Diagnostic value of the perfusion mismatch ratio in identifying intracranial atherosclerotic disease related occlusion.

Abstract

The ability to differentiate intracranial atherosclerotic disease (ICAD) related large vessel occlusion (LVO) from embolism is critical for stroke management. We hypothesized that the mismatch ratio derived from the automated computed tomography perfusion (CTP) could predict underlying ICAD. Patients with acute ischemic stroke (AIS) and LVO from prospective registry databases who underwent CTP were included in the derivation cohort (n=1100). The mismatch ratio, calculated as the ratio of the hypo-perfused volume to the infarct core volume by software, was defined. Receiver Operating Characteristic (ROC) analysis was performed to assess the predictive performance of the mismatch ratio for ICAD, and logistic regression analysis was used to identify independent predictors of LVO associated with underlying ICAD. External validation was conducted using cohorts from two other stroke centers (n=385). In the derivation cohort, 390 patients were classified as ICAD and 720 as embolism. The ICAD group had a higher mismatch ratio (9.8 vs 3.6, P<0.001). The mismatch ratio outperformed age and National Institutes of Health Stroke Scale (NIHSS) score in predicting ICAD (area under the curve (AUC), 0.77 vs 0.36 vs 0.28, P<0.001). The ROC curve had a best cut-off of 7.1 for predicting ICAD, which was an independent predictor of ICAD-related occlusion (adjusted odds ratio (aOR) 5.43, 95% CI 3.68 to 8.03), with 68% sensitivity and 76% specificity. These results were validated in an external cohort (AUC=0.78; 95% CI, 0.73 to 0.83). The perfusion mismatch ratio may be an optimal and simple predictor of anterior circulation ICAD- related LVO before endovascular treatment (EVT). When this ratio was ≥7.1, ICAD was five times more likely than embolism.