小児上衣腫における陽子線治療成績：長期メタアナリシス（PROPEL）

To synthesize existing evidence on pencil-beam and passive-scattering proton therapy outcomes in pediatric intracranial ependymoma and to generate pooled estimates of five-year overall survival (OS), local control (LC), progression-free survival (PFS), and late toxicity profiles.

We performed a random-effects meta-analysis (with fixed- and random-effects estimates converging when between-study heterogeneity was negligible) of eight retrospective single-center cohorts (2000-2025) encompassing 1,100 children treated with 54-60Gy(RBE) proton therapy. Primary endpoints were five-year OS, LC, and PFS. Secondary endpoints included growth hormone deficiency (GHD), hypothyroidism, radionecrosis, and secondary malignancies. Pooled proportions and 95% confidence intervals (CIs) were calculated for each outcome. Between-study heterogeneity was assessed via Cochran's Q and I² statistics. Meta-regression evaluated the influence of surgical extent, proton dose, and delivery technique on oncologic endpoints.

Five-year OS was 83.1% (95% CI 80.7-85.4%), LC 84.9% (95% CI 82.7-87.2%), and PFS 74.1% (95% CI 71.3-76.9%), each with negligible heterogeneity (I² = 0%). Late toxicities were modest: GHD in 21.3% (95% CI 18.8-23.8%), hypothyroidism in 15.0% (95% CI 12.9-17.1%), ototoxicity grade 2 or higher in 5% (95% CI 3.5-6.5%), radionecrosis in 2.0% (95% CI 1.5-3.5%), and secondary malignancy in 0.65% (95% CI 0.17-1.13%). Meta-regression identified subtotal resection as the sole predictor of inferior LC, and PFS (p < 0.01), whereas proton dose range and delivery modality had no significant effect (p > 0.1).

PTB in pediatric intracranial ependymoma achieves excellent five-year OS, LC, and PFS with low rates of serious late effects. Surgical extent remains the principal determinant of outcome, underscoring the need for prospective trials of targeted dose escalation-particularly in cases of subtotal resection-using modern intensity-modulated proton techniques to optimize the therapeutic ratio.