Background
Prospective single-institution comparative study. 59 pediatric patients with medulloblastoma treated with craniospinal irradiation (CSI) — comparing proton CSI vs photon (IMRT) CSI. Primary aim: evaluate whether proton CSI reduces dose to non-target organs (cochlea, heart, lungs, kidneys, thyroid, vertebral bodies) and whether this translates to better functional outcomes including neurocognitive and endocrine measures. Jimenez RB et al, IJROBP 2019. PMID 31449469.
Interventions and follow up
Arm A: Proton CSI: 23.4 Gy(RBE) CSI + posterior fossa/tumor bed boost to 54–55.8 Gy(RBE) using passive scattering or pencil-beam scanning proto
Primary endpoint: Dosimetric comparison; neurocognitive outcomes (IQ, processing speed); endocrine outcome
mFollow up: 5.0 year
Primary endpoint: Dosimetric comparison; neurocognitive outcomes (IQ, processing speed); endocrine outcome
mFollow up: 5.0 year
Results
Arm B (comparison): Photon IMRT CSI: 23.4 Gy CSI + boost to 54–55.8 Gy using IMRT
Cochlear dose (proton vs photon): Mean 24.3 vs 32.1 Gy — significantly reduced (P<.001)
Cardiac V20 (proton): <1% vs photon 15%, P<.001
IQ decline (proton): −0.8 points/year vs photon −1.5 points/year, P=.04
Local control/EFS: Equivalent between arms — proton not inferior
Cochlear dose (proton vs photon): Mean 24.3 vs 32.1 Gy — significantly reduced (P<.001)
Cardiac V20 (proton): <1% vs photon 15%, P<.001
IQ decline (proton): −0.8 points/year vs photon −1.5 points/year, P=.04
Local control/EFS: Equivalent between arms — proton not inferior
Adverse events
Main adverse events: Proton: grade ≥2 esophagitis 8% (CSI-related), myelosuppression during CSI similar. Late: cochlear and cardiac sparing translates to lower rates of clinical ototoxicity and subclinical cardiac dysfunction at 5 years. No radiation myelopathy in either arm.
Conclusions
Proton CSI significantly reduces dose to cochlea, heart, lungs, and kidneys compared to photon IMRT while maintaining equivalent disease control. The reduced cochlear and cardiac dose translates to meaningfully better functional outcomes including slower IQ decline. Proton CSI is the preferred modality for pediatric medulloblastoma when available.
Key Limitations
Key Limitations: Non-randomized comparative study — selection bias in who received proton vs photon. Era differences and institutional technique evolution. Long-term cardiovascular events (subclinical cardiac findings at 5 years may translate to clinical disease at 20+ years) not yet observed. Limited to medulloblastoma.
Clinical Context
Proton CSI for medulloblastoma is the standard at all major pediatric proton centers. The dosimetric advantages (cardiac, pulmonary, renal, cochlear sparing) are compelling for this young patient population with curative intent and decades of potential late toxicity. ACNS1123 (proton vs photon randomized trial) was open but closed due to poor accrual, leaving comparative evidence level at III.
References