Presenter(s)
D. Miles, J. A. Knight II, H. Chen, H. Li, and C. Deville Jr; Department of Radiation Oncology and Molecular Radiation Sciences, Johns Hopkins University School of Medicine, Baltimore, MD
Purpose/Objective(s): Respiratory motion management is critical for renal cell carcinoma (RCC) SBRT, and especially challenging for proton SBRT due to sensitivity to anatomic and range variations. Real-time gated proton therapy (RGPT) uses x-ray fiducial tracking to gate delivery and may improve dosimetric fidelity for moving renal targets. We hypothesized that fiducial-based RGPT, integrated with breath-hold and free-breathing workflows, would improve target coverage and/or OAR sparing in RCC SBRT.
Materials/Methods: Patients were assigned at simulation to breath-hold (BH) or free-breathing (FB) cohorts based on target motion and ability to perform repeated breath holds. Treatment plans were generated using 14 Gy(RBE) x 3 fractions = 42 Gy(RBE) and single-field optimization with 3-5 beam angles. At treatment, targets were localized using an implanted fiducial marker. Orthogonal kV images were acquired at up to 7.5 frames/s, and an automated algorithm determined the fiducial’s real-time position relative to its planned position. For each field, beam delivery was permitted only when the fiducial was within a user-defined uncertainty envelope. Time-resolved logs of fiducial position and the beam-on gating signal were recorded for each fraction.
Motion-perturbed dose reconstruction was performed for 12 representative fractions (n=6 BH, n=6 FB) by offsetting planned beam spots using RGPT log data. Ungated reconstructions applied all measured motion regardless of beam-on status; gated reconstructions used fiducial locations only during beam-on gating. Target coverage and OAR clinical goals were evaluated per fraction for each scenario. Absolute changes in clinical goals were computed per fraction, and relative changes in DVH metrics were computed per arm (gated vs ungated; BH vs FB; reporting mean+std per arm). Statistical comparisons used the Wilcoxon signed-rank test.Results: GTV/ITV coverage improved with gating. For GTV V4200>99%, coverage increased from 92.9±7.1% to 98.1±2.0% in BH (p=0.063) and from 95.5±6.4% to 99.5±0.7% in FB (p=0.063). ITV V4200>99% showed a similar trend in free-breathing (94.5±7.3% to 99.0±1.5%; p=0.063). PTV high-dose coverage improved more robustly: PTV V4200 increased from 83.6±7.6% to 92.8±2.6% in BH (p=0.031) and from 83.5±11.1% to 91.1±6.8% in FB (p=0.031). Small reductions were observed in ipsilateral kidney dose/volume metrics in FB (e.g., V1000: 23.78±10.5% to 22.78±10.1%; p=0.031). Renal cortex and medulla, and adjacent OARs such as the large bowel and duodenum high-dose volumes decreased similarly.
Conclusion: Fiducial-based RGPT provided effective motion management for renal proton SBRT under both BH and FB paradigms, with improved target coverage in both techniques and significant ipsilateral renal dose reductions in FB. These findings support further investigation of RGPT as a strategy to improve delivery fidelity for moving renal targets and to inform future refinement of motion-related robustness and margin design in RCC proton SBRT.