Objective To investigate the coexistence of anterograde and retrograde trans-synaptic axonal degeneration, and to explore the relationship between selective visual pathway damage and global brain involvement in longstanding multiple sclerosis (MS).
Methods In this single-centre, cross-sectional study, patients with longstanding MS (N=222) and healthy controls (HC, N=62) were included. We analysed thickness of retinal layers (optical coherence tomography), damage within optic radiations (OR) (lesion volume and fractional anisotropy and mean diffusivity by diffusion tensor imaging) and atrophy of the visual cortex and that of grey and white matter of the whole-brain (structural MRI). Linear regression analyses were used to assess associations between the different components and for comparing patients with and without optic neuritis and HC.
Results In patients with MS, an episode of optic neuritis (MSON) was significantly associated with decreased integrity of the ORs and thinning of the peripapillary retinal nerve fibre layer (pRNFL) and macular ganglion cell complex (GCC). Lesion volume in the OR was negatively associated with pRNFL and GCC thickness in patients without optic neuritis (MSNON). The pRNFL and GCC showed associations with integrity of the OR, thickness of the primary visual cortex (only in patients with MSON), and also with global white and grey matter atrophy. In HCs, no such relationships were demonstrated.
Interpretation This study provides evidence for presence of bidirectional (both anterograde and retrograde) trans–synaptic axonal degeneration in the visual pathway of patients with MS. Additionally, thinning of the retinal pRNFL and GCC are related to global white and grey matter atrophy in addition to pathology of the visual pathway.
- MULTIPLE SCLEROSIS
Statistics from Altmetric.com
Neuroaxonal degeneration largely contributes to irreversible disability in multiple sclerosis (MS).1–4 It is, however, still not known what mechanism drives neuroaxonal degeneration. To investigate the dynamics of neuroaxonal degeneration, the visual pathway is a particularly suitable model, as it is a hard-wired, single pathway model (figure 1). Previously, we and others have demonstrated that an episode of optic neuritis in patients with MS (MSON) is related to thinning of the inner layers of the retina, most likely as a result of retrograde (towards the retina) axonal degeneration.5–7 Importantly, significant thinning of the innermost retinal layers was also observed in patients without any history of optic neuritis (MSNON), which was suggested to be a result of retrograde trans-synaptic degeneration. More recent studies have suggested that this mechanism is responsible for the spreading of neuroaxonal degeneration in patients with MS and stroke.8–11
In addition to the previously described evidence for retrograde trans-synaptic degeneration (related to thinning of the inner retinal layers),7 there is also evidence for anterograde (directed towards the visual cortex) trans-synaptic degeneration (related to damage of the optic radiations (ORs) and primary visual cortex) in the visual system of patients with MS.10 ,12 Therefore, the unifying term ‘bidirectional (trans-synaptic) axonal degeneration’ was proposed4 to better describe what originally was reported as ‘insidious atrophy’.13 The relation between retrograde and anterograde trans-synaptic degeneration in the visual pathway of patients with MS (as a result of either lesions is the optic nerve or elsewhere in the visual pathway), and its relationship with whole-brain atrophy, is however not entirely clear.
Therefore, the aim of this study was to investigate for the presence of bidirectional (comprising anterograde and retrograde) trans-synaptic degeneration in the visual system of patients with MS, and to explore how damage of the visual system is associated with damage in the rest of the brain.
This study was approved by the medical ethical committee (protocol number 2010/336) and the scientific research committee (protocol number CWO/10-25D) of the VU University Medical Center in Amsterdam, The Netherlands. Written informed consent was obtained from all included subjects.
Study design and patient population
Patients with MS and healthy controls (HC) were enrolled from the VU University medical centre Amsterdam. This cohort has been previously described.7 ,14 From eight of the original 230 patients with MS and one of the 63 HCs described earlier, MRI diffusion tensor imaging (DTI) data was not available. Therefore, the present paper reports the remaining 222 patients and 62 HCs.
All included patients adhered to the following inclusion criteria: aged between 18 years and 80 years, a diagnosis of either relapsing remitting MS (RRMS), secondary progressive MS (SPMS) or PPMS at the time of their assessment (defined by Lublin-Reingold criteria).15 Patients were excluded if they fulfilled any of the following criteria: pregnancy; received a course of steroids or had a relapse within 6 weeks prior to inclusion; HIV or other immunodeficiency syndrome; history of substance abuse (drug or alcohol) and non-MS radiological findings that could interfere with evaluation.
