Abstract TP133: Clot Burden Score Correlates Best With Hypoperfusion Volume and Predicts Final Infarct ASPECTS<=3 in Anterior Circulation Ischemic Strokes

Abstract

Purpose: Clot Burden Score (CBS) is a semiquantitative measure of thrombus extent on CTA for proximal anterior circulation strokes whose relationship with perfusion metrics has not been fully explored. CBS has potential added value during stroke triage if perfusion imaging is not available or suboptimal. We aimed to investigate if CBS is (1) correlated to perfusion estimates of infarct core volume, hypoperfusion volume, or collateral status, and (2) a predictor of final infarct extent without reperfusion therapy. Methods: Retrospective analysis included consecutive patients (1) arriving <24 hours from onset, (2) with occlusion in the intracranial ICA, M1, M2, or A1 or combination, (3) with concurrent CTP, and (4) had followup CT or DWI during hospitalization. CBS is a 10-point negative ordinal scale: 2-points subtracted for occlusion of supraclinoid ICA, proximal M1, or distal M1, and 1-point subtracted for occlusion of infraclinoid ICA, A1, or M2. Correlation of CBS with perfusion metrics was evaluated using Spearman's rank test. Final infarct ASPECTS were scored for those without thrombolysis or thrombectomy. Multivariate logistic regression was used to identify predictors of final ASPECTS<=3 (approximating >=100mL). ROC analysis was evaluated to assess performance and optimal thresholds. Results: 222 patients included with median (IQR) age of 75 (65-84) and NIHSS of 18 (11-24). Median CBS was 7 (6-9) and best correlated to Tmax>6s r=-0.579 (hypoperfusion volume), followed by HIR r=-0.401 (collateral status) and rCBF<30% r=-0.306 (core volume), all p<0.001. 71 (32.0%) untreated patients also had median CBS of 7 (6-9). CBS and Tmax>6s had the strongest correlation to final ASPECTS in this subgroup, r=0.637 and r=-0.686 (both p<0.001). 22 of these (30.1%) had final ASPECTS<=3. Multivariate logistic regression including CBS, Tmax>6s, HIR, baseline ASPECTS, and NIHSS showed only CBS as independent predictor of final ASPECTS<=3 (p<0.001). ROC AUC for CBS was 0.902 [95% CI: 0.819-0.985] with optimal threshold of <=6 for sensitivity 86.4% and specificity 81.6%. Conclusion: CBS as a measure of thrombus extent correlates best with baseline hypoperfusion volume on CTP and independently predicts very large final infarcts (ASPECTS<=3) without treatment.