Cardiovascular risk factors in secondary progressive multiple sclerosis: A crosssectional analysis from the <scp>MSSTAT2</scp> randomized controlled trial

Background and purpose: There is increasing evidence that cardiovascular risk (CVR) contributes to disability progression in multiple sclerosis (MS). CVR is particularly prevalent in secondary progressive MS (SPMS) and can be quantified through validated composite CVR scores. The aim was to examine the cross-sectional relationships between excess modifiable CVR, whole and regional brain atrophy on magnetic resonance imaging, and disability in patients with SPMS. Methods: Participants had SPMS, and data were collected at enrolment into the MS-STAT2 trial. Composite CVR scores were calculated using the QRISK3 software. Prematurely achieved CVR due to modifiable risk factors was expressed as QRISK3 premature CVR, derived through reference to the normative QRISK3 dataset and expressed in years. Associations were determined with multiple linear regressions.


INTRODUC TI ON
There is increasing evidence to support the importance of cardiovascular risk (CVR) factors in the pathogenic evolution of multiple sclerosis (MS).These are prevalent within the MS population, particularly in those with secondary progressive MS (SPMS), and have been reported to be associated with current and future measures of MS severity [1][2][3][4][5][6][7].
Validated composite CVR scores are more accurate in predicting an individual's overall future risk of a cardiovascular event compared to individual risk factors alone [8].The use of such composite scores may therefore be more favourable compared to examining individual risk factors alone in people with MS.
The Framingham Risk Score (FRS) is one such composite measure that has previously been investigated in people with MS [9].
Whilst that analysis did not include biochemical contributions to risk (particularly lipid ratios), the FRS was found to be higher in people with SPMS compared to relapsing-remitting MS (RRMS) and was associated with higher Expanded Disability Status Scale (EDSS) [4].A prospective follow-up of the same cohort over 4.6 years found that those with a higher baseline FRS were more likely to experience relapses, reach EDSS 6.0 or have their disease-modifying therapy escalated [10].A separate study (84% RRMS) reported that higher FRS (again quantified without lipid data) was associated cross-sectionally with smaller whole brain volume [11].Longitudinally, in participants with the highest brain volume at baseline (90th centile or above), higher FRS was associated with a faster subsequent rate of brain atrophy over 1 year [11].
A limitation of using composite CVR scores is their strong association with non-modifiable risk factors, such as age and gender [9].Within National Health Service (NHS) England, it is recommended that CVR is assessed using the QRISK3 composite score [12,13].
This model was derived from a cohort study using data from English general practitioner (GP) practices, including 7.89 million patients in the derivation cohort and 2.67 million patients in the validation cohort [14].QRISK3 quantifies the estimated risk of a cardiovascular event in the next 10 years, but importantly it also allows quantification of prematurely achieved CVR (QRISK3_PCR).This is determined through comparison of patient data to the modelled CVR score derived from age, gender and ethnicity matched healthy controls from the QRISK3 derivation cohorts.The QRISK3_PCR variable therefore provides a quantification (expressed in years) of prematurely achieved CVR attributable to modifiable risk factors, relative to age, gender and ethnicity matched data.
The aims of this study were to examine the cross-sectional relationships between a comprehensive composite score of excess modifiable CVR, using the QRISK3_PCR variable, and imaging and clinical measures of MS disease severity.A clinical trial cohort of people with SPMS was included, in whom CVR was likely to be prevalent [4].As additional exploratory outcomes, which individual components of CVR contribute most to these relationships were then investigated.

ME THODS Participants
Participants included in this analysis were all recruited into the MS-STAT2 randomized controlled trial (NCT03387670) at the lead University College London (UCL) site.All participants gave written informed consent, and MS-STAT2 was approved by the NHS national research ethics committee (London Westminster Research Ethics Committee, 09/10/2017, ref: 17/LO/1509) and conducted according to the Declaration of Helsinki [15].MS-STAT2 is an ongoing multicentre, interventional phase 3 randomized controlled trial assessing high-dose simvastatin Results: For the 218 participants, mean age was 54 years and median Expanded Disability Status Scale was 6.0.Each additional year of prematurely achieved CVR was associated with a 2.7 mL (beta coefficient; 95% confidence interval 0.8-4.7;p = 0.006) smaller normalized whole brain volume.The strongest relationship was seen for the cortical grey matter (beta coefficient 1.6 mL per year; 95% confidence interval 0.5-2.7;p = 0.003), and associations were also found with poorer verbal working memory performance.Body mass index demonstrated the strongest relationships with normalized brain volumes, whilst serum lipid ratios demonstrated strong relationships with verbal and visuospatial working memory performance.

