Safety of thrombolytic agents in acute ischemic stroke: Analysis of randomized controlled trials

PURPOSE: To evaluate the safety of certain thrombolytic agents which can currently be a treatment option for acute ischemic stroke. METHODS: Studies were identified using MEDLINE (1966 to January 2002), the Cochrane Controlled Trial Register, and references of the papers selected. A number of biomedical and stroke related websites were also searched. Randomized-placebo-controlled trials of thrombolytic agents for the treatment of acute ischemic stroke patients were eligible. Streptokinase (SK) trials were excluded since all major SK trials were terminated early and a meta-analysis of individual patient data did not indicate a subgroup of patients for whom SK can be beneficial. Study quality was evaluated by using a previously validated scale. Data was extracted in duplicate by two independent investigators. Symptomatic intracranial hemorrhage (SIH) within the first ten days and mortality from all causes during follow-up were the outcomes. Odds ratio, absolute risk difference, and number needed to harm (NNH) which is the number of patients need to be treated to cause one additional adverse outcome were calculated to evaluate the safety of thrombolytic agents. According to homogeneity test results, a fixed effects model for SIH and a random-effects model for mortality were used to pool the individual effects of trials. RESULTS: 11 randomized controlled trials involving 3,643 patients were included in the analysis. Thrombolytic therapy was associated with a three-fold increase (odds ratio [OR], 3.4; 95% confidence interval [CI], 2.4–4.75; p < 0.0001) in symptomatic intracranial hemorrhage (SIH) and an insignificant increase (OR, 1.07; 95% CI, 0.83–1.39; P = 0.291) in mortality. The treatment was associated with an absolute risk increase of 58 per 1000 persons (95% CI, 43–72; p < 0.0001) for SIH and 11 per 1000 persons (95% CI, (-24)-48; P = 0.261) for mortality. NNH was 17 (95% CI, 13–22) for SIH, and 84 (95% CI, including 0) for mortality. When including only trials using rt-PA, treatment was associated with a significant increase (OR, 3.6; 95% CI, 2.5–5.2; p < 0.0001) in symptomatic SIH and an insignificant increase (OR, 1.25; 95% CI, 0.87–1.78; P = 0.119) in mortality. CONCLUSION: These findings suggest that thrombolytic therapy increases the risk of symptomatic hemorrhage in 10 days when compared with placebo, however there is no significant difference in mortality during follow-up between the groups.