ARTICLE OPEN ACCESS

Chronic demyelination exacerbates neuroaxonalloss in patients with MS with unilateraloptic neuritisYuyi You MD PhD Michael H Barnett PhD FRACP Con Yiannikas FRACP John Parratt FRACP

Jim Matthews MStat Stuart L Graham PhD FRANZCO and Alexander Klistorner MD PhD

Neurol Neuroimmunol Neuroinflamm 20207e700 doi101212NXI0000000000000700

Correspondence

Dr You

yuyiyougmailcom

AbstractObjectiveTo examine the effect of chronic demyelination in the optic nerve of patients with MS onprogressive loss of retinal ganglion cell (RGC) axons

MethodsProgressive retinal nerve fiber layer (RNFL) loss as measured by optical coherence tomog-raphy was longitudinally examined in 51 patients with MS with a history of unilateral opticneuritis (ON) and 25 normal controls Patients were examined annually with amedian of 4-yearfollow-up Pairwise intereye comparison was performed between ON and fellow non-ON(NON) eyes of patients withMS using the linear mixed-effects model and survival analysis Thelatency asymmetry of multifocal visual evoked potential (mfVEP) was used to determine thelevel of demyelination in the optic nerve

ResultsAlthough both ON and NON eyes demonstrate significantly faster loss of RGC axons com-pared with normal subjects ON eyes with severe chronic demyelination show acceleratedthinning in the RNFL in the temporal sector of the optic disc (temporal RNFL [tRNFL])compared with fellow eyes (evidenced by both the linear mixed-effects model and survivalanalysis) Furthermore progressive tRNFL thinning is associated with the degree of optic nervedemyelination and reflects the topography of pathology in the optic nerve More rapid axonalloss in ON eyes is also functionally evidenced by mfVEP amplitude reduction which correlateswith the level of optic nerve demyelination

ConclusionsAlthough the effect of demyelination on axonal survival has been demonstrated in experimentalstudies our results provide first clinically meaningful evidence that chronic demyelination isassociated with progressive axonal loss in human MS

From the Save Sight Institute (YY AK) The University of Sydney Faculty of Medicine and Health Sciences (YY SLG AK) Macquarie University Brain and Mind Centre (MHB)The University of Sydney Sydney Neuroimaging Analysis Centre (MHB AK) Department of Neurology (CY JP) Royal North Shore Hospital and Sydney Informatics and DataScience Hub (JM) The University of Sydney NSW Australia

Go to NeurologyorgNN for full disclosures Funding information is provided at the end of the article

The Article Processing Charge was funded by the authors

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 40 (CC BY-NC-ND) which permits downloadingand sharing the work provided it is properly cited The work cannot be changed in any way or used commercially without permission from the journal

Copyright copy 2020 The Author(s) Published by Wolters Kluwer Health Inc on behalf of the American Academy of Neurology 1

Axonal loss is now accepted as the major cause of irreversibleneurologic disability in MS Acute inflammatory de-myelination is thought to be a principal cause of axonaltransection and subsequent axonal degeneration1 Howeverduring the chronic phase of MS even in the absence of acuteinflammation there is still considerable neuroaxonal lossevidenced by ongoing brain atrophy and worsening of clinicalstatus23 This suggests that there are other mechanisms in-volved in neurodegenerative changes in MS and emphasizesthe shortcomings of the current disease-modifying therapieswhich diminish the formation of new inflammatory lesionsbut have little impact on axonal loss during the progressivestage of the disease2

In addition laboratory studies have indicated that loss ofmyelin can eventually lead to neuroaxonal damage4 Thereis evidence that permanent demyelination may contributeto accelerated axonal degeneration by increasing vulnera-bility of axons to physiologic stress5 Furthermore lack ofneurotrophic support from oligodendrocytes and disrup-tion of normal axon-myelin interactions may also lead todegeneration of chronically demyelinated axons46 Al-though numerous animal and in vitro studies have dem-onstrated negative effects of demyelination on axonalsurvival the clinical significance of chronic demyelinationas a precursor of axonal damage in humans has never beenascertained Myelin pathology has also been identified inmany neurodegenerative diseases other than MS7ndash9 andmay play an important role in their pathogenesis More-over transneuronal spread of degeneration along humanneuronal projections could also be partially mediated bydemyelination9 Therefore an understanding of whetherchronic demyelination can cause neuroaxonal loss inhumans is of paramount importance

The aim of the current study was to investigate the effect ofchronic demyelination on neuroaxonal loss in humans usingthe visual system as a model The degree of demyelinationfollowing acute optic neuritis (ON) can be accurately esti-mated using the latency of visual evoked potentials(VEPs)1011 and neuroaxonal loss of retinal ganglion cells(RGCs) can be monitored with a high level of precision usingoptical coherence tomography (OCT)12 Furthermore usingintrasubject comparison of 2 eyes in patients with a history ofunilateral ON intersubject variability is eliminated and theimpact of other factors (such as age sex disease durationtreatment and the effect of retrochiasmal lesions13) on theevaluation of an effect of chronic demyelination considerablyreduced

