Long-term outcomes after treatment for symptomatic hemorrhage of brain arteriovenous malformations.

OBJECTIVES

To determine the long-term risk factors for rehemorrhage (primary outcome) in patients with ruptured brain arteriovenous malformations (BAVMs) after different treatment strategies and to evaluate the impact of different treatment strategies on lesion obliteration (secondary outcome) over a 10-year follow-up.

METHODS

This single-center retrospective cohort included 645 patients with ruptured brain arteriovenous malformations (BAVMs) treated between 2010 and 2019, with a median follow-up of 96 months. Patients were randomly divided into training (70%) and validation (30%) cohorts to identify and validate predictors of rehemorrhage-free survival (RFS). Cox regression analyses were performed to assess risk factors for RFS, while logistic regression and propensity score matching (PSM) were used to evaluate the impact of treatment strategies on obliteration and rehemorrhage.

RESULTS

Obliteration was observed in 66.0% (384/582) of the patients, and 109 patients (16.9%) experienced rehemorrhage. In the training set, the multivariate Cox regression analysis revealed that, deep location ((hazard ratio (HR) = 2.033, 95% confidence interval (CI) = 1.166-3.545; p = 0.012)/posterior location (HR=2.215, 95% CI=1.213-4.043; p = 0.010), lesion diameter >6 cm (HR = 4.901, 95% CI = 2.046-11.737; p < 0.001), and multiple drain veins (HR = 2.287, 95% CI = 1.207-4.335; p = 0.011), were significant independent risk factors for rehemorrhage-free survival. In the validation set, lesion diameters of 3-6 cm (HR = 4.910, 95% CI = 1.677-14.377; p = 0.004) and >6 cm (HR = 19.416, 95% CI = 5.152-73.173; p < 0.001) remained significant in the multivariate Cox regression analysis. After PSM, univariate logistic analysis revealed that, compared with radiosurgery, preradiosurgery embolization + radiosurgery resulted in a lower non-obliteration rate ((odd ratio) OR = 0.352, 95% CI = 0.134-0.926; p = 0.034).

CONCLUSIONS

Lesion size, lesion location, and multiple draining veins were independently associated with long-term RFS. Preradiosurgical embolization improved the obliteration rate but did not significantly alter the risk of rehemorrhage. These findings directly inform treatment selection for ruptured BAVMs.