Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Jan 1.
Published in final edited form as: Ann Neurol. 2022 Oct 22;93(1):76–87. doi: 10.1002/ana.26529

Trans-synaptic degeneration following acute optic neuritis in multiple sclerosis

Olwen C Murphy 1, Elias S Sotirchos 1, Grigorios Kalaitzidis 1, Elena Vasileiou 1, Henrik Ehrhardt 1, Jeffrey Lambe 1, Ohemaa Kwakyi 1, James Nguyen 1, Alexandra Zambriczki Lee 1, Julia Button 1, Blake E Dewey 2, Scott D Newsome 1, Ellen M Mowry 1, Kathryn C Fitzgerald 1, Jerry L Prince 2, Peter A Calabresi 1, Shiv Saidha 1
PMCID: PMC9933774  NIHMSID: NIHMS1868359  PMID: 36218157

Abstract

Objective:

Explore longitudinal changes in brain volumetric measures and retinal layer thicknesses following acute optic neuritis (AON) in people with multiple sclerosis (PwMS), to investigate the process of trans-synaptic degeneration and determine its clinical relevance.

Methods:

PwMS were recruited within 40 days of AON onset (n=49) and underwent baseline retinal optical coherence tomography (OCT) and brain MRI followed by longitudinal tracking for up to 5 years. A comparator cohort of PwMS without a recent episode of AON were similarly tracked (n=73). Mixed-effects linear regression models were used.

Results:

Accelerated atrophy of the occipital gray matter (GM), calcarine GM, and thalamus was seen in the AON cohort, as compared to the non-AON cohort (−0.76% versus −0.22% per year [p=0.01] for occipital GM, −1.83% versus −0.32% per year [p=0.008] for calcarine GM, −1.17% versus −0.67% per year [p=0.02] for thalamus), while rates of whole brain, cortical GM, non-occipital cortical GM atrophy and T2 lesion accumulation did not differ significantly between the cohorts. In the AON cohort, greater AON-induced reduction in ganglion cell+inner plexiform layer (GCIPL) thickness over the first year was associated with faster rates of whole brain (r=0.32, p=0.04), white matter (r=0.32, p=0.04) and thalamic (r=0.36, p=0.02) atrophy over the study period. Significant relationships were identified between faster atrophy of the subcortical GM and thalamus, with worse visual function outcomes after AON.

Interpretation:

These results provide in-vivo evidence for anterograde trans-synaptic degeneration following AON in PwMS, and suggest that trans-synaptic degeneration may be related to clinically-relevant visual outcomes.

Keywords: multiple sclerosis, optic neuritis, optical coherence tomography, neurodegeneration, trans-synaptic degeneration

Introduction

Neurodegeneration occurs throughout the course of multiple sclerosis (MS), and is accepted as the principal substrate of disability accumulation.1–3 One of the putative mechanisms of neurodegeneration in MS is trans-synaptic degeneration (trans-synaptic degeneration)—whereby injury to a neuron or axon leads to degeneration of synaptically-connected pathways of axons and neurons. This can occur in an anterograde (‘dying forward’) or retrograde (‘dying back’) direction, and accordingly can proceed to areas of the CNS distant to the initial injury. Acute optic neuritis (AON) offers an ideal opportunity to study trans-synaptic degeneration in MS, since the visual pathway exhibits functionally eloquent organization. Thus, the relevant synaptically-connected areas can be defined for evaluation with relative reliability.

Structural changes in the anterior visual pathway during and after AON have already been well-defined by studies employing retinal optical coherence tomography (OCT).4 Increased thickness of the unmyelinated peripapillary retinal nerve fiver layer (pRNFL), comprising axons of the retinal ganglion cells that pass through the lamina cribrosa to form the optic nerve, is detectable within the initial weeks after AON symptom onset, reflecting infiltration of inflammatory cells, interstitial edema, blood-retinal barrier disruption, and proliferation of glial cells.5–7 As the acute inflammatory process subsides, retrograde degeneration results in loss of axons in the pRNFL and loss of retinal ganglion cell bodies in the ganglion cell + inner plexiform layer (GCIPL), demonstrated as rapid pRNFL and GCIPL thinning on OCT.7–9 As the GCIPL thins, concomitant thickening of the inner nuclear layer (INL) and outer nuclear layer (ONL) occurs.7, 8 Dynamic retinal layer thickness changes after AON appear to be limited to the first 4–6 months, after which a new baseline is established and further longitudinal changes in retinal layer thicknesses between the ON eye and the fellow eye are similar, or in the case of GCIPL atrophy perhaps less-pronounced in the ON eye, since there is less remaining tissue to lose.8, 10 Animal studies have also provided important insights into the dynamics of retinal changes after ganglion cell or optic nerve injury, suggesting that one of the earliest changes to occur is shrinkage of the retinal ganglion cell dendrites (which branch within the inner plexiform layer),12, 13 and, moreover, supporting the synchronous occurrence of retrograde and anterograde axonal degeneration of retinal ganglion cells. 11

It can be hypothesized that neuroaxonal degeneration following AON could proceed through the optic tracts, and then trans-synaptically, culminating in loss of third order neurons projecting from the lateral geniculate nucleus of the thalamus to the primary visual cortex of the occipital lobe. Diffusion tensor parameters in the optic radiations have been shown to be altered in patients with a history of ON as compared to healthy controls or PwMS without a history of ON,14–17 and visual cortex volume may be lower in PwMS with a history of ON, as compared to PwMS without a history of ON.18 However, the extent to which anterograde trans-synaptic degeneration occurs and the potential clinical relevance of the phenomenon in MS are uncertain. In this prospective longitudinal study we set out to examine trans-synaptic degeneration by recruiting PwMS at the time of AON alongside a comparator cohort of PwMS without a recent episode of ON, and comparing retinal layer thickness changes (using OCT) and change in brain and relevant brain substructure volumes (using volumetric MRI techniques) between the two cohorts over time.

Materials and methods

Participants

People with relapsing-remitting MS or clinically isolated syndrome (CIS) with a high-risk for conversion to MS were recruited from the Johns Hopkins Multiple Sclerosis Center between 2010 and 2018. Diagnoses were made by the treating neurologist in accordance with the 2010 revised McDonald criteria,19 since study recruitment commenced before the criteria were updated in 2017. Patients with an episode of AON (occurring either as the first attack of MS or as a relapse in a patient with established MS) were recruited and completed baseline study protocols (clinical assessment, OCT and MRI imaging) within 40 days of AON symptom onset, and this group is hereafter referred to as the ‘AON cohort’. A comparator cohort of people with relapsing-remitting MS or high-risk CIS without an episode of AON in the last 3 years was also recruited from the Johns Hopkins Multiple Sclerosis Center, by convenience sampling, and is hereafter referred to as the ‘non-AON cohort’. Only participants with more than 1 year of MRI and OCT follow-up were included. Participants in the non-AON cohort were matched as closely as possible to participants in the AON cohort in terms of age, sex and self-reported race. Historical episodes of ON were recorded for both cohorts. Time since disease onset was defined from the time of onset of first neurological symptoms attributable to MS (in both MS and high-risk CIS). Exclusion criteria included people with moderate to poorly controlled diabetes mellitus or hypertension, glaucoma, refractive errors >6 or <−6 diopters, or other relevant ophthalmological or neurological disorders. Data collected in participants who developed a new episode of AON during the longitudinal follow-up period was censored from the date of onset of new AON symptoms.

