Optical Coherence

Tomography in Neurological
Diseases
Edited by
Peter A. Calabresi, MD, FAAN, FANA
Professor of Neurology, Director of the Richard T. Johnson Division of Neuroimmunology and Neuroinfectious Diseases,
and Director of the Johns Hopkins Multiple Sclerosis Center, Johns Hopkins University School of Medicine, Baltimore, MD, USA

Laura J. Balcer, MD, MSCE, FANA

Professor of Neurology, Population Health and Opthalmology; Vice Chair, Department of Neurology, NYU School of Medicine, NY, USA

Elliot M. Frohman, MD, PhD, FAAN

Professor of Neurology and Ophthalmology
Distinguished Teaching Professor
Kenney-Marie Dixon Pickens Distinguished Professor of MS Research
Irene Wadel & Robert Atha Distinguished Chair in Neurology
Director, Multiple Sclerosis and Neuroimmunology Program and Clinical Center for Multiple Sclerosis
University of Texas Southwestern School of Medicine
University Printing House, Cambridge CB2 8BS, United Kingdom

Cambridge University Press is part of the University of Cambridge.

It furthers the University’s mission by disseminating knowledge in the pursuit of
education, learning and research at the highest international levels of excellence.

www.cambridge.org
Information on this title: www.cambridge.org/9781107041301
© Cambridge University Press 2015
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
Printed in the United Kingdom by Bell and Bain Ltd
A catalogue record for this publication is available from the British Library
Library of Congress Cataloging in Publication Data
Optical coherence tomography in neurological diseases / edited by Peter A. Calabresi, Laura J. Balcer,
Elliot M. Frohman.
p. ; cm.
ISBN 978-1-107-04130-1 (hardback)
I. Calabresi, Peter A., editor. II. Balcer, Laura J., editor. III. Frohman, Elliot M., editor.
[DNLM: 1. Nervous System Diseases – diagnosis. 2. Tomography, Optical Coherence. WN 206]
RC349.T65
616.8ʹ04757–dc23
2014048628
ISBN 978-1-107-04130-1 Hardback
Cambridge University Press has no responsibility for the persistence or accuracy of
URLs for external or third-party internet websites referred to in this publication,
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.
................................................................................................................................

Every effort has been made in preparing this book to provide accurate and
up-to-date information which is in accord with accepted standards and practice
at the time of publication. Although case histories are drawn from actual cases,
every effort has been made to disguise the identities of the individuals involved.
Nevertheless, the authors, editors and publishers can make no warranties that the
information contained herein is totally free from error, not least because clinical
standards are constantly changing through research and regulation. The authors,
editors and publishers therefore disclaim all liability for direct or consequential
damages resulting from the use of material contained in this book. Readers
are strongly advised to pay careful attention to information provided by the
manufacturer of any drugs or equipment that they plan to use.
Contents
List of Contributors vii
Preface ix

1 Introduction to optical coherence 9 Meta-analysis of optical coherence

tomography in neurological tomography in multiple sclerosis 103
diseases 1 Axel Petzold
Peter A. Calabresi, Laura J. Balcer, and
10 Optical coherence tomography and brain
Elliot M. Frohman
magnetic resonance imaging in multiple
2 Basic principles of optical coherence sclerosis 114
tomography 4 Shiv Saidha and Peter A. Calabresi
Sven Schippling
11 Optical coherence tomography in
3 Anatomy of the anterior visual neurodegenerative and other neurologic
pathway 14 diseases 128
Devin D. Mackay, Steven L. Galetta, and Friedemann Paul and Alexander Ulrich
Sashank Prasad Brandt
4 Optical coherence tomography 12 Optical coherence tomography
in acute optic neuritis 28 pathologies to know about in clinical
Andrew P. D. Henderson, S. Anand Trip, practice 145
and David H. Miller Scott D. Newsome and John N. Ratchford
5 Optical coherence tomography 13 Optical coherence tomography
and visual outcomes in acute optic and retinal segmentation in neurological
neuritis 42 diseases 156
Fiona Costello Elias S. Sotirchos and Shiv Saidha
6 Optical coherence tomography 14 Optical coherence tomography and retinal
and low-contrast acuity 61 pathology in neurologic diseases 165
Shin C. Beh and Laura Ari J. Green
J. Balcer
15 Retinal inflammation in multiple sclerosis
7 Optical coherence tomography revealed by optical coherence tomography
and electrophysiology of the visual and ophthalmoscopy 176
pathway 76 Elena H. Martinez-Lapiscina,
A. Klistorner, C. L. Fraser, C. Yiannikas, Bernardo Sanchez-Dalmau, and Pablo
and S. L. Graham Villoslada
8 Optical coherence tomography and 16 Optical coherence tomography
electrophysiology of the optic and optic nerve magnetic resonance imaging
nerve head 89 in demyelinating diseases 184
Shin C. Beh, Zane Schnurman, Darrel Robert T. Naismith
Conger, Amy Conger, Benjamin
17 Optical coherence tomography in neurologic
M. Greenberg, Elliot M. Frohman, and
clinical trials 191
Teresa C. Frohman v
Robert A. Bermel and Peter K. Kaiser
 Contents

18 Optical coherence tomography in a 19 Future technological advances in optical

multi-center setting: quality control coherence tomography 209
issues 198 Hiroshi Ishikawa and Joel Schuman
Axel Petzold, Laura J. Balcer, Peter
A. Calabresi, Fiona Costello, Elliot
M. Frohman, Ari J. Green, A. Klistorner,
Friedemann Paul, Sven Schippling, and Index 218
Pablo Villoslada

vi
Contributors

Laura J. Balcer C. L. Fraser

Professor of Neurology, Population Health and Sydney Eye Hospital, Save Sight Institute, Sydney,
Ophthalmology; Vice Chair, Department of Australia
Neurology, NYU School of Medicine, NY, USA
Elliot M. Frohman
Shin C. Beh Professor of Neurology and Ophthalmology
Assistant Professor of Neurology & Distinguished Teaching Professor Kenney-Marie
Neurotherapeutics Director Vestibular Neurology & Dixon Pickens Distinguished Professor of MS
Neurovisual Disorders Multiple Sclerosis & Research Irene Wadel & Robert Atha Distinguished
Neuroimmunology Program UT Southwestern Chair in Neurology Director, Multiple Sclerosis and
School of Medicine at Dallas, TX, USA Neuroimmunology Program and Clinical Center for
Multiple Sclerosis University of Texas Southwestern
Robert A. Bermel School of Medicine at Dallas, TX, USA
Mellen Center for MS Treatment and Research,
Teresa C. Frohman
Neurological Institute, Cleveland Clinic, Cleveland,
Director Multiple Sclerosis Neuro-Ophthalmology
OH, USA
Research Laboratory Executive Director Multiple
Alexander Ulrich Brandt Sclerosis and Neuroimmunology Program Executive
NeuroCure Clinical Research Center, Director UT Southwestern & National Multiple Sclerosis
Charité – Universitätsmedizin Berlin, Germany Society Comprehensive Treatment Training Program
Multiple Sclerosis Advanced Clinical Practitioner
Peter A. Calabresi Department of Neurology & Neurotherapeutics UT
Professor of Neurology, Director of the Richard Southwestern School of Medicine at Dallas, TX, USA
T. Johnson Division of Neuroimmunology and
Steven L. Galetta, MD
Neuroinfectious Diseases, and Director of the Johns
Professor and Chair, Department of Neurology, New
Hopkins Multiple Sclerosis Center, Johns Hopkins
York University Langone Medical Center, New York,
University School of Medicine, Baltimore, MD, USA
NY, USA
Amy Conger S. L. Graham
Senior Neuro-Ophthalmic Imaging Specialist Clinical Macquarie University, Sydney, Australia
Center for Multiple Sclerosis UT Southwestern
Medical Center Dallas, Tx, USA Ari J. Green
Associate Professor of Neurology and Ophthalmology
Darrel Conger Clinic Medical Director Neurodiagnostics Center
Co-Director, Neuro-Ophthalmology Testing Lab Director UCSF MS Center, Department of Neurology,
Department of Neurology & Neurotherapeutics University of Californa, San Francisco, CA, USA
UT Southwestern School of Medicine at Dallas,
TX, USA Benjamin M Greenberg
Associate Professor of Neurology Director,
Fiona Costello Transverse Myelitis, Neuromyelitis Optical and
Associate Professor of Neurology Department of Pediatric Demyelinating Disease Programs Director,
Clinical Neurosciences, University of Calgary, Neurosciences Clinical Research Center at Dallas, TX, vii
Calgary, Alberta, Canada USA UT Southwestern School of Medicine
 List of contributors

Andrew P. D. Henderson John N. Ratchford

NMR Research Unit, UCL Institute of Neurology, Department of Neurology, The Johns Hopkins
London, UK University School of Medicine, Baltimore, MD, USA
Hiroshi Ishikawa Anand Trip
Ocular Imaging Center, University of Pittsburgh Professor of Clinical Neurology, University Department
School of Medicine, Pittsburgh, PA, USA of Clinical Neurology, National Hospital for Neurology
and Neurosurgery, London, UK
Peter K. Kaiser
Cole Eye Institute, Cleveland Clinic, Cleveland, OH, Shiv Saidha
USA Assistant Professor of Neurology, Department
of Neurology, Johns Hopkins University
A. Klistorner School of Medicine, Baltimore,
The University of Sydney, and Macquarie University, MD, USA
Sydney, Australia
Bernardo Sanchez-Dalmau
Devin D. Mackay Center of Neuroimmunology and Department
Clinical Fellow in Neurology, Department of of Ophthalmology, Institut d’Investigacions
Neurology, Brigham and Women’s Hospital, Harvard Biomèdiques August Pi i Sunyer (IDIBAPS), Hospital
Medical School, Boston, MA, USA Clinic of Barcelona, Barcelona, Spain
Elena H. Martinez-Lapiscina Sven Schippling
Center of Neuroimmunology, Institut d’Investigacions Department of Neurology, University Hospital,
Biomèdiques August Pi i Sunyer (IDIBAPS), Hospital Zurich, Switzerland
Clinic of Barcelona, Barcelona, Spain
Zane Schnurman
David H. Miller
Departments of Neurology, University of Texas
Professor of Clinical Neurology, University
Southwestern School of Medicine at Dallas,
Department of Clinical Neurology, National Hospital
TX, USA
for Neurology and Neurosurgery, London, UK
Robert T. Naismith Joel S. Schuman
Assistant Professor of Neurology, Washington Eye & Ear Foundation Professor and Chairman,
University, Saint Louis, MO, USA Department of Ophthalmology
Director of UPMC Eye Center
Scott Newsome Director, Louis J. Fox Center for Vision Restoration
Assistant Professor of Neurology, Department of Professor of Bioengineering
Neurology, The Johns Hopkins University School of Professor of Clinical and Translational Science
Medicine, Baltimore, MD, USA Glaucoma and Cataract Service
University of Pittsburgh School of Medicine
Friedemann Paul
Eye and Ear Institute, Pittsburgh, PA, USA
Professor of Clinical Neuroimmunology, NeuroCure
Clinical Research Center, Charité – Elias S. Sotirchos
Universitätsmedizin Berlin, Germany Resident Physician, The Johns Hopkins Hospital,
Baltimore, MD, USA
Axel Petzold
Consultant Neurologist, VUMC, MS Center Pablo Villoslada
Amsterdam, NL and UCL Institute of Neurology, Center of Neuroimmunology, Institut d’Investigacions
Department of Neuroinflammation, London, UK Biomèdiques August Pi i Sunyer (IDIBAPS), Hospital
Clinic of Barcelona, Barcelona, Spain
Sashank Prasad
Department of Neurology, Brigham and Women’s C. Yiannikas
viii Hospital, Harvard Medical School, Boston, MA, USA Concord Hospital, Sydney, Australia
Preface

Shortly after the invention of the first two- elegant of the eloquent neuroscience systems from
dimensional optical coherence tomography (OCT) which to dissect pathophysiologic mechanisms of dis-
scans by James Fujimoto and colleagues at the ease, and to detect and monitor the preventative,
Massachusetts Institute of Technology in 1991, protective, and even restorative properties of novel
ophthalmologists recognized the extraordinary neurotherapeutic agents, an exciting dividend of
potential for OCT to facilitate quantitative assessment scientific discovery in the eye.
of the neuroretina. OCT rapidly became a common- Perhaps one of the most gratifying aspects of
place tool in ophthalmologic practice to identify both editing this book has been the collaborative nature
inflammatory and degenerative conditions affecting of the interactions among the authors, as well as the
the optic nerve and retina. The utility of OCT to detect gracious contributions of those who did not have
and quantify sequelae of optic neuropathies, includ- opportunities for their writing to be incorporated
ing glaucoma and those associated with primary neu- into this book. Remarkably, such unprecedented
rological diseases, was soon brought to the forefront. camaraderie reflects the scientific and clinical com-
Rapid advances in the technology have followed and munity’s dedication to the exploration of vision in
have included faster acquisition rates enabling higher- multiple sclerosis (MS) and other neurological dis-
resolution images, eye tracking to minimize motion eases. Indeed, we attribute the exponential expan-
artifacts, and post-processing algorithms that permit sion of OCT in neurological diseases to these
segmentation of the nine layers of the retina. Such exemplary scientific interactions. Extensive docu-
improvements have led to exponential growth in the mentation of reproducibility and validity for OCT
application of OCT in neurological disease, both clini- measurements from independent sites around the
cally and scientifically. world has rendered this technology a meaningful
We now believe that OCT should become a main- and cost-effective contribution to neurologic care
stay of the neurological evaluation, as it is quite simply and scientific investigation. Thanks to the many
a quantitative ophthalmoscope. This technology is scientific and technological advancements, OCT
already commonplace in neuro-ophthalmology measures now provide structural markers for
offices, and facilitates rapid and accurate assessment aspects of neurological disease that are potentially
of the neuroretina for unexplained causes of visual amenable to novel therapies for protection and
loss for many neurologists. This compendium is repair. This body of work, made reality by countless
designed to provide a useful overview for clinicians experts, epitomizes how, in just two decades, we can
and scientists new to the field, and to serve as a revolutionize our ability to practice medicine and
resource for the more seasoned expert. Herein, we underscores the power and productivity of colla-
emphasize the emerging concept that the eye can borative works, with the ultimate shared goal of
indeed be considered a window into the brain, with improving the quality of life for our deserving
the visual networks constituting perhaps the most patients.

ix
 Chapter
Introduction to optical coherence

1 tomography in neurological diseases

Peter A. Calabresi, Laura J. Balcer, and Elliot M. Frohman

While the neurological examination represents a specificity of MRI findings in some cases is helpful
time-honored gold standard for diagnosing and diagnostically. In this regard, OCT continues to
tracking the clinical course of disease, the sensitive evolve in its capacity to identify findings unique to
and accurate assessment of central nervous system any single neurologic or ocular disorder. While the
integrity remains a challenge. Optical coherence architecture of the neural retina has been nearly
tomography (OCT) is an effective tool that has totally inaccessible even with the application of the
emerged over the past decade for measuring struc- most sophisticated and high-precision MRI methods,
ture–function correlations and quantifying patholo- the advent of high-resolution spectral-domain (SD)
gical changes invisible to the human eye. OCT can OCT has fundamentally changed our ability to quali-
complement the clinical neurological examination tatively and quantitatively assess the eye – the most
in ways that are similar to magnetic resonance ima- visible part of the human brain.
ging (MRI), a tool that has given us tremendous Neurological disorders frequently present with,
structural insights into the nervous system. As we or ultimately involve, visual impairment. Greater
begin to understand the remarkable capacity of the than 50% of the brain’s pathways are dedicated to
brain to compensate for neurological injury, the functional networks related to the visual system.
links between structure, function, and electrophy- While ocular motility is germane to foveation and
siology that are provided uniquely by the afferent subsequent image processing, the determination of
visual pathway will be critical. The relative timing of the location and underlying mechanisms that com-
injury to the retinal ganglion cells (neurons), retinal promise vision and visual perception can represent
nerve fiber layer (axons), and other ocular structures one of the most formidably challenging aspects of
may be unique to each neurologic disorder. Such neurologic consultation. This is especially true given
similarities outweigh differences, and the visual the intimate and yet highly distributed expanse of
pathway represents an attractive global model for both the anterior and posterior, as well as afferent
understanding and monitoring disease. and efferent, visual processing networks, all of which
Neurologists acknowledge the capacity for nonin- require careful and systematic examination. Even
vasive imaging technologies to accelerate and opti- the most expert clinical assessment of the visual
mize neurologic diagnosis and follow-up. This is pathway can be augmented by technological tools
perhaps best illustrated by the pivotal role of MRI in such as OCT. When considering all of the highly
the diagnosis and management of multiple sclerosis salient limitations associated with the bedside clin-
(MS). The field of neurology has witnessed a monu- ical neurologic and neuro-ophthalmologic examina-
mental transformation from near complete reliance tions, we can begin to appreciate the impact on
on clinical history and examination to the addition of medical practice and clinical investigation that
MRI and other noninvasive imaging techniques. An OCT has provided.
extension of the clinical examination, MRI has revo- Rapid advances in OCT imaging technology have
lutionized care not only in MS but in the case of other provided unprecedented, noninvasive, in vivo ima-
neurologic disorders. For example, the relative ging of the retina. Spectral-domain OCT has

Optical Coherence Tomography in Neurological Disease, ed. Peter A. Calabresi, Laura J. Balcer, and Elliot M. Frohman. 1
Published by Cambridge University Press. © Cambridge University Press 2015.
 Chapter 1: Introduction to optical coherence tomography in neurological diseases

spectacular image resolution and very low test-retest In this new book we provide an extensive
variability. Furthermore, ultrafast acquisition rates review of the application of OCT across a broad
have minimized motion artifacts, facilitated intra- landscape of neurological diseases. A basic-level
subject/inter-scan co-registration with high precision, introduction to the underlying principles of the
and led to post-processing algorithms for automated technology is provided for the reader. This is fol-
segmentation analyses for measurement of retinal lowed by a review of the neuroanatomy of the
layer thicknesses. visual system, especially the retina, the target tis-
Considerable improvements in both the technical sue for OCT imaging. No imaging device should
and practical performance characteristics of OCT be employed without proper examination of the
have led to an exponential growth in its application critical link with clinical function and outcomes,
to a variety of neurological diseases. This has bene- most notably vision. Several chapters are dedi-
fited clinical practice with respect to the rapid and cated to the clinical validation of the utility of
accurate assessment of the neuroretina, while also OCT in neurological diseases. As expected, there
catalyzing a literal scientific revolution in terms of are significant cross-sectional associations
viewing the eye as a noninvasive “window” into the between retinal pathology detected by OCT and
brain. low-contrast letter acuity, high-contrast visual
The neuroretina is part of the brain, comprised of acuity, and color perception. Perhaps more impor-
first-order neurons that relay visual information to tantly, retinal imaging appears to have predictive
the lateral geniculate nucleus in the thalamus. While value for estimating cerebral pathology in gray
these ganglion cell layer (GCL) neurons are suscepti- matter structures of the central nervous system.
ble to primary pathologies affecting the retinal nerve Involvement of these structures is notoriously
fiber layer (RNFL) axons, secondary pathology that inconspicuous early on in many neurological dis-
arises in other areas of the brain such as the hypotha- eases, and therefore an understanding of gray mat-
lamus and superior colliculus can also be detected in ter pathology may facilitate early detection and
the retina. prognostication for those patients with neurode-
All forms of optic neuropathy have a characteristic generative disorders. The underlying tissue patho-
signature of RNFL thinning and retinal GCL atrophy logical substrates captured with OCT-generated
that can be quantified by OCT. More remarkably, images are limited, and, therefore, several chapters
several neurodegenerative diseases with well-known are dedicated to elucidating the promising
brain pathology, such as Alzheimer’s disease, fronto- correlations between OCT images and histo-
temporal dementia, and Parkinson’s disease, are now pathology, vascular pathology, as well as func-
recognized to have concomitant retinal neuronal layer tional electrophysiology.
degeneration that can be demonstrated by OCT. Because this new technology increasingly takes
Neurological diseases traditionally linked with the its place in the clinic, OCT has also found appli-
peripheral nervous system are increasingly reported cation in clinical trials of putative neuroprotective
to have central nervous system manifestations; these drugs. Indeed, the rapid thinning of the RNFL
include type 2a Charcot-Marie-Tooth disease, in and the GCL following acute optic neuritis is a
which abnormalities in mitofusin can manifest as logical outcome measure for phase II trials of
optic nerve disease. Deeper retinal neuronal layers, drugs that could exert neuroprotective or neuro-
including the inner and outer nuclear layers, have repair properties. The reproducible and accurate
been shown to be abnormal in several diseases classi- quantitation of the axonal and neuronal layers
cally associated only with primary axonal demyelina- that undergo thinning in the three to six months
tion, such as MS. This supports theories that neurons following an acute optic neuritis make OCT an
may be a direct target of neurological disease pro- appealing tool, which is easily employed across
cesses not previously thought to have a primary multiple centers in clinical trials. Therefore, we
degenerative mechanism. As the potential applica- have dedicated several chapters to reviewing the
tions of OCT continue to grow, its role in neurological rationale, approach, and practicality of using OCT
disease is of increasing interest and importance in in clinical research trials. Whether retinal neuro-
discovery biology as well as in the context of clinical nal imaging will find the same application in
2 assessment. neurodegenerative diseases, which manifest with
 Chapter 1: Introduction to optical coherence tomography in neurological diseases

changes that take place over years to decades, retinal imaging devices may facilitate even more
remains to be fully determined, but evidence sup- progress in structural and functional assessments
porting such a role in this context is rapidly of the retinal manifestations of neurological dis-
mounting. Finally, the book concludes with a eases. We now believe that OCT should become a
look to the future and a discussion of how the mainstay of the neurological evaluation, because it
rapid evolution of this technology and associated is, quite simply, a quantitative ophthalmoscope.

3
 Chapter
Basic principles of optical coherence

2 tomography
Sven Schippling

Introduction Basic principles of OCT in retinal

Optical coherence tomography (OCT) uses low-coher- imaging
ent, near infrared light to generate tomographic, two-
OCT is based on the physical principle of partial
dimensional (2-D) structural in vivo images of
interference between light backscattered by tissues
biological tissues based on the physical concept of
and a reference light beam reflected from a reference
time-of-flight delay and intensity of backscattered
mirror [1–3]. In retinal OCT imaging, low-coherent,
light from microscopic constituents within these tis-
near infrared light is emitted from a superluminescent
sues [1, 2]. Therefore, OCT can be considered an
diode and directed onto the retina through the pupil
optical analogue of ultrasound-based tomographic of the person being scanned [2]. The scans can be
imaging [3]. As is the case in ultrasonography, time acquired with or without pupillary mydriasis. The
gating of backscattered reflexes leads to a line profile,
process of generating anteroposterior tomographic
the so-called A-scan [1–3]. The first 2-D or so-called
scans is based on the splitting of this light beam into
B-scan images, repeating the A-scanning process
a sample and a reference fraction by a beamsplitter
while changing a reference mirror´s position, were
(Figure 2.1). Whereas the first is entering the patient´s
provided by Huang and colleagues from the group of
eye, the latter is directed towards a (mobile) reference
James Fujimoto at the Massachusetts Institute of
mirror (Figure 2.1). The light directed onto the retina
Technology (MIT) in 1991 [3]. It took another five crosses the transparent structures of the eye and is
years until the first OCT device with an axial resolution partially absorbed, whereas a larger fraction is back-
of 15 μm became commercially available (OCT 1;
scattered by the different retinal layers [1,2]. Both the
System 2000 Humphrey Instruments, Inc.). Since
reference beam and the sample beam are reflected
then, the technique has undergone a remarkable tech-
backward and simultaneously registered by a photo-
nical development. Also, the traditional application of
detector or spectrometer in which the co-occurrence
OCT in ophthalmological diseases like glaucoma and
of the two wave fronts induces an interference signal
macular edema has more recently been extended and
[2]. The false color or gray contrast OCT image itself
introduced into neurology [4, 5]. Here, within a rea-
is based on the distribution and amplitudes of the
sonably short time period, OCT has inspired a remark- repeatedly incoming interference signals [2, 3].
able body of literature demonstrating both primary
As of this date, two different OCT technologies
and secondary (e.g., following multiple sclerosis-asso-
are still available, the older time-domain (TD) and
ciated optic neuritis) retinal pathology in a number of
the more recent spectral-domain (SD) OCT, also
different disease models [4–9]. Some of this work is
called Fourier-domain (FD) OCT [10]. The latter
described in detail in the following chapters within this
name refers to Jean-Baptiste Joseph Fourier
compendium. This section will describe basic technical
(1768–1830), a French Professor at the École
principles of OCT as well as the development of this Polytechnique, who is known for his work on the
exciting methodology within the last two decades. Fourier Transform [11].

4 Optical Coherence Tomography in Neurological Disease, ed. Peter A. Calabresi, Laura J. Balcer, and Elliot M. Frohman.
Published by Cambridge University Press. © Cambridge University Press 2015.
 Chapter 2: Basic principles of optical coherence tomography

REFERENCE MIRROR

LIGHT SOURCE SCANNING REFERENCE

BEAM
SPLITTER

DETECTOR

SIGNAL PROCESSING

Figure 2.1 Basic principles of time-domain OCT. Interference of reference and reflected light beams from different layers within the tissue
occurs as a function of axial transition of the reference mirror.

Time-domain (TD) optical interference by co-occurrence on the level of the

photodetector. This phenomenon allows accurate
coherence tomography measurement of the echo time delay. Light reflected
The basic components and setup of a TD-OCT system from superficial structures of the retina or uppermost
are detailed in Figure 2.1. As the name of this tech- retinal layers has a shorter echo time delay than light
nology suggests, TD-OCT is based on the difference in reflected from deeper retinal structures (innermost
the time delay of the sample light echoes reflected layers) [2, 5]. The reference mirror is axially translo-
from the different retinal layers as a function of their cated in order to match echo time delays from various
depth within the tissue (here, the retina) and the tissue layers. As the path length of the moving refer-
reference beam echo, a single echo that is varied by ence mirror is known, it is possible to calculate the
changing the position of the mobile reference mirror depth of the tissue from which the fraction of reflected
(i.e., the reference arm-length). light arises based on the specific time delay. Based on
Interference of light beams with low coherence the amplitude of the interference signals that arise
only occurs in cases where the distance traveled by from interference of the reflected light from retinal
the light in both arms of the interferometer is equiva- layers of different depths and the reference light from 5
lent to within the coherence length in order to allow various path lengths, a single axial scan, the so-called
 Chapter 2: Basic principles of optical coherence tomography

A-scan, is deduced. Key to longitudinal scanning is an increasing B-scan density have become possible
the fact that the reference mirror can be mechanically with the latest devices (see Table 2.1). SD-OCT
mobilized, resulting in a shift of the reference beam, achieves an axial resolution in the range of 5–6 µm,
extending the reference path. The reference mirror while digital resolution can be even higher (Table 2.1).
moves with a specific distance- and time-interval con- High-resolution scans together with real-time aver-
stant, leading to multiple adjacent A-scans, which, in aging significantly increase the signal-to-noise ratio
sum, generate the cross-sectional 2-D or longitudinal (SNR), ensuring superior image resolution and qual-
B-scan of the retina (Figure 2.1). In TD-OCT the ity in SD-OCT.
cross-sectional image representing the different ret- Currently, a number of TD- and SD-OCT devices
inal layers is generated as a function of time delay of are commercially available. Technical characteristics
the reflected sample light beams. In turn, the time of different OCT devices are detailed in Table 2.1.
delay depends on the composition and depth of the
different layers and the position or axial translation of Swept source (SS) optical
the reference mirror. These basic principles have led
to the name of this technology: time-domain OCT. coherence tomography
TD-OCT has an axial resolution of approximately The light source in swept source (SS) OCT systems is a
10 µm or less (Table 2.1). Image acquisition speed is tunable narrow bandwidth laser. The interferences at
limited in TD-OCT, however, because the reference different wavelengths are measured over time.
mirror needs to be moved. Through rapid adjustment of the laser, scan speeds
of up to 249,000 A-scans/second can be achieved [12].
Spectral-domain (SD) optical SS-OCT technique was used for the first time in ret-
inal imaging in 2006 [13].
coherence tomography As with TD-OCT, SS-OCT uses a photodetector
The most recent, so-called fourth-generation or instead of the combination of a CCD camera and
spectral-domain (SD) OCT technology, is based on spectrometer, as in SD-OCT. In addition, a light
the mathematical Fourier transform equation. This source with wavelengths around 1,000 nm is applied.
methodology is, therefore, also known as Fourier- The longer wavelengths penetrate deeper into the
domain technology. The Fourier transform eliminates retina and optic nerve head. This might be useful in
the need for a movement of the reference beam mir- assessing the choroid and the lamina cribrosa, as well
ror. SD-OCT replaces the photodetector from TD- as the optic nerve head. A disadvantage of using
OCT with a spectrometer capable of analyzing the longer wavelengths is a reduction of axial resolution
full spectrum of interference signals at one time compared to the SD-OCT (which uses light with a
point generated when the sample and the reference wavelength of around 840–880 nm) [14].
beam meet along the same path. It allows the analysis
of all frequencies simultaneously. As opposed to TD- Advantages of spectral over
OCT, in SD-OCT the interference signal is a function
of the different wavelengths and not of the different time-domain OCT
echo time delays. The whole wavelength spectrum is In TD-OCT depth of the tissue is sampled point by
converted into time delay signals by the Fourier trans- point axially transferring the reference mirror.
form. In retinal OCT imaging, this allows the analysis Compared to SD-OCT this renders TD-OCT ineffi-
of all echoes from the different retinal layers simulta- cient [10,11].
neously. As a consequence, SD-OCT is much more Replacing the photodetector in TD-OCT with a
rapid than its counterpart, while at the same time spectrometer in SD-OCT allows spectral OCT to detect
providing excellent resolution. In TD-OCT, high- the entire wavelength spectrum corresponding to reflec-
resolution imaging can only be achieved at the tions from the entire depth range [11]. SD-OCT allows
expense of an increase in acquisition times. Scanning acquisition speeds of 27,000 to 53,000 A-scans per sec-
speed is 50–100 times faster with SD-OCT than with ond in clinical application as compared to a maximum
TD-OCT [10,11]. of approximately 400 A-scans per second in TD-OCT
Rapid scanning allows larger numbers of B-scans (see Table 2.1) [10–12]. In experimental settings A-scan
6 per time interval, and high-speed macular scans with speeds of even up to 312,500 scans per second have been
Table 2.1 Technical characteristics of commercially available OCT devices

Fundus image OCT image

Scan Max.
Speed Min. no. of
Live Optional Light A-scans/ Transverse Axial Scan pupil A-scans/
Device image Size modes Technology source sec resolution** resolution** depth diameter B-scan Manufacturer Web address

AF, ICGA, FA, Heidelberg www.heidelberg

SPECTRALIS cSLO 30°, 55°, 165° MC SD 870 nm 40000 14 µm 7 µm 1.9 mm 2 mm 1536 Engineering engineering.com

www.meditec.zeiss.
Stratus IR Cam 26° × 20.5° TD 820 nm 400 20 µm 10 µm 2 mm 3.2 mm 768 Zeiss Meditec AG com

www.meditec.zeiss.
CIRRUS 4000 SLO 36° × 30° SD 840 nm 27000 15 µm 5 µm 2 mm 2 mm 4096 Zeiss Meditec AG com

CF Camera, www.topcon-medi
3D OCT 2000 IR Cam 45° AF*, FA* SD 840 nm 50000 20 µm 6 µm 2.3 mm 2.5 mm 1024 Topcon cal.eu

www.topcon-medi
DR 1 SLO 43° SwS 1050 nm 100000 20 µm 8 µm 2.5 mm 1024 Topcon cal.eu

iVue IR Cam 32° × 23° SD 840 nm 25000 15 µm 5 µm 2.3 mm 3 mm 1024 Optovue www.optovue.com

RS-3000 SLO 40°x 30° SD 880 nm 53000 20 µm 7 µm 2.1 mm 2.5 mm Nidek www.nidek-intl.com

OCT SLO cSLO 29° SD 830 nm 25000 20 µm 10 µm 2 mm 3 mm Optos www.optos.com

Copernicus HR IR Cam 30° SD 850 nm 52000 12 µm 3 µm 2 mm 3 mm 20000 Optopol www.optopol.com

Canon OCT www.canon-eur

HS-100 SLO 44° × 33° SD 855 nm 70000 20 µm 3 µm 2 mm 3 mm Canon ope.com

cSLO confocal Scanning Laser Ophthalmoscope

SLO Scanning Laser Ophthalmoscope
AF Auto Fluorescence
ICGA Indocyanin Green Angiographay
FA Fluorescence Angiography
MC Multicolor
* only 3-D OCT FA Plus
** optical resolution
 Chapter 2: Basic principles of optical coherence tomography

REFERENCE MIRROR

LIGHT SOURCE
SCANNING REFERENCE

BEAM
SPLITTER

SPECTROMETER

SIGNAL PROCESSING

Figure 2.2 Basic principles of spectral-domain OCT. Note that in SD-OCT a complete spectrum of interference signals
is processed simultaneously because of the introduction of a spectrometer, allowing faster scan acquisition, while the reference
mirror is fixed.

reported with ultrahigh-speed SD-OCT devices [12]. be an issue during acquisition [11]. A standard
Another advantage that comes with increasing scanning volume scan (200 A-scans/B-scan; 200 B-scans at
speed is that motion artifacts produced by involuntary 27,000 A-scans/sec) takes about 1.48 seconds
eye movements of the person being scanned are (1,480 ms). Involuntary saccades are much faster
reduced [10,11]. Averaging of multiple frames helps and occur at 30–50 µs (micro-saccades) or 150 ms
increasing the signal-to-noise ratio, further improving (macro-saccades). This can result in motion artifacts
image quality [11]. In TD-OCT, scanning of the macu- of the OCT scan. In order to reduce the above men-
lar is based on a lower number of B-scans, and the full tioned motion artifacts induced by physiological sac-
macula scan is extrapolated from these few B-scans. cades and to increase intra-rater and inter-rater
Limited retinal pathology can be missed by this extra- reliability, while at the same time increasing scanning
polation, because it may fall in a gap between two frequency, an active eye-tracking technology has been
neighboring B-scans. introduced in the SPECTRALIS SD-OCT device ®
However, even with the increased speed of SD- (Heidelberg Engineering GmbH, Heidelberg,
8 OCT, physiological involuntary eye movements can Germany). The active eye tracker (TruTrackTM)
 Chapter 2: Basic principles of optical coherence tomography

recognizes eye movements on the infrared fundus scan pattern used in those devices is the peripapillary
image, continuously correcting the OCT scan line circle scan. Boundaries of the RNFL are automatically
and compensating according to the extent of the detected, and the mean overall RNFL thickness is
involuntary saccades. This feature enables the device calculated.
to acquire larger data sets with higher numbers of In TD-OCT macular assessment is performed by
averaged single B-scans (up to 100), while increasing a set of 6 B-scans (128 A-scans each) centered in a
the SNR and avoiding motion artifacts. A further star pattern on the fovea. For analysis the macula is
advantage is the automatic follow-up function that divided into sectors using the ETDRS (Early
comes with the active eye-tracking system. Given Treatment of Diabetic Retinopathy Study) grid
that each retina is unique (like an individual finger- [17]. The grid layout consists of three concentric
print), the scan is exactly repositioned for the follow- circles: one central ring with a diameter of 1 mm,
up scan with only a little variability. As a consequence, an intermediate ring with a diameter of 3 mm, and
it allows monitoring of minimal changes that appear an outer circle with 6 mm. The intermediate and
longitudinally in the context of both pathological outer rings are divided into sectors (superior, infer-
retinal conditions and in neurodegenerative diseases ior, nasal, and temporal). The measures extrapo-
such as multiple sclerosis (MS). lated from this macular scan are the total macular
volume and mean thickness values of the respective
Optic nerve head and macular scan sectors. The boundaries of the inner limiting mem-
brane (ILM), nerve fiber layer (NFL), and Bruch’s
protocols membrane (BM) are already automatically defined.
For the so-called peripapillary ring scan, the A-scans Due to the low axial resolution, in the range of
(Figure 2.3) are centered on a circle around the optic 10 µm, and moderate image quality an automated
nerve head. A circle-diameter of 12° (approximately) segmentation of deeper retinal layers was less
3.4 mm has been shown to provide good reproduci- reliable.
bility [15]. As a consequence, this diameter has widely With the introduction of the Fourier-domain
been used irrespective of the size of the optic nerve OCT technology, axial transition of the reference
head [16]. Standard TD-OCT scanning protocols of mirror to change path lengths was no longer needed
the optic nerve head involve 256 A-scans around the [10, 11]. Scanning speeds increased significantly,
optic disc. To minimize the effect of involuntary eye now allowing scan patterns with a higher number
movements during scanning, a series of three scans is of B-scans and much higher resolution. As a conse-
performed. quence, SD-OCT allows a much more comprehen-
From these peripapillary ring scans the mean sive picture of the entire retina and a more in-depth
overall retinal nerve fiber layer (RNFL) thickness is view of retinal pathology [10,11]. However, the para-
calculated, as are the thicknesses of the four RNFL meters obtained from these scans are still the same.
quadrants (temporal, superior, nasal, and inferior) Depending on the device the circle is calculated
and 12 segments around the optic nerve head from a squared volumetric scan pattern (CIRRUS®,
(Figure 2.3). The optic nerve head or peripapillary Carl Zeiss Meditec AG, Dublin CA), or consists of
ring scan, shows a typical morphology in which up to 1,536 A-scans (SPECTRALIS®, Heidelberg
RNFL thickness is greatest at the superior and infer- Engineering GmbH, Heidelberg) in the form of a
ior poles, whereas the lowest thickness values are circle centered on the optic nerve head. Notably,
registered nasally and temporally. The anatomical the macular volume scans differ in size, spacing
reason behind this distribution is the fact that the between B-scans, and the number of A-scans
majority of ganglion cell axons that – in sum – form acquired. In all devices the 3-D reconstruction of
the optic nerve enter the optic disc from the superior the macular cube scan is still based on extrapolation
and inferior quadrants. from the acquired B-scans.
At the time that OCT was introduced into neurol- It took another several years of software develop-
ogy only TD-OCT technology was been available. As ment until automated or semi-automated algorithms
mentioned above, TD-OCT was capable of acquiring became available to perform segmentation of macular
only 400 A-scans per second. Consequently, the avail- B-scans (see 2.9) that allowed the quantification of
able scan patterns were limited. The most common deeper retinal layers, like the ganglion cell layer, the 9
 Chapter 2: Basic principles of optical coherence tomography

IR 30° ART [HR] IR 30° ART [HR]

OD Asymmetry OS
OD - OS

S
–5

N T
–1 4

I
2

NS TS
–16 6
200 µm 200 µm
N/T N G T PMB
–0.10 –1 0 4 6
OCTART (94) Q: 32 [HR] OCTART (100) Q: 35 [HR]
NI TI
5 –2

ILM ILM
RNFL RNFL

200 µm 200 µm

Above Normal Limits

(p<0.01)

300 300
Borderline Above

Thickness [µm]
Thickness [µm]

240 (p<0.05) 240

180 Within Normal Limits 180
(p>0.05)
120 120
Borderline Below
60 (p<0.05)
60
0 0
Below Normal Limits
–135 –90 –45 0 45 90 135 (p<0.01) –180 –135 –90 –45 0 45 90 135 180
NAS INF TMP SUP NAS NAS INF TMP SUP NAS
Position [°] Position [°]

S
300
OD OS S
132 138
240
Thickness [µm]

T N 180 N T
64 76 120 78 60
60
I I
0
133 132
–135 –90 –45 0 45 90 135
NAS INF TMP SUP NAS
TS NS Position [°] NS TS
155 110 127 149

T G N N/T N/T N G T PMB

PMB
64 102 76 1.20 1.30 78 102 60
53 Classification OD Classification OS 47

TI NI NI TI
137 130 Within Normal Limits Within Normal Limits 125 139

Figure 2.3 Report of a peripapillary spectral-domain OCT scan performed on a normal subject. Top left and right are the infrared fundus
images of the right (OD) and left eye (OS) followed by the B-scans. Automatic segmentation of the retinal nerve fiber layer (inner limiting
membrane (ILM)/red line; retinal nerve fiber layer (RNFL)/blue line).
Individual thickness shown as solid black lines for the left and right eye (within 95% confidence interval/green area). Thickness of the
superior, nasal, inferior, and temporal quadrants.
10
 Chapter 2: Basic principles of optical coherence tomography

(a)

(b)

(c)
C-Scan

B-Scan

A-Scan

Figure 2.4 Spectral-domain OCT scan. Single A-scan profile (a). Central macular B-scan (b). 3-D macular scan generated by
co-registration of fundus image and consecutive macular B-scans (c).

inner and outer nuclear, and the inner and outer

plexiform layers.

3-D visualization
The scanning frequency and the number of B-scans
acquired in SD-OCT allows 3-D reconstruction of
macular and optic nerve head images (Figure 2.4).
Usually 3-D visualization scanning laser ophthalmo-
scopic (SLO) or fundus images and OCT B-scans are
co-registered. Current software allows reconstruction of
3-D images in a frontal or coronal plane (C-scan/en face
mode) (Figure 2.5), which might help to screen and Figure 2.5 Reconstruction of macular B-scans in a frontal plane
identify retinal pathologies in more detail [18]. generates a C-scan perpendicular to the coronal view, as shown in 11
Figure 2.4 (en face image).
 Chapter 2: Basic principles of optical coherence tomography

ILM
NFL
GCL
IPL
INL
OPL
ELM
PR1
PR2
RPE
BM

Figure 2.6 Segmented macular B-scan. Inner limiting membrane (ILM), nerve fiber layer (NFL), ganglion cell layer (GCL), inner
plexiform layer (IPL), inner nuclear layer (INL), outer plexiform layer (OPL), external limiting membrane (ELM), photoreceptor layer 1 (PR1),
photoreceptor layer 2 (PR2), retinal pigment epithelium (RPE), Bruch membrane (BM).

3 Huang D, Swanson EA, Lin CP, et al. Optical coherence

Segmentation of OCT macular tomography. Science 1991; 254: 1178–81.
B-scans 4 Petzold A, de Boer JF, Schippling S, et al. Optical
Most of today´s SD-OCT devices include commer- coherence tomography in multiple sclerosis: a sys-
cially available software capable of segmenting retinal tematic review and meta-analysis. Lancet Neurol 2010;
9: 921–32.
layers beyond the RNFL. However, being able to dis-
criminate the different layers in the retina strongly 5 Jindahra P, Hedges TR, Mendoza-Santiesteban CE,
depends on the quality of the OCT scan, which is also Plant GT. Optical coherence tomography of the retina:
applications in neurology. Curr Opin Neurol 2010; 23:
dependent on signal strength and averaging of the 16–23.
single B-scan.
6 Saidha S, Syc SB, Ibrahim MA, et al. Primary
The quality of all SD-OCT devices is sufficient to retinal pathology in multiple sclerosis as detected
automatically define six to seven boundaries within by optical coherence tomography. Brain 2011; 134:
the retina (ILM, RNFL/GCL, IPL/INL, OPL/ONL, IS/ 518–33.
OS, RPE/BM), but with some of the devices increased 7 Costello F, Coupland S, Hodge W, et al. Quantifying
resolution allows identification of up to 11 retinal axonal loss after optic neuritis with optical coherence
layers (Figure 2.6). Yet again, proper identification tomography. Ann Neurol 2006; 59: 963–9.
of the borders between the layers (especially in the 8 Sotirchos ES, Saidha S, Byraiah G, et al. In vivo identi-
case of GCL/IPL) is only possible in B-scans of suffi- fication of morphologic retinal
ciently high quality. Most device manufacturers are abnormalities in neuromyelitis optica. Neurology 2013;
coming out with automated segmentation software in 80: 1406–14.
addition to manual segmentation. Recent studies sug- 9 Brandt AU, Zimmermann H, Kaufhold F, et al. Patterns
gest that both automated and semi-automated as well of retinal damage facilitate differential
diagnosis between Susac syndrome and MS. PLoS One
as manual segmentation of macular B-scans show
2012; 7: e38741.
excellent intra- and inter-rater reliability when per-
10 Forooghian F, Cukras C, Meyerle CB, et al. Evaluation
formed by experienced OCT readers [19].
of time domain and spectral domain optical
coherence tomography in the measurement of
References diabetic macular edema. Invest Ophthalmol Vis Sci
1 Frohman EM, Fujimoto JG, Frohman TC, et al. Optical 2008; 49: 4290–6.
coherence tomography: a window into the mechanisms 11 Yaqoob Z, Wu J, Yang C. Spectral domain optical
of multiple sclerosis. Nat Clin Pract Neurol 2008; 4: 664– coherence tomography: a better OCT imaging strategy.
75. Biotechniques. 2005; 39: S6–13.
2 Schuman JS, Puliafito CA, Fujimoto JG. Optical coher- 12 Potsaid B, Gorczynska I, Srinivasan VJ, et al. Ultrahigh
12 ence tomography of ocular diseases. Slack, Inc., New speed spectral / Fourier domain OCT ophthalmic
Jersey 2004.
 Chapter 2: Basic principles of optical coherence tomography

imaging at 70,000 to 312,500 axial scans per second. Opt nerve head size: an optical coherence tomography
Express 2008; 16: 15149–69. study. Br J Ophthalmol 2005; 894: 489–92.
13 Lee EC, de Boer JF, Mujat M, Lim H, Yun SH. In vivo 17 Photocoagulation for diabetic macular edema. Early
optical frequency domain imaging of human retina and Treatment Diabetic Retinopathy Study report number
choroid. Opt Express 2006; 14: 4403–11. 1. Early Treatment Diabetic Retinopathy Study
14 Gabriele ML, Wollstein G, Ishikawa H, et al. Optical research group. Arch Ophthalmol 1985; 103: 1796–806.
coherence tomography: history, current status, and 18 Wolff B, Matet A, Vasseur V, et al. En face OCT ima-
laboratory work. Invest Ophthalmol Vis Sci 2011; 52: ging for the diagnosis of outer retinal tubulations in
2425–36. age-related macular degeneration. J Ophthalmol 2012;
15 Schuman JS, Pedut-Kloizman T, Hertzmark E, et al. 542417.
Reproducibility of nerve fiber layer thickness mea- 19 Seigo MA, Sotirchos ES, Newsome S, et al. In vivo
surements using optical coherence tomography. assessment of retinal neuronal layers in multiple
Ophthalmology 1996; 103: 1889–98. sclerosis with manual and automated optical coherence
16 Savini G, Zanini M, Carelli V, et al. Correlation tomography segmentation techniques. J Neurol 2012;
between retinal nerve fibre layer thickness and optic 259: 2119–30.

13
 Chapter
Anatomy of the anterior visual pathway

3 Devin D. Mackay, Steven L. Galetta, and Sashank Prasad

Introduction Anterior chamber

Since its development in the 1990s, optical coher- Anterior chamber depth and anterior chamber angle
ence tomography (OCT) has ushered in a new era (Figure 3.1B) can be assessed with OCT. These mea-
of noninvasive diagnostic imaging of the eye [1]. sures often complement findings that are observed
OCT measures the optical backscattering of near- clinically during gonioscopy. Since the overlying
infrared light in a way that is analogous to B-scan sclera can be highly optically backscattering, detailed
ultrasound, which measures the acoustic backscat- features of the anterior chamber are sometimes diffi-
tering properties of a tissue. OCT imaging of the cult to resolve.
visual pathways is limited to structures accessible The anterior angle (Figure 3.1B) can also be eval-
by light, which include the conjunctiva, cornea, uated with OCT. The corneal epithelium (Figure 3.1C)
lens, layers of the retina, optic disc, and optic continues laterally until it meets with the sclera at the
nerve head. Earlier time-domain OCT technology corneoscleral limbus (Figure 3.1D), just anterior to the
has been largely replaced by high-speed spectral- anterior angle. Schwalbe’s ring is a ring of collagenous
Fourier-domain (SD/FD) methods, which allow fibers found on the posterior corneal surface at the
for much faster image acquisition and improved termination of Descemet’s membrane (the basal
quality. SD-OCT images can have high axial spa- lamina of the corneal endothelium).
tial resolution (2–3 μm), although the resolution The canal of Schlemm is found lateral to Schwalbe’s
of most commercially available machines is 4–7 ring in the depth of the anterior angle and is the
μm. OCT is a powerful tool for neurologists, site of aqueous drainage from the anterior chamber
neuro-ophthalmologists, ophthalmologists, and (Figure 3.1B). Aqueous outflow proceeds through the
radiologists interested in the clinical evaluation trabecular meshwork and into the canal of Schlemm,
of the anterior visual pathways. In this chapter, where aqueous humor enters the aqueous veins.
we illustrate the use of OCT to depict the normal Aqueous veins are distinguished from blood-filled
anatomy of the anterior visual pathway. veins by the lack of optical shadows cast on deeper
structures. From aqueous veins, the aqueous humor
enters the scleral veins, sometimes by way of an inter-
Anterior eye vening vascular plexus. Schlemm’s canal and the aqu-
The anterior eye can be imaged with OCT using eous outflow pathways that lead to the scleral veins can
an adapter module or a dedicated OCT instru- be visualized with commercially available SD-OCT. 3-
ment specifically designed for this purpose. D OCT imaging allows for more detailed visualization
Structures of the anterior eye visible with OCT of the outflow channel networks [2].
include the corneal epithelium, corneal stroma,
corneal endothelium, sclera, trabecular meshwork,
iris, canal of Schlemm, corneoscleral limbus, Sclera
Schwalbe’s ring, and Descemet’s membrane The sclera is the dense, white protective covering
(Figures 3.1A–D and 3.2). of the eye, composed almost entirely of collagen

14 Optical Coherence Tomography in Neurological Disease, ed. Peter A. Calabresi, Laura J. Balcer, and Elliot M. Frohman.
Published by Cambridge University Press. © Cambridge University Press 2015.
 Chapter 3: Anatomy of the anterior visual pathway

A B

Iris
Pupil
Angle

Canal of
S
3
S
2 Schlemm
N T
T N

I
Iris
I

C D

Stroma Cornea Cornea Corneoscleral

Limbus
Corneal Epithelium
Corneal Endothelium Sclera
S

N T
Anterior Chamber
I

Figure 3.1 OCT of the anterior chamber. (A) Image of the pupil demarcated by the iris on either side. (B) Angle of the anterior
chamber and canal of Schlemm. (C) Corneal endothelium, epithelium, stroma, and underlying anterior chamber. (D) The cornea and
sclera meet at the corneoscleral limbus. Images courtesy of Rick Calderon, O.D.

(Figures 3.1D and 3.3). Its density and composition cornea interface, and is reflected back into the
make it highly optically backscattering on OCT. instrument.

Iris Lens
The iris divides the anterior from the posterior cham- The lens is a normally transparent and avascular struc-
ber (Figures 3.1A–B, 3.2, and 3.3). It is composed of ture that focuses light on the retina with the help of its
sphincter and dilator muscles that control the size of biconvex configuration (Figure 3.2). It is about 4 mm
the pupil and the amount of light entering the eye [3]. thick and 9 mm in diameter. Zonule fibers suspend
The iris is configured as a shallow cone that points the lens from the ciliary body, just behind the iris.
anteriorly and is highly pigmented. Because the iris Aqueous humor is present anterior to the lens, while
shadows deeper structures on OCT, images of the vitreous humor is present posteriorly [3].
posterior parts of the eye must be obtained by focus-
ing the OCT beam through the pupillary aperture. Posterior eye
Cornea Choroid and choriocapillaris
The corneal epithelium, Bowman’s layer, and the cor- The choriocapillaris is a network of fenestrated chor-
neal lamellae can be imaged using ultrahigh resolu- oidal capillaries located between the sclera and
tion OCT. Corneal adapter modules focus the OCT Bruch’s membrane that provides nourishment to the
beam on the cornea, while post-processing algorithms nearby retinal pigment epithelium (RPE) and photo-
correct for refraction of the OCT beam at the anterior receptors. In first-generation OCT images, a single
corneal surface. The corneal epithelium is character- layer described as the RPE-choriocapillaris complex
ized by a low scattering layer on OCT (Figure 3.1C). was visible, and visualization of the individual layers
The striated pattern seen in the corneal stroma is of the outer retina beyond the photoreceptor inner
attributed to its collagen structure and organization segments was not possible. With improvements in
[4]. Echo reflection artifacts appear in a plane through axial image resolution, discernment of the compo-
the center of the cornea, with a strong reflection pre- nents of the RPE-choriocapillaris complex became
sent where the OCT beam is orthogonal to the air– possible. 15
 Chapter 3: Anatomy of the anterior visual pathway

Figure 3.2 Detailed anatomy of the anterior segment of the eye and adjacent structures.
Reproduced with permission from Gray’s Anatomy.

Ultrahigh resolution OCT used for research pur- with a mean thickness of 24 μm, with optical proper-
poses provides more detailed imaging of the chorio- ties consistent with those of its blood-filled capillaries
capillaris and choroid than standard commercially (Figure 3.4) [6].
available OCT. While a standard ~800 nm light source
can distinguish the major layers of the retina, the
penetration of this wavelength to the choroid and Bruch’s membrane
choriocapillaris is limited by the optical scattering Bruch’s membrane (BM) is positioned between the
and absorbing properties of the RPE. The chorioca- choriocapillaris and the RPE. It regulates the recipro-
pillaris is best visualized with ~1050 nm light, which cal exchange of nutrients and waste between the retina
undergoes less scattering and absorption by melanin and general circulation. It normally increases in thick-
and can provide additional diagnostic information ness with age [7]. According to some investigators,
despite a lower axial resolution [5]. The choriocapil- BM measures approximately 5 μm in thickness by
16 laris appears on OCT as a highly backscattering layer OCT [8]. Others maintain that there is uncertainty
 Chapter 3: Anatomy of the anterior visual pathway

Superior Rectus Muscle

Retina Choroid Sclera
Posterior Chamber Fovea
Macula
Ora serrata Bulbar sheath
Ciliary Muscle Vorticose Vein

Ciliary Zonules
Optic Nerve

Canal of Schlemm
Central Retinal Vein

Pupil

Anterior Chamber

Cornea

Iris Central Retinal Artery

Lens Cortex Optic nerve sheath

Lens Nucleus Optic disc

Ciliary process Retinal arteries and veins

Conjunctiva Medial rectus muscle

Inferior rectus muscle
Inferior oblique muscle

Figure 3.3 Anatomy of the human eye. Used with permission, Chabacano CC-SA 3.0.

Vitreous

RNFL
GCL
IPL
INL
OPL
ONL

Choriocapillaris
Choroid

ELM

IS/OS
OS
RPE
Rod/cone OS tips

Figure 3.4 OCT of the fovea and identification of retinal layers. RNFL: Retinal nerve fiber layer; GCL: ganglion cell layer; OPL: outer plexiform
layer; INL: inner nuclear layer; IPL: inner plexiform layer; ONL: outer nuclear layer; ELM: external limiting membrane; OS: photoreceptor outer
segment; IS/OS: photoreceptor inner segment/outer segment junction; RPE: retinal pigment epithelium.

17
 Chapter 3: Anatomy of the anterior visual pathway

about the ability of OCT technology to resolve BM demonstrated that the outer HRL has several compo-
because its reported thickness is 1–5 μm and the axial nents. The RPE may have a multilayer appearance on
resolution of ultrahigh resolution OCT is currently OCT due to differences in the length of cone and rod
limited to 2–3 μm [9]. photoreceptor tips (Figure 3.4). Rods are longer and
interdigitate with the RPE in the peripheral retina.
Retinal pigment epithelium (RPE) Variation in the cone and rod distribution in the
retina also influences the RPE’s layered appearance
The retinal pigment epithelium lies just below the
[8]. The highly reflective nature of the RPE allows
photoreceptor layer in the retina, where its function
comparatively little light to pass through to the layers
is to provide metabolic and structural support for the
below. Below the choriocapillaris, the deep choroid
photoreceptors. The RPE-choriocapillaris layer is pre-
and sclera are visible as weakly reflective layers.
sent throughout the retina and terminates at the optic
disc margin, where the choroidal circulation also ter-
minates, at the lamina cribrosa (Figure 3.5). Light that Retina
is not absorbed by the photoreceptor layer is absorbed Once light traverses the eye, it is focused on the
by the RPE, which contains the pigment melanin and neurosensory retina. The retina is traditionally
prevents backscattered light from interfering with divided into ten layers, including four cellular layers
high-acuity vision in humans. The mean RPE layer and two layers of neuronal interconnections, many of
thickness is 9 μm [6]. which can be delineated by OCT (Figure 3.4) [6]. To
On first-generation OCT, the RPE and chorioca- arrive at the photoreceptors, light must first pass
pillaris were seen as indistinguishable components of through the retinal nerve fiber layer (RNFL), ganglion
a highly reflective layer (HRL) because these struc- cell layer (GCL), inner plexiform layer (IPL), the ama-
tures have similar optical properties and are in close crine, bipolar, and horizontal cells of the inner nuclear
proximity [10]. Later generations of OCT, however, layer (INL), and the neuropil of the outer plexiform

36 / 105 Optic disc margin Figure 3.5 Papillomacular bundle OCT

A scan from the optic disc through the fovea
(A) and perifoveal region (B). Note the
increasing thickness of the retinal nerve
fiber layer (RNFL) approaching the optic
disc.

200 µm 200 µm

79 / 105
B

18 200 µm 200 µm
 Chapter 3: Anatomy of the anterior visual pathway

Figure 3.6 Schematic representation of retinal layers and their cellular components. The anterior retina, where
light first reaches the retina, is at the top of the figure. Light passes through each of the layers before
reaching the outer segments of the rod and cone photoreceptors, where the phototransduction cascade is initiated.
Reproduced with permission from Gray’s Anatomy.

layer (OPL) (Figure 3.6). The photoreceptors convert Controversy persists, however, regarding the pre-
the electromagnetic energy of light into an electro- cise correlation between layers of the outer retina
chemical signal. Through a series of cellular connec- visualized on OCT (including the RPE, Bruch’s mem-
tions, the signal is modified and relayed to ganglion brane, and the choriocapillaris) and those demon-
cells, whose axons compose the retinal nerve fiber strated by histology [9]. For example, the strongest
layer (RNFL) and serve as the final common output continuous backscattering signal in the outer retina
pathway of the retina. on ultrahigh resolution OCT images has been inter-
The axial resolution of many commercially preted as the RPE. While the optical properties of this
available OCT instruments is 4–7 μm. Generally, backscattering layer are consistent with the RPE,
layers containing primarily cell bodies are weakly some studies have reported that it appears thicker on
backscattering on OCT (dark), whereas layers con- OCT (up to 20–30 μm) [9] than on histologic sections
taining primarily neuropil are highly backscattering (~9–12 ± 2 μm) [12]. On the other hand, other inves-
(bright). In animal preparations, studies have tigators report less of a discrepancy, with an average
demonstrated that many of the layers viewed on RPE thickness on OCT of 9 μm [6]. In the interpreta-
ultrahigh resolution OCT can be reliably correlated tion of retinal OCT images, it is important to recog-
with histology [11]. The layers of the inner retina nize that OCT relies on differences in the optical
seen on OCT, extending from the vitreoretinal properties of different retinal layers, whereas histol-
interface to the junction of the inner and outer ogy relies on postmortem tissue staining and light
photoreceptor segments (IS/OS junction), correlate microscopy. While OCT provides extremely high spa-
well with histologic preparations of monkey tial resolution, the correlation between these images
retina [11]. and histology is likely to be imperfect.
19
 Chapter 3: Anatomy of the anterior visual pathway

Macula and fovea ending. The membrane of the outer segment has a
stacked-disc configuration that houses visual pig-
The macula is located temporal to the optic nerve and
ments, such as rhodopsin. The inner segment con-
measures approximately 5.5 mm in diameter. The
tains mitochondria and ribosomes, where opsin
macula is further divided into the fovea (1.5 mm in
molecules are assembled to be relayed to the outer
diameter) and foveola (0.35 mm). The fovea has up to
segment [15].
200,000 cones/mm2 (nearly 15-fold higher than in
Ultrahigh resolution OCT has been able to charac-
peripheral retina) so that it can provide excellent
terize outer retinal morphology with impressive detail,
visual acuity [13]. Furthermore, high-acuity vision is
even to the level of visualizing rod and cone outer
subserved by the unique synaptic configurations in
segment tips [8]. A layer of minimal reflectivity (dark
the fovea, where each bipolar cell receives input
on OCT) just anterior to the RPE and choriocapillaris
from a single cone photoreceptor and each cone relays
likely represents the outer segments (OS) of the retinal
information to two ganglion cells [14]. The fovea
photoreceptors. This layer is thickest in the fovea,
contains an avascular center (0.4–0.6 mm in dia-
where there is an abundance of cone photoreceptors.
meter) where photoreceptors can be densely packed
Outside the fovea, the OS layer is less defined, because
without making provision for intervening capillaries.
rods and cones with different outer segment lengths are
The foveal depression reaches its maximum depth at
mixed and give rise to a faintly layered appearance
the center of the macula. Here, the inner layers of the
between the RPE and the inner segment/outer segment
retina, including the RNFL, GCL, OPL, INL, and IPL,
interface (IS/OS). The boundary between the inner and
taper and nearly disappear to allow light to directly
outer segments (IS/OS) is visible on OCT as a highly
stimulate the underlying cone photoreceptors without
backscattering layer (bright on OCT) just anterior to
distortion. The ONL and photoreceptor outer seg-
the RPE (Figure 3.4). The backscattering nature of this
ment layers become thicker in the foveal pit, reflecting
layer suggests there is a difference in the refractive
this region’s specialization for high spatial acuity. The
index of the outer and inner photoreceptor segments
fovea is readily identified using OCT and character-
[8].
ized by thinning of the anterior retina and dense
The distribution of cones and rods across the
aggregation of photoreceptors (Figure 3.4).
retina is highly skewed and directly reflects the spe-
A map illustrating macular thickness in segments
cialized functions of the fovea and retinal periphery
can be generated from a series of scans centered on
[16]. OCT demonstrates that the thickness of the
the fovea. The thickness measurements are typically
photoreceptor layer is increased in the fovea, where
displayed in a segmented circular graph, with each
there is a high concentration of cones with long outer
segment colored to illustrate the relative thickness
segments to support high visual acuity. Photoreceptor
(Figure 3.10A and C).
inner and outer segment length both decrease with
increasing distance from the foveola. The retinal per-
Photoreceptors iphery has lower concentrations of cones and is asso-
After photons pass through the inner layers of the ciated with lower visual acuity [8]. The tips of cone
retina, including the GCL and INL and ONL layers, outer segments extend to the surface of the RPE in the
they reach the photoreceptors in the outer retina. foveola, but separate from the RPE with increasing
There, they initiate a photochemical cascade that distance from the foveal center.
results in a change in the photoreceptor’s firing rate. Rods are virtually absent in the fovea; they are the
Humans possess four photoreceptor types: three dominant photoreceptor in the periphery, where they
cones and a rod. Each type of cone photoreceptor specialize in high sensitivity motion and light detec-
has a unique, optimal response to specific wavelengths tion. Rod outer segment tips are longer than those of
of light, either short (blue), middle (green), or long cones, extend to the RPE surface, and interdigitate
(red). Rods, on the other hand, are saturated at nat- with the RPE in the peripheral retina. The difference
ural light intensities and are incapable of discriminat- in outer segment length between rods and cones is
ing colors; their higher sensitivity to light renders demonstrated by bands visible on OCT just posterior
them effective for night vision (scotopic vision). The to the IS/OS junction band, which correspond to the
retinal photoreceptors are each made up of an outer cone and rod outer segment tips (Figure 3.4). The
20 segment, inner segment, nucleus, and synaptic distance between the IS/OS junction band and outer
 Chapter 3: Anatomy of the anterior visual pathway

segment tips bands is believed to approximate the The inner plexiform layer (IPL) is found between
outer segment length [8]. the GCL and INL and contains neuropil made of
ganglion cell dendrites, bipolar cell axons, and ama-
Nuclear and plexiform layers crine cell axons (Figure 3.6). The IPL is moderately
backscattering on OCT images, producing a light gray
The neurosensory retina consists of three major layers:
layer on grayscale OCT and a blue-green layer on
photoreceptors synapse with bipolar cells, which then
false-color images. The abundance of synapses in the
relay their signal to ganglion cells. In addition, hori-
IPL provides one of the last exchanges of visual infor-
zontal and amacrine cells form lateral connections
mation before being transmitted by the retinal gang-
between elements of these layers via direct excitatory
lion cell axons to the lateral geniculate nuclei of the
or indirect inhibitory connections (Figure 3.6) [17].
thalamus.
While each bipolar cell receives input from a single
Ganglion cells serve as the final output pathway of
photoreceptor in the fovea, bipolar cells in the retinal
visual information from the retina. There is a large
periphery summate the inputs from multiple photore-
convergence of information from the retinal photo-
ceptor cells to support more sensitive detection of light.
receptor cells, which include about 100 million rods
The nuclear layers of the retina exhibit an inter-
and 4 million cones, to the approximately 1.2 million
mediate degree of backscattering of light and appear as
ganglion cells [14]. The majority of ganglion cell
varying shades of gray in grayscale OCT and as blue-
axons project to the lateral geniculate nucleus. There
black in false-color images. The three nuclear layers
are three main types of ganglion cells, each with spe-
that can be discerned include the outer nuclear layer
cialized functions in the detection of visual inputs.
(ONL), inner nuclear layer (INL), and ganglion cell
Anatomical differences underlie important physiolo-
layer (GCL). The inner and outer plexiform layers,
gical and functional differences in these types of cells.
which contain mostly neuropil and dense synaptic con-
Eighty percent of ganglion cells are midget cells, 10%
nections, demonstrate greater optical backscattering.
are parasol cells, and 10% are other types [15]. The
The ONL is composed of cell bodies of rods and
fovea has a high concentration of midget ganglion
cones (Figure 3.5) and shares the optical properties of
cells. Via a narrow dendritic tree, these cells receive
the other nuclear layers. The ONL increases in thick-
signals conveyed by a bipolar cell that has received
ness at the fovea, despite reduced total retinal thick-
inputs from a single cone. The GCL is made of gang-
ness. The external limiting membrane (ELM) is
lion cell bodies that lie just posterior to the retinal
visualized with OCT as a thin, backscattering layer
nerve fiber layer (RNFL). Like other nuclear layers in
posterior to the ONL. It is not a physical membrane
the retina, the GCL is weakly backscattering on OCT,
but probably represents a collection of junctions
giving rise to a dark gray layer on grayscale OCT and
between photoreceptor cells and the supportive
blue-black layer on false-color images. The GCL is
Müller cell processes. The membrane is situated
thickest in the parafoveal region because of the high
between the photoreceptor cell body and the inner/
density of photoreceptors in the fovea, but decreases
outer segments of the photoreceptor.
in thickness toward the optic disc (Figure 3.5A).
The outer plexiform layer (OPL), also known as
Henle’s fiber layer, is somewhat thinner than the inner
plexiform layer (IPL) and contains horizontal cell Retinal nerve fiber layer
dendrites, bipolar cell dendrites, and photoreceptor The axons of ganglion cells travel in the retinal nerve
axons (Figure 3.6). The synapses here represent the fiber layer (RNFL), enter the optic nerve, travel
first integration of visual information in the retina. through the chiasm and tract, and then finally synapse
The OPL backscatters light to a moderate degree, primarily in the lateral geniculate nucleus of the tha-
resulting in a lighter gray layer on grayscale OCT lamus. Foveal ganglion cells send axons directly to the
and a mostly green layer in false-color images. temporal aspect of the optic disc in the papillomacular
The inner nuclear layer (INL) is composed of the bundle. The remaining temporal ganglion cell nerve
cell bodies of bipolar cells, amacrine cells, and horizon- axons are arranged on either side of the horizontal
tal cells (Figure 3.6). This layer is weakly backscatter- raphe and form arcuate bundles that course above and
ing, giving rise to a dark gray layer on grayscale OCT below the fovea and finally enter the superior and
and a blue-black layer on false-color tomograms. inferior portions of the optic nerve. Finally, axons
21
 Chapter 3: Anatomy of the anterior visual pathway

Nasal Inferior Temporal Superior Nasal

–8.0°

200 µm

200 µm

1000
Inferior Arcuate Nerve Fiber Bundle Superior Arcuate Nerve Fiber Bundle
800
Thickness [µm]

600

400

200

0
–180 –135 –90 –45 0 45 90 135 180
NAS INF TMP SUP NAS
Position [°]

Figure 3.7 Circular peripapillary retina OCT scan. A scanning path with a 3.4 mm diameter (green circle) begins at the nasal retina (green
arrowhead) and proceeds in a clockwise direction. The corresponding OCT image is displayed on the right with labeled retinal segments. Note
the relative thickening of the retina superior and inferior to the optic disc, representing the superior and inferior arcuate nerve fiber bundles.
The lower panel graphs the thickness of the peripapillary retina in a clockwise scan starting and ending in the nasal retina. The star (*) denotes
an area of optical shadowing from an overlying blood vessel. NAS: nasal; INF: inferior; TMP: temporal; SUP: superior; INL: internal limiting
membrane; BM: basement membrane.

originating nasal to the disc enter the nasal portion of inferotemporal bundling of nerve fibers are demon-
the optic nerve. strated by the increased RNFL thickness in these
The RNFL is highly backscattering, appearing as a regions observed in circular tomographs (Figure 3.7).
bright layer on grayscale tomograms and as a red layer OCT algorithms are capable of measuring the
in false-color tomograms. It is the first layer of the thickness of the RNFL. The standard for measure-
retina encountered by the OCT beam and is found at ment of RNFL thickness is a circular scan around
the top of the computer-generated tomogram, imme- the optic disc (peripapillary region) (Figure 3.8). A
diately beneath the transparent vitreous. The RNFL diameter of 3.4 mm has been suggested as a standard
varies in thickness with retinal location. As the axons size for the circle, as a diameter of 2.9 mm has been
of retinal ganglion cells travel toward the optic nerve associated with lower reproducibility and a 4.5 mm
head before forming the optic nerve, they accumulate circle captures a thinner part of the RNFL, which
and the RNFL thickness increases from the macula to could potentially reduce the sensitivity of detection
the optic disc (Figure 3.5A) [4]. An increase in the of small RNFL defects [18]. The normative database is
slope of the retinal nerve fibers as they bend to form also based upon 3.4 mm circular OCT scans.
the optic nerve head causes a decrease in OCT signal. However, the optic disc margin is not perfectly circu-
The retinal nerve fibers exhibit direction-dependent lar, so a peripapillary scan path is not equidistant from
reflectance, with maximal reflectance observed when the disc margin at all points, which is a potential
the nerve fibers are perpendicular to the OCT optical source of error in these measurements. The RNFL is
beam, as in the majority of the retina. In the macular well known to increase in thickness approaching the
region, the RNFL is thin because the bulk of the retina optic disc, which may account for an increase in the
is occupied by densely packed photoreceptors. RNFL thickness measured with a standard 3.4 mm
Circular tomographs performed with a 1.5-disc circular scan in persons with larger optic discs [20].
diameter around the optic disc demonstrate a varia- OCT software facilitates RNFL comparisons between
tion in RNFL thickness from 40 μm nasally to 230 μm the two eyes, where even subtle asymmetries can be
22 superotemporally [10]. The superotemporal and detected.
 Chapter 3: Anatomy of the anterior visual pathway

–8.0°

200 µm

200 µm

Peripapillary RNFLT Classification

300

TS NS
144 109 240
(137) (102)
Thickness [µm]

180

T G N N/T
PMB 84 104 75 0.89 120
62 (77) (98) (72) (0.93)
(58)
60

TI NI
158 103 0
(145) (107) –180 –135 –90 –45 0 45 90 135 180
NAS INF TMP SUP NAS
Position [°]
Within Normal Limits
Warning: Classification results valid for Caucasian eyes only.

Figure 3.8 Measurement of peripapillary retinal nerve fiber layer (RNFL) thickness using OCT. Using a circular scan with a diameter of 3.4 mm
centered in the optic disc (green line) measures one of the thickest regions of the RNFL, thereby increasing the scan’s sensitivity. Thickness
measurements can be plotted against age-matched normative data (bottom right). Purple: above normal limits (p<0.01); blue: borderline
above normal limits (p<0.05); green: within normal limits (p>0.05); yellow: borderline below normal limits (p<0.05); red: below normal limits
(p<0.01).

Retinal vasculature the optic disc, which gives rise to the monocular
blind spot [23]. The optic nerve travels posteriorly
The retinal blood supply is derived from the superior
through the lamina cribrosa to exit the back of the
and inferior branches of the central retinal artery.
globe, where it abruptly increases in diameter from
Deoxygenated blood is collected from capillaries into
3 mm to 4 mm.
venules that are posterior to the arteries and drain
The contour of the optic disc is easily visualized
into the main retinal veins. These vessels travel on the
with OCT and demonstrates the normal optic cup and
anterior surface of the retina and are evident in tomo-
neuroretinal rim (Figure 3.9). The RPE, choriocapil-
graphs as focal areas of increased backscatter that cast
laris, and photoreceptor layers terminate near the
an optical shadow on deeper retinal structures
lamina cribrosa, and the site of termination can be
(Figure 3.7, see star). Vascular structures found dee-
used as a marker for the disc margin.
per in the optic nerve head are discussed below.
Because the retina is only about 0.5 mm thick, the
axial dimension of OCT images is expanded to allow
Optic nerve head complex better visualization of the retinal layers (Figure 3.5).
Each optic nerve is composed of approximately 1.2 This expansion exaggerates the features of the optic
million retinal ganglion cell axons (in contrast to the nerve’s head and cup, which must be interpreted with
acoustic nerve, for example, which has only 31,000 the appropriate axial and transverse scales. Disc dia-
axons) [21]. The intraocular segment of the optic meter, cup diameter, neuroretinal rim area, and cup-
nerve head (the optic disc) is typically located to-disc ratio can be measured with OCT.
3–4 mm nasal to the fovea and is 1 mm thick. The Detailed correlation between optic nerve head
optic disc has a central depression, called the optic histology and SD-OCT images has been confirmed
cup, which is typically 1/3 the size of the disc [22]. in monkeys [24]. Landmarks of the neural canal open-
There are no retinal photoreceptors in the region of ing and the anterior lamina cribrosa were also 23
 Chapter 3: Anatomy of the anterior visual pathway

Neuroretinal rim

Optic
cup

200 µm 200 µm

Figure 3.9 Optic nerve head appearance on OCT. The horizontal bold arrow depicts the scanning path used to obtain the image on the right.
The abundance of vessels in the nasal half of the disc and the sharp increase in slope of the retinal nerve fibers entering the optic nerve cause a
degradation of the OCT signal.

confirmed with histology, although deeper optic reflective layers (HRL) on retinal OCT images at the
nerve head structures are not visualized as well by inner and outer retinal boundaries. The first HRL, in
SD-OCT [24]. the inner retina, represents the vitreoretinal surface or
The lamina cribrosa (LC) is a mesh-like segment ILM, and the second HRL, in the outer retina, repre-
of the sclera with pores through which the retinal sents the RPE. The measurement between these layers
ganglion cell axons exit the eye to enter the optic is reported as the retinal thickness (Figure 3.10) [19].
nerve. SD-OCT imaging of the lamina cribrosa has Computer algorithms can automatically detect
recently been greatly improved by the use of adaptive retinal layer boundaries to segment the internal
optics (AO) and computer algorithms [25]. structure of the retina and measure the thickness of
Deeper structures in the optic nerve complex have some layers (Figure 3.8). Errors in boundary detec-
been traditionally difficult to image due in part to tion that occur with signal obscuration by vascula-
optical backscattering from the RPE. In addition, ture or signal dropout are also corrected via
commercially available OCT machines are typically computer algorithm [6]. In previous-generation
set up to preferentially image the retina. The sclera OCT, errors in measuring the true anatomic retinal
and deeper structures are visualized by a defocused, thickness were introduced by the software’s inability
conical OCT imaging beam that widens after passing to reliably distinguish between the RPE and the IS/
through the retina. OS junction as the outer retinal boundary [19].
An imaging technique termed “enhanced depth However, the latest generation of ultrahigh resolu-
imaging” (EDI) SD-OCT involves moving the OCT tion OCT is able to more reliably measure the actual
instrument closer to the eyes, allowing the focused retinal thickness from the RPE [28].
OCT beam to image structures deep to the retina Computer algorithms are now able to reliably seg-
[26]. Deep vascular structures, such as the central ment the retina into at least four layers at the macula:
retinal artery and vein and the short posterior ciliary the RNFL, GCL + IPL, OPL + INL, and ONL + photo-
arteries, can be identified using the EDI SD-OCT receptor layers [29]. Although many of the individual
technique. The peripapillary choroid and sclera, and, component layers may be discerned by close visual
less frequently, the subarachnoid space around the inspection of OCT images, reliable segmentation algo-
optic nerve, are also observed via EDI SD-OCT [27]. rithms are necessary to obtain standardized quantita-
tive measurements for disease monitoring.
Automated quantitative retinal
measurements Three-dimensional OCT
OCT is able to quantify retinal thickness by measuring Improvements in image acquisition speed with
24 the distance between retinal layers. There are highly spectral-Fourier-domain (SD/FD) detection and
 Chapter 3: Anatomy of the anterior visual pathway

13 / 25 800
B
A
700

600

500

Retina Thickness [µm]

400 200 µm

Examina
Retina Thickness [µm]
300
C Average Thickness [µm]
Marker 800
296 MI
226 µm
1.57 700
341 Center
600
0.54 226 µm
200
500
Central Min
309 340 274 324 282
1.64 0.53 0.22 0.51 1.50 225 µm 400

Central Max 300

100
335 315 µm
0.53 200

290 100
200 µm Circle Diameters: 1, 3, 6 mm ETDRS 1.54 200 µm
0 0

Figure 3.10 Measurement of retinal thickness using OCT. (A) A combination of vertical scanning paths (green line) are used to generate a
map of the fovea and surrounding retina. (B) A segmentation computer algorithm measures the distance between the inner and outer
boundaries of the retina: the basement membrane (BM) and the internal limiting membrane (ILM), respectively (red lines). (C) The data is used
to generate a topographic map, with color coding of retinal thickness across various segments.

other advances in OCT technology have paved permeability and leakage, as with traditional fluoros-
the way for three dimensional OCT images. The cein angiography.
50–100-fold increase in speed with SD/FD detection Some of the latest OCT techniques have demon-
allows many closely spaced scans to be performed, strated the ability to assess retinal physiology (termed
which may cover the fovea, for example, and be used “optophysiology”) in addition to structure.
to assemble a volumetric map of the area imaged. Interestingly, physiologic processes that occur at the
Software programs enable 3-D renderings that can cellular level, such as depolarization and changes in
be rotated and viewed from any angle, or even dis- cellular swelling, can cause small changes in a retinal
sected so that individual layers can virtually be cell’s optical properties, which can be detected with
removed. 3-D OCT can also provide data for quanti- OCT. For example, changes in the optical backscatter-
tative mapping of individual retinal layers. ing of the photoreceptor outer segment layer can be
detected following a single flash of light [8]. The
Future OCT applications clinical application of this technology requires further
research and validation.
Variations of the application of OCT technology have
Recent technological developments have intro-
allowed the extraction of functional information from
duced adaptive optics (AO), which allow for correc-
the eye. One example is Doppler OCT, in which
tion of chromatic aberrations inherent to the human
specialized techniques permit the calculation of the
eye that have limited the ability of OCT to resolve
Doppler shift due to moving blood with flow sensitiv-
individual cells in the retina. AO restore the ability of
ities on the order of tens of micrometers/second.
OCT to see even the conic shape of the cone outer
Arteries and veins can be distinguished and quantita-
segments in some cases. Images using AO have esti-
tive imaging of retinal blood flow can be performed,
mated the width of a cone to be ~4 μm, as they are in
even of individual vessels [9]. The term “optical
histologic sections [9]. AO has also allowed visualiza-
coherence angiography” has been given to the use of
tion of the lamina cribrosa with commercially avail-
Doppler OCT to generate 3-D visualization of retinal
able OCT, although further research is needed to
vasculature [30]. Doppler OCT methods enable mea-
establish the utility of AO OCT for other clinical
surements of blood flow within the retinal vascula-
applications.
ture, but they are unable to assess vascular
25
 Chapter 3: Anatomy of the anterior visual pathway

7. Booij JC, Baas DC, Beisekeeva J, et al. The dynamic

Conclusion nature of Bruch’s membrane. Prog Retin Eye Res 2010;
The anatomy of the anterior visual pathways is well- 29: 1–18.
visualized using currently available OCT technology. 8. Srinivasan VJ, Monson BK, Wojtkowski M, et al.
Research has demonstrated good correlation between Characterization of outer retinal morphology with
OCT retinal layers and retinal histopathology [11]. high-speed, ultra-high resolution optical coherence
Nevertheless, as OCT relies on differences in optical tomography. Invest Ophthalmol Vis Sci 2008; 49:
properties and histopathology relies on light micro- 1571–9.
scopy and fixed tissue staining to generate contrast, 9. Drexler W, Fujimoto JG. State-of-the-art retinal opti-
there are small differences. Even with these differ- cal coherence tomography. Prog Ret Eye Res 2008; 27:
ences, OCT is able to discern important features of 45–88.
both the anterior and posterior segments of the eye, 10. Hee MR, Izatt JA, Swanson EA, et al. Optical coherence
including the conjunctiva, cornea, anterior chamber tomography of the human retina. Arch Ophthalmol
1995; 113: 325–32.
and angle, lens, posterior chamber, retina, and optic
nerve head complex. Ultrahigh resolution OCT 11. Anger EM, Unterhuber A, Hermann B, et al. Ultrahigh
resolution optical coherence tomography of the mon-
approaches the ability of histopathology to resolve
key fovea. Identification of retinal sublayers by corre-
retinal details and approximates a “living biopsy” by lation with semithin histology sections. Exp Eye Res
which detailed anatomic information can be obtained. 2004; 78: 1117–25.
Beyond the ability to produce exquisite structural 12. Spraul CW, Lang GE, Grossniklaus HE. Morphometric
images, the adaptation of OCT technology to extract analysis of the choroid, Bruch’s membrane, and retinal
functional, physiologic information from the eyes pigment epithelium in eyes with age-related macular
suggests exciting future applications for OCT in ante- degeneration. Inv Ophth & Vis Sci 1996; 37: 2724–35.
rior visual pathway imaging. 13. Hirsch J,d Curcio CA. The spatial resolution capacity
of human foveal retina. Vision Res 29: 1095–101.
References 14. Rosenzweig MR, Breedlove SM, Leiman AL. Biological
1. Huang D, Swanson EA, Lin CP, et al. Optical coherence Psychology 3rd Edition, Sunderland, MA: Sinauer
tomography. Science 1991; 254: 1178–81. Associates, Inc. 2002; Chapter 10, pp. 288–93.
2. Kagemann L, Wollstein G, Ishikawa H. 15. Prasad S, Galetta SL. Anatomy and physiology of
Visualization of the conventional outflow pathway the afferent visual system. Handb Clin Neurol 2011;
in the living human eye. Ophthalmology 2012; 119: 102: 3–19.
1563–8. 16. Osterberg G. Topography of the layer of rods and
3. Riordan-Eva P. Chapter 1. Anatomy & embryology of cones in the human retina. Acta Ophthalmol 1935;
the eye. In: Riordan-Eva P, Cunningham, Jr. ET, 13: 1–103.
eds. Vaughan & Asbury’s General Ophthalmology. 18th 17. Flores-Herr N, Protti DA, Wassle H. Synaptic
ed. New York: McGraw-Hill; 2011. http://www.access currents generating the inhibitory surround of gang-
medicine.com/content.aspx?aID=55780501. Accessed lion cells in the mammalian retina. J Neurosci 2001;
May 22, 2013. 21: 4852–63.
4. Wollstein G, Folio LS, Nevins JE, et al. Chapter 2. 18. Schuman JS, Pedut-Kloizman T, Hertzmark E, et al.
Interpretation of the optical coherence tomography Reproducibility of nerve fiber layer thickness mea-
image. In: Schuman JS, Puliafito CA, Fujimoto JG, surements using optical coherence tomography.
Duker JS, eds. Optical Coherence Tomography of Ocular Ophthalmology 1996; 103: 1889–98.
Diseases. 3rd ed. Thorofare, NJ: SLACK Incorporated;
19. Costa RA, Skaf M, Melo Jr LAS, et al. Retinal assess-
2013.
ment using optical coherence tomography. Prog Ret
5. Považay B, Herman B, Unterhuber A, et al. Eye Res 2006; 25: 325–53.
Three-dimensional optical coherence
20. Savini G, Zanini M, Carelli V, et al. Correlation
tomography at 1050 nm versus 800 nm in retinal
between retinal nerve fibre layer thickness and
pathologies: enhanced performance and
optic nerve head size: an optical coherence
choroidal penetration in cataract patients. J Biomed
tomography study. Br J Ophthalmol 2005; 89:
Optics 2007; 12: 041211.
489–92.
6. Drexler W, Morgner U, Ghanta RK, et al. Ultrahigh-
21. Bruesch SR, Arey LB. The number of myelinated
resolution ophthalmic optical coherence tomography.
and unmyelinated fibers in the optic nerve of verte-
Nat Med 2001; 7: 502–7 (Erratum in: Nat Med
26 brates. J Comp Neurol 1942; 77: 631–655.
2001; 7: 636).
 Chapter 3: Anatomy of the anterior visual pathway

22. Jonas JB, Gusek GC, Naumann GO. Optic disc, cup 26. Spaide RF, Koizumi H, Pozonni MC. Enhanced depth
and neuroretinal rim size, configuration and correla- imaging spectral-domain optical coherence tomogra-
tions in normal eyes. Invest Ophthalmol Vis Sci 1988; phy. Am J Ophthalmol 2008; 146: 496–500.
29: 1151–8. 27. Park SC, De Moraes CG, Teng CC, Tello C, Liebmann
23. Mariotte LA. A new discovery touching vision. Phil JM, Ritch R. Enhanced depth imaging optical coher-
Trans R Soc Lond 1668; 3: 668–71. ence tomography of deep optic nerve complex struc-
24. Strouthidis NG, Grimm J, Williams GA, Cull GA, tures in glaucoma. Ophthalmology 2012; 119: 3–9.
Wilson DJ, Burgoyne CF. A Comparison of optic 28. Wojtkowski M, Srinivasan VJ, Fugimoto JG, et al.
nerve head morphology viewed by spectral Three-dimension retinal imaging with high-speed
domain optical coherence tomography and by ultrahigh-resolution optical coherence tomography.
serial histology. Invest Ophthalmol Vis Sci 2010; 51: Ophthalmology 2005; 112: 1734–46.
1464–74. 29. Walter SD, Ishikawa H, Galetta KM, et al. Ganglion cell
25. Kim TW, Kagemann L, Girard MJ, et al. Imaging of the loss in relation to visual disability in multiple sclerosis.
lamina cribrosa in glaucoma: perspectives of patho- Ophthalmology 2012; 119: 1250–7.
genesis and clinical Applications. Curr Eye Res 2013; 30. Makita S, Hong Y, Yamanari M, et al. Optical coher-
38: 903–909. ence angiography. Opt Express 2006; 14: 7821–40.

27
 Chapter
Optical coherence tomography in acute

4 optic neuritis
Andrew P. D. Henderson, S. Anand Trip, and David H. Miller

Background Visual acuity is nearly universally diminished by

optic neuritis. The effect may range from mild to no
Optic neuritis usually refers to a symptomatic attack
perception of light, although moderate impairment is
of inflammatory demyelination of the optic nerve.
more common than either severe or mild impairment
This is usually of acute onset and presents with pain-
[5]. Color vision is almost always affected by optic
ful loss of vision, which worsens over several days, and
neuritis, even after visual recovery. In the Optic
recovers (partially or completely) over several
Neuritis Treatment Trial (ONTT), 84% of patients
months. In Western countries, the most common
had detectable abnormalities when using Ishihara
form of optic neuritis is that associated with multiple
pseudoisochromatic plates, and the Farnsworth-
sclerosis (acute idiopathic optic neuritis), but there
Munsell 100 Hue test was abnormal in 94% [6].
are other forms of optic neuritis (such as that
Although the typical description of the visual loss of
associated with neuromyelitis optica (NMO), and
optic neuritis is of a centrocaecal scotoma, any visual
chronic relapsing inflammatory optic neuropathy
field deficit may occur. In the ONTT, the most com-
(CRION) [1, 2]). Optic neuritis is highly associated
mon visual field defect noted by automated perimetry
with multiple sclerosis. Approximately 50% of all
was a diffuse reduction in sensitivity in the affected eye,
patients with multiple sclerosis have optic neuritis at
followed by altitudinal or other nerve fiber bundle
some point in the course of their disease, and for
defects. Central or centrocaecal scotomas were only
many patients, it is the first clinical episode.
seen in 8.2% of patients at presentation [7]. There is
The annual incidence of acute idiopathic optic
almost always a detectable relative afferent pupillary
neuritis in temperate climates is approximately 5 per
defect in acute optic neuritis [8]; the exceptions to
100,000. Although optic neuritis can occur at any
this are if there has been a previous attack of optic
age, the peak incidence is in the fourth decade of
neuritis in the fellow eye and the afferent arms of the
life, and women outnumber men approximately
pupillary light reflex are equally affected.
three to one [3, 4].
Approximately one third of patients with optic neuritis
Most commonly, optic neuritis is unilateral. The
will have detectable swelling of the optic disc on fundo-
usual symptoms of optic neuritis are periorbital pain
scopy (i.e., pappilitis) [5].
(which may precede the onset of visual loss noted by
The prognosis for visual recovery after optic
the patient), pain on eye movement, and reduction in
neuritis is usually good. In the ONTT, the median
vision (both central and peripheral). Patients with optic
visual acuity after twelve months was 20/16, and 95%
neuritis may also experience positive visual phenomena
of patients had a visual acuity of 20/40 or better. Even
(phosphenes), usually provoked by eye movement.
for patients with a visual acuity of 20/200 or worse at
Particularly in the recovery phase of the illness, patients
study entry, 84% had a visual acuity of 20/40 or better
may notice temporary worsening of vision accompanied
at one year [9]. Notwithstanding these favorable prog-
by elevations in body temperature (a type of Uhthoff’s
nostic outcomes, we now know that many people with
phenomenon). The pain typically lasts only a few days,
optic neuritis sustain significant and persistent loss of
and the visual loss may worsen for up to two weeks
low contrast letter acuity, and retinal nerve fiber layer
before recovery ensues.

28 Optical Coherence Tomography in Neurological Disease, ed. Peter A. Calabresi, Laura J. Balcer, and Elliot M. Frohman.
Published by Cambridge University Press. © Cambridge University Press 2015.
 Chapter 4: Optical coherence tomography in acute optic neuritis

thickness (RNFL) reductions, both of which carry In the nine eyes with a past history of optic neuritis,
important ramifications for visual impairment in eight had visible RNFL defects, and five eyes had
low light settings (e.g., driving at night; edge detection visual field defects. MacFadyen and colleagues [18]
such as when walking stairs). described the retinal and electrophysiological find-
In adults, optic neuritis is usually unilateral, but ings in 57 patients with multiple sclerosis, half of
approximately 10% of all optic neuritis is bilateral, whom had a history of optic neuritis. Of the 27 eyes
either simultaneous or occurring in short succession with a clinical history of optic neuritis, 17 had
[10]. The optic neuritis associated with neuromyelitis detectable RNFL defects.
optica is more often bilateral. This should be distin- Unfortunately, although examination of the RNFL
guished from the phenomenon in unilateral acute is possible with a hand-held ophthalmoscope, an
optic neuritis (i.e., unilateral symptoms), where approximately 50% loss of axons is required before
changes in the visual field can be identified on formal RNFL changes are visible [19], and the loss is not
examination in the fellow eye; in the ONTT, 48% of readily quantifiable.
patients had some abnormality detected in the fellow
eye [11]. Retinal nerve fiber layer changes
There are also chronic and subclinical forms of
optic neuritis. The chronic form, which should not be in optic neuritis using optical
diagnosed without excluding compressive lesions on coherence tomography
neuroimaging, can occur as the presenting symptom Parisi and colleagues [20] examined 14 patients with
of multiple sclerosis [12] or develop later in the course multiple sclerosis, all of whom had a prior episode of
of the disease [13]. optic neuritis with complete visual recovery. Using
early generation optical coherence tomography (i.e.,
Retinal nerve fiber layer (RNFL) time-domain OCT), they found that peripapillary
RNFL thickness was significantly reduced in eyes
imaging using fundus photography affected by optic neuritis when compared with
Hoyt and colleagues [14] identified visible nerve fiber healthy control eyes, and was also reduced in clini-
layer defects in the symptomatic eye of a 30-year-old cally unaffected fellow eyes but to a lesser extent.
woman with probable optic neuritis. Following this, This study did not attempt to find an association
Frisén and Hoyt [15] were the first to identify changes with clinical tests of visual function and did not
in the retina in multiple sclerosis; they described two study the macula using OCT.
patients with evidence of retinal nerve fiber layer Trip and colleagues [21] used Stratus OCT (Carl
(RNFL) loss without evidence of clinically overt Zeiss Meditech, Palo Alto, CA) to study 25 patients
acute optic neuritis. They identified these now recog- with a history of a single previous attack of optic
nized cardinal features of optic neuropathy, while neuritis and poor visual recovery, comparing their
utilizing a hand-held ophthalmoscope using red-free affected eye with their clinically unaffected fellow
illumination, and illustrated their findings in a paper eye, and with the eyes of 15 healthy control subjects.
using red-free photography. They found significant reductions of RNFL thick-
Sharpe and Sanders [16] found that retinal nerve ness (68.7 μm in affected eyes vs 94.6 μm in fellow
fiber layer defects, which were more conspicuous eyes vs 102.9 μm in control eyes) and macular
due to hypermyelination of the patient’s retinal volume (6.10 mm3 in affected eyes vs 6.71 mm3 in
nerve fibers (the retinal axons that contribute to fellow eyes vs 6.83 mm3 in control eyes) in the
80% of the composition of the retinal nerve fiber affected eyes of patients, when compared with both
layer are typically unmyelinated, until they make their clinically unaffected fellow eye and the eyes of
their passage through the lamina cribrosa, at which control subjects. Reduced RNFL thickness was sig-
point they achieve oligodendrocyte-derived myeli- nificantly correlated with worse visual function for
nation), in a patient with bilateral optic neuritis in LogMAR acuity, Humphrey visual field mean devia-
the context of multiple sclerosis (see Figure 4.1). tion, and color vision. Reduced macular volume
Elbøl and Work [17] examined red-free retinal correlated with reduced color vision, but not with
photographs of 20 patients with multiple sclerosis, reduced overall field sensitivity, nor with visual
seven of whom had a prior history of optic neuritis. acuity. 29
 Chapter 4: Optical coherence tomography in acute optic neuritis

90° 75°
105°
120° 60°
135° 45°

150° 30°

165° 15°

0° 180°

15° 165°

30° 150°

45° 135°

60° 120°
75° 90° 105°

Figure 4.1 This image demonstrates RNFL defects visible on red-free photography in the left eye of a patient with optic neuritis in association
with multiple sclerosis. There are visual field defects in areas of the visual field corresponding to the visible RNFL defects. From Sharpe and
Sanders [16].

Costello and colleagues studied 54 patients after an attack of optic neuritis, when vision is stable
recruited within one month of onset of an attack (see Table 4.1 and Figure 4.2). Petzold and collea-
of optic neuritis and found reductions in the gues [2] have performed a meta-analysis of most of
retinal nerve fiber layer of the affected eyes when these studies. They found that the mean loss of
compared with that of the fellow eyes of the same RNFL thickness due to optic neuritis was 20.38
group (77.5 μm in the affected vs 99.8 μm for the μm, using the fellow eye as a baseline comparator.
fellow eye) [26]. Lower RNFL values were corre- This may be an overestimate due to ascertainment
lated with worse visual field mean sensitivity, and bias in early studies of acute optic neuritis, and some
those patients with worse vision had significantly patients may have significantly less thinning.
lower RNFL values. The authors hypothesized that In some patients, recurrent attacks of optic neur-
the majority of the RNFL thinning would be evi- itis are associated with further reductions in RNFL
dent by between three and six months after the thickness, and thereby depletion of optic nerve axons
acute attack, although they were not able to ana- [22–24], suggesting that inflammation per se is
lyze their data to calculate the time to first detect- involved in axonal degeneration, in addition to any
able thinning. effects mediated by compression of the optic nerve
Since these papers, several other authors have against the inelastic optic nerve sheath, or within the
30
replicated the findings in the chronic stage, long bony optic nerve canal.
 Chapter 4: Optical coherence tomography in acute optic neuritis

(A)
(B)

30

(D)
(C)

(E) (F)

Figure 4.2 Visual fields (A. and D.), OCT images of the circumpapillary retina (B. and E.), and RNFL thickness maps of the peripapillary retina
(C. and F.) in a 20-year-old woman four months after unilateral left-sided optic neuritis associated with multiple sclerosis.
The visual acuity in the right eye was 6/5 and in the left eye was 6/6. There is reduced RNFL thickness on the left (circumpapillary RNFL
thickness 72 µm compared with 109 µm on the right), despite good visual recovery.

