Stereotactic radiosurgery for arteriovenous malformations, Part 3: outcome predictors and risks after repeat radiosurgery

The object of this study was to evaluate the outcomes and risks of repeat stereotactic radiosurgery (SRS) for incompletely obliterated cerebral arteriovenous malformations (AVMs). Between 1987 and 2006, Gamma Knife surgery was performed in 996 patients with AVMs. During this period, repeat SRS was performed in 105 patients who had incompletely obliterated AVMs at a median of 40.9 months after initial SRS (range 27.5-139 months). The median AVM target volume was 6.4 cm(3) (range 0.2-26.3 cm(3)) at initial SRS but was reduced to 2.3 cm(3) (range 0.1-18.2 cm(3)) at the time of the second procedure. The median margin dose at both initial SRS and repeat SRS was 18 Gy. The actuarial rate of total obliteration by angiography or MR imaging after repeat SRS was 35%, 68%, 77%, and 80% at 3, 4, 5, and 10 years, respectively. The median time to complete angiographic or MR imaging obliteration after repeat SRS was 39 months. Factors associated with a higher rate of AVM obliteration were smaller residual AVM target volume (p = 0.038) and a volume reduction of 50% or more after the initial procedure (p = 0.014). Seven patients (7%) had a hemorrhage in the interval between initial SRS and repeat SRS. Seventeen patients (16%) had hemorrhage after repeat SRS and 6 patients died. The cumulative actuarial rates of new AVM hemorrhage after repeat SRS were 1.9%, 8.1%, 10.1%, 10.1%, and 22.4% at 1, 2, 3, 5, and 10 years, respectively, which translate to annual hemorrhage rates of 4.05% and 1.79% of patients developing new post-repeat-SRS hemorrhages per year for Years 0-2 and 2-10 following repeat SRS. Factors associated with a higher risk of hemorrhage after repeat SRS were a greater number of prior hemorrhages (p = 0.008), larger AVM target volume at initial SRS (p = 0.010), larger target volume at repeat SRS (p = 0.002), initial AVM volume reduction less than 50% (p = 0.019), and a higher Pollock-Flickinger score (p = 0.010). Symptomatic adverse radiation effects developed in 5 patients (4.8%) after initial SRS and in 10 patients (9.5%) after repeat SRS. Prior embolization (p = 0.022) and a higher Spetzler-Martin grade (p = 0.004) were significantly associated with higher rates of adverse radiation effects after repeat SRS. Delayed cyst formation occurred in 5 patients (4.8%) at a median of 108 months after repeat SRS (range 47-184 months). Repeat SRS for incompletely obliterated AVMs increases the eventual obliteration rate. Hemorrhage after obliteration did not occur in thi