The Use of PWI and DWI Measures in the Design of “Proof‐of‐Concept” Stroke Trials

Abstract

The authors used serial magnetic resonance perfusion-weighted imaging (PWI) and diffusion-weighted imaging (DWI) to determine whether major reperfusion and the attenuation of infarct expansion are associated with improved stroke outcome. Forty-nine patients were studied with serial magnetic resonance imaging within 6 hours of stroke onset and again at 4 days (subacute studies) and 3 months (outcome studies). Two imaging parameters were examined: infarct expansion between acute and outcome studies and major reperfusion between acute and subacute studies. Patients with major reperfusion (45% of those with acute PWI lesions) were more likely to have little or no disability at outcome (National Institutes of Health Stroke Scale [NIHSS] score < or = 4, P = .0176; Barthel Index [BI] score > or = 90, P = .0547) after adjustment for baseline differences. In contrast, patients with infarct expansion (48%) were more likely to be dead or dependent at outcome (BI < 90, P = .0414; NIHSS score P = .082; modified Rankin Scale score > 2, P < .0001). These measures were used to generate sample size calculations based on hypothetical treatment effects. Therapies postulated to double the proportion of patients with major reperfusion from one third to two thirds would require 41 patients in each group (treated and untreated) to be sufficiently powered to show a difference. Interventions postulated to halve the number of patients with infarct expansion from 50% to 25% would require 66 patients in each group to show a difference. Infarct expansion and major reperfusion are associated with clinically meaningful changes in stroke outcome. These measures could be used as surrogate markers of outcome in late phase II proof-of-concept stroke studies designed to provide efficacy signals before embarking on large phase III studies with definitive clinical endpoints.