Positioning uncertainty margins for robust optimization in kidney stereotactic body proton therapy

Abstract Introduction Stereotactic body proton therapy (SBPT) has arisen as a promising modality for treatment of primary kidney tumors. However, the sensitivity of proton beams to materials in the beam path limits the use of conventional immobilization devices. The purpose of this study was to assess immobilization using a proton-friendly setup and to develop practical guidance surrounding positioning uncertainty in kidney SBPT robust planning. Materials and methods Data for patients treated between 2024 and 2025 with kidney SBPT were retrospectively reviewed. Patients were treated in the headfirst supine position, with arms in a wing board and legs in a vacuum bag. To minimize beam perturbation, no compression or abdominal immobilization systems were used. Treatment plans used a single-field optimized (SFO) posterior-anterior (P/A) and lateral beam arrangement. The beams were duplicated for volumetric repainting. During treatment, patients were reimaged and repositioned between duplicate beam pairs. Positioning uncertainty parameters for robust optimization were determined using an analytical margin recipe and the measured intrafraction repositioning shifts. The dosimetric effect of intrafraction motion was modeled by applying shifts as isocenter offsets in assessment plans robust optimized using the calculated margin and 3.0-, 4.0-, or 5.0-mm isotropic positioning uncertainties. Results The average 3D shift across 92 included fractions was 2.6 ± 1.3 mm. Calculated margins were 3.0, 4.0, and 2.6 mm in the right-left, inferior-superior, and P/A directions, respectively. Dosimetric modeling demonstrated improved retention of 95%-100% target coverage doses when using intrafraction repositioning and the calculated margin in robust optimization ( p < 0.05). Plans optimized using the analytical, 4.0 mm uniform, or 5.0 mm uniform margins all had 99% coverage dose ≥ 95% of the prescribed dose in 91 of 92 fractions, meeting our institutional acceptance criteria. Increasing margins resulted in higher dose to nontumor kidney tissue ( p < 0.01). Conclusion 4.0 mm inferior-superior (I/S) and 3.0 mm right-left (R/L) and P/A positioning uncertainty appropriately accounts for intrafraction motion measured for our workflow and planning strategy. Performing intrafraction repositioning significantly improves agreement between planned and delivered dose. Though intrafraction shifts were small on average, sporadic instances of large shifts may demonstrate the limitations of using a setup which minimizes beam perturbation. Clinical trial number Not applicable.