MethodsParticipantsFifty-one consecutive patients with relapsing-remittingMS witha history of unilateral ON (reported by patients and confirmedby treating neurologists) from 4 tertiary neuro-ophthalmologyor neurology clinics in Sydney (Royal North Shore HospitalBrain amp Mind Centre Inner West Neurology and the SaveSight Institute) were recruited in this prospective cohort studyMS was diagnosed according to the 2010 revised McDonaldcriteria14 Exclusion criteria included a history of other ocularneurologic or systemic diseases that could affect the resultsOnly patients who did not haveON12months before the studywere included The median post-ON duration in the studycohort was 4 years (range 1ndash26 years) One patient had a re-lapse of ON during the follow-up and only the data collectedbefore the flare-up were included in analysis Patients un-derwent regular ocular examinations (including visual acuityusing the Early Treatment Diabetic Retinopathy Study low-contrast chart) OCT scans and multifocal visual evoked po-tential (mfVEP) recordings and both eyes were analyzedPatients withMSwere examined annually and were followed upfor amedian of 40 years (range 09ndash76 years) (n = 21le2 yearsof follow-up n = 12 3ndash4 years of follow-up n = 18 ge5 years offollow-up) In addition 25 healthy participants with similar sexand age distributions were also recruited as normal controls(median follow-up of 31 years [09ndash78]) The data analyzed inthis study were collected between July 2010 and June 2018

Standard protocol approvals registrationsand patient consentsThe study adhered to the tenets of the Declaration of Helsinkiand was approved by the Human Research Ethics Committeeof the University of Sydney Written consent was signed by allparticipants

OCT scansAll participants underwent a peripapillary ring scan anda macular radial pattern scan using the Heidelberg SpectralisOCT and Eye Explorer software version 6950 (HeidelbergEngineering Germany) as described previously1516 In thisstudy the OCT data were reported according to the APOS-TEL recommendations17 The pupils of participants were notdilated and the scans were performed by the same operatorand device under room light conditions The peripapillaryring scan (manual placement of the ring diameter =350 mm) was used to obtain retinal nerve fiber layer (RNFL)thickness measures Both global and sectoral (includingtemporal nasal superior and inferior) RNFL thicknesses

GlossaryGCIPL = ganglion cell and inner plexiform layer LCVA = low-contrast visual acuity NON = non-ON OCT = opticalcoherence tomographyON = optic neuritis pRNFL = global peripapillary retinal nerve fiber layerRGC = retinal ganglion cellRNFL = retinal nerve fiber layer tRNFL = temporal RNFL VEP = visual evoked potential

2 Neurology Neuroimmunology amp Neuroinflammation | Volume 7 Number 3 | May 2020 NeurologyorgNN

were analyzed The follow-up function was activated whenscans were performed to ensure that the scans were obtainedat the same locations as the baseline scans All images werechecked by 2 investigators (YY and AK) and those withpoor quality and segmentation errors were excluded to ensurethat all OCT images included in the analysis fulfilled theOSCAR-IB criteria18 Baseline RNFL and ganglion cell innerplexiform layer (GCIPL) thicknesses are reported for allcases Due to the fact that substantial thinning of the RNFLwas present in ON eyes at baseline relative change of RNFLwas used for analysis which has also been used by Talmanet al19 previously Longitudinal analysis of relative loss ofGCIPL was not able to be used for this study as the innerplexiform layer is relatively preserved and separating the 2layers by segmentation is not reliable15 A detailed explanationof GCIPL issues and schematic diagram is provided in sup-plementary materials (linkslwwcomNXIA224)

Multifocal VEP recordingsThe level and topography of demyelination in the visualpathways was assessed by the latency delay of mfVEP recor-ded using the Vision Search system (VisionSearch SydneyAustralia) with standard stimulus conditions as describedpreviously1620 In brief 4 gold-disc electrodes (Grass WestWarwick RI) were used for bipolar recording with 2 elec-trodes positioned 4 cm on either side of the inion one elec-trode 25 cm above and another 45 cm below the inion in themidline Electrical signals were recorded along 2 channelsmeasured as the difference between superior and inferior andbetween the left and right electrodes All the VEP traces werechecked by 2 experts (YY and AK) to ensure the quality ofsignal Intereye asymmetry of mfVEP latency was used todetermine the level of ON-related demyelination(postchiasmaloptic radiation lesions are assumed to havesimilar effects on latency delay in both eyes) For VEP anal-ysis16 the largest peak-to-trough amplitude within the intervalof 70ndash210 ms was selected and the second peak of the tracewas used to determine the latency

StatisticsStatistical analysis was performed using SPSS (version 240IBM) and Graphpad Prism (version 70 Graphpad La JollaCA) Baseline characteristics are summarized using de-scriptive statistics Mean and CIs are reported for longitudinalcontinuous variables Frequencies are reported for categoricaldata Longitudinal OCT measures were analyzed in both eyesin the study subjects for intereye comparisons Annual RNFLprogression rates were estimated and compared by usinglinear mixed-effects models that included age sex and diseaseduration (for comparison between ON vs non-ON [NON]eyes) as covariates The models have a multilevel structurewith repeated measures nested within eye (leftright) nestedwithin subjects The frequency distribution of dependentvariables was visually confirmed as normal and no trans-formation was required Boxplots showed that the assumptionof equal variances of groups was valid Subjects had varyingnumbers of repeated measures and this missing data were

handled by the mixed model restricted maximum likelihoodestimation without the need for imputation or deletion ofcases In addition patients were separated into 2 groups fora further subanalysis based on the level of demyelination inthe optic nerve (determined by asymmetry of mfVEP latencylt10 ms and ge10 ms) Kaplan-Meier event rate curves and thelog-rank (Mantel-Cox) test were also used to assess the cu-mulative risk ratio of progression and compared between ONand fellow NON eyes The first time that more than 5 ofRNFL thickness loss was found as compared to the baselinewas defined as the end point in survival analysis21 Five per-cent was selected as a statistically meaningful end point todetermine progression based on previously reported intertestcoefficient of variation of RNFL measures to be between 1and 22223 Correlation between optic nerve demyelination(determined by mfVEP latency asymmetry) and asymmetryin the progression rate in temporal RNFL (tRNFL)mfVEPamplitude between ON and NON eyes was examined usingbivariate correlation and partial correlation controlling forage sex and disease duration A p value of less than 005 wasconsidered significant

