Neurology: A Queen Square Textbook [3 ed.] 1119715539, 9781119715535, 9781119715696, 9781119715672

Citation preview

Neurology A Queen Square Textbook

Neurology A Queen Square Textbook Third Edition Edited by

Robin Howard

National Hospital for Neurology and Neurosurgery UCL Queen Square Institute of Neurology London

Dimitri Kullmann

National Hospital for Neurology and Neurosurgery UCL Queen Square Institute of Neurology London

David Werring

National Hospital for Neurology and Neurosurgery UCL Queen Square Institute of Neurology London

Michael Zandi

National Hospital for Neurology and Neurosurgery UCL Queen Square Institute of Neurology London

This third edition first published 2024 © 2024 John Wiley & Sons Ltd All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/permissions. The right of Robin Howard, Dimitri Kullmann, David Werring and Michael Zandi to be identified as the authors of the editorial material in this work has been asserted in accordance with law. Registered Offices John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com. Wiley also publishes its books in a variety of electronic formats and by print-­on-­demand. Some content that appears in standard print versions of this book may not be available in other formats. Trademarks: Wiley and the Wiley logo are trademarks or registered trademarks of John Wiley & Sons, Inc. and/or its affiliates in the United States and other countries and may not be used without written permission. All other trademarks are the property of their respective owners. John Wiley & Sons, Inc. is not associated with any product or vendor mentioned in this book. Limit of Liability/Disclaimer of Warranty The contents of this work are intended to further general scientific research, understanding, and discussion only and are not intended and should not be relied upon as recommending or promoting scientific method, diagnosis, or treatment by physicians for any particular patient. In view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of information relating to the use of medicines, equipment, and devices, the reader is urged to review and evaluate the information provided in the package insert or instructions for each medicine, equipment, or device for, among other things, any changes in the instructions or indication of usage and for added warnings and precautions. While the publisher and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives, written sales materials or promotional statements for this work. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations it may make. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. Library of Congress Cataloging-­in-­Publication Data applied for [HB ISBN: 9781119715535] Cover Design: Wiley Cover Image: Courtesy of Parashkev Nachev Set in 9/11pt Minion by Straive, Pondicherry, India

Contents

Contributors vii

12

Infection in the Nervous System  369 Robin Howard, Laura Benjamin, Sophia De-Saram, Catherine Houlihan and Hadi Manji

13

Cranial Nerve Disorders  433 Robin Howard, Jeremy Chataway and Paul Jarman

14

Spinal Column and Spinal Cord Disorders  463 Tabish A. Saifee, Simon Farmer, Sachit Shah and David Choi

15

Disorders of the Motor Cells: The Motor Neuron Diseases  499 Andrea Malaspina, Jan Clarke, Pietro Fratta, Robin Howard, Ross Nortley, Richard Orrell, Rickie Patani, Katie Sidle and Michaela Waltho

16

Diseases of the Peripheral Nerves  517 Michael Lunn, Alex Rossor and Mary Reilly

17

Decision Making, Ethics and Law in Neurology  37 Jonathan Martin and Alex Ruck Keene

Disorders of the Neuromuscular Junction  561 Jennifer Spillane, Robin Howard, Dimitri Kullmann and Georgiana Logou

18

Neuropathology: Introduction to History, Diagnostic Approaches, Techniques and their Interpretation  53 Sebastian Brandner

Disorders of Muscle  575 Jasper Morrow, Michael Hanna, Matt Parton and Christopher Turner

19

Multiple Sclerosis and Demyelinating Diseases  603 Siobhan Leary, Wallace Brownlee, Noreen Barker, Declan Chard, Jeremy Chataway, Karen Chung, Olga Ciccarelli, Gavin Giovannoni, Nevin John, Zhaleh Khaleeli, Josephine Swanton, Ahmed Toosy, Anand Trip, Heather Wilson and Alan Thompson

20

Neuro-­Oncology  655 Jeremy Rees, Sebastian Brandner, Naomi Fersht, Joan Grieve, Gary Hotton, Michael Kosmin, Vittorio Russo, Lewis Thorne and Steffi Thust

21

Headache 721 Manjit Matharu, Paul Shanahan, Tim Young and Salwa Kamourieh

22

Neuro-­Ophthalmology  741 James Acheson, Fion Bremner, Sara Ajina, Gordon Plant, Robin Howard, Alexander Leff and Ahmed Toosy

23

Neuro-­Otology: Dizziness, Balance and Hearing  797 Diego Kaski, Doris-­Eva Bamiou, Adolfo Bronstein and Nehzat Koohi

24

Cerebellar Ataxias and Related Conditions  839 Nicholas Wood

25

Restorative Neurology, Rehabilitation and Brain Injury  855 Valerie Stevenson, Sara Ajina, Gerry Christofi, Rachel Farrell, Richard Greenwood, Camille Julien, Siobhan Leary, Alexander Leff, Orlando Swayne, Richard Sylvester and Nick Ward

26

Toxic, Metabolic and Physical Insults to the Nervous System 903 Robin Howard, Talal Al-­Mayhani, Aisling Carr, Alexander Leff, Jasper Morrow and Alexander Rossor

About the Editors  xiii Foreword to the Third Edition  xv Preface to First Edition  xvii Preface to Second Edition  xix Preface to Third Edition  xxi Acknowledgements xxiii 1

2

3 4

Global Burden of Neurological Disease and the Neurology of Climate Change  1 Hadi Manji and Sanjay Sisodiya Approach to the Patient with Neurological Disease  17 Robin Howard, Gerry Christofi, Alex Rossor, Jason Warren and David Werring

5

Neuroimaging 67 Frederik Barkhof and Francesco Carletti

6

Neurogenetics 79 Henry Houlden, Andrea Cortese and Edward J. Wild

7

Neuroimmunology 91 Michael Zandi, Aisling Carr, Rachel Brown and Michael Lunn

8

Stroke and Cerebrovascular Diseases  107 David Werring, Matthew Adams, Laura Benjamin, Martin Brown, Arvind Chandratheva, Peter Cowley, Joan Grieve, Fiona Humphries, Hans Rolf Jäger, Nicholas Losseff, Richard Perry, Robert Simister and Ahmed Toma

9

10

11

Movement Disorders  199 Thomas Foltynie, Kailash Bhatia, Carla Cordivari, Eileen Joyce, Prasad Korlipara, Patricia Limousin, Tabish A. Saifee, Sarah Tabrizi and Thomas Warner Epilepsy and Related Disorders  247 John Duncan, Josemir Sander, Ali Alim-­Marvasti, Simona Balestrini, Sallie Baxendale, Dorothea Bindman, Krishna Chinthapalli, Fahmida Chowdhury, Beate Diehl, Sofia Eriksson, Jackie Foong, Dominic Heaney, Sofia Khan, Matthias Koepp, Dimitri Kullmann, Sanjeev Rajakulendran, Fergus Rugg-­Gunn, Meneka Sidhu, Sanjay Sisodiya, Jane de Tisi, Maria Thom, Matthew Walker and Mahinda Yogarajah Cognitive Impairment and Dementia  319 Jason Warren, John Collinge, Nick Fox, Simon Mead, Catherine Mummery, Jonathan Rohrer, Martin Rossor, Jonathan Schott and Rimona Weil

v

vi

Contents

27

Inherited Disorders of Metabolism  945 Elaine Murphy, Charlotte Ellerton, Simon Heales, Robin Lachmann, David Lynch and Robert Pitceathly

32

Autonomic Aspects of Neurology  1097 Valeria Iodice, Gordon Ingle, Christopher Mathias and Patricia McNamara

28

Disorders of Consciousness and Intensive Care Neurology 987 Robin Howard, Dimitri Kullmann, Andrew Paget, Jennifer Spillane and Manni Waraich

33

Uro-­Neurology  1137 Jalesh N. Panicker, Sara Simeoni and Mahreen Pakzad

34

Systemic Conditions and Neurology  1155 David Werring, Aisling Carr, Robin Howard, Dimitri Kullmann and Michael Zandi

35

Palliative Care in Neurology  1203 Jonathan Martin, Jan Clarke, Robin Howard and Michaela Waltho

29

Disorders of Sleep  1049 Sofia Eriksson, Robin Howard, Sofia Khan and Matthew Walker

30

Neuropsychiatry 1061 Eileen Joyce and Dorothea Bindman

31

Pain in Neurological Disorders  1075 Alan Fayaz and Anupam Bhattacharjee

Index 1219

Contributors

Institution addresses:

National Hospital for Neurology and Neurosurgery, Queen Square, London WC1N 3BG UCL Queen Square Institute of Neurology, Queen Square, London WC1N 3BG

Mr James Acheson MRCP FRCS FRCOphth

Dr Dorothea Bindman MRCP MRCPsych Consultant Neuropsychiatrist National Hospital for Neurology and Neurosurgery

Professor Sebastian Brandner MD FRCPath

Consultant Neuro-­ophthalmologist National Hospital for Neurology and Neurosurgery

Professor of Neuropathology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Matthew Adams FRCR

Mr Fion Bremner PhD FRCOphth

Dr Sara Ajina DPhil MRCP

Professor Adolfo Bronstein PhD FRCP

Consultant Neuroradiologist National Hospital for Neurology and Neurosurgery

Consultant in Rehabilitation Medicine National Hospital for Neurology and Neurosurgery

Dr Talal Al-­Mayhani MRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Dr Ali Alim-­Marvasti PhD MRCP

Honorary Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Simona Balestrini MD PhD

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Professor Doris-­Eva Bamiou PhD FRCP

Consultant Audiovestibular Physician and Professor of Neuro-­otology National Hospital for Neurology and Neurosurgery and Ear Institute, University College London

Ms Noreen Barker BSc RN NMP

Consultant Nurse National Hospital for Neurology and Neurosurgery

Consultant Neuro-­ophthalmologist National Hospital for Neurology and Neurosurgery Consultant Neurologist and Professor of Clinical Neuro-­otology National Hospital for Neurology and Neurosurgery and Imperial College

Professor Martin Brown DM FRCP Emeritus Professor of Stroke Medicine UCL Queen Square Institute of Neurology

Dr Rachel Brown MRCP

Clinical Fellow National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Wallace Brownlee PhD FRACP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Dr Francesco Carletti MD PhD

Consultant Neuroradiologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Aisling Carr PhD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Arvind Chandratheva DPhil MRCP

Professor of Neuroradiology National Hospital for Neurology and Neurosurgery

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Sallie Baxendale PhD FBPS

Dr Declan Chard PhD FRCP

Consultant Clinical Neuropsychologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Principal Research Fellow and Honorary Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Laura Benjamin PhD FRCP

Professor Jeremy Chataway PhD FRCP

Principal Research Fellow and Honorary Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Professor Kailash Bhatia MD DM FRCP

Dr Krishna Chinthapalli PhD FRCP

Professor Frederik Barkhof MD PhD FRCR

Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Dr Anupam Bhattacharjee PhD MRCP

Consultant Neurosurgeon National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Mr David Choi PhD FRCS (SN)

vii

viii

Contributors

Dr Fahmida Chowdhury PhD MRCP

Consultant in Clinical Neurophysiology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Gerry Christofi PhD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Dr Karen Chung MD MRCP

Dr Alan Fayaz MRCP FRCA FFPMRCA Consultant in Anaesthesia and Pain Medicine National Hospital for Neurology and Neurosurgery

Dr Naomi Fersht PhD MRCP(UK) FRCR

Consultant Clinical Oncologist Department of Clinical Oncology, University College London Hospital

Professor Thomas Foltynie PhD MRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Professor Olga Ciccarelli PhD FRCP

Dr Jackie Foong PhD MRCPsych

NHR Research Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Consultant Neuropsychiatrist National Hospital for Neurology and Neurosurgery

Ms Jan Clarke RN MA

Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Consultant Nurse Specialist in Motor Neuron Disease National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Professor John Collinge CBE MD FRCP FRS

Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Institute of Prion Diseases

Dr Carla Cordivari MD FRCP

Consultant in Clinical Neurophysiology National Hospital for Neurology and Neurosurgery

Dr Andrea Cortese MD PhD

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Peter Cowley FRCR

Consultant Neuroradiologist National Hospital for Neurology and Neurosurgery

Ms Jane de Tisi BA

Data Manager, Epilepsy Team National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Sophia De Saram FRCPath MRCP DTM&H Consultant in Microbiology and Infectious Diseases National Hospital for Neurology and Neurosurgery

Professor Beate Diehl MD PhD FRCP

Professor Nick Fox MD FRCP FMedSci

Professor Pietro Fratta MD PhD

Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Professor Gavin Giovannoni PhD FRCP FRCPath Honorary Professor of Neurology UCL Queen Square Institute of Neurology

Dr Richard Greenwood MD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Miss Joan Grieve MD FRCS (SN)

Consultant Neurosurgeon National Hospital for Neurology and Neurosurgery

Professor Michael Hanna MD FRCP

Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Simon Heales PhD FRCPath

Consultant Clinical Scientist and UCL Professorial Research Fellow National Hospital for Neurology and Neurosurgery

Dr Dominic Heaney PhD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Professor of Neurophysiology applied to Epilepsy National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Gary Hotton MD FRCP

Professor John Duncan DM FRCP FMedSci

Professor Henry Houlden PhD MRCP(UK)

Professor of Clinical Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Ms Charlotte Ellerton BSc PGDip

Dr Catherine Houlihan PhD MRCP(UK) FRCPath (Virol)

Specialist Dietetic Practitioner in Metabolics National Hospital for Neurology and Neurosurgery

Dr Sofia Eriksson PhD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Simon Farmer PhD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Rachel Farrell PhD MRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Consultant Clinical Virologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Robin Howard PhD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Fiona Humphries FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Gordon Ingle MD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Contributors  ix

Dr Valeria Iodice MD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Professor Hans Rolf Jäger MD FRCR

Professor of Neuroradiology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Paul Jarman PhD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Nevin John PhD FRACP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Professor Eileen Joyce PhD MRCP(UK) FRCPsych

Professor of Neuropsychiatry National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Camille Julien PhD DClinPsy

Consultant Clinical Neuropsychologist Homerton Healthcare NHS Foundation Trust and Kings College Hospital Foundation NHS Trust

Dr Salwa Kamourieh PhD MRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Diego Kaski PhD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Zhaleh Khaleeli MD (Res) FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Dr Sofia Khan MD (Res) FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Professor Matthias Koepp MD PhD

Professor Alexander Leff PhD FRCP Professor of Cognitive Neurology UCL Queen Square Institute of Neurology

Professor Patricia Limousin MD PhD

Professor of Clinical Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Ms Georgiana Logou RN

Clinical Nurse Specialist Myasthenia National Hospital for Neurology and Neurosurgery

Dr Nicholas Losseff MD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Professor Michael Lunn PhD FRCP

Professor of Clinical Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr David Lynch PhD MRCPI

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Professor Andrea Malaspina PhD FRCP

Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Hadi Manji MD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Jonathan Martin FRCP LLM

Consultant in Palliative Medicine National Hospital for Neurology and Neurosurgery

Professor Manjit Matharu PhD FRCP

Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Professor Christopher Mathias DPhil DSc FRCP FMedSci Honorary Emeritus Professor of Neurology UCL Queen Square Institute of Neurology

Professor of Clinical Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Patricia McNamara PhD FRCP

Dr Nehzat Koohi PhD

Professor Simon Mead PhD FRCP

Senior Audiological Scientist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Consultant Neurologist National Hospital for Neurology and Neurosurgery Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Institute of Prion Diseases

Dr Prasad Korlipara PhD FRCP

Dr Jasper Morrow PhD FRACP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Dr Michael Kosmin FRCP MD (Res)

Dr Catherine Mummery PhD FRCP

Consultant Clinical Oncologist Department of Clinical Oncology, University College London Hospital

Professor Dimitri Kullmann DPhil FRCP FMedSci FRS

Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Robin Lachmann PhD FRCP

Consultant in Adult Inherited Metabolic Disease National Hospital for Neurology and Neurosurgery

Dr Siobhan Leary MD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Elaine Murphy MRCPI FRCPath

Consultant in Adult Inherited Metabolic Disease National Hospital for Neurology and Neurosurgery

Professor Parashkev Nachev PhD FRCP Professor of Neurology UCL Queen Square Institute of Neurology

Dr Ross Nortley PhD MRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

x

Contributors

Dr Richard Orrell MD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Andrew Paget DClinPsy

Consultant Clinical Neuropsychologist National Hospital for Neurology and Neurosurgery

Ms Mahreen Pakzad MD FRCS(Urol) Consultant Urologist National Hospital for Neurology and Neurosurgery

Professor Jalesh N. Panicker MD DM FRCP

Professor in Uro-­neurology and Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Matt Parton PhD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Professor Rickie Patani PhD FRCP

Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Richard Perry PhD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Dr Robert Pitceathly MB ChB PhD

Principal Research Fellow and Honorary Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Gordon Plant MD FRCP FRCOphth

Honorary Associate Professor National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Sanjeev Rajakulendran PhD FRCP Consultant Neurologist National Hospital for Neurology and Neurosurgery

Dr Jeremy Rees PhD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Professor Mary Reilly MD FRCP FRCPI

Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Professor Jonathan Rohrer PhD MRCP

MRC Clinical Scientist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Alex Rossor PhD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Professor Martin Rossor MD FRCP FMedSci

Mr Vittorio Russo FRCS (SN)

Consultant Neurosurgeon National Hospital for Neurology and Neurosurgery

Dr Tabish Saifee PhD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Professor Josemir Sander MD PhD FRCP

Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Professor Jonathan Schott MD FRCP

Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Sachit Shah MD FRCR

Consultant Neuroradiologist National Hospital for Neurology and Neurosurgery

Dr Paul Shanahan MRCPI

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Dr Meneka Sidhu PhD MRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Katie Sidle PhD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Sara Simeoni MD

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Dr Robert Simister PhD MRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Professor Sanjay Sisodiya PhD FRCP

Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Jennifer Spillane PhD MRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Dr Valerie Stevenson MD MRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Dr Josephine Swanton PhD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Orlando Swayne PhD MRCP

Mr Alex Ruck Keene KC (Hon)

Dr Richard Sylvester PhD FRCP

Barrister 39 Essex Chambers and Visiting Professor, King’s College London

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Dr Fergus Rugg-­Gunn PhD FRCP

Professor Sarah Tabrizi PhD FRCP FMedSci

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Contributors  xi

Professor Maria Thom FRCPath MD

Professor Thomas Warner PhD FRCP

Professor Alan Thompson MD FRCPI FRCP FMedSci

Professor Jason Warren PhD FRACP

Professor Lewis Thorne PhD FRCP (SN)

Professor Rimona Weil PhD MRCP

Professor of Neuropathology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology Professor of Neurology National Hospital for Neurology and Neurosurgery

Dr Steffi Thust MD FRCR

Consultant in Diagnostic Neuroradiology National Hospital for Neurology and Neurosurgery

Mr Ahmed Toma MD (Res) FRCS (SN) Consultant Neurosurgeon National Hospital for Neurology and Neurosurgery

Professor Ahmed Toosy PhD MRCP

Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Anand Trip PhD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Dr Christopher Turner PhD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Professor Matthew Walker PhD FRCP

Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Ms Michaela Waltho RN

Clinical Nurse Specialist in Motor Neuron Disease National Hospital for Neurology and Neurosurgery

Dr Manni Waraich MBBS FRCA FFICM Consultant in Neurocritical Care National Hospital for Neurology and Neurosurgery

Professor Nick Ward MD FRCP

Professor of Neurology UCL Queen Square Institute of Neurology and National Hospital for Neurology and Neurosurgery

Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology Professor of Neurology National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Professor David Werring PhD FRCP

Professor of Neurology UCL Queen Square Institute of Neurology and National Hospital for Neurology and Neurosurgery

Professor Edward Wild PhD FRCP

Professor of Neurology UCL Queen Square Institute of Neurology and National Hospital for Neurology and Neurosurgery

Dr Heather Wilson PhD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery

Professor Nicholas Wood PhD FRCP FMedSci

Galton Chair of Genetics National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Dr Mahinda Yogarajah PhD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

Professor Tim Young PhD FRCP Clinical Professor (Teaching) UCL Queen Square Institute of Neurology

Dr Michael Zandi PhD FRCP

Consultant Neurologist National Hospital for Neurology and Neurosurgery and UCL Queen Square Institute of Neurology

About the Editors

Robin Howard PhD FRCP FFICM

Robin Howard trained in medicine in Cambridge and at the Middlesex Hospital. His neurology training was in Oxford, London, and at the National Hospital Queen Square. He undertook his PhD in the Sobell Department of Neurophysiology at the Institute of Neurology. He has been a consultant neurologist at the National Hospital, Queen Square, and St. Thomas’ Hospital, honorary Associate professor at the UCL Queen Square Institute of Neurology, University College London, and honorary civilian adviser to the Royal Navy since 1992. He was head of service for a large general neurological practice at St. Thomas’ Hospital and neurologist to three intensive care units. He was senior neurologist to specialist units for the care of patients with myasthenia gravis, motor neuron disease, Duchenne muscular dystrophy and post-­polio syndrome at both Queen Square and St  Thomas’ hospitals and has written and lectured extensively on each of the subjects. He has been a senior editor and contributor to all three editions of Neurology: A Queen Square Textbook.

Dimitri Kullmann FRCP FMedSci FRS

Dimitri Kullmann trained in medicine in Oxford and London and completed a DPhil in Oxford. Following a postdoctoral fellowship at the University of California San Francisco, he trained in neurology at the National Hospital for Neurology and Neurosurgery, Queen Square, and established a laboratory focusing on synaptic transmission at the Institute of Neurology. He is now a professor of  neurology at the UCL Queen Square Institute of Neurology and  honorary consultant neurologist at the National Hospital for Neurology and Neurosurgery. His research interests include the fundamental mechanisms of synapse function, neurological channelopathies and gene therapy for epilepsy. He was the Editor of Brain from 2014 to 2020 and is on the Editorial Board of Neuron. He was made a Fellow of the Academy of Medical Sciences in 2001 and of the Royal Society in 2018.

David Werring PhD FRCP FESO

David Werring trained in medicine at Guy’s Hospital Medical School and in neurology in London. He was appointed as a consultant neurologist at the National Hospital for Neurology and Neurosurgery, Queen Square in 2005. He is Professor of Clinical Neurology at the Stroke Research Centre, UCL Queen Square Institute of Neurology, and honorary consultant neurologist at the National Hospital for Neurology and Neurosurgery. He delivers acute and outpatient stroke care and leads a research programme focussed on intracerebral haemorrhage and cerebral small vessel disease. He is head of the Research Department of Brain Repair and Rehabilitation, chair of the Association of British Neurologists Stroke Advisory Group, stroke specialty lead for the National Institute for Health Research North Thames Clinical Research Network, President-­Elect of the British and Irish Association of Stroke Physicians, and editorial board member of the European Journal of Neurology, European Stroke Journal, International Journal of Stroke and Practical Neurology. David Werring chaired the UK Stroke Forum 2020–2022.

Michael Zandi PhD FRCP

Michael Zandi trained in medicine in Cambridge, and completed neurology training in Cambridge, Norwich and London, and a PhD  in Cambridge with time in the laboratory of Professor Angela Vincent at the Weatherall Institute of Molecular Medicine in Oxford. He is a ­consultant neurologist at the National Hospital for Neurology and Neurosurgery and Honorary Associate Professor at the UCL Queen Square Institute of Neurology. His research interests include the mechanisms, natural history and clinical treatments of autoimmune encephalitis, cerebral amyloid angiopathy related inflammation, neuroimmunology broadly, and the role of inflammation and autoimmunity in cognitive and psychiatric disorders. He has advised NHS England, the Royal College of Physicians and the National Institute for Health and Care Excellence.

xiii

Foreword to the Third Edition

I am delighted to be asked to introduce the publication of this third edition of Neurology: A Queen Square Textbook. Although the field of neurology continues to become ever more specialised, partly as a result of major progress in neuroscience ­discoveries, this new third edition continues the tradition of presenting a clear useful coherent text that spans the whole of the field. Previously, neurologists dealt with all neurological conditions; today, specialisation is the norm and to be a neurologist without a special interest is unusual. However, it remains very important to have a solid grounding in general neurology, even for those who ultimately become highly specialised. It is thus fitting that the four editors combine very broad general neurology clinical experience with specialist academic expertise. Robin Howard, who works jointly at Guy’s and St. Thomas’ Hospital in London and the National Hospital, is a highly experienced general neurologist; his specialist interests are intensive care neurology and neuromuscular disease. Dave Werring initially practised mainly generally neurology at Watford Hospital but has developed a major interest in stroke and now has an international reputation in clinical research in stroke based at the UCL Queen Square Institute of Neurology. Dimitri Kullmann clinically specialises in intensive care neurology and has made major contributions to the understanding of synaptic physiology in the central and peripheral nervous system, through his laboratory research at the Institute of Neurology. Michael Zandi is an internationally recognised specialist in clinical and laboratory neuroimmunology based at the National Hospital. In each carefully constructed chapter, this book continues to epitomise this combination of practical experience and academic specialisation. All aspects of all neurological disorders are dealt with in a very readable and instructive way. Indeed, the teams that have been assembled to write each chapter are colleagues that ­collaborate between institute and hospital on a daily basis. Together, this whole team of consultants assesses around 200,000 outpatients per year at the National Hospital and are all excellent teachers. Together, they have huge expertise in all aspects of neurological ­disease and are actively engaged in teaching both as a national ­centre and internationally to Queen Square postgraduate students from all over the world. Advances in neuroscience have in part led to a bewildering array of diagnostic tests including specialised MRI, PET imaging,

specialised neurophysiology, liquid biomarkers, gene panels and whole genome sequencing but it remains important that careful clinical assessment and examination of the patient by the neurologist is central to good neurological practice. The tendency and ease of reaching for many elaborate tests is no substitute for careful ­clinical evaluation. Indeed, the neurologist undertaking a careful and detailed history and examination is part of the ‘DNA’ of Queen Square. This principle is evident throughout the book, highlighting that a careful clinical assessment will avoid over- investigation, ­enable the correct investigations and, importantly, allow appropriate correlation between the clinical findings and investigation results. The editors and the authors are experienced and distinguished writers who have devoted time to draw together their practical experience. Each chapter has been very carefully edited, and this is evident in the finished product. One can find here well‐illustrated neuroanatomy, detailed assessments of common conditions such as stroke and dementia, up‐to‐date aspects of neurogenetics and ion channels, the philosophy and practicalities of rehabilitation, and rarities such as metabolic disorders of copper, and even unusual muscle diseases. I know that this book has been produced within an atmosphere of cordiality and friendship at Queen Square, and am pleased to have been able to make a small contribution myself. It is clear from this third edition just how large a subject neurology has become and how scientific advances, many pioneered within the UCL Queen Square Institute of Neurology, have become translated into clinical practice. I believe this edition, and indeed future editions, will continue to be essential reading for trainee and indeed trained neurologists! I am delighted that Neurology: A Queen Square Textbook has become a standard text for the UK neurologists in training exit examination – a measure of its standing, quality and usefulness. I want to sincerely congratulate all the authors on this great achievement and this very important contribution to the clinical practice of neurology globally. Professor Michael Hanna BSc(Hons) MBChB(Hons) MD FRCP(UK) FMedSci Professor of Clinical Neurology Director, UCL Queen Square Institute of Neurology

xv

Preface to First Edition

All Editors, Authors and Specialist Advisory Editors of Neurology: A Queen Square Textbook hold or recently held consultant or equivalent posts at the National Hospital for Neurology and Neurosurgery and/or the Institute of Neurology, Queen Square. The National Hospital is part of University College London Hospitals NHS Foundation Trust, and the Institute of Neurology part of University College London. The 20 co-­ ordinating authors organised individual chapters, encouraged and liaised with over 70 contributors and with them wrote this book. The specialist advisory editors gave invaluable advice and guidance in their respective fields. To ensure a worldwide perspective, the six International Regional Editors, all of whom have had close connections with Queen Square, provided advice and comment. This book is an attempt to provide a fresh and up-­ to-­ date approach to the fascinating subject of neurology. We encouraged each author to relate their own clinical experience but, in order to

achieve a degree of consistency, we took a robust overview of the important specialities within neurology and their relevance. Each chapter has been coordinated by an expert in the field, to give the reader an overall grasp of each major subject, indicating where developments within neurosciences fit into a broader picture. The limited size of this book means that it has not been possible to provide references for all material. With the growth of information technology, a wealth of detailed sources are readily available. We are most grateful to all those who have helped in this joint venture. Charles Clarke Robin Howard Martin Rossor Simon Shorvon Queen Square London WC1

xvii

Preface to Second Edition

All editors, authors and specialist advisory editors of Neurology: A Queen Square Textbook hold or recently held consultant or equivalent posts at the National Hospital for Neurology and Neurosurgery and/ or the UCL Queen Square Institute of Neurology, Queen Square. The National Hospital is part of University College London Hospitals NHS Foundation Trust, and the Institute of Neurology part of University College London. Twenty-­three co-­ordinating authors organised individual chapters, encouraged and liaised with over 70 contributors and with them wrote this book. The specialist advisory editors gave invaluable advice and guidance in their respective fields. To ensure a worldwide perspective, for the first edition our international regional editors, all of whom had close connections with Queen Square, provided guidance and comment. This book is an attempt to provide a fresh and up-­ to-­ date approach to the fascinating subject of neurology. We encouraged each author to relate their own clinical experience but, in order to achieve a degree of consistency, we took a robust overview of the important specialities within neurology and their relevance. Each chapter has been coordinated by an expert in the field, to give the reader an overall grasp of each major subject, indicating where developments within neurosciences fit into a broader picture.

On spelling and use of the English language, whilst appreciating that as a living, multicultural tongue there are wide varieties, we have opted for British English – the sort of way we write our letters, and continue to spell ‘neurone’ with its terminal -­e. On medical conditions named after famous figures, we appreciate that many publishers no longer use the apostrophe to describe the disease named after Alzheimer, Wilson, Parkinson and so on. Our authors by and large did not follow this; thus we have left matters much as they signed off their chapters. The limited size of this book means that it has not been possible to provide references for all material. With the growth of information technology, a wealth of detailed sources is readily available. We are most grateful to all those who have helped in this joint venture. Charles Clarke Robin Howard Martin Rossor Simon Shorvon Queen Square London WC1

xix

Preface to Third Edition

The past five decades have witnessed breathtaking advances in ­clinical medicine and the medical sciences. This rapid change has been particularly striking in the neurosciences, encompassing advances in genetics, neuroimaging, immunology and the interpretation of the molecular and metabolic processes that underlie ­neural function. Many of these dramatic developments have been reflected in our understanding of the mechanism of disease and translated into investigation and treatment. The pace of scientific and clinical innovation has posed challenges of increasing complexity and cost, potentially limiting adequate provision of the highest standard of healthcare. Indeed, the downside of rapid technological advance has been that inequalities of care continue to widen the divide between the more and less economically developed nations. Furthermore, increasing healthcare costs mean that good care becomes vulnerable to interference by government, private companies and entrepreneurs in ways that might threaten scientific and medical independence. The demand for neurological services has also greatly increased in recent years due to ageing populations susceptible to neurodegenerative and cerebrovascular diseases, the burgeoning multiplicity of specific treatments available for neurological diseases and advances in the management of disability. The cost of treating patients has escalated beyond all expectation, leading to a constant demand to balance benefit against the ever-­increasing and often unrealistic expenditure necessary to keep pace with scientific advances. These spiralling costs increasingly fall to society as a whole; many debate individual and collective responsibility for ensuring that the most vulnerable receive appropriate care. There are challenges to neurology that are particular to the more and less economically developed world and many of them have been addressed throughout this book. The newer technologies such as artificial intelligence, machine learning, genetic manipulation and access to large-­scale electronic health records, threaten many of the principles and ethical precepts upon which medical care has been based; their successful implementation requires vigilance and oversight. Their development has led to fundamental changes affecting the clinical and scientific practice of the specialty.

The burgeoning of protocol-­driven care and multiple national and international guidelines is a relatively new phenomenon, which has  rapidly come to dominate clinical practice and is dependent upon the globalisation and interconnection of healthcare. Clinical  and academic neurologists have been in the forefront of scientific advances throughout the history of modern medicine but it could also be argued that, despite the evolution from diagnosis to the treatment, management of neurological conditions and care of the neurological patient, the fundamental clinical approach and work of the clinical neurologist still resemble the practice of Romberg, Duchenne, Charcot, Hughlings Jackson and Gowers. Queen Square, together with other centres in the UK, Europe and throughout the world, was (and is) a cradle of neurology and the neurosciences. The hospital and the institute have made extensive contributions to the remarkable development of the specialty in a bewildering array of disparate fields. Nonetheless, it is essential to recognise that our specialty now exists as a global congress of interested parties. Progress in medicine evolves because of the activities of innumerable friends and colleagues working in specialist and general hospitals and academic institutes in all countries. We will continue to wrestle with the prodigious challenges that face all societies, including the evolution of disease and pandemics, global inequalities of poverty and opportunity, the threat to our environment from climate change and the failings of our politicians. We have tried to address these issues in this book in the context of the extraordinary progress of science and clinical medicine within our specialty. Despite all the developments and challenges of recent decades, it remains self-­evident that the patient lies at the heart of all we do and that excellence of care for those with neurological disease extends beyond treatment alone. Robin Howard Dimitri Kullmann David Werring Michael Zandi Queen Square, London December 2023

xxi

Acknowledgements

This book reflects the efforts of countless clinicians and scientists, most of whom remain unrecognised throughout the text, although their contributions are huge. The editors and authors wish to make it clear that they have only been able to reference a minimal number of those upon whom they have lent heavily and they hope that no offence will be taken where reference to work has not been fully cited because of the lack of space. The editors apologise for any omissions or misrepresentation of fact. They have each maintained a heavy clinical commitment throughout the preparation of this book and have therefore been grateful for the initiative of the lead and contributing authors, to whom they pay tribute. The present editors are particularly grateful to the three ­previous editors who graciously handed over their responsibilities: Dr Charles Clarke, Professor Simon Shorvon and Professor Martin Rossor. Each made an enormous contribution to previous editions of this book. We would like to thank all those who have taught, advised, guided and inspired us to develop our clinical, academic and research studies in neurology and the neurosciences. As noted in previous editions, the list of these individuals is vast and we are unable to mention them all by name. We can simply hope to pass on their wisdom to our own students and colleagues. The authors and editors were distressed to learn of the passing of Linda Luxon. She was an inspiration to patients, students and colleagues. We will miss her wisdom. We thank those who contributed to the second edition and have moved on, reflecting retirement or promotion. We thank our publishers, Wiley, and especially Sophie Bradwell, Bhavya Boopathi, Ella Elliott, Katherine King and Adalfin Jayasingh

all the team who were involved in production. We are particularly grateful to our freelance copyeditor, Jane Moody. The authors have been unstinting in their support of the project throughout the extremely difficult times we experienced during the preparation of this book. Many were redeployed to acute services during the worst of the COVID-19 pandemic, when the demands on their expertise were beyond anything previously experienced. Despite these challenges, they were all entirely supportive and remained patient throughout the numerous and burdensome requests from the editors. The Rockefeller Library provided its valuable resources, both ­historical and current. The Audio Visual Services Unit was most helpful with the sourcing of some figures and photographs. Royalties from Neurology: A Queen Square Textbook are passed directly to the National Brain Appeal (National Hospital Development Foundation), the registered UK charity (No. 290173) that supports projects at Queen Square. We lost many of our vulnerable and disabled patients to the ­ravages of COVID-19 through successive waves of the disease and we remember those who died and those whose lives were changed forever. Robin Howard wishes to put on record his immense gratitude to his many friends and colleagues both past and present at Guy’s and St Thomas’ Hospitals.

Cover illustration (Courtesy of Professor Paraskev Nachev) – Raytraced projection onto a glass sphere of the graph hierarchical community structure of the neural determinants of impaired performance on Advanced Progressive Matrices as derived from graph lesion-deficit mapping of patients with focal brain injury. The

spherical projection onto a glass surface is intended to convey the challenge of obtaining a perspicuous representation of the neural basis of fluid intelligence (Cipolotti L, Ruffle JK, Mole J, Xu T, Hyare H, Shallice T, Chan E, Nachev P. Graph lesion-deficit mapping of fluid intelligence. Brain. 2023 Jan;146(1):167-81.)

Robin Howard Dimitri Kullmann David Werring Michael Zandi November 2023

xxiii

CHAPTER 1

Global Burden of Neurological Disease and the Neurology of Climate Change Hadi Manji1 and Sanjay Sisodiya1,2 National Hospital for Neurology and Neurosurgery, Queen Square, London, UK Epilepsy Department, National Hospital for Neurology and Neurosurgery and Department of Clinical and Experimental Epilepsy, UCL Queen Square Institute of Neurology, Queen Square, London, UK 1 2

Introduction

The global population continues to grow, albeit at a slower pace than previously. According to data from the United Nations (UN) Department of Economic and Social Affairs, it is estimated that population numbers will peak in the 2080s at around 10.4 billion. However, the asymmetry in population growth between different areas of the world is striking: more than half of the projected increase to 2050  will be concentrated in eight low-­income countries: the Democratic Republic of the Congo, Egypt, Ethiopia, India, Nigeria, Pakistan, the Philippines and Tanzania. In contrast, 61 countries and areas with ageing populations, resulting from increased lifespans and reduced fertility rates, the population is projected to decrease by 1% between 2022 and 2050. India is expected to surpass China as the world’s most populous country in 2023 and will have an estimated population of 1.09 billion by 2100. Japan’s population, in contrast, will decline from 128  million in 2017 to 60 million over the same period (Figure 1.1). The other well-­recognised demographic is that the share of the global population aged 65 years and over is projected to rise from 9.7% in 2022 to 16.4% in 2050 (Figures 1.2 and 1.3). In Europe and North America, by 2050 one in four (26.9%, up from 18.7% in 2022) will be aged over 65 years. In Eastern and Southeast Asia, the proportion aged 65 years or older could increase from 13% in 2022 to 26% in 2050. By 2100, the over 60s age group will outnumber the under 5 years age group by two to one. In parallel with the increase of ageing populations there will be an increased incidence and ­prevalence in age-­related neurological disorders including stroke, neurodegenerative conditions such as Alzheimer’s and Parkinson’s diseases, for which there is no prospect of a cure in the immediate future. Governments and planners will need to provide for the long-­ term care of individuals living with cognitive and motor impairment. The COVID-­19 pandemic has highlighted the importance of epidemiological studies for governments, healthcare policy makers, funders and other stake holders in planning and resource allocation. The global unpreparedness for a pandemic resulted in confusing messages from ‘experts’ (mask wearing/no mask wearing; lockdown/no lock down) with the resulting unnecessary loss of life. However, this pandemic has also shown how the medical and scientific community can harness global talents with cooperation and good will to overcome such a once in a lifetime pandemic. The same commitment is required for healthcare planning, resource

allocation and public health education to deal with the global burden of diseases and, in particular, the massive implications of the increasing incidence of neurological disorders. The essential elements of good public health and healthcare are well rehearsed – an adequate health work force; effective, affordable, safe and high-­quality service delivery, access to medicines and diagnostics, health education, good governance and adequate funding. However, health is not simply a question of healthcare provision but must include a broader remit including improving socioeconomic status, gender equality and, most importantly, education, which underpins all these issues. More recently, it has become apparent that topics such as migration, climate change and pollution must also be addressed under the broader heading of health. Gains in life expectancy have not been matched by increases in healthy life expectancy (the number of years a person can expect to live in good health). To improve health, policy makers need up to date data on the challenges faced in their individual countries. Apart from identifying incidence, prevalence and mortality, it is also necessary to identify modifiable risk factors such as smoking, alcohol intake, hypertension and diet. The definition of terms used in epidemiological discussion are listed in Table 1.1.

The Global Burden of Neurological Diseases

The Global Burden of Diseases Injuries and Risk Factors (GBD) study was inaugurated in 1990 with funding from the World Bank. In 2007, the Institute of Health Measurement and Evaluation was established at the University of Washington, Seattle, funded by the Bill and Melinda Gates Foundation, and has continued the study. This iterative process aims to provide a systemic, rigorous and scientific assessment of published, publicly available and contributed data on disease incidence, prevalence and mortality. A study of this vast scope and ambition has many limitations. However, a major strength has been to overcome the inclination of ‘no data, no problem’ by producing a best estimate with an estimate of the degree of uncertainty when data are unavailable or inconsistent. The most recent publication, GBD study 2019, describes 369 diseases and injuries in 204 countries and territories (GBD  2019 Diseases and Injuries Collaborators 2020).

Neurology: A Queen Square Textbook, Third Edition. Edited by Robin Howard, Dimitri Kullmann, David Werring and Michael Zandi. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd. 1

Population shifts Top ten countries by population in

2017

&

2100

Year

Year

2017

2100

1

1.4b China

India 1.09b

1

2

1.38b India

Nigeria 791m

2

3

325m

China 732m

3

4

258m Indonesia

USA 336m

4

5

214m Pakistan

Pakistan 248m

5

6

212m Brazil

DR Congo 246m

6

7

206m Nigeria

Indonesia 229m

7

8

157m Bangladesh

Ethiopia 223m

8

9

146m Russia

Egypt 199m

9

10

128m Japan

Tanzania 186m

...

...

13 14 18

10

...

... ...

103m Ethiopia

Brazil 165m

13

96m Egypt

Russia 106m

19

DR Congo

Bangladesh 60m 25

81m

24

USA

54m

Japan 81m

Tanzania

38

Figure 1.1  Population shifts; top 10 countries by population. Source: reproduced with permission from Vollset et al. (2020). 2017 Male

Female

95 plus 90 to 94 85 to 89

Age group (years)

80 to 84 75 to 79 70 to 74 65 to 69 60 to 64 55 to 59 50 to 54 45 to 49 40 to 44 35 to 39 30 to 34 25 to 29 20 to 24 15 to 19 10 to 14 5 to 9 0 to 4 400

200

0

200

400

Global population (millions)

Figure 1.2  Global population age distribution 2017. Source: reproduced with permission from Vollset et al. 2020).

2100 Male

Female

95 plus 90 to 94 85 to 89

Age group (years)

80 to 84 75 to 79 70 to 74 65 to 69 60 to 64 55 to 59 50 to 54 45 to 49 40 to 44 35 to 39 30 to 34 25 to 29 20 to 24 15 to 19 10 to 14 5 to 9 0 to 4 400

200

0

200

400

Global population (millions)

Figure 1.3  Predicted global population age distribution for 2100. Source: reproduced with permission from Vollset et al. (2020). Table 1.1  Definitions. Term

Definition

Age standardisation

Statistical technique to compare populations with differing age structures in which the characteristics of populations are statistically transformed to match those of a reference population. This is a useful measurement because relative under-­or overrepresentations of different age groups can obscure comparisons of age-­dependent differences across populations

Disability-­adjusted life-­year (DALY)

A measure combining years of life lost due to premature mortality with years of life lost due to time lived in poor health DALY = years of life lost + years lived with disability 1 DALY is a year of perfect health lost Diseases that cause long-­term disability, such as mental health or migraine, may lead to a large proportion of DALYs without being a major cause of mortality The advantage of DALY is that it provides a composite, internally consistent measure of population health, which can be used to evaluate the relative burden of different diseases and injuries and compare population health by geographical region and over time

Incidence

The number of new cases of the disease in a defined population over a defined period

Prevalence

The proportion of people with a disease at any point (point prevalence) or period (period prevalence) in time

Quality-­adjusted life-­year (QALY)

A measure of the state of health of a person or group in which the benefits, in terms of length of life, are adjusted to reflect the quality of life. 1 QALY is equal to 1 year of life in perfect health. QALYs are calculated by estimating the years of life remaining for a patient following a particular treatment or intervention and weighing each year with a quality-­of-­ life score (on a scale of 0–1). It is often measured in terms of a person’s ability to carry out the activities of daily life and freedom from pain and mental disturbance

Sociodemographic index

A composite indicator of a country’s lag distributed income per capita, average years of schooling and the fertility rate in females under the age of 25 years

Years lived with disability

Can also be described as years lived in less than ideal health. This includes conditions such as influenza, which may last for only a few days, or epilepsy, which can be lifelong. It is measured by taking the prevalence of the condition, multiplied by the disability weight for that condition. Disability weights reflect the severity of the different conditions and are developed through surveys of the general public. 1 year lived with disability represents the equivalent of 1 full year of healthy life lost due to disability or ill-­health

Years of life lost (YLL)

Years lost due to premature mortality. YLLs are calculated by subtracting the age at death from the longest possible life expectancy for a person at that age. For example, if the longest life expectancy for men in a country is 75 years, but the man dies at 65 years of cancer, this would be 10 years of life lost due to cancer

4

Neurology: A Queen Square Textbook

Health loss is classified into three broad categories: • Communicable diseases (including HIV, tuberculosis, malaria), maternal, neonatal, nutritional diseases. • Non-­communicable diseases (NCD; heart disease, stroke, diabetes, cancer, depression). • Injuries (including self-­harm, animal bites and vehicle accidents). The overall health of the world’s population has improved; global life expectancy at birth increased from 67.2 years in 2000 to 73.5 years in 2019. The estimated number of deaths in children under the age of 5 years decreased from 9.6 million in 2000 to 5 million in 2019. Analysis of more granular data from the GBD study 2019, between 1990 and 2019, after correcting for the effects of population growth and ageing, shows a decline in the rate of disability-­ adjusted life-­years (DALYs) confirming improvements in overall health. There has also been a shift towards a greater proportion of years lived with disability from NCDs and injuries. In the past decade, the rate of decline of age-­standardised DALY rates was greatest in those younger than 50 years and this was most apparent in the 0-­to 9-­years age group. In those aged 50 years and older, the rate of change was slower from 2010 to 2019 compared with earlier time periods. The most important contributors to the declining burden include nine that predominantly affect children: lower respiratory tract infections, diarrhoeal diseases, neonatal disorders, measles, protein-­energy malnutrition, congenital birth defects, drowning, tetanus and malaria. For example, the age-­standardised DALY rates for measles showed a 90.4% decline (uncertainty interval 87.5– 92.8%) between 1990 and 2019. In addition to these conditions, there was also a decline in tuberculosis, which affects mainly adults. The 10 most important drivers of increasing burden (causing the largest increase in numbers of DALYs) were those affecting mainly older individuals: ischaemic heart disease, diabetes, stroke, chronic kidney disease, lung cancer and age-­ related hearing loss. The remaining four conditions were HIV/AIDS, other musculoskeletal disorders, low back pain and depression. HIV/AIDS peaked in 2004 and dropped significantly after global antiretroviral drug rollout programmes were initiated. The GBD 2019 study analysed the burden for all ages (Figure 1.4); the leading neurological or related disorders were stroke (3rd, 5.7 % of DALYs), road injuries (7th, 2.9%), low back pain (9th, 2.5%), HIV/AIDS (11th, 1.9%), tuberculosis (12th, 1.9%), depressive disorders (13th, 1.8%), headache disorders (15th, 1.8%), age-­related hearing loss (20th, 1.6%), falls (21st, 1.5%), self-­harm (22nd, 1.3%), anxiety disorders (24th, 1.1%), interpersonal violence (26th, 1.1%) and meningitis (40th, 0.6%). Significant increases in the number of DALYs between 1990 and 2019 were noted for all the above disorders except tuberculosis, self-­ harm, interpersonal violence and meningitis, which declined. The burden of disease for children under the age of 10 years declined significantly between 1990 and 2019 by 57.5%. This was primarily due to the decline in infectious diseases, including meningitis. Nevertheless, it remains important to realise that from the neurological perspective both short and long term, neonatal disorders were responsible for 1.88 million deaths, comprising 37.3% of deaths in children under the age of five years. Conditions included under this umbrella term include preterm birth complications (with contributing risk factors such as malnutrition, household air pollution from solid fuels and ambient environmental pollution). In 2019, neonatal disorders were still the highest cause of DALYs (32.4%) in this age group, increasing from 23% in 1990. Meningitis accounted for 2.1% of DALYs and, astonishingly, sexually transmitted infections, accounted for by congenital syphilis in 10th place with 1.4% of DALYs, an increase from 0.7% in 1990. Idiopathic

epilepsy was 23rd, accounting of for 0.5% of DALYs. Sub-­Saharan Africa experienced nearly half of the total DALYs in this age group and, once again, this is a stark reminder of the extent of existing global health inequities. In the adolescents aged 10–24 years, DALYs for NCDs increased by 13.1% with declines in both infectious diseases and injuries. Injuries were more common in adolescent males and accounted for 13.8% of DALYs: road injuries (ranked 1st), self-­harm (3rd) and interpersonal violence (5th). Headache disorders ranked second (5%) followed by depressive and anxiety disorders, which rank 4th and 6th, respectively, were more common in females and accounted for 7% of DALYs. HIV/AIDS ranked 9th (2.6% DALYs), having increased dramatically after 1990 (ranked 33rd, 0.9%) owing to the rapid increase in incidence in the first half of the study period, which was followed by a decline after antiretroviral therapy became more available, especially in sub-­Saharan Africa. This was made possible by the development of highly active antiretroviral drugs in the mid-­1990s and also by the US Presidents Plan for AIDS Relief for developing countries introduced by George W. Bush in 2003. Five causes in the top ten for the 10–24 age group also featured in the 25–49 age group top ten: road injuries (ranked 1st), HIV/AIDS (now ranked 2nd), low back pain (4th), headache disorders (5th) and depressive disorders (6th). Stroke was ranked ninth in this group, which may partly be related to HIV/AIDS as a significant predisposing risk factor. Within the 50–years and the 75 years and older groups (Figure 1.5), there was significant overlap, with ischaemic heart disease and stroke being the leading causes of disease burden. Low back pain (as a symptom but no specific aetiology), age-­related hearing loss, falls, blindness and visual loss also overlapped in the top 20 conditions. In the over 75s, Alzheimer’s disease now ranked fourth and Parkinson’s disease twentieth. Lung cancer and colorectal cancer were also a feature in both age groups, whereas headache disorders, injuries, depression and anxiety were now much lower down the list or not featured at all in the list of most common DALYs. Overall, the 2019 study emphasises both the transition towards NCDs and also the extent to which disability contributes to the overall burden of disease. Low-­and middle-­income countries (LMICs) with significant increases in life expectancy now face the double challenge of communicable diseases (albeit reducing) and a rapid rise in the incidence of NCDs. In the past, most of the focus of global health programmes has been directed towards reducing mortality. The recent change in population demographics and disease burden mean that disability management will demand greater health expenditure and research to identify new effective strategies. National Global Burden of Neurological diseases

United Kingdom

In 2019, the most frequent cause of death in the UK was ischaemic heart disease, with stroke the second most common and Alzheimer’s disease the sixth (Figure 1.6). When comparing the top 10 causes of death and disability to other countries similar in terms of socioeconomic status, the incidence of ischaemic heart disease, neurological and psychiatric conditions including low back pain (2nd), depression (3rd), headache disorders (4th), stroke (8th) and falls (10th) was similar. In the United States, drug use disorders were the highest cause of DALYs, while this was ranked ninth in the UK. In terms of risk factors underlying the most common causes of DALY (Figure  1.7), tobacco use remained the most important despite the reduction in consumption between 2009 and 2019. Other risk factors included high fasting plasma glucose, high body

Global Burden of Neurological Disease and the Neurology of Climate Change  5

All ages Leading causes 1990 1 Neonatal disorders 2 Lower respiratory infections 3 Diarrhoeal diseases 4 Ischaemic heart disease 5 Stroke 6 Congenital birth defects 7 Tuberculosis 8 Road injuries 9 Measles 10 Malaria 11 COPD 12 Protein-energy malnutrition 13 Low back pain 14 Self-harm 15 Cirrhosis 16 Meningitis 17 Drowning 18 Headache disorders 19 Depressive disorders 20 Diabetes 21 Lung cancer 22 Falls 23 Dietary iron deficiency 24 Interpersonal violence 25 Whooping cough 27 Age-related hearing loss 29 Chronic kidney disease 30 HIV/AIDS 32 Gynaecological diseases 34 Anxiety disorders 35 Other musculoskeletal

Percentage of DALYs 1990 10.6 (9.9 to 11.4) 8.7 (7.6 to 10.0) 7.3 (5.9 to 8.8) 4.7 (4.4 to 5.0) 4.2 (3.9 to 4.5) 3.2 (2.3 to 4.8) 3.1 (2.8 to 3.4) 2.7 (2.6 to 3.0) 2.7 (0.9 to 5.6) 2.5 (1.4 to 4.1) 2.3 (1.9 to 2.5) 2.0 (1.6 to 2.7) 1.7 (1.2 to 2.1) 1.4 (1.2 to 1.5) 1.3 (1.2 to 1.5) 1.3 (1.1 to 1.5) 1.3 (1.1 to 1.4) 1.1 (0.2 to 2.4) 1.1 (0.8 to 1.5) 1.1 (1.0 to 1.2) 1.0 (1.0 to 1.1) 1.0 (0.9 to 1.2) 1.0 (0.7 to 1.3) 0.9 (0.9 to 1.0) 0.9 (0.4 to 1.7) 0.8 (0.6 to 1.1) 0.8 (0.8 to 0.9) 0.8 (0.6 to 1.0) 0.8 (0.6 to 1.0) 0.7 (0.5 to 1.0) 0.7 (0.5 to 1.0)

Leading causes 2019 1 Neonatal disorders 2 Ischaemic heart disease 3 Stroke 4 Lower respiratory infections 5 Diarrhoeal diseases 6 COPD 7 Road injuries 8 Diabetes 9 Low back pain 10 Congenital birth defects 11 HIV/AIDS 12 Tuberculosis 13 Depressive disorders 14 Malaria 15 Headache disorders 16 Cirrhosis 17 Lung cancer 18 Chronic kidney disease 19 Other musculoskeletal 20 Age-related hearing loss 21 Falls 22 Self-harm 23 Gynaecological diseases 24 Anxiety disorders 25 Dietary iron deficiency 26 Interpersonal violence 40 Meningitis 41 Protein-energy malnutrition 46 Drowning 55 Whooping cough 71 Measles

Communicable, maternal, neonatal, and nutritional diseases Non-Communicable diseases Injuries

Figure 1.4  Global burden of disease for all ages, 2019; leading 25 level 3 causes of global disability-­adjusted life-­years (DALYs) and percentage of total DALYs (1990 and 2019), and percentage change in number of DALYs and age-­standardised DALY rates 1990–2019 for both sexes combined for all ages. Causes are connected by lines between periods; solid lines are increases in rank and dashed lines are decreases. Age-­related hearing loss = age-­related and other hearing loss. Alzheimer’s disease = Alzheimer’s disease and other dementias. Atrial fibrillation = atrial fibrillation and flutter. Cirrhosis = cirrhosis and other chronic liver diseases. COPD = chronic obstructive pulmonary disease. EMBID = endocrine, metabolic, blood, and immune disorders. DALY = disability-­adjusted life-­year. iNTS = invasive non-­typhoidal salmonella. Haemoglobinopathies = haemoglobinopathies and haemolytic anaemias. Lung cancer = tracheal, bronchus and lung cancer. Other musculoskeletal = other musculoskeletal disorders. Other unspecified infectious = other unspecified infectious diseases. Sudden infant death = sudden infant death syndrome. STI = sexually transmitted infections excluding HIV. Source: reproduced with permission from GBD 2019 Diseases and Injuries Collaborators (2020).

mass index (BMI) and blood pressure and alcohol use (Figure 1.8). Air pollution declined by 22.9% over the 2009–2019 study period but remains an important factor predisposing to neurological disease. In the UK, one in six individuals has a neurological condition, a total of 14,675,879  individuals according to data collated by the Neurological Alliance (2019; Table 1.2). In 2016/17, 12,736,365 bed-­ days were recorded for neurological patients. Between 2001 and 2014 there was a 39% increase in deaths due to neurological disorders.