Changes in the macula and deep was reduced when compared to that of the fellow eye
(6.10 mm3 in affected eyes vs 6.71 mm3 in unaffected
layers of the retina patient eyes) and also when compared to that of
Trip and colleagues [21], in their cohort of optic healthy control eyes (6.83 mm3). Pulicken and collea-
neuritis patients selected for poor recovery, found gues [28] found reduced macular volume in the eyes
that the overall macular volume in the affected eye of subjects with a history of optic neuritis, in a cohort 31
 Chapter 4: Optical coherence tomography in acute optic neuritis

Table 4.1 RNFL findings after optic neuritis

Optic neuritis Fellow eyes Healthy control

Mean n Mean n Mean n
(SD) (SD) (SD)
Trip (2005) [21] 68.7 (18.8) 25 94.6 (14.9) 25 102.9 15
(14.6)
Fisher (2006) [25] 85 (17) 63 –* – 105 (12) 72
Costello (2006) 77.5 54 99.8 54 – –
[26]
Albrecht (2007) [27] 74.47 (22.15) 21 – – 103.4 11
(10.96)
Pulicken (2007) [28] 84.2 (14.7) 62 93.9 (13.1) 62 102.7 47
(11.5)
Pueyo (2008) [29] 84.46 25 (eyes) 94.2 75 (eyes) 104.97 25
(eyes)
Zaveri (2008) [30]** 81.8 (19.3) 41 95.6 (15.0) 48 104.6 43
(10.3)
Klistorner (2008) [31] 84.5 (15.1) 32 103.8 (10.8) 32 104.0 25
(9.2)
Siger (2008) 83.9 (17.6) 20 91.08 (19.3) 20 100.3 12
[32] (12.1)
Frohman (2009) 70.3 (13.4) 12 101.8 (6.0) 12 101.9 8
[33] (8.9)
Ratchford (2009) 88.3 (2009) 155 – – 102.4 77
[34]
Burkholder (2009) ** 85.7 (19.0) 164 (328 95.6 (14.5) 366 (730 104.5 111
[35] eyes) eyes) (10.7)
Bock (2010) 86.2 (16.2) 73 (eyes) 97.0 (13.1) 189 (eyes) 105.2 203
{Bock:2010 da} (9.4)
Garcia-Martin et al. 81.2 34 94.9 34 – –
(2011) [36]
Syc (2012) 78.7 (11.7) 73 (eyes) 84.9 (12.2) 123 (eyes) 93.4 100
[37] (10.4)
Costello et al. (2012) 90.8 (16.9) 105 103.4 (14.0) 105 (F) – –
[38] Females (Females) Females 39 (M)
73.5 (14.0) 39 (Males) 100.1 (13.8)
Males Males
Monteiro et al. (2012) 94.6 (19.0) 45 (eyes) 101.2 (11.6) 74 (eyes) 107.6 82
[39] (9.4) (eyes)

* Direct comparison between patients with one eye affected not published
** Some patients had bilateral optic neuritis; fellow eyes have no clinical history of optic neuritis but the whole cohort had confirmed
multiple sclerosis.

of patients with multiple sclerosis, as did Ratchford We have observed small degrees of whole macu-
and colleagues [34]. Macular volume encompasses the lar swelling in patients in the early phases of acute
whole retina from the vitreoretinal interface to the idiopathic optic neuritis, to a lesser extent than is
retinal pigment epithelium, and a reduction in visible in the RNFL, and it is usually only notable
32 volume likely reflects loss of other components of when the affected eye is compared to the unaffected
the retina rather than simply the RNFL. fellow eye.
 Chapter 4: Optical coherence tomography in acute optic neuritis

Gelfland and colleagues [40] reported the appear- studied a group of patients with acute optic neuritis
ance of macular microcysts in the retina of patients and found that GCL/IPL thinning was evident at three
with multiple sclerosis. Of the 318 patients studied, months after the onset of symptoms. There was no
15 (3.8%) had small lucencies in the deep layers of swelling apparent in these layers of the retina, raising
the retina. The microcysts, which they termed micro- the possibility that atrophy of these layers of the retina
cystic macular edema, occurred principally in the might be detectable earlier than in the RNFL, where
inner nuclear layer of the retina and were more edema confounds early detection of tissue atrophy.
common in eyes with a past history of optic neuritis,
and were associated with greater thinning of the The time course of changes to RNFL
RNFL, and with a lower visual acuity. Microcystic
macular edema was also associated with greater dis-
and macula
ability and more rapid progression. Saidha and col- In the early phases of optic neuritis there may be
leagues [41] reported cystic changes in the macula of swelling of the RNFL [31,49,50] (see Figure 4.3).
some of their patients (10/164, 6% of patients had Although RNFL swelling is not evident in all cases of
microcysts evident on OCT). These changes were optic neuritis, RNFL measurement by OCT is more
also associated with worse visual function and higher sensitive for the detection of RNFL swelling than
MS severity scores (although not higher disability as fundoscopy, and is in keeping with the optic nerve
measured by the Expanded Disability Status Scale; swelling observed on MRI [51]. Studies using scan-
EDSS). The presence of NMO is associated with ning laser polarimetry, which utilizes the birefringent
worse vision, and in multiple sclerosis, with markers properties of the RNFL to calculate its thickness,
of increased disease activity such as increased con- suggest that there is axonal loss evident in the RNFL
trast enhancing and T2 lesion formation. It is not in the early stages of acute optic neuritis [52], which is
clear what microcystic macular edema represents not detectable using conventional OCT because
pathologically or pathophysiologically, as it is seen swelling of the axons of the RNFL obscures early
in other inflammatory optic neuropathies [42,43] axonal loss.
and non-inflammatory optic neuropathies [44]. It The swelling is most commonly followed by loss of
may be that it represents only a marker of severity, RNFL thickness (Figure 4.2). The majority of RNFL
perhaps related to trans-synaptic degenerative thinning occurs in the first three months after symp-
mechanisms [45]. Kaushik and colleagues [46] tom onset, and is near complete by six months
found inner nuclear layer thickening and retinal [22,26]. Using modeled data from a cohort of patients
ganglion cell layer thinning to be inversely propor- with acute optic neuritis, we would predict that
tional, and proposed that inner nuclear layer thick- approximately half the RNFL loss will occur within
ening was on a spectrum which included microcysts one month of symptom onset and 99% of the loss will
of the inner nuclear layer at the more severe end. have occurred by 4.75 months [50].
Pathologic studies of the retina in patients with Atrophy of the RNFL is detectable only after par-
multiple sclerosis demonstrate that, in addition to tial resolution of early edema and is likely to be
atrophy of the RNFL, there is thinning of the ganglion detectable 1.5 to 2 months after the onset of symp-
cell layer and inner nuclear layers [47, 48]. Early forms toms [50] (see Figure 4.4). The time taken for loss of
of OCT lacked sufficient tissue penetration properties macular volume (our modeling suggests that 99% of
to quantify these changes, whereas spectral-domain the loss takes place by approximately 11 months)
OCT allows examination of the outer retinal layers seems to be longer than that of RNFL thickness,
such as the ganglion cell layer and the inner nuclear suggesting that perhaps thinning of deeper layers,
layer. Syc and colleagues [37] examined patients with such as thinning of the ganglion cell layer, or different
multiple sclerosis, patients with neuromyelitis optica, mechanisms, such as trans-synaptic retinal
and healthy controls, and they found reductions in a degeneration, are at play in the macula.
combined measure of the ganglion cell layer and inner
plexiform layer (GCL/IPL) in eyes with a history of Predictors of severity of RNFL loss
optic neuritis in their cohort of patients with multiple The eventual degree of axonal loss in the RNFL is
sclerosis when compared to control eyes and to eyes variable after a single attack of optic neuritis, ranging
without a history of optic neuritis. They also serially from little or minor RNFL loss to significant loss 33
 Chapter 4: Optical coherence tomography in acute optic neuritis

(A) (B)

(C) (D)

Figure 4.3 Asymmetric swelling of the RNFL in a patient with chronic relapsing inflammatory optic neuropathy (CRION[1]). The images were
taken ten days after the onset of bilateral eye pain, pain on eye movement, and development of bilateral reduced visual acuity and visual field
sensitivity. There is significant swelling of both RNFL, which is more marked on the left (average circumpapillary RNFL thickness 146 µm right and
223 µm left). A. and C. are circumpapillary OCT images of the retina of the left and right eye, respectively; the highly reflective RNFL at the top of the
retinal image is markedly swollen. B. and D. are peripapillary thickness maps of the RNFL, again demonstrating marked swelling of the RNFL.

Figure 4.4 Model predicting the time course of

Diseased minus healthy RNFL thickness

150 RNFL thinning following a single episode of optic

neuritis. Red lines indicate individual patient
courses; the overall model is indicated in black; and
100 the dashed line represents the time at which atro-
phy would be detectable using the fellow eye as
comparison.
50

–50

–100
0 1 2 3 6 12 18
months from onset

with associated poor visual recovery. In our cohort color vision, the Farnsworth-Munsell 100 Hue
of patients, we found baseline visual evoked test, best predicted eventual RNFL loss [53],
potential (VEP) latency and amplitude, and baseline suggesting that initial axonal dysfunction is asso-
measures of visual function (full- and low-contrast ciated with later axonal death. Somewhat unex-
visual acuity, and whole field mean deviation) were pectedly, the degree of initial RNFL edema is
34 associated with later RNFL loss. Of these not strongly associated with the eventual RNFL
measures, full-field VEP latency and a measure of atrophy.
 Chapter 4: Optical coherence tomography in acute optic neuritis

(A)

µm OD OS

200

100

0
0 30 60 90 120 150 180 210 240
TEMP SUP NAS INF TEMP

(B) (C)

(D) (E)

Figure 4.5 Left eye visual field (C.right eye not shown as NPL), OCT images of the circumpapillary retina (A. and D.) and RNFL thickness
maps (B. and D.) from a 45-year-old man with neuromyelitis optica after multiple episodes of bilateral optic neuritis. A. shows
circumpapillary RNFL thickness plotted against a normative database (95%, 5%, and 1% of normal eyes lie within the green, yellow, and red
range as indicated).
The right eye had no perception of light and an unreactive pupil, and the corrected visual acuity in the left eye was 6/18.

Costello and others [38] studied a group of patients Changes in the retina in neuromyelitis
with optic neuritis and found there were sex-specific
differences in the degree of RNFL loss after optic neur- optica
itis. Male subjects experienced greater RNFL loss in Neuromyelitis optica (NMO) is an inflammatory
their affected eye. The authors suggested that the demyelinating disease in which the pathology is
mechanism underlying this difference in outcome most often concentrated in the optic nerves and spinal
between genders might be related to the differences cord. It is pathologically distinct from multiple sclero-
that underlie poorer outcome (in terms of destructive sis [56], with greater loss of both oligodendrocytes
lesion load and overall disability) for men when and neurons, and characteristic abnormalities of
astrocytes [57]. In the majority of patients, it is
35
compared with women with multiple sclerosis [54,55].
 Chapter 4: Optical coherence tomography in acute optic neuritis

associated with the presence of antibodies to aqua- Optic neuritis as a model of relapse
porin-4 [58,59]. Neuromyelitis may present with
either optic neuritis (both bilateral and unilateral) or in multiple sclerosis
transverse myelitis, or both [60]. The optic nerve is part of the central nervous system,
The optic neuritis associated with NMO is and, as such, can be considered a white matter tract.
often severe and is associated with poor recovery. Optic neuritis resulting from a demyelinating lesion
Of the 12 patients documented in O’Riordan and of the optic nerve is a common and characteristic
colleagues’ case series [61], 10 had count fingers or initial manifestation of multiple sclerosis. The ante-
worse vision in one or more eyes. Of these 12 rior visual pathway, including the optic nerve and
patients, 10 had a bilateral optic neuropathy due RNFL, can be studied functionally (with quantitative
either to recurrence in the fellow eye or simulta- tests of vision), structurally (with optic nerve and
neous optic neuritis. At the nadir of vision in retinal imaging), and electrophysiologically (by mea-
neuromyelitis associated optic neuritis, visual suring the visual evoked potential). Study of optic
acuity is usually count fingers or worse [60,62], neuritis can give insights into the mechanisms of
and recovery is typically not as complete as in demyelination and remyelination, and loss and
acute idiopathic optic neuritis. Recurrent attacks return of function in the demyelinating lesion in
are associated with further reductions in vision multiple sclerosis [73, 74]. The other significant
[23, 60, 63–65]. advantage in favor of using optic neuritis as a
Ratchford and colleagues [34] studied a small model of acute inflammatory demyelination in
cohort of patients with NMO. When compared proof of principle trials of neuroprotective agents is
with patients with multiple sclerosis, optic neuritis that the events of optic neuritis are complete within
in NMO resulted in more severe RNFL thinning and six months of the onset of symptoms, significantly
macular volume loss, and there was more visual loss curtailing the length of time taken to conduct phase
than in multiple sclerosis-associated optic neuritis. II studies.
Using multivariate linear regression, they estimated For the purposes of clinical trials, sample size esti-
that the initial attack of optic neuritis caused, on mates have been generated for both RNFL thickness,
average, a 31µm reduction in the nerve fiber layer and also for a composite of the ganglion cell layer and
thickness, and each further attack resulted in inner plexiform layer (GCL\IPL). For a six-month trial
approximately 10 µm of thinning. with 80% power, aiming to detect a 50% treatment
Fernandes and colleagues [66] studied a group of effect, the per-arm sample size using RNFL as an end-
patients with optic neuritis due to NMO, and com- point would be 36 [50]. This assumes that patients with
pared them to a group of patients with optic neuritis no prior history of optic neuritis are recruited, so the
due to multiple sclerosis and healthy controls. unaffected fellow eye can be used as a comparator to
Although RNFL and ganglion cell layer atrophy assess visual loss. If the fellow eye is previously affected,
were evident in both multiple sclerosis and in or is not used as a comparator, then per-arm sample
NMO, only patients with neuromyelitis optica had size rises to approximately 58 if RNFL loss is used as an
swelling of the inner nuclear layer. Sotirchos and endpoint. For a similar trial using GCL\IPL as an end-
colleagues [67] examined 39 patients with NMO point sample size would be 44 per-arm [37]. Because
and a group of healthy controls. They found micro- there does not appear to be any swelling of the
cystic macular edema was present in patients with a GCL\IPL complex in optic neuritis, sample sizes are
history of acute optic neuritis and was associated not dependent on using the fellow eye as a comparator.
with more severe thinning of the RNFL and Given that RNFL and GCL\IPL complex measures can
ganglion cell layer, and with poorer vision. Both be acquired on Fourier-domain OCT, it is likely that
Sotirchos and colleagues and Gelfand and colleagues both measures will be used to evaluate the efficacy of
found microcystic macular edema to be more putative treatments, although some devices do not
prevalent in their cohorts of patients with neuro- possess the segmentation algorithm to do so. The uti-
myelitis optica (26% and 20%, respectively) than in lity of a treatment that preserved retinal ganglion cells
previously reported cohorts of patients with multiple but not their axons is questionable.
sclerosis (where it is seen in approximately 5% Sühs and colleagues assessed the efficacy of ery-
36 [40,41]). thropoietin as a neuroprotective agent in acute optic
 Chapter 4: Optical coherence tomography in acute optic neuritis

Table 4.2 Studies of optic neuritis associated with neuromyelitis optica, and comparative studies with multiple sclerosis-associated optic
neuritis

NMO eyes
without optic Healthy
NMO neuritis Multiple sclerosis control
Mean n Mean Mean n Mean n
(SD) (SD) n (SD) (SD)
de Seze et al. 77.9 (22.6) 32 – – – – 102.3 7.4
(2008) [68]
Merle et al. 65.4 (25.2) 15 – – 83.9 (24.1) 15 106.2 23
(2008) [69] (24.5)
Naismith et al. 70.5 (26.2) 22 – – 80.7 (19.3) 47 – –
(2009) [70]
Ratchford et al. 63.6 (20.3) 26 – – 88.3 (16.5) (past 378 102.4 77
(2009) [34] (NMO) ON) (11.0)
96.3 (14.3 96.3 (14.3) (No
(LETM) prior ON)
Green and 59.2 (16.2) 16 – – 82.0 (17.6) 20 – –
Cree (2009)
[71]
Nakamura et al. 63.8 (23.5) 18 106.4 8 84.3 (14.2) 14 – –
(2010) [64] (14.5) (eyes)
Monteiro et al. 82.7 (22.4) 33 97.9 94.6 (19.0) (past 60 107.6 41
(2012) [39] (NMO) (10.7) ON) (9.4)
97.9 (10.7) 101.2 (11.6) (No
(LETM) prior ON)
Bouyon et al. 87.4 (23.3) 30 – – – – – –
(2013) [72] (baseline)
79.7 (22.4)
(follow up)
Lange et al. 63.7 (18.3) 26 97.0 18 73.9 (15.2) (prior 13 98.4 100
(2013) [23] (one episode) 6 (11.1) ON) 37 (14.4)
50.7 (8.1) (>1 93.2 (15.2) (no
episode) prior ON)
Bichuetti et al. 65.3 (18.3) 8 85.7 3 92.4 (14.0) (no 63 – –
(2013) [65] one episode 7 (20.5) prior ON) 28
47.0 (6.1) two 80.1 (14.8) (one 4
episodes episode ON) 3
61.7 (7.1) (two
episodes ON)
81.3(17.5)(three
episodes ON)

neuritis, using RNFL thickness measured with OCT as patients in total, and the primary outcome was mea-
the primary endpoint. They found that it was effective sured at 16 weeks after randomization. Esfahani and
in reducing the RNFL atrophy following optic neur- colleagues [75] measured the effect of memantine
itis: in the treated group the median reduction in given for two weeks after a first-ever attack of optic
RNFL thickness was 7.5 µm whereas in the untreated neuritis. Sixty patients were randomized 1:1 to either
group the median reduction in affected eye RNFL memantine (given as 5 mg daily for one week followed
37
thickness was 16.0 µm. The trial required only 40 by 10 mg daily for two weeks) or a placebo. There was
 Chapter 4: Optical coherence tomography in acute optic neuritis

a significant difference in eventual RNFL thickness in relatively small numbers and are quick to reach the
the memantine treated group (91.3 µm vs. 78.9 µm in final time point.
the placebo group) but no significant effect upon
vision or neurophysiological measures of optic nerve References
function. Tsakiri and colleagues [76] evaluated the 1. Kidd D, Burton B, Plant GT, Graham EM. Chronic
effect of simvistatin (80 mg daily for six months) relapsing inflammatory optic neuropathy (CRION).
upon the visual and neurophysiological measures of Brain. February 2003; 126(Pt 2): 276–84.
optic nerve function, and they found a beneficial effect 2. Petzold A, Pittock S, Lennon V, Maggiore C,
on visual evoked potential latency and amplitude, but Weinshenker BG, Plant GT. Neuromyelitis optica-IgG
it did not quite (p = 0.06) reach the primary outcome (aquaporin-4) autoantibodies in immune mediated
optic neuritis. J Neurol Neurosurg Psychiatr. January
of improving contrast sensitivity using Arden plates. 2010; 81(1): 109–11.
Several other trials using acute optic neuritis as a
3. Rodriguez M, Siva A, Cross SA, O’Brien PC, Kurland
model of relapse and OCT as a primary or secondary LT. Optic neuritis: a population-based study in Olmsted
endpoint are in the recruitment phase as this chapter is County, Minnesota. Neurology [Internet]. February
being written. The RENEW study (ClinicalTrials.gov 1995; 45(2): 244–50. Available from: http://www.neurol
Identifier: NCT01721161) is a trial of an anti-LINGO ogy.org/content/45/2/244.short?sid=5d3bc217-394e-46
compound, BIIB-033. LINGO is an inhibitor of remye- d0-a720-df8704457287//www.neurology.org
lination [77], and inhibition of LINGO promotes 4. Beck RW, Cleary PA, Anderson MM, Keltner JL,
remyelination in animal models [78]. The primary end- Shults WT, Kaufman DI, et al. A randomized, con-
point is the change in optic nerve conduction, but trolled trial of corticosteroids in the treatment of acute
RNFL thickness measured by OCT is a secondary optic neuritis. The Optic Neuritis Study Group. N Engl J
Med. February 27, 1992; 326(9): 581–8.
outcome measure. The Amiloride Clinical Trial In
Optic Neuritis (ACTION) study (ClinicalTrials.gov 5. Optic Neuritis Study Group. The clinical profile of
optic neuritis. Experience of the Optic Neuritis
Identifier: NCT01802489) is a placebo-controlled trial Treatment Trial. Arch Ophthalmol. December 1991;
of amiloride in acute optic neuritis. Amiloride blocks 109(12): 1673–8.
the acid-sensing ion channel and inhibits the influx of 6. Trobe JD, Beck RW, Moke PS, Cleary PA. Contrast
calcium, which can be associated with neuronal death sensitivity and other vision tests in the optic neuritis
[79]. The primary outcome measure is measurement of treatment trial. Am J Ophthalmol. May 1996; 121(5):
the RNFL thickness with scanning laser polarimetry; 547–53.
however, measurement of RNFL thickness by optical 7. Keltner JL, Johnson CA, Spurr JO, Beck RW. Baseline
coherence tomography is a secondary outcome mea- visual field profile of optic neuritis. The experience of
sure. There is also a phase II, double-blind, randomized the optic neuritis treatment trial. Optic Neuritis
controlled trial of phenytoin in acute optic neuritis Study Group. Arch Ophthalmol. February 1993; 111(2):
(ClinicalTrials.gov Identifier: NCT01451593), which 231–4.
is using mean RNFL thickness as the primary outcome 8. Cox TA, Thompson HS, Corbett JJ. Relative afferent
measure. Phenytoin, which blocks voltage-gated pupillary defects in optic neuritis. Am J Ophthalmol.
November 1, 1981; 92(5): 685–90.
sodium channels, reduces the severity of, and preserves
axonal integrity in, experimental allergic encephalo- 9. Beck RW, Cleary PA. Optic neuritis treatment trial.
One-year follow-up results. Arch Ophthalmol. June
myelitis [80]. This study aims to recruit 90 patients 1993; 111(6): 773–5.
and has a primary outcome measure that is measured
10. la Cruz de J, Kupersmith MJ. Clinical profile
at six months. The Fingolimod in Acute Demyelinating of simultaneous bilateral optic neuritis in adults.
Optic Neuritis trial (ClinicalTrials.gov Identifier: British Journal of Ophthalmology. May 2006; 90(5):
NCT01757691) is assessing the neuroprotective effect 551–4.
of fingolimod in acute optic neuritis as measured using 11. Beck RW, Kupersmith MJ, Cleary PA, Katz B. Fellow
OCT of the retinal nerve fiber layer. eye abnormalities in acute unilateral optic neuritis.
Optical coherence tomography has given consid- Experience of the optic neuritis treatment
erable insight into the pathophysiology and clinical trial. Ophthalmology. May 1993; 100(5): 691–8.
characteristics of acute optic neuritis. It also allows 12. Ormerod IE, McDonald WI. Multiple sclerosis pre-
acute optic neuritis to be used as a model of the acute senting with progressive visual failure. J Neurol
38 relapse in trials of neuroprotection, which require Neurosurg Psychiatr. September 1, 1984; 47(9): 943–6.
 Chapter 4: Optical coherence tomography in acute optic neuritis

13. Ashworth B. Chronic retrobulbar and chiasmal neur- Sclerosis. Ophthalmology. February 2006; 113(2):
itis. British Journal of Ophthalmology. October 1967; 324–32.
51(10): 698–702. 26. Costello F, Coupland S, Hodge W, Lorello GR,
14. Hoyt WF, Schlicke B, Eckelhoff RJ. Fundoscopic Koroluk J, Pan YI, et al. Quantifying axonal loss after
appearance of a nerve-fibre-bundle defect. British optic neuritis with optical coherence tomography.
Journal of Ophthalmology. August 1972; 56(8): 577–83. Annals of Neurology. June 1, 2006; 59(6): 963–9.
15. Frisén L, Hoyt WF. Insidious atrophy of retinal nerve 27. Albrecht P, Fröhlich R, Hartung H-P, Kieseier BC,
fibers in multiple sclerosis. Funduscopic identification Methner A. Optical coherence tomography measures
in patients with and without visual complaints. Arch axonal loss in multiple sclerosis independently of
Ophthalmol. August 1, 1974; 92(2): 91–7. optic neuritis. J Neurol. November 9, 2007; 254(11):
16. Sharpe JA, Sanders MD. Atrophy of myelinated nerve 1595–6.
fibres in the retina in optic neuritis. British Journal of 28. Pulicken M, Gordon-Lipkin E, Balcer LJ, Frohman E,
Ophthalmology. Aprl 1, 1975; 59(4): 229–32. Cutter G, Calabresi PA. Optical coherence tomogra-
17. Elbøl P, Work K. Retinal nerve fiber layer in multiple phy and disease subtype in multiple sclerosis.
sclerosis. Acta Ophthalmology. August 1, 1990; 68(4): Neurology. November 27, 2007; 69(22): 2085–92.
481–6. 29. Pueyo V, Martin J, Fernandez J, Almarcegui C, Ara J,
18. MacFadyen DJ, Drance SM, Douglas GR, Airaksinen Egea C, et al. Axonal loss in the retinal nerve fiber layer
PJ, Mawson DK, Paty DW. The retinal nerve fiber in patients with multiple sclerosis. Multiple Sclerosis.
layer, neuroretinal rim area, and visual evoked poten- June 2008; 14(5): 609–14.
tials in MS. Neurology [Internet]. September 1, 1988; 30. Zaveri MS, Conger A, Salter A, Frohman TC, Galetta
38(9): 1353–8. Available from: http://neurology.org/c SL, Markowitz CE, et al. Retinal imaging by laser
ontent/38/9/1353.short polarimetry and optical coherence tomography evi-
19. Quigley HA, Addicks EM. Quantitative studies of ret- dence of axonal degeneration in multiple sclerosis.
inal nerve fiber layer defects. Arch Ophthalmology. May Arch Neurol. July 2008; 65(7): 924–8.
1982; 100(5): 807–14. 31. Klistorner A, Arvind H, Nguyen T, Garrick R, Paine M,
20. Parisi V, Manni G, Spadaro M, Colacino G, Restuccia Graham S, et al. Axonal loss and myelin in early ON
R, Marchi S, et al. Correlation between morphological loss in postacute optic neuritis. Ann Neurol. September
and functional retinal impairment in multiple sclerosis 2008; 64(3): 325–31.
patients. Invest Ophthalmol Vis Sci. October 1999; 32. Siger M, Dziegielewski K, Jasek L, Bieniek M, Nicpan
40(11): 2520–7. A, Nawrocki J, et al. Optical coherence tomography in
21. Trip SA, Schlottmann PG, Jones SJ, Altmann DR, multiple sclerosis: thickness of the retinal nerve fiber
Garway-Heath DF, Thompson AJ, et al. Retinal nerve layer as a potential measure of axonal loss and brain
fiber layer axonal loss and visual dysfunction in optic atrophy. J Neurol. October 2008; 255(10): 1555–60.
neuritis. Ann Neurol. September 2005; 58(3): 383–91. 33. Frohman EM, Dwyer MG, Frohman T, Cox JL, Salter
22. Costello F, Hodge W, Pan Y, Eggenberger E, Coupland A, Greenberg BM, et al. Relationship of optic nerve and
S, Kardon R. Tracking retinal nerve fiber layer loss brain conventional and non-conventional MRI mea-
after optic neuritis: a prospective study using optical sures and retinal nerve fiber layer thickness, as assessed
coherence tomography. Multiple Sclerosis. June 23, by OCT and GDx: a pilot study. J Neurol Sci. July 15,
2008; 14(7): 893–905. 2009; 282(1–2): 96–105.
23. Lange AP, Sadjadi R, Zhu F, Alkabie S, Costello F, 34. Ratchford JN, Quigg ME, Conger A, Frohman T,
Traboulsee AL. Spectral-domain optical coherence Frohman E, Balcer LJ, et al. Optical coherence tomo-
tomography of retinal nerve fiber layer thickness in graphy helps differentiate neuromyelitis optica and
NMO patients. J Neuroophthalmol MS optic neuropathies. Neurology. July 28,
[Internet]. September 2013; 33(3): 213–9. 2009; 73(4): 302–8.
Available from: http://eutils.ncbi.nlm.nih.gov/entrez/ 35. Burkholder BM, Osborne B, Loguidice MJ, Bisker E,
eutils/elink.fcgi?dbfrom=pubmed&id=23863782&ret Frohman TC, Conger A, et al. Macular volume deter-
mode=ref&cmd= prlinks mined by optical coherence tomography as a
24. Bichuetti DB, Oliveira EML, Souza NA, Rivero RLM, measure of neuronal loss in multiple sclerosis. Arch
Gabbai AA. Neuromyelitis optica in Brazil: a study on Neurol. November 2009; 66(11): 1366–72.
clinical and prognostic factors. Multiple Sclerosis. May 36. Garcia-Martin E, Pueyo V, Ara J, Almarcegui C,
2009; 15(5): 613–9. Martin J, Pablo L, et al. Effect of optic neuritis on
25. Fisher J, Jacobs D, Markowitz C, Galetta S, Volpe N, progressive axonal damage in multiple
Nanoschiavi M, et al. Relation of Visual Function to sclerosis patients. Mult Scler [Internet]. June 20, 2011
Retinal Nerve Fiber Layer Thickness in Multiple Jun 20; 17(7): 830–7. Available from: http://eutils.ncbi. 39
 Chapter 4: Optical coherence tomography in acute optic neuritis

nlm.nih.gov/entrez/eutils/elink.fcgi?dbfrom=pubme and inflammation irrespective of disease duration.

d&id=21300734&retmode=ref&cmd=prlinks Brain. June 2010; 133(Pt 6): 1591–601.
37. Syc SB, Saidha S, Newsome SD, Ratchford JN, Levy M, 49. Pro MJ, Pons ME, Liebmann JM, Ritch R, Zafar S,
Ford E, et al. Optical coherence tomography Lefton D, et al. Imaging of the optic disc and retinal
segmentation reveals ganglion cell layer nerve fiber layer in acute optic neuritis. J Neurol Sci.
pathology after optic neuritis. Brain. February 2012; December 1, 2006; 250(1–2): 114–9.
135(Pt 2): 521–33. 50. Henderson APD, Altmann DR, Trip AS, Kallis C,
38. Costello F, Hodge W, Pan YI, Burton JM, Freedman Jones SJ, Schlottmann PG, et al. A serial study of retinal
MS, Stys PK, et al. Sex-specific differences in retinal changes following optic neuritis with sample size esti-
nerve fiber layer thinning after acute optic neuritis. mates for acute neuroprotection trials. Brain.
Neurology. October 2012; 79(18): 1866–72. September 2010; 133(9): 2592–602.
39. Monteiro MLR, Fernandes DB, Apóstolos-Pereira SL, 51. Hickman SJ, Toosy AT, Jones SJ, Altmann DR,
Callegaro D. Quantification of retinal neural loss in Miszkiel KA, MacManus DG, et al. A serial MRI study
patients with neuromyelitis optica and multiple following optic nerve mean area in acute optic neuritis.
sclerosis with or without optic neuritis using Fourier- Brain. November 2004; 127(Pt 11): 2498–505.
domain optical coherence tomography. Invest 52. Garas A, Simó M, Holló G. Nerve fiber layer and
Ophthalmol Vis Sci [Internet]. June 2012; 53(7): macular thinning measured with different imaging
3959–66. Available from: http://eutils.ncbi.nlm.nih.go methods during the course of acute optic neuritis.
v/entrez/eutils/elink.fcgi?dbfrom=pubmed& European Journal of Ophthalmology. October 22, 2010.
id=22589446&retmode=ref&cmd=prlinks
53. Henderson APD, Altmann DR, Trip SA, Miszkiel KA,
40. Gelfand JM, Nolan R, Schwartz DM, Graves J, Green Schlottmann PG, Jones SJ, et al. Early factors asso-
AJ. Microcystic macular edema in multiple sclerosis ciated with axonal loss after optic neuritis. Ann Neurol.
is associated with disease severity. Brain. June 2012; December 21, 2011; 70(6): 955–63.
135(Pt 6): 1786–93.
54. Khaleeli Z, Ciccarelli O, Manfredonia F, Barkhof F,
41. Saidha S, Sotirchos ES, Ibrahim MA, Crainiceanu CM, Brochet B, Cercignani M, et al. Predicting progression
Gelfand JM, Sepah YJ, et al. Microcystic macular in primary progressive multiple sclerosis: a 10-year
edema, thickness of the inner nuclear layer of the multicenter study. Ann Neurol. June 2008; 63(6): 790–3.
retina, and disease characteristics in multiple sclerosis:
55. Koch M, Kingwell E, Rieckmann P, Tremlett H, UBC
a retrospective study. Lancet Neurol. October 4, 2012;
MS Clinic Neurologists. The natural history of sec-
11(11): 963–72.
ondary progressive multiple sclerosis. J Neurol
42. Balk LJ, Killestein J, Polman CH, Uitdehaag BMJ, Neurosurg Psychiatr. September 2010; 81(9): 1039–43.
Petzold A. Microcystic macular edema confirmed, but
not specific for multiple sclerosis. Brain. December 18, 56. Lucchinetti CF, Mandler RN, McGavern D, Brück W,
Gleich G, Ransohoff RM, et al. A role for humoral
2012; 135(12): e226–6.
mechanisms in the pathogenesis of Devic’s neuro-
43. Kaufhold F, Zimmermann H, Schneider E, Ruprecht myelitis optica. Brain. July 2002; 125(Pt 7): 1450–61.
K, Paul F, Oberwahrenbrock T, et al. Optic neuritis is
associated with inner nuclear layer thickening and 57. Parratt JD, Prineas JW. Neuromyelitis optica: a
demyelinating disease characterized by acute destruc-
microcystic macular edema independently of multiple
sclerosis. PLoS ONE. 2013; 8(8): e71145. tion and regeneration of perivascular astrocytes. Mult
Scler. September 30, 2010; 16(10): 1156–72.
44. Abegg M, Zinkernagel M, Wolf S. Microcystic macular
58. Lennon VA, Wingerchuk DM, Kryzer TJ, Pittock SJ,
degeneration from optic neuropathy. Brain. December
Lucchinetti CF, Fujihara K, et al. A serum autoanti-
18, 2012; 135(12): 225–5.
body marker of neuromyelitis optica: distinction from
45. van Buren JM. Trans-synaptic retrograde degeneration multiple sclerosis. Lancet. November 2004; 364(9451):
in the visual system of primates. J Neurol Neurosurg 2106–12.
Psychiatr. October 1963; 26: 402–9.
59. Lennon VA, Kryzer TJ, Pittock SJ, Verkman AS,
46. Kaushik M, Wang CY, Barnett MH, Garrick R, Parratt Hinson SR. IgG marker of optic-spinal multiple
J, Graham SL, et al. Inner nuclear layer thickening is sclerosis binds to the aquaporin-4 water channel. J Exp
inversley proportional to retinal ganglion cell loss in Med. August 15, 2005; 202(4): 473–7.
optic neuritis. PLoS ONE. 2013; 8(10): e78341.
60. Wingerchuk DM, Hogancamp WF, O’Brien PC,
47. Kerrison JB, Flynn T, Green WR. Retinal pathologic Weinshenker BG. The clinical course of neuromyelitis
changes in multiple sclerosis. Retina (Philadelphia, Pa). optica (Devic’s syndrome). Neurology. 1999; 53(5):
1994; 14(5): 445–51. 1107–14.
48. Green AJ, McQuaid S, Hauser SL, Allen IV, Lyness R. 61. O’Riordan JI, Gallagher HL, Thompson AJ, Howard
40 Ocular pathology in multiple sclerosis: retinal atrophy RS, Kingsley DP, Thompson EJ, et al. Clinical, CSF,
 Chapter 4: Optical coherence tomography in acute optic neuritis

and MRI findings in Devic’s neuromyelitis optica. J 70. Naismith RT, Tutlam NT, Xu J, Klawiter EC, Shepherd
Neurol Neurosurg Psychiatr. April 1996; 60(4): 382–7. J, Trinkaus K, et al. Optical coherence tomography
62. Papais-Alvarenga RM, Carellos SC, Alvarenga MP, differs in neuromyelitis optica compared with multiple
Holander C, Bichara RP, Thuler LCS. Clinical sclerosis. Neurology. March 24, 2009; 72(12): 1077–82.
course of optic neuritis in patients with relapsing 71. Green AJ, Cree BAC. Distinctive retinal nerve fibre
neuromyelitis optica. Arch Ophthalmol. January 2008; layer and vascular changes in neuromyelitis optica
126(1): 12–6. following optic neuritis. J Neurol Neurosurg Psychiatr.
63. Rivera JF, Kurtzke JF, Booth VJA, Corona T V. September 2009; 80(9): 1002–5.
Characteristics of Devic’s disease (neuromyelitis 72. Bouyon M, Collongues N, Zéphir H, Ballonzoli L,
optica) in Mexico. J Neurol. May 2008; 255(5): 710–5. Jeanjean L, Lebrun C, et al. Longitudinal follow-up of
64. Nakamura M, Nakazawa T, Doi H, Hariya T, vision in a neuromyelitis optica cohort. Mult Scler
Omodaka K, Misu T, et al. Early high-dose intravenous [Internet]. August 30, 2013; 19(10): 1320–2. Available
methylprednisolone is effective in preserving retinal from: http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elin
nerve fiber layer thickness in patients with neuromye- k.fcgi?dbfrom=pubmed&id=23413296&retmode=r
litis optica. Graefes Arch Clin Exp Ophthalmol. ef&cmd=prlinks
December 2010; 248(12): 1777–85. 73. Compston A. Mechanisms of axon-glial injury of the
65. Bichuetti DB, de Camargo AS, Falcão AB, Gonçalves optic nerve. Eye. November 2004; 18(11): 1182–7.
FF, Tavares IM, de Oliveira EML. The retinal nerve 74. Kolappan M, Henderson APD, Jenkins TM, Wheeler-
fiber layer of patients with neuromyelitis optica and Kingshott CAM, Plant GT, Thompson AJ, et al.
chronic relapsing optic neuritis is more severely Assessing structure and function of the afferent visual
damaged than patients with multiple sclerosis. J pathway in multiple sclerosis and associated optic
Neuroophthalmol [Internet]. September 2013; 33(3): neuritis. J Neurol. March 2009; 256(3): 305–19.
220–4. Available from: http://eutils.ncbi.nlm.nih.gov/ 75. Esfahani MR, Harandi ZA, Movasat M, Nikdel M,
entrez/eutils/elink.fcgi?dbfrom=pubmed&i Adelpour M, Momeni A, et al. Memantine for axonal
d=23917443&retmode=ref&cmd=prlinks loss of optic neuritis. Graefes Arch Clin Exp
66. Fernandes DB, Raza AS, Nogueira RGF, Wang D, Ophthalmol. June 2012; 250(6): 863–9.
Callegaro D, Hood DC, et al. Evaluation of inner 76. Tsakiri A, Kallenbach K, Fuglø D, Wanscher B,
retinal layers in patients with multiple sclerosis or Larsson H, Frederiksen J. Simvastatin improves final
neuromyelitis optica using optical coherence visual outcome in acute optic neuritis: a randomized
tomography. Ophthalmology. February 2013; 120(2): study. Mult Scler. January 2012; 18(1): 72–81.
387–94.
77. Jepson S, Vought B, Gross CH, Gan L, Austen D,
67. Sotirchos ES, Saidha S, Byraiah G, Mealy MA, Ibrahim Frantz JD, et al. LINGO-1, a transmembrane signaling
MA, Sepah YJ, et al. In vivo identification of morpho- protein, inhibits oligodendrocyte differentiation and
logic retinal abnormalities in neuromyelitis optica. myelination through intercellular self-interactions.
Neurology [Internet]. April 8, 2013; 80(15): 1406–14. J Biol Chem. June 22, 2012; 287(26): 22184–95.
Available from: http://eutils.ncbi.nlm.nih.gov/entrez/
78. Mi S, Pepinsky RB, Cadavid D. Blocking LINGO-1 as a
eutils/elink.fcgi?dbfrom=pubmed&id=23516321&ret
therapy to promote CNS repair: from concept to the
mode=ref&cmd=prlinks
clinic. CNS Drugs. July 2013; 27(7): 493–503.
68. de Seze J, Blanc F, Jeanjean L, Zéphir H, Labauge P,
Bouyon M, et al. Optical coherence tomography in 79. Canessa CM, Horisberger JD, Rossier BC. Epithelial
sodium channel related to proteins involved in neu-
neuromyelitis optica. Arch Neurol. July 2008; 65(7):
rodegeneration. Nature. February 4, 1993; 361(6411):
920–3.
467–70.
69. Merle H, Olindo S, Donnio A, Richer R, Smadja D,
80. Black JA, Waxman SG. Phenytoin protects central
Cabre P. Retinal peripapillary nerve fiber layer thick-
axons in experimental autoimmune encephalomyelitis.
ness in neuromyelitis optica. Invest Ophthalmol Vis Sci.
J Neurol Sci. November 15, 2008; 274(1–2): 57–63.
October 2008; 49(10): 4412–7.

41
 Chapter
Optical coherence tomography and visual

5 outcomes in acute optic neuritis

Fiona Costello

Introduction placebo (for 14 days) in 457 patients with acute ON

[4]. From the ONTT, we learned that most ON patients
Optic neuritis (ON) is generally viewed as an inflam-
are young (mean age: 32 years), Caucasian (85%), and
matory optic nerve injury, which is strongly asso-
predominantly women (77%) [3]. Ninety-two percent of
ciated with multiple sclerosis (MS). Optical
patients report pain at the onset of vision loss, which is
coherence tomography (OCT) provides a reliable
often characterized as an “ache” made worse with eye
means of capturing neuronal and axonal structural
movement [4]. Vision loss is generally subacute, pro-
deficits in the afferent visual pathway (AVP), which
gressing over hours to days. The severity of vision loss in
can be compared to tests of visual function, to devise a
ON may range from mild to no light perception initially.
structural–functional paradigm of brain injury. In
In patients with unilateral or asymmetric optic nerve
this respect, the eye provides a unique view into the
involvement, a relative afferent pupil defect (RAPD) can
effects of central nervous system (CNS) inflammation,
be detected in the affected eye, or, in cases of bilateral
which may enhance our understanding of disease
involvement, in the more severely affected eye. Bilateral
mechanisms that contribute to neurological disability
involvement of the anterior visual system, with the
in MS. This chapter addresses the published experi-
magnitude of pathology being equivocal, can represent
ence with OCT in the diagnosis and management of
the setting where sluggish pupillary responses are
acute ON, with particular emphasis on visual
observed, but without evidence of a RAPD. We have
outcomes in patients presenting with this condition.
yet to elucidate a validated method by which we can
detect “non-relative” APDs; an important goal in study-
Optic neuritis ing the relationship between structural and physiologic
Optic neuritis represents the first clinical manifestation measures of tissue injury and cross-correlation.
of MS in approximately 20% of cases and is, therefore, Dyschromatopsia, or decreased color vision, is
one of the most commonly encountered clinically iso- also common in patients with optic neuritis. This
lated syndromes (CIS) that is a harbinger of the disease finding can be particularly helpful in localizing the
[1, 2]. In addition to its being the clinical inception for diagnosis in patients with mild central vision loss,
many, an additional 30% to 70% of affected individuals who have disproportionate deficits in color vision
develop ON over time [3]. Hence, the anterior visual function. In cases of retrobulbar ON, the fundus
pathway is a frequent target in MS, and much of what examination is initially normal, whereas patients
we have come to understand about ON has been with anterior ON or “papillitis” may manifest optic
derived from the long term follow-up of the Optic disc swelling acutely (in conjunction with a corre-
Neuritis Treatment Trial (ONTT) [4]. This rando- sponding compromise in visual acuity, contrasting
mized, multicenter study was initially designed to com- with the disc swelling associated with true papille-
pare the benefits of treatment with either intravenous dema, where a relative preservation of acuity is main-
methylprednisolone (250 mg every 6 hours for 3 days tained until later in the course, and more severe field
followed by oral prednisone (1 mg/kg/day] for 11 days), suppression) [2]. The initial period of visual recovery
oral prednisone (1 mg/kg/day for 14 days), or oral occurs within weeks, and further improvement

42 Optical Coherence Tomography in Neurological Disease, ed. Peter A. Calabresi, Laura J. Balcer, and Elliot M. Frohman.
Published by Cambridge University Press. © Cambridge University Press 2015.
 Chapter 5: Optical coherence tomography and visual outcomes in acute optic neuritis

Table 5.1 Atypical features of acute optic neuritis Table 5.2 Clinical conditions mimicking optic neuritis

Age greater than 50 Neuromyelitis optica

Optic atrophy at acute presentation Steroid responsive optic neuropathies
Orbital signs (proptosis) (including sarcoidosis)
Severe optic disc edema Optic perineuritis
Prominent vitreous cell Neuroretinitis
Hemorrhage affecting the optic nerve or Compressive lesions (pituitary lesions, meningiomas,
surrounding retina aneurysms)
Bilateral simultaneous loss of vision at onset Infectious optic neuropathies (syphilis, tuberculosis,
Lack of recovery after one month Lyme)
Lack of pain at onset or pain that persists more than Ischemic optic neuropathies
two weeks Leber’s hereditary optic neuropathy
Systemic symptoms and signs (fever, weight loss, rash) Big blind spot syndromes

in vision is seen up to a year after the acute episode [4, need for a clinical model amenable to noninvasive
5]. Atypical features of ON (Table 5.1), such as bilat- imaging that captures different mechanisms of tissue
eral involvement at presentation, optic disc hemor- injury contributing to disability progression. As a
rhages, vitreous reaction, absence of pain, and poor putative model of MS, the AVP model, with ON as
clinical recovery, should prompt investigations for its relapse “prototype” lesion, offers several potential
possible clinical mimics (Table 5.2). advantages:
In the post-acute phase, most ON patients develop 1. Localization – Defining Where the Problem Is: In
optic disc pallor as a “footprint” of the original injury. the setting of acute ON, the AVP model provides
The prognosis for recovery after typical ON is gener- objective evidence of a symptomatic lesion, which
ally good, with approximately 95% of patients achiev- can be precisely localized in the CNS [2]. Given a
ing a visual acuity of 20/40 vision in their affected typical history, the presence of an RAPD in
eye a year after clinical presentation [2, 5, 6, 7]. the affected eye, pain on movement, a visual field
Despite regaining “normal” vision, however, many defect that follows the topography of the retinal
ON patients report persistent problems, including nerve fiber layer (RNFL), color vision loss, and, in
fatigue and heat-induced (Uhthoff’s phenomenon) some cases, optic disc hyperemia in the absence of
vision loss, altered motion and depth perception, hemorrhage or vitreous cell production, is highly
and decreased spatial vision at low contrast levels. suggestive of ON. Because the cardinal clinical
There is, therefore, discordance between what manifestations have been so well established,
patients report versus what is captured with standar- ON is the best characterized CIS affecting MS
dized ophthalmic testing in the setting of post-acute patients [1].
ON, indicating a need for more sensitive measures of 2. Time of Onset – Defining When the Problem Starts:
vision loss in this patient population. As foveating animals, humans are “hard-wired”
to seek high-resolution images in the world
Optic neuritis and the afferent visual around them. In light of the highly specialized
pathway: designing a clinical model nature of the AVP, any perturbation in the
system that interferes with visual perception,
of MS particularly central vision, will generally be
Attitudes continue to evolve from viewing MS as a noticed and reported by affected individuals [2].
demyelinating disease to a broader perspective in Because ON is both clinically relevant and
which the relative contributions of inflammation, prevalent, functional recovery can be monitored
axonal loss, and neurodegeneration are weighed in from a relatively precise point of onset in the
the balance [2]. Yet the pathogenesis of MS, and by AVP model of MS.
extension the factors that underpin disability in the 3. Quantifiable – Defining the Relation between 43
disease, remain unknown. Accordingly, there is a Structure and Function: The AVP is a functionally
 Chapter 5: Optical coherence tomography and visual outcomes in acute optic neuritis

eloquent system and deficits can be captured with equally well to the visual system in which anatomical
highly reproducible measures of visual function integrity and clinical function are tightly linked,
including high– and low-contrast visual acuity, allowing precise topographic localization of patho-
automated perimetry, and color vision testing. logical lesions [2]. Since Hemholtz’s invention of the
Furthermore, OCT provides structural measures ophthalmoscope in 1851, the structural conse-
of neuronal and axonal integrity in the AVP. quences of optic nerve injury have been visualized
By pairing OCT measures with quantitative acutely as optic disc edema, followed by the insidious
visual outcomes, we can devise a structural– evolution of optic disc pallor over weeks to months
functional paradigm to elucidate the temporal after the onset of symptoms, and corresponding
evolution and relative contributions of inflam- defects in the retinal nerve fiber layer (RNFL)
mation, axonal loss, neuronal damage, and [2, 7]. The RNFL has the distinct feature of being
cortical compensation to post-ON recovery in an unmyelinated region of the CNS and represents
the AVP model of MS [2]. an accessible substrate through which we can sample
4. The Back of the Eye Is the Front of the Brain: neuro-axonal integrity in the CNS. In the current
Previous pathological studies have shown that tis- ocular imaging era, measure of changes in the peri-
sue specific injury in the AVP mirrors global CNS papillary RNFL, retinal ganglion cell layer plus inner
effects in MS patients [8]. During an acute ON plexifom layer (GCL+IPL), and deeper neuronal
event, cytokine release induces transient conduc- layer in the retina can be quantified with OCT to
tion block, likely induced by nitric oxide [9]. provide an in vivo “optical biopsy” [3] of the AVP.
When myelination and axonal integrity are intact, From these OCT measures of structural integrity
recovery ensues with the removal of inflammatory we can potentially interpret the temporal evolution
mediators. During recovery from ON, remyelina- and functional impact of axonal and neuronal
tion improves saltatory conduction through damage in the CNS.
sodium channels, which are distributed along the
demyelinated optic nerve segment [9].
In conjunction with the characteristic mechan-
Optical coherence tomography
isms described above, cortical plasticity is also in acute and chronic ON
believed to play a role in optimizing function in the At the time of an acute ON event, when vision loss is
more chronic phases of recovery, albeit the timeline at its nadir, patients often manifest peripapillary
and mechanisms involved therein are not well RNFL measurements that are either comparable
understood. The AVP model can be used to detect to or increased in their affected eye (ON eye) relative
and monitor tissue injury, in association of specific to their fellow eye (non-ON eye) [2, 7].
factors that underpin injury. Further, this very Correspondingly, the optic nerve in the ON eye
same approach may be germane to our ability to may be mildly edematous or hyperemic secondary
identify preventative, protective, and perhaps to axoplasmic flow stasis (Case 1). In contrast, OCT-
even restorative properties of neurotherapeutic measured macular volume (MV) and GCL + IPL
strategies [2]. values are similar between ON eyes and fellow eyes
in patients with acute ON [2, 7, 11]. In the ensuing
The role of optical coherence two to three months, optic disc pallor and RNFL
thinning evolve, with earliest signs of significant
tomography in acute ON: quantifying RNFL atrophy manifesting in the temporal RNFL
the structural consequences of an region [2, 7]. The RNFL, MV, and GCL+ IPL values
inflammatory injury in the afferent continue to decrease for six to 12 months after symp-
tom onset, plateauing thereafter [2, 7,11–13]. From
visual pathway studies to date, the inner and outer nuclear layers are
unaffected in acute ON and do not differ between
“Form and function should be one, joined in a spiritual
union.” [10] Frank Lloyd Wright
ON eyes and fellow eyes of ON patients during long-
itudinal follow-up [11]. Visual recovery 12 months
Frank Lloyd Wright was referring to architecture after ON is not related to the extent of peripapillary
44 with his iconic comment, but this principle applies RNFL swelling seen acutely, but ultimately is
 Chapter 5: Optical coherence tomography and visual outcomes in acute optic neuritis

(a) i (a) ii

(b) i (b) ii

(c) i (c) ii

Figure 5.1

associated with the amount of RNFL, MV, and GCL+ event, and 232 eyes without ON. The mean RNFL
IPL thinning observed 6–12 months following the values were significantly lower in recurrent ON eyes
inception of the ON related clinical syndrome [2, 7, (64 μm) relative to single ON eyes (86 μm) and eyes
11–13]. without ON (100 μm) [14]. Yeh and colleagues [15]
noted that average RNFL thickness decreased with an
Interpreting RNFL changes in increasing number of ON episodes in pediatric
patients with demyelinating syndromes, which
patients with recurrent ON supports the premise that recurrent inflammatory
It can be difficult to detect RNFL changes in the events have a cumulative impact in the AVP.
setting of recurrent ON, because the extent of cumu- This creates a practical challenge to using OCT in
lative RNFL atrophy can be severe and the corre- the diagnosis and management of acute ON in a
sponding visual outcomes dire. In a recent OCT patient with preexisting RNFL/MV/GCL+IPL
study of 193 MS patients, peripapillary RNFL values atrophy, because increments of change in OCT mea-
were compared between 29 eyes affected by two or sures decrease over time and may be subtle in the
more ON events, 125 eyes affected by a single ON context of a new ON event. 45
 Chapter 5: Optical coherence tomography and visual outcomes in acute optic neuritis

Figure 5.1 (cont.)

Comparing RNFL changes in ON associated every 1-μm decrease in peripapillary RNFL thickness
with neuromyelitis optica to RNFL changes [16]. In a related study, there was significantly worse
RNFL thinning in NMO ON eyes (64 μm) relative to
in ON associated with MS RRMS ON eyes (88 μm) and control eyes (102 μm) [17].
Neuromyelitis optica (NMO) is a severe inflammatory Ratchford [17] reported that a first episode of ON was
process of the optic nerves and spinal cord estimated to cause 24 μm more RNFL loss in NMO eyes
(with occasional brain involvement) that is generally relative to RRMS ON eyes. Nakamura [18] described
associated with poor clinical recovery. Not surprisingly, lower RNFL values in NMO ON eyes than in MS ON
RNFL atrophy tends to be extensive in ON eyes of eyes (64 μm versus 84 μm; p = 0.0006) and noted that the
patients with NMO. Several studies have explored the frequency of ON relapses and the time to initiate treat-
role of OCT in distinguishing ON associated with NMO ment with high-dose intravenous methylprednisolone
from that associated with MS. Naismith [16] reported significantly affected the preservation of RNFL thickness
lower RNFL values in ON eyes of 22 NMO patients in NMO patients. Syc [11] evaluated 22 NMO patients
compared to ON eyes of 47 MS patients and noted and reported that GCL+ IPL thickness was decreased in
that the superior and inferior RNFL quadrants were NMO ON eyes compared to NMO non-ON eyes, and
more severely affected in the former. In this study, the that NMO eyes with and without remote ON history
46 odds of falling into the NMO group increased by 8% for had reduced GCL+ PIL thicknesses compared with
 Chapter 5: Optical coherence tomography and visual outcomes in acute optic neuritis

Figure 5.1 (cont.)

those of healthy controls. The importance of diagnosing Optical coherence tomography

NMO early cannot be overstated because the natural and visual outcomes in acute
history of ON associated with this disorder is poor and
long-term immunosuppressive therapy is needed. It is optic neuritis
noteworthy that patients who manifest an inter-eye Optical coherence tomography has been viewed as
asymmetry in RNFL thickness of 15 μm or greater a putative marker of neuronal and axonal integrity in
have been deemed more likely to have the diagnosis of the afferent visual pathway. If axonal and neuronal
NMO (75%) than RRMS (24%) [17]. Yet in studies damage contribute to neurological disability, one
showing significantly more RNFL thinning in NMO would infer that OCT-measured changes in peripa-
ON eyes relative to MS ON eyes, there has been con- pillary RNFL, MV, and GCL + IPL should corre-
siderable overlap in the distribution of RNFL values. spond to visual deficits in patients with acute ON.
Hence, in using OCT to differentiate ON associated In this context, OCT would be a useful ancillary
with NMO from ON associated with MS, it is important test that could be used to construct a structural–
to be mindful of issues that could affect RNFL values in functional paradigm, by which factors influencing
either group, including recurrent ON in the same eye injury and repair might be better understood in
and bilateral optic nerve involvement. MS patients. 47
 Chapter 5: Optical coherence tomography and visual outcomes in acute optic neuritis

Figure 5.1 (cont.)

High-Contrast Visual Acuity: High-Contrast Visual superior to Snellen charts because interpatient differ-
Acuity (HCVA) refers to the spatial resolving ability ences are more accurately measured and longitudinal
of the eye and has long been the mainstay of visual follow-up measurements have had more consistent
standard testing. The most common HCVA charts precision, regardless of whether the patients had
employed in clinical research studies include the mild or severe vision loss. [19]. In the ETDRS charts,
Snellen and Early Treatment Diabetic Retinopathy best corrected visual acuity can be converted to
Study (ETDRS) charts. The Snellen chart has letters logMAR HCVA, which converts the geometric
of different sizes arranged from largest at the top to sequence of a traditional chart to a linear scale [19].
smallest at the bottom, which are read, one eye at a In logMAR notation, lower scores correspond to
time, at a distance of six meters (20 feet). Snellen better vision, and as acuity becomes worse, the value
visual acuity is usually expressed as a fraction with of the logMAR increases.
the numerator equal to the distance from the chart In the setting of acute ON, HCVA impairment is
and the denominator being the size of the smallest line generally maximal at onset, and the deficit in HCVA
that can be read. Because there are numerous limita- at one month can predict long-term recovery.
tions to Snellen visual acuity charts [2, 19], the ETDRS Kupersmith and colleagues [20] used the ONTT data-
method has become the “gold standard” HCVA test base to evaluate various “cut points” for baseline
48 for most current clinical trials. The ETDRS charts are and 1-month vision levels that predicted abnormal
 Chapter 5: Optical coherence tomography and visual outcomes in acute optic neuritis

Figure 5.1 (cont.)

six-month vision. Based on their findings, failure to values correlate with worse visual function [2, 7, 11–
reach a one-month HCVA cut off of 20/50 correlated 13]. In a prior prospective OCT study, measures of
with having moderate-to-severe loss in this domain of RNFL thickness were compared with logMAR HCVA
function after six months [20]. On a practical level, if scores for 54 patients observed a minimum of three
HCVA has not improved by one month, it should also months after acute ON [12]. Regression analyses
prompt the physician to consider alternate diagnoses showed that the predictive value for RNFL thickness
or ON mimics (Table 5.2). was significant only for patients with RNFL measures
At initial presentation, OCT-measured peripapil- less than 70 µm; lower RNFL values correlated with
lary RNFL values may be elevated, reflecting optic disc worse visual acuity outcomes after ON [12]. Not
edema in patients with acute ON. In contrast, MV and surprisingly, HCVA is worse in the post-acute phase
GCL +IPL thickness maybe normal (in the case of a of ON for NMO patients [17]. In their study of 26
CIS patient), or perhaps slightly reduced (in the case NMO patients, Ratchford and colleagues [17]
of MS patients with preexisting damage to the AVP). reported a threshold of RNFL thickness measuring
Initially, there is generally poor concordance between 60 µm, below which visual acuity outcomes were
OCT measures of structural integrity and functional poor in the setting of NMO.
outcomes in patients with acute ON. In the post-acute Low-Contrast Letter Acuity: While HCVA has
phase, however, lower RNFL, MV, and GCL + IPL been used as the primary outcome in clinical trials
49
 Chapter 5: Optical coherence tomography and visual outcomes in acute optic neuritis

Figure 5.1 (cont.)

in ophthalmology, it is a relatively crude measure of (lowest contrast level) [22]. Charts are scored letter
afferent visual function in MS. In fact, many patients by letter, and numbers of letters identified correctly
will report significant vision problems after acute ON, constitute the score for each chart. Recently, visual
even in the setting of 20/20 HCVA [21]. High- improvement and loss by the LCLA chart has been
contrast visual acuity is determined under optimal defined as a seven-letter change in score [22]. In the
circumstances of high contrast and high luminance, IMPACT trial and in heterogeneous MS cohorts
and it is therefore not surprising that HCVA tends to LCLA testing has been shown to be a highly reliable
manifest the most obvious early improvement in and practical means of identifying MS-related visual
acute ON patients. Yet the real visual world is one of loss compared to other available tests [22]. Low-
varying spatial and temporal frequencies, contrast, contrast letter acuity scores have also been shown to
color, luminance, and glare [2]. Over the past decade, correlate with quality-of-life outcomes [22], and, in
Balcer and colleagues have spearheaded the imple- many respects, to capture the many shades of gray
mentation of low-contrast letter acuity (LCLA) testing that represent visual perception in our day-to-day
in MS clinical research with the use of Sloan charts lives. In clinical practice, measuring LCLA can be
[22]. Sloan charts have a standardized format based hampered in acute ON patients who present with
on the ETDRS HCVA charts and employ three con- severe vision loss (worse than 20/200) because they
50 trast levels: 100% (high-contrast), 2.5%, and 1.25% will not be able to visualize the charts [21]. However,
 Chapter 5: Optical coherence tomography and visual outcomes in acute optic neuritis

Figure 5.1 (cont.)

post-acutely, LCLA can be used to detect persistent by 4 µm [21]. More recently, Syc [11] followed
visual deficits that may otherwise be missed in 20 acute ON patients with serial HCVA, LCLA, and
patients who have regained “normal” vision on stan- OCT testing at baseline (within four weeks), three
dard HCVA and visual field testing. Previous studies months, and six months of initial presentation. Low-
have shown robust correlations between OCT- contrast visual acuity at three months improved in
measured RNFL, MV, and LCLA outcomes in ON ON eyes compared with baseline (2.5% p = 0.006;
and MS patients [2, 7, 21]. In 2006, Fisher [21] 1.25%, p = 0.02) [11]. In this same study, 22 NMO/
conducted a cross-sectional study comparing RNFL NMO-spectrum patients were studied in cross-
thickness among MS eyes with a history of ON sectional fashion. Ganglion cell layer + IPL, RNFL,
(MS ON eyes), MS eyes without a history of and macular thickness values were decreased in NMO
ON (MS non-ON eyes), and disease-free controls. ON eyes compared to NMO non-ON eyes, and both
They noted that RNFL thickness was reduced signifi- NMO eyes with and without remote ON compared
cantly among MS patients (92 µm) relative to controls with those of healthy controls. Ganglion cell layer+
(105 µm) and that RNFL values were particularly IPL thickness was significantly associated with 1.25%
reduced in MS ON eyes (85 µm) [21]. Furthermore, LCLA in NMO eyes, and 1.25% and 2.5% LCLA in
for every one-line decrease in LCLA or contrast NMO non-ON eyes [11]. Hence, LCLA testing is
sensitivity score, the mean RNFL thickness decreased more sensitive in detecting visual deficits than 51
 Chapter 5: Optical coherence tomography and visual outcomes in acute optic neuritis

Figure 5.1 (cont.)

HCVA testing in MS, and inclusion of these func- (VEP), contrast sensitivity, motion perception, and
tional outcomes in the AVP model will better capture OCT testing. In ON eyes, HCVA, visual field, and
deficits that impact quality of life and day-to-day color perception were significantly impaired at initial
function in these patients. presentation and recovered completely after one
Color Vision: Dyschromatopsia or loss of color month [23]. Contrast sensitivity recovered after four
vision is common in the setting of acute ON, and months [23]. After 12 months, the RNFL values in
improves weeks to months after initial presentation. both ON eyes and non-ON eyes were reduced when
Indeed, the detection of disproportionate color vision compared to the normal controls. The VEP ampli-
loss can be key to making the diagnosis in patients tudes of ON eyes were decreased in the acute phase
with acute ON who manifest mild deficits in HCVA but not subsequently [23]. Both ON eyes and non-ON
function. One of the challenges in interpreting color eyes of ON patients had significantly prolonged VEP
vision outcomes is that various modalities are used, latencies at all testing phases. This study demon-
and there is less standardization relative to HCVA, strated that functional recovery occurs at different
LCLA, and perimetry testing. Raz and colleagues [23] time points, depending on the testing modality used
followed 21 patients with new-onset acute ON with to follow ON patients.
serial HCVA, perimetry, color vision (Ishikawa pseu- Various other testing techniques have been used
52 doisochromatic plates), visual evoked potentials effectively to capture color vision deficits in ON and
 Chapter 5: Optical coherence tomography and visual outcomes in acute optic neuritis

Figure 5.1 (cont.)

MS patients. Hardy-Rand-Rittler (HRR) pseudo- also endeavored to evaluate age-related changes in

isochromatic color plates have demonstrated an chromatic discrimination in both patient groups
advantage over the Ishihara method, because the for- compared to normal control subjects. Color thresh-
mer are more sensitive to red–green and blue–yellow olds for both ON eyes and non-ON eyes in MS
deficits caused by neuro-ophthalmic disorders [2]. patients were significantly higher than controls’
Recently Villoslada and colleagues [24] studied along the protan and tritan axes [25]. In addition,
213 MS patient and 47 healthy controls to determine the ON and non-ON groups differed significantly
the relationship between HCVA, LCVA, color vision along all three-color axes [25]. Multiple sclerosis
(HRR pseudoisochromatic plates and Lanthony D15 patients manifested progressive color discrimination
tests), and OCT measures in MS patients. Multiple impairment with age (along the deutan and tritan
sclerosis patients showed HCVA and LCLA deficits axes) that was almost two times faster than controls,
but exhibited even more profound abnormalities in even in the absence of ON [25]. In a seminal OCT
color discrimination relative to controls [24]. Moura study, Trip [26] selected 25 ON patients in the post-
and colleagues [25] assessed chromatic discrimina- acute phase with incomplete visual recovery, and
tion in 35 MS patients (with and without ON) and performed cross-sectional analyses comparing OCT
74 age-matched controls using the Cambridge color measured peripapillary RNFL thickness and MV,
test (CCT) to determine the magnitude and chromatic VEP, HCVA, perimetry, and color vision testing 53
axes of color vision loss in both patient groups. They (Farnsworth–Munsell). In this study, RNFL thickness
 Chapter 5: Optical coherence tomography and visual outcomes in acute optic neuritis

Figure 5.1 (cont.)

was reduced by 33%, and macular volume by 11% in 100 Hue color vision testing. There was a significant
ON eyes relative to control eyes [26]. Interestingly, correlation between the average peripapillary RNFL
in ON eyes lower MV values were associated thickness, HCVA, LCLA, and visual field scores [27].
with color vision loss. The authors highlighted that Yet, there was no correlation between OCT measures
axons originating in the macula pass in the papillo- and color vision testing. These investigators did not
macular bundle to the temporal side of the optic disc; use temporal RNFL or MV measurements in
consistent with this, MV values were related to their study, which may in part explain the differences
the thickness of the temporal RNFL quadrant in in their results as compared to the findings by Trip
this study [26]. et al. [27]. Future studies could further explore the
Given that NMO patients have generally more optimal color vision test to use in patients with acute
severe ON than that seen in MS patients, RNFL values ON, and determine whether GCL+ IPL thickness as
are reduced in NMO ON eyes, and they correlate a marker of neuronal integrity correlates with color
with more severe color vision deficits than their MS vision deficits in various phases of recovery.
counterparts. In a cross-sectional study of 15 NMO Visual Field Testing: Visual field testing has been
patients, 15 MS patients, and 23 controls, Merle and described as the “cornerstone” of the sensory visual
colleagues [27] reported lower RNFL values in NMO examination and provides invaluable information
patients (65 µm) relative to MS (84 µm) and control about the integrity of afferent visual pathway func-
54 (106 µm) eyes. This group employed OCT, HCVA, tion from the retina to the visual cortex [28].
LCLA (1.25% and 2.5%), perimetry, and Farnsworth Perimetry has evolved through stages since original
 Chapter 5: Optical coherence tomography and visual outcomes in acute optic neuritis

Figure 5.1 (cont.)

confrontation-based techniques to allow quantifica- of the visual fields from the fellow (non-ON) eyes in
tion and statistical analysis in its currently used patients were abnormal [29]. After year one of the
computerized forms. This provides critical informa- study, 50% of the visual fields were abnormal in ON
tion about visual function, including both central eyes, whereas the abnormal visual field frequency in
and peripheral channels. Moreover, many of the non-ON eyes ranged between 34% and 40% [29].
perimeters in common use today are readily avail- Diffuse and central visual field losses were more
able in most cities around the developed world, prominent in ON eyes than non-ON eyes at baseline.
allowing for standardization between centers, and Retinal nerve fiber bundle defects (partial arcuate,
easier comparisons across offices over time. In the paracentral, and arcuate) were the most prominent
recent 15-year follow-up from the ONTT, Keltner localized abnormalities in ON and non-ON eyes
[29] defined visual field characteristics and classifica- during the study [29]. After 15 years, 40% of
tions for the entire cohort, from baseline through 15 abnormalities in the ON eyes and 26% in the non-
years (10,443 visual fields). At initial presentation, ON eyes consisted of localized defects [29]. Foveal 55
100% of the visual fields from the ON eyes and 75% threshold, as measured by Humphrey perimetry,
 Chapter 5: Optical coherence tomography and visual outcomes in acute optic neuritis

Figure 5.1 (cont.)

correlated with HCVA and contrast sensitivity in variability for normal controls [30]. These variations
ON eyes at baseline, six months, and one year [29]. occurred for multiple tests performed on the same day,
For all of the established advantages of automated at specific times, and for tests performed at specific
perimetry, there are potential pitfalls with this testing times on different days [30]. Thus, when using auto-
modality in a patient population that is subject to mated perimetry, or for that matter any subjective
fatigue-related visual dysfunction. Wall and colleagues psychophysical test to follow ON and MS patients,
[30] followed 17 patients with ON and ten healthy distinguishing true change from variability remains a
control subjects with repeat intra-day and interday challenge. In the case of automated perimetry, this
automated perimetry testing (five Humphrey 30–2 requires more than comparing a current visual field
tests were administered during a seven-hour period test with the most recent previous visual field test;
on the same day and at the same period on five separate rather, trends should be observed over time.
56 days). Optic neuritis patients demonstrated variations In 2006, our group followed 54 patients with
in visual field sensitivity outside the entire range of acute ON prospectively over a 12-month period
 Chapter 5: Optical coherence tomography and visual outcomes in acute optic neuritis

with serial OCT, HCVA, and automated perimetry dominated by the macular region due to its cortical
testing [12]. Thinning of the RNFL was seen in the overrepresentation [31, 32]. Moreover, the waveform
majority of patients (74%), and it tended to occur of the ffVEP is prone to cancellation and distortion,
within 3 to 6 months of symptom onset [12]. The which sometimes leads to apparent, rather than real,
average peripapillary RNFL values were thinner in latency delay [31, 32]. In contrast, the multifocal VEP
ON eyes (78 µm) than in non-ON eyes (100 µm) [12]. (mfVEP) allows stimulation of small areas of the visual
Patients with incomplete visual recovery demon- field. The result is a detailed topographical assessment
strated worse RNFL loss after ON. Regression ana- of small groups of axons within the optic nerve and
lyses revealed a threshold of RNFL thickness visual cortex, which is resistant to waveform distortion
(75 µm), below which RNFL measurements pre- [31, 32]. In a recent study, Klistorner and colleagues
dicted persistent visual field dysfunction [12]. In [31] used mfVEP and OCT testing to study 25 subjects
this same study, measures of RNFL thickness were with acute unilateral ON. While mfVEP amplitude
compared with log MAR visual acuity for 54 patients asymmetry at baseline varied significantly among the
observed a minimum of three months after acute patients, it was, on average, very high, indicating con-
ON. The predictive value for RNFL thickness siderable reduction of amplitude in ON eyes [31].
was significant only for patients with measures less The inter-eye asymmetry in mfVEP amplitude
than 70 µm, below which RNFL values were asso- decreased over time, indicating functional recovery
ciated with worse HCVA outcomes after ON [12]. [31]. There was a negative correlation between the
Interestingly, in subsequent work, an OCT- inter-eye asymmetry of OCT-measured RNFL thick-
measured RNFL threshold of 80 µm was found to ness and that of mfVEP amplitude at one month,
distinguish patients with persistent LCLA deficits consistent with vasogenic edema in the acute phase
after acute ON [22]. Given that LCLA, automated (causing an increase in RNFL thickness with a corre-
perimetry techniques, and HCVA are variably sponding reduction in mfVEP amplitude) [31]. Over
sensitive to persistent visual deficits after acute ON, the course of recovery, the correlation became progres-
OCT testing may reveal a range of RNFL sively more robust, suggesting the diminishing role of
(and MV and GCL+IPL) thresholds that predict optic nerve edema in measured RNFL thickness and
visual recovery, depending on visual outcome used. unmasking the association between RNFL atrophy and
Visual Evoked Potential Testing: The VEP is a low mfVEP amplitude [31]. The potential correlation
response of the brain to repeated visual stimulation between OCT-measured RNFL values and mfVEP
and has traditionally been recorded when visual field measures of anterior visual pathway damage was
is stimulated with a single checkerboard pattern in the demonstrated by the same group, who evaluated 32
full-field (ff-VEP). The VEP is known to be generated patients with unilateral ON and 25 control subjects
at the level of striate cortex. The magnitude of the VEP with mfVEP and OCT testing [32]. The mean RNFL
is interpreted to reflect the number of functional affer- thickness in ON eyes (85 μm) was reduced by 19%
ent fibers reaching this region [31, 32]. In ON patients, compared with control eyes (104 μm). There was a
the number of functional afferent fibers is believed to 40% reduction in the amplitude of the mfVEP in ON
be determined by the severity of the inflammation and eyes relative to control eyes [32]. In addition to demon-
axonal degeneration along the visual pathway [31, 32]. strating the utility of mfVEP in tracking optic nerve
Therefore, diminished VEP amplitude indicates injury in ON patients, this study further confirmed the
inflammation-induced conduction block, axonal atro- significant correlations between structural and func-
phy, or a combination of both [31, 32]. Subsequently, tional measures of optic nerve integrity, and showed
an increase in VEP amplitude is a consequence of that demyelination contributes to axonal loss. It may,
resumed conduction in previously blocked fibers, due therefore, be feasible to pair mfVEP and OCT testing to
to resolution of inflammation and edema, or possibly capture the synergistic effects of acute demyelination
expansion of synaptic activity along the visual pathway and axonal loss over time in ON/MS patients.
[31, 32]. Delayed VEP conduction is recognized as one Furthermore, the putative relationship between the
of the earliest features of acute ON, with the subse- VEP latency and axonal loss encourages the notion
quent shortening of latency thought to represent that therapeutic interventions aimed at reducing the
remyelination [31, 32]. Because it is a summation of a effects of demyelination or enhancing remyelination
large number of neuronal elements, the ffVEP is greatly may be given a trial in the AVP model. 57
 Chapter 5: Optical coherence tomography and visual outcomes in acute optic neuritis

Motion Perception Testing: Standard tests of visual ON, suggesting that these visual functions depend on
function generally interrogate our perception a sufficient amount of visual information reaching the
of “form,” and tend to ignore perception of motion, cortex [33]. Yet motion perception was impaired
which has an enormous impact on the day-to-day even in patients with intact VEP amplitudes, indicat-
function. In the convalescent phase, many ing that an intact amount of visual projection alone
ON patients report the “Pulfrich phenomenon” does not impact dynamic visual function [33].
(a binocular perception that a small target moving in Instead, while the magnitude of contrast sensitivity
a frontal plane is traveling in an elliptical path) and improvement related to the extent of VEP amplitude
describe difficulties tracking moving objects. The restoration, the magnitude of motion perception
hypothesis that ON can have prolonged effects on improvement depended on the extent of VEP latency
visual motion processing, which may persist after reduction post-ON [33]. From these findings, it was
there has been an objective return to normal “form” inferred that there is a need for rapid transmission of
perception, was further explored by Raz and collea- visual input to perceive motion. Moreover, motion
gues in a series of elegant papers [23, 33]. They perception testing in concert with VEP may serve to
prospectively followed 21 ON patients over one year assess the processes of demyelination and remyelina-
with tests of spatial and dynamic visual function tion in the AVP model of MS.
including: HCVA, perimetry, contrast sensitivity, Binocular Summation: Binocular summation
color vision, VEP, and OCT testing [23, 33]. In addi- refers to the improved visual function observed
tion, the authors developed a novel set of motion when patients perform threshold tasks such as
perceptual tasks to test dynamic visual deficits as contrast detection with a binocular view as compared
part of their protocol [23, 33]. In ON eyes, visual to monocular vision [34]. Conversely, binocular inhi-
acuity, visual field, and color vision deficits were sig- bition refers to worse outcomes for visual threshold
nificantly impaired in the acute phase, and subse- tasks obtained with binocular vision as compared to
quently improved after one month [23]. Contrast the monocular view obtained when patients use their
sensitivity deficits tended to persist somewhat longer, better seeing eye [34]. The phenomena of binocular
improving four months after symptom onset [23]. summation and inhibition are not well understood,
As compared to tests of spatial visual function, but appear to be related to neural input from both
motion perception was impaired during the full eyes within the post-geniculate visual pathway [34].
follow-up period of one year [23]. Thus, motion Given that binocular summation is most likely pro-
perception testing revealed the most significant and cessed at the cortical level, MS and ON patients may
prolonged impairment after ON and was independent experience functional limitations due to interruptions
of contrast sensitivity levels [23]. in normal cortical signaling. In a recent study of 1,007
In a follow-up study, the same group aimed to patients with MS and 324 disease-free controls, bino-
identify mechanisms underpinning the sustained def- cular summation was substantial for LCLA at the 2.5%
icits in dynamic visual functions following ON. They and 1.25% levels [34]. With HCVA, only 3% of
hypothesized that motion perception may be more patients showed similar degrees of summation [34].
vulnerable to slowed conduction in the optic nerve, Increasing age, greater interocular differences in
which could be measured with VEP testing [33]. To acuity, and prior ON were factors associated with
explore this theory further, they did serial motion lower magnitudes of binocular summation and
perception and ffVEP testing at presentation, one worse binocular inhibition [34]. Greater degrees of
month, four months, and twelve months after acute binocular summation were predictive of better
ON [33]. The VEP amplitudes in ON eyes were sig- quality-of-life outcomes, indicating that the capacity
nificantly reduced compared to fellow eyes in the to use both eyes together is an important factor
acute phase, but these differences resolved in later in determining how well patients with MS perform
phases of recovery [33]. As has been previously daily activities [34]. In future studies, the effects of
reported, visual performance with HCVA, contrast binocular summation and inhibition could potentially
sensitivity, and motion perception at one month was be compared to OCT (as a structural measure of
highly predictive of visual recovery a year after acute neuronal and axonal damage), VEP (as a demyelina-
ON [33]. Intact VEP amplitudes were associated with tion and remyelination marker), and functional
58 recovered visual acuity and contrast sensitivity after MRI (as a means of detecting cortical activation).
 Chapter 5: Optical coherence tomography and visual outcomes in acute optic neuritis

In this way, the relative contribution of cortical 8. Green AJ, McQuaid S, Hauser SL, Allen IV, Lyness R.
compensation to visual recovery, and the factors that Ocular pathology in multiple sclerosis: retinal atrophy
govern the potential for cortical plasticity, may be and inflammation irrespective of disease duration.
Brain 2010; 133: 1591–1.
determined after acute ON in patients with CNS
demyelinating disorders. 9. Miller D, Barkhof F, Montalban X, Thompson A,
Filippi M. Clinically isolated syndromes suggestive of
multiple sclerosis, part 2: non-conventional MRI,
Summary recovery processes, and management. Lancet Neurol.
The AVP model offers an exciting opportunity to 2005; 4: 341–8.
explore disease mechanisms that contribute to neuro- 10. http://www.guggenheim.org/new-york/education/schoo
logical disability in MS patients. Monitoring the acute l-educator-programs/teacher-resources/arts-curriculum-
and chronic consequences of clinically overt ON online?view=item&catid=730&id=120
events may shed light on factors that the govern injury 11. Syc S, Saidha S, Newsome SD, et al. Optical coherence
and repair in acute inflammatory relapse in the CNS. tomography segmentation reveals ganglion cell layer
Highly sensitive and standardized tests of vision can pathology after optic neuritis. Brain. 2012; 135: 521–33.
be paired with high-resolution imaging measures of 12. Costello F, Coupland S, Hodge W, et al. Quantifying
structural integrity in the AVP pathway to develop a axonal loss after optic neuritis with optical coherence
tomography. Ann Neurol. 2006; 59: 963–9.
structural–functional paradigm of CNS injury. In the
future, the AVP model may allow us to determine 13. Henderson APD, Altmann DR, Trip SA, Kallis C,
Jones SJ, Schlottmann PG, et al. A serial study of retinal
how remyelination contributes to axonal and neuro-
changes following optic neuritis with sample size esti-
nal repair. Furthermore, this model may also help us mates for acute neuroprotection trials. Brain 2010;
determine whether retinal damage in MS arises from 133: 2592–602.
retrobulbar optic nerve pathology or primary neuro- 14. Costello F, Hodge W, Pan YI, Metz L, Kardon RH.
nal pathology. Retinal nerve fiber layer and future risk of multiple
sclerosis. Can J Neurol Sci. 2008; 35: 482–7.
References 15. Yeh EA, Weinstock-Guttman B, Lincoff N, et al.
1. Miller D, Barkhof F, Montalban X, Thompson A, Retinal nerve fiber thickness in inflammatory demye-
Filippi M. Clinically isolated syndromes suggestive of linating diseases of childhood onset. Mult Scler. 2009;
multiple sclerosis, Part 1: natural history, pathogen- 15: 802–10.
esis, diagnosis and prognosis. Lancet Neurol. 2005; 16. Naismith RT, Tutlam NT, Xu J, et al. Optical coher-
4: 281–88. ence tomography differs in neuromyelitis optica
2. Costello F. The afferent visual pathway: designing a compared with multiple sclerosis. Neurology. 2009;
structural-functional paradigm of multiple sclerosis. 72: 1077–82.
ISRN Neurol 2013; doi:10.1155/2013/134858. 17. Ratchford JN, Quigg ME, Conger A, et al. Optical
3. Frohman EM, Fujimoto JG, Frohman TC, Calabresi PA, coherence tomography helps differentiate neuromye-
Cutter G, Balcer LJ. Optical coherence tomography: a litis optica and MS optic neuropathies. Neurology.
window into the mechanisms of multiple sclerosis. 2009; 73: 302–8.
Nat Clin Pract Neurol. 2008; 4: 664–75. 18. Nakamura M, Nakazawa T, Doi H, et al. Early high-
4. Beck RW, Cleary PA, Anderson MM, et al. A rando- dose intravenous methylprednisolone is effective in
mized, controlled trial of corticosteroids in the treatment preserving retinal nerve fiber layer thickness in
of acute optic neuritis. The Optic Neuritis Study Group. patients with neuromyelitis optica. Graefes Arch Clin
N Engl J Med. 1992; 326: 581–8. Exp Ophthalmol. 2010; 248: 1777–85.
5. Hickman SJ, Dalton CM, Miller DH, Plant GT. 19. Kaiser PK. Prospective Evaluation of Visual Acuity
Management of acute optic neuritis. Lancet. 2002; 360: Assessment: A Comparison of Snellen versus ETDRS
1953–62. Charts in Clinical Practice (An AOS Thesis). Trans Am
6. Beck RW, Cleary PA, Optic Neuritis Study Group. Ophthalmol Soc. 2009; 107: 311–24.
Optic Neuritis Treatment Trial one year follow-up 20. Kupersmith MJ, Gal RL, Beck RW, et al. Optic Neuritis
results. Arch Ophthalmol. 1993; 111: 773–5. Study Group. Visual function at baseline and 1 month
7. Costello F, Klistorner A, Kardon R. Optical coherence in acute optic neuritis: predictors of visual outcome.
tomography in the diagnosis and management of optic Neurology. 2007 Aug 7; 69(6): 508–14.
neuritis and multiple sclerosis. Ophthalmic Surg Lasers 21. Fisher JB, Jacobs DA, Markowitz CE, et al. Relation
Imaging. 2011; 48: S28-S40. of visual function to retinal nerve fiber layer
59
 Chapter 5: Optical coherence tomography and visual outcomes in acute optic neuritis

thickness in multiple sclerosis. Ophthalmology. 2006; Neuro-Ophthalmology the Practical Guide. Chapter 4,
113: 324–32. Thieme Medical Publishers, 2005.
22. Sakai RE, Feller DJ, Galetta KM, Galetta SL, Balcer LJ. 29. Keltner JL, Johnson CA, Cello KE. Visual field profile
Vision in Multiple Sclerosis: The Story, Structure– of optic neuritis: A final follow-up report from the
Function Correlations, and Models for Optic Neuritis Treatment Trial from baseline through
Neuroprotection. Journal of Neuro-Ophthalmology 15 years. Arch Ophthalmol. 2010; 128(3): 330–7
2011; 31: 362–73. 30. Wall M, Johnson CA, Kutzko KE, Nguyen R, Brito C,
23. Raz N, Dotan S, Benoliel T, et al. Sustained motion Keltner JL. Long– and short-term variability of auto-
perception deficit following optic neuritis. Behavioral mated perimetry results in patients with optic neuritis
and cortical evidence. Neurology; 2011; 76: 2103–11. and healthy subjects. Arch Ophthalmol 1998; 116(1):
24. Villoslada P, Cuneo A, Gelfand J, Hauser SL, Green 53–61.
A. Color vision is strongly associated with retinal thin- 31. Klistorner A, Arvind H, Nguyen T, et al. Axonal loss
ning in multiple sclerosis. Mult Scler. 2012; 18(7): 991–9. and myelin in early ON loss in postacute optic neuritis.
25. Moura AL, Teixeir RA, Oiwa NN, et al. Chromatic Ann Neurol. 2008; 64: 325–31.
discrimination losses in multiple sclerosis patients 32. Klistorner A, Arvind H, Nguyen T, et al. Multifocal
with and without optic neuritis using the Cambridge VEP and OCT in optic neuritis: a topographical study
Colour Test. Vis Neurosci Vis. 2008; 25(3): 463–8. of the structure-function relationship.
26. Trip SA, Schlottmann PG, Jones SJ, et al. Retinal nerve Doc Ophthalmol. 2009; 118: 129–37.
fiber layer axonal loss and visual dysfunction in optic 33. Raz N, Dotan S, Chokron S, Ben-Hur T, Levin N.
neuritis. Ann Neurol. 2005; 58: 383–91. Demyelination Affects Temporal aspects of percep-
27. Merle H, Olindo S, Donnio A, Richer R, Smadja D, tion: An optic neuritis study. Ann Neurol. 2012; 71:
Cabre P. Retinal Peripapillary Nerve fiber layer thick- 531–8.
ness in neuromyelitis optica. Invest Ophthalmol Vis Sci. 34. Pineless SL, Birch EE, Talman LS. One eye or two: A
2008; 49: 4412–7. comparison of binocular and monocular low-contrast
28. M. Wall. Choice and interpretation of visual acuity testing in multiple sclerosis. Am J Ophthalmol
field testing. In: Levin LA, Arnold AC. 2011; 152: 133–40.

60
 Chapter
Optical coherence tomography

6 and low-contrast acuity

Shin C. Beh and Laura J. Balcer

Introduction delineates the existence of a pattern or object [2].

Multiple sclerosis (MS) is perhaps the quintessential More specifically, contrast is defined as the ratio
neurologic disorder for which the structure–function of the difference in the luminance of these two
correlations provided by optical coherence tomography adjacent areas to the lower or higher of these lumi-
(OCT) have allowed the visual system to serve as a nance values [2]. The amount of contrast can
model for understanding the disorder and testing new be quantified using the Weber (used when back-
therapies. Low-contrast letter acuity testing has provided ground luminance remains constant) or Michelson
a sensitive and reliable measure of visual function that (used when both light and dark components change)
completes this model in conjunction with OCT. formulas [3, 4].
However, it was low-contrast letter acuity that first The minimum amount of contrast needed
emerged in the line of investigation that brought vision to visualize a target is referred to as the contrast
to a prominent place within MS research and clinical threshold [2, 3]. In clinical research, it is typically
trials. expressed as contrast sensitivity, where sensitivity is
In the 1990s, in a quest for a more sensitive outcome simply the reciprocal, or inverse, of contrast thresh-
measure, the National MS Society Clinical Outcomes old [2, 3]. Contrast sensitivity is expressed on a
Assessment Task Force developed the MS Functional logarithmic10 scale, to make the values linear and
Composite (MSFC) [1]. Not yet included in the MSFC permit comparison at low- and high-contrast levels
at that time, however, was a measure of visual function; [2, 3].
this was the case since high-contrast visual acuity, mea- The size of a target influences how much contrast
sured most frequently using Snellen charts, did not is required to distinguish it from its background.
show sufficient sensitivity to change over time in the Any target size may be depicted by lines of appropri-
clinical trial datasets used to develop the MSFC. The ate spacing, subtending a specific visual angle. The
ability of low-contrast sensitivity testing to detect visual number of adjacent dark and light lines (cycles)
abnormalities that would otherwise be missed with within a defined visual angle is defined as the spatial
Snellen charts in the Optic Neuritis Treatment Trial frequency. A high spatial frequency (or high number
(ONTT) spurred the use of low-contrast vision as an of cycles per degree) consists of more densely packed
exploratory outcome measure in MS clinical trials. lines, while a low spatial frequency is displayed as
Even prior to the identification of low-contrast letter sparsely packed lines [3]. The contrast-sensitivity
acuity as a potential visual component for the MSFC function represents the relationship between contrast
and test of function for MS research, an extensive sensitivity and spatial frequency [3, 4].
literature on various forms of low-contrast vision test- On the other hand, visual acuity (VA) refers to a
ing provided evidence supporting its role in MS. measure of the spatial-resolving ability of the visual
system, typically under conditions of high contrast
What is low-contrast vision? (black on white, or at least 85% contrast); all the
Spatial contrast can be understood as the light–dark targets are presented at the same contrast level, but
transition at the border or edge of an image that their sizes vary during the test [2].

