SU‐C‐108‐01: Comparison of Fast Rotational IMRT with a 2d Binary Multileaf Collimator (2D‐BMLC) with Helical Tomotherapy (HT)

Purpose: We developed a treatment planning system(TPS) for fast rotational IMRT treatments with a 2D‐bMLC. For the dosimetric evaluation of the achievable dose distributions 2D‐bMLC plans were optimized for two patients and compared to HT plans. Methods: Within the TPS a treatment machine with a FFF 7MV linac source and a 2D‐bMLC was modeled. The total fraction dose is delivered in one gantry rotation within the time TF. 2D‐bMLC plans were generated for TF of 40s(P40), 25s(P25) and 15s(P15). The maximum dose rate was either 2000MU/min(P40, P25) or 4000MU/min(P15). The considered clinical indications were carcinomas of the oral cavity (patient A) and the nasopharyngs(B). For both patients 57.6Gy in 32 fractions were prescribed to the PTV, 70.4Gy to the boost regions. Dose constraints for OARs were the same as for the clinically accepted HT plans (TF=410s and 650s). An additional tool was developed to recompute the 2D‐bMLC plans with a Monte‐Carlo(Geant4.9.5) method, including a realistic model of the collimator absorber and gantry dynamic. Results: Concerning dose to the OARs the 2D‐bMLC plans are in good agreement with the HT plans. Target conformity was increased and the mean dose to healthy tissue was decreased by >10%(A) and >20%(B) respectively. For patient A, 95%‐coverage of the boost regions was the same as for the HT plan. Coverage of PTV was 95%(P40), 94%(P25) and 93%(P15) compared to 97%(HT). Patient B had larger target volumes. Compared to HT the coverage was decreased by ∼4% for the boosts and ∼3% for the PTV. Target dose inhomogeneity(TDI) was increased(diff.<5%). Concerning these parameters, best results were obtained with P40. Conclusion: Clinically acceptable 2D‐bMLC plans were optimized for both patients. For patient A, dosimetric parameters are comparable to the HT plan. For patient B, some minor degradation has to be accepted for the significant reduction in treatment time. This work was partially supported by Siemens Corporate Research.