Neurological Disease in Asia

The burden of diseases varies through the world and is dependent upon variations in age, genetic characteristics, climatic differences, cultural and socioeconomic differences. As an example, the GBD study analysed data for the periods 1990–2019 in Asia. The Asian region is divided into World Health Organization (WHO) Southeast

Asia region (11 countries) and WHO Western Pacific region (31 countries). In the Southeast Asia and the Western Pacific regions, stroke, tetanus and meningitis had the highest age-­standardised DALYs in 1990. By 2019, this hierarchy had changed to stroke, migraine and Alzheimer’s disease and other forms of dementias. Overall, the DALYs of stroke, Alzheimer’s disease and other dementias, Parkinson’s disease, brain cancers, multiple sclerosis, migraine and tension-­type headaches increased in both regions over the 30-­year period (Figure 1.9). Infectious diseases such as encephalitis, tetanus and meningitis decreased mirroring the global disease burden data. The age-­ standardised incidence of stroke and mortality rate decreased, primarily because of the better treatment of hypertension and diabetes between 1990 and 2019. There have been significant consequences of the increased awareness of risk factors for

6

Neurology: A Queen Square Textbook

75 years and older Percentage of DALYs 1990

Leading causes 1990

Leading causes 2019

Percentage of DALYs 2019

Percentage change in numbers of DALYs, 1990–2019

Percentage change in age-standardised DALY rate, 1990–2019

1 Ischaemic heart disease 2 Stroke 3 COPD 4 Alzheimer’s disease 5 Lower respiratory infections 6 Diarrhoeal diseases 7 Diabetes 8 Hypertensive heart disease 9 Age-related hearing loss 10 Lung cancer 11 Falls 12 Tuberculosis 13 Low back pain 14 Chronic kidney disease 15 Stomach cancer

18.6 (17.1 to 19.7) 15.5 (14.3 to 16.7) 9.9 (8.6 to 10.7) 3.8 (1.7 to 8.6) 3.3 (3.0 to 3.6) 3.1 (2.0 to 4.5) 2.6 (2.4 to 2.9) 2.3 (1.9 to 2.5) 2.0 (1.5 to 2.7) 1.9 (1.8 to 2.0) 1.8 (1.6 to 2.1) 1.8 (1.6 to 2.1) 1.7 (1.2 to 2.3) 1.6 (1.5 to 1.8) 1.6 (1.4 to 1.7)

1 Ischaemic heart disease 2 Stroke 3 COPD 4 Alzheimer’s disease 5 Diabetes 6 Lower respiratory infections 7 Lung cancer 8 Falls 9 Chronic kidney disease 10 Age-related hearing loss 11 Hypertensive heart disease 12 Diarrhoeal diseases 13 Low back pain 14 Colorectal cancer 15 Blindness and vision loss

16.2 (14.6 to 17.6) 13.0 (11.7 to 14.0) 8.5 (7.5 to 9.2) 5.6 (2.6 to 12.2) 4.0 (3.6 to 4.3) 3.3 (2.9 to 3.6) 2.6 (2.3 to 2.8) 2.6 (2.2 to 2.9) 2.5 (2.3 to 2.7) 2.5 (1.9 to 3.3) 2.4 (1.8 to 2.7) 1.9 (1.2 to 3.0) 1.8 (1.3 to 2.4) 1.7 (1.5 to 1.8) 1.7 (1.3 to 2.2)

66.6 (57.7 to 74.2) 60.5 (48.7 to 72.5) 63.6 (49.1 to 86.1) 180.0 (168.0 to 194.7) 190.7 (179.4 to 201.0) 87.4 (76.2 to 99.6) 164.3 (143.6 to 183.8) 166.4 (151.1 to 183.4) 196.0 (173.9 to 211.1) 137.8 (132.0 to 143.9) 106.0 (68.5 to 131.7) 15.1 (–16.8 to 65.3) 105.7 (100.2 to 111.4) 126.9 (113.4 to 138.3) 124.7 (119.3 to 130.7)

–32.4 (–35.8 to –29.4) –33.4 (–38.3 to –28.5) –31.0 (–37.1 to –21.9) 2.6 (–2.1 to 6.6) 23.1 (18.6 to 27.5) –25.3 (–29.3 to –20.4) 16.4 (7.4 to 24.9) 6.4 (0.4 to 13.3) 21.6 (12.6 to 27.4) –2.2 (–4.3 to –0.2) –15.1 (–31.5 to –5.0) –51.0 (–64.9 to –30.4) –12.5 (–13.8 to –11.3) –4.5 (–9.7 to 0.1) –7.4 (–9.9 to –4.8)

16 Blindness and vision loss 17 Colorectal cancer 18 Asthma 19 Cirrhosis 20 Prostate cancer 21 Atrial fibrillation 22 Osteoarthritis 23 Oral disorders

1.4 (1.1 to 1.8) 1.4 (1.3 to 1.5) 1.2 (1.0 to 1.7) 1.2 (1.0 to 1.3) 1.0 (0.8 to 1.2) 1.0 (0.8 to 1.2) 0.9 (0.5 to 1.7) 0.8 (0.6 to 1.2)

16 Atrial fibrillation 17 Stomach cancer 18 Prostate cancer 19 Cirrhosis 20 Parkinson’s disease 21 Osteoarthritis 22 Oral disorders 23 Tuberculosis

1.3 (1.1 to 1.5) 1.3 (1.1 to 1.4) 1.1 (1.0 to 1.4) 1.1 (1.0 to 1.2) 1.1 (1.0 to 1.2) 1.1 (06 to 21) 0.9 (0.6 to 1.3) 0.9 (0.8 to 1.0)

148.6 (134.8 to 161.9) 55.0 (43.8 to 66.6) 117.0 (102.1 to 142.3) 82.3 (62.1 to 100.9) 153.7 (138.7 to 166.6) 139.5 (136.5 to 142.6) 112.0 (106.4 to 117.6) –6.3 (–16.9 to 14.6)

–1.8 (–6.9 to 2.5) –32.9 (–37.5 to –28.0) –8.5 (–14.6 to 2.1) –21.3 (–30.2 to –13.5) 6.0 (0.0 to 11.1) 0.8 (–0.4 to 2.1) –10.9 (–12.9 to –8.8) –59.2 (–64.0 to –50.3)

24 Parkinson’s disease 25 Upper digestive diseases

0.8 (0.8 to 0.9) 0.8 (0.7 to 0.9)

24 Asthma 25 Road injuries

0.8 (0.7 to 1.0) 0.8 (0.7 to 0.9)

25.2 (3.2 to 41.2) 110.0 (99.8 to 118.1)

–46.2 (–55.9 to –39.8) –9.3 (–13.5 to –5.9)

26 Road injuries

0.7 (0.6 to 0.8)

32 Upper digestive diseases

0.6 (0.5 to 0.6)

34.0 (22.8 to 46.2)

–43.8 (–48.4 to –38.7)

Communicable, maternal, neonatal, and nutritional diseases Non-Communicable diseases Injuries

Figure 1.5  Leading 25 level 3 causes of global DALYs and percentage of total DALYs (1990 and 2019) and percentage change in number of DALYs and age-­standardised DALY rates from 1990 to 2019 for both sexes combined for all ages 75 years and older. Causes are connected by lines between periods; solid lines are increases in rank and dashed lines are decreases. Age-­related hearing loss = age-­related and other hearing loss. Alzheimer’s disease = Alzheimer’s disease and other dementias. Atrial fibrillation=atrial fibrillation and flutter. Cirrhosis = cirrhosis and other chronic liver diseases. COPD = chronic obstructive pulmonary disease. EMBID = endocrine, metabolic, blood, and immune disorders. DALY = disability-­adjusted life-­year. iNTS = invasive non-­typhoidal salmonella. Haemoglobinopathies = haemoglobinopathies and haemolytic anaemias. Lung cancer = tracheal, bronchus and lung cancer. Other musculoskeletal = other musculoskeletal disorders. Other unspecified infectious = other unspecified infectious diseases. Sudden infant death = sudden infant death syndrome. STI = sexually transmitted infections excluding HIV. Source: reproduced with permission from GBD 2019 Diseases and Injuries Collaborators (2020).

Communicable, maternal, neonatal, and nutritional diseases Non-communicable diseases Injuries 2009

2019

% change, 2009–2019 –0.1%

Ischemic heart disease

1

1

Ischemic heart disease

Stroke

2

2

Stroke

1.6%

Lung cancer

3

3

COPD

17.5%

COPD

4

4

Lung cancer

Lower respiratory infect

5

5

Lower respiratory infect

20.8%

Alzheimer’s disease

6

6

Alzheimer’s disease

16.8%

Colorectal cancer

7

7

Colorectal cancer

17.1%

Breast cancer

8

8

Prostate cancer

26.0%

Prostate cancer

9

9

Breast cancer

Pancreatic cancer

10

10

Pancreatic cancer

9.8%

6.8% 24.3%

Figure 1.6  United Kingdom top 10 causes of death in 2019 and percentage change 2009–19, all ages combined. COPD = chronic obstructive pulmonary disease. Source: data from Institute for Health Metrics and Evaluation.

Global Burden of Neurological Disease and the Neurology of Climate Change  7

Communicable, maternal, neonatal, and nutritional diseases Non-communicable diseases Injuries –30%

–25%

–20%

–15%

–10%

–5%

0%

5%

10%

15%

20%

30%

25%

1 Ischaemic heart disease 2 Low back pain COPD 3 Lung cancer 4 5 stroke Diabetes 6 Depressive disorders 7 Falls 8 Headache disorders 9 Lower respiratory infections 10

Figure 1.7  United Kingdom top 10 causes of death and disability (DALYs) in 2019 and percentage change 2009-­2019, all ages combined. COPD = chronic obstructive pulmonary disease. Source: data from UK Institute for Health Metrics and Evaluation (https://www.healthdata.org/ research-­analysis/health-­by-­location/profiles/united-­kingdom).

Metabolic risks Environmental/occupational risks Behavioural risks 2009

2019

% change, 2009–2019

Tobacco

1

1

Tobacco

–2.9%

High body-mass index

2

2

High fasting plasma glucose

23.5%

High blood pressure

3

3

High body-mass index

13.6%

Dietary risks

4

4

Dietary risks

High fasting plasma glucose

5

5

High blood pressure

Alcohol use

6

6

Alcohol use

High LDL

7

7

High LDL

4.2% –1.1% 3.7% –5.3%

Occupational risks

8

8

Occupational risks

5.6%

Non-optimal temperature

9

9

Non-optimal temperature

4.8%

Air pollution

10

10

Drug use

2.4%

Drug use

11

12

Air pollution

–22.9%

Figure 1.8  United Kingdom top 10 risks contributing to total number of disability-­adjusted life-­years in 2019 and percentage change 2009–19, all ages combined. LDL = low-­density lipoprotein. Source: data from UK Institute for Health Metrics and Evaluation (https://www. healthdata.org/research-­analysis/health-­by-­location/profiles/united-­kingdom).

stroke and the widening implementation of public health policies of screening and prevention. Other risk factors (not documented in the GBD) have been addressed, including treatment of atrial fibrillation with non-­vitamin K antagonists and access to thrombolytic and endovascular treatments, each of which has a role to play in reducing mortality and disability. However, when age-­standardised DALYs were analysed with respect to Sociodemographic Index for each country, higher-­ income countries showed lower DALYs, implying that access to medical care and preventative measures depend on economic status. Thus, it is imperative that strategies are developed that are suited to the reality of each countries’ healthcare needs within the context of the relevant socioeconomic factors.

Specific Neurological Disorders

Stroke

Stroke is a preventable, treatable and manageable disease. In 2019, stroke was the second leading cause of death after ischaemic heart disease and the third leading cause of death and disability combined. There were 12.2 million incident cases of all types of strokes, 101 million prevalent cases, 143 million DALYs due to stroke and 6.55 million deaths. Ischaemic stroke constituted 62.4%. of all incident strokes, intracerebral haemorrhage, 27.9% and subarachnoid haemorrhage 9.7% (GBD  2019 Stroke Collaborators 2021; Figure 1.10). Between 1990 and 2019, the absolute numbers of incident stroke increased by 70%, prevalent strokes increased by 85% and deaths

8

Neurology: A Queen Square Textbook

Table 1.2  United Kingdom incidence and prevalence of neurological disorders (Source: adapted from Neurological Alliance, 2019). Neurological disorder

Prevalence

Ataxia

Incidence

8,600

Brain tumours

85,927

Charcot–Marie–Tooth

19,376

Chronic inflammatory demyelinating neuropathy Dementia

7810

3329

600

759,000

180,000

Encephalitis

5000

Epilepsy

526,000

Essential tremor

842,424

Functional disorders Migraine

32,000

Not known

Not known

7,945,633

Motor neurone disease

3962

1132

Multiple sclerosis

90,000

4120

Parkinson’ disease

121,927

18,461a

Stroke

1,000,000

80,000

Traumatic brain injury

1,095,120

900,000b

 > 45 years age.  Accident and emergency attendances.

a

b

increased by 43%. This increase was especially evident in people younger than 70 years, whereas age-­standardised rates showed a decline in those over the age of 70 years. Although there has been a  trend for a decline in age-­standardised incidence, DALYs and mortality from stroke, there is substantial variation globally, with the lowest rates in high-­income regions. These incremental figures result from a combination of population growth and ageing populations but also increases in exposure to the five leading risk factors for stroke: hypertension, high BMI, high fasting plasma glucose, ambient particulate matter pollution and smoking. Some 89% of stroke DALYs are attributable to 19  modifiable risk factors, most of which are socioeconomically patterned and in part driven by commercial influences (smoking, alcohol, diet and air pollution). The fastest increasing risk factor for stroke between 1990 and 2019  was high BMI (Figure  1.11). The proportion of the world’s population with obesity has tripled since 1975, according to the WHO. It is projected that, by 2025, 1 billion people will have a BMI of more than 30  kg/m2 (normal BMI 18.5–25  kg/m2). In high-­ income Western countries, almost one third of the population can be classified as obese, and 50% if overweight individuals are included (BMI 25–30  kg/m2). In England, 29% of the population were obese in 2019. In the United States, half of the population will be classed as obese by 2030. In 2023, the number of obese youngsters aged five to nine years passed the number who were underweight for the first time in history. These alarming statistics highlight the time-­bomb underlying the global burden of cardio-­ and cerebrovascular diseases. The age-­standardised burden of stroke mortality was 3.6 times higher and the DALY rate was 3.7 times higher in the World Bank

1% 0% 1% 2% 2%

13%

WHO South-East Asia Region

WHO Western Pacific Region

1% 4% 1% 5% 2%

3% 15% 3% 0% 2%

55% 5%

11%

2% 0% 2%

62%

8%

Stroke Motor neuron disease Idiopathic epilepsy Meningitis Other neurological disorders

Alzheimer’s disease and other dementias Multiple sclerosis Migraine Encephalitis

Parkinson’s disease Brain and central nervous system cancer Tension-type headache Tetanus

Figure 1.9  South East Asia and Western Pacific Region 2019 Contribution of neurological disorders to disability-­adjusted life-­years in the World Health Organization South-­East Asia and Western Pacific regions.

All strokes

Incidence rates per 100 000 people

21 units per week Obesity

5% Smoking 4% Depression 4%

2%

Social isolation

Later life

Physical inactivity 2%

Air pollution 1%

Diabetes Potentially modifiable 40%

Risk unknown 60%

Figure 1.13  Preventable risk factors for dementia; population-­attributable fraction of potentially modifiable risk factors for dementia. Source: reproduced with permission from the Lancet Commission (Livingston et al. (2020).

past 30 years has increased by 61% (from 5.5 million in 1990 to 8.8  million in 2019) with the five largest contributors to global neurological DALYs being stroke (42.2%), migraine (16.3%), dementia (10.4%), meningitis (7.9%) and epilepsy (5%). Some 80% of deaths and DALYs due to neurological disorders are in

LMICs. The economic burden is huge; in the United States and Europe alone, the annual cost of neurological disorders is in the region of US$1 trillion. This burden of neurological disease will greatly increase with continuing population growth and increased life expectancy.

Global Burden of Neurological Disease and the Neurology of Climate Change  13

In the past 30 years, great strides have been made in neurogenetics, elucidating the pathophysiological mechanisms and treatments of individual neurological diseases. These include stroke, multiple sclerosis, Parkinson’s disease, epilepsy and neuromuscular disorders. Although useful at an individual level, there is a disconnect between the direction of current neuroscientific research, which is focused on diagnosis and treatment, and the global perspective, which should focus on epidemiology, public health and prevention. The latter could provide much greater improvement in ‘brain health’ at a fraction of the cost.

Neurology and Global Warming

As clinicians providing informed healthcare for people with neurological conditions, we have always responded to manmade and natural environmental events and changes. When neurological challenges, such as HIV, arthropod-­borne neurotropic viruses (e.g. Zika virus), spongiform encephalopathies and the SARS-­2-­CoV pandemic emerged, as a community we sought to map new or spreading diseases, understand potential neurological implications and develop novel treatments and management strategies where needed. The outpouring of neurological publications during the current coronavirus pandemic is the latest example of the neurological community’s response to a changing world. The changing climate, with global warming, is perhaps the biggest challenge now facing humanity. In a series of deeply researched papers, the Lancet Countdown on health and climate change has tracked the health profile of climate change: the 2022 report makes clear the grave risks to human health posed by climate change and the parlous state of readiness of global healthcare systems, compromised by multiple coincident global events (Romanello et al. 2022). In 2021, a joint editorial in 233 leading medical journals called on governments to take emergency action to address the ‘catastrophic harm to health’ from climate change. People with neurological diseases typically have impaired resilience to change, which may arise from acquired or developmental cognitive difficulties, disordered autonomic function, behavioural abnormalities, motor and sensory dysfunction, or combinations of these and other factors. Some neurological conditions may also increase susceptibility to changes due to global warming. As clinicians practising evidence-­based medicine, it is incumbent upon us to consider the implications of global warming for people with neurological diseases and to determine what actions are necessary. Even in the unlikely scenario that greenhouse gas emissions are stopped tomorrow, the emissions already in the atmosphere have already baked in further global temperature rises. We therefore need to act and plan now. Global Warming Impacts on Neurological Diseases Global warming is changing the planetary climate and ecosystems worldwide. Environmental change has always been a fundamental driver of human societal change, sometimes leading to the collapse of civilisations. There has been comparatively little research in neurology and global warming but, from existing knowledge of neurological diseases, an understanding of the limits of physiological compensation, modelling of climate change and its projected consequences and extrapolation from major events, it becomes possible to grasp at least the broad scope of likely outcomes, from the molecular to the epidemiological. The argument is sometimes put that humans can thermoregulate. But human thermoregulatory capacity has limits. Thermoregulation has interacting internal and external components. As ambient

temperature rises, rising skin and core temperatures activate cooling responses through behavioural (e.g. seeking shade) and physiological (e.g. sweating, cutaneous vasodilation) measures, the latter mediated by the autonomic nervous ­system. If air temperature exceeds skin temperature (typically 33–35°C), cooling can only occur through sweat evaporation, itself compromised as ambient humidity rises or when physical activity increases heat generation. Sweat production is limited to 3–4  litres/hour, while maximal gut water absorption is less than 1.5 litres/hour. If heat generation exceeds heat loss, rising body temperature leads to decompensation, including a focus on the nervous system, with heat cramps, heat exhaustion, heat stroke and ultimately death. Observations, both anecdotal and more systematic, already indicate that for some neurological conditions, changes in the ambient temperature affect the disease process. In multiple sclerosis, symptom aggravation by heat is well recognised. In some types of epilepsy, seizures can be precipitated by elevated external temperatures, as they can by fever. Some people with myasthenia gravis report worsening of symptoms with heat. Heat can provoke acute confusion or delirium in those with lower cognitive reserve, and increase rates of hospital admission in those with dementia. Neurological conditions aggravated by disrupted sleep can also be expected to deteriorate with heatwaves when, typically, night-­time temperature elevations occur and can prove more difficult to manage than daytime peaks; elevated night-­time humidity may prove even more challenging. Moreover, global warming has consequences beyond rising temperatures alone. For example, the range of arthropods carrying neurotropic viruses is predicted to expand, while human adaptations to rising temperatures, such as large-­scale national or planetary engineering schemes, will have complex consequences for healthcare, as will urban heat island effects that amplify more widespread or even generate local temperature peaks. Climate change-­related events, such as floods and extreme temperatures, will repeatedly and cumulatively compromise supply chains and infrastructure, with interruption to medication availability and emergency or planned healthcare provision. The scope of consequences for people with neurological diseases can be seen to be wide even from this limited description. The coincidence of multiple challenges, such as heatwaves, compromise of pre-­existing healthcare and armed conflict and migration is likely to be increasingly common, compounding difficulties for those with neurological, and other, conditions. Neurological Diseases may Increase Susceptibility to Consequences of Climate Change Neurological diseases are often multifaceted, with features that can magnify health risks from a changing climate. Developmental or acquired cognitive impairments may compromise the ability to detect, understand and respond to acute and chronic elevated ambient temperature and other warming-­related environmental alterations. Additional involvement of the autonomic nervous system will undermine normal compensatory mechanisms. Physical limitations from motor and sensory disorders may mean that behavioural responses, both acute and in the long term, are blunted or impossible. The very mechanism of disease may put some individuals at heightened risk; for example, for some neurogenetic conditions, the dysfunction of mutant encoded proteins will intrinsically raise susceptibility to increased ambient temperature, as is seen in some channelopathies. Those with neurological disorders causing multisystem disruption are likely to be particularly vulnerable. As an example, individuals with the

14

Neurology: A Queen Square Textbook

rare developmental and epileptic encephalopathy Dravet syndrome, caused by loss-­of-­function variants in the gene SCN1A encoding a temperature-­sensitive sodium channel Nav1.1, typically have significant cognitive impairment, autonomic dysfunction, impaired mobility and fever-­and temperature-­sensitive drug-­resistant seizures. During a heatwave, it is unlikely that those affected would be able to communicate or take evasive action to combat their thermal discomfort. They could have intrinsic aggravation of the underlying disease biology due to mutation-­ related exaggerated channel thermosensitivity and compromised endogenous thermoregulatory responses, aggravated by antiseizure treatments, such as topiramate, that can further reduce sweating capacity; all these risks may be inadvertently worsened by actions of carers who may not be informed about heatwave-­related risks. Similar scenarios can be imagined across the range of neurological diseases, rendering some of those with such conditions additionally at risk from both the acute (e.g. heatwaves, floods) and chronic results of global warming. With experience in disease management, it will fall to neurological experts to interpret for, advise and counsel patients with some conditions about the dangers of global warming. Contributions to Climate Change from Neurology Scientific research forms the basis for much-­needed advances in the treatment of illnesses. Laboratories essential to such work are, however, prolific users of resources and energy; they are responsible for about 2% of global plastic waste and use up to tenfold more energy per unit area than office space, and around 18 times the energy usage compatible with a net-­zero goal. This intensity is generated by many factors; for example, from the mandated number of air changes per hour in a vivarium to the use of chemicals with long atmospheric half-­lives and potent greenhouse capacity. Neuroscience laboratories are not unique in these aspects but they are a sizeable part of the problem, with National Institutes of Health funding in the United States in 2022 amounting to around $11 billion for neuroscience, among the top funded areas. Neuroscience is typically collaborative, and scientists need to discuss and publish their findings. The Society for Neuroscience estimated that travel alone for attendance to its 2014 meeting in Washington produced a carbon footprint roughly equivalent to the annual carbon footprint of 1000  medium-­sized laboratories. As with many climate change related inequalities, most conferences have a disproportionately small representation from LMICs, which are among those countries most likely to experience, or are already experiencing, the largest and earliest effects of climate change. Many neurological conditions are chronic, requiring long-­ term care and account for a significant proportion of the burden of illness globally. The carbon footprint of chronic care is poorly quantified, but if cost to the healthcare system is used as a crude surrogate, then, for example, chronic, drug-­resistant epilepsy is likely to have a larger carbon footprint that easily-­treated epilepsy with seizure freedom. The UK National Health Service is Europe’s biggest institutional greenhouse gas emitter. Together these initial observations suggest that the overall provision of neurological healthcare is likely to be an important contributor to national greenhouse gas emission totals in developed countries. Many component aspects could be rationalised from a sustainability perspective: there is probably something each of us can do. Actions The UN Conference of the Parties (COP) is the name given to the United Nations climate change conferences. The goal of these

conferences is to review progress made by members of the UN Framework Convention on Climate Change, an agreement to ‘stabilise greenhouse gas concentrations at a level that would prevent dangerous anthropogenic (human-­induced) interference with the climate system’. These are new terms and new horizons for neurology. As part of our duty of care to people with neurological conditions, engagement with climate change is essential, because global warming changes the fundamental parameters of our work (and our very existence). At COP27, in 2022, a historic ‘loss and damage’ agreement seems to have been reached to redress the imbalance between nations that are the largest greenhouse gas emitters and those that will suffer the most despite being the smallest emitters. Nevertheless, the 1.5°C limit of the Paris agreement remains in jeopardy. Even if greenhouse gas emissions were to stop now, global warming is still embedded in our future because of the existing excess of long-­lasting greenhouse gas emissions already in the atmosphere and there will, therefore, be health consequences. What, as healthcare professionals, can and must we do? Engagement can be classed into three categories: raising awareness and education; research; and actions. The reaction of the neurological community to the SARS-­2-­CoV provides a contemporary example to which we can relate directly. As a community, we informed ourselves of the new challenge and threat to the health of our patients. We raised our own awareness and educated ourselves to know both how COVID-­19 might affect people with neurological conditions and how we could continue safely to provide neurological services; for example, developing guidance for ophthalmoscopy and adopting telemedicine. We generated recommendations for vaccination in the context of neurological diseases, as experts in the conditions involved. Research capacity was often repurposed; for example, with rapid reorientation to provision of high-­throughput infection screening, which proved essential in the care of those with severe phenotypes living in residential care facilities. Research into neurological consequences of infection and vaccination was also undertaken at impressive speed, benefitting from pre-­existing networks and rapid shifts of focus. We took actions, ranging from advocating for those with neurological diseases to retraining ourselves to provide practical input, such as becoming vaccinators or simply providing manual support in near-­overwhelmed hospitals. The threat was immediate and tangible and the response global and huge, with changes of behaviour at every level. For many across the world, the threat from climate change is also immediate and existential: island nations face extinction, requiring relocation of entire populations; for others, the impacts of climate change have yet to be seriously experienced and the threat therefore apparently less urgent. But the threat is now well documented, with perhaps more evidence available than for any other aspect of human health in history, and is growing everywhere. We have a period of notice within which to take the actions needed. There are pressing needs: for increasing our own knowledge of climate change, of raising awareness of its likely consequences, of informing patients and carers of the risks; for basic research at every level from the molecular to the epidemiological; and for action in our own work, for example considering whether air travel to every meeting is justifiable, whether we can safely move to adopt more telemedicine, and whether laboratory practices are truly sustainable. Reducing greenhouse gas emissions at a massive scale is the essential change needed to avoid irreparable and unmanageable outcomes; such change is achievable only by governmental and intergovernmental commitments. But as a highly trusted constituency, healthcare professionals can have an important influence on national thinking and governmental policy in the face of unprecedented danger. We all have a role to play.

Global Burden of Neurological Disease and the Neurology of Climate Change  15

References

GBD 2019 Diseases and Injuries Collaborators. (2021). Global, regional and national burden of stroke and its risk factors, 1990–2019: a systemic analysis for the Global Burden of Disease Study 2019. Lancet Neurol 20: 795–820. GBD 2019 Diseases and Injuries Collaborators. (2020). Global burden of 369 diseases and injuries in 204 countries and territories, 1990–2019: a systemic analysis for the Global Burden of Disease study 2019. Lancet 396: 1204–1222. Institute for Health Metrics and Evaluation. (2019). Headache disorders – level 3 cause. Lancet 393: 82–83. https://www.healthdata.org/results/gbd_summaries/2019/ headache-­disorders-­level-­3-­cause (accessed 20 July 2023). Livingston G. Huntley J, Sommerlad A et al. (2020). Dementia prevention, intervention and care:2020 report of the Lancet Commission. Lancet 396: 413–446. Neurological Alliance (2019). Neuro Numbers 2019. Watford; Neurological Alliance. https://www.neural.org.uk/publication/mental-­h ealth-­r ightcare-­p athways-­ proposal-­2 (accessed 20 July 2023). Romanello M, Di Napoli C, Drummond P et al. (2022). The 2022 report of the Lancet Countdown on health and climate change: health at the mercy of fossil fuels. Lancet 400: 1619–1654. Vollset SE, Goren E, Yuan C-­W et al. (2020) Fertility, mortality, migration, and population scenarios for 195 countries and territories from 2017 to 2100: a forecasting analysis for the Global Burden of Disease Study. Lancet 396: 1285–1306.

Further Reading

Global Burden of Neurological Disease

Kang S, Eum S,Chang Y et  al. (2022). Burden of neurological diseases in Asia from 1990 to 2019; a systemic analysis using the Global Burden of Disease Study data. BMJ Open 12: e059548. Steiner TJ. (2020). Migraine remains second among the world’s causes of disability, first among young women: findings from GDB2019. J Headache Pain 21: 137. Walsh S, Merrick R, Brayne C. (2022). The relevance of social and commercial determinants for neurological health. Lancet Neurol 21: 1151–1160.

Neurology of Climate Change

Bjurulf B, Reilly C, Hallböök T. (2022). Caregiver reported seizure precipitants and measures to prevent seizures in children with Dravet syndrome. Seizure 103: 3–10.

Caminade C, McIntyre KM, Jones AE. (2019). Impact of recent and future climate change on vector-­borne diseases. Ann N Y Acad Sci 1436: 157–173. Christogianni A, Bibb R, Davis SL et  al. (2018). Temperature sensitivity in multiple sclerosis: An overview of its impact on sensory and cognitive symptoms. Temperature (Austin) 5: 208–223. Clemence M. (2021). Doctors become the world’s most trusted profession. Ipsos 12  October. https://www.ipsos.com/en/global-­trustworthiness-­index-­2021 (accessed 4 August 2023). Diamond J. (2005). Collapse: How societies choose to fail or succeed. New  York, NY: Penguin; 2005. Goldie J, Sherwood SC, Green D, Alexander L. (2015). Temperature and humidity effects on hospital morbidity in Darwin, Australia. Ann Glob Health 81: 333–341. Gulcebi MI, Bartolini E, Lee O et al. (2021). Climate change and epilepsy: Insights from clinical and basic science studies. Epilepsy Behav 116: 107791. Kenny G, Flouris A. (2014). The human thermoregulatory system and its response to thermal stress. In: Wang F, Gao C, eds. Protective Clothing: Managing thermal stress. Woodhead Publishing Series in Textiles 154. Amsterdam: Woodhead Publishing. pp. 319–365. Maslin M, Parikh P, Taylor R, Chin-­Yee S. (2022). COP27  will be remembered as a failure  – here’s what went wrong. The Conversation 21  November. https://the conversation.com/cop27-­will-­be-­remembered-­as-­a-­failure-­heres-­what-­went-­wrong-­ 194982 (accessed 4 August 2023). Nandi R. (2020). Curbing environmental footprint in laboratory environments. Lab Man 15(3). https://www.labmanager.com/lab-­design-­and-­furnishings/curbing-­ environmental-­footprint-­in-­laboratory-­environments-­22146 (accessed 4 August 2023). Nathans J, Sterling P. (2016). How scientists can reduce their carbon footprint. Elife 5: e15928. Paterson RW, Brown RL, Benjamin L et  al. (2020). The emerging spectrum of COVID-­19  neurology: clinical, radiological and laboratory findings. Brain 143: 3104–3120. Sahai N, Bard AM, Devinsky O, Kalume F. (2021). Disordered autonomic function during exposure to moderate heat or exercise in a mouse model of Dravet syndrome. Neurobiol Dis 147: 105154. Wei Y, Wang Y, Lin C-­K et al. (2019). Associations between seasonal temperature and dementia-­associated hospitalizations in New England. Environ Int 126: 228–233.

CHAPTER 2

Approach to the Patient with Neurological Disease Robin Howard1, Gerry Christofi1, Alex Rossor1, Jason Warren2 and David Werring2  National Hospital for Neurology and Neurosurgery, Queen Square, London, UK  UCL Queen Square Institute of Neurology, Queen Square, London, UK

1 2

Invest suffering with the reverence that is due its sad mystery. To relieve suffering, to cure disease and above all to preserve health -­ these are the objectives of your life. (Sir William Gowers, 1884)

What Makes a Neurologist?

A good neurologist is not merely a skilled diagnostician. Diagnosis is futile if it does not in some way help the patient and those close to them to understand and deal with their neurological disease. Those who are experienced in treating patients know that true cures are few and far between, but cure is not the only or, indeed, even the most desired goal of neurological management on some occasions. Despite dramatic advances in supportive technology, the fundamentals of the neurological method have changed surprisingly little over the past century. Because neurological diseases so often threaten the core attributes that make our lives worth living, the patient  – as a rounded human being composed of foibles and inconsistencies  – really does remain absolutely central to the ­practice of neurology. The neurological consultation involves a thorough but also hypothesis-­led approach to eliciting the clinical history and examining the patient, in an attempt to define the ­evolution of neurological symptoms and signs and the pattern of impairment and disability with a view to identifying the cause. Arguably even more than other branches of medicine, neurology remains a quiet art: accurate neurological diagnosis is far more likely to turn on noticing a small detail of the history or examination than on a spectacular intellectual tour de force. The practice of neurology rests on close acquaintance and facility with the scientific principles of nervous system structure, function and pathology. Yet few would argue that among the most important skills a neurologist can acquire is the ability to communicate accurately, kindly and transparently. In formulating a neurological differential diagnosis, it is essential to balance probabilities and priorities, often favouring conditions that are more amenable to management while seeking to minimise harm, distress or indignity. No neurological condition is untreatable, even though many cannot be cured. To be effective, however, treatment must be addressed to the patient as an individual; the neurologist must therefore be an attentive and compassionate student of human behaviour in all its subtlety and diversity.

More than ever before, the delivery of neurological care is a team enterprise. The neurologist is frequently called upon to work closely alongside practitioners with very different backgrounds and skill sets, or indeed to lead a multidisciplinary team. Most doctors are not well trained in management; however, mutual respect and willingness to learn from colleagues are a good starting point from which to build the relevant competency. There is no doubt that neurological disease often bears a freight of stigma and mystique. One of the most useful roles a neurologist can fulfil is that of teacher and debunker of myths – in the consulting room and on the ward, certainly, but also for students of all varieties, for colleagues in other specialties and if called upon, for the public at large.

Approaching the Neurological Patient

Is it important to commence a neurological consultation gently. It is often helpful first to ask the patient to talk about themselves. Patients are not to be approached like crossword puzzles; the neurological consultation is often complex and multifaceted, but never simply about arriving at a diagnosis. It entails: • Establishing a rapport with the patient and gaining their cooperation and confidence. • Digesting and distilling background information from the referrer and other medical consultations. • Obtaining a careful, searching and sensitive history from the patient and/or accompanying informants. • Undertaking a targeted examination. • Formulating a clear and sound differential diagnosis. • Establishing an appropriate programme of investigation. • Selecting, planning and coordinating appropriate management. • Communicating effectively and empathically with the patient, the referrer and relevant others about the illness, its likely short and long-­term consequences and the options for management. • Conveying uncertainties openly and constructively. • Helping the patient and their supporters understand and cope with their illness and/or disability. On first meeting the neurologist, the patient brings a unique set of beliefs and fears about their illness and expectations of the consultation, which will be shaped by their personal circumstances, the views of family and friends, potential misinformation gleaned from the internet and other sources and their past experience of illness

Neurology: A Queen Square Textbook, Third Edition. Edited by Robin Howard, Dimitri Kullmann, David Werring and Michael Zandi. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd. 17

18

Neurology: A Queen Square Textbook

and previous practitioners. The neurologist must try to assess these expectations and manage them accordingly. This is often best done by conveying a genuine interest in the person and their problem – no small undertaking in a busy clinic fraught with computerised medical record systems and other distractions. Particular care is needed in negotiating linguistic and cultural barriers to communication. The consultation should not be hurried and the neurologist must ensure adequate ‘thinking time’. This is increasingly precious as demands on clinical services grow dauntingly, but the neurological examination (rehearsed so well that it flows without deliberation, like playing an instrument) becomes an opportunity to consider the differential diagnosis and plan the next steps. The ‘style’ of the neurological consultation should be flexible and take its cues – verbal and especially, nonverbal – from the patient. Caplan and Hollander (2011) contrast ‘low control’ (passive listening without attempting to influence the patient) with ‘high control’ (directing the course of discussion) approaches; in practice, it is often reasonable to start with a more open-­ended style to build rapport and then switch to a more focused enquiry, as necessary. By ‘focused’ here we do not wish to imply, ‘leading’. Questions that steer or close off the patient’s response are always to be avoided (thus the patient referred with a possible seizure disorder might be asked if they have had any unusual experiences in connection with their episodes, but asking them if they smell burning rubber is likely to be misinformative). If the patient is attending accompanied, a trusted witness may (with the patient’s consent) provide extremely helpful testimony about the illness as well as support. However, the interpretation of other parties (and indeed, the patient’s own interpretations) should not be accepted at face value or uncritically. Much can be gleaned from the patient’s spontaneous bearing, conduct and reactions from the moment they enter the consulting room; this requires active and skilful observation, honed (like other neurological skills) through experience and apprenticeship to more experienced clinicians. The patient should be allowed to express emotions and, indeed, they (together with one’s own emotional reactions to the patient) can be highly informative. This raises the large and important theme of psychological factors in neurological illness. The border between psychiatry and neurology is extensive and ill defended: not only do neurological symptoms and ‘functional’ disorders lacking an adequate organic explanation account for many presentations to neurology clinics, but psychiatric symptoms may signal brain disease, mask it or modify its expression (through mood changes, denial, elaboration or abnormal illness behaviour; see Chapter 30). The neurologist must therefore be comfortable exploring the psychological dimension of the patient’s presentation and interpreting neurological symptoms in psychological terms, as well as asking more expert psychiatry colleagues for help where required. Neurological practice often involves sharing complex or nuanced information or breaking unwelcome news. This should of course always be done sensitively and after sufficient preparation. However, challenging issues around diagnostic labels, prognosis, occupational or social independence should be presented clearly and truthfully, to ensure that the patient and their supporters have clear advice to act upon and also to underline that the neurologist will continue to try to help them manage the illness. Receiving a diagnosis and an indication of its prognosis, even when guarded or unfavourable, may be tremendously significant and valuable for the patient and their loved ones. Where uncertainty exists, this should also be presented honestly. It is useful to summarise the consultation and discussion, and to ask, ‘Is there anything else you wish to tell me?’ or, ‘Do you have any questions for me?’

If the patient’s condition is reviewed at a subsequent appointment, it is often useful to reassess their understanding and to revisit the history, diagnosis and implementation of the management plan; noncompliance is surprisingly common, but the neurologist should recognise and accept gracefully that the patient is the ultimate arbiter and pilot of their own illness. For difficult diagnoses, a second opinion may be entertained and is almost always valuable. However, even patients with ‘incurable’ illnesses often find periodic contact with the neurology clinic helpful and the therapeutic import of this should not be underestimated.

The ‘Neurological Method’

The basic rationale and structure of the neurological consultation as enshrined by Gowers, Holmes and others more than a century ago has proved remarkably resilient and adaptable. Contemporary neurologists, however, have at their disposal a much richer armamentarium of investigations and therapies, and deploying these effectively is just as artful as the initial flourish of clinical deduction. In defining the patient’s problem, the neurologist seeks to localise it within the nervous system, to characterise the process by which symptoms have arisen and to determine the underlying cause; for example, the sudden onset of expressive aphasia (symptom) signalling left inferior frontal lobe damage (localisation) due to a cerebrovascular insult (process) produced by carotid dissection (cause). In any neurological consultation, characterisation of the patient’s symptoms by taking the history should be pre-­eminent. Definition of the disease process relies on the history, which also generally suggests localisation. The neurological examination is guided by tests and corroborates the historical impression. In particular, by delineating a profile of signs it defines the neuroanatomy of the problem more systematically than is often possible from the patent’s account alone. Examination is, of course, not restricted to the couch or bed. Invaluable information (indeed, sometimes the major diagnostic lead) may be gleaned from observing the patient throughout the consultation, especially while they are unaware they are being ‘examined’. Movement and gait disorders, visual disorientation and major neuropsychiatric syndromes not infrequently declare themselves as the patient enters the room. Identifying the cause of the problem then often entails ancillary investigations guided by the history and examination, although sometimes the history alone subsumes the other stages (as in the diagnosis of migraine). The neurological method is much more than the cold working out of clinical logic. The process of deduction, in which sequential clinical hypotheses are made and tested to arrive at a diagnosis is fundamental to the process of neurology but so too is induction – pattern recognition, based on past experience with similar constellations of clinical features. Experienced neurologists learn to tailor the history and examination dynamically, in response to the patient’s presentation and the gestalt of the clinical encounter. Clues furnished by a bespoke history target the examination, so that it becomes an exercise in active hypothesis testing. This saves time in a busy clinic and also avoids perplexing the patient, who may legitimately wonder (to quote Holmes’ example) why the neurologist is embarking on fundoscopy when they have presented with a weak leg. The art of neurology, of course, is to recognise the apparently irrelevant clue that proves decisive (the throw-­away reference to colour desaturation that signals a case of multiple sclerosis). This only becomes possible through seeing many, many patients and by learning the variations of normal as well as the inklings of pathology. All stages of the neurological method depend on an active partnership with the patient. This is more likely to be secured through

Approach to the Patient with Neurological Disease  19

dialogue, trust and compassion. Even where the cause of neurological symptoms cannot be confidently identified, it is important not to lose opportunities to manage the illness. Patients can benefit immensely from the mobilisation of practical and psychological supports and the relief of symptoms; clinical decisions in neurology often must be taken despite limited or incomplete information.

The Neurological History

The history is the cornerstone of neurological diagnosis. In many neurological disorders, the diagnosis rests almost entirely on the history but in all cases it is essential to an understanding of the disease process, its chronology and its setting in the wider life of the patient, which in turn will guide management. As Campbell and Barohn (2020) have observed, ‘The importance of the clinical history cannot be over-­emphasised. History taking is an art. It can be learned partly through reading and study, but it is honed only through experience and practice’. The patient should be invited to state their primary problem in their own words, without interruption; it is often useful to ask, ‘Why have you come to see me now?’ What follows, however must be actively decoded. Patients vary widely in education, eloquence, insight, powers of observation and recall, as well as their expectations based on past contact with other health professionals. The account that is volunteered will reflect what is important to the patient, or what they believe to be important for the doctor to know. It may minimise other (potentially crucial) symptoms, due to fear, embarrassment, ignorance or simply because the patient has become accommodated to them. The label ‘poor historian’ generally indicates the clinician’s inability to elicit required information, impeded by anxiety, misunderstanding, language or cultural barriers. The doctor should be prepared to change tack and reassess why communication is so unsatisfactory. It is often helpful to summarise points of the history back to the patient, to help ensure accuracy and to reassure them that the physician has heard and assimilated the story. Some patients simply cannot present a history coherently or reliably. This may be obvious where the presenting problem is cognitive impairment, but more subtle in other scenarios, such as an incipient confusional state, decompensated psychiatric illness, reduced level of consciousness or in the setting of pain or learning difficulties. Recognising this may be the key that unlocks the illness. An accompanying person, whether a family member or intimate acquaintance, may provide a perspective on the illness of which the patient is unaware. Indeed, it is crucial to seek corroborating testimony when the patient’s insight is suspected to be compromised and/or where the presenting problem is cognitive, psychiatric or behavioural in nature (see Chapter 11) or where the patient is otherwise unable to provide a reliable account (e.g. concerning an episode of altered awareness). If the patient has a bed partner, they may be the only witness to a clinically relevant sleep disorder. While information from previous medical assessments should always be reviewed and synthesised, it is important to avoid endorsing preconceived ideas about the patient’s illness or seeming judgemental or dismissive. A leading priority of the consultation is to ensure the neurologist is speaking the same language as the patient. This is generally not a primarily linguistic challenge; rather, it encompasses the subtler registers of education, social and cultural milieu and above all, the naming of neurological symptoms. Certain terms are notoriously liable to be used differently by doctors and lay people. When such a term is used by the patient, the neurologist must establish what they

mean by it, and if used by the neurologist, they must first ensure that the patient understands them. Examples include: • Dizziness • Weakness • Numbness • Clumsiness • Blackouts • Poor memory • Slurred speech • Confusion. The fundamental considerations in eliciting the history of a neurological presenting complaint include: • The way in which the symptom(s) began – this is crucial. Was the onset abrupt or more insidious, and was there a prodromal illness or other event (allowing that such attributions are susceptible to recall bias)? Keep in mind that ‘sudden’ and ‘gradual’ mean different things to different people – it is helpful to try to pin down the timeframe more explicitly; for example, was the symptom first present when the patient awoke or can they recall what they were doing when it developed? • The tempo and time course of the symptom(s) – how precisely can the onset be dated, has the course been rapid or more insidious? Has it been essentially static, fluctuating, episodic or intermittent? Has the overall trend been for deterioration or improvement? Has there been a recent change? • If intermittent or episodic, what is the character, duration, frequency and severity of the bouts, and what is the relationship to external factors? • Functional milestones and consequences of the complaint  – examples might include when the patient first sought medical assistance (and why), when they last considered themselves well, when they stopped work (and why), or when they first needed to use a stick (or first fell); if there is weakness, what daily life or occupational activities does this affect or preclude? • The character and intensity of the symptom, exacerbating and relieving factors, and any associated features  – and if it can be localised, its focus and any spread? A cardinal example is pain. • It can be helpful to ask what a symptom is like at its worst and at its best, perhaps with a scale of severity. • Are there seasonal, circadian or sleep-­related effects? • Previous events such as neurological symptoms that resolved, and any history of trauma, stroke-­like attacks, headaches, unexplained episodes of altered awareness or psychiatric illness. • What treatments have already been tried and what has been the response, if any? It is not uncommon in neurology for more than one (and sometimes multiple) symptoms to be volunteered. Each of these should be characterised, making special note of their sequence. The ­neurologist thereby builds up a picture of the illness and frames a  differential diagnosis, guided by the recognition of certain key patterns. Symptom localisability and distribution help to determine whether the disease process is focal or diffuse, and its likely site of origin within the nervous system. The character of the complaint frequently signals the major pathway(s) or system(s) predominantly affected (motor, sensory, autonomic, cognitive, etc). It may also hold a clue to the general nature of the disease process; thus, ‘negative’ symptoms (loss of function) tend to result from direct destructive injury (e.g. hemiparesis after stroke); whereas ‘positive’ symptoms (abnormal excess of function) may result from disruption of regulatory controls (e.g. stiffness [spasticity] with upper motor neuron lesions, hallucinations in cerebral disease) or irritation ­

20

Neurology: A Queen Square Textbook

(e.g.  dermatomal pain resulting from spinal nerve root compression). The chronology of the problem, however, is usually the most informative guide to the kind of pathology at work. Symptoms developing acutely (over minutes or hours) are often vascular or infective in nature, while subacute onset (over days to weeks) may signal an inflammatory or neoplastic process and more chronic evolution (many months to years) is more typical of primary neurodegenerative pathologies. Chronic disease processes are also more likely to be associated with aberrant or partially compensated functions, due to behavioural adaptation and neural plasticity. The neurological history is not complete until the presenting complaint has been understood in its medical, social and human context. A demographic sketch of the patient will ideally emerge during the consultation’s opening exchange, but it important to record (alongside their age and sex) such details as occupation (current or previous), handedness, ethnic and cultural background, mother tongue and level of literacy and formal education. The past medical history is sometimes elicited in a cursory fashion, yet it is frequently pivotal in understanding the patient’s current presentation. The ability to rapidly distil essential information from the morass is an enviable skill, and today’s neurologists must find opportunities to hone this with the case file’s digital descendants. It is important not to take previous diagnostic labels at face value. For example, transient ischaemic attacks are frequently ­misdiagnosed, and there is no agreed reference standard for a firm diagnosis. It may be necessary to verify reported or documented diagnoses against the relevant contemporaneous source material relevant to the diagnosis. Recording medications is especially pertinent; many drugs alter neurological function secondarily and older patients, in particular, are often the victims of polypharmacy. The potential for overuse, abuse or intoxication (advertent or inadvertent) should be considered. It is generally relevant to document usage patterns for alcohol, tobacco and other recreational drugs. The family history is a fundamental part of establishing a neurological diagnosis. It is particularly important to record the state of health of the patient’s parents and siblings, their ages and cause of death, as appropriate. It may be necessary to interview family members and to review their records. Uncovering the pattern of manifestation between generations and within family members is essential to establishing the mechanism of genetic transmission. Ethnic and geographical background and any parental consanguinity should be clarified. In younger patients, it may also be pertinent to ask about the circumstances of their birth and early childhood development. In taking the social history, it is often helpful to ask, ‘who is at home with you?’ and to establish marital status, the presence and ages of children or dependents, the nature of the patient’s work, and their daily routine and environment. Eliciting a social history also allows the examiner to enquire about interests, hobbies and holidays. Loss of social facility, awareness or engagement, uncharacteristic faux pas, work conflicts, relationship problems or the development of new interests (particularly where these have an obsessional flavour) can provide vital clues to an unsuspected early dementia or psychiatric problem. Diet and nutrition should not be overlooked and may be highly relevant to some neurological presentations. The sexual history requires tact but should not be eschewed as a source of mutual embarrassment. Recent alterations in sexual interest or performance can be important clues to more pervasive behavioural or autonomic dysfunction or help to localise a spinal lesion; while more longstanding patterns of sexual risk taking may predispose to syphilis or HIV, the neurological manifestations of which are protean.

The Neurological Examination

There are several compelling reasons to examine a patient ­neurologically. Besides assessing the working hypothesis about the ­illness developed during the history, the examination should seek informative negatives (e.g. retained dorsal column sensation in a suspected cord lesion). It also provides the neurologist with a picture of complex functions (such as walking and aspects of cognition) that may be challenging to dissect on the history alone, and may suggest a therapeutic target (e.g. spasticity, ­dystonia). In unusual circumstances (such as an obtunded or unconscious patient), it may be the only avenue open to neurological assessment. As with the history, it is important to note how the patient performs a task, as well as whether they ultimately achieve it; for example, even minor degrees of hesitation may signal pathology when testing proprioception or visual apperceptive function, while psychogenic deficits may be dramatically effortful and marked impulsivity or passivity may betray a frontal lobe lesion. Just as it is helpful to record the patient’s comments verbatim when taking the history, so it is often preferable to record a description of what the patient actually does on examination, rather than risk mislabelling an ‘agnosia’ or ‘rubral tremor’, particularly for inexperienced examiners. The neurologist should take a few moments to stand back and discreetly observe the patient. With practice, much valuable information may be learned from informed inspection alone, such as watching the patient walk into the consulting room or while you take the history. If on the ward, glance round their bed for potential clues to the diagnosis and the wider impact of the illness – walking and hearing aids, numerous pairs of spectacles, packets of sweets, books and photographs. Consider the following: • Bearing and demeanour (posture, eye contact, vigilance vs psychomotor retardation, tearfulness, evidence of continuing visual or auditory hallucinations, dress and grooming). • General habitus (e.g. height, build) and characteristic facies (e.g. Cushingoid appearance, myotonic and facioscapulohumeral dystrophies). • Limb deformities or abnormal posture (e.g. pes cavus of Charcot– Marie–Tooth disease, wasted limb of polio, upper motor neuron contractures). • Facial movement (e.g. hypomimia, ptosis, frontalis overactivity, dysconjugate gaze, impression of upper or lower facial asymmetry or weakness). • Skin (e.g. sweating, zoster, heliotrope rash or other stigmata of connective tissue disease, lupus pernio, erythema nodosum, livedo reticularis, tumours, changes of hypothyroidism). • Scars (e.g. muscle biopsy, nerve biopsy, previous surgery, ­especially cervical and lumbar, burr holes, accident and injury or self-­harm). • Muscle wasting (especially intrinsic hand muscles, global vs proximal vs distal distribution, cachexia). • Fasciculations (distribution  – may be widespread or focal, ­especially over triceps and deltoids). • Abnormal movements (tremor, chorea, ballism, myoclonic jerks, tics). • Voice volume, quality (e.g., hoarseness, nasality, growling) and idiosyncrasies (e.g. tics, hiccough)  – the observant neurologist will listen as well as look, and not only to what the patent says but also how they say it. Having observed, it is time to commence the examination proper. Ensure before beginning that the patient is not in pain. Although the neurological examination can often be targeted, a succinct

Approach to the Patient with Neurological Disease  21

screening neurological survey is a useful tool in the assessment of many patients. This should include the following: • Gait • Mental state  – affect, engagement and orientation in person, place and time • Optic fundi • Cranial nerves: ◦◦ Visual fields ◦◦ Pupils ◦◦ Eye movements ◦◦ Facial sensation and movement ◦◦ Hearing ◦◦ Bulbar function • Motor: ◦◦ Posture and stability ◦◦ Outstretched arms ◦◦ Upper limbs  – deltoid, triceps, biceps, wrist extension, finger abduction, abductor digiti minimi and first dorsal interosseous ◦◦ Lower limbs – iliopsoas, hamstrings, quadriceps, ankle dorsiflexion and plantar flexion, extensor hallucis longus • Tendon reflexes and plantar responses • Coordination • Sensation. A more searching examination may be directed to any of these areas, guided by the history or signs uncovered on the screen. We now consider in more detail the individual components of the full general neurological examination. Detailed examination of the coverings of the nervous system will be guided by the history and may include the following: • Head: skull shape and size (head circumference), injury, deformity, bruising, bleeding and leakages of cerebrospinal fluid (CSF) from the nose and ear, scars, shunts are features indicative of previous neurosurgery or injury, blood vessel (palpate temporal arteries) and listen for bruits over carotid artery and subclavian arteries. • Neck: the neck must not be passively moved following major trauma to head, neck or rest of the body until neck injury has been excluded by radiology. Assess the posture, shape (are there any obvious point of tenderness?), stiffness or axial rigidity (cervical spondylosis, extrapyramidal disorders such as multiple system atrophy or Parkinson’s disease), meningism (due to blood or inflammation in the CSF, cerebellar tonsillar herniation, inflammation of spinal or adjacent structures), and range of mobility. • Spine: assess for shape (scoliosis, lordosis, kyphosis), scars of previous surgery or injury, cutaneous lesions and pressure points (e.g neurofibromatosis), mobility and tenderness. • Eyes (see below): alignment, protrusion (proptosis), conjunctival and periorbital haemorrhage or bruising and vision. • Ears: discharge, Infection and the state of the eardrums. • Face: alignment, jaw appearance, movement and symmetry (temporomandibular joint dysfunction). • Mouth: patency of airways, presence of pooled saliva, oral ­infection, dentition. • Sinuses (Facial tenderness, nasal discharge). Level of Alertness and Orientation Assessment of the level of consciousness is discussed in Chapter 28. Following the history, it will be clear if the patient is alert and able to speak clearly and coherently. General orientation can be established by asking them who they are, where they are, the date and approximate time of day. Although delirium will often be evident or suspected from the clinical context (and may preclude detailed

cognitive assessment), testing working memory (e.g. asking the patient to recite the months of the year backward) may expose an incipient acute brain syndrome; see Chapter 11). Higher Cerebral Function Assessment of higher cerebral function is discussed in detail in Chapter 11. This is relatively time-­consuming even for experienced clinicians and we do not routinely examine higher functions in a patient presenting with an unrelated issue who gives a cogent history that does not reveal any cognitive or behavioural symptoms. However, where such symptoms do emerge or where the presentation is such as to raise suspicion that aspects of cognitive function may be compromised (e.g. accompanying a movement disorder or stroke), the following domains can be targeted. It is very useful to have to hand some ‘prompts’ for eliciting important functions (Figure 2.1), such as a set of familiar pictures to assess naming. A number of brief instruments (e.g. the mini-­mental state examination) are available for screening cognitive function at the bedside; they can be useful adjuncts but should never be used to make or to exclude a diagnosis of dementia. Fundamentally, the formal assessment of higher functions is only reliable in a cooperative patient and after any defects of peripheral vision and hearing are adequately corrected. • Behaviour and executive functions are mediated by frontoparietal and subcortical circuitry but are often poorly localised. Executive dysfunction is often suspected from history but best quantified by a neuropsychologist – it is seldom decisive in clinical diagnosis. Assess: ◦◦ Engagement (e.g. confusion, irritability or distractibility – does the patient behave appropriately to their situation and in their interactions with the examiner?). ◦◦ Affect and conduct (e.g. passivity, utilisation behaviour/environmental dependency, disinhibition, stereotypies). ◦◦ Response inhibition (Stroop task)  – stating the (conflicting) colour of the ink in which a series of colour names are printed (Figure 2.1). ◦◦ Generation of a list of words to criterion specified by the examiner (e.g. category/initial letter (verbal fluency – in non-­aphasic patients). • Spatial attention may be impaired, especially with focal, acute non-­dominant parietal lesions. Assess: ◦◦ Pointing to examples of a particular letter embedded in a visual array – (letters in one half of space are missed in hemispatial neglect). ◦◦ Detection of simultaneous compared with sequentially presented visual or tactile stimuli – one stimulus (usually the left) is missed when presented together with a contralateral stimulus in sensory extinction. • Speech functions should be examined if there is suspicion of a communication disorder, especially with diseases involving the left cerebral hemisphere and, more specifically, the classical language ‘areas’ of Broca (inferior frontal gyrus), Wernicke (temporoparietal junction), anterior temporal cortex and their connections. Aphasia (dysphasia) is inability to produce and/or understand language  – various categorisations based on the most evident impairment are in clinical use (e.g. expressive vs receptive, non-­ fluent vs fluent aphasia), although none is fully satisfactory from a pathophysiological perspective (see Chapter 11). Aphasia is to be contrasted with other kinds of speech disorders that may coexist, such as dysprosody (impaired intonation or ‘melody’ of speech), dysarthria (impaired articulation, manifesting as slurred or uncoordinated speech due to bulbar or cerebellar dysfunction)

22

Neurology: A Queen Square Textbook

(a)

(b)

John Gifford was the kind of person who would go off sailing in his yacht the 'Neva' around the island of Scorba whenever there were signs of chaos at work. He had a thorough knowledge of this area as he had grown up there in his childhood, and this was how he always sought relaxation from the busy routine of the office life at Bergess, Challice & Co. He would sit on deck

(d)

sipping his champagne, as the breeze tangled his normally immaculately combed hair. Gradually his business worries would recede. Dressed in his old, baggy sweater, decrepit suede boots and with his stubbly beard he felt quite the part of an ancient mariner. (c)

(e)

Figure 2.1  Examples of aids for testing cognition at the bedside. These are some examples of materials that we use to sample a range of cognitive domains (see also Table 2.1) when examining a patient with suspected dementia (shown with the permission of Professor E.K. Warrington). (a) A pictorial scene for description, to assess connected language (fluency, word finding and expressive grammar) and ability to parse a complex environment (visuospatial processing). (b) An array of pictures of familiar items for assessing word retrieval (naming), single-­word comprehension (pointing to an item named by the examiner) and subsequent item recall (anterograde episodic memory). (c) A passage for reading aloud to elicit difficulties sounding nonwords (personal names, e.g. ‘Scorba’; phonological dyslexia) or irregular words (e.g. ‘yacht’, regularised as ‘yached’; surface dyslexia) or with tracking lines of text (visuospatial processing). (d) Fragmented letters to assess visual apperceptive function. (e) A Stroop task, requiring the patient to state the (conflicting) colour of the ink in which each colour name is printed (in fluent readers of English, a test of response inhibition: one aspect of executive function). Source: reproduced with permission from Johnson et al. (2021).

and dysphonia (impaired production of vocal sounds due to laryngeal or extrapyramidal dysfunction). Assess: ◦◦ Production of connected speech; for example, elicited through description of a pictorial scene (Figure 2.1), noting abnormalities of content (e.g. word finding pauses, circumlocutions, jargon), structure (agrammatism, often more easily assessed from written sentence production) and execution (speech sound errors – phonological [wrong syllables], speech apraxia [effortful, hesitant, variable misarticulations], dysprosody). ◦◦ Naming of familiar items or pictures (Figure 2.1). ◦◦ Repetition of speech  – may be disproportionately affected in conduction aphasia, spared in transcortical aphasias. ◦◦ Comprehension of single words (pointing to an item or picture nominated by the examiner) and sentences (following spoken commands varying in length and syntactic complexity). ◦◦ Production of syllable strings of varying length (e.g. ‘pa-ke-ta’) – to assess speech rhythm and articulatory agility.

