Retinal neuronal loss and progression independent of relapse activity in multiple sclerosis.

Introduction

Multiple sclerosis (MS) is a chronic disease of the central nervous system (CNS), characterized by inflammatory and neurodegenerative processes [ 1 ]. Disability accumulation in MS can result from residual symptoms after relapses and ongoing progression. Recently, the term “PIRA” was introduced to describe progression independent of relapse activity and was shown to be the main driver of disability in people with MS (pwMS) [ 2 – 4 ].

PIRA may occur early, is notoriously difficult to capture and quantify, thus could remain undetected or be misclassified by clinicians [ 2 ]. Patients with PIRA events within five years after the first demyelinating episode show increased risk for long-term progression [ 5 ]. Thus, early identification of PIRA is crucial for the stratification of patients and possibly application of different treatment algorithms. In this regard, magnetic resonance imaging (MRI) markers like cortical atrophy [ 6 ], spinal cord atrophy and paramagnetic rim lesions (PRL) [ 2 , 7 ] as well as serum-biomarkers like glial fibrillary acidic protein (GFAP) [ 8 ] were shown to be associated with higher PIRA risk. However, not all these markers have been validated for individual use and are practically applicable.

Optical coherence tomography (OCT) is a non-invasive, highly reproducible method [ 9 ] that can quantify neuronal and axonal loss in the retina of pwMS [ 10 ]. Retinal atrophy correlates with diffuse gray matter thinning in MS [ 11 ] and increased risk of disability worsening [ 12 , 13 ]. Thus, OCT could offer additional patient-friendly markers of PIRA in MS. Up to date, a single study investigated the association between OCT and PIRA and found an increased rate of retinal atrophy (thinning of the peripapillary retinal nerve fiber layer, pRNFL and the macular ganglion cell-inner plexiform layers, mGCIPL) in patients with PIRA events, over a 4-year period [ 14 ].

Our aim was to investigate the relationship between OCT measures (pRNFL, mGCIPL as well as the macular inner nuclear layer, mINL) and the risk of PIRA in a well characterized MS cohort, with longer observation time.

Methods

Study participants

We included patients from the MS Center at the University Hospital of Basel (UHB), who participate in the Swiss MS Cohort Study (SMSC), a currently ongoing prospective observational study. Our inclusion criteria were: (1) age ≥ 18 years old (2) MS diagnosis according to the 2017 revised McDonald criteria [ 15 ], and (3) at least one available OCT examination. The exclusion criteria were: (1) presence of any retinal pathology that may interfere with the validity of the OCT, as previously described [ 16 ], (2) history of both eyes being affected clinically by optic neuritis (ON) (3) less than 4 years of observation time, iv) OCT performed > 6 months apart from the clinical visit.

Assessment of PIRA rates

Patients were followed longitudinally in the SMSC and assessed every 6–12 months by certified neurologists in the MS Center at the UHB. Physical disability was assessed using the Neurostatus-EDSS instructions and definitions [ 17 ]. PIRA events were defined as a relevant increase in the EDSS score compared to the previous (reference) EDSS score, confirmed at a subsequent visit at least 6 months apart and with no clinical relapses between the reference and event visit. The criteria for a relevant EDSS increase were as follows: an increase of ≥ 1.5 points if the baseline EDSS is 0, an increase of ≥ 1 point if the baseline EDSS is between 1.0 and 5, or an increase of ≥ 0.5 points if the baseline EDSS is greater than 5 [ 4 ].

We calculated individual PIRA rates, defined as the number of events divided by the total years of observation time per patient.

In addition, we defined PIRMA (Progression Independent of Relapse and MRI Activity) as PIRA events without any new or enlarging lesions on fluid-attenuated inversion recovery (FLAIR) sequences, between reference- and event visits.

