The CSDH Grading Scheme for Middle Meningeal Artery Embolization: Stratified Risk of Surgical Rescue.

Abstract

Prospective stratification of surgical rescue risk after middle meningeal artery embolization (MMAE) for chronic subdural hematomas (CSDHs) can facilitate a follow-up approach and optimize patient expectations. In this study, we have identified factors, identifiable at the time of presentation, for MMAE failure in a summative fashion using a grading scheme. The scheme includes 4 domains defined on presentation (1 point for each): Coagulation/platelet abnormality, Scan findings (midline shift >6 mm, hemispheric hematoma with thickness ≥15 mm, or acute components), Deficits (attributable focal neurological deficits [FNDs]), History (comorbidities requiring antithrombotic agents during the reabsorption of hematomas), making up "CSDH." The predictive performance for risk of surgical rescue was assessed using calibration curve and C-index in an institutional cohort. In a test dataset of 100 cases, its accuracy was further compared with 2 regression models. Of 298 CSDHs undergoing stand-alone MMAE, 9.4% required surgical rescue (n = 28): 5.4% if Grade 1 (0-1 point, 87% cases), 29% if Grade 2 (2 points, 11% of cases), and 67% if Grade 3 (3 points or more, 2% of cases). The CSDH score demonstrated both external accuracy and discrimination (C-index 0.81, 95% CI 0.74-0.87). Accuracy persisted in the test data (area under the receiver-operating characteristic curves [AUROC] 0.85, 95% CI 0.74-0.95) compared with a least absolute shrinkage and selection operator regression model (FNDs, hemispheric hematoma, acute component, midline shift >6 mm, age, and hemodialysis; AUROC 0.79, 95% CI 0.66-0.92, P = .46) and a logistic regression model (FNDs, hemispheric hematoma, acute component, and thrombocytopenia; AUROC 0.75, 95% CI 0.61-0.89, P = .22). Comprehensive assessment of CSDH imaging features, neurological deficits, and predisposing hemorrhagic factors may delineate appropriate candidates for stand-alone MMAE.