Efficacy of Delayed-Release Dimethyl Fumarate vs Glatiramer Acetate on a Novel Composite Outcome Measure of Inflammatory Disease Activity: Post-Hoc Analysis of the CONFIRM Study

Document Type

Conference Abstract

Publication Date


Publication Title

Multiple Sclerosis Journal






Background: Compared with placebo (PBO), delayed-release dimethyl fumarate (DMF) demonstrated efficacy and safety in relapsing-remitting multiple sclerosis (RRMS) in the Phase 3 DEFINE/CONFIRM studies. To more fully examine the effect of DMF on active inflammatory MS, we developed a novel composite endpoint combining relapse, gadolinium-enhancing (Gd+) lesions, and new/enlarging T2 lesions.

Objectives: Compare the efficacy of DMF vs glatiramer acetate (GA) on inflammatory disease activity in a post-hoc analysis of the CONFIRM study.

Methods: Eligibility criteria included age 18-55 years and EDSS score 0-5.0. Patients were randomized to receive DMF 240 mg twice (BID) or thrice daily, PBO, or subcutaneous GA (reference comparator) for up to 2 years (96 weeks). Brain MRI scans were performed in a subset of patients (MRI cohort). Inflammatory disease activity was defined as an event (relapse, Gd+ lesion, or new/enlarging T2 lesion) occurring within a specified interval (week 0−24, 24−48, or 48−96). Patients were considered free of inflammatory disease activity if they did not experience an event within a given interval or any preceding intervals, and were evaluated for inflammatory disease activity as long as they were known to be at risk. Estimate of an underlying proportional hazards model in continuous time was based on a complementary log-log model adjusted for baseline number of relapses (⩽1 vs ⩾2), EDSS score (⩽2.0 vs >2.0), presence or absence of Gd+ lesions, and T2 lesion volume (⩽median vs >median). Results are reported for DMF BID (approved dosing regimen in all regions).

Results: The intent-to-treat population included 359, 350, and 363 patients receiving DMF, GA, or PBO, respectively; among them, 169, 175, and 167 were in the MRI cohort. Life-table estimates of the proportion (standard error) of patients receiving DMF vs GA who were free of inflammatory disease activity were 36% (4%) vs 29% (3%) during week 0−24, 34% (4%) vs 23% (3%) during week 24−48, and 21% (3%) vs 16% (3%) during week 48−96. The overall hazard ratio (HR) (95% confidence interval [CI]) for DMF vs GA was 0.77 (0.59−0.99; P=.0446). To confirm the efficacy of DMF on this novel endpoint, we also compared the effects of DMF vs PBO: the overall HR (95% CI) for DMF vs PBO was 0.60 (0.46−0.79; P=.0002).

Conclusions: DMF significantly reduced the risk of inflammatory disease activity over 2 years compared with GA. The differential treatment effect was seen by 24 weeks.