Review

Optical coherence tomography in multiple sclerosis

Elliot Frohman, Fiona Costello, Robert Zivadinov, Olaf Stuve, Amy Conger, Heather Winslow, Anand Trip, Teresa Frohman, Laura Balcer

We do not have currently satisfactory clinical and anatomical correlates to gauge disability in multiple sclerosis. Lancet Neurol 2006; 5: 853–63
Structural biomarkers (such as MRI) are hindered because they cannot precisely segregate demyelination from axonal Departments of Neurology
elements of tissue injury within the CNS. Axonal degeneration in multiple sclerosis is related to irreversible disability, (E Frohman MD, A Conger,
T Frohman BA, O Stuve MD) and
which suggests that the confirmation of neuroprotective strategies needs highly quantifiable measures of axon loss
Ophthalmology (E Frohman,
that can be correlated with reliable measures of physiological function. The coupling of quantifiable measures of H Winslow MD) University of
visual function with ocular imaging techniques, such as optical coherence tomography, enables us to begin to Texas Southwestern Medical
understand how structural changes in the visual system influence function in patients with multiple sclerosis. In this Center at Dallas, USA;
Department of Neurology,
review, we consider the usefulness of optical imaging of the retina as a biomarker for neurodegeneration in multiple-
Buffalo Neuroimaging Analysis
sclerosis. Center, The Jacobs Neurological
Institute, State University of
Introduction to reduce disease related attacks, CNS lesion development, New York at Buffalo, USA
(R Zivadinov MD); Neurology
Multiple sclerosis is a common disabling progressive and to exert slight effects on disability progression.6–13 Section, Veterans
neurological disorder of young people. Most patients Alternatively, there has been little corresponding success Administration Health Care
present with a relapsing-remitting pattern of acute in changing the secondary-progressive process, System, Medical Service, Dallas
neurological dysfunction, with variable periods of suggesting that distinctively different mechanisms can Texas, USA (O Stuve); NMR
Research Unit, Department of
remission, punctuated by new exacerbations. The early underlie the substrates of inflammation and Neuroinflammation, Institute
phase of the disease process is characterised by these neurodegeneration.14,15 Notwithstanding this hypothesis, of Neurology, University
attacks and associated lesions on MRI of the brain and early untreated or inadequately treated inflammatory College London, UK (A Trip MD);
spinal cord. This early inflammatory phase of the disease events might constitute the early trigger for a series of and Department of Neurology
and Ophthalmology, University
process is the principal focus of most attempts to find injury cascades that drive permanent changes in tissue of Pennsylvania, USA
disease-modifying treatments. By contrast, primary- architecture, including demyelination, axonal injury and (L Balcer MD) and Department
progressive multiple sclerosis is an important subtype of loss, and gliosis (the collection of these components of Neurology, University of
the disease that involves steady progression of disability signifying the multiple sclerosis plaque).16 Ottawa, Canada (F Costello MD)

without evidence of exacerbations. Characterising these In the past few years, progress has been made in Correspondence to:
Elliot M Frohman, Department of
phases may be highly relevant to understanding the understanding some of the molecular mechanisms that Neurology, University of Texas
mechanisms underlying primary-progressive multiple are germane to the process of tissue injury in multiple Southwestern Medical Center at
sclerosis, given that axonal degeneration seems to be the sclerosis.14–16 However, to translate these insights into Dallas, 5323 Harry Hines
primary pathological feature, a process that is also viable treatments for patients with multiple sclerosis, we Boulevard., Dallas, Texas 75235,
USA
prominent in the progressive phase of other subtypes of will need new clinical-trial initiatives to investigate elliot.frohman@
multiple sclerosis. discrete physiological systems that provide rapid, utsouthwestern.edu.
In 1868, Charcot provided the first comprehensive accurate, and meaningful measures of changes secondary
description of both the clinical and histopathological to the disease process. However, application of the
features of multiple sclerosis.1 However, it was not until lessons learned from highly isolated systems can
1993 that the US Food and Drug Adminstration approved ultimately arise only if such changes definitely relate to
the first disease-modifying therapy for relapsing- changes occurring within the CNS in general.
remitting multiple sclerosis. Since this time four We believe that the retina and optic nerve represent a
additional drugs have been approved, all of which have model system from which to test very specific hypothesis-
their most potent effect on exacerbations (rate and driven questions about the viability of neuroprotective
severity) and MRI surrogate markers of disease activity treatment strategies in multiple sclerosis. The development
but have only slight effects on the progressive elements of optical coherence tomography (OCT) in ophthalmology
of the disease.2,3 has resulted in a powerful assessment capability for
Natural history studies have clearly described the nature diabetic optic retinopathy, glaucoma, and macular
of the disease trajectory of multiple sclerosis.4,5 In essence, degeneration. New data now suggest that the application
most patients will, over time, have a transition from a of this same technology to understanding acute and
relapsing form to a more insidious and steadily progressive chronic optic neuropathy in multiple sclerosis can be
phase of the disease, culminating in loss of ambulation, extended to clinical trials, for the ascertainment of tissue
spasticity, sphincter dysfunction, fatigue, weakness, and, preservation, and eventually even tissue restoration.
most importantly, intellectual change. The later, secondary
progressive form of the disease results in a loss of Optic neuropathy in multiple sclerosis
functional capabilities that can affect personal and Acute optic neuritis is the first attack (also known as a
professional life and lead to permanent disability. clinically isolated syndrome) in what eventually is
Most therapeutic clinical trials in multiple sclerosis confirmed as multiple sclerosis, in up to 25–50% of
have shown the capability of anti-inflammatory strategies patients.17,18 As such, this common event of multiple

http://neurology.thelancet.com Vol 5 October 2006 853

 Review

active treatment, did not (even over 5 years of follow up)

