Assessing Stereotactic Body Proton Therapy Treatment Plan Robustness to the Liver
Abstract
Purpose
To assess different proton beam arrangements’ robustness for planning stereotactic body proton therapy (SBPT) to the liver.
Methods
We identified patients (n=19) previously treated at our institution with photon stereotactic body radiation therapy to the liver under breath hold and with CT-on-rails image guidance, for whom SBPT may be indicated based on the laterality of their tumors (right-sided). Using the patients’ clinical photon targets, we planned several scanning beam SBPT treatments in Raystation 12A-SP1 using a validated model of our beamline, including: 2-beam (lateral and anterior/posterior), 3-beam (lateral, anterior/posterior and oblique) and 4-beam (anterior, posterior, and lateral obliques) arrangements. We assessed plan robustness assuming possible ±2mm spatial and ±3.5% density shifts, ensuring 95% coverage of the IGTV under these scenarios. We evaluated D30cc to the chest wall and V15Gy to the liver to assess normal tissue sparing across the plans, with other typical organs-at-risk receiving negligible proton doses due to the targets’ laterality.
Results
Our preliminary results suggest that 2, 3 or 4 beam arrangements can be suitable for planning SBPT for right-sided liver lesions. Compared to the clinical photon plans, the 2,3, or 4 proton plans reduced the liver V15Gy by -26±5%, -28±10% and -28±8%, respectively, and yielded comparable (-3±5%) or reduced (-22±3% and -27±4%) chest wall D30cc, respectively. The worst-performing cases in our SBPT robustness analysis still reduced liver V15Gy by -19±7%, -20±12% and -20±10%, and showed comparable (+2±8%) or reduced (-19±4%, and -23±5) chest wall D30cc , respectively, for 2, 3 or 4-beam arrangements.
Conclusion
Right-sided liver lesions can be planned robustly for SBPT, with 3-beam arrangements offering a good balance of improved normal tissue sparing, plan complexity and robustness. Future work using daily CT-on-rails alignment images will allow us to incorporate daily variations in patient anatomy into our robustness analyses.