exp Ophthalmol. 2017;45(2):192–200. 34. van der Ent W, Burrello C, de Lange MJ, et  al. Embryonic Zebrafish: different phenotypes after injection of human uveal melanoma cells. Ocul Oncol Pathol. 2015;1(3):170–81. 35. Kalirai H, Shahidipour H, Coupland SE, Luyten G.  Use of the chick embryo model in uveal melanoma. Ocul Oncol Pathol. 2015;1(3):133–40. 36. Luyten GP, Mooy CM, De Jong PT, Hoogeveen AT, Luider TM.  A chicken embryo model to study the growth of human uveal melanoma. Biochem Biophys Res Commun. 1993;192(1):22–9. 37. Siolas D, Hannon GJ.  Patient-derived tumor xenografts: transforming clinical samples into mouse models. Cancer Res. 2013;73(17):5315–9. 38. Hidalgo M, Amant F, Biankin AV, et  al. Patient-­ derived xenograft models: an emerging platform for translational cancer research. Cancer Discov. 2014;4(9):998–1013. 39. Heegaard S, Spang-Thomsen M, Prause JU.  Establishment and characterization of human uveal malignant melanoma xenografts in nude mice. Melanoma Res. 2003;13(3):247–51. 40. Kageyama K, Ohara M, Saito K, et al. Establishment of an orthotopic patient-derived xenograft mouse model using uveal melanoma hepatic metastasis. J Transl Med. 2017;15(1):145. 41. Triozzi PL, Aldrich W, Singh A. Effects of interleukin-­1 receptor antagonist on tumor stroma in experimental uveal melanoma. Invest Ophthalmol Vis Sci. 2011;52(8):5529–35. 42. Yang H, Fang G, Huang X, Yu J, Hsieh CL, Grossniklaus HE.  In-vivo xenograft murine human uveal melanoma model develops hepatic micrometastases. Melanoma Res. 2008;18(2):95–103. 43. Barak V, Frenkel S, Valyi-Nagy K, et  al. Using the direct-injection model of early uveal melanoma hepatic metastasis to identify TPS as a potentially useful serum biomarker. Invest Ophthalmol Vis Sci. 2007;48(10):4399–402. 44. Richmond A, Su Y. Mouse xenograft models vs GEM models for human cancer therapeutics. Dis Model Mech. 2008;1(2–3):78–82. 45. Eyles J, Puaux AL, Wang X, et  al. Tumor cells disseminate early, but immunosurveillance limits metastatic outgrowth, in a mouse model of melanoma. J Clin Invest. 2010;120(6):2030–9. 46. Cheng Y, Zhang G, Li G.  Targeting MAPK pathway in melanoma therapy. Cancer Metastasis Rev. 2013;32(3–4):567–84. 47. Albert DM, Kumar A, Strugnell SA, et al. Effectiveness of 1alpha-hydroxyvitamin D2  in inhibiting tumor

J. V. Burnier et al. growth in a murine transgenic pigmented ocular tumor model. Arch Ophthalmol. 2004;122(9):1365–9. 48. Tolleson WH, Doss JC, Latendresse J, et  al. Spontaneous uveal amelanotic melanoma in transgenic Tyr-RAS+ Ink4a/Arf−/− mice. Arch Ophthalmol. 2005;123(8):1088–94. 49. Teh JL, Chen S.  Glutamatergic signaling in cellular transformation. Pigment Cell Melanoma Res. 2012;25(3):331–42. 50. Schiffner S, Braunger BM, de Jel MM, Coupland SE, Tamm ER, Bosserhoff AK.  Tg(Grm1) transgenic mice: a murine model that mimics spontaneous uveal melanoma in humans? Exp Eye Res. 2014;127:59–68. 51. Huang JL, Urtatiz O, Van Raamsdonk CD.  Oncogenic G protein GNAQ induces uveal melanoma and intravasation in mice. Cancer Res. 2015;75(16):3384–97. 52. Chen X, Wu Q, Tan L, et  al. Combined PKC and MEK inhibition in uveal melanoma with GNAQ and GNA11 mutations. Oncogene. 2014;33(39):4724–34. 53. Johansson P, Aoude LG, Wadt K, et al. Deep sequencing of uveal melanoma identifies a recurrent mutation in PLCB4. Oncotarget. 2016;7(4):4624–31. 54. Amaro A, Gangemi R, Piaggio F, et  al. The biology of uveal melanoma. Cancer Metastasis Rev. 2017;36(1):109–40. 55. Feng X, Degese MS, Iglesias-Bartolome R, et  al. Hippo-independent activation of YAP by the GNAQ uveal melanoma oncogene through a trio-­regulated rho GTPase signaling circuitry. Cancer Cell. 2014;25(6):831–45. 56. Patel BR, Tall GG. Ric-8A gene deletion or phorbol ester suppresses tumorigenesis in a mouse model of GNAQ(Q209L)-driven melanoma. Oncogene. 2016;5(6):e236. 57. Mouti MA, Dee C, Coupland SE, Hurlstone AF.  Minimal contribution of ERK1/2-MAPK signalling towards the maintenance of oncogenic GNAQQ209P-driven uveal melanomas in zebrafish. Oncotarget. 2016;7(26):39654–70. 58. Albert DM, Shadduck JA, Craft JL, Niederkorn JY.  Feline uveal melanoma model induced with feline sarcoma virus. Invest Ophthalmol Vis Sci. 1981;20(5):606–24. 59. Pe’er J, Folberg R, Massicotte SJ, et  al. Clinicopathologic spectrum of primary uveal melanocytic lesions in an animal model. Ophthalmology. 1992;99(6):977–86. 60. Braun RD, Vistisen KS.  Modeling human choroidal melanoma xenograft growth in immunocompromised rodents to assess treatment efficacy. Invest Ophthalmol Vis Sci. 2012;53(6):2693–701. 61. Mueller AJ, Folberg R, Freeman WR, et  al. Evaluation of the human choroidal melanoma rabbit model for studying microcirculation patterns with confocal ICG and histology. Exp Eye Res. 1999;68(6):671–8.

9  Animal Models in Uveal Melanoma 62. Ma D, Niederkorn JY.  Efficacy of tumor-infiltrating lymphocytes in the treatment of hepatic metastases arising from transgenic intraocular tumors in mice. Invest Ophthalmol Vis Sci. 1995;36(6):1067–75. 63. Zhang Q, Yang H, Kang SJ, et  al. In vivo high-­ frequency, contrast-enhanced ultrasonography of uveal melanoma in mice: imaging features and histopathologic correlations. Invest Ophthalmol Vis Sci. 2011;52(5):2662–8. 64. Yang H, Grossniklaus HE.  Constitutive overexpression of pigment epithelium-derived factor inhibition of ocular melanoma growth and metastasis. Invest Ophthalmol Vis Sci. 2010;51(1):28–34. 65. Logan PT, Fernandes BF, Di Cesare S, Marshall JC, Maloney SC, Burnier MN Jr. Single-cell tumor dormancy model of uveal melanoma. Clin Exp Metastasis. 2008;25(5):509–16. 66. Musi E, Ambrosini G, de Stanchina E, Schwartz GK.  The phosphoinositide 3-kinase alpha selective inhibitor BYL719 enhances the effect of the protein kinase C inhibitor AEB071  in GNAQ/GNA11-­ mutant uveal melanoma cells. Mol Cancer Ther. 2014;13(5):1044–53. 67. Folberg R, Leach L, Valyi-Nagy K, et  al. Modeling the behavior of uveal melanoma in the liver. Invest Ophthalmol Vis Sci. 2007;48(7):2967–74. 68. Lafreniere R, Rosenberg SA. A novel approach to the generation and identification of experimental hepatic metastases in a murine model. J Natl Cancer Inst. 1986;76(2):309–22. 69. Li H, Alizadeh H, Niederkorn JY. Differential expression of chemokine receptors on uveal melanoma cells and their metastases. Invest Ophthalmol Vis Sci. 2008;49(2):636–43. 70. Dithmar S, Rusciano D, Grossniklaus HE.  A new technique for implantation of tissue culture melanoma cells in a murine model of metastatic ocular melanoma. Melanoma Res. 2000;10(1):2–8. 71. Liu LH, Albert DM, Dohlman HG, Chuo N. Metastasis in a rabbit choroidal melanoma model. Invest Ophthalmol Vis Sci. 1982;22(1):115–8. 72. Notting IC, Buijs JT, Que I, et  al. Whole-body bioluminescent imaging of human uveal melanoma in a new mouse model of local tumor growth and metastasis. Invest Ophthalmol Vis Sci. 2005;46(5):1581–7. 73. Yang H, Cao J, Grossniklaus HE.  Uveal mela noma metastasis models. Ocul Oncol Pathol. 2015;1(3):151–60. 74. Huynh K, Hoon DS.  Liquid biopsies for assessing metastatic melanoma progression. Crit Rev Oncog. 2016;21(1–2):141–54. 75. Amirouchene-Angelozzi N, Schoumacher M, Stern MH, et  al. Upcoming translational challenges for uveal melanoma. Br J Cancer. 2015;113(12):1746. 76. Huang SK, Hoon DS. Liquid biopsy utility for the surveillance of cutaneous malignant melanoma patients. Mol Oncol. 2016;10(3):450–63. 77. Tura A, Luke J, Merz H, et al. Identification of circulating melanoma cells in uveal melanoma patients by

153 dual-marker immunoenrichment. Invest Ophthalmol Vis Sci. 2014;55(7):4395–404. 78. Mazzini C, Pinzani P, Salvianti F, et  al. Circulating tumor cells detection and counting in uveal melanomas by a filtration-based method. Cancers (Basel). 2014;6(1):323–32. 79. Charitoudis G, Schuster R, Joussen AM, Keilholz U, Bechrakis NE.  Detection of tumour cells in the bloodstream of patients with uveal melanoma: influence of surgical manipulation on the dissemination of tumour cells in the bloodstream. Br J Ophthalmol. 2016;100(4):468–72. 80. Caivano A, Laurenzana I, De Luca L, et  al. High serum levels of extracellular vesicles expressing malignancy-related markers are released in patients with various types of hematological neoplastic disorders. Tumour Biol. 2015;36(12):9739–52. 81. Ragusa M, Barbagallo C, Statello L, et  al. miRNA profiling in vitreous humor, vitreal exosomes and serum from uveal melanoma patients: Pathological and diagnostic implications. Cancer Biol Ther. 2015;16(9):1387–96. 82. Eldh M, Olofsson Bagge R, Lasser C, et al. MicroRNA in exosomes isolated directly from the liver circulation in patients with metastatic uveal melanoma. BMC Cancer. 2014;14:962. 83. Stroun M, Lyautey J, Lederrey C, Mulcahy HE, Anker P.  Alu repeat sequences are present in increased proportions compared to a unique gene in plasma/serum DNA: evidence for a preferential release from viable cells? Ann N Y Acad Sci. 2001;945:258–64. 84. Bidard FC, Madic J, Mariani P, et al. Detection rate and prognostic value of circulating tumor cells and circulating tumor DNA in metastatic uveal melanoma. Int J Cancer. 2014;134(5):1207–13. 85. Journee-de Korver JG, Oosterhuis JA, Kakebeeke-­ Kemme HM, de Wolff-Rouendaal D.  Transpupillary thermotherapy (TTT) by infrared irradiation of choroidal melanoma. Doc Ophthalmol. 1992;82(3): 185–91. 86. Rem AI, Oosterhuis JA, Korver JG, van den Berg TJ.  Transscleral laser thermotherapy of hamster Greene melanoma: inducing tumour necrosis without scleral damage. Melanoma Res. 2001;11(5):503–9. 87. Jaffe GJ, Mieler WF, Burke JM, Williams GA.  Photoablation of ocular melanoma with a high-powered argon endolaser. Arch Ophthalmol. 1989;107(1):113–8. 88. Rem AI, Oosterhuis JA, Journee-de Korver HG, de Wolff-Rouendaal D, Keunen JE. Transscleral thermotherapy: short- and long-term effects of transscleral conductive heating in rabbit eyes. Arch Ophthalmol. 2003;121(4):510–6. 89. Kines RC, Varsavsky I, Choudhary S, et al. An infrared dye-conjugated virus-like particle for the treatment of primary uveal melanoma. Mol Cancer Ther. 2018;17(2):565–74.

154 90. Diaz CE, Rusciano D, Dithmar S, Grossniklaus HE. B16LS9 melanoma cells spread to the liver from the murine ocular posterior compartment (PC). Curr Eye Res. 1999;18(2):125–9. 91. Yang H, Dithmar S, Grossniklaus HE. Interferon alpha 2b decreases hepatic micrometastasis in a murine model of ocular melanoma by activation of intrinsic

J. V. Burnier et al. hepatic natural killer cells. Invest Ophthalmol Vis Sci. 2004;45(7):2056–64. 92. Yang W, Li H, Mayhew E, Mellon J, Chen PW, Niederkorn JY. NKT cell exacerbation of liver metastases arising from melanomas transplanted into either the eyes or spleens of mice. Invest Ophthalmol Vis Sci. 2011;52(6):3094–102.

10

Iris Melanoma Arun D. Singh and Bertil E. Damato

Introduction Iris melanoma represents the least frequent of all uveal melanomas. Because of anterior location, iris melanoma is diagnosed and treated when relatively small compared to tumors located in ciliary body and choroid. Iris melanoma also tends to be less aggressive in comparison to melanomas in other uveal locations. Melanoma located in the peripheral iris may represent anterior extension of a ciliary body tumor, and therefore it is of paramount importance to examine the ciliary body in cases of iris melanoma when all iris margins of the tumors are not visualized.

Etiology and Pathogenesis Light-colored irides (blue-gray or green-hazel) are at a higher risk of developing melanoma [1, 2]. Oculo-dermal melanocytosis is also a known risk factor for iris melanoma (Fig. 10.1) [3]. Iris melanoma may arise de novo or from a preexist-

A. D. Singh (*) Department of Ophthalmic Oncology, Cole Eye Institute, Cleveland Clinic, Cleveland, OH, USA e-mail: [email protected] B. E. Damato Nuffield Department of Clinical Neurosciences, University of Oxford, John Radcliffe Hospital, Oxford, UK

ing nevus (Chap. 3). The malignant transformation rate of iris nevus into iris melanoma is estimated to be about 3% in 5 years [4]. Certain clinical features, such as inferior location, diffuse iris growth, episodes of hyphema, and feathery tumor margins predict a higher risk of growth or malignant transformation [4].

Pathology Similar to ciliary body and choroidal melanoma, iris melanoma is composed of spindle and epithelioid cells (Chap. 7). However, iris melanoma is predominantly of spindle-cell type, reflecting less aggressive clinical behavior [2]. On diagnostic cytology, iris melanoma cells have bland features compared with ciliary and choroidal melanomas [5]. In a comprehensive genetic analysis of iris melanomas, using a targeted next-generation sequencing approach, the mutational hotspot regions of nine genes known to be mutated in conjunctival and uveal melanoma (BRAF, NRAS, KIT, GNAQ, GNA11, CYSLTR2, SF3B1, EIF1AX, and BAP1) were assessed [6]. The high frequency of GNAQ and GNA11 mutations in iris melanomas (84%) with lack of BRAF, NRAS, and KIT mutations is similar to those observed in other uveal (ciliary body and choroidal) melanomas. One major difference is the higher rate of EIF1AX mutations in iris melanomas (42%) than in melanomas arising in other parts of the uvea

© Springer Nature Switzerland AG 2019 B. E. Damato, A. D. Singh (eds.), Clinical Ophthalmic Oncology, https://doi.org/10.1007/978-3-030-17879-6_10

155

156

a

A. D. Singh and B. E. Damato

b

c

d

Fig. 10.1  A 71-year-old female with partially discolored (dark) iris was noted to have a new “darker” lesion by her general ophthalmologist, who had periodically examined the patient for more than 5 years and performed her cataract surgery (uncomplicated) in November 2006. The patient had no complaints and had corrected visual acuity of 20/20 OU.  Examination was negative for eyelid and episcleral pigmentation. The left iris was blue. In the right eye, the infero-nasal 70% of the iris was brown, while the remaining iris was blue. Within the brown region of the iris, a darker iris lesion (4.0 × 1.0 mm) extending from 9

to 9:30 o’clock meridian at the pupillary margin was seen (a). The lesion effaced the iris crypts and was associated with ectropion uveae. Anterior segment OCT revealed a denser, circumscribed architecture within the lesion compared with surrounding iris stroma (b). A clinical diagnosis of ocular melanocytosis (nevus of Ota variant) with iris melanoma was made. The tumor was excised by a small incision approach. Histopathology confirmed spindle cell melanoma in association with an iris nevus (c). The surgical margins were clear. One month postoperatively, the eye was settling well (d)

(i.e., ciliary body or choroida) (13%), this mutation correlating with the favorable prognosis of iris melanoma [6]. In another study, the GNAQ A626C mutation (Q209P) was almost exclusively observed in choroidal melanomas (light exposed), whereas A > T mutations were clearly associated with ciliochoroidal location (not light exposed) suggesting light exposure as an etiologic factor [7]. The number of iris melanomas was too low (11) for meaningful statistical analysis. Using MLPA or MSA, iris melanoma displays a low-metastatic-risk chromosomal profile [8]. Similarly, gene expression profile analysis has

revealed that only one third of iris melanomas exhibit the class 2 gene expression profile [9] compared to more than 40% reported with ciliary body tumors [10]. Although such a profile indicates “high risk” of metastases, as yet, at median follow-up of 24 months, none of the 21 patients in the study developed metastases [9].

