Room: AAPM ePoster Library
Purpose: radiotherapy (MBRT) using intensity modulated electron and photon beams collimated by the photon MLC has been shown to give superior plan quality in the treatment of superficial tumours. The purpose of this work was to develop a framework for robust optimization of MBRT plans and to experimentally validate their delivery.
Methods: robust optimization process was implemented within a previously developed hybrid direct aperture optimization framework for MBRT. To make plans robust against setup errors, the optimization minimizes the expectation value of the objective function evaluated across all specified setup error scenarios under simultaneous consideration of both beam types.
A robust plan considering 5 mm systematic translational setup errors was optimized for a brain tumour case designed on an Alderson head phantom. The plan was delivered with and without 5 mm isocenter shifts. Planar dose distributions for the total plan, photon apertures and electron apertures were measured with EBT3 film and compared with calculated dose using gamma criteria of 3%/2mm and a dose threshold of 10%.
To demonstrate the advantages of robust optimized MBRT over conventional PTV-based plans, a PTV plan employing a 5 mm margin was developed for the same case. Dose distributions were calculated for different setup error scenarios and compared with those of the robust plan.
Results: agreement between dose calculations and film measurements was equal or better than 97.7% when comparing the total plan and photon dose distributions. The passing rate was 96.9% for the electron dose distributions. Compared to the PTV plan, the robust plan had a similar CTV D95% and, on average, a 33% reduction in D2% for all the OARs.
Conclusion: work demonstrates, for the first time, experimental validation of the delivery of robust MBRT plans. Robust optimization is a promising alternative to traditional PTV margins to account for setup uncertainties in MBRT.
Funding Support, Disclosures, and Conflict of Interest: This work was supported by grant 200021_185366 of the Swiss National Science Foundation and Varian Medical Systems.