Data availabilityThe authors confirm that the data supporting the findings ofthis study are available within the article and from the cor-responding author on reasonable request

ResultsDemographic data of all the participants are shown in thetable At baseline both ON and NON eyes exhibited signif-icant RGC neuronal and axonal loss as compared to age- andsex-matched control eyes which is consistent with previousOCT studies in MS24 Compared with NON eyes eyes witha history of ON demonstrated significantly worse low-contrast visual acuity (LCVA) and thinner global and tem-poral RNFL Furthermore there was significant prolongationof mfVEP latency in ON eyes compared with NON eyeswhich indicates a considerable degree of chronic ON-relateddemyelination (figure e-1 linkslwwcomNXIA224)

Progressive RNFL thinning inONandNONeyesThe rate of progressive RGC neuronal and axonal loss wasexamined in patients with a history of unilateral ON usingintrasubject between-eye comparison Because we have pre-viously demonstrated that tRNFL is the most sensitive pa-rameter to detect progressive loss of RGC fibers in MS15

longitudinal loss of tRNFL was analyzed first The linearmixed-effects model showed that age and disease duration atbaseline were not associated with progressive tRNFL changeAlthough male subjects had overall thinner tRNFL (p = 002)sex did not have significant effects on the tRNFL progressionrate Both ON (minus13 per year 95CI minus17 to minus08) andNON fellow (minus10 per year 95 CI minus13 to minus07) eyesdisplayed faster rates of tRNFL loss than normal controls(minus06 per year 95 CI minus11 to minus02 p lt 0001 vs ON

NeurologyorgNN Neurology Neuroimmunology amp Neuroinflammation | Volume 7 Number 3 | May 2020 3

and p = 0008 vs NON) Although ON eyes demonstratedfaster tRNFL thinning compared with fellow NON eyes thisdifference was not statistically significant (mean difference =minus03 [95 CI minus07 to 01] p = 01) (figure e-2 linkslwwcomNXIA224 for the trend of tRNFL thinning in μm inON and NON eyes)

Similar analysis applied to the global peripapillary retinalnerve fiber layer (pRNFL) measurement also demonstrateda trend for reduction of RNFL in both ON and NON eyesRelative pRNFL thinning was significantly faster in ON eyes(minus05 per year 95 CI minus08 to minus03) and NON eyes(minus05 per year 95 CI minus07 to minus04) compared withnormal eyes (minus02 per year 95CI minus04 to 00 p = 004vs ON and p = 001 vs NON) However no difference in therate of pRNFL thinning between ON and fellow NON eyeswas detected (p = 063)

Survival analysis demonstrates faster thinningof tRNFL in ON eyesAlthough mean rates of progression were not significantlydifferent between ON and NON eyes time to reach an endpoint did show a difference Because survival analysis has beenshown to demonstrate high sensitivity in detecting subtledifference in RNFL thinning in patients with glaucoma21 itwas applied here to investigate RNFL progression in patientswith MS with unilateral ON More than 5 of RNFL thick-ness loss was defined as the end point The analysis demon-strated significantly faster progressive tRNFL loss in ON eyes

compared with their fellow NON eyes (p = 0001 figure 1A)The difference between ON and NON eyes in pRNFL pro-gression did not reach statistical significance (p = 008 figure1B) To determine whether the above borderline change inpRNFL was mainly driven by tRNFL loss RNFL changes inthe other sectors were also evaluated The subanalysis dem-onstrates no statistical difference in RNFL progression in thenasal (p = 065) superior (p = 095) and inferior (p = 014)disc areas between ON and NON eyes

Faster thinning of tRNFL in ON eyes isassociated with the degree of opticnerve demyelinationTo investigate whether the tRNFL loss in ON eyes was as-sociated with the degree of underlying chronic optic nervedemyelination the patients were subdivided into 2 groupsbased on severity of optic nerve demyelination IntereyemfVEP latency asymmetry was used to characterize the de-gree of demyelination caused by ON1011 There were 24patients with no or mild ON-related chronic demyelination(latency asymmetry lt10 ms average 21 plusmn 46 ms) and 27patients with moderate to severe optic nerve demyelination(latency asymmetry ge10 ms average 219 plusmn 88 ms) In themore severe demyelination group progressive loss of tRNFLwas significantly faster in the ON eyes than in the fellow eyesevidenced by both the linear mixed-effects model (meandifference = minus06 per year 95 CI minus01 to minus10 p =002) and survival analysis (p = 0002) By contrast no sig-nificant intereye difference was observed in the group with

