Retinal nerve fiber and ganglion cell complex layer thicknesses mirror brain atrophy in patients with relapsing-remitting multiple sclerosis

Abstract

Background:

Multiple sclerosis (MS) is associated with progressive brain atrophy, which in turn correlates with disability, depression, and cognitive impairment. Relapsing-remitting multiple sclerosis (RRMS) is a type of MS in which relapses of the disease are followed by remission periods. This is the most common type of the disease. There is a significant need for easy and low-cost methods to these cerebral changes. Changes in retinal layer thickness may reflect alterations in brain white and gray matter volumes. Therefore, this paper aims to determine whether retinal layer thickness, measured using optical coherence tomography (OCT), correlates with volumetric brain assessments obtained by magnetic resonance imaging (MRI).

Methods:

This retrospective cohort study recruited 53 patients with relapsing–remitting MS who underwent MRI and OCT examinations for evaluation of brain compartment volumes and thickness of retinal layers, respectively. OCT parameters, including central retinal thickness; retinal nerve fiber layer thickness (RNFL, peripapillary thickness); ganglion cell complex thickness (GCC, macular thickness); and Expanded Disability Status Scale (EDSS) results were compared with MRI parameters (cerebral cortex; cerebral cortex and basal ganglia combined; brain hemispheres without the ventricular system; and white matter plaques). We also checked whether there is a correlation between the number of RRMS and OCT parameters.

Objective:

Our primary objective was to identify whether these patients had retinal thickness changes, and our secondary objective was to check if those changes correlated with the MRI brain anatomical changes.

Results:

RNFL and GCC thicknesses were strongly (p-value < 0.05) associated with (i) cerebral cortex volume, (ii) combination of brain cortex and basal ganglia volumes, and (iii) the hemispheres but without the ventricular system. White matter plaques (combined) showed only weak or no correlation with RNFL and GCC. There was no correlation between central retinal thickness and brain compartment volumes, and there were weak or no correlations between the summary EDSS scores and OCT results.

Conclusions:

Retinal layer thickness measured by OCT correlates with select volumetric brain assessments on MRI. During the course of RRMS, the anatomo-pathological structure of the retina might serve as a surrogate marker of brain atrophy and clinical progression within selected domains.

Keywords

Multiple sclerosis optical coherence tomography brain atrophy cognitive impairment optic neuritis

1 Introduction

Multiple sclerosis (MS) is associated with progressive multi-domain disability and gradual brain atrophy. During the course of this disease, brain atrophy might correlate not only with progression of focal neurological symptoms but also or even more specifically with multiple non-focal syndromes such as depression and fatigue, and prominently with cognitive impairment (Andravizou et al. 2019). It is possible to assess cerebral atrophy using MRI volumetry (plaque and cortex volume), which is a robust but time-consuming modality that is not universally available. Importantly, MR imaging is especially difficult in certain groups of patients, for example, those suffering from claustrophobia (Enders et al., 2011). Development of new diagnostic tools for evaluating such patients, but also for general disease progression monitoring, including analysis of selected cerebral anatomo-histological structures, that is noninvasive, simple, and low-cost, would help estimate MS progression and monitor therapeutic efficiency in both routine practice and clinical trials.

Therefore, we investigated whether retinal layers’ thickness, measured using optical coherence tomography (OCT), correlates with volumetric brain assessments on MRI in patients with relapsing–remitting multiple sclerosis (RRMS).

2 Methods

2.1 Subjects

We retrospectively identified and analyzed data from 53 adult patients (36 women and 17 men, aged 23–57 years) with RRMS in different clinical stages of the disease, either during or shortly before interferon-beta therapy (Fig. 1 and Table 1, respectively). Patients were divided into two groups, i.e., with and without a history of optic neuritis. Patients previously diagnosed with glaucoma, macular disease, or refractory error higher than 5 Diopters (showed in refractometry after cycloplegia), were excluded from the analysis.

Fig. 1

EDSS scale of the study population –summary result.

Table 1

Demographic data of examined population

Age in years	37.8 (8.4)
Mean (SD)
Optic neuritis in the past (from the medical history)	25 (47.2)
n (%)	Right eye-4 (7.5)	Left eye-12 (22.6)	Both eyes-4 (7.5)	Don’t remember/no data-5 (9.6)
EDSS	2 (1.5)
Median (IQR)

SD, standard deviation; IQR, interquartile range.

All patients underwent a detailed neurological examination, including EDSS (Expanded Disability Status Scale) assessment by an experienced neurologist.

The study was approved by the local institutional ethics committee. Informed consent was obtained from all participants.