NFL
A GCL
catch up in terms of treatment benefit to those receiving
IPL treatment from onset.23 As controversy continues, early
treatment intervention is our best strategy for reducing
INL
OPL disease activity and the related risk of disability.19
Optic neuritis is an early clinical syndrome that could
ONL
ELM identify “benign” phenotypes of multiple sclerosis, with
PR IS substantial visual recovery in most patients.24,25 However,
250 μm PR OS
more sophisticated assessments show that patients with
Temporal Nasal RPE
this syndrome commonly have persistent deficits in low-
500 μm
contrast letter sensitivity and acuity, colour, and overall
visual processing. On the basis of these observations, our
B INL NFL current measures of visual function (at least those we
ONL commonly use in clinical practice) are insensitive measures
IPL of visual impairment. Furthermore, recent data have
OPL ONL
suggested that highly sensitive imaging techniques, such
as OCT, can detect important changes in measures of
IS/OS 250 μm retinal neuronal and axonal structure that are permanent.26,27
Inferotemporal RPE RPE IS/OS Superonasal We suggest that such changes happen over a short time
GCL NFL after acute optic neuritis (probably on an occult basis in the
C ONL INL
chronic optic neuropathy of multiple sclerosis) and might
IPL
OPL ELM therefore be modifiable with acute neuroprotective
ONL
interventions. If the process of neurodegeneration within
the retina is an indication of similar processes occurring
IS/OS 250 μm
more diffusely within the brain and spinal cord of patients
Inferotemporal RPE RPE IS/OS Superonasal
500 μm
with multiple sclerosis, perhaps modelling this process in
the retina will provide us with a powerful biomarker for
Figure 1: Low and high-resolution OCT images of the human retina the screening of possible neuroprotective drugs.
High-magnification ultrahigh-resolution OCT image of the healthy human macula (A). An enlargement of the
image shows the ability to see intraretinal layers that can be correlated with intraretinal anatomy: nerve fibre layer
(NFL), ganglion cell layer (GCL), inner plexiform layer (IPL), inner nuclear layer (INL), outer plexiform layer (OPL),
Retinal neurodegeneration
outer nuclear layer (ONL), external limiting membrane (ELM), photoreceptor inner and outer segments (PR IS, For as long as we have had bedside ophthalmoscopy, the
PR OS), retinal pigment epithelium (RPE). Red labels indicate high-backscattering layers, blue labels, low examiner has had the ability to identify the fundoscopic
backscattering layers. Standard stratus OCT (B) and ultrahigh-resolution OCT (C) images respectively, of the healthy hallmarks of retinal neurodegeneration, pallor, and optic
human macula. Most of the major intraretinal layers can be seen in the stratus OCT image, but the GCL and ELM can
be seen more easily in the ultrahigh-resolution OCT image. Reproduced with permission from Springer-Verlag.34
atrophy. Estimates suggest that optic-disc pallor can be
confirmed in up to 71% of eyes of patients with a history
of optic neuritis.28,29 In 1974, Frisen and Hoyt30 described
sclerosis can be studied in many patients at a time when qualitative changes in the retinal nerve fibre layer (RNFL)
the disease process is early and still a derivative of acute in patients with multiple sclerosis. In a small subsequent
inflammation. The study of patients at this time is study of 14 eyes with optic atrophy due to various causes,
probably the best stage to confirm both the diagnosis and Frisen and Quigley31 reported that visual acuity was
to begin disease-modifying therapy.19 associated with the amount of surviving axons within the
Most patients at the time of the first clinically isolated temporal quadrant of the optic nerve head.
syndrome have occult MRI activity, signifying that A formidable challenge associated with these techniques
sentinel clinical events do not represent the beginning of derives from the fact that, even in the hands of experienced
the disease process.20 Further, the presence of lesion examiners, about 50% of ganglion cells must be lost
dissemination at clinical onset predicts a more accelerated before focal RNFL defects are detectable.32 However, a
transition to clinically definite multiple sclerosis (multiple post-mortem study in 1994 corroborated qualitative
events in space and time) than does the presence of no observations and confirmed atrophy of the RNFL in 35 of
other lesions.20,21 Treatment at the time of a clinically 49 eyes with apparent optic-nerve atrophy.33 MRI has also
isolated syndrome, irrespective of the subtype of shown optic-nerve atrophy after acute optic neuritis.34–36
exacerbation, is associated with substantial benefits in Although most of the lesions that account for acute
terms of the risk of future clinical and radiographical optic neuritis are localised to the retrobulbar distribution
measures of disease activity.22 In one trial, patients who of the optic nerve, the nearly invariable changes within
received treatment with interferon beta-1a at the time of a the RNFL and macula over time, suggest that retrograde
clinically isolated syndrome derived substantial benefits axonal degeneration is an ultimate consequence of such
compared with patients assigned to receive placebo.23 attacks.36,37 Studies with an optic-nerve transection in
Patients receiving placebo for 2 years before changing to non-human primates, confirm the presence of such a

854 http://neurology.thelancet.com Vol 5 October 2006

 Review

μm 157 160 120 148

300 78 111 S
200 OD 52 75 65 T N 98
65 107 I
100 155 99 Signal strength (max 10) 10
149 134
0
0 20 40 60 80 100 120 140 160 180 200 220 240 148 137
133 129
T S N I T 113 74 S
μm 73 53 88 N T 70
300 OS 78 82 I
99 160 135
200 Signal strength (max 10) 10

100 OD (n=3) OS (n=3) OD-OS

Imas/Smax 1·00 1·08 –0·07
0 Smax/Imax 1·00 0·93 0·07
0 20 40 60 80 100 120 140 160 180 200 220 240 Smax/Tavg 2·95 2·31 0·64
T S N I T Imax/Tavg 2·96 2·50 0·47 100%
Smax/Navg 1·97 1·84 0·13 95% Normal
Max–min 148·00 124·00 24·00 5% distribution
Smax 192·00 161·00 31·00 1% percentiles
Imax 193·00 174·00 19·00 0%
Savg 146·00 137·00 9·00
lavg 134·00 135·00 –1·00
Mean thickness 110·70 107·22 3·48

Figure 2: OCT analysis from a healthy person

We show the RNFL thickness. On the left of the diagram we show the distribution of the RNFL thickness measures circumferentially around the retina. Note the
greater thickness of the superior and inferior zones of the retina (the so-called “double hump” histogram). In the centre of the figure we show retinal thickness by
clockface and quadrant sector analyses. The values in green are normal based on information derived from a normative population database. The table provides
complex analysis of these values, however, the last row provides average thickness measures, which are those most commonly used in analysis. On the right side of
the figure we show images of the optic disc, centred within the scan target (an important technical aspect of image analysis). To the right we show the corresponding
OCT generated images of the retinal layers. The top, red layer, constitutes the RNFL. The signal intensity is a measure of scan quality and should be greater than or
equal to level seven. T=temporal. S=superior. N=nasal. I=inferior. OD=right eye. OS=left eye.

dying-back axonopathy, culminating in retinal ganglion and easily measured clinical assessments (such as low
cell loss.38 Alternatively, in patients with papillitis, contrast letter acuity).
inflammation can also affect the most anterior segments
of the optic nerve (the intraocular portion) and retina. In What is OCT?
such cases, ganglion cells within the macula and their OCT is a new technological assessment strategy based on
axonal processes in the RNFL can have Wallerian the principle of echo time delay of back-scattered infrared
degeneration and permanent structural changes.39 Loss light. The process uses an interferometer in conjunction
of retinal ganglion cells has also been identified in an with a low-coherence light source. Ultrasound is an
animal model of inflammatory demyelination, associated method that produces greater tissue penetration
experimental autoimmune encephalomyelitis.40 with lower resolution, whereas OCT is contingent upon
The retina is unique as a model of neurodegeneration tissue-density-dependent reflections of infrared light,
and neuroprotection, because it contains no myelin. This resulting in better anatomical resolution. The resulting
organisation is advantageous because changes in the images allow spectacular differentiation of the major
structure of the RNFL principally represent axonal retinal layers, which can then be analysed for tissue
damage, whereas, application of myelin to the optic nerve thickness and volume, to a resolution of less than 10 μm,
begins behind the eye, at the level of the lamina cribosa. and to about 3 μm with high-resolution OCT (figure 1).41
Hence, the retina can be used to focus on the neuronal The OCT data-analysis program generates values for
and axonal components of multiple-sclerosis pathological average RNFL thickness that can be further divided into
changes. Although MRI has substantially advanced our quadrant and clockface (12 zones) sectors for more localised
ability to measure tissue injury in multiple sclerosis, the analyses (figures 2 and 3). The primary use of this
technology cannot specifically measure changes in axonal technique has been for the characterisation of retinal
integrity. By contrast, OCT (and potentially other retinal changes in patients with glaucoma.42,43 Visual-field
imaging systems) can directly link axonal damage to abnormalities in such patients have been linked to changes
clinically relevant visual outcomes. in RNFL thickness.44
We believe that the development of retinal imaging Normal values for RNFL thickness are available for age
techniques has provided us with a powerful detection stratified populations, with an average decline of about
strategy for the objective measurement of changes in 0·17% per year (table 1).45 The stability of these measures
retinal architecture that can be associated with relevant make OCT an ideal technology for the longitudinal

