Neurodegeneration is the pathological substrate that causes major disability in secondary progressive multiple sclerosis. A synthesis of preclinical and clinical research identified three neuroprotective drugs acting on different axonal pathobiologies. We aimed to test the efficacy of these drugs in an efficient manner with respect to time, cost, and patient resource.
Between Jan 29, 2015, and June 22, 2016, 445 patients were randomly allocated amiloride (n=111), fluoxetine (n=111), riluzole (n=111), or placebo (n=112). The primary analysis included 393 patients who were allocated amiloride (n=99), fluoxetine (n=96), riluzole (n=99), and placebo (n=99). No difference was noted between any active treatment and placebo in PBVC (amiloride vs placebo, 0·0% [95% CI −0·4 to 0·5; p=0·99]; fluoxetine vs placebo −0·1% [–0·5 to 0·3; p=0·86]; riluzole vs placebo −0·1% [–0·6 to 0·3; p=0·77]). No emergent safety issues were reported. The incidence of serious adverse events was low and similar across study groups (ten [9%] patients in the amiloride group, seven [6%] in the fluoxetine group, 12 [11%] in the riluzole group, and 13 [12%] in the placebo group). The most common serious adverse events were infections and infestations. Three patients died during the study, from causes judged unrelated to active treatment; one patient assigned amiloride died from metastatic lung cancer, one patient assigned riluzole died from ischaemic heart disease and coronary artery thrombosis, and one patient assigned fluoxetine had a sudden death (primary cause) with multiple sclerosis and obesity listed as secondary causes.
The absence of evidence for neuroprotection in this adequately powered trial indicates that exclusively targeting these aspects of axonal pathobiology in patients with secondary progressive multiple sclerosis is insufficient to mitigate neuroaxonal loss. These findings argue for investigation of different mechanistic targets and future consideration of combination treatment trials. This trial provides a template for future simultaneous testing of multiple disease-modifying medicines in neurological medicine.
Efficacy and Mechanism Evaluation (EME) Programme, an MRC and NIHR partnership, UK Multiple Sclerosis Society, and US National Multiple Sclerosis Society.
Progressive multiple sclerosis is the major cause of disease-associated costs, both to individuals and health care systems
and, therefore, it is a key target for therapeutic development. However, by contrast with the range of treatments that mitigate inflammatory activity in relapsing-remitting multiple sclerosis, treatments that can slow, stop, or reverse progressive multiple sclerosis are limited.
Evidence before this study
We have previously published a systematic review and synthesis of available evidence for candidate oral neuroprotective drugs tested in clinical trials from patients with multiple sclerosis, dementia, motor neuron disease, Huntington’s disease, and Parkinson’s disease, combined with in-vivo data from experimental autoimmune encephalomyelitis (EAE) studies. We did two further searches in Ovid MEDLINE and Epub Ahead of Print, In-Process & Other Non-Indexed Citations and Daily (from 1946 to Feb 27, 2019), OVID Embase (from 1980 to 2019 week 8), the Cochrane Database of Systematic Reviews, and the Cochrane Central Register of Controlled Trials (CENTRAL). In the first search, we used a combination of keywords and database-appropriate subject headings for the trial drugs: “amiloride” OR “fluoxetine” OR “riluzole” AND “multiple sclerosis” OR “experimental allergic encephalomyelitis” OR “EAE”. We excluded symptomatic human studies. We did not restrict our search by language. In the second search, we used a combination of keywords and database-appropriate subject headings for neurodegenerative neurological diseases, including “multiple sclerosis” OR “Parkinson’s disease” OR “amyotrophic lateral sclerosis-motor neuron disease” OR “Huntingdon’s disease” OR “dementia”, combined with terms to retrieve multiarm drug trials and the Cochrane Highly Sensitive Search Strategy for identifying randomised trials in MEDLINE (sensitivity and precision-maximising version, 2008 revision). We excluded symptomatic, dose-ranging, non-drug trials or studies of relapsing-remitting multiple sclerosis. We did not restrict our search by language. The first search retrieved five experimental studies for amiloride, indicating potential neuroprotection in animal models and human pathological samples. The most likely mechanism was blockage of ASIC1. In a clinical trial of primary progressive multiple sclerosis (n=14), a significant reduction was noted in the rate of whole-brain volume change with amiloride as well as improvements in deep grey and white matter tract diffusion indices. In three EAE studies, fluoxetine partly ameliorated paralysis and reduced inflammatory foci. Findings of two MR spectroscopy studies in humans showed an increase in N-acetylaspartate variables with fluoxetine, suggesting improved neuronal energetics and microstructural integrity by diffusion MRI, although a third study showed no change in markers of phosphocreatine metabolism. In a small placebo-controlled study, patients with relapsing