Clinical Features Iris melanoma may be circumscribed or diffuse. Slit-lamp examination, gonioscopy, and ultrasound biomicroscopy (UBM) allow staging of

10  Iris Melanoma

the tumor to guide the most appropriate treatment (Box 10.1) [11].

Box 10.1 Clinical Features of Iridociliary Melanoma • Circumscribed or diffuse • Secondary glaucoma, cataract, keratopathy, and hyphema • Annular growth • Seeding of cells in anterior chamber

Circumscribed iris melanoma has a nodular shape with variable pigmentation (Fig. 10.2). Iris melanoma tends to arise in the inferior half of the iris. It often has an irregular or rarely a smooth surface, covered by a surface plaque. In lightly pigmented tumors, the vessels are often visible (Fig. 10.3) [12, 13]. Iris melanoma can grow anteriorly into the anterior chamber, posteriorly towards the posterior segment, or in both directions. Iris melanoma involving the anterior chamber angle can subsequently invade the ciliary body, either locally or diffusely. In such cases, gonioscopy and UBM should be performed to examine the entire circumference of the ciliary body before making any treatment decisions. Posterior extension is a

157

limited by the lens, giving a flat, lion’s-paw appearance on UBM examination. Circumscribed iris melanomas can lead to anterior chamber hemorrhage, cataract, and, more rarely, corneal decompensation, with edema and band keratopathy [14]. Tumor progression can also sometimes be complicated by glaucoma, which can be secondary to invasion of the anterior chamber angle by the tumor, tumor necrosis with the accumulation of macrophages in the anterior chamber angle, or mechanical angle closure due to displacement of the lens. Diffuse iris melanoma can develop in two ways. The first consists of primary infiltration of the iris stroma [15]. The iris is thickened without any obvious nodule formation, and such growth is often associated with heterogeneous pigmentation and a deformed pupil. The intraocular pressure is usually increased as a result of invasion of the anterior chamber angle. The second mechanism consists of seeding of tumor cells from a circumscribed iris or ciliary body melanoma [16]. This phenomenon is often associated with progressive iris discoloration, with disappearance of the iris crypts and the accumulation of pigment in the anterior chamber angle. In either case, the onset of acquired hyperchromic heterochromia with ipsilateral secondary glaucoma should raise suspicion of a diffuse iris melanoma [17]. A delay in diagnosis is common, as these patients tend initially to be treated for glaucoma. b

Fig. 10.2  Circumscribed iris melanoma. The tumor is located in the inferior quadrant (a) and is elevated (b)

A. D. Singh and B. E. Damato

158

a

b

c

d

e

f

Fig. 10.3  A 20-year-old man was noted to have an amelanotic iris mass by his optometrist. The lesion was not present on examination a year previously (a). Fine intrinsic vessels and focal melanin pigment were observed upon closer examination (b). Anterior segment OCT showed surface reflectivity, but because of shadowing, the posterior margin of the tumor could not be assessed (c). A diagnosis of iridocilary melanoma was suspected and the

lesion excised by iridocyclectomy (d, H&E low magnification). The melanoma was of mixed cell type (e, H&E high magnification). The tumor stained uniformly with Melan-A (f). Ki-67 was also positive (g). A fresh tumor smear revealed monosomy 3 by fluoroscence in situ hybridization (h) both for centromeric and telomeric probes (i). Six weeks post-op appearance (j). Patient was then entered into adjuvant treatment trial

10  Iris Melanoma

159

g

h

i

j

3p25 Green (1G)

CEP8 Aqua (2A)

CEP3 Red (1R)

Fig. 10.3 (continued)

On gonioscopy, disseminated tumor cells accumulate predominantly in the inferior angle and have a brown, felty appearance, which distinguishes this condition from clumps of fine, dusty pigment observed in some cases of necrotic melanocytoma (Fig. 10.4). Melanomalytic glaucoma [18] responds poorly to medical therapies and tends to cause severe glaucomatous disc cupping and functional loss [19]. Diffuse iris melanoma tends to be of epithelioid cell type, with a higher risk of metastasis than the circumscribed variety [17].

Tapioca Iris Melanoma, a rare variant of diffuse iris melanoma. Tapioca melanoma can arise in tapioca nevus that can be multifocal or diffuse [20]. The description is based upon macroscopic appearance of translucent seeds that resemble tapioca pudding (Fig.  10.5) [21, 22]. The seeds are composed of melanoma cells that may be mixed or predominantly epithelioid cell type [23]. Although, metastasis of tapioca iris melanoma has been reported, it is not known whether this subtype of iris melanoma has a higher risk of metastases because of the tendency for local seeding [24].

160

A. D. Singh and B. E. Damato

Fig. 10.4  Diffuse iris melanoma associated with an annular ciliary body melanoma. Clinical and gonioscopic images

Ciliary Body Melanoma Ciliary body melanoma is often asymptomatic and may go undetected until large enough to cause secondary cataract, localized shallowing of the anterior chamber or glaucoma secondary to pigment shedding [25]. Ciliary body melanomas displace or infiltrate the root of the iris and invade the anterior chamber, where they become visible. At this stage, they can seed cells throughout the anterior chamber, onto the surface of the iris, and into the anterior chamber angle, causing elevated intraocular pressure. There are no specific features that differentiate ciliary body nevi from melanoma. Ciliary body nevi with dimensions ranging from 0.9 × 0.9 × 1.4 mm to 5 × 7 × 8 mm have been reported [26, 27]. Case series have also reported melanocytomas with sizes ranging from 3.5 to 8  mm wide to 1.5–3  mm thick [28, 29]. To date, the smallest ciliary body melanoma to have been reported was 4.5  ×  3.9  ×  2.0  mm [30]. Ciliary

body melanoma may be circumscribed or annular (ring). Slit-lamp examination, gonioscopy, transillumination, and ultrasound biomicroscopy (UBM) allow staging of the tumor to guide the most appropriate treatment (Box 10.1). Circumscribed ciliary body melanoma is a localized tumor that has a nodular shape. At the time of diagnosis, these tumors are generally larger than iris melanomas. In the early stages, they are confined to the ciliary body and are consequently asymptomatic (Fig.  10.6). They are generally brown, corresponding to the color of the overlying pigmented epithelium, unless this has been invaded by the tumor, in which case the true color of the tumor is visible. On UBM, they may have a homogeneous or heterogeneous structure and sometimes appear cavitated. Ring-shaped ciliary body melanoma spreads around the ciliary body in an annular fashion, and this growth pattern must be excluded in every case by UBM (Fig. 10.7) [31–33].

10  Iris Melanoma

161

a

b

c

d

e

f

Fig. 10.5  Whitish gelatinous mutinodular lesion, extending from 3:30 to 5 o’clock with slight peaking of the pupil, is observed (a). Fine intrinsic vascularity was evident on fluorescein angiography (b). Angle extension was suspected but could not be visualized by gonioscopy, anterior segment OCT (c), or ultrasound biomicroscopy (d). Diffuse amelanotic iris melanoma (tapioca variant) was confirmed by incisional biopsy (sector iridectomy) (e). On histopathology, a thin amelanotic spindle-cell neoplasm involving the superficial aspect of the iris comprising spindle cells with variably hyperchromatic nuclei with some nuclear pleomorphism was present (f, hematoxylin and eosin, original magnification × 40). Immunostains for SOX-10, melan A, and HMB-45 were positive (not

shown). The patient underwent proton beam radiation therapy receiving a total dose of 5320 cGy (1330 per fraction daily in 4 fractions). The target volume was the total anterior segment (g, maximum width 12.2  mm, anterior chamber depth 3.5 mm, height 1.0 mm). Staged rescue of a limbal stem cell autograft. Two areas for limbal stem cell and adjacent conjunctival epithelium harvest (h). Four weeks later, the grafts were in place (i and j). By 18  months after radiation therapy (k), pupilloplasty was performed to repair sectoral iridectomy in combination with phacoemulsification and insertion of a PCIOL (radiation-­ induced cataract) and glaucoma procedure (secondary glaucoma). (Reprinted from Singh et al. [112]. With permission from Wolters Kluwer Health, Inc.)

A. D. Singh and B. E. Damato

162

g

i

k

Fig. 10.5 (continued)

h

j

163

10  Iris Melanoma

a

b

c

d

e

Fig. 10.6  A 71-year-old Caucasian female was noted to have a small light-brown mass extending into the angle without iris deformity. On follow-up examination 16  months later a new finding of iris distortion was observed and the patient was referred for treatment. The patient was asymptomatic. The best-corrected visual acuity was 20/25 OD and 20/20 OS. Intraocular pressures and pupils were normal. Melanocytosis heterochromia and sentinel vessels were absent. The irides were blue. Slit-­ lamp examination and gonioscopy of the left eye revealed central displacement of the iris by a brown mass in the angle extending from 2:30 to 3:30 o’clock without intrinsic or feeder vessels (a, b). Localized pigment dispersion was present. A focal area of ciliary body thickening con-

fined to the pars plicata and measuring 2.0 × 1.0 × 1.1 mm was identified by ultrasound biomicroscopy (UBM) (c, d). Given the history of recent growth of the lesion ciliary body melanoma was suspected. Fine-needle aspiration biopsy (FNAB) was performed to confirm the diagnosis (e). Additional cytologic studies for prognostication by fluorescence in situ hybridization (FISH) using a centromeric probe for chromosome 3 (CEP3) revealed absence of monosomy 3. The patient had no evidence of metastatic disease by computed tomography of the chest abdomen and pelvis. Following a discussion of treatment options the patient underwent brachytherapy with an iodine-125 plaque

A. D. Singh and B. E. Damato

164

a c

b

d

Fig. 10.7  A 44-year-old male was followed for asymptomatic amelanotic iris nevus of the right eye (a) that was noted to have a localized ciliary body mass with ring extension along the trabecular meshwork (b). UBM of the right eye revealing a localized ciliary body lesion from 7 to 8 o’clock, with a maximum thickness of 2 mm (c). Fine needle aspiration biopsy was consistent with malignant melanoma. The patient underwent enucleation and remains disease-free at 10 years of follow-up. Histopathology revealed malignant melanoma involving the iris and ciliary body with a 360-degree extension

e

along the trabecular meshwork. The tumor was composed of a mixture of spindled and epithelioid cells with scant pigmentation of the main tumor mass. The melanoma is centered on the iris with extension into the ciliary body and trabecular meshwork (d, arrows, hematoxylin and eosin-stained section, 40× magnification). Microscopic tumor fragments are seen adherent to the angles 180° apart (e, arrows) consistent with the clinical impression of ring-like extension. (Reprinted from Aziz et al. [33]. Copyright © 2015, © 2015 S. Karger AG, Basel)

10  Iris Melanoma

Differential Diagnosis Iris and Ciliary Body Nevus Clinical and diagnostic features of iris and ciliary body nevi including melanocytoma are reviewed elsewhere (Chap. 3).

Iris and Ciliary Body Melanocytoma One feature suggestive of iris melanocytoma is a uniform, black appearance, which is rare in iris melanoma (Fig.  10.8) [34]. Additional features

a

165

differentiating this tumor from melanoma include the absence of intrinsic vascularity, ectropion uveae, and corectopia and the absence of episcleral sentinel vessels [34]. Although benign, ciliary body melanocytomas tend to extend into the anterior chamber and extraocularly, such extension can mimic extra-scleral growth of a ciliary body melanoma.

Iris Pigment Epithelial (IPE) Cyst In a large study of IPE cysts in 672 eyes, the cyst occurred most often in young adults (21–40 years)

b

c

d

Fig. 10.8  A 53-year-old female with a known iris nevus in the left eye for the past 20 years was evaluated as the referring physician noted growth (a). The visual acuity was 20/20 OU with normal intraocular pressure OU. The lesion was dark brown in color with sharp borders (2.2 mm × 2.0 mm × 1.5 mm). Gonioscopy demonstrated absence of angle extension and also absence of pigment dispersion (b). Surface nodularity was documented with anterior segment OCT (c). Ultrasound biomicroscopy confirmed the absence of involvement of the adjacent angle and ciliary body (d). Trans-corneal fine needle aspiration biopsy aspirate samples contain heavily pigmented

e

melanocytic cells (e). The cytoplasmic melanin partially obscures the nuclear detail rendering the distinction between melanoma and melanocytoma problematic. As iris melanoma could not be excluded the lesion was excised via sector iridectomy (f limbal-based scleral flap). Histopathology revealed a circumscribed densely pigmented lesion (g H&E stain low power). The cells have abundant melanin pigment (h H&E stain high power). Bleached section shows polyhedral cells with abundant cytoplasm bland round to ovoid nuclei with inconspicuous nucleoli (i). There is no mitotic activity. (Courtesy of Charles V. Biscotti, MD and Gabriella Yeaney, MD)

A. D. Singh and B. E. Damato

166

f

g

500 µm

h

i

Fig. 10.8 (continued)

with predominant midzonal or peripheral location [35]. Less than 1% of cases were associated with iris nevus, iris melanoma, or ciliary body melanoma. IPE cysts are usually noticed incidentally as an anterior elevation of the iris. Those arising near the pupillary margin may prolapse anteriorly (Fig. 10.9). Larger cysts may be visible as smooth

dome-shaped uniformly pigmented structures, best observed after pupillary dilation (Fig. 10.10). IPE can be readily imaged by ultrasound biomicroscopy [36–39]. The typical ultrasound biomicroscopic appearance is of a thin-walled cyst without internal reflectivity. IPE cysts rarely require intervention because complications such

167

10  Iris Melanoma

a

b

c

d

Fig. 10.9  A pigment epithelial (PE) cyst at the pupil margin (a). The cyst had prolapsed into the anterior chamber as confirmed by ultrasound biomicroscopic features (b).

Histopathology confirmed the clinical diagnosis of a PE cyst (c). At 1  month follow-up visit note residual thickened PE at the pupillary margin (d)

as cataract and glaucoma are infrequent (2, “probable melanoma” (Chap. 6). Uveal tumors are exceptional in the field of oncology in that most of them are treated without prior histologic confirmation of the diagnosis. This practice seems to work well enough in most cases and avoids the difficulties and risks of intraocular biopsy. Some patients demand proof of the diagnosis. Biopsy may also be useful if the patient

a

b

c

f

Dossage Quotient

biopsy in the first instance so as to achieve a diagnosis and commence treatment without delay. Another diagnostic challenge is the differentiation between a large choroidal nevus and a small melanoma, which is commonly labeled as “suspicious nevus” or “indeterminate melanocytic tumor.” Such a lesion is usually managed by sequential examinations, delaying treatment for months or years until growth is unequivocally documented. Ocular oncologists have long debated the safety of this practice, because it is not known when uveal melanomas start to metastasize [1]. When managing a patient with a melanocytic tumor of unknown malignancy, the requirement for informed consent makes it necessary to discuss such matters with the patient. Many patients are reluctant to delay treatment if there is any risk to life, however small, irrespective of any visual consequences, therefore opting for biopsy. The author has seen a patient

3

g

2

1 .5

d

2

e

1

C1 5q31.1 C2 5p13 C3 7q36 C4 7q31.2 C5 21q22.2 C6 12q23 C7 19p13.3 C8 5q35 C9 14q32 C10 12q24.33 C11 18p11.32 C12 20p12.2 1p36.22 1p36.13 1p.34 1p.34 1p.33 1p.31 1p11.2 3p25.3 3p24.3 3p25.3 3p22.1 3p22 3p21.3 3p14.2 3p14.2 3p12.2 3q12 3q21.3 3q25.1 3q29 6p25 6p25.2 6p21.2 6p21 6q23.1 6q26 8p12 8q11.23 8q24.12 8q24.12 8q24.2

.5

Locus

Fig. 11.2  Left fundus photograph of a 65-year-old woman with an inferonasal choroidal tumor measuring 10.1  mm basally with a thickness of 1.6 mm (a). Eight years later, the tumor suddenly grew rapidly to a thickness of 10.5  mm, and the eye was enucleated (b). Light microscopy showed the tumor to have a mushroom shape (c). The base of the

tumor was composed of spindle-B melanoma cells (d) of disomy 3 type (e), whereas the apical region showed epithelioid cells (f) with chromosome 3 loss and chromosome 8q gain (g). This patient subsequently developed fatal metastatic disease. (Reprinted from Callejo et al. [7]. With permission from American Medical Association)

11  Management of Patients with Posterior Uveal Melanoma

is to be left untreated or if an investigational procedure is undertaken. The author had a patient with a small, amelanotic melanoma beneath the papillomacular bundle. Photodynamic therapy was considered, but biopsy showed the tumor to have chromosome 3 loss, indicating metastatic potential. The patient therefore u­ nderwent immediate proton beam radiotherapy. As predicted, he developed metastatic disease, which could not be attributed to any therapeutic delays or failures. The development of the 25- and 27-gauge vitrectomy systems enables trans-­conjunctival, sutureless biopsy of uveal tumors to be performed quickly and easily as an alternative to fine-needle aspiration biopsy (Vol. 1, Chaps. 22 and 23) [8].