http://neurology.thelancet.com Vol 5 October 2006 855

 Review

OD OS

Signal strength (max 10) 7 Signal strength (max 10) 9 Parameter OD OS Diff
(OD-OS)
Thickness Foveal minimum 180 169 11
Foveal 214 215 –1
Temporal inner macula 267 269 –2
Mean Superior inner macula 283 290 –7
retinal Nasal inner macula 281 293 –12
thickness Inferior inner macula 276 283 –7
(μm) Temporal outer macula 223 234 –11
Superior outer macula 242 240 2
Nasal outer macula 260 272 –12
Inferior outer macula 237 247 –10
Superior/inferior outer 1·021 0·972 0·049
Temporal/nasal inner 0·950 0·918 0·032
Temporal/nasal outer 0·858 0·860 –0·002
Fovea 0·168 0·169 –0·001
Temporal inner macula 0·419 0·422 –0·003
Superior inner macula 0·445 0·456 –0·011
242 240 Nasal inner macula 0·442 0·46 –0·018
Volume
Inferior innermacula 0·434 0·445 –0·011
(mm3)
283 290 Temporal outer macula 1·187 1·24 –0·053 100%
Superior outer macula 1·287 1·276 –0·011 99% Normal
223 267 214 281 260 272 293 215 269 234 Nasal outer macula 1·38 1·448 –0·066 95% distribution
5%
Inferior outer macula 1·257 1·313 –0·056 1% percentiles
276 283 Total macula volume 7·023 7·232 –0·208 0%

237 247
μm μm

Figure 3: OCT assessment of a healthy person

We show the corresponding macular data from the same healthy person (in figure 2). The total macular volume and sector measures are all within the green (normal)
zone. A similar sector analysis capability is generated from the OCT software.

assessment of retinal changes, subsequent to the acute method for RNFL in multiple sclerosis. For instance,
or chronic inflammatory processes of multiple sclerosis Heidelberg retinal tomography has a slower acquisition
(figures 4 and 5). Despite the merits of this exciting new time than OCT and provides only an indirect measurement
technology, there are several potential technical artifacts of the RNFL.48 Scanning laser polarimetry can be less
that must be addressed. Most problems relate to sensitive for the detection of regional RNFL loss in the
challenges posed by specific ophthalmological disorders nasal and temporal quadrants, the latter being an area
that can restrict the achievement of adequate signal commonly affected in multiple sclerosis.49 The differences
strength and disc centring by the technician.46 between these capabilities may be relevant in multiple
sclerosis, particularly if longitudinal studies of acute optic
Other retinal-imaging techniques neuritis show anatomical patterns of RNFL loss. Further
Among imaging modalities, OCT is comparable to both studies are underway to examine the role for variable
scanning laser polarimetry (GDx with variable corneal corneal compensation in scanning laser polarimetry
compensation) and confocal scanning laser techniques to ensure uniform detection of RNFL loss.
ophthalmoscopy (Heidelberg retinal tomography) in its
capacity to discriminate between healthy eyes and eyes Use of OCT in multiple sclerosis
with glaucomatous visual-field loss.47 Although the other The first use of OCT in multiple sclerosis was reported in
techniques are comparable for detection of glaucomatous 1999 by Parisi and colleagues.50 This group assessed 14
damage, some data suggest that OCT is the best imaging patients with definite multiple sclerosis with a history of
optic neuritis associated with good visual recovery.
Despite this recovery, the researchers reported a 46%
Age (years) Mean thickness (µm) reduction in the average RNFL in affected eyes compared
16–30 128±11 with healthy eyes, and a 28% reduction in the unaffected
31–50 127±11 eyes of patients with multiple sclerosis.
51–70 120±10 In 2005, Trip and colleagues26 provided a more detailed
>70 114±9 and systematic characterisation of OCT changes
associated with inflammatory demyelinating optic
Table 1: Normal retinal nerve fibre thickness36
neuritis and visual dysfunction. The researchers studied

856 http://neurology.thelancet.com Vol 5 October 2006

 Review

μm 138 117 137

160
300 89 106 S
OD 58 55 69 T N 74
200
61 63 I
100 137
177 136 150 Signal strength (max 10) 10
0
0 20 40 60 80 100 120 140 160 180 200 220 240 84
71 99 83
T S N I T 61 31 S
μm 31 22 42 N T 26
300 OS 25 I
36
76 107 65 83
200 Signal strength (max 10) 9

100 OD (n=3) OS (n=3) OD-OS

Imas/Smax 1·02 0·96 0·06
0 Smax/Imax 0·98 1·04 –0·06
0 20 40 60 80 100 120 140 160 180 200 220 240 Smax/Tavg 2·63 4·77 –2·14 100%
T S N I T Imax/Tavg 2·68 4·58 –1·90 95% Normal
Smax/Navg 2·43 2·93 –0·51 5% distribution
Max–min 138·00 106·00 32·00 1% percentiles
Smax 181·00 124·00 57·00
0%
Imax 184·00 119·00 65·00
Savg 137·00 84·00 53·00
lavg 150·00 83·00 67·00
Mean thickness 107·62 58·83 48·78

Figure 4: OCT assessment from a patient with a remote (>2 years) history of acute optic neuritis
Note the reduction in the average RNFL on the affected side compared with the unaffected eye. T=temporal. S=superior. N=nasal. I=inferior. OD=right eye. OS=left eye.

25 patients (14 with a clinically isolated syndrome and 11 OCT and low-contrast letter acuity
with clinically definite multiple sclerosis). These patients At least 80% of patients with multiple sclerosis will have
had experienced a single episode of acute unilateral optic visual impairment over the course of the disease.51,52 The
neuritis without recurrence. Furthermore, in contrast to routine clinical assessment of suspected optic neuritis
those described in the Parisi study, these patients had involves careful assessment of corrected acuity, visual-
incomplete recovery of visual function. The analysis field analysis, colour perception, assessment for a relative
period, defined as the time from the clinical syndrome to afferent pupillary defect, and ophthalmoscopic inspection
OCT assessment, was a minimum of 1 year (proposed to of the fundus.18 However, the most practical, convenient,
allow for resolution of inflammation and oedema) and and sensitive clinical assessments include low-contrast
ranged from 1 year to 9 years. letter acuity (Sloan chart) and sensitivity (Pelli-Robson
In Trip and colleagues’ study, the patients were assessed charts) measures.53,54 These simple tests have been used
with OCT, Humphrey automated perimetry, visual in multiple sclerosis clinical trials and were primary
evoked potentials, pattern electroretinography, colour outcome measures in the Optic Neuritis Treatment
vision (Farnsworth-Munsell 100 hue test), and high- Trial.55 Monocular assessment of contrast measures
contrast letter acuity (early-treatment diabetic retinopathy provides the best sensitivity for individual deficits in each
study charts). When compared with healthy people, the optic nerve, whereas binocular assessment is focused on
RNFL was reduced by 33% (p<0·001) and the macular capturing overall or optimised visual function (“two eyes
volume was reduced by 11% (p<0·001) in patients with are better than one”; the so-called parallax effect). Data
optic neuritis. When the affected eye was compared with now suggest that visual contrast abnormalities in patients
the unaffected eye, the RNFL thickness was reduced by with multiple sclerosis predict pathophysiological
27% (p<0·001) and macular volume by 9% (p<0·001). changes as shown on visual evoked potentials.56,57
Reduction in RNFL thickness was associated with One could argue that if clinical visual measures are so
worsening in visual function by acuity, visual field, and robust, why not simply use these as the biomarker of
colour. OCT quadrant analysis showed that RNFL neurodegeneration? There is unfortunately no gold
thickness reductions in the superior and inferior zones standard that captures the full effects of an optic-neuritis
predicted corresponding field suppression by automated event. Fatigue itself (a particularly common problem in
perimetry. The researchers also confirmed an association multiple sclerosis) can notably affect the results of any
between RNFL thickness and macular volume, where 1 μm clinical variable, leading to a great deal of test–retest
of thinning in the former resulted in a 0·019 mm³ variability (table 2). Visual field results vary notably on
reduction in the latter (p<0·001). The RNFL and macular repeat testing among patients, and visual acuity measures
changes associated with optic neuritis predicted low commonly do not indicate visual dysfunction. Although
visual amplitudes (consistent with axonal degeneration) low-contrast letter acuity and sensitivity are more sensitive
but not P100 latency (a measure of myelin integrity). assessements, they too rely on an element of patient’s