Clinical assessments

Visual function was assessed monocularly and binocularly with retro-illuminated high-contrast Early Treatment of Diabetic Retinopathy charts at 4m, and low-contrast (2.5% and 1.25%) Sloan letter charts at 2m, using the patient’s usual contact lenses or spectacles. Scores were recorded as the number of letters identified accurately on each chart. Expanded Disability Status Scale (EDSS) scores were determined by certified raters. Multiple Sclerosis Functional Composite (MSFC) scores were derived from the nine-hole peg test, timed 25-foot walk, and paced auditory serial addition test, respectively, according to the MSFC taskforce database.20

Optical coherence tomography

OCT was completed with Cirrus HD-OCT (model 5000, software version 11.5; Carl Zeiss Meditec, Dublin, California) at baseline and then at least annually for up to 5 years in enrolled participants, except in the 12 months after AON onset, during which OCT was completed at baseline, 1 month, 3 months, 6 months, and 12 months (where feasible). Peri-papillary scans were obtained with the Optic Disc Cube 200×200 protocol, and macular scans were obtained with the Macular Cube 512×128 protocols by experienced technicians under low-lighting conditions, as described in detail elsewhere.21 Quality of acquired images was reviewed in accordance with the OSCAR-IB criteria.22 pRNFL thickness was quantified by the conventionally incorporated Cirrus HD-OCT algorithm. Macular retinal layer thicknesses (GCIPL, INL, ONL) were quantified using a validated segmentation algorithm developed at Johns Hopkins University, which has been demonstrated to reliably segment macular images in cross-sectional and longitudinal studies of both PwMS and healthy controls.23–25 Reporting of OCT methods and results is in accordance with the Advised Protocol for OCT Study Terminology and Elements (APOSTEL) criteria.26

MRI

MRI was completed with a 3T Philips Achieva Scanner (Philips Medical System, Best, Netherlands) at baseline and then annually for up to 5 years in enrolled participants. Three axial whole brain sequences without slice gaps were used: multi-slice T2-weighted fluid-attenuated inversion recovery (FLAIR; acquired resolution: 0.8 × 0.8 × 2.2mm; echo time: 68 ms; repetition time: 11 s; inversion time: 2.8 s; SENSE factor: 2; averages: 1); T2-weighted dual-echo turbo spin echo (DE-TSE; acquired resolution: 0.8 × 0.8 × 2.2mm; echo time: 80ms; repetition time: 4170 ms; SENSE factor: 2; averages: 1); and three-dimensional (3d) magnetization prepared rapid acquisition of gradient echoes (MPRAGE; acquired resolution: 0.8 × 0.8 × 1.2mm; echo time: 6ms; repetition time: 10 ms; inversion time: 835 ms; flip angle: 8 degrees; SENSE factor: 2; averages: 1).

A validated multi-step processing pipeline was employed to segment the brain into its component substructures while also outlining MS lesions. MRI images were corrected for intensity bias and all images were co-registered to the MNI-152 atlas using the ANTs software package.27–29 Images were harmonized between different acquisitions using DeepHarmony.30 We performed whole brain segmentation on harmonized images using the SLANT-CRUISE analysis pipeline. Briefly, the images were co-registered and segmented using the SLANT deep learning algorithm.31 Segmentations were then corrected to be consistent with cortical reconstruction using CRUISE, similar to work done in Multi-Atlas CRUISE (Ma-CRUISE).32 Before the SLANT step, automated lesion segmentation was applied using an ensemble-based lesion segmentation algorithm33 and the lesions were filled with values consistent with normal appearing white matter. The final segmentation reincorporates the segmented lesions into the corrected substructure segmentation of the whole brain generated by SLANT-CRUISE. Volumes were normalized by the subjects’ intracranial volume (ICV, to account for individual differences in head size), calculated separately from the whole brain segmentation.34 Thalamic segmentation was also completed using an alternative approach using a trained neural network based on the 3D U-net structure (which includes the lateral geniculate nucleus).35 This network uses T1-weighted structural imaging and diffusion tensor imaging (DTI), which were processed using the Tortoise software package. 36 A sensitivity analysis was completed comparing thalamic atrophy rates derived with the multi-modal approach to thalamic atrophy rates derived with the SLANT-CRUISE approach. After segmentation, ComBat was used to correct calculated volumes for additional scanner changes between scans.37 Due to the longitudinal nature of the data, a time-aware version of ComBat was used. 38 Additonally, ICV was corrected separately using the original ComBat algorithm37 before being used to normalize volumes. All segmentations were assessed qualitatively by experienced reviewers (BD, OCM, ESS) for accuracy, and scans with segmentation errors or failures were excluded.

Statistical analysis

Statistical analyses were completed using STATA version 16 (StataCorp, College Station, TX) and statistical significance was defined as p<0.05. Comparisons of demographic and clinical variables at baseline were completed using the chi-square test for categorical variables, the t-test for normally distributed, and the Wilcoxon rank-sum test for non-normally distributed continuous variables.

Retinal layer thickness reductions over the first year in the AON eyes and fellow eyes of the AON cohort were calculated by comparing OCT measurements acquired at the 12-month visit to the baseline visit, in an unadjusted manner to quantify actual tissue loss most accurately in individual participants over the first year. Mixed-effects linear regression models using time as a continuous covariate were used to derive longitudinal changes in brain volumetric measures in all participants, and longitudinal changes in retinal layer thicknesses in the non-AON cohort and in analyses of the whole study population. Participant-specific and eye-specific random intercepts and slopes were used in these models, which account for baseline brain substructure volume or retinal layer thickness, and intra-participant inter-eye correlations (for OCT measures). Rates of change generated from these models indicate the approximate percentage change per year in brain/brain substructure volume or retinal layer thickness, calculated using logarithmically transformed measures. In our primary analyses, we tested for differential rates of change in MRI measures between the AON and non-AON cohorts with interaction terms including time and cohort, in models adjusted for age, sex and time since disease onset. Cohort sizes were not sufficient and patterns of treatment too heterogeneous to allow adjustment of the statistical models for specific disease-modifying treatments.

For our secondary analyses, Pearson’s correlation analyses were used to assess relationships between participant-specific brain substructure atrophy rates and 1) eye-specific retinal layer thickness reductions in the AON eye over the first year; 2) eye-specific retinal layer thickness atrophy rates over the study period in the whole study population; and 3) visual outcomes in the AON eyes (letter acuity scores recorded in the AON eyes at the final study visit). AON eyes with a pre-existing episode of ON at baseline (n=5) were excluded from correlation analyses, since the previous ON episode would confound the baseline OCT measures and visual function in the affected eye.