TNM Staging Tumor staging is fundamental to patient care in all fields of oncology, and uveal melanoma is no exception (Chap. 18). The TNM (tumor, node, metastasis) staging system of the American Joint Committee on Cancer is based on the anatomic extent of the ocular tumor as well as the presence or absence of nodal and systemic metastases. The AJCC TNM staging is designed to predict survival and not ocular outcomes. The ocular disease is categorized according to tumor size grouping (based on the largest basal tumor diameter and tumor thickness), ciliary body involvement, and extraocular spread [9]. Nodal involvement is extremely rare with uveal melanomas. Conventionally, staging is completed before ocular treatment is undertaken. This approach follows the logic that the detection of any metastases would influence patient care and, in particular, ocular treatment. However, metastases are only rarely detectable when the patient is first seen, except when the ocular tumor is very advanced, in which case enucleation is indicated whatever the staging, either because the eye is painful or because there is a significant risk of the eye becoming painful at a later date, requiring urgent treatment when the patient becomes unwell because of liver disease. Conversely, most patients with abnormal findings on staging turn out not to have metastases, their care and well-­being having been disturbed by the false results [10, 11]. For these reasons, the

189

author prefers to base patient care on prognostication, performed after ocular treatment, when more data are available (see below).

Ocular Treatment Objectives It is generally stated that, when treating uveal melanoma, the main priority is to save life; however, it is not known whether ocular treatment ever influences the patient’s survival, and, if so, in whom [1]. Despite enucleation or successful removal or irradiation of the tumor, almost 50% of patients with uveal melanoma develop metastatic disease [12]. There have been no randomized clinical trials comparing ocular treatment with no treatment with respect to prolonging life. The author suspects that early treatment prevents metastasis in some patients, and this hypothesis is based on his findings that patients who are older at the time of diagnosis and treatment tend to have larger tumors, with a higher prevalence of chromosome 3 loss and a higher metastatic mortality [13]. Although it is possible that uveal melanomas arising later in life are more lethal from their very inception, a more plausible explanation, in the author’s opinion, is that these tumors are more advanced and aggressive in older patients because they have been growing for a longer period. In an effort to conserve vision and the eye, it may be tempting to select radiotherapy instead of enucleation for a large tumor, or low-dose radiotherapy instead of full-dose radiotherapy for a medium-sized tumor, or photodynamic therapy instead of radiotherapy for a small, juxtapupillary tumor. It is widely believed that radiotherapy is as effective at prolonging life as enucleation, because the Collaborative Ocular Melanoma Study (COMS) reported no significant difference in survival between these two forms of treatment [14]. However, the COMS was statistically underpowered because patients developed metastatic disease soon after enrolment, which suggests that the tumor had already disseminated by the time it was treated [15]. Also, it is known that not all uveal melanomas metastasize, even if left untreated. The reported recurrence rate of around 10% after radiotherapy in the COMS cohort indicates that

B. E. Damato

190

Ocular Treatment Selection Previously, debate centered on which ocular treatment was best for all tumors. Today, there is a growing acceptance of the need to choose between the various methods according to the tumor size and location as well as the patient’s needs and wishes [11]. For example, the author’s first choice of treatment for choroidal melanomas is radiotherapy, selecting some form of local resection when any form of radiotherapy is likely to cause the toxic tumor syndrome or excessive collateral damage to healthy tissues. Phototherapy, being relatively unreliable, is rarely administered as a primary therapy, except for very small tumors on the understanding that radiotherapy may s­ubsequently be necessary. The author now treats all iris melanomas with proton beam radiotherapy, reserving resection for small, discrete, peripheral tumors that involve angle or ciliary body and that can be excised without damaging the iris sphincter. There is a growing tendency to follow a multimodality approach wherein different modes of treatment are combined to improve local tumor control while minimizing collateral damage to other parts of the eye (Fig. 11.3). For example, local resection is routinely combined with radiotherapy, to avoid local recurrence, and radiotherapy can be combined with some form of phototherapy, to reduce exudation. The author and associates have developed a scoring system for predicting ocular conservation with such a multimodality approach (Fig. 11.4) [18].

Fig. 11.3  Right fundus after multimodality treatment of a choroidal melanoma measuring 16.5  mm by 15.1  mm basally with a thickness of 12.2 mm. The patient had presented at the age of 46 years with a visual acuity of 20/30 with the affected eye and 20/100 with the left eye, which was amblyopic. The patient underwent exoresection by the author with adjunctive ruthenium plaque radiotherapy followed by transpupillary thermotherapy to the posterior edge of the irradiated area. Laboratory studies showed the tumor to be of mixed-cell type with no evidence of chromosome 3 loss. Eleven years later the patient was well with a visual acuity of 20/25 in the operated eye

a 1

1 2

.8 Eyes conserved

only about 60 patients developed this complication; this number is probably too small to significantly influence mortality in the entire group of patients treated with a plaque. Several studies have shown that local recurrence after radiotherapy is associated with higher metastatic mortality [16, 17]. It is still not known, however, whether metastatic disease is the result of persistent viable tumor or whether the recurrence is merely an indicator of greater tumor malignancy. As with any condition, the objectives of treatment should be defined and agreed with the patient; however, it can be challenging to meaningfully communicate with the patient the current uncertainties regarding the impact of ocular therapy on survival.

3 4

.6

.4

5

.2

0

0

2

4

6

8

10

Follow – up (years)

Fig. 11.4 (a) Kaplan-Meier analyses for the cumulative probability of secondary enucleation after primary conservative therapy, according to predictive score, shown at the end of each curve. The scoresheet (b) is specific to the ocular oncology center in Liverpool, but similar principles could be used to derive scoring systems elsewhere. (Reprinted from Damato and Lecuona [18]. With permission from Elsevier)

191

11  Management of Patients with Posterior Uveal Melanoma Fig. 11.4 (continued)

b Variable

Category

Score

Disk involvement

No

0

Yes

0.8

Temporal

0

Nasal/midline

1

7

0.2

8

0.3

9

0.4

10

0.5

11

0.6

12

0.7

13

0.8

14

0.9

15

1.0

16

1.1

17 18

1.2 1.3

19

1.4

20

1.5

21

1.6

22

1.7

23

1.8

4

1.0

5

1.3

6

1.6

7

1.9

8

2.1

9

2.4

10

2.7

Coronal location

Tumor diameter (mm)

Tumor height (mm)

Total score Each variable is scored by encircling the appropriate row, and scores are then added to provide a total.

Radiotherapy Several forms of radiotherapy have been developed to maximize the dose of radiation to the tumor while reducing collateral damage to the surrounding tissues [19]. Episcleral Plaque Radiotherapy In most centers, when applicable, the first choice of treatment is episcleral brachytherapy, adminis-

tered with a radioactive plaque containing ruthenium-­ 106 or iodine-125 (Chap. 12) [20]. Ruthenium plaques are suitable for uveal melanomas up to 5 mm in thickness, because the beta radiation they emit has a limited range. To reduce collateral damage to the optic disc and fovea, the author has developed a technique for positioning the plaque eccentrically, with its posterior edge aligned with the posterior tumor margin

192

B. E. Damato

(Fig. 11.5) [21]. Iodine plaques emit gamma irradiation and can successfully treat tumors as thick as 10  mm; however, they deliver large doses of radiation to healthy ocular structures [19].

Fig. 11.5  Left fundus of a 49-year-old woman presenting with an inferotemporal choroidal melanoma measuring 9.6 mm by 10.7 mm basally with a thickness of 3.4 mm. The visual acuity with the affected eye was 20/30. The patient was treated with an eccentrically placed ruthenium-106 plaque. Six years after treatment the visual acuity of the treated eye was 20/20-2 with good local tumor control. This is because sufficient radiation was delivered beyond the tumor without damaging the fovea

a

Collimation and dosimetry techniques have been improved to avoid this problem [22, 23]. Proton Beam Radiotherapy Proton beam radiotherapy enables a high dose of radiation to be aimed precisely at a uveal melanoma irrespective of the tumor’s size, shape, and location (Chap. 13). Facilities for this treatment are available in only a small number of centers around the world. Some oncologists use proton beam radiotherapy for all choroidal melanomas; others reserve it for tumors that cannot adequately be treated by brachytherapy, that is, tumors that are large and those extending close to the optic disc or fovea. Conventionally, proton beam radiotherapy is administered with a 2.0–2.5 mm safety margin. However, for many years, the author has treated small, posterior tumors with a narrower posterior safety margin, without recurrence, thereby improving vision (Fig. 11.6). Conversely, large, anterior tumors require a wider safety margin anteriorly because of their tendency to grow circumferentially along the ciliary body. Irradiation of the upper eyelid margin causes painful keratinization of the tarsal conjunctiva. The author and associates have therefore developed protocols for treating superior choroidal melanomas through the closed eye whenever it is

b

Fig. 11.6  Left fundus of a 58-year-old woman with a nasal choroidal melanoma measuring 12.8  mm by 11.4  mm basally with a thickness of 4.4 mm, involving 2 clock hours of the optic disc and causing bullous retinal detachment (a).

The patient was treated with proton beam radiotherapy and transpupillary thermotherapy. More than 6 years later, the visual acuity of the treated eye was 20/40. The tumor thickness was less than 1 mm and the retina was flat (b)

11  Management of Patients with Posterior Uveal Melanoma

not possible to retract the upper eyelid entirely out of the radiation beam [24]. This policy has been followed for more than 20 years without any increase in the local tumor recurrence rate. Proton beam radiotherapy of large tumors is often followed by neovascular glaucoma. Some authors believe that this complication is caused by collateral damage to healthy ocular tissues and have therefore attempted to avoid this complication by hyper-fractionation or by reducing the amount of radiation delivered to the iris and ciliary body [25]. The author considers neovascular glaucoma to be caused by persistence of the irradiated tumor, which becomes ischemic and exudative (“toxic tumor syndrome”) and which is therefore treated by local tumor resection [26, 27]. Others are investigating intraocular injections of antiangiogenic agents administered prophylactically [28]. Prediction of visual outcome after intravitreal therapy has improved, thanks to the use of optical coherence tomography angiography [29]. Stereotactic Radiotherapy With stereotactic radiotherapy, a highly collimated beam of photons or gamma radiation is aimed at the tumor from many different ­directions so that a high dose of radiation is delivered to the melanoma while exposing healthy tissues to small doses of radiation (Chap. 14) [30]. This method is generally used as an alternative to proton beam radiotherapy in centers where a cyclotron unit is not available. In some centers, it is administered as a neoadjuvant therapy before endoresection [31].

193

[32]. This treatment has been advocated for tumors up to 4  mm in thickness. Adjunctive brachytherapy is recommended as a means of avoiding local tumor recurrence, which occurs in approximately 20% of patients after thermotherapy alone [33]. After radiotherapy, transpupillary thermotherapy of the tumor can be an effective treatment for exudative retinal detachment involving macula and macular edema [34, 35]. Photodynamic Therapy Photodynamic therapy using verteporfin has recently been described, but it is still too soon to assess the efficacy of this treatment, both as a primary therapy and as an adjunctive treatment for radiation-induced exudation (Chap. 15) [36]. A novel form of photodynamic therapy under investigation involves intravitreal injection of a recombinant papillomavirus-like particle linked to a phthalocyanine photosensitizer activated by near-­ infrared light [37].

Cryotherapy Cryotherapy has not gained widespread acceptance as a treatment for choroidal melanoma despite some encouraging case reports [38]. Tumor Resection

Photocoagulation Photocoagulation of choroidal melanoma is associated with a high complication rate and has been superseded by transpupillary thermotherapy (Chap. 15).

Trans-scleral Resection Trans-scleral resection of small, ciliary body melanomas has been performed for many years (Chap. 16) [26]. Advances in microsurgery and hypotensive anesthesia have also made it possible to remove large tumors extending as far posteriorly as the fovea (Fig.  11.7). Such surgical procedures are difficult and are therefore performed only in a few centers, where they are reserved for tumors that are considered too large for radiotherapy. Trans-scleral resection of an irradiated melanoma can induce regression of exudative retinal detachment and neovascular glaucoma [27].

Transpupillary Thermotherapy With transpupillary thermotherapy, the tumor is heated by only a few degrees for about 1 min by means of a 3  mm diode laser beam (Chap. 15)

Endoresection With endoresection, the tumor is removed with a vitreous cutter, either through a hole in the retina or after raising a retinal flap (Chap. 16) [39]. This

Phototherapy

194

B. E. Damato

a

Fig. 11.7  Right fundus of a 29-year-old man with an inferonasal choroidal melanoma measuring 13  mm by 11 mm basally with a thickness of 7 mm (a). Exoresection

a

b

was performed by the author in 1988. More than 25 years postoperatively, the patient was well, and the operated eye had a visual acuity of 20/20 with no complications (b)

b

Fig. 11.8  Right fundus appearance after endoresection of a nasal choroidal melanoma by the author. When the patient presented, at the age of 66  years, the tumor had measured 9.7 mm by 8.0 mm basally with a thickness of 6.4 mm and had extended to the optic disc (a). Laboratory

studies showed the melanoma to be of mixed-cell type with no evidence of chromosome 3 loss. More than 6 years after surgery, the visual acuity of his operated eye was 20/40, and there was no evidence of recurrence (b)

is a controversial procedure because of concerns about seeding of malignant cells to other parts of the eye as well as the orbit and systemically [40]. For this reason, endoresection is rarely performed, except perhaps for juxtapapillary tumors when other methods are unlikely to conserve vision (Fig.  11.8). Some surgeons administer adjunctive radiotherapy after endoresection; others prefer neoadjuvant proton beam or stereotactic radiotherapy (Fig. 11.8) [31, 41].

Enucleation Primary enucleation for uveal melanoma is now performed only when other methods are considered unlikely to conserve the eye and useful vision without causing excessive morbidity and/ or if the patient is not motivated to try to save the eye. The author currently performs primary enucleation in about 30% of all patients with uveal melanoma, mostly when a patient presents at a late stage [18]. The enucleation is performed in a

11  Management of Patients with Posterior Uveal Melanoma

standard fashion, using the surgeon’s preferred implant. To ensure that the correct eye is removed, the tumor is visualized by binocular indirect ­ophthalmoscopy, which is done only after draping the patient and covering the other eye. A retrobulbar injection of a long-acting local anesthetic agent with adrenaline (1:80,000) is administered even when the enucleation is performed under general anesthesia as a means of reducing intraoperative hemorrhage and postoperative pain.

Ancillary Treatments Overall outcomes after treatment of uveal melanoma are improving due to advances made in the treatment of cataract, glaucoma, iris coloboma, and rhegmatogenous retinal detachment. Antiangiogenic factors are now widely used to treat macular edema and other radiation-induced vasculopathies [28, 35]. These have largely superseded intraocular steroid injections, which tend to cause cataract and glaucoma.

Prognostication Many clinical, histologic, and genetic factors are associated with an increased risk of metastasis, and the number of known predictors is growing from year to year (e.g., BAP1 abnormality) [42]. Until recently, prognostication was mostly based on one predictive factor (e.g., largest basal tumor diameter, monosomy 3, or class 2 gene expression profile). Such univariate analysis gives only an approximate estimate of the probability of metastatic disease and the likely survival time. The author and associates have developed mathematical methods for integrating the TNM stage with histologic and genetic data, thereby enhancing the reliability of prognostication so that it is relevant to individual patients (Chap. 19) [43]. Others have recently adopted this approach [44, 45]. In patients undergoing radiotherapy, this prognostication requires tumor biopsy, which is performed trans-sclerally or trans-retinally, either before radiotherapy or soon after [46]. A successful biopsy requires surgical expertise as well as a highly skilled laboratory team. The laboratory techniques vary greatly between centers and are developing rapidly.