Conclusions:
Prematurely achieved CVR is associated with lower normalized brain volumes in SPMS.Future longitudinal analyses of this clinical trial dataset will be important to determine whether CVR predicts future disease worsening.

CARDIOVASCULAR RISK IN MULTIPLE SCLEROSIS
versus placebo as a treatment for slowing the progression of disability in patients with SPMS.Eligible participants are 25-65 years of age with EDSS 4.0-6.5, with a confirmed diagnosis of SPMS and evidence of ongoing disability progression.The main exclusions included ongoing use of immunosuppressive disease-modifying therapy (with the exception of siponimod) or current use of a statin.Where a patient's screening assessments suggested significantly elevated CVR (absolute QRISK3 > 10%), they could only be enrolled if, following discussion with their GP, it was agreed that the 50% risk of being on placebo for 3 years would not compromise their general medical care.From a total of 315 participants recruited at the UCL MS-STAT2 site, 218 were additionally recruited into an optional magnetic resonance imaging (MRI) substudy and included in this analysis.

Baseline assessments
Participants underwent a systematic assessment of CVR, together with clinical evaluations of MS severity, before the first dose of trial medication.CVR measures included patient demographics, previous medical history, current medication, blood pressure, body mass index (BMI), smoking status and lipid and renal profiles.Lipid profiles included total cholesterol, low-density lipoprotein and high-density lipoprotein.In line with current recommendations on the assessments of CVR, the lipid profile was expressed as the total cholesterol to high-density lipoprotein ratio and was derived from non-fasting samples [16].Clinical assessments were performed by trained neurologists and included EDSS, timed 25

QRISK3 cardiovascular risk
QRISK3% (the predicted risk of a cardiovascular event in the next 10 years) was calculated using the online QRISK3 software platform, incorporating all of the above stated CVR parameters [12].The QRISK3% therefore represents an estimate of CVR, including modifiable and non-modifiable risk factors.QRISK3_PCR was calculated as the difference between the patient's actual age and the age at which a healthy person of the same sex and ethnicity would be predicted to achieve the same level of QRISK3%.The QRISK3 software defines a healthy person as someone with no adverse clinical indicators (smoking, diabetes etc.), a lipid ratio (total cholesterol/highdensity lipoprotein) of 4.0, a systolic blood pressure of 125 and a BMI of 25, and predicts the age at which such healthy subject would achieve a level of QRISK3% based upon data from its derivation and validation cohorts.Each negative year of QRISK3_PCR therefore represents the subject achieving a level of CVR a year earlier than would be expected through ageing alone, compared to healthy sex and ethnicity matched controls.

Magnetic resonance imaging
The 218 participants enrolled at baseline in the MRI sub-study were all imaged on a 3 T Philips Ingenia CX MR system before the first dose of trial medication.3D sagittal T1-weighted (3DT1) magnetization-prepared rapid acquisition gradient echo and 3D sagittal fluid-attenuated inversion recovery images both with 1 mm 3 isotropic voxels were included.Lesions were automatically segmented via nicMSlesions [16] and quality checked by expert clinicians.These lesion masks were then used for lesion filling on the 3DT1 scans [17,18].Geodesic Information Flows (GIFv2) was then applied to the filled 3DT1 scans to produce brain tissue segmentation and parcellation, which was used to calculate regional volumes [19].Normalized whole brain volume was calculated using SIENAX [20].