http://neurology.thelancet.com Vol 5 October 2006 857

 Review

OD OS

Signal strength (max 10) 10 Signal strength (max 10) 7 Parameter OD OS Diff
(OD-OS)
Thickness Foveal minimum 122 123 –1
Foveal 151 147 4
Temporal inner macula 248 206 42
Superior inner macula 251 221 30
Mean Nasal inner macula 240 220 20
retinal Inferior inner macula 251 214 37
thickness Temporal outer macula 211 204 7
(μm) Superior outer macula 230 205 25
Nasal outer macula 254 199 55
Inferior outer macula 234 193 41
Superior/inferior outer 0·983 1·062 –0·79
Temporal/nasal inner 1·033 0·936 0·097
Temporal/nasal outer 0·831 1·025 –0·194
Fovea 0·118 0·115 0·003
Temporal inner macula 0·39 0·323 0·067
230 205 Superior inner macula 0·395 0·347 0·048
Nasal inner macula 0·378 0·345 0·033
251 221 Volume Inferior inner macula 0·395 0·337 0·058
(cm3) Temporal outer macula 1·122 1·084 0·038 100%
211 248 151 240 254 199 220 147 206 204 Superior outer macula 1·219 1·088 0·131 99% Normal
95% distribution
251 Nasal outer macula 1·351 1·058 0·293 5%
214 Inferior outer macula 1·24 1·023 0·217 1% percentiles
Total macula volume 6·614 5·725 0·889 0%
234 193
μm μm

Figure 5: OCT assessment from a patient with a remote (>2 years) history of acute optic neuritis
Bottom: we present corresponding data on the reduced total macular volume from the same eye, signifying ganglion cell loss. Note that the fovea centralis is reduced
in volume on both sides suggesting photoreceptor loss. The right side shows occult involvement. This patient had severe reduction in low-contrast letter acuity
(Sloan 1·25% chart) on the affected side.

ability, and are subject to fatigue, temperature effects Fisher and colleagues27 recently hypothesised that OCT
(ambient and core body), and other factors that might could be a valid and reproducible structural biomarker
change physiological function (a recognised characteristic for axonal loss in multiple sclerosis. The researchers
of the pathophysiology of demyelination). Ideally, it would assessed 90 patients with multiple sclerosis (with and
be a great advantage to have a structural marker, which is without a history of acute optic neuritis) and 36 disease-
non-invasive, reliable, reproducible, and correlates with free people. Scores for both low-contrast letter acuity and
other measures of visual system function we currently sensitivity were substantially worse among patients.
have at our disposal. There are no reliable imaging or Patients with a history of optic neuritis had substantially
electrodiagnostic techniques (MRI or visual evoked worse visual function than patients with multiple
potential) that we can use to quantify the effects of sclerosis but without a history of optic neuritis. The
regenerative strategies, primarily because these average RNFL thickness was low in the eyes of all patients
techniques do not specifically distinguish between with multiple sclerosis and in those with a history of
demyelinating effects from axonal damage. OCT allows optic neuritis compared with that in controls (p<0·001;
us to quantify RNFL loss in the anterior visual pathways. figure 6). Furthermore, the RNFL thickness in the eyes of
As a measure, this technique and other retinal imaging those with multiple sclerosis and optic neuritis were
assessments are not substitutes for clinical, radiographical, substantially compared with those with multiple sclerosis
or neurophysiological methods, but instead can be seen but without optic neuritis (p<0·001; figure 6). Importantly,
as a complement to these measures. visual function scores were powerful predictors of overall
Axonal loss appears to occur early in multiple sclerosis RNFL thickness among patients with multiple sclerosis
even after a single event of optic neuritis. As such, this (p<0·001). Specifically, a reduction in RNFL thickness of
pathological change may not merely be the consequence about four micrometres (95% CI; 2·7–4·9) predicted a
of the cumulative effect of repeat attacks of inflammatory one line change (worsening) on the low-contrast letter
demyelination.58 Sensitive and specific retinal imaging acuity or sensitivity charts. The Fisher study used
techniques can allow us to quantify these effects in a Spearman’s correlation coefficients (r) between overall
highly discrete system at the earliest phase of the disease average RNFL thickness and visual function scores. The
process and to monitor the effects of prevention (and associations were highly significant but slight in
even axonal preservation) strategies. magnitude, suggesting that visual dysfunction can occur

858 http://neurology.thelancet.com Vol 5 October 2006

 Review

in some patients in the absence of (or perhaps in advance

Information Limitation
of) RNFL axonal loss (r=0·33, p<0·0001 for low-contrast
letter acuity; r=0·31, p<0·0001 for contrast sensitivity; Visual acuity Global high-contrast acuity Insensitive to other visual system abnormalities
Subjective
r=0·26, p=0·0005 for high-contrast visual acuity). Unlike Subject cooperation
generalised estimations, however, simple correlations Potentially affected by fatigue, temperature,
are limited because they do not account for factors such and other factors such as infection, stress
as age and disease duration, and do not allow for Low-contrast Sensitive measure of visual function Subjective
letter acuity/ that relates to physiological and Subject cooperation
adjustment for within-patient, inter-eye correlations. sensitivity structural changes Potentially affected by fatigue, temperature,
A conspicuous outcome of this study was the association and other factors such as infection, stress
between RNFL thickness measurements and multiple Colour vision Cone vision Subjective
sclerosis associated disability, as documented by the Subject cooperation
expanded disability status scale score (p<0·02) and the Potentially affected by fatigue, temperature,
and other factors such as infection, stress
multiple sclerosis functional composite (p<0·001;
Pupillary Retinomesencephlic projections. Information limited to a restricted set of retinal
figure 7).27 RNFL thickness decreased with increasing reflexes Transmission through optic nerve, ganglion cells and their axons
expanded disability status scores (p=0·02 for linear trend, midbrain, and cranial nerve III Potentially affected by temperature and other
accounting for age and adjusting for within-patient, inter- factors such as infection
eye correlations), indicating greater degrees of axonal Funduscopic Evidence of optic pallor and atrophy Subjective
examination Non-quantitative
loss in the anterior visual pathways of patients with
Visual fields Global visual functioning of spatial Subjective
substantial neurological impairment. Furthermore, the
surround Subject cooperation
association between the temporal RNFL and macular Potentially affected by temperature and other
volume was almost identical to that reported in the Trip factors such as fatigue, infection, stress
study.26 These studies provide some data to suggest that Evoked Transmission characteristics of Potentially affected by temperature, infection
OCT is an excellent and reproducible technique for the potentials information within the visual system
Provides measure of magnitude and
quantification of axonal loss in optic neuritis and multiple latencies of responses, reflecting
sclerosis. axonal and myelin integrity
Costello and colleagues59 used OCT to quantify RNFL MRI techniques Various structural measures of tissue Long imaging times
changes among 54 patients followed prospectively after architecture. Sensitive to changes in Not entirely specific for each pathological
myelin, axons, gliosis, and process
an acute optic-neuritis event.59 The researchers looked at
inflammation
whether changes in RNFL thickness after optic neuritis
OCT Measures retinal architecture. RNFL Dependent on accurate disc centring and
predict either visual recovery or impairment. Although contains axons and glia but no myelin adequate signal strength
most patients in this study had good clinical recovery, Patients must be able to fixate
RNFL thinning occured in 40 patients (74%), and tended
Table 2: Methods for assessing the anterior visual system in multiple sclerosis
to occur within 3–6 months of the optic neuritis event.
The average RNFL value was thinner (p<0·0001) in the
affected eyes (78 μm) compared with the unaffected eyes other optic nerve pathologies (eg, tumour).62
(100 μm) of patients. Patients with incomplete visual field Recommended techniques for the detection of
sensitivity (<5 decibels) after recovery showed greater inflammatory lesions in the optic nerve include thin
RNFL loss after optic neuritis than patients with better sliced (eg, 3 mm) fat and cerebrospinal fluid suppressed
visual outcomes. Regression analyses showed a threshold coronal fast spin-echo, fluid attenuated inversion
of RNFL thickness (75 μm), below which RNFL recovery and T1-weighted post-contrast gadolinium
measurements predicted persistent visual field sequences.60,62 The presence of inflammatory lesions on
dysfunction among patients. The results of this study these sequences is slightly associated with decreased
suggest that determination of RNFL thickness can predict visual acuity and colour vision, a relative afferent
visual recovery after optic neuritis and that low RNFL pupillary defect, and reduced P100 latency on the visual
values correlate with impaired visual function. evoked potentials.62,63 Use of these techniques is restricted
by the association between optic-nerve atrophy and
MRI and the optic nerve visual function, which is modest at best. A particularly
MRI is superior to other radiological imaging techniques difficult confounder is the lack of conventional MRI in
in assessing structures associated with the optic nerve.60 its ability to differentiate the composition of tissue injury
Other functional and structural non-radiological imaging (eg, myelin vs axonal disruption).64 Alternately, there is
techniques are emerging as surrogate markers for great promise that non-conventional MRI assessments
monitoring optic-nerve axonal pathology.61 will yield sensitive and specific insights into non-invasive
lesion profiling.
Conventional MRI measures
Optic nerve imaging with conventional MRI techniques Non-conventional MRI measures
in patients with multiple sclerosis can be useful in the The development of optic-nerve atrophy, or structural
diagnosis of acute optic neuritis and the exclusion of changes on magnetisation transfer imaging, subsequent

