Blood–brain barrier vulnerability and microcirculatory dysfunction as predictors of hemorrhagic transformation after endovascular treatment in acute ischemic stroke

Background Hemorrhagic transformation (HT) after endovascular treatment (EVT) for acute ischemic stroke (AIS) remains a major clinical challenge, particularly in populations vulnerable to vascular and endothelial injury. Beyond conventional clinical predictors, microcirculatory dysfunction and blood–brain barrier (BBB) vulnerability may critically influence the risk of hemorrhagic complications following reperfusion. However, existing prediction models often rely on isolated imaging markers or basic clinical variables, failing to capture the integrated pathophysiological processes linking collateral circulation, microvascular perfusion, and BBB integrity. Methods This retrospective study analyzed 202 consecutive AIS patients undergoing EVT (2021–2024). Collateral circulation was assessed via ASITN/SIR scale; CT perfusion (CTP)-derived parameters (rCBF, rCBV, rMTT, rTTP, rPS) quantified microcirculation and BBB integrity. Clinical/laboratory variables (including serum globulin) were collected. Multivariate logistic regression identified HT/parenchymal hematoma (PH) predictors; a nomogram was validated via ROC, calibration, bootstrap, DCA, and CIC. HT subtypes were correlated with functional outcomes. Results HT occurred in 82 patients (40.6%, 37 PH cases). Higher ASITN/SIR grade (OR = 0.616), rCBF (OR = 0.053), and rCBV (OR = 0.204) were protective; rPS (OR = 2.624) and serum globulin (OR = 1.138) were risk factors for HT. The multimodal model (integrating three mechanistic pathways) showed excellent discrimination (AUC = 0.867) and superior net clinical benefit. For PH, ASITN/SIR was protective (AUC = 0.745). HT patients had poorer outcomes ( p 0.001), with PH worse than HI ( p = 0.046). Conclusion This study demonstrates that BBB vulnerability, microcirculatory dysfunction, and collateral status are key determinants of hemorrhagic transformation after endovascular treatment in acute ischemic stroke. By integrating these interrelated vascular and endothelial mechanisms, the proposed model provides insight into susceptibility to reperfusion-related hemorrhagic injury and offers a practical framework for individualized risk stratification in the peri-procedural setting.