2.2 Optical coherence tomography

OCT examination was performed using a Zeiss Cirrus HD-OCT-400/4000 and Macular Cube 512×128 and Disk Cube 200×200 scans were used. Only scans with signal strength equal to or higher than 8 were considered. We analyzed central retinal thickness (CRT), retinal nerve fiber layer thickness (RNFL, peripapillary), and ganglion cell complex thickness (GCC, macular). RNFL thickness was analyzed in four field quadrants, while GCC was evaluated in six quadrants. Images were checked for misalignment of various layers by an experienced OCT technician.

2.3 Magnetic resonance imaging

MR imaging was performed in a 1.5T scanner (Siemens Magnetom Aera) using a 20-channel head/neck coil. The MRI protocol contained T1-weighted (T1w) 3D MPRAGE (high-resolution three-dimensional (3D) magnetization-prepared rapid acquisition with gradient echo) sequence (transverse orientation, TR = 1800 ms, TE = 3.3 ms, TI = 1000 ms, voxel size 1.4 mm×1.4 mm×1.1 mm, FOV 270 mm×270mm, slices 144, NSA = 1) and FLAIR (fluid-attenuated inversion recovery) 3D sequence (sagittal orientation, TR = 6000 ms, TE = 300 ms, TI = 1800 ms, voxel size 1.1 mm×1.1 mm×1.1 mm, FOV 270mm×270 mm, slices 120, NSA = 1). MRI analysis included measurement of white matter, gray matter, whole brain, and lesion volume.

The images were converted to ^*.nii format by MRIConvert. Next, FLAIR images were linearly registered to T1w images using FLIRT (linear registration of 3D scalar volumes) (Jenkinson & Smith, 2001), which is a part of FSL (Smith et al, 2004). MS lesions in white and gray matter were manually outlined by a single rater on transformed FLAIR images using MRIcron [Chris Rorden’s MRIcron, https://people.cas.sc.edu/rorden/mricron/index.html]. Brain tissue volumes were estimated using SIENAX (https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/SIENA, Oxford University, Oxford, UK) after normalization for subject head size (Smith, De Stefano, Jenkinson, & Matthews, 2001; Smith et al., 2002). Analysis was initiated by extracting brain and skull images from the single whole-head input data using the following BET options, namely, brain center estimation and fractional intensity threshold of 0.25 (Smith, 2002). Next, the brain image was affine-registered to the MNI152 space and a volumetric scaling factor was obtained. Subsequently, tissue-type segmentation with partial volume estimation was carried out (Zhang, Brady, & Smith, 2001) and tissue total volumes were calculated for grey matter, white matter, peripheral grey matter, and ventricular CSF. MS lesion masks were used during SIENAX analysis options to remove incorrectly labeled grey matter voxels. Lesion volume was estimated using FSLMATHS and multiplied by the scaling factor.

2.4 Statistical analyses

Statistical analyses and graphs were rendered using statistical software R, version 3.2.3. Normality of distribution was tested using the Shapiro–Wilk test, and Pearson’s correlation coefficient was calculated for quantitative variables showing a normal distribution. In other cases, Spearman’s rank test was used. A scatter plot with a linear regression line was drawn only for results with Pearson’s correlation coefficient greater than 0.45; this value was specified only after correlation analysis. Confidence intervals for the difference in correlations between the two groups (optic neuritis vs. without optic neuritis) were calculated using Zou’s asymptotic method (Zou, 2007), whereas Fisher’s Z transformation (Fisher, 1925) was used to calculate p-values. A significance level of p≤0.05 was set for all two-tailed tests. Significant associations are shown in Table 2.

Table 2
Significant associations(correlations)

OCT result Correlation with EDSS result

Average RNFL (μm) Summary EDSS

RNFL thickness in superior, nasal, inferior and temporal quadrant (μm) EDSS functional system: ambulation, pyramidal, cerebellar, brainstem, sensory, bowel and bladder, visual, cerebral

GCC thickness in superior, superotemporal, inferotemporal, inferior, inferonasal, superonasal quadrant MRI

Brain cortex volume

Brain ventricular volume

White matter volume

Brain cortex and basal ganglia summary volume

CRT Brain hemispheres volume without ventricular system

Demyelinating plaques volume

3 Results

Demographic data of the study cohort are presented in Table 1.

3.1 Correlation between RFNL and brain structure volumes

Volumes of (i) cerebral cortex (Fig. 2), (ii) cerebral cortex and basal ganglia combined, and (iii) brain hemispheres without the ventricular system were associated with average RNFL thickness for the entire retina in each of the eyes examined. Moreover, each volumetric value (cortex, cortex and basal ganglia and hemispheres without ventricular system) correlated with RNFL for each quadrant of the retina of each eye separately, except for the nasal quadrants of both eyes.

