Cohort-specific dosimetric and delivery tradeoffs for proton dynamic arc vs. IMPT for breast and chest wall irradiation.
TL;DR
Compared with conventional intensity-modulated proton therapy for breast/chest wall irradiation across diverse clinical anatomies, DynamicArc achieved target coverage comparable to IMPT and improved robustness pass rates for selected targets.
Abstract
Purpose
To compare proton dynamic arc therapy (DynamicArc) with conventional intensity-modulated proton therapy (IMPT) for breast/chest wall irradiation across diverse clinical anatomies.
Methods
Forty patients were analyzed (10/cohort): tissue expanders, permanent implants, no reconstruction, and bilateral disease. DynamicArc plans used coplanar partial arcs (150-160°) with arc retracing; paired IMPT plans used 2-4 static fields. RayStation 2023BDTK-generated plans shared contours and prescriptions (50.4 Gy[RBE]/28; subset bilateral 42.56 Gy[RBE]/16) and were robustly optimized/evaluated (±5 mm setup,±3.5% range) with Monte Carlo dose calculation (0.5%uncertainty,40000ions/spot) and a 7.5cm range shifter. Target coverage and OAR endpoints were compared using Wilcoxon signed-rank tests with FDR correction. Additionally, robustness pass rates, LETd metrics, and estimated treatment times were compared.
Results
DynamicArc achieved target coverage comparable to IMPT and improved robustness pass rates for selected targets. In the expander cohort, DynamicArc improved coverage posterior to the expander port and increased coverage of the chest wall evaluation volume and nodal targets. Improved D95% was accompanied by increased target V105% in several targets. DynamicArc increased thyroid mean dose and skin V95% in multiple cohorts, although skin D2cc and skin V100% remained low. Focal hotspot metrics for the esophagus, heart, and LAD were largely unchanged, with LAD D0.03cc numerically lower though not statistically significant in the expander cohort. Ipsilateral lung V20 was generally similar, with a reduction in implant cases. Several unadjusted significant findings did not remain significant after FDR correction. High-dose CTV LETd was broadly comparable between techniques with changes in the tails of the distribution. DynamicArc increased spots, energy layers, and beam-on time, but estimated total treatment time was comparable after accounting for IMPT room entry and imaging.
Conclusion
DynamicArc achieved comparable target coverage with improved robustness in challenging anatomies, at the cost of increased V105%, thyroid mean dose, and skin V95% while maintaining comparable treatment times.