ESTRO 2024 - Abstract Book

S4713

Physics - Optimisation, algorithms and applications for ion beam treatment planning

ESTR0 2024

(IMPT OR ) or without (IMPT NOR ) a density override of 1.05 g/m 3 to the internal target volume (ITV) during robust optimisation. The accumulated treatment course dose was reconstructed based on calculations on three to five repeated 4DCTs with robustness settings ±3%-range uncertainty (only for IMPT) and 2-mm residual setup errors. The 3D method included dose calculations on the average CTs of the repeated 4DCTs, and the 4D method included dose calculations on all the phases of the 4DCTs. The MHD was compared between VMAT and the IMPT plans nominally, and using the averaged MHD over all computed scenarios. Subsequently, the MHD was translated into a 2-years mortality probability (2YM) based on the Dutch model-based indication protocol for proton therapy. 1 Target coverage by means of the V 95 was assessed on the voxel-wise minimum dose reconstructed from all scenario doses. 2 In the 3D evaluation, the evaluated target volume was the ITV and in the 4D evaluation, this was the clinical target volume (CTV) on the expiration phase of the 4DCT. Target coverage was adequate (V95 ≥98%) for all VMAT plans, and in 6/8 IMPT OR -plans and in 3/8 IMPT NOR -plans (Figure 1). The underdosage in IMPT OR - and IMPT NOR -plans was caused by inter-CT changes in the superior-inferior position of the diaphragm, causing density changes in the beam paths. In the three patients with the largest target motion amplitudes (≥34 mm) underdosage was observed for all IMPT NOR -plans and one IMPT OR -plan. Target coverage was consistently higher in the 4D-evaluation than in the 3D-evaluation for VMAT, which may be caused by the volume difference of the targets, and was not seen for IMPT. The nominal MHD was substantially higher in VMAT compared to IMPT and was lowest for IMPT NOR (Table 1) (p<0.05). Differences between IMPT OR and IMPT NOR were 0.9 Gy on average, translating into a difference in 2YM of 1.4% on average, which remained similar in the 3D and 4D evaluation (0.8 Gy and 1.3%, 0.9 Gy and 1.3% respectively). However, deviations of the MHD over treatment compared to the planned MHD were larger for both IMPT plans than for VMAT in which both the 3D evaluation and 4D evaluation detected changes of 2.2 Gy-3.4 Gy for patients 7 and 8 (resulting in maximal 4.3% difference in 2YM), which differed maximal 1.0 Gy in VMAT (resulting in maximal 1% difference in 2YM). Results:

Conclusion:

Interfractional shifts of the diaphragm and breathing motion had more impact on target and heart dose robustness in IMPT than in VMAT. Target coverage improved by applying a target density override in IMPT, and only increased the MHD for patients who also gained from it in terms of plan robustness with limited impact on 2YM, while remaining much lower than VMAT. MHD over treatment mostly deviated in IMPT for patients with compromised target coverage, which both may be mitigated by employing plan adaptation.