Management and outcome predictors of patients with ruptured deep-seated brain arteriovenous malformations

OBJECTIVE Decision-making for the management of ruptured deep-seated brain arteriovenous malformations (bAVMs) is controversial. This study aimed to shed light on the treatment outcomes of patients with ruptured deep-seated bAVMs. METHODS Data on bAVM patients were retrieved from the authors' institutional database, spanning 1990-2021. The outcomes were annual hemorrhage risk (before and after intervention), number of follow-up hemorrhages, bAVM obliteration, poor modified Rankin Scale (mRS) score (i.e., mRS score > 2), worsened mRS score, and mortality. Multivariable Cox and logistic regression analyses were conducted to determine predictors of time-to-event and categorical outcomes, respectively. RESULTS Of the 1066 patients in the database with brain bAVM, 177 patients harboring ruptured deep-seated bAVMs were included. The pretreatment annual hemorrhage risk was 8.24%, and the posttreatment risk was lowered to 1.65%. In multivariable Cox regression analysis, a prenidal aneurysm (HR 2.388, 95% CI 1.057-5.398; p = 0.036) was associated with a higher risk of follow-up hemorrhage, while definitive treatment (i.e., either surgery or radiosurgery vs endovascular embolization or conservative management) (HR 0.267, 95% CI 0.118-0.602; p = 0.001) was associated with a lower risk of follow-up hemorrhage. In multivariable logistic regression analysis, Spetzler-Martin grades IV and V (OR 0.404, 95% CI 0.171-0.917; p = 0.033) and brainstem arteriovenous malformation (AVM) (OR 0.325, 95% CI 0.128-0.778; p = 0.014) were associated with lower odds of obliteration, while definitive treatment (OR 8.864, 95% CI 3.604-25.399; p = 0.008) was associated with higher obliteration odds. Controlling for baseline mRS score, cerebellar AVM (OR 0.286, 95% CI 0.098-0.731; p = 0.013) and definitive treatment (OR 0.361, 95% CI 0.160-0.807; p = 0.013) were associated with lower odds of a poor mRS score, and definitive treatment (OR 0.208, 95% CI 0.076-0.553; p = 0.001) was associated with lower odds of a worsened mRS score. Furthermore, smoking (OR 6.068, 95% CI 1.531-25.581; p = 0.01) and definitive treatment (OR 0.101, 95% CI 0.024-0.361; p = 0.007) were associated with higher and lower mortality odds, respectively. CONCLUSIONS A definitive treatment strategy seems to be beneficial in achieving higher obliteration and lower hemorrhage rates while decreasing the odds of a poor mRS score, worsened mRS score, and mortality. In this category of patients, prenidal aneurysms warrant treatment, and smoking cessation should be encouraged.