Intravenously administered tissue plasminogen activator useful in milder strokes? A meta-analysis

The dilemma of whether to treat mild strokes with tPA is a chronic problem. We performed a meta-analysis and metaregression of the published literature to determine the best definition of mild strokes and if intravenously administered tissue plasminogen activator (IV-tPA) is beneficial.

PubMed, Embase, Science Direct, and Cochrane CENTRAL were searched from inception to May 2013. The search terms used were "stroke," "cerebral infarct," "mild stroke," "minor stroke," "small infarct," "modified Rankin scale," "National Institutes of Health Stroke Scale (NIHSS) score," "stroke thrombolysis," and their combinations. Studies were included if they (1) involved 5 or more human patients with stroke; (2) analyzed modified Rankin scale (mRS) scores as the main variables of interest; (3) presented outcomes for NIHSS scores less than 6, 5, 4, or 3 points. Good outcomes were defined as mRS scores 0-1, and other outcomes studied were intracranial hemorrhage and mortality.

Of 894 articles, 30 articles met our criteria. Only 8 articles provided patients arms with and without tPA treatment. A total of 637 patients with IV-tPA treatment and 568 without thrombolysis were included in analysis. Good outcomes were associated with tPA and just reached statistical significance (pooled odds ratio [OR], 1.319; 95% confidence interval [CI], 1.004-1.733; z = 1.987; P = .047). There were moderate levels of heterogeneity between studies (τ(2) = .346; Q = 19.974; df = 7; P = .006; I(2) = 64.954). On metaregression of a-priori sources of heterogeneity within individuals, we found age (B = -.37; z = -2.496; P = .012) to be a significant moderator. Mortality was not significantly different between IV-tPA-treated and nonthrombolyzed groups (pooled OR 1.095; 95% CI, .438-2.738; z = .193; P = .847).

Patients with mild stroke may derive benefit from intravenous thrombolysis without a significant increase in mortality.