Table Participant demographics and OCT data at baseline

MS (n = 51) Normal (n = 25) p Value

Age (yrs mean [SD]) 3930 (930) 3882 (1073) 084a

Sex (malefemale) 1140 718 057b

Disease duration (yrs median [range]) 4 (1ndash23) na na

EDSS score (median [range]) 10 (0ndash6) na na

Eyes ON n = 51 NON n = 51 n = 50 (1)c (2)d (3)d

tRNFL (μm mean [SD]) 4865 (1437) 6155 (1174) 7335 (1247) lt0001 lt0001 lt0001

pRNFL (μm mean [SD]) 7867 (1492) 9059 (1132) 9728 (826) lt0001 lt0001 0001

GCIPL (μm mean [SD]) 4930 (577) 5652 (523) 6111 (387) lt0001 lt0001 lt0001

VEP latency (ms mean [SD]) 1654 (159) 1526 (108) na lt0001 na na

25 LCVA (letter score mean [SD])e 555 (152) 676 (98) na lt0001 na na

125 LCVA (letter score mean [SD])e 438 (133) 548 (121) na lt0001 na na

Abbreviations EDSS = Expanded Disability Status Scale ETDRS = the Early Treatment Diabetic Retinopathy Study GCIPL = ganglion cell inner plexiform layerLCVA = low-contrast visual acuity OCT = optical coherence tomography NON = nonoptic neuritis ON = optic neuritis pRNFL = global peripapillary retinalnerve fiber layer tRNFL = temporal retinal nerve fiber layer VEP = visual evoked potentialComparison of eyes (1) ON vs NON (2) ON vs normal (3) NON vs normal Disease-modifying therapies at baseline included interferon beta-1b (n = 12)glatiramer acetate (n = 9) fingolimod (n = 11) teriflunomide (n = 5) dimethyl fumarate (n = 3) natalizumab (n = 6) alemtuzumab (n = 1) and not on anytreatment (n = 4)a Unpaired t testb Fisher exact testc Paired t testd Generalized estimating equatione ETDRS letter scores

4 Neurology Neuroimmunology amp Neuroinflammation | Volume 7 Number 3 | May 2020 NeurologyorgNN

minimal degree of chronic demyelination (linear mixed-effects model mean difference = minus01 per year 95 CIminus08 to 07 p = 089 survival analysis p = 020 figure 2A)Survival analysis of pRNFL in the severe latency delay groupalso showed a borderline difference in ON eyes in the pro-gression rate compared with NON (p = 008) (figure 2B)

In addition accelerated tRNFL thinning was also significantlyassociated with the degree of ON-related optic nerve de-myelination (r = minus038 p = 0007) (figure 2C) This resultsupports the assumption that a chronic deficit of myelin isassociated with progressive axonal loss in MS

Preferential loss of tRNFL in ON eyes reflectsthe topography of optic nerve demyelinationAs described above only tRNFL demonstrates acceleratedthinning in ON eyes compared with fellow NON eyes Be-cause tRNFL consists of the RGC fibers projecting to thecentral part of the retina it raises the question of whetherthose fibers are more susceptible to demyelination Thereforethe degree of ON-related demyelination at different eccen-tricities of the retina was examined using ring-based analysis ofasymmetry of the mfVEP latency between ON and fellowNON eyes (figure 3A)

The highest level of relative latency delay in the optic nervefibers was found in the optic nerve fibers corresponding to thecentral 2 rings of the mfVEP stimulus which represent thefoveal to parafoveal region of the visual field (figure 3B) Opticnerve fibers projecting to rings 1ndash2 showed significantlyhigher level of latency delay than those projecting to rings 3ndash5(p lt 0001 2-way analysis of variance)

The fact that more severe latency delay is topographicallyrelated to the area demonstrating the most profound RNFLthinning supports the notion that accelerated axonal loss inMS is associated with chronic demyelination

Faster axonal loss in ON eyes is also evidencedby VEP amplitude reductionThe amplitude of the mfVEP represents an objective func-tional measure of the visual pathway While in acute ONreduction of the mfVEP amplitude is a result of conductionblock caused by inflammation in post-acute (chronic) stageON the mfVEP amplitude mainly reflects permanent axonaldamage along the visual pathway25ndash27 Therefore the re-lationship between longitudinal change of mfVEP amplitudeand degree of chronic optic nerve demyelination was in-vestigated Because of variability of mfVEP amplitude onlythe subjects who had at least 3 follow-up visits were included(n = 28) Although 30 of 51 patients had at least 3 follow-upvisits in 2 patients the VEP data for progression analysis werenot available Therefore only 28 patients were included in themfVEP substudy Two patients had very low (nonrecordable)amplitude even at the baseline visit Therefore data from 26patients were analyzed The analysis demonstrated a signifi-cant difference in the rate of amplitude reduction betweenON and fellow NON eyes Annual reduction of mfVEP am-plitude was 13 plusmn 07 faster in ON eyes compared withNON eyes (p = 0007)

Both relative reduction of mfVEP amplitude in ON eyes (vsamplitude of the fellow NON eyes) and increase of intereyeamplitude asymmetry correlated significantly with baselinelatency delay (r = minus050 p = 002 figure 4) In addition the

Figure 1 ON eyes showed a faster rate of RNFL loss in the temporal disc area

Comparisons of progressive tRNFL (A) and pRNFL (B) thinning between ON and fellow NON eyes in patients with unilateral ON (n = 51) using Kaplan-Meiercurves and the log-rank test More than 5 of RNFL thickness loss as compared to the baseline was defined as the end point NON = non-ON ON = opticneuritis pRNFL = global peripapillary retinal nerve fiber layer RNFL = retinal nerve fiber layer tRNFL = temporal RNFL

NeurologyorgNN Neurology Neuroimmunology amp Neuroinflammation | Volume 7 Number 3 | May 2020 5

