Dimethyl Fumarate or Teriflunomide for Relapsing-Remitting Multiple Sclerosis: A Meta-analysis of Post-marketing Studies

Abstract

In the absence of head-to-head comparison trials, we aimed to compare the effectiveness of two largely prescribed oral platform disease-modifying treatments for relapsing-remitting multiple sclerosis, namely, dimethyl fumarate (DMF) and teriflunomide (TRF). We searched scientific databases to identify real-world studies reporting a direct comparison of DMF versus TRF. We fitted inverse-variance weighted meta-analyses with random effects models to estimate the risk ratio (RR) of relapse, confirmed disability worsening (CDW), and treatment discontinuation. Quantitative synthesis was accomplished on 14 articles yielding 11,889 and 8133 patients treated with DMF and TRF, respectively, with a follow-up ranging from 1 to 2.8 years. DMF was slightly more effective than TRF in reducing the short-term relapse risk (RR = 0.92, p = 0.01). Meta-regression analyses showed that such between-arm difference tends to fade in studies including younger patients and a higher proportion of treatment-naïve subjects. There was no difference between DMF and TRF on the short-term risk of CDW (RR = 0.99, p = 0.69). The risk of treatment discontinuation was similar across the two oral drugs (RR = 1.02, p = 0.63), but it became slightly higher with DMF than with TRF (RR = 1.07, p = 0.007) after removing one study with a potential publication bias that altered the final pooled result, as also confirmed by a leave-one-out sensitivity analysis. Discontinuation due to side effects and adverse events was reported more frequently with DMF than with TRF. Our findings suggest that DMF is associated with a lower risk of relapses than TRF, with more nuanced differences in younger naïve patients. On the other hand, TRF is associated with a lower risk of treatment discontinuation for side effects and adverse events.

Keywords: Dimethyl fumarate; Disease-modifying treatments; Head-to-head study; Multiple sclerosis; Teriflunomide.

Fig. 1

PRISMA flow chart [13] for study selection

Fig. 2

Forest plots (left) and funnel plots (right) for the risk of relapse (A), confirmed disability worsening (B), and treatment discontinuation (C). Effect size refers to log-transformed risk ratios (estimates < 1.0 indicate better outcome for DMF versus TRF). Gray diamond in C refers to the pooled effect size by including also one study [22] that, in the funnel plot, is located in the area of statistically significant publication bias (black arrow) and that alters the final results, as indicated by a leave-one-out sensitivity analysis. Please note that some studies were partly or totally funded by the manufacturer of dimethyl fumarate [18, 19, 28] or teriflunomide [24, 30, 31]