http://neurology.thelancet.com Vol 5 October 2006 859

 Review

170 EDSS tertile

Multiple sclerosis ON eyes 110

150 Multiple sclerosis non-ON eyes
Disease-free control eyes †

Average overall RNFL thickness (µm)

130
RNFL thickness µm

90

110
*

90 70

70

50
50 Score 0–1·5 (n=18) Score 2·0–2·5 (n=17) Score 3·0–7·0 (n=19)
Overall Temporal Superior Nasal Inferior
average
Figure 7: Mean values for average overall RNFL thickness (360° around optic
Figure 6: Stratification of the eyes of patients with multiple sclerosis with or without optic neuritis and disc) across categories (tertiles) for patients with multiple sclerosis who
healthy people underwent neurological testing with the expanded disability status scale
The average RNFL thickness is significantly lower in patients with multiple sclerosis (with or without optic neuritis) (EDSS)
when compared with healthy people.24 *p<0·001 RNFL thickness from patients with multiple sclerosis compared Patients with multiple sclerosis were divided into three roughly equal groups to
with healthy people; †p=0·03 RNFL thickness comparing patients with multiple sclerosis but without optic neuritis define EDSS tertiles. RNFL thickness decreased with increasing EDSS scores
with healthy people. Reproduced with permission from Arnold. (p=0·02 for linear trend, accounting for age and adjusting for within-patient,
inter-eye correlations), indicating greater degrees of axonal loss in the anterior
visual pathways of patients with greater degrees of neurological impairment.
to optic neuritis is a very common observation.64–67 EDSS tertile ranges represent minimal abnormalities on neurological
Furthermore, occult optic neuropathy, without evidence examination with no disability (0–1·5), minimal disability in one or two domains
of an acute inflammatory event, is also not rare (as in of function (2·0–2·5), and moderate to severe disability (3·0–7·0). EDSS scores of
6·0, 6·5, and 7·0 are assigned if a patient requires unilateral assistance (cane),
primary-progressive multiple sclerosis). The optic nerve bilateral assistance (walker), or a wheelchair, respectively, for ambulation/
is a very small structure that can be subject to volumetric mobility.
changes (swelling) during the process of acute
inflammation.68 Volume changes within an anatomical structural and functional measures in the detection of
structure, having limited compliance, can be anticipated optic-nerve injury.64 A potential remedy could be
to produce substantial pathophysiological consequences multiparametric, structural–functional models that
(eg, prolongation of P100 visual evoked potential provide more global composite measures of tissue
latencies, and loss of vision; even blindness). These injury within the optic nerve. The pattern of change
symptoms explain the lack of association between early in these measures over time, within and between
optic nerve atrophy and short-term visual outcomes in treatment groups, should be investigated in patients
patients with multiple sclerosis.69 A long time (perhaps with both acute and chronic optic neuropathies. Indeed,
as long as 12 months) is needed for tissue loss to coincide several important questions need to be addressed in
with visual function.59,64,70 future validation studies using these proposed
Among other non-conventional techniques, multiparametric optic-nerve outcomes. For instance, it
magnetisation transfer imaging65,70 diffusion-weighted should be determined whether the degree of axonal loss
and tensor imaging,71,72 and functional MRI73 are the reported during the first attack of optic neuritis, is
most promising techniques for increasing our predictive of a clinically definite multiple sclerosis
understanding of the mechanisms that lead to axonal diagnosis.
optic nerve pathology. Pathological assumptions about The extent of damage to the optic nerve should also be
sensitivity and specificity of non-MRI measures to investigated in patients with different disease types (eg,
indicate tissue damage should ultimately be relapsing-remitting, secondary-progressive, primary-
corroborated with histopathological analysis. As such, progressive, progressive relapsing, Devic’s disease, etc)
we emphasise the notion that systematic banking of and disease durations. The practical use of this approach
CNS tissue is an important initiative that should be can only be realised if optic-nerve imaging variables can
widely available. be associated with global and regional measures of
neurodegeneration within the brain and spinal cord. If
MRI and OCT achieved, this modelling system may provide us with a
There is a correlation between optic-nerve atrophy and highly useful approach for the detection and monitoring
thinning of the RNFL on OCT.64 This finding suggests of the effects of new therapeutic strategies for multiple
the presence of an intrinsic variability among different sclerosis.

