Encrypted login | home

Program Information

Monte Carlo Cumulative 4D and 5D Proton Dose Distribution Computations and Their Comparison with Analytical Dose Computation Model Predictions for Lung Cancer Patients

no image available
U Titt

U Titt*, J Yang , D Mirkovic , P Yepes , A Liu , C Peeler , R Mohan , U.T M.D. Anderson Cancer Center, Houston, TX


MO-RAM-GePD-T-4 (Monday, July 31, 2017) 9:30 AM - 10:00 AM Room: Therapy ePoster Lounge

Purpose: To evaluate differences in dose distributions caused by dose computation algorithms and inter- and intra-fractional changes in patient geometry over the complete treatment course with passively scattered proton therapy.

Methods: Four dimensional CT images (4DCTs), acquired weekly, were used to compute cumulative dose using the MC² Monte Carlo system. All doses were deformed to a reference image (end-exhale image of the planning CT) to yield the “best estimate of dose actually delivered.” The MC² simulations additionally provided LET distributions which were applied to calculate variable RBE weighted doses for comparison to a fixed (1.1) RBE. Results of these computations were compared to predictions of the treatment plans created with a commercial treatment planning software. To evaluate the impact of changes in dose, the dose volume histograms of a number of relevant structures were compared.

Results: Large differences in dose to the CTV were observed in a small number of case studies. The study found possibly significant differences in cumulative dose computations compared to the planning system dose predictions. Dose differences on the order of 10 Gy (RBE) found in one case study indicate the importance of a systematic cumulative dose computation for lung cancer cases.

Conclusion: While a systematic evaluation of a large number of cases is impractical due to enormous CPU-time requirement for MC-runs, the results show that inter- and intra-fractional changes in patients’ geometries may play a significant role, which must be taken into account in accurate accumulative dose predictions of lung cancer treatments. The results of this study will be used to further validate faster radiation transport Monte Carlo codes which will then be used to process large cohorts of patients.

Contact Email: