Author: Ben Archibald-Heeren, Grace Gwe-Ya Kim, Kelly Kisling, Xenia Ray 👨🔬
Affiliation: UC San Diego, Icon Cancer Centres, University of California, San Diego, University of California San Diego 🌍
Purpose: To evaluate improvements in external beam plans for gynecological cancers from a novel planning solution: RapidArc Dynamic (RAD) (Varian Medical Systems, Palo Alto, CA) which optimizes VMAT with dynamic collimator rotation and user-selected STatic Angle Modulated Ports (STAMPs).
Methods: 12 patients previously treated to 50.4Gy/28fx at two clinics (US and Australia-based) were evaluated using a standardized set of dose goals based on EMBRACE-II and additional harmonized metrics from each clinic. For each patient, the treated VMAT plan was re-optimized for the standardized clinical goals using 2-3 arcs (VMAT+). Then three RAD plans were optimized using 2 arcs and 1, 4, or 7 STAMPs. Other RAD-specific plan parameters were held constant per patient. We compared the number of goals within tolerance/within variation/failed and the average metric difference for each plan type. The time required for a single round of optimization and the total MUs required were also assessed.
Results: RAD plans used full dynamic collimator optimization, 1000-4000 iterations, and arc-dominant weighting (-2 for 1 STAMP, -1 for 4 or 7 STAMPs). Of the 282 metrics extracted across patients, RAD plans had more within tolerance vs VMAT+ (230-236 vs 215) and fewer that failed (7-9 vs 13). RAD plans were able to bring PTV V99%, Bladder V30Gy, and Bladder V40Gy to within variation for 3, 2, and 1 patients, respectively. On average RAD also decreased Bowel metrics: D0.03c by 1.2%, V40Gy by 17-22cc, and V30Gy by 28-39cc and decreased the optimization time by approximately two-thirds (6.2 vs 2.3 mins). While RAD plans increased MUs by 32-40%, each required only 2 arcs, while VMAT+ used 3 arcs for 7/12 patients.
Conclusion: 2-arc RAD created GYN plans of equivalent or better quality to VMAT with improved bladder and bowel metrics and decreased optimization time. Future studies will explore potential for further dosimetric improvements.