How Do Patient Outcomes in Mechanical Thrombectomy for Large-Core Stroke Vary Based on Neuroimaging Modalities Used for Patient Selection? A Multicenter Multinational Study

The role of different imaging modalities-non-contrast CT (NCCT), CT perfusion (CTP), and diffusion-weighted imaging (DWI)-in selecting patients with large-core stroke for endovascular thrombectomy (EVT) is a subject of ongoing debate. This study aims to determine whether patients with large-core acute ischemic stroke (AIS) undergoing EVT triaged with CTP or DWI in addition to NCCT had different clinical outcomes compared to those only triaged with NCCT. We queried the Stroke Thrombectomy and Aneurysm Registry (STAR) for patients enrolled between 2014 and 2023 who presented with anterior-circulation AIS and large ischemic core (ASPECTS < 6) who underwent EVT in 41 stroke centers in the USA, Europe, Asia, and South America. Patients were stratified by the imaging used before EVT. Propensity score matching (PSM) was used to compare balanced cohorts of patients with NCCT vs CTP and NCCT vs DWI. The primary outcome was a favorable 90-day functional status (mRS 0-3). Secondary outcomes included intracerebral hemorrhage (ICH) rates, symptomatic ICH (sICH), and successful/complete recanalization, as determined by mTICI score. A total of 403 patients were included, 121 were selected with NCCT alone, 227 with CTP, and 55 with DWI. Before PSM, 90-day mRS 0-3, successful reperfusion mTICI ≥ 2B, and sICH rates were similar across the three imaging modalities. mTICI-2C or greater rates were highest in DWI (50.9%; p < 0.01), followed by NCCT (41.3%) and CTP (27.8%). Patients selected with CTP had the highest ICH incidence (44.1%; p < 0.01). After 1:1 PSM, 104 pairs of NCCT vs CTP and 36 pairs of NCCT vs DWI were compared. There were no significant differences in any procedural or functional outcome measure between the matched groups, including mTICI ≥ 2C recanalization, 90-day mRS 0-3, ICH rates, and sICH rates. In patients with anterior large-vessel occlusion AIS with low ASPECTS, we found that selecting patients for EVT based on NCCT or employing advanced imaging to elucidate collaterals, infarct volume, and ischemic penumbra does not alter procedural or patient outcomes. NCCT alone may be sufficient to select patients for EVT in this patient population, especially in settings with limited resources.