Hypofractionated radiosurgery for intracranial arteriovenous malformations: a systematic review and meta-analysis.

Abstract

Objective: Obliteration of intracranial arteriovenous malformations (AVMs) with radiosurgery is challenging because high radiation doses prescribed for better effectiveness can increase procedural risks. To minimize adverse effects, fractionation protocols ensure delivery of equivalent or higher total radiation in manageable doses. Hypofractionated stereotactic radiosurgery (HF-SRS), a fractionation strategy using doses exceeding 2 Gy per fraction, requires fewer treatment sessions. The authors conducted a systematic literature review to meta-analyze potential benefits of HF-SRS compared with conventional fractionation regimens.

Methods: PubMed and Embase were searched for literature on "hypofractionated," "radiosurgery," and "arteriovenous malformations." Data including patient and AVM characteristics, procedural details, and outcomes were extracted from eligible studies. Meta-analyses were performed on variables reported by ≥ 3 studies. Subgroup analyses were conducted for photon- and proton-based HF-SRS.

Results: The authors included 29 studies comprising 995 patients. For HF-SRS, digital subtraction angiography (DSA) confirmed a 50.1% obliteration rate (95% CI 35.2%-64.9%, I2 = 85.3%) at 41.2 months (95% CI 35.3-48.1 months, I2 = 95%) of follow-up, with an associated 11.1% (95% CI 8.4%-14.5%, I2 = 28.5%) rupture rate, 5.5% (95% CI 3.5%-8.5%, I2 = 0%) new-onset seizure rate, 10.4% (95% CI 6.8%-15.6%, I2 = 41.7%) radionecrosis, and 6% (95% CI 4%-8.9%, I2 = 13.7%) AVM-related mortality. Studies comparing total HF-SRS doses ≥ 35 Gy with < 35 Gy demonstrated trends toward higher rates of DSA-confirmed obliteration (60.9% vs 47.8%, p = 0.57) and radionecrosis (19.9% vs 8.2%, p = 0.02) with higher doses. Similarly, per-fraction doses ≥ 7 Gy showed higher obliteration trends (53.5% vs 44.5%, p = 0.64) and radionecrosis rates (18% vs 6.2%, p = 0.01) compared with < 7-Gy/fraction doses, indicating a dose-response relationship. Photon-based and proton-based HF-SRS resulted in comparable rates of angiographic obliteration (50.7% vs 48%, p = 0.86), follow-up rupture (10.3% vs 9.6%, p = 0.9), and AVM-related mortality (5.4% vs 7.4%, p = 0.5), despite a greater proportion of higher Spetzler-Martin grade AVMs in the photon-based studies (82.2% vs 55.3%, p < 0.01) with lower doses per fraction (7.2 vs 11.1 Gy, p < 0.01).

Conclusions: HF-SRS is a promising radiosurgical strategy for AVMs with no statistically significant differences in effectiveness and adverse effects between photon- or proton-based therapies. There is notable heterogeneity in the application of HF-SRS for AVMs, necessitating further investigations with consistent reporting and incorporation of prior, adjuvant, and salvage treatments for AVMs.