Figure 2 Faster tRNFL thinning in ON eyes is associated with the level of chronic demyelination in the optic nerve

(AandB) Comparisons of progressive RNFL loss betweenONandNONeyes in 2 patient groupswith different levels of demyelination in the optic nerve determinedby asymmetry ofmfVEP latency (n = 27 latency asymmetry ge10ms n = 24 latency asymmetry lt10ms) Faster tRNFL thinningwas only present in the patients withmore severeoptic nerve demyelination (latency asymmetry ge10ms) Therewas also borderline difference (p = 0086) in pRNFL progression betweenONand fellowNON eyes in patients with advanced demyelination by contrast no difference was observed in the latency asymmetry lt10 ms group which also supports anassociation between accelerated RNFL loss in the temporal disc area and the level of optic nerve demyelination (C) There was a significant correlation betweenasymmetry of the tRNFL progression rate (year) and optic nerve demyelination as determined by intereye asymmetry ofmfVEP latency betweenONand fellowNONeyes (r = minus038 p = 0007) which remained significant after controlling for age sex and disease duration by partial correlation (p = minus037 p = 001) NON= non-ON ON = optic neuritis pRNFL = peripapillary retinal nerve fiber layer RNFL = retinal nerve fiber layer tRNFL = temporal RNFL VEP = visual evoked potential

6 Neurology Neuroimmunology amp Neuroinflammation | Volume 7 Number 3 | May 2020 NeurologyorgNN

slope of mfVEP amplitude reduction also correlates withtRNFL thinning (r = 040 p = 0045)

No evidence of new subclinical ON duringthe studyTo exclude the potential effect of new subclinical ON thelongitudinal analysis of mfVEP latency was also performed inthe same cohort of patients There was no increase in theVEP latency in ON eyes detected In fact there was slight butsignificant shortening of the latency in ON eyes between thebaseline and the last visit (16 plusmn 29 ms p = 001 paired ttest) whereas the latency of NON eyes remained stable

(mean difference = 009 plusmn 32 ms p = 08 paired t test)suggesting no active inflammatory demyelination duringfollow-up

DiscussionAlthough chronic (not ON related) axonal and neuronal lossof RGC has been well documented in patients with MS theunderlying mechanisms remain elusive24 Primary neuro-degeneration and transsynaptically mediated damage havebeen suggested as potential causes of chronic axonal loss in

Figure 3 Accelerated tRNFL thinning reflects the topography of optic nerve demyelination in ON eyes

(A) Representative mfVEP traces from a patientwith MS with a history of unilateral ON Signifi-cant latency delay is observed in the ON eyeAsymmetry of the latency of mfVEP was calcu-lated for each of the 5 eccentricity rings to de-termine the pattern of optic nervedemyelination (B) Optic nerve fibers projectingto rings 1ndash2 showed a significantly higher level ofdemyelination than those projecting to rings 3ndash5(p = 00005 2-way ANOVA) The central retinalarea topographically corresponds to the RNFL inthe temporal sector of the optic disc (shown inred) This result could explain the observation ofaccelerated tRNFL thinning inON eyes comparedwith fellowNON eyes Comparedwith traditionalpVEP mfVEP used in this study represents a su-perior technique Because of opposite orienta-tion of upper and lower calcarine cortices thedipoles generating responses from stimulationof the upper and lower hemifields are oppositeand therefore producing oppositely orientedwaveforms which results in a cancelation ef-fect40 Therefore pVEP actually represents thedifference between signals from upper andlower calcarine cortices In addition the majorityof the pVEP power is generated by 3ndash5 centraldegrees of the visual field practically ignoringthe paracentral andmidperipheral fields ANOVA= analysis of variance NON = non-ON ON = opticneuritis pVEP = pattern VEP RNFL = retinal nervefiber layer tRNFL = temporal RNFL

NeurologyorgNN Neurology Neuroimmunology amp Neuroinflammation | Volume 7 Number 3 | May 2020 7

MS91328ndash31 RGC and its axon are thought to be analogous tobrain gray and white matter Therefore studying mecha-nisms of its damage in MS may provide crucial insights intopathogenesis of the disease The current study examined theeffect of chronic demyelination of the optic nerve on loss ofRGC axons in patients with MS and provided the first directin vivo evidence supporting contribution of permanent de-myelination to progressive axonal loss in MS This as-sumption is based on the fact that the faster rate of RNFLthinning in ON eyes is related to both severity and topog-raphy of demyelination in the optic nerve Accelerated lossof axons was predominantly seen in tRNFL which repre-sents RGC axons subserving the macular area and are mostseverely demyelinated

It should be noted that even in tRNFL the preferential lossof axons in ON eyes was only apparent in cases with sub-stantial demyelination This emphasizes the fact thata clinical history of ON by itself is not sufficient to ade-quately estimate the presence of residual demyelinationwhich can vary significantly between patients Thereforea more direct measure must be used to assess the degree ofchronic loss of myelin in the optic nerve This together withthe use (in some studies) of less sensitive time-domainOCT and absence of tRNFL sector analysis may partiallyexplain the fact that previously reported longitudinal studieshave not demonstrated a faster rate of axonal loss in ONeyes1932ndash35 In a 5-year study by Garcia-Martin et al32 nodifferences were detected between eyes with and withoutprevious ON Although VEP was recorded its correlationwith the rate of RNFL thinning was not analyzed Fur-thermore latency asymmetry of only 3 ms between ON and

