Purpose: To evaluate the impact of different beam modeling errors within the treatment planning system (TPS) on patient care.
Methods: The Imaging Radiation Oncology Core (IROC) recently conducted a study to understand suboptimal TPS beam modeling that was associated with IROC phantom failures. The resultant dataset describes TPS parameter selection in the radiotherapy community. The current study evaluated the impact of this spread in TPS parameters on clinical treatment plans. The following parameters were modified in RayStation: multi-leaf collimator offset (MLC(O)), transmission (MLC(T)), gain (MLC(G)),curvature (MLC(C)), source size (SS), leaf tip width (LTW) and tongue and groove (T&G). Each parameter was independently modified to match the community data at the 2.5, 25, 75 and 97.5 percentile levels. The impact of these modifications was evaluated on fifteen patient cases, including prostate, lung and brain plans, generating a total of four hundred and twenty perturbations. The difference in the mean dose delivered to the clinical target volume (CTV) and parallel organs at risk (OAR), and the maximum dose to the serial OAR, were evaluated with respect to the dose delivered using 50-percentile parameter values.
Results: Variation in the MLC(O) and MLC(T) parameters resulted in the greatest dose difference, relative to the 50-percentile level, for all anatomical sites. For the MLC(O) parameter, the greatest impact to the mean CTV and OAR dose was 5.5% and 14.6%, respectively, at the 97.5-percentile. For MLC(T), the greatest CTV and OAR dose perturbations were -2.8% and -26.7%, respectively, at the 2.5-percentile.
Conclusion: MLC(O) and MLC(T) beam modeling parameters play a substantial role in patient dose calculations. Extreme values for these parameters (e.g., 2.5 and 97.5 percentile values), which have previously been associated with failing the IROC phantom, negatively impact the quality of patient care. Careful attention should be given to these beam modeling parameters.
Funding Support, Disclosures, and Conflict of Interest: This work was supported by Public Health Service Grants CA180803 and CA214526, awarded by the National Cancer Institute, United States Department of Health and Human Services.