Robotic radiosurgery for the treatment of pediatric arteriovenous malformations.

Objective: Pediatric intracranial arteriovenous malformations (AVMs) have a greater cumulative lifetime risk of rupture than those in adults. Although obliteration after radiation occurs in a dose-dependent manner, increasing radiation doses must be balanced against the risk of adverse radiation effects (AREs). The authors aimed to assess the efficacy of robotic radiosurgery for pediatric AVMs.

Methods: The authors performed a retrospective review of pediatric patients with AVMs at a single institution who underwent robotic radiosurgery between 2005 and 2021 with one of 3 radiosurgery dosing schedules: 1) single-stage unfractionated (SSU), 2) single-stage fractionated (SSF), and 3) volumetrically multistaged (VMS) treatment. Cox proportional hazards regression was performed to identify predictors of AREs and obliteration.

Results: Ninety-five patients with 100 intracranial AVMs were identified. Median (range) follow-up time was 4.5 (1.8-15.2) years. Forty-four (46.3%) presented with ruptured AVMs. The mean ± SD AVM volume was 10.0 ± 11.88 cm3. A plurality of AVMs were Spetzler-Martin grade III (36.2%). The overall rate of total obliteration was 52.6% (78.8% of SSU-treated, 24.2% of SSF-treated, 10% of VMS-treated patients) with a median (range) obliteration time of 3.25 (2.8-4.1) years. Partial obliteration was achieved in 23.2% of patients. In the univariate analysis, the higher obliteration rate was associated with small volume (HR 0.876, 95% CI 0.812-0.945) (p = 0.001), no prior embolization (HR 0.472, 95% CI 0.254-0.876) (p = 0.017), lower Spetzler-Martin grade (HR 0.437, 95% CI 0.320-0.597) (p ≤ 0.001), and higher single-fraction equivalent dose (HR 1.160, 95% CI 1.020-1.198) (p = 0.015). Pretreatment hemorrhage was found in 51 patients (59.6% of SSU-treated, 45.5% of SSF-treated, and 50% of VMS-treated patients). Thirteen patients experienced posttreatment hemorrhage (3.8% of SSU-treated, 12% of SSF-treated, and 60% of VMS-treated patients). AREs were found afterward in 31.6% of patients. The correlations of male sex (HR 0.447, 95% CI 0.199-1.004) (p = 0.051) and volume of brain tissue that received a single-fraction equivalent dose of 12 Gy or greater (HR 1.020, 95% CI 1.000-1.041) (p = 0.053) with AREs did not reach significance.

Conclusions: SSU treatment was effective for treating smaller AVMs with an obliteration rate of 79%. Although SSF treatment was less effective in achieving total obliteration (24%), this approach significantly reduced the posttreatment hemorrhage rate by nearly 75% (46% of patients had pretreatment hemorrhage vs 12% with posttreatment hemorrhage). Unfortunately, only 10% of AVMs in the VMS cohort were obliterated and posttreatment hemorrhage rates were not reduced.