Brain Behav. 2026 Oct;16(10):e71676. doi: 10.1002/brb3.71676.
ABSTRACT
BACKGROUND: Dural arteriovenous fistulas (dAVFs) carry a significant risk of intracranial hemorrhage (ICH), yet most literature has focused on angiographic features as predictors. The influence of modifiable cardiovascular risk factors (CVRFs) such as hypertension, diabetes, smoking, alcohol use, dyslipidemia, obesity, and use of antithrombotic/anticoagulant agents on hemorrhagic presentation or follow-up outcomes remains largely unexplored.
METHODS: We conducted a systematic review according to PRISMA 2020 guidelines. PubMed, Embase, and Scopus were searched through August 2025 for observational studies including at least 10 patients and reporting CVRFs in dAVFs in relation to hemorrhagic presentation or follow-up hemorrhage. Crude effect estimates were extracted or reconstructed where possible. Quantitative pooling was considered exploratory and performed only when at least two studies reported reconstructable data for the same exposure and outcome.
RESULTS: Of 918 records screened, 31 full texts were reviewed, and 3 studies met inclusion criteria, totaling 543 patients. Two studies (n = 281) provided data for exploratory quantitative synthesis, whereas one hemorrhage-only registry cohort (n = 262) was summarized narratively for post-hemorrhage outcomes. Hypertension was present in 89 of 281 patients (31.7%) and hyperlipidemia in 61 of 281 (21.7%). Hypertension was not associated with hemorrhagic presentation (OR 0.89, 95% CI 0.47-1.68; I2 = 0%). Hyperlipidemia estimates were highly heterogeneous (I2 = 93.6%) and unsuitable for clinical inference. Findings regarding smoking, alcohol consumption, statins, and ACE inhibitors were sparse, unadjusted, and inconsistent.
CONCLUSIONS: Evidence regarding modifiable cardiovascular risk factors in dAVFs is limited, retrospective, and vulnerable to confounding. Angiographic features remain the most reliable predictors of hemorrhagic presentation. Current systemic and therapeutic associations are exploratory and require validation in prospective multicenter studies.
PMID:42855847 | DOI:10.1002/brb3.71676