860 http://neurology.thelancet.com Vol 5 October 2006

 Review

Animal modelling of the molecular

neurobiology of retinal degeneration Search strategy and selection criteria
OCT could become a valuable method for the quantification References for this review were derived from searches in
of axonal loss within the CNS in animal models of multiple MEDLINE, PubMed (without restriction to the year of
sclerosis and for the testing of new therapeutic drugs in publication), and through the authors’ files. The search terms
terms of neuroprotective action. Experimental autoimmune used were “optic neuritis”, “optical coherence tomography”,
encephalomyelitis and Theiler murine encephalomyelitis “GDx”, “Heidelberg retinal tomography”, “retinal nerve fiber
were used to study inflammatory mechanisms underlying layer”, “multiple sclerosis”, “visual evoked potentials”, “low
autoimmune disease in the CNS and to investigate contrast letter acuity” and “sensitivity”. The search was last
infectious and parainfectious events underlying updated in May, 2006.
demyelination, respectively.74,75 Whereas experimental
autoimmune encephalomyelitis is induced by inoculation
of animals with myelin autoantigens in complete Freund’s to track the process of neurodegeneration within the
adjuvant, Theiler murine encephalomyelitis develops in retina of patients with multiple sclerosis with optic
susceptible rodents after intracerebral injection with neuropathy. The successful use of this technology in
Theiler murine encephalomyelitis virus.74,75 Axonal loss is a clinical trials will be predicated on the confirmation that
histopathological feature of experimental autoimmune RNFL thickness and macular volume analyses can serve
and Theiler murine encephalomyelitis.76–85 In some models as a surrogate biomarker and primary outcome measure
of experimental autoimmune encephalomyelitis, axonal to confirm the neuroprotective effects of new drugs. In
loss within the optic nerve has already been reported.74–76,80 the future, it may be feasible for changes within the
Several investigators have successfully used OCT in retina to be linked with corresponding visual field,
rats and mice to image retinal structures.86–90 Measures of multifocal visual evoked potentials, and MRI optic nerve
retinal thickness done by OCT correlate with histological segmentation maps. We believe that the association
analysis.86–89 In addition, loss of retinal fibres is associated between severe visual dysfunction and pathology in the
with decreased neuronal function as measured by retina is now beyond doubt. This conclusion is an
electroretinography.88 Notwithstanding the potential important advance in understanding the pathophysiology
merits of this application to animal modelling of multiple of optic neuritis in particular and potentially the disease
sclerosis, formidable challenges remain, such as process within the brain and spinal cord in general.
increased absorbency of biological material at certain Advances in this area of multiple-sclerosis neurobiology
wavelengths of light and an increase in coherence length will be germane to changing the natural history of the
secondary to spectral dependence of the absorbency, disorder. Success in further development and application
which is substantially diminished in rodent eyes. Optical of OCT and associated technologies can potentially result
imaging of rodent eyes is difficult given the small in substantial validation of the effort; to translate
diameter and correspondingly shorter path length in the hypothesis driven questions about how to manipulate
anterior visual system. Although studies with mice have the mechanisms underlying the pathophysiology of
used custom-designed OCT instruments, there are no disease into effective prevention treatment strategies.
clear physical restrictions that preclude the use of Contributors
commercially available human units. All authors contributed to the editing and critical revision of the review.
OCT in animal models of multiple scleorsis offers EF developed the review and was involved with the writing of the review.
FC contributed to writing the section on the application of OCT in
substantial advances over conventional histopathological multiple sclerosis. RZ contributed to the sections on MRI. OS
methods for the quantification of axonal loss within the contributed to the section on animal modeling and OCT. AC generated
optic nerve and retina. Instead of temporal cross-sectional the normal and multiple sclerosis OCT data sets and contributed to the
assessments of the brain and spinal-cord tissues, writing of the paper. HW generated and analysed the normal multiple
sclerosis OCT data sets and wrote the sections describing this data and
longitudinal assessments can be done in the same the figure legends; HW also added information on the technical errors
animal. This is particularly important in testing the associated with OCT. AT contributed to drafting, editing, and critical
influence of drugs and genetic factors on axonal loss in revision of the entire review. TF contributed to the drafting. LB
autoimmune disease. By contrast to conventional contributed to the writing of the entire review.
histopathological methods to assess neurodegenerative Conflicts of interest
processes in animals, the use of OCT avoids any ex-vivo OS is a speaker for and receives honoraria from Teva Neuroscience and
Serona. RZ is a speaker at Teva Neuroscience, Biogen Idec, Serona, and
manipulation of tissue and does not require active Pizer; RZ also receives grants from Teva Neuroscience, Biogen Idec, and
participation on the part of the anaesthetised animal. Aspreva. EF consults at Contacor, Biogen Idec, Pizer, Serona
Pharmaceuticals.
OCT as a Biomarker Acknowledgments
Although there have been few studies that use OCT in Supported by The Lone Star Chapter of the National Multiple Sclerosis
Society, the “Once Upon A Time”, the Cain/Denius Comprehensive Center
multiple sclerosis, we already have compelling
for Mobility Research, the Irene Wadel and Robert Atha fund, the Kenney
information to suggest that this powerful assessment Marie Dixon Pickens fund, the Jean Ann and Steve Brock Fund for Medical
strategy can provide valid, reliable, and reproducible data Sciences, the Hawn Foundation (EMF), and Karen and Walter Levy.