multiple sclerosis treated with fluoxetine showed a trend towards a reduction in the number of new enhancing lesions over time. In a pilot study in progressive multiple sclerosis (n=42), non-significant benefits were seen in some markers of clinical progression with fluoxetine, although the study was underpowered. Riluzole reduced the severity of inflammation, demyelination, and axonal damage in a myelin oligodendrocyte glycoprotein-induced EAE system. In a run-in study in 16 patients with primary progressive multiple sclerosis, a reduction was noted in the rate of cervical atrophy and new T1 hypointense lesions with riluzole. In another study, no effect was seen on the atrophy rate in early multiple sclerosis with riluzole. The second search for multiarm drug trials in neurodegenerative diseases retrieved three studies. The first study in patients with progressive multiple sclerosis (n=58) assessed cyclophosphamide, adrenocorticotropic hormone, and plasma exchange. In the second study, 782 patients with early Parkinson’s disease were randomly allocated to approved treatments (levodopa and DOPA decarboxylase inhibitor alone; levodopa, DOPA decarboxylase inhibitor, and selegiline; or bromocriptine). The third study was a three-arm trial of tianzhi granule, donepezil, or placebo in improving functional ability in vascular dementia. No other multiarm trials were retrieved by our search.
Added value of this study
In the MS-SMART trial, we chose three agents (amiloride, fluoxetine, and riluzole) to target different mechanistic pathways in patients with secondary progressive multiple sclerosis. We used a multiarm design to enable simultaneous assessment of these drugs at the important phase 2b decision point and, therefore, to accelerate the drug discovery process. The study design was robust with appropriate performance characteristics. We did not find any evidence of a neuroprotective effect of the three agents, despite positive early work in animals and humans.
Implications of all the available evidence
Multiarm trials are feasible and efficient in neurodegenerative diseases. They have the potential to examine promising experimental and early-phase agents in a timely fashion. The results of our study have implications for future experimental paradigms. This style of approach is necessary to accelerate treatment discovery in an area in which limited progress has been made.
In parallel, evidence has been accumulating for a range of candidate therapies that target key molecular processes in the axon itself, representing a more direct or downstream approach to achieve neuroprotection. Over the past three decades, many negative trials have been published, with the findings attributable in part to suboptimum trial design.
However, the fundamental question of whether direct targeting of axonal pathobiological features can be an effective therapeutic strategy remains unanswered.
This approach resulted in a shortlist of seven candidate drugs with different mechanisms of action that might prove beneficial in progressive multiple sclerosis: ibudilast, oxcarbazepine, pirfenidone, polyunsaturated fatty acids (including lipoic acid), amiloride, fluoxetine, and riluzole. This strategy was ultimately confirmed by positive phase 2 results for two of the drugs, ibudilast
and lipoic acid.
we aimed to test the efficacy of targeting axonal pathobiological features as a strategy to achieve neuroprotection in progressive multiple sclerosis. To maximise efficiency of testing and accelerate drug development, we used a multiarm approach that enabled simultaneous assessment of several drugs at the important phase 2b decision point, a strategy that has been used successfully in oncology.
Based on efficacy and drug supply, we selected three drugs from our candidate shortlist (amiloride, fluoxetine, and riluzole) for assessment against placebo.
ASIC1 opens in response to inflammation-induced acidosis, causing sodium and calcium influxes.
This action is associated with axonal injury in post-mortem studies of patients with acute multiple sclerosis,
and blockade of ASIC1 with amiloride reduces axonal damage and improves clinical outcomes in rodent models.
A pilot study in individuals with progressive multiple sclerosis showed a significant reduction in whole-brain atrophy.
In an underpowered negative trial, fluoxetine showed non-significant benefits in some markers of clinical progression.
Glutamate excitotoxicity results in neuronal injury
and its blockade in EAE reduces clinical impairment and axonal damage.
In a pilot study of riluzole in people with progressive multiple sclerosis, a reduction in the rate of cervical cord atrophy and the number of new brain T1 hypointense lesions was recorded,
although findings of another study in individuals with early relapsing-remitting multiple sclerosis or clinical isolated syndrome did not show a reduction in the rate of atrophy.
- Connick P
- De Angelis F
- Parker RA
- et al.
Our aim was to efficiently test the efficacy of targeting axonal pathobiology as a strategy to achieve neuroprotection in progressive multiple sclerosis.
Study design and participants
- Connick P
- De Angelis F
- Parker RA
- et al.
and in the appendix (pp 2–3).
