RESUMEN
BACKGROUND: Repeat stereotactic radiosurgery (SRS) for persistent cerebral arteriovenous malformation (AVM) has generally favorable patient outcomes. However, reporting studies are limited by small patient numbers and single-institution biases. The purpose of this study was to provide the combined experience of multiple centers, in an effort to fully define the role of repeat SRS for patients with arteriovenous malformation. METHODS: This multicenter, retrospective cohort study included patients treated with repeat, single-fraction SRS between 1987 and 2022. Follow-up began at repeat SRS. The primary outcome was a favorable patient outcome, defined as a composite of nidus obliteration in the absence of hemorrhage or radiation-induced neurological deterioration. Secondary outcomes were obliteration, hemorrhage risk, and symptomatic radiation-induced changes. Competing risk analysis was performed to compute yearly rates and identify predictors for each outcome. RESULTS: The cohort comprised 505 patients (254 [50.3%] males; median [interquartile range] age, 34 [15] years) from 14 centers. The median clinical and magnetic resonance imaging follow-up was 52 (interquartile range, 61) and 47 (interquartile range, 52) months, respectively. At last follow-up, favorable outcome was achieved by 268 (53.1%) patients (5-year probability, 50% [95% CI, 45%-55%]) and obliteration by 300 (59.4%) patients (5-year probability, 56% [95% CI, 51%-61%]). Twenty-eight patients (5.6%) experienced post-SRS hemorrhage with an annual incidence rate of 1.38 per 100 patient-years. Symptomatic radiation-induced changes were evident in 28 (5.6%) patients, with most occurring in the first 3 years. Larger nidus volumes (between 2 and 4 cm3, subdistribution hazard, 0.61 [95% CI, 0.44-0.86]; P=0.005; >4 cm3, subdistribution hazard, 0.47 [95% CI, 0.32-0.7]; P<0.001) and brainstem/basal ganglia involvement (subdistribution hazard, 0.6 [95% CI, 0.45-0.81]; P<0.001) were associated with reduced probability of favorable outcome. CONCLUSIONS: Repeat SRS confers reasonable obliteration rates with a low complication risk. With most complications occurring in the first 3 years, extending the latency period to 5 years generally increases the rate of favorable patient outcomes and reduces the necessity of a third intervention.