Optical Coherence Tomography in Neurological Disease, ed. Peter A. Calabresi, Laura J. Balcer, and Elliot M. Frohman. 61
Published by Cambridge University Press. © Cambridge University Press 2015.
 Chapter 6: Optical coherence tomography and low-contrast acuity

The neural substrate for computer-generated visual images as test targets

(usually vertically oriented sine-wave gratings) with
low-contrast vision software-controlled threshold measurement proto-
In the retina, rod photoreceptors specialize in scotopic cols [2]. The use of these devices in a busy clinical
(low-light-level) vision, while cone photoreceptors med- setting was impractical due to a number of factors,
iate chromatic vision in photopic (high-light-level) con- including high cost, a difficult calibration process, the
ditions. Beyond the level of the photoreceptors, there are lack of normative data, and time-consuming testing
predominantly two parallel but morphologically and protocols [2].
physiologically distinct neuronal populations – the As a result, chart-based methods of assessing
large type A (or phasic) retinal ganglion cells (RGC) contrast sensitivity grew more popular in the 1980s
and the smaller type B (or tonic) RGCs [5–7]. [2]. The cycles per degree of the gratings or size of the
The type A RGCs form the magnocellular (M– or letters on these charts determine the spatial frequency
dorsal) pathway that projects to the magnocellular being evaluated [3]. Historically, the first commer-
layers (via thick axons in the optic nerve) of the lateral cially available, printed contrast-sensitivity tests for
geniculate nucleus (LGN), and subsequently the 4C- clinical use were the Arden plates, a booklet consisting
alpha layer of the striate cortex [8–11]. Making up of seven plates, with sine-wave grating printed on
10% of the total RGC population, almost all these each [2]. Early on, there were observations
M cells have large receptive fields, have high-contrast that Arden plate contrast sensitivity was abnormal in
sensitivity but poor chromatic sensitivity, and are various disorders, even in the presence of normal
responsible for detection of low and medium spatial high-contrast VA [2]. However, its weaknesses are
frequencies, motion, and fast flicker [12–17]. In other that the testing method does not involve a forced-
words, M-cells are more sensitive achromatic stimuli choice method and the results are contingent on
of higher temporal, and lower spatial, frequencies [7, the speed of the examiner exposing the contrast
18, 19]. The M-pathway has been shown to be respon- images [3].
sible for frequency-doubling technology (FDT) The VisTech chart and its later incarnation,
contrast sensitivity [20, 21]. the Functional Acuity Contrast Test (FACT) are
The type B RGCs, making up the parvocellular wall-mounted grating charts [2, 3]. Problems with
(P– or ventral) pathway, project via thinner optic these charts include poor test–retest reliability, redun-
nerve axons to the four parvocellular layers of the dant information, and a high probability of correct
LGN, and subsequently to the 4C-beta layer of the guessing (since only three answer choices were avail-
striate cortex [9–11, 22]. About 80% of the RGC able) [2, 3]. The Sine-Wave Contrast Test (SWCT) is
population consists of these P cells. Almost all almost identical to the VisTech charts, and it suffers
P cells receive antagonistic input from red and from similar limitations [2]. The CSV-1000 is another
green cones and, therefore, possess high chromatic distance, grating-based chart that utilizes an internal
sensitivity (particularly red–green color selectivity) retroillumination system and is mounted on the wall
but far lower contrast sensitivity and smaller recep- or a stand [2]. Retroillumination is advantageous,
tive fields compared to the M-cells [13, 14, 17, 23, since it eliminates variation resulting from lighting
24]. P-cells are more sensitive to stimuli of lower conditions, but having only two answer choices
temporal and higher spatial frequencies [18, 19]. creates a 50% chance of guessing correctly [3].
More recently discovered, the koniocellular (K) There are various commercially available letter
pathway conveys information regarding blue–yellow charts, including Pelli-Robson, the most widely used
opponency to cortical layers above 4C, and it is both contrast sensitivity letter chart [30]. Consisting of
morphologically and functionally distinct from the letters of a single large size and varying contrast, the
M- and P-pathways [25–29]. The precise role of the Pelli-Robson chart estimates the peak of the contrast-
K-pathway in vision has yet to be elucidated. sensitivity function. It is a wall-mounted chart con-
sisting of eight rows of letters, with two triplets of
letters per row, placed at 1 meter testing distance.
Measuring contrast sensitivity The contrast in each successive triplet decreases by a
When initially introduced into clinical practice and factor of 0.15 log units, from near 100% at the top to
62 research, contrast-sensitivity testing consisted of less than 1% at the bottom. During testing, the patient
 Chapter 6: Optical coherence tomography and low-contrast acuity

reads from the top of the chart and proceeds down

until he or she can read no more. The Pelli-Robson
charts demonstrate high test–retest reliability, are
relatively unaffected by background luminance, have
a normative database, and are easy and efficiently
administered. As a result, they are a frequent choice
for epidemiologic studies [2, 3]. Pelli-Robson charts
are good measures of medium to low spatial frequen-
cies [31]. However, since the letters are of uniform
size, this method can only detect contrast sensitivity
loss at a single spatial frequency [32]. This may be
problematic in neurologic or ophthalmic disorders
that impair contrast sensitivity for intermediate
spatial frequencies, referred to as “notch” loss of
contrast [33–36].
Figure 6.1 Low-contrast letter acuity chart (low-contrast Sloan
letter chart, Precision Vision, LaSalle, IL). * These charts have a
Measuring low-contrast letter acuity standardized format based on the Early Treatment Diabetic
Retinopathy Study visual acuity charts, the standard used in
Low-contrast letter acuity charts provide diagnostic ophthalmology clinical trials, and they have several advantages over
information qualitatively similar to that of contrast standard Snellen charts or near vision testing cards as traditionally
used in MS trials: (1) letters (Sloan letters) are designed to be equally
sensitivity tests, and they possess comparable detectable for normal observers; (2) each line has an equal number
sensitivity [37, 38]. However, their format differs of letters (five per line); (3) spacing between letters and lines is
from that of contrast sensitivity because they include proportional to the letter size; (4) the change in visual acuity from
one line to another occurs in equal logarithmic steps (a change of
lines of letters of progressively smaller size and a single three lines constitutes a doubling of the visual angle); (5) visual
contrast level. The Low-Contrast Sloan Letter Charts acuity (for high-contrast [black letters on white] chart) may be
(LCSLC) were first used as an exploratory MS clinical specified by Snellen notation for descriptive purposes (i.e., 20/20), by
the number of letters identified correctly. This figure shows the 25%
trial outcome in the IMPACT (International MS contrast level for purposes of illustrating the format; the actual
Progressive Avonex Clinical Trial) study of interferon contrast levels used in these trials, 2.5% and 1.25%, have substantially
beta-1a for secondary progressive MS. While the lighter gray letters. The charts measure 14 × 14 inches for easy use
and portability in the MS clinical trial setting; charts may also be
LCSLC were initially chosen for investigation in MS mounted on a retroilluminated cabinet, thus, eliminating the need
due to a difficulty obtaining the Pelli-Robson contrast for standardization of room lighting levels.
sensitivity charts commercially, the LCSLC charts
actually have several advantages for a clinical trial out- characteristics in clinical trial outcomes), and it can be
come [1, 39]. In addition to potentially capturing easily understood by neurologists [43, 44]. A clinically
“notch” loss of contrast that could theoretically be significant change in vision has been more recently
missed by contrast-sensitivity testing, the LCSLC defined as a seven-letter change for the low-contrast
have a standardized format based on that of the charts and a five-letter change for the high-contrast
Bailey-Lovie and ETDRS (Early Treatment Diabetic charts, which are equivalent to two standard deviations
Retinopathy Study) charts, the standard charts used of inter-rater difference [1, 44]. Chart portability (mea-
for acuity measurement in ophthalmology clinical suring 14 × 14 inches), the use of an retroilluminated
trials (Figure 6.1) [40–42]. The charts are designed cabinet (negating the need for standardizing room
for distance measurement (thus, minimizing the effects lighting levels) high-inter-rater reliability, test time
of presbyopia), and the varying letter sizes allow for the efficiency, and the potential for administration by
testing of multiple spatial frequencies [1, 32]. Three trained, nonphysician professionals make the LCSLC
contrast levels are typically used: 100% (high-contrast), an excellent test in the clinical setting [32, 39, 43].
2.5%, and 1.25% [1]. The number of letters identified Potential limitations of letter charts include uneven-
(from a maximum of 70 per chart) is recorded; this ness of illumination, fading of the print, and reflections
letter-by-letter scoring system is similar to the contin- from the surface [3].
uous scale of the logMAR (log minimal angle of reso- Table 6.1 presents data from studies that have
lution), but it is quantitative and reliable (important examined LCSLC using the Sloan charts, as well as 63
Table 6.1 Mean reference values from recent investigations of vision, QOL, and OCT in MS

Disease- All MS MS, MS, References for

Free No History History of ON Data *
Controls of ON
High-contrast visual acuity (VA), ETDRS, number of 64 ± 5 59 ± 8 60 ± 6 58 ± 9 173 *
letters correct
(n=61 eyes) (n=239 eyes) (n=150 eyes) (n=87 eyes)
Binocular testing 66 ± 5 62 ± 8 63 ± 7 61 ± 10 62 *
(n=324 pts) (n=1,007 pts) (n=544 pts) (n=463 pts)
Low-contrast letter acuity (2.5%), number of letters 34 ± 8 26 ± 11 28 ± 9 22 ± 12 173 *
correct

(n=61 eyes) (n=239 eyes) (n=150 eyes) (n=87 eyes)

Binocular testing 43 ± 6 36 ± 10 38 ± 9 35 ± 11 62 *
(n=324 pts) (n=1,007 pts) (n=544 pts) (n=463 pts)
Low-contrast letter acuity (1.25%), number of letters 25 ± 7 16 ± 10 18 ± 10 11 ± 11 173 *
correct

(n=61 eyes) (n=239 eyes) (n=150 eyes) (n=87 eyes)

Binocular testing 34 ± 8 24 ± 11 26 ± 11 22 ± 12 62 *
(n=324 pts) (n=1,007 pts) (n=544 pts) (n=463 pts)
NEI-VFQ-25 composite score,
best score = 100 96 ± 4 (n=31 pts) 88 ± 13 (n=122 pts) 90 ± 12 (n=111 pts) 85 ± 14 (n=51 pts) 173 *
10-Item Neuro-Ophthalmic
Supplement to the NEI-VFQ-25, 97 ± 3 87 ± 13 88 ± 12 83 ± 14 173 *
best score=100 (n=31 pts) (n=122 pts) (n=111 pts) (n=51 pts)
 Time-domain (TD) OCT
Peripapillary RNFL thickness, μm 104.5 ± 10.7 92.5 ± 16.7 95.6 ± 14.5 85.7 ± 19.0 174 *
(n=219 eyes) (n=1,058 eyes) (n=730 eyes) (n=328 eyes)
Total macular volume, mm3 6.84 ± 0.36 6.54 ± 0.51 6.63 ± 0.48 6.36 ± 0.53 174 *
(n=219 eyes) (n=1,058 eyes) (n=730 eyes) (n=328 eyes)
Spectral-domain (SD) OCT
Peripapillary RNFL thickness, μm 92.9 ± 10.0 84.3 ± 12.8 87.6 ± 11.1 78.4 ± 13.6 173 *
(n=61 eyes) (n=239 eyes) (n=150 eyes) (n=87 eyes)
Ganglion cell + inner plexiform layer (GCL+IPL), μm 88.9 ± 6.9 (n=61 84.1 ± 8.4 (n=239 87.0 ± 6.6 (n=150 79.7 ± 9.2 (n=87 173 *
eyes) eyes) eyes) eyes)
Macular RNFL, μm 29.6 ± 6.0 23.5 ± 8.2 25.5 ± 7.1 20.0 ± 9.0 173 *
(n=61 eyes) (n=239 eyes) (n=150 eyes) (n=87 eyes)
Abbreviations: MS = multiple sclerosis; ETDRS = Early Treatment Diabetic Retinopathy Study; QOL = quality of life; NEI-VFQ-25 = 25-Item National Eye Institute Visual
Functioning Questionnaire; TD = time-domain (OCT-3 platform); SD = spectral-domain (Cirrus platform); OCT = optical coherence tomography; RNFL = retinal nerve fiber
layer
* Reference with asterisk is source of data presented in table; other published studies contain similar data. Table adapted from reference [1].
 Chapter 6: Optical coherence tomography and low-contrast acuity

high-contrast VA by ETDRS, OCT measures, and those for general vision, near and distance activities,
quality-of-life (QOL) scales using MS, optic neuritis role difficulties, and driving. Reduced contrast sensi-
(ON), and disease-free controls as an example [1]. tivity was also associated with poorer scores on other
These data may be used to provide initial reference measures of quality of life, including the Impact of
values and should be viewed in the context of (1) the Visual Impairment Scale (IVIS) and the Short Form
continually evolving field of vision in MS and neuro- 36 Health Survey (SF-36) [60]. Decreased binocular
logic disease, (2) the potential challenge inherent in low-contrast vision has also been shown to correlate
applying group data to individual patients, and with the Extended Disability Status Scale (EDSS) and
(3) the perspective that even the small observed the MS Functional Composite (MSFC) scores, both
differences in mean OCT values have been shown to measures of disability in MS patients [39]. In fact,
correlate with clinically meaningful changes in visual deterioration of low-contrast vision can predict future
function and QOL. EDSS decline [39].
How well a patient sees in high-spatial-frequency
channels, which is captured by standard tests of high-
Low-contrast letter acuity: contrast visual acuity (e.g., Snellen charts), does not
clinical correlates necessarily predict visual function at intermediate or
lower frequencies [2, 3, 61]. Contrast sensitivity test-
On a standard Snellen VA chart, a 20/20 letter occu-
ing can detect apparently occult evidence of visual
pies a very small area (0.083 degree of vision), which is
dysfunction that would otherwise be absent on
equivalent to 18 to 24 cycles per degree, the high-
standard tests of visual acuity. As will be discussed
frequency end of the contrast-sensitivity function
later, a number of disorders can often affect contrast
[3]. On the other hand, the peak of contrast sensitivity
sensitivity and spare high-contrast visual acuity,
occurs between a spatial frequency of 3 and 6 cycles
especially early in the course of the disease. Thus,
per degree (the equivalent visual frequency of the
contrast sensitivity is a sensitive measure of disease
20/200 line on the Snellen chart), a range responsible
progression, and it is a valuable clinical trial outcome
for activities like discerning a bus from a car [3]. This
tool in ophthalmic and neurologic disorders.
peak contrast sensitivity has been shown to be the
most important visual factor in predicting perfor-
mance on activities of daily function [3]. Binocular summation
Far from being an academic exercise without any
real-world applications, low-contrast vision has been and binocular inhibition
shown to play an important role in the ability to carry VA may be measured monocularly (with either eye
out activities of daily living (e.g., pouring liquids, separately) or binocularly (with both eyes together).
using tools) [46, 47], reading [31, 47–49], mobility While measuring binocular acuity saves time and
[50], negotiating stairs [51], driving [52], and facial avoids patient fatigue, comparing monocular to bino-
recognition [53–55], and influences perceived disabil- cular scores is important to ascertain the effects of
ity [56]. Further, diminished contrast sensitivity has binocular summation or inhibition [62]. Binocular
been associated with an increased risk for falls in older summation is a phenomenon in which VA improves
adults [57, 58]. under binocular viewing conditions (i.e., better than
There is evidence that impaired contrast sensitiv- the scores of either eye). In contradistinction,
ity is as disabling as visual field loss and more binocular inhibition results in worse VA than when
disabling than VA loss [59]. Decreased contrast sen- viewing with either eye alone [62, 63]. Low-contrast
sitivity has been shown to negatively impact patients’ acuity has been shown to be more sensitive than high-
quality of life. In a study of multiple sclerosis (MS) contrast VA in assessing the effect of binocular
patients, a two-line difference in low-contrast acuity summation [62, 64].
was associated with a deterioration of greater than Two hypotheses have been proposed to explain the
four points on the NEI-VFQ-25 (25-Item National phenomenon of binocular summation. “Probability
Eye Institute Visual Functioning Questionnaire) Summation” assumes complete independence of the
composite score, indicating an impaired quality of two eyes and predicts enhancement of binocular over
life [60]. The NEI-VFQ-25 domains most associated monocular vision due to the statistical consideration
66 with diminished contrast acuity in this study included that a binocular observer has two opportunities to
 Chapter 6: Optical coherence tomography and low-contrast acuity

detect weak signals. “Neural Summation” posits that impair contrast sensitivity by absorbing light, scatter-
binocular superiority exceeds that which would be ing light, and increasing disability glare [69, 113].
expected from probability summation alone and Notwithstanding the effects of pupillary miosis,
attributes this phenomenon to cortical interactions higher lens density, increased optical aberrations,
(layer VI) [65]. The deterioration in binocular sum- and increased intraocular light scattering on retinal
mation following acute optic neuritis (AON) may be a illuminance, age-related visual pathway changes also
consequence of complex cortical reorganization and conspire to further diminish contrast-sensitivity in
support the neural summation hypotheses [66]. older adults [64, 111]. Rod photoreceptor density is
Decreased binocular summation, or frank binocular particularly affected by aging compared to cone
inhibition, in MS patients is more common in photoreceptor density, which remains relatively stable
older age groups, those with a prior history of AON, [114]. Rod photoreceptor degeneration may result
and those with a greater inter-ocular VA difference from deficits in rhodopsin regeneration [113].
[62]. Better binocular summation was associated Furthermore, at the cortical level, studies in senescent
with higher NEI-VFQ-25 scores (which, in turn, monkeys have demonstrated that cortical neurons
suggests a better quality of life) [62]. Alternately, exhibit lower optimal spatial and temporal frequen-
binocular summation is decreased with age [64] and cies, decreased contrast sensitivity, diminished spatial
in strabismus [65]. resolution, and reduced higher temporal frequency
cut-offs when compared to younger monkeys [115,
Disorders affecting low-contrast vision 116]. FDT perimetry has been shown to be useful in
Contrast sensitivity can be affected as part of the detecting age-related contrast-sensitivity impairment
normal aging process, as well as in various ophthalmic, [112, 117, 118].
neurologic, and psychiatric conditions. Impaired The progressive loss of dopaminergic neurons, the
contrast sensitivity has been observed in myopia [67], hallmark of Parkinson’s disease (PD), also affects
cataract [68, 69], pseudophakia [70], meridional amby- the retina [119]. Retinal dopaminergic deficiency is
lopia [71], strabismus [65], dermatochalasis [72], non- hypothesized to cause physiological dysfunction of
penetrating corneal foreign bodies [73], age-related dopaminergic amacrine, horizontal, and interplexi-
macular degeneration [74–76], xerophthalmia [77], form cells that subsequently compromises the
birdshot chorioretinopathy [78], retinitis pigmentosa receptive field properties of RGCs. Pattern electrore-
[79–80], diabetic retinopathy [81, 82], and glaucoma tinography (PERG) can be used to detect RGC
[83–85]. dopaminergic dysfunction in PD patients, and, indeed,
In addition to MS, other neurologic disorders that levdopa treatment improves delayed PERG latencies
may compromise low-contrast vision include [89, 113, 120, 121]. Compared to the standard,
Parkinson’s disease [38, 86–89], Alzheimer’s disease high-contrast checkerboard pattern, low-contrast
[20, 90–92], Duchenne muscular dystrophy [93], PERG is more sensitive in detecting RGC dysfunction
neurofibromatosis type 1 [94], migraine [95, 96], in PD [88, 89]. Likewise, contrast sensitivity using
pseudotumor cerebri [97–100], Friedreich’s ataxia chart-based testing is diminished in PD [38, 87].
[101], a left parieto-occipital infectious lesion, and a Besides retinal abnormalities, pathologic changes
left occipital meningioma [102]. Reduced contrast elsewhere in the afferent visual pathway contribute
sensitivity has also been observed in schizophrenia to visual dysfunction in PD patients. Specifically, the
[103] and depression [104, 105]. Diminished contrast lateral geniculate nuclei and visual cortices may also
sensitivity is a common, albeit under recognized, side be affected by the neurodegenerative processes of PD
effect of many medications, including ethambutol [122–124]. Orientation-specific loss of contrast sensi-
[106] and tiagabine [107]. Exposure to solvents has tivity in PD may provide evidence for cortical invol-
also been implicated in decreased contrast sensitivity vement [125, 126].
[108–110]. Several studies have reported impaired contrast
Impaired contrast sensitivity under photopic sensitivity in Alzheimer’s disease (AD) [90–92].
conditions with increasing spatial frequencies occurs Using FDT, a more recent study [20] showed that
in older adults, as part of the normal aging process, decreased contrast sensitivity, most notably in
due to both optical and neural changes [111, 112]. the bilateral upper-right visual field quadrant, was
By age 75, almost half of adults have cataracts, which associated with poorer cognitive measures in 67
 Chapter 6: Optical coherence tomography and low-contrast acuity

AD and mild cognitive impairment. In AD, the outer detected by means of sensitive electrophysiological
retina (i.e., photoreceptors) is normal, as evidenced by assessments or tests of visual function. It is important
normal full-field electroretinography [127]. However, to identify such early dysfunction, since it may repre-
like PD, the PERG is abnormal, signifying RGC sent a potentially reversible state of visual compromise;
dysfunction [127, 128], the possible consequence of detecting these abnormalities would permit timely
retinal amyloid deposition [129]. Specifically, electro- therapeutic interventions that would potentially stop
retinographic abnormalities are most pronounced for the disease process, prevent future disability, and even
high-temporal-frequency patterns, indicating selec- reverse the damage.
tive impairment of the large M-type RGCs [128]. Although the EDSS [149] is widely used as
Pathologically, neurodegenerative changes have been an outcome measure for clinical trials in MS, this
noted in the central M-pathway in the lateral genicu- scale is heavily skewed in favor of the motor system
late nucleus and visual cortices [130–134], as well as in involved in ambulation and lacks any sensitivity with
the retina [135–142]. Furthermore, glaucoma, catar- regard to other salient manifestations of MS-related
acts, and macular degeneration (all of which also disability, including fatigue, heat sensitivity, cogni-
compromise contrast sensitivity) are more frequent tion, and, in particular, visual function. The MSFC
among AD patients [143]. was developed to serve as a more sensitive measure of
the multidimensional manifestations of MS-related
Structure meets function: disability and included the quantitative tests of
low-contrast acuity and OCT arm function (the 9-hole Peg Test), leg function (the
timed 25-foot walk), and cognition (the PASAT3)
in multiple sclerosis [150–152], but it is devoid of any measure of visual
MS is arguably the neurologic disorder in which low- function. For a long time, the only measure of visual
contrast acuity, as well as OCT, have been most widely function in MS patients was the hand-held Snellen
tested in recent years. MS frequently affects both the chart, which measured high-contrast visual acuity
afferent and efferent visual pathways. Visual dysfunc- (VA) [153].
tion occurs in 80% of patients and is the presenting While convenient and reproducible, the use of the
feature in 50% [144, 145]. Acute optic neuritis (AON) Snellen chart in MS patients is limited by several
is the presenting feature in 20% of cases, and 50% of inherent problems. First, Snellen VA is tested at the
MS patients develop AON during the course of their near distance, and it may potentially be difficult for
disease [146]. However, virtually all patients have presbyopic patients [32]. Second, despite normal
evidence of optic nerve demyelination at autopsy high-contrast VA, there may be pathologic evidence
[147, 148], indicating that occult, subclinical optic of severe visual pathway demyelination [147] and
nerve demyelination occurs in the vast majority of RNFL thinning [154, 155]. Finally, as discussed ear-
MS patients, whether or not they develop clinical lier, Snellen VA only tests higher-spatial-frequency
symptoms and signs of AON. levels and is insensitive to impairment of intermediate
The eloquence of the afferent visual pathway and to lower spatial frequencies. By contrast (pun
the frequency of AON in MS make it an ideal system to intended), low-contrast vision testing has proven to
use as a model to understand the pathobiological be a more sensitive measure of visual function in MS.
underpinnings of the disease, as well as to develop Many studies have shown that, despite recovery of
sensitive biomarkers that can be applied to clinical standard, high-contrast VA to 20/20 or better follow-
trials of neuroprotective, and even neuro-restorative, ing an attack of AON, patients continue to have
therapies. The robust utility of optical coherence tomo- persistent abnormalities of low-contrast vision
graphy (OCT) in detecting axonal damage in MS is [32, 34–37, 43, 44, 153–171].
explored at length throughout this book, but it remains Decreased contrast acuity correlates with RNFL
possible that early pathologic changes that begin at a thinning and reduced macular volume [44, 154, 155].
cellular level (particularly in the postgeniculate visual Poorer contrast sensitivity has been associated with a
pathway) may easily escape the resolution of even the higher MS lesion load, particularly in the postgenicu-
most sensitive OCT. Before any structural changes late visual pathway [168, 172], suggesting that postge-
become evident, these pathologic perturbations may niculate visual processing plays an important part in
68 manifest as functional abnormalities that may be contrast sensitivity. It would be easy to assume that
 Chapter 6: Optical coherence tomography and low-contrast acuity

RNFL thinning reflects pregeniculate damage while 2. Owsley C. Contrast sensitivity. Ophthalmol Clin N
impaired contrast sensitivity (without OCT changes) Am 2003; 16: 171–7
indicates postgeniculate lesions. Reich et al. [172], how- 3. Richman J, Spaeth GL, Wirotsko B. Contrast sensi-
ever, dispelled this simple dichotomy and showed that tivity basics and a critique of currently available tests.
postgeniculate MS lesions were associated with RNFL J Cataract Refract Surg 2013; 39: 1100–6
thinning as well. The precise mechanisms for retinal 4. Pelli DG, Bex P. Measuring contrast sensitivity.
changes resulting from postgeniculate damage remain Vision Res 2013; 90: 10–4
unclear, but they may be related to trans-synaptic 5. Gouras P. Identification of cone mechanisms in
degeneration. monkey ganglion cells. J Physiol 1968; 199: 533–47
Further, low-contrast acuity been shown to be the 6. Levanthal AG, Roedick RW, Dreher B. Retinal gang-
best measure that distinguishes MS patients from lion cell classes in the Old World monkey: morphology
and central projections. Science 1981; 213: 1139–42
disease-free controls, accounting for age [39, 153].
As a trial outcome for natalizumab (in the AFFIRM 7. Shapley R, Perry VH. Cat and monkey retinal gang-
lion cells and their visual function roles. Trends
and SENTINEL clinical trials), low-contrast acuity
Neurosci 1986; 9: 235
was sensitive enough to detect treatment effects and
8. Dreher B, Fukada Y, Roedick RW. Identification,
visual changes sustained over time, demonstrating the
classification and anatomical segregation of cells with
superiority of natalizumab as a disease-modifying X-like and Y-like properties in the lateral geniculate
agent [43]. As discussed previously, contrast sensitiv- nucleus of old-world primates. J Neurophysiol 1976;
ity correlates well with various measures of visual 258: 433–52
dysfunction and disability in MS. 9. Hubel DH, Wiesel TN. Laminar and columnar dis-
tribution of geniculo-cortical fibers in the macaque
Summary monkey. J Comp Neurol 1972; 146: 421–50
While impaired low-contrast vision is not pathogno- 10. Hendrickson AE, Wilson JR, Ogren MP. The neu-
roanatomical organization of pathways between the
monic for any particular disease, the presence of dorsal lateral geniculate nucleus and visual cortex in
diminished contrast sensitivity can alert the clinician Old World and New World primates. J Comp Neurol
to the presence of a pathologic disorder and help 1978; 182: 123–36
explain the bases of a patient’s visual complaints, 11. Blasdel GG, Lund JS. Termination of afferent axons in
particularly if other tests of visual structure and macaque striate cortex. J Neurosci 1983; 3: 1389–413
function are apparently normal. When used in con- 12. Hubel DH, Wiesel TN. Effects of varying stimulus size
junction with OCT, contrast sensitivity can provide and color on single lateral geniculate cells in Rhesus
useful and valuable information that will allow monkeys. Proc Natl Acad Sci U S A 1966; 55: 1345–6
clinicians and researchers to follow disease 13. Kaplan E, Shapley RM. X and Y cells in the lateral
progression. Contrast sensitivity testing can provide geniculate nucleus of macaque monkeys. J Physiol
information that complements that garnered from 1982; 330: 125–43
structural (e.g., OCT) and functional measures (e.g., 14. Derrington AM, Lennie P. Spatial and temporal
visual fields, color vision, electrophysiologic tests) of contrast sensitivities of neurons in lateral geniculate
the afferent visual system. Contrast sensitivity nucleus of macaque. J Physiol 1984; 357: 219–40
can now be easily tested in busy clinical or research 15. Livingstone MS, Hubel DH. Psychophysical evidence
settings with commercially-available charts as well as for separate channels for the perception of form, color,
established, efficient testing protocols. Evidence from movement, and depth. J Neurosci 1987; 7: 3416–68
MS clinical trials has demonstrated that contrast 16. Lee BB, Martin PR, Valberg A. Sensitivity of macaque
acuity is a reliable, sensitive measure of visual func- retinal ganglion cells to chromatic and luminance
flicker. J Physiol 1989; 414: 223–43
tion that can be used to follow disease evolution and
response to therapy. 17. Lee BB, Pokorny J, Smith VC, et al. Luminance and
chromatic modulation sensitivity of macaque gang-
lion cells and human observers. J Opt Soc Am A 1990;
References 7: 2223–36
1. Sakai RE, Feller DJ, Galetta KM, et al. Vision in 18. Laycock R, Crewther SG, Crewther DP. A role for the
multiple sclerosis: the story, structure-function ‘magnocellular advantage’ in visual impairments in
correlations,and models for neuroprotection. neurodevelopmental and psychiatric disorders.
J Neuroophthalmol 2011; 31: 362–73 Neurosci Biobehav Rev 2007; 31: 363–76 69
 Chapter 6: Optical coherence tomography and low-contrast acuity

19. McKendrick AM, Sampson GP, Walland MJ, 33. Bodis-Wollner I, Diamond SP. The measurement of
Badcock DR. Contrast sensitivity changes due spatial contrast sensitivity in cases of blurred vision
to glaucoma and normal aging: low-spatial fre- associated with cerebral lesions. Brain 1976; 99: 695–
quency losses in both magnocellular and parvocel- 710
lular pathways. Invest Ophthalmol Vis Sci 2007; 48: 34. Regan D, Silver R, Murray TJ. Visual acuity and
2115–22 contrast sensitivity in multiple sclerosis – hidden
20. Risacher SL, Wudunn D, Pepin SM, et al. Visual visual loss: an auxiliary diagnostic test. Brain 1977;
contrast sensitivity in Alzheimer’s disease, mild cog- 100: 563–79
nitive impairment, and older adults with cognitive 35. Kupersmith MJ, Seiple WH, Nelson JI, Carr RE.
complaints. Neurobiol Aging 2013; 34: 1133–44 Contrast sensitivity loss in multiple sclerosis.
21. Anderson AJ, Johnson CA. Frequency-doubling Selectivity by eye, orientation, and spatial frequency
technology perimetry. Opthalmol Clin North Am measured with the evoked potential. Invest
2003; 16: 213–25 Ophthalmol Vis Sci 1984; 25: 632–9
22. De Monasterio FM, Gouras P. Functional properties 36. Nordmann JP, Saraux H, Roullet E. Contrast sensi-
of ganglion cells of the rhesus monkey retina. J Physiol tivity in multiple sclerosis: a study in 35 patients with
1975; 251: 167–95 and without optic neuritis. Ophthalmologica 1987;
23. Wiesel TN, Hubel DH. Spatial and chromatic inter- 195: 199–204
actions in the lateral geniculate body of the rhesus 37. Regan D, Neima D. Low-contrast letter charts as a test
monkey. J Neurophysiol 1966; 29: 1115–56 of visual function. Ophthalmology 1983; 90: 1192–200
24. Hicks TP, Lee BB, Vidyasagar TR. The responses of 38. Regan D, Neima D. Low-contrast letter charts in early
cells in macaque lateral geniculate nucleus to sinu- diabetic retinopathy, ocular hypertension, glaucoma,
soidal gratings. J Physiol 1983; 337: 183–200 and Parkinson’s disease. Br J Ophthalmol 1984; 68:
25. Hendry SH, Yoshioka T. A neurochemically distinct 885–9
third channel in the macaque dorsal lateral geniculate 39. Baier ML, Cutter GR, Rudick RA, et al. Low-contrast
nucleus. Science 1994; 264: 575–7 letter acuity testing captures visual dysfunction in
26. Johnson JK, Casagrande VA. Distribution of calcium- patients with multiple sclerosis. Neurology 2005; 64:
binding proteins within the parallel visual pathways 992–5
of a primate (Galago crassicaudatus). J Comp Neurol 40. Bailey IL, Lovie JE. New design principles for visual
1995; 356: 238–60 acuity letter charts. Am J Optom Physiol Opt 1976; 53:
27. Hendry SH, Casagrande VA. A common pattern for a 740–5
third visual channel in the primate LGN. Soc Neurosci 41. Ferris FL, Kassoff A, Bresnick GH, et al. New visual
Abstr 1996; 22: 1605 acuity charts for clinical research. Am J Ophthalmol
28. Goodchild AK, Martin PR. The distribution of cal- 1982; 94: 91–6
cium-binding proteins in the lateral geniculate 42. Ferris FL, Bailey I. Standardizing the measurement of
nucleus and visual cortex of a New World monkey, visual acuity for clinical research studies: guideline
the marmoset, Callithrix jacchus. Vis Neurosci 1998; from the Eye Care Technology Forum.
15: 625–42 Ophthalmology 1996; 103: 181–2
29. Xu X, Ichida JM, Allison JD, et al. A comparison of 43. Balcer LJ, Galetta SL, Calabresi PA, et al. Natalizumab
koniocellular, magnocellular, and parvocellular reduces visual loss in patients with relapsing multiple
receptive field properties in the lateral geniculate sclerosis. Neurology 2007; 68: 1299–304
nucleus of the owl monkey (Aotus trivirgatus). 44. Talman LS, Bisker ER, Sackel DJ, et al. Longitudinal
J Physiol 2001; 531: 203–18 study of vision and retinal nerve fiber layer thickness
30. Pelli DG, Robson JG, Wilkins AJ. Designing a new in multiple sclerosis. Ann Neurol 2010; 67: 749–60
letter chart for measuring contrast sensitivity. Clin 45. Rubin GS, Roche KB, Prasada-Rao P, Fried LP. Visual
Vision Sci 1988; 2: 187–99 impairment and disability in older adults. Optom Vis
31. Leat SJ, Woo GC. The validity of current clinical Sci 1994; 71: 750–60
tests of contrast sensitivity and their ability to 46. Szylick JS, Seiple W, Fishman GA, et al. Perceived and
predict reading speed in low vision. Eye 1997; 11: actual performance of daily tasks: relationship to
893–9 visual function tests in individuals with retinitis pig-
32. Balcer LJ, Baier ML, Pelak VA, et al. New low-contrast mentosa. Ophthalmology 2001; 108: 65–75
vision charts: reliability and test characteristics in 47. Rubin GS, Legge GE. Psychophysics of reading. VI –
patients with multiple sclerosis. Mult Scler 2000; 6: The role of contrast in low vision. Vision Res 1989; 29:
163–71
70 79–91
 Chapter 6: Optical coherence tomography and low-contrast acuity

48. Whittaker SG, Lovie-Kitchin J. Visual requirements functional contrast sensitivity with decreasing light
for reading. Optom Vis Sci 1993; 70: 54–65 levels in monocular and binocular vision. Invest
49. Crossland MD, Rubin GS. Text accessibility by people Ophthalmol Vis Sci 2013; 54: 6093–103
with reduced contrast sensitivity. Optom Vis Sci 2012; 65. Pineles SL, Velez FG, Isenberg SJ, et al. Functional
89: 1276–81 burden of strabismus: decreased binocular summa-
50. Marron JA, Bailey IL. Visual factors and orientation- tion and binocular inhibition. JAMA Ophthalmol
mobility performance. Am J Optom Physiol Opt 1982; 2013; 131: 1413–19
59: 413–26 66. Gallo A, Esposito F, Sacco R, et al. Visual resting-state
51. Startzell JK, Owens DA, Mulfinger LM, Cavanagh PR. network in relapsing-remitting MS with and without
Stair negotiation in older people: a review. J Am previous optic neuritis. Neurology 2012; 79: 1458–65
Geriatr Soc; 48: 567–80 67. Collins JW, Carney LG. Visual performance in high
52. Freeman EE, Munoz B, Turano KA, West SK. myopia. Curr Eye Res 1990; 9: 217–23
Measures of visual function and their association with 68. Hess R, Woo G. Vision through cataracts. Invest
driving modification in older adults. Invest Opthalmol Ophthalmol Vis Sci 1978; 17: 428–35
Vis Sci 2006; 47: 514–20 69. Abrahamsson M, Sjostrand J. Impairment of contrast
53. Owsley C, Sekuler R, Boldt C. Aging and low-contrast sensitivity function (CSF) as a measure of disability
vision: face perception. Invest Ophthalmol Vis Sci glare. Invest Ophthalmol Vis Sci 1986; 27: 1131–6
1981; 21: 362–5 70. Gil MA, Varon C, Rosello N, et al. Visual acuity,
54. Owsley C, Sloane ME. Contrast sensitivity, acuity, contrast sensitivity, subjective quality of vision, and
and the perception of “real world” targets. quality of life with 4 different multifocal IOLs.
Br J Ophthalmol 1987; 71: 791–6 Eur J Ophthalmol 2012; 22: 175–87
55. McCulloch DL, Loffler G, Colguhoun K, et al. The 71. Freedman RD, Thibos LN. Contrast sensitivity in
effects of visual degradation on face discrimination. humans with abnormal visual experience. J Physiol
Ophthalmic Physiol Opt 2011; 31: 240–8 1975; 247: 687–710
56. Elliott DB, Hurst MA, Weatherhill J. Comparing 72. Rogers SA, Khan-Lim D, Manners RM. Does upper
clinical tests of visual function with the patient’s lid blepharoplasty improve contrast sensitivity?
perceived visual disability. Eye 1990; 4: 712–7 Ophthal Plast Reconstr Surg 2012; 28: 163–5
57. Lord S, Dayhew J. Visual factors for falls in older 73. Coe CD, Bower KS, Brooks DB, et al. Effect of blast
people. J Am Geriatr Soc 2001; 49: 508–15 trauma and corneal foreign bodies on visual perfor-
58. French D, Campbell R, Spehar A, et al. Drugs and falls mance. Optom Vis Sci 2010; 87: 604–11
in community dwelling older people: a national 74. Sjostrand J, Frisen L. Contrast sensitivity in macular
veterans study. Clin Therapeutics 2006; 28: 619–30 disease: a preliminary report. Acta Ophthalmol
59. Leat SJ, Legge GE, Bullimore MA. What is low vision? (Copenh) 197;55: 507–14
A re-evaluation of definitions. Optom Vis Sci 1999; 76: 75. Kleiner RC, Enger C, Alexander MF, Fine SL.
198–211 Contrast sensitivity in age-related macular degenera-
60. Mowry EM, Loquidice MJ, Daniels AB, et al. Vision tion. Arch Ophthalmol 1988; 106: 55–7
related quality of life in multiple scleorsis: correlation 76. Owsley C, Sloane ME, Skalka HW, et al. A compar-
with new measures of low and high contrast letter ison of the Regan Low-Contrast Letter Charts and
acuity. J Neurol Neurosurg Psychiatry 2009; 80: 767–72 contrast sensitivity testing in older patients. Clin Vis
61. Elliot DB, Whitaker D. How useful are contrast sen- Sci 1990; 5: 325–34
sitivity charts in optometric practice? Case reports. 77. Rolando M, Lester M, Macri A, Calabria G. Low
Optom Vis Sci 1992; 69: 378–85 spatial-contrast sensitivity in dry eyes. Cornea 1993;
62. Pineles SL, Birch EE, Talman LS, et al. One eye or two: 17: 376–9
a comparison of binocular and monocular low-con- 78. Kappel PJ, Monnet D, Yu F, et al. Contrast sensitivity
trast acuity testing in multiple sclerosis. among patients with birdshot chorioretinopathy.
Am J Ophthalmol 2011; 152: 133–40 Am J Ophthalmol 2009; 147: 351–6
63. Newman NJ, Wolfe JM, Steward MI, Lessell S. 79. Lindberg CR, Fishman GA, Anderson RJ, Vasquez V.
Binocular visual function in patients with a history Contrast sensitivity in retinitis pigmentosa.
of monocular optic neuritis. Clin Vision Sci 1991; 6: Br J Ophthalmol 1981; 65: 855–8
95–107 80. Yioti GC, Kalogeropoulos CD, Aspiotis MB,
64. Gillespie-Gallery H, Konstantakopoulou E, Harlow Stefaniotou MI. Contrast sensitivity and color vision
JA, Barbur JL. Capturing age-related changes in in eyes with retinitis pigmentosa and good visual
71
 Chapter 6: Optical coherence tomography and low-contrast acuity

acuity: correlations with SD-OCT findings. children with migraine? An exploratory study.
Ophthalmic Surg Lasers Imaging 2012; 43: S44-S53 Cephalalgia 2010; 30: 991–5
81. Midena E, Segato T, Bottin G, et al. The effect on the 96. Shepherd AJ, Beaumont HM, Hine TJ. Motion pro-
macular function of laser photocoagulation for dia- cessing deficits in migraine are related to contrast
betic macular edema. Graefes Arch Clin Exp sensitivity. Cephalalgia 2012; 32: 554–70
Ophthalmol 1992; 230: 162–5 97. Wall M. Contrast sensitivity testing in pseudotumor
82. Greenstein VC, Chen H, Hood DC, et al. Retinal cerebri. Ophthalmology 1986; 93: 4–7
function in diabetic macular edema after focal laser 98. Verplanck M, Kaufman DI, Parson T, et al.
photocoagulation. Invest Ophthalmol Vis Sci 2000; 41: Electrophysiology versus psychophysics in the detec-
3655–64 tion of visual loss in pseudotumor cerebri. Neurology
83. Stamper RL. The effect of glaucoma on central visual 1988; 38: 1789–92
function. Trans Am Ophthalmol Soc 1984; 82: 792–826 99. Buncie JR, Tytla ME. Contrast sensitivity in shunted
84. Johnson CA. Psychophysical measurement of glau- hydrocephalus. Invest Ophthalmol Vis Sci 1989; 30:
comatous damage. Surv Ophthalmol 2001; 45: S313- 407
S318 100. Wall M, Conway MD, House PH, et al. Evaluation of
85. Gandolfi SA. Improvement in spatial contrast sensi- sensitivity and specificity of spatial resolution and
tivity threshold after surgical reduction of intraocular Humphrey automated perimetry in pseudotumor
pressure in unilateral high-tension glaucoma. Invest cerebri patients and normal subjects. Invest
Ophthalmol Vis Sci 2005; 46: 197–201 Ophthalmol Vis Sci 1991; 32: 3306–12
86. Skrandies W, Gottlob I. Alterations of visual contrast 101. Lynch DR, Farmer JM, Rochestie D, Balcer LJ.
sensitivity in Parkinson’s disease. Hum Neurobiol Contrast letter acuity as a measure of visual dysfunc-
1986; 5: 255–9 tion in patients with Friedreich ataxia.
87. Kupersmith MJ, Shakin E, Siegel IM, et al. Visual J Neuroophthalmol 2002; 22: 270–4
system abnormalities in patients with Parkinson’s 102. Bodis-Wollner I. Visual acuity and contrast sensitiv-
disease. Arch Neurol 1982; 39: 284–6 ity in patients with cerebral lesions. Science 1972; 178:
88. Peppe A, Stanzione P, Pierelli F, et al. Low contrast 769–71
stimuli enhance PERG sensitivity to the visual dys- 103. Halasz I, Levy-Gigi E, Kelemen O, et al.
function in Parkinson’s disease. Electroencephalogr Neuropsychological functions and visual contrast
Clin Neurophysiol 1992; 82: 453–7 sensitivity in schizophrenia: the potential impact of
89. Langheinric T, Tebartz van Elst L, Lagreze WA, et al. comorbid posttraumatic stress disorder (PTSD).
Visual contrast response functions in Parkinson’s Front Psychol 2013; 4: 136
disease: evidence from electroretinograms, visually 104. Bubl E, Kern E, Ebert D, et al. Seeing gray when
evoked potentials and psychophysics. feeling blue? Depression can be measured in the eye
Clin Neurophysiol 2000; 111: 66–74 of the diseased. Biol Psychiatry 2010; 68: 205–8
90. Cronin-Golomb A, Corkin S, Rizzo JF, et al. Visual 105. Bubl E, Ebert D, Kern E, et al. Effect of antidepressive
dysfunction in Alzheimer’s disease: relation to nor- therapy on retinal contrast processing in depressive
mal aging. Ann Neurol 1991; 29: 41–52 disorder. Br J Psychiatry 2012; 201: 151–8
91. Gilmore GC, Levy JA. Spatial contrast sensitivity in 106. Santaella RM, Fraunfelder FW. Ocular adverse effects
Alzheimer’s disease: a comparison of two methods. associated with systemic medications: recognition
Optom Vis Sci 1991; 68: 790–4 and management. Drugs 2007; 67: 75–93
92. Sadun AA, Borchert M, DeVita E, et al. Assessment of 107. Sorri I, Kalviainen R, Mantyjarvi M. Color vision and
visual impairments in Alzheimer’s disease. contrast sensitivity in epilepsy patients treated with
Am J Ophthalmol 1987; 104: 113–20 initial tiagabine monotherapy. Epilepsy Res 2005; 67:
93. Barboni MT, Nagy BV, de Araujo Moura AL, et al. 101–7
ON and OFF electroretinography and contrast sen- 108. Boeckelmann I, Pfister EA. Influence of occupational
sitivity in Duchenne muscular dystrophy. Invest exposure to organic solvent mixtures on contrast
Ophthalmol Vis Sci 2013; 54: 3195–204 sensitivity in printers. J Occup Environ Med 2003; 45:
94. Ribeiro MJ, Violante IR, Bernardino I, et al. 25–33
Abnormal achromatic and chromatic contrast sensi- 109. Gong Y, Kishi R, Kasai S, et al. Visual dysfunction in
tivity in neurofibromatosis type 1. Invest Ophthalmol workers exposed to a mixture of organic solvents.
Vis Sci 2012; 53: 287–93 Neurotoxicology 2003; 24: 703–10
95. Braunitzer G, Rokszin A, Kobor J, Benedek G. Is the 110. Costa TL, Barboni MT, Moura AL, et al. Long-term
72 development of visual contrast sensitivity impaired in occupational exposure to organic solvents affects
 Chapter 6: Optical coherence tomography and low-contrast acuity

color vision, contrast sensitivity, and visual fields. 126. Bulens C, Meerwaldt JD, Van der Wildt GJ. Effect of
PLoS One 2012; 7: e42961 stimulus orientation on contrast sensitivity in
111. Owsley C. Aging and vision. Vision Res 2011; 51: Parkinson’s disease. Neurology 1988; 38: 76–81
1610–22 127. Katz B, Rimmer S, Iragui V, et al. Abnormal pattern
112. Mateus C, Lemos R, Silva MF, et al. Aging of low and electroretinogram in Alzheimer’s disease: evidence
high level vision: from chromatic and achromatic for retinal ganglion cell degeneration. Ann Neurol
contrast sensitivity to local and 3D object motion 1989; 26: 221–5
perception. PLoS One 2013; 8: e55348 128. Trick GL, Barris MC, Bickler-Bluth M. Abnormal
113. Jackson GR, Owsley C. Visual dysfunction, neurode- pattern electroretinogram in patients with senile
generative diseases, and aging. Neurol Clin 2003; 21: dementia of the Alzheimer’s type. Ann Neurol 1989;
709–28 26: 226–31
114. Curcio CA, Milican CL, Allen KA, et al. Aging of the 129. Koronyo Y, Salumbides BC, Black KL, Koronyo-
human photoreceptor mosaic: evidence for selective Hamaoui M. Alzeheimer’s disease in the retina: ima-
vulnerability of rods in central retina. Invest ging retinal Aβ plaques for early diagnosis and ther-
Ophthalmol Vis Sci 1993; 34: 3278–96 apy assessment. Neurodegener Dis 2012; 10: 1–9
115. Zhang J, Wang X, Wang Y, et al. Spatial and temporal 130. Armstrong RA. Visual field defects in Alzheimer’s
sensitivity degradation of primary visual cortical cells disease patients may reflect differential pathology in
in senescent rhesus monkeys. Eur J Neurosci 2008; 28: the primary visual cortex. Optom Vis Sci 1996; 73:
201–7 677–82
116. Yang Y, Liang Z, Liang Z, et al. Aging affects contrast 131. Hof PR, Morrison JH. Quantitative analysis of a vul-
response functions and adaptation of middle tem- nerable subset of pyramidal neurons in Alzheimer’s
poral visual area neurons in rhesus monkeys. disease: II. Primary and secondary visual cortex.
Neuroscience 2008; 156: 748–57 J Comp Neurol 1990; 301: 55–64
117. Adams CW, Bullimore MA, Wall M, et al. Normal 132. Leudba G, Saini K. Pathology of subcortical visual
aging effects for frequency doubling technology centres in relation to cortical degeneration in
perimetry. Optom Vis Sci 1999; 76: 582–7 Alzheimer’s disease. Neuropathol Appl Neurobiol
1995; 21: 410–22
118. Anderson AJ, Johnson CA, Fingeret M, et al.
Characteristics of the normal database for the 133. McKee AC, Au R, Cabral HJ, et al. Visual association
Humphrey Matrix Perimeter. Invest Ophthalmol Vis pathology in preclinical Alzheimer disease.
Sci 2005; 46: 1540–8 J Neuropathol Exp Neurol 2006; 65: 621–30
119. Harnois C, Di Paolo T. Decreased dopamine in the 134. Mielke R, Kessler J, Fink G, et al. Dysfunction of
retinas of patients with Parkinson’s disease. Invest visual cortex contributes to disturbed processing of
Ophthalmol Vis Sci 1990; 31: 2473–5 visual information in Alzheimer’s disease.
Int J Neurosci 1995; 82: 1–9
120. Nightingale S, Mitchell KW, Howe JW. Visual evoked
cortical potentials and pattern electroretinograms in 135. Blanks JC, Hinton DR, Sadun AA, Miller CA. Retinal
Parkinson’s disease and control subjects. J Neurol ganglion cell degeneration in Alzheimer’s disease.
Neurosurg Psychiatry 1986; 49: 1280–7 Brain Res 1989; 501: 364–72
121. Gottlob I, Scneider E, Heider W, et al. Alteration of 136. Curcio CA, Drucker DN. Retinal ganglion cells in
visual evoked potentials and electroretinograms in Alzheimer’s disease and aging. Ann Neurol 1993; 33:
Parkinson’s disease. Electroencephalogr Clin 248–57
Neurophysiol 1987; 66: 349–57 137. Hinton DR, Sadun AA, Blanks HC, Miller CA. Optic-
122. Reader TA, Quesney LF. Dopamine in the visual nerve degeneration in Alzheimer’s disease. N Engl J
cortex of the cat. Experimentia 1986; 42: 1242–4 Med 1986; 315: 485–7
123. Parkinson D. Evidence for a dopaminergic innerva- 138. Paquet C, Boissonnot M, Roger F, et al. Abnormal
tions of cat primary visual cortex. Neuroscience 1989; retinal thickness in patients with mild cognitive
30: 171–9 impairment and Alzheimer’s disease. Neurosci Lett
2007; 420: 97–9
124. Papadopoulos GC, Parnavelas JG. Distribution and
synaptic organization of dopaminergic axons in the 139. Parisi V, Restuccia R, Fattapposta F, et al.
lateral geniculate nucleus of the rat. J Comp Neurol Morphological and functional retinal impairment in
1990; 294: 356–61 Alzheimer’s disease patients. Clin Neurophysiol 2001;
112: 1860–7
125. Regan D, Maxner C. Orientation-selective visual loss
in patients with Parkinson’s disease. Brain 1987; 110: 140. Sadun AA, Bassi CJ. Optic nerve damage in
415–32 Alzheimer’s disease. Ophthalmology 1990; 97: 9–17 73
 Chapter 6: Optical coherence tomography and low-contrast acuity

141. Tsai CS, Ritch R, Schwartz B, et al. Optic nerve head 156. Zimmern RL, Campbell FW, Wilkinson IMS. Subtle
and nerve fiber layer in Alzheimer’s disease. disturbances of vision after optic neuritis elicited by
Arch Ophthalmol 1991; 109: 199–204 studying contrast sensitivity. J Neurol Neurosurg
142. Valenti DA. Neuroimaging of retinal nerve fiber layer Psychiatry 1979; 42: 407–12
in AD using optical coherence tomography. 157. Regan D, Raymond J, Ginsburg AP, Murray TJ.
Neurology 2007; 69: 1060 Contrast sensitivity, visual acuity and the discrimi-
143. Valenti DA. Alzheimer’s disease: visual system nation of Snellen letters in multiple sclerosis. Brain
review. Optometry 2010; 81: 12–21 1981; 104: 333–50
144. McDonald WI, Barnes. The ocular manifestations of 158. Patterson VH, Foster DH, Heron J, Mason RJ.
multiple sclerosis. 1. Abnormalities of the afferent Multiple sclerosis. Luminance threshold and mea-
visual system. J Neurol Neurosurg Psychiatry 1992; 55: surements of temporal characteristics of vision. Arch
747–52 Neurol 1981; 38: 687–9
145. Leibowitz U, Alter M. Optic nerve involvement and 159. Kupersmith MJ, Nelson JI, Seiple WH, Carr RE,
diplopia as initial manifestations of multiple sclerosis. Weiss PA. The 20/20 eye in multiple sclerosis.
Clin Neurosci 1994; 2: 180–8 Neurology 1983; 33: 1015–20
146. Balcer LJ. Optic Neuritis. N Engl J Med 2006; 354: 160. Pace R, Woo GC. Contrasta sensitivity function for
1273–80 vision testing in suspected demyelinating disease.
Lancet 1984; 8373: 405–6
147. Ulrich J, Groebke-Lorenz W. The optic nerve in
multiple sclerosis. A morphological study with 161. Ashworth B, Aspinall PA, Mitchell JD. Visual func-
retrospective clinic-pathological correlations. tion in multiple sclerosis. Doc Ophthalmology 1989;
Neuro-ophthalmology 1983; 3: 149–59 73: 209–24
148. Lumsden CE. The neuropathology of multiple 162. Honan WP, Heron JR, Foster DH, et al. Visual loss in
sclerosis. In: Vinken PJ, Bruyn GW, editors. Handbook multiple sclerosis and its relation to previous optic
of Clinical Neurology New York: Elsevier 1970. neuritis, disease duration and clinical classification.
p. 161–216 Brain 1990; 113: 975–87
149. Kurtzke JF. Rating neurologic impairment in multi- 163. Russell MHA, Murray IJ, Metcalfe RA, Kulikowski JJ.
ple sclerosis: an Expanded Disability Status Scale The visual defect in multiple sclerosis and optic
(EDSS). Neurology 1983; 33: 1444–52 neuritis. Brain 1991; 114: 2419–35
150. Rudick R, Antel J, Confavreux C, et al. 164. Leys MJ, Candaele CM, De Rouck AF, Odom JV.
Recommendations from the National Multiple Detection of hidden visual loss in multiple sclerosis.
Sclerosis Society Clinical Outcomes Assessment Task A comparison of pattern-reversal visual evoked
Force. Ann Neurol 1997; 42: 379–82 potentials and contrast sensitivity. Doc Opthalmol
1991; 77: 255–64
151. Miller DM, Rudick RA, Cutter G, et al. Clinical sig-
nificance of the multiple sclerosis functional compo- 165. Trobe JD, Beck RW, Moke PS, Cleary PA. Contrast
site: relationship to patient-reported quality of life. sensitivity and other vision tests in the Optic
Arch Neurol 2000; 57: 1319–24 Neuritis Treatment Trial. Am J Ophthalmol 1996;
121: 547–53
152. Rudick RA, Cutter G, Reingold S. The multiple
sclerosis functional composite: a new clinical out- 166. Beck RW, Gal RL, Bhatti MT, et al. Optic Neuritis
come measure for multiple sclerosis trials. Mult Scler Study Group. Visual function more than 10 years
2002; 8: 359–65 after optic neuritis: experience of the Optic
Neuritis Treatment Trial. Am J Ophthalmol 2004;
153. Balcer LJ, Baier ML, Cohen JA, et al. Contrast letter 137: 77–83
acuity as a visual component for the Multiple
Sclerosis Functional Composite. Neurology 2003; 61: 167. Wender M. Value of Pelli-Robson contrast sensi-
1367–73 tivity chart for evaluation of visual system in mul-
tiple sclerosis patients. Neurol Neurochir Pol 2007;
154. Fisher JB, Jacobs DA, Markowitz CE, et al. Relation 41: 141–3
of visual function to retinal nerve fiber layer
thickness in multiple sclerosis. Ophthalmology 168. Wu GF, Schwartz ED, Lei T, et al. Relation of vision to
2006; 113: 324–32 global and regional brain MRI in multiple sclerosis.
Neurology 2007; 69: 2128–35
155. Bock M, Brandt AU, Kuchenbecker J, et al.
Impairment of contrast visual acuity as a functional 169. Optic Neuritis Study Group. Visual function 15 years
correlate of retinal nerve fibre layer thinning and total after optic neuritis: a final follow-up report from the
macular volume reduction in multiple sclerosis. Optic Neuritis Treatment Trial. Ophthalmology 2008;
Br J Ophthalmol 2012; 96: 62–7 115: 1079–82
74
 Chapter 6: Optical coherence tomography and low-contrast acuity

170. Keltner JL, Johnson CA, Cello KE, et al. Optic 173. Walter SD, Ishikawa H, Galetta KM, Sakai RE, Feller
Neuritis Study Group. Visual field profile of optic DJ, Wilson JA, Maguire MG, Galetta SL, Frohman E,
neuritis: a final follow-up report from the Optic Calabresi PA, Schuman JS, Balcer LJ. Ganglion cell
Neuritis Treatment Trial from baseline through 15 loss in relation to visual disability in multiple sclero-
years. Arch Ophthalmol 2010; 128: 330–7 sis. Ophthalmology 119: 1250–57
171. Graves J, Galetta SL, Palmer J, et al. Alemtuzumab 174. Burkholder BM, Osborne B, Loguidice MJ, Bisker
improves contrast sensitivity in patients with relap- E, Frohman TC, Conger A, Ratchford JN, Warner
sing-remitting multiple sclerosis. Mult Scler 2013; 19: C, Markowitz CE, Jacobs DA, Galetta SL, Cutter
1302–9 GR, Maguire MG, Calabresi PA, Balcer LJ,
172. Reich DS, Smith SA, Gordon-Lipkin EM, et al. Frohman EM. Macular volume determined by
Damage to the optic radiation in multiple sclerosis is optical coherence tomography as a measure of
associated with retinal injury and visual disability. neuronal loss in multiple sclerosis. Arch Neurol
Arch Neurol 2009; 66: 998–1006 2009; 66: 1366–72

75
 Chapter
Optical coherence tomography

7 and electrophysiology of the

visual pathway
A. Klistorner, C. L. Fraser, C. Yiannikas, and S. L. Graham

While modern spectral-domain OCT technology can correlation between the responses measured and the
deliver visualization of individual retinal layers, anatomical appearance on OCT.
allowing for a substantial shift in our knowledge of
both ophthalmic and neurological conditions, it pro- Anatomical origin of the full-field
vides only structural information about the retina. To
fully understand the visual system, structure and electroretinogram
function should be examined together. The electroretinogram (ERG) represents the electrical
Visual electrophysiology tests represent an objective response of different cellular elements of the retina to
way to assess the function of the visual pathway. The light. The ERG is a simple clinical test that can be
visual system can be interrogated using a variety of performed with the use of corneal electrodes. ISCEV
testing paradigms, which have been standardized by full-field ERG stimuli are presented using the ganzfeld
the International Society for Clinical Electrophysiology dome, which results in even illumination across the
of Vision (ISCEV). The most commonly performed tests entire retina in patients with dilated pupils. The
include the full-field electroretinogram (ERG), pattern electrical responses recorded during the ERG are
ERG (PERG), and visual evoked potentials (VEP), which generated by the retinal neurons directly or by
provide a summed response from the retinal elements changes in the retinal glia secondary to the activity
being tested. Additional topographical information can of the retinal neurons.
be assessed through the use of multifocal technologies, The relative contributions of various retinal types
such as a multifocal ERG (mfERG) and multifocal VEP are different under dark-adapted (scotopic) and light-
(mfVEP). The series of tests allow clinical assessment of adapted (photopic) conditions or different levels of
conditions including suspected retinal dystrophy, stimulus intensity. The rod photoreceptors are
neurological disease or unexplained visual loss. involved in the generation of responses under scoto-
By coupling the information from OCT and elec- pic conditions, when very dim blue light is used for
trophysiology, it is possible to perform structural and stimulation. Cone photoreceptors require much
functional assessment of the various cell populations brighter stimulus and are better recorded under
within the retina, as well as the optic nerve fibers. This photopic conditions. Full-field flashes stimulate the
combination of tests is most commonly used in the entire retina and are used to elicit the major waveform
diagnosis and management of retinal disease, which is responses in the typical ERG, providing diagnostic
beyond the scope of this chapter. Instead we will information about rod and cone photoreceptors and
concentrate on the use of these techniques in neuro- inner retinal function across the entire retina, but
logical conditions, predominantly multiple sclerosis. with no localizing information.
A typical ERG protocol includes the following:
Anatomical substrate of clinical after dark-adaptation, a dim scotopic stimulus is
presented, resulting in a pure rod response. This is
electrophysiology recordings the only electrophysiology test that is rod specific.
An understanding of the anatomical origins of the While the patient remains dark-adapted, stimuli of
measured electrophysiological signals allows better increasing intensity generate combined dark-adapted

76 Optical Coherence Tomography in Neurological Disease, ed. Peter A. Calabresi, Laura J. Balcer, and Elliot M. Frohman.
Published by Cambridge University Press. © Cambridge University Press 2015.
 Chapter 7: Optical coherence tomography and electrophysiology of the visual pathway

cone and rod responses. Then, after a period of light illumination), the PERG can be used diagnostically to
adaptation, a single flash stimulus drives a pure differentiate macular damage from diffuse retinal
cone response. pathology.
The PERG consists of two peaks: an initial positive
ERG a-wave deflection (P50) occurs at approximately 50 millise-
conds with a following negative deflection occurring
ERG waveforms show an initial negative deflection
at approximately 95 milliseconds (N95). It is believed
(a-wave) originating from the photoreceptor layer.
that the second component largely arises from retinal
The photoreceptor response to light absorption in
ganglion cells spiking output activity.
the outer segments results in hyperpolarization.
Reduction in the P50 amplitude, or an increase in
The a-wave is not seen, however, in the rod specific
the P50 latency, suggests macular dysfunction ante-
stimulus conditions (fully dark-adapted with dimmer
rior to the ganglion cells [1]. A reduction in P50
flash).
usually leads to a consecutive reduction of the N95
component. The ratio between the P50 and N95 is,
ERG b-wave therefore, not reduced in these cases.
The first positive deflection, called the b-wave, origi- Abnormalities confined to the N95 component,
nates primarily from optic nerve (ON) bipolar cells, leaving P50 unaffected, are seen in diseases of the
with a small contribution from the Muller cells. These optic nerve. Reported examples include optic nerve
ON bipolar cells are driven by the stimulated demyelination, Leber’s Hereditary Optic Neuropathy,
photoreceptors. and autosomal dominant optic atrophy [2].
Oscillatory potentials can be recognized as multiple In summary, the PERG is driven by macular
wavelets on the up-slope of the b-wave (and isolated photoreceptors but its second component arises
with appropriate high pass filters). Their exact origin predominantly from the retinal ganglion cells.
and clinical significance is still debated. The ratio of
b-to a-wave in bright flash conditions can be used as a Anatomical origin of the multifocal
marker of inner retinal pathology, with a reduction
typical of such conditions as melanoma associated electroretinogram
retinopathy (MAR) or carcinoma associated retinopa- The multifocal electroretinogram (mfERG) allows
thy (CAR), but also in more benign conditions such as simultaneous but independent stimulation of multiple
congenital stationary night blindness. retinal areas. The stimulus is presented on a computer
In summary, the standard ERG assesses the photo- monitor and change in illumination of individual areas
receptor and inner nuclear layer, and its integrity is driven by pseudo-random sequences. Responses are
determines the function of the retina through to the derived from a single continuous ERG recording using
ganglion cells. cross-correlation mathematical techniques. The stimu-
lus pattern is designed to produce local responses of
Anatomical origin of the pattern approximately equal amplitudes (i.e., the stimulus area
is scaled to be slightly larger with eccentricity).
electroretinogram The typical mfERG is a cone-driven response from
The overall response from the retina produced by a radius of 20–30 degrees around the fovea. The rods
alternating black and white checkerboard pattern do not contribute to the mfERG. Like the standard
stimulation without a net change in stimulus lumi- ERG, the mfERG is formed by bipolar cell activity
nance (rather than diffuse flash) is called the pattern with smaller contributions from the photoreceptors.
electroretinogram (PERG). It has been suggested that mfERG responses are largely independent of
the retinal ganglion cells are the prime generators of refractive error, but they are critically dependent
this response. on good fixation and patient cooperation. Patients
As the macula comprises less than 5% of the total with poor vision typically have poor fixation, making
retinal area, maculopathy usually does not affect the test interpretation difficult.
full-field ERG response. The PERG stimulus, how- The mfERG is used as an alternative to PERG as a
ever, is projected directly to the macula. With no test of macular function. The information is used in
stray light (since there is no change in total retinal association with full-field ERG and VEP in order to 77
 Chapter 7: Optical coherence tomography and electrophysiology of the visual pathway

localize the cause of visual loss to a widespread retinal less variable between subjects than its amplitude.
abnormality, macular abnormality, or optic nerve There is often a difference in nomenclature between
disease. The localizing information from mfERG is ophthalmologists and neurologists. Ophthalmologists
helpful in diagnosing conditions with specific macular record an initial positive deflection at approximately
involvement such as hydroxychloroquine (Plaquenil) 100 milliseconds, referred to as P100, whereas neurol-
maculopathy. ogists often use reverse polarity and record the same
response as N100. P100/N100 amplitude is reduced in
many different optic nerve disorders. P100 latency
Anatomical origin of the visual delay has been described as typically occurring in
evoked potential demyelination.
The visual evoked potential (VEP) represents the
response of the visual cortex to optical stimulation. Multifocal visual evoked potential
It was proposed as a means of assessing the integrity of The multifocal VEP (mfVEP) is a cortical response to
the visual pathway in optic neuritis (ON) more than simultaneous but independent stimulation of multi-
30 years ago [3], and is understood to be generated at ple areas of the visual field. As mentioned earlier, the
the level of striate cortex by combined activity of post- conventional full-field VEP is greatly dominated by
synaptic potentials [4]. The magnitude of the VEP the macular response due to its cortical overrepre-
reflects the number of functional afferent fibers reach- sentation [9]. Being the vector sum of numerous
ing striate cortex and the degree of synaptic activity in differently oriented dipoles, which are often opposite
V1. Therefore, its amplitude is determined by the due to the cortical anatomy of the striate
severity of the acute inflammation along the visual cortex around the calcarine sulcus, the waveform of
pathway and subsequent axonal degeneration [5]. the full-field VEP is also prone to cancellation and
The timing of the response (latency) reflects the distortion. This can result in under– or overestima-
degree of demyelination along the visual pathway. tion of amplitude and detection of apparent rather
Delayed VEP latency reflects a reduction of conduc- than real latency delay [10]. In contrast, the mfVEP
tion speed due to a change from saltatory to contin- enables simultaneous recording from multiple
uous membrane conduction after demyelination [6]. regions of the visual field, allowing more detailed
Subsequent shortening of VEP latency is thought topographical assessment of small groups of axons
to represent the process of optic nerve remyelination within the optic nerve. Due to the small size of the
[5, 7]. Although remyelination may not be solely stimulated cortical area, mfVEP also eliminates
responsible for the recovery of VEP latency, as other cancellation effects, providing real representation of
factors such as reorganization of the ion channel amplitude and latency [11–15].
distribution and cortical plasticity may also play a The size of individual segments is cortically scaled
role, it is believed to be the major factor [8]. The to stimulate approximately equal cortical areas (simi-
VEP signal is typically recorded at the scalp overlying lar to the mfERG, but with more substantial change in
the occiput. size with eccentricity) (Figure 7.1). Therefore, the
Since central vision is disproportionally repre- mfVEP is ideal to study spatial and temporal evolu-
sented in the striate cortex, the full-field VEP is a tion of optic nerve demyelinated lesions with high
macular-dominated response. A normal macula is precision.
necessary for normal VEP recordings, and it is pru- Similarly to full-field VEP recordings, mfVEP
dent to prove normal macula structure and function responses are dependent on the integrity of the entire
before attributing abnormalities in VEP to optic nerve visual pathway.
pathology.
The two most common ways to elicit a cortical Applications in optic neuritis
response are flash and pattern stimulation. The Flash
VEP (response to bright flash) has significant inter- and multiple sclerosis
individual test variability, but it is useful in infants The advent of VEPs in the 1970s first allowed
and patients with nystagmus. Pattern-reversal (check- clinicians to assess neural conduction in the optic
erboard or gratings) VEP is less variable between nerve. Halliday demonstrated that diagnostic confir-
78 subjects than flash VEP, with the P100 latency being mation of optic neuritis can be made using latency
 Chapter 7: Optical coherence tomography and electrophysiology of the visual pathway

Table 7.1 Summary of electrophysiological tests

Test Stimulus Signal generation Visual region

Full-field Diffuse flash using Ganzfeld Photoreceptors (rods,cones), Pan-retinal
ERG dome bipolar cells
Pattern Reversing black/white Macular photoreceptors, Macula
ERG checkerboard; retinal ganglion cells
15 o and 30 o field size
Pattern Reversing black/white Cortical neurons within Large field, macular
VEP checkerboard; ≥15 o primary visual cortex dominated
Multifocal Flickering black/white hexagons; Cone photoreceptors and Topographical responses
ERG 40o-50o diameter bipolar cells within macula
Multifocal Reversing black/white dartboard; Cortical neurons within Topographical responses
VEP 60o diameter primary visual cortex within visual field

develop MS, with the presence of typical MRI lesions

in the brain at the time of ON onset being a strong risk
factor [17]. The natural history of acute ON mirrors
that of acute MS relapses elsewhere in the central
nervous system [18]. However, in contrast to most
brain lesions, the effects of disease on the optic nerve
are readily clinically apparent and potentially measur-
able [19]. By obtaining mfVEPs as a measure of inflam-
mation and conduction speed, in parallel with detailed
clinical measures of visual function and OCT measure-
ments of retinal layers as a more direct assessment of
Figure 7.1 Multifocal stimulus employed in mfVEP recording. axonal damage, it is possible to develop a picture of
Each sector (area 4 × 4 checks) reverses polarity according to a pathophysiological events produced by individual
pseudo-random sequence. Temporal sector (T) corresponds to lesions as they progress. [20] Therefore, ON represents
central visual field, superior sector (S)-to inferior visual field, and
inferior sector (I)-to superior visual field. an ideal in vivo model for studying the relationship
Reprinted with permission from Klistorner et al., Multifocal VEP between inflammation, axonal loss, and de/remyelina-
And OCT in optic neuritis: a topographical study of the structure- tion following single inflammatory attack [21].
function relationship, Doc Ophthalmol, 2009; 118: 129–37.
It is understood that transection of optic nerve
axons during acute inflammatory demyelination leads
delays of full-field VEP [3, 16]. Due to advances in to retrograde degeneration, which ultimately reaches
neuroimaging techniques, such as magnetic reso- the RNFL and can be accurately measured by OCT
nance imaging (MRI), the VEP was neglected for providing a structural measure of axonal loss. The
many years. However, recently the focus in multiple amplitude of the mfVEP, on the other hand, provides
sclerosis (MS) research is shifting back to structural a functional estimation of preserved optic nerve fibers.
and functional correlation based on OCT and electro- Therefore, application of both techniques allows com-
physiology of the optic nerve and retina. prehensive assessment of the effect of inflammatory
demyelination on visual pathway [5].
Early work exploring the links between optic nerve
Studies of optic neuritis evolution structure and function after acute ON in MS (MSON)
Optic neuritis (ON) is an inflammatory optic neuro- produced inconclusive results.
pathy which is strongly associated with MS. Up to 75% The first study comparing OCT and conventional 79
of ON patients have been reported to eventually VEP was performed by Parisi et al. in 2001 [22], who
 Chapter 7: Optical coherence tomography and electrophysiology of the visual pathway

reported significantly delayed P100 latencies and between RNFL thickness and mfVEP amplitude [26,
reduced N75 to P100 amplitudes in the eyes of 27]. To minimize the effect of inter-subject variability,
MS patients with documented ON (MSON) when between-eye asymmetry of RNFL thickness and
compared with control and fellow eyes. There was amplitude of the mfVEPs were analyzed. There was
also a significant reduction of total and temporal also good topographical association between two
RNFL thickness in MSON eyes compared with the measures (Figure 7.2) [28].
values observed in control eyes. The authors, how- Reduction of mfVEP amplitude, however, was
ever, did not find a significant correlation between considerably greater compared to RNFL thinning,
VEP (both amplitude and latency) and RNFL thick- indicating that the functional deficit may be greater
ness. The study only included 14 patients and an early than structural loss. Some possible explanations were
model of OCT with lower resolution was used. suggested, including:
A study published by Trip et al. [23] in 2005 using a. larger functional loss from channel failure on the
higher-resolution OCT demonstrated a significant denuded, but structurally preserved axons [28];
correlation of RNFL thickness with amplitude of the b. ongoing inflammation in the visual pathway, pos-
full-field VEP. Time since onset of optic neuritis in sibly as part of a more diffuse, slow inflammation
this group was three years, and patients with a in the white matter [29], which may reduce
previous optic neuritis episode with or without MS mfVEP amplitude due to axonal dysfunction or
were included. conduction block in some fibers.
Pueyo et al. [24] in 2008 confirmed a correlation
A longitudinal study performed within the first
between RNFL thickness (temporal sector in particular)
year of acute ON, however, demonstrated a more
and full-field VEP amplitude in a group of 50 MS
complex relationship between functional and struc-
patients. This study also demonstrated a significant
tural measures of optic nerve integrity. This study
correlation of RNFL with VEP latency, which was
showed that while during the acute stage of optic
explained by the fact that, contrary to Trip et al., all
neuritis both RNFL thickness and mfVEP amplitude
study patients had clinically definite multiple sclerosis
are influenced by the severity of inflammation, its
(CDMS), which is more often associated with demyeli-
effect on structural and functional measures is quite
nating processes. Data from the same group published
different. Acute loss of myelin and associated inflam-
two years later (2010) and based on a smaller number of
mation cause conduction block, which reduces or
patients [25] also demonstrated significant latency
completely abolishes mfVEP amplitude. At the same
increase and significant amplitude decrease of VEP
time, inflammatory edema often results in thicker
P100 in MS patients. However, while good correlation
swollen RNFL [30–34]. This leads to a complete struc-
was reported between P100 latency and OCT mean
ture–function dissociation (Figure 7.3a).
RNFL thickness, no correlation was shown between
Conduction block normally recovers within a few
RNFL and VEP amplitude.
weeks, during which inflammation subsides, ion
This discrepancy of structure–function relationships
channels are reconstructed and conduction partially
may be attributed to several factors: first, sensitivity of
resumes, although often in a slower, continuous mode
time-domain OCT devices used in earlier studies was
[28]. This usually leads to recovery of amplitude up to
low; second, there was significant difference in the
the level supported by surviving fibers. Therefore,
cohorts of patients studied, some studies including
after conduction block is over, the magnitude of the
CIS, while others only included CDMS. Additionally,
mfVEP signal becomes mainly dependent on the
and maybe most importantly, full-field VEP recording
degree of axonal transection.
has intrinsic problems described in the section above.
After resolution of edema, axonal degeneration
Therefore, to clarify the structure-function
becomes the predominant pathological feature affect-
relationship in the visual system of MS patients with
ing RNFL thickness. However, it takes time for axons
and without ON, we have undertaken several studies
transected during acute inflammation to degenerate
comparing structural changes in the optic nerve using
back to the retina. Therefore, a mild degree of correla-
OCT RNFL thickness compared with latency and
tion seen at three months is mainly due to time lag
amplitude of mfVEP [26, 27, 30]. An initial cross-
between axonal transection and retrograde degenera-
sectional investigation of patients with post-acute
tion reaching the retina (Figure 7.3b). By six months,
80 optic neuritis demonstrated a strong correlation
however, this process is completed and the direct
 Chapter 7: Optical coherence tomography and electrophysiology of the visual pathway

Microns RNFL thickness RE

300
124
200 S
71 T N 90
100
I
0 134
0 20 40 60 80 100 120 140 160 180 200 210 240
TEMP SUP NAS INF TEMP

RNFL thickness LE
Microns
300 58

200 S
58 N T 51
100 I

125
0
0 20 40 60 80 100 120 140 160 180 200 210 240
TEMP SUP NAS INF TEMP

MfVEP amplitude RE

33° 33° 24°

MfVEP amplitude LE

33° 33° 24°

Figure 7.2 Example of topographic correspondence between RNFL thickness and mfVEP amplitude. Significant thinning of RNFL in upper
sector and reduction of mfVEP amplitude in the lower area of the visual field is detected in left eye. For mfVEP, the first column represents VEP
trace arrays; the second column, the Amplitude Deviation plot; and the third column, the Amplitude Asymmetry Deviation plot.
Reprinted with permission from Klistorner et al., Multifocal VEP and OCT in optic neuritis: a topographical study of the structure–function
relationship, Doc Ophthalmol, 2009; 118: 129–37.

relationship between thinner RNFL and lower mfVEP pathological basis of the RGC axonal and neuronal
amplitudes becomes apparent (Figure 7.3c). loss in the MS-NON eyes is not clear. Therefore, a
combination of functional assessment of the visual
pathway with structural OCT measurement may help
Studies of the fellow eye in MS patients to better understand the nature of this loss. Latency of
An increasing number of studies have demonstrated mfVEP can be used as a marker of subclinical inflam-
significant axonal and neuronal loss of retinal ganglion matory demyelination along the visual pathway.
cells (RGC) in MS patients with no previous ON (MS- Using this approach, we assessed the relationship
NON). A recently published meta-analysis showed an between loss of RNFL and mfVEP in NON-eyes of MS
average RNFL thinning of 7 μm (about 7%) in those patients in early disease. Figure 7.4, which is based on
eyes [35]. Talman et al. [36] presented evidence that a compilation of results from several studies, shows 81
this loss may even be progressive. However, the the percentage of patients with abnormally thin RNFL
 Chapter 7: Optical coherence tomography and electrophysiology of the visual pathway

A B
200 140

mfVEP amplitude asymmetry

mfVEP amplitude asymmetry

180 R2 = 0.01 R2 = 0.19

120
160
140 100
120 80
100
80 60
60 40
40 B
20
20 B A
0 0
0 10 20 30 –20 0 20 40
RNFL thickness asymmetry RNFL thickness asymmetry

C D
120
mfVEP amplitude asymmetry

120
mfVEP amplitude asymmetry
R2 = 0.53 R2 = 0.51
100 100
80 80
60 60
40 40
D
20 20

0 0
0 10 20 30 40 50 –20 0 20 40 60
RNFL thickness asymmetry RNFL thickness asymmetry
Figure 7.3 Correlation between inter-eye asymmetry of RNFL thickness and mfVEP amplitude at (A) 1 month, (B) 3 months, (C) 6 months, and
(D)12 months.
Reprinted with permission from A. Klistorner et al., Interrelationship of optical coherence tomography and multifocal visual-evoked
potentials after optic neuritis. Invest Ophthalmol Vis Sci 2010; 51: 2770–7
The Association for Research in Vision and Ophthalmology is the copyright holder.

and delayed mfVEP latency compared to age- and 70%

mfVEP
gender-matched normal controls. 60%
Only CIS patients (acute unilateral ON) were OCT
50%
examined at one month after the onset of symptoms.
All patients, however, had brain or spinal cord demye- 40%
linating lesions detected by magnetic resonance
30%
imaging (MRI) and were classified as high risk for
development of MS [37]. A mixed group of CIS and 20%
CDMS patients were investigated at 12 and 36 months
10%
after the first demyelinating episode [26, 38], while
only CDMS patients were included in the last group 0%
(54 months) [39]. 1 month 12 36 54
months months months
Taken together, these results demonstrate early
change of conduction along the visual pathway Figure 7.4 Percentage of non-optic neuritis eyes with delayed
mfVEP latency (blue) and reduced RNFL thickness (red).
(which is already obvious in CIS patients) possibly
caused by subclinical inflammation with loss of RGC
axons to follow. temporal RNFL thickness in NON eyes (including
The last study is of particular interest since it fellow eyes of ON patients). Even more interestingly,
82 demonstrated not only significant latency delay of both total and temporal RNFL thickness was asso-
mfVEP, but also significant reduction of total and ciated with latency of the mfVEP (Figure 7.5a, b).
 Chapter 7: Optical coherence tomography and electrophysiology of the visual pathway

A B
190 190
R2 = 0.43 R2 = 0.37
MF VEP latency, ms

MF VEP latency, ms
180 180
170 170
160 160
150 150
140 140
130 130
120 120
50 60 70 80 90 100 110 120 30 40 50 60 70 80 90 100
Total RNFL thickness, µm Temporal RNFL thickness, µm

Figure 7.5 Correlation between mfVEP latency and (A) total RNFL thickness and (B) temporal RNFL thickness.
Reprinted with permission from A. Klistorner et al., Axonal loss in non-optic neuritis eyes of MS patients linked to delayed visual evoked
potential. Neurology 2013; 80: 245.

When grouped according to latency delay, NON eyes may also contribute to degeneration of chronically
with normal latency showed no reduction of RNFL demyelinated axons [33]
thickness compared to controls, while eyes with While remyelination is now accepted as an early
delayed latency demonstrated significantly thinner and frequent phenomenon in MS [41, 42], it is often
RNFL. MS-NON eyes with delayed latency also had incomplete [43] and probably limited in its duration
significantly thinner RNFL compared to those with [44]. It has previously been demonstrated that speed of
normal latencies. In patients with no previous ON in VEP latency recovery significantly slows down within
either eye, delayed latency and reduced RNFL were the first year after ON and practically does not change
bilateral whenever present. The binocular nature of after that [8, 45]. Therefore, latency remains consider-
latency delay suggests a retro-chiasmal location of ably prolonged in a significant proportion of patients
demyelinating lesions. However, whether or not even a few years after the demyelinating episode, indi-
axonal loss is the result of retrograde degeneration cating existence of extensive chronic demyelination.
triggered by optic tract lesions or caused by Using mfVEP to characterize the level of optic
trans-neuronal spread of neurodegeneration from nerve demyelination [46, 47] and optical coherence
LGN neurons to RGC axons following optic radiation tomography to quantify RNFL thickness, it is possible
or cortical lesions, remains to be seen. to investigate the association between the degree of
permanent demyelination and progressive axonal loss
Chronic demyelination and neurodegeneration in a visual system of MS patients. We followed a group
While demyelination is still considered the most of 25 patients with clinically isolated acute unilateral
characteristic histopathological feature of MS, a sig- ON for three years after onset of ON using both tech-
nificant association has been found between perma- niques. All patients had demyelinating MRI lesions of
nent functional deficit in MS and loss of axons, and it the brain or spine on initial MRI scans and were classi-
is now believed that axonal degeneration constitutes fied at enrollment as “possible MS.” During the follow-
the basis for neurological disability. up period, nine patients converted to CDMS.
It was suggested that permanent demyelination in This study demonstrated significant latency delay
the absence of active inflammation may contribute to of mfVEP at 12 months in ON eyes as compared to
axonal degeneration [40] by making axons more normal controls. Latency delay ranged from 4 to
vulnerable to physiological stress. One of the potential 39 ms, indicating a high degree of demyelination
mechanisms of axonal damage due to chronic demye- for some patients. There was a small, but statistically
lination may be related to ischemic insult of white significant, latency recovery over the follow-up
matter and involves an imbalance between axonal period, but it remained extensively delayed even at
energy demand and limited energy supply. Lack of 36 months. Fellow eyes also demonstrated significant
trophic support from myelin or myelin-forming cells latency prolongation at 12 months, although on a
and disruption of normal axon–myelin interaction much lesser scale. Latency of the fellow eye then 83
 Chapter 7: Optical coherence tomography and electrophysiology of the visual pathway

remained stable during entire follow-up period. subclinical inflammation or retro-chiasmal inflamma-
RNFL thickness was significantly reduced in ON tory demyelination with retrograde changes.
eyes at study entry point (12 months) indicating After recovery of acute conduction block, which
substantial axonal loss caused by axonal transection typically takes 4–8 weeks, axonal degeneration
during acute inflammatory demyelination and sub- becomes the predominant pathological process deter-
sequent retrograde degeneration. RNFL thickness of mining amplitude of the VEP, since loss of optic nerve
the fellow eye, however, was not affected. fibers limits the transmission of the afferent stimulus to
There was a small, but significant, reduction of RNFL the visual cortex [51]. It is expected, therefore, that in
thickness between 12 and 36 months for both ON and the post-acute period changes in amplitude of mfVEP
fellow eyes. During the follow-up period RNFL thickness should fall in parallel with changes in the optic nerve
decreased on average by 1.2 µ in ON and 1.3 µ in the structural integrity measure (RNFL thickness). In other
fellow. However, this change in RNFL thickness was not words, that progressive axonal degeneration should
related to the degree of latency delay at 12 months for lead to progressive decline of the mfVEP amplitude.
both ON and fellow eyes, indicating no association A recent study, however, presented a different
between the degree of permanent optic nerve demyeli- picture – contrary to expectation and despite the loss
nation and progressive axonal loss and raising a question of RNFL fibers, significant improvement of the
about any potential role of chronic demyelination in mfVEP amplitude was observed between six and 12
axonal neurodegeneration, at least in the early stages of months after an episode of acute ON. It was suggested
the disease. The fact that fellow eyes demonstrated a that the findings of increased post-synaptic activity in
similar degree of RNFL thinning, while only showing the striate cortex (as measured by improvement of
marginal demyelination, supports this conclusion [48]. the mfVEP amplitude) coupled with simultaneous
reduction of cortical input (RNFL thinning) supports
Pattern ERG studies in MS the concept of continuous cortical reorganization in
Parisi et al. [22] reported significant reduction of PERG post-acute optic neuritis [52].
P50 and N95 amplitudes, as well as a delay of the P50 Neural reorganization has been described after
latency, in MSON eyes when using 15-min arc checks acute optic neuritis and is considered to be one of
stimulation. In addition, the OCT temporal RNFL the mechanisms responsible for recovery of function
thickness was significantly correlated with the PERG from the acute attack. Various mechanisms including
P50 latency and N95 amplitude. The study was small increased expression of sodium channels, neuronal or
(14 patients) and employed an early model of OCT. synaptic changes, increased recruitment of parallel
Correlation of N95 amplitude with RNFL and macular pathways, and cortical reorganization, including
thickness, but not with P50 latency was also reported functional expansion of surviving neurons are
by Almarcegui et al [25]. thought to be involved in this process. While the
In contrast, the study of patients with optic neur- exact site of neuroplasticity in the hierarchy of the
itis by Trip et al. [23] produced quite different results. visual system still remains a matter of controversy [51,
The authors demonstrated no association of RNFL 53, 54], continuous recovery of mfVEP amplitude
thickness with any of the PERG measures. suggests that it may (at least partially) occur below
or at the level of primary visual cortex.
Cortical plasticity in MS
Following optic neuritis, RGC axons show retrograde Applications in generalized
degeneration back to the retina, which results in RNFL
thinning. This process is believed to be completed by
neurological disease
six months [49]. It was, however, demonstrated that in Increasingly, ophthalmic examination is allowing for
MS patients the RNFL continues to deteriorate after greater understanding of the disease processes in neu-
that (and even regardless of a history of ON). A pro- rological conditions that were not classically asso-
gressive loss of optic nerve fibers even one year after the ciated with direct ocular pathology.
attack was also detected in longitudinal MRI studies of
the optic nerve mean area [18, 50]. Parkinson’s disease
The mechanism of RNFL loss is uncertain, but may Parkinson’s disease (PD) is generally considered a
84 be related to ongoing primary axonal degeneration, “movement disorder”; it is now understood that
 Chapter 7: Optical coherence tomography and electrophysiology of the visual pathway

alterations in dopaminergic cells affects the entire clinically [65]. The OCT shows a loss of RPE cells,
visual system from the retina to areas of the central initially as a ring around the macula, and autofluor-
nervous system that subserve vision. In the retina escence patterns are disrupted relatively early.
dopamine is found within amacrine and interplexiform
cells [55]. The PERG specifically records a response Vigabatrin (Sabril)
consequent on the pre-ganglionic retinal circuitry
Vigabatrin is used as an anti-epileptic medication,
that is controlled by these amacrine and interplexiform
particularly for the management of infantile spasms.
cells. Both PERG and VEP abnormalities can be
It works by increasing the concentration of gamma-
detected in PD patients with specific foveal stimulus
amino butyric acid (GABA) in the brain. However, it
parameters. Interestingly, patients treated with L-dopa
also increases GABA in the retina and is associated
show higher PERG amplitudes compared to untreated
with visual field constriction.
patients, but never reach the amplitudes seen in normal
A study of Vigabatrin patients with visual field
controls [56]. The loss of macular PERG correlates with
constriction [66] shows reduced b-wave amplitude
studies showing that on OCT there is loss of the tem-
for the rod responses and the combined rod-cone
poral retinal nerve fiber layer thickness at the optic disc
response. Lesser, but significant reduction of a-wave
[57] and a thinning of the inner plexiform layer at the
was also detected. Significant positive correlations
fovea [58].Other studies have shown alterations in
were found between the total averaged RNFL thick-
EOG [59] and mfERG in correlation with loss of the
ness, superior and inferior RNFL thickness, and
inferior and temporal RNFL thickness [60].
reduced standard ERG parameters. The authors con-
cluded that Vigabatrin is retino-toxic on several
Alzheimer’s disease levels, from photoreceptors to ganglion cells.
Alzheimer’s disease (AD) is a progressive neurodegen-
erative condition resulting in dementia. As in PD, there Using OCT and mfVEP to characterize
is increasing evidence that the retina is also affected in
the pathophysiological process. Studies have documen-
the nature of nonconventional
ted that the mean RNFL thickness at the optic disc is MRI techniques
reduced in AD compared to controls, with particular While MRI has proven to be a very sensitive tool in
loss of the fibers from the superior quadrant [61, 62]. In confirming diagnosis of MS and monitoring of treat-
addition, these OCT changes were shown to correlate ment trials, it is pathologically nonspecific. It has been
with reduced P1 amplitudes of mfERG in both the suggested that nonconventional MRI techniques such
foveal and peri-foveal regions [62]. as the magnetization transfer ratio (MTR) may pro-
vide more specific characterization of underlying
Monitoring medication side effects pathological processes [67–69].
The retina and central nervous system are particularly MTR imaging is a measure of the exchange of
sensitive to the effects of toxic insults. Many medica- protons between free water and macromolecules in
tions in common use have known effects on the elec- membranes. It is believed MTR is affected by either
trophysiology and structural OCT changes. Some dilution of protons caused by edema or loss of tissue
examples are presented below. structure, in particular the structure of myelin. While
animal research has demonstrated a strong relation-
ship between the level of myelination and MTR
Hydroxychloroquine (Plaquenil) [70, 71], human studies are inconclusive.
In the retinal pigment epithelium, chloroquine phos- Our group performed analysis of optic nerve MTR,
phate remains at higher concentration than in the RNFL thickness and latency of mfVEP in 23 patients
liver [63]. Hydroxychloroquine retinopathy typically with a single unilateral episode of ON and ten healthy
results in reduced b-wave amplitudes on ERG with controls [72]. We found a significant reduction of MTR
prolonged latency in advanced cases and a bull’s eye value in ON eyes. A strong correlation of MTR was
pattern of reduced function on mfERG, with localized demonstrated with RNFL thickness and mfVEP ampli-
reduction in response amplitudes in ring 2, best tude, but not with the level of latency delay, suggesting
detected by analyzing ring ratios [64]. In late cases, a that MTR reduction is related to axonal loss, rather 85
corresponding bull’s eye retinal appearance is seen than demyelination.
 Chapter 7: Optical coherence tomography and electrophysiology of the visual pathway

5. Jones SJ, Brusa A. Neurophysiological evidence for

Limitations of electrophysiology long-term repair of MS lesions: implications for axon
testing protection. J Neurolog Sci 2003; 206: 193–8.
The standards of electrophysiological tests published 6. Waxman SG. Altered distributions and functions of
by ISCEV allow for comparison of results between multiple sodium channel subtypes in multiple sclerosis
laboratories. As yet there is no ISCEV standard for and its models. In: Waxman SG, editor. Multiple sclerosis
as a neuronal disease. Amsterdam: Elsevier, 2005.
the mfVEP. Our recommendation is that at least two
7. Klistorner A, Graham SL, Fraser C, Garrick R, Nguyen
orthogonal channels are needed in mfVEP recording
T, Paine M, et al. Electrophysiological evidence for
to ensure a satisfactory detection without cancellation heterogeneity of lesions in optic neuritis. Invest
of the variously oriented dipoles that combine to form Ophthalmol Vis Sci. 2007; 48: 4549–56.
the conventional VEP. By measuring the responses 8. Brusa A, Jones SJ, Plant GT. Long-term remyelination
that are most reproducible between patients, altera- after optic neuritis. Brain. 2001; 124: 468–79.
tions in these results can provide diagnostic informa- 9. Yiannikas C, Walsh JC. The variation of the pattern
tion. However, it is important to note that patient shift visual evoked response with the size of the
compliance influences responses recorded. Frequent stimulus field. Electroencephalography & Clinical
blinking, loss of fixation, and intentional defocusing Neurophysiology 1983; 55: 427–435.
can result in reduced-quality responses, particularly 10. Halliday AM, Darbett G, Blumhardt LD, Kriss A.
for PERG, mfERG, and mfVEP, which have small The macular and submacular subcomponents of the
amplitude signals, requiring considerable signal aver- pattern evoked response. In: B LDaG, editor. Human
aging. Patients who have poor vision and/or amblyo- evoked potentials. New York: Plenum Publishing; 1979.
p. 135–451.
pia may have abnormal responses due to inability to
fixate further complicating test interpretation. The 11. Klistorner AI, Graham SL, Grigg JR, Billson FA.
Multifocal topographic visual evoked potential: improv-
PERG and the mfVEP are acuity dependent, with a
ing objective detection of local visual field defects. Invest
reduction in signals seen in the central rings of the Ophthalmol Vis Sci. 1998; 39(6): 937–50.
mfVEP when acuity falls below 6/12. Lack of concen- 12. Klistorner A, Fraser C, Garrick R, Graham SL, Arvind
tration can affect the mfVEP via the intrusion of alpha H. Correlation between full-field and multifocal VEPs
rhythm, and we have found that a changing fixation in optic neuritis. Doc Ophthal. 2008; 116: 19–27.
target is helpful to engage the patient and reduce 13. Hood DC, Odel JG, Zhang X. Tracking the recovery of
alpha-rhythm contamination. local optic nerve function after optic neuritis: a multifocal
In summary, the use of both electrophysiology as a VEP study. Invest Ophthalmol Vis Sci 2000; 41: 4032–8.
functional measure and OCT as a structural measure 14. Graham SL, Klistorner A, Goldberg I. Clinical
provides better insight into mechanisms of neurode- Application of objective perimetry using multifocal
generative diseases when the visual system is used as a VEP in Glaucoma Practice. Arch Ophthalmol 2005;
model. In clinical trials of potentially neuroprotective 123: 729–39.
and remyelinating drugs an assessment of structural 15. Laron M, Cheng H, Xhang B, Schiffman JS, Tang RA,
and functional changes will be valuable to determine Frishman LJ. Assessing visual pathway function
in multiple sclerosis patients with multifocal
beneficial effects of therapy.
visual evoked potentials. Mult Scler 2009; 15:
1431–41.
References 16. Fredericksen JL, Petrera J. Serial visual evoked poten-
1. Holder GE. Pattern electroretinography (PERG) and tials in 90 untreated patients with acute optic neuritis.
an integrated approach to visual pathway diagnosis. Surv Ophthalmol 1999; 44: S54–62.
Prog Ret Eye Res 2001; 20: 531–61. 17. Ghezzi A, Martinelli V, Torri V et al. Long-term follow-
2. Holder GE, Gale RP, Acheson JF, Robson AG. up of isolated optic neuritis: the risk of developing
Electrodiagnostic assessment in optic nerve disease. multiple sclerosis, its outcome, and the prognostic role
Curr Opin Neurol 2009; 22: 3–10. of paraclinical tests. J Neurol 1999; 246: 770–5.
3. Halliday AM, McDonald WI, Mushin J. Delayed visual 18. Hickman SJ, Toosy AT, SJJ, Altman DR, Mizszkiel KA,
evoked response in optic neuritis. Lancet 1972; i: 982–5. MacManus DG, et al. A serial MRI study following
4. Fahle M, Bach M. Origin of the visual evoked poten- optic nerve mean area in acute optic neuritis. Brain
tials. In: Heckenlively JR, Arden GB, editors. Principles 2004; 127: 2498–505.
and practice of clinical electrophysiology of vision. 19. Youl BD, Turano G, Miller DH, Towell AD,
86 MacManus DG, Moore SG, et al. The pathophysiology
Cambridge: The Mit Press 2006. p. 207–234.
 Chapter 7: Optical coherence tomography and electrophysiology of the visual pathway

of acute optic neuritis. An association of gadolinium Waxman SG, editor. Multiple sclerosis as a neuronal
leakage with clinical and electrophysiological deficits. disease. Amsterdam: Elsevier; 2005. p. 165–184.
Brain 1991; 114: 2437–50. 34. Pro MJ, Pons ME, Liebmann JM, Ritch R, Zafar S,
20. Miller DH. Natural history of multiple sclerosis: when Lefton D, et al. Imaging of the optic disk and retinal
do axons degenerate? In: Waxman SG, editor. Multiple nerve fiber layer in acute optic neuritis. J Neurolog Sci.
sclerosis as a neuronal disease. Amsterdam: Elsevier 2006; 250: 114–9.
2005. p. 185–200. 35. Petzold A, de Boer JF, Schippling S, Vermersch P,
21. Frohman EM, Fujimoto JG, Frohman TC, Calabresi Kardon R, Green A, et al. Optical coherence tomogra-
PA, Cutter GR, Balcer LJ. Optical coherence tomogra- phy in multiple sclerosis: a systematic review and
phy: a window into the mechanisms of multiple meta-analysis. The Lancet (Neurol) 2010; 9: 921–32.
sclerosis. Nature Neurol 2008; 4: 664–775. 36. Talman LS, Bisker ER, Sackel BS, Long DA, Galetta
22. Parisi V, Manni G, Centofanti M, Gandolfi SA, Olzi D, KM, Ratchford JN, et al. Longitudinal study of vision
Bucci MG. Correlation between optical coherence and retinal nerve fiber layer thickness in multiple
tomography, pattern electroretinogram, and visual sclerosis. Ann Neurol 2010; 67: 749–60.
7evoked potentials in open-angle glaucoma patients. 37. Klistorner A, Arvind H, Garrick R, Graham SL, Paine
Ophthalmology 2001; 108(5): 905–12. M, Yiannikas C. Interrelationship of optical coherence
23. Trip A, Schlottmann PG, Jones SJ, Altman DR, tomography and multifocal visual-evoked potentials
Garway-Heath DF, Thompson AJ, et al. Retinal nerve after optic neuritis. Invest Ophthalmol Vis Sci 2010;
fiber layer loss and visual disfunction in optic neuritis. 51(5): 2770–7.
Ann Neurol 2005; 58: 383–91. 38. Klistorner A, Garrick R, Paine M, Graham SL, Arvind H,
24. Pueyo V, Martin J, Fernandez J, Almarcegui C, Ara J, Egea Van Der Walt A, et al. Relationship between chronic
C, et al. Axonal loss in the retinal fiber layer in patients demyelination of the optic nerve and short term axonal
with multiple sclerosis. Mult Scer 2008; 14: 609–14. loss. J Neurol Neurosurg Psychiatry. 2011; Epub ahead of
25. Almarcegui C, Dolz I, Pueyo V, Garcia E, Fernandez print.
FJ, Martin J, et al. Correlation between functional and 39. Klistorner A, Garrick R, Barnett MH, Graham SL,
structural assessments of the optic nerve and retina in Arvind H, Sriram P, et al. Axonal loss in non-optic
multiple sclerosis patients. Clin Neurophysiol 2010; 40: neuritis eyes of patients with multiple sclerosis linked to
129–35. delayed visual evoked potential. Neurol 2013; 15: 242–5.
26. Klistorner A, Arvind H, Nguyen T, Garrick R, Paine M, 40. Bruck W. Inflammatory demyelination is not central
Graham S, et al. Inflammation, demyelination and to the pathogenesis of multiple sclerosis. J Neurol.
axonal loss in post-acute optic neuritis. Ann Neurol 2005; 252(Suppl 5): V/10-V/5.
2008; 61: 325–31. 41. Patrikios P, Stadelmann C, Kutzelnigg A, Rauschka H,
27. Klistorner A, Arvind H, Nguyen T, Garrick R, Paine M, Schmidbauer M, Laursen H, et al. Remyelination is
Graham S, et al. Multifocal VEP and OCT in optic extensive in a subset of multiple sclerosis patients.
neuritis: a topographical study of the structure–function Brain 2006; 129: 3165–72.
relationship. Doc Ophthal 2009; 118: 129–37. 42. Prineas JW, Barnard RO, Kwon EE, Sharer LR, Cho ES.
28. Smith KJ, Waxman SG. The conduction properties of Multiple sclerosis: remyelination of nascent lesions.
demyelinated and remyelinated axons. In: Waxman Ann Neurol 1993; 33: 137–51.
SG, editor. Multiple sclerosis as neuronal disease. 43. Prineas JW, Kwon EE, Goldenberg PZ, Ilyas AA,
Amsterdam: Elsevier Academic Press; 2005, p. 85–100. Quarles RH, Benjamins JA, et al. Multiple sclerosis.
29. Lassmann H. New Concepts on progressive multiple Oligodendrocyte proliferation and differentiation in
sclerosis. Cur Neurol Neurosci Reports 2007; 7: 239–44. fresh lesions. Lab Invest 1989; 61: 489–503.
30. Foster RE, Whalen CC, Waxman SG. Reorganisation 44. Blakemore WF, Chari DM, Gilson JM, Crang AJ.
of the axonal membrane of demyelinated nerve fibres: Modelling large areas of demyelination in the rat reveals
morphological evidence. Science 1980; 210: 661–3. the potential and possible limitations of transplanted glial
31. Bostock H, Sears TA. The internodal axon membrane: cells for remyelination in the CNS. Glia 2002; 38: 155–68.
electrical excitability and continious conduction in 45. Klistorner A, Arvind H, Garrick R, Yiannikas C,
segmental demyelination. J Physiol (London) 1978; Graham SL. Remyelination of optic nerve lesions: spa-
280: 273–301. tial and temporal factors. Mult Scler 2010; 16: 786–95.
32. Shrager P, Simon W, Kazarinova-Noyes K. Na+ 46. Klistorner A, Arvind H, Nguyen T, Garrick R, Paine M,
channel reorganisation in demyelinated axons. In: Graham S, et al. Axonal loss and myelin in early ON
SG W, editor. Multiple sclerosis as a neuronal disease. loss in postacute optic neuritis. Ann Neurol 2008;
Amsterdam: Elsevier; 2005. 64(3): 325–31.
87
33. Peterson J, Kidd D, Trapp BD. Axonal degeneration in 47. You Y, Klistorner A, Thie J, Graham SL. Latency delay
multiple sclerosis: the histopathological evidence. In: of visual evoked potential is a real measurement of
 Chapter 7: Optical coherence tomography and electrophysiology of the visual pathway

demyelination ina rat model of optic neuritis. Invest and functional retinal impairment in patients with
Ophthalmol Vis Sci 2011; 52: 6911–8. Parkinson disease without visual loss. 2011;
48. Raz N, Chokron S, Ben-Hur T, Levin N. Temporal 21: 2429.
reorganization to overcome monocular demyelination. 61. Kirbas S, Turkyilmaz K, Anlar O, Tufekci A, Durmus
Neurology. August 20, 2013; 81(8): 702–9. doi: 10.1212/ M. Retinal nerve fiber layer thickness in patients with
WNL.0b013e3182a1aa3e. Epub July 19, 2013. PubMed Alzheimer disease. J Neuroophthalmol 2013; 33: 58–61.
PMID: 23873970. 62. Moschos MM, Markopoulos I, Chatziralli I, Rouvas A,
49. Costello F, Hodge W, Pan YI, Eggenberger E, Coupland Papageorgiou SG, Ladas I, et al. Structural and functional
S, Kardon R. Tracking retinal nerve fiber layer loss after impairment of the retina and optic nerve in Alzheimer’s
optic neuritis: a prospective study using optical coher- disease. Curr Alzheimer Res 2012; 9: 782–8.
ence tomography. Mult Scler 2008; 14: 893–905. 63. Meier-Ruge W. Drug induced retinopathy. CRC
50. Hickman SJ, Brierley CMH, Brex PA, MacManus DG, Toxicol 1972; 1: 325–60.
Scolding NJ, Compston DAS, et al. Continuing optic 64. Lyons JS, Severns ML. Using multifocal ERG ring
nerve atrophy following optic neuritis: a serial MRI ratios to detect and follow Plaquenil retinal toxicity: a
study. Multiple sclerosis 2002; 8: 339–42. review. Doc Ophthalmol 2009; 118: 29–36.
51. Toosy AT, Hickman SJ, Mizszkiel KA, Jones SJ, Plant 65. Marmor MF, Kellner U, Lai TY, Lyons JS, Mieler WF.
GT, Altman DR, et al. Adaptive cortical plasticity in Revised recommendations on screening for chloro-
higher visual areas after optic neuritis. Ann Neurol quine and hydroxychloroquine retinopathy.
2005; 57: 622–33. Ophthalmology 2011; 118: 415–22.
52. Raz N, Dotan S, Benoliel T, Chokron S, Ben-Hur T, 66. Kjellstrom U, Andreasson S, Ponjavic V. Attenuation
Levin N. Sustained motion perception deficit following of the retinal nerve fibre layer and reduced retinal
optic neuritis: Behavioral and cortical evidence. function assessed by optical coherence tomography
Neurology. June 14, 2011; 76(24): 2103–11. doi: 10.1212/ and full-field electroretinography in patients exposed
WNL.0b013e31821f4602. PubMed PMID: 21670440. to vigabatrin medication. Acta Ophthalmol 2013.
53. Korsholm K, Madsen KH, Fredericksen JL, Skimmingr 67. Tjoa CW, Benedict RH, Dwyer MG, Carone DA,
A, Lund TE. Recovery from optic neuritis: anROI- Zivadinov R. Regional specificity of magnetization
based analysis of LGN and visual cortical areas. Brain transfer imaging in multiple sclerosis. J Neuroimaging
2007; 130: 1244–53. April 2008; 18(2): 130–6. doi: 10.1111/j.1552-
54. Werring DJ, Bullmore ET, Toosy AT, Miller DH, Barker 6569.2007.00198.x.
GJ, MacManus DG, et al. Recovery from optic neuritis is 68. Gallo A, Rovaris M, Benedetti B, Sormani MP, Riva R,
associated with a change in the distribution of cerebral Ghezzi A, et al. A brain magnetization transfer MRI
response to visual stimulation: a functional magnetic study with a clinical follow up of about four years in
resonance imaging study. Journal of Neurology, patients with clinically isolated syndromes suggestive of
Neurosurgery & Psychiatry 2000; 68(4): 441–9. multiple sclerosis. J Neurol. January 2007; 254(1): 78–83.
55. Archibald NK, Clarke MP, Mosimann UP, Burn DJ. The Epub February 14, 2007. PubMed PMID: 17508141.
retina in Parkinson’s disease. Brain 2009; 132: 1128–45. 69. Tortorella P, Rocca MA, Mezzapesa DM, Ghezzi A,
56. Tagliati M, Bodis-Wollner I, Yahr MD. The pattern Lamantia L, Comi G, et al. MRI quantification of gray
electroretinogram in Parkinson’s disease reveals lack of and white matter damage in patients with early-onset
retinal spatial tuning. Electroencephalogr Clin multiple sclerosis. J Neurol. July 2006; 253(7): 903–7.
Neurophysiol 1996; 100: 1–11. Epub March 6, 2006. PubMed PMID: 16511645.
57. Kirbas S, Turkyilmaz K, Tufekci A, Durmus M. Retinal 70. Deloire-Grassin MS, Brochet B, Quesson B, Delalande
nerve fiber layer thickness in Parkinson disease. J C, Dousset V, Canioni P, et al. In vivo evaluation of
Neuroophthalmol 2013; 33: 62–t5. remyelination in rat brain by magnetisation transfer
58. Adam CR, Shrier E, Ding Y, Glazman S, Bodis- imaging. J Neurol Sci 2000; 178: 10–16.
Wollner I. Correlation of inner retinal thickness eval- 71. Zaaraoui W, Deloire M, Merie M, Girard C, Raffard
uated by spectral-domain optical coherence tomogra- G, Biran M, et al. Monitoring demyelination and
phy and contrast sensitivity in Parkinson disease. J remyelination by magnetization transfer imaging in
Neuroophthalmol 2013. the mouse brain at 9.4 T. Magn Res Mat Phys. 2008;
59. Nowacka B, Lubinski W, Karczewicz D. 21: 357–62.
Ophthalmological and electrophysiological features of 72. Klistorner A, Chaganti J, Garrick R, Moffat K,
Parkinson’s disease. Klin Oczna 2010; 112: 247–52. Yiannikas C. Magnetisation transfer ratio in optic
60. Moschos MM, Tagaris G, Markopoulos I, Margetis I, neuritis is associated with axonal loss, but not with
Tsapakis S, Kanakis M, et al. Morphologic changes demyelination. Neuroimage 2011; 56(1): 21–6.
88
 Chapter
Optical coherence tomography and