Ethical approval

Johns Hopkins University institutional review board approval was obtained for all study protocols. Participants provided written informed consent. Data pertaining to this study will be made available at the request of qualified investigators, subject to institutional review board approval.

Results

Baseline demographic and clinical characteristics

Forty-nine patients with AON were recruited (the AON cohort) and 73 patients were recruited for the comparison cohort (the non-AON cohort, Table 1). In the AON cohort, median time from AON symptom onset to baseline evaluation with OCT and MRI was 14 days (range 3 to 40 days). As compared to people in the AON cohort, people in the non-AON cohort were on average older (p=0.001) and had a longer MS disease duration at baseline (p<0.001). The majority of the AON cohort (80%) were not on any disease-modifying treatment (DMT) at baseline (i.e., at AON onset), but at 1-year follow-up 86% of the AON cohort were on a DMT and there was no difference between the AON cohort and non-AON cohort across the categories of low/intermediate/high-potency treatments (p=0.32). In the AON cohort, testing to exclude aquaporin-4 IgG was completed in 39 of 49 participants (80%) and testing to exclude myelin oligodendrocyte glycoprotein-IgG (MOG-IgG) was completed in 9 of 49 participants (18%).

Table 1.

Demographic and clinical characteristics at baseline visit

AON cohort (n=49 participants) Non-AON cohort (n=73 participants) P-value
Age, mean years (SD) 32.5 (9.4) 38.6 (8.1) 0.001 a
Female, n 45 (82%) 43 (75%) 0.32b
Race
 White, n 40 (82%) 54 (74%) 0.51b
 Black, n 11 (21%) 10 (17%)
 Other, n 2 (4%) 1 (2%)
 Not reported, n 2 (4%) 8 (12%) -
Time since disease onset, mean years (SD) 3.2 (5.4) 8.0 (5.4) <0.001 a
Prior history of ON, n 10 (20%) 31 (42%) 0.01 b
Disease-modifying therapy
 None, n 39 (80%) 8 (11%) <0.001 b
 Low-potencyc, n 7 (14%) 43 (59%)
 Intermediate potencyd, n 3 (6%) 3 (4%)
 High potencye, n 0 (0%) 19 (26%)
EDSS score, mean (SD) 1.9 (1.3) 2.3 (1.5) 0.19a
MRI follow-up duration, mean years (SD) 2.5 (1.3) 3.2 (1.3) 0.008 a
OCT follow-up duration, mean years (SD) 3.1 (1.2) 3.6 (1.3) 0.008 a
Monocular visual function 1. AON eye 2. Fellow eye 3. Average of both eyes P-value (1 vs 2) P-value (2 vs 3)
 100% contrast letter acuity, median (IQR) 54 (29–59) 61 (55–65) 60 (56–65) <0.001 a 0.67a
 2.5% contrast letter acuity, median (IQR) 3 (0–14) 29 (20–35) 32 (24–38) <0.001 a 0.14a
 1.25% contrast letter acuity, median (IQR) 0 (0–9) 22 (9–29) 17 (4–24) <0.001 a 0.05a
Open in a new tab
a

Wilcoxon rank sum test

b

Chi-squared test

c

Low-potency: glatiramer acetate, interferons, teriflunamide

d

Intermediate potency: dimethyl fumarate, fingolimod

e

High potency: natalizumab, ocrelizumab, rituximab, alemtuzumab.

Low potency= EDSS=expanded disability status scale, SD=standard deviation, IQR=interquartile range. Where datapoints were not available for all patients, the number of patients with available data is indicated.

Baseline OCT and MRI measures

At baseline in the AON cohort, increased pRNFL thickness was seen in the AON eyes as compared to the fellow eyes (107.6 um [SD 28.0] versus 93.4 um [SD 113.2], p=0.03) while other retinal layer thicknesses (GCIPL, INL, ONL) were not significantly different in AON eyes versus their fellow eyes. Baseline retinal layer thicknesses and monocular visual function scores were also compared according to prior history of ON across the entire cohort, and consistent with previously published data showed reduced pRNFL and GCIPL thicknesses in eyes previously affected by remote ON (p<0.05 for all). Comparisons of baseline MRI measures in the AON versus the non-AON cohorts showed that the following brain and brain substructure volumes were significantly lower in the non-AON cohort at baseline: whole brain, cortical gray matter (GM), non-occipital cortical GM, occipital GM, subcortical GM, and thalamus, while T2 lesion volume was significantly higher in the non-AON cohort at baseline (data not shown).

Longitudinal change in retinal layer thicknesses

AON eyes of the AON cohort exhibited reductions in all retinal layer thicknesses over the first year post-AON (Table 2, Figure 1). Mean reductions were most marked in pRNFL thickness (−19.85% [SD 16.29]) followed by GCIPL thickness (−9.57% [SD 10.18]), INL thickness (3.73% [SD 7.93]), and ONL thickness (−1.65% [SD 2.83]). Reductions in pRNFL, GCIPL and INL thicknesses over the first year post-AON were significantly greater in AON eyes than in their fellow eyes (p<0.001 for pRNFL and GCIPL thicknesses, p=0.02 for INL thickness). Longitudinal changes in retinal layer thicknesses after the first year post-AON until the end of the study were similar between the AON and fellow eyes, with the exception of GCIPL atrophy, which was slower in AON eyes as compared to their fellow eyes (−0.15% per year [95% CI −0.59 to 0.29] versus −0.65% per year [95% CI −1.04 to −0.25], p=0.05). There was a similar trend towards slower pRNFL atrophy in AON eyes as compared to their fellow eyes after the first year post-AON (−0.58% per year [95% CI −1.21 to 0.05] versus −1.10% per year [95% CI −1.55 to −0.65], p=0.15). Longitudinal changes in retinal layer thicknesses in eyes from the non-AON cohort were similar to the fellow eyes of the AON cohort, with the exception of pRNFL thinning which was faster in fellow eyes of the AON cohort (−1.04% per year [95% CI −1.56 to −0.52] versus −0.35% per year [95% CI −0.54 to −0.17, p=0.002).

Table 2.

Change in retinal layer thicknesses over the first year in the AON cohort

AON eyes (n=49) Fellow eyes (n=49) P-valuea (AON eyes vs fellow eyes)
pRNFL thickness, mean % change (SD) −19.85 (16.29) −1.27 (5.89) <0.001
GCIPL thickness, mean % change (SD) −9.57 (10.18) −0.61 (3.05) <0.001
INL thickness, mean % change (SD) −3.73 (7.93) −2.34 (7.87) 0.02
ONL thickness, mean % change (SD) −1.65 (2.83) −1.35 (2.65) 0.44
Open in a new tab
a

Wilcoxon rank-sum test.

pRNFL=peripapillary retinal nerve fiber layer, GCIPL=ganglion cell + inner plexiform layer, INL=inner nuclear layer, ONL=outer nuclear layer.

