SU-E-T-582: Assessment of Improved Critical Structure Sparing Using Biologically- Based Optimization for Volumetric Modulated Arc Therapy of Head and Neck Cancer.

PURPOSE To investigate the critical structure sparing achievable by biological optimization for modulated volumetric arc (VMAT) of head and neck (H&N) cancer. METHODS Ten H&N VMAT patients who were originally planned and treated with simultaneous integrated boost technique using dose-volume based optimization, were selected. For each patient, a new VMAT plan was generated using the generalized equivalent uniform dose (gEUD)-based biological optimization available in Pinnacle3 Smart Arc. While the objectives for targets were defined in terms of minimum and maximum EUDs with 'a' values of -15 and +15, the critical structure objectives were defined in terms of maximum EUDs with 'a' values of'1', '5' and '10'. Both dose-volume and gEUD-based VMAT plans used 6MV, 280 degree dual arcs. The prescription doses for all cases were 70Gy to PTV1, 60Gy to PTV2 and 54Gy to PTV3 delivered in 33 fractions. The critical structures included cord, parotids, brainstem, and esophagus. The critical structure sparing in physical and biological plans were assessed using dose- volume indices, including D2, D50, and Dmean, and MUs and treatment delivery times. RESULTS PTV coverage obtained by both dose-volume and biological-based optimization were generally very similar, although the target conformality was slightly better for dose-volume based VMAT plans. The gEUD-based plans produced superior critical structure sparing, yielding up to 55% reduction in cord D2, 35% reduction in parotids Dmean and 14% reduction in brainstem D2 as compared to dose-volume based plans. On average, MUs for gEUD-based plans increased by 12.1%, 18.8% and 21.3% for 'a' values of'1', '5' and '10' respectively compared to dose-volume based VMAT plans with no change in delivery times. CONCLUSIONS This study showed that the biology-based optimization has ability to generate H&N VMAT plans with significantly improved critical structure sparing with generally similar target(s) coverage as compared to the dose-volume based physical optimization.