• Ability to read, write/spell and calculate is relevant particularly if there is a speech problem and/or suspicion of a dominant parietal lesion. Alexia (dyslexia) is inability to understand written language, not due to visual or general intellectual impairment (beware developmental dyslexia which commonly went undiagnosed in past generations). Agraphia (dysgraphia) is inability to produce written language, not due to motor or sensory loss. Assess: ◦◦ Reading aloud both nonwords (e.g. ‘scorba’) and irregular words (e.g. ‘chaos’) – use of nonwords and irregular words here tests phonological and vocabulary-­based literacy skills, which have different cerebral associations (Figure 2.1; Chapter 11). ◦◦ Spelling both nonwords and irregular words (examples above). ◦◦ Performance of simple mental arithmetic sums (this skill depends on attentional and working memory capacity as well as prior numerical facility). • Working memory is true ‘short-­ term memory’, immediate rehearsal over periods up to around 30 seconds – it has localising

Approach to the Patient with Neurological Disease  23

value for left parietal lesions but also depends on attention and executive capacity. Assess: ◦◦ Repetition of a string of spoken digits of increasing length forward (forward digit span; [normal 7]) and backward (reverse digit span [normal 5] – a measure of executive function). ◦◦ Repetition of phrases (if clearly worse than repetition of single words this points to a deficit of auditory verbal/phonological working memory). • Programming of voluntary action (praxis) is mediated chiefly by the left parietal lobe but with additional, separable anterior control centres, accounting for asymmetric or body-­part specific impairments. Apraxia (dyspraxia) is inability to formulate or execute voluntary movements, not due to weakness, sensory loss or ataxia. Assess: ◦◦ Pantomiming of familiar gestures not shown by examiner (e.g. waving, saluting, using a screwdriver) – ideational apraxia. ◦◦ Imitation of meaningless hand postures shown by examiner – ideomotor apraxia (note that imitation of a sequence of hand positions (Luria sequence) is a searching test of praxis rather than frontal lobe function). ◦◦ Performance of orofacial gestures to command (e.g. yawning, whistling)  – orofacial apraxia (often accompanies speech apraxia). • Recognition of people and objects is mediated chiefly by right  (faces) and left (objects) anterior and inferior temporal lobes. Agnosia is inability to recognise familiar faces (prosopagnosia), objects (visual associative agnosia) or other aspects of the environment, in the absence of explanatory sensory loss. Assess: ◦◦ Recognition of famous faces based either on naming or (if naming is impaired) specific biographical details (e.g. identifying the person’s job or nationality – true prosopagnosia manifests as a loss of knowledge of the person, rather than simply inability to retrieve their name). ◦◦ Ability to indicate what a common object (e.g. hairbrush) would be used for. • Visual apperceptive and visuospatial skills are mediated by the right parietal lobe. Assess: ◦◦ Ability to perceive fragmented letters or pictures (Figure 2.1) – if unable despite normal recognition of non-­fragmented equivalents, this indicates visual apperceptive agnosia. ◦◦ Ability to count dots in an array – visuospatial agnosia. ◦◦ Ability to locate the examiner’s outstretched hand – if there is visual disorientation, the patient will typically touch the doctor’s face or arm while groping clumsily toward the target (note that the deficit may be lateralised but the patient can still perceive movement in the affected field, indicating there is no hemianopia). • Episodic memory is the autobiographical record of everyday events, mediated chiefly by the hippocampi and their connections and with clinically relevant components of encoding, recall and recognition. Memory impairment is often suspected from the history but best quantified by a neuropsychologist. The neurologist can assess: ◦◦ Recall of verifiable recent episodes such as the patient’s route to the hospital (more specifically, this tests topographical recall) or salient events such as procedures during their stay (with/ without cueing). ◦◦ Recall of a name and address supplied by the examiner after a delay of several minutes (with/without cueing). ◦◦ Recall (in non-­ aphasic patients) of previously presented ­pictures from the naming task (with/without cueing).

Gait and Mobility The gait should be observed in all patients capable of independent mobility. Observation of walking can quickly reveal disorders that are not obvious when the patient is examined on the couch, such as hemiparesis, parkinsonism, abnormal movements (tremor, chorea, athetosis, dystonia), proximal muscle weakness and truncal ataxia. Look for the presence of any walking aid, observe the stance and gait for balance, the distance that the feet are apart, the arm swing, general posture symmetry and test the patient’s ability to stand on tiptoes and heels, to walk heel to toe and undertake Romberg test by asking the patient to stand with their feet together and close their eyes. A positive Romberg test suggests that there is a loss of joint position sense (i.e. sensory ataxia). When examining the gait, remember to walk alongside the patient if they appear unsteady. Ask the patient to stand and assess stability and determine if normal or broad-­based, suggesting a cerebellar problem. Assess if the gait is symmetrical or asymmetrical. Neurological gait disturbances are discussed in the appropriate chapters and include: • Spastic – scissoring, narrow-­based, stiff with toe scuffing • Hemiparetic – circumduction with scuffing, which often causes excess wear of the instep of the shoe • Extrapyramidal – shuffling, festinant with poor arm swing, very slow (parkinsonian) gait with freezing and slow turning • Apraxia – gait ignition failure, walking with difficulty, marche à petit pas • Ataxia – broad based and unsteady • High stepping  – may be due to myopathic or neuropathic foot drop or pain (antalgia) • Dystonic, choreiform, myoclonus • Myopathic • Neuropathic • Functional. Cranial Nerves Abnormalities of function are considered in detail in the relevant chapters, particularly Chapter 13.

I Olfactory

Ask the patient whether they have noticed any change in their sense of smell. A simple and quick assessment of olfaction at the bedside can be carried out with the peel of an orange, scented soap or coffee. Check whether the patient can distinguish between rather than recognise odours testing one nostril at a time. However, few neurological offices are equipped to test olfaction reliably, and if this is a relevant clinical issue, a systematic instrument (such as the University of Pennsylvania Smell Identification Test) should be considered.

II Ocular Examination

Before commencing formal examination of the eyes look for: • Ptosis – observe where the upper eyelid bisects the iris. • Pupils – assess pupillary size, asymmetry, pupil irregularity. • Eye position – check whether the eyes are conjugate (point in the same direction) or dysconjugate in the primary position of gaze. • Proptosis – observe the globe from above for globe protrusion. The examination of the eyes should follow a clear routine involving assessment of: • Visual acuity • Colour vision (using Ishihara plates) • Fields to confrontation • Eye movements: tracking and saccadic movements • Pupillary light and accommodation reflexes • Fundoscopy.

24

Neurology: A Queen Square Textbook

Visual Acuity Asking the patient to read small print on a newspaper (with reading glasses) is a convenient basic assessment of reading vision. Distance vision is performed with a Snellen chart read at 6 m. Reduced-­size Snellen chart are available to be read at 2 m or 3 m. The patient must wear their distance glasses and each eye should be tested individually while covering the other. Ask the patient to read down the chart and when they start having difficulties encourage them to guess. If they are able to get more than half the line correct then they have achieved this line on the chart. If the vision is worse than 6/6, a pinhole can be used to correct refractive errors, keeping distance glasses on. If they cannot see the top line (6/60), establish a visual acuity in the following order: • Ability to count fingers at 1 m • Detection of hand movements • Light being shone in the eyes (covering the other eye) -­perception of light and no perception of light. Colour Vision Colour vision is often examined using standard plates (e.g. Ishihara chart) as part of the neurological examination if an optic nerve lesion is suspected. Colour vision has traditionally been tested at the bedside by assessing the extent of desaturation when viewing a red target. Examining the Fundus When examining the patient’s right eye, the observer should hold the ophthalmoscope in the right hand and examine with their right eye and vice versa for the left. The observer should hold the ophthalmoscope vertically with the hand rested on the cheek to steady the ophthalmoscope and with the finger maintained on the focus dial, adjusting the ophthalmoscope to correct for the patient’s refractive error during the examination. The patient should be seated with the observer standing and they should be asked to fix their gaze on a distant object, ideally high on the wall. The red reflex should be assessed at arm’s length from the patient. If the red reflexes are absent or reduced, this indicates an opacity in the ocular media (e.g. cataract or vitreous haemorrhage). Commence the examination approximately 15 degrees temporal to their direction of gaze otherwise the patient may try to accommodate by focusing on the light, making it difficult for the observer to focus on the retina. Alternatively, if the patient is asked to look directly at the light of the ophthalmoscope, then the examiner will be looking directly at the fovea (the centre of the macula). The observer should follow the vessels observing the disc for colour, contour and cup appearance. The superotemporal arcade vessels should be followed away from the disc and then back towards the disc. This should be repeated for the inferotemporal vessels. While following vessels any abnormality, including haemorrhage, exudates, pigment, atrophy, vessel change and laser scars, should be noted. Finally, the appearances of the macula and peripheral retina should be assessed. Visual Fields When assessing the visual fields by confrontation, the aim is to compare the patient’s visual fields with the observer. The assessment should be undertaken exactly halfway between the patient and the observer. The patient should wear their glasses to allow fixation. Assess for visual attention by asking the patient to look at the bridge of the observer’s nose with both eyes open while the

index finger is held just inside the outer limits of the temporal fields. Then move the fingers in turn and then both at the same time and ask the patient to point to the finger that moves. If there is visual inattention, the patient will only point to one finger when you move both at the same time, neglecting the abnormal side, even though the visual fields are normal. Next, the monocular visual fields are assessed by asking the patient to cover their left eye while the observer covers their right eye. The patient is asked to look at the observers uncovered eye. Using a small object (e.g. 5–8 mm white pin head), place it at central fixation (in front of the patient’s pupil) and ask the patient if they can see the target. The target is moved slowly at a 45-­degree angle in a straight line from the periphery to the front of the pupil. When the target appears, continue to the pupil asking the patient if it disappears to prevent missing a central scotoma. Once a field defect has been detected, the observer should assess where the borders of the defect lie, particularly with regard to the vertical and horizontal midline then take the target and place it in the blind area. The target is moved to the seeing area and the patient is asked when they first see it. In a central scotoma, move radially in and out in different directions. If there is a hemianopia, move horizontally from the blind to the seeing area, testing the vertical midline both above and below central fixation but do not move the target too fast, especially in the temporal field. The visual fields are drawn in the notes as seen by the patient (i.e. left eye on left and right eye field on right and labelled appropriately). Eye Position and Movement Assess the eyes for ptosis, proptosis, and lid retraction, the upper lids normally rest 2  mm below the superior limbus (junction between the sclera and upper cornea) while the lower lids rest at the inferior limbus. Consider whether the eyes are straight or if there are features of convergent squint (VIth nerve palsy), ‘down and out’ eye (IIIrd nerve palsy) or vertical deviation (suggesting IVth nerve palsy and associated with head tilt). Assess whether the pupils are equal and look for dilated pupil (compressive IIIrd nerve palsy) or constricted pupil suggesting Horner’s syndrome. Using a pointer or light source, ask the patient to keep their heads still, follow the light with their eyes and report if they see double. Move the pointer up and down looking for lid retraction, lid lag and restricted upgaze. Then move in an H shape to examine the six cardinal positions of gaze while inquiring about the presence of double vision. If the patient sees double (two images of the target), cover either eye in turn and ask the patient which image disappears (cover test). The paretic eye produces the more peripheral and usually fainter of the two images. If the patient sees double even with one eye closed (monocular diplopia), there is no point in continuing with the cover test. Check each extraocular muscle individually by asking the patient to track movements in the shape of an H. Lateral movements are assessed by following elevation or depression to about 30 degrees of lateral gaze. To determine which muscle is weak, the following points should be considered: • The false image is usually less distinct. • Diplopia occurs in positions that depend on contraction of the weak eye muscles. • A false image is projected in the direction of action of the weak muscle. • Image separation increases in the direction of action of a weak muscle. The consequences of individual lesions of III, IV and VI are considered in Chapter 22.

Approach to the Patient with Neurological Disease  25

Ocular Movements The following movements should be examined. • Rapid ‘saccadic’ (voluntary quick movement). • Pursuit (following an object moving in the visual field). • Optokinetic (such as observed when a person looks from a train window at passing objects). Look for nystagmus at rest and during all movements. The ­different patterns of nystagmus are discussed in Chapter 22. Pupils Normal pupils are equal in size and dilate with dim light and constrict in bright light. It is essential to test responses using a bright pen torch in a dimly lit room. Always stand to the side of the patient to avoid inadvertently inducing accommodation pupillary constriction. Light stimulates both temporal and nasal sides of the retina. Afferent impulses travel from the retina by the optic nerve. At the optic chiasm, impulses from the nasal retina cross over, while those from the temporal retina remain ipsilateral, so both optic tracts transmit impulses from one eye to the midbrain. Hence, a bilateral and equal efferent response generated in the Edinger–Westphal nucleus travels in parasympathetic preganglionic fibres within both III nerves to the ciliary ganglion in the orbit. Postganglionic fibres then pass to the pupil sphincters, thus producing pupil constriction in the test eye (direct response) and the contralateral eye (indirect or consensual response). In a unilateral optic nerve lesion, there is typically a loss or reduction of both direct and consensual light response to light shone into the affected eye compared to the brisk response seen when the light is shone into the normal eye, and may be seen more easily by swinging the torch from one pupil to the other shining the light for 3 seconds in each eye. The accommodation reflex is tested by asking the patient to look into the distance then close to a target held 20 cm away at eye level and observe whether the pupil constricts. Loss of pupil constriction to a bright light with preservation of the near responses can imply a lesion in the midbrain. Abnormalities of pupillary responses are discussed in Chapter 22. With a lesion of the sympathetic pathway (Horner syndrome) the changes are best seen in a dim light and consist of ptosis, a pupil that is smaller on the affected side and loss of sweating over the ipsilateral forehead. V (Trigeminal Nerve) There are three divisions of the trigeminal nerve: ophthalmic (V1), maxillary (V2) and mandibular (V3). The motor division innervates the muscles of mastication. Test the bulk and symmetry of the masseter and temporalis muscles by asking the patient to clench their teeth while you palpate the muscles. If the pterygoid muscles are weak on one side, the jaw deviates to the affected side on opening the jaw. Upper motor neuron lesions cause a brisk jaw jerk, elicited by a downward tap on the chin when the mouth is hanging loosely open. The corneal reflex tests the trigeminal and facial nerves. Normally, stimulation of one cornea elicits a simultaneous direct and consensual blink, the efferent limb being via the VIIth nerve. Loss of the corneal reflex is usually the first sensory defect in a lesion of V1. Before testing the corneal reflex, check if the patient wears contact lenses. The reflex is tested with a wisp of cotton wool touched briefly on the cornea, approached from below with the patient looking upward so as not to cause an anticipatory blink. The absence of a corneal reflex infers that the cornea is at risk of injury from an undetected foreign body.

Sensory examination of the trigeminal nerve should include examination of all three divisions and comparison made with the other side. The sensory territory includes the face and the head anterior to the vertex, the eyes, the mucous membranes of the oral, nasal cavity, paranasal sinuses, tongue and the teeth. Central lesions (brainstem) can cause ‘onion skin’ sensory loss; Pinprick and light touch sensation are tested in each of the three divisions. It is important to note the extension of the ophthalmic sensory division to the vertex of the head where it abuts the C2 dermatomes, sparing of the angle of the jaw, which is supplied by the C2 and 3  nerve roots rather than the mandibular division of V (see chapter 13). VII Facial Nerve The facial nerve is tested by observing or testing, eye and mouth closure, facial expression and taste. The patient’s face is observed, noting asymmetry during spontaneous movement or smiling and whether the blink is symmetrical. The patient should be asked to raise the eyebrows, screw up the eyes and bare the teeth. In a lower motor neuron (LMN) lesion, the spontaneous blink is sluggish or absent on the affected side, weakness is obvious in the upper and lower face, and there may be an inability to close the eyes fully. Emotional facial expression, for instance while smiling or laughing, is as severely affected as a volitional movement. Bell’s phenomenon describes the eyeball rolling upwards when the patient tries to blink; it is a hallmark of an LMN lesion since the eyelid is unable to close properly. The VII nerve nucleus receives supraspinal input from both hemispheres (principally to frontalis) and therefore an upper motor neuron lesion of VII has sparing of the muscles of the upper face. Thus, in an upper motor neuron lesion, the lower part of the face is affected more than the upper part, the spontaneous blink is usually normal and emotional facial expression may often be less affected than voluntary movement. A small motor branch supplies the stapedius muscle; hence, Bell’s palsy causes hyperacusis. Taste fibres from the anterior two thirds of the tongue are carried by the cauda tympani but sensation to the tongue surface is supplied by V. VIII Auditory and Vestibular The examination and function of the VIII auditory and vestibular nerve are addressed in detail in Chapter 23. The auditory component can be tested by assessing if the patient can hear a watch ticking or fingers being rubbed together on either side. If necessary, mask one ear by rubbing the tragus and, standing out of the patient’s vision, whisper numbers into the other ear to assess any hearing loss. Rinné’s test is undertaken by placing a 512-­Hz tuning fork on the mastoid process and asking when the patient can no longer hear it. At this point, hold the same tuning fork in front of the ear and they should be able to hear it. Normally, air conduction at external auditory meatus is better than bone conduction. This is also found in deafness due to sensorineural hearing loss. Weber’s test is undertaken by placing a 256-­Hz tuning fork centrally on the forehead. If hearing is normal, the patient will hear the tuning fork equally in both ears. In unilateral conductive deafness, the sound is referred to the affected side and in sensorineural deafness to the unaffected side. The vestibular nerve is assessed by examining balance and eliciting nystagmus; this will have been previously tested during the evaluation of gait and the eye movements. Nystagmus is a pendular or jerking eye movement caused most often by a disorder of the peripheral vestibular apparatus, the vestibular (VIII) nerve or its central connections. It commonly consists

26

Neurology: A Queen Square Textbook

of a drift of the eyes away from the fixation point followed by a fast corrective ‘saccadic’ eye movement (jerk nystagmus). The movement may be horizontal, vertical or rotatory and the same or different in either eye. The direction of nystagmus is described as being in the direction of the fast corrective jerk (see chapter 22). The Hallpike manoeuvre for testing the posterior semicircular canals is performed by moving the patient from the seated to the supine position with the head turned to the side by 45 degrees and the neck extended by 20 degrees. The eyes are observed for nystagmus, which, in benign paroxysmal positional vertigo, characteristically has a latent period and a rotatory component. The test is repeated with the head turned to the other side (see Chapter 23). IX, X Glossopharyngeal, Vagus Nerve The IX, X glossopharyngeal and vagus nerves are generally examined together. The uvula and fauces are illuminated with a pen torch and the patient is asked to say ‘aah’. Check for symmetrical elevation of the soft palate and deviation of the uvula. With unilateral lesions of X, there is weakness of the uvula causing a droop of the palate and flattening of the palatal arch. On phonation, the uvula deviates to the normal side. The voice sounds ‘wet’ in the early stages of bulbar muscle weakness. The gag reflex is tested by using an orange stick to stimulate the posterior pharyngeal wall (IX). The palate should elevate symmetrically, with the stimulation felt similarly on either side. This test is unpleasant and rarely informative. Speech may be nasal in quality due to nasal escape of air with a weak soft palate. Phonation is harsh, hoarse (dysphonia) or reduced in volume (hypophonia) if the larynx or its muscles are not working properly, for instance, in a vocal cord palsy. When the patient is asked to cough with a Xth nerve palsy the vocal cords cannot be fully adducted, the sharp explosive quality of the cough is lost and it becomes ‘bovine’ (see chapter 13).

Swallowing

The patient should be observed swallowing water (assuming they are alert, cooperative, able to sit upright, not ‘chesty’, and that they have not been placed on a ‘nil by mouth’ order). If swallowing is suspected to be abnormal, a small amount of water in a teaspoon should be used initially. The observer should note the following: • Is the swallow slow and cautious? • Does the patient splutter, cough or choke implying penetration of the airway and aspiration? • Do they find water escapes into their nose when they swallow (due to soft palate weakness)? • Is there a ‘double swallow’? Accessory Nerve XI The accessory nerve has a cranial and spinal component. The patient is asked to turn their head to one side, pushing their chin into the examiner’s hand, and the contralateral sternomastoid is palpated for bulk. To test shoulder elevation, the patient is asked to shrug their shoulders against resistance and palpate the trapezius bulk on either side. Winging of the scapula is due to weakness of either serratus anterior or trapezius. Trapezius is supplied by XI, C3 and C4 and, when weak, the scapula winging is more apparent on attempted abduction of the arm rather than forward elevation. Serratus anterior (supplied by the long thoracic nerve originating from C5, C6 and C7) functions during forward arm elevation. When serratus is weakened, then winging is more ­obvious when trying to elevate the arm in front and less obvious when the arm is abducted to the side. This difference is important in distinguishing long thoracic lesions from spinal accessory nerve palsy.

Hypoglossal Nerve (XII) Inspect the tongue at rest on the floor of the mouth. Look at the bulk and observe for wasting, fasciculation and involuntary movement. When the tongue is protruded twitching is commonly seen in normal individuals. Look for wasting of the tongue and deviation to the weak side when protruded. The power of the tongue can be assessed by asking the patient to push against the inside of the cheek. Compare the strength by feeling the cheek on each side in turn. When unilateral weakness is present the tongue deviates to the weak side on protrusion because of the action of the normal genioglossus which protrudes the tongue by drawing the root forward. The patient cannot push the tongue against the cheek on the normal side but is able to push it against the cheek on the side towards which it deviates. Examine tongue movements by asking the patient to say ‘la, la, la’ or to make rapid side to side movements. The speed and amplitude of tongue movements are diminished in bilateral pyramidal lesions and often early in Parkinson’s disease. Muscles of the Neck Multiple muscle groups contribute to flexion and, except for the sternocleidomastoid and trapezius, it is not possible to test these muscles individually. Flexion is tested by asking the patient to place their chin on the chest as the examiner tries to extend the forehead, while the extensors are tested by asking the patient to extend against the examiner’s resistance applied to the occipital. Examination of the neck must be undertaken with care in any patient at risk of cervical spine disease. Limb Examination

Inspection

Observe the patient while they are relaxed at rest, during maintenance of a posture and when performing movements such as walking or other tasks. • Posture: observe the trunk and neck posture on standing and walking. It may be flexed in parkinsonism, with a lack of facial expression, and reduced limb movement. Dystonic movements may be present and there may be severe prolonged spasticity causing contractions of the limbs. • Assess the muscle bulk, looking for global, symmetrical or focal wasting by considering the general build and level of activity of the patient and by comparing sides to look for asymmetry. Muscle atrophy is associated with disease of the anterior horn cell, root, nerve and muscle but not the neuromuscular junction. It also occurs as a result of immobilisation, disuse, inactivity, cachexia, malnutrition and normal ageing. Muscle hypertrophy results from excessive usage, persistent abnormal muscle contraction and myotonia, infiltration or certain forms of muscular dystrophy. On occasion, it may be necessary to measure muscle bulk formally using a tape to assess the circumference at a fixed point above or below a landmark. • Fasciculations are fine, rapid, random, irregular and fleeting flickering movements of muscle, which vary in amplitude and intensity. They are best observed when a muscle and the relevant part of the body are completely at rest. Consistent muscle twitches are caused by irritability or disease in the motor neuron; however, they are commonly seen in the calves of normal subject. Other involuntary movements of muscles such as myokymia, neuromyotonia, spasms, cramps and non-­organic movements should be noted.

Tone

Ask the patient to lie in a relaxed state. Each arm is gently picked up and tone assessed through a range of passive movements; in particular,

Approach to the Patient with Neurological Disease  27

flexion and extension of the wrist and elbow and pronation/supination of the forearm and in the lower limbs. Particularly, test flexion/ extension of the knee and dorsiflexion and plantar flexion of the ankle. For the lower limbs, allow the limbs to rest on the couch, ensure that the patient is relaxed, then roll each limb from side to side at the knee then place the hand under the knee and observe whether the heel remains on the couch (normal) or flicks into the air (spastic). • Spasticity typically develops following an upper motor neuron lesion. It is most obvious with rapid, passive movements of the limb but it is velocity dependant and so may be less obvious with slow movements. Resistance is greater in one direction of movement than another; for instance, on passive elbow extension (due to selectively increased stretch reflexes in elbow flexors) rather than elbow flexion or for arm supination rather than pronation. There may be an impression of a ‘catch’ or ‘clasp knife’ reaction on rapid stretching of the affected muscles. In a spastic lower limb, forced dorsiflexion of the ankle, with the knee flexed, may cause rhythmic contraction of the gastrocnemius/soleus (ankle clonus). Clonus can also be seen at the patella and wrist. A few beats of clonus can occur normally or with anxiety but sustained clonus is abnormal and associated with a pyramidal lesion. In severe long-­ term spasticity, the patient may be unable to straighten the arm/ hand or thigh/leg/foot joints and a posture of fixed flexion or extension may be adopted, due to contractures, which also limits the effectiveness of opposing voluntary movements. • Rigidity is typically associated with extrapyramidal disease (e.g. Parkinson’s disease). Tone is increased in both directions of movement, throughout the range and independent velocity of movement but there may be an impression of cogwheeling or tremor superimposed on the spasticity. Rigidity of one side may be exaggerated by active movement of the contralateral limb (see Chapter 9).

Table 2.2  Nerve and main root supply of muscles (upper limb). Nerve

Muscle

Spinal root

Spinal accessory

Trapezius

C3, C4

Rhomboids

C4, C5

Serratus anterior

C5, C6, C7

Pectoralis major

C5, C6, C7, C8

Supraspinatus

C5, C6

Infraspinatus

C5, C6

Latissimus dorsi

C6, C7, C8

Teres major

C5, C6, C7

Deltoid

C5, C6

Brachial plexus

Axillary

Musculocutaneous Biceps

Radial

Posterior Interosseous

Strength, Power and Movement

It is important to distinguish strength (the absolute force production; i.e. the ability to overcome resistance) and power (the ability to overcome resistance in the shortest period of time). Power is therefore the rate of doing work and the product of force and velocity. The examination should be undertaken with the aim of assessing whether there is weakness and, if so, is it due to a central cause or involvement of the anterior horn cell, nerve roots, peripheral nerves, neuromuscular junction or muscle. The Medical Research Council grading of muscular weakness is shown in Table 2.1. The nerve and main root supply of individual muscles is shown in Tables 2.2 and 2.3 and the most commonly tested movements are summarised in Table 2.4.

Median

Anterior interosseous

Table 2.1  Six UK Medical Research Council grades of weaknessa.

Brachialis

C5, C6

Triceps

C6, C7, C8

Brachioradialis

C5, C6

Extensor carpi radialis longus

C5, C6

Supinator

C6, C7

Extensor carpi ulnaris

C7, C8

Extensive digitorum

C7, C8

Abductor pollicis longus

C7, C8

Extensor pollicis longus

C7, C8

Extensor policies brevis

C7, C8

Extensor indicis

C7, C8

Pronator teres

C6, C7

Flexor carpi radialis

C6, C7

Flexor digitorum superficialis

C7, C8, T1

Abductor pollicis brevis

C8, T1

Flexor policies brevis

C8, T1

Opponents pollicis

C8, T1

Lumbricals I and II

C8, T1

Pronator quadratus

C7, C8

Flexor digitorum profundus I and II

C7, C8

Flexor pollicis longus

C7, C8

Flexor carpi ulnaris

C7, C8, T1

Grade

Definition

5

Normal power

4

Active movement against gravity and resistance

Flexor digitorum profundus III and IV

C7, C8

3

Active movement against gravity

Hypothenar muscles

C8, T1

2

Active movement with gravity eliminated

Adductor pollicis

C8, T1

1

Flicker of contraction

Flexor pollicis brevis

C8, T1

0

No visible muscle contraction

Palmar interossei

C8, T1

Dorsal interossei

C8, T1

Lumbricals III and IV

C8, T1

 These grades were designed to record changes in power during poliomyelitis. Although more applicable to lower motor neuron weakness, they remain widely used throughout neurology.

a

C5, C6

Ulnar

28

Neurology: A Queen Square Textbook

Table 2.3  Nerve and main root supply of muscles (lower limb).

The pattern of weakness may reflect central involvement causing pyramidal weakness or lower motor neuron reflecting involvement of one or multiple roots or peripheral nerves, neuromuscular ­junction or the muscles. The distinction between upper and lower motor neuron ­weakness is summarised in Table 2.5.

Nerve

Muscle

Spinal root

Femoral

Iliopsoas

L1, L2, L3

Rectus femoris

L2, L3, L4

Vastus lateralis

L2, L3, L4

Vastus intermedius

L2, L3, L4

Vastus medialis

L2, L3, L4

Adductor longus

L2, L3, L4

Adductor magnus

L2, L3, L4

Gluteus medius and minimus

L4, L5, S1

Tensor fasciae latae

L4, L5, S1

Inferior gluteal

Gluteus maximus

L5, S1, S2

Sciatic and tibial

Semitendinosis

L5, S1, S2

It is important to establish any history of breathlessness, sleep ­disturbance or orthopnoea. The patient should be assessed for excessive activity in the accessory muscles of respiration. When supine, the anterior abdominal wall should move outwards on quiet inspiration as the diaphragm descends. If there is ­diaphragm weakness, inspiration is associated with inward movement of the anterior abdominal wall (paradoxical movement). Assessment of respiratory muscle weakness in general and diaphragm function, in particular, should be complemented by measurement of the forced vital capacity in both the erect and supine posture.

Biceps

L5, S1, S2

Coordination

Semimembranosus

L5, S1, S2

Gastrocnemius and soleus

S1, S2

Tibialis posterior

L4, L5

Flexor digitorum longus

L5, S1, S2

Abductor hallucis

S1, S2

Abductor digiti minimi

S1, S2

Interossei

S1, S2

Tibialis anterior

L4, L5

Extensive digital longus

L5, S1

Extensor hallucis longus

L5, S1

Extensor digitorum brevis

L5, S1

Peroneus longus

L5, S1

Peroneus brevis

L5, S1

Obturator

Superior gluteal

Sciatic and common peroneal

It is important that the examiner is familiar with major muscle groups and their motor nerves and spinal motor roots. In an upper motor neuron lesion, there is typically loss of dexterity, tested by rapidly tapping thumb and index finger or touching the thumb to all the other fingers in succession or foot tapping. Although the patient may be able to hold the arms outstretched with the palms up, initially there may be a tendency for the hands and the forearm to pronate as the whole arm drifts down (pronator drift). If there is weakness on examination, it is important to ensure this is not due to pain, poor coordination or reduced effort and is  due to true muscle weakness. This can be difficult but it is very important. It is generally necessary to compare an effective muscle group on one side with the same muscle group on the unaffected side. The ability to ‘overcome’ a muscle group is not necessarily an indication that it is weak (e.g. elbow extension and intrinsic hand muscles can generally be overcome). Some muscle groups need to be very weak before examination can detect weakness (e.g. flexors of the hand). If there is weakness, the distribution is important and necessitates a more detailed examination.

Respiratory Muscles

It may be difficult to test coordination in severe weakness and, conversely, severe incoordination may make testing of strength problematic. Observe the patient maintaining posture with the upper limbs outstretched with the palm down and note the presence of tremor. Ask the patient to touch their nose and then reach out fully to touch a target, usually the examiners moving or static finger, with successive ‘to and fro’ movements while looking for intention tremor or ataxia. Ask the supine patient to lift their leg straight, carefully placing the heel on the opposite knee and running the heel up and down the shin. Test rapid alternating movements of the upper limbs by asking the patient to tap the back of one hand with first the palmar and then the dorsal aspect of the fingers of the other hand in quick succession. The dominant hand is usually a little better at this than the non-­dominant. Cerebellar disorders are associated with intention tremor, irregularity of movements in range and strength, incoordination (ataxia), dysdiadochokinesis (impaired rapid alternating movements), unsteady gait and an inability to sit or stand without support (truncal ataxia).

Reflexes

The deep tendon reflexes should be tested with a consistent technique, ensuring that the patient is comfortable and relaxed. Careful positioning of the patient is critical. The examiner should use a good swing of the tendon hammer allowing the head of the hammer to fall using gravity rather than an active striking movement and comparing the respective tendon reflexes on the two sides of the body. If the reflex is absent to an adequate stimulus, reinforce it by asking the patient to clench their fists or their teeth and then immediately retest. The spinal level and grading of reflexes are summarised in Tables 2.6 and 2.7. The most valuable reflexes for clinical diagnosis are the biceps, triceps, brachioradialis, patella and ankle jerks. They are present and symmetrical in most adults. The activity is assessed by the threshold, latency, speed and duration of the contraction and the range of movement. Reflexes may be hypoactive ranging from diminished to absent. Hyperactive reflexes are associated with decreased threshold and latency and increased speed and range of movement with prolonged muscle contraction and extension of the reflex zone with spread of the response. When reflexes are brisk there may be a contralateral response (e.g. crossed adductor reflex) or spread such that the brachioradialis tendon reflex causes finger flexion. Inverted reflexes are less common and represent paradoxical responses (e.g. finger flexion in response to the brachioradialis reflex). These occur when the local reflex arc is disrupted but there

Approach to the Patient with Neurological Disease  29

Table 2.4  Commonly tested movements. Movement

Root

Reflex

Nerve

Muscle

Axillary

Deltoid

Musculocutaneous

Biceps

Upper limb Shoulder abduction

C5

Elbow flexion (supinated)

C5, C6

+

Elbow flexion (pronated)

C6

+

Radial

Brachioradialis

Elbow extension

C7

+

Radial

Triceps

Radial wrist extension

C6

Radial

Extensor carpi radialis longus

Finger extension

C7

Posterior interosseous

Extensor digitorum communis

Finger flexion

C8

Anterior interosseous, median, ulnar

Flexor pollicis longus, flexor digitorum profundus, all other flexors

Finger abduction

T1

Ulnar, median

First dorsal interosseous, abductor pollicis brevis

+

Lower limb Hip flexion

L1,2

Hip adduction

L2,3

Femoral

Iliopsoas

Obturator

Adductors

Hip extension

L5, S1

Sciatic

Gluteus maximus

Knee flexion

S1

Sciatic

Hamstrings

+

Knee extension

L3,4

Femoral

Quadriceps

Ankle dorsiflexion

L4

+

Deep peroneal

Tibialis anterior

Ankle eversion

L5, S1

Superficial peroneal

Peronei

Ankle plantar flexion

S1, S2

Tibial

Gastrocnemius, soleus

Big toe extension

L5

Deep peroneal

Extensor hallucis longus

+

Table 2.5  Lower and upper motor neuron lesions.

Table 2.6  Tendon reflexes.

Feature

Lower motor neuron

Upper motor neuron

Symbol

Description

Inference

Focal muscle wasting

Visible

Absent

0

Absent with reinforcement

Usually pathological

Fasciculation

Visible

Absent

+/−

Fibrillation potentials

Recordable on electromyography

Absent

Present with reinforcement

Sometimes normal, but may be pathological

+

Present

Normal

Tone

Flaccid/normal

Increased/spastic type

++

Brisk

Normal

Pyramidal + reduced dexterity

+++

Very brisk

Pathological if tone is increased but can be a normal finding

Tendon reflexes Depressed/usually absent

Exaggerateda

CL

Clonus

Clonus

Absent

Present

> 3 beats of ankle clonus = pathological; 2 beats may be normal

Abdominal reflexes

Present

Absent

Plantar responses

Flexor (normal)

Extensor

Weakness pattern

Root, nerve or distal

Table 2.7  Spinal levels of tendon reflexes. Spinal level

Reflex

C5–6

Supinator

 Tendon reflexes can be temporarily depressed or absent following an acute upper motor neuron lesion.

C5–6

Biceps

C7

Triceps

is segmental central hyperexcitability. They occur most commonly with radicular compression due to cervical spondylotic myelopathy. The reflex responses should be compared between the two sides and across different sites in the limbs. The knee and ankle reflexes

C8

Finger jerks

L(3)–4

Knee

S1

Ankle

a

30

Neurology: A Queen Square Textbook

are the most important in the lower limbs. Because the responses may be difficult to elicit even in normal subjects, side to side comparison is crucial. It is important to be aware of the slow relaxing reflexes of hypothyroidism. • Plantar response – the patient should be comfortable on a couch and the ankle joint and foot should be relaxed. The examiner must explain what they are going to do. The reflex should be tested using a rigid object (e.g. an orange stick) passing along the lateral aspect of the sole of the foot and then medially across the metatarsal heads. It is often helpful to hold the metatarsophalangeal joint of the hallux. The normal response is plantar flexion of the toes on stimulation of the sole of the foot. An initial extensor (dorsiflexion) movement of the big toe with or without fanning of the other toes is abnormal. Sometimes the reflex is difficult to elicit consistently, especially in individuals who are ticklish and tend to pull their foot away. The Chaddock reflex is elicited by stimulation of the lateral aspect of the foot and not the sole, beginning under the lateral malleolus and drawing the stimulus from the heel forward to the little toe. A pyramidal tract lesion is suggested by extension of the big toe and fanning. The Oppenheim sign is elicited by dragging the hand or the knuckles heavily down the anterior medial surface of the tibia from the infrapatellar region to the ankle. In a patient with a pyramidal lesion both may result in an extensor response when the planter response itself is equivocal. • Abdominal reflexes can be important and are often neglected. The abdominal response is divided into the upper and lower abdominal reflexes, in which brisk stroking of the abdominal wall with an orange stick in each of the four quadrants pulls the umbilicus (which lies at the T10 level) in the direction of the stimulus. The response is mediated in the upper quadrant by T8–T9 and in the lower quadrant by T11–T12. Abdominal operations, obesity and pregnancy may sometimes make the reflexes difficult to obtain. The abdominal reflexes may be absent in an upper motor neuron lesion. • Other reflexes – the pout reflex is elicited by placing a wooden spatula gently on the lips and tapping it with a tendon hammer and observing a pouting movement of the lips. The grasp reflex is apparent on drawing the fingers slowly across the palm of the patient, resulting in involuntary grasping. These reflexes emerge in diffuse cerebral hemisphere disease, particularly involving the  frontal lobes. The palmomental reflex is an involuntary ­contraction of the mentalis muscle of the chin caused by stimulation of the thenar eminence associated with pyramidal tract involvement. Sensory System The sensory examination should be undertaken on an anatomical basis guided by the major root dermatomes and peripheral nerve distribution. These vary quite significantly between individuals, and the area of supply shown in most texts is larger than the area of actual sensory loss if a single sensory root or nerve is damaged due to overlap between adjacent sensory fields. It is also important to consider what pathway is being tested with each modality: • Dorsal column: ◦◦ Joint position sense, two-­point discrimination • Spinothalamic: ◦◦ pain, temperature sensation • Combination: ◦◦ Tactile, light touch ◦◦ Vibration (128-­Hz tuning fork). A thorough sensory examination should be undertaken when a patient complains of relevant sensory symptoms; for instance, numbness, tingling or inability to tell the temperature of water. However, the patient may not identify some types of sensory loss as

such. Detailed sensory testing is appropriate in additional circumstances, such as unsteadiness, falls (loss of proprioception), presence of ulcers on the feet and pressure areas (loss of pain sensation) and the presence of unexplained limb pain. In some disorders (e.g. diabetes mellitus), it may be relevant to screen for sensory loss as an indicator of sensory neuropathy. If there are no relevant symptoms or circumstances as outlined above, it may not be necessary to spend much time formally testing all the sensory modalities, and a check of joint position and vibration sense will be adequate. If the patient does complain of sensory symptoms, the clinician should attempt to map out the abnormal area first and then attempt to compare the sharpness of a pinprick in the numb area with an area the patient indicates as ‘normal’, going from the abnormal area to the normal.

Pinprick and Temperature

Pinprick and temperature appreciation should be tested using a new disposable pin. Ask the patient to compare a pinprick undertaken proximally on the limb with one performed on the distal phalanx of the finger or toe. If they are similar and perceived as sharp, it is unlikely there is major distal sensory loss. If numbness to pinprick appreciation is detected in the feet, document the upper/ proximal level in each leg (i.e. when blunt becomes sharp). If there appears to be a loss of feeling on the abdomen or thorax, this should also be documented on the back. Sensation can be crudely tested by asking the patient to compare the temperature of an object; for instance, a cold tuning fork on an affected or unaffected area. More detailed testing of temperature sensation requires tubes of hot and cold water. A sensory level to pinprick and temperature appreciation most often implies the presence of a spinal cord lesion at or above the documented sensory level on the trunk.

Joint Position Sense

It is important to explain exactly what is required to the patient. Large movements should be used at first to demonstrate that the examiner will move the joint up or down, and ask the patient to describe the movement relative to the starting position (not the absolute position). The distal phalanx of the finger or big toe should be tested, and the joint held from the side to prevent any push–pull stimulus on the joint. The amplitude of the movement should be gradually reduced, ensuring plenty of time for each response. A cooperative alert patient should be able to detect rapidly and accurately the smallest m ­ ovements that the examiner can make. If the patient cannot discern movements distally, the examiner should move to a more proximal joint to quantify joint position sense appreciation.

Light Touch

Screen with a light finger touch and then proceed to a more detailed assessment with touches of cotton wool if there is subjective sensory disturbance to map out areas of that sensory loss.

Vibration Sense

Vibration sense should be tested using a 128-­Hz tuning fork activated by tapping the tines on the forearm. The base of the fork should be applied to a bony prominence on the toe or finger distally, having first established that the patient is familiar with the sensation tested more proximally, for instance on the sternum. In the elderly, vibration sense is commonly reduced in the toes.

Two-­Point Discrimination

The ability to discriminate two points placed closely together is impaired in posterior column disorders, leminiscal or cerebral

Approach to the Patient with Neurological Disease  31

lesions. This is tested by using two orange sticks or points and touching the patient (eyes closed) with one or two of the points and asking how many are felt. On the pulp of the fingers (assuming normal skin), points 2–3 mm apart can normally be resolved, whereas on the back of the hand the distance may be 1–2 cm. If there is sensory loss distally in all four limbs, this is more likely to be due to peripheral nerve disease, although a cervical cord lesion is still possible, and the localisation must be resolved on other criteria.

Dissociated Sensory Loss

Lesions within the spinal cord sometimes result in loss of sensitivity to temperature with preservation of light touch and position sense (dissociated sensory loss) in the same limb. and these patterns are discussed in Chapter 14.

Sensory Inattention

Disturbances of higher sensory pathways, for instance in the parietal lobe, sometimes give rise to sensory abnormalities even when the modalities mentioned above appear intact or are only mildly affected. There may be loss of the ability to detect simultaneous stimuli (e.g. touch) on both sides of the body even though a touch on each side alone can be reliably detected (sensory inattention). This is tested by asking the patient (with eyes closed) to tell you which hand you touch, and then touching left, right and both in a random order.

Astereognosis

The ability to discriminate different shapes and textures and to judge weight is often disordered in parietal lobe disease.

Neglect

A patient with neglected body space may simply ignore the relevant (usually left) side of their body, tend to look to the right and may not even recognise their left hand if it is placed in front of them. This may be exacerbated by a concomitant left homonymous hemianopia.

Autonomic Function

Examination of the autonomic nervous system is discussed in Chapter  32. The following responses are most helpful in routine clinical neurological practice: • Cardiovascular reflexes (faints, dizzy spells, blackouts): ◦◦ Orthostatic hypotension  – measure blood pressure erect and supine. ◦◦ Cardiac rate and blood pressure responses to Valsalva manoeuvre and carotid sinus massage undertaken under careful control circumstances. • Pupillary responses: ◦◦ Light and near responses. ◦◦ Pharmacological tests to evaluate the pupil. • Bladder and bowel: ◦◦ Continence of urine and faeces. ◦◦ Upper motor neuron lesion (spinal cord (including conus) or intracranial) associated with loss of the normal inhibitory input to the reflex emptying mechanism resulting in urgency and urge incontinence. ◦◦ LMN, cauda equina and lumbosacral spine associated with interruption of the reflex arcs that cause impaired bladder or rectal emptying and thus, retention of urine or constipation. ◦◦ If the sensory pathways are injured, there may be painless retention of urine. Palpate for an enlarged bladder. ◦◦ Assess anal sphincter reflex. • Sexual function: ◦◦ Erectile dysfunction with or without impotence.

Systemic Examination

Systemic examination is often neglected during the neurological assessment but it remains of critical importance. Examination of the skin is highly relevant to neurological disease. The examiner should look for cutaneous lesions that might indicate the purpura of meningococcal meningitis, the blisters in a radicular ­distribution to suggest herpes zoster or the characteristic lesions of neurofibromatosis, tuberous sclerosis or Fabry’s disease. Examination of liver and spleen may reveal organomegaly in storage diseases. These issues are discussed in further detail in Chapter 34. A scheme for undertaking a neurological history and examination in shown in Box 2.1.

Remote Neurological Consultation

Remote neurological consultation has been undertaken by telephone and video using internet connection for many years. Video consulting is particularly suited to those areas where the geography means that patients have to travel long distances, but the techniques were rapidly adapted to provide a mechanism for neurology care during the COVID-­19 pandemic. Switching from ‘face to face’ new appointments to remote (telephone or video) consultations allowed continuing provision of neurological services to patients while reducing hospital footfall, thus limiting the spread of COVID-­19 and avoiding exposing vulnerable patients to unnecessary risk. The success of remote outpatient consultations has helped to embed the technique as standard practice which remains suited for some patients following the waning of the pandemic. The initial consultation resembles a conventional face to face consultation but the examination is undertaken differently. For the neurologist, it is important to look into the camera and speak clearly and it may be necessary to share the clinician’s computer screen to review scans, other results or relevant websites. Neurological examination is easier to undertake if there is a far-­end assistant. It  is possible to assess pupillary responses and eye movements as well as lower cranial nerve function. Bradykinesia, tone and power can be tested, together with the reflexes and planter responses. Similarly, the assessment of sensation and coordination is possible with clear instruction. It is possible to assess higher mental function, speech and motor function by observing the patient standing and assessing their gait, including walking on their toes and heels, tandem walk and the Romberg test. A limited examination of the eye, facial and tongue movements is generally possible and finger–nose testing with fast repetitive hand movements and piano playing can be tested. Sensory examination is generally subjective and limited. It is often helpful for the patient to have an advocate or a member of the family present at the consultation and it is important to treat the consultation as a standard clinical encounter with appropriate documentation, follow-­up and safety netting. Some patients will still need to be examined in a face to face clinic after a remote consultation. Considerable expertise in these techniques has been established during the pandemic, in-­ person consultations will always be the gold standard but video consultation is a useful alternative that is generally easy to use where travel time is long, costs are high and the patient might be exposed to risk. However, many patients do struggle to use internet-­based platforms and the quality of the picture will depend upon the patient’s internet connection and device, meaning that the neurological examination will be more limited. Remote consultations are most suitable for follow-­up appointments, particularly where a neurological examination is not

32

Neurology: A Queen Square Textbook

Box 2.1  Summary of neurological history and examination Historical details • Main complaints • History of present illness • Past illnesses • Review of previous opinions, practice notes, etc. • Family history, with details of size of family • Social, personal, alcohol, drugs, tobacco, etc. • Travel, occupation, etc. • Review of systems: ◦◦ cardiovascular ◦◦ respiratory ◦◦ gastro-­intestinal ◦◦ genito-­urinary ◦◦ endocrine ◦◦ psychiatric history ◦◦ allergies – drug and other

Posture of upper limbs – describe • Tone • Power • MRC Scale 0–5 and/or describe wasting; fasciculation, muscle consistency • Assess limbs, neck, diaphragm, abdomen • Coordination (finger–nose, heel–shin, alternating movements, foot tapping, etc.) • Reflexes: ◦◦ jaw jerk ◦◦ biceps jerks ◦◦ supinator jerks ◦◦ triceps jerks ◦◦ finger jerks ◦◦ abdominals (upper and lower) ◦◦ knee jerks ◦◦ ankle jerks ◦◦ plantar responses ◦◦ other reflexes (state)

Examination • Physical appearance, initial appraisai • Mental state and cognition • Speech, language functions, evidence of higher function problems • Skull and spine • Gait, stance, balance • Hand preference – writing, etc.

Sensation • Assess posterior columns: vibration (128 Hz, VS), joint position (JPS), light touch (LT), 2-­point • Assess spinothalamic pathway: pain (PP), hot/cold (TM) • Chart findings: PP, TM, LT, etc.

Cranial nerves • I. Olfaction • II. Visual acuity ◦◦ Visual fields – describe/chart

120

L 90

60 120

R 90 60

150

30

180

0

210

PP

330 240

270

300 240

TM

270

◦◦ Fundus ◦◦ Pupils (mm, shape, reactions) • III, IV, VI. Range of ocular movements, nystagmus • Inspection of each eye, proptosis, cornea, primary position of gaze • V. Facial sensation, corneal reflexes. Power: masseter, temporalis. Jaw jerk • VII. Facial symmetry, weakness, abnormal movements, taste • VIII. Hearing, Rinné, Weber, vertigo, nystgamus • IX, X. Swallowing, phonation, gag reflex, pharyngeal sensation • XI. Sternomastoids, trapezii • XII. Tongue appearance, speed, central protrusion, fibrillation Motor system • Abnormal movements – describe

LT

300

General physical examination • Cardiovascular • Blood pressure • Respiratory • Abdomen • Endocrine • Skin • Nodes • Joints • Summary of findings • Formulation • Provisional diagnosis

Approach to the Patient with Neurological Disease  33

required or where there are visible physical signs of interest. It has clear advantages over telephone consultations and the technique can be effective, easy to use and acceptable to both doctors and patients. Video consultation has the potential for long-­ lasting changes to how we practice medicine and can improve the patient experience in the management of chronic neurological conditions.