OCT

OCT was performed on a Heidelberg Engineering Spectralis device (Heidelberg, Germany), in a dark room, without pharmacologic pupil dilation. The following OCT markers were assessed on both eyes: (1) mean thickness of the pRNFL, (2) mean thickness of the combined mGCL and mIPL (mGCIPL) and (3) mean thickness of the mINL. Strict quality control of the OCT scans was performed applying the OSCAR-IB criteria [ 16 ], which resulted in the exclusion of 37 peripapillary ring scans and 37 macular scans (for details, see supplementary table). Details about the OCT-acquisition settings and scanning protocol according to the APOSTEL 2.0 recommendations [ 18 ] are also summarized in Supplementary Table 1.

The pRNFL was measured using 3.4-mm ring scans around the optic nerve head, while the mGCIPL and mINL were measured using a macular volume scan around the fovea (for scan details see supplementary table). After an initial automated segmentation of all intraretinal layers using the software provided by Heidelberg Engineering (Heyex Version 2.5.3), segmentation results of the different layers were checked and corrected where needed by experienced raters. Mean thicknesses of mGCIPL and mINL were calculated using the 6-mm-diameter cylinder of the 1,3,6-mm ring adjacent to the fovea [ 19 ] and transforming the volume to mean thickness according to the formula below, where for the 6-mm-diameter cylinder r = 3 mm and thus π · r 2  = 28.27433:

The OCT markers used in the study are also depicted in Fig.  1 .

Peripapillary and macular OCT markers used in this study. GCIPL ganglion cell-inner plexiform layer, INL inner nuclear layer, OCT optical coherence tomography, RNFL retinal nerve fiber layer

Patients with history of bilateral ON were not included in this study, since the profound bilateral atrophy would mask any meaningful analysis of the subtle atrophy excepted with progressive retinal inner atrophy. In line with this, we also excluded eyes with possible local damage (optic nerve demyelination), by calculating the inter-eye asymmetry, using previously described thresholds: for pRNFL ≥ 5 µm and for mGCIPL ≥ 4 µm [ 20 – 22 ] and excluding the eye with the lower thickness. For patients with asymmetry in mGCIPL, also mINL of the worse eye was excluded from the analysis. For all other patients, without interocular asymmetry, we used the average of right and left layer thicknesses.

Statistical analysis

All statistical analysis were performed using R [ 23 ] version 4.3.3 with packages: ellipsis, readxl, car and ggplot2. Data distributions were analyzed visually and statistically.

To investigate the association between OCT markers and annualized PIRA rates, we used linear regression with each OCT marker as dependent variable and PIRA rate as independent variable (unadjusted models). Then, we performed linear regression models, adjusted for the following characteristics that were previously reported to affect retinal layers and/or PIRA risk: disease duration, age at disease onset, sex, body mass index (BMI), dominant disease modifying treatment (DMT) during the observation time and annualized relapse rate during the observation time. We used the annualized PIRA rates in the years before the OCT examination, to achieve maximum observation time.

Moreover, we performed a sensitivity analysis, using PIRMA (instead of PIRA) rates, with the same variables as described above. Patients were excluded from this analysis if MRI activity could not be assessed for any of their previous PIRA events due to insufficient imaging data ( n  = 22).

We report the estimated coefficients, 95% confidence intervals (CI), and p values for each model; p values < 0.05 were considered statistically significant.

Results

Patient characteristics

Overall, 171 patients were included. The patients’ characteristics at time of the OCT examination are summarized in Table  1 . Most patients were women; while the median disease duration was 16 years, the EDSS was relatively low (median 2.5). Most patients showed a relapsing course and were treated with DMT (Table  1 ).