195

Screening for Systemic Disease There is no consensus about which patients should be screened for metastatic disease, what investigations should be performed, how frequently they should be undertaken, and for how long (Chap. 22). There is also disagreement as to whether all patients or only high-risk individuals should be screened for systemic metastases after ocular treatment. It is generally accepted that some form of liver imaging is required, ideally every 6–12  months [47]. Opinions are divided as to whether ultrasonography is adequate or whether other methods such as magnetic resonance imaging are superior [48, 49]. There is also disagreement as to whether any MRI should be performed with contrast and diffusion weighting. Repeated computerized tomography may expose patients to dangerous levels of radiation after many years [49]. Some authors have advocated positron emission tomography (PET) [50]; however, others have reported low sensitivity with this imaging [51]. Chest radiography only rarely detects metastases in the absence of liver tumors, and liver function tests become abnormal only when metastases are advanced [52]. With screening, metastases are detected before the onset of symptoms in more than 90% of patients [48]. Such early detection has created opportunities for clinical trials evaluating a variety of novel treatments.

Systemic Therapy Symptomatic metastatic disease from uveal melanoma only rarely responds to therapy [53, 54]. Uveal melanomas are less responsive to immune checkpoint inhibitors than cutaneous melanomas [55]. Encouraging responses have recently been reported with IMCgp100 (Tebentafusp, Immunocore, Oxford, UK). This consists of bispecific “ImmTAC” molecules, which bind to gp100 linked to HLA peptides on the tumor cells and which have an anti-CD3 antibody that recruits circulating T cells, which then release lytic granules that kill the tumor cells. (Carvajal, R. et al. Poster presentation, Society for Immunotherapy

B. E. Damato

196

of Cancer Conference, Maryland, USA, November 2017). There would seem to be scope for investigating adjuvant systemic therapy for high-risk patients, starting this treatment as soon as possible after treatment of the primary tumor. Possible therapies include systemic chemotherapy, ipilimumab, antiangiogenic agents, histone deacetylases, and Cox-2 inhibitors, although each of these treatments is associated with life-threatening complications. Large, multicenter, randomized, prospective studies are required to evaluate the efficacy and safety of such agents [56].

Aftercare After treatment of the primary tumor, patients require continuous ophthalmic care to ensure that any ocular complications are detected and treated without delay. There seems to be a general consensus that follow-up should be lifelong; however, the frequency of ocular examinations varies greatly between centers. Much depends on the perceived risk of complications in each individual case. Apart from surveillance, it is useful to review patients regularly to answer any questions they may have and to provide psychological support if necessary. Follow-up examinations at an ocular oncology center also enhance outcomes analysis; however, patients should be informed if they are being reviewed solely for this purpose.

and these individuals can be helped by a proactive telephone call with advice and support from a psychologist or nurse. Additional psychological help may be required if the patient suffers visual loss, cosmetic deformity, or reduced life expectancy. It is important to provide support to the patient’s close relatives, who play an important role in the patient’s care and who themselves benefit from any encouragement and recognition. In some countries, there are patient organizations such as CureOM, Ocular Melanoma Foundation (OMF), and A Cure In Sight (ACIS) in the United States and OcuMel UK in the United Kingdom. Many patients find it helpful to join such associations and should therefore be informed of the existence of these bodies. These organizations are also proving helpful to ocular oncology centers by sponsoring research, organizing educational events, and undertaking surveys of patients’ satisfaction with their care [5].

Organization of Patient Care Increasingly, patients with uveal melanoma are being managed by a multidisciplinary team, composed of ocular oncologist, general oncologist, radiation oncologist, pathologist, and psychologist. Ocular oncology services may also provide logistic assistance as well as information leaflets, Internet support, and a telephone helpline.

Counseling

Future Trends

Patients with uveal melanoma can have special psychological needs, which may change as they progress through their care pathway [5, 57]. When their tumor is first diagnosed, they may benefit from psychological counseling aimed at strengthening their coping mechanisms. Patients may also need assistance when selecting the most appropriate kind of treatment, if any choice exists [11]. The author routinely gives every new patient an audio-recording of the actual conversation after discussing the diagnosis, prognosis, and therapeutic options [58]. After returning home from their initial treatment, some patients become depressed,

Predictions regarding future developments may in retrospect prove to be quite mistaken. Nevertheless, some trends are worth considering in this chapter.

Non-mydriatic Fundus Cameras The more widespread use of non-mydriatic and wide-angle fundus cameras in the community is likely to increase the detection of melanocytic choroidal tumors, many of which will be of indeterminate malignant potential. Although many

11  Management of Patients with Posterior Uveal Melanoma

patients are observed until growth is documented, the risk of delaying treatment is not known [1, 13]. In such cases, the management of uncertainty will be influenced more strongly by patients, especially if there is shared responsibility for dealing with the unknown risk of delaying treatment. There is likely to be a greater demand for tumor biopsy. With such difficult cases, multicenter randomized and nonrandomized trials are necessary to enroll the large number of patients required to determine the correct management.

Intraocular Tumor Biopsy and Survival Prognostication Tumor biopsy is likely to play a greater role in the management of patients with clinically diagnosed uveal melanoma, this investigation being performed primarily to determine the genetic type and grade of malignancy and hence the prognosis regarding local tumor control and metastasis-free survival [43]. Such information may also influence ocular treatment. For example, radiation safety margins may be adjusted according to the tumor activity.

Adjuvant Systemic Therapy The availability of highly accurate survival prognostication will inevitably influence patient care (Chap. 21). Unless the treatment of detectable metastases improves, patients with a high-grade melanoma will probably create a demand for adjuvant systemic therapy. The development of promising new agents would make such treatment more attractive than it is at present [59]. For these reasons, we are likely to see greater efforts at undertaking multicenter studies evaluating prognostic scores, screening for metastatic disease, and adjuvant systemic therapy.

Improved Visual Outcomes It would be reasonable to expect improved visual results after radiotherapy, due to a better under-

197

standing of the pathophysiology of radiation-­ induced complications as well as greater use of treatments such as phototherapy or local resection of the irradiated “toxic tumor” and intraocular injections of antiangiogenic agents and steroids [28, 34].

Increased Patient Expectations Patients’ expectations regarding local tumor control, visual outcome, and other aspects of care are likely to increase, especially if treatment results and patient satisfaction scores become more readily available on the Internet. This may influence referral practices, with the referring ophthalmologists selecting an ocular oncology center according to the comparative data at their disposal.

Clinical Trials Enrollment of patients in clinical trials not only enhances progress but also tends to improve the quality of care received by the patients participating in such studies. There are many unresolved questions regarding uveal melanomas and their treatment (e.g., juxtapapillary tumors, iris tumors, extraocular extension, etc.), and there is much scope for multicenter collaboration. Participation of a patient in outcome assessments and formal clinical trials would increase if funding for such studies were somehow to be incorporated into treatment fees, thereby reducing the reliance on the whims of grant awarding bodies. The prospects for multicenter studies are improving, thanks to the development of organizations such as the International Society of Ocular Oncology, the European Ophthalmic Oncology Group, the American Association for Ophthalmic Oncologists and Pathologists, and the Asia Pacific Society of Ocular Oncology and Pathology.

Conclusions These trends, together with the formation of new websites and discussion groups on the Internet, will probably raise standards so that patients and

198

B. E. Damato

and prognostic factors: COMS report No. 28. Arch Ophthalmol. 2006;124:1684–93. 15. Damato B. Legacy of the collaborative ocular melanoma study. Arch Ophthalmol. 2007;125:966–8. 16. Gragoudas ES, Lane AM, Munzenrider J, et al. Long-­ term risk of local failure after proton therapy for choroidal/ciliary body melanoma. Trans Am Ophthalmol Soc. 2002;100:43–8; discussion 48–49. References 17. Karlsson UL, Augsburger JJ, Shields JA, et  al. Recurrence of posterior uveal melanoma after 60Co episcleral plaque therapy. Ophthalmology. 1. Damato B. Ocular treatment of choroidal melanoma 1989;96:382–8. in relation to the prevention of metastatic death  - a 18. Damato B, Lecuona K.  Conservation of eyes with personal view. Prog Retin Eye Res. 2018;66:187. choroidal melanoma by a multimodality approach to 2. Damato EM, Damato BE. Detection and time to treattreatment: an audit of 1632 patients. Ophthalmology. ment of uveal melanoma in the United Kingdom: 2004;111:977–83. an evaluation of 2,384 patients. Ophthalmology. 19. Stannard C, Sauerwein W, Maree G, et al. Radiotherapy 2012;119:1582–9. for ocular tumours. Eye. 2013;27:119–27. 3. Bove R, Char DH. Nondiagnosed uveal melanomas. 20. Reichstein D, Karan K. Plaque brachytherapy for posOphthalmology. 2004;111:554–7. terior uveal melanoma in 2018: improved techniques 4. Eskelin S, Kivela T.  Mode of presentation and time and expanded indications. Curr Opin Ophthalmol. to treatment of uveal melanoma in Finland. Br J 2018;29:191–8. Ophthalmol. 2002;86:333–8. 5. Afshar AR, Deiner M, Allen G, et  al. The patient’s 21. Russo A, Laguardia M, Damato B.  Eccentric ruthenium plaque radiotherapy of posterior choroidal experience of ocular melanoma in the US: a survey of melanoma. Graefes Arch Clin Exp Ophthalmol. the ocular melanoma foundation. Ocul Oncol Pathol. 2012;250:1533–40. 2018;4:280–90. 6. Damato BE.  Tumour fluorescence and tumour-­ 22. Astrahan MA, Luxton G, Jozsef G, et  al. An interactive treatment planning system for ophthalmic associated fluorescence of choroidal melanomas. Eye. plaque radiotherapy. Int J Radiat Oncol Biol Phys. 1992;6(Pt 6):587–93. 1990;18:679–87. 7. Callejo SA, Dopierala J, Coupland SE, et al. Sudden growth of a choroidal melanoma and multiplex 23. Le BHA, Kim JW, Deng H, et al. Outcomes of choroidal melanomas treated with eye physics plaques: a ligation-­dependent probe amplification findings sug25-year review. Brachytherapy. 2018;17:981–9. gesting late transformation to monosomy 3 type. Arch 24. Konstantinidis L, Roberts D, Errington RD, et  al. Ophthalmol. 2011;129:958–60. Transpalpebral proton beam radiotherapy of choroidal 8. Sen J, Groenewald C, Hiscott PS, et al. Transretinal melanoma. Br J Ophthalmol. 2015;99:232–5. choroidal tumor biopsy with a 25-gauge vitrector. 25. Daftari IK, Char DH, Verhey LJ, et al. Anterior segOphthalmology. 2006;113:1028–31. ment sparing to reduce charged particle radiotherapy 9. Kivela TT, Piperno-Neumann S, Desjardins L, et al. complications in uveal melanoma. Int J Radiat Oncol Validation of a prognostic staging for metastatic uveal Biol Phys. 1997;39:997–1010. melanoma: a collaborative Study of the European Ophthalmic Oncology Group. Am J Ophthalmol. 26. Damato BE, Stewart JM, Afshar AR, et  al. Surgical resection of choroidal melanoma. In: Schachat AP, 2016;168:217–26. editor. Ryan’s retina, vol. 3. Philadelphia: Elsevier; 10. Feinstein EG, Marr BP, Winston CB, et  al. Hepatic 2018. p. 2591–600. abnormalities identified on abdominal computed 27. Konstantinidis L, Groenewald C, Coupland SE, tomography at diagnosis of uveal melanoma. Arch et  al. Trans-scleral local resection of toxic choroiOphthalmol. 2010;128:319–23. dal melanoma after proton beam radiotherapy. Br J 11. Damato B, Heimann H.  Personalized treatment of Ophthalmol. 2014;98:775–9. uveal melanoma. Eye. 2013;27:172–9. 12. Kujala E, Makitie T, Kivela T. Very long-term progno- 28. Reichstein D. Current treatments and preventive strategies for radiation retinopathy. Curr Opin Ophthalmol. sis of patients with malignant uveal melanoma. Invest 2015;26:157–66. Ophthalmol Vis Sci. 2003;44:4651–9. 1 3. Damato BE, Heimann H, Kalirai H, et  al. Age, sur- 29. Matet A, Daruich A, Zografos L. Radiation maculopathy after proton beam therapy for uveal melanoma: vival predictors, and metastatic death in patients optical coherence tomography angiography alterawith choroidal melanoma: tentative evidence of a tions influencing visual acuity. Invest Ophthalmol Vis therapeutic effect on survival. JAMA Ophthalmol. Sci. 2017;58:3851–61. 2014;132:605–13. 30. Zehetmayer M.  Stereotactic photon beam irra 14. Collaborative Ocular Melanoma Study G. The COMS diation of uveal melanoma. Dev Ophthalmol. randomized trial of iodine 125 brachytherapy for 2012;49:58–65. choroidal melanoma: V.  Twelve-year mortality rates

their families will expect more comprehensive care, which, in addition to treating the ocular tumor, adequately addresses a wide variety of social, spiritual, and psychological needs.

11  Management of Patients with Posterior Uveal Melanoma 31. Schilling H, Bornfeld N, Talies S, et al. [Endoresection of large uveal melanomas after pretreatment by single-­ dose stereotactic convergence irradiation with the leksell gamma knife--first experience on 46 cases]. Klinische Monatsblatter fur Augenheilkunde. 2006;223:513–20. 32. Journee-de Korver JG, Oosterhuis JA, Kakebeeke-­ Kemme HM, et  al. Transpupillary thermotherapy (TTT) by infrared irradiation of choroidal melanoma. Doc Ophthalmol. 1992;82:185–91. 33. Shields CL, Shields JA, Perez N, et  al. Primary transpupillary thermotherapy for small choroidal melanoma in 256 consecutive cases: outcomes and limitations. Ophthalmology. 2002;109:225–34. 34. Damato B. Vasculopathy after treatment of choroidal melanoma. In: Joussen A, Gardner TW, Kirchhof B, Ryan SJ, editors. Retinal vascular disease. Berlin: Springer; 2007. p. 582–91. 35. Groenewald C, Konstantinidis L, Damato B. Effects of radiotherapy on uveal melanomas and adjacent tissues. Eye. 2013;27:163–71. 36. Jmor F, Hussain RN, Damato BE, et al. Photodynamic therapy as initial treatment for small choroidal melanomas. Photodiagn Photodyn Ther. 2017;20:175–81. 37. Kines RC, Varsavsky I, Choudhary S, et al. An infrared dye-conjugated virus-like particle for the treatment of primary uveal melanoma. Mol Cancer Ther. 2018;17:565–74. 38. Wilson DJ, Klein ML.  Cryotherapy as a primary treatment for choroidal melanoma. Arch Ophthalmol. 2002;120:400–3. 39. Konstantinidis L, Groenewald C, Coupland SE, et al. Long-term outcome of primary endoresection of choroidal melanoma. Br J Ophthalmol. 2014;98:82–5. 40. Damato B.  Choroidal melanoma endoresection, dandelions and allegory-based medicine. Br J Ophthalmol. 2008;92:1013–4. 41. Bechrakis NE, Foerster MH. Neoadjuvant proton beam radiotherapy combined with subsequent endoresection of choroidal melanomas. Int Ophthalmol Clin. 2006;46:95–107. 42. Harbour JW, Onken MD, Roberson ED, et al. Frequent mutation of BAP1 in metastasizing uveal melanomas. Science. 2010;330:1410–3. 43. Damato B, Eleuteri A, Taktak AF, et  al. Estimating prognosis for survival after treatment of choroidal melanoma. Prog Retin Eye Res. 2011;30:285–95. 44. Berry D, Seider M, Stinnett S, et al. Relationship of clinical features and baseline tumor size with gene expression profile status in uveal melanoma: a multi-­ institutional Study. Retina. 2018. 45. Vaquero-Garcia J, Lalonde E, Ewens KG, et  al. PRiMeUM: a model for predicting risk of metastasis in uveal melanoma. Invest Ophthalmol Vis Sci. 2017;58:4096–105.