Assessment of the Patient with Hearing Deficits

Deaf patients experience serious communication difficulties when accessing healthcare. For many patients in the UK, British Sign Language (BSL) is their first language, yet few healthcare professionals receive training in deaf awareness or in BSL. BSL is a visual language that combines hand movements, facial expressions and body language to convey meaning. It differs from that used in most other countries, each of which has developed their own sign language. Understanding and communicating with a deaf person also requires appreciation of differences in deaf culture, access to education and lived experience as a deaf person. Prelingual deafness refers to those who have become deaf before acquiring spoken language and may rely entirely on sign language to communicate. General principles for communicating with deaf patients include the requirement for careful preparation before the clinic, close attention to the environment and room setup, appropriate use of  an  interpreter and specific care in optimising the history and examination. A welcoming and accessible environment should be created and administrative staff require training in deaf awareness. The waiting room should have clear visual displays to alert patients when the doctor is ready to see them. It is important for the doctor to meet the patient in the waiting room and introduce themselves. The consulting room needs to have sufficient space, be well lit and to be quiet, without background noise. Interpreters should be registered and familiar with the neurology setting to ensure accurate translation and understanding. The interpreter is placed close to and slightly behind the doctor so the patient and the relative can see both in their same line of sight. This enables them to understand what is being said without having to turn away and therefore allows continuous observation of the speaker’s lips and facial expression. The doctor should always show their name badge, confirm the preferred method of communication at the outset and inform the patient that they can be asked to slow down or pause if they become fatigued. The doctor should gain the patient’s attention before speaking, face the person while speaking and maintain consistent eye contact without looking down at notes or at a computer screen. They should speak clearly, a little more slowly than usual and use short and simple sentences maintaining awareness that the interpretation will not necessarily be temporarily synchronised to the doctor or indeed with the patient’s facial expressions. Questions should be open ended and any cues are useful, including mime, gesture and body language. For patients who can lip read, it is important to look towards the patient, avoid covering the mouth and speak at a normal pace and volume (speaking loudly and overemphasising words can distort the lip pattern, making it difficult to lip read). It is essential to talk to the deaf person rather than to their hearing relatives and to ensure that the deaf person remains fully included in the conversation. The doctor must remain patient and ensure that the questions have been understood. It is often helpful to rephrase questions rather than simply repeat them, and it is important to indicate when there is a change to a new topic. Writing questions

and answers can be a useful adjunct if spoken communication is challenging. Family members should not be asked to interpret, unless this is unavoidable. They are often less well trained in BSL and are unlikely to be impartial. Furthermore, patients may not wish to disclose sensitive information when family members are interpreting. Finally, it can be very traumatic to interpret bad news to a loved one. Using pen and paper is not an adequate method of communication for patients who use BSL; they often have difficulty in interpreting written instructions and may have difficulty reading as a consequence of the educational difficulties they have experienced. It is necessary to remain patient and allow longer than is conventional for the consultation. It is important to establish the level of education and likely comprehension. Deaf patients may provide a response that is tangential and not relevant rather than asking for  clarification, particularly if there are associated language ­difficulties or delay or cognitive impairments such as aphasia. It is important to signpost your next action to prevent any unexpected moves that may provoke anxiety by using phrases such as ‘I am going to ask about your symptoms first and then I will do a physical examination’. When conducting a physical examination, it is necessary to ensure that both the interpreter and the patient understand what is  being tested with short, clear instructions and demonstrating the  movements as necessary. On concluding, the doctor must be confident that the patient has had the opportunity to explain themselves and understands the diagnostic, management and follow-­up plan. The interpreter should be acknowledged and thanked.

Consultations with Patients for Whom English is not Their First Language

The absence of a common language necessarily compromises the good communication essential in the neurological consultation. Migration and globalisation have increased the number of patients for whom English is not their first language and up to 60% of consultations in London are with non-­native English speakers. A wide range of languages are spoken throughout the UK and the extent to which English is spoken and understood varies from those who have none to those who have reasonable or good English. It might be argued that difficulty with communication impinges more on the neurological consultation than any other aspects of medicine and presents challenges that require patience and innovative solutions. Difficulty with communication establishes a barrier between the patient and doctor and is detrimental to the patient’s perception of the consultation. It may lead to misunderstanding, poor rapport and a lack of trust which, in turn, affects the quality of care and the patient’s engagement with the diagnostic and therapeutic process. Interpreters are able to translate directly from the patient’s native tongue and can provide helpful observation concerning the language, mental state and behaviour. However, translation via an interpreter risks missing subtleties of language and interaction. The quality of translation is variable and may be simultaneous, line by line or a summary. The last is least accurate yet is most commonly used and risks the inadvertent admission or omission of important information, as well as subjective interpretation by the translator. It is therefore important to keep sentences short and simple, pausing often to allow the interpreter to translate simultaneously or by the sentence. The doctor should avoid jargon, idioms, acronyms or jokes, which cause confusion. In addition to the pure language barriers, cultural factors may also affect the process

34

Neurology: A Queen Square Textbook

and outcome of the consultation. Patients may be intimidated by the presence of a doctor and be reluctant to speak freely and express their concerns. The importance of gesture and the way in which patients are approached may also be misinterpreted. In some cultures, interpreters may modify the information to minimise the effects of a poor prognosis and there may be issues related to gender, disclosure of personal information and familial concerns. For these reasons, it is important to understand and to take account of individual cultural mores. If the physician is suspicious that the information they are providing is not being appropriately communicated, it is important to  reiterate or reinforce the messages. This may be a particular problem when translation is being provided by friends or family members. It is often helpful to ask the patient to state their own understanding to the interpreter and then to hear this relayed back and it is necessary to encourage the interpreter or family members to intervene if there is any suspicion of misunderstanding. Printed information helps to ensure reliable communication and translated information is increasingly popular for all forms of healthcare communication. Telephone interpretation services provide particular difficulties and consultations are generally performed only at the most basic level with limited discussion and interaction. If an independent interpreter cannot be obtained, then a relative, a family friend or member of staff may be the only option that allows the  consultation to go ahead. The physician must recognise the limitations of translation and allocate appropriate time and effort to the consultation, ensuring they maintain engagement with the linguistic and cultural restrictions experienced by the patient.

Neurological Consultation with an Autistic Patient

Autism is defined by a persistent pattern of social communication difficulties and differences, as well as the presence of restricted and repetitive behaviours. The diagnostic criteria for autism spectrum disorder can be fulfilled by many different presentations and symptom combinations. An autistic person’s cognitive abilities do not necessarily correlate with their ability to function on a day-­to-­day basis. The National Autistic Society (2023) provides helpful guidance that is applicable to communication in all settings. In general, to help an autistic person filter out information that is not directly relevant, it is important to speak slowly and say less, to avoid open-­ ended questions and not ask more than one question at once. Autistic people often need longer than neurotypical people to process information, so it is important to allow enough time for the consultation and longer clinic appointments are generally appropriate. It is necessary to be specific in the choice of words and to avoid idioms, figures of speech, irony and exaggeration. Autistic people can struggle with interpretation of nonverbal communication, such as facial expression and body language. Autistic patients may not respond to questioning in a conversational style and they may show reduced or unusual facial expression and body language and often avoid eye contact as this can feel uncomfortable. Autistic people may say things that can be interpreted as inappropriate or unusually blunt because of difficulties understanding social context. This does not mean that they are being intentionally rude or impolite and offence should not be taken. Prior to the appointment, it is important to ensure that a careful explanation is provided and to remember that patients may be unsettled by any delays or variation from the predicted routine running of the appointment. An autism-­friendly environment avoids

overcrowding, noise and harsh lighting. Many autistic people may prefer remote consultation, although telephone and video appointments can exacerbate communication challenges. During the history taking, it is essential to be particularly clear and specific in questioning and to ensure that the person understands what has been said. During the examination, it is necessary to clearly explain what is going to be undertaken and why. Routine examination procedures, such as testing the tendon reflexes and examining sensation, might cause excessive distress without careful explanation. After the consultation, it is important to ask the patient if they would like you to share information about necessary adjustments and considerations. The letter should be clear, specific and unambiguous. It is essential to increase autism awareness for all clinical, nursing and clerical staff because small changes throughout the process can reduce inequalities in care and remove barriers to seeking care.

Neurological Consultation with a Person Who has a Learning Disability

Learning disability is a broad term that includes a spectrum of people, from those who are highly verbal and able to live with little support and no associated difficulties to those with multiple profound physical impairment and limited communication skills. Many conditions causing learning disability are associated with neurological manifestations or comorbidities. Prior to assessment, it is important that the patient is provided with a brief and intelligible account of the aims of the assessment to help reduce anxiety and agitation. It is essential to ensure that there is a clear request to the family or the care home manager to provide a well-­informed family member or carer to accompany the patient and bring details that are necessary (e.g. medication and observation charts and seizure diaries). In the clinic, it is important to recognise that the patient may be feeling highly anxious, which can result in impairment of existing communication skills. An informed, familiar family member or carer can be crucial, both in supporting communication in the consultation and for providing corroborative information. Many patients with learning difficulties have difficulty in recalling and sequencing major events and relating them to their present condition. The clinic room may contain multiple distractions and difficulties for a person with learning difficulties. These may include interruptions, sunlight through a window, noise from adjoining rooms or building work and chemical odours. Any disturbance to the consultation will increase anxiety and agitation and may make it impossible for the patient to understand the diagnosis or management plan. The examination must be clearly explained with reassurance. Communication should involve a range of techniques adapted to the individual with learning disabilities, which may involve eye contact, eye pointing, facial expression, touch, body language, gestures, signing, use of objects, photographs, symbols as well as through speech and text. It must be recognised that people with a learning disability and suffering from physical health problems often present in a variety of ways, including self-­injurious behaviour and physical aggression which may represent significant ­challenges to investigation and management. It is important that information is sent to the patient and their family or carers following the consultation with clear written reminders of specific actions to be taken. It is also important to ensure that follow-­up of the recommendations is undertaken by  a  learning disability team, either from the hospital or the community.

Approach to the Patient with Neurological Disease  35

Neurological Assessment in an Emergency Setting, Including Hyperacute Stroke

The approach of a neurologist in an emergency setting, in particular a hyperacute stroke assessment, relies on similar principles to those described above, but needs to be modified in line with the urgent time pressure to make an accurate diagnosis to facilitate hyperacute treatment (e.g. revascularisation with thrombolysis or thrombectomy for acute ischaemic stroke). When patients have reduced language function, confusion or reduced conscious level, the history often comes from paramedics or others at the scene of onset; every opportunity should be taken to speak with these and other witnesses in parallel with obtaining rapid vital signs and neurological assessment. The rapidity of onset (immediate, seconds, minutes, hours) and when the person was definitely last known to be well must be established as soon as possible. Any preceding symptoms of relevant illnesses, including infection or trauma, should be determined from all possible sources, including electronic health records. Any clues to risk factors, such as recreational drug use or craniospinal trauma should be sought. The neurological assessment for suspected stroke includes careful but rapid observation for signs of neglect, agitation and paresis. For example, the pattern of a large middle cerebral artery stroke syndrome can be rapidly confirmed within a few seconds, and supplemented by a structured examination, usually involving the National Institutes of Health Stroke Scale. Any possible contraindications to hyperacute treatments such as thrombolysis should be established as soon as possible, often by other team members including specialist nurses. The whole process is usually undertaken while the patient is en route to the emergency computed tomography scanner to seek evidence of large-­vessel occlusion or intracerebral haemorrhage. Once hyperacute treatment has been given, for example on the post-­take ward round, a more detailed history and examination can be undertaken along the lines described in the earlier parts of this chapter.

Conclusion

The history and examination are the foundations of clinical medicine in general and the assessment of the neurological patient in particular. It is only with experience that neurologists develop a sense of what is likely to be wrong, where opportunities for treatment lie and how diagnosis and management will impact the patient as an individual. The seven deadly sins of medicine enumerated by Richard Asher in 1948 – obscurity, cruelty, bad manners, overspecialisation, love of the rare, common stupidity and sloth – remain a pernicious threat in contemporary neurological practice, and are still to be avoided at all costs. The skills of good neurological practice might be compared to the consulting detective, as exemplified by Sherlock Holmes. As Andrew Lees (2015) has described, these shared skills include: • Unhurried observation, guided by a prepared mind • Attention to detail

• Detailed knowledge  – of the discipline and the circumstances surrounding the malady • A sound grasp of probabilities • Absence of prejudice • Regard for negatives (thereby avoiding ‘the crime of Procrustes’ – forcing an interpretation based on false assumptions) • Energetic pursuit of the culprit • Learning from mistakes, so that errors (though inevitable) are not repeated. There are, however, two final attributes in which the complete neurologist must part company with the detective: tolerance of uncertainty and humility. The complexity of the human nervous system defies categorisation and confounds even the most ­experienced of practitioners. As that other Holmes has warned us: ‘A search for absolute and clear-­cut syndromes is foreign to scientific investigation of such biological problems as the [neurologist] meets’ (Holmes 1952).

Acknowledgements

The authors are indebted to the inspirational work of Louis R. Caplan, Joshua Hollander, William W. Campbell, Richard J. Barohn, Michael Aminoff, Allan Ropper, Eelco Widjicks, Anita Harding, Mark C. Wiles, Andrew J. Lees, Geraint Fuller, David Nicholl, Martin Rossor and the many other neurological giants from whom they have borrowed heavily and upon whose shoulders they have stood in preparing this chapter.

References

Campbell WW, Barohn RJ. (2020). DeJong’s The Neurologic Examination, 8th edn. Philadelphia, PA: Wolters Kluwer. Caplan LR, Hollander J. (2011). The Effective Clinical Neurologist, 3rd edn, Shelton, CT: People’s Medical Publishing House. Holmes G. (1952). Introduction to Clinical Neurology, 2nd edn. Edinburgh: E&S Livingstone. Johnson JCS, McWhirter L, Hardy CJD et al. (2021). Suspecting dementia: canaries, chameleons and zebras. Pract Neurol 21(4): 300–312. Lees AJ. (2015). The strange case of Dr William Gowers and Mr Sherlock Homes. Brain 138: 2103–2108.

Further Reading

Cooper M Gale K, Langley K et al. (2022). Neurological consultation with an autistic patient. Pract Neurol 22: 120–125. Duncan C, McLeod AD. (2020). Video consultations in ordinary and extraordinary times. Pract Neurol 20: 396–403. Fuller G. (2019). Neurological Examination Made Easy, 6th edn. Edinburgh: Elsevier. Harris MJ, Atkinson JR, Judd K et al. (2020). Assessing deaf patients in the neurology clinic. Pract Neurol 20: 132–138. Leschziner G, Christophi G, Howard R. (2012). Neurology. In: Bessant, R., ed. The Pocketbook for PACES. Oxford: Oxford University Press. National Autistic Society. (2023). Communication tips. https://www.autism.org.uk/ advice-­and-­guidance/topics/communication/tips (accessed 20 July 2023). Pappworth M. (1971). A Primer of Medicine, 3rd edn. London: Butterworth. Turner B, Madi H. (2018). Consultations with patients for whom English is not their first language. Pract Neurol 19: 536–540. Wiles CM. (2013). Introducing neurological examination for medical undergraduates: how I do it. Pract Neurol 13(1): 49–50.

CHAPTER 3

Decision Making, Ethics and Law in Neurology Jonathan Martin1 and Alex Ruck Keene2  National Hospital for Neurology and Neurosurgery  39 Essex Chambers and Visiting Professor, King’s College London

1 2

Introduction

We use judgement, a process of ‘informally integrating diverse pieces of information into an overall assessment’ (Kahneman et  al.  2021), to inform decision making.1 Our judgements, ­however, are informed by more than simply the clinical facts of a particular decision, especially if the decision is complex. For example, we also bring a moral component (that is, an evaluation of a course of action as being good or bad) to much decision making, even if we  are unaware of doing so, and in this sense clinical decisions are ­usually also moral ones. And, of course, we must incorporate the requirements of the law into decisions. But beyond these we also bring much to decision making that is more to do with us than it is to do with the patient. These latter, extraneous, factors are  important because they increase the risk of unfair decisions, particularly where the patient is from a different cultural background to the decision maker. Extraneous factors include: our assumptions (usually unexamined) about what constitutes acceptable values, such as what makes a life worth living (and whose values should hold sway in decision making); the ­systematic errors we bring to decisions due to our own biases; and the non-­ systematic, sometimes context-­ sensitive, errors ­created by ‘noise’ in the system, such as the ‘HALT’ factors in which our decisions may be influenced by our being hungry, angry, late or tired. How, then, can we optimally integrate the multiplicity of clinical, ethical and legal elements of judgement into a systematic approach to decision making in adults, while mitigating, as far as reasonable, the extraneous factors that we bring as individuals? The intention of this chapter is to offer a practical guide to clinical decision making while directing your attention to some of the less relevant factors that are probably influencing your decisions beneath the level of your awareness. The chapter has two main sections: Part I is about using ethics and law in a systematic way when making a clinical decision; Part II is about the extraneous factors we probably bring into all our decisions and what we could do about them.

1  This textbook is aimed at clinicians, but this chapter is written by both a clinician (JM) and a lawyer (ARK). ‘We’ will therefore sometimes be in reference to the two authors and sometimes to clinicians as a group. We (the authors) believe that this should be obvious from the context but each note that any confusion arising is the fault of the other one.

Fundamentals We begin with an apparently simple question: Why make correct clinical decisions? Doubtless there are many reasons why doing so is a good idea, for example, we are likely to be positively motivated by a desire to do the best for our patient and, not unreasonably, believe that correct clinical decisions support this goal more than incorrect ones would. Or perhaps our motivation may be because not to do so would be to deprive the patient of some fundamental rights. These rights arise, in part, from the fact that the patient is someone to whom we owe a duty of care, recognising, in the inequitable power relationship that is characteristic of patient-­clinician interactions, that the patient is the more vulnerable of the parties involved. But these answers, though accurate, are inadequate. This is because, particularly in the context of ethical or legal complexity, they merely give rise to other important questions, the answers to which are fundamental to wise decisions. For example, who decides what ‘correct’ means when the rightness of a decision depends on factors that lie beyond questions of pathophysiology and therapeutics? Or, how much effort is it reasonable to put into the process of complex decision making? Or, how can we recognise and, to the extent possible, mitigate those factors that impair our ability to make correct decisions? The difficulty is that these subsequent questions, though key, usually remain unexamined. Our suggestion is that the effort spent digging beneath the surface of our assumptions about how complex decision making should operate is rewarded many times over in the quality of subsequent decision making. We hope to persuade you of this contention in this chapter. The Scale of the Challenge of Decision Making In Part I, we outline an approach designed to facilitate a systematic and rational means of decision making, incorporating both relevant law2 and a well-­known framework of ethics. This approach is neither necessary nor sufficient for all decisions and we are certainly

2  We should note at this point that, while one of us (JM) holds a law degree as well as a medical degree and the other (ARK) is a practising lawyer and an honorary KC, the contents of this chapter should not be construed as legal advice. Please also note that, since both authors practise in England, we will almost always refer to the law as it applies to England and Wales.

Neurology: A Queen Square Textbook, Third Edition. Edited by Robin Howard, Dimitri Kullmann, David Werring and Michael Zandi. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd. 37

38

Neurology: A Queen Square Textbook

not claiming that decisions must be made in this way, although we do contend that this approach is useful for very many complex decisions. The converse of a systematic approach to decision making is an unsystematic one; arguably this is frequently the status quo, inasmuch as it is rare for us to be aware of the structures that underpin our decision-­making processes, let  alone to make them explicit. The problem with this latter approach is that it makes us vulnerable to two types of error. The first is that we may ignore or undervalue factors that should be taken into consideration as part of the decision, particularly if it is inconvenient to do so; the second is that we tend to be unprotected from the influence of factors not relevant to the decision. A systematic approach does not negate these errors entirely, but it should make them less influential and thereby reduce the chance of arbitrariness. This is important because many of the decisions that we are asked to take are ones where it could be said that there is more than one ‘correct’ outcome, leading to the challenge that any given outcome is simply arbitrary. Being systematic may not mean that we are ‘objectively’ right in the outcome of the decision, but we can say that we sought to follow a process which objectively seeks to make the right decision (i.e. to reduce the influence of subjectivity). A systematic approach also helps to clarify the particular issues arising within any given decision where disagreement may occur. Perhaps the biggest downside to being systematic is that we may be forced to arrive at decisions we do not want, so the chapter comes with a cautionary note: while, as clinicians, we may believe ourselves to make clinically determined, facts-­based and values-­neutral decisions, in practice, most of our decisions are likely to be very different to this apparent ideal. Indeed, data suggest that most of our moral decisions are made before we are aware of thinking about the question; that is, we usually make moral decisions based on the emotions generated within us by the situation (Haidt 2001); thinking is only subsequently engaged, if at all. When it is, we are very adept at layering rationalisations on top of the emotionally derived decision, to justify it in a way we find acceptable (perhaps because the situation has generated a degree of cognitive dissonance within us) or that we think will be acceptable to others. This process is called ‘motivated reasoning’. We do this frequently and are usually unaware that we are doing it, although we may become aware if we recognise that we have given a particular reason because it seems to be more acceptable than the one we really believe, or because we are unable to articulate the true reason, even to ourselves. For example, a person may say that they do not wish to have a particular medical procedure because it will cause them pain but may then continue to hold the position of not wanting the procedure even when reassured that there will be no pain. A systematic approach that attempts to adopt a critical thinking methodology is a partial mitigation for the problem of our own post-­hoc motivated rationalisations. Perhaps even more important than the acknowledgement that we are typically creatures who make decisions based on habit and emotion, rather than reason, is the recognition that the heuristic shortcuts we rely on may themselves be the products of prejudice. Recent findings suggest that many of us, perhaps all, are subject to implicit biases: systematic ways of thinking that are based in the very beliefs we say (and consciously believe) that we do not hold. Another barrier to taking systematic, values-­ neutral clinical decisions is our adherence to a set of moral values (largely unexamined), which, while tending to be those of the social majority, may not be shared by the patient. What makes your moral values and ethical standards the right ones for another person? Yet another is the ever-­shifting sands of the law. It is very clear that clinicians are highly variable in their ability or willingness to

apply the Mental Capacity Act (MCA)3 and few have the time to be able to take into account the many ‘wrinkles’ in decision making introduced by the senior courts through judge-­made case law. Partly this may reflect the lack of a good enough understanding of the law; partly it may be based on a disagreement with the judgments themselves. Judges, after all, are unlikely to understand a clinical situation in the way a clinician can (any more than a clinician understands the law as a judge does). Of course, none of the factors set out in the preceding sentence are a reason for non-­ compliance with the law, but it would be naïve not to acknowledge their existence. It is important to recognise each of these barriers because not to do so is to acquiesce to a situation in which we allow decision making to continue in a suboptimal manner. Consent Consent lies at the very heart of healthcare decision making and needs to be properly understood. However, this chapter is not about the requirements of the General Medical Council (GMC) in this regard, and we will assume you are aware of, and conform to, the latest GMC guidance on consent and shared decision making (General Medical Council 2020; by doing so you will go a long way to being compliant with the law). Nor is it about various other guidance that may be relevant to particular decisions, such  as that relating to clinically assisted nutrition and hydra­ tion (CANH; Royal College of Physicians and British Medical Association 2018), or that relating to prolonged disorders of ­consciousness (Royal College of Physicians  2020). While such guidance is authoritative to a degree and may be referred to in court as representing an expected standard of medical practice, it does not have binding legal force. So, while useful, we assume in this chapter that you are familiar with those guidelines you rely on, and simply note here that their status from a legal perspective is ambiguous. It is worth noting that this is in sharp contrast to the status of legislation relevant to the provision of healthcare such as the MCA, the Mental Health Act, the Public Health (Control of Diseases) Act, or the status of authoritative judicial interpretations of the law (such as the determination of the meaning of ‘best interests’ by the UK Supreme Court in Aintree v James4). The status of both legislative (i.e. Parliament-­made) and judge-­made law is absolutely clear: this is the law and, whatever we may think of it, our duty as citizens and professionals leaves no scope for ignorance and nor are we entitled to disregard those aspects we find inconvenient.

Part I: Towards Systematic Moral Decision Making

We start this section by outlining the essential features of the approach to decision making that we are suggesting, using a clinical example to illustrate a number of key aspects. A Case Mrs C is a 62-­year-­old woman who is admitted to the emergency department following a haemorrhagic stroke. She has a past medical history of disseminated breast cancer, poorly controlled diabetes and hypertension and usually lives at home with her husband

3 4

 Mental Capacity Act 2005.  Aintree University Hospitals NHS Foundation Trust v James [2013] UKSC 67.

Decision Making, Ethics and Law in Neurology  39

and her 23-­year-­old daughter. She is admitted on a stroke pathway and investigations reveal that she has previously undiagnosed brain metastases that have bled. She makes an initial good recovery but then drops her level of consciousness when she has a further stroke, this time requiring an external ventricular drain to normalise her intracranial pressure. Her initial recovery post procedure is limited and, although the drain is later internalised as a shunt, she appears to remain in a low awareness state. She is being fed via a nasogastric tube. The process of formally assessing her prolonged disorder of consciousness begins and conversations with her family also start to consider potential outcomes. They are told that, while she may go on to make some level of recovery, it is unlikely that she will be able to interact meaningfully with her family and she is also unlikely to ever be well enough for oncological treatment of her brain metastases. As with most people of her age she has not written a lasting power of attorney (LPA) document, nor made any references to how she would like to be looked after in the event of catastrophic illness. The question arises as to whether the nasogastric tube should be removed and a percutaneous endoscopic gastrostomy (PEG) tube inserted, for the purpose of providing nutrition, hydration and medication, or whether, were she able to, Mrs C would refuse to consent to the provision of any form of CANH (considered in law to be a medical treatment5 and therefore an intervention that may be refused). How might decision making be approached in this case? The two questions identified above are clearly linked so the focus here is on the latter: whether CANH should be continued or withdrawn. Note that in law the question is always focused on whether the person consents or, if currently unable to do so, would or not (i.e. whether it is lawful to give a treatment or not), not on whether it is lawful to withdraw the treatment.6 If the conclusion is that there is no consent to provide the treatment then it must not be started or must be stopped. The other important piece of information at this stage is that, being in a low awareness state, Mrs C is found to lack the mental capacity for this (or any other) question at this time and is unlikely to regain capacity in the future. Our suggestion is that decision making in this case should be supported by an ethical analysis of the issues. This is based on an application of the four principles framework, as it is usually understood within medicine, although this is not the only ethical analysis that could be undertaken. In the next section, we outline how the ‘four principles’ framework fits into the wider realm of moral theory; by doing so, we hope  to demonstrate some of the ways we use ethical theory in ­everyday practice, even if we are not aware of it, and to begin to reveal how our unexamined views about what is right may not be shared by others. Ethical Systems There are a number of moral theories that underpin ethical systems that are relevant to the healthcare endeavour. The three main ­theories are known as ‘normative’, inasmuch as they are about the standards we use to justify our actions. They are: • Utilitarianism • Deontological ethics • Virtue ethics.

 Airedale NHS Trust v Bland [1993] AC 789 [804].  Aintree v James (n4) [20].

Utilitarianism

Utilitarianism is a type of consequentialism and was developed by Jeremy Bentham in the 19th century. He thought that an action was morally justified if it brought the greatest happiness to the greatest number of people. In medicine, this is most clearly seen in public health decisions at the population level, which seek to maximise, for the greatest number of people, the positive health consequences of a public health intervention and has the advantage of explicitly prioritising fairness. Public health also intersects with legislation at the individual level; for example, those known to have certain infectious diseases may be compelled, under some circumstances, to remain in hospital against their wishes;7 in such a case, the rights of the individual are considered secondary to those of the community who would otherwise be placed at unacceptable risk.

Deontological Ethics

Whereas utilitarianism focuses on the consequences of an action, deontology is a system of ethics based on duties, rules and obligations (deon is the Greek word for duty). Much as we have had certain rules of grammar instilled into us that we were not explicitly taught (e.g. that size comes before colour in descriptions of objects), so our various cultural backgrounds, at the levels of family, society and profession, have inculcated certain moral rules. These are usually unexamined and, inasmuch as they are thought about at all, are assumed to apply universally. For instance, most people within the dominant culture of this country would believe that it is wrong to use a person for the benefit of another; for example, that it would be wrong to kill one person in order to use that person’s organs to benefit, say, five others.

Virtue Ethics

The virtue ethics theory, developed by Aristotle, suggests that the right course of action is that taken by the virtuous person. This theory therefore focuses not on duties or outcomes but on the person making the decision. We do not explore it further here other than to say that, in developing a systematic approach to, and self-­ awareness in, decision making, the hope is that we will become more virtuous practitioners. Faith-­Based Ethics Faith-­based systems of ethics are considered to focus on the praiseor blameworthiness of actions and incorporate all three of the moral theories outlined above. In keeping with the normative nature of these theories, specific values derived from particular religions inform views that adherents tend to believe to be universally applicable. For example, Catholics are generally portrayed as being opposed to abortion not only for Catholics themselves, but for everyone. It is worth noting, though, such is the variation in beliefs and associated values, that assumptions about what another person believes based on particular aspects of culture are unsafe. The obvious point to note is that your patient’s views may be very different to your own. It is also the case that many of our most cherished values, whether faith-­based or not, are not open to debate, and attempts to dispute them rationally will usually fail. Indeed, sometimes cultural pressures are such that even to question certain values is forbidden; sometimes the views expressed by an individual may not be their own (e.g. if an eldest child is required by his culture to represent a particular view when discussing the best interests of his parent).

5 6

7

 Public Health (Control of Disease) Act 1984 s38(1).

40

Neurology: A Queen Square Textbook

Moral Universalism and Moral Relativism Clinicians are likely to have been socialised into the dominant ethical systems through which our culture tends to view the world. Being dominant (and therefore continuously reinforced by colleagues), clinicians are in danger of assuming that our values are, a priori, correct  – we assume that they apply universally. We may then tend to discount the views of those with other cultural heritages, perhaps considering them unsophisticated or parochial. This can be understood to be part of the process of ‘othering’ the patient, seeing them as ‘not us’, in which we portray ourselves as holders of the scientific high ground; we know better. The problem is that, because we are most comfortable when operating within our own frame of reference, this attitude may shade into intolerance of those who do not share our views (we have a great advantage in this, in that our laws are underpinned by the same dominant values). At the other end of the spectrum to moral universalism is moral relativism, the idea that there is no moral absolute, that what you believe is just as valid within your cultural heritage or social setting as what I believe is in mine, and that therefore everyone ought to respect the views of everyone else, even if we do not agree. Of course, there are limits to this idea; for instance, it is difficult to be respectful of a belief that is based on a factually incorrect idea (such as that autism is caused by the measles, mumps and rubella vaccination or that the COVID-­19 vaccine contains nanoparticles), while it is perhaps easier to be respectful of a belief based on an unprovable proposition (such as that there is a heaven, or that reincarnation repeatedly occurs until nirvana is reached). Despite these limitations, it is an attractive idea; it seems to be inherently respectful of others. But it has a fundamental problem, which is that moral relativism is internally incoherent in that the requirement that everyone ought to hold a relativist view is itself not a relativist view: the term ‘ought’ is absolute (i.e. only relativist views are acceptable) and, ironically therefore, this places it more in the universalist camp. What we can take from this, however, is that the ideas inherent in moral universalism and in moral relativism are in tension, and this highlights an interesting dilemma for clinicians: on the one hand, we may well have no desire to impose our own views on others; on the other, we need to draw the line somewhere because clearly there are some practices that, no matter the degree of apparent acceptability within the cultures in which they originated, still seem to us simply to be unacceptable. Examples, for the authors at least, include enforced marriage and female genital mutilation. So it is with moral decision making in our usual clinical contexts: while it is certainly correct to try to take the point of view of our patient (Lady Hale, speaking in the Supreme Court, famously talked about standing in the shoes of the incapacitous patient8) there must be limits to this. This is where the four principles framework can help. The Four Principles Ethical Framework Sitting somewhere towards the agnostic middle ground of the universalist-­relativist spectrum is perhaps the best-­known approach to health-­related moral decision making in the West, especially in the UK and United States: the four principles approach. Developed by Beauchamp and Childress (2001), this is not a moral theory at all, but rather an attempt to bring a number of moral theories into a  framework of internal relationships, balances and tensions designed to assist a decision maker to reach a decision. The four principles are very well known: respect for autonomy, beneficence,

8

 N v ACCG and ors [2017] UKSC 22 [1].

non-­maleficence and justice. Figure 3.1 represents these in relation to each other, the key to which is ‘balance’. Below, we outline a way in which the four principles can be ­combined into a workable and systematic approach for understanding a complex decision. There may be occasions when the four principles approach is not suitable, but this approach is helpful in many situations and brings the advantages both of providing a means of working through a complex decision in a ‘critiquing’ ­manner and of helping clinicians see where conflicts may arise. Non-­complex decisions, particularly if they represent ‘usual treatment practice’, are unlikely to need a formal application of such an approach, and the potential benefit of applying a critical thinking approach to a decision must be weighed against the time taken to do so. It is also helpful to distinguish between everyday decisions and those relating to serious medical treatments, with the latter needing a more ‘formal’ approach to decision making than the former.9 In the suggested approach outlined below you can see that the MCA is an integral part of the decision-­making process for complex decisions; its inclusion reflects the fact that the clinical conditions that make a decision complex are often also those that may impair a person’s capacity to make that decision for themselves. This is not exclusively the case, of course, and conversely, nor do we mean that the MCA is only applicable in complex decisions (this would be incorrect; the MCA could be in play in very many decisions in relation to the delivery of healthcare to those over 18 years in England and Wales10). In the next section, we outline the factors that we should consider when making a complex decision; that is, the factors that are intrinsic to the decision-­making process. Later, in Part II, we examine some of the factors that are extrinsic (and therefore arguably irrelevant) to the decision-­making process and yet are often influential upon it. This latter section includes ways to mitigate the influence of such extrinsic factors.

The Intrinsic Factors – What We Should Be Thinking About

This section looks at a method by which clinical factors and legal requirements may be integrated with the ‘four moral principles framework’ for ethical decision making. This framework does not stand alone, however, and it is very helpful to understand the context in which it operates. A. Application of the Four Principles Ethical System to Clinical Decision Making The four moral principles, outlined below, may be applied to decision making in various ways. One helpful way is to work through a decision of whether to initiate (or continue) a particular treatment by considering these principles in the order given below; this will help you to determine the justifiability of the decision you ultimately reach. Note that the principle of respect for autonomy (‘self-­ government’) appears both first and last in this approach and encompasses the patient’s own perspective on what constitutes a meaningful quality of life for her and therefore on whether she wants the treatment. 9  See, for example, s4(1) and 4(2) of the Mental Capacity Act 2005 (Independent Mental Capacity Advocates) (General) Regulations 2006. 10  The position in relation to those aged 16–17 years is more nuanced, as both the MCA and the common law relating to parental responsibility may be relevant; the MCA does not apply for these purposes to those under 16 years. This chapter is concerned with those aged 18 years and over.

Decision Making, Ethics and Law in Neurology  41

Each principle should be in balance with (i.e. understood in relation to) the other three, for example: This is about the individual

This is about doing ‘good’

Autonomy

Beneficence

Justice

This is about the community

Non-maleficence

This is about avoiding ‘harm’

v

v

How the principles are balanced is often the point in decision making at which conflicts of values occur Autonomy

Non-maleficence

e.g. Autonomy may conflict with Beneficence (”welfare”)

Beneficence

Justice

Figure 3.1  The four ethical principles and the role of ‘balance’.

As such, respect for autonomy encompasses aspects of decision making that are usually outside the realm of what the clinician can know and therefore outside of what the clinician should consider without the patient’s input. The only exception to this is when clinicians are undertaking a ‘best interests’ decision, in which case we are required to take into account the patient’s wishes and feelings, as far as these can be ascertained. Questions relating to beneficence and non-­maleficence are initially purely clinical enquiries (i.e. considerations of quality of life should not be a factor), but there is a caveat to this, relating to what we mean by a treatment being ‘futile’, which we discuss below. The decision not to provide a given treatment may be reached at any one of the five steps below, in which case continuing with the steps that come after that point is unnecessary. In particular, if the decision is, say, that a certain treatment is unlikely to produce a clinical benefit, then whether to provide this treatment does not later become a best interests decision – it is simply not an available option. This is not to say, however, that the decision is not discussed with and explained to the patient and family. For a treatment which, having considered the first four steps, is determined to be available to the patient, then it is for the patient to decide, on grounds of adequate information, freedom of decision making (i.e. the person is not being coerced) and having mental capacity, if they want it. If they do not have mental capacity, it will then become a best interests decision. There is also the need to consider what clinicians mean by the term ‘benefit’, which may be at odds with the court’s view on

the  meaning of this term. We also discuss this in the section on futility, below. For a treatment to be initiated or continued, all four principles must be sufficiently ‘in balance’, both for the patient and for the clinician (Figure 3.1). This is best understood as seeing each p ­ rinciple in relation to the others, but it is a matter of individual interpretation exactly where the fulcrum of balance between any two of the principles should be; in other words, it is a reflection of the individual’s values, which is why these points are where conflict may arise.

Four Principles in Five Steps

The four principles can be applied to much clinical decision making in the following way. Figure 3.2 summarises the process. 1. Respect for Autonomy (I) This first step applies specifically in the context of a person who does not have the mental capacity for the decision about whether or  not to accept (i.e. consent to) the treatment being offered and focuses on the question of whether the person has made a valid and  applicable advance decision to refuse the treatment (ADRT) being offered. ‘Valid’11 and ‘applicable’12 are terms that have specified meanings in this context, and you need to understand what these meanings

 Mental Capacity Act (n3) s25.  Mental Capacity Act (n3) s25.

11 12

42

Neurology: A Queen Square Textbook

Should this treatment be ‘on offer’ to the patient?

1. Respect for autonomy I

If the patient does not have mental capacity for the specific decision at the time the decision is needed, is there a valid and applicable ADRT?

If no ADRT

2. Beneficence

Based on the benefits of the proposed treatment and the chance that this treatment will achieve its physiological goal within a suitable time frame, should it be offered?

If yes

3. Non-maleficence

Based on the risk(s) that the treatment under consideration may cause significant harm(s), and weighing the benefits of the treatment against these risks, should it be offered?

If yes

4. Justice

Is this treatment a just use of resources?

If yes

5. Respect for autonomy II

Does the patient want the treatment? If the patient lacks relevant capacity, the decision must be taken using a best interests approach

Figure 3.2  Four principles in five steps. ADRT = advance decision to refuse the treatment.

are, as well as how the terms are applied. For example, among other factors, for an ADRT to be valid the person must not have subsequently appointed an attorney under the MCA with powers that cover the same decision (because this implies that the ADRT does not represent the patient’s latest thoughts on the matter). Nor can the person have done anything that is clearly inconsistent with the ADRT remaining their fixed decision13 because this implies a change of mind. A valid and applicable ADRT relating to the treatment under consideration carries the same weight as would a contemporaneous, valid refusal of consent given by someone who has the mental capacity to take the decision and as such, any question of best interests is irrelevant in this situation (regardless of who would otherwise have been the decision maker, be that clinician or attorney under an LPA). On this basis further consideration of the decision in question is not required, although, in keeping with the necessities of consent, if there is more than one potential treatment (including the possibility of doing nothing) this may mean that other related decisions do need further deliberation. If there is no valid and applicable ADRT, a clinician should consider the question of whether they are willing to provide particular treatments by working through steps 2 to 4. If, following these deliberations, the decision is that a given treatment could be provided then whoever is the decision maker at step 5 (patient, clinician under best interests or attorney under best interests) will need the information about benefits, risks and harms (and potentially issues of justice) to help inform the final decision.

2. Beneficence What are the benefits of the proposed treatment and what are the chances that this treatment will achieve its physiological goal within a suitable time frame? Clinicians may fall into the trap of refusing to provide a treatment on the basis that the outcome of a given treatment could, at best, only result in a quality of life that the patient would not want; issues of quality of life, which ought to remain outside of the clinician’s remit at this stage, become folded into the decision making. This is unsafe for reasons inherent to the meaning of autonomy; at this stage, the matter of beneficence is a clinical question. Note, too, that there may be more than one potential benefit and/or physiological goal and that consideration of a suitable time frame for each is also part of this step. For example, while CANH may sustain an unwell person in the short term, it may not have any effect on whether the person eventually dies of their underlying condition; that is, whether CANH is able to meet its physiological goal of sustaining life depends on the period of time you are thinking about (not being explicit about the time frame may be a source of confusion). This distinguishes those who die because CANH was withdrawn, regardless of the underlying disease process, from those who die as a direct result of the underlying disease, regardless of whether CANH was continued or withdrawn. The law looks at these two groups differently, with, as a broad generalisation, a much more critical eye applied to the former.14 Clearly, if the proposed treatment is, as far as it is possible to know, unable to provide the sought-­after clinical benefit then the

 See, for example, Re PW (Jehovah’s Witness: Validity of Advance Decision) [2021] EWCOP 52.

14  For clinical guidance as to the application of the law, see Clinically-­Assisted Nutrition and Hydration (CANH) and Adults who Lack the Capacity to Consent (Royal College of Physicians and British Medical Association 2018).

13

Decision Making, Ethics and Law in Neurology  43

treatment is not usually one that should be initiated (see the section on futility, below). However, the situation is less clear for a treatment already started: while, in theory, withholding and withdrawing treatment have equivalent clinical and moral formulation and significances, withdrawing treatment is much more visible, may contain messages of ‘giving up’ and has the additional hurdle of the need to justify why a treatment, that was once considered reasonable to give, is no longer appropriate. This highlights the need, in practice, for additional conversations with patient/family and brings an added potential for conflict with patient/family or with colleagues. 3. Non-­maleficence What are the risks that the treatment under consideration will cause significant harm and do the potential benefits of the treatment outweigh these risks? This step may give rise to some tricky considerations because the risks of a treatment may simply be unacceptable to the treating clinician (‘first do no harm’). For example, we would usually consider it wrong to ask a surgeon to undertake a procedure in which there is a high chance that the patient will die on the table. However, the greater and arguably more profound ‘share’ of any treatment risk is that carried by the patient. An example of a guideline in which the patient’s desire to undergo a treatment holds considerable sway, despite there being only a very small chance of success, is found in the guidance on cardiopulmonary resuscitation (CPR) produced by the UK Resuscitation Council (British Medical Association et  al. 2016). On the one hand, the guidance says that CPR will almost always be unsuccessful for the person expected to die of an incurable illness within hours or days (which is a very small time window) and therefore it will normally be inappropriate to undertake it. But, on the other hand, where the chance of CPR being successful is not zero, then the guidance suggests that the patient’s own perspective on the degree of risk they are willing to take should be persuasive on the decision whether or not to recommend CPR, even if the chance of success is very small. That is not the same, however, as saying that a clinician can be required to carry out CPR if they conscientiously consider it inappropriate, as no patient (or, if they lack the ability to ask, no person on their behalf) can demand such treatment.15 It is also relevant to note, perhaps because of its representation and place in popular culture, that CPR has been singled out by the courts as an area of particular concern when it comes to conversations between patients and doctors, with a heightened expectation that doctors will discuss with and inform patients of any recommendation that CPR should not be carried out in the event of cardiopulmonary arrest.16 The courts have not set out such expectations in relation to other forms of treatment, although this is not to say that such discussions are not likely to be appropriate, helpful, and even necessary if the real question is about the patient’s willingness to accept a share of treatment risk. 4. Justice: Is This Treatment a Just Use of Resources? This is a question that requires great care, and it is often easier to take the pragmatic view that resource decisions are best taken away from the bedside and at the level of the institution, or of government. This is not always possible, however, and while most clinical decisions are about what is right for the patient, there are some  Aintree v James (n4) [18]. 16  R (Tracey) v Cambridge University Hospital NHS Foundation Trust [2014] EWCA Civ 822; See also: Winspear v City Hospitals Sunderland NHS Foundation Trust [2015] EWHC 3250 (QB). 15

areas of medicine in which decisions about which patients should receive which interventions are often constrained by limited resources, such as the number of available intensive care beds. The COVID-19 pandemic threw into sharp relief the extent to which these decisions take clinicians into difficult waters, because the stretching to near breaking point of resources such as intensive care beds (which, in reality, also meant the availability of appropriately skilled nursing staff and decisions concerning which type of ventilator was available for use) meant that clinicians were confronted with situations in which it was very clear that admission of one patient would deny another patient the chance to benefit. Note that it is one thing to decide that a particular resource cannot be justified for a particular patient on clinical grounds, but it is entirely another to compare one patient against another patient for use of a limited resource when, under circumstances in which that resource was not limited, both patients would receive it. This is not least because, in discriminating between people based on clinical factors, you could easily be discriminating in a social (pejorative) sense without realising it: factors that may disfavour one individual over another for access to a limited resource may have themselves arisen from a past injustice, such as social inequality, causing a ‘corrosive disadvantage’ (De-­ ­ Shalit and Wolff 2007) that piles one injustice on another. It remains to be seen whether a result of the pandemic is a greater recognition that decisions about resource allocation are ones which place individual clinicians in near impossible situations, requiring systems-­ level responses and accountability. 5. Respect for Autonomy (II): Does the Patient Want the Treatment? Having arrived at the point of determining that, all things considered, the treatment under deliberation is indeed one that the clinician is prepared to ‘offer’ the patient, the final step is to ask the patient whether she wants it or not. For consent to be valid, three elements, established in common law, must be optimised: that the individual is properly informed of each available treatment option, including the option of doing nothing; that the person is free from coercion; and that the person has the mental capacity for the decision at the time the decision is needed. Inadequacy of any one of these may invalidate the consent. For patients who lack the requisite mental capacity (and assuming that the situation is not governed by either an ADRT or a decision by an attorney), a best interests decision will be needed on the part of the clinicians to determine whether the treatment may be given to the patient. Where the decision maker is an attorney appointed through an LPA for health and welfare, this person is under the same obligation to reach a best interests decision as a clinician would be, but note that the clinician retains a duty of care,17 as well as a human rights obligation to secure the person’s life,18 so it is important that the clinician both understands the basis for an attorney’s best interests decision and agrees that it is correct. Otherwise, the clinician’s duties require that she challenges the attorney’s decision. If this happens, it may be that the process of challenging leads the attorney to change their mind; it may equally be that the process of  challenging leads the clinician to understand a feature about the patient that they had not previously appreciated, such that the 17  Our duty of care has interesting origins in a case dating from the 1930s about a bottle of ginger ale and a snail: Donoghue v Stevenson [1932] All ER Rep 1. 18  Council of Europe, European Court of Human Rights, ‘Convention for the Protection of Human Rights and Fundamental Freedoms (European Convention on Human Rights)’ (1950), Article 2: The Right to Life.

44

Neurology: A Queen Square Textbook

clinician then agrees with the attorney’s analysis of best interests. Ultimately, however, if the clinician remains of the view that the steps being taken by the attorney are not truly in the patient’s best interests, they will need to seek the assistance of the Court of Protection, the one place where a determinative resolution of the question can be obtained. Note that the current MCA Code of Practice19 allows for three sets of circumstances in which the strong presumption in favour of prolonging life may be displaced on best interests grounds: because treatment would be ‘futile, overly burdensome to the patient20 or where there is no prospect of recovery’.21 At the time of writing, it appears likely that the new iteration of the MCA Code of Practice will not include the term ‘futile’; nevertheless, the concept is of sufficient importance, not least in contexts when the person has relevant mental capacity, that we discuss it in Part IB below. These three sets of circumstances are a potential source of confusion: the question of a treatment being overly burdensome for the patient is a factor for best interests decision making, because this may link to quality of life, but if a treatment is futile or there is no prospect of recovery, then should not either of these have been a reason to reject the treatment at step  2 and not step  5? This is a nuanced point, based on an understanding of the meaning of futility, within which the prospect for recovery will sometimes also sit. Hopefully the section on futility, below, will clarify this but, in short, step 2 is about a rational decision relating to clinical appropriateness, supported by the views of colleagues; the idea of futility is indeed relevant to this ‘pre-­consent’ stage, but there are quality of life elements to futility which therefore may need to be part of best interests deliberations.

Worked Example

How does this approach work in practice? In this section, we apply the five steps outlined above to the case of Mrs C. 1 Respect for autonomy (I) Mrs C has not undertaken any advance care planning, has not made her wishes known and has not made an LPA for health and welfare. Therefore, at this stage, we have no additional information and move to step 2. 2 Beneficence a The CANH, which is providing her with nutrition and fluids, has a very high chance of meeting the physiological goal of keeping her alive for the time being: neither the intracerebral haemorrhage nor the underlying brain metastases will lead to the end of her life in the immediate future, but withdrawal of the CANH would do so. On this basis, the CANH can be said to be of clinical benefit (the question of whether this is of personal benefit to her we must leave for now). Note that this benefit is in keeping with the long-­established legal principle of a presumption in favour of prolonging life. b We should also consider whether the CANH could be an important component of the overall care Mrs C receives which, taken together, has a goal of her recovering, say, to the point of her being able to interact with her family. This is also a clinical question – we are not asking whether this goal is something

19  Department of Constitutional Affairs. Mental Capacity Act 2005 Code of Practice. (Stationery Office, 2007). 20  This, presumably, either means ‘burden’ in terms of, say, the toxicity of chemotherapy being such that the person’s heart would be critically damaged (a clinical decision) or it would need to be question of what the patient is prepared to tolerate (the patient’s decision). 21  Code of Practice (n10) s5.31.

she would value, only whether we could achieve it with the aid, among other treatments, of the CANH. For this second goal the answer is probably going to be no; we think that she has no, or a very low, chance of recovering to this level, and therefore CANH cannot be said to provide a clinical benefit when this goal is in mind. c The converse of providing CANH also needs to be thought about: could withdrawal be considered to provide a clinical benefit? The presumption in favour of prolonging life is not unassailable, either on the grounds that an individual refuses a life-­ sustaining treatment or, if that person lacks capacity, because to continue to provide CANH is found not to be in the person’s best interests. On the basis that, morally and legally, withdrawing a treatment is the same as withholding it, this is equivalent to the ‘doing nothing’ option that should be part of treatment decision making. In Mrs C’s case, it seems a stretch to say that not providing CANH could be of clinical benefit to her. d Step 2 is not only sensitive to the goal we are thinking about but also to the time frame within which we are working. The untreated brain metastases are likely to limit her prognosis, perhaps to a few months, and it goes without saying that her death will happen regardless of the nutrition being provided to her when that time comes. In this sense, over this longer time frame, the CANH does not provide benefit. e If the prognosis were, instead, only a few days then we begin to reach a tipping point at which the withdrawal of CANH would have no impact on the timing of her dying, at which stage the CANH is no longer capable of achieving the physiological goal of keeping her alive. 3 Non-­maleficence a The risks and harms of providing CANH are usually considered minimal but vary between nasogastric feeding and PEG feeding. With respect to PEG tubes, there may be some discomfort from the insertion procedure, and there are risks associated with bleeding and infection, but these are generally regarded as small. As such, should CANH be considered to provide benefit, as in 2(a) above, then usually the risks and harms are insufficient to outweigh this. b However, were Mrs C’s prognosis to be shorter, even small risks and harms take on larger proportions relative to the possible benefits of treatment, until the point is reached when they may outweigh the benefit. c What about the risks and harms of withdrawing the CANH from Mrs C (or not starting it)? This also needs to be considered because this information may prove important in determining her best interests overall. For instance, what symptoms may arise as a result of dehydration and/or inadequate nutrition, and how may these be mitigated? 4 Justice The answer to the question of whether provision of CANH to Mrs C is a just use of resources is linked to the question of beneficence inasmuch as we would normally consider the resource necessary to provide PEG feeding as being reasonable in a person who would benefit from it while not being disproportionately harmed by it. It is possible that questions of resource become critical where beneficence and non-­maleficence are more finely balanced, but such is the problematic nature of applying this principle that such a decision might be better left in the hands of a court. 5 Respect for autonomy (II) For Mrs C, the question of respect for autonomy is worked out through the application of the best interests process by which a

Decision Making, Ethics and Law in Neurology  45

proxy for her consent, or refusal of consent, for CANH is ‘­constructed’. At this point clinicians (or others, if they are the decision makers) are required to take into account all the factors outlined in the framework for this process contained within the MCA.22 Among other things, this requires that the decision maker take into account aspects that were, until now, outside the clinician’s remit. In particular, this includes Mrs C’s wishes and feelings as far as they can be ascertained from those representing her (her husband, her daughter and others, such as close friends), in other words, best interests decision making is a collaborative process. There is a chance that her representatives do not have a clear sense of what Mrs C’s wishes would have been in these particular circumstances, but they are likely to have knowledge of her beliefs and values on which to base a view. It is helpful to reiterate that their role is that of informant, not decision maker (because none is an attorney under an LPA for health and welfare) and that they need only convey what they reasonably believe to be her wishes; they are not being asked for certainty (although some families may well feel certain that they know their loved one’s wishes), although, having said that, it remains to be seen what level of evidence the courts will call for in future cases.

What is the Value of the Four Principles Approach?

As in a case such as that of Mrs C, a working through of the four principles brings clarity to the decision making in a number of ways. The first is that it enables a systematic analysis of the decision, which may highlight aspects of the decision making that might otherwise not have received proper attention. It enables us to say exactly on what basis the decision is being made. The second is that it brings to light possible areas of disagreement and conflict, either between clinicians or between clinicians and the patient (or those representing the patient). These disagreements usually stem from differing values resulting in the prioritisation of different aspects of the four principles, but may also represent difficulties with communication or inadequately defined goals. In the case above, for instance, analysis of the potential benefit of the CANH demonstrates that (a) the goal(s) must be clearly defined and, (b) the time frame under consideration may be an important factor. It is also the case that, even where a clinician agrees that provision of CANH is of clinical benefit, whether the treatment is of value overall to Mrs C cannot be determined until Step 5. Disagreement may arise in any of these aspects; a lack of clarity increases the likelihood of this happening.

Underlying Values

One criticism of the four principles approach is that it is possible for two people, even two clinicians who share a common culture, to apply them to the same decision and arrive at different, even diametrically opposed, conclusions; the framework shows you where the crossroads are, but not which road to take. Why might this be? The reason, as indicated above, may be that other, unstated values are being brought to bear to the analysis, values that lie beneath the surface of the ethical evaluation provided by applying the four principles framework. The key point to note is that, when making apparently facts-­based, values-­neutral clinical decisions, we are in fact doing nothing of the sort. Rather, our own values are influencing decisions and are (for most of us) doing so beneath our awareness. This is a very important component of complex decision making which we explore in depth in Part II.

For now, continuing the examination of factors that we should actively bring to bear in decision making, we turn to the role of the law. B. The Role of the Law The legal landscape in which we practise medicine is constantly changing as statutes are passed that have a direct bearing on healthcare and as jurisprudence develops through the judge-­made law arising from court cases, either through interpretation of statute,23 or through developing responses to situations not covered by ­statute.24 This ingress of the law into medicine adds a layer of ­complexity to how we care for patients but whether you consider this a welcome development, or a hindrance, depends on your ­perspective. With regard to the MCA, for instance, healthcare ­professionals in England and Wales have demonstrated a marked reticence towards embracing it as fully as is intended, with a review by the House of Lords in 2014  noting that clinicians view it as ‘an  optional extra’.25 This is perhaps unsurprising given both the additional work needed to fulfil the requirements of the MCA in an already overstretched system and the fact that our clinical training has left us unprepared for this aspect of care. On the other hand, the obligations that arise from the MCA have the effect of forcing us to consider the role of human rights, and ethics more generally, in our practice of medicine and therefore, in welcome fashion, have given an emphasis to the need to ensure that we do not inadvertently disregard the rights of our patients. High-­profile cases have also started to stimulate discussions between medicine and the law, with a noticeable increase in the number of cases in which medical bodies have intervened to seek to ensure that judges reach their decisions in light of some understanding of clinical realities.

The Legal System

The four countries of the UK are divided into the three legal territories of England and Wales, Scotland and Northern Ireland, and these have, to many, surprisingly separate legislative jurisdictions and court systems. For example, those working in healthcare in England and Wales, while familiar with the MCA, may be unaware of similar (but not identical) law in Scotland (the Adults with Incapacity Act 2000) and in Northern Ireland (currently common law, but, in due course, will be statute law: the Mental Capacity Act (NI) 2016, which is not yet in force). The legal system in the UK follows a system of statutes (Acts), which are laws enacted through Parliament (hence ‘Acts of Parliament’), and which may be interpreted through later court cases, in combination with ‘common’ law, in which judges develop solutions in response to the particular needs of the court proceedings they are overseeing when these are not governed by a particular statute. Statutes, by nature, bring about slow but sometimes profound change and may be accompanied by Codes of Practice (‘Codes’) which themselves will have been subject to parliamentary scrutiny (and ultimately approval) before publication. Codes are not statutes but have a degree of legal standing. For example, the  MCA Code states that healthcare professionals are required (i.e. have a formal duty) to ‘have regard’ to what the Code says. A practitioner cannot do this unless they know what the Code says; ignorance in this regard is no defence.