Figure 1. Longitudinal change in retinal layer thicknesses in AON eyes of the AON cohort over the first year.

Figure 1

Open in a new tab

In the AON eyes of the AON cohort (n=49), changes in retinal layer thicknesses over the course of the first year post-AON are shown here in panels A-D, with each line representing an individual patient. Since pRNFL tends to be swollen at the time of acute optic neuritis (therefore masking a ‘true’ baseline value), the differences in pRNFL between the AON eye and fellow eye at baseline and 1 year are illustrated in panels E and F. Figure was generated using STATA version 16 (StataCorp, College Station, TX). pRNFL=peripapillary retinal nerve fiber layer, GCIPL=ganglion cell + inner plexiform layer, INL=inner nuclear layer, ONL=outer nuclear layer.

Longitudinal change in brain and brain substructure volumes

Within-cohort changes

Over the study period, the AON cohort had significant atrophy of cortical GM (p<0.001), occipital GM (p<0.001), calcarine GM (p=0.001), subcortical GM (p<0.001), and thalamic (p<0.001) volumes, while atrophy was not significant for whole brain (p=0.15), white matter (p=0.53), or non-occipital GM (p=0.10, Table 3) volumes. The non-AON cohort had significant atrophy of whole brain (p<0.001), white matter (p<0.001), cortical GM (p<0.00), non-occipital cortical GM (p=0.001), occipital GM (p=0.02), subcortical GM (p<0.001), and thalamic (p<0.001) volumes. Both the AON and non-AON cohort experienced significant increase in T2 lesion volume over time (p=0.006 for AON cohort and p<0.001 for non-AON cohort).

Table 3.

Longitudinal change in brain and brain substructure volumes

AON cohort (n=49 participants) P-valuea Non-AON cohort (n=73 participants) P-valuea P-value (AON vs non-AON)b
Whole brain, % per year (95% CI) −0.14% (−0.34 to 0.05) 0.15 −0.33% (−0.47 to −0.20) <0.001 0.10
White matter, % per year (95% CI) −0.09% (−0.38 to 0.20) 0.53 −0.35% (−0.54 to −0.16) <0.001 0.13
Cortical GM, % per year (95% CI) −0.30% (−0.45 to −0.14) <0.001 −0.26% (−0.39 to −0.13) <0.001 0.83
Non-occipital cortical GM, % per year (95% CI) −0.14% (−0.30 to 0.03) 0.10 −0.28% (−0.43 to −0.14) 0.001 0.17
Occipital GM, % per year (95% CI) −0.76% (−1.16 to −0.35) <0.001 −0.22% (−0.41 to −0.03) 0.02 0.01
Calcarine GM, % per year (95% CI) −1.83% (−2.88 to −0.77) 0.001 −0.32% (−0.77 to 0.14) 0.18 0.008
Subcortical GM, % per year (95% CI) −0.83% (−1.08 to −0.57) <0.001 −0.54% (−0.75 to −0.34) <0.001 0.10
Thalamus, % per year (95% CI) −1.17% (−1.51 to −0.82) <0.001 −0.67% (−0.90 to −0.43) <0.001 0.02
T2 lesion, % per year (95% CI) +5.6% (1.6 to 9.7) 0.006 +5.8% (3.6 to 7.9) <0.001 0.78
Open in a new tab
a

Mixed-effects linear regression including time

b

Mixed-effects linear regression including time, cohort and interaction between time and cohort, age, sex and time since disease onset.

GM=gray matter, CI=confidence interval.Brain and brain substructure volumes were expressed as a fraction of intracranial volume and logarithmically transformed for longitudinal analyses.

Between-cohort changes

In directly comparing brain volumetric changes between the AON and non-AON cohorts (Table 3, Figure 2), the AON cohort experienced faster atrophy of occipital GM (−0.76% per year [95% CI −1.16 to −0.35] versus −0.22% per year [95% CI −0.41 to −0.03], p=0.01), calcarine GM (−1.83% per year [95% CI −2.88 to −0.77] versus −0.32% per year [95% CI −0.77 to 0.14], p=0.008) and thalamic (−1.17% per year [95% CI −1.51 to −0.82] versus −0.67% per year [95% CI −0.90 to −0.43] p=0.02) volumes. A sensitivity analysis employing a DTI-based approach to thalamic segmentation (which includes the lateral geniculate nucleus) confirmed these results, with accelerated thalamic atrophy seen in the AON cohort as compared to the non-AON cohort (−1.59% per year [CI −2.14 to −1.03) versus −0.30% per year [95% CI −0.96 to 0.37], p=0.01). There were no significant differences in whole brain, white matter, cortical GM, non-occipital cortical GM, and subcortical GM atrophy or T2 lesion accumulation when comparing the AON and non-AON cohorts.

Figure 2. Rates of atrophy of brain and brain substructure volumes in the AON cohort versus the non-AON cohort.

Figure 2

Open in a new tab

Longitudinal changes in brain and brain substructure volumes are demonstrated here in the AON cohort versus the non-AON cohort. Brain and brain substructure volumes were expressed as a fraction of intracranial volume and logarithmically transformed for longitudinal analyses, and direct comparisons in the atrophy rates between the cohorts were made using mixed-effects linear regression including time, cohort and interaction between time and cohort, age, sex and time since disease onset. The AON cohort exhibited significantly faster atrophy of the occipital GM (p=0.01), calcarine GM (p=0.008), and thalamus (p=0.02). There were no significant differences in the rate of whole brain, white matter, cortical GM, non-occipital cortical GM, or subcortical GM between the cohorts. Additionally, the rate of T2 lesion volume accumulation did not differ between the cohorts (+5.6% per year [95% CI 1.6 to 9.7] in the AON cohort versus +5.8% per year [95% CI 3.6 to 7.9] in the non-AON cohort, p=0.78, datapoints not demonstrated on this figure). Figure was generated using GraphPad Prism version 9 (GraphPad Software, La Jolla, CA). GM=gray matter.

Relationships between longitudinal changes in OCT and MRI measures

Relationships were identified between greater reductions in GCIPL thicknesses over the first year post-AON and faster whole brain (r=0.30, p=0.04), white matter (r=0.32, p=0.04), and thalamic atrophy (r=0.28, p=0.05) over the study period (Figure 3). No other significant relationships were identified between longitudinal changes in retinal layer thicknesses and brain volumetric measures in the AON cohort (data not shown). Relationships between longitudinal changes in OCT and MRI measures in the whole study population (availing of the increased statistical power of the larger group for informative purposes and excluding data acquired over the first year post-AON in the AON cohort) were also explored and are outlined in Table 4. Notably, in the analyses of the whole cohort, faster ONL thinning was significantly associated with faster whole brain (r=0.24, p=0.02), cortical GM (r=0.26, p=0.01) and non-occipital cortical GM atrophy (r=0.31, p=0.002), while faster GCIPL thinning was associated with faster whole brain (r=0.29, p=0.005), white matter (r=0.21, p=0.04), cortical GM (p=0.30, p=0.003), non-occipital cortical GM (r=0.23, p=0.02), subcortical GM (r=0.28, p=0.005) and thalamic atrophy (r=0.32, p=0.002).