Fig. 2

Correlation between RNFL thickness (μm) (average for left eye) with brain cortex volume (μl).

The volume of the cerebral ventricular system was inversely correlated with both individual and mean quadrant RNFL thickness.

3.2 Correlation between GCC thickness and volume of brain structures

The volume of the cerebral cortex and the basal ganglia combined (Fig. 3) and of brain hemispheres without the ventricular system were associated with GCC thickness within each macular quadrant in each eye, except for the superior quadrant of the left eye. The volume of the cerebral ventricles was inversely associated with GCC thickness in each eye. There was no association or only a weak association between volumes of all white matter plaques combined and GCC thicknesses. White matter volume was not correlated with GCC thickness.

Fig. 3

Correlation between GCC thickness (μm) (inferonasal quadrant for left eye) with brain cortex and basal ganglia summary volume (μl).

There was no correlation between the number of RRMS and OCT parameters.

3.3 Correlation between CRT and volume of brain structures

CRT, determined separately for each eye, did not correlate with any of the following MRI-based volumes, namely, cerebral cortex, cortex and basal ganglia combined, cerebral hemispheres without ventricular system, white matter plaques combined, or total white matter, in both the groups.

Pearson’s or Spearman’s (where needed) correlations between OCT and MRI measures for the entire study population are shown in Tables 3 –5. Data from patients with a history of optic neuritis was analyzed separately, and the correlation between OCT and MRI parameters was stronger than that seen with the entire study population (Tables 6 7).

Table 3
Pearson’s correlations for OCT and MRI results for the entire study population

Pearson’s correlation Cerebral cortex volume Cerebral cortex and basal ganglia summary volume White matter volume

CRT OD –0.115 –0.114 0.113

Average RNFL OD 0.377^ 0.396^ 0.162

RNFL S OD 0.317^* 0.324^* 0.209

RNFL N OD 0.207 0.245 0.136

RNFL I OD 0.402^ 0.426^ 0.116

RNFL T OD 0.352^** 0.346^* 0.102

GCC S OD 0.397^ 0.401^ 0.077

GCC ST OD 0.362^ 0.372^ 0.048

GCC IT OD 0.400^ 0.411^ 0.055

GCC I OD 0.439^* 0.446^* 0.076

GCC IN OD 0.495^* 0.510^* 0.122

GCC SN OD 0.460^* 0.455^* 0.154

CRT OS –0.120 –0.108 0.085

average RNFL OS 0.489^* 0.518^* 0.068

RNFL S OS 0.440^* 0.458^* 0.085

RNFL N OS 0.141 0.191 –0.181

RNFL T OS 0.43^ 0.446^* 0.221

GCC S OS 0.233 0.251 0.060

GCC ST OS 0.433^* 0.460^* 0.109

GCC I OS 0.441^* 0.445^* 0.140

GCC IN OS 0.469^* 0.475^* 0.063

GCC SN OS 0.420^ 0.416^ 0.132

Pearson’s correlation	Cerebral cortex volume	Cerebral cortex and basal ganglia summary volume	White matter volume
CRT OD	–0.115	–0.114	0.113
Average RNFL OD	0.377^**	0.396^**	0.162
RNFL S OD	0.317^*	0.324^*	0.209
RNFL N OD	0.207	0.245	0.136
RNFL I OD	0.402^**	0.426^**	0.116
RNFL T OD	0.352^**	0.346^*	0.102
GCC S OD	0.397^**	0.401^**	0.077
GCC ST OD	0.362^**	0.372^**	0.048
GCC IT OD	0.400^**	0.411^**	0.055
GCC I OD	0.439^***	0.446^***	0.076
GCC IN OD	0.495^***	0.510^***	0.122
GCC SN OD	0.460^***	0.455^***	0.154
CRT OS	–0.120	–0.108	0.085
average RNFL OS	0.489^***	0.518^***	0.068
RNFL S OS	0.440^***	0.458^***	0.085
RNFL N OS	0.141	0.191	–0.181
RNFL T OS	0.43^**	0.446^***	0.221
GCC S OS	0.233	0.251	0.060
GCC ST OS	0.433^***	0.460^***	0.109
GCC I OS	0.441^***	0.445^***	0.140
GCC IN OS	0.469^***	0.475^***	0.063
GCC SN OS	0.420^**	0.416^**	0.132

^*p-value < 0.05; ^**p-value < 0.01; ^***p-value < 0.001. (OD, right eye; OS, left eye; RNFL, retinal nerve fiber layer; GCC, ganglion cell complex; CRT, central retinal thickness; S, superior quadrant; N, nasal quadrant; T, temporal quadrant; I, inferior quadrant; ST, superotemporal quadrant; IT, inferotemporal quadrant; SN, superonasal quadrant).