Healthy control subjects were eligible for inclusion if they were aged between 18 years and 80 years at the time informed consent was signed, and if they had no history of any neurological, ophthalmological or psychiatric disease. Exclusion criteria for healthy control subjects were similar to those for MS patients, with the addition of 1st-degree or 2nd-degree familial relation to an MS patient.
Retinal imaging was performed with spectral domain optical coherence tomography (SD-OCT, Spectralis, Heidelberg Engineering, Heidelberg, Germany), with the eye tracking function enabled for optimal measurement accuracy.16 A detailed description of the scan protocol has been described elsewhere.7 In brief, peripapillary data were obtained using a 12° ring scan and macular data with a macular volume scan (20×20° field, 25 B-scans, see online supplementary figure S1). Scans were excluded from the analyses if they violated, validated international consensus quality control criteria (OSCAR-IB).17 ,18 In order to assess retinal pathology potentially resulting in exclusion of OCT scans, the personal and medical history of visual symptoms was obtained from all patients and controls and judged in view of the clinical examination. The assessment of history of symptomatic MSON was based on established, phenotypic diagnostic criteria.19 An exclusion criterion was an episode of MSON in the past 6 months.
Image postprocessing was performed for both scans, using commercial segmentation software (Heidelberg Engineering, Heidelberg, Germany). Although the software enabled reliable segmentation of multiple layers in the peripapillary and macular area,20 the mean peripapillary retinal nerve fibre layer (pRNFL) and the macular ganglion cell complex (GCC) (perimacular rim) have shown to be most sensitive to changes in the central nervous system (CNS),5–7 and are, therefore, used in this study (see online supplementary figure S1).
Magnetic resonance imaging
Structural MRI was performed on a 3 T whole-body scanner (GE Signa HDxt, Milwaukee, Wisconsin, USA), using a three-dimensional T1-weighted fast spoiled gradient echo sequence (FSPGR) for volumetics, a three-dimensional fluid-attenuated inversion recovery image (FLAIR) for lesion detection, and DTI for integrity measurements and tractography (see online supplementary e-methods for more details).
Brain and lesion volumes
Normalised grey and white matter volumes (NGMV and NWMV respectively) and lesion volumes were quantified automatically using kNN-TTP,21 and SIENAX (part of the FMRIB Software Library (FSL) 5.0.4, http://www.fmrib.ox.ac.uk/fsl). Lesion filling was applied to minimise the effect of lesions on atrophy measurements.22 FreeSurfer 5.1 was used to quantify the average thicknesses of V1, V2, and V5 (see online supplementary e-methods for more details).
DTI data processing and segmentation of the ORs
The DTI images were corrected for head movement and eddy current distortions using FMRIB's Diffusion Toolbox (FDT; also part of FSL). Subsequently, the diffusion tensor was fitted, from which the fractional anisotropy (FA), mean diffusivity (MD) were calculated.
The diffusion images were then used to segment the ORs and derive tract-specific pathology measures. As tractography in the presence of MS pathology might lead to unreliable results, this was done by means of an atlas based on the healthy controls. Probabilistic tractography was performed between the thalamus and V1 in each hemisphere to obtain the ORs in each healthy control. The tractography results were thresholded, non-linearly registered to a standard space and averaged to obtain an atlas. The atlas was subsequently propagated to each subject to obtain a subject specific OR-segmentation (see online supplementary e-methods for more details).
For each subject and tract separately, the weighted average FA and MD values inside the tract, as well as the lesion volume in the tract, were computed using the atlas probability values as a weighting factor. Weighting was performed to emphasise integrity values in the centre of the tract. Also, the weighted lesion volume inside each tract was computed as described above. Note that the weighted lesion volume is a relative, instead of an absolute value, and cannot be compared with absolute whole-brain lesion volumes.
In order to compare demographic data between patients and healthy controls, independent sample t test and χ2 tests were performed. Next, linear regression analyses with adjustments for age were used to test for differences in the individual structures of the visual pathway between patients with MS and HCs.
