Abstract A026: Posterior circulation Clot Burden Score (pc-CBS) predicts clinical outcome after acute ischemic stroke related to posterior circulation occlusions : An externally validated radiological score

Introduction: Posterior circulation strokes (PCS) account for one-quarter of ischemic strokes and differ from anterior circulation events in presentation, vascular anatomy, and treatment implications. Prognostic tools for PCS are scarce, and existing scores apply mainly to basilar artery occlusion, despite their potential to guide acute treatment and trial design. We aimed to develop and externally validate the Posterior Circulation Clot Burden Score (pc-CBS), a CTA/MRA-based tool to predict 3-month functional outcome across the entire posterior circulation. Methods: This investigator-initiated, retrospective cohort study included 488 consecutive PCS patients with CTA/MRA-confirmed occlusion from the ASTRAL registry (Lausanne, Switzerland) in the derivation/internal validation cohort. External validation used individual-level data from 1,340 patients across 21 stroke centers in 3 continents. Eight arterial segments (extracranial/intracranial vertebral, proximal/mid/distal basilar, P1/P2 posterior cerebral, posterior communicating arteries) were graded as patent or occluded/hairline (<1mm) and interrater reliability was tested both in CTA and MRA. The segments were then weighted via ordinal logistic regression for 3-month modified Rankin Scale (mRS). Predictive accuracy was assessed with Harrell’s C-statistics; independent contribution via likelihood ratio test (LRT). The study was registered on ClinicalTrials.gov as NCT07122934. Results: Interrater agreement was almost perfect for both CTA (n=60, κ=0.89) and MRA (n=60, κ=0.89). The 10-point pc-CBS assigns full points for complete vertebrobasilar patency and deducts for each affected segment. C-statistic were 0.70 (95%CI 0.60–0.80) for internal and 0.72 (95%CI 0.69–0.74) validation. Each 1-point decrease increased odds of unfavorable outcome (mRS 3–6) by 14% (internal) and 8% (external). Adding nine clinical outcome predictors (age, sex, prestroke mRS, NIHSS, ASPECTS, glucose, decreased consciousness, IVT, EVT) improved C-statistic to 0.84 (internal) and 0.85 (external), with pc-CBS retaining independent predictive value (LRT=6.4 and 6.3; p=0.01). Conclusions: pc-CBS is a reproducible CTA/MRA-based score for any occlusive vertebrobasilar stroke, providing independent prognostic value for functional 3-month outcome. Its simplicity, high reliability, and external validation support its use in clinical decision-making and trial stratification for PCS.