http://neurology.thelancet.com Vol 5 October 2006 861

 Review

References 26 Trip SA, Schlottmann PG, Jones SJ, et al. Retinal nerve fiber layer
1 Charcot JM. Histologie de la sclerose en plaques. Gazette des axonal loss and visual dysfunction in optic neuritis. Ann Neurol
Hopitaux, Paris 1868; 41: 554–55. 2005; 58: 383–91.
2 Goodin D, Frohman EM, Garmany GP, et al. Disease modifying 27 Fisher JB, Jacobs DA, Markowitz C, et al. Relation of visual funciton
therapies in multiple sclerosis. Neurology 2002; 58: 169–78. to retinal nerve fiber layer thickness in multiple sclerosis.
3 Corboy J, Goodin DS, Frohman EM. Disease modifying therapy in Ophthalmology 2006; 113: 324–34.
mulitple sclerosis. Curr Treat Options Neurol 2003; 5: 35–54. 28 Trobe JD, Glaser JS, Cassady JC. Optic atrophy: differential diagnosis
4 Weinshenker BG, Bass B, Rice GPA, et al. The natural history of by fundus observation alone. Arch Ophthalmol 1980; 98: 1040–45
multiple sclerosis: a geographically based study I–clinical course 29 MacFadyen DJ, Drance SM, Douglas GR, et al. The retinal nerve
and disability. Brain 1989; 112: 133–46. fiber layer, neuroretinal rim area, and visual evoked potentials in
5 Weinshenker BG, Bass B, Rice GPA, et al. The natural history of MS. Neurology 1988; 38: 1353–58.
multiple sclerosis: a geographically based study II–predictive value 30 Frisen L, Hoyt WF. Insidious atrophy of retinal nerve fibers in
of the early clinical course. Brain 1989; 112: 1419–28. multiple sclerosis: funduscopic identification in patients with and
6 IFNB Multiple Sclerosis Study Group. Interferon beta-1b is effective without visual complaints. Arch Ophthalmol 1974; 92: 91–97.
in relapsing-remitting multiple sclerosis I: clinical results of a 31 Frisen L, Quigley HA. Visual acuity in optic atrophy: a quantitative
multicenter, randomized, double-blind, placebo-controlled trial. clinicopathological analysis. Graefes Arch Clin Exp Ophthalmol 1984;
Neurology 1993; 43: 655–61. 222: 71–74.
7 IFNB Study Group, University of British Columbia MS/MRI 32 Quigley HA, Addicks EM. Quantitative studies of retinal nerve fiber
Analysis Group. Interferon β-1b in the treatment of MS: final layer defects. Arch Ophthalmol 1982; 100: 807–14.
outcome of the randomized controlled trial. Neurology 1995; 33 Kerrison JB, Flynn T, Green WR. Retinal pathologic changes in
45: 1277–85. multiple sclerosis. Retina 1994; 14: 445–51.
8 Jacobs LD, Cookfair DL, Rudick RA, et al. Intramuscular interferon 34 Hickman SJ, Brex PA, Brierley CM, et al. Detection of optic nerve
beta-1a for disease progression in relapsing multiple sclerosis. atrophy following a single episode of unilateral optic neuritis by
Ann Neurol 1996; 39: 285–94. MRI using a fat-saturated short-echo fast FLAIR sequence.
9 Johnson KP, Brooks MD, Cohen JA, et al. Copolymer 1 reduces Neuroradiology 2001; 43: 123–28.
relapse rate and improves disability in relapsing-remitting multiple 35 Inglese M, Ghezzi A, Bianchi S, et al. Irreversible disability and
sclerosis: results of a phase III multicenter, double-blind, placebo- tissue loss in multiple sclerosis: a conventional and magenetization
controlled trial. Neurology 1995; 45: 1268–76. transfer magnetic resonance imaging study of the optic nerves.
10 PRISMS Study Group. Randomised double-blind placebo-controlled Arch Neurol 2002; 59: 250–55.
study of interferon β1a in relapsing/remitting multiple sclerosis. 36 Hickman SJ, Toosy AT, Jones SJ, et al. A serial MRI study following
Lancet 1998; 352: 1498–504. optic nerve mean area in acute optic neuritis. Brain 2004;
11 Jacobs LD, Beck RW, Simon JH, et al. Intramuscular interferon 127: 2498–505.
beta-1a therapy initiated during a first demyelinating event in 37 Garner S. Optic neuropathy in multiple sclerosis. Arch Ophthalmol
multiple sclerosis. N Engl J Med 2000; 343: 898–904. 1953; 50: 718–26.
12 Polman CH, O’Connor PW, Havrdova E, et al. A randomized, 38 Quigley HA, Davis EB, Anderson DR. Descending optic nerve
placebo-controlled trial of natalizumab for relapsing multiple degeneration in primates. Invest Ophthalmol Vis Sci 1977; 16: 841–49.
sclerosis. N Engl J Med 2006; 354: 899–910. 39 Lightman S, McDonald WI, Bird AC, et al. Retinal venous
13 Rudick RA, Stuart WH, Calabresi PA, et al. Natalizumab plus sheathing in optic neuritis: its significance for the pathogenesis of
interferon beta-1a for relapsing multiple sclerosis. N Engl J Med multiple sclerosis. Brain 1987; 110: 405–14.
2006; 354: 911–23. 40 Hobom M, Storch MK, Weissert R, et al. Mechanisms and time
14 Frohman EM, Fillipi M, Stuve O, et al. Characterizing the course of neuronal degeneration in experimental autoimmune
mechanisms of progression in multiple sclerosis. Arch Neurol 2005; encephalomyelitis. Brain Pathol 2004; 14: 148–57.
62: 1345–56. 41 Ko TH, Fujimoto JG, Schuman JS, et al. Comparison of ultrahigh
15 Frohman EM, Stuve O, Havrodova E, et al. Therapeutic and standard resolution optical coherence tomography for imaging
considerations for disease progression in multiple sclerosis. macular pathology. Ophthalmology 2005; 112: 1922–35.
Arch Neurol 2005; 62: 1519–30. 42 Jeoung JW, Park KH, Kim TW, Khwarg SI, Kim DM. Diagnostic
16 Frohman EM, Raine C, Racke MK. Multiple Sclerosis: the plaque ability of optical coherence tomography with a normative database
and its pathogenesis. N Engl J Med 2006; 354: 942–55. to detect localized retinal nerve fiber layer defects. Ophthalmology
17 McDonald WI, Barnes D. The ocular manifestations of multiple 2005; 112: 2157–63.
sclerosis I: abnormalities of the afferent visual system. 43 Sihota R, Sony P, Gupta V, Dada T, Singh R. Diagnostic capability of
J Neurol Neurosurg Psychiatr 1992; 55: 747–52. optical coherence tomography in evaluating the degree of
18 Frohman EM, Frohman TC, Zee DS, McColl R, Galetta S. Neuro- glaucomatous retinal nerve fiber damage. Invest Ophthalmol Vis Sci
ophthalmology of multiple sclerosis. Lancet Neurol 2005; 4: 111–21. 2006; 47: 2006–10.
19 Frohman EM, Havrdova E, Lublin F, et al. Most patients with 44 Kanamori A, Nakamura M, Escano MF, et al. Evaluation of the
multiple sclerosis or a clinically isolated syndrome should be glaucomatous damage on retinal nerve fiber layer thickness
treated at the time of diagnosis. Arch Neurol 2006; 63: 614–19. measured by optical coherence tomography. Am J Ophthalmol 2003;
20 Frohman EM, Goodin D, Calabresi P, et al. The utility of MRI in 135: 513–20.
suspected MS. Neurology 2003; 61: 602–11. 45 Kanamori A, Escano MF, Eno A, et al. Evaluation of the effect of
21 Brex PA, Ciccerelli O, O’Riorda JI, et al. A longitudinal study of aging on retinal nerve fiber layer thickness measured by optical
abnormalities on MRI and disability from multiple sclerosis. coherence tomography. Ophthalmologica 2003; 217: 273–78.
N Engl J Med 2002; 346: 158–164. 46 Ray R, Stinnett SS, Jaffe GJ. Evaluation of image artifact by optical
22 Beck RW, Chandler DL, Cole SR, et al. Interferon beta-1a for early coherence tomography of retinal pathology. Am J Ophthalmol 2005;
multiple sclerosis: CHAMPs trial subgroup analysis. Ann Neurol 139: 18–29.
2002; 51: 481–90. 47 Medeiros FA, Zangwill LM, Bowd C, Weinreb RN. Comparison of
23 Kinkel RP, Kollman C, O’Connor P, et al. IM interferon beta-1a the GDx VCC scanning laser polarimeter, HRT II confocal scanning
delays definite multiple sclerosis 5 years after a first demyelinating laser ophthalmoscope, and Stratus OCT optical coherence
event. Neurology 2006; 66: 678–84. tomography for the detection of glaucoma. Arch Ophthalmol 2004;
24 Ramsaransing G, Maurits N, Zwanikken C, De Keyser J. Early 122: 827–37.
prediction of a benign course of multiple sclerosis on clinical 48 Jaffe GJ, Caprioli J. Optical coherence tomography to detect and
grounds: a systematic review. Mult Scler 2001; 7: 345–47. manage retinal disease and glaucoma. Am J Ophthalmol 2004;
25 Optic Neuritis Study Group. Visual function more than 10 years 137: 156–69.
after optic neuritis: experience of the Optic Neuritis Treatment Trial. 49 Monteiro ML, Medeiros FA, Ostroscki MR. Quantitative analysis of
Am J Ophthalmol 2004; 137: 77–83. axonal loss in band atrophy of the optic nerve using scanning laser
polarimetry. Br J Ophthalmol 2003; 87: 32–37.

