ABSTRACT NUMBER: ESOC2026A2434 PREDICTING FUTILITY IN PATIENTS AGED ≥80 YEARS UNDERGOING MECHANICAL THROMBECTOMY: NOVEL PROGNOSTIC SCORES FOR STANDARD AND EXTENDED TIME WINDOWS

Abstract Background and aims Although mechanical thrombectomy significantly improved outcomes for large vessel occlusion stroke, patients aged ≥80 years remain underrepresented in randomized trials, creating uncertainty in real-world decision-making. We aimed to identify key prognostic predictors in this specific population and develop practical risk scores to guide acute management eligibility. Methods We retrospectively analyzed patients aged ≥80 years undergoing MT, stratified into standard (SW) and extended (EW) time windows. Candidate predictors selected from the literature were analyzed using multivariable logistic regression to identify independent predictors of futility (90-day mRS 5–6). Two specific prognostic scores were constructed by assigning points proportional to the regression coefficients (Exp[B]) and validated using Area Under the ROC curve (AUC) analysis. Results The SW cohort included 157 patients. The derived SW-score assigned 2 points for pre-stroke mRS ≥3, ASPECTS ≤6, and NLR ≥6; and 1 point for NIHSS ≥17, atrial fibrillation, age >85 years, absence of IV-tPA, and poor collateral flow. This model yielded an AUC of 0.764 (C.I. 0,69-0,84). The EW cohort included 87 patients. The EW-score assigned 2 points for ischemic core volume >8 ml, and 1 point for NIHSS ≥14 and age >85 years. This score demonstrated good discriminative performance for predicting futility, with an AUC of 0.777 (C.I. 0,68-0,87). Conclusions Futility in the very elderly is multifactorial, involving specific clinical, radiological, and laboratory markers. The proposed scores provide a standardized tool to refine patient selection, potentially avoiding futile procedures while optimizing the benefit-risk ratio of reperfusion therapies in this fragile cohort. Conflict of interest Nothing to disclose