With the aim of investigating the effect of an episode of optic neuritis (MSON) on anterograde (towards ORs) and retrograde (towards retina) degeneration, integrity of the OR and retinal layer thickness were compared between patients with patients with MSON, MSNON and HCs. The analyses with MD and FA were adjusted for lesion load in the OR. Importantly, for the analyses including pRNFL and GCC thickness, the mean data of both eyes was used. Therefore, these analyses only included patients in whom both eyes had the same history of ON (bilateral MSON or bilateral MSNON). This was done to avoid contamination of the study groups by pooling of MSON eyes (showing a severe degree of pRNFL loss) with MSNON eyes (showing a small degree pRNFL loss).8
Furthermore, it is important to note that an episode of MSON acted as an effect modifier in the multiple regression analyses investigating the associations between retinal layer thickness and other structures of the visual pathway and global brain. Therefore, all these analyses were stratified by MSON, MSNON and HCs.
The effect of lesion volume in the OR on retinal layer thickness (retrograde trans-synaptic degeneration) was investigated using linear regression analyses with adjustments for age.
Next, the associations between retinal layer thickness and integrity of the OR (as measured by weighted FA and MD), NGMV and NWMV were investigated using linear regression analyses with adjustments for age.
In order to increase comparability and ease interpretation of the multiple regression analyses, all values from regression analyses are reported as standardised ß. Statistical analyses were performed using SPSS V.22.0, with a statistical significance level of 0.05.
A total of 222 MS patients (137 RRMS, 57 SPMS, 28 PPMS) and 62 healthy control subjects (HCs) were included in this study. Patients with MS were slightly older than HCs (p<0.05) and had a mean disease duration of 20.3 (±7.0) years (range 8.8–45.9). One hundred and two patients (45.9%) had ever experienced an episode of MSON, and 38 (17.1%) had experienced bilateral MSON. The remaining 103 patients (46.4%) had never experienced an episode of MSON. Only one patient (with previous MSON) showed signs of unilateral microcystic macular oedema (MMO). Exclusion of data from this patient did not change the level of significance for any of the following analyses. An overview of the clinical and demographic data is shown in table 1.
Visual pathway damage in MS
The differences between patients with MS and HCs for all subsequent components of the visual system are shown in table 2. All structural measures were significantly decreased in patients with MS, compared with HC (p<0.005 for all comparisons). Additionally, NWMV and NGMV showed significant reductions in patients with MS, compared with HC (p<0.001 for both comparisons). For comparisons of all components between MSON and MSNON patients, see online supplementary table S1.
Bidirectional trans-synaptic degeneration in the visual system
Anterograde trans-synaptic degeneration
In patients with bilateral MSON, significant loss of integrity of the OR (measured by FA) was demonstrated, compared with patients with bilateral MSNON and HCs (figure 2A, left graph). Consistent with these data, diffusivity (MD) was increased where anisotropy (FA) was decreased. Whereas MD was significantly lowest in HCs, followed by patients with MSNON and MSON, the difference between the MSON and MSNON groups was not statistically significant (figure 2A, right graph).
Retrograde trans-synaptic degeneration
For the comparison of retinal pRNFL and GCC between MSON, MSNON and HCs, a clear trend was observed. Patients with MSON showed the most severe thinning of both retinal layers compared to patients with MSNON and HCs (p<0.001 for all comparisons). Importantly, patients with MSNON also showed significant thinning of both retinal layers, compared with HCs (p<0.001 for both comparisons, figure 2B).
Additionally, the association between weighted lesion volume in the OR and retinal layer thickness was investigated. These analyses were stratified by MSON and MSNON. Increased lesion volume in the OR was related to a significant decrease of the pRNFL and GCC (standardised ß −0.33, p=0.003 and standardised ß −0.32, p=0.005, respectively). This relationship disappeared when patients had experienced MSON, due to the severe damage in the retina (standardised ß −0.15, p=0.520 and standardised ß −0.10, p=0.686, respectively).
Structurally related neurodegeneration in the visual pathway and the global brain
Atrophy of the visual pathway
Table 3 shows the results of the regression analyses with retinal layer thickness (pRNFL and GCC) and integrity of the OR (FA and MD) and visual cortex thickness (V1, V2 and V5). In patients with MSON, the retinal layer thickness was positively associated with FA and negatively associated with MD. The associations were, however, not statistically significant, due to reduced statistical power (n=38). The more substantial axonal loss from MSON was related to a significant degree of localised atrophy of V1. An anatomical relationship exists with V2, to which the OR projects as well, but the study was underpowered to show this. No associations with V5 were demonstrated.