The study was done in accordance with the Declaration of Helsinki and International Council for Harmonisation Good Clinical Practice guidelines. Independent ethics approval for the protocol was granted by REC 13/SS/0007, and all patients provided written informed consent before entering the study. Safety oversight was the responsibility of the Data Monitoring Committee, which reviewed accruing participant-level data every 6 months. Individual site medical monitoring was also mandated.
Randomisation and masking
Within 30 days after screening for enrolment to the study, we randomly allocated patients (1:1:1:1) at baseline either amiloride, fluoxetine, riluzole, or placebo. A research nurse used a centralised web-based service provided by the Edinburgh Clinical Trials Unit (Usher Institute, University of Edinburgh, Edinburgh, UK) to randomly assign interventions, with minimisation by sex, age (<45 years vs ≥45 years), EDSS score at randomisation (4·0–5·5 vs 6·0–6·5), and trial site. The minimisation procedure incorporated a random element whereby the assigned treatment was switched with a probability of 10% from the group that would give greatest balance to one of the other three study groups (with a probability of 3·33% for each of the other study groups). Amiloride, fluoxetine, riluzole, and placebo capsules were over-encapsulated and identical in appearance. Patients and investigators, including MRI analysts, treating clinicians, and independent assessing neurologists were unaware of treatment allocations and had no access to randomisation codes. We asked patients and clinicians to complete a questionnaire at week 96 to assess the validity of the masking procedures.
We initially administered assigned treatments orally once daily, from baseline (week 0) for 4 weeks, then patients received doses twice daily from week 4 until week 96. Doses were amiloride hydrochloride 5 mg, fluoxetine 20 mg, riluzole 50 mg, or matching placebo. After the baseline visit, patients were seen at weeks 4, 8, 12, 24, 36, 48, 72, and 96, with a final safety telephone call at week 100. Brain MRI was done at screening for enrolment, week 24, and week 96. Neurological assessments were done at screening, baseline, week 48, and week 96. Safety blood tests were assessed at every study visit and included full blood count, electrolytes, and liver and renal function tests. Study treatment was discontinued if confirmed repeat measurements showed the following blood test concentrations: potassium less than 2·8 mmol/L or more than 5·5 mmol/L; sodium less than 125 mmol/L; alanine aminotransferase, aspartate aminotransferase, or γ-glutamyl transferase more than five times the upper limit of normal; creatinine more than 130 μmol/L; neutrophil count less than 1·0 × 109 cells per L, platelet count less than 50 × 109 cells per L; or haemoglobin less than 80 g/L. Adverse events were assessed at every study visit.
SIENA is an automated method that registers the follow-up scan to the baseline scan and produces an integral of the edge motion occurring in each voxel between scans and directly calculates the PBVC from those values.
- Connick P
- De Angelis F
- Parker RA
- et al.
MRI data were analysed independently at a central reading site (Queen Square Multiple Sclerosis Centre, University College London, London, UK). Clinical secondary endpoints were changes from baseline to weeks 48 and 96 in EDSS score, the Timed 25-Foot Walk, the 9-Hole Peg Test, the Paced Auditory Serial Addition Test, the Multiple Sclerosis Functional Composite score, the Symbol Digit Modalities Test, high contrast (100%) visual acuity, and Sloan low contrast visual acuity (contrast 5%, 2·5%, and 1·25%).
Time to first relapse was recorded. Patient-reported outcomes were also measured at baseline, week 48, and week 96 using: the Multiple Sclerosis Impact Scale 29 items version 2; the Multiple Sclerosis Walking Scale version 2; the Neurological Fatigue Index; and health-related quality-of-life (measured with EuroQol five dimensions five levels [EQ-5D-5L]). Neuropathic pain scores were also obtained and will be reported separately. The Multiple Sclerosis Functional Composite Z-score was normalised (signed square-root transformed) using participants’ baseline scores.
The percentage of expected total cohort dropouts was based on two phase 2 studies in secondary progressive multiple sclerosis,
from which we calculated that including 110 patients per study group would provide 90% statistical power in the analysis of covariance to detect a 40% reduction in PBVC versus placebo, allowing for 10% dropouts plus 10% of participants discontinuing treatment while remaining in follow-up.
to allow for multiple pairwise comparisons to a common control group and maintain the overall family-wise error rate below 5%. Dunnett-adjusted p values are reported for the primary outcome analyses.
The primary analysis was a complete-case analysis based on the intention-to-treat population. The intention-to-treat population included all patients in the MS-SMART trial who underwent randomisation and had data available at week 96.
