Abstract TP168: Racial and Gender Disparities in Diagnostic Workup of Cryptogenic LVO Stroke in Young Adults: Results from a Multicenter Retrospective Cohort Study

Introduction: Large vessel occlusion acute ischemic stroke (LVO-AIS) in young adults is rare and etiologically diverse, with a third of patients being classified as cryptogenic. Prior studies rely on broad TOAST categories that conflate specific mechanisms with cryptogenic designations, limiting clinical interpretation. We hypothesized that applying a prespecified standard for diagnostic completeness, combined with mechanism-level, clinically adjudicated classification, would clarify whether cryptogenic labeling reflects elusive biology or missed evaluations, and would further identify racial and sex-based disparities in diagnostic workup. Methods: We performed a retrospective cohort study of patients aged 18–50 with AIS across a multi-center, single healthcare system (2017–2021), screening 6,383 cases to identify 214 with CTA/MRA-confirmed LVO. Etiology was classified by TOAST and further adjudicated into a mechanism-level taxonomy. A “complete” evaluation required vascular imaging, transthoracic and transesophageal echocardiography, hypercoagulability testing, and prolonged rhythm monitoring. Demographic, clinical, and outcome data were abstracted for comparative analysis. Results: Among 214 patients (median age 42 [IQR 36–46]; 106 [49.5%] female) with LVO-AIS, adjudication identified 21 distinct mechanisms. TOAST categories (n,%) were: undetermined (74, 34.6%), other determined (64, 29.9%), cardioembolism (51, 23.8%), and large artery atherosclerosis (25, 11.7%). Within the undetermined group, 35 patients (16.4%) had incomplete evaluation and 31 (14.5%) had a negative complete evaluation. Incomplete evaluations were nearly twice as common in non-White patients (49% vs 26%), although they comprised only 36% of the cohort. This subgroup was also more often male (66% vs 48%), presented with higher NIHSS (11 vs 6), and received reperfusion less frequently (34% vs 65%). Conclusions: In young adults with LVO-AIS, mechanism-level adjudication uncovered marked heterogeneity beyond TOAST. Nearly half of cryptogenic cases stemmed from incomplete evaluation, disproportionately affecting non-White and male patients, with associated higher severity and lower treatment rates. These findings point to inequities in diagnostic completeness as a modifiable contributor to outcome disparities, highlight the importance of mechanism-based classification for precision care, and underscore the need for systematic, equitable evaluation standards in young-onset stroke workup and practice.