 Such as in Aintree v James (n4).  For example, DL v A Local Authority & Ors [2011] EWHC 1022 (Fam).  House of Lords Select Committee on the Mental Capacity Act 2005. Mental Capacity Act 2005: Post-­Legislative Scrutiny, 2014. 23 24 25

 Mental Capacity Act (n3) s4; Code of Practice (n10) s5.13.

22

46

Neurology: A Queen Square Textbook

In the case of the MCA, the Code is important not least because the Act is formulated using principles, with relatively little detail about how the principles should be applied, reflecting the broad nature of the concepts of mental capacity and best interests. The detail is then supplied by the Code and subsequent interpretation in the courts. The MCA Code may be rewritten to reflect growing practical experience and case law without requiring that the Act be redrafted. A new edition of the MCA Code, at the time of writing, has undergone public consultation, but we do not know when it will go before Parliament, so the original version of the MCA Code remains in force. In contrast to statute law, decisions reached by judges, either in the interpretation of the statute, or developing the common law, may be responsible for much faster change in the legal standards applied to healthcare provision, but their case-­specific nature may mean that the application of such decisions to other healthcare contexts is not always obvious. Sometimes this requires further court cases to determine the boundaries of preceding decisions, made all the more complicated by the tendency of some judges to make ‘obiter’ (by the way) comments which, while not applying directly to the case before them, provide further fodder for digestion and application to future legal cases. In addition, in England and Wales, the lower courts are bound by decisions made by higher courts. The Court of Protection is its own court, so whether a decision of a judge in the Court of Protection is binding depends upon the seniority of the judge. All judges of the Court of Protection, as with other judges of other lower courts, are bound by the decisions of the Court of Appeal, which itself is bound by decisions taken at the level of the UK Supreme Court. Note that the jurisdiction of the Supreme Court is UK-­wide, as the name implies. However, very many of the cases that you may read, for instance from the Court of Protection, are simply judges of the court applying the MCA to the facts of the individual case before them, and their decisions do not bind the next judge applying the MCA to the facts of a different case. There is a growing sense of the ingress of the law into healthcare. It is not our intention to attempt an up-­to-­date summary of case law as it pertains to healthcare, but only to make the point that it behoves us all to try to keep abreast of the relevant legislation. This protects us as individuals but more than this, ensures that we do not inadvertently disregard the rights of our vulnerable patients. Trusts have access to legal advice and there may be occasions when it is right to access this, but other sources of advice include indemnity organisations, the British Medical Association and the GMC. Knowledgeable clinical colleagues, while unable to provide legal advice (unless appropriately qualified), may be a good source of ­relevant suggestions.

Mental Capacity Act 2005

The MCA came into force in 2007 since which time its reach into day-­to-­day clinical practice has ever increased. We do not review the necessary fundamentals of the MCA here,26 but it is useful to note that there are at least three good reasons why we should closely apply the requirements of the MCA when we need to. First and foremost is the fact that the protections afforded to all citizens (not only to patients) take the form of rights. It is the right of your patient that the MCA is applied properly, which, in practice,

26  For a very helpful guide to best interests decision making, see Clinically-­Assisted Nutrition and Hydration (CANH) and Adults who Lack the Capacity to Consent (Royal College of Physicians and British Medical Association 2018), in particular Appendix 1.

will mean following the guidance contained in the Code unless there is good reason not to. Second, since our patients have these rights, you might expect the courts to take a dim view of responsible clinicians and trusts when these rights are disregarded, and this is indeed what we see. An example is a Court of Protection case in 2021 which involved a patient, ‘GU’, at the Royal Hospital for Neuro-­disability (RHND), Putney, London, who had been in a state of prolonged disorder of consciousness since 2014.27 The question of whether continuation of CANH was in GU’s best interests first arose, at the behest of one of the patient’s relatives, in August 2018, at which time the view of the treating clinicians was that continued treatment was futile and possibly burdensome. Disagreement within the family as to whether CANH should be continued or not, however, led to the continuation of the existing care arrangements, rather than either a best interests decision being made or there being a referral to the Court of Protection for a judgment. This situation continued until the case was brought, again because of the intervention of a relative, to the Court of Protection in 2021, at which time it was concluded that it was not in GU’s best interests to continue to receive CANH. It was withdrawn and he died peacefully. Such were the concerns of the judge (Hayden J), however, about the ‘extraordinary delay’ in reaching a resolution to the decision that, following his judgment, he gave RHND time to prepare an explanation of why the delay had occurred. The hospital’s explanation was deemed inadequate and pointed to a ‘troubling’ prevailing ethos in which the hospital clinicians failed to follow the law or even to work within publicised professional guidance, which would have amounted to the same thing. The judge in this case, as have other judges in other recent court cases, criticised clinicians for failing to follow due process; where this happens, it is not uncommon for the trust to have to give an apology in open court. Obviously, we are more likely to avoid such court appearances, with their associated reputational risks, by adhering to the requirements of the law. Third, as alluded to in the opening of the chapter, when decision making is complex it is useful, on a practical level, to follow the processes laid out in law as part of the means by which a decision is reached. In this regard, the MCA and good clinical practice are – or should be – aligned, rather than working in opposition. It stands to reason, but it is worth making the point that, if the MCA processes are followed, we will sometimes reach best interests decisions with which we are personally uncomfortable. In other words, it is important not to enter into best interests decision-­ making processes, nor, in fact, any of the other decision points within the MCA, thinking of the process as a means by which to rubber stamp the outcome that you wish to see. This is not to say that the patient should receive whatever those speaking on behalf of the patient say the patient would want, any more than this is the case for a patient with capacity. It remains true, for instance, that some treatments and interventions that a patient may wish to have are simply not appropriate or become inappropriate as a clinical picture develops. Decisions regarding such treatments, which we may label futile, should not be subject to a best interests decision unless there is genuine choice to be worked through. At the same time, however, it is important to bear in mind that the courts take a view of futility that differs from the intuitive grasp of the meaning of the term that most clinicians have, so it is worth delineating these issues. We turn to this next.

 North West London CCG v GU [2021] EWCOP 59.

27

Decision Making, Ethics and Law in Neurology  47

Futility

As mentioned above, it is likely that the next iteration of the MCA Code will not make any reference to the term ‘futility’, perhaps reflecting its ambiguous nature and the way in which it may mask the role of value judgements. If the next iteration does remove the term, it will be important not to rely on this expression as justification for clinical decisions in the context of best interests. Nevertheless, futility remains a key ethical idea that is relevant to clinical practice and should be properly understood. Futility is important because it gives not only an ethical justification for the withholding of treatment, even when this is against the wishes of the patient, but it also gives a legal justification. This status is recognised in guidance by the GMC. If the patient (or their representative) disagrees, and in the context of best interests, the final decision rests with the court28 (which is also true for decisions where there is a fine balance between whether or not providing a treatment is in the person’s best interests). Most clinicians think of a treatment as being futile in one of two situations, although there is no consensus regarding definition. The first is where a particular treatment is unable to achieve the physiological goal needed in the time available – futile in a quantitative sense (given the uncertainty that accompanies most predictive judgements about the outcome of treatment, this will usually include circumstances in which there is a low but not zero chance of success). This may be because the treatment is inappropriate for the goal or because, although appropriate under some circumstances, is not appropriate in the particular circumstances of the patient. For example, the chance that CPR may be successful in a frail, elderly person with multiple comorbidities is likely to render this treatment futile in this sense of the word. The second, qualitative, sense is one in which, although the particular treatment may achieve the necessary physiological goal within a reasonable time frame, nevertheless the patient’s resulting clinical state is not one that would be wanted. Making a statement that a treatment is futile is fraught with problems, not least because an apparently clinical judgement in this regard may hide the true, values-­based, nature of the decision. For example, if a treatment has only a 1% chance of success of prolonging life, is it futile? Quite apart from the fact that being able to determine anything like an exact chance of success for a particular patient is unrealistic, even if you have the data at a population level, your response to this question is likely to be influenced by how you value life (of which more in Part II). So, values may well be a component of a quantitative futility decision, although usually unrecognised. Where does that leave us? To begin with, and despite our assertion above, clinicians are qualified (indeed, required) to make judgements in the quantitative sense of futility, but are not qualified to make a judgement regarding whether a patient’s quality of life is one that they would value as being worthwhile. This is a judgement for the patient alone, the only exception to this being when the patient lacks capacity, when it becomes a relevant factor in best interests decision making. Even under these circumstances it may be possible to determine the patient’s wishes and feelings directly from him but, if not, this information should be gathered 28  It is worth noting that, if the clinicians are clear that they are not proposing treatment on the basis that its futility means it is clinically inappropriate, the court will, in any proceedings, consider whether the clinicians’ view is reasonable, rather than whether (irrespective of the clinicians’ view) the treatment should nonetheless be given. A court cannot seek to compel clinicians to provide treatment they properly consider to be inappropriate.

from those representing the patient. The role of the decision maker is to incorporate information regarding the patient’s view of the value of a particular treatment outcome into the best-­ interests decision-­making process. It is helpful to understand that one of the reasons for the differentiation between the quantitative and the qualitative value of the outcome of a treatment, when considering best interests, is that it places a divide between the individual citizen (the patient) and those representing the state (the NHS clinicians) in an attempt to avoid state-­sponsored control of the population. A landmark case in the UK Supreme Court in 2013, Aintree v James (the first case relating to the MCA to come before this court), acknowledges these two distinct meanings of futility,29 but makes the point that the patient’s welfare must be considered in the wider sense (i.e. not limited to the physiological effectiveness, or lack thereof, of a particular treatment). So, there may be circumstances in which a treatment may be futile in the sense of being ineffective but where the patient may still derive a benefit from it; for example, the patient may consider that a short prolongation of her life is worthwhile, even though her life will come to an end soon, and the treatment which, while it is prolonging her life, will not prevent this. The treatment is therefore not futile all things considered. Note that, in this example, the futility of the treatment in terms of its ineffectiveness incorporates an important element of time: the treatment is ultimately ineffective but has some effect in the short-­term. It is important, therefore, when considering futility from a quantitative perspective, to define not only the physiological goal or goals of the treatment, but also the time frame. Failing to do so may give rise to misunderstanding and conflict. It is also important to note that the term ‘futility’, being neither diagnosis nor explanation, should not be used in isolation as a reason for withholding and withdrawing treatment; why the treatment is considered futile should be made explicit.

Disagreements

Disagreements in complex decisions making are to be expected. So far in this chapter, we have suggested that complex decisions are better made when the decision is broken down into its ethical component parts, to clarify the issues at stake and bring to light the areas of potential disagreement between those party to the decision. Doing such an analysis does not make disagreements disappear, but the fact that different interpretations may be placed on the same data, even when evaluated in a systematic manner, highlights the existence of other (usually unconscious) influences being brought to bear on the decision. Partly these influences may reflect different understandings of the law; in theory, at least, most of these may be resolved by better understandings of the law, while others may require the intervention of the courts. But partly these influences are the result of our own biases and assumptions about how the world should work. These are likely to be strongly held convictions and, if we are aware of them at all, we are likely to believe them to be a priori (and obviously) correct. These and other important factors are very likely to be significantly influencing decisions, even though such factors are, to some degree, extrinsic to the issues of the particular decision. One important way to counter the effect of these is to become aware of how we reach decisions at a psychological level and the factors that play into the process. In Part II, therefore, we look at these external factors in detail.

 Aintree v James (n4) [40].

29

48

Neurology: A Queen Square Textbook

Part II: The Extrinsic Factors – the Enemies of Decision Making?

How do any of us make moral judgements? It is possible that, to reach such a decision, a person will deliberate on the relevant ­factors and thereby reach a thoughtful, rational conclusion. This is  indeed what we are suggesting we should do by, for example, ­following the decision-­making steps outlined in Part I. But any conclusion that moral decisions are made through a primarily rational process may well be incorrect. It is likely that other, unconscious but potently influential emotional factors are likely to be at play, important among these being the need for ‘comfort’ in decision making. The role of comfort may, at the very least, be a useful marker for other influences coming into decision making, if we can become aware of it, but it does raise the interesting question of just how much our real-­life moral decision making may be influenced by how comfortable we feel with the proposed action. In other words, how much a moral decision is a reflection of the clinician more than of the patient, and therefore one that is shaped by factors external to the decision, albeit related to it. The worry is that this influence, if unconscious and therefore unexamined, may give rise to unfairness in decision making. Of course, comfort is an experience we have by degrees; the point may come when we are faced with a particular decision outcome with which we feel so uncomfortable that we cannot countenance it, no matter how much rational thinking we apply to the question. So, it is worth noting here that we are not suggesting that we should not have limits. For one thing, to try to operate beyond our moral comfort zone would be to put ourselves at risk of moral distress and even moral injury (Day et  al. 2021); and, as we have suggested above, a relativistic approach is internally incoherent. For another, we are required, both professionally and legally, to care for patients in the way we think is right. This latter point includes not providing a treatment we do not believe to be clinically indicated or to be in the patient’s best interests, as endorsed by Lady Hale in the Aintree v James case.30 What we are suggesting, however, is that, by becoming aware of how we process decisions and just how many of our own assumptions we may be bringing to bear on decisions, we may be able to relax the boundary of our moral comfort zone somewhat; we will begin to see how another’s perspective may be accommodated without compromising ourselves. In the following sections, we discuss three external factors that are likely to inform and influence our decision making and which tend to operate below our active awareness: unexamined assumptions, bias and ‘noise’ in the system. It is very difficult to remove the influence of external factors entirely, nor is it wise, as we have ­indicated above, but this is not to say that we cannot or should not mitigate some of this external influence. A necessary starting point, therefore, is to become aware of some of these influences. Prior to this however we will outline some of the current thinking about the psychology of moral decision making. A. Making Moral Judgements From the time of Plato, the making of moral judgements has been depicted as a struggle between reason and the emotions. For the rationalists of the 17th and 18th centuries, such as Descartes and Kant, however, reason was the prime mechanism of moral decision making. On the other hand, for philosophers such as Hume, who viewed moral judgements as being akin to aesthetic ones, moral knowledge arises from a moral sense, an intuition, by which we  Aintree (n3) [18].

30

r­ ecognise good and evil, rather than being arrived at through logic. For Hume, therefore, reason alone is inadequate for the task of ­making a moral judgement, and he seems to have been correct. Over the last few years, a multiplicity of ‘dual process models’ of cognition have been published, perhaps the most well-­known of which is the fast thinking and slow thinking model developed by Kahneman (Kahneman 2012). This describes cognitive processing as (conceptually) consisting of two systems operating in parallel. System 1 refers to processing that occurs below the level of consciousness, is rapid, with no sense of voluntary control and uses relatively less resource. It gives rise to intuitions, impressions, inclinations or feelings that simply appear in our awareness. System 2 is the processing we engage with much more deliberatively, with relatively high resource use, and is much more analytical. It gives rise to beliefs, intentions and attitudes. Dual process models are sometimes misunderstood to suggest that decision making is a balance between emotion and reason, but Kahneman is suggesting that the two systems are both versions of cognition with one producing its effect as an experience of emotion and the other as thinking. Within this context sits Haidt’s social intuitionist model of moral decision making (Haidt 2001) (developed over 20 years ago, it is still the subject of academic discussion). According to this theory, moral decision making is a process that begins with an intuition. This intuition occurs in the context of the individual’s culture and personality, for example, the degree to which the individual judges certain ‘goods’, such as honesty and fairness, as being important to themselves and to society. This intuition can be thought of as akin to Kahneman’s System 1 cognition, entering into conscious awareness effortlessly and automatically. If we need to, we then expend System 2 energies on defending our intuition through motivated reasoning. In terms of defending a decision related to the prediction of an outcome, for example how well or poorly a patient will recover from a brain insult, we are motivated to defend our underlying moral intuition about what is right because, despite evidence suggesting that we are poor at making predictions, we tend to believe and feel comfortable with the accuracy of our opinions (unless cognitive dissonance has been triggered); we feel ‘satisfied’ with them. This satisfaction and confidence appears to arise from an internal signal, what Kahneman calls a ‘self-­generated reward’ (Kahneman et al. 2021) for making our judgement fit coherently with the story (or, at least, those parts of the story we have paid attention to). However, this process may arise from a flaw in our System 1 activities that would probably only come to light through taking a critical analytical approach to the decision, that is, if we chose to use System 2 energies in a critical way, rather than to use them to defend the position we intuitively prefer. That we are reluctant to do so may reflect the additional cognitive resources required, but it may also reflect something happening within the group, for example, if we are discussing the decisions in a multidisciplinary team setting. Group dynamics are well understood to influence decision making in a number of potentially unhelpful ways, including: the disproportionate impact of the opinion of the first speaker, particularly if that person is a leader; a desire to conform with the group and therefore a reluctance to disagree (group think); and the possibility of polarisation, which is the propensity for a group to reach a more extreme decision than an individual would (rather than the moderating effect we might assume a group to have). There may be other factors at play also, for example, we are much more likely to agree with, and less likely to engage critical analysis of, the point of view of somebody we like. What these ideas suggest is that, in the process of integrating clinical, ethical and legal elements of a given decision, our moral

Decision Making, Ethics and Law in Neurology  49

judgements are likely to be heavily influencing the process. In the next sections, we examine some of the relevant factors which play a  role in our moral intuitions: unexamined assumptions, bias and noise. B. Unexamined Assumptions We all carry assumptions about the way the world works or should work. Some of these have their roots in the various social contexts of our childhoods, not least that of the family life of early childhood. Others derive from the socialisation process that is part of entering one of the healthcare professions. Most of us do not examine these assumptions, not least because to do so may be very uncomfortable if done honestly. How we value life is one of these.

The Value of Life

There are a number of interlocking ideas that are useful to consider at this juncture. One of these is to ask yourself where you sit on a spectrum that has, at one end, the idea that the quality of your life is more important than the length of your life, and, at the other end, the idea that the length of your life is the most important thing, regardless of its quality. Most of us will sit somewhere in the middle for most of our lives, but the approach of the end of life may have the effect of moving us more towards one end or the other, or for us to do so on behalf of a loved one whom we are representing at a best-­interests decision. Most Western clinicians regard quality of life as being a fundamental driver in the decisions we make for ourselves, but while this may also be true for many of our patients, who are therefore likely to agree with our analysis of a situation, it is not true for many others. Many clinicians, however, regard ideas that length of life is of prime importance as being somehow out of step with modern thought, rather than simply representing a legitimately alternative world view. If the family view is that it is in a patient’s best interests that they be kept alive on a ventilator, for example, even though there is no realistic prospect of recovery, we are in danger of supposing that such a view has been appropriated through culture rather than being rational, in other words, we may consider them to be unexamined and unreasonable views. And yet, we are equally guilty of allowing our own unexamined views to influence our decisions. An example of this is our cherished notion of respect for autonomy. The version of self-­government that we usually adhere to in this country, and which is established in both the MCA and common law, is highly individual and atomistic. It is not, however, ­universal: there are some cultures, for example, that hold to a much more relational approach to autonomy and believe that “what is best for me is the decision that most benefits my group”. It is also interesting to note that individualistic autonomy only gets you so far: if you wish to spend your dying days at home, but your family do not want to, or are unable to, support you to do this, the chances are that you will not get your wish. Classically there are three views on the value of life that may be helpful when thinking about our own assumptions. As so often, we will usually hold a mixture of these, with one or another coming to the fore in any given circumstance. • The instrumental view Your life has value inasmuch as you are able to do things, to contribute, say, to your family’s wellbeing. We can detect echoes of this view when family members tell us that their loved one ‘would not want to live like this’: the person cannot do what they used to do, can no longer contribute in the way they once did, and as result their life has less value to the point at which continued existence may not be meaningful to them.

This is not to say, just because there are alternative conceptualisations of the value of life, that the family’s assessment that their loved one would hold this view is incorrect (although we know that people change their minds about what constitutes a worthwhile life), but only to say that such a statement is challengeable from the perspective of other ways in which life may be valued. • The intrinsic view Your life has inherent value, simply on the basis that there is life. This view aligns, to a degree, both with those for whom the sanctity of life is paramount and with the default of a presumption in favour of prolonging life. Unlike a strong version of the sanctity of life view, the presumption in favour of prolonging life is not unassailable, but rather it is the starting point for deliberations about what is the right thing to do when end of life is a possibility; it is the position to be rebutted i.e. you need a good reason to overcome this presumption, in keeping with the European Convention on Human Rights. • The relational view Your life has value because you are valued by others. This is not always a benign position: families may have different reasons to value someone, for example, it may stem from love or from an expectation of service. This factor sometimes places patients in a difficult position, for example, when they no longer wish to continue chemotherapy but their family wish them to. This can be thought of as a vicarious version of sanctity of life, and clinicians need to be aware of the possible role of coercion in decision making stemming from this. Being aware of these ideas enables you to do two things: the first is that, once recognised, you can gently challenge a view point, for example, by asking a family who have asserted that the patient would not want to live like this, what the patient would have thought of the value of life for itself, or by exploring the relational view (e.g. what would the patient say if it was another family member, and not them, in their position). This is not to undermine what the family is saying but to explore and check the strength of that view. Perhaps more importantly, the second is that recognising what your own view would be in a given situation allows you to mitigate the influence of what would otherwise have been an unconscious factor in your discussions and decision making. C. Bias and Implicit Bias Bias is a source of error that operates by introducing a systematic deviation into evaluation, (Kahneman et al. 2021). In dual process theory, biases tend to arise when we engage our intuitive System 1-­type cognition, as might be expected. We may sometimes be cognisant of an idea, feeling, or intuition that has arrived fully formed into our awareness which may reflect a bias, but more usually we are unaware of the operation of bias; nevertheless, biases tend to have a powerful impact on our behaviours. This type of System 1 cognition is often the primary driver in decision making and much more prone to error than the analytical System 2-­type thinking (Croskerry et  al.  2013). Being an unconscious process, when mistakes are made  in our System 1 processes, we are rarely aware of the fact (Kahneman 2012). More than a hundred biases have been characterised, but perhaps the best known of the cognitive biases is ‘confirmation bias’, in which we unconsciously favour data that supports our beliefs and downplay any challenges, perhaps inadequately seeking contrary information. For example, in the context of reaching a diagnosis, flaws tend to be related more to reasoning than to knowledge, and bias plays a role in this. In moral decision making in particular (as  well as in, for instance, risk perception), decisions reached

50

Neurology: A Queen Square Textbook

through System 1 processes often come into our awareness as an emotion, as noted above: we work out what we think based on what we feel. Implicit bias is said to ‘involve associations outside conscious awareness that lead to a negative evaluation of a person on the basis of irrelevant characteristics such as race or gender’ (FitzGerald and Hurst 2017). In other words, this is a social bias, a form of unconscious, prejudicial stereotyping in which a categorical attribute (this person is black, this person is a woman, this person is obese, this person has a mental illness, etc.) is associated with a negative characteristic (perhaps an assumption of intellectual inferiority, or a propensity for violence). It tends to operate against groups who are already disadvantaged and may do so through subtle processes of non-­ verbal communication, such as physical proximity or frequency of eye contact (FitzGerald and Hurst 2017), as well as having the potential to influence decision making. The unconscious element differentiates implicit bias from the more explicit attitudes we associate with terms such as ‘racism’ and interestingly not only is the person unaware of having an implicit bias but often it operates in a direction opposite to the position consciously held by the person. A further helpful concept is that of ‘intersectionality’, which alerts us to the issue of a person belonging to multiple categories for each of which the person in power, such as a clinician, may hold unconscious negative associations. The existence of implicit bias is perhaps not surprising given that we are socialised within cultures where prejudicial stereotyping is common; one systematic review concluded that clinicians are just as likely to be influenced by implicit bias as is the general population (FitzGerald and Hurst  2017). Clinicians in this way may even actively contribute to health inequalities on a larger scale (Smedley et al. 2003). However, holding an implicit bias is not necessarily the same thing as acting on it and indeed there is some evidence to suggest that the influence of implicit bias may be corrected for, under some circumstances. D. Noise This is also a cause of error and is again a source of influence on decisions that arises from irrelevant factors, such as whether your favourite football team has just won or lost a crucial match. Noise is perhaps less well known than bias, is non-­systematic and therefore less predictable than bias, gives rise to variations in judgement that may be additive to those due to bias and may be transient in nature. One particularly interesting category is ‘occasion noise’ and refers to the influence of external phenomena occurring in our lives, such as the football example above. Danziger and colleagues (2011) studied judgments made at parole hearings by eight Israeli judges over ten months and found a clear pattern of association between, from the prisoners’ perspective, a successful judgment and the proximity of the hearing to the judge taking a meal break: those hearings that took place immediately after a food break were more likely to produce an outcome in favour of parole compared to those just before such a break. This study has been criticised on several fronts, but nevertheless reminds us that such external factors may have a profound and unconscious impact on our decision making. This is not new and has attracted the ‘HALT’ acronym, standing for hungry, angry, late (or lonely) and tired. There are many other such factors, some of which have been studied in the healthcare context, giving rise to the recognition, for example, that primary care doctors in the United States are more likely to prescribe opioids to patients attending an afternoon clinic, compared to a morning one, or if the clinic is running late (Neprash and Barnett 2019).

E. Mitigations Given the multiplicity of factors that either ought to bear upon ­decision making and sometimes do not, or are extraneous to the decision but nevertheless do bear upon it, how should we proceed? The first thing, in our view, is to recognise that we cannot get away from our natural ways of thinking in order to adopt an entirely neutral approach to decision making, even if, sometimes, it appears that the courts seem to adopt the position that this is possible.31 Not only is it not possible, it would, indeed, be likely to be an actively bad thing to do. This is because, as we alluded to above, taking a wholesale relativist approach is logically incoherent and risks moral self-­injury. Having said that, the second thing to note is that we can probably do much more to mitigate the impact of our own world view on decisions that affect others. Where to draw the line between these two positions we cannot know, as it will be different for us all, but this section looks at some approaches that may help to reset the line closer to the ideal.

The Self

There is no doubt that skill and intelligence play a part in how well we manage to navigate the various threats to good decision making, and, of course, the good news is skills can be learnt if they are perceived to matter. But it is also likely that adopting the right attitude and cognitive style will make a considerable difference; Kahneman recommends being ‘actively open minded’ (Kahneman et al. 2021). There are some practical steps to take: as advertisers know well, what we pay attention to matters because it alters behaviours. On this basis, paying attention to the various types of thinking error to which we are each prone begins to bring to awareness of how these impact upon our decisions. In particular, becoming aware of how often we rationalise, post-­hoc, our affectively mediated moral intuitions may allow us to consider why we have the intuitions we do, and to ask ourselves if these are reasonable. One helpful approach to adopt in this regard is to resist reaching a premature conclusion (Kahneman et  al.  2021) by becoming aware of our self-­generated reward signals and distrusting them. And keep the HALT factors in mind – could the decision you are making be influenced by the fact that it is time for lunch?

The Group

Group dynamics play a very important role in decision making and therefore the role of the chair of a meeting is vital to facilitating an open forum for discussions. It is helpful to recognise and mitigate the features of group interactions we mentioned at the beginning of Part II, section A, and for the chair to actively welcome contrary opinions. We need to learn how to hear a disagreeing viewpoint in the spirit of reaching a better decision (the goal, after all, is accuracy, not individual expression; Kahneman et al. 2021) and, to this end, team members should be encouraged to articulate the reasons behind their views clearly, perhaps using the four principles framework, so that others may be able to see any errors of cognition (or fact). Apart from anything else, this has the effect of ‘stress testing’ a given decision. And it may sometimes be helpful, in the course of  decision making, to identify which moral theory is in the ascendancy for any particular component of a decision  – in so doing ­cultural biases may become apparent. 31  See Institute for Criminal Policy Research. Judging Values And Participation In  Mental Capacity Law, https://www.icpr.org.uk/judging-­values-­and-­participation-­ mental-­capacity-­law for a research project investigating the influence of judge and legal practitioner values on aspects of mental capacity law.

Decision Making, Ethics and Law in Neurology  51

Conclusion

In the attempt to reach less arbitrary decisions, how do we optimally integrate clinical, ethical and legal elements when making complex decisions, while mitigating the influence of less relevant, extraneous factors? We hope we have demonstrated the potential of taking a systematic approach that incorporates the twin aspects of ethics and law, as well as raising awareness of some of the factors in decision making that often go undetected and yet are profoundly influential. Our contention is that, by becoming more aware of these factors and by taking personal and group-­level steps to mitigate them, we can move further towards the patient’s perspective for a given decision, while at the same time recognising, and being able to justify, where the limits to this lie. At the end of the day, a decision must be taken in a timely manner so that the time taken over the decision does not itself become an issue of the just use of a scarce resource. This chapter, despite appearances, is not a counsel of perfection, only a counsel for improvement. In our experience, decisions taken using some of the tools outlined in this chapter both bring their own sense of satisfaction related to having reached a clearly reasoned decision and become easier and quicker to apply with practice.

References

Beauchamp T, Childress J. (2001). Principles of Biomedical Ethics, 5th edn. Oxford: Oxford University Press. British Medical Association, Resuscitation Council (UK), Royal College of Nursing. (2016). Decisions Relating to Cardiopulmonary Resuscitation: Guidance from the British Medical Association, Resuscitation Council (UK) and the Royal College of Nursing (previously known as the ‘Joint Statement’). London: British Medical Association, Resuscitation Council (UK) and Royal College of Nursing. https://www. resus.org.uk/library/publications/publication-­decisions-­relating-­cardiopulmonary (accessed 21 July 2023).

Croskerry P, Singhal G, Mamede S. (2013). Cognitive debiasing 1: origins of bias and theory of debiasing. BMJ Qual Saf 22: ii58–ii64. Danziger S, Levav J, Avnaim-­Pesso L. (2011). Extraneous factors in judicial decisions. Proc Natl Acad Sci U S A 108: 6889–6892. Day P, Lawson J, Mantri S et al. (2022). Physician moral injury in the context of moral, ethical and legal codes. J Med Ethics 48: 746–752. De-­Shalit A, Wolff J. (2007). Disadvantage. Oxford: Oxford University Press. Department of Constitutional Affairs. (2007). Mental Capacity Act 2005 Code of Practice. London: Stationery Office. FitzGerald C, Hurst S. (2017). Implicit bias in healthcare professionals: a systematic review. BMC Med Ethics 18(1): 19. General Medical Council. (2020). Decision Making and Consent. London: General Medical Council. https://www.gmc-­uk.org/ethical-­guidance/ethical-­guidance-­for-­ doctors/decision-­making-­and-­consent (accessed 21 July 2023). Haidt J. (2001). The emotional dog and its rational tail: a social intuitionist approach to moral judgement. Psychol Rev 108: 814–834. Kahneman D. (2012). Thinking, Fast and Slow. London: Penguin. Kahneman D, Sibony O, Sunsyein C. (2021). Noise. London: Collins. Neprash H, Barnett M. (2019). Association of primary care clinic appointment time with opioid prescribing. JAMA Netw Open 2(8): e1910373. Royal College of Physicians. (2020). Prolonged Disorders of Consciousness following Sudden Onset Brain Injury: National Clinical Guidelines. London: Royal College of Physicians. Royal College of Physicians, British Medical Association. (2018). Clinically-­Assisted Nutrition and Hydration (CANH) and Adults who Lack the Capacity to Consent: Guidance for Decision-­Making in England and Wales. London: British Medical Association. Smedley B, Stith A, Nelson A, eds. (2003). Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Washington DC: National Academies Press.

Further Reading

Kong C, Ruck Keene A. (2018). Overcoming Challenges in the Mental Capacity Act 2005: Practical Guidance for Working with Complex Issues. London: Kingsley. Danbury C, Newdick C, Ruck Keene A, Waldmann C, eds. (2020). Law and Ethics in Intensive Care. Oxford: Oxford University Press.

CHAPTER 4

Neuropathology: Introduction to History, Diagnostic Approaches, Techniques and their Interpretation Sebastian Brandner UCL Queen Square Institute of Neurology

Introduction

Neuropathology as a discipline is an integral and essential ­component of the clinical neurosciences. This chapter introduces the h ­ istory of the development of neuropathological techniques such as the use of tinctorial stains, immunohistochemical techniques and a selection of molecular pathology techniques that are relevant in neuropathology diagnostics. This chapter is illustrated with images explaining the pathway of a tissue biopsy through the laboratory, resulting in a histological slide, a molecular test result, and ultimately, an integrated diagnosis. It is therefore a useful reference for the readership who are not directly familiar with techniques and workflow in cellular pathology laboratories and helpful to those who want to refresh their knowledge of the technological background in neuropathology.

Neuropathological Techniques

The principle of tissue analysis in neuropathology is the same as in histopathology, using identical reagents and instruments for tissue fixation, processing into paraffin wax, sectioning and staining (Figure 4.1). Likewise, molecular genetic techniques are the same in general histopathology and neuropathology disciplines. These two disciplines are set apart only by the tissues examined and the commonly applied tinctorial and immunohistochemical stains, and the specific genetic alterations examined.

Processing and Analysis of Histological Samples Samples can originate from the living (biopsy) or from the dead (autopsy or postmortem). Samples from both sources can be very small (e.g. biopsy or autopsy of a nerve) or very large (e.g. resection of a temporal lobe as part of epilepsy surgery or removal of an entire brain or spinal cord at autopsy). Likewise, autopsy samples can be entire organs (brain) or very small (e.g. a peripheral nerve or a small Historical Aspects of Neuropathology The neuropathology discipline has often developed independently sample of a whole organ). All samples, regardless of their size, from anatomical pathology. It was in fact often more closely associ- are first assessed microscopically. Such examination can be on the ated to clinical neurosciences, such as neurology, neurosurgery and native material (i.e. fresh) or after immersion in formalin fixative. psychiatry. During its foundation over 100 years ago, neuropathol- Macroscopic examination is of course essential for the examination ogy had arguably more input from psychiatrists and neurologists of whole brains, to assess alterations such as atrophy, tissue defects (Alois Alzheimer, Franz Nissl, Arnold Pick) or neurosurgeons (infarctions), neoplasia, demyelination or neoplasia. However, mac(Harvey Cushing and Percival Bailey leading the pathway to modern roscopic inspection and determination of weight and size is also neuro-­oncology) than from general histopathology. During these essential for surgical neuropathology, for example to assess size and times, the development of neuropathological techniques represented shape of the sections of tumours or lobes, or of small biopsy is such powerful analytical methods for the discovery and understanding of as stereotaxic brain biopsies, nerve biopsies or muscle biopsies. Ideally, all samples are directly delivered as fresh specimens into a range of neurological disorders. While technological developments in histopathology/anatomical pathology directly benefited neuro- the neuropathology laboratory. Fresh tissue is essential for the followpathology, it was often the neuropathology discipline that led the ing scenarios: (i) brain biopsy, delivered from the neurosurgical theafield in terms of innovation; for example, resulting in evidence-­ tre to the neuropathology laboratory to prepare an intraoperative based molecular classifications, such as biomarker-­defined brain squash or smear preparation to rapidly diagnose a neoplastic lesion in tumour classes, epigenetic profiling with DNA methylation arrays brain biopsy; (ii) muscle biopsy, to preserve enzymatic activity, morand the algorithmic classification of the methylome data or an inte- phological features and integrity of antigens for most diagnostic grated anatomical and molecular classification of neurodegenerative immunohistochemical stains; and (iii) banking of diagnostic tissue disorders. Such innovative approaches were sometimes led by neu- for molecular pathology applications such as whole genome sequencropathologists who were recruited from neighbouring disciplines ing, and for possible research applications in the future. The majority of samples are fixed in formalin, an aqueous s­ olution such as neurology, neurosurgery or histopathology, creating a unique mix of skill set, knowledge, research interest and diagnostic of formaldehyde, which irreversibly crosslinks primary amino groups abilities. Such diversity is not always achievable, and sometimes of proteins and also preserves nucleic acids. Typically, biopsy material overtly prescriptive recruitment schemes can present an obstacle to is placed into small, standardised tissue cassettes (Figure 4.1Ib–e) and requires fixation usually for four hours, while a whole brain requires maintaining such a small, varied and unique specialty.

Neurology: A Queen Square Textbook, Third Edition. Edited by Robin Howard, Dimitri Kullmann, David Werring and Michael Zandi. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd. 53

I Cut-up Specimen cut-up

(a) Specimen inspection

(b) Wrapped in gauze

II Processing

III Sectioning

IV Staining

Tissue processor

Microtome

Autostainer H&E

(a)

(a)

Embedding station

(b)

(a) Sectioning

(b)

Embedding mould

H&E slides

(b) Section floating

IHC instruments (c) Tissue-tek cassette

(c)

c

(c)

Slides to dry

Sample in wax block

(c) Antibody dispenser (d) Basket in processor

(d)

(d)

Wax blocks, storage (d) IHC - slides

(e)

(e)

(e)

Figure 4.1  Pathway of a tissue specimen from fresh tissue to stained slides. (I) Specimen cut-­up and fixation: (a) The tissue arrives in a container with buffered saline and is put on to a Petri dish for dissection. (b) Specimens are examined and size is determined. (c) Small samples such as this one are wrapped into thin gauze (blue) and placed into a cassette. The gauze is used to avoid escape of small specimens through the openings of the cassette during the tissue processing. (d) The cassette is then closed with a lid. (e) The cassette is then transferred to in a basket and immersed in formalin fixative. All cassettes collected during the day are placed in this basket. (II) Processing: (a) The basket is placed into a tissue processor, which circulates fixative, graded alcohols, xylene and eventually hot paraffin. This process is enhanced by vacuum infiltration and is precisely timed. (b) After completion of the processing cycles, usually in the next morning, the tissue cassettes are taken out from the processor and transferred to an embedding station. (2c) The embedding station dispenses hot wax into metal moulds into which the tissue specimen is temporarily placed. (d) The same cassette is placed on top of the mould and, once the paraffin wax is cooled, it forms a paraffin block. (e) Paraffin blocks can be stored at room temperature and thus can be conveniently and cost-­effectively archived in boxes or drawers. (III) Sectioning: (a) The next step, which can be immediately afterwards or at any later time, is the microtome sectioning. (b) The detail shows the placement of the block in the microtome holder and how it is moved up and down, shearing off thin paraffin slices (3–5 μm). (c) The slices are then placed on a warm water bath. The section floats on the water bath and is picked up with a glass slide. (d) Wet glass slides with freshly mounted sections are temporarily stored on a rack to dry. Subsequently, they are placed in an oven to remove residual humidity and to ensure best adhesion to the glass slide. Again, these slides can be used immediately for staining or can be stored in an archive for months or years. (IV) Staining of slides: (a) Tinctorial staining is done in most diagnostic pathology laboratories with automated staining machines to remove the paraffin wax from the glass slide and then transfer the slides into vessels containing the haematoxylin and eosin (H&E) or other dyes, as necessary. After staining, the slides are dehydrated again in ascending series of ethanol and xylene solvents, a drop of coverslipping medium is added and a coverslip is placed on the slide. (b) The resulting H&E-­stained slides are placed on a tray for examination. (c–e) Automated immunostaining is performed in a similar way using automated machines (c), which dewax the slide, place antibodies (d) and wash them off, followed by counterstaining. The last step (coverslipping) is done in a machine (coverslipper) as shown in (a, left part of the instrument). Finally, the coverslipped slides (e) are placed on a tray for microscopic examination or, in some laboratories, scanned at high-­resolution for examination with an imaging software. CGH, comparative genomic hybridisation; EGFR, epidermal growth factor receptor; IDH, isocitrate dehydrogenase; IHC, immunohistochemistry; JC, John Cunningham; MAPK, Mitogen-­activated protein kinase; NADH, nicotinamide adenine dinucleotide + hydrogen; PCR, polymerase chain reaction.

Neuropathology: Introduction to History, Diagnostic Approaches, Techniques and their Interpretation  55

two to three weeks in formalin for adequate fixation. Following fixation, the tissue cassettes are then placed in an automated tissue processor (Figure 4.1Ie, IIa) where they are gradually dehydrated, using alcohols of ascending concentrations, followed by xylene and eventually, hot (liquid) paraffin wax. In a subsequent manual process, the tissue is cast into paraffin blocks (Figure 4.1IIb,c), which is the universal, permanent storage format of tissue in histopathology and neuropathology laboratories worldwide (Figure  4.1IId,e). In routine settings, these blocks are then sectioned to obtain paraffin sections of 3–5 μm thickness using a microtome (Figure 4.1IIIa,b). These sections are spread out on a warm water bath (Figure  4.IIIc) and mounted on standardised glass slides (25 × 75 mm), on which they are left to dry (Figure  4.1IIId). These are so-­called ‘unstained sections’, which can be either stored in an archive or, more typically, further processed subsequently, to remove the paraffin c­oating in ­solvents, followed by rehydration in decreasing concentrations of ethanol, and phosphate buffered saline. This leaves a dewaxed tissue section on the slide, which now can be used for tinctorial stains (Figure 4.1IVa,b), immunohistochemical stains (Figure 4.1IVc,d,e) or for the extraction of nucleic acids for molecular tests. Histological Techniques A range of techniques have been developed over the last 120 years, including histochemical stains (Figure  4.2a,b,c,e), immunohistochemical stains (Figure  4.2d,f) and nucleic acid-­based detection methods. The histological techniques aim at the visualisation of the components of the central nervous system, such as cell nuclei, the cytoplasm including processes such as axons (Figure 4.2c,d), dendrites, glial cell processes, myelinated fibres (Figure 4.2e,f), and to detect alterations and pathologies, such as intracellular or extracellular inclusions, protein deposits (Figure 4.3i,k,l), abnormally phosphorylated proteins (Figure 4.3j), loss of expression (Figure 4.3p), protein overexpression, point mutations (Figure  4.3m,n,o) and even gene fusions. Tinctorial Stains The most commonly used stain in pathology departments worldwide is a combination of haematoxylin, which stains cell nuclei in deep purple, and eosin, a pink-­red dye, staining cytoplasm, connective tissue, and many other structures (Figures 4.2a, 3a,b). It can be prepared rapidly (within 30  minutes manually or using an automated staining equipment) and is relatively inexpensive. It is established as a de facto standard for all histological specimens, which is helpful when pathologists exchange samples for consultation, but also for modern applications, such as image analysis. A large number of more specialised tinctorial stains were used in the past, in particular in neuropathology. These were developed in the first part of the 20th century, at a time when anatomical and histological structures were discovered and characterised and needed adequate visualisation. Typical examples are Nissl (cresyl violet) staining for rough endoplasmic reticulum, visualised in neuronal cell bodies (Figure 4.2b), silver staining methods for axons (Figure  4.2c), dendrites, intracellular neurofibrillary tangles or myelin stains for myelinated fibres of the white matter (Figure 4.2e, i-­l) or the peripheral nerve. Pathological aggregates, such as amyloid, can also be detected with tinctorial stains. Still relevant nowadays are dyes to visualise for example elastic fibres and collagen, glycogen-­rich inclusions or myxoid deposits. Immunohistochemical Techniques The principle of immunohistochemistry (IHC) was established in the 1930s, and the American biochemist Coons published the first

IHC study using antibodies conjugated with fluorescein isothiocyanate in 1941 (Coons et al. 1941). This conjugate resulted in covalent binding of the fluorescent molecules to an antibody, which Coons used to localise pneumococcal antigens in infected tissues. A significant milestone was the discovery and development of monoclonal antibodies in the 1970s, which led to the introduction of immunohistochemistry into histology laboratories in the 1980s. As with all methods introduced into clinical practice, IHC was initially an experimental method and was gradually standardised, optimised and became mainstream, and thus became affordable first in industrialised countries and subsequently also in emerging economies. Nowadays, IHC is in most laboratories implemented as a fully automated process using highly sophisticated staining instruments (Figure  4.1), which perform all steps from dewaxing, antigen retrieval, antibody application, washing steps and colorimetric detection (usually diaminobenzidine), including the application of a counterstain (haematoxylin) (Table 4.1). Antibodies are nowadays available against virtually any antigen and, if they are not commercially available (e.g. for specific research applications), they can be custom-­made within weeks. Nearly all diagnostically relevant antibodies are commercially available and have been optimised and modified to detect formalin-­crosslinked antigens on tissue sections. The first generation of neuropathology-­ relevant diagnostic antibodies in in the 1980s detected ‘biomarkers’ to identify cell or tissue types; for example, synaptophysin, a synaptic marker, for neuronal tissues, glial fibrillary acidic protein (GFAP) for glial lineages (Figure 4.3c, d), including glial tumours. Immunostaining against cytokeratins can help detecting cells of epithelial lineage, in the context of diagnostic neuropathology, for example for cyst linings or metastatic carcinomas or stainings for lymphoid markers identify T cells and B cells in the context of neuroinflammation or haematological disorders involving the central nervous system. In neuromuscular diagnostics, the development of antibodies against membrane proteins in the 1990s (dystrophins, dystroglycans, sarcoglycans) were a significant step towards ‘molecular’ diagnostics and precision diagnosis. The discovery of tumour-­specific mutations in the 2010s has led to the development of antibodies that react specifically with a mutated epitope, for example the specific amino acid change R132H of the isocitrate dehydrogenase 1 (IDH1; Figure  4.3a), which later became a diagnostic marker for astrocytoma and oligodendroglioma, the histone K27M mutation for diffuse midline gliomas (Figure 4.3n), histone H3 G34R mutation for rare hemispheric high-­grade gliomas in young adults (Figure 3o) or the detection of the BRAF V600E mutation in melanomas, pleomorphic xanthoastrocytoma or ganglioglioma. Detection of the loss of expression is a routine test in glioma diagnostics, for example the ATRX protein (Figure 4.3p). Even fusions can be detected, for example through the translocation of a fusion partner from the cytoplasm to the nucleus, as it is the case in solitary fibrous tumours, where the STAT6 protein fuses with NAB1 and results in a diagnostic nuclear location of STAT6. The alteration of epitopes, such as the phosphorylation of the tau protein (Figure 4.3j), is a commonly used method, mainly in the diagnosis of tauopathies, a group of neurodegenerative diseases, and for the diagnosis of Alzheimer’s disease (Table 4.1). Molecular Pathology Techniques The nucleic acids DNA and RNA are remarkably well preserved in formalin-­ fixed, paraffin-­ embedded (FFPE) tissues. Even though DNA is broken down into relatively small fragments of a few hundred base pairs, it is suitable for a wide range of applications. FFPE-­ derived DNA can be used for detection of point mutations, using

Haematoxylin & Eosin

Nissl

Exemplar magnifications

General morphology

(a) Bielschowsky

Neurofilament IHC

Axons

(b) Luxol fast blue

Myelin basic protein IHC

Myelin

(c) Electron microscopy

Confocal microscopy

Myelin

(d)

Figure 4.2  Staining methods and visualisation of structures of the nervous system (See also Table 4.1). (top row) The most widely used tinctorial stain in diagnostic histopathology and neuropathology is haematoxylin and eosin. It shows the general morphology (here, cerebral cortex) with nuclei in purple and the neuropil in pink. The Nissl staining (cresyl violet) preferentially labels nucleic acids (negatively charged molecules) and, in particular, in neurons visualises ribosomes (arrowhead) (cerebellum, Purkinje cells). (second row) The Bielschowsky silver impregnation has been extensively used to demonstrate neuronal processes including dendrites, axons (shown here, subcortical white matter) and neurofibrillary tangles. With the increasingly available immunohistochemical (IHC) technique, Bielschowsky silver staining has been superseded by more specific immunostainings, such as the neurofilament IHC (peripheral nerve). (third row) The Luxol® fast blue (Merck, Rahway, NJ, USA) staining for myelin is still popular for assessment of the myelination in autopsy material (here, pons). In biopsy samples, such as nerve and brain biopsies, myelin basic protein IHC is preferred (brain biopsy), as it is more specific and can be done on automated instruments. (bottom row) Electron microscopy is the method of choice to analyse the ultrastructure of myelinated fibres (sural nerve biopsy). Research sample to detect myelin basic protein (green) and abnormal prion protein (red) in myelinated fibres of the subcortical white matter. The right-­hand column illustrates different magnification steps in diagnostic neuropathology. (a) At 0.5×, the entire anatomical structure (inferior olive, staining with Luxol fast blue and cresyl violet or Luxol–Nissl) is visible. (b) At 2.5×, individual anatomical structures such as the olive nuclei and myelinated fibre tracts can be identified; however, not at detail of individual cells. (c) With a 10× lens, individual neurons can be identified but subcellular structures are not yet visible. (d) With a 40× lens, individual neurons are identified, with visible nucleus, nucleolus and individual myelinated fibres. Images were taken at corresponding magnifications using a calibrated digital slide. Scale bar corresponds to 5 mm (a), 1 mm (b) 250 μm (c) and 60 μm (d).

Pathological reactions in the CNS Normal

(a)

Pathological

(b) Cortex (H&E)

(c)

(g) Cortex CD68

α-synuclein

Demyelination

Histone H3 G34R

(p)

(l) Macrophage

Histone H3 K27M

(o)

(k)

(h)

IDH1 R132H

(n) Phospho-tau

Gliosis

White matter (MBP)

(m)

(j)

(f)

Mutations

Amyloid β

Ischaemic infarction

Cortex (GFAP)

Neurooncology

Protein aggregates

(i)

(d)

(e)

Neurodegeneration

Prion protein

ATRX, loss of express.

Figure 4.3  Examples of pathologies in the central nervous system (CNS) and methods of their detection (See also Table 4.1). The left and left-­centre panels show a juxtaposition of normal (left) and pathological (centre left) tissue. (a, b) Normal cortex and ischaemic infarction, with vacuolisation (red arrow) and hypoxic damage of neurons (blue arrow). (c) CNS cortex with minimal gliosis, visualised with an immunostaining for glial fibrillary acidic protein. (d) Corresponding area in a brain with advanced prion disease. (e) Myelin basic protein staining of normal subcortical white matter. (f) Demyelination in a biopsy sample with an acute demyelinating encephalomyelitis. The microglia and macrophage marker CD68 shows very few microglia all cells in a normal cortex (g) but widespread activated microglia/macrophages, here in a sample with a CNS vasculitis (h). The centre-­right column shows protein aggregates in neurodegeneration. (i) Amyloid-­β plaques in the neocortex of a brain with Alzheimer’s disease, and (j) immunostaining for hyperphosphorylated tau showing neurofibrillary tangles (blue arrow) and neuritic plaques (red arrow), in the hippocampus of the same brain. (k) Alpha-­synuclein in a midbrain with Parkinson’s disease and (l) plaques in a brain with inherited prion disease. The right-­hand column shows antibodies to detect point mutations and loss of expression in neuro-­oncology diagnostics. (m) Immunostaining against IDH1 R132H, the most common isocitrate dehydrogenase (IDH) mutation in astrocytoma and oligodendroglioma, highlights the cytoplasm of tumour cells. (n) Histone H3 K27M, and (o) histone H3 G34R detected in the nucleus of tumour cells using mutation-­specific antibodies. (p) ATRX, a protein involved in chromatin remodelling, can be lost in a range of CNS neoplasms. In combination with an IDH mutation, it is diagnostic for the lineage of IDH-­mutant astrocytomas of all grades. The dark brown nuclei are non-­neoplastic (blue arrow); that is, they have retained the expression, while the blue counterstain indicates tumour cells with loss of protein expression (i.e. no brown staining detected, red arrow).

Table 4.1 Test methods for the examination of neuropathological specimens. The range of tests spans conventional tinctorial or histochemical stains, the very commonly used immunohistochemical (IHC) staining variety of molecular tests and electron microscopy. Each test is listed with typical examples for its use, the most commonly used technology, structures or genetic changes detected and the typical advantages or disadvantages of the methodology. The table is complemented by Figure 4.4. Detected feature, mutation or pathology

What is detected

Examples

Tissue type and cellular components, cellular content

Tinctorial properties of tissue and cell structures

Nuclei: haematoxylin Cytoplasm: eosin Myelin: Luxol fast blue Connective tissue: Van Gieson Glycogen: periodic acid Schiff Phosphorylated neurofilaments and tau filaments: axonal stains (e.g. Bielschowski) Fungi: silver stain Bacteria: Gram stain

Cell types and their Enzymatic activity metabolic state or activity

Cell of origin, lineage

Technique or technology

Advantages of test method

Disadvantages of test method

Chemical reaction with tissue components or cellular structures (Membrane, microtubuli, myelin, DNA, ribosomes, intermediate filaments)

Simple readout, robust methodology, established in most pathology laboratories. Mostly very cost-effective, fast and reproducible

Usually not specific to a target (neurofilaments, myelin etc.). Some of the more complex methods (silver stain) can be more time-consuming and costly than antibodies (e.g. phosphorylated tau)

Muscle: ATPase, succinate dehydrogenase

Histochemical (enzymatic) reaction

Sensitive, cell specific, enzyme-specific

Requires frozen material, limited to a few enzymatic tests; requires highly specialised technical skills

Protein, differentiation marker

Astroglial: GFAP Oligodendroglial: Olig2 Neural, neuroendocrine: synaptophysin Epithelial: cytokeratin B cells: CD20 T cells: CD3

IHC

Simple readout/interpretation, affordable, scalable (IHC machine)

Can have limited lineage selectivity; may capture aberrant expression

Exogenous microorganisms

Expression of antigen

Toxoplasma Cytomegalovirus JC virus (nuclear accumulation)

IHC

Detection of single affected cells, Robust validation essential; control rare presence of microorganism. tissue may be difficult to source Very easy readout, reliable, usually specific

Protein folding and metabolism

Accumulation of misfolded protein

Amyloid-β: plaques in Alzheimer’s disease, vessel walls in cerebral amyloid angiopathy

IHC

Simple, affordable

Hyperphosphorylated tau: Neurofibrillary tangles in Alzheimer’s disease, neurons or glial cells in other tauopathies

Semiquantitative; difficult to validate. In neurodegenerative diseases; not specific to misfolded forms of disease proteins

Alpha-synuclein: neurons in Parkinson’s disease, dementia with Lewy bodies, multiple system atrophy Prion protein: synaptic and extracellular deposits in all forms of prion diseases TDP 43: Intracellular inclusions in certain forms of frontotemporal dementias and motor neuron disease

0005727399.INDD 58

02-12-2024 17:58:47

Detected feature, mutation or pathology

What is detected

Examples

Deletion, point mutation

Loss of protein expression

ATRX (a proportion of IDH, histone, BRAF mutant tumours) SMARCB1 (AT/RT) SMARCB4 (AT/RT)

Gene amplification

Increase of protein expression

Gain of genetic material; copy number changes, gene amplification

Gene amplification

Loss of expression

Point mutation

Technique or technology

Advantages of test method

Disadvantages of test method

IHC

Simple readout/interpretation, affordable, scalable (IHC machine)

Robust validation essential; control tissue may be scarce

EGFR in glioblastoma MYC in various tumours

Immunohistochemistry (IHC)

Simple, affordable

Imprecise, Semiquantitative Difficult to validate

PDGFR EGFR MYC CDK4

Array comparative genomic hybridisation (CGH)

Easy interpretation

Outdated methodology

Array output 850k

Easy interpretation. Part of other investigations

Limited resolution

Quantitative (q)PCR

Easy readout, inexpensive

Validation for each target required

Next-generation sequencing

Usually part of wider investigation

Expensive

MLPA

Easy to set up the laboratory

Limited flexibility

FISH

Relatively easy readout, spatial accuracy

Expensive, single target

Loss of protein expression

Ini1 immunostaining to detect SMARCB1 loss IHC p16 immunostaining to detect CDKN2A/B deletion ATRX immunostaining to detect mutations Histone H3 K27me3 loss of trimethylation (ependymoma subtypes A, histone mutant gliomas)

Usually simple readout, easy to implement in pathology test laboratory, cost-effective, fast

Not all mutations result in loss of protein expression, leading to possible false negative results

Amino acid change

IDH1 R132H (astrocytoma, oligodendroglioma) H3 K27M (midline glioma) BRAF V600E (ganglioglioma, PXA)

Usually simple readout, easy to implement in pathology test laboratory, cost-effective, fast

Not all mutations detectable (IDH); antibody quality/affinity essential to ensure robust result

Loss of function

ATRX (see above), INI-1

No detection of silent mutations

Gain of function

β-catenin (mutation in exon 3 results in stabilisation of protein and nuclear localisation)

Specificity not always unequivocal

Loss of genetic material

CDKN2A/B deletion SMARCB1 loss

IHC

(continued )

0005727399.INDD 59

02-12-2024 17:58:47

Table 4.1 (continued) Detected feature, mutation or pathology

Fusion mutations

What is detected

Examples

Nucleotide change

IDH1, 2 Histone H3 K27, G34 BRAF V600 TERT promoter FGFR1 HIST1H3B K27

Gene fusion

BRAF:KIAA C11orf95-RELA fusion C11orf95-YAP fusion FGFR fusions

Technique or technology

Advantages of test method

Disadvantages of test method

Sanger sequencing

Robust, inexpensive

Limited to small numbers of target

Pyrosequencing

Technology slowly phased out; expensive sequencer

lengths of individual reads of DNA sequence less then Sanger sequencing

MLPA

Test large panel of target

Set up and validation more complex and test kits are not flexible

Next-generation sequencing

Covers a large number of targets Requires significant infrastructure and (often customised, 200–500 bioinformatics support targets); comprehensive readout

Exome sequencing

Returns information on all coding DNA sequences

Requires library preparation and multiplexing (simultaneous sequencing of large numbers) making it less suitable for diagnostic purposes

RT-PCR

Easy to set up, robust, can be established in small facilities

Limited to a few targets, requires validation of individual targets

Next-generation sequencing

See above

See above

FISH

Easy to set up, robust, can be established in histopathology laboratories

Expensive, time-consuming, variability in readout, low throughput

Set up in his pathology laboratory workflow

Specificity of some translocated proteins may be limited in readout (e.g. P 65) may require further testing

Fusion protein

Fusion product

STAT6-NAB1 fusion: nuclear translocation of STAT6 RELA (P65) nuclear expression

IHC

Genome wide changes

Methylation profile

Established methylation classes, based on combinations of cell of origin, and tumour driver mutations

Methylation array (Illumina Robust technology, works on 450 K, 850 K); FFPE material, identifies alternatively: bisulphite histologically ambiguous cases sequencing, nanopore sequencing

Relatively costly, requires batching, not possible without access to algorithmic classification tool

Subcellular structures, virus particles, organelles

Existence of Viral inclusions structures, Mitochondrial structure, mitochondrial morphological aggregates changes, quantitative changes

Electron microscopy

Lack of specificity (viral particles); expensive equipment, decreasing use

Precise morphological recognition and location

AT/RT, atypical teratoid rhabdoid tumour; FFPE, formalin-fixed, paraffin-embedded; FISH, fluorescence in situ hybridisation; GFAP, glial fibrillary acidic protein; JC, John Cunningham; MLPA, multiplex ligation-dependent probe amplification; PXA, pleomorphic xanthoastrocytoma; RT-PCR, reverse transcriptase polymerase chain reaction; TERT, telomerase reverse transcriptase.