In patients with MSNON, the GCC and, to a lesser extent, the pRNFL, were directly associated with the FA (standardised ß 0.72, p<0.001 and 0.42, p=0.064, respectively). Consistently, both were, although not statistically significant, inversely correlated with the MD (standardised ß −0.46, p=0.110 and standardised ß −0.11, p=0.715, respectively).
In the visual pathway of HCs, there were no relationships observed between pRNFL or GCC thicknesses and MRI metrics of the OR, V1, V2 or V5 in HCs.
The associations between retinal layer thickness and brain volume structures are illustrated in table 3. In patients with MSON, the pRNFL and the GCC tended to be positively associated with NWMV and NGMV, but these relationships were not statistically significant, probably due to reduced statistical power. Interestingly, in the group of patients with MSNON, similar positive associations with both NWMV and NGMV were observed, all statistically significant. No such quantitative relationships were found for HCs.
This study provides evidence for the presence of bidirectional trans-synaptic axonal degeneration in the visual pathway of patients with MS. Degeneration of neurons and axons disseminates by means of both anterograde (towards the visual cortex) and retrograde (towards the retina) trans-synaptic axonal degeneration. Importantly, retinal layer thickness did also reflect global brain atrophy of both white and grey matter.
First, the present data shows that a history of MSON is associated with decreased integrity of the OR. This decrease in integrity of the OR is thought to be caused by anterograde trans-synaptic degeneration.
Second, besides decreased integrity of the OR, patients with MSON also showed significantly more thinning of the pRNFL and GCC compared with patients with MSNON or HCs. These findings showed that episodes of optic neuritis are not only suggestive of anterograde degeneration of the OR, but also of retrograde degeneration of the optic nerve and retinal layers. Importantly, whereas the presence of MSON was related to thinning of the retinal layers by retrograde degeneration, there was also significant retinal thinning observed in patients with MSNON. In these patients, with no history of lesions in the optic nerve, retinal layer thinning is most probably caused by retrograde trans-synaptic degeneration, caused by local damage in other structures of the optic pathways.
Likewise, lesion volume in the OR was also related to retinal layer thickness, suggesting that lesions in the OR may lead to thinning of the retina by retrograde trans-synaptic degeneration. Interestingly, this relationship was less clear in patients with MSON, which was most probably caused by the fact that this subtle relationship was masked by the severe retinal thinning in patients with MSON.
Third, there were clear structural relationships between retinal layer thickness and other components of the visual pathway. Interestingly, the severe axonal loss in the optic nerve as a result from MSON, seemed to result in a significant degree of localised atrophy of V1. Although a weak relationship may still exist with V2, no relationship whatsoever was found with V5. In patients with MSNON however, retrograde trans-synaptic axonal degeneration is held responsible for the relationship between damaged OR and atrophy of the pRNFL and GCC. Importantly, in HCs there was no relationship whatsoever between pRNFL or GCC thicknesses and MRI metrics of the OR, V1, V2 or V5. This implies that the aforementioned relationships in patients with MS were atrophy based, and therefore, only present in case of neuroaxonal damage.
The presence of retrograde trans-synaptic degeneration in the visual pathway of MS has been described before.7 ,10 ,11 ,23 In a previous study, we described the presence of retrograde trans-synaptic degeneration in the eyes of patients with MS. An episode of MSON was strongly associated with severe atrophy of the inner layers of the retina, but not the outer layers. This suggested that retrograde (trans-synaptic) axonal degeneration is halted at an anatomical structure capable of neuroplasticity, the inner nuclear layer (INL) of the retina.7 The present data suggest that the secondary visual cortex, like the INL, highly capable of plasticity,24 may act as a physiological barrier to anterograde trans-synaptic degeneration, but this should be further investigated. In order to test this hypothesis, one would need to make use of detailed voxel-based morphometry of different areas of the higher visual cortex.
Consistent with the present study, others have found that in patients with optic neuritis, the OR,25–27 but also the visual cortex,28 ,29 revealed damage after episodes of MSON. A more recent study investigating multiple components of the visual pathway, reported that pRNFL thinning was associated with atrophy of the visual cortex, but this relationship disappeared when MSON and MSNON patients were analysed separately (N=100).10 Gabilondo et al10 also demonstrated a relationship between retinal layer thickness and lesion volume in the OR.