The effects of missing data or outliers on the primary outcome findings were investigated by doing three sensitivity analyses. These entailed excluding outliers more than 4 SD away from the mean, imputing missing data using a standard multiple imputation method, and imputing missing data under a missing-not-at-random assumption, whereby a constant value was added to the values imputed using multiple imputation equal to the observed SD in the primary outcome at 96 weeks.
A further primary outcome analysis was done in the per-protocol population, which included participants who were adherent to the protocol and compliant with the originally assigned treatment throughout the duration of follow-up. Patients were judged compliant with the assigned treatment if they reported taking, on average, 90% or more of their prescribed medication (taking account of planned down-titrations and deferred up-titrations) in the 30 days preceding every clinic visit.
Secondary outcome analyses were not adjusted for multiple comparisons and used 5% as the nominal significance level. The number of new or enlarging T2 lesions detected at the 96-week MRI scan was compared between each of the three active treatment groups and placebo by means of an overdispersed Poisson regression model fitted to the number of new or enlarging T2 lesions at 96 weeks, with adjustment for minimisation variables.
For continuous or ordinal outcomes measured at 96 weeks (ie, the EDSS score, the 9-Hole Peg Test, the Paced Auditory Serial Addition Test, the Multiple Sclerosis Functional Composite score, the Symbol Digit Modalities Test, Sloan low contrast visual acuity, the Multiple Sclerosis Impact Scale 29 items version 2, the Multiple Sclerosis Walking Scale version 2, the Neurological Fatigue Index, and health-related quality of life), we used a multiple linear regression method to calculate adjusted mean differences and 95% CIs for the individual comparisons between each active treatment and placebo. Regression models were adjusted for baseline and the minimisation variables.
For the EDSS outcome only, 95% CIs were calculated using a bootstrap method due to the ordinal nature of the outcome variable. Cox proportional-hazard models adjusting for the minimisation variables were used for time-to-first relapse and the Timed 25-Foot Walk test at 96 weeks. Analyses were done using SAS version 9.4 (SAS Institute, Cary, NC, USA).
Role of the funding source
The funder had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all data in the study and had final responsibility for the decision to submit for publication.
Table 1Baseline characteristics of randomised participants
Data presented are n (%), mean (SD), or median (IQR). No data were missing for the 9-Hole Peg Test, Multiple Sclerosis Impact Scale 29 items version 2, and EQ-5D-5L (VAS score); for the remaining variables, varying amounts of data were missing up to a maximum of 12 patients overall (maximum five for amiloride, two for fluoxetine, three for riluzole, and four for placebo). EQ-5D-5L=EuroQol five dimensions five levels. VAS=visual analogue scale.
Table 2Primary outcome analysis of PBVC at 96 weeks, intention-to-treat population
Data are mean (SD) unless otherwise stated. PBVC=percentage brain volume change.
At 96 weeks, the mean number of new or enlarging T2 lesions in patients assigned placebo was 3·0 (SD 6·9; median 0 [IQR 0–2]). Compared with placebo, a similar number of new or enlarging T2 lesions were detected at the 96-week MRI scan for amiloride (rate ratio [RR] 1·2, 95% CI 0·8–1·8; p=0·29) and for riluzole (1·0, 0·6–1·5; p=0·81). Fluoxetine had fewer new or enlarging T2 lesions compared with placebo (RR 0·5, 95% CI 0·3–0·9; p=0·012).
Table 3Secondary outcomes at 96 weeks
Data are mean (SD). Results are derived from a model analysing data for 393 participants who had at least some 96-week outcome data. Numbers of patients are the maximum per group; the minimum sample size was n=380 for each secondary outcome (minimum per group: amiloride, n=98; fluoxetine, n=94; riluzole, n=92; placebo, n=94). The multiple regression model for each outcome included randomised treatment as an explanatory factor variable (with placebo as the reference category), the baseline measurement, and minimisation variables (age, sex, treatment centre, and Expanded Disability Status Scale score at baseline). EQ-5D-5L=EuroQol five dimensions five levels. VAS=visual analogue scale.
51 (11%) of 445 patients had at least one relapse during the study, with 16 (14%) having a relapse in the amiloride group, ten (9%) in the fluoxetine group, 11 (10%) in the riluzole group, and 14 (12%) in the placebo group. Compared with placebo, time to first relapse did not differ for amiloride (hazard ratio [HR] 1·14, 95% CI 0·56–2·35), fluoxetine (0·74, 0·33–1·66), or riluzole (0·78, 0·35–1·73). Similarly, compared with placebo, the Timed 25-Foot Walk test did not differ for amiloride (HR 0·82, 95% CI 0·61–1·12), fluoxetine (0·81, 0·59–1·10), or riluzole (0·84, 0·61–1·13).
