2200 - Free-Breathing Spirometer-Gated Proton Pencil Beam Scanning Radiotherapy - Maintain Tumor Coverage with Improved Organ-at-Risk Sparing
Presenter(s)
H. Qi1, A. Zheng1, S. Huang2, U. Aura3, F. Yu1, Q. Chen1, M. W. Ho1, S. Wei1, M. Rohman1, H. Zhai1, M. Kang4, Y. Lei5, S. Lazarev5, N. Y. Lee6, C. Guha7, I. Yacoub1, A. M. Chhabra1, I. J. Choi1, C. B. Simone II1, and H. Lin1; 1New York Proton Center, New York, NY, 2Tianjin Medical University Cancer Institute & Hospital, Tianjin, China, 3DYN'R Medical Systems, 13290 Aix-en-Provence, France, 4University of Wisconsin, Madison, WI, 5Icahn School of Medicine at Mount Sinai, New York, NY, 6Memorial Sloan Kettering Cancer Center, New York, NY, 7Albert Einstein College of Medicine and Montefiore Medical Center, Bronx, NY
Purpose/Objective(s): Managing tumor motion during pencil beam scanning proton therapy remains challenging, particularly for patients with large respiratory motion amplitudes who cannot tolerate breath-hold techniques. Despite existing mitigation strategies, significant motion-induced dose uncertainties persist. Free-Breathing Respiratory-Gated Therapy (FBRGT) has the potential to be a universal strategy, enhancing patient comfort without disrupting a patient’s natural breathing. Here, we demonstrated a practical FBRGT solution for proton therapy and investigated its dosimetric benefits.
Materials/Methods: The interplay effect was assessed for FBRGT against a repainting approach for large-motion uncertainties. A dynamic lung phantom was used to mimic lung patient with 14 cc spherical tumor moving 12/20 mm (A12/20), representing cases that cannot tolerate breath-hold. 4DCT images of the moving phantom were acquired at 12 breaths/min. Gating windows (GW) were selected at the end of exhalation. Three-field proton plans (30Gy/5fx) were robustly optimized using the clinical protocol on average 4DCT with 3 (A12) or 4 (A20) repainting, or on average CT within different amplitude GW (6 mm for A12, 5/10 mm for A20). Interplay effects were assessed via film measurements and compared to Monte Carlo simulations, considering spot-phase delivery time sequences. Beam on/off latency was investigated and the impact of gating latency was evaluated by repeating measurements with phantom motion restricted within GW. Delivery times were extracted from machine logs.
Results: The internal target volume (ITV) was 22-28 cc for repainting strategy and reduced to 17-21 cc for gating strategy. Comparison of DVHs from interplay simulations show that all plans achieved comparable target coverage, ITV D95% and the ratio of V95%/ITV+5mm varied between 99.9%-101.3% and 1.2-1.4. The lung dose was high for repainting plans due to larger ITVs (Dmean: 20.0%-22.2%, V50%: 14.7%-16.1%) and decreased with smaller GW and breathing amplitude (Dmean: 14.8%-18.8%, V50%: 10.7%-13.5%). Film measurements showed high agreement with the interplay simulation, with gamma passing rates (GPR; 3%/2mm, 10% threshold) of 92.0%-98.2%, indicating the effectiveness of gated delivery. No significant latency effect was observed (GPR: 93.9%-94.6%) with a measured beam on/off latency of 200ms/2ms. A significant improvement in delivery efficiency was achieved in FBRGT compared to the free-breathing repainting approach, and treatment times increased with more repainting (214-264s) and narrow GW (126-165s).
Conclusion: FBRGT was successfully demonstrated through a motion-phantom study. Dosimetric consistency and target coverage were maintained. Improved OAR-sparing and shorter treatment times were achieved with gated delivery compared to the repainting strategy. More validation is warranted for future clinical implementation.