199

46. Hussain RN, Kalirai H, Groenewald C, et  al. Prognostic biopsy of choroidal melanoma after proton beam radiation therapy. Ophthalmology. 2016;123:2264–5. 47. Eskelin S, Pyrhonen S, Summanen P, et al. Screening for metastatic malignant melanoma of the uvea revisited. Cancer. 1999;85:1151–9. 48. Marshall E, Romaniuk C, Ghaneh P, et al. MRI in the detection of hepatic metastases from high-risk uveal melanoma: a prospective study in 188 patients. Br J Ophthalmol. 2013;97:159–63. 49. Wen JC, Sai V, Straatsma BR, et al. Radiation-related cancer risk associated with surveillance imaging for metastasis from choroidal melanoma. JAMA Ophthalmol. 2013;131:56–61. 50. Patel P, Finger PT.  Whole-body 18F FDG posi tron emission tomography/computed tomography evaluation of patients with uveal metastasis. Am J Ophthalmol. 2012;153:661–8. 51. Strobel K, Bode B, Dummer R, et al. Limited value of 18F-FDG PET/CT and S-100B tumour marker in the detection of liver metastases from uveal melanoma compared to liver metastases from cutaneous melanoma. Eur J Nucl Med Mol Imaging. 2009;36:1774–82. 52. Diener-West M, Reynolds SM, Agugliaro DJ, et  al. Screening for metastasis from choroidal melanoma: the Collaborative Ocular Melanoma Study Group Report 23. J Clin Oncol Off J Am Soc Clin Oncol. 2004;22:2438–44. 53. Augsburger JJ, Correa ZM, Trichopoulos N.  Surveillance testing for metastasis from primary uveal melanoma and effect on patient survival. Am J Ophthalmol. 2011;152:5–9 e1. 54. Kim IK, Lane AM, Gragoudas ES. Survival in patients with presymptomatic diagnosis of metastatic uveal melanoma. Arch Ophthalmol. 2010;128:871–5. 55. Komatsubara KM, Carvajal RD. Immunotherapy for the treatment of uveal melanoma: current status and emerging therapies. Curr Oncol Rep. 2017;19:45. 56. Whitehead J, Tishkovskaya S, O’Connor J, et  al. Devising two-stage and multistage phase II studies on systemic adjuvant therapy for uveal melanoma. Invest Ophthalmol Vis Sci. 2012;53:4986–9. 57. Damato B, Hope-Stone L, Cooper B, et  al. Patient-­ reported outcomes and quality of life after treatment of choroidal melanoma: a comparison of enucleation vs radiotherapy in 1596 patients. Am J Ophthalmol. 2018;193:230. 58. Ah-Fat FG, Sharma MC, Damato BE. Taping outpatient consultations: a survey of attitudes and responses of adult patients with ocular malignancy. Eye. 1998;12(Pt 5):789–91. 59. Yang J, Manson DK, Marr BP, et  al. Treatment of uveal melanoma: where are we now? Ther Adv Med Oncol. 2018;10:1758834018757175.

Uveal Melanoma: Brachytherapy

12

Gustav Stålhammar, Stefan Seregard, and Bertil E. Damato

Introduction In 1930, Sir Foster Moore first used brachytherapy for uveal melanoma by inserting radon-222 seeds directly into the tumor [1]. This technique was later modified by Stallard and eventually further refined using radioactive plaques containing cobalt-60 anchored to the episcleral surface [2, 3]. In the United States, this radionuclide was gradually replaced by plaques loaded with iodine-125 seeds, as this provided less radiation to surrounding tissues [4, 5]. In Europe, the pioneering work of Lommatzsch in the 1970s led to the introduction of ruthenium-106 as a radioactive source for episcleral brachytherapy of uveal melanoma [6]. Although observational data sug-

gested that there was no survival difference compared to patients enucleated for uveal melanoma, it had not been confirmed in a randomized control trial until the Collaborative Ocular Melanoma Study (COMS) was launched in the mid-1980s. This included patients with medium-sized uveal melanoma (2.5–10  mm in thickness and basal diameter  ≤  16  mm), who had equal 5-year and 10-year survival rates in the enucleation and iodine brachytherapy groups [7, 8]. Ever since, brachytherapy has been a mainstay treatment within this size group (Box 12.1). Smaller tumors ( 10 mm and/or basal diameter > 16 mm) still undergo enucleation. Box 12.1 Brachytherapy of Uveal Melanoma

G. Stålhammar (*) · S. Seregard Ophthalmic Pathology and Oncology Service and Department of Clinical Neuroscience, St. Erik Eye Hospital and Karolinska Institutet, Stockholm, Sweden e-mail: [email protected] B. E. Damato Nuffield Department of Clinical Neurosciences, University of Oxford, John Radcliffe Hospital, Oxford, UK

• Administered with plaques containing radioactive isotopes such as iodine-125 or ruthenium-106. • Minimum dose to tumor apex of 70–100 Gy. • May be combined with transpupillary thermotherapy or tumor resection. • Can cause collateral damage to surrounding ocular tissues, especially with

© Springer Nature Switzerland AG 2019 B. E. Damato, A. D. Singh (eds.), Clinical Ophthalmic Oncology, https://doi.org/10.1007/978-3-030-17879-6_12

201

G. Stålhammar et al.

202

large tumors requiring high radioactive doses to tumor base. • The most common complications include radiation retinopathy, radiation optic neuropathy, cataract, and neovascular glaucoma. • For medium-sized tumors (2.5–10  mm in thickness and  ≤  16  mm in largest basal diameter), survival is not significantly worse than that after enucleation.

mechanisms. Further, the tissue absorption of ionizing radiation breaks chemical bonds and forms free radicals, causing oxidative stress, inflammation, edema, neural and vascular damage, and loss of reproductive capacity. If the cell’s repair mechanisms are overwhelmed, death occurs by necrosis or apoptosis. The absorbed dose is usually measured in Grays (Gy), with 1  Gy being equal to 1  J of energy absorbed by 1 kg of tissue.

Episcleral Radioactive Plaque Radiation Brachytherapy (after the Greek brachy, meaning a short distance) refers to the implantation of radioactive material either within or close to a tumor [9]. A nuclide is a type of atom specified by its unique combination of protons and neutrons and its energy state. Isotopes of the same element differ in the number of neutrons. Radionuclides and radioisotopes decay into more stable forms, emitting ionizing radiation, which has the capacity to displace electrons from atoms and molecules and thereby create ions. This causes double-stranded DNA breaks, which in turn do relatively more harm to rapidly proliferating tumor cells with malfunctioning DNA repair

The radionuclides used for episcleral brachytherapy of uveal melanoma include cobalt-60 (now obsolete), ruthenium-106 [10, 11], iodine-125 [4], palladium-103 [12], gold (aurum)-198 [13], iridium-192 [14], and strontium-90 [15] (Table  12.1) [9]. Some radionuclides almost exclusively emit β particles with a minimal γ component (e.g., ruthenium 106), and others emit β particles plus a significant high-energy γ component (e.g., cobalt-60) or a pure low-energy γ component (e.g., iodine-125).

Plaque Design Most episcleral plaques are bowl-shaped and usually about 15–20 mm in diameter. The inner part

Table 12.1  Physical properties including decay energies in kiloelectronvolt (keV) of various nuclides used for ophthalmic brachytherapy Element Cobalt Iodine Ruthenium Rhodium Iridium Palladium Gold Strontium

Nuclide Co 125 I 106 Ru 106 Rh 192 Ir 103 Pd 198 Au 90 Sr 60

Mean γ energy (keV) 1252 36 0 602 372 360 415 0

Mean X-ray energy (keV) 0.5 26 0 2 4 2 4 0

Mean β energy (keV) 96 0 10 1410 181 0 312 196

Half-life 5.3 years 60 days 374 days 30 seconds 74 days 17 days 3 days 28.9 years

Note that some radionuclides emit almost only β-particles, while others emit γ-rays, X-rays, or a combination thereof. Radioactive nuclides generally decay to the ground state via intermediate isotopes. For example, iodine-125 undergoes electron capture to tellurium-125, which is the true emitter of the stated γ- and X-ray energies. Ruthenium disintegrates to palladium-103 via the γ- and highly β-emitting but short-lived nuclide rhodium-106, which is why the average decay energy of a ruthenium-106 plaque of 97 keV is higher than that of the ruthenium-106 nuclide isolated. Generally, a shower of internal conversion and Auger electrons are also emitted with the decay of each nuclide, but at low and clinically insignificant energies

203

12  Uveal Melanoma: Brachytherapy

a

Silver window (0.1 mm) Ru-106 coating Target foil (0.2 mm) Silver backing (0.7 mm)

b

risk of secondary enucleation [18]. Dosimetry for episcleral plaques is calculated by different methods, but the results are reasonably consistent. There have been shortcomings in plaque manufacture and dosimetry (e.g., radiation leakage and inaccurate dose specifications), and so it is important for hospitals to check each new plaque before it is used.

Treatment Preoperative Assessment Fig. 12.1 A ruthenium-106 coated foil encapsulated within an episcleral silver applicator (a), and a COMS plaque shell made of a gold alloy and a silicone insert for positioning of iodine-125 seeds (b). (Courtesy of Eckert & Ziegler BEBIG, Berlin, Germany)

contains the radioactive source, either integrated in the plaque or held in place by glue or a silicon mold (Fig. 12.1). The outer surface is lined by a heavy metal, such as silver or gold, to prevent the radiation of tissues external to the eye. Two or more eyelets near the edge of the plaque allow the plaque to be sutured to the episcleral surface.

Dosimetry Typically, the total dose provided and the radioactive dose per time unit of exposure (dose rate) are given at various distances from the radioactive source. According to the international consensus guidelines, the recommendable tumoricidal dose at the tumor apex ranges between 70 and 100 Gy [16]. A higher dose tends to increase ocular morbidity, whereas a lower dose may cause inadequate local tumor control. It remains however to be shown that a lower dose entails a higher recurrence rate and worse survival than a higher dose, within a reasonable interval. In a systematic review of 14 retrospective and 1 prospective study, no statistical significant decrease in tumor recurrences with increased average study doses was found (range 62.5– 104 Gy) [17]. The dose rates do not influence the

The basal diameter and height of the tumor are measured by funduscopy, fundus photography, ultrasonography, and/or transillumination. A correct estimate of the largest basal diameter is important for the selection of an appropriate-­ sized plaque, and conventionally a 2 mm safety margin around the tumor is added. The height measurement is usually obtained by ultrasonography and is critical for the calculation of the appropriate delivery time and hence the radiation dose. Most centers deliver an apex dose of 80–100  Gy so that the sclera receives a much higher dose, particularly if the tumor is thick and a nuclide with a steep radioactive dose fall-off is used (Fig.  12.2). There is no agreed maximum scleral dose, with up to 1500  Gy using ruthenium-­ 106 having been administered without scleral necrosis [10]. Some centers administer a minimum scleral dose of 350 Gy so that choroidal atrophy becomes visible within 6 months of treatment, thereby providing ophthalmoscopic evidence of the adequacy of plaque placement.

Plaque Positioning Correct plaque positioning at the time of surgery is essential for a good clinical outcome. The impact of experience is significant with an ­estimated cumulative experience of some 1275 cases required to achieve an adequate tumorplaque apposition in >90 % of cases, according to one publication. [19]. General anesthesia is believed by some to facilitate plaque positioning

204 Fig. 12.2 Approximate dose fall-off curves for Ruthenium 106 (Ru-106) and Ioidine 125 (I-125) plaques when 85 Gy is prescribed at the apex +1 mm of a 5 mm thick and 7 mm thick (dotted lines) tumor (a). Illustration of the 2D dose distribution of Ru-106 and I-125 planned with a dose of 85 Gy to the apex of a thin tumor. I-125 shown on the left and Ru-106 on the right (b). (Courtesy of Eckert & Ziegler BEBIG Berlin Germany)

G. Stålhammar et al.

a 1500 Gy Ru-106

I-125

1250 Gy

1000 Gy

750 Gy

500 Gy

250 Gy 85 Gy Scleral surface 0

1 mm 2 mm 3 mm 4 mm 5 mm 6 mm 7 mm 8 mm 9 mm Tumor apex height +1 mm

b

[11], but local anesthesia is more widely used. Figure  12.3 shows Damato’s technique for inserting a ruthenium plaque (Fig.  12.3). The ­ tumor margins are localized by transpupillary and transocular illumination using a 20-gauge transilluminator, and these margins are marked on the sclera with a pen. If necessary, any overlying extraocular muscles are disinserted, ­ after measuring the knot-to-limbus distances. In

10 Gy 20 Gy 30 Gy 40 Gy 50 Gy 60 Gy 70 Gy 80 Gy 85 Gy 90 Gy 100 Gy 150 Gy 200 Gy 300 Gy 400 Gy

many centers, a template is sutured to the sclera, and once it is well placed in relation to the tumor, it is replaced with the radioactive plaque. In several centers, the position of the template or plaque in relation to the tumor is checked by intraoperative ultrasonography. Damato has developed a ­ template with grooves and perforations to ­ facilitate transillumination through the ­perforations using a right-angled transillumina-

12  Uveal Melanoma: Brachytherapy

a

205

c

b Date: ___________ Patient • Name: _______________ • Surname: _______________ • Number: _______________ • DOB: __________ Eye: ________ • Fovea to disc: _____ (mm) • Disc meridian: _____ (mins) Tumour • Lat. Meridian 1: _____ (mins) Lat. Meridian 2: _____ (mins) • • Long. Diameter: _____ (mm) • Trans. Diameter: _____ (mm) • Thickness: _____ (mm) • Dist. to fovea: _____ (mm) • Dist. to disc: _____ (mm) Dist. to limbus: _____ (mm) Plaque • Diameter: _____ (mm) • Meridian: _____ (mins) • Post. edge to tumour: _____ (mm) Ant. edge to tumour: _____ (mm) Post. edge to disc: _____ (mm) Post. edge to fovea: _____ (mm) Ant. edge to limbus: _____ (mm) Dosimetry • Apex dose: ____ (Gy) Scleral dose: ____ (Gy) • Dose to disc: ____ (Gy) Dose to fovea: ____ (Gy) Practitioner: _____________________

Fig. 12.3  Technique for placement of ruthenium plaque, developed by the third author (BD). The longitudinal basal tumor diameter is measured with B-scan echography so that the intended location of the plaque can be estimated. For example, if the tumor diameter is 14  mm and the plaque diameter is 20 mm, positioning the plaque with its anterior edge 6 mm anterior to the anterior tumor margin will result in its posterior edge aligned with the posterior tumor margin (a). A fundus drawing is made, indicating the meridian of the tumor and the distances between the posterior tumor margin and the fovea and optic disk (b). The tumor is localized by transocular illumination, using a 20-G fiber optic transilluminator (c). The tumor margins and the intended location of the anterior plaque edge are marked on the sclera with a pen (d). The plaque template is placed on the sclera so that its anterior edge lies over the

relevant ink mark (e). The template is held with Moorfields forceps, which have pointed tips, which are pressed against the eye to create two dimples in the sclera, which indicate the entry and exit points for the lug suture (f). The plaque template has grooves and perforations to guide and position the transilluminator (g). The position of the plaque in relation to the tumor is checked by performing binocular indirect ophthalmoscopy while performing trans-scleral illumination through each perforation in turn (h). The template is replaced by the radioactive plaque, which is secured with the same eyelet sutures and which is firmly apposed to the eye by a mattress suture (i). The extraocular muscles are reinserted or attached to the sclera or plaque by slings. When reinserting rectus muscles, care is taken to ensure that the knot-to-limbus distances are the same as those measured before muscle disinsertion (j)

G. Stålhammar et al.

206

d

e

f

g

h

i

j

Fig. 12.3 (continued)

207

12  Uveal Melanoma: Brachytherapy

tor, the tip of this instrument being inserted in a hole in the template while performing binocular indirect ophthalmoscopy [11]. Clinically visible extraocular tumor extension can be treated with a radioactive plaque [20]. Alternatively, the extraocular nodule can be excised together with the surrounding episclera and with the superficial lamella of the adjacent sclera [21]. Care must be taken to avoid tilting of the plaque, especially with juxtapapillary tumors, because the tissue can become wedged between the plaque and the sclera [22]. A mattress suture helps to prevent the plaque from tilting away from the sclera. The importance of checking the plaque’s position at a

the time of surgery was demonstrated by studies in which 24–36% of plaques required repositioning when their location was checked by intraoperative ultrasonography [23, 24]. To reduce radiation to the optic disk and fovea, one might deliberately position a ruthenium plaque eccentrically in relation to the tumor, aligning the posterior plaque edge with the posterior tumor margin and relying on the side-scatter of radiation to treat any lateral tumor extension (Fig. 12.4) [11, 20, 25]. With this technique, 75% of patients with the posterior tumor edge at least 3 mm from the fovea may retain 20/40 visual acuity at 4 years after brachytherapy (Fig. 12.5) [19, 24]. Once the b

c

Fig. 12.4  Right fundus photograph showing an inferotemporal choroidal melanoma extending close to the fovea (a). Fundus appearance 2 years after brachytherapy

with an eccentrically positioned ruthenium plaque. The visual acuity was 20/20 (b). OCT of the same eye showing a healthy fovea (c)

208

G. Stålhammar et al.

a

b

c

d

Fig. 12.5  Kaplan-Meier curves showing conservation of visual acuity of 20/40 or better according to (a) posterior tumor extension, (b) coronal tumor location, (c) tumor height, and (d) the sum of these three risk factors. Each

step represents the last date when vision was known to be good (unlike conventional Kaplan-Meier curves, which indicate the first date when vision is found to be poor)

plaque is in place, any rectus muscles are repositioned, using slings if necessary, and the conjunctiva is closed. When the prescribed dose of radiotherapy has been delivered, usually after 2–7 days, the plaque is removed by a second procedure. Any disinserted rectus muscles are replaced, ensuring that the knot-to-limbus distances are the same as before. If the inferior oblique muscle is disinserted, it is usually left unattached. Care should be taken to ensure that adequate postoperative analgesia is prescribed.

initially every 3–6 months, then every 6 months for about 5 years, and eventually once every year. The comparison of ophthalmoscopic appearances with a baseline color photograph or serial fundus photography should reveal any marginal recurrence at an early stage. Ultrasonography is especially useful for measuring changes in tumor thickness. Tumor regression is usually not apparent for the first 3–6 months after brachytherapy. The rate of regression varies significantly between tumors. There is disagreement as to whether tumor regression is more rapid and complete in patients who subsequently develop metastatic disease [26–28]. Recurrence should be suspected if any apparent growth exceeds 0.5 mm and if a trend is confirmed by repeated examination.

