ESTRO 2024 - Abstract Book

S4158

Physics - Intra-fraction motion management and real-time adaptive radiotherapy

ESTRO 2024

Material/Methods:

The multi-target dose-optimised MLC tracking algorithm used a simplified real-time dose calculation to accumulate the planned and delivered doses at 50 ms timepoints during treatment to volumes assigned to each target. At each timepoint the dose volumes were translated according to the motion of the corresponding target. The difference between the planned and delivered doses was calculated and a real-time volumetric dose optimisation algorithm was used to determine the MLC leaf positions that would minimise the cost function associated with dosimetric error. To test multi-target MLC tracking for the lung, three patient cases with distinct anatomies were generated using the 4D-XCAT computational phantom [3]. Each test case had two spherical lesions 2 cm in diameter placed in the lungs. To test the effect of motion for a range of lesion locations, the patient cases included lesions placed in the lower, middle, and upper lungs. A 4DCT was generated for each patient case where the magnitude of the diaphragm motion for each patient was based on motion observed during 4DCT for three patients treated in the LIGHT SABR clinical trial [4]. Two SABR treatment plans were generated for each patient case: a tracking-based plan, where the planning target volumes (PTVs) were created by adding 5 mm margins to the gross tumour volumes (GTVs) delineated on the end-of exhale phase CT, and an internal target volume (ITV)-based plan, where the GTVs were delineated across all phases on the 4DCT and expanded by 5 mm to create the PTV. Each treatment plan consisted of two volumetric modulated arc therapy (VMAT) fields with a prescription of 54 Gy in 3 fractions. Treatment delivery was simulated by inputting diaphragm motion traces measured using a bellows belt during treatment for the LIGHT SABR patients into the XCAT phantom. The motion induced in each lesion was adapted to using multi-target MLC tracking with the tracking-based treatment plan. Three fractions were simulated for each patient case. Performance of the multi-target MLC tracking method was compared to treatment simulated without tracking using the ITV-based plan. The dosimetric errors to the GTVs were evaluated using a 3D gamma comparison between the planned and delivered doses with a 2%/2 mm pass criterion.