Figure 3. Relationships between reduction in GCIPL thickness in the AON eye over the first year, and rates of brain and brain substructure atrophy over the study period, in the AON cohort.

Figure 3

Open in a new tab

Scatterplots demonstrate relationships between change in GCIPL thickness in the AON eyes over the first year, and rates of brain and brain substructure atrophy, in the AON cohort. R-values and P-values were calculated with Pearson’s correlation, using percentage change in GCIPL thickness in the AON eye over the first year, and participant-specific brain substructure atrophy rates generated from mixed-effects linear regression models. In AON eyes, significant relationships were seen between greater reductions in GCIPL thickness over the first year, and faster whole brain (A), white matter (B), and thalamic (C) atrophy over the study period. Figure was generated using STATA version 16 (StataCorp, College Station, TX).

Table 4.

Relationships between rates of change in brain substructure volumes and retinal layer thicknesses in the whole study population (n=122 participants, excluding data acquired over the first year post-AON in AON cohort)

RNFL thickness (% per year) GCIPL thickness (% per year) INL thickness (% per year) ONL thickness (% per year)
R-value P-value R-value P-value R-value P-value R-value P-value
Whole brain (% per year) 0.20 0.05 0.29 0.005 0.09 0.41 0.24 0.02
White matter (% per year) 0.21 0.04 0.21 0.04 0.07 0.47 0.16 0.12
Cortical GM (% per year) 0.12 0.25 0.30 0.003 0.07 0.51 0.26 0.01
Non-occipital cortical GM (% per year) 0.12 0.24 0.23 0.02 0.07 0.48 0.31 0.002
Occipital GM (% per year) 0.01 0.92 0.20 0.05 0.01 0.94 -0.04 0.70
Calcarine GM (% per year) −0.15 0.17 0.04 0.73 −0.07 0.50 −0.25 0.02
Subcortical GM (% per year) 0.20 0.05 0.28 0.005 0.03 0.76 0.20 0.06
Thalamus (% per year) 0.23 0.03 0.32 0.002 0.03 0.74 0.16 0.11
T2 lesion (% per year) 0.02 0.87 −0.02 0.82 0.02 0.83 −0.01 0.92
Open in a new tab

R-values and P-values were calculated with Pearson’s correlation, using participant-specific brain substructure atrophy rates and participant-specific retinal layer thickness atrophy rates generated from mixed-effects linear regression models adjusted for intra-subject inter-eye correlations.pRNFL=peripapillary retinal nerve fiber layer, GCIPL=ganglion cell + inner plexiform layer, INL=inner nuclear layer, ONL=outer nuclear layer, GM= gray matter.

Relationships between longitudinal change in brain and brain substructure volumes and visual outcomes

Relationships were identified between worse 100% contrast letter acuity outcomes in AON eyes and faster atrophy of the subcortical GM (r=0.31, p=0.04), and between worse 1.25% contrast letter acuity outcomes and faster atrophy of the thalamus (r=0.31, p=0.04, Figure 4). These same relationships were not seen between rates of brain substructure atrophy and letter acuity scores of the fellow eyes.

Figure 4. Relationships between visual outcomes and rates of brain and brain substructure atrophy over the study period, in AON eyes of the AON cohort.

Figure 4

Open in a new tab

Scatterplots demonstrate relationships between monocular letter acuity at the final study visit, and rates of brain and brain substructure atrophy, in AON eyes of the AON cohort. R-values and P-values were calculated with Pearson’s correlation, using participant-specific brain and brain substructure atrophy rates generated from mixed-effects linear regression models, and letter acuity scores recorded at the final study visit. In AON eyes, significant relationships were seen between worse 100% contrast letter acuity outcomes and faster atrophy of subcortical GM (A), and between worse 1.25% contrast letter acuity and faster thalamic atrophy (B). By comparison, in the fellow eye, there were no significant relationships between letter acuity scores and atrophy of brain or brain substructures. Figure was generated using STATA version 16 (StataCorp, College Station, TX). GM=gray matter.

Discussion

Results of this study support that anterograde trans-synaptic neuroaxonal degeneration occurs in the posterior visual pathway after AON in PwMS, as evidenced by faster rates of atrophy of posterior visual pathway structures in the AON cohort, as compared to the non-AON cohort. The disproportionate acceleration in atrophy of the thalamus and visual cortex seen in the AON cohort appears to support the hypothesis that neuro-axonal degeneration after AON proceeds trans-synaptically to the posterior visual pathway (and is at least detectable at a cohort level). Furthermore, in the AON cohort greater reductions in GCIPL thicknesses over the first year post-AON were associated with faster rates of thalamic atrophy over the study period. These findings suggest that a larger magnitude of retinal tissue loss after AON appears to be linked to a greater extent of subsequent anterograde trans-synaptic neuroaxonal degeneration. Importantly, results of this study also support the clinical relevance of trans-synaptic degeneration in PwMS, since worse visual outcomes in the AON eye were also related to faster rates of thalamic atrophy. This finding is particularly notable since imaging correlates of visual outcomes have been relatively elusive in prior studies of AON in MS.

Anterograde trans-synaptic degeneration has long been hypothesized in MS, but direct evidence in support of the phenomenon has been lacking. Our findings provide compelling evidence for accelerated atrophy of the posterior visual pathway structures in PwMS after AON. More specifically, thalamic atrophy was almost twice as fast, occipital GM atrophy was three times as fast, and calcarine GM atrophy was five times as fast in the AON cohort, as compared to the non-AON cohort, while atrophy rates of the whole brain and other brain substructures did not differ between the cohorts. Previous work has suggested that volume of the visual cortex is lower in PwMS with a history of severe ON as compared to those without a history of ON, and that visual cortex volume is associated with RNFL thinning (independently of ON history).18 Another study reported DTI changes in the posterior visual pathway after ON, as compared to healthy controls.17 Our prospective findings build on this prior work by demonstrating atrophy of both the thalamus and visual cortex occurring in ‘real time’ after AON in PwMS, as compared to patients with relapsing MS without a recent episode of AON. While our findings relating to the visual cortex are important in confirming hypothesized mechanisms of neuroaxonal degeneration in MS, the identification of accelerated thalamic atrophy after AON is particularly notable. In recent years, thalamic atrophy has been consistently identified as one of the most robust MRI metrics of neurodegeneration throughout the MS disease course, with strong clinical correlates.1, 39–41 The thalamus has a wealth of synaptic connections to other areas of the CNS, acting as a relay center for sensory, motor and cognitive functions, which may all be affected in MS. However, our findings suggest that trans-synaptic degeneration following AON is a driver of greater thalamic atrophy than may be related to relay effects from extra-visual pathways, since the non-AON cohort in this study in general demonstrated similar rates of whole brain, white matter and non-occipital cortical GM atrophy, and similar T2 lesion volume accumulation to the AON cohort. This suggests that the accelerated thalamic atrophy seen in the AON cohort is more likely attributable to AON itself, rather than simply an active inflammatory disease phenotype with incidental multi-focal inflammation. Interestingly, since the lateral geniculate nucleus is one of the smaller thalamic nuclei,42 one question that does arise is which other thalamic nuclei contributed to the accelerated thalamic atrophy seen in the AON cohort. Given the markedly accelerated occipital and calcarine GM atrophy observed in the AON cohort, a likely plausible candidate for the observed thalamic atrophy in this cohort may be the pulvinar nucleus (the largest thalamic nucleus), which plays an important role in higher visual processing—demonstrating high levels of bidirectional connectivity with the visual cortex in humans,43 and has also been shown in primate studies to receive direct projections from retinal ganglion cells and the superior colliculus.44–46 Pulvinar atrophy might occur after AON as a downstream effect of trans-synaptic degeneration in the visual pathway, and perhaps explains a proportion of the thalamic atrophy seen in the AON cohort.

