Serial Computed Tomography Perfusion and Diffusion Imaging After Complete Recanalization in Large-Vessel Occlusion

Infarct evolution on diffusion-weighted imaging (DWI) after complete recanalization and its relationships to computed tomography perfusion core and outcome remain uncertain.

From a multicenter registry (2014-2024), we identified 189 patients with anterior-circulation large-vessel occlusion, modified Thrombolysis in Cerebral Infarction grade 3 recanalization, prethrombectomy computed tomography perfusion, immediate DWI within 24 hours, and delayed DWI at 24 to 168 hours. Robust linear regression identified predictors of DWI volume change; logistic regression assessed 90-day modified Rankin Scale scores of 0 to 2.

DWI volume increased in 162 of 189 patients (85.7%), from a median of 4.8 mL (interquartile range, 1.1-16.4) immediately after thrombectomy to 11.7 mL (interquartile range, 3.8-40.2) on delayed imaging; the median change was 4.4 mL (interquartile range, 0.7-14.8). Baseline computed tomography perfusion core was method-sensitively associated with greater change in the full model without immediate DWI (β=0.154 [heteroscedasticity-consistent 95% CI, -0.024 to 0.332]; P=0.09; bootstrap 95% CI, 0.018-0.336; P=0.014) and attenuated after immediate DWI adjustment (β=0.100; P=0.22; bootstrap P=0.077). Each decile increase in volume change was associated with lower odds of functional independence (adjusted odds ratio, 0.81 [95% CI, 0.68-0.97]; P=0.021). Delayed DWI showed the best outcome discrimination.

DWI-defined infarct volume increased after complete recanalization and was associated with functional outcome. The core association was method-sensitive and attenuated after accounting for immediate DWI, supporting a possible mediated rather than independent pathway.