Abstract TP303: Refining EVT Benefit in MeVO Through AI-Based Imaging and Clinical Enrichment

Introduction: Recent randomized trials failed to demonstrate benefit of endovascular treatment (EVT) in patients with medium vessel occlusions (MeVO). We aimed to identify subpopulations in which successful recanalization yields a meaningful clinical benefit. Methods: We included consecutive patients with MeVO treated at two comprehensive stroke centers. Complete recanalization was defined as TICI 2c–3. The treatment effect was defined as the difference in the proportion of patients achieving mRS 0–2 at 90 days between those with and without complete recanalization. Predicted infarct core was calculated with an AI-based software (AI-core, Methinks), using CTP maps (CBF<30%), and hypoperfused volume was calculated on Tmax>6s maps (Rapid AI). We assessed treatment effect in the whole population and after applying population enrichment strategies (clinical and imaging-based). Results: A total of 103 patients were analyzed; the rate of complete recanalization was 62.1%. In the whole cohort, the treatment effect of recanalization was 7.5%. Population enrichment improved treatment effect substantially: baseline NIHSS >10: 17.5% (28% of all patients), (AI-core > 0: 22.7% (33%), Tmax 6s> 40 ml: 25.8% (39%), Age < 80: 18.9% (60%), CBF30 > 0 ml: 26.2% (30%). Combination of different criteria yielded to higher treatment effects (figure). Conclusion: Although EVT for MeVO showed modest benefit in unselected patients, treatment effect increases markedly in enriched populations defined by baseline imaging and clinical variables. Combinations of selection criteria identify subgroups with treatment effects above 30%, suggesting that refined patient selection may unlock the clinical benefit of EVT in MeVO.