A salient finding of the current study includes the relationships identified between worse visual outcomes at high- and low-contrast after AON and accelerated atrophy of the thalamus and subcortical GM. Low-contrast visual acuity captures visual dysfunction particularly well in PwMS (since high-contrast letter acuity is similar in PwMS and controls) and correlates with structural measures (OCT) and functional measures (visual evoked potentials) of visual pathway injury as well as global disability measures.47 These findings are clinically important and support the hypothesis that greater trans-synaptic degeneration occurs in patients with poorer clinical outcomes after an inflammatory relapse.

In recent years, there has been increased interest in how changes in the deeper layers of the retina may reflect MS disease activity. Cross-sectionally, higher INL thickness is associated with subsequent clinico-radiologic inflammatory activity in PwMS,48 and higher INL thickness in MS eyes without an ON history is associated with higher white matter lesion volume and lower normal-appearing white matter volume.49 Additionally, the INL appears to transiently thicken in both eyes after AON,8 and even after MS relapses outside the visual pathway,50, 51 which has been hypothesized to reflect increased vascular permeability, or a reactive phenomenon with Müller cells proliferating following loss of retinal ganglion cells due to subclinical optic neuropathy. In the current study, we found greater reductions in INL thickness in the AON eye as compared to the fellow eye over the first year. This finding is novel and was not identified in our prior study of post-AON retinal changes in a smaller cohort.8 Intriguingly, the differential INL reductions between the AON eye and fellow eye suggests that these changes cannot be solely attributed to transient global inflammatory responses, and may actually support the occurrence of retrograde trans-synaptic degeneration in the bipolar cells of the inner nuclear layer. Regarding the ONL (containing the photoreceptor cell bodies), recent work suggests that thinning of both the INL and ONL is faster in progressive MS as compared to relapsing-remitting MS and unlike pRNFL and GCIPL thinning does not seem to be modulated by disease modifying treatments.52 In this context, a notable finding from our study was that faster ONL thinning in the whole cohort was associated with faster whole brain, cortical GM, and non-occipital GM atrophy, and this was distinct from faster GCIPL thinning, which was also associated with faster thalamic and subcortical GM atrophy. ONL thinning may be a putative marker of global neurodegeneration occurring in the MS brain that is not necessarily driven by the same pathobiological processes occurring at the same point in the disease process as GCIPL thinning.

Our study represents an important advance in the exploration of trans-synaptic degeneration in MS through detailed concurrent longitudinal monitoring of retinal layer thicknesses and brain volumetric measures following AON. However, our findings should be considered in the context of several limitations. Recruitment of patients following AON eye should ideally be as soon as possible following onset of symptoms, and for pragmatic purposes we allowed up to 40 days after symptom onset for completion of baseline study protocols. Importantly, the lack of difference in GCIPL thickness between the AON eyes and the fellow eyes at initial assessment suggests that we have captured patients before neuroaxonal loss was clearly established in the anterior visual pathway. While the AON and non-AON cohorts were well-matched in terms of disability status (EDSS and visual function scores), the non-AON cohort were on average older and had a longer time since disease onset than the AON cohort (since AON was the first clinical presentation of MS in many patients in the AON cohort). To minimize the impact of the inter-cohort differences, we did utilize mixed-effects linear regression models with random intercepts and slopes to account for inherent differences in baseline measures. The non-AON cohort also exhibited similar levels of T2 lesion accumulation over the study period, suggesting similar levels of inflammatory disease activity between the two cohorts. Additionally, rates of thalamic atrophy in PwMS have been shown to be relatively constant across different disease durations in MS,1 suggesting that the accelerated thalamic atrophy seen in the AON cohort cannot be attributed to shorter disease duration. Additionally, we did not segment individual thalamic nuclei, which is certainly an area of interest for future studies. The size of each cohort was modest, and we may have been under-powered to detect certain longitudinal relationships. Race may have an impact on MS disease severity and propensity to neurodegeneration,53 and DMT choice may have an impact on rates of brain and brain substructure atrophy,54 but our study was not powered to examine the impact of either race or DMT on trans-synaptic degeneration. However, while many of the non-AON cohort were not on a DMT at baseline evaluation, most (60%) of the non-AON cohort were taking a low-potency DMT at baseline, and differences in treatment status between the cohorts were not present by 1-year follow-up, suggesting that any effect of DMT over the first year is unlikely to be the sole driver of the differences in brain/brain substructure atrophy rates seen between the cohorts. Additionally, our analyses were adjusted for age, sex, and time since disease onset but not for non-AON MS relapses, which could also impact both regional and global MRI metrics. A final notable limitation is that testing rates for MOG-IgG were low in the AON cohort, since most study recruitment took place prior to the availability of commercial testing for MOG-IgG at our institution (2017). It is possible that some patients with MOG-IgG associated ON were erroneously diagnosed with MS at the outset, although no participants were noted to have a disease course atypical for MS over the longitudinal follow-up period.

Conclusions

Our study findings provide compelling in-vivo evidence for the occurrence of trans-synaptic degeneration following AON in PwMS, with atrophy of the occipital GM, calcarine GM, and thalamus being significantly accelerated in the AON cohort. Moreover, greater retinal tissue injury after AON (evidenced by greater GCIPL reductions within the first year post-AON) was associated with faster subsequent thalamic atrophy, suggesting that trans-synaptic degeneration is indeed a driver of thalamic atrophy in relapsing MS. The occurrence of trans-synaptic degeneration also appears to be clinically relevant, since poorer visual outcomes in the AON eyes were associated with faster atrophy of the thalamic and subcortical GM. Trans-synaptic degeneration appears to be an important pathobiological mechanism of neurodegeneration in MS, and the extent of trans-synaptic degeneration after a relapse may be a marker of more aggressive disease with greater tissue destruction, poorer tissue repair, a propensity to trans-synaptic pathology, or a combination of these mechanisms.