Table 4

Spearman’s correlations between OCT and MRI results for the entire study population

Spearman’s correlation	RNFL I-OS
Cerebral cortex volume	0.449^***
Ventricular volume	–0.431^**
Cerebral cortex and basal ganglia summary volume	0.480^***
White matter volume	0.133
Cerebral hemispheres without ventricles summary volume	0.413^**
Plaques volume	–0.350^*

^*p-value < 0.05; ^**p-value < 0.01; ^***p-value < 0.001. (OS, left eye; RNFL, retinal nerve fiber layer; I, inferior quadrant). Statistically significant data are red.

Table 5

Spearman’s correlations for OCT and MRI results for the study population

Spearman’s correlation
	Ventricular volume
Average RNFL- OD	–0.38^**
RNFL S-OD	–0.315^*
RNFL N-OD	–0.177
RNFL I-OD	–0.336^*
RNFL T-OD	–0.383^**
GCC S-OD	–0.400^**
GCC ST OD	–0.354^**
GCC IT OD	–0.399^**
GCC I-OD	–0.458^***
GCC IN OD	–0.499^***
GC C SN OD	–0.418^**
average RNFL OS	–0.530^***
RNFL S-OS	–0.391^**
RNFL N-OS	–0.317^*
RNFL I-OS	–0.431^**
RNFL T-OS	–0.455^***
GCC S-OS	–0.397^**
GCC ST OS	–0.479^***
GCC IT-OS	–0.409^**
GCC I-OS	–0.416^**
GCC IN OS	–0.356^**
GCC SN OS	–0.362^**

Table 6

Spearman’s correlation between OCT and EDSS functional system in two groups (optic neuritis vs. without optic neuritis)

EDSS functional system	EDSS- summary	“Cerebellar”	“Bowel and bladder”	“Cerebral”
RNFL T	–0.42/–0.077	–0.499/0.083	–0.643/0.089	–0.374/0.049
	(–0.832;0.196)	(–1.033;–0.038)^*	(–1.152;–0.226)^**	(–0.907;0.137)
GCC S	–0.485/–0.132	–0.615/–0.043	–0.526/–0.087	–0.4/0.102
	(–0.832;0.166)	(–1.018;–0.072)	(–0.906;0.079)	(–0.972;0.058)
GCC ST	–0.636/–0.129	–0.745/–0.053	–0.661/0.064	–0.507/0.275
	(–0.956;–0.026)^*	(–1.115;–0.235)^**	(–1.146;–0.226)^**	(–1.19;–0.237)^**
GCC IT	–0.742/–0.192	–0.774/–0.114	–0.635/–0.007	–0.509/0.096
	(–0.983;–0.12)^*	(–1.086;–0.226)^**	(–1.065;–0.128)^*	(–1.052;–0.063)^*
GCC I	–0.558/–0.196	–0.613/–0.106	–0.571/–0.138	–0.483/–0.012
	(–0.829;0.128)	(–0.959;–0.016)^*	(–0.896;0.064)	(–0.94;0.066)
GCC IN	–0.449/0.064	–0.547/0.082	–0.539/–0.152	–0.494/0.048
	(–0.979;0.037)	(–1.07;–0,097)^*	(–0.857;0.116)	(–1;–0.001)^*
GCC SN	–0.477/–0.052	–0,586/0.037	–0.529/–0.214	–0.474/0.159
	(–0.9;0.108)	(–1.064;–0.105)^*	(–0.787;0.179)	(–1.076;–0.082)^*

^*p-value < 0.05; ^**p-value < 0.01. Data shown in schema: r1/r2 (a;b), where, r1-correlation coefficient for data without optic neuritis, r2- correlation coefficient for data with optic neuritis, (a;b)- values in brackets show the 95% confidence interval (r1–r2).

