2308 - Characterization of a Treatment Plan Optimization Algorithm for VMAT with Integrated Static-Angle Ports

E. Hubley¹, B. Koger¹, M. Salerno¹, R. M. Scheuermann¹, T. Li¹, L. Dong², and B. K. K. Teo¹; ¹Department of Radiation Oncology, University of Pennsylvania, Philadelphia, PA, ²Department of Radiation Oncology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA

Purpose/Objective(s): A new optimizer allows for the automatic addition of static-gantry ports during VMAT delivery (VMATp). The optimizer also allows for dynamic collimator rotation, with either full optimization or optimization between static ports. Plan quality for whole breast and chestwall planning using the new optimizer with a variety of port geometries was evaluated against conventional VMAT to determine optimal use of the new system. Materials/

Methods: For 13 patients (9 left-sided and 4 right-sided), 7 plans were created: (1) conventional VMAT with two partial arcs (2) VMATp without static ports, (3,4,5) VMATp with 2, 4, and 8 static ports, (6) VMATp with two static ports and dynamic collimator rotation between static ports, and (7) VMATp with two static ports and fully dynamic collimator. All VMATp plans included a single partial arc, and static ports were placed to emulate tangent fields. All arcs spanned from approximately 60° across midline (as medial as possible without entering through the contralateral breast), to 150° (left-sided) or 210° (right-sided). For plans without dynamic collimator rotation, collimator angles were selected to allow for the leaf ends to align with the breast/lung interface (approximately 10-350°). In the optimizer, the maximum number of iterations was set to 1000 and static port weight was set to 50. All plans were created with 6MV, a prescription dose of 4005cGy in 15 fractions, and were normalized to PTV D_95%=95% to facilitate comparison. Evaluation metrics included PTV D_95% and V_105%, heart mean, lungs V_4Gy, V_16Gy, and means, and contralateral breast mean. A two-tailed paired t-test was used to determine if DVH metrics were significantly (p=0.05) different between VMAT and VMATp plans.
Results: Compared to a two-arc VMAT plan, a single arc VMATp plan with a static port at each tangent angle (plan 3) significantly reduced the mean PTV V_105% (p<0.01) from 0.8±2.3% to 0.4±0.8%. The heart mean was reduced (p<0.01) from 1.3±0.4Gy to 1.1±0.4Gy. Ipsilateral and contralateral lung means were reduced (both p<0.01) from 5.8±0.9Gy to 5.4±0.9Gy and from 1.0±0.2Gy to 0.6±0.1Gy, respectively. Ipsilateral lung V_4Gy was reduced (p<0.01) from 31.7±5.3% to 28.8±5.3%. Contralateral breast mean dose decreased from 1.1±0.3 to 1.0±0.4 Gy, contralateral lung V_4Gy was reduced from 1.2±1.0% to 0.2±0.2%, and ipsilateral lung V_16Gy was reduced from 12.2±2.7% to 12.0±3.0% (NS). Plans with 4 or 8 ports or dynamic collimation resulted in similar plan quality to the RAD plans with two static ports.
Conclusion: A single-arc, 2-port VMATp plan resulted in significantly decreased heart and ipsilateral and contralateral lung doses compared to conventional VMAT while maintaining PTV coverage.