Table 4Adverse events, safety population
The findings of MS-SMART, a large, multicentre, multiarm randomised trial in patients with secondary progressive multiple sclerosis, showed that none of the three study drugs (amiloride, fluoxetine, and riluzole) had any effect on the primary outcome of PBVC over 96 weeks or any secondary outcomes, compared with placebo. Significant efficiency gains with respect to the study design were, however, seen in terms of people, time, and economic resources. The study cohort was typical of patients with non-relapsing secondary progressive multiple sclerosis, with a median disease duration of 21 years (IQR 15–29), median secondary progression of 6 years (3–10), use of unilateral support when walking (median EDSS score 6·0 [5·5–6·5]), and more than 90% of patients were relapse-free for at least 2 years before recruitment. High levels of retention (>90%) and adherence were reported, and on-trial behaviour was also typical of patients with non-relapsing secondary progressive multiple sclerosis, with low levels of inflammatory disease activity clinically (11% of patients had an on-study relapse) and radiologically (mean number of new or enlarging lesions was 3·0 [SD 6·9], median 0 [IQR 0–2]). Clear radiological progression was seen, with PBVC of roughly −0·7% per year.
Taken together with the effectiveness of our trial masking and the meeting of all previous assumptions underlying our power calculations, MS-SMART was sensitive to detect any neuroprotective effect of the three drugs tested. Some data were missing, 12% for the primary endpoint and as high as 15% for some secondary endpoints; although this shortfall brings some potential for bias, sensitivity analyses were supportive of the findings of the primary analysis.
first, we looked at preclinical studies of demyelination, inflammation, axonal loss, and neurobehavioural changes in EAE models of multiple sclerosis; second, we assessed clinical evidence at phase 2a in patients with multiple sclerosis and other neurodegenerative disorders that share common pathways of neurodegeneration. We did not select agents with clinically significant safety profile issues, those most likely to produce only symptomatic benefit, drugs with an immunosuppressive mechanism of action, agents with limited efficacy data or biological plausibility, and those that had been assessed previously in patients with relapsing-remitting multiple sclerosis.
Pharmacokinetic considerations (eg, limited bioavailability in the CNS) are also of potential relevance to the results of our trial. However, riluzole and fluoxetine have known CNS effects at the doses used and are established drugs for treatment of motor neuron disease and depression, respectively. Amiloride has a primary non-CNS target; however, the dose used was based on an earlier phase 2a trial.
Although all these processes represent biologically plausible targets for neuroprotection, many additional pathological processes can be implicated in disease progression, including innate or adaptive mediated inflammation and inadequate remyelination. Thus, several explanations could account for why amiloride, fluoxetine, and riluzole did not show efficacy in our study, including (but not restricted to) the ultimate relevance of the dynamic in-vivo disease processes primarily targeted by these three drugs. Indeed, our findings expose our incomplete knowledge of the pathobiology of secondary progressive multiple sclerosis.
In view of positive results from the ocrelizumab
phase 3 trials, future work will most likely stratify use of anti-inflammatory compounds in conjunction with possible neuroprotective agents.
The findings that fluoxetine was associated not only with a significantly higher PBVC at 24 weeks compared with placebo but also a concomitant decrease in the number of new or enlarging T2 lesions at 96 weeks should be interpreted with caution. These outcomes were secondary outcomes and the number of participants developing new or enlarging T2 lesion was, overall, low (mean 3·0 [SD 6·9] lesions; median 0 [IQR 0–2]).
in both immunomodulatory (eg, natalizumab)
and sodium-channel blockade (eg, lamotrigine)
studies. Fluoxetine might, therefore, exert a degree of immunomodulatory activity (eg, decreased lymphocyte proliferation and suppressed interferon-γ production) by acting on astrocytes;
findings of a small study showed reduction of new enhancing lesions in patients with relapsing-remitting multiple sclerosis.
However, the primary outcome of our study at week 96 was negative and accords with the lack of therapeutic effect seen in the FLUOX-PMS trial of 137 patients with progressive multiple sclerosis, which also did not show a beneficial effect on brain atrophy in a subgroup.
The MS-SMART trial was restricted to patients with secondary progressive multiple sclerosis, with inclusion criteria as described. Generalisation to a wider population (eg, greater disability, older, or with primary progressive disease) is uncertain. Nonetheless, overall, we feel our study population will have mostly captured the secondary progressive clinical phenotype.