Table 1 Patient characteristics at time of OCT Parameter People with MS ( n  = 171) Age, median years [IQR] 51.0 [42.4, 59.8] Sex, female n (%) 110 (64.3) % Disease duration, median years [IQR] 16 [10.5, 23.1] EDSS, median [IQR] 2.5 [1.5, 4.0] Relapsing course, n (%) 148 (86.5%) DMT, n (%) 145 (84.8%)  Interferon-beta agents − 8 (4.7%)  Glatirameracetate − 3 (1.8%)  Fumarates (dimethyl fumarate, diroximel fumarate) − 14 (8.2%)  Teriflunomide − 7 (4.1%)  S1P-receptor modulators (fingolimod, ozanimod) − 60 (35.1%)  Anti-CD20 B-cell-depleting (ocrelizumab, ofatumumab, rituximab) − 43 (25.2%)  Natalizumab − 10 (5.8%) Untreated n (%) 26 (15.2%) pRNFL thickness median [IQR], µm 94.0 [85.8, 102.0] mGCIPL thickness median [IQR], µm 67.2 [60.1, 70.7] mINL thickness median [IQR], µm 35.4 [31.8, 35.4] DMT disease modifying Treatment, EDSS expanded disability status scale, mGCIPL macular ganglion cell-inner plexifom layer, mINL macular inner nuclear layer, IQR interquartile range, MS multiple sclerosis, OCT optical coherence tomography, pRNFL peripapillary retinal nerve fiber layer

Patient characteristics at time of OCT

DMT disease modifying Treatment, EDSS expanded disability status scale, mGCIPL macular ganglion cell-inner plexifom layer, mINL macular inner nuclear layer, IQR interquartile range, MS multiple sclerosis, OCT optical coherence tomography, pRNFL peripapillary retinal nerve fiber layer

During the follow-up period of at least 4 years before OCT, 67 patients (39%), experienced at least one PIRA event (Table  2 ).

Table 2 PIRA events during the time before OCT Observation time, years; median [IQR] 8.1 [5.8, 9.4] Patients with PIRA during observation time (%)  1 event 55 (32%)   ≥ 2 events 12 (7%) IQR interquartile range, OCT optical coherence tomography, PIRA progression independent of relapse activity

PIRA events during the time before OCT

IQR interquartile range, OCT optical coherence tomography, PIRA progression independent of relapse activity

Associations between OCT markers and annualized PIRA rates

In the unadjusted models, we found associations of pRNFL (beta = − 2.24, 95% CI [− 4.28; − 0.20], p  = 0.031) as well as mGCIPL (beta = − 1.89, 95% CI [− 3.16; − 0.62], p  = 0.004) with PIRA rates. In contrast, mINL thickness was not associated with PIRA (beta = − 0.35, 95% CI [− 0.76; 0.06], p  = 0.092).

In the analysis that was adjusted for multiple patient characteristics (Table  3 ), the associations of pRNFL and mGCIPL with PIRA rates remained significant. Mean pRNFL thickness decreased by 2.28 µm (95% CI [− 4.32; − 0.24], p  = 0.029; Table  3 ). and mean mGCIPL thickness decreased by 1.70 µm for each PIRA event per decade (95% CI [− 2.97; − 0.42], p  = 0.010; Table  3 ). Figure  2 summarizes the results of unadjusted and adjusted models for all OCT markers, while Fig.  3 depicts the associations of retinal layers with PIRA rates in the models that were adjusted for multiple characteristics.