862 http://neurology.thelancet.com Vol 5 October 2006

 Review

50 Parisi V, Manni G, Spadaro M, et al. Correlation between 71 Iwasawa T, Matoba H, Ogi A, et al. Diffusion-weighted imaging of
morphological and functional retinal impairment in multiple the human optic nerve: a new approach to evaluate optic neuritis in
sclerosis patients. Invest Ophthalomol Vis Sci 1999; 40: 2520–27. multiple sclerosis. Magn Reson Med 1997; 38: 484–91.
51 Leibowitz U, Alter M. Optic nerve involvement and diplopia as 72 Trip SA, Wheeler-Kingshott C, Jones SJ, et al. Optic nerve diffusion
initial manifestations of multiple sclerosis. Acta Neurol Scand 1968; tensor imaging in optic neuritis. Neuroimage 2006; 30: 498–505.
44: 70–80. 73 Rombouts SA, Lazeron RH, Scheltens P, et al. Visual activation
52 Sorensen TL, Frederiksen JL, Bronnum-Hansen H, Petersen HC. patterns in patients with optic neuritis: an fMRI pilot study.
Optic neuritis as onset manifestation of multiple sclerosis: Neurology 1998; 50: 1896–99.
a nationwide, long-term survey. Neurology 1999; 53: 473–78. 74 Zamvil SS, Steinman L. The T lymphocyte in experimental allergic
53 Balcer LJ, Maier ML, Cohen JA, et al. Contrast letter acuity as a encephalomyelitis. Annu Rev Immunol 1990; 8: 579–621.
visual component for the multiple sclerosis functional composite. 75 Rodriguez M, Oleszak E, Leibowitz J. Theiler’s murine
Neurology 2003; 61: 1367–73. encephalomyelitis: a model of demyelination and persistence of
54 Baier ML, Cutter GR, Rudick RA, et al. Low-contrast letter acuity virus. Crit Rev Immunol 1987; 7: 325–65.
testing captures visual dysfunction in patients with multiple 76 Hayreh SS, Massanari RM, Yamada T, Hayreh SM. Experimental
sclerosis. Neurology 2005; 64: 992–95. allergic encephalomyelitis I: optic nerve and central nervous system
55 Trobe JD, Beck RW, Moke PS, Cleary PA. Contrast sensitivity and manifestations. Invest Ophthalmol Vis Sci 1981; 21: 256–69.
other vision tests in the Optic Neuritis Treatment Trial. 77 Guy J, Ellis EA, Kelley K, Hope GM, Rao NA. Quantitative analysis
Am J Ophthalmol 1996; 121: 547–53. of labelled inner retinal proteins in experimental optic neuritis.
56 Sisto D, Trajano M, Vetrugno M, et al. Subclinical visual Curr Eye Res 1989; 8: 253–60.
involvement in multiple sclerosis:a study by MRI, VEPs, frequency 78 Guy J, Ellis EA, Tark EF III, Hope GM, Rao NA. Axonal transport
doubling perimetry, standard perimetry, and contrast sensitivity. reductions in acute experimental allergic encephalomyelitis:
Invest Ophthalmol Vis Sci 2005; 46: 1264–68. qualitative analysis of the optic nerve. Curr Eye Res 1989; 8: 261–69.
57 Weinstock-Guttman B, Baier M, Stockton R, et al. Pattern reversal 79 Maehlen J, Olsson T, Zachau A, Klareskog L, Kristensson K. Local
visual evoked potentials as a measure of visual pathway pathology enhancement of major histocompatibility complex (MHC) class I
in multiple sclerosis. Mult Scler 2003; 9: 529–34. and II expression and cell infiltration in experimental allergic
58 Trapp BD, Peterson J, Ransohoff RM, Rudick R, Mork S, Bo L. encephalomyelitis around axotomized motor neurons.
Axonal transection in the lesions of multiple sclerosis. N Engl J Med J Neuroimmunol 1989; 23: 125–32.
1998; 338: 278–85. 80 White SR, Samathanam GK, Bowker RM, Wessendorf MW.
59 Costello F, Coupland S, Hodge W, et al. Quantifying axonal loss Damage to bulbospinal serotonin-, tyrosine hydroxylase-, and TRH-
after optic neuritis with optical coherence tomography. Ann Neurol containing axons occurs early in the development of experimental
2006; 59: 963–69. allergic encephalomyelitis in rats. J Neurosci Res 1990; 27: 89–98.
60 Zivadinov R, Bakshi R. Optic neuritis diagnosis and prognosis of 81 Zhu B, Moore GR, Zwimpfer TJ, et al. Axonal cytoskeleton changes
multiple sclerosis. In: Minagar A, ed. Inflammatory, Pathology, in experimental optic neuritis. Brain Res 1999; 824: 204–17.
Etiology, Diagnosis, and New Treatment Strategies. Totowa, NJ: 82 Mancardi G, Hart BA, Capello E, et al. Restricted immune
Humana Press, 2005: 55–90. responses lead to CNS demyelination and axonal damage.
61 Sergott RC. Optical coherence tomography: measuring in-vivo J Neuroimmunol 2000; 107: 178–83.
axonal survival and neuroprotection in multiple sclerosis and optic 83 Kornek B, Storch MK, Weissert R, et al. Multiple sclerosis and
neuritis. Curr Opin Ophthalmol 2005; 16: 346–50. chronic autoimmune encephalomyelitis: a comparative quantitative
62 Rocca MA, Hickman SJ, Bo L, et al. Imaging the optic nerve in study of axonal injury in active, inactive, and remyelinated lesions.
multiple sclerosis. Mult Scler 2005; 11: 537–41. Am J Pathol 2000; 157: 267–76.
63 Simon JH, McDonald WI. Assessment of optic nerve damage in 84 Mancardi G, Hart B, Roccatagliata L, et al. Demyelination and
multiple sclerosis using magnetic resonance imaging. J Neurol Sci axonal damage in a non-human primate model of multiple
2000; 172 (suppl 1): S23–26. sclerosis. J Neurol Sci 2001; 184: 41–49.
64 Trip SA, Schlottmann PG, Jones SJ, et al. Optic nerve atrophy and 85 Wujek JR, Bjartmar C, Richer E, et al. Axon loss in the spinal cord
retinal nerve fibre layer thinning following optic neuritis: evidence determines permanent neurological disability in an animal model
that axonal loss is a substrate of MRI-detected atrophy. Neuroimage of multiple sclerosis. J Neuropathol Exp Neurol 2002; 61: 23–32.
2006; 31: 286–93. 86 Fukuchi T, Takahashi K, Shou K, Matsumura M. Optical coherence
65 Hickman SJ, Toosy AT, Jones SJ, et al. Serial magnetization transfer tomography (OCT) findings in normal retina and laser-induced
imaging in acute optic neuritis. Brain 2004; 127: 692–700. choroidal neovascularization in rats. Graefes Arch Clin Exp
66 Hickman SJ, Toosy AT, Jones SJ, et al. A serial MRI study following Ophthalmol 2001; 239: 41–46.
optic nerve mean area in acute optic neuritis. Brain 2004; 87 Sho K, Takahashi K, Fukuchi T, Matsumura M. Quantitative
127: 2498–505. evaluation of ischemia-reperfusion injury by optical coherence
67 Hickman SJ, Brierley CM, Brex PA, et al. Continuing optic nerve tomography in the rat retina. Jpn J Ophthalmol 2005; 49: 109–13.
atrophy following optic neuritis: a serial MRI study. Mult Scler 2002; 88 Li Q, Timmers AM, Hunter K, et al. Noninvasive imaging by optical
8: 339–42. coherence tomography to monitor retinal degeneration in the
68 Youl BD, Turano G, Towell AD, et al. Optic neuritis: swelling and mouse. Invest Ophthalmol Vis Sci 2001; 42: 2981–89.
atrophy. Electroencephalogr Clin Neurophysiol Suppl 1996; 46: 173–79. 89 Horio N, Kachi S, Hori K, et al. Progressive change of optical
69 Hickman SJ, Brex PA, Brierley CM, et al. Detection of optic nerve coherence tomography scans in retinal degeneration slow mice.
atrophy following a single episode of unilateral optic neuritis by Arch Ophthalmol 2001; 119: 1329–32.
MRI using a fat-saturated short-echo fast FLAIR sequence. 90 Grieve K, Paques M, Dubois A, Sahel J, Boccara C, Le Gargasson JF.
Neuroradiology 2001; 43: 123–28. Ocular tissue imaging using ultrahigh-resolution, full-field optical
70 Inglese M, Ghezzi A, Bianchi S, et al. Irreversible disability and coherence tomography. Invest Ophthalmol Vis Sci 2004; 45: 4126–31.
tissue loss in multiple sclerosis: a conventional and magnetization
transfer magnetic resonance imaging study of the optic nerves.
Arch Neurol 2002; 59: 250–55.

http://neurology.thelancet.com Vol 5 October 2006 863