Statistical analysis
Inspection of the following variables demonstrated a skewed distribution which could be normalized through log 2 transformation: QRISK3%, T2 hyperintense lesion volume (T2LV), systolic blood pressure, lipid ratio and BMI.In all analyses, a 1 unit increase in these variables therefore represents a doubling of their value.
To assess whether our participants had achieved a level of CVR earlier than would be predicted in healthy sex and ethnicity matched controls through ageing alone, each participant's age was compared with the paired age at which QRISK3 predicted such controls would achieve the same QRISK3%.This was assessed with a paired twosample t test with unequal variance and Wilcoxon exact matched pairs signed-rank test.
The pre-specified primary outcome of this analysis was to assess the relationship between QRISK3_PCR as the predictor and normalized whole brain volume as the dependent variable.This was assessed through multiple linear regression.The following covariates were included, because of their established relationship with normalized brain volumes: T2LV, MS duration (time since symptom onset) and age.Sex was not included as a covariate in the analysis as QRISK3_PCR is derived relative to sex matched control data from the QRISK3 population.To confirm that the relationship between QRISK3 predictor variables and normalized brain volumes persisted independently of the established relationship between sex and normalized brain volumes, sensitivity analyses were conducted where normalized brain volumes were expressed as sex-specific t scores.To test for collinearity between predictors and covariates, variance inflation factors were calculated for all multivariate models, and univariate linear regression models or two-sample t tests with unequal variance were constructed to directly examine the relationship between individual predictors.It is well established that covariates can introduce substantial confounding even with a low degree of collinearity, producing unreliable estimates of coefficients [21].Where relationships between individual predictors were identified, the multivariate models were therefore repeated following the exclusion of each predictor in turn, in order to assess its effect on model coefficients.
Pre-specified secondary analyses included repeating the above models but with normalized regional brain volumes (normalized grey matter volume, cortical grey matter volume, deep grey matter volume, white matter volume) as the dependent variable.
Exploratory analyses included repeating the above models but with clinical outcomes as the dependent variable.Additional exploratory analyses then included repeating the above models including each of the individual CVR components from the QRISK3 composite scores as predictors, instead of the composite scores themselves.In these models, a reverse stepwise multiple linear regression approach was taken, excluding variables until all variables made statistically significant contributions, removal of further variables caused >1% loss of R 2 , or when removal was judged to meaningfully change the coefficients of the other remaining variables [22].In these models, ethnicity was dichotomized into white or non-white due to the small number of participants in each individual non-white category.Similarly, smoking status was dichotomized into never regularly smoked versus ever regularly smoked.In all models assessing cognitive outcomes, participant years in education was included as an additional covariate, in accordance with established practices within the literature [23,24].
Whilst the inclusion of multiple secondary and exploratory analyses will increase the family-wise type-I error rate, it does not increase the individual per-comparison-wise error rate of each test [25].The results of the secondary and exploratory analyses should not, however, be generalized beyond the precise relationships assessed [25].
For all models, assumptions were assessed through examination of the normality of residuals and residual versus fitted plots to exclude heteroscedasticity.Data were inspected for particularly influential outliers via analysis of datapoint leverage.Where evidence was identified of datapoints with high leverage or outlying residuals, data were inspected for accuracy, and sensitivity analyses were performed following the exclusion of these participants [26].Models with EDSS as the dependent variable violated assumptions on the normality of residuals.Estimates were therefore derived through bias corrected and accelerated bootstrap with 10,000 replications.

Cohort characteristics
Characteristics of the cohort are shown in Table 1, including variables contributing to CVR.

Degree of prematurely achieved cardiovascular risk
Eighty of 941 (8.5%) patients considered for the MS-STAT2 trial at UCL were excluded prior to formal screening due to already taking a statin.487 were excluded due to other eligibility issues (see Williams et al. [27]).Of the remaining 374 candidates, six (1.6%) were excluded from randomization as their GP felt their high QRISK3% meant they required treatment with a statin outside of the trial.
Despite this potentially biasing the cohort towards those with a TA B L E 1 Characteristics of the UCL MS-STAT2 MRI cohort at baseline-demographics, MS variables, QRISK3 variables and components of cardiovascular risk.

