Home | Publications | CHF+26

Heterogeneity of Treatment Effects of Preoperative Embolization in Patients With Brain Arteriovenous Malformations

MCML Authors

Link to Profile Stefan Feuerriegel

Stefan Feuerriegel

Prof. Dr.

Core PI

Artificial Intelligence in Management

Abstract

Background/Introduction: Preoperative embolization is widely used as an adjunct to microsurgical resection of brain arteriovenous malformations (AVMs). While a recent meta-analysis found no significant difference in functional outcomes, subgroup studies and clinical experience suggest that the treatment effect of preoperative embolization may vary across different patients.<br>Objective/Purpose: To analyze the treatment effect heterogeneity of preoperative embolization in patients with AVMs and identify subgroups who benefit from preoperative embolization.<br>Methods/Case Description: A total of 487 patients from the Multicenter International Study for Treatment of Brain AVMs consortium were included. We developed a causal forest model to estimate individualized treatment effects (ITEs) in patients who underwent preoperative embolization (n=230) compared to microsurgery alone (n=257). The ITE was defined as the difference in probability of favorable outcome (modified Rankin Scale (mRS ≤ 2) at last follow-up) between the combined treatment group and the microsurgery group.<br>Results/Findings: There was a substantial heterogeneity in ITEs, ranging between −38% and +31%. The average treatment effect was −1% (95% confidence interval: −7 to +5%). A total of 41.5% of patients had a predicted positive ITE, and 11.5% had an ITE > 10%. Our treatment rule targeting the top 30% of patients by predicted ITE improved favorable outcome rates from 83.4% to 90.1% (p=0.04). SHAP analysis identified several factors associated with treatment benefit, including higher age, larger nidus volume, higher preoperative mRS, multiple feeders and diffuse morphology.<br>Discussion/Conclusions/Learning: Points Our study demonstrated a substantial heterogeneity in the treatment effect of preoperative embolization, despite the absence of an average treatment effect on a population level. Our findings underscore the role of preoperative embolization in selected patients with larger, higher-flow AVMs.

article CHF+26

Journal of NeuroInterventional Surgery

18.A32-A33. Sep. 2026.
Top Journal

Authors

O. Ciobanu-Caraus • P. Heesen • S. Feuerriegel • N. Adeeb • M. Bendszus • M. A. Möhlenbruch • J.-K. Burkhardt • S. Kandregula • V. M. Srinivasan

Links

DOI

Research Area

 A1 | Statistical Foundations & Explainability

BibTeXKey: CHF+26

#p-feuerriegel
Back to Top