Acta Neurochir (Wien). 2025 Dec 20. doi: 10.1007/s00701-025-06737-8. Online ahead of print.
ABSTRACT
PURPOSE: to comprehensively and hierarchically assess risk factors for recurrence requiring reoperation (RrR) in chronic subdural hematoma (cSDH) in the era of middle meningeal artery embolization (MMAE).
METHODS: Patients treated for a cSDH from January 2019 to October 2024 at Fondazione Gemelli research hospital were considered for inclusion. Clinical, coagulation, radiological, and treatment factors were recorded. MMAE was performed systematically from October 2022, using polyvinyl alcohol (PVA) particles injected directly from the main trunk of MMA. The dataset comprised 45 quantitative and qualitative variables for each cSDH. Variables showing statistical significance (p-value < 0.05) were selected as covariates in two supervised learning frameworks to predict the RrR (outcome, Y): (i) Classification and Regression Tree (CART) and (ii) Random Forest (RF) classifier.
RESULTS: 500 patients were eligible and 233 were included, resulting in 283 treated cSDHs (mean follow-up: 119 days); 129 underwent adjuvant MMAE. 50 cSDH had a RrR (mean time to recurrence: 47 days), of which 41 (82%) in the non-embolized group and 9 (18%) in the embolized group (p-value < 0.001). Adjuvant embolization was the strongest factor associated with RrR, significantly reducing the risk for reintervention. Markwalder grading scale, preoperative cSDH volume, and platelet count (PLT) are strong predictors in non-embolized patients. A critical PLT cut-off of 229 × 109/L strongly impacts RrR risk for substantial cSDH volumes.
CONCLUSIONS: The present results support the routine use of MMAE and the correction of PLT in relation to cSDH volume.
PMID:41420704 | DOI:10.1007/s00701-025-06737-8
