Purpose: A TrueBeamSTx 6FFF beam model geared toward treating small targets with VMAT and dynamic conformal arcs (DCA) was created. However, a revision was necessitated due to IMRT QA failures. The purpose of this abstract is to describe the needed changes and required validation, as well as to discuss the challenges encountered in creating an adequate model for two distinct treatment techniques.
Methods: Since its clinical release, our beam model was predominately used for small VMAT targets (2-4 cm in diameter). However, IMRT QA failures, where the calculated dose was higher than measured, were occurring. Of 70 total plans, 15 (21.4%) plans exhibited median dose differences of >3.0%, with an average median dose difference of -2.3%. To improve absolute dose agreement, several MLC parameters were adjusted. Subsequent validation work included film, ion chamber, and Delta4 measurements for representative DCA and VMAT plans.
Results: Post-revision, profile agreement in the low-dose shoulders for simple DCA plans worsened based on film measurements. Original DTA values were 0.1-0.2 mm and the average 5%/0.5mm gamma passing rate was 95.5% for a range of target sizes. Corresponding values for the revised model were 0.1-0.3 mm and 90.9%. For VMAT deliveries, the absolute dose agreement within targets improved, with the average discrepancy decreasing from -1.7% to -0.6% for TG-119 type validation plans. Post-revision, the average median dose difference for IMRT QA was -0.5% (59 plans), with all passing clinical criteria.
Conclusion: Our experiences highlight the challenges with creating a single beam model that is accurate for a range of target sizes and treatment techniques. For small targets, a compromise between DCA profile accuracy and VMAT absolute dose accuracy was needed. Ultimately, a model revision was performed to optimize the model based on actual clinical utilization.