Acknowledgments

This study was funded by the National MS Society (RG-1606-08768 & RG-1907-34405 to S.S), Race to Erase MS (to S.S.), and NIH/NINDS (R01NS082347 to P.A.C.).

Footnotes

Potential Conflicts of Interest

The authors have nothing to report.

REFERENCES

  • 1.Azevedo CJ, Cen SY, Khadka S, et al. Thalamic atrophy in multiple sclerosis: A magnetic resonance imaging marker of neurodegeneration throughout disease. Ann Neurol 2018;83:223–234. doi: 10.1002/ana.25150. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Azevedo CJ, Overton E, Khadka S, et al. Early CNS neurodegeneration in radiologically isolated syndrome. Neurol Neuroimmunol Neuroinflamm 2015;2:e102. doi: 10.1212/NXI.0000000000000102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Dutta R, Trapp BD. Mechanisms of neuronal dysfunction and degeneration in multiple sclerosis. Prog Neurobiol 2011;93:1–12. doi: 10.1016/j.pneurobio.2010.09.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Lambe J, Saidha S, Bermel RA. Optical coherence tomography and multiple sclerosis: Update on clinical application and role in clinical trials. Mult Scler 2020;26:624–639. doi: 10.1177/1352458519872751. [DOI] [PubMed] [Google Scholar]
  • 5.Kallenbach K, Simonsen H, Sander B, et al. Retinal nerve fiber layer thickness is associated with lesion length in acute optic neuritis. Neurology 2010;74:252–258. doi: 10.1212/WNL.0b013e3181ca0135. [DOI] [PubMed] [Google Scholar]
  • 6.Manogaran P, Samardzija M, Schad AN, et al. Retinal pathology in experimental optic neuritis is characterized by retrograde degeneration and gliosis. Acta Neuropathol Commun 2019;7:116. doi: 10.1186/s40478-019-0768-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Gabilondo I, Martínez-Lapiscina EH, Fraga-Pumar E, et al. Dynamics of retinal injury after acute optic neuritis. Ann Neurol 2015;77:517–528. doi: 10.1002/ana.24351. [DOI] [PubMed] [Google Scholar]
  • 8.Al-Louzi OA, Bhargava P, Newsome SD, et al. Outer retinal changes following acute optic neuritis. Mult Scler 2016;22:362–372. doi: 10.1177/1352458515590646. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Syc SB, Saidha S, Newsome SD, et al. Optical coherence tomography segmentation reveals ganglion cell layer pathology after optic neuritis. Brain 2012;135:521–533. doi: 10.1093/brain/awr264. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Saidha S, Al-Louzi O, Ratchford JN, et al. Optical coherence tomography reflects brain atrophy in multiple sclerosis: A four-year study. Ann Neurol 2015;78:801–813. doi: 10.1002/ana.24487. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Leung CK, Weinreb RN, Li ZW, et al. Long-term in vivo imaging and measurement of dendritic shrinkage of retinal ganglion cells. Investigative ophthalmology & visual science 2011;52:1539–1547. doi: 10.1167/iovs.10-6012. [DOI] [PubMed] [Google Scholar]
  • 12.Miller NR, Subramanian PS, Patel VR. Walsh and Hoyt’s Clinical Neuro-Ophthalmology: The Essentials, 3rd Edition ed. Philadelphia, USA: Lippincott Williams & Wilkins; 2016. [Google Scholar]
  • 13.Kanamori A, Catrinescu M, Belisle JM, et al. Retrograde and wallerian axonal degeneration occur synchronously after retinal ganglion cell axotomy. Am J Pathol 2012;181:62–73. doi: 10.1016/j.ajpath.2012.03.030. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Rocca MA, Mesaros S, Preziosa P, et al. Wallerian and trans-synaptic degeneration contribute to optic radiation damage in multiple sclerosis: A diffusion tensor MRI study. Mult Scler 2013;19:1610–1617. doi: 10.1177/1352458513485146. [DOI] [PubMed] [Google Scholar]
  • 15.Kolbe S, Bajraszewski C, Chapman C, et al. Diffusion tensor imaging of the optic radiations after optic neuritis. Hum Brain Mapp 2012;33:2047–2061. doi: 10.1002/hbm.21343. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.You Y, Joseph C, Wang C, et al. Demyelination precedes axonal loss in the transneuronal spread of human neurodegenerative disease. Brain 2019;142:426–442. doi: 10.1093/brain/awy338. [DOI] [PubMed] [Google Scholar]
  • 17.Tur C, Goodkin O, Altmann DR, et al. Longitudinal evidence for anterograde trans-synaptic degeneration after optic neuritis. Brain 2016;139:816–828. doi: 10.1093/brain/awv396. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Gabilondo I, Martínez-Lapiscina EH, Martínez-Heras E, et al. Trans-synaptic axonal degeneration in the visual pathway in multiple sclerosis. Ann Neurol 2014;75:98–107. doi: 10.1002/ana.24030. [DOI] [PubMed] [Google Scholar]
  • 19.Polman CH, Reingold SC, Banwell B, et al. Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria. Ann Neurol 2011;69:292–302. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Fischer JS, Rudick RA, Cutter GR, Reingold SC. The multiple sclerosis functional composite measure (MSFC): An integrated approach to MS clinical outcome assessment. national MS society clinical outcomes assessment task force. Mult Scler 1999;5:244–250. [DOI] [PubMed] [Google Scholar]
  • 21.Syc SB, Warner CV, Hiremath GS, et al. Reproducibility of high-resolution optical coherence tomography in multiple sclerosis. Mult Scler 2010;16:829–839. doi: 10.1177/1352458510371640. [DOI] [PubMed] [Google Scholar]
  • 22.Tewarie P, Balk L, Costello F, et al. The OSCAR-IB consensus criteria for retinal OCT quality assessment. PLoS ONE 2012;7:e34823. doi: 10.1371/journal.pone.0034823. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Bhargava P, Lang A, Al-Louzi O, et al. Applying an open-source segmentation algorithm to different OCT devices in multiple sclerosis patients and healthy controls: Implications for clinical trials. Mult Scler Int 2015;2015 doi: 10.1155/2015/136295. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Lang A, Carass A, Al-Louzi O, et al. Combined registration and motion correction of longitudinal retinal OCT data. Proc SPIE Int Soc Opt Eng 2016;9784 doi: 10.1117/12.2217157. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Lang A, Carass A, Hauser M, et al. Retinal layer segmentation of macular OCT images using boundary classification. Biomed Opt Express 2013;4:1133–1152. doi: 10.1364/BOE.4.001133. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Cruz-Herranz A, Balk L, Oberwahrenbrock T, et al. The APOSTEL recommendations for reporting quantitative optical coherence tomography studies. Neurology 2016;86:2303–2309. doi: 10.1212/WNL.0000000000002774. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Tustison NJ, Avants BB, Cook PA, et al. N4ITK: Improved N3 bias correction. IEEE Trans Med Imaging 2010;29:1310–1320. doi: 10.1109/TMI.2010.2046908. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Avants BB, Tustison NJ, Stauffer M, et al. The insight ToolKit image registration framework. Front Neuroinform 2014;8:44. doi: 10.3389/fninf.2014.00044. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Fonov V, Evans A, McKinstry R, et al. Unbiased nonlinear average age-appropriate brain templates from birth to adulthood. NeuroImage 2009;47:S102. [Google Scholar]
