Abstract TMP71: Optimizing Stroke Outcome Prediction: Derivation And Validation Of The Charlotte Large Artery Occlusion Endovascular Therapy Outcome Score

Abstract

Introduction: Prognostic scores for large artery occlusion patients treated with endovascular intervention have not incorporated CTP parameters. We derived the Charlotte Large artery occlusion Endovascular therapy Outcome Score (CLEOS) to optimize outcome prediction after mechanical thrombectomy. Methods: Patients presenting with an ICA or MCA occlusion treated with thrombectomy from November 2016-July 2020 were included from our health system’s prospectively collected, code stroke registry in the derivation analysis. Prognostic factors independently associated with poor outcomes (90-day mRS score 4-6) were combined with CTP parameters in a regression model. Internal cross-validation was performed. A separate, prospective validation cohort was included from the registry with patients from December 2020-April 2021. The final scale was compared to previously reported scales using Area Under the Curve-Receiver Operator Characteristic analysis. Results: 646 patients were included in the derivation analysis and 60 patients in the separate validation cohort. CLEOS = (6 x Age) + (13 x NIHSS) + Glucose - (162 x CBV Index). CLEOS predicted poor 90-day outcomes (AUC 0.74, 95% CI 0.71-0.78) better than the Stroke Prognostication using Age and National Institute of Health Stroke Scale-100 Index (AUC 0.63, p < 0.0001), the Houston Intra-Arterial Therapy 2 score (AUC 0.69, p = 0.0029), and the Pittsburgh Response to Endovascular therapy score (AUC 0.71, p = 0.0240) in a combined analysis of the derivation and validation cohorts. Patients with CLEOS scores ≥ 850 did not have a lower risk of poor 90-day outcomes despite excellent revascularization (modified TICI 2c-3). Conclusions: CLEOS can predict poor 90-day outcomes after an ICA or MCA occlusion and help identify patients unlikely to benefit from endovascular thrombectomy despite excellent revascularization.