Relationship between cardiovascular risk and normalized brain volumes
The relationship between prematurely achieved CVR (QRISK3_PCR) and normalized whole and regional brain volumes is shown in Table 2.
The primary analysis revealed that each additional year of prematurely achieved CVR was associated with a 2.73 (95% CI 0.77-4.68)mL reduction in normalized whole brain volume (p = 0.006).
Secondary analyses suggested that this relationship was strongest for the cortical grey matter, where each additional year of prematurely achieved CVR was associated with a 1.60 (95% CI 0.54-2.67)mL reduction in normalized volume (p = 0.003).For comparison, the same models suggested that each doubling of T2LV was associated with an 11.1 (6.8-15.3)mL reduction in normalized cortical grey matter volume (p < 0.001; Table 2).All significant relationships persisted after normalized brain volumes were expressed after adjusting for sex (see Table S1).

TA B L E 2
Relationship between the degree of prematurely achieved cardiovascular risk and normalized whole and regional brain volumes.Note: Results are derived from five separate multiple linear regression models each including baseline data from the 218 participants in the UCL MS-STAT2 MRI sub-study.The relationship between QRISK3_PCR and nWBV was the primary analysis; all other analyses were pre-specified secondary analyses.For each model, normalized whole or regional brain volume is the dependent variable, and QRISK3_PCR is the predictor.In all models, age, MS duration (from symptom onset) and baseline log 2 T2LV are included as covariates.For each predictor and covariates, the coefficient estimate and 95% confidence interval, together with the p value, are presented, plus the R 2 value for the overall model.All volumes are reported as mm 3 .For QRISK3_PCR, a more negative value represents the patient achieving a greater degree of prematurely achieved CVR; hence the positive coefficients between QRISK3_PCR and normalized volumes represent a greater degree of prematurely achieved CVR being associated with smaller normalized volumes.

Relationship between cardiovascular risk and clinical disability
Exploratory analyses assessing the relationship between prematurely achieved CVR (QRISK3_PCR) and measures of clinical disability are shown in Table 3.There was little evidence to support any relationships between prematurely achieved CVR and measures of physical disability.
Greater prematurely achieved CVR, however, was associated with poorer verbal working memory: each additional year of prematurely achieved CVR was associated with a 0.45 (0.13-0.77) points worse CVLT-II performance (p = 0.006).A relationship of borderline significance was also present between excess prematurely achieved CVR and poorer Brief Visuospatial Memory Test performance (0.22, 0.00 to +0.44, p = 0.052).

Relationship between individual components of cardiovascular risk and both clinical and imaging measures of disease severity
Final models assessing the relationships between individual components of CVR and imaging or clinical MS severity variables are shown in Tables 4 and 5.When CVR components were included separately, they explained a greater degree of variability in normalized brain volumes compared to including the composite QRISK3 scores alone.
Beyond T2LV, age and sex, BMI and ethnicity were consistently associated with normalized brain volumes.BMI was also significantly associated with measures of lower limb physical disability (EDSS and timed 25-ft walk).Lipid ratio, in contrast, was associated with verbal and visuospatial working memory performance.

