Efficacy and safety of immunotherapy in pediatric malignant brain tumors: a systematic review and exploratory meta-analysis
Abstract
Background Immunotherapy for pediatric brain tumors has expanded rapidly, yet its effects on survival and neurologic safety remain unclear. A systematic review and meta-analysis was conducted to quantify efficacy and toxicity across immunotherapy modalities. Methods Following PRISMA 2020 guidelines, prospective pediatric immunotherapy studies were identified. Random-effects meta-analyses were performed using vaccine-based immunotherapy as the reference. Primary endpoints were overall survival (OS) and 12-month overall survival (OS12); grade ≥3 neurotoxicity was the primary safety outcome. Meta-regression and sensitivity analyses explored modality-specific effects. Results Twenty studies were included. Median OS ranged from 3.6 months in heavily pretreated recurrent cohorts to 17.8 months in selected intratumoral oncolytic virus trials, while OS12 ranged from 8–13% to 75%. Compared with vaccines, non-vaccine immunotherapy did not improve OS (pooled HR 1.14, 95% CI 0.99–1.31; I² 43%) or OS12 (HR 1.13, 95% CI 0.95–1.34; I² 65%). Subgroup analyses showed comparable OS for CAR-T therapies (OS HR 1.09, 95% CI 0.89–1.34) and checkpoint/systemic therapies (OS HR 1.14, 95% CI 0.96–1.36), or oncolytic/viral therapies (OS HR 1.51, 95% CI 0.81–2.84). Severe neurotoxicity differed substantially by modality. No grade ≥3 neurotoxicity occurred in vaccine studies. In contrast, pooled non-vaccine therapies showed a markedly increased risk (OR 19.66, 95% CI 6.78–57.02; I² 60%). CAR-T therapies (OR 57.95, 95% CI 16.15–207.98) and oncolytic/viral approaches (OR 69.41, 95% CI 17.00–283.33) had the highest risk, whereas checkpoint/systemic therapy showed a lower, non-significant increase (OR 2.26, 95% CI 0.77–6.60). Conclusions Across pediatric immunotherapy trials, survival outcomes were comparable to vaccine-based approaches, but neurotoxicity varied markedly by modality. Cellular and oncolytic therapies carry a substantial neurologic risk that must be balanced against uncertain survival benefit in future trial designs.