This study is now finally published. The initial findings were presented virtually at the last ECTRIMS and although the study was negative (i.e. the primary outcome was not met) it is still interesting to read the whole caboodle in the final article.
The road to remyelination has been a tough one, and this one in particular is underpinned by at least a decade of basic science preceding it. Bexarotene, a retinoic acid receptor agonist (with activity against the α, β, and γ isoforms) has the potential to promote remyelination. In the final study 52 participant were randomised 1:1 to bexarotene vs. placebo (dummy tablet).
Although, the study did not meet its primary endpoint, which was to show a difference in the mean adjusted submedian lesional magnetisation transfer ratio (MTR – an MRI marker purported to reflect remyelination) between bexarotene and placebo, it did on further data evaluation demonstrate a difference on lesional MRI in in cortical grey matter, deep grey matter, and the brainstem (see figure below). The primary outcome here is a summary score based on imaging assumptions of remyelination that may be taking place, whilst the latter is more selective of MS lesions in pre-specified territories. It is always tricking deciding on a correct primary outcome for work that has not been done before, and maybe trials that follow of this ilk may use predefined lesional analysis from the outset. They also looked at visual evoked potentials (an assessment of visual pathway conduction), which again was negative but I would say this was not the correct outcome for this study.
Moreover there were no visible improvements in disability, and again the study was not powered to demonstrate this. If the initial findings had been promising I’m sure a larger study to check this would have been undertaken.
However, this is also a safety study, and unfortunately adverse events occurred more frequently in the bexarotene arm compared to the placebo arm; 159 event in total vs. 39. Things that would be considered problematic was that all 26 individuals on treatment became hypothyroid, with 24 starting thyroid tablets. The other was that supranormal elevations in triglyceride levels (a type of body fat) occurred in a majority of individuals on treatment with 6 displaying of readings of 10 mmol/l or more, which is in the familial hypertriglyceridemia range (normally high levels are between 2-6 mmol/l).
The final word from the authors was as follows: “We do not recommend the use of bexarotene in people with multiple sclerosis. Bexarotene was poorly tolerated and the primary efficacy outcome, which comprised an MRI endpoint untested in previous trials, was not met. Nonetheless, converging evidence from several other MRI and electrophysiological outcomes, in this trial not powered to detect a treatment difference in these outcomes, suggests that bexarotene might have a small biological effect, promoting remyelination in some demyelinated lesions.” Which, I have to agree with. But, the question remains whether it is worth tinkering away at more selective RXR agonists or dumping the project in entirety.
Safety and efficacy of bexarotene in patients with relapsing-remitting multiple sclerosis (CCMR One): a randomised, double-blind, placebo-controlled, parallel-group, phase 2a study
J William L Brown 1 , Nick G Cunniffe 2 , Ferran Prados 3 , Baris Kanber 4 , Joanne L Jones 5 , Edward Needham 5 , Zoya Georgieva 5 , David Rog 6 , Owen R Pearson 7 , James Overell 8 , David MacManus 9 , Rebecca S Samson 9 , Jonathan Stutters 9 , Charles Ffrench-Constant 10 , Claudia A M Gandini Wheeler-Kingshott 11 , Carla Moran 12 , Paul D Flynn 13 , Andrew W Michell 5 , Robin J M Franklin 14 , Siddharthan Chandran 15 , Daniel R Altmann 16 , Declan T Chard 17 , Peter Connick 10 , Alasdair J Coles 5
Background: Progressive disability in multiple sclerosis occurs because CNS axons degenerate as a late consequence of demyelination. In animals, retinoic acid receptor RXR-gamma agonists promote remyelination. We aimed to assess the safety and efficacy of a non-selective retinoid X receptor agonist in promoting remyelination in people with multiple sclerosis.
Methods: This randomised, double-blind, placebo-controlled, parallel-group, phase 2a trial (CCMR One) recruited patients with relapsing-remitting multiple sclerosis from two centres in the UK. Eligible participants were aged 18-50 years and had been receiving dimethyl fumarate for at least 6 months. Via a web-based system run by an independent statistician, participants were randomly assigned (1:1), by probability-weighted minimisation using four binary factors, to receive 300 mg/m2 of body surface area per day of oral bexarotene or oral placebo for 6 months. Participants, investigators, and outcome assessors were masked to treatment allocation. MRI scans were done at baseline and at 6 months. The primary safety outcome was the number of adverse events and withdrawals attributable to bexarotene. The primary efficacy outcome was the patient-level change in mean lesional magnetisation transfer ratio between baseline and month 6 for lesions that had a baseline magnetisation transfer ratio less than the within-patient median. We analysed the primary safety outcome in the safety population, which comprised participants who received at least one dose of their allocated treatment. We analysed the primary efficacy outcome in the intention-to-treat population, which comprised all patients who completed the study. This study is registered in the ISRCTN Registry, 14265371, and has been completed.
Findings: Between Jan 17, 2017, and May 17, 2019, 52 participants were randomly assigned to receive either bexarotene (n=26) or placebo (n=26). Participants who received bexarotene had a higher mean number of adverse events (6·12 [SD 3·09]; 159 events in total) than did participants who received placebo (1·63 [SD 1·50]; 39 events in total). All bexarotene-treated participants had at least one adverse event, which included central hypothyroidism (n=26 vs none on placebo), hypertriglyceridaemia (n=24 vs none on placebo), rash (n=13 vs one on placebo), and neutropenia (n=10 vs none on placebo). Five (19%) participants on bexarotene and two (8%) on placebo discontinued the study drug due to adverse events. One episode of cholecystitis in a placebo-treated participant was the only serious adverse event. The change in mean lesional magnetisation transfer ratio was not different between the bexarotene group (0·25 percentage units [pu; SD 0·98]) and the placebo group (0·09 pu [0·84]; adjusted bexarotene-placebo difference 0·16 pu, 95% CI -0·39 to 0·71; p=0·55).
Interpretation: We do not recommend the use of bexarotene to treat patients with multiple sclerosis because of its poor tolerability and negative primary efficacy outcome. However, statistically significant effects were seen in some exploratory MRI and electrophysiological analyses, suggesting that other retinoid X receptor agonists might have small biological effects that could be investigated in further studies.