DISCUSS ION
The investigating which aspects of CVR may be the most influential, higher BMI was found to be strongly associated with lower brain volumes and a greater degree of lower limb disability, whilst higher lipid ratios, as for greater prematurely achieved CVR, were associated with poorer memory performance.
In all, 10% of potential trial participants were excluded from the UCL MS-STAT2 cohort due to pre-existing statin use or unacceptably high CVR.The overall SPMS population is therefore expected to have a higher burden of CVR than those included in this studyalthough our cohort, with a median QRISK_PCR value of −1 years (interquartile range −4 to +1 years), still maintained a significantly greater excess CVR than healthy control data from the QRISK3 population.One might therefore expect the impact of prematurely achieved CVR across the whole SPMS population to be even greater than estimated here, although this will require confirmation with further research in less selective observational cohorts.
It should be recognized that whilst the relationships described between QRISK3_PCR and MS severity variables (such as normalized cortical grey matter volume or CVLT-II) were statistically significant, the degrees of variability accounted for by excess CVR were modest.For example, in univariate analyses without other covariates, QRISK3_PCR only explained 3.5% of the variability in cortical grey matter volume.This is to be expected, however, given the established relationship between other quantifiable aspects of MS severity, such as neuroinflammatory variables (e.g., T2LV, T1gadolinium enhancing lesions, neurofilament light concentrations) and subsequent brain atrophy [28,29].In terms of magnitude, however, our data suggest that 10 additional years of prematurely achieved CVR were associated with a similar degree of cortical grey matter volume loss as a doubling of T2LV: 16.0 (95% CI 5.4-26.7)mL per 10 years of prematurely achieved CVR, compared to 11.1 (95% CI 6.8-15.3)mL for each doubling of T2LV.The finding that excess CVR may be associated with poorer working memory performance, whilst exploratory, lends support to the potential clinical significance of the relationships between prematurely achieved CVR and brain volumes.
The models including individual components of CVR explained more variability in brain volumes than the models including the composite QRISK3_PCR scores.This suggests that whilst the QRISK3derived scores have been extensively validated against future cardiovascular events, they may need to be optimized for prediction of outcomes in people with SPMS.Our exploratory analyses therefore suggest that the value of the QRISK3 score, in terms of predicting relevant MS outcomes, may be less than the sum of its parts, with higher BMI and potentially higher lipid ratios being particularly important aspects.

TA B L E 4
Final models following reverse stepwise multiple linear regression with individual cardiovascular risk components as predictors, against the dependent variable of normalized whole and regional brain volumes.Note: Results are derived from the five final separate multiple linear regression models, each with a different normalized brain volume region as the dependent variable.All cardiovascular risk variables contained within the QRISK3 composite score were initially included as predictors, and reverse stepwise removal of non-contributory variables was performed as described.In the final models, all variables were either significantly associated with the dependent variable or their removal resulted in >1% loss in R 2 or was deemed to materially affect the relationship of the remaining predictors.T2LV was included as an additional covariate in all models.
a When combined with BMI, lipid ratio was significantly associated with nWBV and nWMV, but the two variables introduced multicollinearity.BMI remained independently associated with nWBV without lipid ratio and when examined in a univariate linear regression model; lipid ratio was only significant when included with BMI, and displayed no evidence of a univariate relationship with nWBV or nWMV.Lipid ratio was therefore excluded from the final model.