Table 3 Associations of PIRA rates with the OCT markers, in linear regression models, adjusted for multiple patient characteristics pRNFL mGCIPL mINL β CI (95%) p β CI (95%) p β CI (95%) p Past PIRA rate per decade − 2.28 [− 4.32;−0.24] 0.029 − 1.70 [− 2.97; − 0.42] 0.010 − 0.33 [− 0.75;0.1] 0.133 Disease duration − 0.47 [− 0.67;−0.26]  < 0.001 − 0.31 [− 0.45;−0.17]  < 0.001 − 0.01 [− 0.06;0.03] 0.552 Age at onset 0.11 [−0.10;0.31] 0.304 − 0.04 [− 0.17;0.09] 0.541 − 0.04 [− 0.08;0.01] 0.105 Sex: Female vs. male 0.66 [− 3.27;4.59] 0.740 − 1.57 [− 4.03;0.89] 0.210 −  0.89 [ −  1.71; −  0.07] 0.034 BMI 0.36 [− 0.02;0.75] 0.065 0.17 [− 0.07;0.41] 0.162 0.05 [− 0.03;0.13] 0.255 DMT  Platform vs. untreated 2.70 [− 5.55;10.95] 0.518 2.20 [− 3.08;7.48] 0.413 − 1.27 [− 3.02;0.49] 0.157  Orals vs. untreated − 1.50 [− 7.59;4.58] 0.627 0.19 [− 3.66;4.04] 0.923 0.08 [− 1.21;1.36] 0.906  Monoclonals vs. untreated 0.19 [− 6.75;7.13] 0.958 − 0.26 [− 4.72;4.20] 0.909 − 0.23 [− 1.71;1.26] 0.765 Past annualized relapse rate 288.83 [− 969.4;1547.1] 0.651 − 32.18 [− 838.7;774.3] 0.937 26.90 [− 242.9;296.7] 0.844 R 2 adj = 0.153 R 2 adj = 0.156 R 2 adj = 0.05 Note that three models were performed, one for each OCT marker. Significant associations are marked in bold. Regarding DMT, platform summarizes interferon beta agents and glatirameracetate, orals summarize S1P receptor modulators, fumarates and teriflunomide, monoclonals summarize B-cell depleting agents and natalizumab (see Table  1 ) β estimate, BMI body mass index, CI confidence interval, DMT disease modifying treatment, mGCIPL macular ganglion cell-inner plexiform layer, mINL macular inner nuclear layer, OCT optical coherence tomography, PIRA progression independent of relapse activity, pRNFL peripapillary retinal nerve fiber layer

Associations of PIRA rates with the OCT markers, in linear regression models, adjusted for multiple patient characteristics

Sex:

Female vs. male

Note that three models were performed, one for each OCT marker. Significant associations are marked in bold. Regarding DMT, platform summarizes interferon beta agents and glatirameracetate, orals summarize S1P receptor modulators, fumarates and teriflunomide, monoclonals summarize B-cell depleting agents and natalizumab (see Table  1 )

β estimate, BMI body mass index, CI confidence interval, DMT disease modifying treatment, mGCIPL macular ganglion cell-inner plexiform layer, mINL macular inner nuclear layer, OCT optical coherence tomography, PIRA progression independent of relapse activity, pRNFL peripapillary retinal nerve fiber layer

Associations of PIRA rates with the different OCT markers in unadjusted and adjusted models. Forest plot summarizing the results of the linear regression models (unadjusted and fully adjusted for multiple covariates, as described in methods and shown in Table  3 ). For each OCT measure, the estimated coefficient of PIRA rate (each event per decade) as a point, together with the 95% CI, as a line, are shown. A dashed line at 0 is added. mGCIPL macular ganglion cell-inner plexiform layer, mINL macular inner nuclear layer, OCT optical coherence tomography, PIRA progression independent of relapse activity, pRNFL peripapillary retinal nerve fiber layer

Associations between PIRA rate and OCT markers. Note that the effect sizes of the associations are marginalized over multiple covariates: disease duration, age at onset, sex, body mass index, disease modifying treatment (dominant treatment during observation time) and annualized relapse rate during observation time. Each dot represents the raw data (i.e., combination of OCT parameter and PIRA rate) for a single individual. mGCIPL ganglion cell inner plexiform layer (in the macula). mINL inner nuclear layer (in the macula), PIRA progression independent of relapse activity, pRNFL peripapillary retinal nerve fiber layer

Sensitivity analysis with PIRMA rates

Overall, 11 patients with PIRA showed MRI activity between reference and event visits and were thus classified as not having PIRMA. Both pRNFL and GCIPL were associated with PIRMA rates per decade in the adjusted models (pRNFL: beta = − 3.70, 95% CI [− 6.23; − 1.17], p  = 0.005; mGCIPL: beta = − 2.49, 95% CI [− 4.12; − 0.87], p  = 0.003) (Supplementary Table 1). Interestingly, also mINL thickness was associated with PIRMA (beta = − 0.58, 95% CI [− 1.11; − 0.05], p  = 0.031, Supplementary Table 2).