Follow-Up As with other forms of conservative therapy, lifelong surveillance is indicated, with assessment

12  Uveal Melanoma: Brachytherapy

209

Radiation Safety

Collateral Damage to Ocular Tissues

Radiation safety guidelines based on dosimetry modeling indicate that a surgeon can safely perform approximately 100 ruthenium-106 plaque procedures or 50 iodine-125 plaque operations each year. The silver shielding of ruthenium plaques and the gold shielding used for iodine-125 plaques absorb 99% or more of radiation [29]. Once the plaque is inserted, visitors and healthcare personnel working by the bedside should receive minimal doses of radiation. Each hospital has its own safety rules, which must be strictly enforced.

Using simulation software programs, the risk of radiation-related side effects can usually be estimated before brachytherapy [37]. At least 50% of patients with a large uveal melanoma experience significant ocular morbidity [38]. Some radionuclides like iodine-125 are associated with frequent and often significant complications [39]. With small tumors, ocular adverse effects are less common and less severe. The steep dose fall-off of β-emitting sources such as ruthenium-106 reduces radiation doses to the retina and other surrounding tissues [10, 20]. Most radiation-­ induced ocular morbidities occur within the first postoperative years, but adverse effects may develop after a prolonged period.

Plaque Modifications Ruthenium plaques are available in over 15 different shapes and sizes, with diameters of 11–25 mm to facilitate individualized treatment. Circular plaques are used for most peripheral and posterior tumors. Notched plaques are suitable for juxtapapillary tumors, and bean-shaped plaques for ciliary body tumors or tumors close to the iris. Custom-designed plaques are sometimes used for nonresectable iris melanoma [30]. Binuclide plaques combine ruthenium-106 and iodine-125 in a single applicator, so that tumors with a thickness of 6.5–9 mm can be treated adequately while minimizing collateral damage to uninvolved ocular tissues [31]. The standardized COMS plaque has been modified so that the iodine seeds are held in place with metal cutouts instead of a solidified acrylic gel and are held farther from the scleral surface than before; these alterations have improved dosimetry and facilitated seed loading [32].

Combined Treatment Adjunctive transpupillary thermotherapy (TTT) may prevent juxtapapillary tumor growth [33, 34]. In addition, adjunctive TTT reduces exudation from the irradiated tumor, thereby minimizing visual loss due to macular edema. Adjunctive plaque radiotherapy after trans-scleral local resection is used to reduce the risk of local recurrence [35, 36].

Intraoperative Complications Ocular perforation when suturing can cause subretinal or vitreous hemorrhage as well as retinal detachment. This complication is treated by immediate cryotherapy, performed before the radioactive plaque is inserted, or binocular indirect laser photocoagulation. Occasionally, the surgical manipulation of the eye may cause small hemorrhages in and around the tumor, which resolve after a few weeks [40]. Choroidal detachments can occur in the immediate postoperative period if a vortex vein is compressed or when a new plaque with high dose rate (Gy/h) is used. As with uveal effusion of any cause, this can cause angle-closure glaucoma, RPE hypertrophy, RPE hyperplasia, non-rhegmatogenous retinal detachment, and vision loss. Troublesome diplopia can develop if any disinserted muscles are not correctly replaced after plaque removal.

Cataract Cataract is a very common consequence of any ocular radiotherapy and develops more frequently after the treatment of anterior and large tumors [41]. By 5  years, 83% of the iodine-treated patients in COMS had developed cataract and 12% had undergone cataract surgery [38].

G. Stålhammar et al.

210

Ruthenium plaques have a relatively low incidence of cataract. When needed, phacoemulsification is performed in the standard fashion.

Optic Neuropathy Collateral damage to the optic nerve is almost inevitable with juxtapapillary melanoma, but may occur after treatment of any posteriorly located tumor. Nearly half (46%) of the patients treated with iodine-125 brachytherapy for large uveal melanoma develop optic neuropathy during the subsequent 5  years [38]. When using ruthenium for smaller tumors, optic neuropathy develops in only 12% of patients at 5  years [42]. Eccentric plaque positioning reduces this complication [23]. Observations after external beam radiotherapy indicate that a total cumulative dose of >50 Gy or ≥10 Gy at a single occasion is necessary to produce radiation-induced optic neuropathy.

Retinopathy While the risk for radiation retinopathy and the severity of this complication increase with a higher radiation exposure, no thresholds delineating a safe dose have been identified. Patients who are pregnant at the time of brachytherapy as well as those with comorbidities such as hypertension and diabetes and those receiving systemic chemotherapy are at increased risk of developing retinopathy after receiving a normal dose of 80–100 Gy to the tumor apex [42, 43]. In the immediate postoperative period, exudative retinal detachment may develop, which can be severe and bullous, taking several weeks or months to resolve. Radiation damage to the macula is more likely in eyes with a large posterior tumor [41]. The 5-year cumulative incidence of radiation maculopathy is 30–52% depending not only on the size and location of the tumor but also on the radionuclide used [38, 42]. Also, comparative dosimetry calculations suggest that ruthenium has a dose distribution more confined to the tumor than iodine [44]. Lower incidences of retinopathy and maculopathy should be expected

with small- and medium-sized tumors, especially with eccentric plaque positioning [23, 25]. Scatter laser photocoagulation may induce regression of radiation retinopathy [45]. Maculopathy caused by exudates and edema can be treated by administering transpupillary thermotherapy to the tumor [23, 46]. Such maculopathy can also respond transiently to intravitreal or periocular triamcinolone injection [47, 48]. Some centers offer regular injections with VEGF inhibitors to patients that have undergone plaque brachytherapy. Significantly lower incidences of macular edema, radiation maculopathy and poor visual acuity have been reported [49], but in many patients, improvements are modest and transitory [50–52].

Neovascular Glaucoma Iris rubeosis and neovascular glaucoma have been reported in 12–69% of patients, the incidence increasing with tumor size [42, 53]. This is a major complication which may cause uncontrolled hemorrhage, glaucoma, and pain. Panretinal photocoagulation can be successful in decreasing the ischemic drive of neovascular proliferations, and trans-scleral diode laser cyclophotocoagulation may control the intraocular pressure. Local resection of the irradiated tumor induces the regression of exudative retinal detachment and neovascular glaucoma after proton beam radiotherapy and may well be effective after brachytherapy [36]. If these measures, in addition to pressure lowering drugs, fail to relieve hypertension and pain, enucleation can be the last resort even in cases with full tumor regression.

Scleral Melting This severe radiation-related complication is rare [54]. This complication is more likely to develop when scleral exposure occurs because of conjunctival wound dehiscence. With ruthenium brachytherapy, scleral doses as high as 1500 Gy are usually tolerated [10]. Management may include scleral grafting or enucleation. If scleral

12  Uveal Melanoma: Brachytherapy

grafting is performed, the graft must be larger than the offending plaque so that it can be sutured to nonirradiated sclera.

Choroidal Atrophy

211

brachytherapy for large uveal melanoma, depending on baseline visual acuity (Fig. 12.8) [56].

Local Tumor Recurrence

Local tumor recurrence is the main reason for secondary enucleation following episcleral brachytherapy for uveal melanoma [10, 57]. However, many eyes with local recurrence can be retained after further radiotherapy or by performing transpupillary thermotherapy. The overall tumor recurrence rate is approximately 10% at 5 years, and treatment failure is associated with greater size and posterior extension of the tumor [57]. Although tumor recurrence after brachytherapy for juxtapapillary tumors is more frequent, a local control rate of 80% at 10 years following iodine brachytherapy has been reported [58]. Local control rate is significantly better for smaller tumors, with a recurrence rate of only 3% at 7 years [11]. There are observational data suggesting that with tumors exceeding 5 mm in thickness, local recurrence may be Results more common with ruthenium compared to iodine [59, 60]. Some use ruthenium for choroiVisual Outcome dal melanoma up to 7 mm or more in thickness with good local control [10]. An increase in Although plaque radiotherapy of anterior mela- tumor size after initial tumor regression may be nomas is more likely to cause reversible visual caused by intratumoral hemorrhage and does loss secondary to cataract, the treatment of poste- not necessarily indicate local recurrence [61]. In rior tumors is more likely to be associated with 40% of eyes enucleated for suspected failure of irreversible loss caused by retinopathy [40]. local control after brachytherapy, subsequent Approximately 3–5  years after brachytherapy, histopathological examination could not conhalf of the patients (49–55%) maintain a best cor- firm an increase in tumor size [62]. rected visual acuity of 20/200 or better, and one-­ Small, dark, episcleral deposits commonly third (31–33%) have 20/50 visual acuity or better appear after brachytherapy for uveal melanoma. in the affected eye [10, 55]. By treating patients These macrophage-associated deposits should not with juxtapapillary tumors using eccentric plaque be confused with extraocular tumor recurrence [63]. fixation to minimize radiation to the fovea, as Although most recurrences occur within the many as 57% of patients may have preservation first few postoperative years, regrowth after of 20/200 visual acuity or better 9 years follow- 10–15  years has been reported, indicating the ing ruthenium brachytherapy for uveal melanoma need for lifelong follow-up [10]. The value of (Figs.  12.6 and 12.7) [23]. A significant loss of prolonged surveillance is further suggested by vision following brachytherapy is associated the finding that eyes enucleated for neovascular with greater tumor height and shorter distance of and other complications have been found to conthe tumor to the fovea. Useful vision is usually tain cycling tumor cells without any clinical evionly maintained for 1–2  years following iodine dence of residual disease [64]. Choroidal atrophy has generally been regarded as an inevitable consequence of brachytherapy, occurring because of the high basal radiation dose required for the delivery of a tumoricidal dose to the tumor apex. Such choroidal atrophy undoubtedly causes a severe defect in the corresponding part of the visual field. The introduction of eccentric plaque placement has led to the discovery that marginal recurrence does not occur even when the base of the tumor extends slightly beyond the area of visible atrophy. This has led to the clinical impression that local tumor control and visual conservation can be achieved simultaneously, even with tumors that extend far posteriorly.

212

G. Stålhammar et al.

a

c

b

d

e

Fig. 12.6  Clinical features (a) and imaging by optical coherence tomography (OCT) (b) of macular edema 9  months after ruthenium brachytherapy for a medium-­ sized uveal melanoma. Visual acuity was reduced from

a

Fig. 12.7  Right fundus of a 61-year-old man with a temporal choroidal melanoma measuring 11.0 mm in the largest basal diameter and 5.0 mm in height (a). The patient was treated with an eccentrically placed ruthenium plaque delivering a scleral dose 612  Gy and an apex dose of

20/20 before radiotherapy to 20/50. Intravitreal injection of triamcinolone (c) induced prompt resolution of edema, both clinically (d) and by OCT (e). Visual acuity improved to 20/30 within 2 weeks

b

90 Gy. Thirty-two months later, his vision was 20/20. The tumor appeared atrophic and had regressed to a thickness of 1.7 mm (b). There was no evidence of lateral growth. A small retinal hemorrhage indicated a significant dose of radiation beyond the visible choroidal atrophy

213

12  Uveal Melanoma: Brachytherapy Fig. 12.8  Mean visual acuity with 95% confidence intervals (indicated by error bars) after iodine brachytherapy of uveal melanoma (COMS 16). (Adapted from Melia et al. [55]. With permission from Elsevier)

Local tumor recurrence after plaque radiotherapy is associated with reduced survival [57]. In turn, high-risk features such as large size, iris root involvement, and gene expression class 2 are overrepresented in tumors that eventually relapse. Further, once a tumor has been irradiated, the tumor size becomes prognostically irrelevant in the sense that a small relapse may not be associated with a smaller risk for metastasis than a larger relapse or larger nonirradiated tumor [65]. Local recurrence might therefore be an indicator of aggressive disease rather than the cause of metastasis. Further studies are needed.

baseline visual acuity in the affected eye [57]. Generally, eyes are enucleated following episcleral brachytherapy in 12–17% of patients at 3–5  years follow-up [10, 57], but eyes with smaller tumors have a much lower risk of secondary enucleation [23]. The reasons for enucleation vary from one study to another and include local tumor recurrence [10, 57], recurrent vitreous hemorrhage [23], and painful neovascular glaucoma [57]. Even eyes with large uveal melanoma may be retained after brachytherapy, although visual results are poor and cosmetic results variable (Fig. 12.9) [55].

Ocular Conservation

Quality of Life

The probability of ocular conservation depends on many factors [10]. The factors associated with secondary enucleation are large tumor size, collar-­stud shape (presumably because of retinal invasion), posterior tumor extension, and poor

Quality of life is often measured according to visual function and moods such as depression and anxiety. Specific instruments to assess cognitive functioning after brachytherapy for choroidal melanoma have been developed, and observa-

214

G. Stålhammar et al.

Fig. 12.9  Cumulative proportion of patients undergoing secondary enucleation (red line) or with local treatment failure (green line) after iodine brachytherapy (COMS No

19). (Adapted from Jampol et  al. [57]. With permission from Elsevier)

tional data suggest that patients with brachytherapy have functioning similar to patients with enucleation [66, 67]. In a large questionnaire-­ based study, patients that had undergone enucleation had higher rates of depression and reduced physical and functional well-being compared with patients treated with plaque brachytherapy [68]. This may however be a reflection of the fact that patients selected for enucleation are generally older with more advanced disease, rather than an effect of enucleation itself.

iodine-125 brachytherapy arm and 11, 17, and 17%, respectively, in the enucleation arm [8]. Although the COMS data are solely based on patients treated with iodine-125, the observational data suggest that survival after ruthenium brachytherapy is similar for medium-sized tumors [10, 60]. The observational data also suggest that survival after brachytherapy for large uveal melanoma is comparable to that after enucleation [69]. There are no gender-based differences in melanoma-related mortality after brachytherapy [70]. Further studies are needed to determine whether failure of plaque radiotherapy enables the expression of immune checkpoints, novel mutations and increased metastatic growth.

Survival The COMS medium-sized choroidal melanoma trial concluded that patient survival following iodine brachytherapy is not statistically significantly different from that after enucleation [8]. The 5-, 10-, and 12-year rates of death with histopathologically confirmed melanoma metastases were 10, 18, and 21%, respectively, in the

Summary In most centers, plaque radiotherapy is the first choice of treatment for uveal melanoma. This is more reliable than phototherapy, less expensive

12  Uveal Melanoma: Brachytherapy

than proton beam radiotherapy, and less invasive than local resection; however, it can be difficult to position a plaque accurately over a small, posterior tumor (in which proton beam radiotherapy may be preferable, given the choice). With proper case selection, a team experienced in brachytherapy can achieve rates of local tumor control that match those of proton beam radiotherapy but without the ocular surface, lacrimal drainage, and eyelid complications of teletherapy. As with any other treatment, conservation of vision depends greatly on the distance between the tumor and the optic disk and macula. Collateral radiation damage to these two areas can be minimized by collimating iodine-125 radiation or by using lower-range isotopes such as ruthenium-106 and strontium-90 and, in the case of ruthenium, by eccentric plaque placement. Despite insignificant doses of radiation to the optic disk and macula, many patients lose vision after brachytherapy because of exudation from the irradiated tumor (“toxic tumor syndrome”). Such “secondary” or “indirect” radiation maculopathy can be treated successfully with transpupillary thermotherapy to the tumor or by antiangiogenic or intravitreal steroid injection. Brachytherapy is increasingly used as an adjunctive form of treatment, after transpupillary thermotherapy or local resection. Several studies have shown that survival after brachytherapy for medium-sized tumors is not significantly worse than that after enucleation. For all the reasons mentioned above, it is likely that brachytherapy, with or without adjuvant thermotherapy, will continue to be the first line of treatment for smalland medium-sized uveal melanomas.

References 1. Moore RF.  Choroidal sarcoma treated by the intraocular insertion of radon seeds. Br J Ophthalmol. 1930;14(4):145–52. 2. Stallard HB.  Radiotherapy of malignant intra-ocular neoplasms. Br J Ophthalmol. 1948;32(9):618–39. 3. Stallard HB. Radiotherapy for malignant melanoma of the choroid. Br J Ophthalmol. 1966;50(3):147–55. 4. Sealy R, le Roux PL, Rapley F, et  al. The treatment of ophthalmic tumours with low-energy sources. Br J Radiol. 1976;49(582):551–4. 5. Packer S, Rotman M.  Radiotherapy of choroidal melanoma with iodine-125. Ophthalmology. 1980;87(6):582–90.