8 electrophysiology of the optic nerve head

Shin C. Beh, Zane Schnurman, Darrel Conger, Amy Conger, Benjamin M.
Greenberg, Elliot M. Frohman, and Teresa C. Frohman

Introduction commensurate clinical concomitants, the so-called

It is reasonable to predict that functional disruptions clinic-radiologic paradox [3]. In contradistinction,
precede any measurable structural changes in the ner- extremely modest alterations in tissue architecture
vous system and retina. Early pathologic changes that (e.g., very few brain hyperintensities on T2 or FLAIR-
begin at a cellular level may easily evade the resolution weighted MR imaging sequences, or subtle to no iden-
of highly sensitive structural metrics but may be tifiable changes in the thickness of retinal layers), can
detected by electrophysiologic means. It is possible nonetheless be associated with disabling physiologic
that early dysfunction represents a potentially reversi- changes within the central nervous system (CNS).
ble state of visual compromise, and, therefore, timely In the elegant study by Naismith and colleagues,
detection of such dysfunction would allow therapeutic the authors demonstrated discrete changes in visual
interventions that would halt the pathological process, evoked potential (VEP) response metrics in patients
prevent future disability, and potentially even reverse with acute optic neuritis (ON) at a time when corre-
the accrued damage, even before structural changes of sponding abnormalities of the retinal nerve fiber layer
retinal nerve fiber layer or ganglion cell layer thinning thickness (RNFL) had not yet been sufficiently
are evident on OCT. manifested to document the changes, even with the
Visual perimetry, visual acuity, contrast vision application of high-definition spectral-domain OCT
testing, and color testing (time-tested neuro- imaging techniques [4]. This important observation
ophthalmic methods of investigating visual function) confirms the principle that pathophysiologic mechan-
are contingent upon the active participation of a isms, such as changes in axonal transmission proper-
cooperative patient and are, therefore, prone to varia- ties (due to demyelination, ion channel perturbations,
bility. However, these assessments themselves are intra-axonal neurofilamentous, and microtubular
considered valid and of adequate utility to have ren- deconstruction; changes in the thresholds for mem-
dered them the gold standards for clinical evaluation. brane depolarization; and the commensurate require-
Alternately, each of these tests are lacking in sufficient ments for meeting an escalated requirement for intra-
sensitivity such that a more than insignificant burden axonal ATP) may antedate the detection of changes in
of disease activity could have been sustained within retinal architecture. The compromise in signal fidelity
the visual system before such changes translate into fails to provide the brain with critical visual informa-
corresponding functional consequences. For instance, tion and actually serves to predict the localization and
evidence confirms the hypothesis that, in glaucoma, magnitude of tissue injury. Both the localization and
significant damage to the retinal ganglion cells (RGC) magnitude of tissue injury are germane to influencing
may occur before visual field deficits are detectable the integrity of normal visual system circuit capabil-
[1, 2]. An analogue of this observation reflects a ities (i.e., vision, adjustment of the pupillary aperture
similar discordance between changes in the burden across various lighting conditions to optimize
of radiographic measures of disease activity in foveation while reducing chromatic aberration asso-
multiple sclerosis (MS) and the surprising lack of ciated with peripheral lens distortions, and for the

Optical Coherence Tomography in Neurological Disease, ed. Peter A. Calabresi, Laura J. Balcer, and Elliot M. Frohman. 89
Published by Cambridge University Press. © Cambridge University Press 2015.
 Chapter 8: Optical coherence tomography and electrophysiology of the optic nerve head

transmission of light information into the hypothala- of dysfunction [7]. Second, approximately 65% of the
mus, which relates to a myriad of important cortical response network is generated by the central
physiologic processes). Given the evidence-based 2° of the visual field [8]. Further, since the macular
observations underscored earlier, we are at a cross- response (particularly driven by the lower field) dom-
roads with respect to the need for objective, conveni- inates the VEP, spatially prominent dysfunctional
ent, tolerated, and time-efficient electrophysiologic areas confined to the upper field distribution may
techniques that provide for the capability to precisely still evade capture by conventional VEP techniques
define both stimulus and response characteristics in [7, 9]. Third, inter-individual heterogeneity of striate
conjunction with a highly stringent low test–retest cortical anatomy, as well as the contribution of extra-
variability. striate regions to the VEP response, combine to inten-
Ultimately, the principal thrust of such testing sify the broad variation in what constitutes “normal”
paradigms must be the ability to accurately, objec- response metrics [10–18]. Finally, VEP abnormalities
tively, and reproducibly monitor the physiologic sig- may arise from lesions anywhere along the retino-
natures associated with RGC neurobiology. A cortical pathway, rendering the technique lacking in
sensitive electrophysiologic technique that allows the all important localization aspect of systems
early detection of RGC or optic nerve dysfunction in analysis in neurology and neurobiology.
neurologic diseases like MS would permit timely insti-
tution of immunomodulatory agents, or neuroprotec- Multifocal visual evoked potentials (mfVEP)
tive and even neuro-restorative therapies that might
The application of multifocal VEP (mfVEP) over-
serve to arrest or at least attenuate the inflammatory
comes the poor spatial resolution, as well as the dom-
cascades affiliated with MS, particularly, that culmi-
inance of the central and lower fields inherent to
nate in progressive disability.
conventional VEPs. Instead mfVEP parcellates the
central 24°– 32° of vision into 60 segments (each of
Electrophysiologic investigations of which has a 4 × 4 black-and-white checked grid that
the afferent visual system contrast-reverses according to a pseudo-random
m-sequence) and allows evoked responses to be gen-
Visual-evoked potentials (VEPs) erated simultaneously from multiple areas of the
visual field. This permits topographical analysis of
A well-established technique in neuro-ophthalmology,
the afferent visual system (thus improving the detec-
VEPs refer to electrical potentials believed to be gener-
tion of small, localized lesions) without a significant
ated in the primary visual cortex (recorded on the scalp
increase in recording time. Additionally, the contri-
in close proximity to the calcarine fissure) and are
bution from extra-striate regions is smaller to that in
elicited by a brief photic stimuli [5–7]. Visually evoked
VEP [7, 19–23].
potential cortical responses reflect the functional integ-
The utilization of mfVEP has its most powerful
rity of the retino-geniculo-calcarine pathway; hence,
application for the purpose of characterizing the
any lesion along the afferent visual pathway (e.g.,
intra-individual interocular mfVEP comparison tech-
optic neuritis, chiasmitis, tract, geniculate, optic radia-
nique, with evidence demonstrating clear superiority
tion, or occipital lesions) may produce abnormalities in
over the monocular mfVEP technique [7; 24–27]. This
magnitude or timing responses associated with VEP
interocular technique is suitable for monocular/uni-
responses.
lateral pathologies (e.g., non-arteritic ischemic ante-
rior optic neuropathy). Notwithstanding the inherent
Visual evoked potentials: limitations advantages of mfVEP technology, this advanced
There are several limitations to the application of method of visual system physiology is not without
conventional VEP studies in clinical practice and in important limitations. First, a major drawback of
clinical investigation. First, the full-field illumination mfVEPs is the lack of normative values [7], a signifi-
used in VEP elicits a compound waveform that repre- cant obstacle to the wide employment of this techni-
sents a consolidated global-field response, which is a que in clinical investigation and principally related to
derivative of both abnormal and normal cortical gen- the wide inter-individual variability in cortical anat-
erators and thereby lacks the spatial resolution neces- omy. Further, diseases like MS often affect both reti-
90 sary in order to identify geographically discrete areas nas and optic nerves, and this may confound accurate
 Chapter 8: Optical coherence tomography and electrophysiology of the optic nerve head

interocular comparisons. Second, while the mfVEP is disrupted in diseases affecting the inner retina and
sensitive for detection of central defects, areas of dys- optic nerve (as a consequence of retrograde RGC
function outside the central 10° may be missed due to degeneration) [30; 36–42]. However, there is evidence
sparser sampling [25]. Lastly, like conventional VEPs, that the PERG response may not be purely contingent
mfVEP abnormalities may occur from lesions any- upon nonlinear (i.e., RGC) components [43].
where along the retino-cortical pathway. Additionally, PERG responses are dominated by
macular RGC contributions [44, 45]; hence, periph-
Electroretinography (ERG) eral areas of dysfunction may escape detection.
Finally, the PERG is also limited by poor spatial reso-
Conventional ERG lution; like the cERG, the massed response may easily
miss localized areas of dysfunction.
Conventional full-field ERG (cERG) generates a single
compound waveform (recorded over the cornea) in
response to a brief flash delivered via a Ganzfeld bowl Photopic negative response (PhNR)
to represent the activity of the entire retina [28, 29]. It The PhNR is a component of the photopic ERG that
is useful for evaluating retinal disorders, in particular represents the activity of the RGCs and their axons, as
those that affect the health of the rods and cones (e.g., well as the integrity of K+ currents in retinal glial cells
retinitis pigmentosa, toxic retinopathies). [40, 46–47]. In primates, the PhNR is eliminated by
The cone response may be isolated by recording intravitreal tetrodotoxin (which abolishes RGC activ-
the cERG under light-adapted (photopic) conditions, ity) and in glaucoma models [46]. In human beings, it
by using a stimulus wavelength above 680 nm or is diminished in diseases that affect the RGCs and
with a 30 Hz flicker. Alternately, the rod response their axons, including MS, glaucoma, non-arteritic
may be isolated by scotopic conditions, that is, acute ischemic optic neuropathy (NAION), compres-
dark-adaptation for 45 minutes, followed by a short- sive optic neuropathy, traumatic optic neuropathy,
wavelength stimulus as a single flash, or a 10 Hz autosomal dominant optic atrophy, and central
flicker [30,31]. retinal artery occlusion [40; 48–53]. Interestingly,
Photopic responses generate small b-wave ampli- epilepsy ataxia sensorineural deafness (EAST) syn-
tudes with short latencies (~30 ms). Scotopic condi- drome, an autosomal recessive mutation affecting
tions result in larger b-wave amplitudes and longer the inward-rectifying potassium channel, is also asso-
latencies (~60 ms) [30]. There is little contribution ciated with PhNR abnormalities [47]. Retrobulbar
from the RGC population to the cERG [32, 33], limit- optic nerve disruption without RGC perturbation
ing its utility in assessing pathologies of the inner (i.e., before retrograde degeneration occurs) does
retinal layers (those layers closest to the vitreous) not appear to affect the PhNR, implicating an intrao-
and optic nerve. Additionally, since the cERG repre- cular RGC source for this waveform [50]. Therefore, it
sents a global summed retinal response, small areas of is possible that early optic nerve disease may not be
dysfunction are subordinated by the principal retinal detected until intraocular RGC dysfunction begins.
response. Unlike the VEP (where the central 2° of Further, since the PhNR utilizes a full-field
vision dominates the response), the foveal response stimulus like the cERG, it may fail to capture small
represents less than 2% of the ERG response [34]; yet discretely localized definitive defects.
therefore, central scotomas may easily be missed.
Multifocal ERG (mfERG)
Pattern ERG (PERG) The application of mfERG helps overcome the limited
PERG is a biphasic response of smaller amplitude spatial resolution of the cERG. In mfERG, the retina is
than the cERG and is a useful method of assessing stimulated (under photopic conditions) by a hexago-
inner retinal function. Utilizing an achromatic, 0.5– nal array subtending the central 20°–25° of vision. The
1 s phase reversing checkerboard pattern (similar to stimulus array is scaled in size to elicit responses that
VEP) that maintains constant luminance, the PERG are equal in amplitude across the retina, with the
isolates nonlinear (i.e., inner retinal) responses [30, central hexagons being smaller than the peripheral
35]. PERG generation is hypothetically contingent ones [30]. The mfERG takes a single, continuous
upon the functional integrity of the RGCs and is electrical signal from the eye and mathematically 91
 Chapter 8: Optical coherence tomography and electrophysiology of the optic nerve head

extracts and isolates a response profile (i.e., a “mini and ONHC) [57–58]. The RC represents the contribu-
ERG”) for each element of the display, thereby oper- tions of diversified elements across different layers of
ationally executing a topographical interrogation of the retina. Alternately, the normal ONHC response is
the retina, in conjunction with a corresponding non- derived from the inner retinal, with an onset latency
linear analysis of retinal responses [20, 29, 54–55]. that is related to the distance from the retinal patch of
The first-order kernel responses of the mfERG (the stimulation to the optic nerve head.
most commonly measured parameter), consists of an While the RC response is directly derived from the
initial negative deflection (N1), a positive component patch of retinal stimulation and recorded by the corneal
(P1), and a second negative deflection (N2). It has electrode, the ONHC response is contingent on the
been shown that N1 and P1 are analogous to the a- integrity of the RGC, its axon, and the associated pro-
and b-waves of the cERG, respectively [29]. While the pagation of the action potential from the patch of retinal
first-order kernel responses are devoid of significant stimulation, beyond which it travels to the optic nerve
contributions from the RGC pool, higher-order ker- head, and subsequently across the lamina cribrosa.
nels (the largest of which is the second-order kernel With these principles in mind, we can understand
first slice, K2.1) are generated by flash interactions how those patches of retinal stimulation that are closest
(i.e., modification of responses induced by a to the optic nerve head will reveal ONHC responses of
preceding photic stimulus) and are purely nonlinear, shortest latency, while those patches stimulated
representing predominantly inner retinal (in particu- furthest away will generate the longest latency responses
lar, RGC) function [56–57, 61]. (resembling a chevron-like pattern; Figure 8.2).

The optic nerve head component (ONHC) The ONHC in animal studies
When retinal responses are derived by means of the Pig eyes are morphometrically quite similar to human
standard (fast) multifocal flicker protocol, analysis of eyes, and also demonstrate a late positive mfERG
the K2.1 revealed several induced components, waveform (referred to as the P3 peak) [59] that bears
including a unique, and long-latency, electrophysio- a striking resemblance to the ONHC response in
logic signature waveform (with respect to the retinal human beings. As in the human ONHC response,
component (RC), the so-called optic nerve head com- the P3 peak is predominantly related to RGC activity,
ponent (ONHC) response [56]. The origination of the a hypothesis that has been corroborated by studies
ONHC responses have been hypothesized to be gen- that demonstrated abolishment of these late responses
erated in the vicinity of the optic nerve head (ONH). following the administration of tetrodotoxin, isoflur-
The stereotyped geometric configuration of the RGC ane, and NMDA, and by its time-related deterioration
axons within the RNFL traverse the neural retinal rim following optic nerve section [59–60]. While the P3
(at which point they sharply bend out of the plane of did not demonstrate the expected changes in latency
the inner retina and into the optic cup), concomitant with distance from the optic nerve head (as in human
with their physiologic transformation from unmyeli- beings and primates), this may be due to interspecies
nated axonal membrane conduction to saltatory con- morphometric differences [59].
duction properties, as they penetrate the lamina Macaques also demonstrate a very prominent
cribrosa and obtain oligodendrocyte-derived myelin ONHC (due to the high number of RGCs) [61]. Its
internodes (Figure 8.1) [56]. elimination by intravitreal TTX [62–64] and disrup-
Subsequent modifications in the mfERG stimulus tion in advanced glaucoma [33] indicates the promi-
paradigm (specifically, by appending one or more glo- nent contribution of RGC activity to this waveform.
bal flashes after the m-frame to emphasize the fast
adaptive mechanisms that give rise to the induced The normal human ONHC
components) have improved the ability to elucidate Our group sought to study the characteristics of the optic
RGC activity in the first-order kernel mfERG responses nerve head component response (ONHC) responses in a
[57]. As stated previously, it was with the second-order cohort of normal human subjects. We demonstrated two
first slice K2.1 that enabled us to elucidate the ONHC. principal response waveforms: the retinal component
Response waveforms obtained by these global flash (RC) and ONHC (Figure 8.2). The RC is a conspicuous,
paradigms consist of a direct component and several large-amplitude waveform consisting of an initial nega-
92 induced components (which are dominated by the RC tive trough, a positive deflection, followed by a second
 Chapter 8: Optical coherence tomography and electrophysiology of the optic nerve head

LEGEND
Stimulated Retinal Patch
RETINAL LAYERS LIGHT CELLS
Retinal Component
Optic Nerve Head Component
Inner Retina Disrupted Optic Nerve Head Component
Retinal Nerve
Fiber Layer
Ganglion Cell Layer Burrian-Allen Electrode
Retinal
Ganglion Cell Cornea
Inner Plexiform Layer
Anterior Chamber

Amacrine Cell
Inner Nuclear Layer
Bipolar Cell Iris

Horizontal
Outer Plexiform Layer
Cell

Outer Nuclear Layer Pupil

Rod

Photoreceptor Layer Lens

Cone
Retinal Nerve Fiber Layer
Retinal Pigment
Retina
Epithelium Layer
Outer Retina Choroid
Sclera
Macula

Unmyelinated
Retinal Ganglion Unmyelinated Retinal
Cell Axon Ganglion Cell Axon

Lamina Cribrosa Lamina Cribrosa

Myelin Sheath
Oligodendrocyte

Oligodendrocyte Membrane Conduction

along Demyelinated
Axon Axonal Segment

Demyelinated
Saltatory Axonal Axons
Conduction
Node of Ranvier Myelin Sheath

Node of Ranvier

Retinal Component
Retinal Component
Waveform
Unmyelinated Retinal Waveform
Optic Nerve Head Nerve Fibers Absent
Component
Lamina Cribrosa Optic Nerve Head
Waveform Component
Waveform
Myelinated Axons of the
Optic Nerve

Central Retinal Artery

Figure 8.1 Generation of the optic nerve head component response

This illustration details how the retinal component (RC) and optic nerve head component (ONHC) responses are generated within the retina
and subsequently detected and characterized by multifocal electroretinographic (mfERG) techniques. The patient is exposed to stimuli
consisting of 103 hexagons, the map of which we illustrate within the posterior part of the globe of the eye. Each stimulus consists of five
frames – the first contains focal flashes (controlled by the VERIS pseudo-random m-sequence), the second and fourth contain global flashes,
and the third and fifth frames are dark.
The photic stimulus generates response first in the photoreceptors and then in the retinal ganglion cells. In this illustration, the retinal
patches of stimulation (orange hexagons) will yield electrical responses that are detected at the corneal surface with a Burian-Allen electrode.
The large-amplitude RC (teal arrow) is directly derived from the stimulated retinal patch, representing the response of multiple cell types within
the retina, with only a modest contribution made by the retinal ganglion cells. However, a smaller, later waveform following the RC is
designated as the ONHC waveform. The action potential that is generated at the stimulated retinal patch is propagated (membrane
conduction) via the retinal ganglion cell axons in the retinal nerve fiber layer to the optic nerve head.
The ONHC response (solid green arrow) is generated as this action potential traverses the lamina cribrosa and its conduction transforms
from membrane to saltatory conduction (as the optic nerve axons obtain oligodendrocyte myelination). As the ONHC is generated at the optic
nerve head region, its latency is contingent upon the distance of the stimulated retinal patch from this region (i.e., the farther the patch, the
longer the latency) and the thickness of the axons transmitting the action potential (i.e., the thicker the axons, the shorter the latency). In the
context of ocular or neurologic disease that affects the RGC, RGC axons, or optic nerve myelin, the ONHC would appear abnormal or be
completely lost (broken green arrow).

negative trough. The RC response demonstrated con- The ONHC is the positive waveform occurring
stant peak latency, irrespective of distance or eccentricity immediately following the second negative trough of
from the optic nerve head. This finding supports the the RC. Its peak latency varied with distance and eccen-
hypothesis that the RC is composed of various cellular tricity from the ONH. Within rings of equal eccentricity,
responses (and not purely RGC activity) and is directly the ONHC peak latency was directly proportional to the
derived from each stimulated retinal patch [56, 65]. distance of the stimulated retinal patch from the optic 93
 Chapter 8: Optical coherence tomography and electrophysiology of the optic nerve head

Ring 1 Ring 2 Ring 3 Ring 4 Ring 5

1 1 1 1 1
2 2 2 2 2
3 3 3 3 3

4
4 4 4 4
5 5 5 5 5
6 6 6 6 6
7 7 7 7
8 8 8 8
60 70 80 90 ms 9 9 9 9
10 10 10 10
11 11 11 11
12 12 12 12
13 13 13
14 14 14
60 70 80 90 ms
15 15 15
16 16 16
17 17 17
18 18 18
19 19
20 20
60 70 80 90 ms 21 21
22 22
23 23
24 24
Ring 1
25
Ring 2 26
60 70 80 90 ms
Ring 3 27
28
Ring 4 5 4 3 29
Ring 5 6 2 30

1 31
7 Optic Nerve Head
32
8 12
33
9 10 11 34
35
36
37
38
39
40
41
42

60 70 80 90 ms

Figure 8.2 The appearance of the normal optic nerve head component response.
The stimulus array consists of hexagonal patches arranged in a series of concentric rings, each progressively more eccentric from the fovea
centralis. Ring 1 is the closest to the fovea, and Ring 4 the farthest away. By convention, the stimulation pattern begins at the peripapillary zone,
then proceeds superiorly, temporally, and inferiorly, and ultimately returns to the peripapillary zone. This results in the progressive increase
and subsequent decrease in the ONHC latency (for patches of equal eccentricity from the fovea centralis) and gives rise to the chevron-like
pattern (marked by the dashed lines) of the ONHC response when displayed in the trace array. This chevron-like pattern is an important feature
seen in normal ONHC responses.

nerve head. This observation supports the hypothesis axons transmitting the stimulus to the optic nerve
that the ONHC is ultimately generated in the region of head. The shortest latency responses were those clo-
the optic nerve head and is, therefore, dependent on the sest to the ONH, while those further away were
time needed for an action potential to travel via mem- delayed commensurate with their distance from the
brane conduction from the RGC to the optic nerve head, optic nerve head.
at which point the axons bend into the optic cup, pene- Of interest, we also found that ONHC peak
trate the lamina cribrosa, and become myelinated. latency was inversely related to eccentricity away
Within rings of equal eccentricity (meaning a cir- from the central macula across successive rings
cular pattern of retinal patch stimulation, with each (i.e., the ONHC peak latency decreased with increas-
patch equidistant to the center point of the ring), the ing eccentricity from the central macula, in the verti-
94 ONHC latency is contingent upon the length of the cal meridian). While this may seem counterintuitive
 Chapter 8: Optical coherence tomography and electrophysiology of the optic nerve head

(since the distance of the stimulation patch is farther makes it more difficult to establish normal parameters
away from the ONH), this finding is corroborated by for both the RC and ONHC peak amplitudes, which is
changes that occur in retinal anatomy and physiology reflected by the wide amplitude variability.
with eccentricity, and it is in keeping with principles That variability was reduced when we examined
of cable theory. the ONHC:RC peak amplitude ratio, given that both
Sutter et al. showed that the calculated propaga- components are affected similarly by a number of
tion velocities derived from concentric rings, charac- confounding factors. Pathological processes extend-
terized by increasing eccentricity from about 40 cm/s ing beyond the inner retina would be expected to
in Ring 1 (smallest diameter ring of retinal patch affect both the RC and ONHC (while minimally
stimulation) to 120 cm/s in Ring 4 [56]; these propa- impacting the ONHC:RC ratio), while a purely inner
gation velocities are comparable to those found by retinal disorder would predominantly affect the
means of antidromic stimulation in feline [66] and ONHC (decreasing the ONHC:RC ratio). In this cir-
primate [67–68] optic nerves. Axonal propagation cumstance, the ONHC:RC ratio may have utility for
velocities in unmyelinated fibers are linearly corre- the identification of patients with dysfunction
lated to the square root of their axon diameters restricted to the inner retina.
(v = k √D; where 1.4 < k < 1.8) [56, 67]; the faster Another characteristic we observed was that both
conduction speed in the outer rings would imply that the RC and ONHC peak amplitudes decreased with
the axonal fibers of these rings have a greater diameter growing eccentricity from the optic nerve head, as a
than those situated within the inner rings (consistent result of a decrease in retinal cell densities (both RGC
with axonal cable theory). Primate and human studies and cones) with increasing eccentricity [56, 65, 73]. The
have shown that the mean RGC axon diameter is ONHC:RC peak amplitude ratio diminished with
smallest in the immediate vicinity of the central eccentricity as a consequence of greater ONHC peak
fovea (in particular, within the papillomacular bun- amplitude diminution [65]; this is corroborated by the
dle; PMB) and increases as one travels eccentrically declining RGC:cone density ratio with eccentricity [73].
towards the superior and inferior arcuate bundles
[67–72]. Even within the proximal 20° of angular The ONHC in neurologic and ophthalmic
distance from the fovea (i.e., the area of the retina
stimulated by the hexagonal array), there is a gradual disease
increase in axonal diameter with growing eccentricity Our investigative team has proceeded to evaluate
[70]. Notably, the mean axon diameter of the inferior ONHC abnormalities in a small group (n = 7) of
arcuate bundle is larger than the superior bundle [67– MS patients with and without prior ON, and to
72], in keeping with our ONHC latency profiles, with compare the ONHC changes with those of structural
a significantly lower latency associated with the infer- (OCT, fundus photography) and functional (visual
ior outer ring responses when compared with posi- perimetry, infrared pupillometry, contrast acuity,
tionally analogous (along the vertical extent of the mfVEP, and mfERG) metrics [65]. We demonstrated
concentric rings) patches in the superior retinal field correspondence between abnormal visual field loss,
(data in preparation). reduced contrast acuity, reduced RNFL thickness,
While the peak latencies of both the RC and ONHC mfVEP cortical response asymmetry, the presence
demonstrated consistent characteristics between sub- of an objective RAPD, and abnormalities of the
jects, the amplitude of both waveforms were more ONHC [65]. Using a larger cohort of MS patients
variable, leading us to instead analyze the ONHC:RC with prior unilateral ON (n=18), we demonstrated a
peak amplitude ratio. The wide variation we found in robust relationship between ONHC abnormalities
the absolute values of the RC and ONHC peak ampli- and low contrast visual acuity and RNFL thinning
tudes is not surprising since there is a twofold range in as measured by spectral-domain OCT. An example
total RGC number in normal eyes, reflected by the of abnormal ONHC responses in an MS patient
number of axons in the optic nerve (between 846,000 with prior optic neuritis is provided in Figures 8.3
and 1.7 million) [73]. Further, ERG amplitudes vary and 8.4.
between healthy subjects and are influenced by factors There is evidence that abnormalities of the ONHC
such as fundus pigmentation [74] and globe axial are present in early glaucoma in human beings [76]
length, as in high myopes [75]. This broad range and in nonhuman primates (as discussed earlier). 95
 Chapter 8: Optical coherence tomography and electrophysiology of the optic nerve head

Ring 1 Ring 2 Ring 3 Ring 4 Ring 5

Contrast Acuity

Contrast Level OD

100.00% 67

2.50% 35
1.25% 24

HVF OCT ONHC Scoring

TS NS
118 68

T G N N/T
PMB 57 82 56 0.98
40

TI NI
136 109

Number of abnormal traces = 17

Number of very abnormal traces = 0

Figure 8.3 The optic nerve head component response in the eye of an MS patient without a history of optic neuritis.
Here we present data from the unaffected (historically) right eye from a patient with multiple sclerosis who had prior left acute optic neuritis.
The upper-left text box indicates the number of correct letters identified on contrast acuity charts (at 100%, 2.5%, and 1.25% levels). Below, we
show the normal pattern-deviation plot from Humphrey automated perimetry, using the 30–2 test. In the lower-left aspect of the figure, we
present the retinal nerve fiber layer (RNFL) thickness analysis by spectral-domain optical coherence tomography (OCT; Spectralis, Heidelberg,
Germany). The average RNFL thickness is mildly reduced (at 82 microns for the “unaffected” right eye), suggesting the presence of occult
disease activity. On the right aspect of the figure, we present the concentric rings of retinal patch stimulation, with the ONHC responses
aligned vertically. The waveforms traced in red are those where the ONHC is either abnormal or absent. The retinal patch topography map
(bottom middle part of the figure) indicates the location of the abnormal or absent responses.

Additionally, there is anecdotal evidence that ONHC Given the constellation of complex and time-
perturbation occurs in Leber’s hereditary optic neuro- intensive investigations, we cannot underscore
pathy (LHON) [77]. enough the importance of working with highly
trained and experienced neuro-ophthalmic techni-
Limitations cians. The technician is responsible for ensuring
proper scientific protocol is observed to ensure data
A major drawback to the application of ONHC
integrity as well as safeguarding patient safety and
responses is the lack of an objective method for
dignity.
accurately quantifying abnormalities of this RGC-
Finally, since the subject is required to fixate on a
specific response. With existing technology, we are
target during recording, uncooperative patients and
already able to quantify the amplitude and peak
those with pendular nystagmus or severely impaired
latency of the waveform. However, in disease
visual acuity are typically excluded from participating
states, the magnitude of response disorganization
in studies examining the ONHC.
may preclude the accurate identification of latency
or amplitude measures. Given this, there is a need
to develop techniques that permit accurate quan- Future directions
titative methods for the assessment of the ONHC Any objective assessment strategy to characterize the
96 response. ONHC response is likely to be contingent upon the
 Chapter 8: Optical coherence tomography and electrophysiology of the optic nerve head

Ring 1 Ring 2 Ring 3 Ring 4 Ring 5

Contrast Acuity
Contrast Level OD
100.00% 9
2.50% 0
1.25% 0

HVF OCT ONHC Scoring

TS NS
48 68

T PMB
N/T N G 28 28
1.55 43 53

NI TI
96 74
Number of abnormal traces = 20
Number of very abnormal traces = 6

Figure 8.4 The optic nerve head component response in the eye of an MS patient previously affected by optic neuritis.
Here we present data from the same patient in Figure 8.3, but derived from the left eye that was previously affected by optic neuritis.
Bedside examination revealed a severe left relative afferent pupillary defect. On ophthalmoscopic examination, there was diffuse optic disc
pallor (signifying chronic changes, compositionally most consistent with astrogliosis, a cardinal histopathologic feature of chronic MS-related
optic neuropathies). Note the severe loss of acuity (in both high- and low-contrast acuity levels), the broad suppression of the corresponding
Humphrey visual field, and the optical coherence tomography (OCT) retinal nerve fiber layer (RNFL) topography map demonstrating diffuse
thinning of the RNFL (both average and several sectors; red indicates RNFL thickness levels below 1% of predicted for a matched population).
On the right aspect of the figure, note the more diffuse nature of the abnormal or absent optic nerve head component (ONHC) responses.
Compared to the right eye, the waveforms are more poorly defined or absent.

high precision, valid, and reproducible methods for The ultimate goal is to explore the relationship
dissecting the mechanisms that impact the RGC from between structural architecture and corresponding
the cell body, the anatomic course of the RGC axons measures of neurophysiology in the visual system,
as they “sweep” through the RNFL and into the optic with an eye toward elucidating a surrogate biomarker
nerve head (at the neural retinal rim), and ultimately (s) for the CNS in degenerative disorders in general,
exit the eye through the lamina cribrosa en route via and for MS in particular. We propose that the ONHC
the retino-geniculo-calcarine pathways for vision to response is a viable, sensitive, and objective method for
the rostral midbrain for eliciting the pupillary light assessing and monitoring RGC structure and function,
reflexes, with the smallest number of projections as well as anterior optic nerve myelin integrity. Diseases
navigating their way from the eye to the hypothala- that may be expected to result in ONHC response
mus, and thereby contributing to a number of abnormalities include glaucoma, ischemic optic neuro-
important bodily processes including mood pathies, hereditary optic neuropathies (e.g., LHON),
regulation, cognitive vigilance, hunger, thirst, and and demyelinating disorders (in particular, MS and
satiety, thermoregulation, the neuroendocrine reflex neuromyelitis optica, NMO).
arcs, sleep–wake cycle transitions. In this analysis Early detection and treatment of glaucoma, one of
scheme, we will need to precisely identify the response the leading causes of blindness in the world, helps
latency (and timing responses) in conjunction with arrest subsequent progression to irreversible visual
the magnitude of ONHC responses. loss [76]. The ONHC response may, therefore, help 97
 Chapter 8: Optical coherence tomography and electrophysiology of the optic nerve head

Table 8.1 How lesions along the afferent visual pathway may affect electrophysiologic tests of visual function

CVEP MfVEP ERG PERG MfERG ONHC

Outer Retina Abnormal Abnormal Abnormal Normal Abnormal Normal
(widespread
defect)
Outer Retina (focal Likely Abnormal Possibly Normal Abnormal Normal
defect) normal normal
Inner Retina Abnormal Abnormal Likely Abnormal Abnormal Abnormal
(widespread normal
defect)
Inner Retina (focal Likely Abnormal Normal Possibly Possibly Abnormal
defect) normal normal normal
Macula Abnormal Abnormal Likely Likely Abnormal Abnormal
normal normal
Optic Nerve Abnormal Abnormal Normal Normal Normal Abnormal
Chiasm Abnormal Abnormal Normal Normal Normal Normal
Retrochiasmatic Abnormal Abnormal Normal Normal Normal Normal
This table provides some general guidelines about how various lesions along the afferent visual pathway may affect electrophysiologic
tests. CVEP: conventional pattern reversal visual evoked potential; MfVEP: multifocal visual evoked potential; ERG: electroretinography; PERG:
pattern electroretinography; MfERG: multifocal electroretinography; ONHC: optic nerve head component response.

Table 8.2 Summary of salient characteristics, potential applications, and limitations of the optic nerve head component
(ONHC) response

Retinal Component • Directly generated at the stimulated retinal patch from the responses of various cell
Characteristics types.
• RC peak latency remains constant regardless of distance from the optic nerve head (ONH)
and eccentricity.
ONHC • Generated in the vicinity of the ONH from inner retinal activity from the transformation of
Characteristcs membrane to saltatory conduction (as unmyelinated RGC axons acquire
oligodendrocyte myelin at the level of the lamina cribrosa).
• Within rings of equal eccentricity (and across the horizontal meridian), ONHC peak latency is
contingent upon the distance of the stimulation patch from the ONH; this is responsible for
the “chevron-like” pattern on the trace analysis.
• Across rings of increasing eccentricity in the vertical meridian, the ONHC peak latency
diminishes; this is a result of increasing RGC axonal diameter as one moves from the
papillomacular bundle to the superior and inferior arcuate fibers. Of note, ONHC peak
latencies for retinal areas corresponding with the inferior arcuate are lower than those of the
superior arcuate since axonal caliber is greater in the inferior fibers.
• ONHC peak amplitude diminishes to a greater degree than the RC with growing
eccentricity since RGC density decreases more steeply than those of other retinal cell
types.
Potential applica- • Any disorder that perturbs inner retinal, RGC axons, or anterior optic nerve myelin/axonal
tions of the ONHC integrity may disrupt the ONHC. Therefore, it may be useful in monitoring disease activity
and treatment responses in the following disorders: multiple sclerosis, neuromyelitis optica,
Leber’s hereditary optic neuropathy, glaucoma, ischemic optic neuropathies, and toxic-
metabolic optic neuropathies.
98
 Chapter 8: Optical coherence tomography and electrophysiology of the optic nerve head

Table 8.2 (cont.)

ONHC compared to • Compared to conventional and multifocal ERG, the ONHC assesses inner retinal and anterior
other electrophy- optic nerve integrity.
siologic tests • ONHC abnormalities indicate more localized damage (inner retinal, RGC axon, or anterior
optic nerve) compared to visual evoked potentials (VEP), which assess the entire retino-
geniculo-calcarine pathway.
• Unlike multifocal VEP studies, ONHC responses do not require interocular comparison.
• ONHC responses provide superior spatial resolution compared to conventional and pattern
ERG as well as conventional VEP.
Limitations of the • The ONHC is a novel electrophysiologic metric and as such, inter-rater, intra-rater, and test–
ONHC retest variability studies, as well as reliable quantification methods, are needed.
• ONHC recording requires the use of mydriatic agents.
• The Burian-Allen bipolar contact lens electrode is the current standard for ONHC recording
and carries a small risk of corneal abrasion. Its application requires experience as well as
topical corneal anesthesia. Ideally, patients should undergo slit lamp examination prior to
and after ONHC recording.
• As accurate recordings requires adequate fixation for the entire duration, severe visual loss
or pendular nystagmus, and an inability to cooperate (e.g., children, demented patients)
precludes testing.

provide localized and early evidence of RGC dysfunc- neuron and its processes. Further, this technique
tion in glaucoma (particularly low-tension glaucoma) may be of sufficient utility to ascertain the health of
and hence permit early intervention and prevention optic nerve axons and may be used to monitor the
of visual loss. longitudinal evolution of optic neuritis, thereby
With the advent of idebenone as a treatment representing a useful marker of visual recovery (or
for LHON [78], there is a need for electrophy- lack thereof). Ultimately, preservation and/or recon-
siologic methods that can reliably monitor RGC stitution of the ONHC response latency and wave-
degeneration and recovery. Larger numbers of form morphology would confirm neuroprotective
patients and longitudinal studies are needed to and/or neurorestorative efficacy of novel neuro-
evaluate if the ONHC response characteristics therapeutic agents; a prospect that could not be
may have utility as a sensitive means of seriously considered in neurodegenerative disorders
assessing RGC function, neuroprotection strate- just a few years ago.
gies, and even to detect and monitor
neurorestoration. References
The afferent visual system is commonly affected 1. Quigley HA, Kerrigan–Baumrind LA, Pease ME,
in demyelinating diseases like MS; abnormalities of Kerrigan DF, Mitchell RS. The number of retinal
retinal structural metrics can be objectively quanti- ganglion cells in glaucoma eyes compared to
fied with techniques like OCT. ONHC metrics may threshold visual field data in the same eyes. Invest
Ophthalmol Vis Sci 1999;40:S582
provide a means of detecting early disruption of
optic nerve myelin integrity in MS patients. Our 2. Kerrigan-Baumrind LA, Quigley HA, Pease ME,
Kerrigan DF, Mitchell RS. Number of ganglion cells in
studies have shown that ONHC response disruption glaucoma eyes compared with threshold visual field
is associated with perturbations of other visual tests in the same persons. Invest Ophthalmol Vis Sci
metrics. The sensitivity of the ONHC response for 2000;41:741–748
detecting occult optic neuropathy would make it an 3. Barkhof F. The clinico-radiological paradox in
ideal test for those patients with subclinical or mild multiple sclerosis revisited. Curr Opin Neurol
disease in whom other metrics may not yet reveal 2002;15:239–245
any obvious abnormality. The ONHC response may 4. Naismith RT, et al. Optical coherence tomography
potentially serve as an important electrophysiologic is less sensitive than visual evoked potentials in 99
biomarker signifying the integrity of the RGC optic neuritis. Neurology 2009;73:46–52
 Chapter 8: Optical coherence tomography and electrophysiology of the optic nerve head

5. Halliday AM, McDonald WI, Mushin, J. Delayed visual field. Electroencephalogr Clin Neurophysiol
visual evoked response in optic neuritis. Lancet 1994;90:65–81
1972;1:982–985 22. Baseler HA, Sutter EE. M and P components of the
6. Halliday AM, McDonald WI, Mushin J. Visual evoked VEP and their visual field distribution. Vision Res.
response in diagnosis of multiple sclerosis. Br Med J 1997;37:675–90
1973;1:661–664 23. Slotnick SD, Klein SA, Carney T, Sutter E, Dastmalchi
7. Hood DC, Odel JG, Winn BJ. The multifocal visual S. Using multi-stimulus VEP source localization to
evoked potential. J Neuro-Ophthalmol obtain a retinotopic map of human primary visual
2003;23:279–289 cortex. Clin Neurophysiol 1999;110:1793–1800
8. Weinstein GW, Odom JV, Cavender S. Visually 24. Grover LK, Hood DC, Ghadiali Q, et al. A comparison
evoked potentials and electroretinography in of multifocal and conventional visual evoked potential
neurologic evaluation. Rev Neurol Clin techniques in patients with optic neuritis/multiple
1991;9:225–242 sclerosis. Doc Ophthalmol 2008;117:121–128
9. Fortune B, Hood DC. Conventional pattern-reversal 25. Hood DC, Greenstein VC. Multifocal VEP and gang-
VEPs are not equivalent to summed multifocal VEPs. lion cell damage: applications and limitations for the
Invest Ophthalmol Vis Sci 2003;44:1364–1375 study of glaucoma. Prog Retin Eye Res 2003;22:201–251
10. Halliday AM, Michael WF. Changes in pattern-evoked 26. Hood DC, Greenstein VC, Odel JG, et al. Visual field
responses in man associated with the vertical and defects and multifocal visual evoked potentials: evi-
horizontal meridians of the visual field. J Physiol dence of a linear relationship. Arch Ophthalmol
1970;208:499–513 2002;120:1672–1681
11. Brindley GS. The variability of the human striate cor- 27. Hood DC, Zhang X, Greenstein VC, et al. An intero-
tex J Physiol 1972;225:1P-3P cular comparison of the multifocal VEP: a possible
12. Stensaas SS, Eddington DK, Dobelle WH. The topo- technique for detecting local damage to the optic
graphy and variability of the primary visual cortex in nerve. Invest Ophthalmol Vis Sci 2000;411:580–1587
man. J Neurosurg 1974;40:747–755 28. Frishman, LJ. Origins of the electroretinogram. In: John
13. Victor JD. Isolation of components due to intracortical R. Heckenlively and Geoffrey B. Arden, editors.
processing in the visual evoked potential. Proc Natl Principles and practice of clinical electrophysiology of
Acad Sci USA 1986;83:7984–7988 vision. 2nd ed. Cambridge, Massachusetts and London,
England: Massachusetts Institute of Technology (MIT);
14. Srebro R. The topography of scalp potentials evoked by
2006.
pattern pulse stimuli. Vision Res 1987;27:901–914
29. Hood DC, Odel JG, Chen CS, Winn BJ. The
15. Schroeder CE, Tenke CE, Givre SJ, et al. Striate cortical
multifocal electroretinogram. J Neuroophthalmol
contribution to the surface-recorded pattern-reversal
2003;23:225–235
VEP in the alert monkey. Vision Res 1991;31:1143–
1157 30. Young B, Eggenberger E, Kaufman D. Current elec-
trophysiology in ophthalmology: a review. Curr Opin
16. Rademacher J, Caviness VS Jr, Steinmetz H, Galaburda
Ophthalmol 2012;23:497–505
AM. Topographical variation of the human primary
cortices: implications for neuroimaging, brain map- 31. Marmor M, Fulton A, Holder G, et al. ISCEV standard
ping, and neurobiology. Cereb Cortex 1993;3:313–329 for full-field clinical electroretinography (2008
update). Doc Ophthalmol 2008;118:69–77
17. Givre SJ, Schroeder CE, Arezzo JC. Contribution of
extrastriate area v4 to the surface-recorded flash VEP 32. Hood DC, Greenstein V, Frishman L, et al. Identifying
in the awake macaque. Vision Res 1994;34:415–428 inner retinal contributions to the human multifocal
ERG. Vision Res 1999;39:2285–2291
18. Aine CJ, Supek S, George JS, et al. Retinotopic orga-
nization of human visual cortex: departures from the 33. Hare WA, Ton H, Ruiz G, et al. Characterization of
classical model. Cereb Cortex 1996;6,354–361 retinal injury using ERG measures obtained with both
conventional and multifocal methods in chronic
19. Fortune B, Hood DC. Conventional pattern-reversal ocular hypertensive primates. Invest Ophthalmol Vis
VEPs are not equivalent to summed multifocal VEPs. Sci 2001;42:127–136
Invest Ophthalmol Vis Sci 2003;44:1364–1375
34. Graham SL, Klistorner A. Electrophysiology: a review
20. Sutter EE. The fast m-transform: a fast computation of of signal origins and applications to
cross correlations with binary m-sequences. Soc Ind investigating glaucoma. Aust N Z J Ophthalmol
Appl Math 1991;20:686–694 1998;26:71–85
21. Baseler HA, Sutter EE, Klein SA, et al. The topogra- 35. Holder GE. Electrophysiological assessment of optic
phy of visual evoked response properties across the
100 nerve disease. Eye 2004;18:1133–1143
 Chapter 8: Optical coherence tomography and electrophysiology of the optic nerve head

36. Maffei L, Fiorentini A. Electroretinographic responses optic nerve atrophy. Arch Ophthalmol
to alternating gratings in the cat. Exp Brain Res 2004;122:341–346
1982;48:327–334 51. Miyata K, Nakamura M, Kondo M, et al. Reduction of
37. Maffei L, Fiorentini A, Bisti S, Hollander H. Pattern oscillatory potentials and photopic negative response
ERG in the monkey after section of the optic nerve. Exp in patients with autosomal dominant optic atrophy
Brain Res 1985;59:423–425 with OPA1 mutations. Invest Ophthalmol Vis Sci
38. Sieving PA, Steinberg RH. Proximal retinal contribu- 2007;48:820–824
tion to the intraretinal 8-Hz pattern ERG of cats. 52. Machida S, Gotoh Y, Tanaka M, Tazawa Y.
J Neurophysiol 1987;57:104–120 Predominant loss of the photopic negative response in
39. Baker CL Jr, Hess RR, Olsen BT, Zrenner E. Current central retinal artery occlusion. Am J Ophthalmol
source density analysis of linear and non-linear com- 2004;137:938–940
ponents of the primate electroretinogram. J Physiol 53. Shinoda K, Yamada K, Matsumoto CS, Kimoto K,
1988;407:155–176 Nakatsuka K. Changes in retinal thickness are corre-
40. Wang J, Cheng H, Hu Y, et al. The photopic negative lated with alterations of electroretinogram in eyes with
response of the flash electroretinogram in multiple central retinal artery occlusion. Graefes Arch Clin Exp
sclerosis. Invest Ophthalmol Vis Sci 2011;53:1315–1323 Ophthalmol 2008;246:949–954
41. Froehlich J, Kaufman DI. The pattern electroretino- 54. Sutter EE, Tran D. The field topography of ERG com-
gram: N95 amplitudes in normal subjects and optic ponents in man – I. The photopic luminance response.
neuritis patients. Electroencephalogr Clin Neurophysiol Vision Res 1992;32:433–446
1993;88:83–91 55. Bearse MA Jr, Sutter EE. Imaging localized retinal
42. Holder GE. The incidence of abnormal pattern dysfunction with the multifocal electroretinogram.
electroretinography in optic nerve demyelination. J Opt Soc Am A Opt Image Sci Vis 1996;13:634–640
Electroencephalogr Clin Neurophysiol 1991;78:18–26 56. Sutter EE, Bearse MA Jr. The optic nerve head com-
43. Vaegan, Sutter EE. Fundamental differences between ponent of the human ERG. Vision Res 1999;39:419–436
the nonlinearities of pattern and focal electroretino- 57. Shimada Y, Bearse MA Jr, Sutter EE. Multifocal elec-
grams. Doc Ophthalmol 1990;76:13–25 troretinograms combined with periodic flashes: direct
44. Lawwill T. The bar-pattern electroretinogram for responses and induced components. Graefes Arch Clin
clinical evaluation of the central retina. Am J Exp Ophthalmol 2005;243:132–141
Ophthalmol 1974;78:121–126 58. Luo X, Patel NB, Harwerth RS, Frishman LJ. Loss of
45. Arden GB, Carter RM, Macfarlan A. Pattern and the low-frequency component of the global-flash
Ganzfeld electroretinograms in macular disease. Br J multifocal electroretinogram in primate eyes with
Ophthalmol 1984;68:878–884 experimental glaucoma. Invest Ophthalmol Vis Sci
2011;52:3792–3804
46. Viswanathan S, Frishman LJ, Robson JG, Harwerth RS,
Smith EL, III The photopic negative response of the 59. Lalonde MR, Chauhan BC, Tremblay F. Retinal gang-
macaque electroretinogram: reduction by lion cell activity from the multifocal electroretinogram
experimental glaucoma. Invest Ophthalmol Vis Sci in pig: optic nerve section, anesthesia and intravitreal
1999;40:1124–1136 tetrodotoxin. J Physiol 2006;570:325–338
47. Thompson DA, Feather S, Stanescu HC, et al. Altered 60. Ng YF, Chan HH, Chu PH, et al. Pharmacologically
electroretinograms in patients with KCNJ10 defined components of the normal porcine multifocal
mutations and EAST syndrome. J Physiol ERG. Doc Ophthalmol 2008;116:165–176
2011;589:1681–1689 61. Hood DC, Bearse MA Jr, Sutter EE, Viswanathan S,
48. Viswanathan S, Frishman LJ, Robson JG, Walters JW. Frishman LJ. The optic nerve head component of the
The photopic negative response of the flash electrore- monkey’s (Macaca mulatta) multifocal electroretino-
tinogram in primary open angle glaucoma. Invest gram (mERG). Vision Res 2001;41:2029–2041
Ophthalmol Vis Sci 2001;42:514–522 62. Hood DC, Frishman LJ, Viswanathan S, et al. Evidence
49. Rangaswamy NV, Frishman LJ, Dorotheo EU, et al. for a ganglion cell contribution to the primate electro-
Photopic ERGs in patients with optic neuropathies: retinogram (ERG): effects of TTX on the multifocal ERG
comparison with primate ERGs after pharmacologic in macaque. Vis Neurosci 1999;16:411–416
blockade of inner retina. Invest Ophthalmol Vis Sci. 63. Frishman LJ, Saszik S, Harwerth RS, et al. Effects of
2004;45:3827–3837 experimental glaucoma in macaques on the multifocal
50. Gotoh Y, Machida S, Tazawa Y. Selective loss of the ERG. Multifocal ERG in laser-induced glaucoma. Doc
photopic negative response in patients with Ophthalmol 2000;100:231–251
101
 Chapter 8: Optical coherence tomography and electrophysiology of the optic nerve head

64. Fortune B, Cull G, Wang L, Van Buskirk EM, Cioffi 72. Fitzgibbon T, Taylor SF. Mean retinal ganglion
GA. Factors affecting the use of multifocal electroreti- cell axon diameter varies with location in the
nography to monitor function in a primate model of human retina. Jpn J Ophthalmol
glaucoma. Doc Ophthalmol 2002;105:151–178 2012;56:631–637
65. Frohman TC, Beh SC, Saidha S, et al. Optic nerve head 73. Curcio CA, Allen KA. Topography of ganglion
component responses of the multifocal electroretino- cells in human retina. J Comp Neurol
gram in MS. Neurology 2013;81:545–551 1990;300:5–25
66. Stanford LR. Conduction velocity variations minimize 74. Wali N, Leguire LE. Fundus pigmentation and the
conduction time differences among retinal ganglion dark-adapted electroretinogram. Doc Ophthalmol
cell axons. Science 1987;25:225–231 1992;80:1–11
67. Fukuda Y, Watanabe M, Wakakuwa K, Sawai H, 75. Westall CA, Dhaliwal HS, Panton CM, et al. Values
Morigawa K. Intraretinal axons of ganglion cells in the of electroretinogram responses according
Japanese monkey (Macaca fuscata): conduction velo- to axial length. Doc Ophthalmol
city and diameter distribution. Neuroscience Research 2001;102:115–130
1988;6:53–71 76. Fortune B, Johnson CA, Cioffi GA. The topo-
68. Ogden TE, Miller RF. Studies of the optic nerve of graphic relationship between multifocal electroreti-
the rhesus monkey: nerve fiber spectrum and nographic and behavioral perimetric measures
physiological properties. Vision Res of function in glaucoma. Optom Vis Sci
1966;6:485–506 2001;78:206–214
69. Ogden TE. Nerve fiber layer of the primate retina: 77. Sadun AA, Salomao SR, Berezovsky A, et al.
morphometric analysis. Invest Ophthalmol Vis Sci Subclinical carriers and conversions in Leber heredi-
1984;25:19–29 tary optic neuropathy: a prospective
70. Hebel R, Hollander H. Size and distribution of gang- psychophysical study. Trans Am Ophthalmol Soc
lion cells in the human retina. Anat Embryol (Berl) 2006;104:51–61
1983;168:125–136 78. Klopstock T, Yu-Wai-Man P, Dimitriadis K, et al. A
71. Roedick RW, Binmoeller KF, Dineen J. Parsaol and randomized placebo-controlled trial of idebenone in
midget ganglion cells of the human retina. J Comp Leber’s hereditary optic neuropathy. Brain
Neurol 1985;233:115–132 2011;134:2677–2686

102
 Chapter
Meta-analysis of optical coherence

9 tomography in multiple sclerosis

Axel Petzold

Introduction sclerosis, MS, optic neuritis, ON, optical coherence

tomography, OCT, retinal nerve fiber layer, and
This chapter aims to help with a holistic interpreta-
RNFL. All references were reviewed in order to identify
tion of optical coherence tomography (OCT) data in
those investigations where Stratus OCT was utilized
multiple sclerosis (MS) research [34]. A meta-analysis
(Zeiss, software version 3.0 & 4.0), given that it was
will be used to combine some of the data also
the most broadly used system at time of this review.
presented in other chapters of this book. The pooled
Sixty–two of ninety-nine studies were excluded because
data will give you an overview on the level of evidence
the data were not detailed enough, were based on single
on two related hypotheses:
cases, or used different OCT technology, all precluding
1. Retinal axonal loss follows MS optic neuritis any further meta–analysis. Others were excluded
(MSON) and can be quantified by OCT because they were double publications of data with a
2. Retinal axonal loss in MS occurs independently change of emphasis in the interpretation, reviews, or
from MSON and is less severe than MSON comments. Thirty–five studies were included in
These two hypotheses will be further developed using the meta–analysis [1, 3, 5, 7–13, 15–18, 21, 23, 25–28,
the meta–analysis to show important limitations and 30, 31, 36–44, 46–49].
highlight pitfalls of OCT methodology. The chapter
will conclude with a model of the development of Meta-analysis
retinal axonal loss in MS and MSON based on the The Cochrane Collaboration’s Review Manager
present meta-analysis. software package (RevMan5) was used for performing
the meta–analysis. Data on RNFL thickness was used
Method as a continuous variable. Random effects and inverse
variance were used in the model. The selected effect
Definitions measure was the mean difference of RNFL thickness
In this chapter MSON is defined by having at least one in the eyes between: (1) MSON patients and control
episode of optic neuritis as typically seen in MS [35]. subjects, (2) MS patients who did not have MSON and
In order to distinguish, within a given patient with control subjects and (3) from the same MS patients, a
MS, which eye had MSON and which one did not, the comparison of the MSON-eye with the MS-eye. Next,
first will be called MSON-eye and the latter MS-eye. pooled regression analyses on cross-sectional data
were performed using SAS software (version 9.1.3).
Literature search A p-value of ≤ 0.05 was regarded as significant.
The literature review was conducted on all studies using
OCT in MS patients between the first description of Evidence for the use of OCT in MS
the method by Huang [20] and May 2010. Four data-
bases were searched: Pubmed, EMBASE, Medline, Web OCT in MSON-eyes
of Science, and the Cochrane Register of Diagnostic There were 14 studies with data on patients with
Test Accuracy Studies. The search terms were: multiple MSON-eyes and normal control eyes [1, 3, 11, 12,

Optical Coherence Tomography in Neurological Disease, ed. Peter A. Calabresi, Laura J. Balcer, and Elliot M. Frohman. 103
Published by Cambridge University Press. © Cambridge University Press 2015.
 Chapter 9: Meta-analysis of optical coherence tomography in multiple sclerosis

MSON Control Mean Difference Mean Difference

Study or Subgroup Mean SD Total Mean SD Total Weight IV, Random, 95% CI IV, Random, 95% CI
Albrecht 2007 74.47 22.15 21 103.4 10.96 11 3.6% –28.93 [–40.41, –17.45]
Bock 2010 86.2 16.2 73 105.2 9.4 406 11.6% –19.00 [–22.83, –15.17]
Burkholder 2009 85.7 19 328 104.5 10.7 219 13.8% –18.80 [–21.30, –16.30]
Fisher 2006a 85 17 63 105 12 72 9.7% –20.00 [–25.03, –14.97]
Frohman 2009 70.3 13.4 12 101.9 8.9 8 4.6% –31.60 [–41.37, –21.83]
Klistorner 2008 84.5 15.1 32 104 9.2 25 7.8% –19.50 [–25.85, –13.15]
Merle 2008 83.85 24.12 30 106.24 12.46 46 4.9% –22.39 [–31.74, –13.04]
Pueyo 2008 (1) 84.46 0 25 104.97 0 25 Not estimable
Pulicken 2007 84.2 14.7 82 102.7 11.5 94 11.4% –18.50 [–22.44, –14.56]
Ratchford 2009 88.3 16.5 157 102.4 11 77 12.1% –14.10 [–17.66, –10.54]
Sepulcre 2007 (2) 0 0 24 92.3 16.7 58 Not estimable
Siger 2008 83.92 17.63 40 100.3 12.1 24 6.7% –16.38 [–23.68, –9.08]
Trip 2005 68.7 18.8 25 102.9 14.6 15 4.2% –34.20 [–44.64, –23.76]
Zaveri 2008 81.8 19.3 68 104.6 10.3 85 9.6% –22.80 [–27.88, –17.72]

Total (95% CI) 956 1107 100.0% –20.38 [–22.86, –17.91]

Heterogeneity: Tau2 = 9.91; Chi2 = 28.03, df = 11 (P = 0.003); I2 = 61%
–50 –25 0 25 50
Test for overall effect: Z = 16.14 (P < 0.00001)
Favours experimental Favours control
(1) Standard deviation not available from author
(2) Data (mean +/ –SD) of ON eyes not published and not available from author.

Figure 9.1 Meta-analysis of OCT studies in MS patients who did suffer from MSON. The overall averaged RNFL data (mean ±SD) and number
of MSON-eyes investigated are shown for patients and normal subjects. To the right are the data expressed as micron difference in RNFL
thickness of eyes that had optic neuritis compared to normal eyes; the length of the horizontal bar is the 95% confidence interval for each
study. Figure reprinted with permission from [34]

23, 28, 38–40, 44, 48, 49], of which 12 could be meta-analysis from 15 of these studies [1, 3, 11, 12,
included in the meta-analysis [1, 3, 11, 12, 23, 28, 17, 21, 23, 38–40, 44, 48, 49]. In total 3,154 eyes were
38–40, 44, 48, 49]. In total, data from 2,063 eyes investigated.
were included [34]. Figure 9.2 illustrates that the estimated average
Figure 9.1 clearly demonstrates a highly significant RNFL loss in MS-eyes was only about a third (–7.08
loss of RNFL in MSON-eyes. The estimated average μm, [–8.65, –5.52]) compared to what was seen for
RNFL loss was –20.38 μm [–22.86, –17.91]. This is MSON (–20.38 μm [–22.86, –17.91]). In other words,
about a fifth of all axons leaving the eye. the presumed healthy eyes of patients with MS had
Importantly, as discussed by Andrew Henderson in lost approximately a tenth of their retinal axons. This
Chapter 4 and Fiona Costello in Chapter 5, there are a still substantial amount of unexpected axonal loss will
number of other pathologies to be considered which require explanation.
can mimic MSON (see Table 5.2). Particularly severe
RNFL loss, frequently combined with microcystic OCT in MS-eyes with contralateral MSON
macular edema (MMO), is seen following neuromyeli-
The reader may ask at this point if the data
tis optica (NMO) and chronic relapsing inflammatory
presented in the meta-analysis adds up. Can the
optic neuropathy (CRION) [6, 15, 33, 46]. Therefore,
gap of about 13 μm RNFL thickness between
an important limitation of these studies is that infor-
MSON eyes, MS-eyes and control eyes be
mation on previous episodes of optic neuritis was fre-
explained by data derived solely from MS
quently sought retrospectively. In the absence of an
patients? Or may an inclusion or selection bias
international consensus on how to investigate patients
have been introduced on a group level such that
with suspected optic neuritis, it is probable that the
the above two meta-analyses were flawed? To
studies pooled for this meta-analysis were rather
address this question a third meta-analysis was
heterogeneous.
conducted. This time only MS patients with uni-
lateral MSON were included. The OCT data from
OCT in MS-eyes, never MSON the MSON-eyes were compared to the data from
Sixteen studies investigated the RNFL thickness in the presumed unaffected eye from 27 studies [1,
patients with MS who did not suffer from MSON [1, 3, 5, 7–9, 11–13, 16, 23, 25–27, 30, 31, 38–44, 46,
3, 11, 12, 17, 18, 21, 23, 37–40, 44, 48, 49]. Data were 47, 49]. In total, data from 4,199 eyes were
104 sufficiently detailed to permit inclusion into the included into the meta-analysis.
 Ms–non–ON Control Mean Difference Mean Difference
Study or Subgroup Mean SD Total Mean SD Total Weight IV, Random, 95% CI IV, Random, 95% CI
Albrecht 2007 84.59 16.03 27 103.4 10.96 11 2.6% –18.81 [–27.67, –9.95]
Bock 2010 97 13.1 189 105.2 9.4 406 12.1% –8.20 [–10.28, –6.12]
Burkholder 2009 95.6 14.5 730 104.5 10.7 219 12.9% –8.90 [–10.66, –7.14]
Fisher 2006a 96 14 108 105 12 72 8.0% –9.00 [–12.83, –5.17]
Frohman 2009 101 6 12 101.9 8.9 8 3.7% –0.90 [–7.94, 6.14]
Gundogan 2007 107.6 16.3 78 110.9 10.3 76 7.1% –3.30 [–7.60, 1.00]
Henderson 2008a 91.12 12.6 50 98.8 10.5 20 5.0% –7.68 [–13.46, –1.90]
Jeanjean 2008 88.58 13.39 7 102.34 7.47 15 1.9% –13.76 [–24.38, –3.14]
Klistorner 2008 103.8 10.8 32 104 9.2 25 5.7% –0.20 [–5.40, 5.00]
Pueyo 2008 (1) 94.2 0 75 104.97 0 25 Not estimable
Pueyo 2009 97.93 9.08 40 105.37 9.48 20 6.0% –7.44 [–12.46, –2.42]
Pulicken 2007 95.9 14 202 102.7 11.5 94 9.8% –6.80 [–9.82, –3.78]
Ratchford 2009 97.4 13.9 338 102.4 11 77 10.2% –5.00 [–7.87, –2.13]
Siger 2008 94.38 15 62 100.3 12.1 24 4.6% –5.92 [–12.03, 0.19]
Trip 2005 94.6 14.9 25 102.9 14.6 15 2.3% –8.30 [–17.72, 1.12]
Zaveri 2008 95.6 15 87 104.6 10.3 85 8.0% –9.00 [–12.84, –5.16]

Total (95% CI) 1987 1167 100.0% –7.08 [–8.65, –5.52]

Heterogeneity: Tau2 = 4.13; Chi2 = 29.66, df = 14 (P = 0.008); I2 = 53%
–20 –10 0 10 20
Test for overall effect: Z = 8.87 (P < 0.00001)
Favours experimental Favours control
(1) Standard deviation not available from author.

Figure 9.2 Meta-analysis of OCT studies in MS patients who did not suffer from ON, by history. The overall averaged RNFL (mean ±SD) and number of eyes
investigated is shown, similarly to Figure 9.1 Note that the bar graph to the right, which summarizes the difference in RNFL in the asymptomatic eyes compared
to normal eyes, shows there is loss of RNFL even in eyes supposedly not suffering from previous optic neuritis. Figure reprinted with permission from [34]
 Chapter 9: Meta-analysis of optical coherence tomography in multiple sclerosis

Figure 9.3 convincingly corroborates the hypoth- (MS-eye) approaches ≈80 μm (which is about what
esis that a history of optic neuritis results in a greater is seen within one year following MSON) after about
magnitude of axonal loss when compared to MS eyes 20 years (Figure 9.4). Taking away the effect of normal
without such a history. Indeed, the estimated average ageing (about 2 μm per decade [4]), leaves a consider-
RNFL loss between MSON-eyes and MS-eyes was able amount of axonal loss of an estimated 7–9 μm per
close to 13 μm (–13.84 μm [–15.37, –11.72]; please decade of MS disease duration.
note that, for mathematical reasons and internal con-
sistency, the averaged amount of RNFL loss, “x”, is Time in MS-eye, never MSON
presented as “–x”).
With the advent of high-speed, high-definition,
Taken together, our findings necessarily under-
spectral-domain OCT, we are finally in a position
score an important lingering question for the field;
to address one of the remaining and formidably
why and how does a substantial amount of axonal
challenging facets of retinal neurodegeneration in
loss occur in the presumed unaffected eye of a
MS; to dichotomize MS eyes with a history of MS
patient suffering from MS? A possible explanation
ON, and MS eyes with no history of MS ON
will be presented at the end of the chapter, but first,
(at least according to the patient and their family).
the likely time course of development of axonal loss
The principal objective of our most recent work
in the MS eye is reviewed.
has been to delineate the longitudinal impact
upon RNFL thickness in patients without a history
Time course of RNFL loss in MS of MSON. From the three studies we were able
As a rule of thumb, RNFL loss becomes readily to identify [17, 18, 44], a similar picture emerged
detectable to OCT about three months after an epi- (Figure 9.5). Conspicuously, when the data is
sode of acute MSON. Earlier detection of the reduc- analyzed cross-sectionally, we found that, within
tion in RNFL thickness from axonal atrophy is about 12 years, thinning of the RNFL is approach-
difficult to distinguish from a reduction resulting ing 80 μm. Of course, it is not surprising that
from resolution of axonal swelling, which is common there are outliers that collectively, upon analysis,
in acute MSON, as discussed by Andrew Henderson exhibit considerable heterogeneity, with individual
and colleagues in Chapter 4 by Fiona Costello in MS eyes distributed to either fast, slow, or average
Chapter 5 and Robert Bermel et al in Chapter 17 of axonal degeneration. Pragmatically, the identifica-
this book. Their combined data demonstrates that tion of novel protective and/or restorative
about 99% of axonal loss will occur within four to six neurotherapeutic properties would have poten-
following MSON. tially important treatment implications for the
A correlation between disease duration and the MS retina and optic nerve in particular, and for
averaged overall RNFL was found by some [11, neurodegenerative disorders in general.
37, 44] but not by others [18, 23] included in the
meta-analysis (Figure 9.2). The divergent results Time in MSON-eyes
may in part be explained by the variation in aver-
Using a similar cross-sectional meta–analysis it can
age disease duration and a bias in the population
be demonstrated that following the acute axonal
of MS patients studied. A meta-analysis of the
loss associated with MSON [7, 8, 10], little pro-
published cross-sectional data remains statistically
gression seems to be detectable thereafter (Figure
questionable, but may help to shed light on the
9.6). In fact, the RNFL thickness appears to settle
presumed association between RNFL loss and dis-
around ≈80 μm within the first year. This is con-
ease duration. Clearly, accurate information will
sistent with the clinical experience that vision
need to come from carefully conducted longitudi-
remains relatively stable even many years after
nal studies that are executed by multiple colla-
MSON and that the disability outcome is relatively
borative Center teams.
good in patients with optic neuritis. The relevant
clinical point here is that progressive visual
Time in MS-eyes with contralateral MSON loss following presumed MSON is hardly ever
A graph of cross-sectional data analysis illustrates that observed and alternative diagnoses need to be con-
106 the average RNFL thickness in the unaffected eye sidered [35].
 MSON eye MS non–OM eye Mean Difference Mean Difference
Study or Subgroup Mean SD Total Mean SD Total Weight IV, Random, 95% CI IV, Random, 95% CI
Albrecht 2007 74.47 22.15 21 84.59 16.03 27 2.2% –10.12 [–21.36, 1.12]
Bock 2010 86.2 16.2 73 97 13.1 189 4.6% –10.80 [–14.96, –6.64]
Burkholder 2009 85.7 19 328 95.6 14.5 730 5.2% –9.90 [–12.21, –7.59]
Cheng 2007 76.12 14.92 28 96.45 11.73 33 3.5% –20.33 [–27.15, –13.51]
Costello 2006 77.5 29.9 54 99.8 32.5 54 3.0% –22.30 [–34.08, –10.52]
Costello 2008 (1) 89 7.2 21 105.1 3.6 21 4.8% –16.10 [–19.54, –12.66]
Costello 2008a (2) 78.3 15.74 27 104.4 11.29 27 3.3% –26.10 [–33.41, –18.79]
Costello 2009 (3) 82.3 19.7 33 103.7 15.5 45 3.1% –21.40 [–29.50, –13.30]
Fisher 2006a 85 17 63 96 14 108 4.2% –11.00 [–15.96, –6.04]
Frohman 2009 70.3 13.4 12 101.8 6 12 3.0% –31.50 [–39.81, –23.19]
Garcia–Martin 2010 83.27 9.5 20 92.86 4.01 61 4.5% –9.59 [–13.87, –5.31]
Grazioli 2008 81.7 19.2 29 93.6 15.3 31 2.8% –11.90 [–20.72, –3.08]
Klistorner 2008 104 9.2 32 103.8 10.8 32 4.3% 0.20 [–4.72, –5.12]
Kochkorov 2009 89 18 16 95 14 24 2.4% –6.00 [–16.45, –4.45]
Laron 2010 79.1 2.5 47 96.3 1.4 65 5.5% –17.20 [–17.99, –16.41]
Merle 2010 80.81 18.4 31 96.7 15.8 29 2.9% –15.89 [–24.55, –7.23]
Nakamura 2010 84.28 14.18 19 109.45 12.78 9 2.4% –25.17 [–35.68, –14.66]
Oreja–Guevara 2010 76.42 16.87 18 85.52 18.62 18 2.1% –9.10 [–20.71, 2.51]
Pulicken 2007 84.2 14.7 82 93.9 13.1 42 4.2% –9.70 [–14.78, –4.62]
Quelli 2010 78.01 17.43 51 95.24 11.64 65 4.0% –17.23 [–22.79, –11.67]
Ratchford 2009 88.3 16.5 157 97.4 13.9 338 5.0% –9.10 [–12.08, –6.12]
Sepulcre 2007 (4) 85.8 13.9 122 92.3 16.7 58 4.2% –6.50 [–11.46, –1.54]
Siepman 2010 71.15 13.46 27 90.39 13.46 38 3.6% –19.24 [–25.88, –12.60]
Siger 2008 83.92 17.63 20 91.08 19.3 20 2.1% –7.16 [–18.62, 4.30]
Spain 2009 75.81 5.85 24 90.93 2.95 24 5.1% –15.12 [–17.74, –12.50]
Talman 2010 83 18 208 96 13 381 5.0% –13.00 [–15.77, –10.23]
Zaveri 2008 81.8 19.3 68 95.6 15 87 4.0% –13.80 [–19.37, –8.23]

Total (95% CI) 1631 2568 100.0% –13.84 [–15.97, –11.72]

Heterogeneity: Tau2 = 21.31; Chi2 = 173.33, df = 26 (P < 0.00001); I2 = 85%
Test for overall effect: Z = 12.76 (P < 0.00001) –20 –10 0 10 20
Favours experimental Favours control
(1) Data on one year after ON. The authors also present 2-year data which is not included into this analysis.
(2) The one year after (13–18 months follow-up) data is taken. Data on standard deviation kindly provided by Dr Fiona Costello.
(3) The one year data on RRMS is presented.
(4) This study included 24 ON eyes and 98 eyes without ON. More details not available from authors.

Figure 9.3 Meta-analysis of OCT studies comparing only MS patients: MSON-eyes and MS-eyes (never affected by MSON). The overall averaged RNFL (mean ±SD) and number of
eyes investigated is shown. The difference between the MSON- and MS-eyes shown in this figure is less than the difference between MSON-eyes and normal control eyes Figure 9.1
(note that the scale of the x-axis differs between the two figures). Figure reprinted with permission from [34]
 Chapter 9: Meta-analysis of optical coherence tomography in multiple sclerosis

120 Figure 9.4 Cross-sectional, averaged longitudinal

RNFL changes in the unaffected eye of MS patients
who suffered from MSON (blue dots). The relatively
preserved RNFL (thickness ≈ 105 μm) early in the
[10] [10] disease may allow detection of subtle loss of retinal
[24] [7] [7] [7] [7] axons during the disease course of MS, at least on a
100 [8] [35] group level. The numbers adjacent to the dots
[5] [11] indicate the respective references.
[36] [47] [16] [21]
RNFL [µm]

[43]

[7] [7]
80

60
0 50 100 150 200 250
Averaged months from onset of contralateral MSON

120 Figure 9.5 The averaged longitudinal RNFL changes in the eyes
of MS patients who never suffered from MSON (green dots)
[17] suggests that subtle retinal axonal loss may be a feature of
insidious axonal attrition in MS, as proposed by Hoyt et al. [19], at
100 least on a group level. The numbers adjacent to the dots indicate
RNFL [µm]

[18] the respective references.

[43]
[18]
80

60
0 50 100 150 200 250
Averaged months from onset of MS

One needs to be very careful drawing conclusions In this context, it should be highlighted that the
on the time course of RNFL loss in an individual eye current duration of phase II trials in MS is fre-
from cross-sectional meta-analyses of averaged data. quently around 4–6 months, thus it is not likely
Extrapolation from this data is risky because of a that OCT will provide a reliable outcome over this
possible inclusion bias of patients with early visual time scale. From the preceding data, one would
problems having OCT and likely exclusion of expect that MS patients without MSON will require
patients with severe, late progressive MS not follow-up for at least two years. Again, the reader is
included in an observational research study. referred to the excellent chapters by Andrew
Nevertheless, the results suggest that, in MS without Henderson and colleagues (Chapter 4) and Robert
MSON, there is an estimated yearly thinning of the Bermel (Chapter 17).
overall RNFL in the range of ≈ 1–2 μm. Reassuringly,
the data from the meta–analysis is consistent with Limitations and pitfalls
the longitudinal data by Talman and colleagues Taken together from Evidence for the use of OCT in
108 suggesting a yearly loss of around 1 μm [47]. MS (page 103) and Time course of RNFL loss in MS
 Chapter 9: Meta-analysis of optical coherence tomography in multiple sclerosis

120 Figure 9.6 Averaged longitudinal RNFL changes in the eye

affected by MSON (red dots). The degree of the damage caused to
the RNFL by MSON appears cross-sectionally to be so large (RNFL
thickness ≈ 80 μm) that any subtle damage that may be attributed
to trans-synaptic axonal degeneration could be masked. The
100 numbers adjacent to the dots indicate the respective references.
RNFL [µm]

[10]
[37][24] [7] [7] [7] [43][11] [28] [7]
[10] [16] [7]
80 [8] [7]
[5] [29]

[47]

60
0 50 100 150 200
Averaged months from onset of MSON

(page 106) in this Chapter the averaged RNFL loss is hours [2]. Larger changes of RNFL thickness measures
largest in MSON-eyes with –20.38 μm, followed by can be introduced by artifacts related to the method
presumably normal MS–eyes of –7.08 μm. In addi- itself. Therefore, rigorous OCT quality control criteria
tion, most damage following MSON happens rapidly are required to safeguard accuracy and validity of the
within 4–12 months following the event. In contrast, in data. This will be discussed in chapter 18.
presumably normal MS-eyes, RNFL loss seems to man-
ifest itself at a rate of around 1–2 μm per year. Taken Putting it together: a model
together, these two data sets imply an accelerated rate At the beginning of this chapter, it was suggested that a
of axonal degeneration, whereas we have another meta-analysis may help to enable a holistic interpreta-
group that instead exhibits an insidiously slow and tion of OCT data in MS research. A possible interpreta-
steady degeneration attributed to trans-synaptic axonal tion of the data of the meta-analysis is summarized in
degeneration. These two presentations (rapid vs slow Figure 9.7. The conceptual framework of the visual
or retrograde axonal vs trans-synaptic retrograde axo- pathway model has not only inspired the present, but
nal degeneration), on opposite ends of the injury spec- also other Chapters in this book. It serves as a template
trum (optic nerve vs posterior visual pathways), may be stimulating formulation of hypotheses which will ulti-
instructive with respect to the application of specific mately help us to link OCT as a tool to other means of
treatment modalities that may be aimed at one set of assessing structure and function with the overall aim to
mechanisms, but not necessarily both. understand and treat neurodegeneration.
Importantly, any form of optic neuritis will mask, The anatomical basis is discussed in detail by Devin
subtle and slow axonal loss in the eye of patients with Mackay et al in chapter 3. In brief, retinal axons origi-
MS. Therefore, any study and analysis on a presumed nate from ganglion cells. Retinal axons travel through
relationship of RNFL loss with pathology or treatment the optic nerve to the chiasm, where those supplying the
need to carefully distinguish MS-eyes from MSON- temporal field of vision cross to the other side. From
eyes. there, axons reach the dorsal lateral geniculate nucleus
In the acute phase of MSON, optic disc edema and (LGN), where they end and synapse. The second axon
axonal swelling will mask RNFL thinning. This may within the optic radiations project to the primary visual
considerably limit the prognostic value of OCT in the cortex (V1, see Figure 9.7A). In addition, to these major
acute phase. Whether or not MMO, inner nuclear projections, axons from the retina also reach the tec-
layer thickening and/or periphlebits might be of addi- tum, V2 and V5. It is suggested that any lesion within
tional prognostic value remains to be seen and will be this trajectory can cause thinning of the RNFL, as
discussed in Chapter 4, 10 and 13. revealed by the meta-analysis. The data of the meta-
The rate of RNFL loss in MS-eyes is slow. An annual analysis are further supported by histological the data
decrease of around 1–2 μm is in the range of the presented in Chapter 3 and Chapter 14. Logically, there
physiological variation in humans seen with a few are three potential locations were a lesion may occur in 109
 Chapter 9: Meta-analysis of optical coherence tomography in multiple sclerosis

A C D
Eye RNFL RNFL RNFL

RGC ON RGC

LGN

Optic
Optic Radiations
Radiations

Occiptial

Time

Time
cortex

B RNFL RNFL RNFL

ON RGC ON ON

LGN

Optic
Radiation

Figure 9.7 A model of the presumed relationship between RNFL thickness and MS pathology. (A) A simplified sketch of the normal human
visual pathways. The retinal ganglion cells (RGC) send their unmyelinated axons into the eye, where they form the retinal nerve fiber layer (RNFL,
gray inlay), travel to the optic disc, and leave the orbit. Once the axons pass the sclera, they become myelinated and form the optic nerve (ON).
After passing through the chiasma, where the temporal fibers cross (not shown), they are called the optic tract. The optic tract wraps around
the midbrain and enters the lateral geniculate nucleus (LGN), where all axons must synapse. After the LGN the axons fan out through the deep
white matter (optic radiations) to reach the occipital cortex. (B) MSON (red) directly causes acute axonal loss in the ON (red dotted line), leading
to marked thinning of the RNFL (small gray box). (C) MS lesions within the optic radiations (blue dotted line) do not immediately result in RNFL
thinning. This is thought to be a chronic consequence of trans-synaptic axonal loss through the LGN. With time (black arrow) trans-synaptic
axonal degeneration causes a relative smaller degree of axonal loss in the ON (red dashed-dotted line) with a quantifiable degree of RNFL loss
(gray box). (D) Progressive loss of RGC (yellow dot) is a likely result of chronic changes in the anterior visual pathways themselves and causes a
small degree of RNFL loss (gray box). Figure reprinted with permission from [34]

this trajectory: within the optic nerve (Figure 9.7B), at immediate, severe thinning of the RNFL
the level of the optic radiations (Figure 9.7C) or at the (Figure 9.7B).
level of the ganglion cell (Figure 9.7D). 2. Axonal transsection in the optic radiations caus-
Possible underlying pathophysiolgoical mechan- ing retrograde axonal degeneration (towards the
isms are: dorsal LGN). In the dorsal LGN degeneration of
1. Axonal transsection in the optic nerve during a damaged axon (in the optic radiation) is passed
MSON causing retrograde axonal degeneration on to a previously healthy axon (in the optic
(that is, towards the retinal ganglion cell) and nerve) by a process called trans-synaptic axonal
Wallerian (anterograde) axonal degeneration degeneration [22]. The likely cause for axonal
(towards the LGN). Because axons are densely transsection in the optic radiations is a MS
packed in the optic nerve, a lesion is likely lesion. Because axons of the optic radiations are
to affect a large number of axons, causing more widespread in the brain, a single MS lesion
110
 Chapter 9: Meta-analysis of optical coherence tomography in multiple sclerosis

is likely to affect less axons, resulting in less 8 Fiona Costello et al. “Quantifying axonal loss after optic
severe RNFL thinning compared to MSON, neuritis with optical coherence tomography.” eng. In:
which also takes more time to establish (Figure Ann Neurol 59.6 (June 2006), pp. 963–969.
9.7C). 9 Fiona Costello et al. “Retinal nerve fiber layer and
3. Primary neurodegeneration originating from future risk of multiple sclerosis.” eng. In: Can J Neurol
Sci 35.4 (Sept. 2008), pp. 482–487.
the retinal ganglion cell which subsequently
leads to anterograde axonal degeneration 10 F. Costello et al. “Tracking retinal nerve fiber layer loss
after optic neuritis: a prospective study using optical
(Figure 9.7D).
coherence tomography.” eng. In: Mult Scler 14.7 (Aug.
2008), pp. 893–905.
It seems conceivable that neurodegeneration of
the second order neuron in the LGN and the third- 11 Jennifer B Fisher et al. “Relation of visual function
order neuron in the visual cortex also may result in to retinal nerve fiber layer thickness in multiple
sclerosis.” eng. In: Ophthalmology 113.2 (Feb. 2006),
RNFL thinning by retrograde trans-synaptic degen-
pp. 324–332.
eration. This will need to be tested by combining
12 Elliot M Frohman et al. “Relationship of optic
structural and functional brain imaging techniques
nerve and brain conventional and non-conventional
with OCT in longitudinal studies. Likewise, a holistic MRI measures and retinal nerve fiber layer
model may include biomarkers able to identify axonal thickness, as assessed by OCT and GDx: a pilot
damage in the acute phase of inflammatory demyeli- study.” eng. In: J Neurol Sci 282.1–2 (July 2009),
nating syndromes [33, 35]. pp. 96–105.
13 Elena Garcia-Martin et al. “Progressive changes in
the retinal nerve fiber layer in patients with multiple
References sclerosis.” eng. In: Eur J Ophthalmol 20.1 (2010),
1 Philipp Albrecht et al. “Optical coherence tomography pp. 167–173.
measures axonal loss in multiple sclerosis 14 Jeffrey M. Gelfand et al. “Microcystic Inner Nuclear
independently of optic neuritis.” eng. In: J Neurol Layer Abnormalities and Neuromyelitis Optica.” eng.
254.11 (Nov. 2007), pp. 1595–1596. In: JAMA Neurol (Mar. 2013), pp. 1–5.
2 Lisanne J. Balk et al. “The physiological variation 15 E. Gordon-Lipkin et al. “Retinal nerve fiber layer
of the retinal nerve fiber layer thickness and macular is associated with brain atrophy in multiple
volume in humans as assessed by spectral sclerosis.” eng. In: Neurology 69.16 (Oct. 2007),
domain-optical coherence tomography.” eng. pp. 1603–1609.
In: Invest Ophthalmol Vis Sci 53.3 (Mar. 2012), 16 Erica Grazioli et al. “Retinal nerve fiber layer thickness
pp. 1251–1257. is associated with brain MRI outcomes in multiple
3 Markus Bock et al. “Patterns of retinal nerve fiber layer sclerosis.” eng. In: J Neurol Sci 268.1–2 (May 2008),
loss in multiple sclerosis patients with or without optic pp. 12–17.
neuritis and glaucoma patients.” eng. In: Clin Neurol 17 Fatih C Gundogan, Seref Demirkaya, and Gungor
Neurosurg (May 2010). Sobaci. “Is optical coherence tomography really a new
4 Donald L Budenz et al. “Determinants of normal retinal biomarker candidate in multiple sclerosis?–A
nerve fiber layer thickness measured by Stratus OCT.” structural and functional evaluation.” eng. In:
eng. In: Ophthalmology 114.6 (June 2007), pp. 1046– Invest Ophthalmol Vis Sci 48.12 (Dec. 2007), pp. 5773–
1052. 5781.
5 Han Cheng et al. “The relationship between visual field 18 Andrew P D Henderson et al. “An investigation of
and retinal nerve fiber layer measurements in patients the retinal nerve fibre layer in progressive multiple
with multiple sclerosis.” eng. In: Invest Ophthalmol Vis sclerosis using optical coherence tomography.” eng.
Sci 48.12 (Dec. 2007), pp. 5798–5805. In: Brain 131.Pt 1 (Jan. 2008), pp. 277–287.
6 Stacey S Choi et al. “Changes in cellular structures 19 W. F. Hoyt, B. Schlicke, and R. J. Eckelhoff.
revealed by ultra-high resolution retinal imaging in “Fundoscopic appearance of a nerve-fibre-bundle
optic neuropathies.” eng. In: Invest Ophthalmol Vis Sci defect.” eng. In: Br J Ophthalmol 56.8 (Aug. 1972),
49.5 (May 2008), pp. 2103–2119. pp. 577–583.
7 Fiona Costello et al. “Differences in retinal nerve 20 D. Huang et al. “Optical coherence tomography.” eng.
fiber layer atrophy between multiple sclerosis In: Science 254.5035 (Nov. 1991), pp. 1178–1181.
subtypes.” eng. In: J Neurol Sci 281.1–2 (June 2009), 21 L. Jeanjean et al. “[Retinal atrophy using optical
pp. 74–79. coherence tomography (OCT) in 15 patients with 111
 Chapter 9: Meta-analysis of optical coherence tomography in multiple sclerosis

multiple sclerosis and comparison with healthy sub- meta-analysis.” eng. In: Lancet Neurol 9.9 (Sept. 2010),
jects].” fre. In: Rev Neurol (Paris) 164.11 (Nov. 2008), pp. 921–932.
pp. 927–934. 35 A. Petzold et al. “The investigation of acute optic
22 P. Jindahra, A. Petrie, and G. T. Plant. “Retrograde neuritis: a review and proposed protocol.” Nature Rev
trans-synaptic retinal ganglion cell loss identified by Neurol 2014; 10: 447–458.
optical coherence tomography.” eng. In: Brain 132
36 Victoria Pueyo et al. “Sub-clinical atrophy of the retinal
(2009), pp. 628–634.
nerve fibre layer in multiple sclerosis.” eng. In: Acta
23 Alexander Klistorner et al. “Axonal loss and myelin in Ophthalmol (June 2009), pp. 1–5.
early ON loss in postacute optic neuritis.” eng. In: Ann
37 V. Pueyo et al. “Axonal loss in the retinal nerve fiber
Neurol 64.3 (Sept. 2008), pp. 325–331.
layer in patients with multiple sclerosis.” eng. In: Mult
24 Alexander Klistorner et al. “Multifocal VEP and OCT Scler 14.5 (June 2008), pp. 609–614.
in optic neuritis: a topographical study of the structure-
function relationship.” eng. In: Doc Ophthalmol 118.2 38 M. Pulicken et al. “Optical coherence tomography and
(Apr. 2009), pp. 129–137. disease subtype in multiple sclerosis.” eng. In:
Neurology 69.22 (Nov. 2007), pp. 2085–2092.
25 A. Kochkorov et al. “Rigidity of retinal vessels in
patients with multiple sclerosis.” eng. In: Klin Monatsbl 39 Amanda Quelly et al. “Comparison of Optical
Augenheilkd 226.4 (Apr. 2009), pp. 276–279. Coherence Tomography and Scanning Laser
Polarimetry Measurements in Patients with Multiple
26 Michal Laron et al. “Comparison of multifocal visual
Sclerosis.” eng. In: Optom Vis Sci (May 2010).
evoked potential, standard automated perimetry and
optical coherence tomography in assessing visual 40 J. N. Ratchford et al. “Optical coherence tomography
pathway in multiple sclerosis patients.” eng. In: Mult helps differentiate neuromyelitis optica and MS optic
Scler 16.4 (Apr. 2010), pp. 412–426. neuropathies.” eng. In: Neurology 73.4 (July 2009), pp.
302–308.
27 Harold Merle et al. “Retinal nerve fiber layer thickness
and spatial and temporal contrast sensitivity in multi- 41 A. R. Salter et al. “Retinal architecture predicts pupil-
ple sclerosis.” eng. In: Eur J Ophthalmol 20.1 (2010), pp. lary reflex metrics in MS.” eng. In: Mult Scler 15.4 (Apr.
158–166. 2009), pp. 479–486.
28 Harold Merle et al. “Retinal peripapillary nerve fiber 42 Jorge Sepulcre et al. “Diagnostic accuracy
layer thickness in neuromyelitis optica.” eng. In: of retinal abnormalities in predicting disease
Invest Ophthalmol Vis Sci 49.10 (Oct. 2008), pp. 4412– activity in MS.” eng. In: Neurology 68.18
4417. (May 2007), pp. 1488–1494.
29 R. T. Naismith et al. “Optical coherence tomography 43 Theodora A M Siepman, Marijke Wefers Bettink-
differs in neuromyelitis optica compared with multiple Remeijer, and Rogier Q Hintzen. “Retinal nerve fiber
sclerosis.” eng. In: Neurology 72.12 (Mar. 2009), layer thickness in subgroups of multiple sclerosis,
pp. 1077–1082. measured by optical coherence tomography and
30 Masahiko Nakamura et al. “Early high-dose scanning laser polarimetry.” eng. In: J Neurol (May
intravenous methylprednisolone is effective in 2010).
preserving retinal nerve fiber layer thickness in patients 44 Malgorzata Siger et al. “Optical coherence
with neuromyelitis optica.” eng. In: Graefes Arch Clin tomography in multiple sclerosis: thickness of the
Exp Ophthalmol (Mar. 2010). retinal nerve fiber layer as a potential measure of
31 C. Oreja-Guevara et al. “Optic neuritis, multiple axonal loss and brain atrophy.” eng. In: J Neurol
sclerosis-related or not: structural and functional 255.10 (Oct. 2008), pp. 1555–1560.
study.” engspa. In: Neurologia 25.2 (Mar. 2010), 45 Elias S. Sotirchos et al. “In vivo identification of
pp. 78–82. morphologic retinal abnormalities in neuromyelitis
32 A Petzold and GT Plant. “Chronic Relapsing optica.” eng. In: Neurology 80.15 (Apr. 2013),
Inflammatory Optic Neuropathy: a Systematic Review pp. 1406–1414.
of 122 cases reported.” In:J Neurol ([Epub ahead of 46 Rebecca I Spain et al. “Thickness of retinal nerve fiber
print]). layer correlates with disease duration in parallel
33 Axel Petzold. “Neurofilament phosphoforms: surrogate with corticospinal tract dysfunction in untreated mul-
markers for axonal injury, degeneration and loss.” eng. tiple sclerosis.” eng. In: J Rehabil Res Dev 46.5 (2009),
In: J Neurol Sci 233.1–2 (June 2005), pp. 183–198. pp. 633–642.
34 Axel Petzold et al. “Optical coherence tomography 47 Lauren S Talman et al. “Longitudinal study
in multiple sclerosis: a systematic review and of vision and retinal nerve fiber layer thickness in
112
 Chapter 9: Meta-analysis of optical coherence tomography in multiple sclerosis

multiple sclerosis.” eng. In: Ann Neurol 67.6 (June 49 Maulik S Zaveri et al. “Retinal imaging by laser
2010), pp. 749–760. polarimetry and optical coherence tomography
48 S. Anand Trip et al. “Retinal nerve fiber layer axonal evidence of axonal degeneration in multiple
loss and visual dysfunction in optic neuritis.” eng. In: sclerosis.” eng. In: Arch Neurol 65.7 (July 2008),
Ann Neurol 58.3 (Sept. 2005), pp. 383–391. pp. 924–928.