Table 7

Spearman’s correlation between OCT and radiological data in two groups (optic neuritis vs. without optic neuritis)

	Cerebral cortex volume	Ventricules volume	Cerebral cortex and basal ganglia summary volume	White matter volume	Cerebral hemispheres without ventricles summary volume	Plaques volume
CRT	–0.109/–0.052	0.235/0.153	–0.099/0.018	0.162/0.11	–0.009/0.121	0.009/0.132
	(–0.622;0.517)	(–0.476;0.636)	(–0.675;0.465)	(–0.514;0.615)	(–0.682;0.456)	(–0.675;0.462)
Average RNFL	0.386/0.446	–0.666/–0.281	0.39/0.467	0.459/0.021	0.482/0.353	–0.314/–0.222
	(–0.531;0.42)	(–0.829;0.053)	(–0.544;0.397)	(–0.102;0.913)	(–0.347;0.602)	(–0.608;0.434)
RNFL S	0.232/0.426	–0.506/–0.192	0.219/0.447	0.404/0.07	0.324/0.362	–0.269/–0.22
	(–0.682;0.32)	(–0.791;0.19)	(–0.713;0.284)	(–0.209;0.825)	(–0.535;0.468)	(–0.574;0.481)
RNFL N	0.156/0.32	–0.265/–0.307	0.176/0.296	0.069/–0.318	0.133/–0.023	–0.191/–0.25
	(–0.675;0.376)	(–0.48;0.555)	(–0.636;0.419)	(–0.18;0.878)	(–0.41;0.688)	(–0.479;0.585)
RNFL I	0.4/0.375	–0.471/–0.197	0.398/0.424	0.383/0.018	0.513/0.35	–0.502/–0.177
	(–0.46;0,513)	(–0.759;0.236)	(–0.5;0.456)	(–0.187;0.856)	(–0.306;0.63)	(–0.803;0.182)
RNFL T	0,393/0,353	–0.679/–0.213	0,386/0,302	0.512/0.139	0.514/0.276	–0.181/–0.178
	(–0,451;0.533)	(–0.91;–0.018)^*	(–0.419;0.582)	(–0.138;0.848)	(–0.246;0.712)	(–0.544;0.535)
GCC S	0.513/0.445	–0.458/–0.41	0,513/0,379	0.544/0.005	0.672/0.303	–0.219/–0.195
	(–0.378;0.521)	(–0.519;0.419)	(–0.328;0.598)	(0.014;0.994)^*	(–0.061;0.81)	(–0.557;0.512)
GCC ST	0.441/0.456	–0.385/–0.415	0.403/0.403	0.379/0.052	0.53/0.353	–0.436/–0.197
	(–0.476;0.45)	(–0.455;0.509)	(–0.478;0.484)	(–0.221;0.822)	(–0.287;0.641)	(–0.73;0.278)
GCC IT	0.432/0.449	–0.455/–0.509	0.427/0.406	0.229/0.007	0.508/0.299	–0.496/–0.221
	(–0.481;0.452)	(–0.396;0.499)	(–0.453;0.499)	(–0.342;0.742)	(–0.272;0.683)	(–0.755;0.224)
GCC I	0.463/0.416	–0.505/–0.503	0.475/0.372	0.39/–0.064	0.615/0.234	–0.286/–0.186
	(–0.419;0.515)	(–0.443;0.432)	(–0.369;0.576)	(–0.106;0.932)	(–0.085;0.836)	(–0.622;0.435)
GCC IN	0.492/0.605	–0,551/–0,54	0.492/0.614	0.43/–0.264	0.608/0.268	–0.198/–0.482
	(–0.529;0.297)	(–0.431;0.402)	(–0.536;0.285)	(0.136;1.123)^*	(–0.12;0.796)	(–0.226;0.761)
GCC SN	0.513/0.443	–0.529/–0.383	0.524/0.393	0.489/–0.058	0.638/0.267	–0.185/–0.175
	(–0.377;0.523)	(–0.606;0.311)	(–0.325;0.591)	(0.004;1.005)^*	(–0.08;0.821)	(–0.549;0.53)

^*p-value < 0.05. Data shown in schema: r1/r2 (a;b), where r1- correlation coefficient for data without optic neuritis, r2- correlation coefficient for data with optic neuritis, (a;b)- values in brackets show the 95% confidence interval (r1–r2).

3.4 Correlation between OCT and EDSS

There were only weak or no correlations between EDSS and OCT parameters, i.e., RFNL, GCC, and CRT. However, there were some significant associations between the individual “cerebellar” and “bowel and bladder” functional systems and RNFL thickness. GCC thickness was associated with “vision,” “cerebellar,” “bowel and bladder,” “sensory,” and “cerebral” functional systems (Table 6).

4 Discussion

Here, in the largest study of its kind to-date, we describe significant and clinically relevant associations between the RNFL and GCC thicknesses, along with both cortical and deep gray matter volumes assessed by MRI. In contrast, correlations between these OCT measures and brain volumetric parameters that included white matter were either much weaker or were non-existent. Notably, combined white matter lesion volume was not associated with any of the basic OCT parameters.

