Strategic modulation to avoid radiation toxicities for brain integrity (SMART-BRAIN).

Abstract

Patients with medulloblastoma require craniospinal irradiation (CSI) to optimize disease control; however, cranial irradiation is associated with significant morbidity. We sought to develop a novel target volume (CTV_2340 cGy) and explore dosimetric benefits to minimize cognitive toxicity. We conducted a literature review to confirm patterns of failure for average risk medulloblastoma to develop the CTV_2340 definition approach. Five patients (ages 4-16 years) were identified and proton and photon plans were generated for the new CTV_2340. Planning target goals included V100% ≥ 95% (the entire volume receiving at least 95% prescription dose) and V98% ≥ 97%. Proton planning included 3D robust optimization +/- 2 mm and +/- 2% range uncertainties. Photon plans included a 1 mm planning target volume. Medulloblastoma principally recurs in the ventricular system, periventricular region, cortical surface, or posterior fossa, supporting potential omission of remaining non-target-brain (NTB, e.g. white matter tracts, thalami). The novel CTV_2340 included these regions with a 1 mm brain expansion. All proton-based plans achieved appropriate target coverage. Median whole brain volumes were 1317.1 cc (range: 1116.0-1442.1), reflecting brain size over the age spectrum. Median NTB was 15% of the brain volume (9-31%), whereas the hippocampi were < 1%. Median NTB_V12Gy and V18Gy were: 91.7% (79.3-92.3) and 70.2% (59.8-78.2). Median NTB_minimum and NTB_mean were: 2.1 Gy (0.8-2.9) and 19.4 Gy (18.0-20.2). CTV_2340 overlapped with the hippocampi, limiting hippocampal sparing (median V18 Gy 99.1%). Photon-based planning showed inferior target coverage with minimal sparing. This study introduces a strategic approach to reduce radiotherapy-related neurotoxicity in medulloblastoma through proton-based sparing of critical brain regions. Multi-institutional studies are warranted to further advance this approach.