  • 30.Dewey BE, Zhao C, Reinhold JC, et al. DeepHarmony: A deep learning approach to contrast harmonization across scanner changes. Magn Reson Imaging 2019;64:160–170. doi: 10.1016/j.mri.2019.05.041. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Huo Y, Xu Z, Xiong Y, et al. 3D whole brain segmentation using spatially localized atlas network tiles. NeuroImage 2019;194:105–119. doi: 10.1016/j.neuroimage.2019.03.041. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Huo Y, Plassard AJ, Carass A, et al. Consistent cortical reconstruction and multi-atlas brain segmentation. Neuroimage 2016;138:197–210. doi: 10.1016/j.neuroimage.2016.05.030. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Tohidi P, Remedios SW, Greenman DL, et al. Multiple sclerosis brain lesion segmentation with different architecture ensembles. Presented at the Medical Imaging 2022: Biomedical Applications in Molecular, Structural, and Functional Imaging; 2022, 2022;. [Google Scholar]
  • 34.Singh M, Pahl E, Wang S, Carass A, Lee J, Prince JL. Accurate estimation of total intracranial volume in MRI using a multi-tasked image-to-image translation network. Presented at the Medical Imaging 2021: Image Processing; 2021, 2021;. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Shao M, Zuo L, Carass A, Zhuo J, Gullapalli RP, Prince JL. Evaluating the impact of MR image harmonization on thalamus deep network segmentation. Presented at the Medical Imaging 2022: Image Processing; 2022, Apr 4, 2022;. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Pierpaoli C, Walker L, Irfanoglu MO, et al. TORTOISE: An integrated software package for processing of diffusion MRI data. Presented at the ISMRM 18th Annual Meeting; 2010, 2010;. [Google Scholar]
  • 37.Fortin J, Cullen N, Sheline YI, et al. Harmonization of cortical thickness measurements across scanners and sites. NeuroImage 2018;167:104–120. doi: 10.1016/j.neuroimage.2017.11.024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Beer JC, Tustison NJ, Cook PA, et al. Longitudinal ComBat: A method for harmonizing longitudinal multi-scanner imaging data. NeuroImage 2020;220:117129. doi: 10.1016/j.neuroimage.2020.117129. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Henry RG, Shieh M, Okuda DT, et al. Regional grey matter atrophy in clinically isolated syndromes at presentation. J Neurol Neurosurg Psychiatry 2008;79:1236–1244. doi: 10.1136/jnnp.2007.134825. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Eshaghi A, Marinescu RV, Young AL, et al. Progression of regional grey matter atrophy in multiple sclerosis. Brain 2018;141:1665–1677. doi: 10.1093/brain/awy088. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Eshaghi A, Prados F, Brownlee WJ, et al. Deep gray matter volume loss drives disability worsening in multiple sclerosis. Ann Neurol 2018;83:210–222. doi: 10.1002/ana.25145. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Li M, He HG, Shi W, et al. Quantification of the human lateral geniculate nucleus in vivo using MR imaging based on morphometry: Volume loss with age. American journal of neuroradiology : AJNR 2012;33:915–921. doi: 10.3174/ajnr.A2884. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Lambert C, Simon H, Colman J, Barrick TR. Defining thalamic nuclei and topographic connectivity gradients in vivo. Neuroimage 2017;158:466–479. doi: 10.1016/j.neuroimage.2016.08.028. [DOI] [PubMed] [Google Scholar]
  • 44.Grünert U, Lee SCS, Kwan WC, et al. Retinal ganglion cells projecting to superior colliculus and pulvinar in marmoset. Brain Struct Funct 2021. doi: 10.1007/s00429-021-02295-8. [DOI] [PubMed] [Google Scholar]
  • 45.Kaas JH, Lyon DC. Pulvinar contributions to the dorsal and ventral streams of visual processing in primates. Brain Res Rev 2007;55:285–296. doi: 10.1016/j.brainresrev.2007.02.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Lin C-, Kaas JH. The inferior pulvinar complex in owl monkeys: Architectonic subdivisions and patterns of input from the superior colliculus and subdivisions of visual cortex. J Comp Neurol 1979;187:655–678. doi: 10.1002/cne.901870403. [DOI] [PubMed] [Google Scholar]
  • 47.Balcer LJ, Raynowska J, Nolan R, et al. Validity of low-contrast letter acuity as a visual performance outcome measure for multiple sclerosis. Mult Scler 2017;23:734–747. doi: 10.1177/1352458517690822. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Saidha S, Sotirchos ES, Ibrahim MA, et al. Microcystic macular oedema, thickness of the inner nuclear layer of the retina, and disease characteristics in multiple sclerosis: A retrospective study. Lancet Neurol 2012;11:963–972. doi: 10.1016/S1474-4422(12)70213-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Saidha S, Sotirchos ES, Oh J, et al. Relationships between retinal axonal and neuronal measures and global central nervous system pathology in multiple sclerosis. JAMA Neurol 2013;70:34–43. doi: 10.1001/jamaneurol.2013.573. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Pisa M, Croese T, Dalla Costa G, et al. Subclinical anterior optic pathway involvement in early multiple sclerosis and clinically isolated syndromes. Brain 2021;144:848–862. doi: 10.1093/brain/awaa458. [DOI] [PubMed] [Google Scholar]
  • 51.Balk LJ, Coric D, Knier B, et al. Retinal inner nuclear layer volume reflects inflammatory disease activity in multiple sclerosis; a longitudinal OCT study. Mult Scler J Exp Transl Clin 2019;5:205521731987158–2055217319871582. doi: 10.1177/2055217319871582. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Sotirchos ES, Caldito NG, Filippatou A, et al. Progressive multiple sclerosis is associated with faster and specific retinal layer atrophy. Annals of Neurology 2020. doi: 10.1002/ana.25738. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Caldito NG, Saidha S, Sotirchos ES, et al. Brain and retinal atrophy in african-americans versus caucasian-americans with multiple sclerosis: A longitudinal study. Brain 2018;141:3115–3129. doi: 10.1093/brain/awy245. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Sotirchos ES, Gonzalez-Caldito N, Dewey BE, et al. Effect of disease-modifying therapies on subcortical gray matter atrophy in multiple sclerosis. Multiple sclerosis (Houndmills, Basingstoke, England: ) 2019:1352458519826364. doi: 10.1177/1352458519826364. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES