Impact of Nuclear Halo Modeling on Secondary Cancer Risk Estimates In Proton Craniospinal Irradiation
Poster
aapm_poster_shin.pdf Hosted by AAPM · 1.2 MBAbstract
Purpose
To quantify the impact of proton dose calculation methodology, including nuclear halo contributions, on secondary cancer risk for near-field organs following proton craniospinal irradiation.
Methods
Lifetime attributable risk at age 70 (LAR₇₀) for lung and kidney was estimated in nineteen patients treated with proton pencil beam scanning using three dose calculation approaches: an analytical algorithm with simple halo correction (AA), TOPAS Monte Carlo simulation with constant relative biological effectiveness (MC-1.1), and TOPAS Monte Carlo simulation incorporating a variable RBE model based on dose-averaged linear energy transfer (MC-LETd). Lung LAR₇₀ was estimated using a BEIR VII–based framework with organ equivalent dose, while kidney LAR₇₀ was estimated using the RadRAT kidney model based on mean organ dose. For each patient, LAR₇₀ estimates and relative differences for MC-1.1 and MC-LETd were calculated with respect to AA.
Results
Compared with AA, Monte Carlo–based dose calculations resulted in higher estimated lifetime attributable risk at age 70 (LAR₇₀) for both lung and kidney. For lung cancer, LAR₇₀ increased from 1.07% (95% CI: 0.84–1.29) with AA to 1.28% (1.02–1.54) using MC-1.1 and 1.35% (1.08–1.61) using MC-LETd, corresponding to relative increases of 21.7% (17.8–25.6) and 28.9% (23.9–33.9), respectively. Kidney cancer LAR₇₀ increased from 0.41% (0.21–0.62) with AA to 0.46% (0.24–0.68) with MC-1.1 and 0.50% (0.27–0.74) with MC-LETd, corresponding to relative increases of 24.2% (11.5–36.9) and 40.5% (23.5–57.5). Across both organs, MC-LETd yielded the highest risk estimates, due to the elevated dose-averaged LET (>2 keV/µm) near the field edge.
Conclusion
The findings highlight the importance of accurately characterizing dose in near-field organs when estimating secondary cancer risk. Increases in estimated risk for near-field organs were non-negligible, suggesting that Monte Carlo–based dose calculation may be important for secondary cancer risk assessment in proton craniospinal irradiation.