Discussion

We investigated the relationship between retinal layers and PIRA in a well characterized MS cohort with long observation time (median of approximately 8 years). Our main finding is, that CNS-neuronal and axonal loss in the retina, as reflected by thinner mGCIPL and pRNFL, is associated with higher PIRA rates. In contrast, mINL did not show associations with PIRA. The relationship between pRNFL and mGCIPL with PIRA remained, after adjustment for multiple patient characteristics, including disease duration, DMT and annualized relapse rates.

Associations between OCT markers and PIRA were previously described in a single study [ 13 ], which reported that patients experiencing PIRA events showed higher rates of pRNFL- and mGCIPL thinning over 4 years. Our results are in line with these prior findings and confirm an association between OCT and PIRA over a longer observation time.

Interestingly, a recent multicenter study [ 29 ] found a significant association between thinning of the pRNFL and higher likelihood of relapses among patients with RRMS, while mGCIPL was not associated with clinical disease activity [ 29 ]. We found that the relationship between both OCT measures and PIRA remained after adjustment for characteristics that are related to inflammatory activity, like annualized relapse rate and DMT. In addition, mGCIPL thinning was related to subclinical (MRI) disease activity in the previous multicenter study [ 29 ], underlying the complex relationship between retinal thinning, inflammatory- and neurodegenerative processes in MS. In our sensitivity analysis, we still found significant associations between both OCT measures (pRNFL, mGCIPL) and PIRMA events. This suggests that the relationship between retinal layers and PIRA was not driven by subclinical inflammatory activity in our study. Taking into account the high reproducibility of OCT [ 9 ] and the patient-friendly, quick implementation, our findings support the role of mGCIPL and pRNFL as useful clinical markers of progression independent of disease activity in pwMS.

It has to be noted, that although both mGCIPL and pRNFL were associated with PIRA and PIRMA in our study, mGCIPL may be less prone to axonal swelling and inflammatory processes in the visual pathway and may thus be a more specific marker for neurodegeneration in MS [ 25 – 28 ]. Thus, it is probably meaningful to include macular —not only peripapillary— scans in the OCT protocol, when aiming at capturing neurodegenerative processes and PIRA. The macular scans have also the advantage of less anatomic variability than the papillary/optic-disk area [ 30 ].

Compared to the layers containing the axons (pRNFL) and CNS-neurons (GCL/GCIPL) of the retina, the role of the mINL as a marker of clinical disease activity or progression in MS is less understood. In a prior longitudinal multicenter study, mINL thickening was shown to be indicative of adjacent optic nerve inflammation, as well as to correlate with the occurrence of relapses not involving the optic nerve, but not with increase of disability over time [ 31 ]. Others showed that high mINL volume may be associated with MRI- and clinical inflammatory activity [ 24 , 31 – 33 ]. A recent multicenter study suggested that mINL thickening may occur in the early and more inflammatory stages in RRMS, followed by mINL atrophy in later stages or progressive forms of MS [ 29 ]. We did not find an association of mINL with PIRA in our study, although our patients had overall a long disease duration. However, in the sensitivity analysis taking MRI activity into account, thinner mINL was indeed associated with PIRMA rates. This suggests that subclinical inflammatory activity may affect the INL thickness in certain patients, thereby “masking” an association between mINL thinning and PIRA in the entire group. Due to this “double” role of mINL, and based on all our results, pRNFL and mGCIPL seem more suitable OCT markers than mINL to reflect progression in pwMS.

