Perfusion MR predicts outcome in high-risk transient ischemic attack/minor stroke: a derivation-validation study.

BACKGROUND AND PURPOSE Transient or minor ischemic stroke is associated with an early risk of deterioration. Baseline perfusion-diffusion mismatch may predict clinical deterioration and infarct growth in this population. METHODS High-risk transient ischemic attack and minor stroke (National Institutes of Health Stroke Scale ≤3) subjects were prospectively enrolled and imaged with MRI within 24 hours of symptom onset as part of sequential derivation and validation cohorts. Baseline diffusion-weighted imaging, perfusion-weighted imaging (Tmax≥4 s), mismatch (Tmax≥4 s-diffusion-weighted imaging), and follow-up fluid-attenuated inversion recovery infarct volumes were measured. Primary outcome was infarct growth on fluid-attenuated inversion recovery, and secondary outcome was symptom progression. RESULTS One hundred thirty-seven and 281 subjects were included in the derivation and validation cohorts, respectively. Infarct growth occurred in 18.5% of the derivation and 5.5% of the validation cohorts. Symptom progression occurred in 9.5% of the derivation and 4.5% of the validation cohorts. In the derivation cohort, subjects with baseline mismatch were significantly more likely to show infarct growth on fluid-attenuated inversion recovery (relative risk [RR], 13.5; 95% confidence interval [CI], 4.2-38.9) and symptom progression (RR, 7.0; 95% CI, 2.0-7.3). A baseline mismatch volume of 10 mL in the derivation cohort was the optimal threshold to predict infarct growth (area under the curve, 0.89; 95% CI, 0.80-0.98). This threshold was highly predictive of infarct growth in the validation cohort (P=0.001). Baseline mismatch was associated with clinical deterioration in the derivation (area under the curve, 0.81; 95% CI, 0.67-0.96) and validation cohorts (area under the curve, 0.66; 95% CI, 0.46-0.85). CONCLUSIONS Among subjects with high-risk transient ischemic attack and minor stroke, diffusion-weighted imaging-perfusion-weighted imaging mismatch predicts infarct growth and clinical deterioration. These findings suggest that reperfusion strategies would be beneficial in this population.