No significant associations were found between CVR variables and T2LV.A previous cross-sectional observational study in a mixed cohort (60% RRMS) found that the presence of hypertension, smoking, heart disease and a BMI >25 kg/m 2 , in various combinations, was associated with higher T2LV, but predominantly in people with RRMS [30].Another longitudinal study in those with a clinically isolated syndrome found higher lipid profiles were associated with increased risk of new lesion formation [31].Other studies, however, have not replicated these relationships [3].Certainly, the absence of significant relationships between CVR variables and T2LV in our study does not refute these previous findings, and the study may be underpowered to detect modest relationships.
In healthy controls, CVR has an established association with white matter hyperintensities, presumably mediated via cerebral small vessel disease [32].One possible interpretation is therefore that whilst CVR appears to be associated with T2 lesion variables earlier in the course of disease, by the advanced disease stage included in this study (SPMS, disease duration 24 years) T2LV predominantly represents neuroinflammatory lesions, as shown in previous histological studies [33].Further studies are warranted, however, to further explore this relationship.
Interestingly, given our finding of a relationship between greater prematurely achieved CVR or higher lipid ratios and worse verbal working memory performance, one previous study of predominantly RRMS found that higher FRS is also associated with poorer CVLT-II performance [49].Whilst lipid ratios were not included, in further support of our findings, higher total cholesterol was also associated with poorer CVLT-II performance [49].This again may not be an MS-specific effect, as studies in people without MS have reported a relationship between mid-and late-life CVR, Note: Results are derived from the seven final separate multiple linear regression models, each with a different clinical outcome measure as the dependent variable.As for the MRI analysis, all cardiovascular risk variables contained within the QRISK3 composite score were initially included as predictors, and reverse stepwise removal of non-contributory variables was performed as described.In the final models, all variables were either significantly associated with the dependent variable or their removal resulted in >1% loss in R 2 or was deemed to materially affect the relationship of the remaining predictors.T2LV was included as an additional covariate in all models.
Abbreviations: 25FW, timed When combined with lipid ratio, BMI was significantly associated with BVMT-R, but the two variables introduced multicollinearity.Lipid ratio remained independently associated with BVMT-R without BMI, but BMI did not when lipid ratio was excluded.BMI was therefore excluded from the final model.
including cholesterol and smoking, with future incidence of dementia and cortical amyloid burden [50,51].Our findings, and those of others, may therefore represent an extension of these same processes in people with SPMS, or they may represent a particular additional vulnerability of people with SPMS to the cognitive impacts of such CVRs.The possibility that people with MS may be particularly vulnerable to cerebral small vessel disease has been suggested by a histology case-control study [52].
People with MS had higher rates of small vessel disease, and a stronger relationship between measures of peripheral cardiovascular disease and cerebral small vessel disease, compared to controls [52].One possible interpretation of this literature and our data is therefore that greater prematurely achieved CVR may contribute to accelerated brain atrophy and poorer memory performance due to a particular vulnerability of people with SPMS to the mechanisms of cerebral small vessel disease.This interpretation is speculative, however, and requires exploration in further research.
The main limitation of this work is the cross-sectional design.At present, it is not therefore possible to comment on causality and whether the observed relationships are relevant to ongoing changes in our participants' MS severity.A major advantage of investigating CVR within this MS-STAT2 clinical trial cohort, however, is first that longitudinal clinical and MRI follow-up is ongoing, with the 3-year dataset due to be completed in 2025.Secondly, the randomization between high-dose simvastatin and placebo will allow further investigation of causality, given that simvastatin is an established and probably pleiotropic modifier of CVR [53].