0005727399.INDD 60

02-12-2024 17:58:47

Neuropathology: Introduction to History, Diagnostic Approaches, Techniques and their Interpretation  61

various sequencing methods (single-­target Sanger sequencing as well as panel sequencing, a next-­generation sequencing method) and for detection of changes in DNA methylation, which has developed into a mainstream neuropathology brain tumour diagnostics in developed countries, used in combination with the use of a methylome classifier. FFPE-­derived RNA can be used for expression profiling, for example by exome sequencing, or for targeted RNA sequencing panels for diagnostic purposes. The archiving of fresh frozen tissue samples is, however, essential if long-­read sequencing approaches are pursued, for example for whole genome sequencing or the nanopore technology. The latency between obtaining a tissue sample and its frozen archiving is an increasingly important factor when in particular RNA quality and integrity is paramount, for emerging technologies such as spatial transcriptomics and similar in situ RNA detection techniques. The most common molecular diagnostic techniques in neuro-­oncology are explained below. Techniques for the Visualisation of Tissue Components The purpose of staining techniques is to visualise structures in tissue samples (See also Table 4.1). The most common diagnostic tool in diagnostic cellular pathology is the light microscope, while research applications often also use fluorescent techniques to visualise multiple stains simultaneously on a single section, typically to identify co-­localisation or spatial relationship of different antigens. Electron microscopy is a superb analytical method in diagnostic neuropathology to identify cellular ultrastructure, but a wide range of diagnostic antibodies and molecular genetic tests (e.g. for the diagnosis of peripheral neuropathies, muscle disorders or brain tumours) have emerged to replace this technique for diagnostic purposes.

Diagnostic Light Microscopy

Diagnostic light microscopy uses upright microscopes in which a white light source transilluminates the object (i.e. the light path passes through the object, a tissue section, and is collected by an objective lens). Diagnostic microscopes have dry lenses and cover a wide range of magnifications. The 1×, 2× or 2.5× and 4× or 5× objectives are ideal for specimen overview (Figure 4.2i,j) and are typically used to identify anatomical structure, and the context of pathologies such as demyelination, necrosis, patterns of inflammation, macrophages and tumour architecture. The 10× and 20× objectives are best for studying histological patterns such as cellular processes, nuclear shapes and larger cellular inclusions (Figure 4.2k). The 40× and 60× objectives are necessary for a detailed study of cellular and subcellular structures and subtle pathologies, such as demyelination, or individual unmyelinated axons on resin semi-­ thin sections, nuclear detail or cytoplasmic inclusions (Figure 4.2l). An additional magnification is required for the actual visualisation to the human eye and this is achieved with eyepieces, with 10× or 12× magnification in most diagnostic microscopes. The total magnification, as seen by the pathologist, is calculated by multiplying the objective and eyepiece magnification; for example, a 12× eyepiece and a 20× objective result in a visible magnification of 240×. This is often incorrectly applied to printed photographs of histological images, where the same combination of objective and eyepiece is assumed, whereas a scale bar is necessary to correctly identify the scale of a reproduced image, in particular when publications are viewed in a browser on a computer screen (Figure 4.2i-­l).

Fluorescence Microscopes

Fluorescence microscopes are typically used in research ­applications, and they have a role in diagnostic histopathology or

neuropathology only in very specific circumstances. Fluorescence microscopes are engineered to detect fluorescent dyes, applied either directly on tissues (e.g. nuclear detection) or linked to antibodies used to detect specific antigens. The light source in fluorescence microscopes produces a broad range of wave lengths, from which light of a specific wavelength is filtered with specifically engineered filter sets. The light is conducted through the objective (unlike the light path in light microscopy, where the light passes through the specimen) and is then projected on to the specimen (‘epifluorescence’). On the specimen, it excites the fluorophore, which emits light of a longer wavelength (emission wavelength), and this light is returned through the light path in the objective, but is diverted by a dichroic filter set to the eyepiece or a digital camera. For example, the fluorescent dye fluorescein is excited by a blue light and emits green light (i.e. a longer wavelength). The main purpose of fluorescence microscopy is the detection of multiple distinct structures, such as nucleus, cytoplasm and cell processes, with a blue, green and red fluorophore, to identify spatially distinct or co-­localising structures (e.g. Figure 4.2h). A technical modification of the fluorescence light microscope is the confocal laser scanning microscope, which uses laser beams to generate fluorescent signals.

Electron Microscopy

Electron microscopy can be used for diagnostic and research applications. It requires tissue fixation with glutaraldehyde, a much stronger crosslinking agent than formaldehyde, and subsequent processing of the fixed tissue into a resin polymer and the preparation of initially semithin resin sections (0.5  μm thick) to identify relevant tissue structures, and then ultrathin sections (100 nm) that are contrasted with heavy metal and placed on a grid. This grid is placed into the electron microscope, where an electron beam is transmitted through the specimen and generates an image. The image is magnified and focused on to an imaging device, which in the past was photographic film and today is a charged-­coupled device camera. This technique has declined in popularity because of relatively high technical infrastructure costs and the time needed for of specimen preparation. It still has a firm place in nerve biopsy diagnostics, usually for the assessment of myelin sheaths (Figure 4.2g) and a few other selected applications. Much of electron microscopy has been replaced by the increasing number of excellent antibodies and by the recent advances in molecular diagnostics. For example, electron microscopy has a resolution that can identify viral particles, but its relevance has declined as next-­generation sequencing techniques are able to detect these pathogens more rapidly, with lower costs and far greater precision. Another relevant example is the near obsolescence of electron microscopy in the diagnosis of genetic neuropathies, which are now diagnosed by genetic testing of the patients. In research, multicolour fluorescence and recent development in high-­resolution fluorescence microscopy have replaced electron microscopy in many applications. In tumour diagnostics, the detection of subcellular structures by electron microscopy has been replaced by next-­generation sequencing panels for the detection of diagnostic mutations or by the use of methylation arrays to determine a tumour type. In nerve biopsy diagnostics, electron microscopy is still used to detect subtle changes in the myelin sheath (Figure 4.2g), for example in the context of immunoglobulin-­ related neuropathies. Digital Pathology Digital pathology is used for research, teaching and clinical diagnostic applications. It requires the photographic capturing

62

Neurology: A Queen Square Textbook

of histological slides using a specialised scanning device (‘slide scanner’) and their subsequent transformation into digital data files. This entails stitching together thousands of individual tiles and computing into additional layers to simulate intermediate magnification steps (pyramid file format). This postprocessing generates images that can be interactively viewed at different magnifications, ranging from a low-­power overview to a highly detailed image (the different magnifications in Figure 4.2i–l were generated with such processed images). Currently, with the acquisition of the digital image with a 40× objective, the typical resolution of digital images correspond to 0.25  μm per pixel. Depending on the image compression (e.g. using JPEG), the image of a 10 × 10 mm histological sample creates data volumes of 500  Mbyte to 1.5  Gbyte. A large pathology department produces between 250,000–500,000 slides annually, and the transformation of these slides into digital data files can create very substantial volumes of data, in the range of hundreds of terabytes annually. Such demands on data storage, as well as comparably expensive scanning devices and a complex infrastructure for the viewing software and their integration into clinical patient record and pathology databases, require substantial investments into digital infrastructure. The advantages of digital pathology are the facilitation of remote working, remote diagnosis and conferencing (telepathology), rapid access to past specimens from the same patient, automated digital slide analysis (for example Ki-­67 labelling index or identification of mitotic figures), data generation for research (artificial intelligence and machine learning algorithms) or for teaching applications. These benefits have to be weighed against high cost for hardware, software and storage. It is thought that with decreasing cost for data storage, affordable computer monitors for adequate, clinically safe viewing of the images and decreasing cost for slide scanners, such solutions can be gradually introduced into clinical practice.

Interpretation of Histological Samples and Systematic Approach to Diagnosis

The diagnostic approach can be summarised in a flow diagram, as shown in Figure 4.4, linking a cellular component, tissue structure or pathological feature with a detection method, resulting in a morphological or molecular (in silico) readout. Morphological assessment of a tissue section is always the first and essential step in diagnostic neuropathology. Initial examination at low-­ power magnification (Figure  4.2i,j) shows the anatomical structure, overall cellular density, and also pathologies in tissue substructures such as white matter, grey matter, abnormalities in blood vessels, oedema or haemorrhages. In the diagnostic approach to neoplasms, at low-­ power magnification pathological hallmarks such as papillae, rosettes, infiltration margins and perivascular structures become visible. In neuromuscular diagnostics, patterns of muscle fibres, atrophy, degeneration and inflammation are assessed. At intermediate (Figure  4.2j,k) and high magnification (Figure  4.2l), the neuropathologist identifies intracellular inclusions, extracellular protein deposits or, in tumour diagnostics, nuclear or cellular inclusions, morphology of tumour cell processes, mitotic figures and vascular abnormalities. In neuromuscular diagnostics (e.g. mitochondrial or tubular aggregates) can be identified at high magnification. Detection of such patterns and pathologies allow a preliminary diagnosis in most instances, helps in formulating a hypothesis and informs the next diagnostic steps. The most common subsequent examination in diagnostic neuropathology is the preparation of an immunohistochemical stain.

The choice of immunohistochemical stains depends on the diagnostic differential diagnosis. To perform such stains, the laboratory has to prepare additional (usually serial) sections, each mounted on a separate glass slide (Figure 4.1IIIc). Each slide therefore contains an adjacent section from the tissue block. Individual sections are stained with antibodies directed against specific antigens. For example, the diagnosis of low-­grade gliomas usually involves an antibody detecting a mutant protein (IDH1 R132H, Figure 4.3m) and the assessment of retention or loss of another diagnostic protein ATRX (Figure  4.3p). Midline gliomas will be usually tested with an antibody against histone H3 K27M (Figure 4.3n). Markers for the diagnosis of neurodegenerative disease include amyloid β aggregates (Figure 4.3i) and hyperphosphorylated tau (Figure 4.3I) and, ideally, also TDP 43, p62, α-­synuclein (Figure 4.3k) and abnormal prion protein (Figure  4.3l). Inflammatory lesions instead are examined with antibodies against T cells, B cells, plasma cells and macrophages (Figure 4.3g) to establish a profile of the inflammatory infiltrate. Molecular tests can comprise a wide spectrum of methods interrogating genetic and epigenetic changes in a tissue sample. A number of recurring diagnostic and prognostic biomarkers can already be tested with mutation-­specific antibodies; for example, IDH1 R132H (Figure 4.3m) in the majority of astrocytoma and oligodendroglioma, BRAF V600E in a range of low-­grade gliomas such as ganglioglioma and pleomorphic xanthoastrocytoma, histone H3  K27M (Figure  4.3n) in midline gliomas or histone H3 G34R (Figure 4.3o) for hemispheric high-­grade gliomas in teenagers and young adults. However, such antibodies are available only for common mutations, while less common mutations will need detection either by targeted Sanger sequencing or by next-­ generation sequencing using larger panels to detect multiple distinct mutations. Certain types of mutations, such as gene fusions and copy number changes, require FISH, real-­time polymerase chain reaction (PCR) or an array technology (see also Table 4.1). A relatively recent development is the assessment of epigenetic changes in brain tumours. Epigenetic changes are caused by the methylation of neighbouring nucleotides C (cytosine) and G (guanine), separated by only one phosphate group linked by a phosphodiesterase bond, a so-­called CpG site. Clusters of CpG sites are called CpG islands. CpG sites can be methylated or unmethylated, and this feature can be interrogated with a methylation-­sensitive microarray. The global distribution of methylated or unmethylated CpG sites is not only different between normal tissue and cancer tissue but also shows characteristic patterns depending on the brain tumour type, which is linked to a cell of origin and influenced by the type of mutation. The readout and computational analysis using machine learning algorithms has led to a revolutionary change to the way brain tumours are diagnosed. Again, this is an example how neuropathology leads the field and is currently ahead of other pathology subdisciplines in applying novel technologies for advanced diagnostics.

Common Techniques in Molecular Diagnostics in Neuro-­Oncology

Sanger Sequencing Sanger sequencing is an economical, easy to implement and informative test, particularly to identify a small number of defined point mutations of diagnostic or prognostic relevance. This test usefully complements the use of mutation-­ specific antibodies, where multiple variants can exist on a single codon, as it is the case for the IDH1 or IDH2 genes. Another example is the telomerase

Structures or pathological features

Specific example

Detection method

Diagnostic readout

Nuclei Cytoplasm Axons

Myelin

Cellular components

Tinctorial

Amyloid beta Neurofibrillary tangles Histochemical Fungi

Protein misfolding

alpha-Synuclein Bacteria

Electron microscopy

Prion protein Viral particles Micro-organisms

Morphological

JC Virus Herpes simplex virus Toxoplasma

Immunohistochemistry

Glial origin Neuronal origin Cell lineage

Enzymic activity Gene fusion

Epithelial origin ATPase, NADH RELA fusion

Enzymatic reaction

Methylation array

ATRX mutation Point mutation

Fluorescence in situ hybridisation (FISH) IDH mutation

In silico Next generation sequencing Sanger sequencing

Gene amplification

EGFR amplification

Epigenetic changes

MAPK alteration

Multiplex ligation-dependent probe amplification (MLPA) Array CGH Quantitative PCR

Figure 4.4  Alluvial diagram showing structures or alterations defining a pathological feature in different ways, on the left. Each type of cellular property, component, lesion or other abnormality is then linked to a typical example (See also Table 4.1). This further feeds into the category of detection methods, which can be used to detect such abnormalities. For example, the cellular component ‘axons’, or ‘myelin’ can be detected by three methods: histochemistry, ultrastructural (electron microscopy) and immunohistochemistry. Another example is the determination of the cellular lineage, where a glial or neuronal origin of a tumour can be determined by immunohistochemistry (GFAP, synaptophysin) or by means of a methylation array that also incorporates the tumour cell of origin or lineage. The detection methods shown here show two distinct readouts, either morphological (i.e. visible on a tissue section, through a microscope or digital image) or molecular, in the form of data readout from a sequencing or array-­based technology (in silico). Importantly, some molecular tests can be read by morphological means; for example, detection of a point mutation with a mutation-­specific antibody or gene losses or amplifications with fluorescence in situ hybridisation. The purpose of this relationship diagram is the illustration of common detection methods but it may not be always exhaustive.

64

Neurology: A Queen Square Textbook

reverse transcriptase promoter mutation, which does not encode for a protein and thus cannot be detected with an antibody. This mutation is highly informative in the context of other genetic alterations for the diagnosis of oligodendroglioma or glioblastoma. However, Sanger sequencing is economical and useful only for a small number of targets. If mutations are not found in a hotspot but can be disseminated across the open reading frame or a larger number of mutations need to be interrogated, next-­generation sequencing is a better choice. Fluorescence or Chromogenic In Situ Hybridisation Fluorescence (FISH) and chromogenic in situ hybridisation (CISH) techniques were established several decades ago and use DNA probes labelled with a fluorochrome (to be detected on a fluorescence microscope) or a chromogenic dye (visualised on a light microscope). These probes hybridise to target DNA in tissue sections or cell preparations. This tissue-­based method is still in use in many laboratories and particularly popular with pathologists as it can be read out by microscopy. The FISH or CISH probes are comparably expensive and the visual screening time consuming, making this method uneconomical to screen multiple targets. Therefore, FISH or CISH are increasingly replaced by more informative copy number assays, which can be read out in panel sequencing (next-­generation sequencing) assays or from methylation arrays. Multiplex Ligation-­Dependent Probe Amplification The multiplex ligation-­dependent probe amplification PCR method is suitable for the detection of copy number changes and of single nucleotide polymorphism or mutations. This technique is relatively popular in many pathology laboratories, as it requires a thermocycler (‘PCR machine’) and capillary electrophoresis equipment. It is an economical alternative for array-­based techniques, in particular for recurring, defined mutations and copy number changes. The disadvantage of this essay is its lack of flexibility, as the multiplex assays are predeveloped and cannot be easily modified without major validation steps. Next-­Generation Sequencing Next-­ generation sequencing technology is based on Sanger sequencing and results in single-­base resolution of DNA samples. In contrast to traditional Sanger sequencing, next-­ generation sequencing covers millions of fragments in parallel. A disadvantage is the necessity for a time-­consuming and relatively costly library preparation. The technology generates a significant amount of data, requires high-­volume data storage and expert knowledge and software to read and interpret the results. Therefore, this technology is usually set up in core facilities, such as the genome laboratory hubs established by the NHS in England to provide widespread and equal access to high-­quality molecular diagnostics for all patients. The advantage of next-­generation sequencing is a wealth of data, which can also be reviewed and analysed at later stages. Pyrosequencing Pyrosequencing is a method of DNA sequencing relying on the detection of pyrophosphate release and the generation of light on nucleotide incorporation, rather than chain termination with dideoxynucleosides, used in Sanger sequencing. The advantage is a more quantitative readout and a superior limit of detection of mutant alleles, whilst the read length is shorter, typically 100–400 bases. The most common application in neuropathology molecular ­laboratories is MGMT (O6-­methylguanine-­DNA methyl-­transferase)

promoter methylation quantification. This method is not widely used, however, and relies on comparably expensive equipment. Methylation Array Methylation array-­based technology interrogates in its current format (Illumina Infinium methylationEPIC 850K) approximately 850,000 CpG sites. It has become a tool of major importance for brain tumour diagnostics. The approach is fundamentally different from sequencing, as it interrogates surrogate markers of mutations in combination of cell of origin. A key advantage is the simplicity of the sample preparation (bisulphite conversion), the small amount of starting material needed (starting from 200  ng DNA) and the wealth of information that can be derived from the data. These arrays generate comparably small volumes of data (30 MB per sample). These data need to be further processed through a classifier to compare an individual sample with a reference dataset. This step can be performed by technical staff or pathologists within less than 30 minutes for a batch of eight samples. The disadvantage for many laboratories is the expensive chip reader, requiring a core facility. However, the infrastructure is the same as those required for next-­ generation sequencing technologies. A further disadvantage is the time to prepare samples, hybridise the arrays and complete the classification step, which often can take two to three weeks, even in high-­throughput laboratories. Nanopore Technology Nanopore technology is a relatively new technology that is in development and may be ready for clinical application in a few years. It is based on a long-­read approach whereby high molecular weight DNA is sequenced in real time as it passes through a nanometre-­ sized pore. Samples can be run in real time until sufficient data are captured. Reading long sections of DNA makes assessment of structural variants more accessible and it can be used flexibly, either as a single ad hoc sample or as multiplex sample. Unlike methylation arrays, DNA does not require bisulphite conversion, as CpG methylation status can be read directly from native DNA. A proof of principle has been published and can under optimal experimental settings provide a diagnosis after 6–12 hours. Currently, this technology is being refined and developed in multiple centres, aiming at introduction into clinical practice.

Reference

Coons AH, Creech HJ, Jones RN. (1941). Immunological properties of an antibody containing a fluorescent group. Exp Biol Med 47: 200–202.

Further Reading

Bechet D, Gielen GG, Korshunov A et  al. (2014). Specific detection of methionine 27 mutation in histone 3 variants (H3K27M) in fixed tissue from high-­grade astrocytomas. Acta Neuropathol 128: 733–741. Capper D, Jones DTW, Sill M et al. (2018). DNA methylation-­based classification of central nervous system tumours. Nature 555: 469–474. Capper D, Preusser M, Habel A et al. (2011). Assessment of BRAF V600E mutation status by immunohistochemistry with a mutation-­specific monoclonal antibody. Acta Neuropathol 122: 11–19. Capper D, Zentgraf H, Balss J et  al. (2009). Monoclonal antibody specific for IDH1 R132H mutation. Acta Neuropathol 118: 599–601. Coons AH, Kaplan MH. (1950). Localization of antigen in tissue cells; improvements in a method for the detection of antigen by means of fluorescent antibody. J Exp Med 91: 1–13. Euskirchen P, Bielle F, Labreche K et  al. (2017). Same-­day genomic and epigenomic diagnosis of brain tumors using real-­time nanopore sequencing. Acta Neuropathol 134: 691–703. Fagreus A, Harboe M, Natvig JB. (1979). Albert Coons, 1912–78. Scand J Immunol 9: 395–396.

Neuropathology: Introduction to History, Diagnostic Approaches, Techniques and their Interpretation  65

Hancock WW, Becker GJ, Atkins RC. (1982). A comparison of fixatives and immunohistochemical technics for use with monoclonal antibodies to cell surface antigens. Am J Clin Pathol 78: 825–831. Jaunmuktane Z, Capper D, Jones DTW et  al. (2019). Methylation array profiling of adult brain tumours: diagnostic outcomes in a large, single centre. Acta Neuropathol Commun 7: 24. Köhler G, Milstein C. (1975). Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 256: 495–497. Kvastad L, Carlberg K, Larsson L et al. (2021). The spatial RNA integrity number assay for in situ evaluation of transcriptome quality. Commun Biol 4: 57.

Louis DN, Wesseling P, Brandner S et al. (2020). Data sets for the reporting of tumors of the central nervous system: recommendations from the International Collaboration on Cancer Reporting. Arch Pathol Lab Med 144: 196–206. Schweizer L, Koelsche C, Sahm F et al. (2013). Meningeal hemangiopericytoma and solitary fibrous tumors carry the NAB2–STAT6 fusion and can be diagnosed by nuclear expression of STAT6 protein. Acta Neuropathol 125: 651–658. Suvarna KS, Layton C, Bancroft JD. (2012). Bancroft’s Theory and Practice of Histological Techniques, 8th edn. Amsterdam: Elsevier Health Sciences.

CHAPTER 5

Neuroimaging Frederik Barkhof1,2 and Francesco Carletti1 1 2

National Hospital for Neurology and Neurosurgery, London, UK Brain Repair and Rehabilitation, University College London, London, UK

Clinical Scenarios in Neuroradiology

The starting point of the neuroradiological consultation often is the  request form. This should provide detailed clinical data on ­history, symptoms and a suspected (differential) diagnosis. When the ­information is incomplete, additional information may need to be requested from the referring physician and might unnecessarily delay scheduling or lead to a suboptimal protocol choice. In certain cases, it is sufficient for the request form to only provide limited information such as ‘trauma’, ‘stroke’ in the acute setting, or in the non-­acute setting ‘worsening of headache’, ‘first insult’ or ‘memory clinic’. More commonly, specific symptoms and signs should be provided that are needed to determine a specific location or type of lesion, even if this can be misleading as the culprit may be located elsewhere (e.g. in the cerebral hemispheres rather than the brainstem, or in the cervical rather than the thoracic cord). The source of referral (primary care physician, medical specialist) and age/­ gender, as well as the medical history of the patient also provide information that will help to determine the differential diagnosis. The collective information available will determine whether the chance of finding pathology is low or high; in other words, whether the goal of the examination is to ‘rule out’ or to narrow down a specific differential diagnosis.

Neurological Imaging

Plain Film Radiography Plain film radiography uses x-­rays, ionising radiation with a short wavelength but high energy, to produce images of the body. When x-­rays travel through the tissues of the body, they absorb and attenuate the primary x-­ray beams depending on the electron density of the tissue. A digital detector on the other side of the body picks up the x-­rays and turns them into an image based on attenuation. When x-­rays pass through air, they are minimally attenuated and produce a black image. X-­rays going through bone are heavily attenuated, giving a white appearance to bones and metal objects. X-­rays are cheap, widely available, easy and quick to perform, and the radiation dose is (much) lower than for computed tomography (CT). Spinal x-­rays are frequently used in preoperative or postoperative assessment, as they can show the integrity and position of metalwork, prostheses or shunts. They are also used in the evaluation of spinal deformities, fractures, congenital abnormalities or destructive lesions. Skull x-­rays have been largely superseded by

CT and magnetic resonance imaging (MRI), which offer a greater soft-­tissue contrast. They are still sometimes used to rule out the presence of foreign bodies or in the management of programmable shunt valves or in shunt series. Computed Tomography Computed tomography uses x-­rays to give cross-­sectional images of the body. The patient is moved slowly on a bed through the centre of the scanner while x-­rays are being emitted by a rotating tube around the patient. A CT scan may have one or several rings of detectors around the patient which collect (attenuation of) the transmitted radiation. A computer then processes the data acquired by the detectors to form an image. Modern CT scanners are open and quiet. The duration of the scan varies depending on the area of the body to be scanned but is  generally quick. Multislice or multidetector CT scanners have ­multiple rows of detector rings and allow faster scanning (seconds only for a brain scan). Radiation doses depend on the examination protocol but are much higher than plain films. Some examinations require the intravenous administration of iodine-­based contrast media, which aims to improve the contrast between tissues of the body by increasing the density (attenuation) of tissues they opacify. CT angiograms or venograms use intravenous iodine contrast material to visualise arteries or veins and allow identification of the occlusion or stenosis of arteries and veins and the characterisation of aneurysms and vascular abnormalities (Figure 5.1). CT cisternograms and myelograms use iodine-­based contrast agents to depict the subarachnoid space around the brain or spine (e.g., to detect dural leaks). The intravenous administration of iodine-­based contrast media comes with potential risks of renal impairment and allergy. Magnetic Resonance Imaging MRI is a non-­invasive medical imaging technique that produces images of the body using a large magnet and radiofrequency waves. MRI is particularly suited to image the nervous system because of its great soft-­tissue contrast resolution, high spatial resolution and multiplanar capability. Unlike CT, MRI does not use ionising radiation and is far superior to CT in the assessment of the posterior fossa structures, meninges and cranial nerves and spinal cord. The downside of MRI is that each examination is long, some patients might find the scanner claustrophobic or move too much during the examination.

Neurology: A Queen Square Textbook, Third Edition. Edited by Robin Howard, Dimitri Kullmann, David Werring and Michael Zandi. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd. 67

68

Neurology: A Queen Square Textbook

CT

CTA

Figure 5.1  Plain computed tomography (CT) showing hyperdense subarachnoid haemorrhage caused by a basilar top aneurysm revealed by subsequent CT angiography (CTA).

Our body is almost entirely made of water molecules, each of which contains two hydrogen atoms, each containing a single ­proton. When a patient is placed in the MRI scanner, the protons (hydrogen nuclei) align with the magnetic field. During an MRI scan a radiofrequency pulse is transmitted, causing the protons to alter their alignment relative to the field and then return to the original alignment with the magnetic field. These alignment changes create a magnetic vector which is detected by the receiver coil. With the help of additional magnetic fields generated by switching on and off gradients coils (this causes loud knocking sounds), the MRI scanner records the locations of signals with high precision. The captured signal is processed by a computer and converted to a ­greyscale image. By changing the parameters on the scanner, one can create ­contrast between different types of body tissue, using so-­called pulse-­sequences. The most common ones are T1- and T2-weighted sequences. One can discriminate these two easily because fluid in the cerebrospinal fluid (CSF) will appear hypointense (dark) on ­T1-­weighted images and hyperintense (bright) on T2-­weighted images, while adipose tissue will display high signal intensity on both T1- and T2-weighted images. The subject of MRI sequences is broad, and a detailed explanation of each sequence is beyond the scope of this chapter. It is worth mentioning here that sometimes the signal from fat is suppressed (saturated) to allows fluid to stand out. Quite frequently, we use sequences like fluid attenuated inversion recovery (FLAIR) where the signal from CSF is suppressed, to highlight parenchymal lesions. Diffusion-­weighted imaging (DWI) may be used to identify areas of the nervous system that have recently been damaged or injured, for instance by ischaemia or infective process. One important property of DWI is that it is much more sensitive than T2-­weighed images (or CT) in detecting cytotoxic oedema caused by acute ischaemia and therefore is used routinely in patients with transient ischaemic attacks or suspected stroke. Susceptibility-­weighted images (SWI) are used to investigate the presence of intracranial haemorrhages

of  variable size, thrombi, vascular malformations and venous congestion. Most MRI protocols include T1-­and T2-­ ­ weighted imaging, FLAIR, DWI and SWI. Contrast agents (often containing a paramagnetic substance called gadolinium) may be given to a patient intravenously before or during the MRI to increase the speed at which protons realign with the magnetic field thereby increasing the T1 signal and highlight disruption of the blood– brain barrier. Magnetic resonance angiography (MRA) may be used in patients with a suspected stroke who cannot have a CT angiogram or require follow-­up imaging. MRA is an important non-­invasive tool in the evaluation of the vasculature of the nervous system. In the past two decades, there has been significant development in the field of MRA that goes beyond the time-­of-­flight MRA to include contrast-­ enhanced MRA and time-­resolved (four-­dimensional) MRA for the study of fistulas and vessel wall imaging for vasculitis. To image the venous system, magnetic resonance venograms and postcontrast T1-­weighted images can provide information on venous thrombosis, and are most frequently used in non-­acute settings. Additional complex techniques include magnetic resonance spectroscopy (MRS), perfusion-­weighted (PW) MRI, diffusion tensor imaging (DTI) and functional MRI (fMRI). These non-­invasive magnetic resonance techniques enable detailed physiological and functional imaging of the brain. MRS measures concentrations of various compounds in the brain. It may be a useful tool in the ­investigation of tumorous lesions, metabolic disorders or brain infections. DTI and tractography investigate the brain and cord microstructure and aims to identify the white matter tracts that could not be seen with conventional sequences. Functional MRI visualises salient cortical areas (e.g. motor, language, visual, memory based) and provides a map of cortical activity. The clinical applications of fMRI and DTI include presurgical assessment in the case of a brain tumour or epilepsy; for instance, in the assessment of language centre localisation and hemispheric dominance. Currently, resting-­state fMRI remains predominantly a research tool.

Neuroimaging  69

PW imaging enables the assessment of brain tissue perfusion. Two of the techniques require intravenous contrast administration: dynamic susceptibility contrast (DSC) magnetic resonance perfusion and dynamic contrast-­enhanced (DCE) perfusion. A third technique, called arterial spin-­ labelled (ASL), is completely non-­ invasive because measures tissue perfusion using a magnetic proton-­labelling technique rather than gadolinium. DSC-­MRI uses gadolinium and a T2*-­weighted echo-­planar imaging technique to estimate cerebral blood volume and flow, and the p ­ erfusion parameters such as time to peak and mean transit time. In DCE-­MRI (sometimes called permeability MRI), the time course of the enhancement is used to calculate perfusion parameters (e.g. in tumours), such as ktrans and time to peak enhancement. Perfusion imaging can be used in the presurgical assessment of brain tumours or in the differentiation between a recurrent tumour from treatment-­related changes in patients with glioma. Emerging c­ linical applications of ASL are in cerebrovascular diseases, d ­ ementia and epilepsy. Ultrasound Imaging A device called a transducer is used to emit high-­frequency sound waves (1–20 MHz) and record their echoes (reflected sound waves) from tissues in the body. Echoes are then converted into electric signals and amplified to form an image. Ultrasound is used in newborns (e.g. to investigate hypoxic brain lesions, haemorrhage or congenital malformations) but not so much in adults because the skull and spine prevent the propagation of sound waves towards the brain and spinal cord. Ultrasonography is used as a firstline imaging investigation for neck lumps and allows a safe tissue sampling, offering an attractive alternative to more invasive excision biopsies. A form of ultrasound, called Doppler, is used for the examination of blood flow through cerebral and neck blood vessels. Doppler ultrasound is a quick and non-­invasive method of screening patients

with carotid disease. Doppler ultrasonography can provide  very accurate information about flow inside neck and skull vessels. Transcranial Doppler is also a tool for the assessment and follow-­up of intracranial arterial vasospasm after subarachnoid haemorrhage. The main limitation of this modality arises from its operator-­ dependent nature: ultrasound has high diagnostic accuracy in suitably trained hands but may lead to an incorrect diagnosis if not done properly. Another drawback is that it is difficult to document an ultrasound examination and that a revision of external images cannot be requested. Catheter Angiography Catheter angiography is an invasive imaging technique used to evaluate the intraluminal aspect of vessels of the neuraxis. Sedation is rarely required. A catheter is introduced through the groin in the femoral artery access to reach proximal vessels using fluoroscopic guidance. An iodine-­based contrast material is injected and images in multiple projections are acquired to delineate anatomy and abnormalities. Digital subtraction angiography (DSA) images are generated by ­subtracting native from contrast-enhanced images. Since multiple images can be acquired during injection, DSA has high temporal ­resolution (in contrast to MRA and CT angiography [CTA]) allowing the dynamics of blood flow to be examined. Although non-­invasive imaging (CTA or MRA) often allows a precise diagnosis of a wide spectrum of neurovascular disorders, diagnostic angiography still has several indications and may be required to confirm suspicion of a dural fistula. Cerebral angiography may be used to further investigate patients with vascular ­stenosis or occlusion, aneurysm, vascular malformations, vasculitis and vascular tumours. Catheter angiography allows introduction of microcatheters towards the pathology with subsequent interventions such as embolisation of aneurysms (Figure  5.2) or clot retrieval in stroke patients.

PICA

Vertebral artery

CT (plain)

DSA Vertebral injection

X-ray Coil material

DSA Post-intervention

Figure 5.2  Plain computed tomography (CT) shows hyperdense subarachnoid haemorrhage. Cather angiography was performed with injection of the vertebral artery. Digital subtraction angiography (DSA) image shows an aneurysm (circled) at the origin of the posterior interior cerebellar artery (PICA). Through a microcatheter, coils were inserted that can be seen on the x-­ray image. Post-­intervention DSA image confirms exclusion of the aneurysm with maintained patency of the vertebral artery and PICA.

70

Neurology: A Queen Square Textbook

Spinal angiography is performed to rule ‘in’ or ‘out’ spinal dural arteriovenous fistula, in case of spontaneous spinal haemorrhage and to characterise the vascular supply of tumours prior to surgery and plan preoperative embolisation. Albeit uncommon, angiography might have complications related to the administration of contrast agents (hypersensitivity reactions and contrast-­induced nephropathy), to the percutaneous access (hematomas, femoral pseudoaneurysms, retroperitoneal haemorrhages and infections) or to the intravascular manipulation (stroke, transient ischaemic attacks, haemorrhage, dissections).

Choosing the Right Imaging Modality

The choice of the most appropriate (and cost-­efficient) imaging modality and protocol will depend on the information on the request form, availability of previous examinations (plus laboratory data and clinical letters), the clinical setting (emergency department vs outpatient) and mode of presentation, and local availability of the different radiological techniques. CT is the first-line examination for several acute disorders such as trauma and stroke, while for subacute and chronic modes of presentation, MRI in general is the preferred modality (Figure  5.3). For acute vascular disorders (e.g.  subarachnoid haemorrhage [SAH], acute ischaemic stroke), additional CT angiography is often performed; in the non-­acute setting, MRA can suffice, certainly for follow-­up examinations (e.g. screening for aneurysms or following them after coiling). The indications for diagnostic catheter angiography have decreased over the last few decades and are now confined to a few scenarios such as CTA-­ negative SAH, confirmation of arteriovenous malformation (AVM) or dural arteriovenous fistula (dAVF), and DSA is now almost exclusively used in the interventional setting. Interventional neuroradiology techniques include thrombectomy (stroke), coiling and stenting (aneurysm), glue/particle injection (AVM/dAVF), blood patch injection (CSF leaks) and

nerve root injection. Intrathecal contrast injection these days is usually f­ ollowed by CT (myelography and cisternography) and can be used to detect CSF leaks if MRI fails to reveal them. Ultrasound has very limited value examining the brain and spine in adults and is used for examination of plexus and peripheral nerves. Additionally, ultrasound can be used to monitor for intracranial vasospasm following subarachnoid haemorrhage. Nuclear medicine techniques such as positron emission tomography and single-­photon emission CT can be used to map receptor density (e.g. dopamine) or to visualise abnormal protein deposition (e.g. amyloid). Plain x-­rays are still used in the assessment of spinal deformities and suspicion of drain disconnection. Radiographs with lateral views in flexion and extension are often used to assess for stability of the cervical or lumbar spine. CT is the first choice for the evaluation of spinal fractures in the trauma setting and is widely used in surgical planning or instead of MRI when this is contraindicated or not available. MRI remains the investigation of choice for the spine in a wide range of scenarios, including demonstration of cord lesions in disorders like multiple sclerosis (MS). The advent of open MRI scanners allows a dynamic evaluation of the spine with ­functional loading and provides an insight on in-­vivo spinal biomechanics.

Safety and Contraindications

Radiation Protection CT, fluoroscopy, radiography and nuclear medicine examinations use ionising radiation. Patients undergoing frequent repeat imaging are at risk of radiation-­induced malignancies. To minimise the risk of radiation injury, the diagnostic studies and interventional procedures need to be justified, optimised to obtain diagnostic images using the lowest possible dose, and the number of examinations limited over a given period of time. As patients and referrers are largely unaware of the potential risks associated with medical

Figure 5.3  This 28-­year-­old man had unexplained persistent cognitive disturbances after a car accident. Computed tomography (left) and T2-­weighted magnetic resonance imaging (middle) were normal. Susceptibility-­weighted images (right) revealed numerous cerebral ­microhaemorrhages (arrows).

Neuroimaging  71

i­maging, in the past years, the European Commission, the Royal College of Radiologists and other radiological societies have ­developed appropriateness criteria to guide referrers in choosing radiological examinations for specific clinical scenarios. Whenever possible, MRI should be used for frequent and repeated imaging. Safety of Magnetic Resonance Imaging MRI is generally considered one of the safest medical procedures available as it does not use x-­rays. This does not mean that MRI scans are without risks or that anyone can undergo an MRI scan safely. Most contraindications to MRI are relative precautions. The principal contraindications to MRI are related to the presence of metallic or electrical implants, devices or foreign bodies in a patient. In general, implants are becoming increasingly magnetic resonance safe and an individual risk assessment is recommended. The ASTM International (2023) F2503 guidelines classify medical implants and devices under one of these three categories: 1 MRI safe – the device poses no hazard in all MRI environments. 2 MRI unsafe – the device is not safe in any MRI environment (e.g. ferromagnetic objects such as scissors or electric appliances such as mobile phones). 3 MRI conditional – the device is safe in a specific MRI environment under certain conditions of use: specific magnetic field strength, maximum magnetic field gradient, maximum specific absorption rate and description of the testing conditions. The guidelines for safe MRI of conditional devices are often available on the manufacturer’s website. Nearly all known MRI conditional devices can be identified by model name, number and manufacturer at www.MRIsafety.com. When the implant’s safety profile is unknown (e.g. in the case of legacy implants and devices), the referrer should discuss the indications for examination and alternative diagnostic procedures with the radiologist when the examination is requested. The potential risks of using MRI relate to the use of large static and changing magnetic fields and high-­powered radiofrequency coil systems used in MRI scanners. When metallic or electrical implants, devices or foreign bodies are near an MRI scanner, the patient is exposed to four different kinds of hazards: 1 Translational force and rotational force (torque) injuries. MRI exerts strong forces on ferromagnetic objects such as some ­vascular stents. These forces increase with ferromagnetic composition and mass of objects or devices and with the gradient of the magnetic field strength. Some vascular stents cannot be imaged until six weeks after implantation (when they become securely embedded into the vessels), because they could translate or rotate within the magnetic field. 2 Projectile injuries: even when the scanner is not acquiring images, the main magnetic field is always on and can pull ferromagnetic objects external to the patient into the bore. Examples of objects that can become projectiles are medical support ­equipment (e.g. cylinders filled with anaesthetic gas) or non– MRI-­compatible beds or chairs. 3 Thermal injury: direct skin contact with receiver coil cables or wire leads from cardiac devices can cause burns through ­different mechanisms. Clothing containing metal fibres, body piercing and tattoos with very dark inks (rich in iron oxide) can cause thermal injuries from electromagnetic induction. Dermal drug patches which contain microscopic amounts of conductive material can produce enough heat to cause a burn and should be removed unless declared MRI safe by the manufacturer.

4 Interference with electrical or mechanical components of medical devices. The main magnetic field, the radiofrequency coil and the gradient coils can interact with medical devices’ electrical or mechanical parts. Cardiac pacemakers, implantable cardioverter defibrillators, cochlear or otologic implants and drug infusion pumps are examples of devices that can be damaged during an MRI scan. In the past, cardiac pacemakers were an absolute ­contraindication to MRI. This is because radiofrequency could cause inappropriate pacing and atrial and ventricular leads could cause burns. Patients with magnetic resonance-­conditional pacemakers can nowadays have safe MRI scans with the support of both cardiologists and radiologists. Although this is great news for patients, there are still reasons for caution because the documentation on the device implanted is not always present in the patient’s notes and many magnetic resonance-­unsafe pacemakers are still being implanted today. Therefore, it is worth remembering that these devices require radiologists and referrers to follow strict safety measures to the letter. If the scan is not safe, it must be avoided.

Intravenous Contrast Material

Iodinated Intravenous Contrast for CT Iodinated contrast agents are indicated for a wide spectrum of head and neck pathologies, such as tumours and infections, or to ­visualise arteries (CTA) and veins (CT venogram). Contrast agents may be injected into the thecal sac (CT myelogram or cisternography) or intravenously to enhance the visualisation of the spinal cord, nerve roots and vascular structures. Intravenous iodinated contrast can be safely administered to patients with an estimated glomerular filtration rate (eGFR) of ≥  30  ml/min/1.73  m2 and in those who are anuric on chronic ­ haemodialysis. Patients with severely impaired renal function (i.e., eGFR 200

Cognitive impairment

FRAXE

309548

(CCG)n

FMR2

4–39

200–900

Cognitive impairment

FRDA

229300

(GAA)n

Frataxin

6–32

200–1700

Sensory ataxia, cardiomyopathy

RNA or repeat peptides mediated toxicity DM1

160900

(CTG)n

DMPK

5–37

56–10 000

Myotonia, weakness, cardiac involvement, diabetes, cataracts

DM2

602668

(CCTG)n

ZNF9

10–26

75–11 000

Similar to DM1, more proximal weakness

FXTAS

309550

(CGC)n

FMR1

6–60

60–200

Ataxia, tremor, parkinsonism

C9ORF72

105550

(GGGGCC)n

C9ORF72

1–29

Usually > 1000

ALS, FTD can present as CBD

Unknown pathogenic mechanism SCA8

608768

(CTG)n

ATXN8OS

16–34

> 74

Ataxia, cognitive decline

SCA10

603516

(ATTCT)n

ATXN8

10–20

500–4500

Ataxia, tremor, dementia

SCA12

604326

(CAG)n

PPP2R2B

7–45

55–78

Ataxia and seizures

HDL2

606438

(CTG)n

Junctophilin 7–28

66–78

Similar to HD

SCA31

117210

(TGGAA)n

BEAN

1

2.5–3.8 kb

Ataxia and neuropathy

SCA36

614153

(GGCCTG)n

NOP56

1

650–2500

Ataxia, tongue fasciculations, deafness

SCA37

615945

ATTTC(n)

DAB1

250 AAGGG

Ataxia, sensory neuropathy, vestibular areflexia, chronic cough

FGF14 SCA27B 620174

(AAG)n

 250 (> 300 with complete penetrance)

ALS, amyotrophic lateral sclerosis; CBD, corticobasal degeneration; DM1/2, myotonic dystrophy types 1 and 2; DRPLA, dentatorubral pallidoluysian atrophy; FRAXA/E, fragile X mental retardation; FTD, frontotemporal dementia; FXTAS, fragile X tremor/ataxia syndrome; HD, Huntington’s disease; SBMA, spinobulbar muscular atrophy; SCA, spinocerebellar atrophy; RFC1, replication factor complex subunit 1.

84

Neurology: A Queen Square Textbook

is in some areas, very effective, but it is very expensive and time-­consuming. Recently, many of the technical barriers to identifying disease-­ Concordance causing mutations and rare variants have been overcome with next-­ Disorder rate (%) Type of study generation sequencing, which initially consisted of whole-­exome sequencing of all protein-­encoding regions (exomes) and intron Alzheimer dementia 60–80 Twin studies boundaries. Even more complete is whole-­ genome sequencing Autism 70–90 Twin studies (introns and exons and intergenic regions), which is now taking over from exome sequencing in most centres. Next-­generation Epilepsy 80 Twin study sequencing platforms produced by various companies use different Frontotemporal 42 Family history data technologies, but in general are high-­throughput, producing thoudementia sands or millions of sequences at the same time. In finding mutations and risk variants, this method offers four critical advantages Multiple sclerosis 25–76 Twin studies over the traditional linkage and positional cloning approaches: cost, Migraine with aura 68 Twin study speed, the use of relatively little DNA and the capacity to use genetiRestless legs syndrome 40–90 Twin studies cally less informative samples. A large number of samples that were previously unsuitable for gene identification because of poor quality material and the availability of single affected individuals, can schizophrenia. However, the fact that the monozygotic twin of an now be used to find novel mutations. In the past 10 years, whole-­ affected individual does not develop the disease 100% of the time exome and whole-­genome sequencing have been successfully used indicates that other factors are involved. In migraine with aura the to identify hundreds of novel genes associated with Mendelian correlation of liability has been estimated at 68% in monozygotic neurological diseases (Figure 6.1). Although this technique genertwin pairs, with no significant difference between males and females. ates enormous amounts of data, it does not overcome the challenge Similar twin studies have provided evidence that genetic factors of differentiating between normal genetic variability and pathohave a role in many other neurological disorders (Table 6.3). genic mutations, so it is important to analyse a large number of affected individuals, compare them with unaffected individuals, and interpret the data in the context of gene expression and gene Mutation Versus Polymorphism network data (Figure 6.2). Every individual has over 20,000 genetic variants in the protein-­ Next-­generation sequencing typically involves exome sequencencoding exons and flanking intronic regions. The vast majority of ing of several affected and unaffected family members, where posthese variants are thought to have no deleterious consequences sible including affected cousins or other distant relatives. In the first (benign polymorphisms), but some of these genetic changes can be instance, DNA is extracted and a genome library made from each associated with human disease. Disease-­causing genetic mutations individual; after this, libraries may be enriched for exome regions are rare and proving the pathogenicity of a genetic variant is often (e.g. in whole exome sequencing), pooled and sequenced. The most difficult. The criteria to identify a sequence variant as a pathogenic difficult part of next-­generation sequencing is the data analysis, mutation include: which involves the need for a powerful computer cluster to align • The variant causes a change to the encoded protein, through an sequences, remove duplicates and annotate the data sequence and amino acid substitution, insertion or deletion, or through a non- variants. When investigating disease-­causing mutations, common sense change (frameshift and premature stop codon). polymorphisms and non-­ coding amino acid changes are then • The variant occurs in more than one family and segregates with removed to leave a list of unique or very rare (0.01%)

Several putative disease genes

Post-exome filtering in list of putative genes 1 Sanger sequence to confirm variants 2 Screen ethnically matched controls 3 Segregation in families 4 Investigate mutation mechanism 5 Sequence in other disease series Functional work 1 Over-express the mutation in mammalian cells 2 Knock down with RNAi if recessive 3 Fibroblast and lymphoblast studies 4 iPSC, invertebrate and animal models

Disease specific filters Select by inheritance, gene expression and function

Figure 6.2  Filtering and functional analysis strategy for next generation sequencing data. SNP = single nucleotide polymorphism. iPSC = induced pluripotent stem cells. RNAi = RNA interference.

86

Neurology: A Queen Square Textbook

Clinical history, detailed family tree and examination

Investigations to exclude acquired causes

Examine other family members where possible

High-yield, core, single gene/type neurogenetic tests such as Huntington’s disease expansion and CMT1A 17p11.2 duplication

Advanced Genetic Testing

Comparative Genomic Hybridization (arrayCGH). First line in ID

Mitochondrial (mtDNA) sequencing. Greater yield in muscle compared with blood DNA

Extended repeat expansion testing such as C9orf72 and RFC1; Southern blot sizing required

Exome sequencing to investigate a differential diagnosis gene panel

Future directions First line genome sequencing for all to investigate diagnostic gene panel and identify the location of repeat expansions and copy number variation

Difficult to interpret genome sequencing results, validate with second diagnostic test such as repeat expansion analysis with optical genome mapping or long-read sequencing

Figure 6.3  General current strategy and future direction of genetic testing. CGH = comparative genomic hybridisation.

Total variants for each individual

Filter in very rare non-benign variants: 1 MAF  70% not enough data = 1–10% = AChR, acetylcholine receptor; AQP4, aquaporin 4; CSF, cerebrospinal fluid; GBS, Guillain–Barré syndrome; Ig, immunoglobulin; IMNM, immune-mediated necrotizing myopathy; LGI1, leucine-rich glioma inactivated 1; MFS, Miller Fisher syndrome; MG, myasthenia gravis; MOG, myelin oligodendrocyte glycoprotein; NMDAR, N-methyl D-aspartate receptor; NMOSD, neuromyelitis optica spectrum disorder. Source: adapted from Volkov et al. (2022). Used under Creative Commons CC-BY licence 4.0.

0005737021.INDD 97

02-16-2024 10:31:38

98

Neurology: A Queen Square Textbook

colourless, translucent (‘gin clear’) fluid have often consigned it to the position of a ‘less useful’ body fluid in the diagnosis of disease. This was perhaps true when analytical techniques were only able to identify basic cell types and molecules in relatively large concentrations (‘CSF total protein’). Advances in qualitative and quantitative techniques of CSF analysis are beginning to reveal the CSF as an increasingly useful fluid from which to identify biomarkers with relevance to diagnosis, prognosis and response to treatment. Discussion of most of this emerging field is beyond the remit of this chapter. However, there is a very solid basis for the use of the CSF as  a biomarker in many diseases of the CNS and PNS that are ­discussed here. Analysis of Cerebrospinal Fluid Solutes In health, the CSF is maintained as a protein-­poor solvent in which are dissolved proteins derived mainly from brain tissue. These are produced in the most part from physiological and pathophysiological processes of release and cellular decay and recycling. Perhaps of most interest in the clinical investigation of the CSF are normal levels of CSF antibodies, amyloid beta, tau and phosphotau proteins (as a baseline from which change indicates disease), and asialotransferrin (as an easy analyte as a marker of CSF as a fluid). Many other proteins contribute to ‘total’ CSF protein (e.g. transthyretin) but have less use, at present, in the assessment of health or disease. In the most basic of clinical CSF analysis, CSF total protein is measured by one of a large number of methods. However, ‘CSF protein’ contains contributions from CNS and systemic compartments (if there is a dysfunctional blood–brain barrier) and is a very non-­ specific measure. The measurement of specific protein is much more useful. No albumin is produced in the CNS and so any albumin detected in the CSF is there as the result of direct blood–brain barrier transport and control or as a result of ‘leakage’ through a dysfunctional barrier. CSF albumin is easily measured and its ratio to the serum albumin is constant in health. This ratio (CSFalb : serumalb) is known as the albumin quotient (Qalb) and is a method-­ independent measure of blood–brain barrier permeability. In health, all the IgG in the CSF is actively transported across the blood–brain barrier (as are other immunoglobulins). A raised CSF IgG in comparison to the serum (raised CSFIgG : serumIgG) or the IgG quotient (QIgG) can be used similarly. The IgG index (QIgG : Qalb), which is the ratio of these ratios, can define the permeability characteristics of the blood–brain barrier, helping to define the pathogenesis of processes occurring inside or outside the CSF space with, or without, humoral drive. Other CSF solutes can also be measured. Most are present in very small quantities and the current dynamic range of most assays as well as inter-­individual differences in response is unable to distinguish solute profiles on an individual patient basis. However, some solutes released in large quantities (e.g. a-­beta amyloid proteins, tau, neurofilament proteins, S-­100β and 14-­3-­3 protein) are easily measured. The profiles of these solutes, assayed in the correct clinical context, can provide biomarker support for the diagnosis of Alzheimer’s disease versus frontotemporal lobar dementias and in identifying neurodegenerative brain disease, or brain injury. The collection of CSF samples and their immediate handling are crucial to these assays as considerable pre-­process changes can occur as a result of interaction with plastic ware and at room temperature. Furthermore, assays should be carried out in recognised quality-­ controlled laboratories where variances in assays are minimised.

Platforms for assays of low concentration solutes continue to develop. Enzyme-­ linked immunosorbent assay (ELISA) can be both sensitive and specific. However, multiplex assay systems with enhanced development techniques, mass spectrometry and ‘single molecule detection’ techniques will advance the field rapidly in the near future.

Cerebrospinal Fluid Cells

The CSF is essentially an acellular fluid. The CNS is patrolled by a tightly regulated system of microglia generally maintained in a state of active patrol but hyporesponsiveness. Very few cells are found circulating in normal CSF. Most laboratories will quote less than five white cells per cubic millimetre as a normal range, but any cells in the CSF should provoke suspicion of a continuing process, even if very non-­specific. Cells are nearly always lymphocytes, and the identification of macrophages or neutrophils is almost certainly disease related and useful in diagnosis. Cell counts are not reliable for the identification of other cell types and formal cytology or flow cytometry studies are necessary to provide positive identification. In malignant meningitis, cell type can both make the diagnosis and prompt a search for the primary process or other disseminated disease. In haematological malignancies, where the cell type may be lymphocytic, and where accurate diagnosis depends upon the lymphocytic type, clonal molecular flow cytometry can provide an accurate diagnosis as long as sufficient CSF is provided containing sufficient cells. Red cells in the CSF are always abnormal, although the most common cause for these is a traumatic lumbar CSF tap.

