2363 - Local Control After Stereotactic Radiosurgery for Pediatric Brain Tumors: Updated Systematic Review and Meta-Analysis

Purpose/Objective(s): Pediatric cranial stereotactic radiosurgery (SRS) experience is rapidly expanding, yet pooled efficacy benchmarks may be inflated by earlier, lower-volume datasets. We updated the most recent pediatric SRS meta-analysis by incorporating approximately three additional years of published evidence and hypothesized that continued evidence accrual would refine and not necessarily improve pooled local control estimates across tumor histologies and arteriovenous malformations (AVMs).

Materials/Methods: Per PRISMA guidelines, we updated the prior meta-analysis with a PubMed/MEDLINE search (2022 - 2025) plus manual reference review. Eligible studies were English-language original reports including =5 patients =18 years treated with cranial SRS (any platform, any fractionation within SRS paradigms) for benign/malignant brain tumors or AVMs, reporting post-SRS outcomes. Overlapping cohorts were managed by retaining the largest or most comprehensive dataset. The primary endpoint was pooled local control, for tumors, or obliteration for AVMs, with 95% confidence intervals using random-effects models (R, metafor package). Heterogeneity was assessed with I². Updated pooled estimates were compared with previously reported pooled values by histology.

Results: Seventeen new studies met inclusion criteria, increasing included studies from 67 to 84. Overall pooled local control/obliteration declined from 0.69 (95% CI 0.65-0.74) to 0.66 (95% CI 0.61-0.70), with high heterogeneity (I²=84.3%). By histology, pooled local control decreased for medulloblastoma 0.73 to 0.70 (95% CI 0.52-0.84; 3 studies), craniopharyngioma 0.89 to 0.81 (95% CI 0.36-0.97; 5 studies), ependymoma 0.75 to 0.56 (95% CI 0.25-0.83; 6 studies), and glioma 0.89 to 0.81 (95% CI 0.60-0.93; 6 studies). In contrast, AVM obliteration remained stable at 0.65 to 0.64 (95% CI 0.60-0.69) despite a marked increase in included AVM studies (50 to 64).

Conclusion: Incorporation of approximately three additional years of pediatric SRS literature produced systematically lower pooled local control estimates for most tumor histologies, while high-volume AVM indications demonstrated stable pooled obliteration. These findings suggest earlier pooled analyses likely overestimated local control and that continued evidence accrual primarily refines efficacy benchmarks rather than improving observed outcomes, informing realistic counseling, trial design, and quality metrics in pediatric SRS.