Selection of anterior circulation acute stroke patients for mechanical thrombectomy

Background The use of mechanical thrombectomy (MT) for acute ischemic stroke (AIS) patients has increased with a parallel burden in procedural costs. We tested whether a new prognostic score could identify patients who are unlikely to benefit from MT. Methods Patients from our endovascular stroke registry were assessed for imaging and clinical outcome measures and randomly divided into two subsets for derivation and validation. We created a new prognostic score based on clinical and radiological prognostic factors of poor outcome (mRS score >= 3) from the derivation cohort. Receiver operating characteristics curve analysis was used to assess the discrimination ability of the score. The score was then validated and compared to the MR PREDICTS score. Results The derivation/validation included 270/116 patients, respectively. After multivariate logistic regression analysis, pre stroke mRS, age, admission glycaemia, admission NIHSS, collateral flow, Clot Burden Score, Alberta Stroke Program Early CT score were used to create a new prognostic scoring system called Tor Vergata Stroke Score (TVSS). TVSS revealed a good prognostic accuracy with an AUC of 0.825 [95% CI 0.77-0.88] in the derivation cohort and an AUC of 0.820 [95% CI 0.74-0.90] in the validation cohort. When compared to the MR PREDICTS in the validation cohort, TVSS demonstrated higher prediction ability which was, however, not statistically significant (0.80 vs 0.78; P = 0.26). Conclusions TVSS is a reliable tool for selection of AIS candidates for MT and optimization of transfer to comprehensive stroke centers.