Of note, we found almost no associations between patient characteristics like sex, BMI and DMT and retinal layers in our multivariate models, although this could be due to the stronger effects of PIRA and disease duration on retinal thinning. The negative association between disease duration and retinal layers (pRNFL and mGCIPL) is in line with previous reports [ 29 , 34 ].

It is also interesting to comment on the proportion of patients with PIRA in our study: 39% experienced at least one PIRA event during 8 years of observation time, while, as expected [ 35 ], the proportion was higher in patients with progressive disease courses (78% vs. 33% in patients having officially a relapsing course). Our PIRA rates are overall slightly higher than previously reported (e.g., 25% in a study with median follow-up of approximately 7 years [ 5 ] and 28% in another study with follow-up of approximately 12 years [ 2 ]). This could be attributed to differences among the study populations or slightly modified PIRA definitions used in the different studies. Noteworthy, we applied a relatively strict definition of PIRA, not allowing any relapses between reference- and event visits. Overall, the high incidence of PIRA in our- and previous studies highlights its relevance as main driver of disability in pwMS [ 2 , 4 ].

Strengths of our study are the long clinical observation period of approximately 8 years (compared to the previous shorter study [ 14 ]), as well as the well characterized large MS cohort, including accurate assessment of disability every 6–12 months, based on the Neurostatus-EDSS definitions and information on MRI activity for most patients. Regarding limitations, this study was monocentric and included predominantly white pwMS, which may limit the generalizability of our results. We did not include healthy control data in this analysis, however previous studies already demonstrated that the rate of atrophy in retinal layers is significantly lower in healthy individuals compared to pwMS [ 36 , 37 ]. We evaluated PIRA events in time before OCT, since this significantly increased the length of observation time and the number of PIRA events. Since the assessment of disability and definition of PIRA events were performed prospectively, we do not believe that this is a major drawback. Prospective studies, using OCT to predict future risk of PIRA, including multi-ethnic population and combination with other biomarkers are needed to advance our understanding of PIRA in MS and establish the utility of different measures for patient stratification. Furthermore, prospective diagnosis of ON, according to recently published criteria, which also included the rare form of primary progressive ON [ 20 ], as a “pure” form of PIRA of the optic nerve, would contribute to our understanding of the complex relationship between the visual system and progression in pwMS.

To conclude, our findings underline the role of OCT-based neuronal- and axonal loss in the CNS (mGCIPL, pRNFL) as sensitive, patient-friendly markers of PIRA and PIRMA. Prospective larger studies are needed to define the relative value of retinal biomarkers in patient stratification and prognostic algorithms in pwMS.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary file1 (DOCX 33 KB)

Supplementary file2 (DOCX 21 KB)

Acknowledgements

We extend our gratitude to Marguerite Limberg for her exceptional efforts as an OCT technician, whose dedication greatly contributed to the success of this study. We also thank the entire SMSC Team for their invaluable support and collaboration throughout this project. Finally, we are deeply grateful to the people with MS who participated in this study, whose contributions were vital to advancing our research.

Author contributions

Federico Burguet Villena led the conceptualization of the study, as well as the writing of the original draft, including review and editing. Lisa Hofer, Sabine Schädelin, and Paskal Benkert were responsible for the methodology, data curation, and visualization. Nuria Cerdá-Fuertes, Shaumiya Sellathurai, and Kean Schoenholzer contributed to the conceptualization and data curation. Alessandro Cagol provided the data on MRI activity (data acquisition, analysis and interpretation). Marcus D’Souza, Johanna Oechtering, Henner Hanssen, Konstantin Gugleta, Cristina Granziera, Ludwig Kappos, Axel Petzold, and Jens Kuhle played a key role in validation, conceptualization, and manuscript review and editing. Athina Papadopoulou designed the study, provided supervision, validation, project administration, and contributed to drafting, reviewing and editing of the manuscript. All authors drafted the work or revised it critically for important intellectual content and all approved the submitted version of the manuscript.