113
 Chapter
Optical coherence tomography and

10 brain magnetic resonance imaging

in multiple sclerosis
Shiv Saidha and Peter A. Calabresi

Introduction paradox) [17]. Similarly, T2 lesion quantification at

Multiple sclerosis (MS) is considered an immune- baseline in MS has been found to have modest pre-
mediated disorder of the central nervous system dictive value for the development of disability [18].
(CNS) and is the most common nontraumatic cause On the other hand, MRI measures of neurodegenera-
of neurologic disability in early to middle adulthood tion and atrophy seem to correlate better with dis-
throughout the developed world [1]. Although the ability progression. Numerous studies have shown
precise etiology of MS remains to be completely that MRI measures of whole brain atrophy, such as
elucidated, known pathologic hallmarks of MS lesions brain parenchymal fraction (BPF), are inversely
include the breakdown of the blood-brain barrier, related to disability progression in MS [19, 20]. In
demyelination, axonal degeneration, neuronal loss, addition to offering the capability of whole-brain
and astrogliosis [2, 3]. Despite being conventionally volume measurement, MRI segmentation techniques
designated as primarily an inflammatory demyelinat- enable parcellation of the brain into its component
ing disorder of the CNS, early descriptions of MS substructures, allowing the measurement of brain
highlight the presence of prominent axonal and neu- substructure volumes (Figure 10.1). Gray matter
ronal pathology [4, 5]. In recent years, axonal and (GM) degeneration in particular has been shown to
neuronal pathology in MS have regained considerable be extremely common in MS, beginning early in the
attention. Investigation of neurodegeneration in MS course of MS, and may perhaps be more closely
has not only led to improvements in our understanding linked with disability than white matter (WM)
of the biological underpinnings of the disease, but also degeneration [11, 12, 21–23]. Histopathological
led to a wide recognition that axonal and neuronal examination of MS brain tissue has shown that
degeneration are the principal pathological substrates intense inflammation occurs within the cortical GM
of permanent disability in MS [6–12]. Although axonal in MS, a process that appears to begin early in the
and neuronal degeneration may be the sequelae of course of the disease [24]. The primary antigenic
inflammatory demyelination in MS [13–16], there target of GM inflammation in MS remains to be
may be other pathobiological processes operating as definitively established and may not be restricted
part of the disease process that contribute to their to myelin targets.
etiology. Although cortical demyelinating lesions account
Magnetic resonance imaging (MRI) of the brain is for a large proportion of all brain lesions in
one of the most widely established modalities for detect- MS (approximately 26–59%) [25, 26], the precise
ing disease activity in MS. Although conventional pathophysiologic basis underlying neuronal injury and
MRI may be sensitive enough to capture inflammatory neuronal loss in MS requires determination. While
disease activity in MS by way of new or enlarging neuronal loss in MS may be caused by retrograde
T2/FLAIR lesions or contrast-enhancing lesions, [13, 15] or anterograde degeneration [14], it remains
the association between MRI markers of inflamma- unclear whether neuronal degeneration may also result
tion and disability progression during the course of from processes independent of demyelination or axonal
MS is modest at best (i.e., the clinico-radiologic injury in MS. Prior observations raise the possibility that

114 Optical Coherence Tomography in Neurological Disease, ed. Peter A. Calabresi, Laura J. Balcer, and Elliot M. Frohman.
Published by Cambridge University Press. © Cambridge University Press 2015.
 Chapter 10: Optical coherence tomography and brain magnetic resonance imaging in multiple sclerosis

Figure 10.1 The top-left panel represents an

axial MP-RAGE image in which gray–white differ-
entiation is clear. The top-right panel represents an
axial FLAIR scan in which MS lesions are obvious.
The bottom-left panel represents segmentation of
the brain MRI into cortical GM, cerebral WM, deep
GM structures, and lesions (all of which are color-
coded) by the LESION-TOADS segmentation
algorithm. The bottom-right panel depicts the
same axial FLAIR scan as located in the top-right
panel, but with the lesions identified by the
segmentation algorithm (colored red) [72].

in some instances the mechanism of GM tissue injury in also found in the brain parenchyma of MS patients [29].
MS may differ from that of WM tissue injury, and Such findings serve to remind us not only of the
that GM injury may be the derivative of a primary complexity but also of the heterogeneity of MS.
neuronal pathobiology [27, 28]. This contemporary Examples of nonconventional MRI techniques for
principle should emphasize the inherent potential dissecting mechanisms of neurodegeneration in MS
advantages of assessing brain substructure volumes, include diffusion tensor imaging (DTI), magnetiza-
such as cortical GM and cerebral WM volumes, rather tion transfer imaging (MTI), and magnetic resonance
than just whole-brain volumes in MS. For instance, spectroscopy (MRS). DTI and MTI are sensitive to
measures of neurodegeneration derived from differing microstuctural tissue properties [30]. DTI utilizes
brain compartments could potentially reflect different information pertaining to the orientation and magni-
pathobiological processes, which are operative in the tude of water diffusion within tissue. This provides
disease process. Along this same line of reasoning, the insight into the integrity of the examined tissue.
emerging role of B-cells in the immunopathogenesis of DTI indices include fractional anisotropy (FA; the
MS has also lead to the reappraisal of MS as a primarily extent to which water diffuses along the principal
T-cell mediated disorder, and it is now recognized that eigen vector divided by the sum of diffusion in all
antigen-experienced B-cell clones, which are found in the orthogonal plans), mean diffusivity (MD; an over- 115
lymphoid follicles in the meninges of MS patients, are all measure of water diffusion), perpendicular
 Chapter 10: Optical coherence tomography and brain magnetic resonance imaging in multiple sclerosis

diffusivity (λ⊥; water diffusion perpendicular to and reproducibility, with highly satisfactory test–retest
intact axon fibers), and parallel diffusivity (λ∥; water variability [32–35].
diffusion along or parallel to intact axon fibers). MTI The innermost layer of the retina is called the
relies on measuring the transfer of longitudinal magne- RNFL and is principally composed of unmyelinated
tization between rotationally restricted hydrogen pro- axons. In fact, under normal circumstances the retina
tons within macromolecules (bound to lipids or constitutes an unmyelinated CNS structure, making it
proteins) and the free water pool. The magnetization an ideal structure for investigating neurodegeneration
transfer ratio (MTR) is, therefore, thought to be more in MS, since axonal and neuronal measures derived
sensitive to myelin than conventional MRI. MRS has from the retina are not confounded by the concomi-
emerged as a technique to assess neuronal and axonal tant presence of myelin. This in itself represents a
damage by measuring the concentration of brain meta- major advantage of examining the process of neuro-
bolites such as N-acetyl-aspartate (NAA), which is a degeneration, as well as the impact of factors that may
neuronal marker [31]. Even though nonconventional alter neurodegeneration (such as potentially neuro-
MRI techniques, such as brain substructure volumetrics, protective and neurorestorative therapies), in the
provide measures that may correlate well with disability retina as opposed to the brain, where myelin is abun-
in MS, and may therefore be informative, performing dant, even in the cortical GM. The axons of the RNFL
these techniques can be time-consuming, and they are (about 1.2 million per eye) are derived from the gang-
generally not available in routine clinical practice. lion cell neurons, which are located in the ganglion
Moreover, there are a multitude of factors which may cell layer (GCL) below the RNFL (Figure 10.2). These
influence these measures, such as inflammation, age, axons coalesce at the optic discs to form the optic
disease modifying therapies, and hydration status. nerves and then exit posteriorly through the lamina
Consequently, there has been a drive to develop alter- cribrosa, an important zone of demarcation beyond
native techniques, including additional nonconven- which oligodendrocyte-derived myelin is applied to
tional MRI techniques, as well as techniques unrelated these axons retro-orbitally, thereby signifying a
to MRI for assessing neurodegeneration in vivo in MS. transition from membrane (slow) to saltatory (fast)
Optical coherence tomography (OCT) is a high- conduction properties [36, 37].
resolution imaging technique that conventionally There is a clear predilection for affliction of the
enables the quantitative estimation of peripapillary- optic nerves in MS, not just clinically such as from
retinal nerve fiber layer (p-RNFL) thickness, as well as acute optic neuritis (AON), which occurs during the
total or average macular thickness (a nonspecific course of the disease in 30–70% of patients, but also
measure of the combined thickness of all of the layers subclinically, reflecting subclinical or occult optic
of the retina). OCT is relatively inexpensive, reprodu- neuropathy. Optic nerve pathology could be
cible, easily repeatable, noninvasive, and well toler- regarded as virtually ubiquitous in the MS disease
ated. Recent refinements in OCT technology have led process, since 94–99% of MS patients are found to
to the powerful and highly precise capability to also exhibit demyelinating optic nerve lesions upon post-
automate the segmentation of the distinctive layers mortem examination [38, 39]. Optic nerve demyeli-
comprising the retina in the macular region, for the nation (whether a derivative of acute or occult optic
purpose of investigating the integrity of its architec- neuropathy) has long been recognized to result in
ture in normal subjects, and in those afflicted with axonal dysfunction, transection, and ultimately ret-
disorders that target the visual system such as MS. In rograde degeneration toward the retina (along with
particular, modern, high-speed, high-definition, Wallerian degeneration distally), eventually produ-
spectral-domain OCT is capable of rendering high- cing atrophy of the RNFL and GCL compartments
resolution images, from which the individual retinal within the retina [15, 40, 41]. In keeping with evi-
layers can be discriminated and objectively quantified dence-based characterizations of postmortem visual
(Figure 10.2). For instance, the macular-RNFL system analysis in MS [40, 41], OCT has provided
(m-RNFL), combined ganglion cell layer and inner replete in vivo evidence of p-RNFL and GCL thin-
plexiform layer (GCIP), combined inner nuclear layer ning in MS eyes, irrespective of a history of AON or
and outer plexiform layer (INL), and the outer nuclear not [33, 34, 42 43].
layer, including the photoreceptor segments (ONL), Since the RNFL represents the axonal tracts of the
116 can now be segmented with great facility, accuracy, ganglion cell neurons, GCL reduction is believed to be
 Chapter 10: Optical coherence tomography and brain magnetic resonance imaging in multiple sclerosis

C
ILM
RNFL
B
GCL
ILM
INL GCL
IPL
ONL RNFL
OPL
IPL INL

Fovea OPL

ONL

ELM

IS/OS PR
IS ELM
RPE

OS RPE

Figure 10.2 Panel A represents a fundus photograph from a healthy control. Panel B is a 3-D macular volume cube generated by Cirrus HD-
OCT from the macular region denoted by the red box from the same healthy control. Note the individual layers of the retina are readily
discernible, except for the ganglion cell layer (GCL) and inner plexiform layer (IPL), which are difficult to distinguish. During segmentation of the
OCT image, segmentation software identifies the outer boundaries of the macular retinal nerve fiber layer (RNFL), inner plexiform layer (IPL),
and outer plexiform layer (OPL), as well as the inner boundary of the retinal pigment epithelium (RPE), which is identified by the conventional
Cirrus HD-OCT algorithm. The identification of these boundaries facilitates OCT segmentation, enabling the determination of the thicknesses of
the macular-RNFL, GCL+IPL (GCIP), the inner nuclear layer (INL)+OPL, and the outer nuclear layer (ONL), including the inner and outer
photoreceptor segments. Panel C illustrates the cellular composition of the retinal layers depicted in panel B.
Abbreviations: IS: inner photoreceptor segments; OS: outer photoreceptor segments; IS/OS: IS/OS junction; PR: photoreceptors; ILM: inner
limiting membrane; ELM: external limiting membrane
Reproduced with permission from JAMA Neurology

the result of retrograde axonal degeneration of the potentially superior structure–function relationships
retinal nerve fibers, i.e., reduction in both RNFL and of GCIP over p-RNFL thickness measurements. GCIP
GCL are thought to be derivative of the same patho- thickness measurements appear to have superior
logic processes, namely optic neuropathy (and asso- reproducibility than RNFL thickness measurements
ciated tissue injury mechanisms). Despite the biologic [32]. In addition, it has been proposed that the GCL
similarity of the both the RNFL and GCL, OCT may not be prone to edema, unlike the RNFL, during
derived measures of GCIP thickness seem to have optic nerve inflammation [34]. Microinflammatory
superior structure–function relationships than those processes, if operative in the optic nerve(s), may
of RNFL thickness [33]. This appears to be the case therefore not confound GCIP measurements similar
not only with visual measures acquired at high to RNFL measurements [32, 33].
contrast (100%) and low contrast (2.5% and 1.25%), Notwithstanding the importance of the recent and
but also with expanded disability status scale (EDSS) abundant evidence corroborating that both RNFL
estimates of global MS disability. Given the biologic and GCL pathology are common and germane to
similarity of both the RNFL and GCIP, and that these visual system pathophysiology in MS, there is
two retinal compartments may be regarded as mere coincident evidence that deeper retinal (inner nuclear
extensions of one another, a number of potential layer and outer nuclear layer) pathology also occurs
explanations have been proposed to explain the in MS. Consistent with electroretinographic and 117
 Chapter 10: Optical coherence tomography and brain magnetic resonance imaging in multiple sclerosis

Figure 10.3 An optical coherence tomography (OCT) peripapillary retinal nerve fiber layer (p-RNFL; Panel A) and macular (Panel B) report
generated by Cirrus HD-OCT (Carl Zeiss Meditec, Dublin, California) in a patient with multiple sclerosis without history of optic neuritis. The
upper-middle section of Panel A displays the average p-RNFL thickness for the right eye (OD) and the left eye (OS), as well as the quadrant and
clockhour measures of p-RNFL thickness for each eye. Note that the average p-RNFL thickness, as well as the quadrant and clockhour measures,
is represented in colors that correspond to the normal distribution of p-RNFL thickness values. The average p-RNFL thickness (as well as
quadrants and sectors) in each eye is represented in green (indicating values within normal range).
The top-right section of panel B displays quadrant measurements of retinal thickness. Again, these are represented in colors that
correspond to the normal distribution of macular thickness values. The central macula represents the foveola, with the four quadrants
immediately surrounding this (inner macula) representing the parafoveola. Note that the average macular thickness (cube average thickness)
indicated in the bottom-right chart (as well as all of the macular quadrant thicknesses) are represented in red, indicating values less than 1% of
what would be expected compared to an age-matched reference population. The macular scan of the left eye in the same patient (not shown)
is similar to that of the right eye in panel B. The combination of OCT findings described in this figure fulfills our criteria for a macular-thinning-
predominant (MTP) patient. Please note that this represents a patient with an average macular thickness < 1st percentile. Those with a normal
average p-RNFL thickness and average macular thickness between the 1st and 5th percentile also fulfill our criteria for an MTP patient.
Reproduced with permission from Brain

postmortem findings [40, 44–47], OCT-segmentation phenotype of MS [32] (Figure 10.3). The MTP phe-
demonstrates quantitative inner nuclear layer (INL) notype may be associated with more rapid accumula-
and outer nuclear layer (ONL) abnormalities in MS tion of sustained disability in MS [32], and the
[32, 33, 35]. Although the definitive etiological rudi- mechanisms of INL and ONL changes may be patho-
ments of INL and ONL pathology in MS remains to be biologically distinct from those that mediate tissue
determined, recent observations support the possibi- injury within the RNFL and GCL, although this
lity that a primary retinal neuronal mechanism of remains unclear and requires further research.
pathology may be operative in anterior visual system Although oculohistopathological examination
pathophysiology in a limited but not insignificant reveals dropout of INL neurons in up to 40% of MS
proportion of MS patients. OCT identification of eyes [40], qualitative assessment of OCT scans derived
INL and ONL thinning in MS eyes without a prior from MS patients in vivo reveal microcysts in a small
history of optic neuritis, in which there is relative proportion, which are predominantly localized to the
preservation of the RNFL and GCIP, has been referred INL [48, 49] (Figure 10.4). Again, the etiology and
118 to as the macular thinning predominant (MTP) significance of these microcysts in MS eyes has been
 Chapter 10: Optical coherence tomography and brain magnetic resonance imaging in multiple sclerosis

Figure 10.4 A–C: All images are from the same patient in chronological order over a three-year period. (A) Microcystic macular edema (MME)
of the inner nuclear layer (INL) (red arrows) was present at baseline as well as a foveal cyst of the outer nuclear layer (ONL). (B-C) The foveal ONL
cyst progressively resolved during the follow-up period.
(D-F) All images are from the same patient in chronological order over a two-year period. (D) A single INL cyst (red arrow) was present at
baseline. (E) The cyst spontaneously resolved after one year. (F) Following fingolimod treatment (initiated after scan E), the patient developed
new cystic changes of the INL (red arrow). An epiretinal membrane is noted (white arrow) that had been present on previous scans as well.
(G-I) Three different patients with MME of the INL (red arrows). Vessel artifacts (black arrows) are demonstrated for comparison.
Reproduced with permission from Lancet Neurology

the source of lively scientific debate. Some proponents following optic nerve transaction dating back to
have raised the possibility these microcysts may repre- the 1960s [55]. Moreover, microcysts may simply
sent sites of primary retinal inflammation, on the represent mechanical changes resulting from vitreo-
basis that 25% of patients with macular edema macular traction, although this is not universally
demonstrating fluorescein leakage examined in an agreed upon [53].
ophthalmological study exhibited similar appearing Although the etiology, relevance, and significance
microcysts [50], microcysts in MS eyes may be of INL and ONL changes in MS remain to be estab-
dynamic in nature over time, and they may be a lished, there is an abundance of information highlight-
harbinger of more aggressive MS. Given that micro- ing the possibility of primary retinal inflammation in
cysts may be dynamic in nature, some authors have MS that is difficult to ignore. Despite being an unmye-
suggested that increased INL thickness in the absence linated structure, retinal periphlebitis occurs in up to
of visible microcysts may enable capturing of the same 20% of MS patients [41], highlighting that myelin does
process. As such, increased INL thickness at baseline not need to be present for instigating or propagating
has been shown to predict clinical and radiologic inflammation in MS. The antigenic target of this
disease activity during follow-up [49]. On the other inflammation remains unknown. Interestingly, pre-
hand, some investigators have proposed that micro- vious studies have shown that, during periods of active
cysts in MS eyes are of little significance, since they retinal periphlebitis, there may be concomitant break-
may be seen in other neuroinflammatory disorders down of the blood-brain barrier in MS [56]. Moreover,
(including neuromyelitis optica and chronic relapsing retinal periphlebitis in MS has been shown to be a risk
inflammatory optic neuropathy) as well as non- factor for the development of relapses, contrast enhan-
inflammatory disorders such as neurofibromatosis, cing lesions, as well as an increase in T1 lesion volume
Leber’s hereditary optic neuropathy and glaucoma during follow-up [57, 58]. In addition to retinal per-
[51–54]. Given the array of conditions in which macu- iphlebitis, up to 16% of MS patients experience uveitis,
lar microcysts have been described, some consider particularly intermediate uveitis of the pars planitis
them a final common pathway of retrograde degen- subtype [59]. Activated microglia have also been
eration, especially since there are histopathological demonstrated in MS retinas at postmortem examina-
descriptions of cavitation occurring within the INL tion [40]. Despite ongoing debate, all combined, these 119
 Chapter 10: Optical coherence tomography and brain magnetic resonance imaging in multiple sclerosis

Table 10.1 Relation between OCT-derived measures of p-RNFL thickness and MRI-derived brain volumetrics in MS

Study n OCT measure MRI Correlations

segmentation
Yes No

Time-domain OCT studies

Sepulcre et al. 61 Minimum p-RNFL VBM method WM, GM T2 LV, T1 LV
[57]
Gordon-Lipkin 40 Minimum p-RNFL SIENAX BPF, CSF GM, WM
et al [60]
Grazioli et al. 18/ Average p-RNFL of both SIENAX NBV, WM, GM, T1 LV
[61] 30 eyes T2 LV
Siger et al.a [62] 51 p-RNFL in ON vs. non-ON SPM99 BPF, GM, WM
eyes T2 LV, T1
LV
Dorr et al. [63] 104 p-RNFL of each eyeb SIENAX BPF
b
Pfueller et al. 83 p-RNFL of each eye SIENAX BPF
[65]
Spectral-domain OCT studies
Young et al. 44 p-RNFL of each eyeb & JIM, FAST, FSL BPF, WM GM, BHFr, T2 LV
[68] minimum RNFL
Zimmermanna 63 p-RNFL of ON vs. non-ON SIENAX NBV, WM,
et al. [69] eyesb GM
Saidha et al. 84 p-RNFL of each eye & LESION-TOADS GM, NAWM, thalamus,
[70] p-RNFL of ON vs. non-ON caudate brainstem, T2 LV
eyesb
Abbreviations: OCT: optical coherence tomography; p-RNFL: peripapillary retinal nerve fiber layer; MRI: magnetic resonance
imaging; VBM: voxel based morphometry; WM: white matter; GM: gray matter; LV: lesion volume; BPF: brain parenchymal
fraction; CSF: cerebrospinal fluid; NBV: normalized brain volume; BHFr: black hole fraction; NAWM: normal appearing white
matter
a
Presented relationships are those from eyes without a prior history of ON.
b
Statistical models were used to account for within-subject, inter-eye correlations in analyses from these studies.

observations may be diametrically in opposition to the parenchymal fraction (BPF) in MS [60–63], suggest-
long-recognized pathological polarity involving retro- ing p-RNFL thickness, at least in part, reflects global
bulbar inflammation and its sequelae [32–35]. CNS pathology, associations between average macular
thickness (or total macular volume), a nonspecific
Time-domain OCT and the brain measure of the composite thickness or volume of all
of the intervening layers of the retina in the macular
MRI in MS region, and BPF have been less consistent [60, 63].
There have been a number of studies investigating the Some studies have shown relationships between total
relationships between time-domain OCT derived macular volume and BPF [63], while others have not
measures and brain MRI in MS (Table 10.1). [60]. Given the lack of specificity of total macular
Although several studies have consistently found volume or average macular thickness, however, the
120 that p-RNFL thickness correlates with brain potential utility of this measure in MS is questionable,
 Chapter 10: Optical coherence tomography and brain magnetic resonance imaging in multiple sclerosis

and thus some investigators did not examine addition to these differences in the utilization of
the relationships between total macular volume p-RNFL thickness, there were also differences in the
and brain MRI, restricting their analyses to measures statistical approaches and models adopted in each
of p-RNFL thickness only [57]. More important is study. Finally, there were differences in the brain seg-
the lack of consistency observed in these studies mentation techniques employed between the studies to
between p-RNFL thickness measurements and brain- determine brain substructure volumes. Some studies
substructure volumes [57, 60–62]. Some investigators utilized SIENAX, while others used different techni-
found relationships between p-RNFL thickness and ques. Moreover, not all studies investigated lesion mor-
BPF, but not GM or WM volumes [60], while in phometry volumes, or indeed even brain substructure
some studies p-RNFL thickness and both GM and volumes. For instance, in some studies only relation-
WM volumes were related [57], while in others ships between OCT measures and BPF were assessed.
p-RNFL thickness was associated with either GM The lack of consistency for relationships of p-
volume only [62] or WM volume only [61]. RNFL thickness and brain substructure volumes
Relationships between p-RNFL thickness and T1 lesion observed in the above studies does not mitigate the
volume, as well as T2 lesion volume, were similarly consistent and encouraging observation that p-RNFL
inconsistent. For example, in one study p-RNFL thick- thickness and BPF appear to be related in MS, sug-
ness correlated with neither T1 lesion volume nor T2 gesting p-RNFL thickness, at least to some degree may
lesion volume [57], while in another it correlated with reflect global brain atrophy and global CNS processes
both T1 lesion volume and T2 lesion volume [62]. in MS. Investigations of relationships between time-
More specifically, on the basis of these studies, it is domain OCT derived measures, principally reflecting
unclear whether p-RNFL thickness, at least as esti- the anterior visual pathway, and measures of the pos-
mated with time-domain OCT technology, correlates terior visual pathways derived from nonconventional
with measures of the normal-appearing cerebral white MRI techniques have also yielded interesting results
matter (NAWM), white matter (WM) lesional mor- in MS. In one such study, DTI indices along the optic
phometry, or cortical gray matter (GM) volumes. radiation were correlated with p-RNFL thickness
There may be several potential explanations underlying measures in MS, controlling for age, sex, and corre-
the inconsistency of the results observed in these var- sponding MRI indices along a portion of the corti-
ious studies, especially with respect to the relationships cospinal tract (to attempt to at least partially account
of p-RNFL thickness measures with brain substructure for global CNS changes driving relationships) [64].
volumes, which makes the comparability of the results Moderate correlations of average RNFL thickness
derived from these different studies limited. with FA and λ⊥ in the OR were observed, suggesting
The patient populations included in these studies an association between p-RNFL thickness and OR
varied quite a bit with respect to sample size, MS damage that may not be primarily driven by abnorm-
subtype, as well as the degrees of disability and disease alities outside of the visual system. Interestingly,
duration of the patients included. Treatment effects p-RNFL thickness in the nasal retinal quadrant
were also minimally accounted for across these various was specifically correlated with FA and λ⊥ in the
studies. There are several major methodological differ- synaptically connected contralateral OR. Since this
ences between the studies. Some studies utilized mini- was a cross-sectional study, the results cannot be
mum p-RNFL thickness (lowest p-RNFL thickness of over-interpreted as substantiated evidence of transy-
either eye) in analyses [57, 60], some studies utilized naptic degeneration. Even though these results point
average p-RNFL thickness of both eyes in analyses [61], toward a relationship between changes in the anterior
some studies utilized the p-RNFL thickness of both visual pathway and posterior visual pathway in MS,
eyes but accounted for within-subject inter-eye corre- the study does not address or help determine how the
lations [63], while in some other studies comparisons changes may be connected, that is, retrograde
were of p-RNFL thickness in eyes with and without a vs. anterograde degeneration, or coincidental conco-
history of AON [62]. Interestingly, in the latter study, mitant pathology simultaneously afflicting
p-RNFL thickness in MS eyes without a history of AON both the anterior and posterior visual pathways.
correlated with BPF and GM fraction, while in MS eyes Another very interesting cross-sectional study utiliz-
with a history of AON p-RNFL thickness correlated ing time-domain OCT studied the relationships
with neither BPF nor brain substructure volumes. In between p-RNFL thickness, global brain atrophy 121
 Chapter 10: Optical coherence tomography and brain magnetic resonance imaging in multiple sclerosis

(as estimated by BPF), and MRS-derived NAA con- (Table 10.1). In a recent study including participants
centrations in the visual cortex, as well as the with clinically isolated syndrome (CIS) or early relap-
NAWM in MS [65]. In this study the investigators sing remitting MS (RRMS), not on active treatment,
found that p-RNFL thickness measures correlated in which brain volumetrics were assessed with a 1.5 T
with NAA concentrations in the visual cortex, but MRI, p-RNFL thickness and total macular volume
not the NAWM, and were consistent with prior independently predicted WM fraction, but not GM
observations that RNFL thickness correlated with fraction or black hole fraction, correcting for age and
BPF in MS. The multivariate statistical model history of AON [68]. However, in another recent
employed revealed that the correlation between study of just RRMS patients, in which brain volu-
RNFL thickness and visual cortex NAA concentra- metrics were also assessed with a 1.5 T MRI,
tion was unlikely to simply represent the conse- p-RNFL thickness and TMV (albeit less so) were
quence of global brain atrophy, as one might both significantly associated with normalized WM
assume on the basis of the relationship between volume and normalized GM volume [69]. However,
p-RNFL thickness and BPF, since BPF and visual in this study, the authors found that the significant
cortex NAA concentrations seemed to be both inde- association of OCT measures with normalized GM
pendently associated with p-RNFL thickness. In fact, volume was abolished in eyes with a history of AON
visual cortex NAA concentrations and BPF did not (while the relationship with normalized WM volume
seem to be correlated with one another in this study. was essentially unaffected). On the basis of this find-
The investigators’ findings thus highlight a potential ing, which was similar to that observed by Siger et al.
interconnection between changes in the anterior and [62] in their study of time-domain OCT–brain MRI
posterior visual pathways, beyond that which may be relationships, the authors proposed the possibility
expected due to simple global neurodegeneration. that excessive tissue damage resulting from AON
Consistent with this observation, the study also may mask the underlying relationship between OCT
revealed lower NAA visual cortex concentrations in measures and GM volume driven by global neurode-
subjects with a prior history of AON. Again, how- generation in MS, as observed in eyes without a his-
ever, this was a cross-sectional study and, therefore, tory of AON. In this study, the authors also included
the results should not be over-interpreted to sub- limited OCT segmentation data. Specifically, they
stantiate evidence of transynaptic degeneration. assessed macular GCL thickness measures, as esti-
mated by the composite thickness of the GCL and
inner plexiform layer (GCIP). As might be expected,
Spectral-domain OCT, segmentation, given the biological similarity of the GCL and RNFL
and brain MRI in MS (they could be regarded as extensions of one another),
The first analyses attempting to correlate changes in GCIP thickness measures exhibited relationships with
retinal architecture and global CNS pathology in MS, brain volumetrics similar to those observed with p-
as described previously, utilized time-domain OCT, RNFL thickness measures. On the basis of previous
which has lower reproducibility and resolution than work in which GCIP thickness measures appeared to
the high-speed, high-definition spectral-domain have superior structure–function relationships to p-
devices, which enable the rapid and near-automated RNFL thickness measures (33), possibly related to
acquisition of high-resolution images, in conjunction superior reproducibility among other potential rea-
with a very low test–retest variability. [66, 67] sons, one might have expected that GCIP thickness
Notwithstanding these impressive technical advance- measures would reflect global CNS processes more
ments, there has been a paucity of studies involving strongly than p-RNFL thickness measures. However,
the utilization of spectral-domain OCT, for the pur- this was not observed in this study.
pose of studying the relationship between retinal and A more detailed assessment of the relationships
brain metrics in MS. While these latest studies show a between spectral domain OCT segmentation derived
similar relationship between spectral-domain derived retinal measures and 3 T MRI-derived brain substruc-
p-RNFL thickness measurements and time-domain ture volumes has also been performed in MS [70].
derived p-RNFL thickness measurements, with BPF, Since brain substructure volumes correlate with intra-
results of relationships with brain substructure cranial volume (a surrogate of head size), these
122 volumes are also unfortunately conflicting measures are conventionally normalized or adjusted
 Chapter 10: Optical coherence tomography and brain magnetic resonance imaging in multiple sclerosis

for intracranial volume in order to account for these study. Overall, this represents a preliminary study,
normal relationships. In undertaking this study, the since detailed investigations of the relationships
authors also investigated whether OCT (including between OCT segmentation derived measures and
OCT segmentation) derived retinal measures might brain MRI measures are lacking. The findings of this
also correlate with head size (using intracranial study need to be validated and the basis for the
volume as a surrogate) in order to potentially opti- observed relationships need to be explored and
mize the accuracy of their results. Indeed, the investi- determined.
gators of this study found that OCT measures in One of the most recent studies utilizing spectral-
healthy controls and MS patients appear to correlate domain OCT in MS investigated the association
with intracranial volume (or head size) and, therefore, between damage to the anterior visual pathway
not only adjusted brain substructure volumes, but (as assessed by OCT-derived measures of p-RNFL
also OCT measures for intracranial volume, as part thickness and total macular volume) and posterior
of their regression models when assessing retinal- visual pathway (as assessed by 3 T MRI-derived
brain relationships. In this study, p-RNFL and GCIP volumetry and spectroscopy), both at baseline and
thickness measures exhibited relationships similar to after one year of follow-up [71]. Independent of
one another with brain substructure volumes, with AON history (as well as other confounders), baseline
perhaps the relationships being interestingly slightly analyses revealed that visual cortex volume, visual
stronger with p-RNFL than GCIP measures. Both p- cortex NAA, and OR lesion volumes were associated
RNFL and GCIP thickness measures were associated with p-RNFL thickness. Repeating the same analyses
with cortical GM volume (as well as caudate volume), after one year of follow-up similarly revealed that
but only in the eyes of MS patients without a history of visual cortex volume and OR lesion volumes were
AON, again raising the possibility that excessive local associated with p-RNFL thickness. These findings
tissue injury following AON may mask underlying raise the possibility of retrograde transynaptic
relationships that would otherwise be seen. Neither degeneration similar to prior studies. Conversely,
p-RNFL nor GCIP thickness measures were asso- baseline analyses revealed that patients with a prior
ciated with either NAWM volume or T2 lesion history of severe AON had significantly lower visual
volume. However, an interesting finding to emerge cortex volumes than MS patients without a prior
from this study was the unexpected and apparently history of AON. Although the finding was not sig-
strong relationship between INL thickness and T2 nificant, MS patients with a prior history of mild
lesion volume, with a corresponding inverse relation- AON also exhibited lower visual cortex volumes
ship with NAWM volume in RRMS. In other words, than MS patients without previous AON, and again
higher INL thickness in this MS cohort appeared to be although the finding was not significant, MS patients
associated with higher T2 lesion volume and lower with a prior history of severe AON seemed to have
NAWM volume (Figure 10.5). The basis for this find- reduced visual cortex volume relative to MS patients
ing could not be determined as part of this cross- who had previously experienced mild AON. Similar
sectional study, although the possibility that increased to observations raised by other investigators, these
INL thickness may represent part of the spectrum of findings raise the possibility of anterograde transy-
macular microcysts (when the microcysts are not naptic degeneration in MS.
visible to the naked eye, for example) or even poten-
tially retinal periphlebitis have been raised. Another Summary
interesting finding to emerge from this study was that Although results of some investigations assessing the
INL and ONL changes in MS may not reflect similar relationships between OCT derived measures and
global CNS processes, unlike the RNFL and GCL. For brain MRI in MS are conflicting, the consistently
instance, ONL thickness in MS eyes only with a his- observed relationship between p-RNFL thickness
tory of AON was associated with cerebellar WM and BPF is encouraging, and it provides evidence
volume and tended to towards being associated with that, at least to a degree, p-RNFL thickness measures
cortical GM, caudate, thalamus, and T2 lesion reflect global CNS processes in MS. That this cheap,
volume, raising the possibility that it may be more a easily obtainable, and repeatable, reproducible, and
marker of global neurodegeneration in MS. However, noninvasively acquired retinal measure may provide
the basis for these findings was not explored in this insight into global changes within the brains of MS 123
 Chapter 10: Optical coherence tomography and brain magnetic resonance imaging in multiple sclerosis

A 20 B 10

10
GCIP thickness**

GCIP thickness**
5
0
0
–10
–5
–20

–30 –10
–400000 –200000 0 200000 400000 –200000 –100000 0 100000 200000
ICV** ICV**

C 20 D
40

RNFL thickness**
10
INL thickness**

20

0 0

–10 –20

–20 –40
–100000 –50000 0 50000 100000 –40000 –20000 0 20000 40000
NAWM volume** Cortical GM volume**

Figure 10.5 **Residual values from multivariate regression models

Panel A represents an adjusted variables plot of ganglion cell layer + inner plexiform layer (GCIP) thickness and intracranial volume (ICV) in
multiple sclerosis (MS), adjusted for age, sex, and disease duration. The solid red line illustrates the independent relationship between GCIP
thickness and ICV in MS. Note that, as ICV increases, GCIP thickness similarly increases, consistent with the detection of significant associations
between GCIP thickness and ICV in MS (p = 0.008). Panel B represents an adjusted variables plot of ganglion cell layer + inner plexiform layer
(GCIP) thickness and intracranial volume (ICV) in healthy controls (HCs), adjusted for age and sex. The solid red line graphically illustrates the
independent relationship between GCIP thickness and ICV in HCs. As ICV increases, GCIP thickness similarly increases. Panel C represents an
adjusted variables plot of inner nuclear layer (INL) thickness and normal-appearing white matter (NAWM) volume in MS, adjusted for age, sex,
disease duration, and ICV. The solid red line shows the independent relationship between INL thickness and NAWM volume in MS. Note that, as
INL thickness increases, NAWM volume decreases. Panel D depicts an adjusted variables plot of peripapillary retinal nerve fiber layer (p-RNFL)
thickness and cortical gray matter (GM) volume in RRMS, adjusted for age, sex, disease duration, and ICV. The solid red line graphically illustrates
the independent relationship between p-RNFL thickness and cortical-GM volume in RRMS.
Reproduced with permission from JAMA Neurology

patients is compelling and represents an attractive brain MRI in MS, since this remains an understudied
advantage of OCT relative to other emerging imaging area. Specifically, there is a need for longitudinal
techniques, including nonconventional MRI techni- assessments of OCT and MRI in MS, which are cur-
ques, for assessing neurodegeneration in MS. rently lacking. Moreover, future studies need to be
Moreover, the possibility that one may glean informa- more homogenous in terms of their patient inclusion,
tion regarding different processes operative within OCT and MRI devices utilized, OCT and MRI seg-
different compartments of the brains of MS patients mentation techniques employed, and their approach
through the incorporation of OCT segmentation to analyses, in order to make the results of studies
measures not only may be helpful in shedding light more comparable. Although much work is needed in
on the pathobiology of MS, but could turn out to be this front, preliminary results are exciting, and if
the greatest advantage of its application in MS. indeed they turn out to be correct, through the use
Nonetheless, there is need for further study of the of OCT, physicians may be able to gain deeper insight
relationships between OCT-derived measures, in par- into ongoing processes within the brains of their MS
124 ticular OCT-segmentation-derived measures and patients at the bedside.
 Chapter 10: Optical coherence tomography and brain magnetic resonance imaging in multiple sclerosis

Disclosures: Dr. Saidha has been the recipient of a 14. Madigan MC, Rao NS, Tenhula WN, Sadun AA.
Race to Erase MS grant. Preliminary morphometric study of tumor necrosis
factor-alpha (TNF alpha)-induced rabbit optic neuro-
References pathy. Neurol Res June 1996;18(3):233–6.
1. Anderson DW, Ellenberg JH, Leventhal CM, Reingold 15. Shindler KS, Ventura E, Dutt M, Rostami A.
SC, Rodriguez M, Silberberg DH. Revised estimate of Inflammatory demyelination induces axonal injury and
the prevalence of multiple sclerosis in the united states. retinal ganglion cell apoptosis in experimental optic
Ann Neurol 1992 Mar;31(3):333–6. neuritis. Exp Eye Res September 2008;87(3):208–13.
2. Prineas J. Pathology of multiple sclerosis. in Cook S, 16. Frischer JM, Bramow S, Dal-Bianco A, Lucchinetti CF,
ed. Handbook of Multiple Sclerosis In: New York: Rauschka H, Schmidbauer M, et al. The relation
Marcel Dekker; 2001. p. 289–324. between inflammation and neurodegeneration in
multiple sclerosis brains. Brain May 2009;132(Pt
3. Frohman EM, Racke MK, Raine CS. Multiple sclerosis 5):1175–89.
– the plaque and its pathogenesis. N Engl J Med 2006
Mar 2;354(9):942–55. 17. Kappos L, Moeri D, Radue EW, Schoetzau A,
Schweikert K, Barkhof F, et al. Predictive value of
4. Marburg O. Die sogennate akute multiple sklerose. gadolinium-enhanced magnetic resonance imaging for
Jahrb Psychiatrie 1906;27:211–312. relapse rate and changes in disability or impairment in
5. Putnam T. Studies in multiple sclerosis. Arch Neurol multiple sclerosis: A meta-analysis. gadolinium MRI
Psych 1936;35:1289–308. meta-analysis group. Lancet March 20, 1999;353
6. van Waesberghe JH, Kamphorst W, De Groot CJ, van (9157):964–9.
Walderveen MA, Castelijns JA, Ravid R, et al. Axonal 18. Fisniku LK, Brex PA, Altmann DR, Miszkiel KA,
loss in multiple sclerosis lesions: Magnetic resonance Benton CE, Lanyon R, et al. Disability and T2 MRI
imaging insights into substrates of disability. lesions: A 20-year follow-up of patients with relapse
Ann Neurol November 1999;46(5):747–54. onset of multiple sclerosis. Brain March 2008;131(Pt
7. De Stefano N, Narayanan S, Francis GS, Arnaoutelis R, 3):808–17.
Tartaglia MC, Antel JP, et al. Evidence of axonal 19. De Stefano N, Matthews PM, Fu L, Narayanan S,
damage in the early stages of multiple sclerosis and its Stanley J, Francis GS, et al. Axonal damage correlates
relevance to disability. Arch Neurol January 2001;58 with disability in patients with relapsing-remitting
(1):65–70. multiple sclerosis. results of a longitudinal magnetic
8. Compston A, Coles A. Multiple sclerosis. Lancet April resonance spectroscopy study. Brain August 1998;121
6, 2002;359(9313):1221–31. ( Pt 8)(Pt 8):1469–77.
9. Miller DH. Biomarkers and surrogate outcomes in 20. Fisher E, Rudick RA, Simon JH, Cutter G, Baier M, Lee
neurodegenerative disease: Lessons from multiple JC, et al. Eight-year follow-up study of brain atrophy in
sclerosis. NeuroRx April 2004;1(2):284–94. patients with MS. Neurology November 12, 2002;59
(9):1412–20.
10. Minneboo A, Uitdehaag BM, Jongen P, Vrenken H,
Knol D, van Walderveen MA, et al. Association 21. Calabrese M, De Stefano N, Atzori M, Bernardi V,
between MRI parameters and the MS severity scale: A Mattisi I, Barachino L, et al. Detection of cortical
12 year follow-up study. Multiple Sclerosis May 2009; inflammatory lesions by double inversion recovery
15(5):632–7. magnetic resonance imaging in patients with multiple
sclerosis. Arch Neurol October 2007;64(10):1416–22.
11. Calabrese M, Atzori M, Bernardi V, Morra A,
Romualdi C, Rinaldi L, et al. Cortical atrophy is rele- 22. Inglese M, Oesingmann N, Casaccia P, Fleysher L.
vant in multiple sclerosis at clinical onset. J Neurol Progressive multiple sclerosis and gray matter pathol-
September 2007;254(9):1212–20. ogy: An MRI perspective. Mt Sinai J Med March–April
2011;78(2):258–67.
12. Calabrese M, Agosta F, Rinaldi F, Mattisi I, Grossi P,
Favaretto A, et al. Cortical lesions and atrophy asso- 23. Simon JH. Brain atrophy in multiple sclerosis: What
ciated with cognitive impairment in relapsing-remit- we know and would like to know. Multiple Sclerosis
ting multiple sclerosis. Arch Neurol September 2009; December 2006;12(6):679–87.
66(9):1144–50. 24. Lucchinetti CF, Popescu BF, Bunyan RF, Moll NM,
13. Rawes JA, Calabrese VP, Khan OA, DeVries GH. Roemer SF, Lassmann H, et al. Inflammatory cortical
Antibodies to the axolemma-enriched fraction in demyelination in early multiple sclerosis. N Engl J Med
the cerebrospinal fluid and serum of patients with December 8, 2011;365(23):2188–97.
multiple sclerosis and other neurological 25. Brownell B, Hughes JT. The distribution of plaques in
diseases. Multiple Sclerosis December 1997;3(6): the cerebrum in multiple sclerosis. J Neurol Neurosurg
363–9. Psychiatry November 1962;25:315–20. 125
 Chapter 10: Optical coherence tomography and brain magnetic resonance imaging in multiple sclerosis

26. Lumsden C. The neuropathology of multiple sclerosis. 39. Ikuta F, Zimmerman HM. Distribution of plaques in
In: Vinken PJ, bruyn GW, eds. Handbook of Clinical seventy autopsy cases of multiple sclerosis in the united
Neurology. Amsterdam: Elsevier; 1970. pp. 217–309. states. Neurology June 1976; 26(6 PT 2): 26–8.
27. Bo L, Vedeler CA, Nyland HI, Trapp BD, Mork SJ. 40. Green AJ, McQuaid S, Hauser SL, Allen IV, Lyness R.
Subpial demyelination in the cerebral cortex of multi- Ocular pathology in multiple sclerosis: Retinal atrophy
ple sclerosis patients. J Neuropathol Exp Neurol July and inflammation irrespective of disease duration.
2003;62(7):723–32. Brain April 21, 2010;133(6):1591–601.
28. Moll NM, Rietsch AM, Ransohoff AJ, Cossoy MB, 41. Kerrison JB, Flynn T, Green WR. Retinal pathologic
Huang D, Eichler FS, et al. Cortical demyelination in changes in multiple sclerosis. Retina 1994;14(5):
PML and MS: Similarities and differences. Neurology 445–51.
January 29, 2008;70(5):336–43. 42. Naismith RT, Xu J, Tutlam NT, Trinkaus K, Cross AH,
29. Lovato L, Willis SN, Rodig SJ, Caron T, Almendinger Song SK. Radial diffusivity in remote optic neuritis
SE, Howell OW, et al. Related B cell clones populate the discriminates visual outcomes. Neurology May 25,
meninges and parenchyma of patients with multiple 2010;74(21):1702–10.
sclerosis. Brain. February 2011;134(Pt 2):534–41. 43. Talman LS, Bisker ER, Sackel DJ, Long DA,Jr, Galetta
30. Oh J, Zackowski K, Chen M, Newsome S, Saidha S, KM, Ratchford JN, et al. Longitudinal study of vision
Smith SA, et al. Multiparametric MRI correlates of and retinal nerve fiber layer thickness in multiple
sensorimotor function in the spinal cord in multiple sclerosis. Ann Neurol June 2010;67(6):749–60.
sclerosis. Multiple Sclerosis. April 2013;19(4):427–35. 44. Papakostopoulos D, Fotiou F, Hart JC, Banerji NK.
31. Filippi M, Rocca MA, De Stefano N, Enzinger C, Fisher The electroretinogram in multiple sclerosis and
E, Horsfield MA, et al. Magnetic resonance techniques demyelinating optic neuritis. Electroencephalogr Clin
in multiple sclerosis: The present and the future. Neurophysiol January–February 1989;74(1):1–10.
Arch Neurol December 2011;68(12):1514–20. 45. Forooghian F, Sproule M, Westall C, Gordon L,
32. Saidha S, Syc SB, Ibrahim MA, Eckstein C, Warner CV, Jirawuthiworavong G, Shimazaki K, et al.
Farrell SK, et al. Primary retinal pathology in multiple Electroretinographic abnormalities in multiple sclero-
sclerosis as detected by optical coherence tomography. sis: Possible role for retinal autoantibodies.
Brain February 2011;134(Pt 2):518–33. Doc Ophthalmol September 2006;113(2):123–32.
33. Saidha S, Syc SB, Durbin MK, Eckstein C, Oakley JD, 46. Gundogan FC, Demirkaya S, Sobaci G. Is optical
Meyer SA, et al. Visual dysfunction in multiple coherence tomography really a new biomarker candi-
sclerosis correlates better with optical coherence date in multiple sclerosis? – A structural and functional
tomography derived estimates of macular ganglion cell evaluation. Invest Ophthalmol Vis Sci December
layer thickness than peripapillary retinal nerve fiber 2007;48(12):5773–81.
layer thickness. Multiple Sclerosis. August 24, 2011; 47. Gills JP,Jr. Electroretinographic abnormalities and
34. Syc SB, Saidha S, Newsome SD, Ratchford JN, Levy M, advanced multiple sclerosis. Invest Ophthalmol
Ford E, et al. Optical coherence tomography segmen- December 1966;5(6):555–9.
tation reveals ganglion cell layer pathology after optic 48. Gelfand JM, Nolan R, Schwartz DM, Graves J, Green
neuritis. Brain October 17, 2011; AJ. Microcystic macular edema in multiple sclerosis is
35. Seigo MA, Sotirchos ES, Newsome S, Babiarz A, associated with disease severity. Brain June 2012;135
Eckstein C, Ford E, et al. In vivo assessment of retinal (Pt 6):1786–93.
neuronal layers in multiple sclerosis with manual and 49. Saidha S, Sotirchos ES, Ibrahim MA, Crainiceanu CM,
automated optical coherence tomography segmenta- Gelfand JM, Sepah YJ, et al. Microcystic macular
tion techniques. J Neurol March 15, 2012; edema, thickness of the inner nuclear layer of the
36. Frohman EM, Fujimoto JG, Frohman TC, Calabresi PA, retina, and disease characteristics in multiple sclerosis:
Cutter G, Balcer LJ. Optical coherence tomography: A A retrospective study. Lancet Neurol November
window into the mechanisms of multiple sclerosis. 2012;11(11):963–72.
Nat Clin Pract Neurol December 2008;4(12):664–75. 50. Brar M, Yuson R, Kozak I, Mojana F, Cheng L, Bartsch
37. Saidha S, Eckstein C, Ratchford JN. Optical coherence DU, et al. Correlation between morphologic features
tomography as a marker of axonal damage in multiple on spectral-domain optical coherence tomography and
sclerosis. CML – Multiple Sclerosis 2010;2(2):33–43. angiographic leakage patterns in macular edema.
38. Toussaint D, Perier O, Verstappen A, Bervoets S. Retina March 2010;30(3):383–9.
Clinicopathological study of the visual pathways, eyes, 51. Sotirchos ES, Saidha S, Byraiah G, Mealy MA, Ibrahim
and cerebral hemispheres in 32 cases of disseminated MA, Sepah YJ, et al. In vivo identification of morpho-
sclerosis. J Clin Neuroophthalmol September 1983;3 logic retinal abnormalities in neuromyelitis optica.
126 (3):211–20. Neurology April 9, 2013;80(15):1406–14.
 Chapter 10: Optical coherence tomography and brain magnetic resonance imaging in multiple sclerosis

52. Kaufhold F, Zimmermann H, Schneider E, Ruprecht layer as a potential measure of axonal loss and brain
K, Paul F, Oberwahrenbrock T, et al. Optic neuritis is atrophy. J Neurol October 2008;255(10):1555–60.
associated with inner nuclear layer thickening and 63. Dorr J, Wernecke KD, Bock M, Gaede G, Wuerfel JT,
microcystic macular edema independently of multiple Pfueller CF, et al. Association of retinal and macular
sclerosis. PLoS One August 6, 2013;8(8):e71145. damage with brain atrophy in multiple sclerosis.
53. Barboni P, Carelli V, Savini G, Carbonelli M, La PLoS One April 8, 2011;6(4):e18132.
Morgia C, Sadun AA. Microcystic macular degen- 64. Reich DS, Smith SA, Gordon-Lipkin EM, Ozturk A,
eration from optic neuropathy: Not inflammatory, Caffo BS, Balcer LJ, et al. Damage to the optic radiation
not trans-synaptic degeneration. Brain July 2013; in multiple sclerosis is associated with retinal injury
136(Pt 7):e239. and visual disability. Arch Neurol August 2009;66
54. Wolff B, Basdekidou C, Vasseur V, Mauget-Faysse M, (8):998–1006.
Sahel JA, Vignal C. Retinal inner nuclear layer micro- 65. Pfueller CF, Brandt AU, Schubert F, Bock M, Walaszek
cystic changes in optic nerve atrophy: A novel spectral- B, Waiczies H, et al. Metabolic changes in the visual
domain OCT finding. Retina November–December cortex are linked to retinal nerve fiber layer thinning in
2013;33(10):2133–8. multiple sclerosis. PLoS One April 6, 2011;6(4):e18019.
55. Gills JP,Jr, Wadsworth JA. Degeneration of the inner 66. Syc SB, Warner CV, Hiremath GS, Farrell SK,
nuclear layer of the retina following lesions of the optic Ratchford JN, Conger A, et al. Reproducibility of high-
nerve. Trans Am Ophthalmol Soc 1966;64:66–88. resolution optical coherence tomography in multiple
56. Engell T, Hvidberg A, Uhrenholdt A. Multiple sclero- sclerosis. Multiple Sclerosis. July 2010;16(7):829–39.
sis: Periphlebitis retinalis et cerebro-spinalis. A corre- 67. Warner CV, Syc SB, Stankiewicz AM, Hiremath G,
lation between periphlebitis retinalis and abnormal Farrell SK, Crainiceanu CM, et al. The impact of uti-
technetium brain scintigraphy. Acta Neurol Scand May lizing different optical coherence tomography devices
1984;69(5):293–7. for clinical purposes and in multiple sclerosis trials.
57. Sepulcre J, Murie-Fernandez M, Salinas-Alaman A, PLoS One 2011;6(8):e22947.
Garcia-Layana A, Bejarano B, Villoslada P. Diagnostic 68. Young KL, Brandt AU, Petzold A, Reitz LY, Lintze F,
accuracy of retinal abnormalities in predicting disease Paul F, et al. Loss of retinal nerve fibre layer axons
activity in MS. Neurology May 2007;68(18):1488–94. indicates white but not grey matter damage in early
58. Ortiz-Perez S, Martinez-Lapiscina EH, Gabilondo I, multiple sclerosis. Eur J Neurol. May 2013;20(5):
Fraga-Pumar E, Martinez-Heras E, Saiz A, et al. Retinal 803–11.
periphlebitis is associated with multiple sclerosis 69. Zimmermann H, Freing A, Kaufhold F, Gaede G, Bohn
severity. Neurology. September 3, 2013;81(10):877–81. E, Bock M, et al. Optic neuritis interferes with optical
59. Donaldson MJ, Pulido JS, Herman DC, Diehl N, coherence tomography and magnetic resonance ima-
Hodge D. Pars planitis: A 20-year study of incidence, ging correlations. Multiple Sclerosis April 2013;19
clinical features, and outcomes. Am J Ophthalmol (4):443–50.
December 2007;144(6):812–7. 70. Saidha S, Sotirchos ES, Oh J, Syc SB, Seigo MA, Shiee
60. Gordon-Lipkin E, Chodkowski B, Reich DS, Smith SA, N, et al. Relationships between retinal axonal and
Pulicken M, Balcer LJ, et al. Retinal nerve fiber layer is neuronal measures and global central nervous system
associated with brain atrophy in multiple sclerosis. pathology in multiple sclerosis. JAMA Neurol January
Neurology October 16, 2007;69(16):1603–9. 2013;70(1):34–43.
61. Grazioli E, Zivadinov R, Weinstock-Guttman B, 71. Gabilondo I, Martinez-Lapiscina EH, Martinez-Heras
Lincoff N, Baier M, Wong JR, et al. Retinal nerve fiber E, Fraga-Pumar E, Llufriu S, Ortiz S, et al. Trans-
layer thickness is associated with brain MRI outcomes synaptic axonal degeneration in the visual pathway in
in multiple sclerosis. J Neurol Sci May 15, 2008;268 multiple sclerosis. Ann Neurol October 2, 2013;
(1–2):12–7. 72. Shiee N, Bazin PL, Zackowski KM, Farrell SK,
62. Siger M, Dziegielewski K, Jasek L, Bieniek M, Nicpan Harrison DM, Newsome SD, et al. Revisiting brain
A, Nawrocki J, et al. Optical coherence tomography in atrophy and its relationship to disability in multiple
multiple sclerosis: Thickness of the retinal nerve fiber sclerosis. PLoS One 2012;7(5):e37049

127
 Chapter
Optical coherence tomography in

11 neurodegenerative and other neurologic

diseases
Friedemann Paul and Alexander Ulrich Brandt

Introduction substantial overlaps in clinical presentation and paracli-

nical findings. In NMO, the disease course is usually
In recent years, OCT has been used in studies across
more aggressive and irreversible neurological disability
a wide range of neurodegenerative diseases, from
accrues more rapidly than in MS [3]. Considering that it
highly prevalent conditions such as Alzheimer’s dis-
is less common than MS, yet may have worse visual
ease to rare diseases like spinocerebellar ataxias.
consequences, the number of OCT studies on NMO is
OCT assessment was initially introduced in
impressive, and several groups have published similar
response to clinical observation of symptoms and
and consistent data on the disease’s retinal pathology [4].
paraclinical findings (e.g., visual evoked potentials,
Previous research has focused on quantifying retinal
brain MRI) suggesting that the visual system was
damage in comparison to MS – the most relevant clinical
involved in many of these conditions, and was also
differential diagnosis – and healthy subjects, on the
first studied to examine validity for clinical visual
one hand, and the suitability of OCT as a differential
outcome measures such as low-contrast letter acuity
diagnosis tool for distinguishing between NMO and
for clinical trials and research in multiple sclerosis
MS, on the other.
(MS; see Chapter 6). Since then, its easy applicability
Based on long-standing clinical experience, NMO-
and noninvasiveness have also contributed to its
induced ON is understood to cause more severe visual
increasing popularity. However, our understanding
impairment than a “classical” MSON attack and tends
of retinal pathology measured by OCT in most neu-
to show poorer recovery, even after treatment with
rodegenerative diseases is still lacking compared to
high-dose corticosteroids or plasma exchange. In an
the plethora of literature published on MS and optic
NMO cohort characterized at the Mayo Clinic, 60% of
neuritis (ON) over the past 15 years. Many studies
relapsing patients were functionally blind (acuity
are limited by small cohort sizes, lack of reproduci-
20/200 or worse) in at least one eye after a mean
bility by independent research groups, and the unas-
disease duration of 7.7 years [5].
sessed influence of comorbidities such as glaucoma,
In line with this, some studies have reported more
diabetic retinopathy, and macular disease, which all
pronounced thinning of the retinal nerve fiber layer
may be present in elderly patients with neurodegen-
(RNFL) in NMO-ON eyes compared to MSON eyes
erative diseases.
[6]. Naismith and colleagues found RNFL thicknesses
in ON eyes from NMO patients to be on average 10
Neuromyelitis optica µm or more thinner (depending on the statistical
Neuromyelitis optica (NMO) is a rare autoimmune model) than those from MS patients, and 39% of
inflammatory CNS disease that predominantly pre- NMO eyes versus 10% of MS eyes had RNFL values
sents with clinical attacks of optic neuritis and myelitis < = 39 µm, indicating that NMO leads to more pro-
[1, 2]. Following the detection of a highly specific nounced axonal damage in the optic nerve. Similarly,
biomarker, a serum antibody to the most abundant Ratchford and colleagues reported substantial RNFL
CNS water channel aquaporin-4 (AQP4), in 60–80% thinning in NMO-ON eyes versus MSON eyes and
of patients with NMO, the latter was classified as healthy controls (63.6 µm, 88.3 µm, and 102.4 µm,
an autoimmune condition distinct from MS but with respectively) [7]. In this study, a first episode of ON

128 Optical Coherence Tomography in Neurological Disease, ed. Peter A. Calabresi, Laura J. Balcer, and Elliot M. Frohman.
Published by Cambridge University Press. © Cambridge University Press 2015.
 Chapter 11: Optical coherence tomography in neurodegenerative and other neurologic diseases

was assumed to lead to 24 µm more RNFL loss than microcystic pattern with lower reflectance
in MS, and visual impairment was more severe. (Figure 11.2) and is mainly located in the inner
Similar results were found for macular volume. A nuclear layer (INL) in small, discrete patches [16].
French study reported a strong inverse correlation Both the pathogenesis and association with various
between visual field defects and RNFL thickness and nosologic entities have yet to be clarified. One pos-
between the Expanded Disability Status Scale (EDSS) sibility is that MME is a symptom of chronic
as a global measure of neurological disability and inflammation in the INL, because this layer was
RNFL thickness [8]. More recent studies have shown to be a prominent site of inflammation and
applied modern spectral-domain (SD) technology microglia activation in a recent autopsy study [18].
that offers higher resolution at reduced scan times A further argument for an inflammatory pathogen-
[9] and allows segmentation of the different intra- esis is provided by Saidha et al., who showed that
retinal layers [10] (see Chapter 13). Both Syc et al. the INL thickening, which usually accompanies
[11] and Sotirchos et al. [12] reported thinning of the MME, may be indicative of a more severe disease
ganglion cell layer (GCL) and inner plexiform layer progression in MS, as indicated by a higher number
(IPL) in NMO-ON eyes versus MS-ON eyes or of inflammatory lesions on brain MRI, accelerated
healthy controls, which confirmed earlier data sug- EDSS progression, and a higher frequency of
gesting a more severe retinal involvement in NMO. relapses [19]. A study by our group further revealed
Similar results are reported by Monteiro et al. and that MME is not specific to MS or NMO, but is
Fernandes et al. [13,14]. Schneider et al. detected a strongly linked to a history of ON: 95% of eyes with
more pronounced thinning of the RNFL and the MME (in total 22 eyes) from a cohort of 255
GCL in NMO-ON eyes as opposed to MS and, inter- patients with MS or clinically isolated syndrome
estingly, a more pronounced association of struc- (CIS), 20 NMO spectrum disorder patients, and 9
tural morphological damage and impaired visual patients with chronic inflammatory optic neuropa-
function: GCL thickness was a stronger predictor of thy had previously experienced ON. In all three
visual acuity, including low-contrast acuity, in NMO groups, patients with unilateral ON and without
than in MS [15]. This may indicate that, below a visible MME showed an increased INL thickness
certain threshold of neuroaxonal loss, retinal neu- of the ON-affected eyes compared to the contralat-
rons and axons are no longer able to sufficiently eral eyes, suggesting that INL thickening and MME
maintain visual function. The latter study also sug- may both be symptoms of underlying retinal
gested that the pattern of retinal damage may differ inflammation [20]. Additionally, Gelfand et al.
between NMO and MS. While MS-ON eyes tended only detected MME in eyes with a prior history of
to show the typical preponderance of damage loca- ON in a cohort of 25 NMO patients [17]. Both studies
lized to the temporal quadrant, which is in line with showed that MME may be more frequent in NMO
clinical observations of optic disc pallor and optic (approx. 20%) as compared to MS (approx. 5%); how-
atrophy predominantly in this sector, NMO-ON eyes ever, this finding has to be interpreted with caution
exhibited a broader distribution of RNFL thinning given the low number of the subjects characterized in
across all quadrants, albeit with the highly salient the two NMO cohorts. In contradistinction to the
observation that the nasal quadrant RNFL thickness premise that MME represents a cardinal feature of
was more strongly affected than the temporal quad- retinal edema and inflammation restricted to a few
rant (Figure 11.1). Notwithstanding the differen- diagnostic entities, recent case reports on a patient
tiated patterns of retinal tissue damage when with optic nerve glioma and a patient with Leber’s
considering MS vs NMO, the construct validity of hereditary optic neuropathy (LHON), both of whom
these observations has yet to be confirmed, exhibited MME, strongly contested the idea that MME
thereby precluding the application of these pattern has an inflammatory pathogenesis, and instead sug-
differences as an objective criterion for differential gested that MME may be a consequence of any optic
diagnosis. nerve pathology, irrespective of their pathobiological
In 2012, microcystic macular edema (MME) was basis. [21–23]
reported in both MS and NMO in association with The literature to date is inconclusive regarding the
reduced visual acuity and retinal nerve fiber layer detection of retinal damage by OCT in NMO non-ON
thinning [12, 15–17]. MME presents as a eyes. In some studies, NMO non-ON eyes exhibited 129
 Chapter 11: Optical coherence tomography in neurodegenerative and other neurologic diseases

MSON NMOSD-ON MME- NMOSD-ON MME+

(F 29 y) (F 44 y) (F 34 y)

NS TS NS TS NS TS
102 96 58 63 38 70
(102) (137) (102) (134) (102) (136)
Within Normal Limits
(p>0.05)
RNFL

N/T N G T N/T N G T N/T N G T Borderline Below

2.51 88 82 35 PMB 1.04 37 49 36 PMB 0.77 21 42 27 PMB
(0.93) (72) (98) (77) 29 (0.96) (72) (97) (74) 33 (0.94) (72) (98) (76) 22 (p<0.05)
(58) (57) (58)
Below Normal Limits
(p<0.01)
NI TI NI TI NI TI
121 94 55 69 41 95
(107) (145) (106) (142) (106) (144)
GCL
INL

Figure 11.1 RNFL thinning in NMO vs. MS.

Three examples of OCT findings from an MS patient’s eye with previous ON (left), a neuromyelitis optica spectrum disorder (NMOSD)
patient’s eye with previous ON without microcystic macular edema (MME) (center), and an NMOSD patient’s eye with previous ON and MME
(right). The first row shows peripapillary RNFL thickness data for average RNFL (G) and sectors. The second and third rows show thickness maps
of the retinal ganglion cell layer (GCL) and inner nuclear layer (INL) (adapted from [15]).

SLO OCT

Figure 11.2 Microcystic macular edema (MME) in NMO

Examples of images from one NMO patient’s eye showing MME. In the scanning laser ophthalmoscopy (SLO) image on the left, MME is
visible as the darker crescent around the fovea. In the OCT B-scan on the right, MME can be identified by the small hypointense lesions
130 exclusively in the INL (green arrow).
 Chapter 11: Optical coherence tomography in neurodegenerative and other neurologic diseases

no apparent retinal damage, as all OCT measure- corresponding relationship to clinical disability, along
ments closely approximated those derived from with emerging utility as a tool for differentiating those
healthy control subjects [7, 8, 15]. This supports the pathophysiologic mechanisms that help distinguish
notion that retinal damage in NMO is linked to clinically NMO from MS and other conditions. Not surprisingly,
manifest ON attacks, and does not occur progressively the features that characterize high precision, high-
or as a consequence of subclinical optic neuropathy, speed, high-definition spectral-domain OCT have ush-
as may be the case in many MS patients [15]. ered in a new and exciting era, one where this powerful
Accordingly, NMO has rarely been found to exhibit technique may have utility in helping to advance treat-
features that would confirm the presence of a second- ing neuroinflammatory and neurodegenerative disor-
ary progressive course [24]. However, Sotirchos et al. ders by being endowed with the capability to detect and
and Syc et al. have reported both GCL and IPL thin- monitor preventative, protective, and even restorative
ning in NMO non-ON eyes, a highly conspicuous and properties of novel neurotherapeutic agents.
controversial finding that must be confirmed or
refuted in future studies [11,12]. Parkinson’s disease
In summary, OCT data on morphological retinal Parkinson’s disease (PD) is a neurodegenerative con-
damage in NMO-ON eyes and analysis of the correla- dition that typically presents with motor symptoms
tion between this damage and poor visual function are such as akinesia, tremor, and rigidity, and olfactory
abundant and largely consistent. In contrast, the patho- dysfunction. However, PD has also been widely
physiology of retinal damage in a disease characterized reported to include visual symptoms like decreased
by serum antibodies against the CNS water channel contrast sensitivity and color discrimination, visuos-
AQP4 has not been fully elucidated. Mechanistic stu- patial deficits, visual hallucinations [28–30], and some
dies have shown that AQP4-antibodies not only serve electrophysiological studies of visual evoked poten-
as a biomarker for an NMO diagnosis, but also are tials (VEPs) and pattern electroretinograms suggest
pathogenically relevant in terms of binding to astrocytic pathological signatures within the retina and/or the
membranes, the putative scaffolding upon which tar- anterior visual pathways [30–32]. While PD has long
geted damage to the aquaporin water channel can been associated with a neuropathological hallmark,
materialize [25, 26]. Could an inflammatory attack the accelerated loss of dopaminergic (DA) neurons
targeting the AQP4-expressing myelinated retrolami- within the substantia nigra, dopaminergic neurons
nar optic nerve conceivably cause retinal atrophy via a (amacrine cells) have also been detected in the INL
dying-back mechanism. A further and perhaps more and IPL of the human retina, thereby suggesting that
intriguing scientific question arises from evidence that the pathobiological basis for PD involves, at least in
the retina also contains a plethora of astrocytes that part, the specific targeting of DA neurons by disease-
express AQP4 at high levels (e.g., Müller cells are but specific injury cascades [33].
one of a number of astrocytic subtypes capable of cell- Retinal DA deficiency may contribute to visual
surface expression of this key target antigen within the impairment in PD, as DA is involved in visual proces-
retina). Taken together, the evidence would suggest that sing by modulating the organization of receptive fields
the retinal Müller cell represents a bona fide second of the retinal ganglion cell neurons, whose axons ulti-
ocular target of pathogenic AQP4 antibodies [27]. mately coalesce into the optic nerve [34, 35]. A post-
Confirmation of this potentially seminal observation mortem study showed decreased DA in retinas from
represents a formidable challenge, particularly given PD patients who had not been treated with levodopa
that adequate animal models for human NMO have [36], whereas an in vivo study involving PD patients
yet to be developed. reported improvement of spatial contrast sensitivity
In the clinical setting, the use of OCT in the after treatment with levodopa [37].
diagnosis and treatment of NMO is proliferating Several OCT studies have investigated retinal atro-
throughout the neurological and ophthalmological phy in patients with PD with diverging results, prob-
communities, with the powerful advent of performing ably as a result of differences between the cohorts and
a simple, rapid, convenient, reproducible, and nonin- the OCT devices used. The validity of most OCT
vasive assessment of an elegantly eloquent CNS studies in PD is hampered by small cohort size and
compartment, with compelling ramifications for long- the insufficient or ambiguous consideration of con-
itudinal ascertainment of retinal damage, and its founding factors prevalent among PD patients, 131
 Chapter 11: Optical coherence tomography in neurodegenerative and other neurologic diseases

including glaucoma and diabetic retinopathy, to name results, our study, in which we had applied modern
a few. Altintas et al. found a significant reduction in SD OCT technology with intra-retinal segmentation,
average RNFL thickness and TMV in 17 PD patients as previously used in numerous studies [49], clearly
vs. healthy controls, and an inverse correlation showed that retinal DA deficiency may cause struc-
between foveal retinal thickness and the Unified tural retinal damage beyond the RNFL that is detect-
Parkinson Disease Rating Scale (UPDRS), a clinical able by OCT. The cause for this observation remains
scale to assess neurological impairment in PD [38]. largely unexplained, and we have already formulated
In another investigation, Inzelberg et al. reported a strategic plans to fully characterize this pathological
reduction of RNFL thickness in inferior and temporal substrate. For example, the PRL contains rods and
quadrants in 10 PD patients [39], while a another study cones that receive input from dopaminergic inter-
found a decreased foveal thickness in 9 PD patients plexiform and amacrine cells via the inner and outer
[40]. Hajee et al. demonstrated significant thinning of plexiform layers [50, 51]. Thus, we may assume a
the inner retinal layers in 24 PD patients [41], and process of transsynaptic degeneration of ONL neurons
Aaker et al. found differences in macular thickness as a consequence of altered synaptic input from dopa-
(thinning and thickening), but no differences in minergic neurons. The fact that atrophy of the ONL/
RNFL thickness in 9 PD patients vs. 16 controls [42]. PRL was not associated with deficits in color discrimi-
Shrier et al. found increased intraocular differences nation may be – at least in part – explained by the
in macular volume between 23 PD patients and 18 observation that FMT is prone to various confounding
controls [43]. In contrast, Archibald et al. did not elements such as motor and cognitive impairment as
find any alteration in RNFL thickness or TMV in 37 well as dopaminergic treatment.
PD patients [44]. Likewise, application of modern SD In summary, OCT data in PD are highly hetero-
OCT technology in another study showed no differ- geneous, and the value of OCT in the clinical assess-
ences in RNFL and TMV between 40 PD patients and ment of patients with PD remains to be established. A
25 healthy controls [45]. Interestingly, manual intra- complicating factor is that many PD patients have
retinal segmentation revealed a thicker INL in PD vs. difficulties complying with the experimental require-
controls in this study. No correlation between OCT ments of OCT examination (thereby leading to a clear
measures and visual acuity, disease duration, or disease acquisition bias) due to head or neck tremor, axial
severity according to the UPDRS was observed. In a rigidity, or cognitive impairment. Recognizing these
recent study using SD OCT, Garcia-Martin et al. inves- important and test-limiting factors may represent at
tigated 75 PD patients in comparison with 75 matched least part of the bases for the discrepant outcomes
healthy controls and found subtle RNFL reduction across previous studies. Ultimately, OCT image quality
using three different RNFL scan protocols [46]. from PD patients has been demonstrated to be of lower
Our group recently completed a spectral-domain than the image characteristics derived from healthy
OCT study on 97 patients with idiopathic PD and 32 controls. Image quality cannot be ignored, especially
healthy controls [47]. Intra-retinal segmentation given that OCT generated data can easily be skewed in
revealed significant thinning of the combined outer the setting of poor-quality images [52].
nuclear (ONL) and photoreceptor layer (PRL) in PD
vs. controls (118.6 vs. 123.5 µm, p = 0.001). Neither Alzheimer’s disease
RNFL thickness nor TMV or any of the other layers Alzheimer’s disease (AD), which predominantly
differed from those of controls. We investigated affects episodic memory, is the most common neuro-
whether retinal damage is associated with impaired degenerative cause of dementia in the elderly. The
color vision, as assessed by the Farnsworth-Munsell pathophysiological cause is unclear, but the deposi-
color discrimination test (FMT), as color perception tion of Aβ plaques in the brain has been suggested as
has been reported to be altered in PD, a salient feature having a major role and is linked – along with other
that may be associated with disease progression [48]. biomarkers – to clinical disease severity [53].
We identified no correlation between performance in Recently, the presence of Aβ plaques in the retina
the FMT, and any of the OCT measures. Furthermore, was reported in a murine Alzheimer’s disease model,
no association was found between OCT measures and and confirmed in postmortem tissue from
clinical parameters such as UPDRS or disease dura- Alzheimer’s patients. Importantly, retinal Aβ plaques
132 tion. Despite the heterogeneity of previous OCT correlated with brain pathology and clinical severity,
 Chapter 11: Optical coherence tomography in neurodegenerative and other neurologic diseases

and were detectable in suspected AD patients [54]. addition to AD patients also investigated cohorts
These recent observations might lead to AD-specific with mild cognitive impairment (a precursor symp-
retinal imaging, once labeling of the pathognomonic tom of AD), reported a similar finding in 49 cogni-
Aβ plaque becomes feasible in human subjects. tively impaired patients. Here, RNFL reduction was
The authors of this study (ref 54) have provided the more prominent in the more severely affected
first in vivo data in mice using systemically adminis- patients [68]. Lu et al. reported RNFL thinning in
tered curcumin for the specific labeling of Aβ the superior and inferior quadrants in 22 patients
plaques. compared to age-matched healthy controls [69].
Currently, researchers and clinicians seeking to Berisha and colleagues studied nine early stage AD
focus on the retina of AD patients face two chal- patients vs. matched healthy subjects and found that
lenges. First, although visual symptoms in AD the AD patients showed significant RNFL reduction
patients are widely reported, they could easily be in the superior quadrant (AD, superior RNFL 92 ±22
the result of impairment to higher cognitive µm compared to the controls’ superior 114 ±1 µm)
functions necessary for visual processing [55, 56]. [70]. In one of the largest AD cohorts investigated
Second, the high coincidence of other pathologies to date, which included 40 patients and 40 well-
commonly associated with the aging retina proble- matched healthy subjects, Kirbas and colleagues
matizes analysis. Age-related macular dystrophy reported significant RNFL reductions (AD, average
(AMD), vascular changes, wide-angle glaucoma, RNFL 65 ±6 µm vs. average RNFL 75±4 µm). Again,
cataract, and diabetic retinopathy are common RNFL thinning was predominantly located in the
comorbidities that partially account for the reported superior quadrant [71].
visual dysfunction in AD (see also below). In summary, these OCT studies suggest signifi-
Consequently, distinguishing between retinal cant RNFL reduction is a common feature of AD.
changes that are potentially AD-related and those As in PD, owing to comorbid physical and cogni-
caused by comorbidities is difficult. tive symptoms, performing OCT examinations reli-
Studies investigating retinal changes in AD have ably on AD patients can represent a formidable
yielded conflicting results: A reduction of retinal challenge, one that markedly limits obtaining data
ganglion cells (RGC) in comparison to healthy con- of acceptable quality, whether used in clinical prac-
trols was reported in an early postmortem study tice or as outcome measures in clinical research.
investigating 10 AD patients [57], as well as in a The recognized high standard deviations of
further study comprised of 9 AD patients with the data collected from small cohorts of patients
matched healthy subjects [58]. RGC reduction was afflicted by neurodegenerative disease may reflect
confirmed by investigating postmortem optic nerve the broad pathological heterogeneity across patients
tissue from AD patients, what showed prominent carrying the same diagnosis. Further, the age
loss of M-cells, the largest class of RGCs contribut- matching in most of the referenced studies lacked
ing large calibre fibers to the optic nerve [59]. Other stringency, adding age as a potentially important
histopathological studies reported no differences in confounder.
the layout in the AD patient’s retina [60–62]. The Of note in this regard is another important
latter is supported by several studies using scanning finding in AD: the high co-occurrence with open-
laser polarimetry, retina tomography, and pattern angle glaucoma (OAG). In comparison to healthy
electroretinography [62–66]. subjects, AD patients show a higher occurrence
Several studies have applied OCT to investigate of OAG [72], and subjects with OAG show a
retinal changes in AD patients. Iseri and colleagues higher risk to develop AD in later life [73]. The
investigated 14 AD patients vs. 15 matched healthy link between OAG and AD is currently unclear and
subjects. They reported a reduction of RNFL (AD, needs to be investigated further in the future.
average RNFL 87±24 µm (mean ± standard devia- Conspicuously, the preponderance of RNFL loss
tion) vs. controls, average RNFL 113 ±7 µm) and in the superior quadrant in most of the studies
total macula volume (AD, TMV 6.8 ±0.4 mm3 vs. interrogated may also represent a potential
controls, TMV 7.1 ±0.2 mm3). Reduction in TMV confounding effect, namely, undiagnosed glau-
correlated with clinical severity expressed by mini- coma, which also tends to affect the superior quad-
mental state examination [67]. Paquet et al., who in rant [74]. 133
 Chapter 11: Optical coherence tomography in neurodegenerative and other neurologic diseases

M-1 M0 M2 700

500

Retinal Thickness (µm)

300

200 µm
100

200 µm

Figure 11.3 Formation of a Susac syndrome retinal lesion.

Examples of findings from a Susac syndrome patient’s eye. One month prior to the acute lesion (M-1 on the left), the retina appears mostly
normal in the nasal-inferior area around the macula. During the acute event (M0), massive swelling can be observed around the retinal branch
vessel directly inferior to the macula, indicated by the red arrows. After two months (M2 on the right), the swelling has decreased and the inner
retinal layers at the site of prior inflammation and in all of the area supplied by this vessel are almost completely degenerated, indicated by the
red arrows.

Figure 11.4 Papilledema in idiopathic intracranial hypertension.

HC IIH
Comparison of a 3-D optic nerve head scan from a healthy
subject (HC on the left) and a patient with idiopathic intracranial
hypertension (IIH) with increased ICP (adapted from [79]).

Other neurologic diseases sectorial damage that can be distinguished from

MS patients [77]. During an acute event of Susac
Susac syndrome syndrome, the affected areas show profound swel-
Susac syndrome presents as a combination of ling, which later leads to almost complete loss of all
encephalopathy, hearing loss, and visual deficits inner retinal layers supplied by the occluded branch
due to acute branch retinal artery occlusions [75]. artery (Figure 11.3). This presentation during the
The etiology of this rare disease and its exact acute phase of visual symptoms clearly differenti-
prevalence are unknown [76]. Differentiating it ates Susac syndrome from optic neuritis resulting
from MS can be challenging since the correspond- from MS or NMO, where acute and chronic damage
ing clinical presentation and diagnostic findings is much more broadly distributed and certainly not
may significantly overlap. However, in OCT, a consequence of a single vessel distribution of
134 eyes from patients with Susac syndrome show localized tissue damage.
 Chapter 11: Optical coherence tomography in neurodegenerative and other neurologic diseases

Table 11.1

Susac FRDA SCA-1 SCA-2

Study Brandt et al. Fortuna et al. Noval et al. Stricker et al. Pula et al. [90] Pula et al.
[77] [84] [86] [89] [90]
Year 2011 2009 2012 2010 2011 2011
Device Stratus Stratus Stratus Spectralis Spectralis Spectralis
TD-OCT TD-OCT TD-OCT SD-OCT SD-OCT SD-OCT
Cohort 9 (33±11 26 (32±8 23 (25±7 y) 9 (52±9 y, 4/ 7 (n=29, 58 y, 7 (n=29,
(age, y, 6/3) y, 8/18) 5) 18/11)* 58 y, 18/
F/M) 11)*
Matched matched HC 48 HC (33±8 none 9 HC (51±9 y, 27 HC (56 y, 18/ 27 HC
cohorts and MS y, 20/28) 4/5) 9)* (56 y,
(age 18/9)*
F)/M
Main Significant Significant RNFL Significant Non-significant Significant
results RNFL RNFL reduction in reduction of RNFL reduction RNFL
reduction reduction (76 comparison average (93±8 µm) vs. reduction
(81±18 µm) ±12 µm) vs. to norma- RNFL (84±13 HC (98±9 µm). (84±12
vs. HC (107 HC (100±9 tive data µm) and µm) vs. HC
±9 µm). in µm). and with temporal (98±9 µm).
comparison Description good RNFL (62±8
to MS sec- of 3 distinct correlation µm) vs. HC
torial loss types to visual average
damage correlating function RNFL (97±8
both in RNFL visual tests. µm) and
and TMV. dysfunction. temporal
RNFL (74±9
µm).
Summary Sectorial RNFL reduc- RNFL RNFL No changes in Significant
damage tion in 3 reduction. reduction RNFL. RNFL
separates distinct types with reduction.
Susac retinal correlating temporal
damage with visual quadrant
from ON function. focus.
related
retinal
damage.

In differential diagnosis – for example, when clinicians, can also be measured using OCT
acute visual symptoms present in combination with (Figure 11.4) [78]. In idiopathic intracranial hyper-
cognitive changes or hearing loss – OCT should tension (IIH), our group could show that ONH
always be performed to exclude possible Susac syn- volume measurements can reliably quantify the papil-
drome. A suspected diagnosis of Susac syndrome can ledema and discriminate between healthy subjects,
then be substantiated using fluorescence angiogra- IIH patients undergoing clinically effective therapy
phy (FAG). but who nevertheless exhibited increased ONH
volume, and untreated patients, who showed the lar-
Intracranial hypertension gest ONH volume [79]. This parameter (i.e., ONH
Elevated intracranial pressure can present as papille- volume) correlated highly with intracranial pressure
dema. This edema of the optic nerve head (ONH), (ICP), suggesting that OCT is also of benefit in the
long familiar to neurology and ophthalmology diagnosis and monitoring of patients affected by other
135
 Chapter 11: Optical coherence tomography in neurodegenerative and other neurologic diseases

Wilson’s
SCA-3 SCA-6 SCA-7 MSA-C ARSACS ALS Disease
Pula et al. Pula et al. Manrique Pula et al. Vingolo Roth et al. Albrecht
[90] [90] et al. [91] [90] et al. [95] [101] et al.
2011 2011 2009 2011 2011 2013 2012
Spectralis Spectralis Stratus Spectralis not Cirrus Spectralis
SD-OCT SD-OCT TD-OCT SD-OCT reported HD-OCT SD-OCT
5 (n=29, 58 6 (n=29, 58 y, 7 (5 from 5 (n=29, 58 y, 5 (34±8 y, 76 (56±11 y, 42 (40±14
y, 18/11)* 18/11)* one family) 18/11)* 5/3) 26/50) y, 24/18)
27 HC (56 27 HC (56 y, none 27 HC (56 y, 5 HC (62±8 54 HC (56 76 HC (43
y, 18/9)* 18/9)* 18/9)* y, 4/1) ±11 y, 26/28) ±13 y, 35/
29)
Significant Non-signifi- Case-wise Non-signifi- Case-wise No changes Significant
RNFL cant RNFL discussion. cant RNFL discussion. in RNFL (94 RNFL
reduction reduction RNFL reduc- changes RNFL thick- ±9 µm) vs. reduction
(85±9 µm) (95±5 µm) tion with (100±11 µm) ening with HC (93±10). (95±9 µm)
vs. HC (98 vs. HC (98±9 focus on the vs. HC (98±9 myelinated No vs. HC (97
±9 µm). µm). superior and µm). appearance. differences ±10).
Significant in more Significant in GCL, INL/ Reduction
reduction of severe cases reduction of OPL and in GCL and
macular also on nasal macular ONL/PRL. INL.
thickness and inferior thickness Only 15
(310±5 µm quadrants. (314±12 µm) definite ALS
vs. HC (339 vs. HC (339 according to
±17 µm). ±17 µm). El Escorial
criteria.
Significant Significant Mild to No changes. RNFL thick- No changes. RNFL, RGC
RNFL macular severe RNFL ening with and INL
reduction. thickness loss myelinated thinning.
reduction focussed on appearance.
but no superior,
changes in nasal and
RNFL. inferior
quadrants.

* Pula et al [90]. only reported cohort statistics for the full cohort, not for disease subtypes

diseases with elevated ICP. Another study using the Hereditary and other rare neurode-
Frisén scale, which quantifies edema using fundus
photography, also showed good correlation between generative diseases
an alternative ONH quantification approach and OCT has been applied in several other hereditary or
Frisén scale derived papilledema grades [80]. rare neurodegenerative diseases. With a few excep-
Currently, further studies are needed on the temporal tions, most published OCT studies show varying
dynamics in conjunction with inter- and intra- levels of RNFL reduction in the peripapillary ring
individual changes in ONH volume, and visual func- scan that is often accentuated in the temporal quad-
tion loss and ICP. Such studies will determine whether rant. Fibers of the temporal quadrant RNFL comprise
OCT has satisfactory utility as a surrogate biomarker in primarily parvo-cellular axons from the papillo-
the assessment of ICP. macular bundle. These axons consist of small and
136
 Chapter 11: Optical coherence tomography in neurodegenerative and other neurologic diseases

thinly myelinated axons with rapid firing rates that challenging because patients often present with only
connect macular ganglion cells via the lateral genicu- a subset of the symptoms known for SCA diseases as
late nucleaus with the visual cortex. As such, they are a whole [88]. OCT can aid in differential diagnosis
highly relevant to high-resolution visual acuity and by identifying patients with optic nerve atrophy
high spatial frequency of contrast sensitivity and color especially without apparent visual symptoms.
vision. Degeneration of these macular fibers has a However, in many diseases OCT changes might be
much greater effect on visual quality of life than statistically significant but hardly relevant on an
defects in the peripheral visual areas. Unfortunately, individual level.
precisely these fibers are affected by neurodegenera- Retinal changes identified by OCT have been
tion in many mitochondriopathies and other optic reported in SCA patients with and without visual
neuropathies, such as Leber’s hereditary optic neuro- symptoms. SCA-1 presents with a mild and tem-
pathy (LHON) and OPA1-related dominant optic porally focused RNFL reduction, and exhibits fea-
nerve atrophy (DOA) [81]. One reason for this may tures similar to other mitochondria-related optic
be their small volume and fast firing, which might make neuropathies [89]. SCA-2, SCA-3 and SCA-6
them more susceptible to energy depletion by defective patients also show RNFL or macular thickness
cellular metabolism and, thus, neurodegeneration [82]. reductions [90]. The focal distribution of changes
Consequently, OCT assessment of RNFL focuses on the in many SCAs is currently unclear, as is a poten-
peripapillary ring rather than the macula. An overview tial correlation with clinical scores, because most
of current OCT studies is given in Table 11.1, and a studies have involved only a few patients, espe-
summary of key findings is provided in the following cially in the rare SCAs, which limits their useful-
paragraphs. ness for detailed review.
Of the forms of SCA, SCA-7 is unique in hav-
Friedreich’s ataxia ing a high prevalence of retinal photoreceptor
One of the first hereditary ataxias investigated in abnormalities that appear to be directly related to
detail using OCT was Friedreich’s ataxia (FRDA) disease pathology. Consequently, SCA-7 patients
[83]. Visual symptoms and optic neuropathy are regularly present with visual symptoms. However,
regularly found in FRDA [84] and mitrochondrio- current OCT technology can only detect these
pathy is likely to directly contribute to the disease photoreceptor anomalies with difficulty and pri-
pathophysiology [85]. In OCT, FRDA patients marily in advanced disease stages when RGCs
present with widespread RNFL reduction. and retinal axons in the RNFL undergo secondary
Notably, visually asymptomatic patients also neurodegeneration [91]. In contrast to the tem-
show optic neuropathy, the level of which corre- poral quadrant RNFL effect usually found in
lates to general clinical severity and progression mitochondria-related disorders and other SCAs
[84, 86, 87]. Fortuna et al. proposed three different [92, 93], in SCA-7 patients’ eyes, the superior,
types of RNFL degeneration in FRDA: Type 1 with inferior, and nasal quadrants are preferentially
diffuse and severe reduction of RNFL thickness in affected, whereas the temporal quadrant is mainly
all quadrants, Type 2 with diffuse RNFL reduction affected during late stages of disease [91].
that is more marked in the superior quadrant, Preliminary data on OCT findings further exist in
and Type 3 with diffuse but mild reduction of SCA-14. In the largest cohort of SCA-14 patients
RNFL thickness in all quadrants [84]. However, investigated to date (n = 15) – a SCA that does not
the clinical and pathophysiological significance commonly present with visual symptoms – there were
of the individual categories remains to be no SCA-14 specific retinal changes detectable on test-
elucidated. ing (unpublished data).

Spinocerebellar ataxias Autosomal recessive spastic ataxia

Spinocerebellar ataxias (SCAs) comprise a group of of Charlevoix-Saguenay
over 30 genetically distinct hereditary ataxias with Autosomal recessive spastic ataxia of Charlevoix-
similar clinical features. They are mostly numbered Saguenay (ARSACS) is a rare disorder that was first
according to the date of their first description. identified in the Quebec region but can today be
Differential diagnosis of these disorders can be found throughout Northern America and Europe 137
 Chapter 11: Optical coherence tomography in neurodegenerative and other neurologic diseases

A OD OS complex forms of the disease but not in clinically

pure forms [100].

Amyotrophic lateral sclerosis

Amyotrophic lateral sclerosis (ALS) is a neurodegen-
erative disorder affecting upper and lower motor
neuron tracts. Although visual symptoms are uncom-
mon, VEP abnormalities have been reported.
However, there is currently no evidence that OCT
B
can detect retinal involvement in amyotrophic lateral
sclerosis (ALS) [101]. This requires confirmation in
further studies, because genetically distinct subtypes
of this heterogeneous disease, which has several
different clinical phenotypes, might lend themselves
to assessment by OCT.

Wilson’s disease
Wilson’s disease (WD) is a hereditary disorder of
copper accumulation that leads to hepatic damage
and neurodegeneration [102]. Although only few
patients report visual symptoms, many patients
show pathological findings in electroretinography
and VEP [103]. In OCT, patients with WD show
reduction in RNFL and RGC thickness [104].