In conclusion, we have shown that a multiarm approach can be used successfully to expedite drug discovery in patients with progressive multiple sclerosis. Such trial designs will be highly relevant to future therapeutic development in brain medicine in general. They are necessary to confirm or refute postulated pathways and give insight into the pathobiology of progressive multiple sclerosis.
Our results do not support the effectiveness of amiloride, fluoxetine, and riluzole in reducing disease progression for secondary progressive multiple sclerosis, and they indicate that exclusive targeting of axonal pathobiology is an inadequate strategy to achieve neuroprotection in progressive disease. In view of the need to develop disease-modifying treatments for progressive multiple sclerosis, this finding challenges the area of research to shift towards combinatorial strategies or stratification based on greater resolution of relevant pathobiology at the level of individual patients.
Jeremy Chataway, Claudia A M Gandini Wheeler-Kingshott, Floriana De Angelis, Domenico Plantone, Anisha Doshi, Nevin John, Thomas Williams, Jonathan Stutters, Ferran Prados Carrasco, David MacManus, Frederik Barkhof, Sebastien Ourselin, Marie Braisher, Tiggy Beyene, Vanessa Bassan, Alvin Zapata (Queen Square Multiple Sclerosis Centre, University College London and University College London Hospitals NHS Foundation Trust, London, UK); Siddharthan Chandran, Peter Connick, Dawn Lyle, James Cameron, Daisy Mollison, Shuna Colville, Baljean Dhillon (Anne Rowling Regenerative Neurology Clinic, The University of Edinburgh, Royal Infirmary of Edinburgh, NHS Lothian, Edinburgh, UK); Christopher J Weir, Richard A Parker, Moira Ross, Gina Cranswick, Allan Walker, Lorraine Smith (Edinburgh Clinical Trials Unit [ECTU], Usher Institute, University of Edinburgh, Edinburgh, UK); Gavin Giovannoni, Sharmilee Gnanapavan (Blizard Institute, Barts and The London School of Medicine and Dentistry, Queen Mary University, Barts Health NHS Trust, London, UK); Richard Nicholas (Imperial College Healthcare NHS Trust, London, UK); Waqar Rashid, Julia Aram (Brighton and Sussex University Hospitals NHS Trust, Brighton, UK); Helen Ford (Leeds General Infirmary, Leeds Teaching Hospitals NHS Trust, Leeds, UK); Sue H Pavitt (Dental Translational and Clinical Research Unit, University of Leeds, Leeds, UK); James Overell (The Queen Elizabeth University Hospital Glasgow, NHS Greater Glasgow and Clyde, Glasgow, UK); Carolyn Young, Heinke Arndt (The Walton Centre NHS Foundation Trust, Liverpool, UK); Martin Duddy, Joe Guadagno (Royal Victoria Infirmary, The Newcastle upon Tyne Hospital NHS Foundation Trust, Newcastle, UK); Nikolaos Evangelou (Queens Medical Centre, Nottingham University Hospital NHS Trust, Nottingham, UK); Matthew Craner, Jacqueline Palace (John Radcliffe Hospital, Oxford University Hospitals NHS Foundation Trust, Oxford, UK); Jeremy Hobart (Derriford Hospital, University Hospitals Plymouth NHS Trust, Plymouth, UK); Basil Sharrack, David Paling (Royal Hallamshire Hospital, Sheffield Teaching Hospitals NHS Foundation Trust, Sheffield, UK); Clive Hawkins, Seema Kalra (Royal Stoke University Hospital, University Hospitals of North Midlands NHS Trust, Stoke-on-Trent, UK); Brendan McLean (Royal Cornwall Hospitals NHS Trust, Truro, UK); Nigel Stallard (Statistics and Epidemiology, Division of Health Sciences, Warwick Medical School, University of Warwick, Coventry, UK); and Roger Bastow (patient representative).
JC had the idea for the study. JC, SC, CJW, SHP, NS, CAGW-K, CH, GG, BS, and GC designed the study. JC, FDA, PC, DP, AD, and NJ obtained data. MR did data management. MB coordinated the study at the London site. DMcM, JS, FDA, and FB did MRI quality control, post-processing, analysis, and oversight. FPC, SO, AD, and NJ contributed to MRI analysis. RAP and CJW were the trial statisticians and did all statistical analyses. RB was the patient representative. JC, FDA, RAP, and CJW wrote the first draft of the report. All authors contributed to writing, interpretation, critical review of data, and approved the final version.
