Optimal outcome measures for detecting clinical benefits of early reperfusion: insights from the DEFUSE Study.

BACKGROUND: There is no consensus regarding which clinical outcome scales are the most sensitive indicators of early reperfusion in patients with acute stroke.
METHODS: Patients with acute stroke enrolled in the Diffusion and Perfusion Imaging Evaluation for Understanding Stroke Evolution (DEFUSE) study with a perfusion-/diffusion-weighted imaging mismatch at baseline magnetic resonance imaging were studied. Prespecified secondary outcome measures were evaluated at both 30 and 90 days after treatment with intravenous tissue plasminogen activator. A nonparametric recursive partitioning algorithm was also used to identify the optimal dichotomous splits for differentiating patients who experienced early reperfusion from those who did not.
RESULTS: In all, 34 of the 74 patients enrolled in DEFUSE met the inclusion criteria for this study. Statistically significant benefits of reperfusion were documented on multiple outcome measures. The most powerful predefined outcome measure was improvement in the National Institutes of Health Stroke Scale (NIHSS) score of greater than or equal to 11 points between baseline and day 90 and/or a day-90 NIHSS score of 0 to 1 (odds ratio 22.5, P = .0021). The recursive partitioning algorithm analysis identified an improvement of greater than or equal to 10 on the NIHSS score between baseline and 30 days and an NIHSS score of less than or equal to 2 at 30 days as optimal end points.
CONCLUSIONS: For patients with stroke and a perfusion-/diffusion-weighted imaging mismatch treated with intravenous tissue plasminogen activator at 3 to 6 hours, a substantial change in the baseline NIHSS score (> or =10 points) is a potent discriminator of patients who experience early reperfusion from those who do not. In addition, an NIHSS score of less than or equal to 2 appears to be an excellent end point for phase II studies of reperfusion therapies.