The present study extends on previous studies, by investigating individual retinal layers, the complete visual pathway and the global brain in a larger number of subjects. We observed a clear association between retinal layer thickness and global NWMV and NGMV, which is consistent with previous studies in MS30 ,31 and NMO.32 These data suggest that trans-synaptic degeneration might not be restricted to the visual pathway but could disseminate through the whole-brain. However, in our opinion, it is most unlikely that this process can be solely explained by trans-synaptic degeneration, as structures capable of plasticity act as a barrier to this process. Besides, if trans-synaptic degeneration would not be stopped, it would spread through the brain like a domino effect. Therefore, the observed damage to the global brain structures are more likely to be caused by a more global mechanism. In this systemic model, neuroaxonal degeneration is driven by a global mechanism, affecting the entire brain, including the visual system. Although hypothetical, a possible explanation for this systemic model could be mitochondrial failure, also known as the ‘virtual hypoxia hypothesis’ in MS research.33
An important limitation of this study was its cross-sectional nature. We were, therefore, unable to determine whether the degree of atrophy may already have reached a plateau after an average of 20 years of disease duration, or continues to progress. To address this question, longitudinal data is needed. Next, the HCs were significantly younger than our patient cohort. This was however taken into account by age-adjusted statistical tests. Finally, the assessment of history of MSON was based on clinically confirmed episodes and patient-reported history on earlier events during the disease. Although this approach is consistent with other studies in the field, it may have introduced a bias towards more severe episodes of MSON and subclinical episodes may have gone unnoticed.
In summary, this study provides evidence for presence of bidirectional trans–synaptic axonal degeneration in the visual pathway of patients with MS. Additionally, thinning of the retinal pRNFL and GCC reflect visual pathway damage, and also global white and grey matter atrophy.
We would like to thank the patients and healthy controls for participating, and their substantial time commitment. We also want to thank the staff of the Departments of Ophthalmology and Neurology, the data unit of the MS Center VUmc and the Dutch MS research foundation for supporting this study.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
Contributors LJB contributed to designing the study concept, performing data collection, revision and QC assessment of OCT scans, automated and manual layer segmentation, statistical analyses and prepared the first draft of the manuscript. MDS and MD and assisted with data collection, MRI processing and revised the manuscript. PT performed data collection, revision and QC assessment of OCT scans and revised the manuscript. JK, BMJU and CHP contributed to designing the study concept and revised the manuscript. FB, MPW and HV assisted with MRI processing, QC assessment of MRI scans and revised the manuscript. AP contributed to designing the study concept, revision and QC assessment of OCT scans, automated and manual layer segmentation, statistical analyses and revising the manuscript. All authors gave final approval of the version submitted.
Competing interests LJB, MDS, MD and AP are supported by the Dutch MS Research Foundation through a programme grant to the VUmc MS Center Amsterdam (grant number 09-538d). MD was supported by a private sponsorship to the VUmc MS Center Amsterdam. JK has received speaker and consulting fees from Merck-Serono, Biogen-Idec, Novartis, Teva, Genzyme and Novartis. FB has received consultation fees from Bayer-Schering Pharma, Sanofi-Aventis, Biogen-Idec, Teva, Novartis, Roche, Synthon BV and Jansen Research. HV receives research support from the Dutch MS Research Foundation, grant numbers 05-358c, 09-358d and 10-718MS and has performed sponsored contract research projects for Pfizer, Novartis, and Merck-Serono. MPW has received consultation fees from consultant for Roche and Biogen Idec. BMJU has received consultation fees from Novartis, Merck Serono, Biogen Idec, and Danone Research. CHP received an institutional grant from Novartis; consultancy fees from Actelion, Biogen Idec, Bayer Schering, TEVA, Merck-Serono, Novartis, Glaxo SK, UCB, Roche, Antisense Ther; expert testimony: Biogen. AP sits on the Novartis steering committee for a multicentre observational study: ‘A 3-year, open-label, multi-centre, multi-cohort, parallel-group study to validate optical coherence tomography in patients with multiple sclerosis’.
Ethics approval Medical ethical committee (protocol number 2010/336) of the VU University Medical Center in Amsterdam, the Netherlands.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.