2629 - Impact of Manual and Auto Dynamic Collimator and Static Angle Modulated Port Configurations on Treatment Plan Quality and Efficiency on C-Arm and Ring Gantry Linacs

Purpose/Objective(s): To evaluate the impact of GPU-accelerated inverse optimization for Volumetric Modulated Arc Therapy incorporating dynamic collimator rotation and Static Angle Modulated Ports (STAMPs), termed VMATp, within a commercial planning system. To determine whether new features —including ring-gantry (RG) platform compatibility, automated STAMP angle detection, and automated STAMP weight optimization—improves radiotherapy plan quality and planning efficiency.

Materials/Methods: Seven cases (prostate, breast, stereotactic spine) with prior VMATp plans were re-optimized with the new available functionality. For each case, seven new plans were generated to isolate the effect of different settings for each feature: (1) application of automatic STAMP configuration including angle (AD) and weight (AW) and, (2) implementation on RG. All other planning parameters were held constant. Plan quality was evaluated by comparing STAMP angles, target coverage (e.g., PTV D95%) and organ-at-risk (OAR) dose–volume metrics (mean, max doses) against a reference C-arm (C) manual VMATp plan planned by an experienced VMATp planner. A threshold of 1Gy or 1% was set on all dosimetric goals to be considered improved or degraded. Visual assessment of isodose distributions (e.g., low-dose spill, hotspots) was also performed. Optimization time per plan was recorded to assess changes in planning efficiency.

Results: In total, 49 plans were included in this study across C and RG delivery systems. AD and AW demonstrated variable effects across breast, prostate node, and spine sites. AD did not agree with manual planning angles within +_5 degrees. Across automated VMATp settings, individual clinical goals varied across cases. Automated settings resulted in an increase of improved clinical goals by 0.6 +/- 1.0 and 0.8 +/- 1.1 for C and GR linacs respectively. The settings also resulted in an increase in degraded clinical goals for both cohorts, by (C= 0.9 +/- 1.1, GR = 1.1 +/- 1.4). Overall, no systematic changes in plan quality were observed across the patient cohort.

Conclusion: New VMATp features (ring-gantry support and automated STAMP angle/weighting) preserved plan quality with similar optimisation times to the manual VMATp planning process. These features allowed for comparable plan quality in terms of clinical goals achieved without the need for an experienced VMATp planner present.