A phase III randomized efficacy trial of 2000 mg citicoline in acute ischemic stroke patients

Abstract

Background: Citicoline may reduce CNS ischemic injury by stabilizing cell membranes and reducing free radical generation. Previous safety and efficacy trials in patients who have had acute strokes suggested that citicoline may improve neurologic outcome with minimal side effects.

Objective: To determine the safety and efficacy of citicoline treatment in acute stroke patients.

Method: An 118-center, randomized, double-blind, efficacy trial in 899 patients compared placebo (n = 446) with citicoline (n = 453) (1000 mg PO twice a day) for 6 weeks, with a 6-week post-treatment follow-up period. Patients with acute (< or =24 hours) ischemic strokes clinically thought to be in the middle cerebral artery territory with NIH Stroke Scale (NIHSS) scores > or =8 were enrolled.

Results: Mean time to treatment was 13 hours for both groups and mean age was 67 years for those receiving placebo and 68 years for those receiving citicoline. Mean baseline NIHSS scores were 14.5 for placebo and 13.9 for citicoline (p = 0.06); medians were 14 for placebo and 13 for citicoline (p = 0.04). The incidence and type of side effects were similar between the groups. There were no between-group differences on the planned primary analysis, percent of patients with a > or =7-point NIHSS score change at 90 days (placebo 51%, citicoline 52%). There were no between-group differences on the other planned secondary analyses at 90 days, including mortality. However, post hoc analyses using standard "excellent recovery" measures suggested a possible treatment effect on the modified Rankin 0 or 1 (last observation carried forward: placebo 20%, citicoline 26%; p = 0.025) as well as a global outcome statistic.

Conclusions: Citicoline was safe but ineffective in improving the outcome of patients with acute ischemic stroke as measured by the planned analyses. Post hoc analyses suggest that a modest treatment effect may have been seen if more traditional analyses had been used.