FLAIR Vascular Hyperintensities/DWI Mismatch Predicts Penumbra in Proximal Middle Cerebral Artery Occlusion.

BACKGROUND

Perfusion imaging used to select patients for extended-window stroke therapy is not widely available. We investigated whether fluid-attenuated inversion recovery vascular hyperintensities (FVHs) on routine magnetic resonance imaging could serve as a reliable surrogate for perfusion imaging by predicting the mismatch between hypoperfused tissue volume (perfusion-weighted imaging [PWI]) and infarct core volume (diffusion-weighted imaging [DWI]) in patients with proximal middle cerebral artery occlusion.

METHODS

We included, in a retrospective Bordeaux Hospital study (2021-2022), 200 consecutive proximal middle cerebral artery occlusion patients and developed 2 semiquantitative scoring systems-FVH-c-ASPECTS (cortical-Alberta Stroke Program Early Computed Tomography Score) and DWI-c-ASPECTS-based on the 7 c-ASPECTS regions. We assessed the correlation between the FVH/DWI-c-ASPECTS ratio and the true PWI/DWI mismatch and analyzed its association with 24-hour and 3-month outcomes. Predictive performance for the PWI/DWI ratio was tested using the area under the curve-receiver operating characteristic curve. Two thresholds were evaluated-the cohort median PWI/DWI ratio and PWI/DWI ≥1.8 with mismatch >15 mL in patients imaged ≥6 hours after onset-and sensitivity, specificity, positive and negative predictive value with 95% CIs were calculated.

RESULTS

The FVH/DWI-c-ASPECTS ratio strongly correlated with the PWI/DWI mismatch (ρ=0.77; P<0.001). A ratio ≥1.6 accurately predicted a high PWI/DWI ratio ≥5 (area under the curve, 0.90; sensitivity, 82.1%; specificity, 85.2%; positive predictive value, 85.7%; negative predictive value, 81.5%). Among patients imaged ≥6 hours after onset, a ratio ≥0.8 identified a moderate PWI/DWI ratio ≥1.8 with a mismatch >15 mL (area under the curve, 0.92 [95% CI, 0.84-1.00]; sensitivity, 97.8%; specificity, 73.3%; positive predictive value, 91.8%; and negative predictive value, 91.7%). The FVH/DWI-c-ASPECTS ratio was independently associated with early major neurological improvement, 24-hour National Institutes of Health Stroke Scale score decrease ≥8 or at 0 to 1 (odds ratio, 1.31; P=0.004) and inversely associated with hemorrhagic transformation (P=0.002).

CONCLUSIONS

The FVH/DWI-c-ASPECTS ratio is a potential surrogate imaging marker of perfusion-diffusion mismatch in patients with proximal middle cerebral artery occlusion. It may offer a practical and widely accessible alternative to perfusion imaging for identifying candidates for reperfusion therapy beyond the standard time window.