Declaration of interests
MB has received funding from the UK Multiple Sclerosis (MS) Society and the UK National Institute of Health Research (NIHR) Local Clinical Research Network. FB is a consultant for Bayer, Biogen-Idec, Teva, Merck, Novartis, Roche, Apitope, IXICO, and Lundbeck. CAGW-K has received research grants as principal investigator and co-applicant from Spinal Research, Craig H Neilsen Foundation, EPSRC, Wings for Life, UK MS Society, Horizon2020, and the NIHR/Medical Research Council. GG reports personal fees from AbbVie, GW Pharma, Fiveprime, Synthon BV, Eisai, Elan, Genentech, GlaxoSmithKline, and Pfizer; grants and personal fees from Biogen, Canbex, Merck-Serono, Novartis, Teva, Roche, Bayer-Schering, Ironwood, and Genzyme/Sanofi; and grants from UCB Pharma. SHP has received grant income from the NIHR, UK MS Society, and the NIHR Local Clinical Research Network; is Director of Dental Translation and Clinical Research Unit (part of the NIHR Leeds Clinical Research Facility); serves on the NIHR Clinical Trials Unit Standing Advisory Board; and is a former NIHR Efficacy and Mechanism Evaluation (EME) board member. JC has received support from the EME Programme and Health Technology Assessment Programme (NIHR), the UK MS Society, and the US National MS Society. In the past 3 years, JC has been a local principal investigator for trials in MS funded by Receptos, Novartis, Roche, and Biogen Idec; has received an investigator grant from Novartis; and has taken part in advisory boards and consultancy for Roche, Merck, MedDay, Biogen, and Celgene. SC declares collaborative grants through the University of Edinburgh with Biogen; and consultancy through the University of Edinburgh with Novartis, Merck, and Roche. FDA, PC, RAP, DP, AD, NJ, JS, DMcM, FPC, SO, BS, MR, GC, RB, CJW, NS, and CH declare no competing interests.
MS-SMART is an investigator-led project sponsored by University College London (UCL). This project (reference 11/30/11) is funded by the Efficacy and Mechanism Evaluation (EME) Programme, a Medical Research Council (MRC) and National Institute for Health Research (NIHR) partnership. The views expressed in this publication are those of the author(s) and not necessarily those of the MRC, NIHR, or the Department of Health and Social Care. This research is also funded by the UK Multiple Sclerosis (MS) Society and the US National MS Society. Additional support came from the University of Edinburgh, the NIHR-UCL Hospitals Biomedical Research Centre and UCL; and the NIHR Leeds Clinical Research Facility (Dental Translation and Clinical Research Unit). CJW and RAP were supported in this work by the UK National Health Service Lothian via the Edinburgh Clinical Trials Unit. FPC has a non-clinical Guarantor of Brain fellowship. Riluzole was provided without charge by Sanofi-Genzyme who was not involved in either the trial design, running of the trial or analysis. We thank David Miller, Lorraine Smith, Marios Yiannakas, Marcello Moccia, Almudena Garcia, Alberto Calvi, Tim Friede, Michael Hutchinson, Stanley Hawkins, Robert Swingler, John Thorpe, John O’Brien, and Allan Walker; we acknowledge the support of Alan Thompson; and we thank those people with multiple sclerosis who took part in this trial.
Progressive multiple sclerosis: prospects for disease therapy, repair, and restoration of function.
Lancet. 2017; 389: 1357-1366
Cost of illness of multiple sclerosis: a systematic review.
PLoS One. 2016; 11e0159129
Siponimod versus placebo in secondary progressive multiple sclerosis (EXPAND): a double-blind, randomised, phase 3 study.
Lancet. 2018; 391: 1263-1273
Ocrelizumab versus placebo in primary progressive multiple sclerosis.
N Engl J Med. 2017; 376: 209-220
Clinical trials in progressive multiple sclerosis: lessons learned and future perspectives.
Lancet Neurol. 2015; 14: 208-223
Drug repurposing: a systematic approach to evaluate candidate oral neuroprotective interventions for secondary progressive multiple sclerosis.
PLoS One. 2015; 10: 1-18
Phase 2 trial of ibudilast in progressive multiple sclerosis.
N Engl J Med. 2018; 379: 846-855
Lipoic acid in secondary progressive MS: a randomized controlled pilot trial.
Neurol Neuroimmunol Neuroinflamm. 2017; 4: e374
Celecoxib plus hormone therapy versus hormone therapy alone for hormone-sensitive prostate cancer: first results from the STAMPEDE multiarm, multistage, randomised controlled trial.
Lancet Oncol. 2012; 13: 549-558
Acid-sensing ion channel-1 contributes to axonal degeneration in autoimmune inflammation of the central nervous system.