215 6. Lommatzsch P.  Treatment of choroidal melanomas with 106Ru/106Rh beta-ray applicators. Surv Ophthalmol. 1974;19(2):85–100. 7. Collaborative Ocular Melanoma Study Group. The COMS randomized trial of iodine 125 brachytherapy for choroidal melanoma. III.  Initial mortality findings. COMS report No. 18. Arch Ophthalmol. 2001;119:969–82. 8. Collaborative Ocular Melanoma Study Group. The COMS randomized trial of iodine 125 brachytherapy for choroidal melanoma. V.  Twelve-year mortality rates and prognostic factors. COMS report No. 28. Arch Ophthalmol. 2006;124:1684–93. 9. Stannard C, Sauerwein W, Maree G, et al. Radiotherapy for ocular tumours. Eye (Lond). 2013;27(2):119–27. 10. Bergman L, Nilsson B, Lundell G, et al. Ruthenium brachytherapy for uveal melanoma, 1979–2003: survival and functional outcomes in the Swedish population. Ophthalmology. 2005;112(5):834–40. 11. Damato B, Patel I, Campbell IR, et al. Local tumor control after 106Ru brachytherapy of choroidal melanoma. Int J Radiat Oncol Biol Phys. 2005;63(2):385–91. 12. Finger PT, Chin KJ, Duvall G, et  al. Palladium-103 ophthalmic plaque radiation therapy for choroidal melanoma: 400 treated patients. Ophthalmology. 2009;116(4):790–6. 796.e1. 13. Karvat A, Duzenli C, Ma R, et  al. The treatment of choroidal melanoma with 198 Au plaque brachytherapy. Radiother Oncol. 2001;59(2):153–6. 14. Shields CL, Shields JA, Cater J, et al. Plaque radiotherapy for uveal melanoma: long-term visual outcome in 1106 consecutive patients. Arch Ophthalmol. 2000;118(9):1219–28. 15. Missotten L, Dirven W, Van der Schueren A, et  al. Results of treatment of choroidal malignant melanoma with high-dose-rate strontium-90 brachytherapy. A retrospective study of 46 patients treated between 1983 and 1995. Graefes Arch ClinExpOphthalmol. 1998;236(3):164–73. 16. The American Brachytherapy Society consensus guidelines for plaque brachytherapy of uveal melanoma and retinoblastoma. Brachytherapy. 2014;13(1):1–14. 17. Echegaray JJ, Bechrakis NE, Singh N, et  al. Iodine-125 brachytherapy for uveal melanoma: a systematic review of radiation dose. Ocul Oncol Pathol. 2017;3(3):193–8. 18. Fili M, Lundell G, Lundell M, et  al. High dose rate and low dose rate ruthenium brachytherapy for uveal melanoma. No association with ocular outcome. Br J Ophthalmol. 2014;98(10):1349–54. 19. Shah NV, Houston SK, Murray TG, et al. Evaluation of the surgical learning curve for I-125 episcleral plaque placement for the treatment of posterior uveal melanoma: a two decade review. Clin Ophthalmol. 2012;6:447–52. 20. Gunduz K, Shields CL, Shields JA, et  al. Plaque radiotherapy for management of ciliary body and choroidal melanoma with extraocular extension. Am J Ophthalmol. 2000;130(1):97–102. 21. Muen WJ, Damato BE.  Uveal malignant melanoma with extrascleral extension, treated with plaque radiotherapy. Eye (Lond). 2007;21(2):307–8.

216 22. Williams DF, Mieler WF, Jaffe GJ, et  al. Magnetic resonance imaging of juxtapapillary plaques in cadaver eyes. Br J Ophthalmol. 1990;74(1):43–6. 23. Damato B, Patel I, Campbell IR, et  al. Visual acuity after Ruthenium(106) brachytherapy of choroidal melanomas. Int J Radiat Oncol Biol Phys. 2005;63(2):392–400. 24. Tabandeh H, Chaudhry NA, Murray TG, et  al. Intraoperative echographic localization of iodine-125 episcleral plaque for brachytherapy of choroidal melanoma. Am J Ophthalmol. 2000;129(2):199–204. 25. Russo A, Laguardia M, Damato B.  Eccentric ruthenium plaque radiotherapy of posterior choroidal melanoma. Graefes Arch Clin Exp Ophthalmol. 2012;250:1533–40. 26. Augsburger JJ, Gamel JW, Shields JA, et  al. Post-­ irradiation regression of choroidal melanomas as a risk factor for death from metastatic disease. Ophthalmology. 1987;94(9):1173–7. 27. Rashid M, Heikkonen J, Singh AD, et  al. Clinical predictors of regression of choroidal melanomas after brachytherapy: a growth curve model. Ophthalmology. 2018;125(5):747–54. 28. Damato B.  Predicting choroidal melanoma regression after brachytherapy. Ophthalmology. 2018;125(5):755–6. 29. Laube T, Fluhs D, Kessler C, et  al. Determination of surgeon’s absorbed dose in iodine 125 and ruthenium 106 ophthalmic plaque surgery. Ophthalmology. 2000;107(2):366–8; discussion 368–9. 30. Shields CL, Naseripour M, Shields JA, et  al. Custom-designed plaque radiotherapy for nonresectable iris melanoma in 38 patients: tumor control and ocular complications. Am J Ophthalmol. 2003;135(5):648–56. 31. Fluhs D, Anastassiou G, Wening J, et al. The design and the dosimetry of bi-nuclide radioactive ophthalmic applicators. Med Phys. 2004;31(6):1481–8. 32. Astrahan MA, Szechter A, Finger PT.  Design and dosimetric considerations of a modified COMS plaque: the reusable “seed-guide” insert. Med Phys. 2005;32(8):2706–16. 33. Seregard S, Landau I.  Transpupillary thermother apy as an adjunct to ruthenium plaque radiotherapy for choroidal melanoma. Acta Ophthalmol Scand. 2001;79(1):19–22. 34. Bartlema YM, Oosterhuis JA, Journee-De Korver JG, et al. Combined plaque radiotherapy and transpupillary thermotherapy in choroidal melanoma: 5 years’ experience. Br J Ophthalmol. 2003;87(11):1370–3. 35. Damato BE, Paul J, Foulds WS. Risk factors for residual and recurrent uveal melanoma after trans-scleral local resection. Br J Ophthalmol. 1996;80(2):102–8. 36. Damato BE, Groenewald C, Foulds WS.  Surgical resection of choroidal melanoma. In: Ryan SJ, editor. Retina, vol. 3. 5th ed. London: Elsevier; 2013. p. 2298–306. 37. Puusaari I, Heikkonen J, Kivela T. Effect of radiation dose on ocular complications after iodine brachytherapy for large uveal melanoma: empirical data and

G. Stålhammar et al. simulation of collimating plaques. Invest Ophthalmol Vis Sci. 2004;45(10):3425–34. 38. Collaborative Ocular Melanoma Study. Incidence of cataract and outcomes after cataract surgery in the first 5 years after iodine 125 brachytherapy in the Collaborative Ocular Melanoma Study: COMS report No. 27. Ophthalmology. 2007;114(7):1363–71. 39. Wen JC, Oliver SC, McCannel TA.  Ocular complications following I-125 brachytherapy for choroidal melanoma. Eye (Lond). 2009;23(6):1254–68. 40. Robertson DM, Fuller DG, Anderson RE.  A technique for accurate placement of episcleral iodine-125 plaques. Am J Ophthalmol. 1987;103(1):63–5. 41. Finger PT.  Tumour location affects the incidence of cataract and retinopathy after ophthalmic plaque radiation therapy. Br J Ophthalmol. 2000;84(9):1068–70. 42. Summanen P, Immonen I, Kivela T, et  al. Radiation related complications after ruthenium plaque radiotherapy of uveal melanoma. Br J Ophthalmol. 1996;80(8):732–9. 43. Krema H, Xu W, Payne D, et  al. Factors predictive of radiation retinopathy post (125)Iodine brachytherapy for uveal melanoma. Can J Ophthalmol. 2011;46(2):158–63. 44. Mourtada F, Koch N, Newhauser W. 106Ru/106Rh plaque and proton radiotherapy for ocular melanoma: a comparative dosimetric study. Radiat Prot Dosimetry. 2005;116(1–4 Pt 2):454–60. 45. Finger PT, Kurli M. Laser photocoagulation for radiation retinopathy after ophthalmic plaque radiation therapy. Br J Ophthalmol. 2005;89(6):730–8. 46. Damato B. Vasculopathy after treatment of choroidal melanoma. In: Joussen A, Gardner TW, Kirchhof B, Ryan SJ, editors. Retinal vascular disease. Berlin: Springer; 2007. p. 582–91. 47. Shields CL, Demirci H, Dai V, et al. Intravitreal triamcinolone acetonide for radiation maculopathy after plaque radiotherapy for choroidal melanoma. Retina. 2005;25(7):868–74. 48. Horgan N, Shields CL, Mashayekhi A, et  al. Periocular triamcinolone for prevention of macular edema after plaque radiotherapy of uveal melanoma: a randomized controlled trial. Ophthalmology. 2009;116(7):1383–90. 49. Shah SU, Shields CL, Bianciotto CG, et al. Intravitreal bevacizumab at 4-month intervals for prevention of macular edema after plaque radiotherapy of uveal melanoma. Ophthalmology. 2014;121(1):269–75. 50. Khan N, Khan MK, Bena J, et  al. Plaque brachytherapy for uveal melanoma: a vision prognostication model. Int J Radiat Oncol Biol Phys. 2012;84(3):e285–90. 51. Groenewald C, Konstantinidis L, Damato B. Effects of radiotherapy on uveal melanomas and adjacent tissues. Eye (Lond). 2013;27(2):163–71. 52. Finger PT.  Radiation retinopathy is treatable with anti-vascular endothelial growth factor bevacizumab (Avastin). Int J Radiat Oncol Biol Phys. 2008;70(4):974–7.

12  Uveal Melanoma: Brachytherapy 53. Detorakis ET, Engstrom RE Jr, Wallace R, et  al. Iris and anterior chamber angle neovascularization after iodine 125 brachytherapy for uveal melanoma. Ophthalmology. 2005;112(3):505–10. 54. Kaliki S, Shields CL, Rojanaporn D, et  al. Scleral necrosis after plaque radiotherapy of uveal melanoma: a case–control study. Ophthalmology. 2013;120:1004–11. 55. Melia BM, Abramson DH, Albert DM, et  al. Collaborative ocular melanoma study (COMS) randomized trial of I-125 brachytherapy for medium choroidal melanoma. I. Visual acuity after 3 years COMS report no. 16. Ophthalmology. 2001;108(2):348–66. 56. Puusaari I, Heikkonen J, Summanen P, et  al. Iodine brachytherapy as an alternative to enucleation for large uveal melanomas. Ophthalmology. 2003;110(11):2223–34. 57. Jampol LM, Moy CS, Murray TG, et al. The COMS randomized trial of iodine 125 brachytherapy for choroidal melanoma: IV.  Local treatment failure and enucleation in the first 5 years after brachytherapy. COMS report no. 19. Ophthalmology. 2002;109(12):2197–206. 58. Sagoo MS, Shields CL, Mashayekhi A, et  al. Plaque radiotherapy for juxtapapillarychoroidal melanoma: tumor control in 650 consecutive cases. Ophthalmology. 2011;118(2):402–7. 59. Papageorgiou KI, Cohen VM, Bunce C, et  al. Predicting local control of choroidal melanomas following 106Ru plaque brachytherapy. Br J Ophthalmol. 2011;95(2):166–70. 60. Filì M, Trocmé E, Bergman L, et al. Ruthenium-106 versus iodine-125 plaque brachytherapy of 571 choroidal melanomas with a thickness of ≥5.5 mm. Br J Ophthalmol. 2019; https://doi.org/10.1136/bjophthalmol-2018-313419. [Epub ahead of print] 61. Seregard S, Lundell G, Lax I, et al. Tumour cell proliferation after failed ruthenium plaque radiotherapy for posterior uveal melanoma. Acta Ophthalmol Scand. 1997;75(2):148–54.

217 62. Avery RB, Diener-West M, Reynolds SM, et  al. Histopathologic characteristics of choroidal melanoma in eyes enucleated after iodine 125 brachytherapy in the collaborative ocular melanoma study. Arch Ophthalmol. 2008;126(2):207–12. 63. Toivonen P, Kivela T.  Pigmented episcleral deposits after brachytherapy of uveal melanoma. Ophthalmology. 2006;113(5):865–73. 64. Seregard S, aft Trampe E, Lax I, et  al. Results following episcleral ruthenium plaque radiotherapy for posterior uveal melanoma. The Swedish experience. Acta Ophthalmol Scand. 1997;75(1):11–6. 65. Arnljots TS, Al-Sharbaty Z, Lardner E, et al. Tumour thickness, diameter, area or volume? The prognostic significance of conventional versus digital image analysis-based size estimation methods in uveal melanoma. Acta Ophthalmol Scand. 2018;96(5):510–8. 66. Brandberg Y, Kock E, Oskar K, et al. Psychological reactions and quality of life in patients with posterior uveal melanoma treated with ruthenium plaque therapy or enucleation: a one year follow-up study. Eye (Lond). 2000;14(Pt 6):839–46. 67. Brandberg Y, Damato B, Kivela T, et al. The EORTC ophthalmic oncology quality of life questionnaire module (EORTC QLQ-OPT30). Development and pre-­ testing (Phase I-III). Eye (Lond). 2004;18(3):283–9. 68. Damato B, Hope-Stone L, Cooper B, et  al. Patient-­ reported outcomes and quality of life after treatment of choroidal melanoma. A comparison of enucleation versus radiotherapy in 1596 patients. Am J Ophthalmol. 2018;193(1):230–51. 69. Shields CL, Naseripour M, Cater J, et  al. Plaque radiotherapy for large posterior uveal melanomas (> or =8-mm thick) in 354 consecutive patients. Ophthalmology. 2002;109(10):1838–49. 70. Stålhammar G, See TRO, Filì M, et al. No gender differences in long-term survival after brachytherapy of 1,541 patients with uveal melanoma. Ocul Oncol Pathol. 2019; https://doi.org/10.1159/000497186. [Epub ahead of print].

Uveal Melanoma: Proton Beam Radiation Therapy

13

Anne Marie Lane, Ivana K. Kim, and Evangelos S. Gragoudas

Introduction Radiotherapy (RT) is the standard of care for the vast majority of malignant melanomas of the uveal tract. Eye conservation and, in many cases, maintenance of useful vision are possible. Two major radiotherapeutic techniques for the treatment of uveal melanomas are available, brachytherapy [1] with radioactive plaques, which are sutured on the sclera over the area of the tumor, and external beam irradiation using charged particles [2, 3]. Both plaque radiotherapy and external beam irradiation are effective for treating small, medium, and most large uveal melanomas; high rates of local control are achieved, and survival rates are similar to those observed after enucleation [4–6], a modality that is reserved for very large tumors, tumors with large extrascleral extensions, or patients with neovascular glaucoma in a blind, painful eye. The physical properties of charged particles such as protons allow for localized dose distributions [2, 7, 8]. Stereotactic radiosurgery [9, 10] does not appear to have as favorable dose distri-

A. M. Lane · I. K. Kim · E. S. Gragoudas (*) Ocular Melanoma Center, Department of Ophthalmology, Massachusetts Eye and Ear Infirmary/Harvard Medical School, Boston, MA, USA e-mail: [email protected]

butions. Most published studies have reported results in small patient cohorts. Findings thus far indicate higher rates of complications and visual loss when compared to brachytherapy [11] or proton beam radiotherapy [12].

Proton Beam Radiotherapy In 1946 Robert R. Wilson was the first to recognize that the physical properties of protons [13– 15] (Box 13.1) could be advantageous for medical purposes [16], including treating tumors. Today proton radiation is the most widely used form of EBRT. Box 13.1. Proton Beam Radiotherapy of Uveal Melanoma • Delivery of maximum density of ionization as protons stop (Bragg peak) • Sharp reduction of the dose outside the treated area • Modulation of beam energy: Bragg peak can be broadened to cover a tumor at any depth

Because of these properties, proton radiation is particularly advantageous for tumors that are large and/or located near the optic nerve or macula.

© Springer Nature Switzerland AG 2019 B. E. Damato, A. D. Singh (eds.), Clinical Ophthalmic Oncology, https://doi.org/10.1007/978-3-030-17879-6_13

219

220

Proton therapy in these cases can provide improved local control and reduced radiation-­induced damage to normal ocular tissue. Patients with large tumors located at the peripheral fundus are able to maintain vision because the beam can enter the eye directly in the area of the tumor and noninvolved ocular structures need not be exposed. The availability of proton facilities has increased over the past decade or more. However, at the present time in the United States, only a few of these centers treat patients with ocular melanoma including the Francis H.  Burr Proton Therapy Center at Massachusetts General Hospital, the University of Florida Proton Therapy Institute, and the University of California, San Francisco. There are also facilities in Canada, England, Germany, Russia, Switzerland, France, and Italy. With this increased access to proton therapy, more patients with uveal melanoma can now consider proton irradiation as a treatment option.