These results suggest that changes in RFNL and GCC mirror neurodegenerative components of progressing MS pathology rather than inflammatory associated white matter pathology, and for the same reason, OCT may be more useful for monitoring progression of non-focal neurological deficits characteristic of MS, and especially cognitive deterioration.

Very few reports have described an association between brain atrophy and retinal nerve fiber thickness, and as these suffer from substantial methodological variations(for example, different stages of the disease, different methodology of brain volumetric assessment) and limitations, they have yielded divergent conclusions. For example, Grazioli et al. (2008) found a few associations between average RNFL thickness, normalized brain volume, normalized white matter volume, and T2-lesion volumes (Grazioli et al., 2008), whereas Siger et al. (2008) found that RNFL thickness correlates with brain parenchymal fraction and lesion volume. Many others have reported varying results as well (Abalo-Lojo et al., 2014; Cilingir et al., 2017; Di Filippo et al., 2010; Gordon-Lipkin et al., 2007; Lichtman-Mikol et al., 2019; Saidha et al., 2013; Siepman, Bettink-Remeijer, & Hintzen, 2010; Stellmann et al., 2017).

Although volumes of different brain compartments can be measured using MRI, these assessments are not included in routine examination protocols, and detailed post-processing of such data remains time-consuming. Moreover, MR images are difficult or even impossible to acquire in certain groups of patients, for example, in those with claustrophobia, and this phenomenon will persist till there is easy access to open MR machines. Thus, OCT can play a significant role in such situations and may even replace MRI in some selected and specific cases.

One advantage of our study is that we included a relatively homogenous patient population. In addition, to the best of our knowledge, no other studies have systematically evaluated potential correlations between brain anatomical structures and RNFL thickness in individual optic disc and macular quadrants. Therefore, we think that this is the first study to show that CRT does not correlate with brain compartment volume.

Published data indicate that the GCC best reflects gray matter atrophy in MS. Interestingly, GCC thickness in N-ON eyes decreases long before RNFL, and this cell layer is more resistant to inflammatory processes in the optic nerve (Merle et al., 2008). We demonstrate the strongest correlation between gray matter brain atrophy and GCC in nasal quadrants, which is well explainable since they are located close to the optic disc, due to the accumulation of nerve fibers (Talebi et al., 2013).

We did not find a significant correlation between OCT and the number of MS relapses or disease activity, and a possible explanation for this observation is that MS relapses are caused by focal white matter demyelination. In contrast, damage to the brain cortex, the basal ganglia, and nerve fibers, which are mirrored by OCT, is caused by other mechanisms. Further, as disease progress in white and gray matter in MS remains associated, it is realistic to expect a relationship between OCT parameters and demyelination.

We found only weak or no correlation between the EDSS and RNFL or GCC. There were, however, some associations between optic disc and macula quadrants and EDSS functional systems (FS). The strongest of these were that between RNFL and “bowel and bladder” or “cerebellar” FS and between GCC and “visual,” or “cerebellar.” Further, associations were observed between some quadrants and “bowel and bladder,” “sensory,” or “cerebral” FS. These results represent hitherto unknown results that should be confirmed in a larger patient sample.

In general, there are significant discrepancies between studies on EDSS and OCT parameters that might be due to a plethora of factors, including methodological and demographic limitations and obviously EDSS imperfections. Several groups have reported widely divergent data on this subject (Grazioli et al., 2008; Abalo-Lojo et al., 2014; Di Filippo et al., 2010; Talebi et al., 2013; Toledo et al., 2008; Sepulcre et al., 2007).

Only two previous publications have analyzed OCT parameters with both EDSS and brain atrophy elements (MRI) (Grazioli et al., 2008; Abalo-Lojo et al., 2014). Our study is also the first to describe associations between the retinal layers’ thickness in an individual disc, macular quadrants, and different EDSS functional systems.

However, this was a retrospective study, and therefore, we could not verify patients’ best-corrected low-contrast visual acuity and visual fields, which constitutes an important limitation.

In summary, we propose that OCT might be a useful, practical, and relatively easy technique for monitoring disease progression in MS, including brain atrophy. Thus, OCT might help better evaluate responses to therapies in MS under conditions where frequent systemic and routine MRI are limited or in individual cases where MRI scanning is contraindicated or difficult. However, due to the limitations of this particular study and similar works from other groups, large-scale studies are needed to confirm the findings described here.

