Thromobolysis for acute ischemic stroke: is intra-arterial better than intravenous? A treatment effects model

Abstract

Background: Three randomized trials of intra-arterial thrombolysis (IAT) for acute ischemic stroke ≤ 6 hours were conducted without intravenous tissue plasminogen activator (IV-tPA) treatment of patients in the control groups now known to benefit.

Methods: The effect of IV-tPA treatment on 130 control subjects in the Prolyse in Acute Cerebral Thromboembolism (PROACT), PROACT II, and Middle Cerebral Artery Embolism Local Fibrinolytic Intervention Trial (MELT) studies was modeled using linearly weighted time-dependent odds ratios (ORs) from pooled IV-tPA trials. In the PROACT trials, the model assumed that 50% (36/73) were treated at 4.5 hours, the median time to arteriography. For MELT, the model assumed treatment at arrival plus 90 minutes based on hospital arrival times obtained from the principal investigator. The OR of 1.31 for all 130 controls (91 presumed treated ≤ 4.5 hours; OR 1.44) was applied to the original control data to derive the adjusted control outcome, and this was compared to the IAT group. Sensitivity analyses were performed.

Results: Meta-analysis of the original data revealed a statistically significant benefit for IAT (P = .03). After adjustment for the effect of IV-tPA in controls, there was no longer a significant treatment benefit for IAT (P = .26). Loss of significant IAT treatment benefit persisted if either the OR for benefit of IV-tPA or the number of treated controls was more than halved. These 3 randomized trials of IAT for acute ischemic stroke ≤ 6 hours would not likely have shown a benefit if eligible controls had been treated with IV-tPA.

Conclusions: Whether IAT is superior to IV-tPA in IV-tPA-eligible patients or better than placebo in IV-tPA-ineligible patients remains to be determined.