Limitations
Figure 11.5 OCT findings in ARSACS.
While OCT is likely helpful to further elucidate the
Examples of an OCT report from an ARSACS patient. (A) RNFL
thickness maps around the optic nerve head show RNFL thickening; natural or treated history of neurodegenerative
(B) RNFL thickening can also be detected using the peripapillary disease, or to complete models of structure–function
RNFL ring scan. Comparison to RNFL distribution of normal correlation for treatment, OCT measurements are not
controls shows elevated RNFL thickness in the inferior and superior
quadrants (kindly provided by Dr. Elena Garcia-Martin, Zaragoza, specific for any particular disorder.
Spain). First, RNFL changes with age. In healthy sub-
jects, annual RNFL decrease is reported as between
[94]. In OCT, patients with ARSACS present with –0.16 µm/year [105] and –0.44 µm/year [106].
increased RNFL thickness [95], which has been However, this decrease is not evenly distributed
proposed as due to pathological myelination over age but is initially slow and accelerates
of retinal axons [96] or axonal hypertrophy after the age of 50 [105, 106]. A similar process
(Figure 11.5) [97]. takes place in the macula and affects mainly the
RNFL, RGC, IPL, and INL but not, it appears,
Hereditary spastic paraplegia the outer plexiform and nuclear layers, including
Hereditary spastic paraplegias (HSPs) are progres- the photoreceptors [107]. Age-related accelerated
sive neurodegenerative disorders with several dis- loss of RNFL and macular thickness in older patients
tinct genetic phenotypes [98]. Common features makes OCT assessment of optic nerve or retinal
are a dysfunction of axonal transport, predomi- neurodegeneration in neurodegenerative diseases
nantly leading to axonal length-dependent neurode- more suitable in younger patients. In clinical
generation of corticospinal tracts. Visual symptoms research, exact age matching of cohorts is, therefore,
are rarely reported and mostly are limited to com- essential to ensure the validity of results. Minor
plex forms of the disease [99]. OCT shows RNFL differences in age can produce differences in results
reduction in temporal quadrants in clinically unrelated to the neuropathology being studied.
138
 Chapter 11: Optical coherence tomography in neurodegenerative and other neurologic diseases

Additionally, the coincidence of other eye diseases, differential diagnosis and follow-up of patients
particularly glaucoma (GD) but also diabetic retino- with (suspected) MS and related neuroimmunolo-
pathy (DR) and AMD, increases with age. Above all, gic diseases like NMO or Susac syndrome. Above
the risk of GD and here open-angle glaucoma rises all, OCT should always be performed as part of
considerably after the age of 50 [108]. While open- the differential diagnosis of patients presenting
angle glaucoma prevalence among 50-year-olds is with symptoms indicating acute optic neuritis
only 0–3%, in 80-year-olds it rises to 5–15%. Its [112].
incidence is also strongly ethnicity-dependent, Furthermore, using OCT to detect optic nerve
with the disorder being more common among atrophy, which is characteristic of some, but not all,
African, Asian, and Latin American groups than neurodegenerative diseases is of benefit in the differ-
Caucasians [109]. Moreover, GD often also remains ential diagnosis of other complex, often-related dis-
undiagnosed [110] because retinal changes due eases. It provides an objective means of detecting
to GD are sometimes difficult to differentiate visual system involvement, which is challenging to
from reported findings in several neurodegener- assess and often overlooked, particularly in patients
ative diseases. This makes GD an important con- with cognitive impairment. In our experience,
founder that should be considered in both research OCT is a means of tracing visual symptoms in
and clinical routine. Although some OCT studies these patients to retinal or optic nerve abnormalities.
discussed here went to considerable lengths Therefore, we suggest routinely performing OCT if
to minimize potential confounder effects especially at least one important differential diagnosis invol-
all of the studies investigating older cohorts ving the retina is under discussion for a patient.
should be interpreted with these limitations in Moreover, several rare diseases show distinct retinal
mind. pathologies that can easily be identified by means of
Second, the majority of existing studies, espe- OCT.
cially in rarer neurodegenerative diseases, are The current value of OCT in AD and PD is far
cross-sectional studies, which often included only a more difficult to assess. Both diseases show clear ret-
small number of subjects. Only in very few diseases, inal pathologies – in the case of AD the formation of
such as NMO and SCA-1, have findings been possible retinal Aβ plaques and in the case of PD
reported and confirmed by several independent neurodegeneration of dopaminergic ganglion cells
groups. In all cases, the cross-sectional nature of and potentially also photoreceptors. However, state-
the studies makes validating the disease specificity of-the-art OCT is currently unable to trace these
of many findings difficult. Longitudinal studies are changes, which are probably simply too subtle.
urgently required to better implement disease- Higher imaging resolution or functional OCT macu-
specific and neurodegeneration-generic OCT find- lar scan applications that specifically target the
ings into clinical routine. affected structures in these diseases might well solve
Finally, OCT examinations are more demand- this limitation in the future.
ing in severely clinically affected patients, who Currently, both diseases show a high
have difficulty complying with the needs of the co-occurrence with other diseases or changes of the
experimental setting due to cognitive or motor aging retina, such as GD DR, AMD, and vascular
impairment. In clinical routine, and especially in changes. This makes applying OCT challenging, par-
research, strictly adhering to the quality guidelines ticularly when attempting to track disease-specific
given in literature is vital to preventing artifacts changes. More importantly, such confounders
and diagnostic errors [111]. Any significant varia- increase the quality requirements for OCT studies
tion in image quality should be investigated and on AD and PD tremendously and hamper assessing
reported [52]. the validity of published studies. Current data do
suggest OCT is a valuable tool for studying AD and
PD in research settings. However, clinician research-
Summary ers should interpret possible findings cautiously,
OCT has potential benefit in the assessment of keeping the preceding issues in mind.
several neurodegenerative diseases. In routine clin- We expect OCT will prove a valuable follow-up
ical practice, OCT is highly recommended for both marker in many of the diseases discussed previously. 139
 Chapter 11: Optical coherence tomography in neurodegenerative and other neurologic diseases

However, more research is required into its role as a optical coherence tomography in multiple sclerosis: a
disease monitoring tool before we can support its comparative cross-sectional study. Mult
integration into everyday routine clinical practice. Scler Houndmills Basingstoke Engl. July
2010;16(7):893–6.
10. Oberwahrenbrock T, Ringelstein M, Jentschke S,
Acknowledgments Deuschle K, Klumbies K, Bellmann-Strobl J, et al.
We thank Ella Maria Kadas and Timm Retinal ganglion cell and inner plexiform layer
Oberwahrenbrock for assistance with preparing the thinning in clinically isolated syndrome. Mult
figures. This work was supported by the German Scler December 2013;19(14): 1885–95.
Research Council (DFG Exc 257 to FP) and the doi: 10.1177/1352458513489757. Epub May 23,
2013.
German Ministry for Education and Research
(KKNMS Competence Network Multiple Sclerosis 11. Syc SB, Saidha S, Newsome SD, Ratchford JN, Levy
M, Ford E, et al. Optical coherence tomography
to FP). segmentation reveals ganglion cell layer pathology
after optic neuritis. Brain J Neurol [Internet].
References October 17, 2011 [cited December 6, 2011];
1. Jarius S, Ruprecht K, Wildemann B, Kuempfel T, Available from: http://www.ncbi.nlm.nih.gov/pubme
Ringelstein M, Geis C, et al. Contrasting disease d/22006982
patterns in seropositive and seronegative 12. Sotirchos ES, Saidha S, Byraiah G, Mealy MA,
neuromyelitis optica: A multicentre study Ibrahim MA, Sepah YJ, et al. In vivo identification
of 175 patients. J Neuroinflammation of morphologic retinal abnormalities in
2012;9:14. neuromyelitis optica. Neurology April 9,
2. Wingerchuk DM, Lennon VA, Pittock SJ, Lucchinetti 2013;80(15):1406–14.
CF, Weinshenker BG. Revised diagnostic criteria for 13. Monteiro MLR, Fernandes DB, Apóstolos-Pereira
neuromyelitis optica. Neurology May 23, SL, Callegaro D. Quantification of retinal neural
2006;66(10):1485–9. loss in patients with neuromyelitis optica and
3. Jarius S, Paul F, Franciotta D, Waters P, Zipp F, multiple sclerosis with or without optic neuritis
Hohlfeld R, et al. Mechanisms of disease: using Fourier-domain optical coherence
aquaporin-4 antibodies in neuromyelitis optica. Nat tomography. Invest Ophthalmol Vis Sci
Clin Pract Neurol April 2008;4(4):202–14. 2012;53(7):3959–66.
4. Bennett JL, Seze J de, Lana-Peixoto M, Palace J, 14. Fernandes DB, Raza AS, Nogueira RGF, Wang D,
Waldman A, Schippling S, et al. Neuromyelitis optica Callegaro D, Hood DC, et al. Evaluation of
and multiple sclerosis: Seeing differences through inner retinal layers in patients with multiple
optical coherence tomography. Mult Scler February 6, sclerosis or neuromyelitis optica using optical
2015. pii: 1352458514567216. coherence tomography. Ophthalmology.
5. Wingerchuk DM, Hogancamp WF, O’Brien PC, October 17, 2012;
Weinshenker BG. The clinical course of neuromyeli- 15. Schneider E, Zimmermann H, Oberwahrenbrock T,
tis optica (Devic’s syndrome). Neurology September Kaufhold F, Kadas EM, Petzold A, et al. Optical
22, 1999;53(5):1107–14. coherence tomography reveals distinct patterns of
6. Naismith RT, Tutlam NT, Xu J, Klawiter EC, retinal damage in neuromyelitis optica and
Shepherd J, Trinkaus K, et al. Optical coherence multiple sclerosis. PLoS ONE. June 21,
tomography differs in neuromyelitis optica compared 2013;8(6):e66151.
with multiple sclerosis. Neurology. March 24, 16. Gelfand JM, Nolan R, Schwartz DM, Graves J,
2009;72(12):1077–82. Green AJ. Microcystic macular edema in multiple
7. Ratchford JN, Quigg ME, Conger A, Frohman T, sclerosis is associated with disease severity.
Frohman EM, Balcer LJ, et al. Optical coherence Brain [Internet]. April 25, 2012 [cited
tomography helps differentiate neuromyelitis optica May 30, 2012]; Available from: http://brain.oxfor
and MS optic neuropathies. Neurology July 28, djournals.org/content/early/2012/04/25/brain.
2009;73(4):302–8. aws098
8. De Seze J, Blanc F, Jeanjean L, Zéphir H, Labauge P, 17. Gelfand JM, Cree BA, Nolan R, Arnow S, Green
Bouyon M, et al. Optical coherence tomography in AJ. Microcystic inner nuclear layer abnormalities
neuromyelitis optica. Arch Neurol July and neuromyelitis optica. JAMA Neurol. March
2008;65(7):920–3. 25, 2013;1–5.
9. Bock M, Brandt AU, Dörr J, Pfueller CF, Ohlraun S, 18. Green AJ, McQuaid S, Hauser SL, Allen IV, Lyness
140 R. Ocular pathology in multiple sclerosis: retinal
Zipp F, et al. Time domain and spectral domain
 Chapter 11: Optical coherence tomography in neurodegenerative and other neurologic diseases

atrophy and inflammation irrespective of disease. Loss in spatiotemporal contrast

disease duration. Brain J Neurol June 2010;133 sensitivity. Brain J Neurol December 1987;110
(Pt 6):1591–601. (Pt 6):1675–98.
19. Saidha S, Sotirchos ES, Ibrahim MA, Crainiceanu 31. Nightingale S, Mitchell KW, Howe JW. Visual
CM, Gelfand JM, Sepah YJ, et al. Microcystic evoked cortical potentials and pattern
macular edema, thickness of the inner nuclear electroretinograms in Parkinson’s disease
layer of the retina, and disease characteristics and control subjects. J Neurol
in multiple sclerosis: a retrospective study. Neurosurg Psychiatry November
Lancet Neurol October 4, 2012; 1986;49(11):1280–7.
20. Kaufhold F, Zimmermann H, Schneider E, Ruprecht 32. Tagliati M, Bodis-Wollner I, Yahr MD. The pattern
K, Paul F, Oberwahrenbrock T, et al. Optic Neuritis Is electroretinogram in Parkinson’s disease
Associated with Inner Nuclear Layer Thickening and reveals lack of retinal spatial tuning.
Microcystic Macular Edema Independently of Electroencephalogr Clin Neurophysiol January
Multiple Sclerosis. PLoS ONE Aug 6, 2013;8(8): 1996;100(1):1–11.
e71145. 33. Frederick JM, Rayborn ME, Laties AM, Lam DM,
21. Barboni P, Carelli V, Savini G, Carbonelli M, Hollyfield JG. Dopaminergic neurons in the
La Morgia C, Sadun AA. Microcystic macular human retina. J Comp Neurol September 1,
degeneration from optic neuropathy: not inflamma- 1982;210(1):65–79.
tory, not trans-synaptic degeneration. Brain J Neurol 34. Djamgoz MB, Hankins MW, Hirano J, Archer
February 8, 2013; SN. Neurobiology of retinal dopamine in
22. Abegg M, Zinkernagel M, Wolf S. Microcystic relation to degenerative states of the
macular degeneration from optic neuropathy. Brain tissue. Vision Res December
J Neurol September 7, 2012; 1997;37(24):3509–29.
23. Brandt AU, Oberwahrenbrock T, Kadas EM, Lagrèze 35. Jackson GR, Owsley C. Visual dysfunction,
WA, Paul F. Dynamic formation of macular micro- neurodegenerative diseases, and aging. Neurol
cysts independent of vitreous traction changes. Clin August 2003;21(3):709–28.
Neurology July 1, 2014;83(1):73–7. doi: 10.1212/ 36. Harnois C, Di Paolo T. Decreased dopamine in the
WNL.0000000000000545. Epub May 23, 2014. retinas of patients with Parkinson’s disease.
24. Wingerchuk DM, Pittock SJ, Lucchinetti CF, Lennon Invest Ophthalmol Vis Sci November
VA, Weinshenker BG. A secondary progressive 1990;31(11):2473–5.
clinical course is uncommon in 37. Hutton JT, Morris JL, Elias JW. Levodopa improves
neuromyelitis optica. Neurology February 20, spatial contrast sensitivity in Parkinson’s disease.
2007;68(8):603–5. Arch Neurol July 1993;50(7):721–4.
25. Jarius S, Wildemann B, Paul F. Neuromyelitis optica: 38. Altintaş O, Işeri P, Ozkan B, Cağlar Y. Correlation
clinical features, immunopathogenesis and treat- between retinal morphological and functional find-
ment. Clin Exp Immunol May 2014;176(2):149–64. ings and clinical severity in Parkinson’s disease. Doc
doi: 10.1111/cei.12271. Review. Ophthalmol Adv Ophthalmol March
26. Hinson SR, Pittock SJ, Lucchinetti CF, Roemer SF, 2008;116(2):137–46.
Fryer JP, Kryzer TJ, et al. Pathogenic potential of 39. Inzelberg R, Ramirez JA, Nisipeanu P, Ophir A.
IgG binding to water channel extracellular Retinal nerve fiber layer thinning in
domain in neuromyelitis optica. Neurology Parkinson disease. Vision Res November
December 11, 2007;69(24):2221–31. 2004;44(24):2793–7.
27. Levin MH, Bennett JL, Verkman AS. Optic neuritis 40. Cubo E, Tedejo RP, Rodriguez Mendez V, López Peña
in neuromyelitis optica. Prog Retin Eye Res March MJ, Trejo Gabriel Y Galán JM. Retina thickness in
30, 2013; Parkinson’s disease and essential tremor. Mov
28. Biousse V, Skibell BC, Watts RL, Loupe DN, Drews- Disord Off J Mov Disord Soc. October 30,
Botsch C, Newman NJ. Ophthalmologic features of 2010;25(14):2461–2.
Parkinson’s disease. Neurology January 27, 41. Hajee ME, March WF, Lazzaro DR, Wolintz AH,
2004;62(2):177–80. Shrier EM, Glazman S, et al. Inner retinal layer thin-
29. Archibald NK, Clarke MP, Mosimann UP, Burn DJ. ning in Parkinson disease. Arch Ophthalmol. June
The retina in Parkinson’s disease. Brain J Neurol. May 2009;127(6):737–41.
2009;132(Pt 5):1128–45. 42. Aaker GD, Myung JS, Ehrlich JR, Mohammed M,
30. Bodis-Wollner I, Marx MS, Mitra S, Bobak P, Mylin Henchcliffe C, Kiss S. Detection of retinal changes in
L, Yahr M. Visual dysfunction in Parkinson’s Parkinson’s disease with spectral-domain optical 141
 Chapter 11: Optical coherence tomography in neurodegenerative and other neurologic diseases

coherence tomography. Clin Ophthalmol Auckl NZ 53. Jack CR Jr, Knopman DS, Jagust WJ, Petersen RC,
2010;4:1427–32. Weiner MW, Aisen PS, et al. Tracking pathophysio-
43. Shrier EM, Adam CR, Spund B, Glazman S, Bodis- logical processes in Alzheimer’s disease: an
Wollner I. Interocular asymmetry of foveal thickness updated hypothetical model of dynamic
in Parkinson disease. J Ophthalmol biomarkers. Lancet Neurol February
2012;2012:728457. 2013;12(2):207–16.
44. Archibald NK, Clarke MP, Mosimann UP, Burn 54. Koronyo-Hamaoui M, Koronyo Y, Ljubimov AV,
DJ. Retinal thickness in Parkinson’s Miller CA, Ko MK, Black KL, et al. Identification of
disease. Parkinsonism Relat Disord July 2011;17 amyloid plaques in retinas from Alzheimer’s
(6):431–6. patients and noninvasive in vivo optical
imaging of retinal plaques in a mouse
45. Albrecht P, Müller A-K, Südmeyer M, Ferrea S,
model. NeuroImage January 2011;54 Suppl 1:S204–217.
Ringelstein M, Cohn E, et al. Optical coherence
tomography in parkinsonian syndromes. PloS One 55. Cronin-Golomb A, Rizzo JF, Corkin S, Growdon
2012;7(4):e34891. JH. Visual function in Alzheimer’s disease and nor-
mal aging. Ann N Y Acad Sci 1991;640:28–35.
46. Garcia-Martin E, Satue M, Fuertes I, Otin
S, Alarcia R, Herrero R, et al. Ability and 56. Rizzo JF 3rd, Cronin-Golomb A, Growdon JH,
reproducibility of Fourier-domain optical Corkin S, Rosen TJ, Sandberg MA, et al.
coherence tomography to detect retinal nerve Retinocalcarine function in Alzheimer’s disease. A
fiber layer atrophy in clinical and electrophysiological study. Arch Neurol
Parkinson’s disease. Ophthalmology October January 1992;49(1):93–101.
2012;119(10):2161–7. 57. Hinton DR, Sadun AA, Blanks JC, Miller CA. Optic-
47. Roth NM, Saidha S, Zimmermann H, Brandt AU, nerve degeneration in Alzheimer’s disease. N Engl J
Isensee J, Benkhellouf-Rutkowska A, et al. Med August 21, 1986;315(8):485–7.
Photoreceptor layer thinning in 58. Blanks JC, Torigoe Y, Hinton DR, Blanks RH. Retinal
idiopathic Parkinson’s disease. Mov Disord April 1, pathology in Alzheimer’s disease. I. Ganglion cell loss
2014;n/a–n/a. in foveal/parafoveal retina. Neurobiol Aging June
48. Müller T, Woitalla D, Peters S, Kohla K, Przuntek H. 1996;17(3):377–84.
Progress of visual dysfunction in Parkinson’s 59. Sadun AA, Bassi CJ. Optic nerve damage in
disease. Acta Neurol Scand April Alzheimer’s disease. Ophthalmology January
2002;105(4):256–60. 1990;97(1):9–17.
49. Saidha S, Syc SB, Durbin MK, Eckstein C, Oakley JD, 60. Curcio CA, Drucker DN. Retinal ganglion cells in
Meyer SA, et al. Visual dysfunction in multiple Alzheimer’s disease and aging. Ann Neurol March
sclerosis correlates better with optical coherence 1993;33(3):248–57.
tomography derived estimates of macular ganglion 61. Davies DC, McCoubrie P, McDonald B, Jobst KA.
cell layer thickness than peripapillary retinal Myelinated axon number in the optic nerve is unaf-
nerve fiber layer thickness. Mult Scler J [Internet] fected by Alzheimer’s disease. Br J Ophthalmol June
August 24, 2011 [cited September 12, 1995;79(6):596–600.
2011]; Available from: http://www.ncbi.nlm.nih.gov/
pubmed/21865411 62. Justino L, Kergoat M, Bergman H, Chertkow H,
Robillard A, Kergoat H. Neuroretinal function
50. Nguyen-Legros J. Functional neuroarchitecture is normal in early dementia of the
of the retina: hypothesis on the dysfunction of Alzheimer type. Neurobiol Aging August
retinal dopaminergic circuitry in Parkinson’s 2001;22(4):691–5.
disease. Surg Radiol Anat SRA 63. Danesh-Meyer HV, Birch H, Ku JY-F, Carroll S,
1988;10(2):137–44. Gamble G. Reduction of optic nerve fibers in
51. Wulle I, Wagner HJ. GABA and tyrosine patients with Alzheimer disease identified by
hydroxylase immunocytochemistry reveal different laser imaging. Neurology November 28, 2006;67
patterns of colocalization in retinal neurons of (10):1852–4.
various vertebrates. J Comp Neurol June 1, 64. Kergoat H, Kergoat MJ, Justino L, Robillard A,
1990;296(1):173–8. Bergman H, Chertkow H. Normal optic nerve
52. Wu Z, Huang J, Dustin L, Sadda SR. Signal strength is head topography in the early stages of dementia of
an important determinant of accuracy of nerve the Alzheimer type. Dement Geriatr Cogn Disord
fiber layer thickness measurement by optical December 2001;12(6):359–63.
coherence tomography. J Glaucoma March 65. Kergoat H, Kergoat MJ, Justino L, Chertkow H,
142 2009;18(3):213–6. Robillard A, Bergman H. An evaluation of the
 Chapter 11: Optical coherence tomography in neurodegenerative and other neurologic diseases

retinal nerve fiber layer thickness by scanning laser reported cases. Nat Rev Neurol June
polarimetry in individuals with dementia of the 2013;9(6):307–16.
Alzheimer type. Acta Ophthalmol Scand April 77. Brandt AU, Zimmermann H, Kaufhold F,
2001;79(2):187–91. Promesberger J, Schippling S, Finis D, et al. Patterns of
66. Kergoat H, Kergoat M-J, Justino L, Chertkow H, retinal damage facilitate differential diagnosis between
Robillard A, Bergman H. Visual retinocortical func- Susac syndrome and MS. PloS One 2012;7(6):e38741.
tion in dementia of the Alzheimer type. Gerontology 78. Scott CJ, Kardon RH, Lee AG, Frisén L, Wall M.
August 2002;48(4):197–203. Diagnosis and grading of papilledema in patients with
67. Iseri PK, Altinaş O, Tokay T, Yüksel N. Relationship raised intracranial pressure using optical coherence
between cognitive impairment and retinal tomography vs clinical expert assessment using a
morphological and visual functional abnormalities clinical staging scale. Arch Ophthalmol June
in Alzheimer disease. J Neuro-Ophthalmol Off J 2010;128(6):705–11.
North Am Neuro-Ophthalmol Soc March 79. Kaufhold F, Kadas EM, Schmidt C, Kunte H,
2006;26(1):18–24. Hoffmann J, Zimmermann H, et al. Optic nerve head
68. Paquet C, Boissonnot M, Roger F, Dighiero P, Gil R, quantification in idiopathic intracranial
Hugon J. Abnormal retinal thickness in patients hypertension by spectral domain OCT. PloS One
with mild cognitive impairment and Alzheimer’s 2012;7(5):e36965.
disease. Neurosci Lett June 2007;420(2):97–9. 80. Wang J-K, Kardon RH, Kupersmith MJ, Garvin MK.
69. Lu Y, Li Z, Zhang X, Ming B, Jia J, Wang R, et al. Automated quantification of volumetric optic disc
Retinal nerve fiber layer structure abnormalities in swelling in papilledema using spectral-domain opti-
early Alzheimer’s disease: evidence in optical coher- cal coherence tomography. Invest Ophthalmol Vis Sci
ence tomography. Neurosci Lett August 9, June 2012;53(7):4069–75.
2010;480(1):69–72. 81. Carelli V, Ross-Cisneros FN, Sadun AA. Optic nerve
70. Berisha F, Feke GT, Trempe CL, McMeel JW, degeneration and mitochondrial dysfunction: genetic
Schepens CL. Retinal abnormalities in early and acquired optic neuropathies. Neurochem Int May
Alzheimer’s disease. Invest Ophthalmol Vis Sci May 2002;40(6):573–84.
2007;48(5):2285–9. 82. Carelli V, Ross-Cisneros FN, Sadun AA.
71. Kirbas S, Turkyilmaz K, Anlar O, Tufekci A, Durmus Mitochondrial dysfunction as a cause of optic neu-
M. Retinal nerve fiber layer thickness in patients ropathies. Prog Retin Eye Res January
with Alzheimer disease. J Neuro-Ophthalmol Off 2004;23(1):53–89.
J North Am Neuro-Ophthalmol Soc March 83. Harding AE. Friedreich’s ataxia: a clinical and genetic
2013;33(1):58–61. study of 90 families with an analysis of early diag-
72. Bayer AU, Ferrari F, Erb C. High occurrence rate of nostic criteria and intrafamilial clustering of
glaucoma among patients with Alzheimer’s disease. clinical features. Brain J Neurol September
Eur Neurol 2002;47(3):165–8. 1981;104(3):589–620.
73. Helmer C, Malet F, Rougier M-B, Schweitzer C, Colin 84. Fortuna F, Barboni P, Liguori R, Valentino ML,
J, Delyfer M-N, et al. Is there a link between open- Savini G, Gellera C, et al. Visual system
angle glaucoma and dementia? The 3C-alienor involvement in patients with Friedreich’s
cohort. Ann Neurol May 18, 2013; ataxia. Brain J Neurol January
74. Bock M, Brandt AU, Dörr J, Kraft H, Weinges-Evers 2009;132(Pt 1):116–23.
N, Gaede G, et al. Patterns of retinal nerve fiber 85. Alldredge CD, Schlieve CR, Miller NR, Levin LA.
layer loss in multiple sclerosis patients with or Pathophysiology of the optic neuropathy associated
without optic neuritis and glaucoma patients. with Friedreich ataxia. Arch Ophthalmol November
Clin Neurol Neurosurg October 2003;121(11):1582–5.
2010;112(8):647–52. 86. Noval S, Contreras I, Sanz-Gallego I, Manrique RK,
75. Dörr J1, Radbruch H, Bock M, Wuerfel J, Arpa J. Ophthalmic features of Friedreich
Brüggemann A, Wandinger KP, Zeise D, Pfueller CF, ataxia. Eye Lond Engl February 2012;26(2):315–20.
Zipp F, Paul F. Encephalopathy, visual disturbance 87. Seyer LA, Galetta K, Wilson J, Sakai R, Perlman S,
and hearing loss-recognizing the symptoms of Mathews K, et al. Analysis of the visual
Susac syndrome. Nat Rev Neurol December system in Friedreich ataxia. J Neurol June
2009;5(12):683–8. doi: 10.1038/nrneurol.2009.176. 18, 2013;
76. Dörr J, Krautwald S, Wildemann B, Jarius S, 88. Klockgether T. The clinical diagnosis of autosomal
Ringelstein M, Duning T, et al. Characteristics dominant spinocerebellar ataxias. Cerebellum Lond
of Susac syndrome: a review of all Engl 2008;7(2):101–5. 143
 Chapter 11: Optical coherence tomography in neurodegenerative and other neurologic diseases

89. Stricker S, Oberwahrenbrock T, Zimmermann H, 101. Roth NM, Saidha S, Zimmermann H, Brandt AU,
Schroeter J, Endres M, Brandt AU, et al. Temporal Oberwahrenbrock T, Maragakis NJ, et al. Optical
retinal nerve fiber loss in patients with coherence tomography does not support optic nerve
spinocerebellar ataxia Type 1. PLoS ONE July 29, involvement in amyotrophic lateral sclerosis. Eur J
2011;6(7):e23024. Neurol Off J Eur Fed Neurol Soc April 14, 2013;
90. Pula JH, Towle VL, Staszak VM, Cao D, Bernard JT, 102. Ala A, Walker AP, Ashkan K, Dooley JS, Schilsky ML.
Gomez CM. Retinal nerve fibre layer and macular Wilson’s disease. Lancet February 3, 2007;369
thinning in spinocerebellar ataxia and cerebellar (9559):397–408.
multisystem atrophy. Neuro-Ophthalmol Aeolus Press 103. Satishchandra P, Ravishankar Naik K. Visual pathway
Jun 2011;35(3):108–14. abnormalities Wilson’s disease: an electrophysiologi-
91. Manrique RK, Noval S, Aguilar-Amat MJ, Arpa J, cal study using electroretinography and visual
Rosa I, Contreras I. Ophthalmic features of spino- evoked potentials. J Neurol Sci May 1,
cerebellar ataxia type 7. J Neuro-Ophthalmol Off J 2000;176(1):13–20.
North Am Neuro-Ophthalmol Soc September 104. Albrecht P, Müller A-K, Ringelstein M, Finis D,
2009;29(3):174–9. Geerling G, Cohn E, et al. Retinal neurodegeneration
92. Hugosson T, Gränse L, Ponjavic V, Andréasson S. in Wilson’s disease revealed by spectral domain
Macular dysfunction and morphology in spinocere- optical coherence tomography. PloS One 2012;7(11):
bellar ataxia type 7 (SCA 7). Ophthalmic Genet. March e49825.
2009;30(1):1–6. 105. Parikh RS, Parikh SR, Sekhar GC, Prabakaran S, Babu
93. Miller RC, Tewari A, Miller JA, Garbern J, Van JG, Thomas R. Normal Age-Related Decay of Retinal
Stavern GP. Neuro-ophthalmologic features of Nerve Fiber Layer Thickness. Ophthalmology May
spinocerebellar ataxia type 7. J Neuro-Ophthalmol Off 2007;114(5):921–6.
J North Am Neuro-Ophthalmol Soc September 106. Alamouti B, Funk J. Retinal thickness decreases with
2009;29(3):180–6. age: an OCT study. Br J Ophthalmol July
94. Bouchard JP, Barbeau A, Bouchard R, Bouchard RW. 2003;87(7):899–901.
Autosomal recessive spastic ataxia of Charlevoix- 107. Ooto S, Hangai M, Tomidokoro A, Saito H, Araie M,
Saguenay. Can J Neurol Sci J Can Sci Neurol February Otani T, et al. Effects of age, sex, and axial length on
1978;5(1):61–9. the three-dimensional profile of normal macular layer
95. Vingolo EM, Di Fabio R, Salvatore S, Grieco G, structures. Invest Ophthalmol Vis Sci
Bertini E, Leuzzi V, et al. Myelinated retinal fibers in 2011;52(12):8769–79.
autosomal recessive spastic ataxia of Charlevoix- 108. Rudnicka AR, Mt-Isa S, Owen CG, Cook DG, Ashby
Saguenay. Eur J Neurol Off J Eur Fed Neurol Soc D. Variations in primary open-angle glaucoma pre-
September 2011;18(9):1187–90. valence by age, gender, and race: a Bayesian meta-
96. Desserre J, Devos D, Sautière BG, Debruyne P, analysis. Invest Ophthalmol Vis Sci October
Santorelli FM, Vuillaume I, et al. Thickening of 2006;47(10):4254–61.
peripapillar retinal fibers for the diagnosis of autoso- 109. Quigley HA, Broman AT. The number of people
mal recessive spastic ataxia of Charlevoix-Saguenay. with glaucoma worldwide in 2010 and 2020. Br J
Cerebellum Lond Engl December 2011;10(4):758–62. Ophthalmol March 2006;90(3):262–7.
97. Garcia-Martin E, Pablo LE, Gazulla J, Polo V, 110. Varma R, Ying-Lai M, Francis BA, Nguyen BB-T,
Ferreras A, Larrosa JM. Retinal nerve fibre layer Deneen J, Wilson MR, et al. Prevalence of open-
thickness in ARSACS: myelination or hypertrophy? angle glaucoma and ocular hypertension in
Br J Ophthalmol February 2013;97(2):238–41. Latinos: the Los Angeles Latino Eye
98. Blackstone C, O’Kane CJ, Reid E. Hereditary spastic Study. Ophthalmology August
paraplegias: membrane traffic and the motor path- 2004;111(8):1439–48.
way. Nat Rev Neurosci January 2011;12(1):31–42. 111. Tewarie P, Balk L, Costello F, Green A, Martin R,
99. Carelli V, La Morgia C, Valentino ML, Barboni P, Schippling S, et al. The OSCAR-IB Consensus
Ross-Cisneros FN, Sadun AA. Retinal ganglion cell Criteria for Retinal OCT Quality
neurodegeneration in mitochondrial inherited Assessment. PLoS ONE April 19,
disorders. Biochim Biophys Acta May 2012;7(4):e34823.
2009;1787(5):518–28. 112. Petzold A, Wattjes MP, Costello F, Flores-Rivera J,
100. Wiethoff S, Zhour A, Schöls L, Fischer MD. Retinal Fraser CL, Fujihara K, et al. The investigation of acute
nerve fibre layer loss in hereditary spastic paraplegias optic neuritis: a review and proposed protocol. Nat
is restricted to complex phenotypes. BMC Neurol Rev Neurol August 2014;10(8):447–58. doi: 10.1038/
144 2012;12:143. nrneurol.2014.108. Epub Jul 8, 2014.
 Chapter
Optical coherence tomography pathologies

12 to know about in clinical practice

Scott D. Newsome and John N. Ratchford

Introduction Epiretinal membrane

In neurology, OCT has primarily been used as a Epiretinal membranes (ERM), also known as macular
research tool to quantitatively assess the peripapillary pucker, typically occur as a response to changes within
retinal nerve fiber layer (RNFL) and macular volume the vitreous humor and proliferation of glial tissue
in patients with inflammatory optic neuropathies and within the RNFL [1]. The underlying retinal structures
demyelinating disorders. However, more recently are affected by traction forces from ERM and appear
OCT is being used in neurology clinical practices to thicker than the normal fellow eyes [2]. Macular
help aid with specific neurological diagnoses (i.e., degeneration usually does not occur with ERM because
optic neuritis vs. macular disease) and determine to photoreceptor, rod, and cone cells are not damaged.
what extent an individual’s anterior visual pathway is However, in rare instances the ERM can become very
structurally compromised (as in multiple sclerosis or thick and hard, resulting in damage to these cells.
neuromyelitis optica). Neurologists utilizing OCT for On OCT, an ERM can be identified as a thin
research and clinical purposes should be aware of the hyperreflective band that lies between the vitreous
more common retinal abnormalities that may be and retina (Figure 12.1). This is one of the more
detected, many of which will be reviewed in this common findings on OCT, especially in older popula-
chapter. These incidental findings are important to tions. Cystic-like gaps between the RNFL and inner
recognize since detailed ophthalmologic evaluation limiting membrane (ILM) often occur with ERM
may be warranted to further characterize these (Figure 12.2) [1]. In addition, ERM often accompa-
abnormalities; this differential of optic nerve vs. ret- nies a posterior vitreous detachment.
inal disease is one of the most important potential Common predisposing factors for ERM include
uses of OCT in neurological practice. aging and diabetes. Even though ERM does not

Figure 12.1 OCT image of a patient with a simple epiretinal membrane (arrow).

Optical Coherence Tomography in Neurological Disease, ed. Peter A. Calabresi, Laura J. Balcer, and Elliot M. Frohman. 145
Published by Cambridge University Press. © Cambridge University Press 2015.
 Chapter 12: Optical coherence tomography pathologies to know about in clinical practice

Figure 12.2 OCT image of a patient with an epiretinal membrane (yellow arrow) with cystic gaps (red arrow) and vitreomacular traction.

Figure 12.3 OCT images of a patient with retinal drusen

(arrow).

threaten vision loss, in severe cases it can cause progressing more rapidly to advanced age-related
a decrease in visual acuity and metamorphopsia macular degeneration (AMD) and is an early clinical
(i.e., distorted vision). Surgical intervention is usually manifestation of AMD [3–5]. Age-related drusen are
not warranted in ERM unless visual symptoms are distinct from optic disc drusen, which result in calcific
impeding individuals’ daily activities. degeneration in some of the axons of the optic nerve.
Both of these entities can be seen via ophthalmoscopy
Drusen and visualized with OCT.
Drusen typically occur in eyes due to accumulations On OCT, retinal drusen are visualized as hyperre-
of extracellular substances between Bruch’s mem- flective protuberances underneath the RPE layer
brane and the retinal pigment epithelium (RPE) [3]. (Figure 12.3). The retinal tissue can be displaced if
It is suspected that drusen decrease oxygen and nutri- the drusen are large enough. Optic disc drusen can
ents to the RPE and photoreceptor cells, which in turn displace the RNFL, which can be associated with a
leads to the pathologies discussed in the paragraph thickening of the disc tissue (Figure 12.4).
below. Most patients with drusen are asymptomatic;
Drusen most commonly occur with aging and, however, it can mimic symptoms of papilledema
when severe, are associated with macular degenera- and is associated with AMD [6, 7]. Optic disc drusen
tion. Rarely, drusen are related to dominant familial can ultimately lead to loss of visual field due to
drusen, which is a comparatively benign condition. displacement of the nerve fiber layers. OCT is a
In contrast, the presence of large, soft, and numerous useful tool to monitor the progression of drusen
146 drusen appear to increase an individual’s risk for over time.
 Chapter 12: Optical coherence tomography pathologies to know about in clinical practice

Figure 12.4 OCT images of a patient with optic disc

drusen (arrow).

Central serous chorioretinopathy resolution of the subretinal fluid collection, visual

acuity can improve almost back to patients’ pre-CSR
Central serous chorioretinopathy (CSR) develops
baseline, although some individuals will have perma-
when a leakage of fluid occurs in the subretinal
nent impaired night vision, color desaturation, meta-
space or in conjunction with a pigment epithelium
morphosis, and poor contrast sensitivity [11] CSR can
detachment (PED). This fluid leakage is thought to
reoccur, resulting in progressive vision loss, and can
develop due to dysfunction of the RPE and/or
be associated with subretinal neovascularization and
choroid. Fluid accumulation can involve the central
pigment epitheliopathy.
macula which can ultimately lead to vision loss that
CSR treatment should be considered if it has not
mimics acute optic neuritis.
spontaneously resolved in a few months and
CSR is most commonly seen in young and middle-
especially if an individual is still symptomatic. The
aged patients, with a higher proportion of men being
therapy chosen depends on whether there is fluid
affected. The majority of CSR cases are idiopathic in
leakage into the central macula. Laser photocoagula-
nature; however, there are several potential triggering
tion is done if the CSR is distant from the central
factors. Specifically, previous studies have noted that
macula because of risk for causing permanent central
stress, Cushing’s syndrome, corticosteroid therapy,
vision loss with this treatment. In CSR cases involving
and anti-anxiety medications are independent risk
or within close proximity of the central macula, trans-
factors for developing CSR [8–10].
pupillary thermotherapy is preferred [12]. There are
CSR is best detected by OCT and fluorescein
other treatments that demonstrate promise with
angiography. On OCT, CSR is identified as subretinal
minimal complications that are becoming more pop-
clear fluid accumulation in the macular region, which
ular [13].
leads to elevation of the retinal layers above
(Figure 12.5). If the RPE is intact then there is a
hyperreflective band at the base of the fluid cavity Pigment epithelium detachment
(Figure 12.5). However, if the RPE is detached then Pigment epithelium detachment (PED) occurs when
a hyperreflective band is seen over the fluid cavity there is disruption of the normal junction between
(Figure 12.6). the basement membrane of the RPE and Bruch’s
The majority of patients with CSR will experience membrane. Subretinal fluid accumulates from the
some form of visual disturbance. It can be acute in choriocapillaries, resulting in PED. This fluid collec-
onset with a visual acuity as severe as 20/200 with tion and the resultant separation of membranes
accompanying central and paracentral relative can also occur in conjunction with a CSR and retinal
scotomas. An urgent evaluation is needed to rule out detachment.
retinal detachment, which requires a detailed ophthal- PEDs are most commonly seen in males between
mological evaluation, including OCT. However, upon the ages of 20 and 60 years. There are several conditions 147
 Chapter 12: Optical coherence tomography pathologies to know about in clinical practice

Figure 12.5 OCT image of a patient with central serous chorioretinopathy.

Figure 12.6 OCT image of a patient with central serous chorioretinopathy (red arrow) with retinal pigment epithelium detachment
(yellow arrow).

that predispose individuals to developing PEDs, the fovea is not involved. If the fovea is involved
including AMD, choroidal tumors, high myopia, patients may report blurred vision, visual distortion,
choroidal neovascular membranes (CNM), ocular micropsia, or metamorphopsia [14]. In older patients
histoplasmosis, and hereditary choroidal degenera- with PED that does not have accompanying CNM,
tion [14]. In some cases, PEDs are idiopathic in close monitoring is required because these mem-
nature. However, in clinical practice PEDs are branes can develop, resulting in significant visual
mostly seen in association with eyes that are treated impairment, including loss of central vision and less
for neovascular AMD [15]. than 20/200 visual acuity [16]. Older patients with
PEDs can be seen ophthalmoscopically in the PEDs that are associated with CNM have a poor visual
posterior fundus, although they are best visualized prognosis; therefore, focal laser photocoagulation
with fluorescein angiography and OCT. On OCT, therapy is warranted in these patients. Serious retinal
PEDs appear as a dome-shaped elevation with a detachments will often occur concurrently with PEDs
hyperreflective red band overlying the dome which requires treating the underlying precipitating
(Figure 12.7). The PED causes shadowing of the chor- disease process.
oid below, which is not seen in CSR without accom-
panying PED. Macular edema
The majority of patients are < 55 years old with Macular edema occurs when deposits of protein and
PED present with small detachments that are not fluid accumulate within or under the macula. The
associated with other chorio-retinal disease processes. macula is located near the center of the retina, where
In these cases, the prognosis is excellent and the the cones are located, which become affected with
148 patient requires no specific intervention, especially if macular edema.
 Chapter 12: Optical coherence tomography pathologies to know about in clinical practice

Figure 12.7 OCT image of a patient with pigment epithelium detachment.

Figure 12.8 OCT image of a patient with macular

edema.

There are two main categories of macular edema: of macular edema can manifest as just blurred vision or
cystoid macular edema and diabetic macular result in severe central vision loss. Colors can also
edema. Both types appear to occur due to abnormal appear altered or washed out. Peripheral vision is
retinal/macular capillary permeability. There are spared with macular edema, which in combination
several causes of cystoid macular edema, including with central vision loss and changes in color can be
diabetes, pars planitis, specific medications (i.e., fin- mistaken as acute optic neuritis.
golimod), post-cataract or retinal detachment sur- The treatment for macular edema varies depend-
gery, retinitis pigmentosa, and venous occlusion, ing on the cause and severity of the condition. The
among other causes. Diabetic macular edema leads primary goal of treating macular edema in diabetic
to the most common cause of vision loss in diabetic patients is to stabilize their vision with focal laser
retinopathy. treatment. This treatment seals off the leaky blood
Macular edema is well visualized on OCT. On OCT, vessels that are causing disruption in macular func-
macular edema results in thickening of the retina and tion. More recently, novel medications, including
loss of the normal foveal contour, which is best seen on anti-VEGF, are being used to help slow the leakage
the macular tomograms (Figure 12.8). Continued and growth of abnormal blood vessels [17]. Cystoid
intraretinal fluid accumulation eventually leads to the macular edema may need nonsteroidal anti-
visible cystic spaces in cystic macular edema. inflammatory or steroid eye drops and, in more
The presence of macular edema from any cause can refractory cases, intraocular steroid injections, vitrect-
result in central vision distortion. The initial symptoms omy, or lens replacement. 149
 Chapter 12: Optical coherence tomography pathologies to know about in clinical practice

hypertensive retinopathy, Coats’ disease, optic nerve

pit, retinal coloboma, and toxemia of pregnancy [18].
A serous retinal detachment can be asymptomatic until
it involves the macula. Symptoms include photopsias,
floaters, a curtain across the visual field, or decreased
vision. OCT will show a fluid collection between the
neurosensory retina and the RPE without a break in the
retina (Figure 12.10). This may manifest as an area of
elevated macular thickness when the macula is involved.
Treatment is focused on the underlying cause, and
surgery is usually not needed.
A traction retinal detachment occurs when a
fibrotic proliferation causes the retina to pull off of
the RPE without evidence of a retinal break.
Predisposing etiologies include diabetic retinopa-
thy, sickle cell retinopathy, retinopathy of prema-
turity, venous occlusions, proliferative
vitreoretinopathy, toxocariasis, and familial exuda-
tive vitreoretinopathy.[18] When a traction retinal
detachment involves the macula, symptoms are
similar to other types of retinal detachment.
Figure 12.9 OCT images of the macula in two patients with
rhegmatogenous retinal detachments. Vitreous fluid is present in the Though often managed conservatively, surgery is
subretinal space [30]. considered if it threatens the macula.

Posterior vitreous detachment and

Retinal detachment vitreomacular traction syndrome
Retinal detachment occurs when the neurosensory
A posterior vitreous detachment occurs when the vitr-
retina separates from the retinal pigment epithelium
eous separates from the retina throughout
(RPE). This separation creates a potential space where
the peripheral fundus but remains adherent to it
subretinal fluid accumulates. There are several types of
posteriorly. This can produce traction on the region
retinal detachment: rhegmatogenous, serous, and
containing the macula and optic nerve. This
traction.
syndrome is called vitreomacular traction syndrome.
A rhegmatogenous retinal detachment occurs
Vitreomacular traction is believed to contribute to
when a full thickness retinal break allows vitreous
multiple macular diseases, including cystoid macular
fluid to access the subretinal space. It will typically
edema and macular holes [19]. Vitreomacular traction
present with photopsias, floaters, a curtain across the
syndrome is almost invariably accompanied by an
visual field, or decreased visual acuity, or it could rarely
epiretinal membrane, suggesting that the membrane
be asymptomatic. OCT will show a hyporeflective dark
contributes to the pathogenesis (Figure 12.11) [20].
space under the neurosensory retina (Figure 12.9). A
The individual OCT findings of the various man-
rhegmatogenous retinal detachment is treated with
ifestations of vitreomacular traction syndrome are
urgent surgery (within 24 hours if central vision is
described individually below. OCT is used by ophthal-
threatened), so if there is a suspicion of this, urgent
mologists to evaluate the macula in detail in order to
referral to a retina specialist is recommended.
determine the type of maculopathy in vitreomacular
A serous retinal detachment is caused by subretinal
traction syndrome.
transudation of fluid without a retinal break. Fluid
can be from an inflammatory condition, vascular lesion,
tumor, or degenerative process. Predisposing conditions Macular hole
include Vogt–Koyanagi–Harada syndrome, Harada’s A macular hole is a full-thickness hole in the retina
disease, idiopathic uveal effusion syndrome, choroidal that involves the fovea. It can be distinguished clini-
150 tumors, central serous retinopathy, posterior scleritis, cally from a macular pseudohole or lamellar hole by
 Chapter 12: Optical coherence tomography pathologies to know about in clinical practice

Macular pseudoholes
A macular pseudohole can resemble a full-thickness
macular hole clinically, but it does not have loss of
foveal tissue (Figure 12.14). The pathogenesis is
different, because these lesions have been attributed
to the centripetal contraction of a surrounding
epiretinal membrane. Since they appear similar to
lamellar macular holes on biomicroscopy, OCT can
help distinguish the two entities. Vision typically
remains relatively good with macular pseudoholes.
On OCT, pseudoholes are found to have a
steepened foveal pit with thickened edges contracted
by an epiretinal membrane, reduced foveal pit
diameter, and normal or slightly increased central
foveal thickness [23]. Most pseudoholes are followed
conservatively [24].

Lamellar macular hole

The term lamellar macular hole was first used to
describe a macular lesion resulting from the opening
of the central cyst of a cystoid macular edema [25].
This term is also used to describe a macular lesion that
Figure 12.10 OCT image of a patient with serous retinal results from the same process that leads to a full-
detachment
thickness macular hole when the process does not
go to completion. By definition, a lamellar macular
hole is a partial thickness macular hole where the
the presence of significant central vision loss and by inner layers of the fovea are detached from the under-
other testing methods such the Watzke-Allen and lying cellular layers of the fovea (Figure 12.15).
laser aiming beam tests. Patients with lamellar holes have relatively well-
Macular holes are often idiopathic, but risk factors preserved vision, and the macula contains a round,
include cystoid macular edema, vitreomacular well-circumscribed reddish lesion [26].
traction, inflammation, trauma, ocular surgery, and Proposed OCT criteria for diagnosis of a lamellar
laser treatment [18]. Presenting symptoms typically hole include an irregular foveal contour, a break in the
include a sudden-onset metamorphopsia (i.e., abnormal inner fovea separation of the inner from the outer foveal
Amsler grid test), decreased vision, and, less commonly, retinal layers leading to an intraretinal split, and absence
central scotoma. of a full-thickness foveal defect with intact photorecep-
An isolated foveal cyst can be seen as an early tors posterior to the area of foveal dehiscence [26]. OCT
stage of the development of a macular hole will show a high central foveal thickness. Epiretinal
(Figure 12.12). With a macular hole, OCT will typi- membranes are present in 50–89% of lamellar macular
cally show a full-thickness retinal defect involving holes [26, 27], suggesting that they may be involved in
the fovea on a macular scan (Figure 12.13). the pathogenesis of some lamellar macular holes. Many
Treatment depends on the severity of the hole and patients are followed conservatively, but surgery with
visual defect. An impending hole is followed con- vitrectomy and epiretinal membrane peeling has also
servatively, but more significant macular holes are been used with success [28, 29].
typically treated by a retina specialist with vitrect-
omy, detachment of the posterior vitreous cortex, Conclusion
peeling of any epiretinal membrane around the OCT is becoming more widely used by neurologists in
hole, gas–fluid exchange, and postoperative prone clinical practice for the evaluation of suspected or
positioning [21, 22]. established neurological disease. This technology is 151
 Chapter 12: Optical coherence tomography pathologies to know about in clinical practice

Figure 12.11 Two examples of vitreomacular traction with associated epiretinal membranes. [31, 32]

Figure 12.12 OCT image of a patient with a foveal cyst. [31]

generating great interest in clinical practice because it is It is crucial for neurologists using OCT to be aware
noninvasive, relatively inexpensive, and quickly of the common qualitative retinal abnormalities that
acquires quantitative information that has structure- can be detected with these devices. More importantly,
152 function correlations. neurologists need to know which incidental findings
 Chapter 12: Optical coherence tomography pathologies to know about in clinical practice

Figure 12.13 OCT image of a patient with a macular hole. [32]

Figure 12.14 OCT image of a patient with a macular pseudohole with a steepened foveal contour. [32]

153
Figure 12.15 OCT image of two examples of lamellar macular holes. [31, 32]
 Chapter 12: Optical coherence tomography pathologies to know about in clinical practice

Table 12.1 Common retinal abnormalities needing 6. Sahin A, Cingü AK, Ari S, et al. Bilateral optic disc
ophthalmological/retinal referral drusen mimicking papilledema. J Clin Neurol 2012;
8(2): 151–4.
Complicated epiretinal membrane
7. Kinori M, Moroz I, Zolf R, et al. Pseudopapilledema–
optic disc drusen. Harefuah 2013; 152(3): 154–7.
Central serous chorioretinopathy
8. Carvalho-Recchia C, Yannuzzi L, Negra˜o S, et al.
Pigment epithelium detachment Corticosteroids and central serous chorioretinopathy.
Ophthalmology 2002; 109: 1834–7.
Macular edema 9. Tsai D-C, Chen S-J, Huang C-C, et al. Epidemiology of
idiopathic central serous chorioretinopathy in Taiwan,
Retinal detachment 2001–2006: A population-based study. PLoS ONE
2013; 8(6): e66858.
Macular hole 10. Haimovici R, Koh S, Gagnon D, et al. Risk factors for
central serous chorioretinopathy: a case–control study.
Macular pseudohole
Ophthalmology 2004; 111: 244–9.
Lamellar macular hole 11. Baran N, Gürlü V, and Esgin H. Long-term macular
function in eyes with central serous chorioretinopathy.
Any incidentally identified condition that may be Clinical and Experimental Ophthalmology 2005; 33:
associated with visual dysfunction 369–72.
12. Wei SY and Yang CM. Transpupillary thermotherapy
in the treatment of central serous chorioretinopathy.
Ophthalmic surgery, lasers & imaging 2005; 36(5):
412–5.
need appropriate referral to an ophthalmologist and/ 13. Karakus SH, Basarir B, Pinarci EY, et al. Long-term
or retinal specialist (Table 12.1). Independent of these results of half-dose photodynamic therapy for chronic
referrals, it is typically also recommended for patients central serous chorioretinopathy with contrast sensi-
with diseases that can affect the anterior visual path- tivity changes. Eye 2013; 27: 612–20.
way to undergo yearly ophthalmological evaluation as 14. Zayit-Soudry S, Moroz I, and Loewenstein A. Retinal
part of their general care. Moreover, thorough review pigment epithelial detachment. Surv Ophthalmol 2007;
of research OCT scans is vitally important since unex- 52: 227–43.
pected retinal pathologies can impact accurate data 15. Nagiel A, Bailey Freund K, et al. Mechanism of retinal
acquisition. pigment epithelium tear formation following intravi-
treal anti–vascular endothelial growth factor therapy
revealed by spectral-domain optical coherence tomo-
References graphy. Am J Ophthalmol 2013 [Epub ahead of print].
1. Reed B, Glabe D, and Yudcovitch L. Ocular Coherence
16. Chang L and Sarraf D. Tears of the retinal pigment
Tomography Guide. Faculty Scholarship 2011; Paper 12.
epithelium. RETINA 2007; 27: 523–4.
2. Koo H, Rhim W, and Lee E. Morphologic and func-
17. Kiire CA, Porta M, and Chong V. Medical management
tional association of retinal layers beneath the epiret-
for the prevention and treatment of diabetic macular
inal membrane with spectral-domain optical
edema. Surv Ophthalmol 2013; 58(5): 459–65.
coherence tomography in eyes without photoreceptor
abnormality. Graefes Arch Clin Exp Ophthalmol 2012; 18. Friedman NJ and Kaiser PK. Essentials of
250: 491–8. Ophthalmology. Philadelphia: Elsevier Inc., 2007.
3. Garcia-Garcia J, Ruiz-Moreno J, Holm K, et al. 19. Shechtman DL and Dunbar MT. The expanding spec-
Macular dysfunction in drusen maculopathy assessed trum of vitreomacular traction. Optometry 2009; 80:
with multifocal electroretinogram and optical coher- 681–7.
ence tomography. Clinical Ophthalmology 2013: 7; 20. Smiddy W. Vitreomacular traction syndrome. In:
1303–9. Yanoff M., Duker J.S., eds. Ophthalmology. Third Ed.
4. Klein R, Klein B, Tomany S, et al. Ten-Year incidence Philadelphia: Elsevier, Inc., 2009.
and progression of age-related maculopathy; The Beaver 21. Kelly NE and Wendel RT. Vitreous surgery for idio-
Dam Eye study. Ophthalmology 2002; 109: 1767–79. pathic macular holes. Results of a pilot study.
5. Klein R, Peto T, Bird A, et al. The Epidemiology of Arch Ophthalmol 1991; 109: 654–9.
Age-Related Macular Degeneration. Am J Ophthalmol 22. Wendel RT, Patel AC, Kelly NE, et al. Vitreous surgery
154 2004; 137: 486–95. for macular holes. Ophthalmology 1993; 100: 1671–6.
 Chapter 12: Optical coherence tomography pathologies to know about in clinical practice

23. Haouchine B, Massin P, Tadayoni R, et al. Diagnosis of retrospective comparison study. Clin Ophthalmol
macular pseudoholes and lamellar macular holes by 2013; 7: 1843–8.
optical coherence tomography. Am J Ophthalmol 2004; 29. Sun JP, Chen SN, Chuang CC, et al. Surgical treatment
138: 732–9. of lamellar macular hole secondary to epiretinal
24. Garcia-Fernandez M, Navarro JC, Sanz AF, et al. Long- membrane. Graefes Arch Clin Exp Ophthalmol 2013
term evolution of idiopathic lamellar macular holes May 17 (Epub).
and macular pseudoholes. Can J Ophthalmol 2012; 47: 30. Huang C, Fu T, Zhang T, et al. Scleral buckling
442–7. versus vitrectomy for macula-off rhegmatogenous
25. Gass JD. Lamellar macular hole: a complication of retinal detachment as accessed with spectral-domain
cystoid macular edema after cataract extraction: a optical coherence tomography: a retrospective
clinicopathologic case report. Trans Am Ophthalmol observational case series. BMC Ophthalmol 2013;
Soc 1975; 73: 231–50. 13: 12.
26. Witkin AJ, Ko TH, Fujimoto JG, et al. Redefining 31. Sotirchos ES, Vidal-Jordana A, Saidha S, et al.
lamellar holes and the vitreomacular interface: an Unexpected retinal abnormalities detected upon
ultrahigh-resolution optical coherence tomography evaluation for neurological disease. Abstract,
study. Ophthalmology 2006; 113: 388–97. presented at the 2013 American Committee for
27. Chen JC and Lee LR. Clinical spectrum of lamellar Treatment and Research in Multiple Sclerosis
macular defects including pseudoholes and pseudo- meeting, p. 47.
cysts defined by optical coherence tomography. 32. Yong SO and Lam WC. Differential diagnosis of
Br J Ophthalmol 2008; 92: 1342–6. macular holes and pseudoholes. Retinal Physician
28. Sanisoglu H, Elbay A, Sevim S, et al. Surgical therapy 2011; 8: 30–4.
versus observation for lamellar macular hole: a

155
 Chapter
Optical coherence tomography and retinal

13 segmentation in neurological diseases

Elias S. Sotirchos and Shiv Saidha

Introduction measured; this process is known as segmentation.

Furthermore, in the case of multiple adjacent cross-
The advent of optical coherence tomography (OCT)
sectional images (i.e., B-scans), such as in the case of
has provided the ability to rapidly and noninvasively
a macular cube scan, three-dimensional reconstruc-
acquire micrometer-resolution cross-sectional or 3-D
tions can be created from the segmentation-derived
images of the retina. OCT has revolutionized the
boundaries by interpolating them to the unsampled
study, diagnosis, and monitoring of retinal disease
areas between B-scans. The quantitative OCT mea-
by providing both qualitative (morphology) and
sures that predominantly have been utilized include
quantitative (thickness and volume) information.
borders that can be easily and reliably detected due
The ability to easily and cost-effectively acquire
to significant differences in the reflectivity of tissues
images of a component of the central nervous system
at their interface; these measures have mainly
is especially attractive for neurologists and has
included the peripapillary retinal nerve fiber layer
spurred significant interest in the potential research
(RNFL) thickness and the average macular thickness
and clinical applications of retinal imaging with OCT
or total macular volume.
in neurological diseases. The main focus of interest in
Advances in segmentation algorithms and the
OCT in neurology has been on quantitative retinal
development of modern, spectral-domain OCT that
measures, since neurological conditions are predomi-
allows for the three-dimensional acquisition of ultra-
nantly thought to be associated with atrophy of retinal
high-resolution images have vastly expanded the
structures rather than morphologic abnormalities,
potential of OCT by enabling the isolation of more
although emerging evidence suggests that qualitative
specific regions of interest, including individual ret-
aberrations may also occur in neurological disease.
inal layers. In contrast to gross quantitative measures,
OCT images are generated by scanning the retina
such as the average macular thickness, this refined
with a beam of near-infrared light and analyzing the
approach generates quantitative information about
backscattered light, in a fashion analogous to ultra-
specific neuronal and axonal subpopulations of the
sonography [1]. Contrast between retinal layers is
retina and provides the possibility of pinpointing spe-
generated by differences in their optical properties,
cific retinal areas and structures that are differentially
which result from their varying histological compo-
affected by disease processes. The study of these mea-
sition. High-reflectivity layers correspond to areas
sures is currently an area of intensive research and
where retinal elements are arranged horizontally,
holds promise for the development of novel biomar-
such as the retinal nerve fiber layer (RNFL) and
kers of neurological disease and for promoting our
plexiform layers, whereas the nuclear layers exhibit
understanding of the pathological mechanisms
relatively lower reflectivity [2]. In order to derive
underlying various neurological diseases.
quantitative measures from OCT scans, post-acqui-
sition processing is required and involves the iden-
tification of boundaries between areas of interest, Retinal layers and cell populations
thus allowing the partitioning of the acquired images The retina is structured in a multilayered fashion and
into specific regions that may subsequently be is traditionally described as being composed of the

156 Optical Coherence Tomography in Neurological Disease, ed. Peter A. Calabresi, Laura J. Balcer, and Elliot M. Frohman.
Published by Cambridge University Press. © Cambridge University Press 2015.
 Chapter 13: Optical coherence tomography and retinal segmentation in neurological diseases

following ten layers (beginning from the inner por- Functionally, the retina may be considered as
tion of the retina, i.e., closest to the vitreous): being composed of three nuclear layers (where the
1. Inner limiting membrane (ILM): This layer con- cell bodies of the retinal neurons are located) that
stitutes the boundary between the retina and the are separated by the two plexiform layers, where the
vitreous and is formed primarily by end-feet of the synaptic connections between the retinal neurons are
Müller cells, which are located in the inner nuclear located. The flow of information occurs in the follow-
layer. ing direction: Photoreceptors –> Bipolar Cells –>
2. Retinal nerve fiber layer (RNFL): This layer Ganglion Cells. Furthermore, retinal interneurons
consists of unmyelinated axons of the ganglion include the horizontal and amacrine cells, which
cells that coalesce at the optic disc to form the modify the signals between the photoreceptors and
optic nerve, which becomes myelinated posterior bipolar cells, and bipolar and ganglion cells,
to the lamina cribrosa. In addition to the nerve respectively.
fibers, fibrous astrocytes are present in the RNFL.
3. Ganglion cell layer (GCL): This layer contains
primarily the cell bodies of the ganglion cells.
Retinal segmentation
4. Inner plexiform layer (IPL): This layer contains The discussion of the technical details of segmen-
the synaptic connections between ganglion cell tation is beyond the scope of this chapter, but in
dendrites, bipolar cell axons, and axons and/or general this process may be performed either
dendrites of amacrine cells. manually by an individual visually identifying
the boundaries between retinal layers and using
5. Inner nuclear layer (INL): This layer comprises
specialized software in order to trace them, or by
the cell bodies of the bipolar cells, horizontal cells,
applying computerized algorithms that automati-
amacrine cells, and Müller cells (primary glial cell
cally detect retinal boundaries. Manual techniques
of the retina).
have been employed in small-scale studies and for
6. Outer plexiform layer (OPL): This layer is the
the validation of automated segmentation algo-
location of synaptic connections between the
rithms, and have provided important information
photoreceptors and the dendrites of the bipolar
in studies of retinal involvement in neurological
cells.
disease. As can be expected, however, manual
7. Outer nuclear layer (ONL): This layer consists of
segmentation is an extremely tedious and time-
the cell bodies of the photoreceptors (rods and consuming process and, thus, cannot be used for
cones). large-scale clinical studies or in everyday clinical
8. External limiting membrane (ELM): This is a practice. Automated segmentation algorithms
line of intercellular junctions between photo- have been commercially available for many years
receptor cells and Müller cells that separates and have primarily focused on the segmentation
the photoreceptor segments from the cell of the RNFL in a ring scan around the optic disc;
bodies. thus, the peripapillary RNFL thickness has been
9. Photoreceptor layer: This layer contains the inner the most extensively utilized OCT measure in
and outer (photosensitive) segments of the rods both research studies and clinical practice. Other
and cones. measures that may be quantified with commer-
10. Retinal pigment epithelium (RPE): This is a sin- cially available software include the average macu-
gle layer of pigmented hexagonal cells that plays a lar thickness and total macular volume.
multifunctional supportive role, including absorp- These measures have been shown to demonstrate
tion of excess light, synthesis of growth factors, excellent reproducibility, especially with modern
and phagocytosis of photoreceptor waste pro- spectral-domain OCT, and their reliability can be
ducts, among other roles. Beneath the RPE lies maximized by ensuring optimal scan acquisition.
Bruch’s membrane, which separates the RPE [3, 4].
from the choroid. All quantitative OCT measures are obviously
The layers of the retina can be easily discerned in derived from segmentation of OCT images; how-
images acquired with SD-OCT (Figures 13.1 and ever, we make the distinction between “conven-
13.2). tional” OCT measures, which include the 157
 Chapter 13: Optical coherence tomography and retinal segmentation in neurological diseases

C
ILM
RNFL
B
GCL
ILM
GCL
INL IPL
ONL RNFL
OPL
INL
IPL
Fovea OPL

ONL

ELM

PR
IS/OS
IS ELM
RPE

RPE
OS

Figure 13.1 Eye of a healthy control subject.

(A) Fundus photograph from a healthy control subject. (B) A three-dimensional macular volume cube generated by spectral-domain
OCT (Cirrus HD-OCT) from the macular region denoted by the red box in panel A from the same healthy control subject. Note the
individual layers of the retina are readily discernible, except for the ganglion cell layer (GCL) and inner plexiform layer (IPL), which are
difficult to distinguish. During the segmentation process (performed in 3-dimension), the segmentation software identifies the outer
boundaries of the macular retinal nerve fiber layer (RNFL), IPL, and outer plexiform layer (OPL), as well as the inner boundary of the retinal
pigment epithelium (RPE), which is identified by the conventional Cirrus HD-OCT algorithm. The identification of these boundaries
facilitates OCT segmentation, enabling determination of the thicknesses of the macular RNFL, GCL+IPL, the inner nuclear layer (INL) + OPL,
and the outer nuclear layer (ONL), including the inner and outer photoreceptor segments. (C) Illustration of the cellular composition of the
retinal layers depicted in panel B. ELM indicates the external limiting membrane; ILM, the inner limiting membrane; IS, the inner
photoreceptor segments; and OS, the outer photoreceptor segments.
Reproduced with permission from Saidha, Shiv, Sotirchos, Elias S., et al. Relationships Between Retinal Axonal and Neuronal Measures and
Global Central Nervous System Pathology in Multiple Sclerosis Retinal Measures and Global CNS Pathology in MS. American Medical
Association, Archives of Neurology, 2012.

peripapillary RNFL thickness, the average macular reflectivity. Of note, however, is the fact that the
thickness and the total macular volume, and “intra- contrast difference between the GCL and IPL is
retinal segmentation derived” measures, which small, and consequently the boundary between
include the thicknesses of individual or composite these layers is difficult to assess (Figure 13.2) [5, 6].
layers in the macular area. Intra-retinal segmenta- The same, but to a lesser extent, is also true of the
tion has been made possible by advances in scan INL and OPL. Thus, distinguishing these layers has
acquisition with the advent of high-resolution spec- been challenging and the vast majority of studies
tral-domain OCT and also by the development of utilizing intra-retinal segmentation in the study of
sophisticated automated algorithms. As previously neurological disease have utilized composite mea-
mentioned, contrast is generated between retinal sures, including the combined GCL and IPL (GCL
layers on OCT by differences in their optical proper- +IPL) and combined INL and OPL (INL+OPL). In
ties, which result from their varying histological addition to these composite measures, the macular
composition, and high reflectivity layers correspond RNFL and the ONL, including the photoreceptor
to areas where retinal elements are arranged hori- segments, have also been utilized, and all these layers
158 zontally, such as the RNFL and plexiform layers, may now be isolated with great accuracy and relia-
whereas the nuclear layers exhibit relatively lower bility with fully automated segmentation algorithms.
 Chapter 13: Optical coherence tomography and retinal segmentation in neurological diseases

INL IPL GCL RNFL

1
2
3
4
5
6
7
8

ONL OPL
RPE OS IS

Figure 13.2 Example of a manually segmented spectral-domain OCT (Spectralis) macular B-scan traversing the fovea.
(1) inner limiting membrane, (2) outer boundary of retinal nerve fiber layer, (3) outer boundary of inner plexiform layer, (4) outer boundary of
inner nuclear layer, (5) outer boundary of outer plexiform layer, (6) external limiting membrane, (7) inner/outer photoreceptor segment
junction, (8) Bruch’s membrane. Layers: (1–2) = RNFL; (2–3) = GCL+IPL (note the minimal difference in contrast between the two layers, the
border has not been segmented); (3–4) = INL; (4–5) = OPL; (5–6) = ONL; (6–7) = IS; and (7–8) = OS + RPE (OS/RPE border has not been
segmented)
Reproduced with permission from Siego, Micahaela A., In vivo assessment of retinal neuronal layers in multiple sclerosis with manual and
automated optical coherence tomography segmentation techniques. Journal of Neurology; Volume 259, Issue 10, pp 2119–2130.

Retinal segmentation in demyelinating INL and ONL dysfunction in MS [13–15].

Furthermore, retinal perivascular inflammation
diseases of the central nervous (periphlebitis), suggesting disruption of the blood–
system retina barrier, occurs in up to 20% of patients with
MS. Active retinal periphlebitis tends to occur
Multiple sclerosis simultaneously with disruption of the blood–brain
Multiple sclerosis (MS) is an inflammatory disorder of barrier in these patients, and may be a risk factor
the central nervous system that has a predilection to for relapses and gadolinium-enhancing lesions [16,
affect the optic nerves clinically and subclinically. 17]. Intermediate uveitis, especially pars planitis,
Acute optic neuritis (ON) is the initial demyelinating also occurs in up to 15% of patients with MS [12].
event in 15–20% of patients with MS and occurs at Consistent with these clinical observations, postmor-
some time during the course of the disease in 50% of tem analyses show retinal inflammation with activated
patients [7]. Moreover, postmortem studies have microglia in the eyes of people with MS [12].
revealed that 94–99% of MS patients have demyelinat- Collectively, these findings show that the retina is
ing plaques in their optic nerves, irrespective of ON a common site of inflammation, disruption of the
history [8, 9]. Retinal pathology is thought to occur blood-retina barrier, and neuronal-axonal loss in
primarily secondary to optic nerve involvement in MS. Thus, in the study of MS the retina represents a
MS; optic nerve inflammation and demyelination unique, unmyelinated, and easily accessible part of the
results in axonal transection and retrograde degenera- CNS within which to study the neurodegeneration
tion of its constituent fibers, leading to atrophy of the and inflammation associated with the disease. This
RNFL, the innermost retinal layer, where these axons has fueled extensive interest in utilization of OCT
originate. Degeneration of the retinal nerve fibers in with macular intra-retinal segmentation in order to
turn leads to death of retinal ganglion cells (the neu- study in vivo the effects of MS-related disease pro-
rons from which these axons are derived) via apopto- cesses in the retina.
tic mechanisms [10]. Pathologic studies examining The primary OCT measure that has been utilized
the retinas of patients with MS have demonstrated in the study of MS eyes is the peripapillary RNFL
atrophy of the inner retinal layers (RNFL and GCL) thickness. The use of macular intra-retinal segmenta-
in 71–79% of eyes of patients with MS [11, 12]. tion has also allowed the measurement of the macular
In addition to the inner retinal layer pathology RNFL and GCL+IPL thicknesses, measures that in
observed in MS, evidence exists that deeper retinal general are considered to convey similar information.
layers may also be affected in MS, with extensive These measures have been shown across OCT plat-
qualitative atrophy of the INL identified in over 40% forms and segmentation algorithms to exhibit excel-
of MS eyes [12]. Also, studies employing electroreti- lent reproducibility, and both measures have been
shown to be decreased in MS eyes relative to healthy 159
nography have identified abnormalities suggestive of
 Chapter 13: Optical coherence tomography and retinal segmentation in neurological diseases

controls irrespective of ON history, but the decreases studies have been unable to recapitulate these findings
are significantly greater in eyes with a clinical history [28–30]. One longitudinal study utilizing macular intra-
of ON [5, 6, 18–20]. retinal segmentation showed that accelerated thinning
Following ON, the thickness of the peripapillary of the GCL + IPL also occurs in MS in the absence of
RNFL and GCL + IPL have been shown to exhibit a clinical episodes of ON and that the rate of thinning is
decrease over time, with most thinning occurring significantly increased in those patients with more
during the first three months and continuing for six active disease, as defined by clinical relapses, presence
months to one year after onset [21, 22]. During the of gadolinium-enhancing lesions and/or new T2-hyper-
initial phase of ON, however, the two measures exhi- intense lesions [30].
bit different characteristics: the GCL + IPL thickness As previously mentioned, abnormalities of deeper
appears to be unaffected initially, whereas the peripa- retinal layers have also been reported in MS on the basis
pillary RNFL thickness may exhibit increased thick- of pathologic and electrophysiologic studies [12–15].
ness, most likely due to axonal edema. Thus, when Quantitative abnormalities of deeper retinal layers in
utilizing peripapillary RNFL measurements in acute OCT studies utilizing macular intra-retinal segmenta-
ON to follow patients longitudinally, it is inherently tion have also been reported in MS; however, results
difficult to differentiate between resolution of edema vary. In one study, thinning of the deeper retinal layers
versus true atrophy of the peripapillary RNFL. (INL + OPL and ONL + PR) was demonstrated to occur
Approaches to this problem have included utilizing in a subset of MS patients in eyes without a history of
the fellow “unaffected” eye to establish baseline mea- ON that had normal peripapillary RNFL thickness [5].
sures; however, this approach has significant short- Patients with eyes exhibiting these characteristics were
comings due to the fact that often patients with determined to exhibit a macular thinning predominant
multiple sclerosis have experienced ON in their fellow (MTP) retinal phenotype and were shown to have
eye and also since it is known that, even in the absence accelerated accumulation of global disability, as
of a clinical episode of ON, peripapillary RNFL thin- demonstrated by higher Multiple Sclerosis Severity
ning occurs in MS. This is an important factor and Scores (MSSS), a measure of disability adjusted for
makes the GCL + IPL appear to be a more useful disease duration. It was suggested that these retinal
measure in acute ON since establishing with confi- abnormalities may be due to a primary retinal process
dence a baseline measurement, since establishing with that occurs independent of optic nerve pathology, but
confidence a baseline measurement that can be fol- this thinning of deeper retinal layers has not been
lowed longitudinally is extremely important in order consistently shown to occur in other studies.
to reliably employ a measure in clinical practice as Furthermore, an entity that has been termed microcys-
well as in research studies. tic macular edema (MME; Figure 13.3), which predo-
Multiple studies have demonstrated the functional minantly affects the INL, has been recently described in
relevance of RNFL and GCL + IPL atrophy by showing association with optic neuropathies of various etiolo-
that thinning of these layers correlates with visual dys- gies, including MS, NMO, CRION, optic nerve glioma
function in MS, as measured by assessment of high- and associated with neurofibromatis type I, and Leber’s and
low-contrast visual acuity [6, 19, 23]. Additionally, these Tanzanian hereditary optic neuropathy [31–38]. MME
layers have been shown to correlate with the Expanded was first described in association with MS, and in
Disability Status Scale (EDSS; a global measure of dis- studies that have systematically examined for the pre-
ability in MS) and with MRI measures of brain atrophy, sence of MME in MS (of which one employed macular
correlations that are predominantly observed in the intra-retinal segmentation), this pathology was
absence of a history of ON, suggesting that the atrophy reported to occur in 4.7–6% of MS patients [15, 31,
of inner retinal layers observed in non-ON eyes may 32]. MME eyes had more frequently experienced a
reflect global neurodegenerative processes in MS [18, prior episode of symptomatic ON, had worse letter
24, 25]. Longitudinal studies in MS have shown that the acuity at both high- and low-contrast and correspond-
peripapillary RNFL exhibits accelerated thinning in MS ingly more severe thinning of the RNFL and GCL +
compared to healthy controls, even in the absence of IPL, and patients exhibiting MME had experienced
ON, and this has been shown to correlate with clinically more rapid progression of their disease (i.e., higher
significant visual loss and with relapses during follow- MSSS scores). MME was found to be related to
160 up [26, 27]. However, data are conflicting and other increased INL + OPL thickness, which was detectable
 Chapter 13: Optical coherence tomography and retinal segmentation in neurological diseases

A B C

200 µm

D E F

G H I

Figure 13.3 Microcystic macular edema of the inner nuclear layer as identified by spectral-domain optical coherence tomography, with
automated segmentation lines displayed.
Optical coherence tomography (OCT) segmentation done in three dimensions shows the inner limiting membrane, the outer
boundaries of the retinal nerve fiber layer (RNFL), inner plexiform layer, outer plexiform layer, and the inner boundary of the retinal pigment
epithelium. The identification of these retinal boundaries enables the determination of the thicknesses of the macular RNFL, the ganglion
cell layer and inner plexiform layer (GCL + IPL; labeled * in panel A), the INL (including the outer plexiform layer; labeled † in panel A), and
the ONL (including the inner and outer photoreceptor segments; labeled ‡ in panel A). For panels A to C, all images were acquired from the
same patient during a three-year period and are presented in chronological order. Microcystic macular edema (MME; green arrows) of the
INL was present at baseline, as well as a foveal cyst of the outer nuclear layer (ONL; white arrow A). The ONL cyst progressively resolved
during follow-up (B–C). For panels D to F, all images were acquired from the same patient during a two-year period and are presented in
chronological order. A single INL cyst (green arrow) was present at baseline (D). The cyst spontaneously resolved after one year (E). After
fingolimod treatment (initiated after scan E), the patient developed new cystic changes of the INL (green arrow; F). Panels G to I show three
different patients with MME of the INL (green arrows). Vessel artifacts (arrow heads) are shown for comparison.
Reproduced with permission from Lancet Neurology

in eyes prior to the development of visible MME,

highlighting the potential for earlier and broader Neuromyelitis optica
identification of this process with utilization of OCT Neuromyelitis optica (NMO) is an inflammatory dis-
segmentation [32]. Additionally, increased INL thick- order of the central nervous system, the cardinal
ness has been shown to occur in MS eyes after ON in manifestations of which are ON and longitudinally
the absence of MME, and this increase has been extensive transverse myelitis (LETM). Autoantibodies
shown to correlate with the atrophy observed in the (NMO–immunoglobulin G [IgG]) targeting aqua-
GCL + IPL, suggesting that MME may represent the porin-4 are found in the sera of the majority of
extreme end of a spectrum of INL pathology [32, 39]. patients with NMO. In approximately 50% of NMO
Interestingly, one cross-sectional study showed that patients isolated ON is the initial manifestation of the
in RRMS the INL + OPL thickness exhibits a positive disease. ON associated with NMO carries a worse
correlation with T2 lesion volume and an inverse prognosis than in MS with 60% of NMO patients
correlation with normal appearing white matter experiencing unilateral or bilateral blindness at a
[24]. Moreover, in the one longitudinal study of median of 7.7 years after disease onset, compared
MME that employed macular intra-retinal segmenta- with 4% of MS patients with ON at 15-year follow-
tion, increased INL thickness at baseline was asso- up [40]. Histopathologically, NMO lesions are not
ciated with increased clinico-radiological disease uncommonly associated with necrosis and cavitation,
activity during follow-up (relapses, EDSS progres- features that are not typically encountered in MS [41].
sion, new T2 and contrast-enhancing lesions) [32]. Unfortunately, retinal pathologic studies have not
This finding must be validated by further studies but been performed in NMO and the data regarding ret-
highlights the potential of OCT segmentation for inal involvement in NMO is almost exclusively
identifying novel biomarkers for neurological disease. derived from studies utilizing OCT. 161
 Chapter 13: Optical coherence tomography and retinal segmentation in neurological diseases

OCT studies utilizing conventional OCT measures Retinal segmentation in other

have shown that the decrease that occurs in the peripa-
pillary RNFL and macular volume following NMO- neurological diseases
associated ON is substantially greater than in MS, and
this has been proposed as an important differentiating Alzheimer’s disease
factor between the two conditions [42–44]. Also, the Visual dysfunction may occur early in Alzheimer’s
pattern of peripapillary RNFL atrophy appears to differ, disease (AD) and includes deficits in visual acuity,
with NMO-ON resulting in severe thinning of all quad- contrast sensitivity, color vision, motion perception,
rants, whereas MSON is predominantly associated with visuospatial construction and visual memory [49].
temporal thinning [42, 44]. Intra-retinal segmentation Furthermore, histopathologic studies have found axo-
has allowed for more specific observations to be made nal degeneration in the optic nerves, and reductions
and studies have shown that not only the peripapillary in the number of retinal ganglion cells and in RNFL
RNFL, but also the macular-RNFL and GCL + IPL are thickness in eyes of patients with AD. Studies utilizing
more severely affected in NMO [22, 45]. OCT have confirmed in vivo that thinning of the
Furthermore, MME of the INL has been shown to peripapillary RNFL and reduction of macular volume
occur in 18–26% of patients with NMO-spectrum dis- occurs in AD eyes; however, OCT studies employing
orders, a prevalence substantially greater than that intra-retinal segmentation have not been conducted
observed in MS (~5%).[33, 34, 45] Moreover, MME in and this is an important future direction for retinal
NMO appears to be more closely linked with ON, with research in AD [49, 50].
MME being exclusively identified in eyes with a history
of symptomatic ON. As in MS, MME in NMO has been Parkinson’s disease
shown to be associated with increased thickness of the Visual symptomatology is common in Parkinson’s
INL and poor visual function and is present in eyes that disease (PD), as demonstrated by reports of abnormal
have severe retinal axonal and neuronal loss, as demon- visual acuity, contrast sensitivity, color vision motion
strated by greater decreases in the thickness of the RNFL perception, and the presence of visual hallucinations
(peripapillary and macular) and GCL + IPL when com- [51]. Furthermore, of interest is the fact that some of
pared to NMO-ON eyes that have not developed MME. these abnormalities have been shown to improve with
The etiology of the higher prevalence of MME in L-dopa administration. Electrophysiologic studies,
NMO relative to MS has been proposed to be due to the including electroretinograms and visual evoked
increased severity of NMO-ON, with MME represent- potentials, have also identified abnormalities in PD.
ing a marker of axonal and neuronal injury. The preceding evidence has provided the impetus for
Alternatively, it has been thought that a primary retinal OCT studies to be performed in PD. Decreases of the
process occurring in conjunction with ON may be peripapillary RNFL and macular thickness have been
responsible. Aquaporin-4 (the target of the immune identified in some studies in PD; however, other stu-
response in NMO) is highly expressed in the retina by dies have been unable to reproduce these findings.
Müller glial cells (the cell bodies of which are located in Larger-scale studies employing high-definition OCT
the INL), especially in end-feet membranes facing blood with intra-retinal segmentation are indicated in order
vessels [40]. Thus, a plausible and intriguing hypothesis to further study retinal involvement in PD [51, 52].
is that following acute ON and breakdown of the blood–
retinal barrier, an immune response directed against
aquaporin-4 in the retina may contribute to the devel- Amyotrophic lateral sclerosis
opment of MME and account in part for the poor visual Although visual disturbances are not recognized to
outcomes in NMO-ON. In any case, further studies and occur in amyotrophic lateral sclerosis (ALS), growing
characterization of MME is necessary in order to clarify evidence supports that non-motor systems may be
the pathogenesis of this retinal phenotype. involved in amyotrophic lateral sclerosis. One study
Additionally, subclinical involvement of the visual has utilized OCT with automated intra-retinal seg-
pathway has been suggested to occur in NMO spectrum mentation to examine the eyes of ALS patients, but
disorders, with some studies identifying decreased no differences were found relative to healthy controls
RNFL and GCL + IPL thicknesses in NMO non-ON in the macular GCL + IPL, INL + OPL, ONL + PR and
162 eyes, but data are conflicting [22, 33, 42, 43, 46–48]. peripapillary RNFL thicknesses [53].
 Chapter 13: Optical coherence tomography and retinal segmentation in neurological diseases

15. Gundogan FC, Demirkaya S, Sobaci G. Is optical

References coherence tomography really a new biomarker candi-
1. Frohman EM, Fujimoto JG, Frohman TC, Calabresi date in multiple sclerosis?–A structural and functional
PA, Cutter G, Balcer LJ. Optical coherence tomogra- evaluation. Invest Ophthalmol Vis Sci 2007; 48:5773–81.
phy: a window into the mechanisms of multiple
16. Engell T, Hvidberg A, Uhrenholdt A. Multiple sclero-
sclerosis. Nat Clin Pract Neurol 2008; 4:664–75.
sis: periphlebitis retinalis et cerebro-spinalis. A corre-
2. Debuc DC. A review of algorithms for segmentation of lation between periphlebitis retinalis and abnormal
retinal image data using optical coherence tomography. technetium brain scintigraphy. Acta Neurol Scand
In: Ho P-G, editor. Image Segmentation: InTech; 2011. 1984; 69:293–7.
3. Cettomai D, Pulicken M, Gordon-Lipkin E, Salter A, 17. Sepulcre J, Murie-Fernandez M, Salinas-Alaman A,
Frohman TC, Conger A, et al. Reproducibility of Garcia-Layana A, Bejarano B, Villoslada P. Diagnostic
optical coherence tomography in multiple sclerosis. accuracy of retinal abnormalities in predicting disease
Arch Neurol 2008; 65:1218–22. activity in MS. Neurology 2007; 68:1488–94.
4. Syc SB, Warner CV, Hiremath GS, Farrell SK, 18. Petzold A, de Boer JF, Schippling S, Vermersch P,
Ratchford JN, Conger A, et al. Reproducibility of high- Kardon R, Green A, et al. Optical coherence tomogra-
resolution optical coherence tomography in multiple phy in multiple sclerosis: a systematic review and
sclerosis. Mult Scler 2010; 16:829–39. meta-analysis. Lancet Neurol 2010; 9:921–32.
5. Saidha S, Syc SB, Ibrahim MA, Eckstein C, Warner CV, 19. Walter SD, Ishikawa H, Galetta KM, Sakai RE, Feller
Farrell SK, et al. Primary retinal pathology in multiple DJ, Henderson SB, et al. Ganglion cell loss in relation
sclerosis as detected by optical coherence tomography. to visual disability in multiple sclerosis. Ophthalmology
Brain 2011; 134:518–33. 2012; 119:1250–7.
6. Seigo MA, Sotirchos ES, Newsome S, Babiarz A, 20. Warner CV, Syc SB, Stankiewicz AM, Hiremath G,
Eckstein C, Ford E, et al. In vivo assessment of retinal Farrell SK, Crainiceanu CM, et al. The impact of uti-
neuronal layers in multiple sclerosis with manual and lizing different optical coherence tomography devices
automated optical coherence tomography segmenta- for clinical purposes and in multiple sclerosis trials.
tion techniques. J Neurol 2012; 259:2119–30. PLoS ONE 2011; 6:e22947.
7. Balcer LJ. Clinical practice. Optic neuritis. N Engl J 21. Henderson AP, Altmann DR, Trip AS, Kallis C, Jones
Med 2006; 354:1273–80. SJ, Schlottmann PG, et al. A serial study of retinal
8. Toussaint D, Perier O, Verstappen A, Bervoets S. changes following optic neuritis with sample size esti-
Clinicopathological study of the visual pathways, eyes, mates for acute neuroprotection trials. Brain 2010;
and cerebral hemispheres in 32 cases of disseminated 133:2592–602.
sclerosis. J Clin Neuroophthalmol 1983; 3:211–20. 22. Syc SB, Saidha S, Newsome SD, Ratchford JN, Levy M,
9. Ikuta F, Zimmerman HM. Distribution of plaques in Ford E, et al. Optical coherence tomography segmen-
seventy autopsy cases of multiple sclerosis in the tation reveals ganglion cell layer pathology after optic
United States. Neurology 1976; 26:26–8. neuritis. Brain 2012; 135:521–33.
10. Shindler KS, Ventura E, Dutt M, Rostami A. 23. Saidha S, Syc SB, Durbin MK, Eckstein C, Oakley JD,
Inflammatory demyelination induces axonal injury Meyer SA, et al. Visual dysfunction in multiple
and retinal ganglion cell apoptosis in experimental sclerosis correlates better with optical coherence
optic neuritis. Exp Eye Res 2008; 87:208–13. tomography derived estimates of macular ganglion cell
11. Kerrison JB, Flynn T, Green WR. Retinal pathologic layer thickness than peripapillary retinal nerve fiber
changes in multiple sclerosis. Retina 1994; 14:445–51. layer thickness. Mult Scler 2011; 17:1449–63.
12. Green AJ, McQuaid S, Hauser SL, Allen IV, Lyness R. 24. Saidha S, Sotirchos ES, Oh J, Syc SB, Seigo MA, Shiee
Ocular pathology in multiple sclerosis: retinal atrophy N, et al. Relationships between retinal axonal and
and inflammation irrespective of disease duration. neuronal measures and global central nervous system
Brain 2010; 133:1591–601. pathology in multiple sclerosis. JAMA Neurol 2013;
70:34–43.
13. Papakostopoulos D, Fotiou F, Hart JC, Banerji NK.
The electroretinogram in multiple sclerosis and 25. Zimmermann H, Freing A, Kaufhold F, Gaede G, Bohn
demyelinating optic neuritis. Electroencephalogr Clin E, Bock M, et al. Optic neuritis interferes with optical
Neurophysiol 1989; 74:1–10. coherence tomography and magnetic resonance ima-
ging correlations. Mult Scler 2013; 19:443–50.
14. Forooghian F, Sproule M, Westall C, Gordon L,
Jirawuthiworavong G, Shimazaki K, et al. 26. Talman LS, Bisker ER, Sackel DJ, Long DA, Jr., Galetta
Electroretinographic abnormalities in multiple sclero- KM, Ratchford JN, et al. Longitudinal study of vision
sis: possible role for retinal autoantibodies. Doc and retinal nerve fiber layer thickness in multiple
Ophthalmol 2006; 113:123–32. sclerosis. Ann Neurol 2010; 67:749–60. 163
 Chapter 13: Optical coherence tomography and retinal segmentation in neurological diseases

27. Garcia-Martin E, Pueyo V, Almarcegui C, Martin J, 40. Levin MH, Bennett JL, Verkman AS. Optic neuritis in
Ara JR, Sancho E, et al. Risk factors for progressive neuromyelitis optica. Prog Retin Eye Res 2013.
axonal degeneration of the retinal nerve fibre layer in 41. Lucchinetti CF, Mandler RN, McGavern D, Bruck W,
multiple sclerosis patients. Br J Ophthalmol 2011; Gleich G, Ransohoff RM, et al. A role for humoral
95:1577–82. mechanisms in the pathogenesis of Devic’s neuro-
28. Henderson AP, Trip SA, Schlottmann PG, Altmann myelitis optica. Brain 2002; 125:1450–61.
DR, Garway-Heath DF, Plant GT, et al. A preliminary 42. Naismith RT, Tutlam NT, Xu J, Klawiter EC, Shepherd
longitudinal study of the retinal nerve fiber layer in J, Trinkaus K, et al. Optical coherence tomography
progressive multiple sclerosis. J Neurol 2010; differs in neuromyelitis optica compared with multiple
257:1083–91. sclerosis. Neurology 2009; 72:1077–82.
29. Serbecic N, Aboul-Enein F, Beutelspacher SC, Vass C, 43. Ratchford JN, Quigg ME, Conger A, Frohman T,
Kristoferitsch W, Lassmann H, et al. High resolution Frohman E, Balcer LJ, et al. Optical coherence tomo-
spectral domain optical coherence tomography (SD- graphy helps differentiate neuromyelitis optica and MS
OCT) in multiple sclerosis: the first follow up study optic neuropathies. Neurology 2009; 73:302–8.
over two years. PLoS ONE 2011; 6:e19843.
44. Green AJ, Cree BA. Distinctive retinal nerve fibre layer
30. Ratchford JN, Saidha S, Sotirchos ES, Oh JA, Seigo and vascular changes in neuromyelitis optica following
MA, Eckstein C, et al. Active MS is associated with optic neuritis. J Neurol Neurosurg Psychiatry 2009;
accelerated retinal ganglion cell/inner plexiform layer 80:1002–5.
thinning. Neurology 2013; 80:47–54.
45. Schneider E, Zimmermann H, Oberwahrenbrock T,
31. Gelfand JM, Nolan R, Schwartz DM, Graves J, Green AJ. Kaufhold F, Kadas EM, Petzold A, et al. Optical
Microcystic macular edema in multiple sclerosis is asso- coherence tomography reveals distinct patterns of
ciated with disease severity. Brain 2012; 135:1786–93. retinal damage in neuromyelitis optica and multiple
32. Saidha S, Sotirchos ES, Ibrahim MA, Crainiceanu CM, sclerosis. PLoS ONE 2013; 8:e66151.
Gelfand JM, Sepah YJ, et al. Microcystic macular 46. de Seze J, Blanc F, Jeanjean L, Zephir H, Labauge P,
edema, thickness of the inner nuclear layer of the Bouyon M, et al. Optical coherence tomography in
retina, and disease characteristics in multiple sclerosis: neuromyelitis optica. Arch Neurol 2008; 65:920–3.
a retrospective study. Lancet Neurol 2012; 11:963–72.
47. Merle H, Olindo S, Donnio A, Richer R, Smadja D,
33. Sotirchos ES, Saidha S, Byraiah G, Mealy MA, Ibrahim Cabre P. Retinal peripapillary nerve fiber layer thick-
MA, Sepah YJ, et al. In vivo identification of morpho- ness in neuromyelitis optica. Invest Ophthalmol Vis Sci
logic retinal abnormalities in neuromyelitis optica. 2008; 49:4412–7.
Neurology 2013; 80:1406–14.
48. Monteiro ML, Fernandes DB, Apostolos-Pereira SL,
34. Gelfand JM, Cree BA, Nolan R, Arnow S, Green AJ. Callegaro D. Quantification of retinal neural loss in
Microcystic inner nuclear layer abnormalities and patients with neuromyelitis optica and multiple sclero-
neuromyelitis optica. JAMA Neurol 2013; 70:629–33. sis with or without optic neuritis using optical coher-
35. Balk LJ, Killestein J, Polman CH, Uitdehaag BM, ence tomography. Invest Ophthalmol Vis Sci 2012.
Petzold A. Microcystic macular edema confirmed, but 49. Ikram MK, Cheung CY, Wong TY, Chen CP. Retinal
not specific for multiple sclerosis. Brain 2012; pathology as biomarker for cognitive impairment and
135:e226. Alzheimer’s disease. J Neurol Neurosurg Psychiatry
36. Abegg M, Zinkernagel M, Wolf S. Microcystic 2012; 83:917–22.
macular degeneration from optic neuropathy. Brain 50. He XF, Liu YT, Peng C, Zhang F, Zhuang S, Zhang JS.
2012; 135:e225. Optical coherence tomography assessed retinal nerve
37. Barboni P, Carelli V, Savini G, Carbonelli M, La fiber layer thickness in patients with Alzheimer’s dis-
Morgia C, Sadun AA. Microcystic macular ease: a meta-analysis. Int J Ophthalmol 2012; 5:401–5.
degeneration from optic neuropathy: not inflamma- 51. Archibald NK, Clarke MP, Mosimann UP, Burn DJ.
tory, not trans-synaptic degeneration. Brain 2013; The retina in Parkinson’s disease. Brain 2009;
136:e239. 132:1128–45.
38. Kisimbi J, Shalchi Z, Mahroo OA, Mhina C, Sanyiwa 52. Archibald NK, Clarke MP, Mosimann UP, Burn DJ.
AJ, Mabey D, et al. Macular spectral domain optical Retinal thickness in Parkinson’s disease. Parkinsonism
coherence tomography findings in Tanzanian endemic Relat Disord 2011; 17:431–6.
optic neuropathy. Brain 2013.
53. Roth NM, Saidha S, Zimmermann H, Brandt AU,
39. Kaushik M, Wang CY, Barnett MH, Garrick R, Parratt Oberwahrenbrock T, Maragakis NJ, et al. Optical
J, Graham SL, et al. Inner Nuclear Layer Thickening Is coherence tomography does not support optic nerve
Inversley Proportional to Retinal Ganglion Cell Loss in involvement in amyotrophic lateral sclerosis. Eur J
164 Optic Neuritis. PLoS ONE 2013; 8:e78341. Neurol 2013; 20:1170–6.
 Chapter
Optical coherence tomography and retinal

14 pathology in neurologic diseases

Ari J. Green

Pathology is the foundation of clinical medicine. Since than the rule, and tissue samples from the living retina
the nineteenth century, diseases have come to be are exceptionally uncommon indeed.
understood not simply in the context of the symptoms As a consequence, our understanding of the
they produce but in consideration of the cellular dis- pathological underpinnings of neurodegenerative
ruption that accompanies them. We, therefore, define conditions comes principally from pathological speci-
diseases in reference to their particular pattern of mens obtained at the end of life – and not from tissue
cellular injury and molecular dysfunction. samples obtained when the disease is developing or is
Knowledge concerning cellular pathology guides our most active. Even these end-of-life samples are in
understanding of the biological mechanisms that short supply because of a combination of cultural
underpin disease development and progression. In and social considerations and the difficulty with
the age of immunohistochemistry, histopathology procuring neuronal tissues in a timely manner before
has been extended to help unfold the specific mole- tissue degradation associated with death has signifi-
cular pathways that are dysregulated in each disease cantly compromised tissue quality. This poses a
process. challenge for our understanding of these disease
The principal limitation of standard pathology is processes because the end stage of tissue injury may
that it only provides a window on the specific aberrant not reflect the mechanisms that led to this ultimate
cellular and molecular processes at a single point in state. In some instances, by the time tissue can be
time (when the tissue is obtained). In diseases outside obtained, the distortion caused by the disease leaves
of the central nervous system, the capacity to clinically us without insights regarding the mechanisms
analyze biopsy specimens allows for the study of and processes that resulted in that injury. As a field,
cellular dysregulation on a semi-dynamic scale. this has left us with [1] a limited supply of tissue
However, in tissues where maintenance of intact with [2] a restricted view of disease processes reflect-
cellular architecture is crucial to retained organ func- ing the end result of years or even decades of disease
tion (such as in the central nervous system [CNS]) progression.
biopsy has a limited role for diagnosing disease and Biomedical imaging provides the capacity to study
essentially no role for monitoring and tracking patho- pathological processes more dynamically, especially in
logical progression. This is an obvious limitation in the tissue types that cannot be sampled without
the brain – where critical periods of highly dynamic disrupting critical cellular function. However, findings
development hardwire the system’s functionality – from biomedical imaging must be grounded in a
and dysfunction caused by disruption from tissue pathology-based understanding of the disease under
acquisition would be difficult to recover. This circum- study. Therefore, imaging is principally useful in refer-
stance is accentuated in the retina because the ence to known pathological processes, which allow us
complex cellular architecture of the retina is engraved to contextualize our observations in regard to the stan-
on a tissue that is paper thin and not amenable to easy dard methods for categorizing and understanding
tissue sampling. Therefore biopsies of the brain to disease. This is particularly true for imaging methods
diagnose neurological disease are the exception rather that are optimized in terms of scale and resolution for

Optical Coherence Tomography in Neurological Disease, ed. Peter A. Calabresi, Laura J. Balcer, and Elliot M. Frohman. 165
Published by Cambridge University Press. © Cambridge University Press 2015.
 Chapter 14: Optical coherence tomography and retinal pathology in neurologic diseases

monitoring tissue morphology. Although their relative pathological tissue or at least by reference to animal
resolutions differ substantially, this limitation applies models of disease with fidelity for the disease feature
across a broad selection of biomedical imaging techni- under study.
ques including standard radiography, computerized MRI and to a lesser extent CT scans have come to
tomography, magnetic resonance imaging, and even prominence in clinical neuroscience in large part
optical coherence tomography. because the images can be represented in such a
This condition constitutes the principal challenge way that they show good fidelity for known gross
in using retinal imaging to understand and track dis- pathology of brain substructures and disease pathol-
eases of the central nervous system. Namely, we must ogy (i.e., lesions in MS and localized atrophy in neu-
demonstrate that we have identified the specific sig- rodegeneration). There is, therefore, tremendous
nificance of any imaging abnormality identified. value to correlative studies between pathology and
For example, we must determine if volume loss biomedical imaging. At the spatial resolution of
detected on imaging reflects cell loss, tissue atrophy, OCT, that association would best be represented by
or loss of tissue water. Furthermore, in a pathological histopathology.
state we don’t know that volume loss per se reflects In total, imaging is most meaningful in reference
tissue injury to a particular class of cells unless we to known underlying histopathology, but advancing
know the cell types that inhabit that tissue in both the field is likely an iterative process requiring
health and disease. The second major assumption of human subject imaging, acquisition, and analysis of
retinal imaging in CNS diseases is that retinal mani- carefully ascertained human tissue samples when
festations of the disease meaningfully represent some available, and additional cautious reference to ani-
broader aspect of the disease process. It has been mal tissue samples with matched imaging acquired
suggested that retinal imaging must recapitulate all for further refinement.
the features of the broader CNS disease to be useful.
While desirable to some degree, this contention is not
true. For example, in multiple sclerosis, a single lesion Multiple sclerosis
in the subcortical white matter may not reflect all the Multiple sclerosis is a common neurodegenerative
elements of the broader disease, but studying it in disease of the central nervous system, characterized
detail is still profoundly valuable – especially if we by robust inflammation with an adaptive immune
possess the capacity to study lesion dynamics at a response that appears to target myelin. It effects nearly
cellular level. The same is true of retinal manifesta- 2.5 million people worldwide and frequently leads to
tions of neurological disease. Understanding the the development of relentless disability within a few
disease in the retina has the potential to provide us decades of disease onset. The disease is pathologically
important insights about the disease in general if it manifested by a number of principal features. These
represents some fundamental aspect of pathogenesis include lymphocyte and monocyte infiltration into
or progression– it does not necessarily need to the central nervous system, especially in the area
manifest the full spectrum of the disease process. around blood vessels, microglial activation, substan-
Unfortunately, given the challenges posed with tial myelin loss – with less pronounced loss of neurons
obtaining tissue, much of what we know about retinal and their axons – and an astrocytic response with
manifestations of brain diseases is based on a limited gliosis and scarring. As it is the most common
sample. Very few cases of neurological disease have acquired disease involving CNS myelin and given
undergone ophthalmic pathological analysis. that the principal adaptive immune response appears
Furthermore, in many of the cases that have come to to target myelin, it is considered the canonical demye-
autopsy there is limited data to determine if these linating disease of the central nervous system.
cases constitute typical examples of disease. Even Under normal circumstances the human retina
fewer have received systematic ophthalmic pathologi- does not contain any myelin. Therefore, it may be
cal review – meaning that certain pathological pro- surprising that the retina could serve as a model for
cesses may have been missed or overlooked, especially understanding or studying MS. However, the retinal
if they are not evident in all patients with the disease. ganglion cell – which processes visual information
Therefore, new discoveries on retinal imaging may transmitted from the outer retinal layers and trans-
166 need to be contextualized by returning to human mits it the brain (including targets in the thalamus,
 Chapter 14: Optical coherence tomography and retinal pathology in neurologic diseases

midbrain tectum and hypothalamus) – is myelinated normal axonal constituents of the nerve. It was also
along most of its distal portion (through the optic noted that retinal atrophy was irregular but was
nerve, chiasm, and tract). The pathology detailed concentrated in the papillomacular bundle. Ganglion
below highlights that all the canonical features of MS cell loss was described as significant, including
other than demyelination can be detected in the retina marked loss of perifoveal ganglion cells and atrophic
[1–7, 10–13]. This includes axonal and neuronal loss, shrunken appearance to the surviving ganglion cells.
microglial and astroglial activation, lymphocyte infil- Furthermore, perivascular lymphocytes and increased
tration with perivascular cuffing, and gliosis. In fact, glial populations and fibroblasts were described as
the presence of the injury observed in the absence of well as thickened appearance to some blood vessels.
myelin suggests that injury and inflammation in MS In some cases, this gliosis was extensive with fibrosis
can both proceed in the absence of a local myelin that deformed the architecture of the retina and prox-
target. Theoretically, it also means that by studying imal optic nerve [4].
the retina these other processes could potentially be Toussaint conclusively demonstrated that demye-
better understood absent the sometimes overwhelm- linating injury to the anterior visual pathway could be
ing presence of myelin injury. found in virtually all patients with MS in a larger
Given the high frequency of visual dysfunction in series of 32 cases [5]. This was the first investigation
MS, it is not at all surprising that scientists have to include some clinical characterization of source for
undertaken pathological investigations of the visual the tissue sample and there was surprisingly little
system. The most common and well-recognized clin- correlation between the degree of demyelination
ical syndrome involving the visual system in MS is seen pathologically and the amount of visual loss
optic neuritis (ON). Although the episode is classi- that the patient experienced during life [5].
cally thought to initiate and focus injury in the retro- Subsequent studies have also confirmed the unifor-
bulbar optic nerve, the cell that suffers functional mity of optic nerve atrophy and demyelination on
impairment resides in the retina. In ON, the absence pathology in subjects with MS at end of life [6, 7].
of retinal findings defines one of the clinical aphor- Further small series confirmed the preference for
isms most commonly associated with the disease injury of the papillomacular bundle [7] and suggested
(“The patient sees nothing and the doctor sees noth- that the clinical periphlebitis described by Tar Braak
ing.”) Unsurprisingly, there have been no recorded and Rucker was in fact constituted by inflammatory
cases of pathology during acute optic neuritis. Optic infiltrate and was similar to perivascular lymphocyte
nerve head swelling is seen in a minority of cases but cuffing seen in the brain [8–12]. An elegant modern
associated retinal abnormalities are unusual and study demonstrated size selective injury involving
would lead a clinician to question the diagnosis. smaller retinal ganglion cells by studying their projec-
On the other hand, tissue pathology has been far tions in the lateral geniculate nucleus. This provided a
better characterized in MS and demonstrates unequi- context for understanding the injury of the papillo-
vocally that injury to the anterior visual pathways is macular bundle (which is predominantly constituted
near universal in the disease. In fact, involvement of by the smaller parvocellular retinal ganglion cells that
the visual pathway was reputed to be one of the have both smaller axons and dendritic arbors). The
defining pathological characteristics of the disease authors demonstrated that even among parvocellular
for early neuropathologists [2] including Charcot. In RGCs, smaller cells were more prone to injury and
the 1930s Lisch described visual pathway demyelina- loss [13]. This study also documented that optic nerve
tion in 12/12 MS pathology samples studied (only six atrophy extended into the tract and that optic nerve
of which came from patients who had a history of ON) volume loss was at least in part the consequence of
[3]. In the middle part of the last century, Gartner axonal loss. [13]
described retinal and optic nerve atrophy in ten MS The largest and most definitive pathological
cases and documented the significant inner retinal description of retinal injury in MS included 132 eyes
injury seen in the disease [4]. He described demyeli- from 82 subjects with the disease (and 10 embedded
nating plaques in 100% of the optic nerves of patients reference subjects) [1]. Cases were examined in a
with MS. All but one of these patients had significant standardized fashion with evaluators blind to the
nerve atrophy and the exception was noted to have a underlying diagnosis to reduce the influence of poten-
marked increase in reactive glial cells replacing the tial bias. Disease duration ranged from four months to 167
 Chapter 14: Optical coherence tomography and retinal pathology in neurologic diseases

more than five decades. Retinal injury was detected in retinal cell loss and broader neurodegeneration in
all subtypes at all time points of the disease, but the the rest of the CNS, brain volume was correlated
degree of injury was variable. The study conclusively with the degree of retinal atrophy identified in
demonstrated that ganglion cells and their axons are the series and there was trend detected between longer
lost in the large majority of retinas and optic nerves of disease duration and qualitative assessment of retinal
patients with multiple sclerosis. Although difficult to atrophy.
study in the retinal nerve fiber layer itself, neurofila- Interestingly, significant inflammation was
ment staining (which labels axons) demonstrated sig- detected in the retinas of these subjects even though
nificant loss of axons of the optic nerve, and this the samples were taken at the end of life. Many of
appeared to be partially dependent on the age of the the subjects had secondary progressive disease and the
patient and length of time that the patient had suf- principle inflammatory response would be presumed
fered from MS [1]. The series highlighted that retinal to have abated. The study demonstrated further
injury can be detected earlier in the course of disease evidence of retinal inflammation in a subset of cases
by showing that patients documented to have relap- subjected to additional immunohistochemical assess-
sing remitting multiple sclerosis and acute fulminant ment. This inflammation was characterized by the
disease at the time of death exhibited ganglion cell loss appearance of large numbers of cells expressing
and loss of axons as well (Figure 14.1). Overall, 80% of MHC Class II molecules (defining them as of immune
subjects demonstrated qualitative cell loss in the lineage) throughout the retina, some of which were
retina and 71% demonstrated optic nerve atrophy concentrated in the space around blood vessels
[1]. Lastly, as an indication of the association between (Figure 14.2). The rate of frank retinal periphlebitis
was similar between the patients with progressive
and relapsing MS. Foamy appearing cells that
appeared to be of innate immune lineage were also
seen, many of which were outside of the normal
microglial network surrounding the inner nuclear
layer of the retina (Figure 14.3). The finding of infil-
trating and migrating native immune cells distributed
throughout the retina means that some of the retinal
layers seen using standard retinal imaging protocols
may be constituted by cells that are not found in
the healthy retina [1].
The retinal nerve fiber layer is defined at its inner
edge by the internal limiting membrane and at its
outer edge by the ganglion cell layer. Although the
inner edge is a true biological interface, the outer
edge is in large part conceptual. It is defined by
the theoretical boundary between the cell soma of
retinal ganglion cells and their axonal projections.
Figure 14.1 Axon loss in the optic nerve in multiple sclerosis. The boundary appears sharp on optical coherence
These images show significant reduction in normal axonal density in tomography images in part because of how OCT
the optic nerves of patients with MS using a neurofilament stain
(NF). Panel A shows loss of normal fiber bundles at 10x, whereas
signal is produced – axonal microtubules reflect
Panel B show loss of fibers themselves (40x) accompanied by an light differently when they are perpendicular to the
increased glial response. light source.

