Prognostic factors and treatment effect in the CHIMES study

Abstract

© 2015 by National Stroke Association. Background: Stroke trials often analyze patients with heterogeneous prognoses using a single definition of outcome, which may not be applicable to all subgroups. We aimed to evaluate the treatment effects of MCL601 among patients stratified by prognosis in the Chinese Medicine Neuroaid Efficacy on Stroke Recovery (CHIMES) study. Methods: Analyses were performed using data from the CHIMES study, an international, randomized, placebo-controlled, double-blind trial comparing MLC601 with placebo in patients with ischemic stroke of intermediate severity in the preceding 72 hours. All subjects with baseline data and the modified Rankin Scale (mRS) score at 3 months were included. Results: Data from 1006 subjects were analyzed. The predictive variables for mRS score greater than 1 at month 3 were age older than 60 years (P <.001), baseline National Institutes of Health Stroke Scale score 10-14 (P <.001), stroke onset to initiation of study treatment of more than 48 hours (P <.001), and female sex (P =.026). A higher number of predictors was associated with poorer mRS score at month 3 for both placebo (P <.001) and treatment (P =.001) groups. The odds ratio (OR) for achieving a good outcome increased with the number of predictors and reached statistical significance in favor of MLC601 among patients with 2 to 4 predictors combined (unadjusted OR = 1.44, 95% confidence interval, 1.02-2.03; adjusted OR = 1.60, 95% confidence interval, 1.10-2.34). Conclusions: Age, sex, baseline National Institutes of Health Stroke Scale score, and time to first dose are predictors of functional outcome in the CHIMES study. Stratification by prognosis showed that patients with 2 or more predictors of poorer outcome have better treatment effect with MLC601 than patients with single or no prognostic factor. These results have implications on designing future stroke trials.