ESTRO 2024 - Abstract Book

S4710

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

ESTR0 2024

isocentricity and CBCT isocentricity) and an additional 0.5mm error to compensate for neglected errors such as deformable registration uncertainties. The random error consists of patient positioning errors found by analysing the inter and intrafraction motion using a camera system (AlignRT) during daily treatment alignment and surveillance. The 3% systematic range error distribution was determined using the results from a previously published range verification study [4]. Treatment courses were simulated by sampling one systematic error for the whole treatment and a random error every fraction. Each simulated treatment course consisted of 15 daily dose distributions calculated on the verification CT closest in time to the fraction while taking the systematic and daily random error into account. The daily dose distributions were deformed to the planning CT and accumulated to obtain the simulated course dose. A total of 25 treatment courses were simulated for each patient for all 5 plans. The accumulated dose of each simulation was analysed in terms of target and dose on the organs-at-risk: Heart, lungs and contralateral mamma. The excess risk was estimated from the average OAR dose according to the National Indication Protocol Proton therapy on breast cancer (2022) [5].

Results:

The systematic error was estimated to have a uniform standard deviation of 1.0mm. The random error was found to be 1.2mm lateral, 1.3mm longitudinal and 1.4mm in the vertical direction.

Plans with a 3mm/3% setup/range uncertainty adequately covered the thoracic wall CTV and the regional lymph nodes (D99>95% in 100% and 96.5% of the treatment scenarios respectively). The reduction in margin size has the most significant impact on the lymph node target coverage, where plans with a 5mm margin already did not consistently achieve a D99 of 95% (fig.1). The average OAR dose decreased with 0.18 Gy(RBE)/mm, 0.05 Gy(RBE)/mm and 0.02 Gy(RBE)/mm set up uncertainty reduction for the lungs, heart and contralateral breast respectively. This results in an absolute reduction of the excess risk of developing secondary tumours in the lungs and contralateral breast of 6.80±0.95% and 0.35±0.05% when comparing the 3mm to the 5mm plan (fig.2).

Figure 1: Probability of D99>95% of the prescribed dose for IMPT plans robustly optimized with a 3% range uncertainty and decreasing set-up uncertainty.