Figure 14.2: Immune activation in the retina in MS.

HLA-DR staining showing MHC class II + cells in the inner
retina, including in the retinal nerve fiber, ganglion cell,
and inner nuclear/plexiform layers at 20x (Panel A) and
inner nuclear layer at 40x (Panel B).

168
 Chapter 14: Optical coherence tomography and retinal pathology in neurologic diseases

Figure 14.3 Reactive microglia in the retina in MS. Foamy reactive Figure 14.4 Immune cells in perivascular space. HLA-DR staining
innate immune cells identified in the retina near optic nerve head in showing MHC class II positive immune cells adjacent to blood
a subject with SPMS (HLA-DR Stain). These may constitute a reaction vessels in the nerve fiber layer of a subject with MS. Given the
to degenerating neurons. morphology, many of these cells appear to be lymphocytes.

because most perivascular sheathing has been identi-

fied to occur in the mid-periphery and not proximate
to the optic disc [9, 11].
On the other hand, fibrosis especially at the optic
nerve head was another frequently observed patholo-
gical phenomena seen in the pathology series
Figure 14.5 Gliotic scar in the retina and optic nerve. Perivenular [Figure 14.5]. This fibrosis was at times severe enough
gliosis near optic disc (Panel A) and within optic nerve (Panel B) in that appeared to lead to deformation of the normal
subject with SPMS (GFAP stain) architecture of the optic nerve head and could con-
ceivably impact the shape of the cup and the volume
Between the ILM and the GCL inner boundary, of the RNFL [1, 4]. Additional, gliotic changes were
the RNFL is constituted by RGC axons but also detected using GFAP staining for astrocytes that
contains other cellular elements, including the inner showed reactive astrocytes next to vessels exhibiting
retinal vascular branches of the central retinal artery blood retinal barrier disruption and located within the
and vein. The other constituents of the “RNFL” could layer that constitutes RNFL [1].
include displaced amacrine cells, and Müller cell Immune cells were not restricted to the RNFL in
processes in the healthy state. (In fact, recent data the study and HLA-DR positive cells with an activated
suggests that up to half the cells of the GCL in the appearance were found in deeper layers of the retina
mouse may be amacrine cells rather than RGCs) [14]. including the inner nuclear layer and inner plexiform
In disease, infiltrating cells trapped anywhere between layer [1]. Under normal circumstances microglial
the ILM and GCL as well fibrosis/gliosis could also cells of the retina are tiled through the IPL and OPL
contribute volume to the RNFL[15]. In the pathology surrounding the INL; however, in the state of activa-
series 15% of patients had HLA-DR + cells found in tion they have the potential to migrate to sites of
the perivascular cuff (which would presumably be injury or inflammation [16].
detected as RNFL) [Figure 14.4] [1], Outside of the One of the striking findings of the larger systematic
peripapillary RNFL this would contaminate estima- evaluation of retinal samples from MS subjects was the
tions of nerve fiber layer thickness of volume. identification of extensive injury involving the inner
However, this would only have limited impact on nuclear layer of the retina [1]. The population of neu-
routine peripapillary RNFL thickness assessments rons found in this layer (horizontal and bipolar cells) 169
 Chapter 14: Optical coherence tomography and retinal pathology in neurologic diseases

are not myelinated and, therefore, their loss suggests Both lipopigments and ceroid have the tendency to
that there is either direct inflammatory mediated injury be endogenously autofluorescent, meaning that light
in the retina or that secondary trans-synaptic processes of an appropriate frequency will elicit the release of
underlie cell loss in cells connected to a disrupted photons with lower energy (i.e., a longer wavelength)
pathway. [18]. In many instances, these autofluorescent materi-
In terms of the specificity of the findings this study als are sensitive to a wide spectrum of stimulating
had the added benefit that the evaluating pathologist light and, given their varied molecular structure, also
was blinded to the underlying diagnosis (given emit a broad spectrum of photons in response.
embedded pathological controls). This served to The diagnosis of NCL is established through the
reduce potential bias on the part of the evaluating identification of Periodic Acid Schiff and Luxol fast
pathologist in determining the frequency of the blue positive autoflourescent material in the brain and
pathological features observed. retina [17]. In these tissues, as well as in the muscle,
subcutaneous fat, and skin, electron microscopy exhi-
Neuronal ceroid lipofuscinosis bits curvilinear bodies, fingerprint bodies, and
Granular osmophilic deposits. Today, molecular diag-
The neuronal ceroid lipofuscinoses are a set of inher-
nosis is the gold standard for confirming disease and
ited neurodegenerative diseases that exhibit signifi-
establishing disease subtype. Given tissue accessibility
cant progressive injury to the brain and retina.
and low morbidity of biopsy specimens, skin samples
Considered together, they constitute the most
are typically the tissue of choice for disease identifica-
common pediatric neurodegenerative disorders in
tion and subtype confirmation can now frequently be
humans with a cumulative prevalence of 1 per
confirmed via genetic testing.
10,000. Clinically, the diseases are characterized by
Unfortunately, by the time of autopsy the disease
developmental regression, visual disturbances,
has reached its end stage so that damage to neuronal
stimulus-induced myoclonus, seizures, and early
and retinal populations is widespread. By some
death. The diseases typically manifest in the first dec-
accounts, there is predominant injury to smaller
ade of life – and the most common forms present with
cortical cells early, and injury in the retina and
visual dysfunction – although later onset forms have
damage proceeds from the photoreceptor layers out-
been identified [17]. On ophthalmologic examination,
wards. Clinical studies demonstrate retinal atrophy
most patients with NCL demonstrate significant optic
and pigmentary retinopathy that precede the devel-
atrophy and at least 70% have significant visual
opment of nerve fiber layer loss of disc atrophy, sug-
impairment at the time of death. Visual findings are
gesting the spread of disease to the inner retina
particularly prominent in patients with infantile
[19, 20]. Histopathology from retinas on autopsy in
(CLN1) and juvenile (CLN3) onset NCL, where the
Batten’s disease (now known as NCL type 3) showed
injury is early and frequently profound. By contrast
total loss of photoreceptors, including both rods and
some of the other NCL subtypes, such as late infantile
cones and extending into the outer plexiform layer.
NCL (also known as CLN2), manifest visual symp-
Pigmentary changes were less uniform and the degree
toms and presumably retinal injury later in the
of injury to the ganglion cell layer were also less
disease. Most of the other NCL subtypes have also
consistent [20]. Electron microscopy also demon-
been described to have significant retinal injury –
strated curvilinear and fingerprint bodies in the gang-
but are less common and the spectrum of disease
lion cells of NCL subjects.
has been less well described [17].
In almost all described histopathology cases of NCL
On brain imaging, NCL patients exhibit extensive
type 3 (Batten’s disease) there is accumulation of auto-
confluent subcortical white matter injury along with
fluorescent pigment in both retinal ganglion cells and
severe brain atrophy and thalamic T2 hypointensities.
cells of the inner nuclear layer. This distinguishes NCL
The pathological characteristic feature of all these
from other ophthalmologic diseases with lipofuscin
diseases is the intracellular accumulation of lipid-
such as age-related macular degeneration in which
and protein-derived pigments called ceroid. Ceroid
lipofuscin accumulation is predominantly seen in the
and lipofuscin are typically distinguished in that lipo-
retinal pigment epithelium at the outer edge of the
fuscin is lipopigment that accumulates as part of nor-
retina. Furthermore, lipopigment is not always found
mal senescence but ceroid is lipopigment that
170 in cells that are degenerating – suggesting that
accumulates as a consequence of a disease state.
 Chapter 14: Optical coherence tomography and retinal pathology in neurologic diseases

lipofuscin/ceroid accumulation is not a necessary A seminal paper in the field compared ten AD
prerequisite for cell death [17, 20]. It should be subjects with ten age-matched controls and described
noted, however, that only one autopsy case of NCL moderate levels of axonal degeneration in the optic
has been evaluated in the age of modern immuno- nerves of subjects with AD. The investigators reported
histochemistry and that additional analyses and eva- relatively selective loss of larger diameter axons in AD
luations would be appropriate especially given the subjects (although very small axons of 0.5–1 micron
variety of genetic mutations described to date – even diameter appeared to potentially be lost as well).
within a single disease causing gene. Furthermore, Reactive gliosis was described in both the retina and
given the timing of tissue acquisition neither the the optic nerve [28]. The authors described particular
precise evolution of tissue injury nor the cell types diminution of the thickness of the nerve fiber layer.
that are most susceptible to tissue injury is clear. However, they specifically describe an absence of both
Indeed, in CLN5, for example, there is not reported intraneuronal tau and amyloid beta (as assessed by
to be the “spread” of injury from the outer retina to Thioflavin S staining) [29–31]. Ultrastructural assess-
the inner retina as described for CLN3 [17]. ments in a related set of samples reported a vacuolated
Furthermore, the number of retinas that have been frothy appearance to RGC cytoplasm. Using addi-
pathologically described in the literature is small and tional stains the investigators determined that in one
a full characterization of the nature and extent of case 20% of RGCs appeared to be actively degenerat-
retinal injury remains to be performed. ing [29]. Although the bulk of the literature in this
area has suggested that inner retinal atrophy is detect-
Alzheimer’s disease able in AD, at least one well-documented series was
unable to demonstrate any loss of RGCs or axonal loss
Alzheimer’s disease (AD) is the most common dement-
in the retina [32].
ing illness of the elderly and accounts for more than half
The presence of amyloid beta fibrils in the retinae
of dementia cases with pathological confirmation at
of AD patients is controversial and unclear. Most
autopsy. AD is pathologically characterized by neuron
investigators have been unable to find amyloid depos-
loss in specific areas of the brain as well as the presence
its in the retinae of patients with AD. Age-related
of intraneuronal inclusions of hyperphosphorylated tau
accumulation of amyloid beta has been described in
in vulnerable cell populations (known as neurofibrillary
the normal human and mouse retina [33], but
tangles) and plaques of amyloid beta (neuritic plaques).
only one group has reported the capacity to detect
These amyloid plaques constitute the other pathological
human retinal amyloid deposition in AD subjects
hallmark of AD and are composed of sheets of beta-
[34, 35]. Others have found no evidence for either
amyloid peptides (these peptides in turn are 36–43
amyloid plaques or hyperphosphorylated tau in the
amino acids in length and are principally described by
AD retina [36].
the peptide length detected).
In animal models of AD (mice genetically
Patients with AD have been reported to manifest
manipulated to recapitulate one or some of the
visual dysfunction [21, 22], although the extent of
major genetic defects associated with nonsporadic
these visual deficits has been difficult to characterize
AD) pathological descriptions have suggested that
because of patients concomitant alterations in mem-
Abeta deposition can be detected in the inner retina
ory and behavior. There have been consistent reports
around blood vessels, and that this may be associated
documenting that patients with AD demonstrate
with ganglion cell loss and thinning of the retinal
significant inner retinal atrophy, including descrip-
nerve fiber [34, 37]. This observation requires further
tions of disc pallor and nerve fiber loss documented
replication and analysis.
with fundus photos and OCT [23–27]. Although most
of the OCT studies were done on older technology the
general impression has been that thinning predomi- Friedreich’s ataxia
nantly involves the arcuate bundles rather than the Friedreich’s ataxia (FA) is an autosomal recessive
papillomacular bundle. However, these studies have neurodegenerative disease caused by excessive GAA
generally lacked pathological confirmation of diagno- repeats in intron 1 of the frataxin gene on chromo-
sis (and predated the use of amyloid PET-based some 9. This leads to increased susceptibility to oxi-
imaging) and the total numbers of patients studied dative stress and mitochondrial dysfunction induced
has been small. by aberrant iron influx. The disease is characterized 171
 Chapter 14: Optical coherence tomography and retinal pathology in neurologic diseases

by progressive leg weakness with ataxia and gait Despite interest in the field, and the recognition of
disturbance as well as loss of deep tendon reflexes visual functional deficits believed to be retinal in ori-
and dysarthria. Up to 30% of patients with FA have gin in subjects with IPD, there are no published
been noted to manifest significant optic neuropathy descriptions of retinal histopathology in patients
[38, 39]. In addition, much less frequently patients with IPD [36, 41]. A single pathological series includ-
exhibit progressive pigmentary retinopathy that ing three patients has been the subject of two papers in
appears similar to standard retinitis pigmentosa, patients with another major synucleinopathy –
and a different subset of patients has been reported dementia with Lewy bodies. This analysis reported
to have more fuliminant and dramatic visual decline the identification of pale intracellular inclusions in
[38]. Although nearly all patients have evidence of the outer plexiform layer and cytoskeletal disorgani-
injury to the anterior visual pathway, including thin- zation with enlarged rod inner segments. These inclu-
ning of nerve fiber, very few have significant visual sions were not synuclein, but their molecular
symptoms. To date, there are no reported cases of structure was not identified [42].
pathological evaluation of the retina in patients
with FA [38, 39]. Huntington’s disease
Huntington’s disease (HD) is a relentlessly progres-
Parkinson’s disease and sive monogenic autosomal dominant hereditary neu-
rodegenerative disease caused by expanded
synucleinopathies trinucleotide repeats in the gene encoding the protein
Idiopathic Parkinson’s disease (IPD) is the second huntingtin (Htt) on chromosome 4. The expanded
most common neurodegenerative disease in the CAG repeat gives rise to an excessively long polyglu-
industrialized world and is pathologically manifested tamine region in mutant Htt. Disease onset is most
by injury to dopaminergic cells of the central nervous typically in middle age and characterized by defects in
system – most characteristically in the substantia motor function and psychiatric disturbances followed
nigra pars compacta of the midbrain. However, injury by the ultimate development of dementia and death.
to the dopaminergic system is widespread in the dis- Principal injury occurs to the indirect pathway of the
ease and includes areas of the brainstem and cerebral striatum including the caudate nucleus. A study of
hemispheres. retinal function in living HD patients showed higher
Induction of Parkinsonism in mice with the thresholds to blue light following adaptation to yellow
administration of 1-methyl-4-phenyl– 1,2,3,6- light, suggesting deficient neural processing in the
tetrahydropyridine (MPTP) produces a model of dis- retina compared to both normal controls as well as
ease by selectively targeting dopaminergic function in with patients with Tourette’s syndrome and schizo-
the central nervous system. Use of MPTP causes a phrenia [43].
reversible but dose-dependent reduction in tyrosine Murine and drosophila models of HD exhibit clear
hydroxylase in amacrine cells. In the absence of evidence of retinal degeneration as part of the disease
tyrosine hydroxylase, neurons cannot convert phenotype. The R6/1 mouse model expresses exon 1
L-Tyrosine into L-DOPA (which is itself a necessary of the HD gene with 115 CAG repeats driven by the
precursor for dopamine production) [40]. Decreases human HD promoter. Similarly, the R6/2 model
in retinal dopaminergic amacrine function might be expresses the same exon but with 150 repeats (and
predicted to result in failure of center surround antag- thereby has accelerated and more severe disease). The
onism in ganglion cells. Patients with IPD are known R6/1 model shows cone dysfunction and injury start-
to have deficits detectable on flash electroretinograms ing around the same age as other manifestations of
and pattern electroretinograms – especially mani- disease in the brain. Ultimately, these mice showed
fested in tests done with a high spatial frequency – as complete loss of cone opsin and transducin. The more
would be predicted by injury to amacrine cells. severe R6/2 model demonstrates significant injury on
However, TH immunostaining in human subjects light and electron microscopy affecting the entire
with IPD has been reported to be normal. retina, especially concentrated in the center [44].
Furthermore, intracellular aggregations of alpha- The drosophila model of HD also has evidence of
synuclein have been reported to be absent in the photoreceptor neuron degeneration [45]. Only a sin-
172 retina [41]. gle subject with HD appears to have ever undergone
 Chapter 14: Optical coherence tomography and retinal pathology in neurologic diseases

detailed ophthalmic pathological analysis. The inves- evaluation and demonstrate profound loss of retinal
tigators failed to find any clear retinal injury, but ganglion cells and their axons of the temporal nerve
examination of the photoreceptor layer, which is fiber and replacement of normal RNFL with dense
affected in the drosophila model, was limited by tech- collagenous tissue, fibroblasts and glial cells [48, 49].
nical difficulties because it peeled off the retina during Methanol toxicity is associated with a more precipi-
processing [45]. tous decline in vision but with the same pattern of
injury – absent the gliosis [50]. Given the rarity of
such fulminant cases in the industrialized world, addi-
Spinocerebellar ataxia type VII tional pathological descriptions to confirm these find-
Spincerebellar ataxia type VII is an autosomal domi- ings are unlikely. However, the degree to which
nant progressive neurological syndrome character- ganglion cell loss was accompanied by fibrotic
ized by gait ataxia, dysarthria, ophthalmoplegia, response is an important caveat when considering
dysphagia pyramidal impairment, and visual loss. It retinal imaging findings.
is caused by a variable but expanded CAG (polyglu-
tamine) repeat near the N-terminal end of the Ataxin-
7 gene. As with other polyglutamine repeat disorders, Aging
the disease manifests genetic anticipation with repeat In studying neurodegenerative disease, aging is both
length inversely proportional to age of disease onset. an important parameter for understanding the pro-
Most patients ultimately progress to complete blind- gression of the underlying disease and a confounder.
ness. It has been reported that impairment of yellow– There are changes in tissue architecture, cellular
blue color vision may precede the onset of the disease appearance, and tissue volume with normal aging.
by many years and may be one of the earliest signs of However, these diseases often manifest atrophy and
disease [46]. cellular changes that reflect accelerated aging. It is,
Descriptions of pathological injury in SCA-VII are therefore, important to consider carefully the impact
sparse. However, one series described detailed retinal of aging on any pathological retinal specimen evalu-
histopathology from five members of a single family ated – especially as these samples in neurological
and described severe inner and outer retinal abnorm- disease almost uniformly are assessed at the end of
alities, including near complete loss of photoreceptors life when pathological injury is most profound.
and ganglion cells. In addition, there was RPE degen- The total number of retinal neurons declines with
eration and abnormal presence of pigment detected age including loss of photoreceptors and retinal gang-
into the retina [46]. lion cells [51, 52]. Ubiquitin and alpha synuclein were
both observed with high frequency in subjects older
than age 80 and were principally found in the inner
Toxic and metabolic states nuclear layer [53]. Tau as a cytoskeletal protein with
Vitamin B12 deficiency is manifest by loss of both high affinity for microtubules is found in highest
central nervous system and peripheral nervous system concentration in the terminal end of developing
myelin (in addition to macrocytosis of erythrocytes). In axons and is involved in determining axon polarity
the retina, B12 deficiency is characterized by loss of and is required for normal axon development. It is,
retinal ganglion cells in the macula and thinning of the therefore, not surprising that unphosphorlyated tau
temporal RNFL. There are no human cases of retinal can be found in the human retina. In one study,
pathology in subjects with documented B12 deficiency. diffuse staining for unphosphorylated tau was
On pathology, loss of axons, particularly in the tem- observed in the inner plexiform layer of the retina in
poral retina and retinal ganglion cells, is readily man- all subjects, including many with diseases having no
ifest in monkeys with induced cobalamin deficiency relation to neurodegeneration; however, phosphory-
[47]. Other animal models of B12 deficiency are afove- lated tau was not detected. Unphosphorylated tau was
ate and the retina has not been studied. also observed to a lesser extent in photoreceptors [53].
In general, patients with nutritional deficiencies
have appeared to follow the same pattern with pre-
dominant injury to the papillomacular bundle. Summary
Isolated cases of subjects with nutritional (tobacco) In total, retinal pathology is a common feature of
amblyopia have undergone retinal pathological neurodegenerative diseases of the central nervous 173
 Chapter 14: Optical coherence tomography and retinal pathology in neurologic diseases

system. However, the pathological mechanisms 13. Evangelou N, Konz D, Esiri MM, et al. Size-selective
underlying this injury remain elusive given the neuronal changes in the anterior optic pathways sug-
absence of conclusive histopathological evaluations gest a differential susceptibility to injury in multiple
sclerosis. Brain 2001;124:1813–20.
in a wide variety of patients – with a few exceptions.
Additional pathological evaluations in these diseases 14. Schlamp CL, Montgomery AD, Mac Nair CE, et al.
Evaluation of the percentage of ganglion cells in the
will help with understanding findings from the bur-
ganglion cell layer of the rodent retina. Mol Vis June
geoning field of retinal imaging and visual electrophy- 27, 2013;19:1387–96.
siology in the neurology clinic. Further studies in
15. Wang L, Cioffi GA, Cull G, Dong J, Fortune B.
patients and animal models of disease, especially Immunohistologic evidence for retinal glial cell
when done with correlative imaging, will help changes in human glaucoma. Invest Ophthalmol Vis Sci
advance the field significantly and resolve lingering April 2002;43(4):1088–94.
controversies. 16. Liu S, Li ZW, Weinreb RN, Xu G, et al. Tracking retinal
microgliosis in models of retinal ganglion cell damage.
References Invest Ophthalmol Vis Sci September 2012;53
1. Green, AJ. Mcquaid S, Hauser SL, et al Ocular pathol- (10):6254–62.
ogy in multiple sclerosis: retinal atrophy and inflam- 17. Anderson GW, Goebel HH, Simonati A. Human
mation irrespective of disease duration. Brain 2010; pathology in NCL. Biochim Biophys Acta November
133 (Pt 6): 1591–1601. 2013;1832(11):1807–26.
2. Oppenheim H. Zur pathologie der disseminirten 18. Seehafer SS, Pearce DA. You say lipofuscin, we say
sklerose. Berliner Klinische Wochenschrift 1887; ceroid: defining autofluorescent storage material.
48:904–7. Neurobiol Aging April 2006;27(4):576–88.
3. Lisch K. Die Veranderung der peripheren schbahn bei 19. Bensaoula T, Shibuya H, Katz ML, Smith JE, Johnson
Mulipler Sklerose Arch f. Augehn 1933;107:380. GS, John SK, Milam AH. Histopathologic and immu-
4. Gartner S. Optic neuropathy in multiple sclerosis; nocytochemical analysis of the retina and ocular tis-
optic neuritis. AMA Arch Ophthalmol December sues in Batten disease. Ophthalmology September
1953;50(6):718–26. 2000;107(9):1746–53.
5. Toussaint D, Périer O, Verstappen A, Bervoets S. 20. Goebel HH, Fix JD, Zeman W. The fine structure of the
Clinicopathological study of the visual pathways, eyes, retina in neuronal ceroid-lipofuscinosis. Am J
and cerebral hemispheres in 32 cases of disseminated Ophthalmol January 1974;77(1):25–39.
sclerosis. J Clin Neuroophthalmol September 1983;3 21. Cogan DG. Visual disturbances with focal progressive
(3):211–20. dementing disease. Am J Ophthalmol July 15, 1985;100
6. Kerrison JB, Flynn T, Green WR. Retinal pathologic (1): 68–72.
changes in multiple sclerosis. Retina 1994;14(5): 22. Tzekov R, Mullan M. Vision function abnormalities in
445–51. Alzheimer disease. Surv Ophthalmol October 22, 2013.
7. Mogensen PH. Histopathology of anterior parts of the pii: S0039–6257(13) 00237–3.
optic pathway in patients with multiple sclerose. Acta 23. Hedges TR 3rd, Perez Galves R, Speigelman D, et al.
Ophthalmol (Copenh) April 1990;68(2):218–20. Retinal nerve fiber layer abnormalities in Alzheimer’s
8. ter Braak JWG, van Herwaarden A. Ophthalmo ence- disease. Acta Ophthalmol Scand June 1996;74(3):271–5.
phalomyelitis. Klin Monatsbl Augenheilkd 1933;91: 24. Tsai CS, Ritch R, Schwartz B, Lee SS, et al. Optic nerve
316–43. head and nerve fiber layer in Alzheimer’s disease. Arch
9. Rucker CW. Sheathing of the retinal veins in multiple Ophthalmol February 1991;109(2):199–204.
sclerosis. Review of pertinent literature. Mayo Clin Proc 25. Lu Y, Li Z, Zhang X, Ming B, et al. Retinal nerve fiber
1972;47:335–40. layer structure abnormalities in early Alzheimer’s dis-
10. Toussaint D. Perivenous sheathing in multiple sclero- ease: evidence in optical coherence tomography.
sis. Bull Soc Belge Ophtalmol. November 1983;208 Pt Neurosci Lett August 9, 2010 Aug 9;480(1):69–72.
1:369–74. 26. Kesler A, Vakhapova V, Korczyn AD, et al. Retinal
11 Sanders MD. Retinal arteritis, retinal vasculitis and thickness in patients with mild cognitive impairment
autoimmune retinal vasculitis. Eye 1987;1:441–65 and Alzheimer’s disease. Clin Neurol Neurosurg
September 2011;113(7):523–6.
12. Arnold AC, Pepose JS, Hepler RS, Foos RY. Retinal
periphlebitis and retinitis in multiple sclerosis. I. 27. Berisha F, Feke GT, Trempe CL, et al. Retinal
Pathologic characteristics. Ophthalmology 1984; abnormalities in early Alzheimer’s disease. Invest
174 91:255–62. Ophthalmol Vis Sci May 2007;48(5):2285–9.
 Chapter 14: Optical coherence tomography and retinal pathology in neurologic diseases

28. Hinton DR, Sadun AA, Blanks JC, Miller CA. Optic- hydroxylase-containing retinal amacrine cells. Brain
nerve degeneration in Alzheimer’s disease. N Engl J Res September 10, 1990;527(1):21–31.
Med August 21, 1986;315(8):485–7. 41. Archibald NK, Clarke MP, Mosimann UP, Burn DJ.
29. Blanks JC, Torigoe Y, Hinton DR, Blanks RH. Retinal The retina in Parkinson’s disease. Brain May 2009;132
degeneration in the macula of patients with (Pt 5):1128–45.
Alzheimer’s disease. Ann N Y Acad Sci 1991;640:44–6. 42. Maurage CA, Ruchoux MM, de Vos R, et al. Retinal
30. Blanks JC, Torigoe Y, Hinton DR, Blanks RH. Retinal involvement in dementia with Lewy bodies: a clue to
pathology in Alzheimer’s disease. I. Ganglion cell loss hallucinations? Ann Neurol October 2003;54(4):542–7.
in foveal/parafoveal retina. Neurobiol Aging May-June 43. Paulus W, Schwarz G, Werner A, et al. Impairment of
1996;17(3):377–84. retinal increment thresholds in Huntington’s disease.
31. Blanks JC, Schmidt SY, Torigoe Y, et al. Retinal Ann Neurol October 1993;34(4):574–8
pathology in Alzheimer’s disease. II. Regional neuron 44. Petrasch-Parwez E, Saft C, Schlichting A, et al. Is the
loss and glial changes in GCL. Neurobiol Aging. May- retina affected in Huntington disease? Acta
June 1996;17(3):385–95. Neuropathol November 2005;110(5):523–5.
32. Curcio CA, Drucker DN. Retinal ganglion cells in 45. Jackson GR, Salecker I, Dong X, et al Polyglutamine-
Alzheimer’s disease and aging. Ann Neurol March expanded human huntingtin transgenes induce
1993;33(3):248–57. degeneration of Drosophila photoreceptor neurons.
33. Hoh Kam J, Lenassi E, Jeffery G. Viewing ageing eyes: Neuron September 1998;21(3):633–42.
diverse sites of amyloid Beta accumulation in the age- 46. Martin JJ, Van Regemorter N, Krols L, et al. On an
ing mouse retina and the up-regulation of macro- autosomal dominant form of retinal-cerebellar degen-
phages. PLoS One October 1, 2010;5(10). eration: an autopsy study of five patients in one family.
34. Koronyo-Hamaoui M, Koronyo Y, Ljubimov AV, Acta Neuropathol 1994;88(4):277–86.
et al. Identification of amyloid plaques in retinas 47. Agamanolis DP, Chester EM, Victor M, et al.
from Alzheimer’s patients and noninvasive in vivo Neuropathology of experimental vitamin B12 defi-
optical imaging of retinal plaques in a mouse ciency in monkeys. Neurology October 1976;26
model. Neuroimage January 2011; 54 Suppl 1: (10):905–14.
S204–17.
48. Smiddy WE, Green WR. Nutritional amblyopia.
35. Koronyo Y, Salumbides BC, Black KL, Koronyo- A histopathologic study with retrospective clinical cor-
Hamaoui M. Alzheimer’s disease in the retina: imaging relation. Graefes Arch Clin Exp Ophthalmol 1987;225
retinal aβ plaques for early diagnosis and therapy (5):321–4.
assessment. Neurodegener Dis 2012;10(1–4):285–93.
49. Victor M, Dreyfus PM. Tobacco-alcohol amblyopia.
36. Ho CY, Troncoso JC, Knox D, et al. Beta-amyloid, Further comments on its pathology. Arch Ophthalmol.
phospho-tau and alpha-synuclein deposits similar to November 1965;74(5):649–57.
those in the brain are not identified in the eyes of
50. Sharpe JA, Hostovsky M, Bilbao JM, Rewcastle NB.
Alzheimer’s and Parkinson’s disease patients. Brain
Methanol optic neuropathy: a histopathological study.
Pathol January 2014;24(1):25–32.
Neurology October 1982;32(10):1093–100.
37. Alexandrov PN, Pogue A, Bhattacharjee S, Lukiw WJ.
51. Curcio CA. Photoreceptor topography in ageing
Retinal amyloid peptides and complement factor H in
and age-related maculopathy. Eye(Lond) June 2001;15
transgenic models of Alzheimer’s disease. Neuroreport
(Pt 3):376–83.
August 24, 2011;22(12):623–7
52. Pitts DG. The effects of aging on selected visual
38. Fortuna F, Barboni P, Liguori R, et al. Visual system
functions: dark adaptation, visual acuity,
involvement in patients with Friedreich’s ataxia. Brain
stereopsis and brightness contrast. In: Sekuler R,
January 2009;132(Pt 1):116–23.
Kline DW, Dismukes K, editors. Aging and
39. Noval S, Contreras I, Sanz-Gallego I, et al. Ophthalmic Human Visual Function. New York: 1982;
features of Friedreich ataxia. Eye (Lond) February p. 131–159.
2012;26(2):315–20.
53. Leger F, Fernagut PO, Canron MH, et al. Protein
40. Tatton WG, Kwan MM, Verrier MC, et al. MPTP aggregation in the aging retina. J Neuropathol Exp
produces reversible disappearance of tyrosine Neurol January 2011;70(1):63–8.

175
 Chapter
Retinal inflammation in multiple sclerosis

15 revealed by optical coherence tomography

and ophthalmoscopy
Elena H. Martinez-Lapiscina, Bernardo Sanchez-Dalmau, and Pablo
Villoslada

Retinal abnormalities associated with multiple frequently found around veins of the retinal periphery,
sclerosis (MS) are receiving increasing interest. The which cannot be reached by OCT, it is highly recom-
most significant are retinal vasculitis (periphlebitis) mended to assess it using pupil dilatation and indirect
and pars planitis. Retinal vasculitis represents a or slit-lamp ophthalmoscopy. However, retinal periph-
group of diseases characterized by inflammation lebitis is frequently undiagnosed in the clinical setting.
affecting the retinal vasculature. They represent a Although some authors have tried to improve diagno-
group of uncommon disorders that may occur as sis by using retinal angiography in order to confirm the
an isolated disease or more commonly in association presence of retinal vessel leakage (typical of active
with other ocular diseases or a variety of systemic retinal periphlebitis), angiography does not increase
diseases, including MS. Table 15.1 lists the different the identification of retinal periphlebitis [1]. The
conditions that can be associated with retinal vascu- ultra-wide-field retinal imaging devices that provide a
litis. However, some other findings include presence high-resolution image of up to 200 degrees of the retina
of edema in different layers of the retina associated in a single capture may help to identify retinal periph-
with different processes. lebitis, allowing also a wide field fluorescein angiogra-
phy (Figure 15.1C).
Retinal periphlebitis Most patients with retinal vascular sheathing have
Retinal periphlebitis is an inflammatory disorder of no visual complaints referable to the condition, but
the retina characterized by focal or diffuse perivascu- some patients develop recurrent episodes that may
lar inflammatory infiltrates around retinal vessels. eventually become associated with severe retinal dis-
The pathology may be located anywhere on the retina, ease that ultimately results in retinal ischemia, retinal
the peripheral retina being the most commonly detachment, or other retinal dysfunction with conse-
affected site. Retinal periphlebitis affects retinal veins quent permanent visual loss.
with normal retinal arteries and is asymptomatic. It is The etiological process underlying retinal
characterized by retinal venous sheathing (Rucker´s periphlebitis remains unclear. Retinal periphlebitis
sign), especially on sites of arterio-venous crossover affects multiple veins with different evolution times
and focal perivenous exudation and hemorrhage and often results in relapses. This course is typical
(Figure 15.1A-B). of immune-mediated disorders. Additionally, the
Retinal periphlebitis can be present in two forms. presence of perivascular inflammatory infiltrate
The first one is active periphlebitis, with venous sheath- with mononuclear cells composed of lymphocytes,
ing, which may be focal or diffuse, and perivenous plasmacytes, and monocytes [2, 3] also argues in
exudation and hemorrhage. The second is inactive favor of an immune-mediated disorder. Moreover,
periphlebitis, with vascular sclerosis and retinal pig- retinal periphlebitis is associated with auto-immune
mentary changes around the vein. The diagnosis of disorders such as sarcoidosis [4], Behçet’s disease
retinal periphlebitis is based on the retinal findings by [5], lupus erythematous [6], and mainly with MS
ophthalmoscopy. Since retinal periphlebitis is most [1, 2, 7–9].

176 Optical Coherence Tomography in Neurological Disease, ed. Peter A. Calabresi, Laura J. Balcer, and Elliot M. Frohman.
Published by Cambridge University Press. © Cambridge University Press 2015.
 Chapter 15: Retinal inflammation in MS revealed by optical coherence tomography and ophthalmoscopy

Table 15.1

Infectious disorders Bacterial: tuberculosis, syphilis, Lyme disease, Whipple’s disease, brucellosis, cat scratch disease,
endophthalmitis, post-streptococcal syndrome, Mediterranean spotted fever, Rocky
Mountain spotted fever
Viral: human T cell lymphoma virus type 1, cytomegalovirus, herpes simplex virus, varicella zoster
virus, Epstein-Barr virus, Rift Valley fever virus, hepatitis, acquired immunodeficiency
syndrome, West Nile virus infection, dengue fever virus
Parasitic: Toxoplasmosis
Neurologic disorders Multiple sclerosis
Microangiopathy of the brain, retina, and cochlea (Susac’s syndrome)
Malignancy Paraneoplastic syndromes
Ocular lymphoma
Acute leukemia
Systemic inflammatory Behçet’s disease, sarcoidosis, systemic lupus erythematosus, Wegener’s granulomatosis, polyar-
diseases teritis nodosa, Churg-Strauss syndrome, relapsing polychondritis, rheumatoid arthritis, HLA-B27-
associated uveitis, Crohn’s disease, postvaccination, dermatomyositis, Takayasu’s disease,
Buerger’s disease, polymyositis
Primary Ocular disorder Frosted branch angiitis, Idiopathic retinal vasculitis, aneurysms, and neuroretinitis (IRVAN), acute
multifocal hemorrhagic retinal vasculitis, idiopathic recurrent branch retinal arterial occlusion,
pars planitis, birdshot chorioretinopathy.

Retinal periphlebitis in multiple sclerosis quadrant [1]. However, the small sample size and the
relatively low incidence rate of retinal periphlebitis
Retinal periphlebitis can be found in 6–36% of patients
during the study period may precluded our ability to
with MS [1, 7–11]. These differences in prevalence
establish a significant association. Moreover, retinal
rates may be explained by several factors related to
periphlebitis might damage retinal layers other than
MS subtype (relapsing-remitting or progressive
the RNFL.
forms), disease duration, time of monitoring of MS
Although the clinical impact of the retinal periph-
patients, and use of immunomodulatory therapy, as
lebitis on visual function in MS patients might not be
well as the instruments used for retinal ascertaining
relevant, its pathogenic significance or its use as a
periphlebitis (direct ophthalmoscopy with or without
biomarker is raising the interest in this finding. The
mydriasis or ultra-wide-field retinal imaging devices).
perivascular inflammatory infiltrate with lymphocytes
Retinal periphlebitis in patients with MS may cause
and monocytes [2, 3] is nearly indistinguishable from
blurred vision [10] or a very mild decrease in visual
the histopathological signature of perivenular cuffs of
acuity [9], but most frequently this condition is asymp-
inflammation within the central nervous system
tomatic. Retinal periphlebitis most often affects the
(CNS) of patients with MS, particularly in the cerebral
peripheral retina without relevant damage to the
white matter. Thus, the pathological substrate of both
macula responsible for the most detailed central vision.
disorders is similar. This raises two questions. The
This may explain why retinal periphlebitis associated
first one is a theoretical issue. Myelin is the putative
with MS is rarely accompanied by clinical abnormal-
principal target of immune activation in MS, and the
ities. However, it is still unknown whether retinal per-
retina lacks myelin. However, the retina is a common
iphlebitis may induce local damage in the retinal tissue.
site of inflammation in MS [2]. Therefore, it will be
We have analyzed the effect of retinal periphlebitis in a
important to evaluate whether the immune response
given quadrant (temporal, nasal, superior, or inferior)
in MS patients with retinal periphlebitis differs from
and the corresponding retinal nerve fiber layer (RNFL)
that of patients without it. The second question is a
thickness in that quadrant [1]. The presence of retinal
practical issue. The similarities in the pathological
periphlebitis in a given quadrant was not associated
processes associated with MS and periphlebitis raise
with a significantly thinner RNFL in the corresponding 177
the possibility that retinal inflammation parallels the
 Chapter 15: Retinal inflammation in MS revealed by optical coherence tomography and ophthalmoscopy

A Besides inflammation, MS causes neurodegeneration,

which determines disease severity and impacts quality
of life. Identifying appropriate biomarkers of axonal
damage that can be easily assessed at the bedside is a
key point for the development of neuroprotective
therapies. Inflammation and neurodegeneration are
directly correlated in MS [12]. Thus, retinal periph-
lebitis may be regarded also as biomarker of neurode-
generative damage in MS. We have recently shown
that patients with retinal periphlebitis had a trend
toward a higher EDSS score at baseline and acceler-
B
ated disability progression after one year of follow-up
when compared to patients without retinal inflamma-
tion [13]. Moreover, these patients had higher lesion
volume and reduced T1 normalized brain parenchy-
mal volume and normalized gray matter volume than
patients without it. In comparison with MS patients
without retinal inflammation, patients with retinal
periphlebitis had also a thinner RNFL and smaller
macular volumes, both surrogate retinal imaging
markers of axonal damage [13].
C
In conclusion, retinal periphlebitis is associated
with both the inflammatory and neurodegenerative
processes in MS in terms of clinical parameters
(relapses and disability), surrogate markers of inflam-
mation (gadolinium-enhancing lesions, increased
CSF IgG index, and intrathecal IgG synthesis rate),
and features characteristic of neurodegeneration
(lesion load, brain atrophy, and reduced RNFL thick-
ness, and MV) (Table 15.2).

Figure 15.1 Retinal periphlebitis: (A) wide field retinography

Edema of retinal layers in multiple sclerosis
showing active retinal periphlebitis associated with MS; (B) non-active The development of spectral-domain (SD) OCT
retinal periphlebitis, with pigmentary changes around the vein; (C) devices has allowed studying in great detail the retina
wide field fluorescein angiography in a case of diffuse periphlebitis.
of patients with MS. The possibility of segmenting the
different layers of the retina provides the opportunity
inflammatory activity in the CNS in patients with MS for quantifying differences in layer thickness and
as a part of the overall brain inflammatory activity. identifying subtle abnormalities related with the
We have shown previously that patients with retinal inflammatory process associated with MS. Based on
periphlebitis had a 54% increased risk of new relapses these enhanced capabilites, recent reports have
[OR=1.52 95% IC (1.36–1.64)] in comparison with described the presence of retinal edema, which can
those without retinal inflammation [1]. Retinal per- be localized to the inner nuclear layer, within the
iphlebitis is associated with both imaging and mole- RNFL or in a pattern consistent with macular edema
cular biomarkers of disease activity, since it has been (e.g., as associated with fingolimod therapy) [15].
reported that MS patients with retinal periphlebitis
have a higher gadolinium-enhancing lesion load [1, 9] 1. Microcystic Macular edema
and higher values of IgG index and intrathecal IgG Microcystic macular edema (MMO) refers to the pre-
synthesis [9] compared with patients without it. sence of multiple retinal microcysts, preferentially
Consequently, retinal periphlebitis may become a sui- involving the inner nuclear layer of retina, which
178 table biomarker of inflammatory activity in MS. can be revealed with SD-OCT [14] (Figure 15.2).
 Chapter 15: Retinal inflammation in MS revealed by optical coherence tomography and ophthalmoscopy

Table 15.2 Association between retinal periphlebitis and inflammatory and neurodegenerative parameters in multiple sclerosis

Parameters Reference Conclusion

Relapses Sepulcre et al. RP is a risk factor of new Relapses over 2-years follow-up OR = 1.52
Neurology 95% CI (1.36–1.64) p = 0.002
2007;68(7):544–49
MRI gadolinium- Sepulcre et al. Patients with RP had larger gadolinium-enhancing lesion volume
enhancing lesions Neurology than patients without RP at baseline p = 0.003
2007;68(7):544–49
Stamenkoviæ M Patients with RP had gadolinium-enhancing lesions more com-
et al. monly than patients without RP at baseline p = 0.012
Vojnosanit Pregl
2011;68(7):544–9

CSF IgG index and Stamenkoviæ M Patients with RP had increased values of IgG Index (1.39 ±0.42) and
intrathecal IgG et al. intrathecal IgG synthesis (26 ±20.3) than patients without RP [IgG
synthesis Vojnosanit Pregl Index:0.82 ±0.49 Intrathecal IgG synthesis: 9.6 ±13] p <0.001
2011;68(7):544–9
Disability Sepulcre et al. RP is not associated with disability progression over two-year follow-up.
Neurology
2007;68(7):544–49

Ortiz Pérez S et al. Patients with RP had a nonsignificant increased EDSS score at base-
Neurology 2013;81 line [2.38 95%IC (1.44–3.32)] and EDSS progression after one-year
(10):877-81 follow-up [0.34 95%IC(0–0.75)] compared to patients without RP:
EDSS score at baseline [1.87 95%IC (1.61–2.12)] and EDSS progression
[0.17 95%IC(0–0.29)]
MRI total lesion volume Ortiz Pérez S et al.
Patients with RP had higher lesion volume at baseline [19.0 95%IC (9.3–
Neurology 2013;81
28.8)] compared to patients without RP [8.6 95%IC (6.0–11.3)] p = 0.038
(10):877-81
MRI brain atrophy Ortiz Pérez S et al. Patients with RP had lower normalized brain parenchymal volume
Neurology 2013;81 [1481 95% IC (1410–1552)]compared to patients without RP [1550
(10):877-81 95%IC (1530–1569)] p = 0.059
Patients with RP had lower normalized gray matter volume [772
95%IC (733–811)] compared to patients without RP [806 95%IC
(796–817)] p = 0.085
OCT RNFL thickness Ortiz Pérez S et al. Patients with RP had thinner RNFL [79.2 95%IC (68.2–90.2)] com-
Neurology 2013;81 pared to patients without RP [92.6 95%IC (89.6–95.5)] p = 0.018
(10):877-81
OCT macular volume Ortiz Pérez S et al. Patients with RP had a nonsignificant smaller MV [8.43 95%IC (8.07–
Neurology 2013;81 8.80)] compared to patients without RP [8.55 95%IC (8.45–8.64)] p =
(10):877-81 0.563
RP: Retinal periphlebitis; MRI: Magnetic resonance imaging; CSF: cerebrospinal fluid; EDSS: Expanded disability status scale; OCT:
optical coherence tomography; RNFL: Retinal nerve fiber layer; MV: Macular volume.

This retinal abnormality was first identified as a com- or in those treated with fingolimod, as well as in a
plication of uveitis in MS patients [16]. However, variety of other diseases that affect the visual pathways
recent studies have shown that MMO can be found [13, 17–19]. Eyes with MMO had higher impaired
in about 0.5–6% of patients with MS without uveitis visual acuity than eyes without MMO [17, 18].
179
 Chapter 15: Retinal inflammation in MS revealed by optical coherence tomography and ophthalmoscopy

Figure 15.2 Microcystic macular edema associated

with multiple sclerosis. A) OCT (Spectralis®) of the macula:
the green arrow indicates the position of the OCT scans;
B) Microcystic macular edema associated with Multiple
Sclerosis: MMO is seen as small, round, and empty spaces
in the inner nuclear layer: the red stars indicate the inner
plexiform – retinal ganglion cell complex just above the
cysts.

(a) (b)

Visual complaints are more frequently found in and with neurodegenerative activity in terms of dis-
patients with MMO than in those with retinal periph- ability progression over two years of follow-up [17].
lebitis. However, eyes with MMO have thinner RNFL MMO has been identified in 26% patients with
than eyes without it and MMO occurs more com- neuromyelitis optica [20]. The pathogenic mechan-
monly in eyes with prior optic neuritis [17, 18]. isms explaining MMO in patients with neuromyelitis
Further, visual acuity remained significantly more optica are not completely understood, but they may
abnormal among patients with MMO than those be similar to those related to MS. MMO in patients
patients without MMO, even after adjustment for with neuromyelitis optica is associated with higher
potential confounders such as age, sex, disease dura- visual disability and more severe retinal axonal
tion, and prior optic neuritis [17, 18]. The pathogenic damage regardless of prior optic neuritis episodes.
mechanism underlying MMO in MS is not clear and Therefore, the presence of MMO may be a marker of
different mechanisms may contribute to MMO such poor visual outcome among patients with neuromye-
us Müller cell dysfunction [19], and blood-retinal litis optica [20].
barrier disruption and retinal inflammation [17, 18].
Clinically, MMO can be monitored by SD-OCT 2. RNFL edema due to interferon
and this has led to the suggestion that MMO could
be used as a suitable biomarker of MS disease activity treatment
in a subset of patients. MMO was associated with Ikebe and collegues described the first case of retinal
disease severity in an MS population investigated hemorrhages and cotton-wool spots following inter-
using cross-sectional analyses [13, 17, 18], whereas feron-alfa or -beta treatment in 1990 [21]. Since then,
such an association was not identified when the ophthalmological complications due to interferon
patients were followed with a longitudinal strategy administration have been reported especially among
[17]. MMO is detected only among a small proportion patients with hepatitis [22], but also among patients
of patients with MS (3–6%) and this low incidence with MS [23–25]. The main fundoscopic sign of
may limit its value as a biomarker of disease activity. retinopathy due to interferon is cotton-wool spots
This hurdle can be overcome considering the follow- (soft retinal exudates) [22]. SD-OCT has identified
ing hypothesis postulated by Saidha and colleagues this finding as focal retinal nerve fiber layer edema
[17]. The increased thickness of the inner nuclear with subsequent focal retinal thickening over the
layer demonstrated by SD-OCT in MS patients may follow-up [26] (Figure 15.3). Cotton-wool spots repre-
be a minor form of MMO. Retinal inflammation may sent retinal swellings resulting from focal occlusion of
cause thickening of the inner nuclear layer in the retinal arterioles that caused ischemic damage of the
absence of visible MMO [17]. These authors showed retinal nerve fiber layer. Cotton-wool spots are a com-
that inner nuclear layer thickness was associated with mon finding in patients with hypertension and diabetes
both inflammatory activity in terms of relapses and mellitus, two well-recognized causes of ischemic
180 gadolinium-enhancing lesions and new T2 lesions, microvascular damage by atherosclerosis. Interferon
 Chapter 15: Retinal inflammation in MS revealed by optical coherence tomography and ophthalmoscopy

A
C

200 µm

Figure 15.3 RNFL edema associated with interferon-beta treatment in MS. (A) OCT (Spectralis®): the black line indicates the position of the
OCT scans. The white arrow points the RNFL edema as a grayish lesion in the funduscopic image. (B) The black star indicates the RNFL edema in
OCT. Observe the RNFL thickening next to the focal edema. (C) SD-OCT monitoring of RNFL defect performed monthly over four-month follow-
up. Notice the progressive resolution of the RNFL thickening.

can induce the production of autoantibodies leading to a monitoring even in absence of visual symptomatology.
local inflammatory infiltrate with deposition of cyto- More importantly, SD-OCT evaluates not only the OCT
kines, immune complexes, and leucocytes that finally correlate of these retinal soft exudates but also the sub-
occludes small retinal vessels leading to a focal retinal sequent RNFL thickening that might be relevant for MS
microinfarct [27–30]. RNFL defects due to interferon patients.
treatment may result as immune-mediated microvascu-
lar damage but the specific pathogenic mechanism of References
damage is not completely understood. Visual acuity 1. Sepulcre J, Murie-Fernandez M, Salinas-Alaman A,
impairment is the most common complaint found Garcia-Layana A, Bejarano B, Villoslada P. Diagnostic
among patients with cotton-wool spots [22]. accuracy of retinal abnormalities in predicting disease
Improvement of visual complaints, fundoscopic exam, activity in MS. Neurology May 1, 2007;68(18):1488–94.
and SD-OCT abnormalities has been reported after 2. Green AJ, McQuaid S, Hauser SL, Allen IV, Lyness R.
interferon discontinuation [24] and also in patients Ocular pathology in multiple sclerosis: retinal atrophy
and inflammation irrespective of disease duration.
that continue with this therapy [23]. However, the Brain June 2010;133(Pt 6):1591–601.
patient described by Sommer and colleagues suffered
3. Engell T. Neurological disease activity in multiple
from new bilateral cotton-wool spots over the follow- sclerosis patients with periphlebitis retinae. Acta
up period, which finally disappeared after interferon Neurologica Scandinavica February 1986;73(2):168–72.
disruption [23]. There is not enough evidence to deter- 4. Herbort CP, Rao NA, Mochizuki M. International
mine whether treatment should be permanently discon- criteria for the diagnosis of ocular sarcoidosis: results
tinued or not. This interferon-associated retinopathy is a of the first International Workshop On Ocular
dynamic process, so if interferon is not discontinued, Sarcoidosis (IWOS). Ocular Immunology and
SD-OCT offers the possibility of a strict ophthalmologic Inflammation May–June 2009;17(3):160–9. 181
 Chapter 15: Retinal inflammation in MS revealed by optical coherence tomography and ophthalmoscopy

5. Mili-Boussen I, Letaief I, Zbiba W, Trabelsi O, Ben 17. Saidha S, Sotirchos ES, Ibrahim MA, Crainiceanu CM,
Younes N, Abid J, et al. [Retinal vasculitis. Gelfand JM, Sepah YJ, et al. Microcystic macular
Epidemiological, clinical and etiological features]. edema, thickness of the inner nuclear layer of the
Journal Francais d’Ophtalmologie October 2010;33 retina, and disease characteristics in multiple sclerosis:
(8):529–37. Les vascularites retiniennes: profils epide- a retrospective study. Lancet Neurol November
miologiques, cliniques et etiologiques. 2012;11(11):963–72.
6. Quillen DA, Stathopoulos NA, Blankenship GW, 18. Gelfand JM, Nolan R, Schwartz DM, Graves J, Green
Ferriss JA. Lupus associated frosted branch periphlebi- AJ. Microcystic macular edema in multiple sclerosis is
tis and exudative maculopathy. Retina (Philadelphia, associated with disease severity. Brain. June 2012;135
Pa) 1997;17(5):449–51. (Pt 6):1786–93.
7. Rio J, Colin A, Salvador F, Tintore M, Viguera ML, 19. Balk LJ, Killestein J, Polman CH, Uitdehaag BM,
Montalban J, et al. [Retinal periphlebitis in multiple Petzold A. Microcystic macular edema confirmed, but
sclerosis. A prospective study]. Neurologia (Barcelona, not specific for multiple sclerosis. Brain December
Spain). October 1993;8(8):252–5. 2012 Dec;135(Pt 12):e226; author reply e7.
8. Tola MR, Granieri E, Casetta I, Monari P, Scorrano R, 20. Sotirchos ES, Saidha S, Byraiah G, Mealy MA,
Mazzeo V, et al. Retinal periphlebitis in multiple Ibrahim MA, Sepah YJ, et al. In vivo identification
sclerosis: a marker of disease activity? European of morphologic retinal abnormalities in neuromye-
Neurology 1993;33(2):93–6. litis optica. Neurology April 9, 2013 Apr 9;80
9. Stamenkovic M, Obradovic D. Retinal periphlebitis in (15):1406–14.
patients with multiple sclerosis. Vojnosanitetski Pregled 21. Ikebe T, Nakatsuka K, Goto M, Sakai Y, Kageyama S. A
Military – medical and pharmaceutical review July case of retinopathy induced by intravenous administra-
2011;68(7):544–9. tion of interferon. Folia Ophthalmol Jpn (Ganka-Kiyo).
10. Schmidt S, Wessels L, Augustin A, Klockgether T. 1990;41:6.
Patients with multiple sclerosis and concomitant 22. Fragoso YD, Paggiaro MS, Mastromauro R, Jacondino
uveitis/periphlebitis retinae are not distinct from those Gda S, Wilson HM. Literature systematic review on the
without intraocular inflammation. Journal of the ophthalmological side effects of interferons. Arquivos
Neurological Sciences June 15, 2001;187(1–2):49–53. Brasileiros de Oftalmologia. July–August 2011;74
11. Quentin CD, Altenhoff M, Schipper HI, Vogel M. (4):306–10.
[Retinal periphlebitis and multiple sclerosis. Incidence, 23. Sommer S, Sablon JC, Zaoui M, Rozot P, Hosni A.
cerebrospinal fluid and results of visual evoked poten- [Interferon beta-1 b retinopathy during a treatment for
tials]. Fortschritte der Ophthalmologie : Zeitschrift der multiple sclerosis]. Journal Francais d’Ophtalmologie.
Deutschen Ophthalmologischen Gesellschaft 1990;87 May 2001;24(5):509–12.
(4):359–61. 24. Mallada-Frechin J, Abellan-Miralles I, Alfaro-Beltra
12. Barkhof F, Calabresi PA, Miller DH, Reingold SC. ML, Medrano V, Munoz-Gil MB, Fernandez-
Imaging outcomes for neuroprotection and repair in Izquierdo S, et al. [Retinopathy secondary to treatment
multiple sclerosis trials. Nature reviews Neurology May with Interferon beta-1a in a patient with multiple
2009;5(5):256–66. sclerosis]. Revista de Neurologia. April 16–30,
13. Ortiz-Perez S M-LE, Gabilondo I, Fraga-Pumar E, 2005;40(8):482–4.
Martínez-Heras E, Saiz A, Sánchez-Dalmau B, Villoslada 25. Hara K, Tanito M, Ohira A. Retinal nerve fibre layer
P. Retinal periphlebitis is associated with multiple sclerosis defects associated with cotton-wool spots in patients
severity. Neurology 2013;81(10):877–81. with interferon retinopathy. Acta Ophthalmologica.
14. Mackenzie S, Schmermer C, Charnley A, Sim D, Vikas March 2012;90(2):e158–60.
T, Dumskyj M, et al. SDOCT imaging to identify 26. Ioannides A, Georgakarakos ND, Elaroud I,
macular pathology in patients diagnosed with diabetic Andreou P. Isolated cotton-wool spots of unknown
maculopathy by a digital photographic retinal screen- etiology: management and sequential spectral
ing programme. PloS One 2011;6(5):e14811. domain optical coherence tomography documenta-
15. Kappos L, Radue EW, O’Connor P, Polman C, tion. Clinical Ophthalmology (Auckland, NZ).
Hohlfeld R, Calabresi P, et al. A placebo-controlled 2011;5:1431–3.
trial of oral fingolimod in relapsing multiple sclerosis. 27. Taylor JL, Grossberg SE. The effects of interferon-
The New England Journal of Medicine. February 4, alpha on the production and action of other cytokines.
2010;362(5):387–401. Seminars in Oncology. Febrary 1998;25(1 Suppl 1):
16. Malinowski SM, Pulido JS, Folk JC. Long-term visual 23–9.
outcome and complications associated with pars pla- 28. Sugano S, Suzuki T, Watanabe M, Ohe K, Ishii K,
182 nitis. Ophthalmology June 1993;100(6):818–24. Okajima T. Retinal complications and plasma C5a
 Chapter 15: Retinal inflammation in MS revealed by optical coherence tomography and ophthalmoscopy

levels during interferon alpha therapy for chronic interferon alfa-2 b treatment for malignant melanoma.
hepatitis C. The American Journal of Gastroenterology. Lancet. April 17, 1999;353(9161):1326.
December 1998;93(12):2441–4. PubMed PMID: 30. Guyer DR, Tiedeman J, Yannuzzi LA, Slakter JS,
9860406. Epub 1998/12/22. eng. Parke D, Kelley J, et al. Interferon-associated reti-
29. Lohmann CP, Kroher G, Bogenrieder T, Spiegel D, nopathy. Archives of Ophthalmology. March
Preuner J. Severe loss of vision during adjuvant 1993;111(3):350–6.

183
 Chapter
Optical coherence tomography and optic

16 nerve magnetic resonance imaging

in demyelinating diseases
Robert T. Naismith

Introduction which result in a particular vulnerability for attack [1].

Studies of optical coherence tomography (OCT) and Because MS white-matter plaques are often centered
optic nerve magnetic resonance imaging (MRI) pro- upon blood vessels, the dual arterial supply, high
vide opportunities to cross-validate very distinct tools metabolic demands of the retina, and prominent
to assess visual pathway integrity. MRI of visual path- venous drainage may contribute to potential for
ways provides distinct advantages for testing quanti- blood brain barrier breakdown within the optic
tative sequences and relating these to other measures nerve. The posterior visual pathways are also fre-
of clinical function, structure, and physiology. These quently affected in MS due to close proximity of
techniques are not in competition, as they can offer optic radiations to the ventricles, a prime location
complementary information. Selection of one or both for the classic periventricular plaque.
techniques would depend upon the investigative Regulatory agencies that approve disease-mod-
question, the timing of study, and other design fea- ifying therapies are most interested in surrogate
tures, and the advantages and disadvantages of each measures with high clinical relevance toward dis-
imaging method as it relates to a particular patient ability and patient reported outcomes. Because
population. Indeed, OCT has clear advantages in imaging events are much more frequent than
acquisition time, equipment expense, operator exper- relapses and disability, MRI has been transforma-
tise, and patient convenience. One major advantage tional for evaluating new therapeutics over a rea-
for MR sequences is that, once validated on visual sonable time frame for phase I and phase II studies.
systems, they can then be implemented within other MRI metrics in corroboration with standard clin-
regions of the central nervous system directly hidden ical endpoints (i.e., relapses, sustained disability
from view by OCT (e.g., spinal cord, brain). progression) have been of major importance in
phase III trials of MS therapeutics. Because the
Optic nerve imaging and the clinical- visual pathways represent a well-circumscribed
structural system that is frequently involved in MS
radiologic paradox and have clinical outcome measures that relate to
The optic nerve has major relevance in multiple quality of life, and with little redundancy or con-
sclerosis (MS) as a prime site for clinical and subcli- founding purpose, the optic nerve MRI serves as a
nical inflammatory demyelinating events that result model system for investigating how imaging tech-
in axon injury and disability. Optic “nerve” is a mis- niques correspond with clinical dysfunction. The
nomer, because the structure is a direct “tract” of the relation of imaging to a clinically meaningful
central nervous system and has myelin derived from event and quality of life is paramount for regulatory
oligodendrocytes. While the propensity for the optic agencies to consider approval of new MS medica-
nerve to become involved in MS is clear, the rationale tions. This relation has been elusive for MRI
for the predominance of optic nerve involvement in sequences in clinical practice but, based on optic
MS is not. Perhaps the proclivity is conferred by a nerve imaging, it can be improved upon with more
combination of vascular and immunologic features, specialized sequences.

184 Optical Coherence Tomography in Neurological Disease, ed. Peter A. Calabresi, Laura J. Balcer, and Elliot M. Frohman.
Published by Cambridge University Press. © Cambridge University Press 2015.
Chapter 16: Optical coherence tomography and optic nerve magnetic resonance

Figure 16.1 Axial optic nerve slice represent-

ing T2-weighted and diffusion-weighted
sequences. (A) Optic nerve FLAIR MRI demon-
strates increased T2 signal and swelling within
the left nerve. (B) B0 image used for generating
region of interest (yellow) consisting of 22 voxels
within the nerve center. (C) Fractional anisotropy
map demonstrating high directionality of the
optic nerve on the left (right optic nerve cut by
plane of slice).
 Chapter 16: Optical coherence tomography and optic nerve magnetic resonance

Optic nerve MRI has high sensitivity and speci- of blood–brain barrier breakdown, frequently seen in
ficity for diagnosing optic neuritis and evaluating the acute setting of optic neuritis [5, 6].
for other etiologies in clinical practice. The addition T1-weighted sequences after administration of
of optic nerve MRI to select clinical trials utilizing gadolinium reveals optic nerve enhancement with a
OCT and other visual outcomes can help support a sensitivity of 66–94%, typically lasting 1–2 months
treatment effect, if all measures are beneficially [7]. While optic nerves with more extensive enhance-
affected in a positive manner [2, 3]. Specialized ment had worse vision acutely, the association of final
MR sequences may have potential to evaluate clinical outcome to enhancement length and presence
acute axon injury, susceptibility to axonal loss, the of enhancement in the intracanicular portion of
composition of inflammatory infiltrates, and the the optic nerve has conflicted among investigations
potential for remyelination and efficacy of neural [7–10]. Owing to the high sensitivity for contrast
repair (Figure 6). enhancement in the setting of acute optic neuritis,
the absence of enhancement should at least make
one consider an alternate diagnosis, such as ischemic
Challenges of optic nerve MRI optic neuropathy [11].
Optic nerve measures 3–4 mm in diameter in a healthy T2-weighted signal abnormalities can be
subject, less in someone with long-standing MS due to observed in both the acute and chronic setting of
atrophy. High spatial resolution is a requisite for deriv- optic neuritis. The length of chronic optic nerve
ing reliable imaging measures of the optic nerve. The hyperintensity has been demonstrated to correlate
optic nerve is surrounded by different imaging inter- with thinning of the retinal nerve fiber layer
faces, such as cerebral spinal fluid, orbital fat, air within (RNFL) by OCT [12]. However, MRI hyperintense
sinuses, and bone. These different tissue interfaces lesion length did not appear to correlate with clinical
create potential for imaging artifacts to negatively outcomes. The decision to treat optic neuritis with
impact scan quality. Additionally, optic nerve imaging anti-inflammatory glucocorticoids should be made
is impacted by eye or head movement, both of which on clinical and individual patient circumstances, as
can degrade image quality. Even arterial pulsations intravenous methylprednisolone did appear to
may negatively impact high-resolution quantitative shrink the resultant length of MRI hyperintense
techniques. The fast acquisition times and immediate T2-signal within the optic nerve [13].
processing of OCT does have an advantage for patient
optimization and high-quality scans.
Optic nerve imaging in clinical practice is utilized
Optic nerve volumes
frequently and with a high degree of success. Volumetric measures of the CNS can serve as a sur-
However, investigational techniques require high- rogate of axonal loss and perhaps demyelination.
resolution and high signal-to-noise ratio (SNR), two Unfortunately, tissue volume measures also involve
qualities that require time and increase the chances a number of confounding pathologic processes such
for excessive movement. As the resolution increases, as edema, cellular infiltration, and gliosis [14–16]. The
so must the scanning time and propensity to move- time course of optic nerve volume changes has rele-
ment and other potential artifacts. vance for interpreting studies of OCT with regard to
timing. Due to the inflammation and edema, the optic
nerve can have prominent swelling during the acute
Optic nerve MRI in clinical practice period, with a mean 20% expansion over control
Standard optic nerve imaging includes high-resolu- nerve, and remain swollen for up to 60 days [17].
tion sequences in axial, sagittal, and coronal views. One year after optic neuritis, the nerve becomes
T2-weighted short tau inversion recovery (STIR) atrophic by 14% compared to control.
sequence is utilized to assess the presence of abnormal While the more severely swollen optic nerves had
T2 hyperintense signal within the nerve, can provide worse baseline visual function, neither the amount of
information about whether the nerve is swollen or swelling nor the amount of atrophy predicted clinical
atrophic, and can assess for other impinging or infil- visual outcomes at one year. Similar to what was
trating structures on the optic nerve [4]. A fat-sup- observed with intravenous methylprednisolone and
pressed T1-weighted sequence after administration chronic T2-weighted lesion length, glucocorticoids
186 of Gadolinium contrast can assess for the presence did not appear to impact the degree of optic nerve
 Chapter 16: Optical coherence tomography and optic nerve magnetic resonance

atrophy when administered in the acute setting [18]. demyelination and tissue integrity [24–27]. DTI sum-
This implies that anti-inflammatory agents have little mary parameters include mean diffusivity (lambda1 +
effect on axonal injury, even when administered lambda2 + lambda 3/3) and fractional anisotropy, a
within days of optic neuritis clinical onset [19]. This scalar value of a directional ellipsoid between zero to
also illustrates how the visual pathway can provide one. Axial diffusivity and fractional anisotropy both
important information about MS pathogenesis and imply a principal direction in their measure.
treatment effect (or lack thereof). In remote optic neuritis, radial diffusivity and
fractional anisotropy of the optic nerve have estab-
lished strong correlations with RNFL by OCT, high-
Optic nerve diffusion tensor imaging and low-contrast vision, and visual-evoked potentials
Diffusion tensor imaging (DTI) is a quantitative tech- (Figure 16.2) [26, 28–31]. Similarly, FA and radial
nique that can assess tissue directionality and integ- diffusivity of the optic radiations have been demon-
rity. Because the visual system is a tightly packed, strated to correlate with RNFL, multifocal VEPs, and
structured tract of axons and myelin, DTI can be visual acuity [31–33]. These studies help to cross-
utilized to define the location and trajectory of the validate OCT and DTI because they hold high con-
visual pathways. Diffusion imaging is based upon the cordance among all outcome measures despite asses-
Brownian motion of protons, which will diffuse along sing two different areas of the optic nerve, with vastly
the trajectory of the axon. Mean diffusion will increase different techniques and methodologies.
when there is loss of tissue. In contrast to remote optic neuritis studies pre-
DTI is acquired in six or more directions, and the viously discussed, acute optic neuritis must be inter-
largest direction of diffusion within a voxel is called the preted through a different set of DTI parameters,
principal eigenvector. The principal eigenvector is underscoring the importance of clinical correlation
referred to as lambda-1. By convention, eigenvectors and a temporal evaluation. After acute optic neuritis,
perpendicular to lambda –1 include lambda-2 and axial diffusivity has been found to be depressed for
lambda-3. From these eigenvectors, one can derive one to three months, followed by a period of apparent
several interrelated DTI parameters to summarize and “normalization” around six months, and finally eleva-
describe the diffusion within a specified voxel. Lambda- tion of axial diffusivity after one year [34, 35, http://
1, or the principal eigenvector, is sometimes referred to journals.plos.org/plosone/article?id=10.1371/journal.
as axial diffusivity and has been proposed to be a surro- pone.0083825]. The precise etiology for this decrease
gate for axonal injury during acute inflammation [20– is not clear, but an explanation can include axonal
23]. Radial diffusivity is the mean of the two perpendi- swelling and beading as a result from mitochondrial
cular diffusivities (lambda-2 plus lambda-3 divided by dysfunction, cytotoxic edema, and aggregation of
2) and has been proposed as a marker of both intracellular proteins with increased intracellular

Figure 16.2 Correlation of optic nerve radial diffusivity

2.00 with retinal nerve fiber layer thickness. Retinal nerve fiber
R2 linear = 0.579
layer thickness has a striking correlation with radial dif-
fusivity within the optic nerve. The correlation cross-
1.75
Radial diffusivity (µm2/ms)

validates the techniques for assessing tissue integrity

within the anterior visual pathway when also considering
1.50 the robust relationship to clinical outcomes for each
methodology. Radial diffusivity measures included
1–2 mm of the optic nerve, a location contiguous but
1.25 separated in space from the RNFL.
(Reproduced with permission from Naismith et al,
Neurology, 2009)
1.00

0.75

40.00 60.00 80.00 100.00 120.00

RNFL by optical coherence tomography (microns)
187
 Chapter 16: Optical coherence tomography and optic nerve magnetic resonance

viscosity and impaired fast transport through micro- trend towards increasing MTR after 6 months has
tubules. This initial depression in axial diffusivity has been hypothesized to represent remyelination post
been found to correspond to long-term outcomes of inflammation, and these post-6 month changes have
RNFL by OCT, high- and low-contrast sensitivities, been found to correspond to less volume loss in 12
and visual evoked potentials at 6 and 12 months. The month RNFL compared with those whose MTR con-
other DTI parameters, mean and radial diffusivities, tinued to have a late decline [40].
and fractional anisotropy, do not appear to convey
this acute alteration. Three months after optic neuritis Conclusions
onset, radial diffusivity begins to increase and reaches
Optic nerve MRI opens new opportunities to test
a plateau between 6 to 12 months. This increase in
novel imaging techniques and provide critical struc-
radial diffusivity has also been shown to correspond to
tural and metabolic information about the central
RNFL and visual function parameters. The increases
nervous system in demyelinating diseases. One of the
in axial and radial diffusivities between three and six
major goals of MR imaging is to develop a sequence
months might be explained by axonal drop-out and
with pathologic specificity to axon injury, myelin
loss of tissue integrity, both of which would result in
loss, and inflammation. Incorporation of OCT and
an increased total diffusivity within the affected voxel.
VEPs with imaging helps to validate the degree
The MS lesions thus have a variety of pathologies
to which the MRI technique discerns axonal and
which can change over time.
myelin pathologies. While the perfect technique
for accomplishing this specificity has not been
Optic nerve magnetization transfer definitively established, optic nerve imaging has
imaging most certainly brought the MS field closer to this
Magnetization transfer imaging (MTI) is a quantita- realization. Similarly, OCT in MS seeks to meet
tive technique proposed for the assessment of tissue some of these needs to study inflammation and
macromolecules such as myelin. The technique uses neurodegeneration.
an off-resonance radiofrequency pulse to saturate
macromolecular protons rather that the standard References
pulse that saturates protons within the liquid pool.
1 Schmidt S. Candidate autoantigens in multiple sclero-
This hydrogen saturation is transient as the energy is sis. Multiple Sclerosis 1999; 5: 147–60.
transferred from the immobile macromolecular
2 Sühs K-W, Hein K, Sättler MB, et al. A randomized,
hydrogen pool to the free-water hydrogen pool, thus double-blind, phase 2 study of erythropoietin in optic
decreasing the free water signal in a predictable man- neuritis. Annals of Neurology 2012; 72: 199–210.
ner. By evaluating the ratio (MTR) of the saturated 3 Connick P, Kolappan M, Crawley C, et al. Autologous
and the unsaturated sequences, one can obtain an mesenchymal stem cells for the treatment of secondary
indirect measure of macromolecular concentration progressive multiple sclerosis: an open-label phase 2a
and, thus, myelin [36]. MTR is beginning to be uti- proof-of-concept study. The Lancet Neurology 2012; 11:
lized to assess therapies designed to preserve myelin 150–6.
or enhance remyelination in MS. 4 Miller DH, MacManus DG, Bartlett PA, Kapoor R,
On a cross-sectional basis in remote optic neuritis, Morrissey SP, Moseley IF. Detection of optic nerve
optic nerve MTR has correlated with RNFL thickness lesions in optic neuritis using frequency-selective fat-
saturation sequences. Neuroradiology 1993; 35: 156–8.
and macular volumes by OCT, and visual evoked
potential latency and amplitude [37, 38]. In longitu- 5 Guy J, Mao J, Bidgood Jr WD, Mancuso A, Quisling
RG. Enhancement and demyelination of the intraorbi-
dinal studies, MTR does not appear to be substantially
tal optic nerve: Fat suppression magnetic resonance
impacted in acute optic neuritis [39, 40]. Rather, MTR imaging. Ophthalmology 1992; 99: 713–9.
appears to decrease between 3 and 6 months after
6 Guy J, Mancuso A, Quisling RG, Beck R, Moster M.
optic neuritis clinical onset and, at that time, has Gadolinium-DTPA-enhanced magnetic resonance
been found to correlate with 12-month RNFL by imaging in optic neuropathies. Ophthalmology 1990;
OCT, high- and low-contrast visual acuity, and multi- 97: 592–9.
focal VEP. Subsequent to 6 months, MTR can appear 7 Kupersmith MJ, Alban T, Zeiffer B, Lefton D. Contrast-
188 to increase and improve for some patients, remain enhanced MRI in acute optic neuritis: Relationship to
stable, or continue to drop through 12 months. The visual performance. Brain 2002; 125: 812–22.
 Chapter 16: Optical coherence tomography and optic nerve magnetic resonance

8 Dunker S, Wiegand W. Prognostic value of magnetic the Experimental Autoimmune Encephalomyelitis

resonance imaging in monosymptomatic optic neuritis. Spinal Cord: A Quantitative Pixelwise Analysis. The
Ophthalmology 1996; 103: 1768–73. Journal of Neuroscience 2009; 29: 2805–13.
9 Miller DH, Newton MR, Van der Poel JC, et al. 22 Song SK, Sun SW, Ju WK, Lin SJ, Cross AH,
Magnetic resonance imaging of the optic nerve in optic Neufeld AH. Diffusion tensor imaging detects and
neuritis. Neurology 1988; 38: 175–9. differentiates axon and myelin degeneration in mouse
10 Hickman SJ, Toosy AT, Miszkiel KA, et al. Visual optic nerve after retinal ischemia. Neuroimage 2003;
recovery following acute optic neuritis. J Neurol 2004; 20: 1714–22.
251: 996–1005. 23 Naismith RT, Xu J, Tutlam NT, Lancia S, Trinkaus K,
11 Rizzo Iii JF, Andreoli CM, Rabinov JD. Use of magnetic Song S-K, Cross AH. Diffusion tensor imaging in acute
resonance imaging to differentiate optic neuritis and optic neuropathies: Predictor of clinical outcomes.
nonarteritic anterior ischemic optic neuropathy. Arch Neurol 2011: archneurol.2011.243.
Ophthalmology 2002; 109: 1679–84. 24 Song S-K, Sun S-W, Ramsbottom MJ, Chang C, Russell
12 Kallenbach K, Simonsen H, Sander B, Wanscher B, J, Cross AH. Dysmyelination revealed through MRI as
Larsson H, Larsen M, Frederiksen JL. Retinal increased radial (but unchanged axial) diffusion of
nerve fiber layer thickness is associated with water. Neuroimage 2002; 17: 1429–36.
lesion length in acute optic neuritis. Neurology 25 Song S-K, Yoshino J, Le TQ, Lin S-J, Sun S-W, Cross
2010; 74: 252–8. AH, Armstrong RC. Demyelination increases radial
13 Kapoor R, Miller DH, Jones SJ, et al. Effects of intra- diffusivity in corpus callosum of mouse brain.
venous methylprednisolone on outcome in MRI-based Neuroimage 2005; 26: 132–40.
prognostic subgroups in acute optic neuritis. Neurology 26 Naismith RT, Xu J, Tutlam NT, Trinkaus K, Cross AH,
1998; 50: 230–7. Song S-K. Radial diffusivity in remote optic neuritis
14 Frisén L, Quigley HA. Visual acuity in optic atrophy: A discriminates visual outcomes. Neurology 2010; 74:
quantitative clinicopathological analysis. Graefe’s Arch 1702–10.
Clin Exp Ophthalmol 1984; 222: 71–4. 27 Naismith RT, Xu J, Tutlam NT, et al. Increased
15 Hickman SJ, Brex PA, Brierley CMH, et al. Detection diffusivity in acute multiple sclerosis lesions predicts
of optic nerve atrophy following a single episode of - risk of black hole. Neurology 2010; 74: 1694–701.
unilateral optic neuritis by MRI using a fat-saturated 28 Kolbe SC, Marriott M, van der Walt A, et al. Diffusion
short-echo fast FLAIR sequence. Neuroradiology 2001; tensor imaging correlates of visual impairment in
43: 123–8. multiple sclerosis and chronic optic neuritis.
16 Hickman SJ, Brierley CM, Brex PA, MacManus DG, Investigative Ophthalmology & Visual Science 2012; 53:
Scolding NJ, Compston DA, Miller DH. Continuing 825–32.
optic nerve atrophy following optic neuritis: A serial 29 Smith SA, Williams ZR, Ratchford JN, et al. Diffusion
MRI study. Multiple Sclerosis 2002; 8: 339–42. tensor imaging of the optic nerve in multiple sclero-
17 Hickman SJ, Toosy AT, Jones SJ, et al. A serial MRI sis: Association with retinal damage and visual dis-
study following optic nerve mean area in acute optic ability. American Journal of Neuroradiology 2011; 32:
neuritis. Brain 2004; 127: 2498–505. 1662–8.
18 Hickman SJ, Kapoor R, Jones SJ, Altmann DR, Plant 30 Dasenbrock HH, Smith SA, Ozturk A, Farrell SK,
GT, Miller DH. Corticosteroids do not prevent optic Calabresi PA, Reich DS. Diffusion tensor imaging of the
nerve atrophy following optic neuritis. Journal of optic tracts in multiple sclerosis: Association with ret-
Neurology, Neurosurgery & Psychiatry 2003; 74: inal thinning and visual disability. Journal of
1139–41. Neuroimaging 2011; 21: e41-e9.
19 Beck RW, Cleary PA, Anderson MM, et al. A 31 Li M, Li J, He H, et al. Directional diffusivity
Randomized, Controlled Trial of Corticosteroids in the changes in the optic nerve and optic radiation in
Treatment of Acute Optic Neuritis. New England optic neuritis. British Journal of Radiology 2011; 84:
Journal of Medicine 1992; 326: 581–8. 304–14.
20 Budde MD, Kim JH, Liang HF, Russell JH, Cross AH, 32 Reich DS, Smith SA, Gordon-Lipkin EM, et al. Damage
Song SK. Axonal injury detected by in vivo diffusion to the optic radiation in multiple sclerosis is associated
tensor imaging correlates with neurological disability with retinal injury and visual disability. Archives of
in a mouse model of multiple sclerosis. NMR Biomed Neurology 2009; 66: 998–1006.
2008; 21: 589–97. 33 Kolbe S, Bajraszewski C, Chapman C, et al. Diffusion
21 Budde MD, Xie M, Cross AH, Song S-K. Axial tensor imaging of the optic radiations after optic
Diffusivity Is the Primary Correlate of Axonal Injury in neuritis. Hum Brain Mapp 2012; 33: 2047–61. 189
 Chapter 16: Optical coherence tomography and optic nerve magnetic resonance

34 Naismith RT, Xu J, Tutlam NT, et al. Disability in optic loss and demyelination in optic neuritis. Multiple
neuritis correlates with diffusion tensor-derived direc- Sclerosis 2007; 13: 875–9.
tional diffusivities. Neurology 2009; 72: 589–94. 38 Klistorner A, Chaganti J, Garrick R, Moffat K,
35 Naismith RT, Xu J, Tutlam NT, et al. Diffusion tensor Yiannikas C. Magnetisation transfer ratio in optic
imaging in acute optic neuropathies: Predictor of clin- neuritis is associated with axonal loss, but not
ical outcomes. Archives of Neurology 2012; 69: 65–71. with demyelination. Neuroimage 2011; 56: 21–6.
36 Inglese M, Ghezzi A, Bianchi S, et al. Irreversible dis- 39 Hickman SJ, Toosy AT, Jones SJ, et al. Serial magneti-
ability and tissue loss in multiple sclerosis: A conven- zation transfer imaging in acute optic neuritis. Brain
tional and magnetization transfer magnetic resonance 2004; 127: 692–700.
imaging study of the optic nerves. Archives of Neurology 40 Wang Y, van der Walt A, Paine M, et al. Optic Nerve
2002; 59: 250–5. Magnetisation Transfer Ratio after Acute Optic
37 Trip S, Schlottmann P, Jones S, et al. Optic nerve Neuritis Predicts Axonal and Visual Outcomes. PLoS
magnetization transfer imaging and measures of axonal One 2012; 7: e52291.

190
 Chapter
Optical coherence tomography in neurologic

17 clinical trials
Robert A. Bermel and Peter K. Kaiser

Introduction widespread availability of the equipment in either

ophthalmology or neurology departments of most
Clinical trials of any new therapeutic agent depend on
academic medical centers, excellent reproducibility,
sensitive indices of disease activity or progression to
and the advantage of rapid, painless testing without
detect benefit. Surrogate measures, which ideally are
any significant medical risks [2–4].
directly linked to the mechanism of disease, are not
available for many neurological diseases. In multiple
sclerosis (MS) clinical trials, MRI markers of inflam- Possible designs for trials
matory disease activity have been crucial to rapid Trials in neurological disease may be designed with
acceleration in development of therapies that impact OCT as either a primary or key secondary outcome
the inflammatory mechanisms of the disease, which measure. Given the lack of an established precedent
manifest as new MRI lesions and relapses. The devel- and ineligibility for imaging measures to serve as the
opment of new imaging measures that quantify primary outcome measure of phase III trials to sup-
axonal loss or neurodeneration in MS would likely port definitive regulatory approval, the current focus
facilitate acceleration in the discovery and testing of is on phase II trial designs. Some examples of trial
new therapeutic options, especially for progressive designs which might be designed with OCT as pri-
MS. Optical coherence tomography (OCT) is mary or key secondary outcome measure include an
currently being applied and evaluated in the context acute neuroprotection design, in acute optic neuritis,
of neurologic clinical trials to fill that need. and a chronic neuroprotection design, in secondary
progressive MS (SPMS) (Figure 17.1).
Advantages to using OCT as an Advantages of the acute neuroprotective study
design include the relative large magnitude of pRNFL
outcome measure loss that occurs over a short period of time following
OCT is an appealing outcome measure to include in optic neuritis, which averages 10–20 uM and is largely
clinical trials, for multiple reasons. First, as the fields complete by six months after the attack [5]. Challenges
of multiple sclerosis and neurodegenerative disease for trials in acute optic neuritis include the narrow
move toward a goal of neuroprotection, having a window of opportunity for action (making recruitment
measure of unmyelinated axonal integrity (as can be difficult) and also the difficulty calculating change in
obtained from the peripapillary retinal nerve fiber pRNFL when the baseline pRNFL in the affected eye
layer [pRNFL]) is of great appeal. Second, OCT also may be increased due to pRNFL swelling. Requiring a
offers the ability to specifically examine the preserva- narrow window (<14 days) from optic neuritis onset to
tion of first-order sensory neurons, at the macula, randomization poses a challenge to enrollment but
either by total macular volume or by quantifying the maximizes the chance of starting the drug during a
ganglion cell layer (GCL) thickness. Third, OCT links theoretical window of opportunity for neuroprotection.
very intuitively to a sensitive clinical outcome within Possible solutions to the recruitment issue involve
the same pathway, namely low-contrast letter acuity inclusion of sites who have an established rapid referral
[1]. Finally, feasibility of OCT is extremely high, with network composed of optometrists, ophthalmologists,

Optical Coherence Tomography in Neurological Disease, ed. Peter A. Calabresi, Laura J. Balcer, and Elliot M. Frohman. 191
Published by Cambridge University Press. © Cambridge University Press 2015.
 Chapter 17: Optical coherence tomography in neurologic clinical trials

(A) Optic Figure 17.1 Commonly proposed

neuritis multi-center study designs that utilize OCT as
a key outcome measure for parallel-group,
Study drug (active or placebo) placebo-controlled multi-centered trials.
(A) acute optic neuritis; (B) progressive MS.

active

placebo

Baseline Month 2 Month 4 Month 6

OCT Safety
OCT OCT OCT
follow-up
(B)
active

placebo

Baseline Month 6 Month 12 Month 18 Month 24

OCT OCT OCT OCT OCT

and neuroophthalmologists, who may represent “first and derivation of these measures vary across technol-
responders” for acute optic neuritis. To combat the ogies and manufacturers. Other variables of potential
problem of pRNFL swelling, the fellow (unaffected) interest include the macular RNFL, and subsets of the
eye can be utilized to statistically control for swelling pRNFL measured in individual quadrants or specific
of the pRNFL in the affected eye, an analysis which regions such as the papillomacular bundle. Utilization
assumes that both eyes were equal without prior occur- of the latter set of measures in multi-center clinical
rence of optic neuritis in the fellow eye. This requires a trials has not been extensively explored. At present, no
prior history of optic neuritis be added to the exclusion multi-center clinical trial has conducted an OCT ana-
criteria of the trial, with the associated potential of lysis using third-party or platform-agnostic retinal
limiting the eligible population. Alternatively, para- segmentation and quantification algorithms. Thus,
meters measured at the macula, such as the GCL thick- instrument-specific algorithms are currently relied
ness, are not increased due to swelling in acute optic upon, which introduces inherent incompatibilities
neuritis and may be used as the outcome of interest in when combining data across platforms. Retinal
acute optic neuritis trials. Challenges for OCT-focused segmentation algorithms that can analyze data from
trials in SPMS include the required length of the trial different OCT scanner platforms are being developed.
(2–3 years, which is longer than is optimal for phase II A set of common data elements (CDE) is provided by
trials) given the generally slow rate of decline of OCT the U.S. National Institute for Neurological Disorders
measures in purely progressive MS [6]. and Stroke (NINDS), to guide the collection of core
OCT variables in MS clinical trials [7]. The CDE
Key variables of interest provide example source documents and key compo-
nent variables for the two most commonly utilized
For the purposes of neurological studies, both
spectral-domain OCT platforms in neurological trials
peripapillary and macular scans are obtained. Key
(Zeiss Cirrus and Heidelberg Spectralis). Beyond this,
variables of interest for OCT trials in neurological
there is no standard data dictionary agreed upon for
disease include the pRNFL, total macular volume
use when integrating OCT into clinical trials.
(TMV), and the ganglion cell/inner plexiform layer
(GCL/IPL) thickness (obtained at the macula). The
primary reason for using these measures is that they Technical and practical issues
are readily available using instrument-derived One of the factors that makes implementation of any
192 algorithms, although the scan patterns, availability, imaging modality in multi-center clinical trials a
 Chapter 17: Optical coherence tomography in neurologic clinical trials

challenge is the issue of different technologies, differ- analogous to different sites using different MRI plat-
ent manufacturers, and upgrades or changes to the forms. The same was not true of TD-OCT (Zeiss
equipment or software during the course of a trial. Stratus) values, which differ significantly from corre-
The MS clinical trials community has experience with sponding SD-OCT-measured values for pRNFL. The
these issues from years of using MRI in multi-center ability to combine the data across SD-OCT platforms
trials [8]. Technical issues with MRI have been dealt is one of the primary advantages to utilizing pRNFL as
with by restricting the technologies accepted for use a key outcome measure in large multi-center clinical
by sites in the trial (by manufacturer, magnetic field trials. For trials which restrict OCT acquisition to one
strength, etc.), restricting the outcomes and pulse platform, other platform-specific algorithm-derived
sequences utilized (advanced MRI outcomes have measures such as GCL/IPL provided by the Cirrus
been slow to move into multi-center trials for this or papillomacular bundle thickness provided by
reason), placing restrictions on equipment or soft- Spectralis may also prove to be useful outcome
ware upgrades during the trial, and employing a measures.
central reading center to ensure quality control and Specific acquisition protocols are delineated at the
uniform implementation of procedures. There are outset of the trial, and if instruments from multiple
analogous solutions that are utilized in trials driven manufacturers are being utilized in the trial, the
by OCT, and the community’s prior experience with acquisition protocol and manual of operations
MRI somewhat eases the adoption of some of these (MOOP) are designed to provide data that are as
solutions for OCT. analogous as is possible across manufacturers.
Deciding which technologies to allow in a trial A general requirement to reduce variability is that
always requires a compromise between uniformity of any one patient be scanned on the same instrument
data (highest if only a single technology or manufac- for all study visits. The “repeat scan” (co-registration
turer is included) and feasibility (more sites can be of scan pattern based on fundus photograph) feature
easily included if there are fewer restrictions). For available on most SD-OCT scanners can be employed
OCT, time-domain OCT (TD-OCT) is being rapidly to maximize reproducibility of placement of the
supplanted in the neurological community by spec- pRNFL circle. Changes to equipment or software
tral-domain OCT (SD-OCT), which is now widely during the course of a trial are discouraged.
available. Even within SD-OCT however, there are
multiple manufacturers, each of which has its own Sample sizes for trials using OCT
unique implementation of the variables of interest.
For any clinical trial, the sample size and power of the
Strategies to deal with the inclusion of partially
trial is dependent largely on the anticipated effect size
incompatible platforms include restricting certain
of the treatment of interest and the variability of the
analyses to a subset of patients scanned on a single
measurement. For trials employing OCT as an
platform, or transforming the data in a validated way
outcome measure, this can be understood as percent
to improve compatibility across platforms. Such a
preservation of the OCT variable of interest (such as
transformation has been described for total macular
pRNFL) compared to the natural history, placebo-
volume, which is calculated differently by the
treated state, or active comparator depending on the
Spectralis (restricted to the ETDRS circle) than by
trial design.
the Cirrus (calculated for the entire 6 × 6 × 2 mm
For trials in acute optic neuritis, the natural his-
acquired macular cube).[9]
tory of pRNFL loss following acute optic neuritis is
Combining pRNFL data from Cirrus and
well described, and detailed sample size calculations
Heidelberg devices in the context of a multi-center
are published for a range of statistical techniques,
clinical trial is supported by one validation study [9],
anticipated effect sizes, and statistical power levels
as long as any individual patient within the trial is
[10]. Common analyses rely on the fellow-eye-
scanned consistently on the same platform (ideally the
controlled change in pRNFL, as previously discussed.
same exact instrument) throughout the entire trial. In
Although these sample sizes were derived using data
that study, overall agreement for pRNFL between SD-
from TD-OCT, the overall results are felt likely to be
OCT platforms Cirrus and Spectralis was high,
applicable to SD-OCT, though SD-OCT affords
suggesting the two platforms could be utilized in
greater inter-scan reliability, likely lower variance,
parallel in the context of a multi-center clinical trial, 193
 Chapter 17: Optical coherence tomography in neurologic clinical trials

and therefore slightly smaller sample sizes when collection of study results could be justified as essen-
deployed in a clinical trial compared to TD-OCT tial on a logistical basis alone. Oversight by an experi-
[4]. Depending on the chosen power level, duration enced OCT reading center also confers advantages
of follow-up, and anticipated treatment effect, that result in improved data quality and usability.
between 40 and 100 participants are required per Some of these advantages were highlighted in a recent
arm to demonstrate differences in pRNFL between report that examined the difference in the quality of
groups [10]. TD-OCT data obtained in the context of a study
For trials in progressive MS, sample size estimates planned without the use of an OCT reading center,
are not well established. Quantifying the natural his- with OCT obtained at individual sites without a struc-
tory of OCT in progressive MS requires longitudinal tured protocol or central oversight compared to OCT
measurement over time, which is complicated by data obtained in the context of a similar study except
changing technology and the need for large cohorts for centrally specified and coordinated OCT data
given the heterogeneity of disease. The most robust collection [13]. Inclusion of the OCT reading center
estimates for rate of decline come from one study with preplanned OCT study protocol was associated
pooling TD-OCT data across three centers that exam- with more usable data, fewer errors, and higher signal
ined 299 patients with at least two time points spread strengths [13].
over ≥ 6 months’ time, showing pRNFL thinning that Favorable study characteristics facilitated by an
averaged 2.9 uM over 2–3 years of follow-up [11]. OCT reading center that likely improve data quality
This study excluded patients who experienced acute include implementation of a standard OCT protocol
optic neuritis during the course of follow-up, but was with well-defined quality standards, certification of
composed mostly of patients with relapsing MS, with OCT technicians, and real-time data transmission
median disease duration of nine years. Most of the with quality assurance feedback to sites [14].
patients (87%) were on standard MS disease modify- Figure 17.2 shows how an OCT reading center inte-
ing therapies. In the absence of more robust published grates with overall study architecture. Specific roles of
data on purely progressive MS, these are the most the OCT reading center throughout the course of a
analogous natural history data on which to base sam- trial are outlined in Table 17.1.
ple size calculations for progressive MS trials.
Although groups of patients with secondary progres- Differences between OCT-driven trials
sive MS (SPMS) have thinner pRNFL than those with
CIS or RRMS when evaluated in cross-sectional stu-
in ophthalmology and neurology
dies [12], this does not necessarily mean that the Because early experience with OCT in clinical trials
yearly rate of thinning is greater in SPMS. It is likely, was in the ophthalmology space, OCT instrument
based on the Talman study, that there is substantial vendors and OCT reading centers may have substan-
inter-individual variability in the rate of pRNFL thin- tial experience with ophthalmological clinical trials.
ning. Patients with shorter disease duration and more Important differences exist, however, between trials
active inflammatory characteristics such as new T2 or conducted in ophthalmology and those in neurologic
gadolinium-enhancing lesions appear to have more disease. Ophthalmology trials (especially those in ret-
rapid neurodegeneration [6]. Clinical trials may, inal disorders) rely heavily on categorical or subjective
therefore. potentially be enriched by including grading of OCT scans to classify outcomes, whereas
patients with these characteristics. These enrichment the focus in neurological trials to date is on quantita-
tactics may be particularly useful if macular GCL/IPL tive data. Although there is less of a focus on catego-
is utilized as an outcome measure, as disease activity rical grading in neurological trials (and commonly
and shorter disease duration have been shown to instrument-generated quantitative data could be
correlate with rapidity of GCL/IPL thinning more so created for each scan at the site), it is still important
than pRNFL thinning [6]. for independent raters to look at each scan in a trial to
assure accurate collection of the variables of interest,
especially if multiple scans are obtained at each visit,
Role of the OCT reading center as is common practice to maximize opportunity for
Given the complexities involved in the incorporation quality control. Optimally, each scan is evaluated by
of OCT in multi-center clinical trials, use of an OCT two or more independent raters, and if there is dis-
194 reading center to provide expertise and facilitate the agreement between the two raters that exceeds a given
 Chapter 17: Optical coherence tomography in neurologic clinical trials

Table 17.1 Potential roles of an OCT reading center

Potential Roles of an OCT Reading Center:

1. Provide input into protocol development with respect to OCT
2. Draft instrument-specific manual of operations
3. Site and technologist training, qualification and certification
4. Handling, storage, archiving of source images
5. Quantitative data management using regulatory-compliant methods
6. On-study individual scan review and grading
7. Quality assurance with real-time feedback to sites
8. Study closeout, query resolution, OCT database lock

conducted locally at the sites and not centrally coor-

Site Master
OCT dinated or aggregated. Because of the time frame in
Study
Reading which those trials were conducted, time-domain
Database
Center
OCT was the available technology. No specific OCT
Quality protocol was mandated, rather this was left up to the
feedback
treating ophthalmologist at the site. Quantitative
Incidental OCT results from this trial were not systematically
findings Manual collected, are not available in aggregate, and would
review/
have marginal utility given the lack of standardized
grading
acquisition or analysis protocol. FREEDOMS II
Figure 17.2 Interaction of the OCT reading center in the context of included oversight from the UC Davis OCT reading
multisite clinical trials of neurological disease center, and baseline OCT results from that trial were
recently reported [15]. In that multi-center trial
cohort, pRNFL and TMV thinning was common at
threshold, then a third rater will examine the scan. baseline, with 34% of the patients having a mean
Neurological trials always collect binocular data (even pRNFL value below the fifth percentile in at least
in monocular pathologies such as unilateral optic one eye compared to a normative database.
neuritis), whereas sometimes monocular data is uti- The OCTAGON trial was a randomized double-
lized in ophthalmology trials. It is important to recog- blind parallel group trial of glatiramer acetate (GA)
nize that neurological patients may not have had a vs. placebo in acute optic neuritis, conducted at multi-
dilated ophthalmoscopic exam prior to trial entry; ple sites in the United States. Change in peripapillary
therefore, the possibility of incidental ophthalmologi- RNFL over six months (measured by time-domain
cal findings is greater in neurological trials, and a OCT) was the primary outcome measure. The study
workflow must be built to effectively convey those is notable because it was the first planned acute optic
findings back to the sites. neuritis study utilizing OCT outcomes. With a
planned enrollment of 100 patients per arm, the
study was designed to have 80% power to detect a
Examples of neurologic clinical trials difference of 5 uM in baseline to six-month RNFL
utilizing OCT between groups [16]. After enrollment occurred
The first large-scale trials to utilize OCT at all were more slowly than planned, the trial was closed in late
the phase III trials of fingolimod, FREEDOMS I and 2010, with 44 patients having enrolled in total. The
FREEDOMS II. OCT was initially included in these OCTAGON study confirmed two anticipated issues
trials as a key safety outcome, due to the potential with acute optic neuritis trials driven by OCT: difficult
for fingolimod to cause macular edema. In enrollment within an acute neuroprotective window
FREEDOMS I, OCT was employed by most sites as of opportunity, and the confound of baseline pRNFL
a safety screening test and to monitor on-study for swelling (baseline pRNFL = 130.5 ±8.9 uM in the
macular edema, but scan protocol and analysis was placebo arm) [16]. 195
 Chapter 17: Optical coherence tomography in neurologic clinical trials

The first multi-center neuroprotective trial in Randomized placebo-controlled phase II clinical

acute optic neuritis to show benefit on OCT was a trials of multiple agents in acute optic neuritis are
trial of erythropoietin versus placebo as add-on to currently ongoing. A trial of phenytoin versus placebo
intravenous methylprednisolone, enrolling at three is being conducted in the United Kingdom, and plans
centers in Germany between August 2006 and for enrollment of 90 total participants within 14 days
February 2011. Measured by TD-OCT, pRNFL of onset of visual loss, followed for six months, with
decreased by a median of 7.5 μm in the erythropoietin final data collection planned in April 2014 [22]. The
group compared to a median of 16.0 μm in the pla- amiloride clinical trial in optic neuritis (ACTION)
cebo group by week 16 (p = 0.0357) [17]. Other out- study estimates enrollment of 46 participants, with
comes including retrobulbar optic nerve diameter and primary outcome being the difference in pRNFL at 6
VEP latency showed benefit of erythropoietin com- months and 12 months, with data collection antici-
pared to placebo. The trial is also notable because they pated to be complete by September 2014 [23]. A
had to screen 97 participants to include 40, likely a multi-center trial of fingolimod in acute demyelinat-
factor of the stringent requirement of a seven-day ing optic neuritis (ADON), which aimed to recruit
window from optic neuritis onset to randomization. 126 total participants with a first episode of ADON
A follow-up study, Treatment of Optic Neuritis with andrandomize to fingolimod or placebo for 18 weeks,
Erythropoetin (TONE), is planned, randomizing was planned but closed prior to completing enroll-
between erythropoietin or placebo, with pRNFL ment [24]. A randomized study of BIIB033, an anti-
measured using SD-OCT as the primary outcome LINGO monoclonal antibody, is being conducted in
measure, planning to enroll 100 participants with acute optic neuritis, with planned recruitment of 80
onset of first symptoms less than 10 days prior to total participants. Nerve conduction velocity on full-
randomization [18]. field visual evoked potential is the primary outcome
A single-center study conducted in Iran in 60 measure, although both pRNFL and GCL/IPL at week
participants with visual symptoms eight days or less 24 are key secondary outcome measures [25].
in duration showed a benefit of memantine versus Multiple other trials have incorporated OCT or
placebo on preservation of pRNFL at three months utilized it as a secondary outcome measure. In a phase
(78.9 uM in placebo group vs. 91.3 uM in memantine IIa open-label study of autologous mesenchymal stem
group, p = 0.01), with no difference in full-field visual cells in 10 patients with SPMS, there was no signifi-
evoked potentials or full-contrast visual acuity found cant change in pRNFL on follow-up compared to
between treatment groups [19]. baseline [26]. A phase II trial of riluzole in early MS
There are two notable large-scale multi-center employed TD-OCT as a secondary outcome measure
trials currently planned or in process which are found no significant difference between treatment
attempting to utilize OCT to support a neuroprotec- arms in pRNFL over 36 months in 43 participants,
tive effect of experimental therapy in progressive MS. consistent with the negative result on the primary
The Secondary and Primary Progressive Ibudilast outcome measure, percent brain volume change [27].
NeuroNext Trial in MS (SPRINT-MS) will randomize
250 patients with either primary or secondary pro- Conclusions
gressive MS to receive either ibudilast or placebo [20].
As the field of multiple sclerosis clinical trials enters
The primary outcome of the trial is whole brain atro-
into an era of testing treatments for neuroprotection,
phy measured on MRI, though pRNFL on OCT is a
OCT will play a key role as an outcome measure given
key secondary outcome measure. All 19 sites in the
its feasibility and early track record in acute optic
trial are providing OCT data on all patients. Either
neuritis of showing proof of concept for neuroproten-
Zeiss Cirrus or Heidelberg Spectralis OCT is being
tion in a short time frame.
accepted by the central OCT reading center. The
Multiple Sclerosis-Secondary Progressive Multi-Arm References
Randomisation Trial (MS-SMART) anticipates
1. Balcer LJ, Baier ML, Cohen JA, Kooijmans MF,
enrolling 440 participants randomized to amiloride, Sandrock AW, Nano-Schiavi ML, et al. Contrast
riluzole, ibudilast, or placebo and followed for two letter acuity as a visual component for the Multiple
years. OCT is a tertiary/supportive outcome measure Sclerosis Functional Composite. Neurology 2003;
196 in MS-SMART [21]. 61:1367–1373.
 Chapter 17: Optical coherence tomography in neurologic clinical trials

2. Frohman EM, Costello F, Stuve O, Calabresi P, Miller function correlations, and models for neuroprotection.
DH, Hickman SJ, et al. Modeling axonal degeneration J Neuroophthalmol 2011; 31:362–373.
within the anterior visual system: implications for 15. Winges KM, Werner JS, Harvey DJ, Cello KE, Durbin
demonstrating neuroprotection in multiple sclerosis. MK, Balcer LJ, et al. Baseline retinal nerve fiber layer
Arch Neurol 2008; 65:26–35. thickness and macular volume quantified by OCT in
3. Cettomai D, Pulicken M, Gordon-Lipkin E, Salter A, the North American Phase 3 Fingolimod Trial for
Frohman TC, Conger A, et al. Reproducibility of Relapsing-Remitting Multiple Sclerosis. Journal of
optical coherence tomography in multiple sclerosis. Neuro-Ophthalmology 2013; 33: 322–329 10.1097/
Arch Neurol 2008; 65:1218–1222. WNO.0b013e31829c51f7.
4. Syc SB, Warner CV, Hiremath GS, Farrell SK, 16. OCTAGON trial. http://clinicaltrials.gov/ct2/show/re
Ratchford JN, Conger A, et al. Reproducibility of high- sults/NCT00856635; Accessed 12/2013.
resolution optical coherence tomography in multiple 17. Suhs KW, Hein K, Sattler MB, Gorlitz A, Ciupka C,
sclerosis. Mult Scler 2010; 16:829–839. Scholz K, et al. A randomized, double-blind, phase 2
5. Costello F, Coupland S, Hodge W, Lorello GR, study of erythropoietin in optic neuritis. Ann Neurol
Koroluk J, Pan YI, et al. Quantifying axonal loss after 2012; 72:199–210.
optic neuritis with optical coherence tomography. Ann 18. Treatment of Optic Neuritis With Erythropoietin
Neurol 2006; 59:963–969. (TONE). http://clinicaltrials.gov/ct2/show/results/NC
6. Ratchford JN, Saidha S, Sotirchos ES, Oh JA, Seigo T00856635; Accessed 01/2014.
MA, Eckstein C, et al. Active MS is associated with 19. Esfahani MR, Harandi ZA, Movasat M, Nikdel M,
accelerated retinal ganglion cell/inner plexiform layer Adelpour M, Momeni A, et al. Memantine for axonal
thinning. Neurology 2013; 80:47–54. loss of optic neuritis. Graefes Arch Clin Exp
7. Multiple Sclerosis Standards – NINDS Common Data Ophthalmol 2012; 250:863–869.
Elements.; 2014. 20. Safety, Tolerability and Activity Study of Ibudilast in
8. Bermel RA, Fisher E, Cohen JA. The use of MR ima- Subjects With Progressive Multiple Sclerosis. http://cl
ging as an outcome measure in multiple sclerosis inicaltrials.gov/show/NCT01982942; Accessed 12/
clinical trials. Neuroimaging Clin N Am 2008; 18:687– 2013.
701, xi. 21. MS-SMART: Multiple Sclerosis-Secondary Progressive
9. Warner CV, Syc SB, Stankiewicz AM, Hiremath G, Multi-Arm Randomisation Trial. http://clinicaltrials.g
Farrell SK, Crainiceanu CM, et al. The impact of uti- ov/show/NCT01910259; Accessed 01/2014.
lizing different optical coherence tomography devices 22. Neuroprotection with phenytoin in optic neuritis. htt
for clinical purposes and in multiple sclerosis trials. p://clinicaltrials.gov/show/NCT01451593; Accessed
PLoS One 2011; 6:e22947. 12/2013.
10. Henderson APD, Altmann DR, Trip AS, Kallis C, 23. Amiloride clinical trial in optic neuritis.
Jones SJ, Schlottmann PG, et al. A serial study of retinal http://www.clinicaltrials.gov/show/NCT01802489;
changes following optic neuritis with sample size Accessed 12/2013.
estimates for acute neuroprotection trials. Brain 2010;
133:2592–2602. 24. Fingolimod (FTY720) in Acute Demyelinating Optic
Neuritis (ADON). http://www.clinicaltrials.gov/show/
11. Talman LS, Bisker ER, Sackel DJ, Long DA,Jr, Galetta NCT01802489; Accessed 01/2014.
KM, Ratchford JN, et al. Longitudinal study of vision
25. 215ON201 BIIB033 In Acute Optic Neuritis (AON)
and retinal nerve fiber layer thickness in multiple
(RENEW). http://clinicaltrials.gov/show/NC
sclerosis. Ann Neurol 2010; 67:749–760.
T01721161; Accessed 01/2014.
12. Costello F, Hodge W, Pan YI, Eggenberger E,
26. Connick P, Kolappan M, Crawley C, Webber DJ,
Freedman MS. Using retinal architecture to help
characterize multiple sclerosis patients. Can J Patani R, Michell AW, et al. Autologous
mesenchymal stem cells for the treatment of
Ophthalmol 2010; 45:520–526.
secondary progressive multiple sclerosis: an open-
13. Keltner JL, Cello KE, Balcer LJ, Calabresi PA, label phase 2a proof-of-concept study. Lancet Neurol
Markowitz CE, Werner JS. Stratus OCT Quality 2012; 11:150–156.
Control in Two Multi-Centre Multiple Sclerosis
27. Waubant E. A phase II trial of neuroprotection with
Clinical Trials. Neuro-Ophthalmology 2011; 35:57–64.
riluzole in early relapsing-remitting MS. ECTRIMS
14. Sakai RE, Feller DJ, Galetta KM, Galetta SL, Balcer LJ. 2013, Copenhagen, Denmark. Late Breaking News
Vision in multiple sclerosis: the story, structure- Session 10/05/2013.