Patient Evaluation Prior to treatment, patients have a complete ophthalmological examination as well as fundus photography, and A-scan and B-scan ultrasonography. Staging imaging is performed before treatment to rule out metastasis. Proton radiation is planned for all patients who are free of metastasis and have tumors that occupy up to 30% of the ocular volume.

Pre-radiation Surgery During a surgical procedure, the tumor is localized by transillumination and/or indirect ophthalmoscopy, and the episcleral tissues over the tumor are examined for evidence of extrascleral extension. The edges of the tumor are marked with a surgical marking pen, and four 2.5  mm radiopaque tantalum rings are sutured on the sclera around the borders of the tumor (Fig. 13.1 [17]), which serve as reference points for the placement of the proton beam at the time of treatment. For tumors that extend into the ciliary body and iris, rings are placed at the posterior margin of the tumor, and the distance from the rings to

A. M. Lane et al.

Fig. 13.1  Tantalum rings sutured to the sclera at the edges of the tumor seen by transillumination (Reprinted from [17]. With permission from Elsevier)

the anterior margin of the lesion is measured. If tumors are in contact with the optic nerve, rings are placed only at the anterior and lateral margins of the lesion, and the distance from the rings to the posterior margin is estimated from fundus photographs. To further ensure the accuracy of placement of the radiation beam, the tumor is transilluminated again, and the distances from the rings to the edges of the tumor are measured on the sclera. Drawings are then made to document the size and shape of the lesion as well as the location of the tantalum rings. Some patients elect to have fine needle aspiration biopsy (FNAB) of the tumor for molecular prognostic testing. Surgery is not necessary for patients with tumors involving only the ciliary body or iris because transillumination and/or photographs can be used to define tumor margins in relation to the anatomic landmarks of the iris and conjunctiva.

Proton Treatment Planning The interactive treatment planning system developed for ocular melanomas [EYEPLAN, Martin Sheen, Clatterbridge Centre for Oncology,

13  Uveal Melanoma: Proton Beam Radiation Therapy

221

Bebington, UK] and used for all patients seen at our site has been previously described [2, 18]. A team of ocular oncologists, radiation oncologists, and physicists use axial eye length and tumor height, determined by A- and B-mode ultrasonography; tumor basal dimensions measured using ophthalmoscopy, transillumination, and ultrasound; and location of the tantalum rings, determined by orthogonal X-rays taken in the treatment position during a simulation session, to develop a specific treatment plan for each patient. These data are used to create a three-dimensional model of the tumor that is superimposed on a model of a normal eye, scaled to the ultrasonically determined length of the patient’s eye. The fundus photographs are important when creating models for very posterior tumors or tumors abutting the optic nerve because the tumors are not fully surrounded with marker rings. Two tumors can be created when tumors are shaped irregularly. Iris tumors and ciliary body tumors are drawn from clinical and ultrasound information. The computer can be made to rotate the eye so that it follows a user-controlled fixation point in near to real time, which allows the planner to choose a fixation angle that will minimize radiation exposure of the lens, optic disk, and fovea to the extent possible [18]. This program also automatically designs an aperture that approximates the shape of the tumor

and gives a 3 mm margin (can vary between 2.5 and 4 mm if needed) around the tumor at the 50% dose level. It calculates the maximum and minimum depths of the tumor, and the beam range and modulation width, based on user-selected proximal and distal margins. Program-generated fundus dose distributions are displayed in the geometry of a wide-angle fundus photograph and in any plane through the eye (Fig.  13.2 [17]). Dose-volume histograms for the tumor and many structures of the eye (e.g., globe, lens, ciliary body, retina, macula, and disk) can be generated for each treatment plan.

a

Fig. 13.2  Graphic depiction of dose distribution from treatment plan for 13-mm-diameter tumor. Isodose contours shown in a plane through the eye. (a) Isodose con-

Irradiation Procedure Prior to the start of irradiation, a positioning procedure is performed during which the patient must be immobilized and the eye properly aligned with the proton beam. The patient is treated in the seated position, with immobilization of the head accomplished by using an individually contoured face mask and bite block, which are attached to a head holder. The head holder is attached to the proton beam collimator, and the aperture is mounted in the end of the collimator nearest to the eye (Fig.  13.3 [17]). Orientation of the patient’s eye is established by voluntary fixation of the eye on a flashing light set at a position determined by

b

tours shown on the retinal surface (b) (Reprinted from [17]. With permission from Elsevier)

222

A. M. Lane et al.

mounted on the collimator, is viewed onscreen in the control area to monitor the eye position. This system allows for the interruption of treatment and repositioning of the patient if unacceptable eye movement (more than 0.5 mm) is observed. Each treatment takes approximately 1  min, and almost all treatments are completed without interruption. The standard total dose is 70 cobalt Gy equivalents (63.6 proton Gy times 1.1 relative biologic effectiveness equals 70 cobalt Gy equivalents (CGE)) administered in daily fractions of 14 CGE over 5  days. We chose this fractionation scheme based on favorable clinical results demonstrated with the use of a small number of relaFig. 13.3  Patient immobilized for treatment with the use tively large dose fractions [8, 21] in patients with of head holder, face mask, and bite block (Reprinted from skin melanomas. At the prescribed dose of 70 [17]. With permission from Elsevier) CGE, we estimate that the optic nerve and macula receive the full dose when the tumor is less the planning program. If the vision is poor in the than 1 mm from these structures and half the dose eye to be treated, the other eye, which is usually (35 CGE) when the tumor is located 3 mm from covered, can be used for fixation. these structures. Using a fluoroscopic system that provides a The maximum tolerated and minimum effecvirtually instantaneous picture held on an image-­ tive doses of proton radiation have yet to be storage device, the alignment of the tumor rela- established. However, the findings of a dose tive to the beam axis and the edges of the reduction trial [22] indicated that treatment with beam-defining aperture is achieved by moving 50 CGE did not compromise tumor control in the head holder until the surgically placed tanta- patients with small- or medium-sized tumors lum rings that spatially define the tumor are in the located near the optic nerve or fovea. As a result, desired position relative to the beam axis. For in this select group of patients, we administer a tumors treated without surgical localization, a dose of 50 CGE to potentially reduce vision-­ light beam coaxial with the central axis of the threatening radiation-induced complications. proton beam is used to position the tumor relative Unfortunately, there was no benefit found in to the beam during treatment [19, 20]. This sys- terms of visual acuity in a more recent evaluation tem also assists in confirming eye immobilization of patients with paramacular tumors treated with during treatment. Finally, a beam-simulation the 50 CGE dose [23]. At several facilities in field light is projected through the aperture onto Europe, a total dose of 60 CGE is given in four the eye to be treated, to ensure that the light field equal fractions over 4 days [24–26]. falls on the eye in relation to the edge of the limbus. The alignment procedure lasts approximately 15  minutes. After successful completion Follow-Up Protocol of this procedure, treatment can begin. During treatment, the eyelids are retracted by Ophthalmological examinations are performed the ophthalmologist using a lid speculum to 6  weeks, 6  months, and 12  months after treatreduce radiation exposure to the eyelids, and the ment. Thereafter, these examinations are perpatient is asked to fixate. Using the same fluoro- formed biannually up to 5  years after treatment scopic system used for patient positioning, a and then annually to monitor the tumor and magnified picture of the eye, taken by a camera patient status.

13  Uveal Melanoma: Proton Beam Radiation Therapy

223

Visual function, tumor recurrence, and ocular complications are assessed at every follow-up examination. Fundus photography and ­ultrasonography are performed at varying intervals to document tumor regression, tumor recurrence, and radiation-induced complications. Metastasis surveillance routinely includes liver function tests and liver imaging and is now tailored to molecular prognostic information obtained through fine needle aspiration biopsy if available.

proton facilities following essentially the same or a somewhat modified planning protocol as that developed at the Harvard Cyclotron [24–26].

Ophthalmic Outcomes Local Control

Disappearance of the tumor or formation of a flat scar occurs infrequently, and the vast majority of lesions continue to regress years after therapy. Patient Characteristics Regression (Fig. 13.4) is most likely due to both direct cell death from irradiation, achieved by The largest series of uveal melanoma patients has damage to chromosomal DNA when the cell been treated at the Massachusetts Eye and Ear enters mitosis, and damage to the vasculature that Infirmary (MEEI)/MGH Francis H. Burr Proton carries nutrients to the tumor cells. Delayed Therapy Center. Over 5000 patients have been regression in some patients may be due to the treated beginning in 1975 at MEEI, and long-­ prolonged intermitotic phases of melanoma cells. term follow-up continues to accrue. Local recurrence after charged particle irradiaAt MEEI [27] approximately equal numbers tion is observed in 2–5% [24, 26, 28] of patients. In of men and women have been treated, and the the MEEI series [27], approximately 3% of tumors mean age of these patients is 61  years. The exhibited growth, confirmed by ultrasonography median basal diameter and median tumor height and fundus photography, and just under one-half of the treated tumors were 13.2 and 5.3 mm, and were marginal recurrences. The highest annual rate 68% of these were located within 3  mm of the of recurrence, 1%, occurred 1  year after proton optic nerve or macula [27]. therapy, with the earliest occurrence observed Patients with comparable demographic and 5 months and the latest 11 years after irradiation. tumor characteristics have been treated at other The 15-year probability of local tumor control,

a

b

Fig. 13.4  Large ciliochoroidal melanoma extending up to the optic nerve and associated with serous retinal detachment. Before proton irradiation (a). Approximately 1 year posttreatment, significant reduction of the tumor is seen (b)

A. M. Lane et al.

224

based on confirmed and possible recurrences combined (N   =   60), was 95% (95% CI, 93–96%). Similar control rates were reported in a series of patients treated at Center Hôpital Ophtalmique Jules-Gonin/Paul Scherrer Institute (PSI) [24]. In some cases, eyes with local recurrence can be retreated successfully with repeat proton irradiation [29] or – for cases in which the recurrence is marginal and flat – laser photocoagulation. Higher recurrence rates after brachytherapy have been observed, and suboptimal plaque positioning [30] in posteriorly located tumors has been proposed as one explanation for poorer outcomes. Proton irradiation may be a better choice of treatment for such tumors, and a recent study of long-term outcomes after proton irradiation for patients with tumors contiguous to the optic nerve (within 1 disk diameter) supports this view. The 5-year tumor recurrence rate was approximately 3%, virtually identical to the rate seen in patients treated during the same period with tumors farther away from the nerve [31]. In contrast, tumor recurrence was reported in 10% of patients who received brachytherapy for tumors overhanging the disk [32]. Recent advances in plaque design may improve tumor control in these cases [32–34]. Inability to achieve local control is associated with an increased risk of metastasis and poorer survival [24, 26, 35], suggesting that tumor recurrence is a marker of a particularly aggressive tumor type. In one study, approximately one-half of patients who experienced a tumor recurrence were alive 10  years after proton therapy, compared to almost three-quarters of patients whose tumors were controlled [24].

Eye Loss The probability of retaining the eye for patients treated at the MEEI was 91% at 5 years, 88% at 10  years, and 84% at 15  years after irradiation, and the rates at these time points were virtually identical in patients treated at PSI [36]. Removal of the eye may become necessary after treatment if the tumor recurs or complications develop. Neovascular glaucoma is the most

common complication leading to enucleation [27, 35]. Tumor characteristics that increase the risk of enucleation include proximity of the tumor to critical structures [27, 36, 37], larger tumors (tumor height [26, 27, 35, 36] and tumor diameter [25–27]). In an analysis of patients with large tumors (defined as tumors  ≥  10  mm in height or > 16 mm in diameter or a height greater than 8  mm when the optic nerve was involved) treated at MEEI, enucleation was necessary in approximately 23% of patients 5 years after treatment, and this increased to almost 30% at 10 years after PBI. Five-year rates of enucleation were higher (31%) for tumors located near the optic nerve and fovea and lower (12%) for tumors located farther away from both structures [38]. In another study of patients with large tumors, defined as T3-T4 tumors, 19.5% of the cohort underwent enucleation [39]. Eyes enucleated after proton therapy and examined histopathologically exhibit degenerative and vascular changes [40]. Thickening or thrombosis of the tumor vasculature is a hallmark effect of radiation; in one study, vascular damage was identified over ten times more often in irradiated tumors than non-irradiated tumors [40]. Consistent with these findings, fewer vascular regions were identified by color Doppler imaging in proton-irradiated tumors versus pre-irradiation tumors [41].

Vision Loss Visual acuity after proton beam irradiation depends on tumor height and proximity to the fovea and the optic nerve [42]. In eyes with tumors located farther than 3  mm from these structures, tumor destruction usually occurs without functionally significant radiation vasculopathy [42]. In a group of 558 patients with small- to moderate-sized tumors located within 4 disk diameters of the optic nerve or macula, increased dose to the macula, increased tumor height, poorer baseline vision, and a history of diabetes elevated the risk of vision loss, which was 68% at 5  years after proton therapy [43]. In contrast, the 5-year

225

13  Uveal Melanoma: Proton Beam Radiation Therapy

Table 13.1  Cumulative rates (percent and 95% confidence intervals) of enucleation, vision loss, and melanoma-­related death, by risk groupa and years after proton irradiation Endpoint All patients Loss of eye 5 years 9 (7, 10) 10 years 12 (10, 14) 15 years 16 (13, 20) Vision loss to 20/200 or worse 5 years 52 (50, 55) 10 years 65 (62, 68) 15 years 71 (66, 75) Melanoma-related death 5 years 14 (13, 16) 10 years 23 (21, 25) 15 years 27 (24, 29)

Low risk

Low to medium risk

Medium to high risk

High risk

2 (1, 3) 3 (2, 6) 3 (2, 6)

6 (4, 8) 8 (6, 11) 11 (7, 16)

10 (7, 14) 14 (10, 20) 23 (15, 35)

27 (22, 33) 33 (27, 41) 48 (33, 65)

9 (7, 15) 16 (11, 22) 24 (14, 39)

40 (35, 45) 58 (52, 65) 65 (55, 75)

71 (67, 76) 84 (79, 88) 88 (81, 93)

95 (91, 98) 99 (96, 100) 100 (−,-)

2 (1, 4) 3 (2, 5) 5 (3, 9)

7 (5, 9) 13 (11, 17) 16 (13, 19)

19 (16, 22) 31 (27, 35) 35 (31, 40)

40 (35, 47) 57 (51, 64) 63 (56, 71)

Risk groups derived from values of statistically significant prognostic factors and their coefficients in multivariate Cox regression models (Reprinted from [44]. With permission from Elsevier.)

a

rate of vision loss when all patients were evaluated was 52% (Table 13.1 [44]). Fuss et  al. found that radiation dose greater than 35 CGE to the optic disk or macula was associated with visual deterioration [37], which is similar to the findings of our group [43], and suggest that there may be a threshold dose for deleterious effects. In a series of 349 patients treated with proton therapy with tumors considered unsuitable for brachytherapy or other conservative modalities, tumor height, initial visual acuity, and retinal invasion were identified as risk factors for vision loss to 20/200 or worse, and proximity to the optic nerve and/or macula predicted vision loss to worse than 20/40 [25]. In a multivariate regression model, tumor location within 2 DD of both the optic nerve and macula was the strongest predictor of poor visual outcome; increased risk of vision loss was also associated with baseline visual acuity of 20/50 or worse, history of diabetes, degree of retinal detachment, increased tumor height, and increased tumor diameter [27]. Based on risk scores derived from this model, the 10-year probabilities of visual deterioration (20/200 or worse) were 16% for patients in the “low-risk” category and 99% for patients in the “high-risk” group. In patients with macular tumors (within 1 DD of the fovea and at least 1 DD away from the optic disk), 5-year rates of vision retention were 35.5% (at least 20/200) and 16.2% (at least 20/40), and

visual prognosis was significantly better for patients with less elevated tumors and good baseline visual acuity [23]. Tumor location is often correlated with size, with large tumors often in close proximity to the optic nerve and/or macula. Large tumors are often associated with exudative retinal detachments which further increase risk of visual impairment. At MEEI, almost two-thirds (61%) of tumors classified as large were located within 1 DD of the optic nerve. Although 73% of these patients had a baseline visual acuity of 20/200 or better, this declined to 16% of patients by 5 years after irradiation [38]. In contrast, a 5-year rate of 37% was reported in a small (N  =  77) patient cohort with large non-peripapillary tumors treated with proton therapy [45].

Complications The most serious anterior segment complications are rubeosis iridis and neovascular glaucoma, which increase the risk of vision loss and loss of the eye. In a series of patients with tumors too large to be treated with plaque radiotherapy, larger tumor diameter and presence of retinal detachment were highly significant risk factors (p