Using Scorecards to Tune Ethos Directive Templates: An ARTIA Cervix Dosimetric Planning Study

Abstract

The Adaptive Radiation Therapy Individualized Approach (ARTIA) Cervix clinical trial uses predefined clinical directive templates (CDT) combined with RapidPlan DVH estimations (DVHe) to guide plan optimization in the Ethos treatment planning system. The dosimetric scorecard tool (DST) quantifies improvements in plan quality. This is the first study to utilize the DST to tune an Ethos CDT to improve resulting plan quality. Iterative replanning was used to modify the draft CDT (CDT-1) in Ethos to generate a new CDT (CDT-2) that maximized the clinical consensus scorecard's total score compared toCDT-1. CDT-2 was established, and resulting plans were compared with and without a DVHe. Additional fixed field IMRT beam geometries were compared between CDT-1 and CDT-2, both with DVHe. After obtaining favorable results when comparing CDT-1 verses CDT-2 for two test cases, 10 additional cases were retrospectively identified and tested. For the initial test cases, CDT-2 decreased OAR doses without compromising PTV coverage (No DVHe). While both plans met the protocol target guidelines and OAR constraints, the scorecard was able to quantify the improvement with CDT-2 on a test case with a score of 166.1 (78.7%) vs CDT-1, 163.87 (77.6%). When CDT-2 was combined with the DVHe, it still marginally outperformed CDT-1: 168.73 (76%) versus 166.13 (74.8%). Plan quality was further improved by increasing the total number of fields to 19. Combining CDT-2 and DVHe with a 19-field geometry resulted in the greatest benefit at 184.6 (83.2%). This scored higher than the ARTIA-Cervix defined delivery technique of CDT-1 and DVHe, with a 9-field geometry 166.1 (74.8%). The study was expanded to a separate analysis on 10 new cases. The 19-field approach was superior for all 10 cases and CDT-2 achieved a higher score in 7/10 cases. When comparing 9 versus 19 fields, the total optimization and calculation time increased by an average of 1.9 minutes while the beam delivery time increased by an average of 2.8 minutes (+/- 0.1). The average MU/field was 174.3 (total 1568.3) and 129.9 (total 2468) for 9 and 19 fields, respectively. Two test plans were re-optimized and calculated with Ethos 1.1 maintenance release (MR) 1 with both 9 and 19 fields. For case 1, MR 1 resulted in an 8.4% and 6.9% decrease in MU and scored -0.6% and +0.5% for 9 and 19 fields, respectively. For case 2, MR 1 resulted in a 0.8% and 3.1% decrease in MU and scored -3.4% and -0.3% for 9 and 19 fields, respectively. The scorecard allows for easy evaluation of the dosimetric impact of other planning parameters (beam arrangements and use of DVHe) to identify the best approach. Using a scorecard to finely-tune a CDT is expected to improve planning efficiency, decrease intra-institutional plan quality and variability, improve the average calculated plan quality and benefit CBCT-guided ART.