NON eyes reported by the authors indicates minimal de-gree of residual demyelination in the cohort of patients withON Although the study by Talman et al19 also failed toshow a difference in the relative rate of RNFL loss betweenON and NON eyes time since the episode of acute ONappeared to be a major determinant of RNFL thinningsuggesting that the duration of chronic demyelination mayhave some effects on axonal loss A similar result was re-cently reported by Behbehani et al33 however only totalRNFL thickness was measured and no estimation of thedegree of demyelination was performed

One limitation of this study is a potential deleterious effect ofsubclinical ON on axonal integrity during the course of thefollow-up Although recurrence of ON is usually reported bypatients and detected by worsening of VEP latency VEP hasbeen described to be highly sensitive even in subclinicalcases36 Longitudinal latency analysis in the current study didnot reveal any evidence of subclinical inflammatory de-myelination in the optic nerve Conversely slight latencyshortening in ON eye was observed which can be explainedby preferential axonal loss of the most extensively demyeli-nated axons leading to paradoxically improved nerve con-duction Furthermore the potential effect of disease-modifying therapy was not examined due to the relativelysmall sample size for each medication group which representsanother potential limitation of the study In addition in thisstudy we were not able to identify any difference in LCVAdeterioration between ON and NON eyes which was possi-bly due to low sensitivity of the subjective test Many ON eyesalready had significant LCVA loss at baseline which mightalso mask subtle further deterioration during follow-up Al-though using VEP latency alone as the measurement of de-myelination may by itself be a limiting factor we are notfamiliar with any other techniques which would allow toprecisely measure degree of demyelination In additioncompared with traditional pattern VEP which has been usedin MS remyelination trials with variable results mfVEP usedin the current study represents a superior technique (figure 3)Future MS clinical trials could consider using mfVEP toprovide more accurate measure of remyelination Althoughslightly more time consuming the latency results of mfVEPare more meaningful and reliable which results in significantlysmaller sample size needed to demonstrate remyelination37

Although we did not perform mfVEP in control subjects(which may be considered as another limitation) the con-clusions were mainly drawn from intrasubject analysis be-tween ON and NON in patients with MS

To the best of our knowledge this study provides the firstclinical evidence supporting the hypothesis that chronic de-myelination is associated with progressive axonal damage inMS Although the result of the study indicates that the short-term (3ndash5 years) effect of chronic demyelination on axonalsurvival is small it is clearly measurable suggesting its po-tentially detrimental role in disease progression and func-tional deterioration during long course of the disease The

Figure 4 Correlation between baseline latency delay andasymmetry of the rate of mfVEP amplitude re-duction (r = minus047 p = 002)

The correlation remained significant after adjusted for age sex and diseaseduration (r = minus050 p = 002) VEP = visual evoked potential

8 Neurology Neuroimmunology amp Neuroinflammation | Volume 7 Number 3 | May 2020 NeurologyorgNN

current study therefore strongly supports the need for de-velopment of new pro-remyelination therapies a rapidlyemerging research area in MS3839 In addition given the factthat demyelination has also been observed as a pathologiccomponent in other neurodegenerative disorders78 as well asbeing implicated in transneuronal degeneration9 the revealedassociation between chronic demyelination and axonal lossmay have broader implications for human neurodegenerativediseases

Study fundingThis study was funded by the National Multiple SclerosisSociety (NMSS) the National Health and Medical ResearchCouncil (NHMRC) the Ophthalmic Research Instituteof Australia (ORIA) and the Sydney Medical SchoolFoundation

DisclosureDisclosures available NeurologyorgNN

Publication historyReceived by Neurology Neuroimmunology amp NeuroinflammationOctober 7 2019 Accepted in final form January 30 2020

References1 Trapp BD Peterson J Ransohoff RM Rudick R Mork S Bo L Axonal transection in

the lesions of multiple sclerosis N Engl J Med 1998338278ndash2852 Beck ES Reich DS Brain atrophy in multiple sclerosis how deep must we go Ann

Neurol 201883208ndash209

3 Wang C Barnett MH Yiannikas C et al Lesion activity and chronic demyelinationare the major determinants of brain atrophy in MS Neurol Neuroimmunol Neuro-inflamm 20196e593 doi 101212NXI0000000000000593

4 Nave KA Myelination and support of axonal integrity by glia Nature 2010468244ndash252

5 Kornek B Storch MK Weissert R et al Multiple sclerosis and chronic autoimmuneencephalomyelitis a comparative quantitative study of axonal injury in active inactiveand remyelinated lesions Am J Pathol 2000157267ndash276

6 Wilkins A Majed H Layfield R Compston A Chandran S Oligodendrocytes pro-mote neuronal survival and axonal length by distinct intracellular mechanisms a novelrole for oligodendrocyte-derived glial cell line-derived neurotrophic factor J Neurosci2003234967ndash4974

7 Fornari E Maeder P Meuli R Ghika J Knyazeva MG Demyelination of superficialwhite matter in early Alzheimerrsquos disease a magnetization transfer imaging studyNeurobiol Aging 201233428e7ndash419

8 Dean DC III Sojkova J Hurley S et al Alterations of myelin content in Parkinsonrsquosdisease a cross-sectional neuroimaging study PLoS One 201611e0163774

9 You Y Joseph C Wang C et al Demyelination precedes axonal loss in thetransneuronal spread of human neurodegenerative disease Brain 2019142426ndash442