Nat Med. 2007; 13: 1483-1489
Acid-sensing ion channel 1 is involved in both axonal injury and demyelination in multiple sclerosis and its animal model.
Brain. 2011; 134: 571-584
Targeting ASIC1 in primary progressive multiple sclerosis: evidence of neuroprotection with amiloride.
Brain. 2013; 136: 106-115
Up-regulation of 5-HT2B receptor density and receptor-mediated glycogenolysis in mouse astrocytes by long-term fluoxetine administration.
Neurochem Res. 2002; 27: 113-120
Fluoxetine regulates the expression of neurotrophic/growth factors and glucose metabolism in astrocytes.
Psychopharmacology. 2011; 216: 75-84
The effect of fluoxetine on progression in progressive multiple sclerosis: a double-blind, randomized, placebo-controlled trial.
ISRN Neurol. 2013; 2013370943
A review of the neural mechanisms of action and clinical efficiency of riluzole in treating amyotrophic lateral sclerosis: what have we learned in the last decade?.
CNS Neurosci Ther. 2011; 17: 4-31
Mechanisms of glutamate toxicity in multiple sclerosis: biomarker and therapeutic opportunities.
Lancet Neurol. 2016; 15: 1089-1102
Glutamate excitotoxicity in a model of multiple sclerosis.
Nat Med. 2000; 6: 67-70
Autoimmune encephalomyelitis ameliorated by AMPA antagonists.
Nat Med. 2000; 6: 62-66
Riluzole suppresses experimental autoimmune encephalomyelitis: implications for the treatment of multiple sclerosis.
Brain Res. 2003; 989: 196-204
Glutamate inhibition in MS: the neuroprotective properties of riluzole.
J Neurol Sci. 2005; 233: 113-115
A randomized controlled phase II trial of riluzole in early multiple sclerosis.
Ann Clin Transl Neurol. 2014; 1: 340-347
Multiple Sclerosis-Secondary Progressive Multi-Arm Randomisation Trial (MS-SMART): a multiarm phase IIb randomised, double-blind, placebo-controlled clinical trial comparing the efficacy of three neuroprotective drugs in secondary progressive multiple sclerosis.
BMJ Open. 2018; 8e021944
Defining the clinical course of multiple sclerosis: the 2013 revisions.
Neurology. 2014; 83: 278-286
Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria.
Ann Neurol. 2011; 69: 292-302
Diagnostic criteria for multiple sclerosis: 2005 revisions to the “McDonald criteria”.
Ann Neurol. 2005; 58: 840-846
Accurate, robust, and automated longitudinal and cross-sectional brain change analysis.
Neuroimage. 2002; 17: 479-489
Advances in functional and structural MR image analysis and implementation as FSL.
Neuroimage. 2004; 23: S208-S219
Sample sizes for brain atrophy outcomes in trials for secondary progressive multiple sclerosis.
Neurology. 2009; 72: 595-601
Effect of high-dose simvastatin on brain atrophy and disability in secondary progressive multiple sclerosis (MS-STAT): a randomised, placebo-controlled, phase 2 trial.
Lancet. 2014; 383: 2213-2221
Lamotrigine for neuroprotection in secondary progressive multiple sclerosis: a randomised, double-blind, placebo-controlled, parallel-group trial.
Lancet Neurol. 2010; 9: 681-688
Multiple testing problems in pharmaceutical statistics.
1st edn. Chapman & Hall/CRC Biostatistics Series,
A multiple comparison procedure for comparing several treatments with a control.
J Am Stat Assoc. 1955; 50: 1096-1121
The factor analytic approach to simultaneous inference in the general linear model.
J Comput Graph Stat. 1992; 1: 151-168
MRI outcomes in a placebo-controlled trial of natalizumab in relapsing MS.
Neurology. 2007; 68: 1390-1401
Mechanisms of action of disease-modifying agents and brain volume changes in multiple sclerosis.
Neurology. 2008; 71: 136-144
Antidepressants suppress production of the Th1 cytokine interferon-γ, independent of monoamine transporter blockade.
Eur Neuropsychopharmacol. 2006; 16: 481-490
Effects of fluoxetine on disease activity in relapsing multiple sclerosis: a double-blind, placebo-controlled, exploratory study.
J Neurol Neurosurg Psychiatry. 2008; 79: 1027-1031
Fluoxetine in progressive multiple sclerosis: the FLUOX-PMS trial.
Mult Scler J. 2019; 25: 1728-1735
Published: January 22, 2020
© 2020 The Author(s). Published by Elsevier Ltd.