SU-E-T-495: Monte Carlo Dose Verification of Passive Scattering Proton Therapy for Prostate Cancer.

PURPOSE To verify the clinical pencil beam dose calculation algorithm for passive scattering proton therapy using field with large range in tissue, i.e. in prostate cancer, using a Monte Carlo (MC) simulation system. METHODS Previously treated prostate cancer cases were randomly selected from our patient database. All patients received the same dose prescription of 50Gy (25 fractions) to planning treatment volume including the seminal vesicles (PTV1), followed by 28Gy (14 fractions) boost to the prostate gland only (PTV2). Patient and beam geometry were imported to our MC simulation platform (TOPAS - TOol for PArticle Simulation) and the dose of each individual beam, as well as their weighted sum, were calculated and compared to pencil beam algorithm-based calculation from the clinical treatment planning system (XiO). RESULTS Preliminary results from four patient cases show overall good agreement between the pencil beam and MC calculations. However, a small but systematic overestimation of the dose, as calculated by the pencil beam calculation algorithm, was noticed for the target structures (<2% difference in D95 for PTV1 and PTV2), compared to the MC calculation. The inverse was observed for the OARs (rectum and bladder) for which the dose seems to be somewhat underestimated by the pencil beam calculation algorithm (up to 3.75% difference in the volume covered by the 70Gy and 75Gy isodose lines). Furthermore, systematic difference in the range calculation was noticed: the pencil beam calculation algorithm results in larger proton range, in the order of 3-4mm, compared to the MC calculations for all beams and patients studied. This can be attributed to the bone anatomy in the path of the beams (femoral heads). CONCLUSIONS Routine MC dose calculation has the potential to improve delivery accuracy in proton therapy of prostate cancer and influence the analysis of currently ongoing clinical trials of protons versus IMRT. Funded by NIH/NCI R01 CA140735.