Other Constituents of Cerebrospinal Fluid

Pathogens can be isolated from the CSF by a variety of methods. Direct visualisation of bacteria or fungi can be immediately useful. Polymerase chain reactions (PCR) for viruses are highly specific and variably sensitive. It must be remembered that PCR depends upon the primers chosen, the strain of the virus and the copy number within the fluid. A negative result should not always be taken as negative and alternative or repeated tests might be useful. This is a common situation in herpes simplex virus (HSV) encephalitis, for instance. Viral serology in CSF (IgM and IgG) has increasing utility in clinical diagnosis; experimentally, next-­generation sequencing has started to enter routine clinical practice, although it has a higher yield in brain biopsy samples. Culture studies can grow both bacteria (especially tuberculosis) and fungi at a later date, and the possibility of such pathogens should be considered at the time of sampling; the request for special cultures will be required as well as adequate volumes of fluid. The use of immunoglobulin in the CSF covers total IgG, the comparison of serum and CSF oligoclonal banding patterns and specific immunoglobulins in disease. The total IgG level was discussed earlier and is quickly measured. The oligoclonal banding pattern of the CSF in relation to the serum is a more lengthy process but can identify the relative production of monoclonal or oligoclonal humoral responses from within the CSF. This is useful in diagnosis of inflammatory CNS diseases. However, it remains a mystery as to the antigenic targets of oligoclonal bands in any disease, and particularly in multiple sclerosis; the identification of a target antigen in this and other diseases would be a huge contribution towards explaining pathogenesis. Attempts have been made to identify virus-­specific oligoclonal bands in the CSF with limited success. These techniques have been superseded by PCR and whole-­genome, next-­generation sequencing techniques with greater and highly specific coverage for expected or unexpected pathogens.

Neuroimmunology  99

Autoantibody Testing in Central Nervous System Disease in Serum and Cerebrospinal Fluid Auto-­antibodies to brain proteins can be measured using several techniques. Those of most use to the neurologist in the diagnosis of autoimmune encephalitis, for example, require specific and accurate tests which minimise the risk of non-­specific binding and false negatives and positives. In current practice, such testing is done in most parts of the world in a dedicated neuroimmunology laboratory with expertise of the tests. The history of such antibody testing has moved away from radioimmunoprecipitation assays  – for example the voltage-­gated potassium channel autoantibody test, and Western blots (e.g. basal ganglia antibody testing), which have been hampered by binding of intracellular proteins and therefore often demonstrate non-­specific irrelevant binding. Such tests can lead to misdiagnosis, even with results showing high levels of antibody, which can often lead a clinician or patient to overinterpret their value. In the past 15 years, cell-­based assays (CBAs), in which the target protein is expressed on a human embryonic kidney cell line to try to allow relevant autoantibodies to the protein in its native confirmation to be tested for. In a research testing these can be done live (Figure 7.4), thus interfering minimally with the structure of the protein and using a fixation method after the human sample of either serum plasma or CSF was applied to the cell line, allowing relatively pure demonstration of antibody binding to protein in its native state. However, such an assay is time intensive and has also been shown to find low levels of antibody in the healthy population so on its own does not demonstrate a diagnostic blood test. For clinical diagnostics, most centres use a combination of immunohistochemistry to look for patterns of binding useful in the diagnosis of paraneoplastic syndromes, and prefixed CBAs, which can be tested rapidly and give a result within hours. In the long term, these technologies can generate ELISAs for rapid throughput testing in local laboratories, and also the rollout of preprepared slides of fixed cells to smaller hospitals. Testing both serum and

Figure 7.4  Live cell-­based assay for NMDAR antibodies. Human embryonic kidney cells transfected with NMDAR subunits, stained with serum (anti human IgG, shown as red).

cerebrospinal fluid is useful generally to rule out other processes in the cerebrospinal fluid but also clinically useful in N-­methyl-­D-­ aspartate receptor (NMDAR) encephalitis diagnosis in particular. IgG s­ubtype testing is not usually carried out clinically at the ­current time.

Immunological diseases of the Nervous system

The immune system probably has roles in the causation, pathogenesis, resistance to and recovery from a vast number of diseases in the nervous system. Furthermore, clinical diagnosis is greatly assisted by the analysis of ‘immune’ biomarkers (antibodies, cells, cell-­related factors and aberrantly released molecules), which are measured in immunology laboratories. The immune system is critically involved in the resolution (and occasionally the worsening) of infectious diseases, but this is not described further here. Immunology is a part of every chapter of this textbook and diseases may be described as B-­or T-­cell mediated, complement-­ dependent or cytokine-­driven. It is not possible to make clean divisions of diseases in relation to their involvement of the immune system, as each involves the interaction of many players in the ‘immunological soup’. As duplication and redundancy of function both exist, the analysis of a single component of the immune system is unlikely to provide the explanation for any one disease. In treatment options, the balance of immune components in disease can lead to non-­intuitive pathology (e.g. the provocation or worsening of demyelinating diseases with inhibitors of TNF-­α). There follow some brief examples of diseases that are discussed later in the book with delineation of the predominant immune pathogenesis. We pay particular attention to autoimmune encephalitis in this chapter as an example of an immune-­mediated neurological disorder with increasingly well-­defined phenotypes, together with good theoretical, in  vitro and ex  vivo models of disease, and here we discuss this subset of disorders in more detail, the understanding of which has deepened significantly in the past decade. T-­Cell-­Mediated Neurological Disease T-­cell-­mediated neurological disease is difficult to define as the specific targets of T-­cell receptors are challenging to isolate. In clinical practice, T-­cell-­mediated diseases are generally less reversible than B-­cell mediated diseases and, perhaps because of this less, are responsive to current immunotherapy. However, increasing numbers of therapies are emerging that target key components of the T-­cell immune response, especially cytokines. The final common pathway to tissue damage in cerebral and ­peripheral nerve vasculitis involves T-­cell infiltration of tissue and destruction of tissues associated with the T-­cell infiltrate and the inflammatory milieu elaborated by those cells. The targeting of ­particular blood vessels has yet to be explained. Multiple sclerosis is probably the most common and best described of the CNS neurological disorders, once again with perivascular T-­cell infiltrates and myelin, and subsequently axonal destruction in plaques of inflammation. Whether this process or a neurodegenerative underlying process is the primary mechanism of disease and whether underlying viral pathogens drive the initiation of inflammatory responses remains a continuing debate. In the PNS, there is evidence that CIDP is a T-­cell-­mediated disease with evidence for deficiencies in the autoimmune regulator

100

Neurology: A Queen Square Textbook

Antibody-associated Neurological Diseases Antibody-associated neurological diseases can be divided into those where the antibody defines the disease and probably is primarily responsible for the pathogenesis of that condition, and diseases where antibodies are markers for the disease without necessarily being causative.

diseases are caused by immune factors presumably released by, or in response to the tumour which cause damage to the end tissue. Those agents are presumed to be antibodies, T cells or other soluble molecules but these are often difficult to demonstrate. A number of well-­known paraneoplastic antibodies are described (e.g. anti-­Hu, anti-­Yo and anti-­Ri) which are associated with a group of phenotypically recognisable disorders such as sensory ganglionopathies, limbic encephalitis and the opsoclonus myoclonus syndrome. They are more useful in their association with a number of underlying tumour types such as small cell cancers of breast and ovary but neither the antibody nor the tumour types co-­segregate perfectly with one another (Figure 7.5). Many of the antibodies first described are to intracellular antigens and mechanisms of pathogenesis remain unexplained; in these syndromes, the antibody may be an epiphenomenon and the disease mechanism T-­cell mediated. Syndromes with intracellular targets tend to be difficult to treat, and although identification and successful resection of the tumour may result in limitation of further progression, reversal or recovery is seldom seen. In paraneoplastic syndromes where antibodies to surface antigens and receptors are present, the antibody appears to have a direct effect. Lambert–Eaton myasthenic syndrome typifies these diseases, where voltage-­gated calcium channel antibodies are pathogenic. More recently, antibodies to surface epitopes have been identified in the encephalitides associated with antibodies to leucine-­rich glioma-­inactivated 1 (LGI1), NMDAR and glycine receptors, and other synaptic proteins with a number of distinct phenotypic presentations associated with them.

Neurological Disease Associated with Systemic Disorders and Systemic Autoantibodies Several diseases are associated with antibodies that may have ­systemic effects, including effects on the nervous system, or have antibodies that are central to the diagnosis but which have questionable relevance in the pathogenesis. The connective tissue diseases have a number of neurological manifestations which may or may not be related to the presence of the antibodies described as a part of the syndromes and sometimes more or less essential in the diagnostic criteria. The antineutrophil cytoplasmic antibodies are associated with some vasculitides. Although they may activate neutrophils and are necessary for the diagnosis in some diseases, for example granulomatosis with polyangiitis, they are not essential to the pathogenesis or the pathology. Antibodies to the extractable nuclear antigens are associated with a number of other connective tissue diseases such as primary and secondary Sjögren’s syndrome, which have associations with some neurological syndromes (e.g. sensory neuronopathy). Antibodies to phospholipids (cardiolipin, beta 2  microglobulin and detection of the lupus anticoagulant) are associated with the antiphospholipid syndrome, which may cause a catastrophic disorder of coagulation that may involve the CNS. In some cases, this resembles MS and making the distinction between causation and association may be very difficult and lead to inappropriate therapies. These antibodies are commonly present at low levels after infection and require measurement to demonstrate sustained positivity (e.g. after three months) and an association with thrombotic disease to be deemed clinically relevant. Paraneoplastic diseases overlap between humoral and T-­ cell mediated disease with the presence of antibodies defining the syndromes. Paraneoplastic syndromes occur as an immunological response to the presence of cancer, but the phenotypic manifestations are not directly caused by the presence of cancer cells. The

Autoimmune Encephalitis The autoimmune encephalitides are the archetypal example in the CNS of how highly specific components of the adaptive immune response may target specific neuronal or glial molecules, including pre-­or post-­synaptic neurotransmitter receptors, ion channels and structural proteins, leading in turn to specific clinical syndromes. While relatively in its infancy (potassium channel – now known to be LGI1 and CASPR2 – antibody CNS syndromes first described by Angela Vincent in 2001), high levels of momentum and interest in this field mean that new antibodies and syndromes are described regularly. Seronegative autoimmune encephalitis remains a challenge and likely represents a heterogeny of syndromes, some with antibodies not yet described, and others in which other immune mechanisms play a principal role. The most common forms of autoimmune encephalitis include NMDAR encephalitis, leucine-­rich glioma inactivated 1 (LGI1) antibody associated disease, contactin-­associated protein-­2 (CASPR2) antibody-­associated encephalitis and then a list of rarer syndromes. The diagnosis remains clinical, and there are research criteria which can be helpful for diagnosis, but there are many mimics, including neurodegenerative, psychiatric, and rare genetic disorders (including genetic Alzheimer’s disease, cytotoxic T lymphocyte antigen 4  haploinsufficiency, X-­ linked adrenoleukodystrophy, prion disease – all of which may present in adulthood). The cell surface autoimmune encephalitis syndromes generally have symptoms which are either highly specific, for example faciobrachial dystonic seizures with LGI1 antibody disease, or crossover between neurology and psychiatry as in the NMDAR encephalitis, which may present most often to psychiatrists first. Patients often have an anterograde amnesia which can be quite marked in LGI1 disease, with disruptions in the perception of time and space, and in addition is associated with frontal dysexecutive features in NMDAR encephalitis. Fever is not predictive of infectious verses autoimmune

(AIRE) protein, reduced T-­regulatory mechanisms, failure of Fas-­ Fas ligand lymphocyte downregulation and mixed Th1 and Th2 cytokine profile up-­regulation in the serum and the endoneurium. There is also evidence for B-­cell mediated pathways. Cytokine-­Driven Processes Primary cytokine processes are less well described. A clear example of cytokine dysregulation occurs in the POEMS (polyneuropathy, organomegaly, endocrinopathy, M-­protein and skin changes) syndrome. Although there is a paraprotein in this disorder, it does not appear to play an obvious part in the pathogenesis. In addition, although plasma cells are frequently expanded, it appears that the central processes driving the disease are unregulated vascular endothelial growth factor (VEGF) and IL-­6 production, possibly exacerbated by hypoxia-­ induced factor 1α. This unregulated cytokine drive results in proliferation, migration and maturation or B cells and increasing amounts of cytokine production resulting in multisystem symptomatology. Once again, interference with these processes within the pathway can result in stabilisation of the disease as long as there is not an essential requirement for the ablated aspect of the pathway (e.g. VEGF).

Neuroimmunology  101

Cancer association

Clinical phenotype

Ri

RPCS

KLHL-11 Tr (DNER)

AMPAR GABAbR P/Q VGCC mGluR5 CASPR2 LGI1 GAD65 mGIuR1 GFAP DPPX

Testicular HL Breast

NMDAR Encephalitis

NSCLC

SCLC

Ovarian

Ovarian teratoma Malignant thymoma

LE

NMDAR

Breast NSCLC Neuroendocrine Neuroblastoma

Diencephalic

SCLC

RPCS HL

Encephalitis

Malignant thymoma

LE

(70% cancer)

HIGH-RISK ABS

PCA2 (MAP1B) Hu (ANNA-1) Ma/Ma2

Thymoma EM

CV2/CRMP5 Ampiphysin

RPCS

Neuroendocrine SCLC Hematologic malignancies

SPS

Encephalitis

Ovarian teratoma

Other carcinomas

B-cell malignancies

Figure 7.5  Main clinical phenotype and associated cancer in paraneoplastic encephalitis. Abs, antibodies; ANNA, antineuronal nuclear antibody; AMPAR, α-­amino-­3-­hydroxy-­5-­methyl-­4-­isoxazolepropionic acid receptor; CRMP5, collapsin response-­mediator protein 5; CASPR2, contactin-­ associated proteinlike 2; DPPX, dipeptidyl peptidase-­like protein; DNER, delta/notch-­like epidermal growth factor-­related receptor; EM, encephalomyelitis; GABAaR, gamma-­aminobutyricacid-­A receptor; GABAbR, gamma-­aminobutyric acid-­b receptor; GAD, glutamic acid decarboxylase; GFAP, glial fibrillar acidic protein; HL, Hodgkin lymphoma; KCTD16, potassium channel tetramerisation domain containing; KLHL11, Kelch like protein 11; LGI1, leucine rich gliomainactivated protein 1; LE, limbic encephalitis; MAP1B, microtubule-­associated protein 1B; mGluR1, metabotropic glutamate receptor type 1; mGluR5, metabotropic glutamate receptor type 5; NMDAR, n-­methyl-­D-­aspartate receptor; NSCLC, non-­small cell lung cancer; OMS, opsoclonus-­myoclonus syndrome; PCA, Purkinje cell antibody; RPCS, rapidly progressive cerebellar syndrome; SCLC, smallcell lung cancer; SPS, Stiff Person Syndrome; VGCC, voltagegated calcium channel. Source: Vaišvilas et al. (2022) Used under Creative Commons licence CC BY 4.0.

encephalitis, but a prodrome of psychiatric features and hyperkinetic movement disorder makes NMDAR encephalitis more likely than infection. Brain imaging may show signal abnormalities in the mesial temporal lobes in up to 50% of cases of LGI1 disease but these can be subtle. Imaging can be normal in NMDAR encephalitis. An electroencephalogram (EEG) can be more useful particularly in patients with psychiatric symptoms at onset but is often non-specific. The delta brush pattern is mostly seen in patients on the intensive care unit. Autoantibodies should be examined for in both blood and CSF. In some cases, antibodies may be found in the CSF only, whereas others including LGI1 antibodies may be more readily identifiable in blood. Early treatment is associated with better response and outcomes though trials comparing efficacy of specific agents are lacking. First-­ line treatments include high-­dose corticosteroids, often with plasma exchange and/or IVIg, and a prolonged corticosteroid taper. Second-­ line treatment is often required in patients without sufficient response after two to four weeks. Second-­line therapies, including rituximab and cyclophosphamide, may be tried, with careful consideration of adverse effect profiles. Rituximab is now approved as a second-­ line therapy for NMDAR encephalitis within the NHS. Other agents including bortezomib have been used but evidence for use is currently limited to case reports only. Several trials are in

progress at the time of writing, but these are complicated by small patient numbers, the need for multicentre involvement and often international collaboration and funding. Multidisciplinary input both in the acute and rehabilitative phases is key. The striking reversibility of some syndromes with appropriate treatment may indicate the reversibility of antibody binding effects to receptors or channels, recycling of receptors to the surface at the synapse or the potential for CNS plasticity. For example, following (often prolonged) rehabilitation, patients with NMDAR encephalitis may achieve very good or good recovery despite a prolonged stay in intensive care with coma. But in those with apparently good recovery, detailed clinical assessment including neuropsychometry may identify subtle residual deficits. However, outcome varies across the autoimmune encephalitis spectrum. Overall, these syndromes can be associated with significant long-­term morbidity (cognitive, neuropsychiatric, physical or epileptic) and, in some cases, mortality. Relapse of symptoms may occur with some syndromes after treatment, in some cases corresponding to antibody titres. The possibility for underlying malignancy should be carefully considered in all cases. While autoimmune encephalitides with cell surface antigen targets are often considered to be associated with lower risk of malignancy than those with anti-­neuronal antibodies targeting intracellular antigens, this division is not completely

102

Neurology: A Queen Square Textbook

accurate. Some syndromes with surface receptor antibodies, including to AMPA-­R, GABABR, and CASPR2 in the context of Morvan ­syndrome are associated with intermediate (30–70%) risks of cancer, while some against intracellular antibodies are associated with low risk (e.g. against glutamic acid decarboxylase). Careful consideration of the clinical phenotype, antibody specificity and associations and patient factors including age, gender and systemic symptoms should be considered when assessing risk of underlying malignancy. Follow-­up screening may be required where suspicion is high but initial investigations have been unrevealing. The co-­ occurrence of more than one antibody may be an additional red flag for an underlying tumour. A range of other rarer syndromes have been described associated with antibodies to inhibitory and excitatory proteins on the cell surface (Table 7.2). In general, the antibodies have evidence of blocking receptor function. An unmet need is tracking disease activity over time and the development of bespoke treatments to try to prevent or minimise long-­term sequelae, for example cognitive impairment or epilepsy, often associated with atrophy on a magnetic resonance imaging. CNS targeted antibodies generally block functions but can exert their effects in numerous ways including allosteric modulation of receptors or alteration of receptor expression impacting on synaptic signalling (the ‘acquired synaptopathy’). They can also recruit other immune factors to the cell surface leading to neuronal or Table 7.2  Autoantibody associations of autoimmune encephalitis. Autoantibody

Clinical syndrome

N-­methyl-­D-­ aspartate receptor

Psychosis, cognitive syndrome with disinhibition, seizures, coma in severe cases with concomitant rhythmic movement disorder, association with ovarian teratoma mainly in young women

LGI

Anterograde amnestia, faciobrachial dystonic and related complex or simple partial seizures, hyponatraemia

CASPR2

As per LGI1 plus acquired neuromytonia, autonomic features, cancer association

GAD

Epilepsy, cerebellar syndromes – not all immunotherapy responsive

MOG

ADEM-­like acute encephalitis

Glycine receptor, DPPX

Progressive encephalomyelitis with rigidity and myoclonus, significant anxiety, weight loss

GABA(A)R and GABA(B)R, neurexin 3α

Rare encephalitides associated with status epilepticus and thymoma

AMPAR

Limbic encephalitis with seizures and psychiatric features

IgLON5

Encephalitis with prominent sleep disorder – REM and non-­REM

ADEM, acute disseminated encephalomyelitis; AMPAR, α-­amino-­3-­hydroxy-­5-­ methyl-­4-­isoxazolepropionic acid receptor; CASPR2, contact-­associated protein 2; DPPX, dipeptidyl peptidase-­like protein 6; GABA, gamma aminobutyric acid; GAD, glutamic acid decarboxylase; GlyR, glycine receptor; ; IgLON5, immunoglobulin-­like cell-­adhesion molecule 5; LGI1, leucine-­rich glioma inactivated protein 1; MOG, myelin-­oligodendrocyte glycoprotein; REM, rapid eye movement.

glial injury, for example when complement fixing IgG subtypes are present. In NMDAR encephalitis, antibody binding causes cross-­ linking and reversible internalisation of receptors, leading to reduction in post-­synaptic currents. By contrast, glycine receptor antibodies (progressive encephalomyelitis with rigidity and myoclonus) appear to reduce glycinergic transmission through direct antagonistic effects on the receptor. Antibodies acting on metabotropic receptors may also exert effects by disrupting receptor function and signalling. Recruitment of other immune components may in some cases lead to highly selective lesioning within the CNS, for example selective cornu ammonis 3 atrophy in LGI1 encephalitis. Any immune event in the relatively protected CNS or PNS space must result from a disorder of normal immune tolerance. The events leading to tolerance breakdown vary and almost all remain unknown. However, many of these syndromes sit at the interface of adaptive immunity and infection, neoplasia or neurodegenerative disease, providing clues as to the immune processes involved. Immunological signatures including human leukocyte antigens (HLA) associations may confer increased risk in certain individuals or indirectly implicate a role for T cells. NMDAR encephalitis is associated with ovarian teratoma, albeit with lower frequency than initially reported. Abnormal neuronal tissue and germinal centres are found in disease associated teratomas. Furthermore, NMDAR antibodies have been isolated both from aspirates of cystic structures within them, and from supernatants of cultured B cells aspirated from teratoma tissue, suggesting teratomas are a site of immune priming in the periphery, with subsequent migration of this response to the CNS. NMDAR antibodies can also arise after HSV encephalitis or other CNS infections or rarely after iatrogenic manipulation of immunological tolerance, though the exact sequence of events in both cases remains unclear. Antibodies to IgLON5, a neuronal cell adhesion molecule, are associated with a syndrome in individuals in the sixth decade or older, of chronic onset (over several years) of sleep disorder, gait disturbance and brainstem signs including dysarthria, dysphagia and central hypoventilation. The insidious onset, presence of tau/TDP-­ 43 pathology and poor response to immunotherapy supports a neurodegenerative process, with secondary antibody generation, perhaps as a consequence of neuronal breakdown in the context of predisposing HLA alleles. However, there has been some more recent evidence that these antibodies can induce neurodegenerative changes in neural stem cells, indicating a need for further study. Autoimmune and therapy associations of autoimmune encephalitis are growing (e.g. arising in patients taking TNF-­blockade therapies, checkpoint inhibitor therapies and novel immune therapies). Other Neurological Disease with Pathogenic Autoantibodies There are a range of diseases of the PNS and CNS exist with good evidence for predominant antibody-­mediated pathogenesis. These include myasthenia gravis, GBS, Lambert–Eaton myasthenic syndrome and perhaps stiff-­person syndromes have more or less evidence for a predominant humoral pathogenesis. The fulfilment of the Witebsky postulates is only partial in most of these diseases, largely because the generation of antibodies in animal models in response to immunisation is limited because of self-­tolerance. The neurological disease with the best evidence for an antibody pathogenesis is autoimmune myasthenia gravis. Antibodies to the post-­ synaptic acetylcholine receptor result in a complement-­ dependent disruption of the post-­synaptic neuromuscular junction complex which results in fatigable weakness. However, the

Neuroimmunology  103

initiation of the autoimmune acetylcholine receptor antibody response remains obscure. In the acute motor axonal form (AMAN variant) of GBS, the  ­initiation of the antibody response resulting in the disease is  understood. Some strains of Campylobacter jejuni elaborate human ganglioside-­like epitopes on their lipooligosaccharide coat. Infection in individuals who probably both have impaired self-­ tolerance mechanisms and in whom sufficient adjuvant stimulation exists are able to make antibodies to their peripheral nerve gangliosides. Subsequently, these antibodies have both complement dependent and independent effects to alter electrophysiological membrane characteristics at the node of Ranvier and neuromuscular junction, to damage cell structure of axons and myelin and to impair the normal repair mechanisms that result in recovery. Similar mechanisms may exist in the more common demyelinating GBS, acute inflammatory demyelinating polyradiculoneuropathy. In the CNS, antibodies to the water channel protein, aquaporin­4, and possibly to myelin oligodendrocyte glycoprotein, are associated with neuromyelitis optica. The support for an antibody pathogenesis for these diseases largely rests upon there being a putative antigen at a feasible site, antibody binding experiments and response to immunotherapy. Interfering with the Immune System as a Treatment for Disease The immune-­mediated neurological diseases present opportunities for effective therapy. Immunotherapies work by a number of different mechanisms and sometimes more than one is required to effect a response Table  7.3). Sometimes, immunosuppression may be required in infection to modulate an overactive and damaging immune response. For example, in the treatment of HIV, tuberculosis, meningitis or cysticercosis, antimicrobial treatment can result in a devastating immune response resulting in more damage than the primary infection. Broadly, the mechanisms of action of drugs that modulate the immune system can be classified as follows: • Immunomodulatory • Immunosuppressive

• Replacement • Specifically targeted ablative therapies • Specifically targeted molecules interfering with selective immune elements.

Immunomodulatory

IVIg acts as an immunomodulatory agent by multiple mechanisms, plasma exchange by non-­ specific removal of soluble immune factors and interferons by supplementation and modulation of the cytokine network. The mechanisms of IVIg action are poorly understood. Mechanisms interfering with B-­and T-­cell interactions, macrophage activity and complement are highly important but anti-­idiotype activity is unlikely to be a major mechanism. IVIg may work in part through sialylation of the Fc region of IgG affecting its downstream binding and effects. Another mode of action of IVIg is likely to be mediated by saturation of the neonatal Fcgamma receptor (FcRn). FcRn extends the half-­life of IgG and serum albumin by reducing lysosomal degradation of these proteins in endothelial cells (Figure 7.6). Saturation of this receptor removes the protection of native IgG molecules from degradation and reduces the circulating levels of potentially pathogenic antibodies. Plasma exchange non-­specifically removes low molecular weight solutes, including cytokines and antibodies (especially IgG) and can be quick and effective in its action in appropriate diseases (e.g. GBS, systemic vasculitides and the antibody-­ mediated autoimmune encephalitides). Interferons modulate the immune response by unknown mechanisms.

Immunosuppressive

Steroids, oral immunosuppressants (azathioprine, methotrexate, mycophenolate, ciclosporin) or intravenous agents such as cyclophosphamide, depress immune responses more or less non-­ selectively. These agents are the backbone of the current immunotherapeutic pharmacopoeia. Steroids work quickly and effectively to suppress B-­and T-­cell mediated responses and sensitise cells to the effects of other immunosuppressants, particularly cyclophosphamide. However, their medium and long-­term adverse event profiles limit their long-­term use and in most circumstances

Table 7.3  Mechanism of action of common immunomodulatory agents. Drug

Mechanism of action

Immune consequences

Corticosteroids

Inhibition of gene transcription for secretion of inflammatory cytokines

Reduction of leukocyte migration, phagocytic function of neutrophils and monocytes, and T-­cell function

Azathioprine

Purine antimetabolite: inhibits resting (G1) and DNA synthesis (S) phases of the cell cycle

Apoptosis of T lymphocytes

Methotrexate

Folic acid antagonist; inhibition of purine synthesis

Specific immune cell targets unknown

Mycophenolate

Blocks de novo purine synthesis

Anti-­lymphocyte (T-­and B-­cell) action. Less toxicity than azathioprine

Cyclophosphamide

DNA-­alkylation: blocks all phases of cell cycle

Anti T-­cell and B-­cell activity

Rituximab

Monoclonal anti-­CD20 antibody

Reduces pathogenic antibody production by reducing CD20 positive B-­cells and the number of new plasma cells (plasma cells are CD20 negative but develop from lineage). Pathogenic antibodies reduced and disruption of other roles of B-­cells (e.g. as antigen-­presenting cells) in the immune system

Source: Foster et al. (2023).

104

Neurology: A Queen Square Textbook

(a)

(b)

Figure 7.6  Proposed role of the neonatal Fc receptor for IgG (FcRn) in the vascular endothelial cells. Many cells express the neonatal Fc receptor for IgG (FcRn), including vascular endothelial cells. Once taken up by endothelial cells, the Fc portion of IgG binds FcRn with high affinity within the acidic endosome. When the endosome fuses back to the plasma cell membrane, the neutral pH causes dissociation of Fc from FcRn, thereby recycling IgG back into the circulation and avoiding lysosomal degradation. Source: with permission from Macmillan Publishers Ltd: Nature Reviews Immunology; Roopenian DC, Akilesh S. FcRn: the neonatal Fc receptor comes of age. 2007;7(9):715-­725. 40 Copyright © 2007. Fc, dimeric constant region.

the clinician is looking to stop or substitute these agents as quickly as possible. Azathioprine, methotrexate and mycophenolate are easy to use with appropriate regular blood monitoring for adverse effects, but they are rather slow to work. Cyclophosphamide is an excellent immunosuppressant and with cautious dose restriction, pulsed therapy and metabolite protection of the bladder epithelium by mesna toxicities are much less than with oral medication.

Replacement

For example, IVIg in hypogammaglobulinaemic disorders. RNA enteroviruses are lytic and immunity relies on humoral rather than cytotoxic clearance. Patients with hypo-­or agammaglobulinaemic disorders are susceptible to systemic RNA virus infection that may result in meningoencephalitis. Stabilisation relies upon supplementation of polyvalent immunoglobulins from donors with the normal commensal exposure to common viruses to maintain clearance.

Specifically Targeted Ablative Therapies

Specifically targeted ablative therapies (e.g. anti-­CD20 and anti­CD52  monoclonal antibodies) are increasingly important in the treatment of immune-­mediated disease, sometimes in combination with other immunosuppressants (e.g. rituximab and methotrexate in rheumatoid arthritis, rituximab in paranodal antibody-­mediated neuropathies). The ability to directly interfere with, ablate or block a specific molecule in the immune system as a very powerful tool. Idiosyncratic and predictable but rare complications are the limiting factor in both the risks of initiating therapy and in long-­term use. For example, PML is a known, usually fatal, complication of rituximab, natalizumab and other drugs. Appropriate predictive testing and limitation of the length of exposure reduce the risks. However, for some diseases where death or severe disability is not an outcome then the use of these medications is probably not ethical, but in diseases with potentially worse prognosis they can be very powerful.

Specifically Targeted Molecules Interfering with Selective Immune Elements

Specifically targeted molecules interfere with selective immune elements (e.g. complement inhibitors, FcRn inhibitors). Eculizumab

inhibits the cleavage of C5 by the C5 convertase into C5a, a potent anaphylatoxin with prothrombotic and proinflammatory properties, and C5b, which then forms the terminal complement complex C5b-­9 which also has prothrombotic and proinflammatory effects. It is licensed in treatment of generalised myasthenia gravis and neuromyelitis optica with some promise in GBS. Efgartigimod and rozanolixizumab are neonatal Fc receptor blockers, the antibody fragment binds to the neonatal Fc receptor (FcRn), preventing FcRn from recycling immunoglobulin G (IgG) back into the blood. The medication causes a reduction in overall levels of IgG, including the abnormal acetylcholine receptor antibodies that are present in myasthenia gravis. FcRn blockage has potential in the treatment of other IVIg and plasma exchange-­responsive conditions such as CIDP and autoimmune encephalitis. Bortezomib interferes with plasma cell metabolism and has had use in small open studies in NMDAR encephalitis. There are active randomised control trials exploring efficacy of these ­molecules in some of these disorders.

Specifically Targeted Molecules Interfering with Cytokine and Chemokine Networks

Specifically targeted molecules interfere with cytokine and chemokine networks (e.g. anti-­TNF, anti-­VEGF and anti-­IL-­6 therapies). Some of these therapies are targeted toxins (e.g. the ‘-­cept’ drugs, such as etanercept). Others are antibodies with anti-­cytokine therapeutic activity, such as bevacizumab, which acts as an anti-­ VEGF agent, tocilizumab which is anti-­IL-­6 and anakinra which is ant-­IL1.

Naming of Biological Agents

The naming of the biological agents follows an organised schema. The last consonant describes the activity of the agent; for example, ‘mab’ is a monoclonal antibody and ‘cept’ a conjugated toxin. The penultimate syllable describes the structure and origin, such that ‘xi’ is a chimeric and ‘zu’ a humanised monoclonal. The second syllable describes the target activity, thus ‘tu’ is a tumour, li(m) the immune and ‘ci’ the circulatory system. Lastly, the first syllable is any euphonious collection of letters to make the drug name recognisable.

Neuroimmunology  105

Therapeutic Approach to Immune-­Mediated Neurological Disorders The decision to treat with immunosuppressive or immunomodulatory agents and the selection of specific therapy is influenced by the characteristics of the individual case. It should be based on the likely treatment efficacy in relation to the disease mechanisms, individual clinical features of the patient and their disease, and any the risk of complications (Figure 7.7). The diagnosis of an immune-­ mediated neurological disease should always be made as thoroughly as possible, with appropriate and ample laboratory support before treatment is considered. This includes a tissue diagnosis where it is possible and relevant, especially in vasculitis or where first-­line response has not been as expected. Once treatment is initiated, retrospectively collecting such

pathological data with diagnostic relevance is virtually impossible. The diagnosis and supporting investigations should be clearly ­documented, preferably alongside diagnostic criteria where available. The two essential elements of treatment induction and maintenance are: • Drug efficacy monitoring, which should be disease and patient centred. • Drug safety screening and monitoring, which should be drug and patient centred. Treatment efficacy or failure is primarily a clinical decision in neurological disease, as there are few reliable serological biomarkers of disease activity. To establish objective evidence of clinical change, the use of disease-­and symptom-­specific outcome measurements is recommended at pre-­and post-­treatment assessments.

References

∙ Diagnosis ∙ Treatment decision

Eligibility ∙ Discuss alternatives ∙ General

Informed consent

immunosuppressionrelated risks ∙ Drug-specific risks

∙ Select

Treatment Induction

outcome measurements ∙ Screening ∙ Dose

Further Reading ∙ Efficacy: clinical outcome ∙ Safety: side effects

Monitoring

∙ Not tolerated ∙ Ineffective

Dalakas MC. (2008). B cells as therapeutic targets in autoimmune neurological disorders. Nat Clin Pract Neurol 4(10): 557–567. De Haes W, Pollard C, Vanham G, Rejman J. (2012). ‘Wrapped up’ vaccines in the context of HIV-­1 immunotherapy. In: Metodiev K, ed. Immunodeficiency. London: InTechOpen. http://dx.doi.org/10.5772/2994. Foster MA, Lunn MP, Carr AS. (2023). First-­line immunosuppression in neuromuscular diseases. Pract Neurol 23: 327–338. Kanda T. (2013). Biology of the blood–nerve barrier and its alteration in immune mediated neuropathies. J Neurol Neurosurg Psychiatry 84(2): 208–212. Roopenian DC, Akilesh S. (2007). FcRn: the neonatal Fc receptor comes of age. Nat Rev Immunol 7(9): 715–725. Vaišvilas M, CianoPetersen NL, Macarena Villagrán-­García MD et al. (2022). Paraneoplastic encephalitis: clinically based approach on diagnosis and management. Postgrad Med J [Online ahead of print]. https://doi.org/10.1136/postgradmedj-­2022-­141766. Volkov M, Coppola M, Huizinga R et al. (2022). Comprehensive overview of autoantibody isotype and subclass distribution. J Allergy Clin Immunol 150(5):999–1010.

∙ Disease remission

Discontinue

Figure 7.7  Approach to immunotherapy. Source: Foster et al. (2023).

Ayrignac X, Carra-­Dallière C, Marelli C, et al. (2022). Adult-­onset genetic central nervous system disorders masquerading as acquired neuroinflammatory disorders: a review. JAMA Neurol 79(10):1069–1078. Antel J, Birnbaum G, Hartung H, Vincent A. (2006). Clinical Neuroimmunology, 2nd edn. Oxford: Oxford University Press. Crisp SJ, Kullmann DM, Vincent A. (2006). Autoimmune synaptopathies. Nature Rev Neurosci 17: 103–117. Dalmau J, Graus F. (2022). Autoimmune Encephalitis and Related Disorders of the Nervous System. Cambridge: Cambridge University Press. Mackay I, Rose NR. (2014). The Autoimmune Diseases, 5th edn. San Diego, CA: Academic Press. Male D, Brostoff J, Roth D, Roitt I. (2012). Immunology, 8th edn. New York, NY: Elsevier. Ransohoff RM, Engelhardt B. (2012). The anatomical and cellular basis of immune surveillance in the central nervous system. Nat Rev Immunol 12: 623–635. Vincent A, Martino G, eds. (2012). Autoantibodies in Neurological Diseases. Topics in Neuroscience. Milan: Springer. Witebsky E, Rose NR, Terplan , et  al. (1957). Chronic thyroiditis and auto-­ immunization. JAMA 164: 1439–1447. Zandi MS, Paterson RW, Ellul MA et al. (2015). Clinical relevance of serum antibodies to extracellular N-­ methyl-­ D-­ aspartate receptor epitopes. J Neurol Neurosurg Psychiatry 86: 708–713.

CHAPTER 8

Stroke and Cerebrovascular Diseases David Werring1,2, Matthew Adams3, Laura Benjamin1,2, Martin Brown1, Arvind Chandratheva1,2, Peter Cowley3, Joan Grieve4, Fiona Humphries1,2, Hans Rolf Jäger1,3, Nicholas Losseff2, Richard Perry1,2, Robert Simister1,2 and Ahmed Toma4  Stroke Research Centre, UCL Queen Square Institute of Neurology, London, UK  Comprehensive Stroke Service, National Hospital for Neurology and Neurosurgery, London, UK 3  Lysholm Department of Neuroradiology, National Hospital for Neurology and Neurosurgery, London, UK 4  Department of Neurosurgery, National Hospital for Neurology and Neurosurgery, London, UK 1 2

Clinical Approach, Terminology and Classification of Stroke

Stroke is a clinically defined syndrome of acute, focal neurological deficit attributed to vascular injury (infarction or haemorrhage) of the central nervous system (CNS; brain, spinal cord or retina). In modern clinical practice, neuroimaging is increasingly used to confirm the exact pattern of tissue injury. Older definitions requiring that stroke symptoms should last 24 hours or longer are now obsolete because treatments for stroke require diagnosis and treatment as soon as possible: nervous tissue is rapidly and irretrievably lost in acute stroke, leading to the widely quoted maxim ‘time is brain’. More recently proposed definitions suggest that the duration of the clinical event is of secondary importance to the evidence of ischaemia. However, in practice, the duration of symptoms is mainly important to determine the most appropriate clinical pathway for the patient: those with persistent disabling symptoms need inpatient care whereas others can usually be managed via rapid-­access outpatient clinics. Ischaemic stroke is defined as ‘an episode of neurological ­dysfunction caused by focal brain, spinal or retinal infarction’; infarction is defined as ischaemic injury, based on either pathological, imaging or other objective evidence in a defined vascular ­distribution or on clinical evidence of ischaemic injury lasting more than 24 hours or until death, with other aetiologies excluded. ‘Silent’ CNS infarction has imaging or neuropathological evidence of infarction without a history of acute neurological dysfunction attributable to the lesion. Transient ischaemic attack (TIA) applies to clinical events lasting less than 24 hours without evidence of CNS infarction, but in many countries transient events are classified as TIA even if there is ischaemic tissue injury on neuroimaging (around one third of clinical TIAs lasting less than one hour show one or more areas of acute infarction on diffusion-­ weighted ­imaging (DWI), increasing to over 50% in those lasting over three hours). These infarcts are often small and not seen on computed tomography (CT) but are nevertheless associated with an increased risk of subsequent ischaemic stroke. Thus, magnetic resonance imaging (MRI) with DWI sequences is the imaging modality of choice for ­suspected TIA or minor stroke. Intracerebral haemorrhage is defined as a focal collection of blood within the brain parenchyma or ventricular system and stroke caused by intracerebral haemorrhage as ‘rapidly developing clinical signs of neurological dysfunction attributable to a focal collection

of blood within the brain parenchyma or ventricular system that is not caused by trauma’. Silent intracerebral haemorrhage is defined as ‘a focal collection of blood products within the brain parenchyma or ventricles without history of acute neurological dysfunction’. Subarachnoid bleeding is also often classified as a type of stroke, although focal brain injury does not always occur; subdural and extradural bleeding are not usually considered as stroke because they frequently result from traumatic injury (see later sections on intracerebral haemorrhage and subarachnoid haemorrhage). The term ‘haemorrhagic stroke’, used variably to refer to a range of ­different types of intracranial bleeding including secondary haemorrhage into an infarct, is imprecise and unhelpful. The term ‘stroke’ should be restricted to a description of the clinical event experienced by the patient; appearances on brain imaging should not be described as showing a ‘stroke’ because a scan cannot show a clinical syndrome. Rather, the scan may show infarction or haemorrhage, which may have been either associated with a stroke syndrome or have been asymptomatic. The term ‘spinal stroke’ is often used to describe the occurrence of infarction or haemorrhage in the spinal cord. The mechanisms include vascular occlusion and haemorrhage from structural lesions of the cord (see also Chapter 14). Stroke is not a single disease but is rather a clinical presentation of cerebral ischaemia or haemorrhage, each with varying underlying vascular pathologies. It is important to recognise the differences between risk factors, disease processes and mechanisms (Figure 8.1). Simply identifying a risk factor (e.g. hypertension, smoking), although critically important for rational stroke prevention, does not mean that the cause and mechanism underlying a stroke have been adequately identified. The well-­known vascular risk factors (e.g. hypertension, diabetes and smoking) are not in themselves direct causes of stroke, but instead promote the development of the underlying pathological processes (e.g. atherosclerosis or arteriolosclerosis) and mechanisms (embolism, occlusion or rupture) that are responsible for stroke. A full diagnosis of stroke requires the cerebral circulatory disturbance to be defined in terms of: (1) pathology (ischaemia, infarction or haemorrhage); (2) location within the brain in relation to vascular anatomy (e.g. left middle cerebral artery [MCA] superior division cortical branch territory); (3) mechanism (e.g. artery-­to-­artery embolism); (4) relevant disease processes (e.g. atherosclerosis); and (5) relevant vascular risk factors (e.g. hypertension). A key task of the stroke physician is to make an accurate diagnosis in these terms

Neurology: A Queen Square Textbook, Third Edition. Edited by Robin Howard, Dimitri Kullmann, David Werring and Michael Zandi. © 2024 John Wiley & Sons Ltd. Published 2024 by John Wiley & Sons Ltd. 107

108

Neurology: A Queen Square Textbook

Genetic factors Ischaemic stroke

Disease processes Atherosclerosis Small vessel diseases Cardiac disease (e.g., atrial fibrillation) Arterial dissection Aneurysms Arteriovenous malformations (AVMs)

Risk factors • • • • • • •

Hypertension Diabetes Smoking Obesity Lipid profile Haematological Others

Intracerebral haemorrhage

Mechanisms • Large artery-to-artery embolism • Small vessel occlusion or rupture • Aneurysmal or AVM rupture • Cardiac embolism • Haemodynamic • Other

Figure 8.1  Risk factors, disease processes and mechanisms relevant to stroke.

to guide appropriate management, in particular rational stroke prevention tailored to the underlying mechanism. This requires a basic knowledge of the clinical and radiological patterns that different stroke syndromes produce and the underlying pathophysiology of stroke. Stroke is a medical emergency. In ischaemic stroke due to large-­vessel occlusion, it is estimated that for every minute’s delay in  revascularisation, 1.9  million neurons, 14 billion synapses and 12 km of myelinated fibres are destroyed. Patients with suspected stroke therefore require immediate assessment and rapid transfer to  hospital for consideration of therapy (e.g. thrombolysis or mechanical thrombectomy for ischaemic stroke, rapid preventive treatment for TIA and anticoagulant reversal, blood pressure ­management, neurosurgical intervention, or neurointensive care for intracerebral haemorrhage). The older term, cerebrovascular accident, is outdated, inaccurate, encourages therapeutic nihilism, and should no longer be used. Stroke is not an accident: it results from well-­defined vascular pathologies and mechanisms. Some key clinical concepts in stroke are listed in Box 8.1. From the above discussion, it will be clear that strokes do not occur at random. The underlying pathology responsible for the persistent symptoms of stroke is either infarction or haemorrhage, each due to a defined range of cerebrovascular pathologies and mechanisms (Table 8.1). In about 85% of cases, stroke is secondary to acute ischaemia and subsequent infarction. A map showing the anatomical distribution of common acute ischaemic brain lesions in patients with ischaemic stroke is shown in Figure 8.2. When infarction involves only a small volume of the tissue in the subcortical white matter, basal ganglia, or brainstem (typically  99th percentile) 3 Anti-­β2 glycoprotein-­I antibody of IgG and/or IgM isotype (in titre > the 99th percentile)

Nephrotic syndrome COVID-­19 and other viral infections Disseminated intravascular coagulation

a permanently damaged vessel wall or premature formation of atheroma in the injured vessel. If a patient develops bone marrow necrosis during a severe sickle cell crisis, fat globules released into the circulation may cause strokes. This rare but devastating complication is, paradoxically, more common in milder forms of sickle cell disease such as HbSC. Sickle cell disease is also an important risk factor for cerebral venous sinus thrombosis and this diagnosis should be considered when a patient with this condition has an unusually long and severe headache, particularly if there are focal neurological features. However, this clinical scenario more commonly turns out to be a manifestation of migraine, a syndrome that is associated with sickle cell disease but not specifically with sickle cell crises. Migraine is managed in the same way as in patients without sickle cell disease, with acute and preventive treatments, although some experts caution against the use of triptans for acute headache because of a theoretical concern that triptan-­ induced vasoconstriction may risk causing cerebral ischaemia in sickle cell patients, particularly in those with intracranial stenosis.

Source: adapted from Miyakis et al. (2006).

The risk factors for cerebral venous thrombosis, listed in Table  8.4, are rarely relevant in arterial stroke and should not be tested for routinely, even in younger patients. However, in a patient with a known PFO, venous risk factors may become relevant through the mechanism of paradoxical embolism, and so testing is more easily justified. CVT, however, is a more likely mechanism of stroke in individuals with these risk factors. The most common detectable abnormality in the population is activated protein C resistance; the others are very rare. Resistance to activated protein C has no clear relationship to arterial stroke. In patients with arterial stroke, it is much more important to pay attention to the full blood count, looking for polycythaemia and thrombocythaemia than a thrombophilia screen. Be aware of marginally high red blood cells and platelet counts, which may be easy to dismiss but are common in JAK2 positive myeloproliferation. Antiphospholipid syndrome (APLS; Box  8.4) is associated with arterial (ischaemic) stroke, although high-­quality prospective epidemiological studies are lacking. APLS is not the same as having antiphospholipid antibodies, which may occur in association with other vasculopathies including atherosclerosis. In most cases, the finding of antiphospholipid antibodies in a patient with ischaemic stroke is unrelated to the cause of the stroke (so may be a secondary phenomenon elicited by arterial thrombosis) and follow-­up studies

124

Neurology: A Queen Square Textbook

show antiphospholipid antibodies are not associated with an increased risk of recurrent stroke. However, in a small proportion of  patients, antiphospholipid antibodies appear to have a pathogenic role in arterial and venous thrombosis. Ischaemic stroke is the  most common arterial thrombotic event with APLS; 10–30% develop stroke or TIA over 10 years. Other neurological manifestations include neuropsychiatric features, movement disorders and  migrainous headaches, with aura and cognitive impairment reported in 11–60%. There may be recurrent foetal loss and livedo reticularis. Antiphospholipid antibodies may be inferred by the presence of a circulating lupus anticoagulant (a functional test of coagulation) or immunoglobulin M/G cardiolipin or beta 2 glycoprotein 1 antibodies (immunological assays). These antibodies may be present in patients with SLE, chronic infection and neoplasia, but a rarer subgroup has primary APLS with no other associated disease. In SLE 30–40% of patients have antiphospholipid antibodies and 15% have APLS. The clinical significance of antiphospholipid antibodies in COVID-­19 remains undefined. Antiphospholipid antibody tests can be elevated in viral and bacterial infections such as hepatitis B/C, HIV and syphilis; in these settings they are typically transient and not associated with thrombosis. As part of an APLS screen, anticardiolipin antibodies, anti-­beta-­ 2-­glycoprotein antibodies and lupus anticoagulant all need to be tested for, as relevant positivity may occur in just a single test with potential relevance for arterial thrombosis. In one series, the ­frequencies of a single persistent antiphospholipid test were as ­follows: beta-­2-­glycoprotein 29%, anti-­cardiolipin 10% and lupus anticoagulant 15%. Levels of antibodies can fluctuate considerably, and the diagnosis of antiphospholipid syndrome cannot be made unless the antibodies remain present for three months or more. Current guidelines recommend testing young patients ( 50 cm/second in 24 hours

Advantage – non-­invasive and can be done at bedside. Disadvantage of interindividual variation and is a snapshot. Acoustic windows not possible in 20% population (robotic TCD overcomes some of these disadvantages)

Cerebral micro dialysis

Increase in lactate:pyruvate; glycerol and glutamate

May predict DCI before clinical appearance of deficit

Brain tissue PO2

PbtO2  30%) and high rate of severe functional deficit (50–70%), the duration of status epilepticus does not necessarily predict the outcome and some patients do make a good functional recovery, particularly if they are young and there is no structural brain lesion. Therefore, treatment should not be abandoned prematurely, particularly when there is no clear underlying cause.

New-­Onset Refractory Status Epilepticus

NORSE is a recently described clinical presentation that can affect all ages and arises without a previous history of epilepsy or any other significant neurological or systemic disorder. An underlying cause is found in about 50% of patients. The most frequent cause is autoimmune encephalitis, which may be paraneoplastic. Other causes include infectious encephalitis and rare genetic (SCN1A, PCDH19) and mitochondrial (MELAS, POLG1) disorders. In children, the related condition of febrile infection-­related epileptic syndrome (FIRES) is thought to be due to a fulminant inflammatory response in the central nervous system (CNS). By definition, NORSE is particularly resistant to conventional antiepileptic medication and often requires prolonged treatment with multiple agents to induce burst suppression coma, and seizures may recur with weaning. There is increasing evidence to support an autoimmune aetiology, even in the apparently idiopathic group of patients and, therefore, to suggest an important role for intravenous steroids and immunomodulatory treatment including intravenous immunoglobulin and plasma exchange. Immunosuppressant medication including rituximab, cyclophosphamide, tocilizumab (an IL-­6 receptor inhibitor), anakinra (a human recombinant IL-­1 receptor antagonist which inhibits biological actions of IL-­1β) and bortezomib (a proteosome inhibitor which decreases plasma cells) have also been used with varying benefit in anecdotal series.

Non-­Convulsive Status Epilepticus

NCSE is a condition of continuing or intermittent seizure ­activity which causes impairment of consciousness without overt convulsive activity. It may present as abrupt and unexplained ­deterioration in the conscious level leading to coma, mutism, amnesia, catatonia and focal limb weakness. There may be subtle motor manifestations such as rhythmic twitching of one or more muscle groups, tonic eye deviation, hippus or nystagmoid eye ­jerking in the absence of classical ictal signs. The late stages of convulsive status epilepticus can resemble NCSE. Focal movements and impaired consciousness may be dismissed as being due to metabolic derangement including hepatic encephalopathy unless an EEG is undertaken. The underlying causes of NCSE are varied but in the critically ill they may occur following hypoxic–ischaemic or traumatic brain injury, ischaemic stroke, SAH, encephalitis (infective or autoimmune) and metabolic disorders. NCSE in coma should be managed in the ICU whether the underlying abnormality is generalised or focal. The condition is under-­recognised in the critically ill and any unexpected impairment in the level of consciousness should be investigated by routine EEG, MRI and cEEG monitoring, as indicated. Recent studies have led to a consensus that paroxysmal discharges at a frequency greater than 2.5 Hz are a signature of NCSE, with the diagnosis supported by associated clinical manifestations and improvement after intravenous antiseizure medication. Encephalitis Encephalitis is a common cause for admission to critical care. Autoimmune encephalitis is increasingly recognised and may have a similar incidence and prevalence to viral encephalitis. The presentation of both conditions may be similar and difficult to distinguish. The most common cause is herpes simplex virus (42%) or varicella zoster. Anti-­ N-­ methyl-­ D-­ aspartate receptor (NMDAR) encephalitis is the most frequent cause of autoimmune encephalitis. The principal causes are summarised in Table 28.18.

Table 28.18 Causes of encephalitis leading to NCCU admission. Anatomical localisation of primary infection/ inflammation

Clinical features

MRI appearances

Infective causes

Autoimmune causes

Other causes

Comments

Temporal lobe/limbic encephalitis

Psychiatric symptoms Cortical atrophy; deep (mood disorder, grey nuclei and FLAIR hallucination) sleep hyperintensity dysfunction, seizures, short-term memory loss

HSV-1, HSV-2, VZV, CJD, Mycobacterium tuberculosis

NMDA, LGI-1, AMPA, GABA B, ampiphysin, ANNA1 – Hu, CRMP5, Ma2

Susac syndrome; vasculitis

HSV is favoured over autoimmune aetiology by acute onset, fever, a failure, high level of CSF protein and erythrocytes with asymmetrical temporal lobe abnormalities. Bilateral symmetrical temporal lobe involvement favours an autoimmune aetiology as do lesions outside the limbic system or temporal lobe

Meningeal (pachymeningitis and leptomeningitis), meningoencephalitis

Headache, fever, neck Meningitis: thin linear stiffness, altered mental leptomeningeal state enhancement. Pachymeningitis: localized multiple or diffuse enhancing dural thickening commonly forming mass like lesions

Acute bacterial meningitis (meningococcus, pneumococcus, Listeria); Gram-negative bass line; coagulase-negative staphylococcus; H influenza; HSV-2; Cryptococcus; histoplasmosis; coccidiomycosis; Mycobacterium tuberculosis

Glial fibrillary acidic protein

Neurosarcoidosis; IgG4

Associated with extension of primary meningitis or encephalitis

Thalamus/basal ganglia

Movement disorders, Deep grey matter reduced level of change consciousness, thalamic aphasia

Respiratory virus (influenza, parainfluenza, adenovirus, respiratory syncytial virus); arbovirus; CJD; Mycobacterium tuberculosis; toxoplasmosis/Cryptococcus

NMDA, CRMP5, ANNA1, neurexin 3, LGI-1, GAD

APLS; Sjögren syndrome; vascular (top of basilar)

In children, respiratory viruses are the most common cause but in adults CJD, TB and arbovirus are more common. It is important to emphasise that the top of the basilar syndrome can mimic encephalitis affecting the thalamus

Brainstem

Mesencephalon, (refers to midbrain and upper pons), rhombencephalitis (refers to medulla, pons and cerebellum)

Cranial nerve deficits; cross signs; intractable nausea/vomiting; hiccup (area postrema); reduced level of consciousness hypoventilation

Listeria, Mycobacterium tuberculosis, trip pallidum, Brucella, Whipple’s disease, HSV1/2, VZV, HIV, PML, enterovirus 71, arbovirus

NMO, ANNA1, ANNA2, Ma2, CLIPPERS, GQ1b, Fisher–Bickerstaff syndrome, DPPX

Multiple sclerosis, Behçet syndrome, neurosarcoidosis

Cerebellum

Ataxia

Mass effect due to cerebellar swelling

VZV, EBV, PML, influenza, listeria, CJD

PCA1, PCA2, GAD, see chapter 24 VGCC, post infection, EBV, influenza A, B, mumps, VZV, Mycoplasma pneumoniae

AMPA, α-amino-3-hydroxy-5-methyl-4-isoxazoleproprionic acid; ANNA, antineuronal nuclear antibody; APLS, antiphospholipid syndrome; CJD, Creutzfeldt–Jakob disease; CLIPPERS, chronic lymphocytic inflammation with pontine perivascular enhancement; CRMP, collapsin response-mediator brain protein; CSF, cerebrospinal fluid; DPPX, dipeptidyl peptidase-like protein 6; EBV, Epstein–Barr virus; FLAIR, fluid attenuated inversion recovery; GABA, γ aminobutyric acid; GAD, glutamic acid decarboxylase; HSV, herpes simplex virus; IgG, immunoglobulin G; LGI1, leucine-rich glioma inactivated protein 1; NMDA, N-methyl-D-aspartate; PCA, Purkinje cell cytoplasmic antibody; PML, progressive multifocal leukoencephalopathy; TB, tuberculosis; VGCC, voltage-gated calcium channels; VSV, varicella zoster virus.

