Sub-Phase Optimized Spot-Scanning Proton Therapy for Mitigating Respiratory Interplay Effects Under Free Breathing
Abstract
Purpose
Respiratory-induced interplay effects in spot-scanning proton therapy can lead to dose degradation for lung cancer patients treated under free-breathing conditions. While motion mitigation techniques such as deep-inspiration breath-hold or abdominal compression can be effective, they are often infeasible for patients with compromised pulmonary function. Volumetric repainting is therefore commonly used, but substantially prolongs treatment time. This work proposes a sub-phase–optimized proton delivery strategy that eliminates repainting while improving dosimetric plan quality under free breathing.
Methods
The patient respiratory cycle was discretized into 40 temporal sub-phases (0.1 s resolution over a 4.0 s breathing cycle), each representing a candidate delivery start time. For each sub-phase, motion-induced interplay effects were simulated and accumulated using 4D deformable dose accumulation implemented in OpenREGGUI. For each treatment fraction, one sub-phase–specific delivery plan was selected. A multi-island genetic algorithm was used to optimize the combination of delivery start sub-phases across fractions by jointly optimizing target coverage, dose homogeneity, and robustness against respiratory motion. The proposed strategy was evaluated in two clinical scenarios: lung stereotactic body radiotherapy (SBRT, 5 fractions) and conventionally fractionated lung treatments (30 fractions).
Results
Compared with conventional free-breathing delivery using volumetric repainting, the proposed strategy substantially reduced treatment duration by completely eliminating repainting. In the 5-fraction lung SBRT case, target D95 improved from 91.5% with four-times repainting to 95.1% with the proposed sub-phase–optimized strategy. In the 30-fraction case, conventional random combinations of delivery start times did not consistently yield optimal dose distributions, whereas the proposed optimization framework identified an optimal sub-phase combination with reduced delivery time.
Conclusion
Sub-phase–optimized spot-scanning proton delivery under free breathing effectively mitigates respiratory interplay effects while substantially reducing treatment time and improving plan quality. The method requires only respiratory phase identification at treatment initiation and does not rely on continuous motion monitoring or gating, facilitating straightforward clinical implementation.