Comparison of multiple disease modifying therapies in multiple sclerosis with marginal structural models. [BMJ Neurology Open]

Abstract

Background Because of methodological challenges comparisons of multiple treatments in multiple sclerosis cohorts have been limited to pairwise and triple comparisons. Objectives Extend marginal structural models (MSM) to allow simultaneous comparisons of multiple MS treatments. Methods We selected patients from the MSBase registry with Clinically Isolated Syndrome and Relapsing-Remitting MS followed for -1 year, with -3 visits, -1 visit per year and exposed to a MS therapy. MSMs were used to compare cumulative hazards of 6-month confirmed worsening and improvement of disability, and the incidence of relapses between treatments. MSMs were continuously re-adjusted for patient age, sex, pregnancy, date from first symptom, prior relapse history and MRI activity. We used MSMs to compare the Average Treatment Effect (ATE), the effect a treatment would have had if the entire study population had been treated with this treatment vs. another treatment. We also estimated the Average Treatment Effect Among the Treated (ATT): comparison an observed effect of a treatment with a counterfactual (not observed) effect of another treatment in the same study population. Results Among 23687 patients, we compared ATE of glatiramer acetate (reference), interferon b, natalizumab, fingolimod, dimethyl fumarate, and teriflunomide. In ATE, a reduction of relapse frequency was more prominent on natalizumab, followed by fingolimod (47% and 24% respectively, reference: glatiramer acetate) when compared with the other treatments. The ATT models confirmed these observations. Conclusions Compared to other DMTs natalizumab and fingolimod were associated with superior reduction in relapse frequency than glatiramer acetate, interferon beta, teriflunomide and dimethyl fumarate.