SU‐FF‐T‐178: Optimization of Internal Target Margins for Dynamic IMRT and RapidARC

Purpose: To determine the optimal Internal Margin (IM) for targets undergoing respiratory motion during treatment with dynamic IMRT and RapidARC. Methods and Materials: Dynamic IMRT and RapidArc plans were created for two targets (3 cm and 5 cm diameter) in the exhale phase with 6 different IM expansions (0,2,4,6,8,10 mm). The plans were delivered to a stationary two‐dimensional ion chamber array (Matrixx, IBA) using a 0.2 second sampling rate. Breathing motions of sin, sin², sin⁴, and sin⁶, with amplitudes of 0 to 4 cm, were simulated by shifting the dose frames by the respiration trace. The resulting simulated motion‐blurred dose planes were randomly sampled from ten starting breathing phases and summed for 30 fractions. The summed dose planes were then compared to the dose delivered to a stationary target with a 0 mm. The optimal IM was calculated as that which resulted in the Equivalent Uniform Dose (EUD) or minimum dose to 95% of the target (D95) closest to that of the stationary case. The optimal margins were fit to the following linear equation: IM = C₁*Amp + C₂ using a least‐squares fit. Results: The value of C1 ranged from 0.55 to 0.98, depending on the target size, type of target motion, and treatment plan. The optimal IM was ∼2mm larger for sin motion compared with sin⁶ motion. In all cases the optimal IM calculated using the two different criteria (EUD and D95) agreed within 1mm. Conclusion: Optimal Internal Margins are a function of target size and target motion. In many cases the IM is smaller than the peak‐to‐peak motion. We have developed formulae that correlate motion with the necessary IM.