Optimization of LET distribution in proton beam therapy for brain tumors

This thesis explores the potential clinical impact of integrating LET (Linear Energy Transfer) optimization into the treatment planning workflow for proton therapy of brain tumors.. Recent advances in treatment planning systems (TPS) have introduced the ability not only to visualize LET distributions, but also to directly incorporate LET optimization into the planning process. Despite increasing awareness of LET’s biological significance, there is still no consensus on reference LET values to support clinical decisions. In the first phase, a normal tissue complication probability (NTCP) model was developed using LET values from a retrospective cohort of brain tumor patients treated at CNAO. Then, 20 patients with similar anatomical and dosimetric features were selected. Their approved clinical plans served as baselines for multiple re-optimizations using LET-based strategies. Optimization was performed for both the current fixed-beam geometry at CNAO and a simulated gantry-based setup, anticipating the future clinical integration of a gantry system. The goal was to reduce LET in healthy brain structures while preserving acceptable dose distributions. All plans were evaluated in terms of: dose quality (trade-off between tumor coverage and OAR sparing), LETd, and robustness to clinical uncertainties. This made it possible to identify clinically feasible compromises among these competing factors. Importantly, the LETd reduction was translated into a lower predicted risk of radiation-induced side effects, using the developed NTCP model. The findings showed that optimized plans—particularly gantry-based plans with stronger LET constraints—achieved reductions in LETd of up to 5.6% in the brain, without compromising target coverage, robustness, or LET distribution. These findings support the feasibility and clinical value of LET-guided proton therapy planning and offer practical strategies to improve treatment safety and biological effectiveness in brain tumor patients.