10 You Y Klistorner A Thie J Graham SL Latency delay of visual evoked potential isa real measurement of demyelination in a rat model of optic neuritis Invest Oph-thalmol Vis Sci 2011526911ndash6918

11 van der Walt A Kolbe S Mitchell P et al Parallel changes in structural and functionalmeasures of optic nerve myelination after optic neuritis PLoS One 201510e0121084

12 Martinez-Lapiscina EH Arnow S Wilson JA et al Retinal thickness measured withoptical coherence tomography and risk of disability worsening in multiple sclerosisa cohort study Lancet Neurol 201615574ndash584

13 Klistorner A Graham EC Yiannikas C et al Progression of retinal ganglion cell loss inmultiple sclerosis is associated with new lesions in the optic radiations Eur J Neurol2017241392ndash1398

14 Polman CH Reingold SC Banwell B et al Diagnostic criteria for multiple sclerosis2010 revisions to the McDonald criteria Ann Neurol 201169292ndash302

15 Graham E You Y Yiannikas C et al Progressive loss of retinal ganglion cells andaxons in non-optic neuritis eyes in multiple sclerosis a longitudinal optical coherencetomography study Invest Ophthalmol Vis Sci 2016572311ndash2317

16 Shen T You Y Arunachalam S et al Differing structural and functional patterns ofoptic nerve damage in multiple sclerosis and neuromyelitis optica spectrum disorderOphthalmology 2019126445ndash453

17 Cruz-Herranz A Balk LJ Oberwahrenbrock T et al The APOSTEL recom-mendations for reporting quantitative optical coherence tomography studies Neu-rology 2016862303ndash2309

18 Tewarie P Balk L Costello F et al TheOSCAR-IB consensus criteria for retinal OCTquality assessment PLoS One 20127e34823

19 Talman LS Bisker ER Sackel DJ et al Longitudinal study of vision and retinal nervefiber layer thickness in multiple sclerosis Ann Neurol 201067749ndash760

20 Triplett JD Yiannikas C Barnett MH et al Pathophysiological basis of low con-trast visual acuity loss in multiple sclerosis Ann Clin Transl Neurol 201851505ndash1512

21 Shen T Gupta VK Klistorner A Chitranshi N Graham SL You Y Sex-specific effectof BDNF Val66Met genotypes on the progression of open-angle glaucoma InvestOphthalmol Vis Sci 2019601069ndash1075

22 Vizzeri G Weinreb RN Gonzalez-Garcia AO et al Agreement between spectral-domain and time-domain OCT for measuring RNFL thickness Br J Ophthalmol200993775ndash781

23 WadhwaniM Bali SJ Satyapal R et al Test-retest variability of retinal nerve fiber layerthickness and macular ganglion cell-inner plexiform layer thickness measurementsusing spectral-domain optical coherence tomography J Glaucoma 201524e109ndashe115

24 Petzold A Balcer LJ Calabresi PA et al Retinal layer segmentation in multiplesclerosis a systematic review and meta-analysis Lancet Neurol 201716797ndash812

25 You Y Thie J Klistorner A Gupta VK Graham SL Normalization of visual evokedpotentials using underlying electroencephalogram levels improves amplitude re-producibility in rats Invest Ophthalmol Vis Sci 2012531473ndash1478

26 You Y Klistorner A Thie J Gupta VK Graham SL Axonal loss in a rat model of opticneuritis is closely correlated with visual evoked potential amplitudes using electro-encephalogram based scaling Invest Ophthalmol Vis Sci 2012533662

27 Klistorner A Arvind H Garrick R Graham SL Paine M Yiannikas C In-terrelationship of optical coherence tomography and multifocal visual-evokedpotentials after optic neuritis Invest Ophthalmol Vis Sci 2010512770ndash2777

28 Petzold A de Boer JF Schippling S et al Optical coherence tomography in mul-tiple sclerosis a systematic review and meta-analysis Lancet Neurol 20109921ndash932

29 Klistorner A Sriram P Vootakuru N et al Axonal loss of retinal neurons in multiplesclerosis associated with optic radiation lesions Neurology 2014822165ndash2172

30 Green AJ McQuaid S Hauser SL Allen IV Lyness R Ocular pathology in multiplesclerosis retinal atrophy and inflammation irrespective of disease duration Brain20101331591ndash1601

31 You Y Graham EC Shen T et al Progressive inner nuclear layer dysfunction in non-optic neuritis eyes in MS Neurol Neuroimmunol Neuroinflamm 20185e427 doi101212NXI0000000000000427

Appendix Authors

Name Location Contribution

Yuyi YouMD PhD

Save Sight Institute TheUniversity of Sydney NSWAustralia

Study design acquisitionof data analysis orinterpretation of datadraftingrevising themanuscript andobtaining funding

Michael HBarnettPhD FRACP

Brain and Mind Centre TheUniversity of Sydney NSWAustralia

Acquisition of data anddraftingrevising themanuscript

ConYiannikasFRACP

Royal North Shore HospitalSydney NSW Australia

Acquisition of data anddraftingrevising themanuscript

JohnParrattFRACP

Royal North Shore HospitalSydney NSW Australia

Acquisition of data

JimMatthewsMStat

Sydney Informatics andData Science Hub TheUniversity of Sydney NSWAustralia

Analysis or interpretationof data

Stuart LGrahamPhDFRANZCO

Macquarie UniversitySydney NSW Australia

Draftingrevising themanuscript

AlexanderKlistornerMD PhD

Save Sight Institute TheUniversity of Sydney NSWAustralia

Study design analysis orinterpretation of datadraftingrevising themanuscript andobtaining funding