Funding

Open access funding provided by University of Basel. This study was supported by the Swiss National Science Foundation (Project number PZ00P3_216468; “Ambizione” Grant to: AP) and from the Swiss Academy of Medical Sciences and the G. & J. Bangerter-Rhyner Foundation (Grant “Young talents in clinical research” to: FBV).

Data availability

Fully anonymized (encoded) data of this study could become available from the corresponding author, upon reasonable request from qualified investigators.

Declarations

Conflict of interest

FBV received travel support by TEVA and a research grant (“Young talents in clinical research”) by the Swiss Academy of Medical Sciences and of the G. & J. Bangerter-Rhyner Foundation. KS and SS are supported by the Swiss National Science Foundation (PZ00P3_216468). NCF is an employee of Neurostatus-UHB MD is CEO of Neurostatus AG and has received travel support from Bayer AG, Biogen, Teva Pharmaceuticals and Sanofi Genzyme and research support from the University Hospital of Basel. JO served on advisory boards for Roche and Merck. KG is supported by the Swiss National Science Foundation Grant No SNF212656. MD’S is an employee of the University Hospital Basel and the CEO of Neurostatus-UHB Ltd. He has received research support from the University Hospital Basel. LK has received no personal compensation. His institutions (University Hospital Basel/Foundation Clinical Neuroimmunology and Neuroscience Basel) have received and used exclusively for research support: payments for steering committee and advisory board participation, consultancy services, and participation in educational activities from: Actelion, Bayer, BMS, df-mp Molnia & Pohlmann, Celgene, Eli Lilly, EMD Serono, Genentech, Glaxo Smith Kline, Janssen, Japan Tobacco, Merck, MH Consulting, Minoryx, Novartis, F. Hoffmann-La Roche Ltd, Senda Biosciences Inc., Sanofi, Santhera, Shionogi BV, TG Therapeutics, and Wellmera, and license fees for Neurostatus-UHB products; grants from Novartis, Innosuisse, and Roche. AC: has received speaker honoraria from Novartis and Roche. CG: The University Hospital Basel (USB), as the employer of C.G., has received the following fees which were used exclusively for research support: (i) advisory board and consultancy fees from Actelion, Genzyme-Sanofi, Novartis, GeNeuro and Roche; (ii) speaker fees from Genzyme-Sanofi, Novartis, GeNeuro and Roche; (iii) research support from Siemens, GeNeuro, Roche. JK received speaker fees, research support, travel support, and/or served on advisory boards by Swiss MS Society, Swiss National Research Foundation (320030_189140/1 and 320030_212534/1), University of Basel, Progressive MS Alliance, Bayer, Biogen, Bristol Myers Squibb, Celgene, Merck, Novartis, Octave Bioscience, Roche, Sanofi. AP received speaker fees/fees for advisory boards/for consulting from Sanofi Genzyme, Eli Lilly, AbbVie, Lundbeck, Pfizer and TEVA (all used for her research in the University of Basel) and travel support from Bayer AG, Abbvie, Teva and Hoffmann-LaRoche. Her research was supported by the University- and University Hospital of Basel, the Swiss Multiple Sclerosis Society, the “Stiftung zur Förderung der gastroenterologischen und allgemeinen klinischen Forschung sowie der medizinischen Bildauswertung”, the “Freie Akademische Gesellschaft Basel” and the Swiss National Science Foundation (Project numbers: P300PB_174480 and currently: PZ00P3_216468). The other co-authors have nothing to disclose.

Ethical approval and consent to participate

The study was approved by the local ethics committee (Ethikkommission Nordwest- und Zentralschweiz (EKNZ); under the Project ID: 2023–02144) and conducted in accordance with the declaration of Helsinki. All participants gave written informed consent before inclusion in the study.

References

Associated Data

Supplementary Materials

Supplementary file1 (DOCX 33 KB)

Supplementary file2 (DOCX 21 KB)

Data Availability Statement

Fully anonymized (encoded) data of this study could become available from the corresponding author, upon reasonable request from qualified investigators.