Occurrence of hemorrhage or radiological progression of residual cerebral arteriovenous malformations after incomplete surgical resection

BACKGROUND: While evidence suggests that incomplete obliteration of AVMs following radiosurgery or embolization does not alter the natural history of those lesions, such evidence is lacking for incomplete treatment with microsurgery. Additionally, there are case reports of spontaneous obliteration of post-surgical residual AVM, which may complicate decision-making. METHODS: A retrospective review was conducted of prospectively maintained institutional databases of AVMs undergoing microsurgical resection at Stanford Medicine from 2000 to 2025 and Queensland Children's Hospital from 2014 to 2025. Clinical and radiological variables were collected from medical record reviews and radiology reports. Residual AVM after microsurgery was defined as the presence of both a residual visible nidus and an early draining vein on digital subtraction angiography. The primary outcomes were hemorrhage from postoperative residual or radiological progression. RESULTS: During the study period, 637 patients with AVMs were treated using microsurgery, with postoperative residual AVM identified in 87 patients (57 ruptured, 30 unruptured) and followed for a total of 108 patient-years prior to retreatment. In the ruptured cohort, 5/57 (9%) patients had an interval hemorrhage prior to retreatment, while no hemorrhages occurred in the previously unruptured cohort. The hemorrhage from postoperative residual in the previous ruptured cohort was 1/57 (2%) at 1 year and 4/57 (7%) at 5 years. 5/57 (9%) of residual nidus in the ruptured cohort and 2/30 (7%) in the unruptured cohort increased in size during follow-up. No cases of spontaneous obliteration were identified. CONCLUSIONS: Incomplete surgical resection of cerebral AVMs does not eliminate the risk of hemorrhage.