Footnotes

Acknowledgments

None.

Conflict of interest

No conflicts of interest to disclose.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Ethics statement

The research has been carried out in accordance with The Code of Ethics of the World Medical Association (Declaration of Helsinki) and has been approved by the Ethics Committee Review Board of Medical University of Gdańsk. Informed consent was obtained from all participants.

References

Abalo-Lojo,

J.M.

, Limeres,

C.C.

, Gómez,

M.A.

, Baleato-González,

, Cadarso-Suárez,

, Capeáns-Tomé,

, & Gonzalez,

(2014). Retinal nerve fiberlayer thickness, brain atrophy, and disability in multiple sclerosispatients. Journal of Neuro-Ophthalmology, 34(1), 23–28. doi: 10.1097/WNO.0000000000000057

Andravizou,

, Dardiotis,

, Artemiadis,

, Sokratous,

, Siokas,

, Tsouris,

,. . . Hadjigeorgiou,

G.M.

(2019). Brain trophy in multiple sclerosis: mechanisms, clinical relevance and treatment options. Autoimmunity Highlights, 10, 7. https://doi.org/10.1186/s13317-019-0117-5

Cilingir,

, Batur,

, Bulut,

M.D.

, Milanlioglu,

, Yılgor,

, Batur,

, & Tombul,

(2017). The association between retinalnerve fibre layer thickness and corpus callosum index in differentclinical subtypes of multiple sclerosis. NeurologicalSciences, 38(7), 1223–1232. doi: 10.1007/s10072-017-2947-0

Di Filippo,

, Anderson,

V.M.

, Altmann,

D.R.

, Swanton,

J.K.

, Plant,

G.T.

, Thompson,

A.J.

, & Miller,

D.H.

(2010). Brain atrophy andlesion load measures over year relate to clinical status after 6 years in patients with clinically isolated syndromes. Journalof Neurology, Neurosurgery, and Psychiatry, 81(2), 204–208. doi: 10.1136/jnn2009.171769

Enders,

, Zimmermann,

, Rief,

, Martus,

, Klingebiel,

, Asbach,

, Klessen,

, Diederichs,

, Bengner,

, Teichgräber,

, Hamm,

, & Dewey,

(2011). Reduction ofclaustrophobia during magnetic resonance imaging: methods and designof the “CLAUSTRO” randomized controlled trial. BMC MedicalImaging, 10(11), 4. doi: 10.1186/1471-2342-11-4

Fisher,

R.A.

(1925). Statistical methods for researchworkers. Edinburgh, Scotland: Oliver & Boyd.

Gordon-Lipkin,

, Chodkowski,

, Reich,

D.S.

, Smith,

S.A.

, Pulicken,

, Balcer,

L.J.

, & Calabresi,

P.A.

(2007). Retinal nervefiber layer is associated with brain atrophy in multiple sclerosis. Neurology, 69(16), 1603–1609. doi: 10.1212/01.wnl.0000295995.46586.ae

Grazioli,

, Zivadinov,

, Weinstock-Guttman,

, Lincoff,

, Baier,

, Wong,

J.R.

, & Ramanathan,

(2008). Retinal nerve fiberlayer thickness is associated with brain MRI outcomes in multiplesclerosis. Journal of the Neurological Sciences, 268(1–2), 12–17. doi: 10.1016/j.jns.2007.10.020

Jenkinson,

, & Smith,

(2001). A global optimisation method forrobust affine registration of brain images. Medical ImageAnalysis, 5(2), 143–156. doi: 10.1016/S1361-8415(01)00036-6

10.

Lichtman-Mikol,

, Razmjou,

, Yarraguntla,

, Bao,

, Santiago-Martinez,

, Seraji-Bozorgzad,

, & Bernitsas,

(2019). Racial differences in retinal neurodegeneration as a surrogatemarker for cortical atrophy in MS. Multiple Sclerosis andRelated Disorders, 31, 141–147. doi: 10.1016/j.msard.2019.04.001

11.

Merle,

, Olindo,

, Donnio,

, Richer,

, Smadja,

, & Cabre,

(2008). Retinal peripapillary nerve fiber layer thickness inneuromyelitis optica. Investigative Ophthalmology and VisualScience, 49(10), 4412–4417. doi: 10.1167/iovs.08-1815

12.

Saidha,

, Sotirchos,

E.S.

, Oh,

, Syc,

S.B.

, Seigo,

M.A.

, Shiee,

, & Calabresi,

P.A.

(2013). Relationships between retinal axonal and neuronal measures and global central nervous system pathology inmultiple sclerosis. JAMA Neurology, 70(1), 34–43. doi: 10.1001/jamaneurol.2013.573.

