Outcomes of single-session gamma knife radiosurgery for deep-seated arteriovenous malformations.

Background: Deep-seated arteriovenous malformations (AVMs) pose unique surgical challenges due to their proximity to critical structures. Gamma knife radiosurgery (GKRS) thus serves a promising alternative that may reduce treatment-related morbidity of deep-seated AVMs. However, the efficacy of GKRS in achieving obliteration in deep AVMs is not established. Therefore, we examined our experience with GKRS in deep-seated AVMs.

Methods: We retrospectively reviewed all deep-seated (thalamic, basal ganglia and brainstem AVMs) that underwent GKRS at our center between 2006 and 2022. Factors that predicted obliteration were assessed, and the incidence of various long-term complications was recorded.

Results: The cohort comprised 137 patients, 74 males (54 %) and 63 females (46 %). The mean age was 25.1 ± 11.9 years. The mean dose delivered was 21.7 ± 2.6 Gy. The median volume of AVMs in our cohort was 3.2 cc (IQR: 1.5, 6.3 cc). The median follow-up was 36 months (IQR: 24, 42 months), following which 66 AVMs (48.2 %) were obliterated. Radiation-induced edema occurred in 8 cases (5.8 %), 7 of whom were symptomatic with neurological deficits (5.1 %). However, all patients improved on further follow-up with no residual neurological deficits. The median volume was significantly greater among cases with residual AVM (3.45 cc, IQR: 1.77, 6.85 cc) than those that were obliterated (2.77 cc, IQR: 1.09, 5.10 cc, p = 0.049). The mean Pollock Flickinger score was significantly greater among non-obliterated cases, as was the mean dose delivered (22.3 ± 2.7 Gy vs. 21.5 ± 2.8 Gy, p = 0.044). Multivariate analysis revealed that a nidus volume less than 3 cc was the only factor that predicted obliteration of the nidus (HR: 4.994, 2.139, 9.166, p = 0.016).

Conclusions: GKRS is an effective and safe treatment option for deep-seated AVMs. Obliteration occurred in 48.2% of cases, with complications in 5.8%. Nidus volume, rather than other clinical or radiological factors, appears to be the most significant predictor of successful obliteration. Despite the challenges posed by the complex anatomy of deep AVMs, including variable venous drainage and the potential for radiation-induced edema, GKRS is similarly effective in AVMs located in the thalamus, basal ganglia, and brainstem.