DWI Positivity in Mild, Nondisabling Acute Cerebral Ischemia: Data From the PRISMS Trial.

Background: In patients with acute cerebral ischemia (ACI) and mild deficit, acute infarction on magnetic resonance imaging (diffusion-weighted imaging [DWI]-positivity) is inconsistently detected. Our objective was to characterize the rate of detection in the first dedicated trial of a population with mild, nondisabling stroke and assess the detection and its association with baseline clinical and imaging characteristics. We hypothesized that white matter hyperintensity (WMH) would independently predispose patients to DWI-positivity. Methods: The phase 3b, randomized, double-blinded PRISMS (Potential of Alteplase for Ischemic Strokes with Mild Symptoms) trial compared administration of intravenous alteplase to aspirin for mild (National Institutes of Health Stroke Scale score 0-5), nondisabling stroke at ≤3 hours from onset. For the current, prespecified study, patients with a final diagnosis of neurovascular mimic were excluded. Central readers, blinded to the treatment arm, assessed day 2 magnetic resonance imaging for prior infarct, WMH burden (graded quantitatively [volume] and qualitatively [(Fazekas]), and presence of DWI-positivity (primary outcome). Alteplase treatment, demographics, and clinical covariates were prespecified for adjustment in regression modeling. Results: Of 313 patients enrolled, 273 (87%) had a final diagnosis of ACI. Of 212 (77%) cases of ACI with magnetic resonance imaging, 109 (51%) had DWI positivity (median lesion volume 1.20 cc, interquartile range 0.57-2.50 cc). Univariately, the proportion with higher Fazekas grade (grade 2-3: 27% vs 41%; P = 0.02) and prior infarct (45% vs 27%, P<0.01) were associated with DWI positivity. WMH was not associated with DWI positivity in adjusted analysis (adjusted odds ratio [aOR]: 1.01, 95% CI: 0.99-1.03). The multivariable model also showed alteplase treatment inversely associated with DWI-positivity (aOR: 0.48, 95% CI: 0.25-0.93). Conclusions: Among patients presenting with mild, nondisabling ACI, WMH is not associated with DWI-positivity. A hypothesis emerges that alteplase treatment may reduce DWI positivity in mild, nondisabling ACI.