Quantitative validation of a CT perfusion-derived corrected CBV index as a surrogate of angiographic collateral status(ASITN/SIR grading) in anterior circulation ischemia.

Collateral circulation critically influences tissue perfusion not only in hyperacute large-vessel occlusion (LVO) stroke but also in non-intervened anterior circulation ischemia, including ICAS, ECAS, moyamoya disease, and selected LVO cases not undergoing emergent endovascular therapy (EVT). In such patients, elective diagnostic DSA remains the reference standard for evaluating collateral status and cerebral hemodynamic reserve. However, noninvasive CT perfusion (CTP)–derived indices have not been systematically validated against angiographic collateral grading in this mixed ischemia population. To quantitatively validate a corrected cerebral blood volume (CBV) index, derived from CTP, as a surrogate marker of angiographic collateral status (ASITN/SIR scale) in patients with anterior circulation ischemia who underwent elective DSA. We retrospectively analyzed 123 patients presenting with anterior circulation ischemia who did not receive emergent endovascular treatment and subsequently underwent elective DSA. The corrected CBV index was defined as mean CBV within the Tmax > 6 s region normalized to contralateral cortical CBV. Collateral status was graded using ASITN/SIR by two blinded neurointerventionalists. ROC analyses, DeLong pairwise comparisons, bootstrap confidence intervals, logistic regression, and 10-fold cross-validation were performed. Subgroup analyses were conducted across etiologies (large-vessel occlusion LVO without EVT, extracranial atherosclerotic stenosis (ECAS), intracranial atherosclerotic stenosis (ICAS), moyamoya disease (MMD)). The corrected CBV index demonstrated the highest discriminative performance for good versus poor collaterals (AUC 0.83; 95% CI 0.75–0.90), outperforming conventional CBV index (AUC 0.81), PVT (0.60), PRR (0.53), and HIR (0.48). Cross-validation confirmed robustness (mean AUC 0.84). Subgroup analyses showed consistent accuracy across LVO, ECAS, ICAS, and moyamoya, with higher optimal thresholds in moyamoya reflecting distinct collateral physiology. Among the evaluated perfusion metrics, the corrected CBV index demonstrated consistently favorable overall performance. The corrected CBV index provides a quantitative, non-invasive marker of cerebral hemodynamics that correlates with angiographic collateral status across diverse vascular pathologies. This index may assist in assessing hemodynamic reserve, identifying collateral insufficiency, and supporting revascularization strategies in selected clinical contexts involving ICAS, ECAS, moyamoya disease, and non-intervened LVO.