Author: Awens Alphonse, Nebi Demez, Michael Kasper, Noufal Manthala Padannayil, Shyam Pokharel, Suresh Rana, Hina Saeed, Gagandeep Saini, Nishan Shrestha, Somol Sunny 👨🔬
Affiliation: Lynn Cancer Institute, Boca Raton Regional Hospital, Baptist Health South Florida 🌍
Purpose: This study evaluates the robustness of DynamicARC proton therapy plans for lung and prostate cancer, comparing small spot size (SS) and large spot size (LS) plans under nominal and worst-case conditions.
Methods: DynamicARC plans were generated using a single partial arc technique and robustly optimized with 5 mm positional and ±3% range uncertainty. All plans were normalized so 99% of the CTV received 99% of the prescription dose (5000 cGy(RBE) for lung, 3625 cGy(RBE) for prostate). Robustness was evaluated across 28 scenarios (14 positional variations, 2 range perturbations) using the same uncertainty magnitudes as in robust optimization. Plans were analyzed under nominal conditions, worst-case scenario (WCS), and voxel-wise minimum (VWMin) evaluation using key dosimetric metrics.
Results: For lung cancer, SS showed greater robustness than LS. In WCS, D99% and D95% decreased by 2.4% and 2.5%, respectively, in LS, while SS maintained higher coverage. VWMin confirmed this trend, with LS losing 1.8% (D99%) and 2.1% (D95%). Additionally, LS had worse dose homogeneity (HI) in all scenarios, indicating greater heterogeneity. For prostate cancer, SS and LS performed similarly. Under nominal conditions, D95% was only 0.2% lower in LS. In WCS, D99% was 0.7% higher in LS, and D95% increased by 0.6%, suggesting LS retained better robustness. The HI difference was slightly reduced in WCS and negligible in VWMin, indicating minimal impact on dose uniformity.
Conclusion: SS demonstrated greater robustness than LS in lung cancer, where LS exhibited dose degradation under uncertainties. In contrast, LS performed comparably in prostate cancer. These preliminary findings of the novel DynamicARC technique underscore its feasibility in utilizing small spot size to preserve target coverage under perturbed conditions when compared to large spot size. However, further research is needed to assess the impact of small spot size on the interplay effects in DynamicARC lung plans.