ABSTRACT NUMBER: ESOC2026A2498 CEREBROVASCULAR MORPHOLOGY AS A PREDICTOR OF EARLY NEUROLOGICAL DETERIORATION AFTER THROMBECTOMY

Abstract Background and aims Early neurological deterioration (END) affects 10–30% of patients with acute ischemic stroke (AIS) despite successful recanalization. END impacts mortality and long-term outcomes. We hypothesize that collateral cerebrovascular circulation impacts END occurrence and outcomes following endovascular thrombectomy (EVT). Methods We performed a retrospective analysis of AIS patients who underwent EVT 2019-2024 at our Institution. END was defined as a ≥4 points increase in the total NIHSS score within 24 hours after the initial stroke. Collateral circulation was quantified using our validated fully automated machine-learning algorithm applied to presentation computed tomography angiography (CTA). Results Among patients with END, admission NIHSS scores were significantly lower compared to those without END (mean NIHSS 10 vs. 16; p < 0.0001). No statistically significant differences were observed between groups in terms of age, race, sex, core/penumbra volume, time from symptom onset to recanalization, ASPECTS, successful recanalization (TICI ≥2B). Quantitative collateral index was significantly higher in the group without END as compared to patients with END (p = 0.021). Multivariate logistic regression analysis identified lower NIHSS at admission (OR = 1.114; 95% CI: 1.019–1.218; p = 0.0174) and lower collateralization (OR = 1.27; 95% CI: 0.99-1.62; p = 0.05) as independent predictors of END. Conclusions Among AIS patients undergoing EVT, an underdeveloped collateral network was independently associated with the occurrence of early neurological deterioration. These findings suggest that clinical characteristics and collateral circulation, may play a greater role in predicting END, and further studies are needed to refine risk stratification in this population. Conflict of interest Pouya Tahsili Fahadan: Nothing to disclose