NeurologyorgNN Neurology Neuroimmunology amp Neuroinflammation | Volume 7 Number 3 | May 2020 9

32 Garcia-Martin E Ara JR Martin J et al Retinal and optic nerve degeneration inpatients with multiple sclerosis followed up for 5 years Ophthalmology 2017124688ndash696

33 Behbehani R Adnan H Al-Hassan AA Al-Salahat A Alroughani R Predictors of retinalatrophy in multiple sclerosis a longitudinal study using spectral domain optical coherencetomography with segmentation analysis Mult Scler Relat Disord 20182156ndash62

34 Zhostkova MA Davydovskaya MV Boyko AN Akopyan VS The three-year follow-up study of retinal changes in patients with multiple sclerosis [in Russian] Zh NevrolPsikhiatr Im S S Korsakova 201611635ndash41

35 Garcia-Martin E Pueyo V Almarcegui C et al Risk factors for progressive axonaldegeneration of the retinal nerve fibre layer in multiple sclerosis patients Br J Oph-thalmol 2011951577ndash1582

36 Yang EB Hood DC Rodarte C Zhang X Odel JG Behrens MM Improvement inconduction velocity after optic neuritis measured with the multifocal VEP InvestOphthalmol Vis Sci 200748692ndash698

37 Klistorner A Chai Y Leocani L et al Assessment of opicinumab in acute optic neuritisusing multifocal visual evoked potential CNS Drugs 2018321159ndash1171

38 You Y Gupta V The extracellular matrix and remyelination strategies in multiplesclerosis eNeuro 20185e0435ndash0417

39 Plemel JR Liu WQ Yong VW Remyelination therapies a new direction and chal-lenge in multiple sclerosis Nat Rev Drug Discov 201716617ndash634

40 Klistorner AI Graham SL Grigg JR Billson FA Multifocal topographic visual evokedpotential improving objective detection of local visual field defects Invest Oph-thalmol Vis Sci 199839937ndash950

10 Neurology Neuroimmunology amp Neuroinflammation | Volume 7 Number 3 | May 2020 NeurologyorgNN

DOI 101212NXI000000000000070020207 Neurol Neuroimmunol Neuroinflamm

Yuyi You Michael H Barnett Con Yiannikas et al unilateral optic neuritis

Chronic demyelination exacerbates neuroaxonal loss in patients with MS with

This information is current as of March 13 2020

Academy of Neurology All rights reserved Online ISSN 2332-7812Copyright copy 2020 The Author(s) Published by Wolters Kluwer Health Inc on behalf of the AmericanPublished since April 2014 it is an open-access online-only continuous publication journal Copyright

is an official journal of the American Academy of NeurologyNeurol Neuroimmunol Neuroinflamm

ServicesUpdated Information amp

httpnnneurologyorgcontent73e700fullhtmlincluding high resolution figures can be found at

References httpnnneurologyorgcontent73e700fullhtmlref-list-1

This article cites 40 articles 10 of which you can access for free at

Subspecialty Collections

httpnnneurologyorgcgicollectionoptic_neuritisOptic neuritis see Neuro-ophthalmologyOptic Nerve

httpnnneurologyorgcgicollectionoptic_nerveOptic nerve

httpnnneurologyorgcgicollectionmultiple_sclerosisMultiple sclerosis

httpnnneurologyorgcgicollectionevoked_potentials-visualEvoked PotentialsVisual

httpnnneurologyorgcgicollectionall_imagingAll Imagingfollowing collection(s) This article along with others on similar topics appears in the

Permissions amp Licensing

httpnnneurologyorgmiscaboutxhtmlpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in

Reprints

httpnnneurologyorgmiscaddirxhtmlreprintsusInformation about ordering reprints can be found online

Academy of Neurology All rights reserved Online ISSN 2332-7812Copyright copy 2020 The Author(s) Published by Wolters Kluwer Health Inc on behalf of the AmericanPublished since April 2014 it is an open-access online-only continuous publication journal Copyright

is an official journal of the American Academy of NeurologyNeurol Neuroimmunol Neuroinflamm

ServicesUpdated Information amp

httpnnneurologyorgcontent73e700fullhtmlincluding high resolution figures can be found at

References httpnnneurologyorgcontent73e700fullhtmlref-list-1

This article cites 40 articles 10 of which you can access for free at

Subspecialty Collections

httpnnneurologyorgcgicollectionoptic_neuritisOptic neuritis see Neuro-ophthalmologyOptic Nerve

httpnnneurologyorgcgicollectionoptic_nerveOptic nerve

httpnnneurologyorgcgicollectionmultiple_sclerosisMultiple sclerosis

httpnnneurologyorgcgicollectionevoked_potentials-visualEvoked PotentialsVisual

httpnnneurologyorgcgicollectionall_imagingAll Imagingfollowing collection(s) This article along with others on similar topics appears in the

Permissions amp Licensing

httpnnneurologyorgmiscaboutxhtmlpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in

Reprints

httpnnneurologyorgmiscaddirxhtmlreprintsusInformation about ordering reprints can be found online

Academy of Neurology All rights reserved Online ISSN 2332-7812Copyright copy 2020 The Author(s) Published by Wolters Kluwer Health Inc on behalf of the AmericanPublished since April 2014 it is an open-access online-only continuous publication journal Copyright

is an official journal of the American Academy of NeurologyNeurol Neuroimmunol Neuroinflamm