Abstract DP295: Site Specific Performance of Thrombectomy Techniques for First Pass Effect from the SVIN Registry

Background and Aims: The optimal initial mechanical thrombectomy (MT) technique for large vessel occlusion (LVO) by occlusion site remain unknown. This study aimed to investigate the interplay between occlusion location and MT technique on first pass effect (FPE) rates. Methods: Data were analyzed from the SVIN Registry+, a prospective, multicentre U.S. registry capturing endovascular stroke procedures (January 2012–January 2024) to study patients with LVO treated with stent-retriever only (SR classic), direct aspiration, or a combined strategy. FPE was defined as eTICI 2c/3 reperfusion on the initial device pass. Associations between technique, occlusion site and FPE were explored with multivariable logistic regression that incorporated balloon-guide catheter use, age and vascular comorbidities as covariates. Results: Of the 6,413 MT procedures, 4,140 cases contained complete covariate data and were entered into the multivariable models; 33.5 % were treated with SR classic, 30.0 % with aspiration, and 36.5 % with combination therapy. Compared with SR classic, aspiration was less likely to achieve FPE in intracranial ICA occlusions (aOR 0.53, 95 % CI 0.31–0.92, p = 0.024) and in M2 segments (aOR 0.66, 0.46–0.94, p = 0.025). No technique demonstrated clear superiority for basilar artery occlusions. Conclusions: These real-world data reinforce that MT efficacy is not one-size-fits-all. SR classic yielded higher FPE in intracranial ICA and M2 occlusions and a modest edge over combination therapy in M1 lesions, while no approach was clearly superior for basilar strokes. Prospective, occlusion-stratified trials are needed to validate these signals and refine device-selection algorithms aimed at maximising FPE and functional recovery.