13.

Sepulcre,

, Murie-Fernandez,

, Salinas-Alaman,

, García-Layana,

, Bejarano,

, & Villoslada,

(2007). Diagnostic accuracy of retinal abnormalities in predicting diseaseactivity in MS. Neurology, 68(18), 1488–1494. doi: 10.1212/01.wnl.0000260612.51849.ed

14.

Siepman,

T.A.

, Bettink-Remeijer,

M.W.

, & Hintzen,

R.Q.

(2010). Retinal nerve fiber layer thickness in subgroups of multiplesclerosis, measured by optical coherence tomography and scanninglaser polarimetry. Journal of Neurology, 257(10), 1654–1660. doi: 10.1007/s00415-010-5589-1

15.

Siger,

, Dzięgielewski,

, Jasek,

, Bieniek,

, Nicpan,

, Nawrocki,

, & Selmaj,

(2008). Optical coherence tomographyin multiple sclerosis. Thickness of the retinal nerve fiber layer apotential measure of axonal loss and brain atrophy. Journal ofNeurology, 255(10), 1555–1560. doi: 10.1007/s00415-008-0985-5.

16.

Smith,

S.M.

(2002). Fast robust automated brain extraction. Human Brain Mapping, 17(3), 143–155. doi: 10.1002/hbm.10062

17.

Smith,

S.M.

, De Stefano,

, Jenkinson,

, & Matthews,

P.M.

(2001). Normalised accurate measurement of longitudinal brain change. Journal of Computer Assisted Tomography, 25(3), 466–475. doi :10.1097/00004728-200105000-00022

18.

Smith,

S.M.

, Jenkinson,

, Woolrich,

M.W.

, Beckmann,

C.F.

, Behrens,

T.E.

, Johansen-Berg,

, & Matthews,

P.M.

(2004). Advances infunctional and structural MR image analysis and implementation asFSL. NeuroImage, 23(Suppl. 1), S208–S219. doi: 10.1016/j.neuroimage.2004.07.051

19.

Smith,

S.M.

, Zhang,

, Jenkinson,

, Chen,

, Matthews,

P.M.

, Federico,

, & De Stefano,

(2002). Accurate, robust andautomated longitudinal and cross-sectional brain change analysis. NeuroImage, 17(1), 479–489. doi: 10.1006/nimg.2002.1040

20.

Stellmann,

J.P.

, Cetin,

, Young,

K.L.

, Hodecker,

, Pöttgen,

, Bittersohl,

, & Siemonsen,

(2017). Pattern of gray mattervolumes related to retinal thickness and its association withcognitive function in relapsing-remitting MS. Brain andBehavior, 7(2), e00614. doi: 10.1002/brb3.614

21.

Talebi,

, Nikanfar,

, Sorkhabi,

, Sharifipour,

, Bahrebar,

, Kiavar,

, & Khanli,

H.M.

(2013). Optic coherence tomographyfindings in relapsing-remitting multiple sclerosis patients of thenorthwest of Iran. Iranian Journal of Neurology, 12(3), 81–86.

22.

Toledo,

, Sepulcre,

, Salinas-Alaman,

, García-Layana,

, Murie-Fernandez,

, Bejarano,

, & Villoslada,

(2008). Retinalnerve fiber layer atrophy is associated with physical and cognitivedisability in multiple sclerosis. Multiple Sclerosis, 14(7), 906–912. doi: 10.1177/1352458508090221

23.

Zhang,

, Brady,

, & Smith,

(2001). Segmentation of brain MRimages through a hidden Markov random field model and theexpectation maximization algorithm. IEEE Transactions onMedical Imaging, 20(1), 45–57. doi: 10.1109/42.906424

24.

Zou,

G.Y.

(2007). Toward using confidence intervals to comparecorrelations. Psychological Methods, 12(4), 399–413. doi: 10.1037/1082-989X.12.4.399

OCT result	Correlation with	EDSS result
Average RNFL (μm)		Summary EDSS
RNFL thickness in superior, nasal, inferior and temporal quadrant (μm)		EDSS functional system: ambulation, pyramidal, cerebellar, brainstem, sensory, bowel and bladder, visual, cerebral
GCC thickness in superior, superotemporal, inferotemporal, inferior, inferonasal, superonasal quadrant		MRI
		Brain cortex volume
		Brain ventricular volume
		White matter volume
		Brain cortex and basal ganglia summary volume
CRT		Brain hemispheres volume without ventricular system
		Demyelinating plaques volume