CON CLUS IONS
In this cross-sectional analysis of people with SPMS from the UCL MS-STAT2 trial cohort, a greater degree of prematurely achieved CVR was cross-sectionally associated with lower normalized brain volumes, particularly for cortical grey matter, and worse working memory performance.Raised BMI and lipid ratios appeared to be key contributors to the relationship between CVR and MS severity measures, and it is suggested that composite CVR scores may require optimizing to better predict important MS severity endpoints.Whilst this cross-sectional analysis cannot comment on causality, longitudinal follow-up of the cohort is ongoing, within which the randomization between high-dose simvastatin versus placebo will allow important future conclusions to be drawn on whether prematurely achieved CVR is a relevant modifiable target for treatment in SPMS.

ACK N O WLE D G E M ENTS
The authors thank the participants of the MS-STAT2 14681331, 2023, 9, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ene.15924by Test, Wiley Online Library on [16/10/2023].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Relationship between the degree of prematurely achieved cardiovascular risk and clinical disability measures.Results are derived from seven separate multiple linear regression models each including baseline data from the 218 participants in the UCL MS-STAT2 MRI sub-study.All models should be considered exploratory.For each model a separate clinical measure is the dependent variable as indicated, and QRISK3_PCR is the predictor.In all models, age, MS duration (from symptom onset) and baseline log 2 T2LV are included as covariates.
25r all cognitive measures, educational years was additionally included as a covariate.For each predictor and covariates, the coefficient estimate and 95% confidence interval, together with the p value, are presented, plus the R 2 value for the overall model.EDSS estimates are produced via bias corrected and accelerated bootstrap; hence p values cannot be calculated but may be inferred from the 95% confidence interval of the coefficients.Both 25FW and 9HPT are expressed as a speed (with units ft/s × 1000 and s −1 × 1000, respectively).All other variables are included as raw scores.As previously for QRISK3_PCR, a more negative value represents the patient achieving a greater degree of prematurely achieved CVR.Abbreviations: 25FW, timed25-ft walk; 9HPT, timed nine-hole peg test; BVMT-R, Brief Visual Memory Test, Revised; CVLT-II, California Verbal Learning Test II; EDSS, Expanded Disability Status Score; log 2 T2LV, log 2 of T2-weighted lesion volume; MRI, magnetic resonance imaging; MS, multiple sclerosis; QRISK3_PCR, QRISK3 premature cardiovascular risk; SDMT, Symbol Digit Modalities Test; SLCLA 2.5%, Sloan Low Contrast Letter Acuity 2.5%; UCL, University College London.
Final models following reverse stepwise linear regression with individual cardiovascular risk components as predictors, against the dependent variables of MS clinical and cognitive disability.
This project was funded by the MS-STAT2 trial (NCT03387670), which is an investigator-led project sponsored by University College London and funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment, UK Multiple Sclerosis Society, US National Multiple Sclerosis Society and the Rosetrees Trust.Thomas Williams has received honoraria for educational talks from Novartis and Merck; he is funded by the MS-STAT2 trial grant.Nevin John is a local principal investigator for trials in multiple sclerosis funded by Sanofi, Novartis and Biogen Idec.Alberto Calvi is supported by the ECTRIMS postdoc fellowship (2022), previously received a UK MS Society PhD studentship (2020), a Guarantors of Brain 'Entry' clinical fellowship (2019) and an ECTRIMS-MAGNIMS fellowship (2018).Alessia Bianchi has received a research grant from the Italian Society of Neurology.Floriana De Angelis has received speaker honoraria from Neurology Academy, Janssen, Merck, Novartis, Sanofi and served on an advisory board for Novartis.She received congress fees from Janssen, Novartis and Roche.She is regional coordinator for the Oratorio Hand Trial (Hoffmann-La Roche) and PI of commercial and academic trials including CHARIOT-MS, ALITHIOS (Novartis), O'HAND (Roche).Madiha Shatila was supported by the MENACTRIMS research grant 2021.Ferran Prados received a Guarantors of Brain fellowship 2017-2020 and is supported by NIHR, Biomedical Research Centre initiative at University College London Hospitals (UCLH).David MacManus is a share-holder in Queen Square Analytics Ltd.Anisha Doshi, Sarah Wright, Marios Yiannakas, Fatima Chowdhury, Jon Stutters, Antonio Ricciardi, Marie Braisher, James Blackstone: nothing to disclose.Olga Ciccarelli received research funding from the NIHR Biomedical Research Centre initiative at UCLH, UK and National MS Societies, and Rosetrees Trust; serves as consultant for Novartis, Roche and Teva; and is an Associate Editor for Neurology®.Gandini Wheeler-Kingshott receives research funding from the MS Society (77), Wings for Life (169111), Horizon2020 (CDS-QUAMRI, 634541), BRC (RC704/CAP/CGW), UCL Global Challenges Research Fund (GCRF) and MRC (MR/S026088/1); and is a share-holder in Queen Square Analytics Ltd.Frederik Barkhof is supported by the NIHR Biomedical Research Centre at UCLH; serves on the editorial boards of Brain, European Radiology, Journal of Neurology Neurosurgery and Psychiatry, Neurology, Multiple Sclerosis and Neuroradiology; steering committee or iDMC member for Biogen, Merck, Roche, EISAI and Prothena; consultant for Roche, Biogen, Merck, IXICO, Jansen, Combinostics; research agreements with Merck, Biogen, GE Healthcare, Roche; co-founder and shareholder of Queen Square Analytics Ltd.Jeremy Chataway: In the last 3 years, JC has received support from the Health Technology Assessment (HTA) Programme (NIHR), the UK MS Society, the US National MS Society and the Rosetrees Trust.He is supported in part by the NIHR UCLH Biomedical Research Centre, London, UK.He has been a local principal investigator for a trial in MS funded by the Canadian MS Society; a local principal investigator for commercial trials funded by Ionis, Novartis and Roche; and has taken 14681331, 2023, 9, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ene.15924by Test, Wiley Online Library on [16/10/2023].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License trial, together with their families and carers; the wider MS-STAT2 investigator group (see Appendix); and the National Institute for Health and Care Research Biomedical Research Centre at University College London Hospitals for their support.CO N FLI C T O F I NTER E S T S TATEM ENT