Effect of glatiramer acetate on primary progressive multiple sclerosis: Initial analysis of the completed PROMiSe trial

OBJECTIVE:To determine if glatiramer acetate (GA) slows the progressive accumulation of disability in multiple sclerosis (MS) in the absence of the confounding effects of relapses. BACKGROUND:Immunomodulators including GA have proven efficacy in relapsing forms of MS. Their effects on more progressive forms of the disease is less certain, in part clouded by their impact on attacks. A smaller, two center study of GA in "chronic progressive" MS provided conflicting data on the effect of GA on accumulating disability. DESIGN/METHODS:943 PPMS subjects (455 males) were randomized to treatment with either GA or placebo at a 2:1 assignment until the last enrolled subject completed the 3-year, double-blind trial core. The primary efficacy endpoint was an intention to treat analysis of the time to confirmed disease progression, defined as a change of ≥ 1 EDSS point for an entry EDSS of 3.0 - 5.0, or ≥ 0.5 EDSS for an entry EDSS of 5.5 - 6.5, either sustained for 3 months. The second interim analysis included a preplanned principal efficacy analysis that was based on 935 subjects with post baseline EDSS data of whom 757 had completed ≥ 2 years on study or had terminated the study prematurely. The independent data safety monitoring committee (DSMC) could discern no treatment effect on the primary outcome measure, nor project that a treatment effect could be expected by the planned trial end. Following recommendations of the DSMC, all patients were taken off trial medication and offered continued participation as part of a natural history extension to capture additional information on this unique PPMS cohort until October 2003. An intention to treat analysis was performed for all patient data obtained through the conclusion of the 3-year trial. RESULTS:A strong trend for delayed time to and lower proportion with progression was found for the GA treated group compared to the placebo assigned cohort (p=0.071; hazard ratio 0.826, CI: 0.672 - 1.016). Most sensitivity and subcohort analyses supported the primary analysis, and some strengthened it. The survival curves for males assigned to GA diverged from placebo assigned patients early and the gap widened over time (p=0.004; hazard ratio 0.659, CI: 0.495 - 0.877). Analyses of an on drug cohort supported a treatment effect based on MRI monitored enhancements and plaque burden, but not clinical endpoints. This presumably was related to the different time dependent effects of GA on these endpoints. CONCLUSIONS:The premature discontinuation of study medication and unanticipated very low event rate greatly complicate interpretation of this trial. However, it appears that GA had beneficial impact on clinical progression, most evident in those subcohorts of patients where the placebo group showed more rapid progression.