Hemorrhage following Complete Arteriovenous Malformation Resection with No Detectable Recurrence: Insights from a 27-Year Registry

BACKGROUND: Although recurrence and de novo formation of arteriovenous malformations (AVMs) have been reported following complete resection, the occurrence of hemorrhage in the same location of an AVM with no detectable lesion (lesion-negative hemorrhage) has not been described after microsurgery. OBJECTIVE: To characterize the incidence and properties of lesion-negative hemorrhage following complete microsurgical resection METHODS: A prospectively maintained registry of AVM patients seen at our institution between 1990 and 2017 was used. Microsurgically treated patients were selected, and the incidence of a lesion-negative hemorrhage was calculated and described with a Kaplan-Meier curve. Baseline characteristics as well as functional outcome at last follow-up were compared between patients with and without a lesion-negative hemorrhage. RESULTS: From a total of 789 AVM patients, 619 (79%) were treated, and 210 out of 619 patients (34%) underwent microsurgery with or without preoperative embolization or radiosurgery. The microsurgically treated cohort was followed up for a mean of 6.1 ± 3.0 yr after surgery with 5 (2.4%) patients experiencing postresection lesion-negative hemorrhage (3.9 per 1000 person-years) at an average of 8.6 ± 9.0 yr following surgery. Follow-up angiograms after hemorrhage (up to 2 mo posthemorrhage) confirmed the absence of a recurrent or de novo AVM in all cases. All patients with a lesion-negative hemorrhage initially presented with rupture before resection (Fisher P =. 066; log-rank P =. 057). The occurrence of a lesion-negative hemorrhage was significantly associated with worse modified Rankin scale scores at last follow-up (P =. 031). CONCLUSION: A lesion-negative hemorrhage can occur following complete microsurgical resection in up to 2.4% of patients. Exploration of possible underlying causes is warranted.