OBJECTIVE: The risk of delayed hemorrhage occurring greater than 2 years after treatment in brain arteriovenous malformations (AVMs) rarely is reported. In this study, we compare the risk of delayed hemorrhage across different treatment modalities. METHODS: We performed a retrospective chart review of treated patients with a single intracranial AVM seen at our institution from 1990 to 2013. Delayed hemorrhage was defined as hemorrhage occurring at least 2 years after last treatment. Survival analysis was used to assess risk of delayed hemorrhage by treatment modalities. RESULTS: Our study included 420 patients. Spetzler-Martin grades were as follows: I (12.6%), II (36.2%), III (32.6%), IV (15.0%), and V (3.6%). Average follow-up time is 5.1 years. Twenty-two patients (5.2%) were found to have 28 delayed hemorrhages. Average interval between last treatment and delayed hemorrhage was 7.6 years, with the longest being 24.2 years. Proportions of delayed hemorrhages by treatment modalities were as follows: surgery ± embolization (group I, 9.1%), radiosurgery ± embolization (group II, 63.6%), embolization only (group III, 22.7%), and surgery + radiosurgery ± embolization (group IV, 4.5%). Annualized hemorrhage risk after 2 years for each treatment group was as follows: group I (0.4%), group II (1.2%), group III (3.7%), and group IV (1.7%). Survival analysis demonstrated lowest risk of delayed hemorrhage for group 1 (P < 0.01). CONCLUSIONS: This study is the first to compare the risk of delayed hemorrhage across different treatment modalities. Surgical resection is associated with the lowest risk for delayed hemorrhage compared with other treatment modalities. Patients with partially embolized AVMs should seek timely definitive treatment to decrease the risk of delayed hemorrhage.
OBJECTIVE: The risk of delayed hemorrhage occurring greater than 2 years after treatment in brain arteriovenous malformations (AVMs) rarely is reported. In this study, we compare the risk of delayed hemorrhage across different treatment modalities. METHODS: We performed a retrospective chart review of treated patients with a single intracranial AVM seen at our institution from 1990 to 2013. Delayed hemorrhage was defined as hemorrhage occurring at least 2 years after last treatment. Survival analysis was used to assess risk of delayed hemorrhage by treatment modalities. RESULTS: Our study included 420 patients. Spetzler-Martin grades were as follows: I (12.6%), II (36.2%), III (32.6%), IV (15.0%), and V (3.6%). Average follow-up time is 5.1 years. Twenty-two patients (5.2%) were found to have 28 delayed hemorrhages. Average interval between last treatment and delayed hemorrhage was 7.6 years, with the longest being 24.2 years. Proportions of delayed hemorrhages by treatment modalities were as follows: surgery ± embolization (group I, 9.1%), radiosurgery ± embolization (group II, 63.6%), embolization only (group III, 22.7%), and surgery + radiosurgery ± embolization (group IV, 4.5%). Annualized hemorrhage risk after 2 years for each treatment group was as follows: group I (0.4%), group II (1.2%), group III (3.7%), and group IV (1.7%). Survival analysis demonstrated lowest risk of delayed hemorrhage for group 1 (P < 0.01). CONCLUSIONS: This study is the first to compare the risk of delayed hemorrhage across different treatment modalities. Surgical resection is associated with the lowest risk for delayed hemorrhage compared with other treatment modalities. Patients with partially embolized AVMs should seek timely definitive treatment to decrease the risk of delayed hemorrhage.