0005737023.INDD 1022

02-16-2024 12:59:17

Disorders of Consciousness and Intensive Care Neurology  1023

Limbic Encephalitis Associated with Neuronal Surface Antigens Limbic encephalitis typically presents with acute to subacute onset of memory loss, confusion and seizures. There may be psychiatric/ behavioural disturbances. Limbic encephalitis is associated with intracellularly targeted antibodies (particularly SOX 1, anti Hu [ANNA-­1], CRMP-­5, Ma1, Ma2) and cell surface antibodies (α-­ amino-­3-­hydroxy-­5-­methyl-­4-­isoxazoleproprionic acid receptor [AMPAR], γ-­aminobutyric acid [GABA] B, LGI-­1). (Voltage-­gated potassium channel complex antibodies recognise either LGI-­1 or CASPR2, with the former more commonly causing limbic encephalitis.) Patients with limbic encephalitis may present to intensive care because of worsening encephalopathy, often with hyponatraemia due to inappropriate antidiuretic hormone secretion, or because of seizures. MRI typically shows high signal in the medial temporal lobes and hippocampal swelling (Figure 28.20). The EEG is abnormal with diffuse slowing and occasional temporal epileptogenic foci and CSF may be normal or show a mild lymphocytosis with unmatched oligoclonal bands. Encephalopathy due to contactin-­ associated protein-­2 (CASPR2) antibodies may be amenable to immunotherapy and some patients respond well to high-­ dose oral corticosteroids, pulsed methylprednisolone, plasma exchange and/or immunoglobulin. The seizures often cease and hyponatraemia resolves but patients may be left with prolonged encephalopathy and late-­onset hippocampal sclerosis, especially if immunomodulation is delayed. Limbic encephalitis associated with AMPAR and  GABA  B antibodies may relapse and require longer-­ term immunosuppression. Brainstem Encephalitis Brainstem encephalitis is seen in autoimmune neurological ­disorders but it rarely occurs in isolation. It is characterised by involvement of: • Cranial nerves – extraocular movements, facial palsy, dysphagia • Area postrema – intractable nausea, vomiting or hiccup • Cerebellar peduncles or ascending proprioceptive pathways  – cerebellar or sensory ataxia

• Reticular formation or respiratory centre  – depressed level of consciousness and hypoventilation • Usually associated with sensory neuropathy and/or limbic encephalitis. ANNA-­1 (anti Hu) is associated with brainstem dysfunction including ataxia, nystagmus, dysarthria, dysphagia, laryngoscope, central hypoventilation and depressed level of consciousness while ANNA-­2 (anti-­Ri) causes limbic encephalitis with cerebellar ataxia and cranial neuropathy. Ma-­2 is associated with diencephalic dysfunction, brainstem encephalitis and limbic encephalitis progressing to complete ophthalmoplegia and vertical gaze palsy. Opsoclonus-­myoclonus syndrome is characterised by unpredictable multidirectional saccades, myoclonus and ataxia. Nausea, vomiting, visual abnormalities and altered speech due to myoclonus develop early in the illness. It is often paraneoplastic (ANNA-­2) occurring in association with breast, ovarian, small-­cell lung and neuroblastoma in children but may be parainfectious. Morvan Syndrome Morvan syndrome is characterised by muscle fasciculation and cramp. However, the presence of autonomic disturbance and CNS involvement manifests as sleep disturbance, particularly insomnia, and hallucinations may require admission to intensive care. It is associated with CASPR2 antibodies and there is a significant incidence of underlying tumour (often thymoma). Encephalitis Due to N-­Methyl-­D-­Aspartate Receptor (NMDAR) Antibodies Encephalitis due to NMDAR antibodies is distinct from limbic encephalitis with much more extensive involvement of brain ­function. The onset is generally with psychiatric or behavioural ­disturbances. Seizures develop and worsen over two to three weeks before evolving into severe and, occasionally life-­threatening, dyskinesia (orofacial grimacing, dystonic posturing and choreoathetoid movements). There is often autonomic instability and reduced consciousness, often fluctuating with agitation and catatonia, which may require prolonged ventilatory support. The CSF may show pleocytosis, intrathecal synthesis of oligoclonal bands and positive antibodies. Seizures are common and the EEG may show the extreme delta brush pattern (Figure 28.21), MRI is generally normal but may show mild abnormalities (non-­specific T2 hyperintense lesions) in the cerebral cortex, meninges or basal ganglia. The condition is often associated with ovarian tumour (usually teratoma) in women or occasionally other tumours in men and women (e.g. lymphoma) but is often non-­paraneoplastic. The syndrome often responds to removal of the tumour if present and immunotherapy, but recovery may be slow and there is significant mortality amongst patients who require prolonged intensive care. Treatment is with immunomodulation with steroids, plasma exchange or intravenous immunoglobulin but long-­term immunosuppression with rituximab is often necessary. Relapses can occur in up to 25% of non-­paraneoplastic cases, although are usually less severe than the initial episode. Other Immune-­mediated Encephalopathies

Acute Disseminated Encephalomyelitis

Figure 28.20  Limbic encephalitis. Fluid attenuated inversion recovery coronal image showing swelling and abnormally high signal within the right hippocampal formation (arrow).

Acute disseminated encephalomyelitis (ADEM) is characterised by multifocal demyelination, and presents with acute encephalopathy and focal neurological deficits including weakness, ataxia, cranial nerve involvement and optic neuritis. It is often associated with preceding infection or vaccination. MRI shows FLAIR/T2 hyperintensities involving the periventricular region or deep grey

1024

Neurology: A Queen Square Textbook

Figure 28.21  Extreme delta brushes over the left cerebral hemisphere in a patient with N-­methyl-­D-­aspartate receptor encephalitis.

nuclei with corpus callosum and spinal cord involvement. There may be mild lymphocytic pleocytosis and elevated CSF protein. Treatment includes high-­ dose steroids, plasma exchange and IVIg. The disorder is usually monophasic.

Acute Haemorrhagic Leukoencephalitis

Acute haemorrhagic leukoencephalitis (Hurst disease) is a severe form of autoimmune demyelinating encephalopathy associated with a rapid decline in the level of consciousness with or without seizures, it is associated with prior viral illness (particularly HSV, EBV, influenza A, measles and mumps), mycoplasma infection and malaria. MRI is characterised by punctate or more focal haemorrhage and the key neuropathological finding is the presence of necrotising small vessel vasculitis. The condition carries a grave prognosis. Treatment with aggressive immunosuppression is rarely effective but occasional recovery with or without severe residual disability has been described.

Acute Necrotising Encephalopathy

Acute necrotising encephalopathy mainly occurs in children and is  characterised by a rapidly deteriorating neurological course ­associated with seizures and focal neurological deficit, systemic inflammatory response syndrome (SIRS), multiple organ failure and disseminated intravascular coagulation. There is generally a preceding viral illness. MRI shows multifocal symmetrical lesions, in particular, hyperintense T2 signal bilateral thalamic and brainstem involvement. Rarely lesions are seen in the cervical spinal cord, cerebellum, medial temporal lobe, insular cortex or sub-­cortical regions including the mamillary bodies, involving both grey and white matter distributed particularly in the thalamus, brainstem, cerebral white matter and cerebellum. Treatment is based on aggressive immunosuppression but there is a high mortality and patients are often left with neurological sequelae. A rare genetic form (relapsing necrotising encephalopathy) is associated with a pathogenic mutation in the RANBP2 gene on chromosome 2q. RANBP2 is a nuclear pore protein involved in numerous cellular functions.

Brainstem (Bickerstaff) Encephalitis

Bickerstaff encephalitis is a form of rhombencephalitis is characterised by a subacute onset of ataxia, ophthalmoplegia and impaired level of consciousness. Focal brainstem signs include pupillary abnormalities, facial weakness, bulbar symptoms or signs and

generalised weakness mimicking Guillain–Barré Syndrome (GBS). MRI may show brainstem fluid attenuated inversion recovery (FLAIR)/T2 hyperintensity and elevated CSF white cell count is also present in about half the patients. The condition is related to Miller Fisher syndrome and GQIb ganglioside antibodies are often positive in both conditions. The condition is monophasic and the outcome is generally good although full recovery may be prolonged.

Distinction Between Bickerstaff Encephalitis and Miller Fisher Syndrome

Both conditions describe patients with ataxia and ophthalmoplegia preceded by infection. Bickerstaff described a group of patients manifesting altered consciousness and hyperreflexia supporting a brainstem origin. Miller Fisher’s patients were areflexic, consistent with a peripheral localisation. The finding of GQIb antibodies in both groups suggests that GQIb and encephalitis should be considered a spectrum disorder of peripheral and central involvement. Metabolic Encephalopathy Acute metabolic encephalopathy can be considered as a condition of global cerebral dysfunction in the absence of primary structural brain disease. The diagnosis includes delirium and acute confusional states. The pathophysiology of metabolic encephalopathy varies according to the cause but all forms interfere with the function of the ascending RAS and its cortical projections leading to impairment of arousal and awareness. Metabolic encephalopathy can occur as a result of organ failure (respiratory, cardiac, hepatic, renal), homeostasis disturbances (abnormalities in electrolytes, blood glucose, osmolarity and acid-­base balance and nutritional causes), or exposure to endogenous or exogenous toxins and drugs (medicinal or recreational) that can affect the CNS. It can also be complicated by, or associated with, epileptic seizures that can further complicate the clinical metabolic picture, however, seizures in this toxic-­metabolic context are usually symptomatic and less likely to develop into epilepsy. The causes are summarised in Box 28.10. All causes of metabolic encephalopathy are characterised by impaired and fluctuating attention and cognition from mild memory loss to severe delirium or coma. Acute metabolic encephalopathy may present as an acute confusional state or delirium characterised by impaired attention ranging from subtle cognitive disturbance to florid delirium or coma with marked fluctuations in the level of alertness and consciousness varying from apathy and withdrawal to anxiety and agitation. There may be a disturbed sleep–wake cycle,

Disorders of Consciousness and Intensive Care Neurology  1025

Box 28.10  The causes of metabolic encephalopathy requiring admission to a neurocritical care unit. Systemic organ failure • Hepatic encephalopathy • Renal failure/dialysis • Pulmonary failure causing hypoxaemia and/or hypercarbia • Cardiac failure • Haematological – thrombocytosis, polycythaemia, blast crisis, hypereosinophilia Metabolic derangement • Electrolytes: elevated or depressed levels of sodium, calcium, magnesium, phosphate • Endocrine disturbances • Thyroid • Parathyroid • Pancreas • Pituitary • Adrenal • Hyperglycaemia/hypoglycaemia • Hypoxaemia/hypercarbia • Hyper-­and hypo-­osmolar states Nutritional • Wernicke, B12 deficiency • Niacin deficiency (Pallegra) • Folate deficiency • Acute refeeding syndrome (hypophosphataemia) Inborn errors of metabolism Infection • Septic encephalopathy Drugs • Prescription medication – opioids, sedatives, hypnotics, antipsychotics, lithium, anticholinergics • Ethanol • Heroin • Cocaine • Amphetamine • Hallucinogens • Medication adverse effects • Hyperammoniacal states • Neuroleptic malignant syndrome • Serotonin syndrome • Cholinergic crisis Toxins • Heavy metal – arsenic, lead • Organic poison • Ethylene glycol, methanol • Cyanide • Carbon monoxide • Organophosphate Other brain disorders • Hypertensive crisis • Posterior reversible encephalopathy syndrome • Osmotic demyelination

sensory misperception, hallucinations, rambling speech, disorganised thought or paranoid ideation, impaired memory and disorientation. There may be a lack of fixed focal deficit, tremor, asterixis, brisk reflexes, bilateral extensor plantar responses, multifocal myoclonus and relative sparing of the cranial nerves. Autonomic instability is associated with worsening delirium. Seizures may occur and can be focal, multifocal or complex partial. It is important to note that seizures may be non-­convulsive in acute metabolic encephalopathy. Marked variation in the mental status over times is characteristic. While metabolic encephalopathy is often treatable, the clinical course can be protracted, with neurological recovery often lagging behind recovery of the underlying condition. The impact of all these causes is more pronounced in unwell patients with underlying acute and systemic medical/surgical ­conditions, the elderly with urinary retention and constipation or those with poor cognitive reserve. Patients with vascular, neurodegenerative or certain genetic disorders are particularly vulnerable, such as patients with small vessel disease, Alzheimer’s or Parkinson’s disease, and mitochondrial disorders. The pathophysiology of metabolic encephalopathy is complex and varies depending on the nature of the underlying cause(s). Metabolic encephalopathy is commonly encountered in acute medical setting, constituting a substantial proportion of referrals to neurology. Its cause, however, can be unclear at presentation and often requires systematic approach, and detailed clinical and laboratory analyses before the correct diagnosis is made and appropriate therapy instituted. Septic Encephalopathy Septic encephalopathy is the most common form of encephalopathy encountered in intensive care medicine and is present in 50–70% of septic patients. It is characterised by disorientation, delirium, impaired consciousness or coma. There may be rigidity, tremors and seizures. Usually, an extracranial site of infection can be identified and appropriate treatment commenced, but in fewer than 50% of cases is an organism isolated from blood cultures. The presence of septic encephalopathy has been particularly associated with COVID-­19 infection and correlates with morbidity and mortality. The pathophysiology of septic encephalopathy is multifactorial and includes focal ischaemia related to micro-­circulatory abnormalities, altered blood-­brain barrier permeability, inflammatory cytokines and neurotransmitter perturbation. EEG may show triphasic waves and burst suppression, although these features are relatively non-­ specific. EEG abnormalities ­however correlate well with the severity of encephalopathy and morbidity and mortality. There is a close association between the  development of septic encephalopathy and critical illness neuromyopathy. Uraemic Encephalopathy Uraemic encephalopathy, dialysis disequilibrium and dialysis dementia are distinct conditions. Uraemic encephalopathy can occur as a consequence of renal failure particularly if the onset is acute or if there is coexisting hepatic failure. Clinical features of uraemic encephalopathy are non-­specific with fatigue, insomnia, pruritus and progressive cognitive impairment culminating in asterixis, tetany, myoclonus, confusion, seizures, stupor and coma. Progression mirrors the severity of uraemia although coexisting metabolic and endocrine factors may also be present (e.g. hypocalcaemia, hyperphosphotaemia, hypokalaemia and metabolic acidosis).

1026

Neurology: A Queen Square Textbook

Uraemic encephalopathy is easily reversible with dialysis. Imaging in uraemic encephalopathy is frequently normal but on T2 and FLAIR images there may be be a bilateral symmetrical or asymmetrical hyperintense rim delineating the white matter surrounding the basal ganglia, and the internal and external capsules resembling a fork (the lentiform fork sign). Dialysis disequilibrium syndrome generally follows the initiation of treatment with haemodialysis and is probably caused by the development of cerebral oedema resulting from a urea gradient in which brain urea concentration exceeds that in the blood. It may be associated with milder manifestations including disorientation, tremor or more severe features such as seizures and coma. Dialysis dementia is now extremely rare. It was thought to be related to aluminium toxicity caused by dialysis. This led to progressive and permanent memory loss, dysarthria, facial grimacing, myoclonus and psychiatric sequelae. Hepatic Encephalopathy Hepatic encephalopathy may develop due to acute liver failure associated with cerebral oedema or as a consequence of portosystemic shunting in chronic liver disease. Common precipitants of chronic hepatic encephalopathy include high-­protein diet, diuretics, gastrointestinal bleeding and medication including benzodiazepines and opiates. Hepatic encephalopathy is primarily associated with ammonia-­induced neurotoxicity but multiple mechanisms may be implicated. Hepatic encephalopathy presents with lethargy, reversed sleep– wake cycle and somnolence, inattention, disorientation and apathy. More severe hepatic encephalopathy may manifest as anxiety with paranoia and agitation, culminating in a progressive decline in the level of consciousness associated with raised ICP due to cerebral oedema leading to coma. On examination there may be disorientation, inattention and difficulties with progressive incoordination, tremor and asterixis with dysarthria. Worsening raised ICP may lead to tentorial herniation with brainstem (ocular bobbing, dysconjugate eye movements and tonic downward deviation) and focal limb signs including hemiparesis and extensor posturing. Cerebral oedema is caused by ammonia-­induced vasodilation and leads to impaired autoregulation. Features of raised ICP and seizures may occur before the development of deep coma. There are no specific diagnostic liver function test abnormalities although elevated blood ammonia levels and an abnormal EEG (bilateral synchronous delta waves and triphasic waves) indicate hepatic encephalopathy and portal–systemic shunting. CT scan may show diffuse cerebral oedema. On MRI manganese accumulation can be shown on T1-­weighted sequences owing to its paramagnetic effect as bilateral, symmetrical high signal intensity of the globus pallidus and substantia nigra, and less frequently, in the putamina, subthalamic region and adenohypophysis. Less often the white matter of both cerebral hemispheres may be affected. MR imaging features of acute hyperammonaemia include symmetric high signal within the insula (most common), thalamus, and posterior limbs of the internal capsule, and cingulate gyrus on T2/FLAIR, which may be reversible. More severe involvement is associated with diffuse cortical oedema and hyperintensity, typically sparing perirolandic and occipital regions. There may be symmetrical signal hyperintensity with restricted diffusion in the insular and cingulate gyrus and high signal affecting the hemispheric white matter and corticospinal tracts. On susceptibility-­weighted imaging, microhaemorrhages may be seen in the white matter or cortex. In fulminant hepatic failure, the apparent diffusion coefficient is decreased, suggesting intramyelinic and intracellular oedema with

acute astrocytic swelling, and magnetic resonance spectroscopy shows specific patterns reflecting intra-­ astrocyte glutamate accumulation. The management of chronic hepatic encephalopathy requires avoidance of precipitating factors and the use of lactulose and antibiotics, particularly neomycin. Severe acute hepatic encephalopathy caused by fulminant hepatic failure is an ICU emergency involving the management of cerebral oedema, acute elevated ICP and seizures. It requires monitoring, ventilation and osmotherapy with mannitol. Hypothermia and hyperbaric oxygen have also been used. The only proven therapy to improve outcome is orthotopic liver transplantation. Hypernatraemia Hypernatraemia may be caused by unreplaced water loss (e.g. insensible sweat loss), gastrointestinal loss, cranial or nephrogenic diabetes insipidus, osmotic diuresis (e.g. glucose in uncontrolled diabetes mellitus or mannitol). It may also be caused by sodium overload (intake of hypertonic sodium solution). Slowly developing mild hypernatraemia is not a concern to cerebral function because brain cells maintain volume but patients with serum osmolality greater than 350 mOsm/kg or with serum sodium above 160 mEq/l may develop lethargy, confusion, decreased mental status and occasionally seizures. Severe hypernatraemia causes water movement out of the brain, resulting in a decrease in brain volume, which may cause new focal intracerebral or subarachnoid haemorrhage and venous sinus thrombosis. The rate of correction should be slow. Hyponatraemia Hyponatraemia is considered acute if it has developed over a period less than 48 hours. Acute hyponatraemia usually results from parenteral fluid administration in postoperative patients or from self-­induced water intoxication (e.g. competitive runners, psychiatric patients with extreme polydipsia and users of ecstasy). It is considered chronic if it has been present for more than 48 hours. This is more common and hyponatraemia should be presumed to be chronic when the duration is unclear. The classification and causes of hyponatraemia are shown in Table 28.19. Encephalopathy resulting from hyponatraemia depends on the rate of development and the absolute sodium levels. Clinical manifestations include nausea and malaise, headache, muscle cramp, weakness, lethargy, obtundation and eventually seizures, coma and respiratory arrest due to cerebral oedema and tentorial herniation. Presentation may be less severe with chronic hyponatraemia because cerebral adaptation prevents rapid changes in cell volume, even when the sodium concentration is low. The usual physiological defence against hyponatraemia is the capacity to excrete large volumes of urine with low concentration of sodium and potassium associated with suppression of the release of antidiuretic hormone. Patients who develop hyponatraemia typically have impairment in renal water excretion, most often due to an inability to suppress ADH secretion. An exception is in patients with primary polydipsia, who can become hyponatraemic because they rapidly drink such large quantities of fluid that they overwhelm the excreting capacity of the kidney even though ADH release is appropriately suppressed.

Syndrome of Inappropriate Antidiuretic Hormone Secretion

The syndrome of inappropriate ADH (SIADH) secretion causes normovolaemic hypotonic hyponatraemia due to excessive water retention by the kidneys leading to dilutional hyponatraemia. The secretion of ADH is not appropriately suppressed by low plasma

Disorders of Consciousness and Intensive Care Neurology  1027

Table 28.19  Classification of hyponatraemia. Class

Symptoms

Causes

Hypovolaemic

Renal sodium loss (urine sodium > 20 mmol/l)

Diuretics, osmotic diuresis, adrenocortical deficiency, cerebral salt wasting, salt-­losing nephritis, renal tubular acidosis, post relief of urinary obstruction

Extrarenal sodium loss (urine sodium  100 Osm/kg) • elevated urinary sodium (> 20 mmol/l) • low serum uric acid concentration due to urea wasting in the urine. In clinical practice, the only reliable way of distinguishing SIADH from CSW is assessment of the volume status. Hypovolaemia, as

seen in CSW, is associated with orthostatic hypotension, tachycardia, dry mucous membranes, poor skin turgor and reduced central venous pressure, but none of these signs is entirely reliable.

Management of Hyponatraemia

Rapid correction of hyponatraemia is a hypotonic stress to ­astrocytes which triggers apoptosis, disruption of the blood–brain barrier and eventually, demyelination. Fatal brain swelling has only rarely been reported in acute hyponatraemia and empirical observation indicates that, although some correction of hyponatraemia is usually indicated in patients with severe acute hyponatraemia, the goal of therapy to raise the serum sodium concentration by 4–8 mg/l over 24 hours is generally adequate to reverse impending brain

1028

Neurology: A Queen Square Textbook

herniation or prevent active seizures in patients with severe acute hyponatraemia. Such an increase can be reliably achieved 100 ml of bolus infusion of 3% saline administered at 10-­minute intervals, if necessary, to a total of three doses. The subsequent limit of replacement should remain 4-8 mmol/l/day, erring to the lower side for those patients at high risk of osmotic demyelination syndrome (alcoholism, malnutrition, hypokalaemia or liver disease). Acute symptomatic hyponatraemia is often due to hypervolaemic states such as congestive cardiac failure or nephrotic syndrome. Treatment is therefore with loop diuretics to promote renal loss of water in excess of sodium. The distinction between SIADH and CSW is essential because the management of SIADH involves reduction in the intake of electrolyte-­free water to reduce the expansion of intracellular volume. The administration of saline would simply increase the extracellular fluid as sodium continues to be lost as a consequence of continuing ADH secretion and the shift of sodium from the extracellular fluid to the intracellular fluid would cause progressive cerebral oedema. Conversely, in CSW, where there is a hypovolaemic state, treatment is the administration of both salt and water with isotonic or hypertonic saline to counteract the primary abnormality of renal salt wasting. In this situation, fluid restriction actually worsens the hypovolaemia and hyponatraemia as the salt wasting continues and may increase the risk of cerebral infarction. The situation is particularly complex in SAH, where there appear to be features of both SIADH and CSW. However, the hypovolaemic state exacerbates vascular spasm and therefore it is now widely accepted that hypervolaemic therapy is appropriate to prevent further vascular spasm.

Osmotic Demyelination

Osmotic demyelination (previously called central pontine myelinolysis) is a demyelinating syndrome, initially described in patients with alcoholism or who are malnourished. The condition is associated with over-­rapid correction of hyponatraemia, which leads to

(a)

shift of fluid from the intracellular to the extracellular compartments causing dehydration of the brain resulting in non-­inflammatory damage to the myelin sheath and the oligodendrocytes with relative sparing of neurons and axons. The clinical course is biphasic, beginning with encephalopathy caused by hyponatraemia, which improves after initial elevation of sodium. This is followed by the development of new neurological signs after two to three days, characterised by seizures and spastic tetraparesis leading to coma or death without medical intervention. The condition may present with a severe and unexplained encephalopathy or seizures occurring secondary to hyponatraemia but it may progress to flaccid quadriparesis, pseudobulbar palsy manifest as dysarthria, dysphagia with pupillary and oculomotor abnormalities and, occasionally, patients may present with a locked-­ in syndrome. The presence of extrapontine myelinolysis leads to a more diffuse neurological pattern with movement disorders including parkinsonism, choreoathetosis, dystonia and spasmodic dysphonia. Demyelination is most frequently seen in the central pons, but may extend into the midbrain; however, extrapontine myelinolysis is often symmetrical and may occur with concomitant or isolated involvement of extra-­pontine sites, including basal ganglia, midbrain, thalamus, cerebellum and cerebral white matter. Imaging changes may be delayed by one to two weeks after symptom onset. MRI confirms the presence of hyperintense, symmetrical ‘bat-­wing’ lesions on T2-­weighted images, which do not enhance and are seen primarily in the central pons (Figure 28.22). Diffusion-­ weighted imaging is a more sensitive technique to show these abnormalities. The outcome is variable and the prognosis depends on the severity and extent of demyelination in the pons. Some patients may recover from the condition despite extensive radiological or clinical involvement. Osmotic demyelination syndrome does not occur if acute hyponatraemia which has been present for less than 24 hours is rapidly corrected (e.g. marathon runners, psychogenic polydipsia, users of ecstasy).

(b)

Figure 28.22  Osmotic demyelination (central pontine myelinolysis). T2-­weighted (a) and diffusion-­weighted (b) axial images demonstrating abnormally high signal and restricted diffusion (white arrow) within the central pons in a typical trident pattern.

Disorders of Consciousness and Intensive Care Neurology  1029

Posterior Reversible Encephalopathy Syndrome Posterior reversible encephalopathy syndrome (PRES) is a clinical radiological syndrome characterised by reversible subcortical vasogenic cerebral oedema predominantly involving the parieto-­occipital region. It occurs in various conditions including acute hypertensive encephalopathy, eclampsia, renal failure, sepsis, autoimmune disorders and immunosuppressive and cytotoxic drugs (Box 28.13). The differential diagnosis is reviewed in Table 28.20. Onset is with headache, dizziness, vomiting and a progressive encephalopathy characterised by confusion and disorientation. There may be seizures (convulsive and non-­convulsive), visual ­disturbance (hemianopia, blurred vision, visual neglect, visual hallucinations, cortical blindness), altered mental state (confusion, drowsiness, coma), hemiparesis, aphasia/dysarthria and ataxia. It is characterised by consecutive disruption of the blood–brain barrier and vasogenic oedema. It is suggested that a precipitate increase in blood pressure leads to impaired cerebral autoregulation followed by hypoperfusion, endothelial injury, breakdown of the blood–brain barrier and secondary vasogenic oedema. In general, changes are confined to the posterior circulation territory, perhaps because the vascular blood in this area has lower pressure and less sympathetic innervation compared with the anterior circulation and is therefore more easily affected by systemic hypertension. However, anterior territory and asymmetrical changes are increasingly recognised. Intracranial haemorrhage is a frequent complication in PRES and may be characterised by intraparenchymal or lobar haemorrhage, sulcal SAH and microhaemorrhages.

CT may be normal or may show scattered hypodense areas involving bilateral cerebral hemispheres in up to half of cases. Haemorrhagic change may be seen in 15% of patients. MRI shows characteristic high intensity white matter abnormalities on T2-­weighted sequences in the parieto-­occipital territory. FLAIR sequences show a similar predominantly white matter pattern of hyperintensity consistent with vasogenic oedema (Figure 28.23), which may be localised to a or generalised throughout the hemisphere. Changes are most frequently seen in the parietal-­occipital white matter (98%), followed by the frontal lobe (68%), inferior temporal region (40%) and cerebellum (32%). Vasogenic oedema can also involve the deep white matter, splenium, basal ganglia, thalami, brainstem and pons.

Table 28.20  Differential diagnosis of posterior reversible encephalopathy syndrome. System

Diagnoses

Vascular

Cerebral ischaemia Cerebral venous sinus thrombosis Intracerebral haemorrhage Central nervous system vasculitis Reversible cerebral vasoconstriction syndrome

Autoimmune, demyelination, degenerative

Multiple sclerosis/acute disseminated encephalomyelitis Autoimmune encephalitis

Box 28.13  Causes of posterior reversible encephalopathy syndrome.

Limbic encephalitis

• Immunosuppression/chemotherapy: ◦◦ Tacrolimus ◦◦ Cyclosporin ◦◦ Mycophenolate ◦◦ Cisplatin ◦◦ 5 fluorouracil ◦◦ Methotrexate ◦◦ Interferon α • other drugs: ◦◦ Amphotericin B • Autoimmune disorders: ◦◦ Systemic lupus erythematosus ◦◦ Thrombotic thrombocytopenic purpura ◦◦ Primary sclerosing cholangitis ◦◦ Rheumatoid arthritis ◦◦ Sjögren syndrome ◦◦ Polyarteritis nodosa ◦◦ Systemic sclerosis ◦◦ Granulomatous with polyangiitis ◦◦ Crohn’s disease ◦◦ Neuromyelitis optica ◦◦ Infection/sepsis • Pre-­eclampsia/eclampsia • Toxic: ◦◦ Cocaine • Metabolic disorders: ◦◦ Acute renal failure

Rasmussen’s

Hashimoto’s

Systemic lupus erythematosus Behçet syndrome CADASIL MELAS Creutzfeldt–Jakob disease Infection

Herpes simplex encephalitis Progressive multifocal leukoencephalopathy Other viral encephalitis

Cancer

Lymphoma, gliomatosis cerebri, metastatic disease Paraneoplastic encephalitis Chemotherapy Radiotherapy

Metabolic

Metabolic encephalopathy Osmotic demyelination syndrome Toxic encephalopathy

CADASIL, cerebral autosomal dominant arterial the with subcortical infarcts and leukoencephalopathy; MELAS, mitochondrial encephalopathy, lactic acidosis with stroke-­like episodes.

1030

Neurology: A Queen Square Textbook

(a)

(b)

Figure 28.23  Posterior reversible leukoencephalopathy in a patient who was hypertensive following the administration of intrathecal methotrexate. magnetic resonance T2-­weighted image: (a) following intrathecal methotrexate; and (b) three months later showing complete resolution. Source: courtesy of Dr Paul Holmes, St Thomas’ Hospital, London.

Restricted diffusion may occur in a smaller number of cases and represents the earliest sign of non-­reversibility as severe vasogenic oedema progresses to cytotoxic oedema and therefore indicates a poor prognosis. Diffusion-­weighted imaging is important in helping to distinguish PRES from posterior circulation stroke. Management involves aggressive lowering of the elevated blood pressure with intensive care monitoring, withdrawal of immunosuppression and supportive care. This condition is usually reversible with treatment and the MRI may normalise but there may be residual focal cerebral haemorrhage or permanent injury. Generally, PRES has a good short-­and long-­term prognosis if there is rapid treatment and clinical symptoms in MRI appearances are reversed within hours or days. However, cerebral haemorrhage and ischaemia may lead to irreversible neurological deficits.

(a)

(b)

Mortality and morbidity are also dependent on the underlying ­condition and patients receiving chemotherapy for cancer have a particularly poor prognosis.

Hypoxic–Ischaemic Brain Injury

HIBI usually follows cardiac arrest and carries a very poor prognosis. The history of the event and the presence of prearrest morbidity are important prognostic factors but detailed examination of the patient remains the mainstay of assessment. The assessment of HIBI remains primarily clinical and is made more difficult by the use of more sophisticated techniques of sedation, ventilation, hypothermia, neuromuscular blockade and haemodynamic management (Figure 28.24).

(c)

Figure 28.24  After hypoxic–ischaemic brain injury, magnetic resonance image showing changes including high signal in the caudate and putamen, less so in the thalami. (a) T1. (b & c) Diffusion-­weighted sagittal image with restricted diffusion on apparent diffusion coefficient in the occipital regions and peri-­rolandic cortex.

Disorders of Consciousness and Intensive Care Neurology  1031

Hypoxia and ischaemia should be considered as pathologically and clinically distinct patterns of brain injury although they usually coexist. Ischaemia describes a reduction in blood supply leading to decreased oxygen delivery and limited or no removal of damaging cellular metabolites, which therefore accumulate (e.g. lactate, H+, glutamate) leading to severe brain injury. Hypoxaemia refers to a reduction of oxygen supply while hypoxia refers to reduced oxygen availability for use by tissues and in respiration. It can develop as a direct consequence of reduced oxygen supply, reduced ambient partial pressure of oxygen, low haemoglobin or impaired tissue use following poisoning of the mitochondrial cytochrome enzymes (e.g. due to cyanide). The pathophysiology of carbon monoxide poisoning is complex, with components of both anaemic and histotoxic hypoxia, and occasionally global ischaemia caused by cardiac failure. Following isolated hypoxia, there is an increase in CBF, which allows continuing delivery of glucose to the brain and clearance of toxic metabolites from the brain. Therefore, hypoxia that occurs in isolation, even if prolonged, rarely causes severe brain injury, provided the systemic circulation is preserved. Prognostic Factors Following Cardiac Arrest The long-­term outcome is critically dependent on the cause of arrest and the periarrest management. The outcome is best for patients with ventricular tachycardia or fibrillation, in whom resuscitation is commenced rapidly. The likelihood of successful resuscitation in asystole or pulseless electrical activity is much lower. In primary hypoxic arrest immediate restoration of oxygenation is of overwhelming importance to outcome. The most important prognostic factors following out-­of-­hospital cardiac arrests are: • Age • Comorbidity • Circumstances of arrest • Witness/immediate bystander resuscitation • Effective cardiopulmonary resuscitation • Early attendance of paramedics • Cardiac rhythm • Resuscitation • Fever within the first 48 hours • Duration of cardiac arrest • Advanced directive. The current recommendations for initial post-­resuscitation care include artificial ventilation, sedation and targeted therapeutic management at 33–36°C for 24 hours, initiated within 6 hours of cardiac arrest. Recent studies have suggested that induced controlled hypothermia has no effect on outcome and the treatment is now being abandoned in both TBI and following HIBI. The outcome of HIBI worsens if: • The patient has been in coma (i.e. unresponsive) for more than six hours. • Lack of spontaneous limb movements or failure to localise painful stimuli in the initial stages. • Prolonged loss of pupillary responses (if atropine has not been administered). • Sustained conjugate eye deviation (upgaze or downgaze). • Specific forms of abnormal eye movements (including upbeat and downbeat nystagmus, ping pong gaze or period alternating nystagmus). • Myoclonic seizures. • Involvement of lower cranial nerve function.

However, these signs are variable and are dependent on the medication used during resuscitation. Investigations The EEG has been widely used over many years to assess the level of consciousness and to guide prognosis after HIBI. A number of patterns suggest a poor prognosis: • Generalised electrical suppression • Generalised burst suppression • Unresponsive alpha, theta or delta rhythms • Overt clinical seizures, status epilepticus or epileptiform EEG changes • Periodic patterns: ◦◦ Periodic lateralised epileptiform discharges ◦◦ Bilateral independent or synchronous. Focal or generalised convulsive tonic–clonic seizures are relatively unusual in the initial stages following HIBI but may evolve during the recovery period. Status epilepticus occurring after HIBI responds poorly to conventional antiepileptic drugs including phenytoin. The return of continuous baseline activity is a favourable sign. Post-­ hypoxic myoclonic status may develop immediately after resuscitation, when it typically causes bilateral synchronous local or generalised multifocal jerking of the face, limbs, trunk or diaphragm. The EEG shows limited background activity with a burst suppression pattern and intermittent generalised periodic complexes with no cortical focus. It responds poorly to medication and carries a uniformly bad prognosis. A different form of myoclonus (Lance–Adams syndrome) comes on after a latent period of 24–48 hours after resuscitation. It often follows a primary respiratory arrest or anaesthetic event and usually occurs in a younger age group. Consciousness is usually less deeply impaired, and focal myoclonus is often action or startle sensitive. The prognosis of the Lance–Adams form of post-­hypoxic myoclonus is generally favourable and these patients continue to improve over time, although cerebellar signs including ataxia, dysarthria and intention tremor may persist. The EEG shows a focal cortical origin with responsive cortical rhythms which progressively regain normal patterns. The condition responds reasonably well to antiepileptic drugs including valproate, piracetam, levetiracetam and clonazepam.

Neuroimaging

In the first two days after HIBI, CT may show diffuse swelling with effacement of the basal cisterns, ventricles and sulci, attenuation of the grey–white matter interface and hypodensity of the cortical grey matter and basal ganglia (caudate, lenticular nucleus, thalamus and putamen) resulting from cytotoxic oedema. Focal areas of infarction may develop in the basal ganglia or cortical watershed territories. MRI is undertaken less commonly following HIBI because patients often require sedation, ventilation and airway protection. In the acute stages (first few days) after severe HIBI, diffusion-­ weighted and FLAIR images show widespread hyperintensity initially involving the basal ganglia, caudate, striatum and thalamus followed by the cortex and subcortical white matter, cerebellum and hippocampus. Conventional T1-­and T2-­weighted scans are normal. Characteristic patterns are shown in Figures 28.25 and 28.26. The pattern of change on MRI involves and can help to guide prognosis. Patterns associated with a poor clinical outcome include involvement of: • Diffuse cortical and deep grey matter with and without perirolandic sparing • Medial occipital territory including perirolandic area

1032

Neurology: A Queen Square Textbook

(a)

(b)

(c)

Figure 28.25  After severe circulatory compromise, for example following problematic cardiac surgery, a pattern of border-­zone ischaemia may occur. Magnetic resonance image showing high signal on T2 (a) and fluid attenuated inversion recovery (b) in the deep border-­zone regions, as well as high signal on diffusion-­weighted imaging (c).

(a)

(b)

(c)

Figure 28.26  Diffuse white matter high signal following hypoxic–ischaemic brain injury (HIBI). T2 (a), DWI (b), ADC map (c).

• • • • •

Precentral gyrus Diffuse white matter Brainstem Cerebellar Hippocampus. A more favourable clinical outcome is associated with border-­ zone patterns (see below) and isolated basal ganglia involvement without cortical change. Changes of cortical laminar necrosis (involvement of the entire cortex) pseudolaminar necrosis (involvement of mid and deep layers 3, 4 and 5) may be seen as a consequence of neurons within the cortex being far more metabolically active than glial cells or adjacent white matter stop early cytotoxic oedema causes high signal on diffusion-­weighted imaging corresponding low ADC values in the affected cortex. Cortical enhancement may be seen later but intrinsic T1 signal increases the most specific imaging finding, becoming apparent after several days and fading over months. T2 images show either increased signal or isointensity. Diffusion tensor imaging can also show reduced axial diffusion in

central regions and cerebral hemispheres and magnetic resonance spectroscopy may show elevated lactate and reduced N-­ acetyl aspartate in affected areas.

Somatosensory Evoked Potentials

Somatosensory evoked potentials are valuable in assessing the prognosis following HIBI. Bilateral absence of the cortical N20 response represents widespread cortical injury after cardiac arrest with a high specificity but low sensitivity for poor outcome. Consideration of the N20  amplitude (range less than 0.65  μV) might improve sensitivity, but this still remains unreliable.

Blood Biomarkers

Blood biomarkers break down products from neurons; astrocytes measured in the serum after cardiac arrest are of uncertain value. High serum neuron-­specific enolase concentration within the first 48–72 hours may be more valuable than S100. Neurofilament light chain is promising as a biomarker.

Disorders of Consciousness and Intensive Care Neurology  1033

Prognosis of Brain Injury After Hypoxic–Ischaemic Brain Injury Assessment involves balancing concerns that an overly conservative prognosis could leave patients in a severely disabled state with the possibility that an inaccurately pessimistic prognosis could lead to  the withdrawal of life-­ sustaining treatment in patients who might  otherwise have a good functional outcome. It is important to recognise that delivering a poor prognosis typically leads to the withdrawal of life-­sustaining therapy and death. Hence, the specificity of the prediction should be maximal. Clinical Examination Clinical examination should be undertaken serially, where possible, without residual sedation and conclusions about prognosis should be delayed by at least 72 hours after a rest to allow clearance of sedative drugs. Assessing the prognosis for patients who have sustained HIBI due to cardiac arrest is increasingly guided by multimodal algorithms which are rapidly developing to reflect advances in clinical assessment and investigation. At present, it is recommended that the diagnosis of irreversible brain damage cannot be made at less than 72 hours from the cardiac arrest and require at least two of the following: • Absent brainstem reflexes • Myoclonus developing   30% suggests diaphragm impairment) • Inspiratory and exploratory mouth or sniff pressures • Peak cough flow – ability to clear airway through measurement of cough flow Critical values to consider intubation according to the 20/30/40 rule are: • FVC less than 20 ml/kg body weight • maximum inspiratory pressure  90%

Light sleep (stage 1 and 2)

50–70%

Deep sleep (stage 3)

15–25%

REM sleep

15–25%

Sleep onset

> 10 minutes

REM sleep onset

60–90 minutes

REM, rapid eye movement.

Sleep and Breathing Slow-­wave sleep leads to an increase in parasympathetic tone and decrease in sympathetic tone, which in turn leads to a reduction in heart rate, blood pressure and respiratory rate. REM sleep is associated with autonomic instability causing fluctuations in heart rate and blood pressure. The muscles of the upper airway relax during

all stages of sleep (especially in REM sleep) causing snoring and predisposing to the significant airway obstruction that underlies sleep apnoea or hypopnoea syndrome. Respiratory failure resulting from neuromuscular weakness is also exacerbated during REM sleep because of muscular relaxation involving upper airway, intercostal muscles and the diaphragm.

Classification of Sleep Disorders

Sleep medicine is an evolving field with wide variability in ­knowledge of each individual sleep disorder. Box  29.1 lists the most  recent classification that is based on common complaints (e.g.  insomnia or hypersomnias) as well as presumed aetiological basis (e.g. circadian rhythm disorders) or organ system from which problems arise (e.g. sleep-­related breathing disorders). The classification system is currently under review.

Insomnia

Insomnia is the complaint of ‘poor sleep’. Insomnia is defined as difficulty initiating or maintaining sleep. These difficulties lead to  daytime consequences such as fatigue, cognitive dysfunction, low mood, accidents and/or dissatisfaction with sleep. Insomnia is

Disorders of Sleep  1051

Box 29.1  Classification of sleep disorders. 1 Insomnia 2 Sleep-­related breathing disorders 3 Central disorders of hypersomnolence 4 Circadian rhythm sleep-­wake disorders 5 Parasomnias 6 Sleep-­related movement disorders 7 Other sleep disorders

In addition to reducing quality of life, insomnia and poor overnight sleep often contribute to the worsening of daytime neurological symptoms and seizures. Insomnia can, exceptionally, result from prion disease. Fatal familial insomnia is a very rare hereditary prion disease caused by a mutation in codon 178 of the prion protein gene. Fatal familial insomnia can also occur sporadically. It is characterised by insomnia, loss of slow-­ wave sleep, oneiric episodes (daytime dream ­episodes and enactment) and progressive neurological decline.

Management A critical aspect of management is to treat the underlying cause (e.g. anxiety, depression). In addition, the first line is to teach, ­establish and maintain good sleep hygiene with regular sleep times, the avoidance of nocturnal stimulants, stressful activity or exercise close to bedtime, prolonged daytime napping and establishing extremely common and is said to merit treatment in 5–15% of the ­comfortable and regular sleep arrangements. Regular exercise not normal population. It is often associated with daytime fatigue but close to bedtime can increase and consolidate deep sleep. Cognitive rarely daytime sleepiness and patients will usually have difficulties behavioural therapy (CBT) has been shown to be the most effective falling asleep in the day, even if given the opportunity to nap. treatment for chronic insomnia and is first-­line treatment. This can Insomnia can have pronounced psychiatric consequences. For be administered either individually or in group sessions. Access to example, insomnia, in adults without prior psychiatric symptoms, appropriate CBT for insomnia may be limited and a number of increase the risk for developing depression, anxiety disorders and internet-­based applications have been developed and the benefit of substance abuse. Interestingly, treating depression does not neces- these have been shown in a recent review. Most drugs used in insomnia act as agonists at the benzodiazesarily result in relief from insomnia and residual insomnia is a risk factor for depressive relapse. However, if patients have insomnia pine site of the γ-­aminobutyric acid type A (GABAA) receptor; these and depression, treating insomnia can effectively treat depression drugs have effects other than their sedating action, including muscle as well. Although insomnia can be idiopathic, in many instances relaxation, antiepileptic effects, anxiolytic effects, memory impairment, behavioural disturbance (especially in children) and ataxia. there is an underlying cause. Several factors may contribute to insomnia: psychological diffi- Some drugs have a more specific action (e.g. zolpidem, ­zaleplon and culties, a variety of prescribed medications (particularly beta-­ zopiclone) by targeting only specific GABAA receptor subtypes. blockers) and non-­prescribed drugs including caffeine, alcohol and Drugs with longer duration of action may affect psychomotor persubstances of abuse, particularly cocaine and amphetamine. Sleep is formance, memory and concentration, and also have prolonged also disrupted by depression, anxiety, stress, pain, cognitive impair- anxiolytic and muscle relaxing effects. Alternatively, benzodiazment, pregnancy, nocturia and other medical disorders (especially epines with too short a half-­life may result in rebound insomnia and respiratory and cardiac disorders). Insomnia may occur as a conse- daytime hyperexcitability. Hypnotic drugs such as zolpidem and quence of other sleep disorders (e.g. restless legs syndrome or sleep specific benzodiazepines are indicated for the management of acute apnoea) causing interrupted sleep architecture. Other sleep disor- insomnia or in the short-­term management of chronic insomnia. ders such as circadian rhythm disorders, in particular, delayed sleep Long-­acting melatonin is licensed for treatment of insomnia in phase syndrome, may mimic insomnia as the natural sleep time is people over 55 years of age. Recently, daridorexant, a dual orexin receptor antagonist, has been licensed for treatment of insomnia. delayed relative to the ‘norm’. Insomnia is common in patients with neurological conditions. Antidepressants are commonly prescribed for treatment of insomUp to 40% of patients with multiple sclerosis experience insomnia. nia but there is limited published evidence. Of the antidepressants, Sleep disturbances in individuals with multiple sclerosis can be the best evidence is for trazodone (anti-­noradrenergic), doxepin and caused by demyelination or degeneration of areas of the brain that mirtazapine but all prescription is off-­label. Additional sleep disorders contributing to sleep initiation (RLS, control sleep or by physical and psychological factors such as pain, spasticity, medication, anxiety, depression and bladder problems. delayed sleep phase syndrome) or sleep maintenance insomnia Insomnia is commonly reported by patients with Parkinson’s dis- (periodic limb movements of sleep [PLMS], obstructive sleep ease, regardless of the severity of the disease, with a reported preva- apnoea [OSA]) should be sought and treated if identified. For patients with neurological disorders, it is important to review lence of 37–83%. All subtypes of insomnia have been reported in Parkinson’s disease, with variations across stages, but sleep mainte- regular medication and the possibility that symptoms due to the nance insomnia due to disrupted sleep appears to be most com- underlying disorder (i.e. dystonia, pain, seizures) may contribute to mon. Worsening of motor as well as non-­motor symptoms (such as insomnia and may warrant interventions. nocturia), psychological factors or medication may contribute to insomnia in Parkinson’s disease. Sleep disorders (REM sleep behaviour disorder, restless legs syndrome [RLS] or periodic limb move- Sleep-­Related Breathing Disorders ments of sleep) may lead to sleep disruption. Insomnia has been Obstructive Sleep Apnoea/Hypopnoea Syndrome reported in 36–74% of adult patients with epilepsy. In addition to OSA/hypopnoea syndrome (OSAHS) is the most common medical psychological factors, poor seizure control, polytherapy and devel- cause of daytime hypersomnolence and is associated with signifiopmental delay are associated with worse insomnia. Some antisei- cant morbidity including an increased incidence of hypertension, zure medications can contribute to insomnia. Nocturnal seizures cerebrovascular disease and road traffic accidents. It is associated may disrupt sleep and fear of seizures overnight may lead to sleep with nocturnal sleep disturbance, unrefreshing sleep, difficulty in initiation insomnia. concentration and nocturnal choking. Sleep apnoea refers to a Source: American Academy of Sleep Medicine (2014). Reproduced with permission.

1052

Neurology: A Queen Square Textbook

c­ essation in airflow during sleep for longer than 10 seconds, while hypopnoea is defined as greater than 50% reduction in airflow associated with oxygen desaturation above 4% or arousal on EEG. More than five apnoeas/hypopnoeas per hour (apnoea/hypopnoea index [AHI]) is considered significant with AHI 5–15 events/hour representing mild, AHI 15–30  events/hour moderate and AHI over 30 events/hour severe OSA. Overnight oximetry is sometimes used but the specificity and sensitivity of the test have been questioned and in general, polygraphy of polysomnography are required for the diagnosis. Obstruction of the upper airway usually occurs between the caudal soft palate and epiglottis. It is worsened during sleep when reduced upper airway muscle tone leads to collapse of the upper airway and obstruction. Risk factors for OSA include body mass index greater than 30 kg/m2, neck circumference greater than 43 cm for men or 41 cm for women, crowding of the oropharynx and jaw or facial structural anomalies, nasal abnormalities or a narrow palate. Patients can have loud snoring, stridor, coughing spells during sleep and restless sleep with frequent awakenings. Severe obstructive sleep apnoea is associated with morning headache, impaired memory, nocturia and anxiety. OSA can be seen in conjunction with neurological disorders and often exacerbate symptoms such as worsening seizure control in epilepsy, worsening cognitive symptoms in dementia and motor symptoms in multiple sclerosis and Parkinson’s disease.

Management

The first-­line management is weight loss, establishing regular sleep patterns and avoidance of alcohol, nicotine, caffeine and sedatives in the evening. Investigation and appropriate treatment of comorbidity, particularly hypothyroidism, is mandatory. Continuous ­positive airway pressure by nasal or face mask is effective in OSAHS. Compliance may be variable in the less severe forms of the condition and its use requires considerable educational support and technical back-­up. Oral appliances such as mandibular advancement splints may be valuable in treating mild OSAHS. They work by holding the mandible forward, thus increasing the upper airway space by advancing the tongue and possibly changing genioglossus activity. Surgical treatment (uvulopalatopharyngoplasty) does not have a role. Tracheostomy may be necessary in patients with incipient cor pulmonale. Central Sleep Apnoea The mechanisms of central respiratory rhythm generation critically depend on chemoreceptor inputs, particularly partial pressure of carbon dioxide in arterial blood (PaCO2) sensitivity. These effects of central respiratory insufficiency are particularly manifest in sleep because, at sleep onset, there is a loss of the wakefulness stimulus and the volitional behavioural influences that affect respiration. In addition, several respiratory control mechanisms are impaired at sleep onset. The muscle tone of the upper airway and respiratory muscles, particularly the diaphragm, is reduced. There is an accompanying increase in upper airway resistance reducing ventilation for a given level of drive. There is also a reduction in chemosensitivity at the onset of sleep which may be manifest as a decrease in the PaCO2 apnoea threshold. These changes are present in normal individuals but become clinically apparent with the development of apnoea if there is a failure in the central drive or weakness of the respiratory muscles. Central sleep apnoea is caused by a lack of drive to breathe during sleep, resulting in insufficient or absent respiratory effort and compromised gas exchange. It is due to a failure in the central generation

of the respiratory rhythm by the brainstem. Central sleep apnoea is associated with both a failure to generate the central respiratory rhythm and progressive hypoventilation due to respiratory muscle weakness, particularly affecting the diaphragm. Ventilation during sleep is characterised by a normal respiratory rate but diminished tidal volume with episodes of prolonged apnoea without respiratory effort. In contrast, OSA is caused by obstruction to airflow and there is continuing respiratory effort despite the apnoea. Considerable overlap exists between obstructive and central apnoea. The clinical manifestations of nocturnal desaturation and hypercapnia are similar in both conditions and include frequent night-­time awakenings, excessive daytime sleepiness and an increased risk of adverse cardiovascular complications and sudden death. A variety of conditions may affect central respiratory drive by causing structural or functional damage to brainstem nuclei (see Chapter  28). A range of neuromuscular disorders are associated with hypoventilation and these may involve the phrenic and intercostal nerves, neuromuscular junction or muscle. Congenital central hypoventilation syndrome is a rare condition in children. A variety of medications have a central respiratory depressant effect, particularly opioids and barbiturates. There are several manifestations of impaired and unstable central respiratory drive. These include high altitude induced periodic breathing, which can occur in healthy individuals, idiopathic central sleep apnoea, drug-­induced central apnoea, obesity hypoventilation syndrome, Cheyne–Stokes, apnoeic, periodic and cluster breathing (Chapter 28). Investigation necessitates polysomnography with respiratory muscle inductance bands or surface EMG, to assess chest wall and abdominal excursion. Newer and more invasive techniques of diaphragm function assessment are now available. Treatment involves management of the underlying condition such as control of obesity and specific therapy for nerve, neuromuscular junction or muscle disorders. Ventilatory support may be necessary using continuous positive airway pressure or non-­invasive bilevel positive airway pressure. Newer techniques of implantable pacing devices have not proved to be sustainable for patients with diaphragm weakness.

Central Disorders of Hypersomnolence

Narcolepsy Narcolepsy is the most common cause of central hypersomnia with an estimated prevalence of 3–5/10,000. It may develop at any age but the peak onset is between 15 and 30 years, with a secondary peak in the fourth decade. Narcolepsy can rarely be familial; a clear Mendelian pattern of inheritance occurs in fewer than 5% of all those affected. However, the lifetime risk for developing narcolepsy is increased in first degree relatives of narcoleptic patients to 1%. The excessive daytime somnolence (EDS) presents with irresistible sleep attacks during the day often occurring at inappropriate times. These naps are usually brief (less than 20 minutes), associated with dreaming and on waking the person no longer feels sleepy. Such naps can occur many times per day. In contrast to idiopathic hypersomnia (see below), the total sleep time over 24 hours is within normal limits but disrupted nocturnal sleep due to narcolepsy is very common. Other symptoms of this syndrome are cataplexy, sleep paralysis, hypnogogic hallucinations or vivid dream-­like images, which characteristically occur at sleep onset. REM sleep behavioural disorder and short periods of automatic behaviour during daytime (micro-­sleeps) are also common. Narcolepsy is divided into type 1 (presence of cataplexy) and type 2 (lack of cataplexy).

Disorders of Sleep  1053

There may be a range of secondary symptoms related to s­ leepiness including difficulties with memory and concentration. In combination, the symptoms of narcolepsy often have a major impact on relationships, education, employment, driving, mood and quality of life.

Investigations

Polysomnography is important in excluding other or coexistent causes of EDS including obstructive sleep apnoea, periodic limb movement disorder; however, all are more common in people with  narcolepsy. Polysomnography should be preceded by one to two weeks of actigraphy to exclude sleep inefficiency as a cause Cataplexy of hypersomnolence. Polysomnography may also demonstrate Cataplexy is the occurrence of brief episodes of muscle weakness or early onset sleep (