Proton therapy in children under three shows 85% two-year survival but substantial late toxicity

Among 76 children aged three years or younger, proton therapy achieved 85.2% two-year local control, with 14.9% severe late toxicity.

KEY POINTS

  • This retrospective study from The Christie NHS Foundation Trust, Manchester, evaluated 76 consecutive children aged ≤3 years treated with proton beam therapy (PBT) through the UK national proton pathway between December 2018 and January 2025. Median age at treatment was 28.1 months (range 9.4–36.0). Diagnoses included 33 CNS tumors, predominantly ependymoma (n=20), and 43 non-CNS malignancies, including sarcoma (n=20), neuroblastoma (n=12), and renal tumors (n=11).
  • All patients received robustly optimized pencil-beam scanning PBT with daily image guidance under general anesthesia. CNS tumors were predominantly treated postoperatively, with a median prescription of 59.4 Gy(RBE) in 33 fractions; non-CNS tumors received a median 36.0 Gy(RBE) in 20 fractions, generally following induction chemotherapy. Treatment was completed as prescribed in 72/76 children (94.7%). During PBT, 23.7% required hospitalization and 38.2% required enteral feeding support, although most feeding tubes had been placed before irradiation.
  • Overall survival and local control were encouraging but varied substantially by diagnosis. With median follow-up of 2.9 years, one-, two-, and three-year OS were 87.9%, 85.1%, and 81.1%, respectively. Two-year progression-free survival was 65.5% (95% CI 54.8–78.3%), while two-year local control was 85.2% (95% CI 76.1–95.5%). The survival analyses included 75 patients, excluding one child who discontinued treatment after five fractions because of an anesthetic complication.
  • All nine local recurrences occurred within the high-dose target volume, suggesting that geographic miss was not the predominant mechanism of treatment failure. In the ependymoma subgroup, three of 19 evaluable patients developed local recurrence, all after salvage irradiation for previously recurrent disease; two-year local control was 79.0%. CNS embryonal tumors had 91.7% two-year local control but only 40.0% two-year PFS, reflecting frequent progression outside the primary treatment site.
  • Severe late toxicity occurred in 11/74 evaluable children (14.9%). Nine experienced grade 3 events, six grade 4 events, and one grade 5 event, with multiple severe toxicities in some patients. Median time to grade ≥3 late toxicity was 1.9 years (range 0.4–5.2), while two-year freedom from severe late toxicity was 90.4%. The short observation period means that neurocognitive, endocrine, growth-related, and other delayed complications remain incompletely characterized.
  • One child died from treatment-related brainstem radionecrosis despite meeting the institutional dose constraints. The patient had posterior fossa ependymoma and received 54 Gy(RBE) in 30 fractions at 17 months of age. Brainstem worst-case D0.1cc was 54.8 Gy(RBE) and D50% was 50.3 Gy(RBE). Within the necrotic region, mean dose-averaged LET was 3.7 keV/μm, with a maximum of 5.0 keV/μm. The necrosis became symptomatic four months after PBT, and the child died 14 months after treatment without disease progression.
  • The study establishes feasibility but not superiority over photon radiotherapy. Patients were selected through a national multidisciplinary pathway, and high-risk disease biology, previous surgery, and concurrent systemic treatment influenced outcomes. No photon-treated comparison cohort was included, while late effects were assessed over a relatively short period for children expected to survive many decades. The fatal brainstem event also illustrates that conventional fixed-RBE dose constraints may not fully characterize risk in very young patients.

CLINICAL TAKEAWAY

Proton therapy can be delivered to infants and very young children with acceptable treatment completion and encouraging early disease control, but severe late toxicity remains a substantial concern. The fatal brainstem injury despite compliant dose constraints emphasizes the importance of age, tumor location, and potentially variable biological effectiveness when evaluating risk. Long-term surveillance and prospective pediatric proton data are essential before reduced integral dose can be equated with improved clinical safety.

SOURCE

Practical Radiation Oncology

Browse more research Suggest a correction