197
 Chapter
Optical coherence tomography in a

18 multi-center setting: quality control issues

Axel Petzold, Laura J. Balcer, Peter A. Calabresi, Fiona Costello,
Elliot M. Frohman, Ari J. Green, Alexander Klistorner, Friedemann
Paul, Sven Schippling, and Pablo Villoslada

Introduction algorithm failures and imperfect centering of the

ring-scan around the head of the optic nerve [4]. In
The many advantages of retinal OCT imaging
some cases there will be obvious protocol violations
described in this book do not come without
such as scanning with the wrong protocol at the
attention to quality control (QC). Qualitatively
wrong retinal location. Any such scans should be
poor OCT scans may cause loss of data or wrong
recognized early and excluded if the error cannot
data due to measurement artifacts [13, 4]. This
be repaired. Still, there are more subtle pitfalls that
remains of particular concern for longitudinal
can lead to inappropriate quantitative data if left
data in a multi-center setting typical for rando-
unchecked.
mized controlled trials (RCT). Therefore, the
authors of this chapter convened in 2012 with
the aim to establish robust and transparent QC Quantitative measurement artifacts
criteria for the use of retinal OCT in multiple due to QC issues
sclerosis (MS) research and trials. To illustrate the size of the problem, the annual
In this chapter we will first review the published loss of the peripapillary retinal nerve fiber layer
evidence on the effect of poor-quality scans. Next, we (RNFL) in MS is about 1–2 μm per year. This is
will illustrate how substantial measurement artifacts over 10 times the expected 0.1 μm observed in
can be introduced by breaking a few rules with an healthy subjects [12]. It is relevant to notice that
otherwise well-acquired OCT scan. This is followed the small quantity of localized changes caused by
by a detailed description of seven QC criteria which neurodegeneration may easily be masked by mea-
permit one to identify scans that should be rejected surement artifacts. For example, a localized mea-
from analysis in a multi-center setting. For didactic surement artifact up to 40 μm can be the result of
reasons, this is followed by a number of examples off-center placement of the measurement beam
seen in an OCT reading center. For brevity, this [2], but this will not readily be visible to an
chapter predominantly deals with the peripapillary OCT reading center. Poor placement of the
ring scan, but points relevant for macular volume peripapillary ring-scan results in artifacts around
scans will also be discussed. The chapter concludes the 3.4 μm range of the global average [5]. Poor
with pragmatic advice on how to obtain high quality signal is a problematic issue, and measurement
OCT scans. artifacts in the 10 μm range can readily arise [1].
For this reason, OCT scans containing such mea-
The problem of poor-quality surement artifacts may contaminate the data and
OCT scans mask the detection of relevant retinal layer
atrophy. We are the opinion that such artifactual
Recognized reasons for poor quality OCT scans
OCT ought to be excluded from clinical studies or
are poor illumination, boundary line errors, or
trials.

198 Optical Coherence Tomography in Neurological Disease, ed. Peter A. Calabresi, Laura J. Balcer, and Elliot M. Frohman.
Published by Cambridge University Press. © Cambridge University Press 2015.
 Chapter 18: Optical coherence tomography in a multi-center setting: quality control issues

Reliable QC assessment in a 15 dB [13] based on our experience and published

evidence suggesting that poor signal strength
multi-center setting decreases the signal-to-noise ratio of RNFL measures
A set of seven quality control (QC) criteria which (Figure 18.2) [14, 8]. However, scans with poor signal
address all of the above discussed issues was validated strength (defined as < 15 dB for the Heidelberg
in a previous study [13]. These criteria could be reliably Spectralis and < 10 dB for the Cirrus) but relatively
used with a substantial inter-rater agreement in a multi- good contrast between layers allowing for proper
center setting. For mnemonics reasons the set was image post-processing and segmentation do exist. In
called “OSCAR-IB,” where each of the letters indicates absence of a transparent way to judge signal strength
one of seven QC criteria which have to be met [13]. other than dB, we recommend to aim for a signal of at
least 15 dB for the Heidelberg Spectralis and 10 dB for
The OSCAR-IB criteria the Cirrus (the cut-off may be different for different
A summary of the seven OSCAR-IB criteria is shown devices entering the ever-increasing OCT market),
in Table 18.1. Each of the seven criteria is discussed in but allow for a reasonable margin of judgment in
more detail later in this chapter. For didactic reasons, OCT scans that give a good inter-layer contrast
representative examples that have been rejected in a despite a lower signal strength. In order words, the
reading center setting are presented for illustration. S-criterion should be used as a soft criterion.

O-criterion C-criterion
For the first of the OSCAR-IB criteria, “obvious” (O), The third of the OSCAR-IB criteria, correct centration
scans will be rejected for highly apparent issues, of the ring scan (C), is a hard criterion (Figure 18.3).
including severe lens opacities, vitreous hemorrhage, Even a small degree of displacement between peripa-
or poor focus of the lens. See Figure 18.1 for a typical pillary ring scans can lead to a measurement artifact
real-life example of a poorly focused image. Rejection above 3.4 μm [5]. It should be highlighted that any
also occurred if there was a protocol violation such as change of position between a baseline and follow-up
deliberately not averaging several B-scans by turning scan should be rejected for this reason. Likewise, any
off the ART mode. In our experience, the rejection large level of imperfect centering at baseline, which in
rate based on this criterion is low, but the reasons the worst-case scenario may cut through the optic nerve
remain unpredictable. head itself, should be rejected (Figure 18.4). To judge
what a “large displacement” means, we proposed to use
the Royal Air Force (RAF) logo as a guide and reject
S-criterion peripapillary ring scans that cross two of the RAF logo
For the second of the OSCAR-IB criteria, signal rings [13]. What might be permissible is a slight dis-
strength (S), we had arbitrarily chosen a cut-off of placement that is taken into consideration at follow-up

Table 18.1 The OSCAR-IB quality-control criteria for retinal OCT scans. Table modified with permission from reference [13].

Item criteria
O Obvious problems (e.g., unfocused image, wrong scan protocol, etc.).
S Is the OCT signal sufficient to give a sharp contrast between layers needed for layer segmen-
tation?
We recommend a signal > 10 for the Cirrus and > 15 for the Heidelberg Spectralis OCT.,
appropriate B–scan averaging (ART function activated).
C Is the scan correctly centered?
A Is there an algorithm failure?
R Is there visible retinal pathology which may potentially impair the RNFL reading? (minor floaters
casting a shadow are not relevant)
I Is the fundus well illuminated?
B Is the measurement beam placed centrally? 199
 Chapter 18: Optical coherence tomography in a multi-center setting: quality control issues

ILM

RNFL

200 µm

200 µm

Peripapillary RNFLT Classification

300
TS NS
117 96
240
Thickness [µm]

(136) (102)
180
44
T G N N/T
PMB 87 62 0.75 120
83
55 (98) (72) (0.94)
(56) (76)
60
TI NI
125 72 0
(144) (107) –180 –135 –90 –45 0 45 90 135 180
NAS INF TMP SUP NAS
Position [°]
With in Normal Limit

Figure 18.1 This OCT scan is poorly focused, as illustrated by the blurred cSLO image to the left. This scan was QC rejected according to the
[O] criterion for this reason alone, but also conflicts with the [C] and [S] criteria.

ILM

RNFL

200 µm 200 µm

Peripapillary RNFLT Classification

300
NS TS
96 123 240
Thickness [µm]

(102) (133)
180 65
N/T N G T
1.18 63 86 53 PMB 120
(0.97) (72) (96) (73) 46
(56)
60
NI TI
107 127 0
(105) (140) –180 –135 –90 –45 0 45 90 135 180
NAS INF TMP SUP NAS
Position [°]
Borderline Below

Figure 18.2 The poor signal strength in this image (5 dB) causes a grainy and very noisy image. In addition, the (C, B) criteria are
violated. The contrast levels between retinal layers are so poor that image post-processing will become questionable. This scan was QC
rejected.
200
 Chapter 18: Optical coherence tomography in a multi-center setting: quality control issues

ILM

RNFL

200 µm

200 µm

Peripapillary RNFLT Classification

300
TS NS
98 72 240
Thickness [µm]

(136) (102)
180
53
T G N N/T
PMB 102 99 79 0.78 120
69 (76) (98) (72) (0.94)
(57)
60
TI NI
157 99 0
(144) (106) –180 –135 –90 –45 0 45 90 135 180
NAS INF TMP SUP NAS
Borcle rllie Above Position [°]
Borcle rllie Below

Figure 18.3 Poor centering of the peripapillary ring scan. The cSLO image to the left shows that the green circle is not placed centrally over
the optic nerve head, but is shifted to the right. As a result, the thickness profile of the temporal quadrant (T, TMP, −45–45 degree position)
appears to be artificially thickened. This scan was QC rejected.

200 µm 200 µm

Figure 18.4 Another example of a poorly centered OCT technician ring scan submitted to a reading center in a multi-center setting. In this
example, the nasal quadrant cuts through the cup of the optic disc. This scan was QC rejected.

scans by means of scan registration. Having some flex- In selected cases, it may be possible to correct such
ibility here is relevant because of the variable anatomy errors by image post-processing techniques, including
of the optic nerve head. manual correction. This procedure is extremely time-
consuming. A decision may be made on individual
A-criterion scans. While this criterion should be handled rather
strictly on the baseline scan in order to ensure optimal
The fourth of the OSCAR-IB criteria, algorithm fail-
settings for longitudinal studies, one may accept that a
ures (A), also present a serious issue. Any localized
minor, manually correctable algorithm failure on a
algorithm failure invalidates data taken from the
valuable follow-up scan should not necessarily lead to
respective retinal region. Such scans will need to be
immediate rejection. An OCT technician should be
rejected. Importantly, this can be corrected by the
aware of this fact, because manual correction of algo-
reading center and in contrast to the other criteria
rithm failures may be possible prior to uploading a
re-assessment of a patient may not be needed. 201
scan to a reading center.
 Chapter 18: Optical coherence tomography in a multi-center setting: quality control issues

Table 18.2 Pathology of the retina to be considered by the OSCAR-IB criteria. Table reproduced with permission from reference [13].

Summary Diseases
Structural Drusen, Cysts, Detachment, Large discs, Small crowed discs, presence of myelinated axons, naevus,
tumor, peripapillary atrophy, optic disc edema, more than six diopters of myopia or hyperopia
Vascular AION & PION, NA-AION & NA-PION, GCA, CRO, CRBO, AVM, Cotton-wool spots, CVA affecting the optic
pathways
Immune paraneoplastic, MAR, NMO, CAR, SLE, uveitis, birdshot retinochoroiditis
Infectious Viral, bacterial, fungus, HIV, Lyme disease, secondary syphilis
Hereditary Leber’s, DOA, albinism, cone dystrophy, retinitis pigmentosa
Iatrogenic Retina surgery, photocoalgulation, solar retinopathy, central serous chorioretinopathy, Purtscher’s
retinopathy, optic nerve sheet fenestration, brain surgery affecting the optic pathways
Metabolic/ Diabetes, Vit A deficit, alcohol-, tobacco– and malnutrition-induced amblyopia, amiodarone,
toxic chloroquine, vigabatrin
Other Glaucoma, macular degeneration, acute posterior multifocal placoid pigment epitheliopathy, acute
macular neuroretinopathy

R-criterion In contrast, poor illumination by a suboptimal

placement of the laser beam causes partial illumina-
The fifth of the OSCAR-IB criteria, retinal pathology
tion of the retina with just about acceptable B-scan
(R), unrelated to MS can independently influence
quality in some areas and very dark B-scans from
RNFL measures (see Figure 18.5) and ought to be
poorly illuminated areas. Typically, the poorly illumi-
taken into account in studies focused on MS. A list of
nated areas are located at the border of the IR image.
pathologies to be considered is presented in Table 18.2.
Clearly, in poorly illuminated areas the signal drops to
At present, a number of recent OCT studies have
such low levels that no image post-processing can be
provided evidence that retinal pathology other than
performed. In summary, a shadow caused by floaters,
RNFL thinning exists in MS [11, 6, 10, 3], including
cataracts, long eye lashes, and the like should not
microcystic macular edema (MME) predominantly
be considered a general cause of exclusion for the
affecting the inner nuclear layer (INL) of the retina.
I-criterion in cases where A- and B-scan quality is
We, therefore, do not recommend rejecting OCT
good.
scans showing evidence of MME because this may
be a clinically relevant finding in the disease.
B-criterion
I-criterion The last of the OSCAR-IB criteria, the beam place-
ment criterion (B), may not be intuitively obvious to
The sixth of the OSCAR-IB criteria refers to poor
everyone. To put it simply, it is important to keep the
illumination of scans (I). Poor or unequal illumina-
OCT scan in the live window horizontally orientated
tion had been made one of the first QC criteria in
whenever possible.
other studies but remains difficult to judge on its own.
Problems with the B-criterion arise when a scan is
In most cases poor illumination will also cause a loss
tilted in one direction at baseline and to the opposite
of signal and related problems (see Figure 18.6).
direction at follow-up, because measurement errors in
Therefore, a scan may be rejected on the basis of a
the range of 9–40 μm can be introduced [2]. The
combination of several OSCAR-IB criteria.
reason for this is illustrated in Figure 18.8.
What remains difficult is how to judge poor illumi-
Therefore, scans with suboptimal beam placement as
nation in isolation. To give one example, Figure 18.7
described, should be rejected.
shows a shadow cast on the retina by a floater. Should
In some cases the anatomy of the retina may make
this scan be rejected? We would not do so because the
it difficult to achieve a perfectly horizontally aligned
scan quality is such that all retinal layers can clearly be
OCT. In this context, the QC rating of the follow-up
distinguished and the scan would not fail because of
scan will ensure that the same direction of tilting is
202 any of the other OSCAR-IB criteria.
 Chapter 18: Optical coherence tomography in a multi-center setting: quality control issues

ILM

RNFL

200 µm

200 µm

Peripapillary RNFLT Classification

300
NS TS
145 149 240
Thickness [µm]

(102) (134)
180
67
N/T N G T
93 177 267 PMB 120
0.35
(0.96) (72) (97) (74) 291
(56)
60
NI TI
149 255 0
(105) (142) –180 –135 –90 –45 0 45 90 135 180
NAS INF TMP SUP NAS
Position [°]
Above Normal Limits

Figure 18.5 In this patient central serous edema extends to the area captured by the peripapillary ring scan. Consequently the temporal
quadrant of the RNFL appears to be thickened artificially. This scan was QC rejected.

ILM

RNFL

200 µm 200 µm

Peripapillary RNFLT Classification

300
TS NS
132 103 240
Thickness [µm]

(136) (102)
180 65
T G N N/T
PMB 67 99 84 0.25 120
51 (76) (96) (72) (0.94)
(58)
60
TI NI
135 115 0
(144) (106) –180 –135 –90 –45 0 45 90 135 180
NAS INF TMP SUP NAS
Position [°]
Within Normal Limits

Figure 18.6 A poorly illuminated OCT scan, which also causes a low signal such that it is impossible to recognize any of the retinal layers but
the RNFL in the temporal quadrant. The (B) criterion is also violated. This scan was QC rejected. 203
 Chapter 18: Optical coherence tomography in a multi-center setting: quality control issues

ILM

RNFL

–4.8°

Left mouse button moves the vertical marker.

ALT to change contrasr.

200 µm

200 µm

Peripapillary RNFLT Classification

300
NS TS
76 126 240
Thickness [µm]

(102) (132)
180
42
N/T N G T
1.13 65 79 58 PMB 120
(0.97) (72) (96) (73) 42
(56)
60
NI TI
63 122 0
(105) (139) –135 –90 –45 0 45 90 135 180
NAS INF TMP SUP NAS
Position [°]
Borcle rllie Below

Figure 18.7 This OCT scans appears to be unequally illuminated. What happened here is that a floater casts a shadow in the lower-left part of
the IR image. The scan does not violate any of the OSCAR-IB criteria. This scan was QC accepted.

repeated, thus ensuring consistency. The core of the measurement artifacts are introduced [9]. The
B-criterion is that the direction of tilt causes a char- possibility of post hoc repositioning of the
acteristic, reproducible, robust signal pattern in the EDTRS grid needs to be considered.
outer plexiform (OPL) and outer nuclear layers (ONL, 4. A: Algorithm failures are more frequent in areas
Figure 18.9). This sign allows tracing possible errors with very thin retinal layers. Because the foveola
due to off-center beam placement. itself does not contain any measurable degree of
We recommend the rejection of any scans that several retinal layers, algorithm failures in this
combine an obvious violation of the B-criterion with region are common, but they should not lead to
an associated change of retinal layer thickness values rejection if image post-processing remains possi-
in the corresponding retinal area. ble for relevant B-scans.
5. R: Any retinal pathology other then MME affect-
Macular volume scans ing the macula should be discussed for rejection in
Validated QC criteria for macular volume scans do MS research.
not exist at present. We propose that the same seven 6. I: Poor and unequal illumination should be used
OSCAR–IB criteria be used the same way as for the as a soft criterion for rejection of macular scans if
peripapillary ring scan for the following reasons: present in isolation of on a small number of less
1. O: Any obvious reason including protocol viola- relevant (peripheral) B-scans.
tions which led to exclusion of a peripapillary ring 7. B: Some degree of off-center beam placement is
scan very likely compromise macular volume likely to occur in parts of the large area scanned
scans as well. and should not be used as a hard criterion for
2. S: Signal strength may be used as a soft criterion if rejection if not combined with corresponding
image post-processing is possible such that reli- measurement artifacts on longitudinal scans.
able quantitative data can be obtained from rele- There is published evidence demonstrating that
vant B-scans. off-center beam placement affects retinal thickness
3. C: Poor centering of a macular scan is a clear measurements [7] as it did for peripapillary ring
204 reason for rejection because considerable scans [2].
 Chapter 18: Optical coherence tomography in a multi-center setting: quality control issues

B C

temporal nasal temporal nasal

D E

Figure 18.8 (A) The OCT measurement beam is focused on the dilated right eye of a healthy control subject. (B) temporal off-center
placement of the measurement beam results in a shorter light path to the temporal part of the optic nerve head (dotted line) and a
longer pathway from the nasal part of the optic nerve head (dotted-dashed line). The difference in path length results in a tilted
appearance of the B-scan. (C) Nasal off-center placement of the measurement beam results in a mirror pattern. The resulting
averaged OCT image is of good quality (ART 25, signal strength 35 dB) for both (D) temporal off-center placement and (E) nasal
off-center placement. The quantification of the RNFL thickness by the algorithm is, however, clearly different (Global
average OD with temporal off-center placement 106 μm and nasal off-center placement 103 μm). Figure reprinted with permission from
reference [2].

205
 Chapter 18: Optical coherence tomography in a multi-center setting: quality control issues

A Figure 18.9 Inhomogeneous reflectivity of

the outer part of the ONL indicates off-center
ILM RNFL placement of the OCT measurement beam. (A)
RNFL
The averaged summary scan obtained from the
ONL correctly, horizontally orientated live images of
the reference scan. This images shows a homo-
geneous reflectivity of the outer ONL (black
200 µm
arrow). The automated segmentation identifies
the borders of the RNFL (red/gray lines). Note,
this is the image that is sent to the reading center
B
and used for automated calculation of the RNFL
ILM thickness. (B) temporal off-center placement
RNFL results in a inhomogeneous outer-ONL reflectiv-
ONL ity. The ONL reflectivity is increased for the cen-
trally elevated part in the live image (white
arrows) and decreased in the periphery (gray
arrows). (C) nasal off-center placement, (D) rostral
200 µm
off-center placement, (E) caudal off-center
placement.
C
ILM

RNFL

ONL

200 µm

D
ILM

RNFL
ONL

200 µm

E
ILM

RNFL
ONL

200 µm

Several times we referred to “relevant B-scans” in Prior to scanning Before you start imaging make
this proposed QC list. The rationale is that QC failure sure that you use the right OCT scan protocol and
of B-scans taken through the foveola or perimacular obtain a clearly visible OCT image (O). If available
rim posed a proportional greater risk than QC failure activate automated B-scan averaging and/or the
of a B-scan taken at the extreme edges (Figure 18.10). eye-tracking function.
Therefore, one may handle the QC criteria softer at While scanning During imaging, center your
the edge of a macular volume scan but remain peripapillary ring scan and/or macular scan well (C);
rigorous in the center. take care of good and equal illumination (I); keep the
B-scan horizontally aligned in the live window to
Practical advice for obtaining ensure central beam placement (B).
After scanning After scanning briefly revise your
high-quality OCT scans scans asking two questions. Is there any obvious
As a rule of thumb we advise to keep the mne- algorithm failure (A)? Is there any retinal pathology
206 monics of the OSCAR-IB criteria in mind. visible that may lead to exclusion (R)?
 Chapter 18: Optical coherence tomography in a multi-center setting: quality control issues

ILM

BM

200 µm 200 µm

Figure 18.10 This OCT scan is cut on the edges. Potential rejection of such a scan taken at the very edge of the macular volume scan should
be carefully considered if all other, more relevant, OCT B-scans of the entire volume are acceptable.

Conclusion 6 Jeffrey M. Gelfand, Rachel Nolan, Daniel M. Schwartz,

This chapter showed that transparent, reliable, and Jennifer Graves, and Ari J. Green. “Microcystic macular
edema in multiple sclerosisis associated with disease
validated QC criteria are available for retinal OCT in
severity.” Brain (2012), pp. 1786–93.
MS research. It is advisable to take these criteria into
7 Amirhossein Hariri, Sun Young Lee, Humberto
account, particularly in a multi-center setting, in
Ruiz-Garcia, Muneeswar Gupta Nittala, Florian
order to maximise the efficacy of this new and pro- M. Heussen, and Srinivas R. Sadda. “Effect of angle of
mising technology. incidence on macular thickness and volume
measurements obtained by spectral-domain optical
coherence tomography.” Invest Ophthalmol Vis Sci
References 53 (2012), pp. 5287–5291.
1 Madhusudhanan Balasubramanian, Christopher Bowd, 8 Christopher Kai-shun Leung, Carol Yim-lui Cheung,
Gianmarco Vizzeri, Robert N. Weinreb, and Linda M. Dusheng Lin, Chi Pui Pang, Dennis S C Lam, and
Zangwill. “Effect of image quality on tissue thickness Robert N Weinreb. “Longitudinal variability of
measurements obtained with spectral domain-optical optic disc and retinal nerve fiber layer measurements.”
coherence tomography.” Opt Express 17 (2009), pp. Invest Ophthalmol Vis Sci
4019–4036. 49 (2008), pp. 4886–4892.
2 Lisanne J. Balk, Willemien A E J. de Vries-Knoppert, 9 Jeong W. Pak, Ashwini Narkar, Sapna Gangaputra,
and Axel Pet– zold. “A Simple sign for recognizing Ronald Klein, Barbara E K.Klein, Stacy Meuer, Yijun
off-axis OCT measurement beam placement in the Huang, and Ronald Danis. “Effect of optical coherence
context of multicentre studies.” PLoS One 7 (2012), tomography scan decentration on macular center
e48222. subfield thickness measurements.” Invest Ophthalmol
3 Lisanne J. Balk, Joep Killestein, Chris H. Polman, Vis Sci (2013).
Bernard M J. Uit– dehaag, and Axel Petzold. 10 Shiv Saidha, Elias S. Sotirchos, Mohamed A. Ibrahim,
“Microcystic macular edema confirmed, but Ciprian M. Crainiceanu, Jeffrey M. Gelfand, Yasir
not specific for multiple sclerosis.” Brain J. Sepah, John N. Ratchford, Jiwon Oh, Michaela
135 (2012), e226. A. Seigo, Scott D. Newsome, Laura J. Balcer, Elliot
4 Amitha Domalpally, Ronald P Danis, Baoyan Zhang, M. Frohman, Ari J. Green, Quan D. Nguyen, and
Dawn Myers, and Christina N Kruse. “Quality issues in Peter A. Cal– abresi. “Microcystic macular
interpretation of optical coherence tomograms in edema, thickness of the inner nuclear layer of the
macular diseases.” Retina 29 (2009), pp. 775–781. retina, and disease characteristics in multiple
5 Michelle L Gabriele, Hiroshi Ishikawa, Gadi Wollstein, sclerosis: a retrospective study.” Lancet Neurol 11
Richard A Bilonick, Kelly A Townsend, Larry (2012), pp. 963–972.
Kagemann, Maciej Wojtkowski, Vivek J Srinivasan, 11 Shiv Saidha, Stephanie B. Syc, Mohamed A. Ibrahim,
James G Fujimoto, Jay S Duker, and Joel S Schuman. Christopher Eckstein, Christina V. Warner, Sheena
“Optical coherence tomography scan circle location K. Farrell, Jonathan D. Oak– ley, Mary K. Durbin, Scott
and mean retinal nerve fiber layer measurement A. Meyer, Laura J. Balcer, Elliot M. Frohman, Jason
variability.” Invest Ophthalmol Vis Sci 49 (2008), M. Rosenzweig, Scott D. Newsome, John N. Ratchford,
pp. 2315–2321. Quan D. Nguyen, and Peter A. Calabresi. “Primary 207
 Chapter 18: Optical coherence tomography in a multi-center setting: quality control issues

retinal pathology in multiple sclerosis as detected by 13 Prejaas Tewarie, Lisanne Balk, Fiona Costello,
optical coherence tomography.” Brain 134 (2011), Ari Green, Roland Martin, Sven Schippling,
pp. 518–533. and Axel Petzold. “The OSCAR-IB Con–
12 Lauren S Talman, Esther R Bisker, David J Sackel, David sensus Criteria for Retinal OCT
A Long, Kristin M Galetta, John N Ratchford, Deacon J Quality Assessment.” PLoS One 7
Lile, Sheena K Far– rell, Michael J Loguidice, Gina (2012), e34823.
Remington, Amy Conger, Teresa C Frohman, Dina A 14 Gianmarco Vizzeri, Christopher Bowd, Felipe
Jacobs, Clyde E Markowitz, Gary R Cutter, Gui– Shuang A. Medeiros, Robert N. Weinreb, and Linda
Ying, Yang Dai, Maureen G Maguire, Steven L Galetta, M. Zangwill. “Effect of signal strength and improper
El– liot M Frohman, Peter A Calabresi, and Laura J alignment on the variability of stratus optical
Balcer. “Longitudinal study of vision and retinal nerve coherence tomography retinal nerve fiber layer
fiber layer thickness in multiple sclerosis.” Ann Neurol 67 thickness measurements.” Am J Ophthalmol 148
(2010), pp. 749–760. (2009), 249–255.e1.

208
 Chapter
Future technological advances in optical

19 coherence tomography
Hiroshi Ishikawa and Joel Schuman

Since its commercial introduction in 1996, optical coronal view of the 3-D OCT data) permitted us to
coherence tomography (OCT) has gone through pinpoint the sampling location within the 3-D cube
numerous iterations of technological improvement. OCT data relative to the structures that can be used as
The most dramatic and important change was the references (e.g., major retinal blood vessels, optic
transition from time-domain (TD) to commercially nerve head (ONH), etc.) (Figure 19.2). This change
available spectral-domain (SD) OCT in 2006. SD- can be recognized as analogous to the transition in
OCT eliminated the need for the physically moving radiology from the simple x-ray to CT scanning.
reference mirror that dictated the scan speed, so The transition to SD-OCT also changed how the
that commercial SD-OCT devices achieved 40 times OCT technology was commercialized. With the
faster scanning compared to even the fastest TD-OCT apparent public domain nature of the SD-OCT tech-
systems. This enabled three-dimensional (3-D) nology, multiple companies jumped into the medical
sampling of the retinal tissue in vivo rather than OCT market. Currently, there are at least nine com-
the two-dimensional (2-D) scanning formerly per- panies manufacturing SD-OCT. This competition
formed. In addition to the faster scanning, SD-OCT accelerates technological improvement and innova-
also provided higher resolution (4–5 μm compared tion in a very positive way, as well as driving prices
to 8–10 μm). down for consumers.
3-D scanning allowed visualization of retinal tis- In this chapter, many new technological advance-
sue in coronal sections in addition to the conventional ments in the ocular OCT imaging field are discussed.
sagittal and transverse (axial) sections. Coronal sec- Some are still in purely experimental stages, but
tioning confirmed that the seemingly random reflec- others are on the verge of clinical commercialization
tions below the retina in TD-OCT sagittal or or even already on the market. All provide particular
transverse images were the choroidal vasculature advantages in a variety of aspects of OCT imaging that
(Figure 19.1). Also, the OCT fundus image (summed are useful for both clinical and research purposes.

Figure 19.1 The coronal sectioning of 3-D SD-OCT

data showed the seemingly random noise below
the retina in cross-sectional images to be choroidal
vasculature. (A) C-mode image of choroidal vasculature,
which was generated by summing up the information
within the slab of thickness indicated by the three
horizontal light blue lines shown on the cross-sectional
image (B).

Optical Coherence Tomography in Neurological Disease, ed. Peter A. Calabresi, Laura J. Balcer, and Elliot M. Frohman. 209
Published by Cambridge University Press. © Cambridge University Press 2015.
 Chapter 19: Future technological advances in optical coherence tomography

A B

C D

Figure 19.2. The optical coherence tomography (OCT) fundus image (summed coronal view of the 3-D OCT data), as shown in
(A) and (C), permits pinpointing of the sampling location within the 3-D cube OCT data. Red lines on the fundus image indicate where
the corresponding cross-sectional images on the right were sampled. The vertical yellow lines point the same location in the
corresponding fundus and cross-sectional images. In (B), the vertical yellow line points out the position of the foveola, while in (D) it
indicates the location of the blood vessel.

Eye-tracking system Eye-tracking OCT is not really a new technological

advancement as of June 2014, because there are some
Eye movement artifacts on OCT images have been
commercial SD-OCT devices that already have this
one of the biggest problems that affect both qualitative
feature. They send two beams of light to the target eye
and quantitative assessment from the very beginning.
simultaneously: one beam for tracking and the other
When SD-OCT was commercialized, we hoped that
beam for OCT scanning. With eye tracking, it not only
the faster scanning speed of SD-OCTs would have
resolves eye movement or blink artifacts occurring
solved the eye movement artifact problem.
during scanning, but also enables repeated scanning
Ironically, the faster speed was used to increase sam-
at the same location. This allows signal averaging to
pling density (from 768 samplings on six radial 2-D
improve the signal-to-noise ratio, resulting in greater
slices to 40,000 samplings on a 3-D cube covering 6 ×
image quality (Figure 19.4) [1–5]. It is also possible to
6 mm area for approximately the same scan time of
register the scanning location to the previous scan so
1.2 to 1.4 seconds) instead of decreasing scanning
that scan-to-scan measurement variability can be mini-
time, resulting in the same eye movement artifacts in
mized. This is particularly useful for longitudinal study
SD-OCT exhibited with TD-OCT (Figure 19.3). On
to assess changes in a given subject [6].
the other hand, it is easier to catch such artifacts on
The one practical drawback of eye-tracking OCT
SD-OCT than TD-OCT because the 3-D scanning
is its prolonged scanning time. The fast eye move-
allows an en face view of the image. Also, eye move-
ment and blinking during scanning forces the
ments including micro-saccades are so fast that even
machine to discard the frame just taken and simply
near instantaneous scanning might still demonstrate
repeat scanning at the same location until there is no
some eye movement artifacts, although the magnitude
detectable motion for that particular frame. This
becomes small with extremely fast scanning.
means that the overall scanning time is a function
Therefore, it seems common sense to add eye tracking
of the number of eye movement/blinking events
210 to OCT.
during scanning. Unfortunately, as the number of
 Chapter 19: Future technological advances in optical coherence tomography

A C Figure 19.3. Eye motion artifacts during scanning. (A)

and (C) present good scans, without eye motion artifacts
or blinking, of macular region. (B) and (D) show scans with
eye movement during the scans, as indicated by the red
arrows. Eye motion artifacts can be observed as the dis-
continuity of the blood vessels on the fundus images and
significantly affect the analysis results.

B D

A Swept-source OCT
Swept-source (SS) OCT is similar to SD-OCT in the
use of detected wavelength to determine reflection
depth. In SS-OCT, an A-scan line is obtained by
scanning with a starting frequency and then “sweep-
ing” the frequency over a period of time to obtain the
different spectral components of the detected signal,
B using a tunable laser source with a continuous
“sweeping” mode. Instead of using a spectrometer to
determine different spectral components, which is the
speed bottle neck in SD-OCT, a single detector gath-
ers all the information of the “sweep” [7]. The acqui-
sition speed of SS-OCT is dictated by the tuning speed
of the laser light source. It reaches 100,000 to 400,000
Figure 19.4 Cross-sectional images through the fovea from the
same eye using OCT device with an eye-tracking system. (A) no
A-scans in one second, substantially increasing the
frame averaging is applied. (B) 100 frames averaged. 100-frame scanning speed to 4 to 8 times faster than SD-OCT.
averaged scan reveals stronger and clearer definition of the retinal SS-OCT has been reported to scan at MHz rates as
layer structures with less speckle noise compared to a non-frame-
averaged scan.
well [8–10].
With the dramatically improved scanning speed,
regular 200 × 200 A-scans covering a 6 × 6 mm area
eye movement/blinking events increases with age can be done within a fraction of a second, and there-
and severity of vision impairment, the overall scan fore we are getting one step closer to motion-free 3-D
time may become sometimes intolerably long for OCT images even without an eye-tracking system.
such cases if a dense 3-D volume scan is needed. However, the faster speed is generally used for denser
This also indicates that the scanning speed of the sampling so that the overall scanning time for a 3-D
current SD-OCT is still not fast enough to capture cube scan remains about one second. This is because 211
“motion-free” 3-D OCT image. the sampling density has not reached the optimal
 Chapter 19: Future technological advances in optical coherence tomography

number for both qualitative and quantitative assess- visualization of individual retinal cells and generates
ment yet. Even with potential eye movement artifacts, 3-D images of photoreceptors in vivo. Since AO-OCT
denser sampling gets higher priority in OCT imaging has better focus on the retina and more efficient col-
for both clinical and research purposes. lection of backscattered light by reducing the diffrac-
Tunable laser light sources dramatically improve tion effect, it enhances image contrast and quality as
the scanning speed; however, they also bring disad- well as transverse resolution. With a pupil diameter of
vantages. The major disadvantage is the lower axial > 6 mm, a transverse resolution of 2–3 μm can be
resolution than SD-OCT (5.3 to 8 μm (SS-OCT) vs. 4 achieved, which enables visualization of individual
to 5 μm (SD-OCT)). [11] Ironically, the narrow band- rods and cones as well as individual nerve fiber bun-
width of the swept-source laser, which achieves faster dles [15,16].
scanning, decreases the axial resolution because reso-
lution is proportional to the bandwidth. Therefore, a Doppler OCT
narrower-bandwidth light source results in worse
Assessment of retinal vasculature and blood flow is an
axial resolution
important clinical observation for evaluating various
On the other hand, SS-OCT does not exhibit some
pathological conditions and understanding patho-
of the problems seen with SD-OCT. Unlike with SD-
physiology. Conventionally, fluorescein angiography
OCT, neither the sensitivity nor the resolution of SS-
(FA) and indocyanine green angiography (ICGA)
OCT diminishes with increasing distance from the
have been used to observe the retinal blood vessel
“zero-delay.” The zero-delay can be thought of as
morphology, and Doppler ultrasound has been used
the point in space representing the beginning of
for blood flow measurement. However, FA and ICGA
each given A-scan, such that the farther a point is
imaging are invasive due to the requirement of dye
from the zero-delay, the deeper into the tissue the
injection [17, 18], and Doppler ultrasound can target
scan point is acquired.
only one or two vessels at a time and its measurement
is largely operator dependent.
Adaptive optics optical coherence Doppler OCT (D-OCT) was developed for obser-
vation of both the vasculature morphology and blood
tomography flow measurement in a noninvasive fashion. The prin-
OCT technology is known to provide great axial reso- ciple behind D-OCT is similar to the Doppler ultra-
lution (4 to 5 μm with SD-OCT). However, the trans- sound. It is based on the frequency shift caused by the
verse, or lateral, resolution is still limited to the range moving blood cells. When the incident light signal is
of 15 to 20 μm. This asymmetric proportion of the scattered or reflected back from moving particles, a
resolution in different dimensions hinders its visuali- shift in frequency will be observed due to the Doppler
zation clarity in the coronal plane. For example, inner effect. The shift in the central frequency of the
and outer borders of the photoreceptor layer together reflected light can be detected as the phase change
with the inner/outer segment junction can be recog- between sequential A-scans [19].
nized in axial slices, while the photoreceptor mosaic With the ability to detect the frequency shifts or
cannot ordinarily be visualized except in high A-scan phase changes, D-OCT allows imaging blood vessels
density coronal slices. [12,13] in vivo with high resolution and sensitivity and offers
Adaptive optics (AO) was originally developed valuable information regarding the blood flow direc-
and used in astronomical telescopes in order to tion and velocity, as well as detailed structural
improve the resolution of astronomical imaging and investigation of the vasculature [20–28]. With sophis-
remove the effects of atmospheric distortion. [14] It is ticated image-processing techniques, quantitative
now incorporated into OCT devices to correct ocular flow investigation of retinal perfusion is now possible
aberrations. AO systems measure the distortion in the with commercially available OCT devices, though it
wave shape of the reflected signal and compensate for still remains challenging. The most important contri-
them with a deformable mirror. In this way, AO fixes bution of D-OCT is to provide 3-D tissue retinal
the ocular aberrations caused by wavefront distor- microvascular information in a noncontact and non-
tions when the signal passes through the eye and, invasive fashion compared to FA and ICGA; however,
therefore, is able to markedly improve transverse D-OCT cannot identify leakage of fluid from vessels.
212 resolution. Combined with OCT, it provides detailed D-OCT uses Doppler to generate image contrast
 Chapter 19: Future technological advances in optical coherence tomography

without the use of dye through a technique called structures with birefringence properties. RPE exhibits
optical coherence angiography (OCA) and, therefore, depolarization, so it can also be a good target for PS-
enables segmentation and display of vasculature. OCT, especially in eyes with pathologies involving
Enabling the 3-D visualization of the vasculature, RPE (e.g., macular degeneration, pigment epithelial
especially retinal and choroidal vasculature, provides detachments, pseudo vitelliform dystrophy, etc.) [46].
an alternative means for diagnosis and management With potentially more accurate and detailed assess-
of many retinal pathologies related to vasculature ment of birefringent tissues using PS-OCT, we may be
damage [29, 30]. able to detect minute changes in the integrity of nerve
fiber bundles at much earlier stages than currently
Polarization-sensitive OCT possible.
Conventional OCT visualizes different tissues (mainly
various retinal layers) based on the differences in their High penetration OCT
optical indices of refraction. However, the intensity- One of the limitations of conventional OCT is its
based contrasts can be less optimal for distinguishing relatively short penetration depth. Although the scan-
certain adjacent tissues (e.g., boundary between gang- ning window covers ~2mm in the axial direction,
lion cell and inner plexiform layers), especially under OCT signals quickly become attenuated below the
pathological conditions. There are other types of opti- RPE. In the ONH area, visibility of lamina cribrosa
cal properties that the conventional OCT is not depends on the thickness of the pre-lamina tissue.
detecting, such as absorption, fluorescence, lumines- Additionally, SD-OCT exhibits gradual image
cence, and polarization. Polarization sensitive (PS) dynamic range loss along the depth. In other words,
OCT can assess birefringence and depolarization of images look sharp on the top but blurry on the bot-
target tissues with depth information. tom. This is an inherent problem stemming from the
Fibrous tissues (e.g., nerve fiber bundles) that con- SD-OCT principle. In cases in which the region of
sist of long parallel fibrils surrounded by tissues with interest is a deep tissue (e.g., choroid, lamina cri-
different refractive indices exhibit birefringence. brosa), the so-called enhanced depth imaging (EDI)
Multiple scattering at large particles or scattering at technique can be applied, in which the dynamic range
nonspherical particles causes depolarization of deterioration can be reversed (blurry top and sharp
incoming light. PS-OCT visualizes tissue specific con- bottom) [47–63]. However, there are other ways to
trast by detecting these optical properties. Initially, overcome this limitation, such as by using a longer
PS-OCT was only implemented in TD-OCT but now wavelength light source (so-called high penetration
is adapted to SD-OCT and able to provide a contrast- (HP) OCT).
enhanced OCT image with better sensitivity in real Conventional OCT devices use infrared light with
time [31–35]. center wavelengths around 830 nm as the light source
Retinal tissues show both good birefringence and because signals around 830 nm are widely available
depolarization effects, so they can be good targets for and are well transmitted by water, the main content of
PS-OCT. Scanning laser polarimetry (SLP or GDx) the vitreous [64]. However, signals with wavelengths
has been used for glaucoma assessment by measuring of 830 nm have limited reflected signal strength
the retardation amount of the laser light (RNFL bire- beyond the RPE and, therefore, limit the ability to
fringence) that is supposed to linearly correlate with visualize the deeper structures. Instead of using
RNFL thickness [36–41]. Several studies suggest that 830 nm, HP-OCT seeks an alternative local minimum
RNFL birefringence changes may precede RNFL on the spectrum of the water absorption, which is
thickness changes that can be measured using con- located around 1,050 nm, and uses it as the light
ventional OCT [42–45]. The biggest limitation of SLP source. Besides minimal absorption by the water,
is that it measures tissue retardation as a whole. In signals with wavelength around 1,050 nm exhibit
other words, the RNFL measurements are con- less scattering in the choroid and less attenuation in
founded by other birefringent tissues in the eye (e.g., the RPE. Therefore, signals can penetrate beyond the
cornea, vitreous, sclera, etc.). RPE and enable the visualization of tissue structure
Since PS-OCT allows a depth-resolved optical of choroid and sclera with better image quality and
probe of birefringence, RNFL birefringence can contrast [65–67]. This approach is different from EDI
be separately measured from other confounding in using a longer wavelength light source. 213
 Chapter 19: Future technological advances in optical coherence tomography

HP-OCT has been employed in several studies to Therefore, JA-OCT is especially suitable for ultra-
evaluate the high-penetration effects and clinical ben- high-speed in vivo imaging [77]. Based on the angle
efits [67–76]. HP-OCT is able to provide information diversity concept, JA-OCT is able to generate a high-
posterior to the retina, like choroid, choroidal vascu- quality image with reduced speckle noise without
lature, and sclera, to observe choroid or sub-RPE sacrificing imaging speed by organizing multiple
damage or morphological changes caused by various channels in an asymmetric configuration and varying
pathologies, such as the development of choroidal beam sizes. With multiple channels, JA-OCT provides
neovascularization (CNV) beneath the RPE induced angular scattering information about the sample in a
by age-related macular degeneration (AMD) and single acquisition, which may be useful for automatic
polypoidal choroidal vasculopathy (PCV) [67, 72], tissue identification detection [78].
and to evaluate choroidal thickness with better repro- Klein et al. demonstrated that, scanning with JA-
ducibility. Some statistically significant differences in OCT, the acquisition time for human retina with
choroidal thickness or morphological changes of the 2,112 × 258 axial scans for each channel is 0.46 sec-
ocular tissues between healthy and pathological cases onds [77]. The acquired data showed significantly
have been reported [73, 75, 76]. Furthermore, reduced speckle noise, increased collection efficiency,
combining the OCA technology with an HP-OCT higher signal-to-noise ratio (SNR), reduction of spec-
engine, high-penetration optical coherence angiogra- ular and parasitic reflections, and better visualization
phy (HP-OCA), enables vascular imaging of the deep of retinal layers. With all these advantages, JA-OCT
posterior eye in a noninvasive manner. A recent study successfully compensates the dim signal level problem
showed that HP-OCA revealed depth-resolved abnor- of ultrahigh-speed OCT devices, and provides high-
mal vasculature in exudative macular disease and the quality images with substantially reduced speckle
en face HP-OCA images showed high similarity with noise at megahertz imaging speed. JA-OCT may be
FA and ICGA images. This suggests that HP-OCA able to provide more ocular tissue structural informa-
may be clinically useful in noninvasive 3-D angiogra- tion, originally hidden because of limited image
phy [74]. acquisition speed and image quality. Furthermore,
JA-OCT also offers new perspectives for OCT func-
tional imaging: increased channels may allow for
Joint aperture OCT reconstruction of the true Doppler vector and the
As we have seen so far, improving the scanning speed angular scattering characteristics may make tissue
has been one major focus of OCT technology identification feasible.
advancements. However, it is known that higher
speeds may degrade image quality due to shorter
exposure time. One of the key components that dic-
Conclusions
tate overall image quality is sensitivity, the minimum OCT is a rapidly evolving technology. The advances
detectable fraction of back reflected light. Naturally, described in this section represent only a small num-
under the same illumination condition (light source ber of the many areas in which OCT is continuing to
power in OCT domain), shorter exposure lowers sen- develop. As lasers improve and prices drop, other new
sitivity. One way to overcome this law of physics is a OCT iterations will appear, such as is possible in
new approach called joint aperture (JA) OCT. laboratories today with vertical cavity lasers permit-
JA-OCT is an angle-resolved OCT method, in ting OCT imaging of the entire length of the eye in a
which illumination from an active channel is simulta- single A-scan. The excitement and possibilities in
neously probed by several passive channels. In other OCT development are palpable.
words, JA-OCT combines multiple imaging channels
that are targeting the same subject from different References
angles that are strategically placed. JA-OCT increases
1. Sander, B., M. Larsen, L. Thrane, et al., Enhanced
the collection efficiency and effective sensitivity of the optical coherence tomography imaging by multiple
OCT system without increasing the light source scan averaging. Br J Ophthalmol 2005; 89(2): 207–12.
power on the sample, hence no further exposure or 2. Sakamoto, A., M. Hangai, and N. Yoshimura, Spectral-
safety risk. Additionally, JA-OCT provides angular domain optical coherence tomography with multiple
214 scattering information about the sample in a single B-scan averaging for enhanced imaging of retinal dis-
acquisition, so the OCT imaging speed is not reduced. eases. Ophthalmology, 2008;115(6): 1071–1078 e7.
 Chapter 19: Future technological advances in optical coherence tomography

3. Pappuru, R.R., C. Briceno, Y. Ouyang, et al., Clinical with adaptive optics. Biomed Opt Express, 2011.
significance of B-scan averaging with SD-OCT. 2(4): 748–63.
Ophthalmic Surg Lasers Imaging, 2012;43(1): 63–8. 17. Yannuzzi, L.A., K.T. Rohrer, L.J. Tindel, et al.,
4. Jorgensen, T.M., J. Thomadsen, U. Christensen, et al., Fluorescein angiography complication survey.
Enhancing the signal-to-noise ratio in ophthalmic Ophthalmology, 1986. 93(5): 611–7.
optical coherence tomography by image 18. Hope-Ross, M., L.A. Yannuzzi, E.S. Gragoudas, et al.,
registration–method and clinical examples. J Biomed Adverse reactions due to indocyanine green.
Opt, 2007;12(4): p. 041208. Ophthalmology, 1994. 101(3): 529–33.
5. Wu, W., O. Tan, R.R. Pappuru, et al., Assessment of 19. Drexler, W. and J.G. Fujimoto, State-of-the-art retinal
frame-averaging algorithms in OCT image analysis. optical coherence tomography. Prog Retin Eye Res,
Ophthalmic Surg Lasers Imaging Retina, 2013;44(2): 2008. 27(1):45–88.
168–75.
20. Wehbe, H., M. Ruggeri, S. Jiao, et al., Automatic retinal
6. Chin, E.K., R.W. Sedeek, Y. Li, et al., Reproducibility of blood flow calculation using spectral domain optical
macular thickness measurement among five OCT coherence tomography. Opt Express, 2007. 15(23):
instruments: effects of image resolution, image regis- 15193–206.
tration, and eye tracking. Ophthalmic Surg Lasers
21. Makita, S., T. Fabritius, and Y. Yasuno, Quantitative
Imaging, 2012; 43(2) 97–108.
retinal-blood flow measurement with three-dimen-
7. Drexler, W. and J.G. Fujimoto, Optical Coherence sional vessel geometry determination using ultrahigh-
Tomography: Technology and Applications. 2008. resolution Doppler optical coherence angiography.
New York:Springer. Opt Lett, 2008. 33(8): 836–8.
8. Blatter, C., T. Klein, B. Grajciar, et al., Ultrahigh-speed 22. Wang, Y., B.A. Bower, J.A. Izatt, et al., Retinal blood
non-invasive widefield angiography. J Biomed Opt, flow measurement by circumpapillary Fourier domain
2012. 17(7): 070505. Doppler optical coherence tomography. J Biomed Opt,
9. Goda, K., A. Fard, O. Malik, et al., High-throughput 2008. 13(6): 064003.
optical coherence tomography at 800 nm. Opt Express, 23. Szkulmowska, A., M. Szkulmowski, D. Szlag, et al.,
2012. 20(18): 19612–7. Three-dimensional quantitative imaging of retinal and
10. Klein, T., W. Wieser, C.M. Eigenwillig, et al., choroidal blood flow velocity using joint Spectral and
Megahertz OCT for ultrawide-field retinal imaging Time domain Optical Coherence Tomography. Opt
with a 1050 nm Fourier domain mode-locked laser. Express, 2009. 17(13): 10584–98.
Opt Express, 2011. 19(4): 3044–62. 24. Makita, S., Y. Hong, M. Yamanari, et al., Optical
11. Potsaid, B., B. Baumann, D. Huang, et al., Ultrahigh coherence angiography. Opt Express, 2006. 14(17):
speed 1050nm swept source/Fourier domain OCT 7821–40.
retinal and anterior segment imaging at 100,000 to 25. An, L. and R.K. Wang, In vivo volumetric imaging of
400,000 axial scans per second. Opt Express, 2010. vascular perfusion within human retina and choroids
18(19): 20029–48. with optical micro-angiography. Opt Express, 2008.
12. Zawadzki, R.J., S.S. Choi, S.M. Jones, et al., Adaptive 16(15): 11438–52.
optics-optical coherence tomography: optimizing 26. Kim, D.Y., J. Fingler, J.S. Werner, et al., In vivo volu-
visualization of microscopic retinal structures in three metric imaging of human retinal circulation with
dimensions. J Opt Soc Am A Opt Image Sci Vis, 2007. phase-variance optical coherence tomography. Biomed
24(5): 1373–83. Opt Express, 2011. 2(6): 1504–13.
13. Zawadzki, R.J., S.S. Choi, A.R. Fuller, et al., Cellular 27. Makita, S., F. Jaillon, M. Yamanari, et al.,
resolution volumetric in vivo retinal imaging with Comprehensive in vivo micro-vascular imaging of
adaptive optics-optical coherence tomography. the human eye by dual-beam-scan Doppler optical
Opt Express, 2009. 17(5): 4084–94. coherence angiography. Opt Express, 2011. 19(2):
14. Ragazzoni, R., E. Marchetti, and G. Valente, 1271–83.
Adaptive-optics corrections available for the whole 28. Miura, M., S. Makita, T. Iwasaki, et al., Three-dimen-
sky. Nature, 2000. 403(6765): 54–6. sional visualization of ocular vascular pathology by
15. Miller, D.T., O.P. Kocaoglu, Q. Wang, et al., Adaptive optical coherence angiography in vivo. Invest
optics and the eye (super resolution OCT). Eye (Lond), Ophthalmol Vis Sci, 2011. 52(5): 2689–95.
2011. 25(3): 321–30. 29. Hwang, J.C., R. Konduru, X. Zhang, et al., Relationship
16. Kocaoglu, O.P., S. Lee, R.S. Jonnal, et al., Imaging cone among visual field, blood flow, and neural structure
photoreceptors in three dimensions and in time using measurements in glaucoma. Invest Ophthalmol Vis Sci,
ultrahigh resolution optical coherence tomography 2012. 53(6): 3020–6.
215
 Chapter 19: Future technological advances in optical coherence tomography

30. Adhi, M. and J.S. Duker, Optical coherence tomogra- without altering RNFL thickness. Invest Ophthalmol
phy–current and future applications. Curr Opin Vis Sci, 2008. 49(1): 255–61.
Ophthalmol, 2013. 24(3): 213–21. 43. Zaveri, M.S., A. Conger, A. Salter, et al., Retinal ima-
31. Pircher, M., E. Goetzinger, R. Leitgeb, et al., ging by laser polarimetry and optical coherence
Transversal phase resolved polarization sensitive optical tomography evidence of axonal degeneration in mul-
coherence tomography. Phys Med Biol, 2004. 49(7): tiple sclerosis. Arch Neurol, 2008. 65(7): 924–8.
1257–63. 44. Bagga, H., D.S. Greenfield, W. Feuer, et al., Scanning
32. Pircher, M., E. Gotzinger, R. Leitgeb, et al., Imaging of laser polarimetry with variable corneal compensation
polarization properties of human retina in vivo with and optical coherence tomography in normal and
phase resolved transversal PS-OCT. Opt Express, 2004. glaucomatous eyes. Am J Ophthalmol, 2003. 135(4):
12(24): 5940–51. 521–9.
33. Gotzinger, E., M. Pircher, and C.K. Hitzenberger, High 45. Sehi, M., S. Ume, and D.S. Greenfield, Scanning laser
speed spectral domain polarization sensitive optical polarimetry with enhanced corneal compensation and
coherence tomography of the human retina. Opt optical coherence tomography in normal and glauco-
Express, 2005. 13(25): 10217–29. matous eyes. Invest Ophthalmol Vis Sci, 2007. 48(5):
34. Zotter, S., M. Pircher, E. Gotzinger, et al., Measuring 2099–104.
retinal nerve fiber layer birefringence, retardation, and 46. Michels, S., M. Pircher, W. Geitzenauer, et al., Value of
thickness using wide-field, high-speed polarization polarisation-sensitive optical coherence tomography
sensitive spectral domain OCT. Invest Ophthalmol Vis in diseases affecting the retinal pigment epithelium. Br
Sci, 2013. 54(1): 72–84. J Ophthalmol, 2008. 92(2): 204–9.
35. Gotzinger, E., M. Pircher, B. Baumann, et al., Three- 47. Merchant, K.Y., D. Su, S.C. Park, et al., Enhanced
dimensional polarization sensitive OCT imaging and depth imaging optical coherence tomography of optic
interactive display of the human retina. Opt Express, nerve head Drusen. Ophthalmology, 2013.
2009. 17(5): 4151–65. 48. Hedels, C. and J. Krohn, Enhanced depth imaging
36. Windisch, B.K., P.J. Harasymowycz, J.L. See, et al., optical coherence tomography of optic disc maculo-
Comparison between confocal scanning laser tomo- pathy without a visible optic pit. Clin Experiment
graphy, scanning laser polarimetry and optical coher- Ophthalmol, 2013.
ence tomography on the ability to detect localised 49. Skondra, D., T. Papakostas, and D.G. Vavvas,
retinal nerve fibre layer defects in glaucoma patients. Enhanced depth imaging optical coherence tomogra-
Br J Ophthalmol, 2009. 93(2): 225–30. phy in age-related macular degeneration. Semin
37. Ferreras, A., L.E. Pablo, A.B. Pajarin, et al., Scanning Ophthalmol, 2012. 27(5–6): 209–12.
laser polarimetry: logistic regression analysis for peri- 50. Rahman, W., F.K. Chen, J. Yeoh, et al., Enhanced
metric glaucoma diagnosis. Eye (Lond), 2009. 23(3): depth imaging of the choroid in patients with neovas-
593–600. cular age-related macular degeneration treated with
38. Yamanari, M., M. Miura, S. Makita, et al., Phase anti-VEGF therapy versus untreated patients. Graefes
retardation measurement of retinal nerve fiber layer by Arch Clin Exp Ophthalmol, 2013. 251(6): 1483–8.
polarization-sensitive spectral-domain optical coher- 51. da Silva, F.T., V.M. Sakata, A. Nakashima, et al.,
ence tomography and scanning laser polarimetry. Enhanced depth imaging optical coherence tomogra-
J Biomed Opt, 2008. 13(1): 014013. phy in long-standing Vogt-Koyanagi-Harada disease.
39. Moon, B.G., K.R. Sung, J.W. Cho, et al., Glaucoma Br J Ophthalmol, 2013. 97(1): 70–4.
progression detection by retinal nerve fiber layer 52. Nakayama, M., H. Keino, A.A. Okada, et al., Enhanced
measurement using scanning laser polarimetry: depth imaging optical coherence tomography of the
event and trend analysis. Korean J Ophthalmol, 2012. choroid in Vogt-Koyanagi-Harada disease. Retina,
26(3): 174–81. 2012. 32(10): 2061–9.
40. Hoffmann, E.M. and A. Schulze, [Glaucoma diagnosis 53. Querques, G., R. Lattanzio, L. Querques, et al.,
using scanning laser polarimetry]. Ophthalmologe, Enhanced depth imaging optical coherence tomogra-
2009. 106(8): 696–8, 700–1. phy in type 2 diabetes. Invest Ophthalmol Vis Sci, 2012.
41. Gotzinger, E., M. Pircher, B. Baumann, et al., Retinal 53(10): 6017–24.
nerve fiber layer birefringence evaluated with polari- 54. Shields, C.L., S. Kaliki, D. Rojanaporn, et al., Enhanced
zation sensitive spectral domain OCT and scanning depth imaging optical coherence tomography of small
laser polarimetry: a comparison. J Biophotonics, 2008. choroidal melanoma: comparison with choroidal
1(2): 129–39. nevus. Arch Ophthalmol, 2012. 130(7): 850–6.
42. Fortune, B., L. Wang, G. Cull, et al., Intravitreal 55. Shah, S.U., S. Kaliki, C.L. Shields, et al., Enhanced
216 colchicine causes decreased RNFL birefringence depth imaging optical coherence tomography of
 Chapter 19: Future technological advances in optical coherence tomography

choroidal nevus in 104 cases. Ophthalmology, 2012. coherence tomography. Invest Ophthalmol Vis Sci,
119(5): 1066–72. 2009. 50(1): 405–13.
56. Park, H.Y., S.H. Jeon, and C.K. Park, Enhanced depth 68. Maruko, I., T. Iida, Y. Sugano, et al., Morphologic
imaging detects lamina cribrosa thickness differences analysis in pathologic myopia using high-penetration
in normal tension glaucoma and primary open-angle optical coherence tomography. Invest Ophthalmol Vis
glaucoma. Ophthalmology, 2012. 119(1): 10–20. Sci, 2012. 53(7): 3834–8.
57. Park, S.C., C.G. De Moraes, C.C. Teng, et al., Enhanced 69. Nagase, S., M. Miura, S. Makita, et al., High-penetra-
depth imaging optical coherence tomography of deep tion optical coherence tomography with enhanced
optic nerve complex structures in glaucoma. depth imaging of polypoidal choroidal
Ophthalmology, 2012. 119(1): 3–9. vasculopathy. Ophthalmic Surg Lasers Imaging, 2012.
58. Wong, I.Y., H. Koizumi, and W.W. Lai, Enhanced 43 Online: p. e5–9.
depth imaging optical coherence tomography. 70. Usui, S., Y. Ikuno, A. Miki, et al., Evaluation of the
Ophthalmic Surg Lasers Imaging, 2011. 42 Suppl: choroidal thickness using high-penetration optical
S75–84. coherence tomography with long wavelength in highly
59. Ikuno, Y., I. Maruko, Y. Yasuno, et al., Reproducibility myopic normal-tension glaucoma. Am J Ophthalmol,
of retinal and choroidal thickness measurements in 2012. 153(1): 10–6 e1.
enhanced depth imaging and high-penetration optical 71. Jaillon, F., S. Makita, E.J. Min, et al., Enhanced imaging
coherence tomography. Invest Ophthalmol Vis Sci, of choroidal vasculature by high-penetration and dual-
2011. 52(8): 5536–40. velocity optical coherence angiography. Biomed Opt
60. Imamura, Y., T. Iida, I. Maruko, et al., Enhanced depth Express, 2011. 2(5): 1147–58.
imaging optical coherence tomography of the sclera in 72. Nakai, K., F. Gomi, Y. Ikuno, et al., Choroidal obser-
dome-shaped macula. Am J Ophthalmol, 2011. 151(2): vations in Vogt-Koyanagi-Harada disease using high-
297–302. penetration optical coherence tomography. Graefes
61. Margolis, R. and R.F. Spaide, A pilot study of enhanced Arch Clin Exp Ophthalmol, 2012. 250(7): 1089–95.
depth imaging optical coherence tomography of the 73. Chung, S.E., S.W. Kang, J.H. Lee, et al., Choroidal
choroid in normal eyes. Am J Ophthalmol, 2009. thickness in polypoidal choroidal vasculopathy and
147(5): 811–5. exudative age-related macular degeneration.
62. Spaide, R.F., Enhanced depth imaging optical coher- Ophthalmology, 2011. 118(5): p. 840–5.
ence tomography of retinal pigment epithelial detach- 74. Hong, Y.J., M. Miura, S. Makita, et al., Noninvasive
ment in age-related macular degeneration. Am J investigation of deep vascular pathologies of exudative
Ophthalmol, 2009. 147(4): 644–52. macular diseases by high-penetration optical coher-
63. Spaide, R.F., H. Koizumi, and M.C. Pozzoni, Enhanced ence angiography. Invest Ophthalmol Vis Sci, 2013.
depth imaging spectral-domain optical coherence 54(5): 3621–31.
tomography. Am J Ophthalmol, 2008. 146(4): 496–500. 75. Koizumi, H., T. Yamagishi, T. Yamazaki, et al.,
64. Hale, G.M. and M.R. Querry, Optical Constants of Subfoveal choroidal thickness in typical age-related
Water in the 200-nm to 200-microm Wavelength macular degeneration and polypoidal choroidal vas-
Region. Appl Opt, 1973. 12(3): 555–63. culopathy. Graefes Arch Clin Exp Ophthalmol, 2011.
249(8): 1123–8.
65. de Bruin, D.M., D.L. Burnes, J. Loewenstein, et al., In
vivo three-dimensional imaging of neovascular age- 76. Rishi, P., E. Rishi, G. Mathur, et al., Ocular perfusion
related macular degeneration using optical frequency pressure and choroidal thickness in eyes with poly-
domain imaging at 1050 nm. Invest Ophthalmol Vis Sci, poidal choroidal vasculopathy, wet-age-related macu-
2008. 49(10): 4545–52. lar degeneration, and normals. Eye (Lond), 2013.
66. Lee, E.C., J.F. de Boer, M. Mujat, et al., In vivo optical 77. Klein, T., R. Andre, W. Wieser, et al., Joint aperture
frequency domain imaging of human retina and detection for speckle reduction and increased collec-
choroid. Opt Express, 2006. 14(10): 4403–11. tion efficiency in ophthalmic MHz OCT. Biomed Opt
Express, 2013. 4(4): 619–34.
67. Yasuno, Y., M. Miura, K. Kawana, et al., Visualization
of sub-retinal pigment epithelium morphologies of 78. Wax, A., C. Yang, and J.A. Izatt, Fourier-domain low-
exudative macular diseases by high-penetration optical coherence interferometry for light-scattering spectro-
scopy. Opt Lett, 2003. 28(14): 1230–2.

217
 Index

Note: locators followed by ‘f’ and ‘t’ indicate figures and tables.
A-criterion, OSCAR-IB, 201–202 cornea, 15
activity of RGCs, axons, 91 iris, 15
acute optic neuritis (AON), 68. See also optic neuritis (ON) lens, 15
background, 28–29 sclera, 14–15
binocular summation, 58–59 anterior visual pathway, anatomy, 14–26
chronic and subclinical forms of, 29 anterior chamber, 14
color vision and, 52–54 anterior eye, 14–15
high contrast visual activity (HCVA), 47–49 automated quantitative retinal measurements, 24
incidence of, 28 Bruch’s membrane, 15–16
low contrast letter activity, 49–52 cornea, 15
macula and retina, changes in, 31–33 iris, 15
microcystic macular edema, 33, 36 lens, 15
motion perception testing and, 58–59 macula, fovea, 20
and multiple sclerosis (MS), 28, 29–30, 36–38 nuclear, plexiform layers, 21
and neuromyelitis optica, 29, 35–36, 37t and OCT, 14, 25
and OCT, 44–47 three-dimensional, 24–25
role in acute, chronic ON, 44–45 optic nerve head complex, 23–24
visual outcomes in, 47–59 photoreceptors, 20–21
recurrent attacks of, 30 posterior eye, 15–18
as relapse model in multiple sclerosis (MS), 36–38 retina, 18–24
and retinal nerve fiber layer (RNFL), 29–31 retinal nerve fiber layer, 21–23
fundus photography, 29 retinal vasculature, 23
loss severity predictors, 33–35 retina pigment epithelium (RPE), 18
subclinical forms of, 29 sclera, 14–15
symptoms, 28 AO. See adaptive optics (AO)
time course of RNFL changes, macula, 33 AON. See acute optic neuritis (AON)
visual acuity, 28 Arden plates, 62
visual evoked potential (VEP) testing, 57 assessing inner retinal function, 91
visual field testing, 54–57 automated quantitative retinal measurements, 24
visual recovery and, 28–29 autosomal recessive spastic ataxia of Charlevoix- Saguenay
AD. See Alzheimer’s disease (AD) (ARSACS), 137–138
adaptive optics (AO), 24, 212 axonal damage in MS, 68
afferent visual pathway (AVP), and optic neuritis, 43–44 axon loss, 168f
age-related macular degeneration/dystrophy (AMD),
133, 214 B-criterion, OSCAR-IB, 202–204
aging, 173 binocular summation
Alzheimer’s disease (AD), 162, 171 acute optic neuritis, 58–59
and dementia, 85, 132 probability and neural summation, 66–67
impaired contrast sensitivity in, 67 biomedical imaging, retinal pathology, 165–166
Aβ plaques, 132 Bowman’s layer, 15
retina challenges, 133 brain parenchymal fraction (BPF), 120
AMD. See age-related macular degeneration/dystrophy brain segmentation techniques, 121
(AMD) Bruch’s membrane, 16–18
amiloride clinical trial in optic neuritis (ACTION), 195
amyotrophic lateral sclerosis (ALS), 138, 162 C-criterion, OSCAR-IB, 199–201
anterior eye, 25 central nervous system, and retinal segmentation, 159–162
218 anatomy of, 16f central serous chorioretinopathy (CSR), 147
anterior chamber, 14, 15f choriocapillaris, 15
 Index

choroidal neovascularization (CNV), 214 pattern electroretinogram (PERG), anatomical origin, 77

choroid and choriocapillaris, posterior eye, 15 pattern ERG studies, MS, 84
chronic demyelination, neurodegeneration, 83–84 vigabatrin (Sabril) monitoring, 85
chronic relapsing inflammatory optic neuropathy visual-evoked potentials (VEPs), 90
(CRION), 28 anatomical origin, 78
clinically isolated syndrome (CIS), 122 electroretinography (ERG), 76–78
clinic-radiologic paradox, 89 conventional full-field ERG (cERG), 91
CNV. See choroidal neovascularization (CNV) definition, 77
color vision, 52–54 multi-focal ERG (mfERG), 91–92
contrast sensitivity, 67. See also low-contrast acuity; specific pattern ERG (PERG), 77, 91
types photopic negative response (PhNR), 91
chart-based methods of assessing, 62 enhanced depth imaging (EDI) SD-OCT, 24, 213
frequency-doubling technology (FDT), 62 epilepsy ataxia sensorineural deafness (EAST), 91
functional acuity contrast test (FACT), 62 epiretinal membranes (ERM), 145–146
Pelli-Robson chart, 62 ERG a-wave, 77
risk in older adults, 66 ERG b-wave, 77
Sine-Wave Contrast Test (SWCT), 62 Extended Disability Status Scale (EDSS), 66
testing, 66 clinical trials in MS, 68
vs. spatial frequency, 61 eye-tracking OCT, 210–211
conventional full-field ERG (cERG), 91
cornea, 15 FA. See fluorescein angiography (FA)
cortical plasticity, MS, 84 fellow eye in MS patients, studies, 81–84
cortical response, visual-evoked potentials, 90 fluorescein angiography (FA), 212
Fourier, Jean-Baptiste Joseph, 4
diffusion tensor imaging (DTI), 115, 187–188 Fourier domain optical coherence tomography (FD-OCT).
diminished contrast sensitivity, 67 See spectral domain optical coherence tomography
dopaminergic (DA), 85, 131 (SD-OCT)
Doppler OCT (D-OCT), 25, 212–213 Friedreich’s ataxia (FRDA), 137, 171–172
drusen, 146 full-field electroretinogram (ERG), anatomical origin,
76–77
early treatment of diabetic retinopathy study (ETDRS), 9 Functional Acuity Contrast Test (FACT), 62
EDI SD-OCT. See enhanced depth imaging (EDI) SD-OCT
electrophysiologic signature waveform, 92 ganglion cell layer and inner plexiform layer (GCIP), 36, 116
electrophysiology testing. See also electroretinography vs. RNFL measurements, 117f
(ERG) ganglion cell layer (GCL), 21
effect of visual function, 99t GCIP. See ganglion cell layer and inner plexiform layer
Alzheimer’s disease (AD), 85 (GCIP)
chronic demyelination, neurodegeneration, 83–84 glatiramer acetate (GA), 195
clinical recordings, anatomical substrate of, 76–78 glaucoma, 89, 97
cortical plasticity, MS, 84 gliotic scar, retinal pathology, 169f
ERG a-wave, 77 global brain atrophy, 121–122
ERG b-wave, 77
fellow eye in MS patients, studies, 81–84 Henle’s fiber layer. See plexiform layer
full-field electroretinogram (ERG), anatomical origin, hereditary, other rare neurodegenerative diseases, 136–138
76–77 amyotrophic lateral sclerosis, 138
generalised neurological disease, applications, 84 autosomal recessive spastic ataxia of Charlevoix-
hydroxychloroquine (Plaquenil) monitoring, 85 Saguenay, 137–138
limitations, 86 Friedreich’s ataxia, 137
medication side effects monitoring, 85 hereditary spastic paraplegia, 138
multi-focal electroretinogram (mfERG), anatomical ori- spinocerebellar ataxias, 137
gin, 77–78 Wilson’s disease, 138
multi-focal visual evoked potentials (mfVEP), 78, 90–91 hereditary spastic paraplegias (HSPs), 138
OCT, mfVEP nonconventional MRI techniques, 85 high-contrast visual acuity (HCVA), 47–49
OCT and, 76 highly reflective layer (HRL), 18
optic nerve head component (ONHC), 92 high penetration (HP-) OCT, 213–214
optic neuritis (ON) high-quality OCT scans, advice for, 206–207
evolution, studies, 79–81 histopathological examination of MS, 114
multiple sclerosis applications, 78 HP-OCT. See high penetration (HP-) OCT
origin, 77–78 HRL. See highly reflective layer (HRL)
Parkinson’s Disease (PD), 84–85 human eye, anatomy of, 17f 219
 Index

Huntington’s disease (HD), 172 medication side effects monitoring, 85

hydroxychloroquine (Plaquenil) monitoring, 85 methanol toxicity, retinal pathology, 173
microcystic macular edema (MME), 104, 179–180, 202
ICGA. See indocyanine green angiography (ICGA) fluorescein leakage, 118
I-criterion, OSCAR-IB, 202 microcysts, 33
idiopathic Parkinson’s disease (IPD), 172 RNFL loss with, 104
ILM. See inner limiting membrane (ILM) “motion-free” 3-D OCT image, 211
Impact of Visual Impairment Scale (IVIS), 66 motion perception testing, 58–59
impaired contrast sensitivity, 67 MRI. See magnetic resonance imaging (MRI)
indocyanine green angiography (ICGA), 212 MRS. See magnetic resonance spectroscopy (MRS)
INL. See inner nuclear layer (INL) MS. See multiple sclerosis (MS)
inner limiting membrane (ILM), 9 MS–eyes with contralateral MSON, time course of RNFL
inner nuclear layer (INL), 21, 202 loss, 106
microcysts, 33, 118 MS Functional Composite (MSFC), 68
inner plexiform layer (IPL), 21 MSON. See multiple sclerosis optic neuritis (MSON)
International MS Progressive Avonex Clinical Trial MSON-eyes, time course of RNFL loss in MS, 106
(IMPACT), 63 multi-focal electroretinogram (mfERG), anatomical
intracranial hypertension, 135–136 origin, 77–78
intra-individual interocular mfVEP technique, 90 multi-focal visual evoked potential (mfVEP), 78, 90–91
intra-retinal segmentation, 132 multiple sclerosis (MS), 159–161, 166–170, 198
SD OCT technology, 132 axonal damage in, 68
IPL. See inner plexiform layer (IPL) central nervous system (CNS), 124
iris, 15 etiology, 114
gray matter (GM) degeneration, 114
joint aperture (JA-) OCT, 214 low-contrast acuity in, 66, 68–69
microcystic macular edema (MMO), 178–180, 180f
lamellar macular hole, 151, 152f MRI segmentation and, 114
lamina cribrosa (LC), 24 non-conventional MRI and, 115–116
lateral geniculate nucleus (LGN), 62, 109 and optic neuritis, 37t
Leber’s hereditary optic neuropathy (LHON), 96 patients, use of Snellen chart, 68
treatment for, 99 QOL and OCT in, 63–66
lens, 15 retinal pathology, 166–170
low-contrast acuity, 66 retinal segmentation, 159–61
binocular summation and inhibition, 66–67 and spectral domain OCT, 122–123
clinical correlates, 66 and time domain OCT, 120–122
contrast sensitivity measurement of, 62–63 multiple sclerosis (MS), retinal inflammation
defined, 61 microcystic macular edema, 179–180
disorders affecting, 67–68 edema of retinal layers, 178
low-contrast letter activity, 49–52 retinal periphlebitis, 177
measurement, 63–66 in MS, 178
neural substrate for, 62 RNFL edema, interferon treatment, 181–182
and OCT in multiple sclerosis (MS), 68–69 multiple sclerosis optic neuritis (MSON),
spatial contrast, target size influences and, 61 meta-analysis, 103
low-contrast letter acuity, 49–52 definitions, 103
low-contrast sloan letter charts (LCSLC), 63 hypotheses, 103
limitations, pitfalls, 108–109
macula, fovea, 20 literature search, 103
macular B-scans, segmentation, 12 loss of RNFL, 103–104
macular changes, deep layers of retina, 31–33 time course of, 106
macular edema, 148–150 microcystic macular edema (MME), 104
macular hole, 150–151 model, 109–111
lamellar, 151 OCT in MS-eyes
macular pseudoholes, 151 with contralateral MSON, 104–106
macular scan protocols, 9–11 never MSON, 104
macular thinning predominant (MTP), 118 OCT in MSON-eyes, 103–104
macular volume scans, QC criteria, 204–206 OCT studies in MS patients, 104f, 105f
magnetic resonance imaging (MRI), 1, 114 time in MS-eyes
magnetic resonance spectroscopy (MRS), 115 with collateral MSON, 106
magnetization transfer imaging (MTI), 115, 188 never MSON, 106
220 M-cells, 62 time in MSON-eyes, 106–108
 Index

N-acetyl-aspartate (NAA), 116 and neurodegenerative diseases, 128–139

in visual cortex, 122 and neuroprotection, 191
National Institute for Neurological Disorders and Stroke optic nerve head and macular scan protocols, 9–11
(NINDS), 192 and optic nerve magnetic resonance imaging
nerve fiber layer (NFL), 9 (MRI), 184–188
neurodegeneration in MS, 115, 116 overview, 4
neurodegenerative diseases, 128–139 potential applications of, 2
neurological diseases reading center, roles, 194–195, 195t
hereditary and neurodegenerative, 136–138 retinal imaging, basic principles, 4–5
intracranial hypertension, 135–136 retinal inflammation
neuroretina and, 1, 2 multiple sclerosis (MS), 173, 177t
and optic nerve head component (ONHC), 95–96 retinal periphlebitis, 176–178, 178f, 179t
optic neuropathies detection, quantification and, 1 spectral domain (SD) vs. time domain (TD), 6–9
visual pathway assessment and, 1 swept source optical coherence tomography (SS-OCT), 6
neurology, transformation of, 1 3-D visualization, 11
neuromyelitis optica (NMO), 128–131, 161–162 time domain optical coherence tomography
retina changes, 35–36 (TD-OCT), 5–9
neuronal ceroid lipofuscinosis, 170–171 visual outcomes, 47–59
neuroretina, 1, 2 optical coherence tomography (OCT), future advances
NFL. See nerve fiber layer (NFL) adaptive optics, 212
noninflammatory disorders, 119 Doppler OCT (D-OCT), 212–213
normal-appearing cerebral white matter (NAWM), 121 eye-tracking system, 210–211
nuclear, plexiform layers, 21 high penetration (HP-) OCT, 213–214
joint aperture (JA-) OCT, 214
OAG. See openangle glaucoma (OAG) overview, 209
OCA. See optical coherence angiography (OCA) polarization-sensitive OCT (PS-OCT), 213
O-criterion, OSCAR-IB criteria, 199 swept-source (SS-) OCT, 211–212
OCT. See optical coherence tomography (OCT) optical coherence tomography (OCT), in MS meta–analysis,
edema of retinal layers, 178 103–111
ON. See optic neuritis evidence for usage of, 103–106
ONHC and RC ratio, 95 literature search, 103
ONL. See outer nuclear layers (ONL) method, 103
Open angle glaucoma (OAG), 133 in MSON-eyes, 103–104
ophthalmic disease in, optic nerve head component optical coherence tomography (OCT), multi-centre setting
(ONHC), 95–96 quality control (QC) issues
ophthalmology clinical trials, 63 high-quality scans, advice for, 206–207
OPL. See outer plexiform layer (OPL) mascular volume scans, 204–206
optical coherence angiography (OCA), 213 OSCAR-IB criteria, 199–204
optical coherence tomography (OCT) A-criterion, 201–202
in acute and chronic ON, 28–38, 44–45 B-criterion, 202–204
and afferent visual pathway, 44–47 C-criterion, 199–201
neuromyelitis optica vs. MS, 46–47 I-criterion, 202
recurrent ON, RNFL changes interpretation, 45 O-criterion, 199
brain magnetic resonance imaging in MS, 114 pathology of retina, 202
in clinical practice, 2–3, 145–153 R-criterion, 202
commercial devices, technical characteristics, 6 S-criterion, 199
derived measures, 117, 120t overview, 198
emergence of, 1 poor-quality scans, 198
future applications, 25 QC assessment reliability, 199
limitations, neurodegenerative diseases, 138–139 quantitative measurement artifacts, 198
low-contrast acuity, 61–69 (See also low-contrast acuity) optical coherence tomography (OCT), neurological practice
macular B-scans, segmentation, 12 central serous chorioretinopathy (CSR), 147
measurement beam, 205 drusen, 146
off-center placement of, ONL, 206 epiretinal membranes (ERM), 145–146
mfVEP nonconventional MRI techniques, 85 foveal cyst, 152f
in MS-eyes lamellar macular hole, 151, 152f
with contralateral MSON, 104–106, 104t macular edema, 148–50
never MSON, 104, 105t macular hole, 150–151
in multiple sclerosis, 68–69 pigment epithelium detachment (PED), 147–8
neural retina structure and, 1 posterior vitreous detachment, 150 221
 Index

optical coherence tomography (cont.) low-contrast letter acuity, 49–52

retinal detachment, 150 motion perception testing, 58
retinal nerve fiber layer (RNFL), 145 optic neuritis, 42–43
vitreomacular traction syndrome., 150 afferent visual pathway, MS clinical model, 43–44
optical coherence tomography (OCT), neurologic clinical visual evoked potential testing, 52–55, 57
trials optic nerve head, 9–11, 10f
designs for trials, 191–192 appearance on OCT, 24f
examples, 195–196 complex, 23–24
key variables of interest, 192 optic nerve head component (ONHC) response, 95
multi-center study designs, 192f in animal studies, 92
OCT as outcome measure, advantages, 191 application limitations, 96
OCT-driven trials, ophthalmology vs. neurology differ- characteristics, potential applications, 99t
ences, 194–195 eye of MS patient, 96f
overview, 193 future research, 96–100
potential roles of an OCT reading center, 195t glaucoma, 95
role of the OCT reading center, 194 Leber’s hereditary optic neuropathy (LHON), 96
sample sizes for trials, 193–194 macular scan protocols, 9–11
technical and practical issues, 192–193 in neurologic, ophthalmic disease, 95–96
optical coherence tomography (OCT), retinal pathology normal human, 92–95
aging, 173 other visual metrics, 100
Alzheimer’s disease, 171 response origination, 92
axon loss, 168f waveforms, 93f, 94f
biomedical imaging, 165–166 optic nerve imaging
Friedreich’s ataxia (FA), 171–172 clinical-radiological paradox and, 184–186
gliotic scar, 169f diffusion tensor imaging (DTI), 187–188
Idiopathic Parkinson’s disease (IPD), 172–173 magnetization transfer imaging (MTI), 188
immune cells, 169f optic nerve MRI
inner nuclear layer, 168f challenges, 186
methanol toxicity, 173 in clinical practice, 186
multiple sclerosis, 166–170 optic nerve volumes, 186–187
neuronal ceroid lipofuscinoses, 170–171 overview, 184
reactive microglia, 169f optic nerve volumes, 186–187
spincerebellar ataxia type VII, 173 optic neuritis (ON), 66. See also acute optic neuritis (AON)
vitamin B12 deficiency, 173 and afferent visual pathway, 43–44
optical coherence tomography (OCT), retinal segmentation evolution, studies, 79–81
algorithms, 157–158 microcysts, 33
Alzheimer’s disease (AD), 162 multiple sclerosis applications, 78
amyotrophic lateral sclerosis (ALS), 162 overview, 42
eye of a healthy control subject., 158f recurrent ON, RNFL changes interpretation, 45
manually segmented spectral-domain, example, 159f RNFL changes, ON with neuromyelitis optica (NMO) vs.
microcystic macular edema, 161f ON associated with MS, 46–47
multiple sclerosis (MS), 159–161 Optic Neuritis Treatment Trial (ONTT), 28
neuromyelitis optica (NMO), 161–162 Snellen charts in, 61
Parkinson’s disease (PD), 162 OSCAR-IB criteria
retianl layers, 156–157 A-criterion, 201–202
3-D images, 156 B-criterion, 202–204
optical coherence tomography (OCT), scans C-criterion, 199–201
blurred infrared (IR) image, 200 I-criterion, 202
edge of the macular volume, 206 O-criterion, 199
high-quality, practical advice for, 206–207 pathology of retina, 202
OSCAR-IB quality criteria for retinal, 199 R-criterion, 202
poor-quality of, 198 S-criterion, 199
protocol violations, 198 outer nuclear layers (ONL), 21, 204
QC rejected, 200, 201 outer plexiform layer (OPL), 21, 204
ring-scan, imperfect centering of, 198
optical coherence tomography (OCT), visual outcomes in Parkinson’s disease (PD), 131–132
acute optic neuritis idiopathic Parkinson’s disease (IPD), synucleinopathies,
binocular summation, 58 172
consequences of inflammatory injury, 44–47 movement disorder, 84–85
222 high-contrast visual acuity (HCVA), 47–49 visual and motor symptoms, 131, 162
 Index

pathology, pathological processes, 165–166 changes in optic neuritis using OCT, 29–31
pattern electroretinogram (PERG), anatomical origin, 67, edema, interferon treatment, 181–182
77, 91 imaging, fundus photography, 29
pattern ERG studies, MS, 84 innermost layer of retina, 116
P-cells, 62 loss in MS, time course, 106
PCV. See polypoidal choroidal vasculopathy (PCV) loss severity predictors, 33–35
Pelli-Robson chart, 62–63 swelling of RNFL, 33, 34f
peripapillaryretinal nerve fiber layer (p-RNFL), 116 thickness, MS pathology relationship, 110f
peripapillary ring scan, 201 thinning of, 109
photopic negative response (PhNR), 91 retinal patches, of stimulation, 92
photoreceptors, 20 retinal periphlebitis, 176–178, 179t
pigment epithelium detachment (PED), 147–148 in MS, 178
plexiform layer, 21 retinal pigment epithelium (RPE), 18
polarization-sensitive OCT (PS-OCT), 213 retinal segmentation, 157–158
polypoidal choroidal vasculopathy (PCV), 214 and central nervous system, 159–162
poor-quality OCT scans, 198 multiple sclerosis (MS), 159–161
posterior eye, 15–18 neuromyelitis optica (NMO),
Bruch’s membrane, 16–18 161–162
choroid and choriocapillaris, 15–16 and other neurological diseases, 162
retinal pigment epithelium (RPE), 18 Alzheimer’s disease (AD), 162
posterior vitreous detachment, 150 amyotrophic lateral sclerosis (ALS), 162
PS-OCT. See polarization-sensitive OCT (PS-OCT) Parkinson’s disease (PD), 162
retinal vasculature, 23
QC assessment, multi-center setting in, 199 retina stimulation, 91
quality-of-life (QOL), MS in, 64t–65t retroillumination, 62
quantitative measurement artifacts, QC issues, 198 RNFL. See retinal nerve fiber layer (RNFL)
rod photoreceptor density, 67
randomized controlled trials (RCT), 198 RPE. See retinal pigment epithelium (RPE)
R-criterion, OSCAR-IB criteria, 202
reduced contrast sensitivity, 67 scanning laser ophthalmoscope (SLO), 11
relapsing remitting MS (RRMS), 122 scanning laser polarimetry (SLP/GDx), 213
reliable QC assessment, in OCT, 199 sclera, 14–15
retina S-criterion, OSCAR-IB, 199
macula and fovea, 20 SD-OCT. See spectral domain optical coherence
nuclear and plexiform layers, 21 tomography (SD-OCT)
optic nerve head complex, 23–24 secondary and primary progressive ibudilast neuroNext
photoreceptors, 20 trial in MS (SPRINT-MS), 195
retinal layers, 19f segmentation, of OCT macular
retinal nerve fiber layer, 21–23 B-scans, 12
retinal vasculature, 23 signal-to-noise ratio (SNR), 6
retinal abnormalities, 154t Sine-Wave Contrast Test (SWCT), 62
retinal Aβ plaques, 132–133 SLO. See scanning laser ophthalmoscopic (SLO)
retinal axonal loss in MS, 103 Sloan charts, 63
retinal component (RC), 92–95 SLP/GDx. See scanning laser polarimetry (SLP/GDx)
retinal DA deficiency, 131 Snellen charts, 67, 68
retinal detachment, 150 examined LCSLC using, 63, 65
retinal disorders, evaluation, 91 in Optic Neuritis Treatment Trial (ONTT), 61
retinal dopaminergic deficiency, 67 spectral domain optical coherence tomography
retinal ganglion cells (RGCs), 21, 62 (SD-OCT), 6
in AD patients, reduction of, 133 Fourier domain technology, 6
axons of, 21–23 high resolution images, 116
damage to, 89 OCT devices, technical characteristics, 7
type A, 62 scan report, 10f, 11f
type B, 62 segmentation, brain MRI in MS, 122–123
types of, 21 spectrometer signal processing, 8f
retinal imaging, OCT, 4 studies, 120t
retinal layers, cell populations, 156–157 vs. time domain OCT, 6–9
retinal nerve fiber layer (RNFL), 9, 21–23, 145, 198 SPECTRALIS SD-OCT device, 8–9

223
 Index

spinocerebellar ataxias (SCAs), 137 Unified Parkinson Disease Rating Scale (UPDRS), 132
type VII, 137
SS-OCT. See swept-source (SS-) OCT vigabatrin (Sabril) monitoring, 85
Susac’s syndrome, 134–135 visual acuity, 28
swept-source (SS-) OCT, 6, 211–212 visual dysfunction, 68
visual evoked potential (VEP), 90
TD-OCT. See time domain optical coherence tomography anatomical origin, 78
(TD-OCT) cortical responses, 90
3-D optical coherence tomography, 11–12, 24–25 limitations, 90
time-domain OCT, brain MRI in MS, 120–122 testing, 57
time domain optical coherence tomography visual field testing, 54–57
(TD-OCT), 5–6 visual pathway, electrophysiology. See electrophysiology
light beams process, 5f testing
vs. spectral domain OCT, 6–9 vitreomacular traction syndrome, 150
total macular volume (TMV), 192
toxic and metabolic states, nutritional deficiencies, 173 Wilson’s disease (WD), 138
treatment of optic neuritis with erythropoetin
(TONE), 195 zero-delay, 212

224