ESTRO 2024 - Abstract Book

S4621

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

ESTR0 2024

Results:

For the whole patient cohort, all plans met clinical requirements for target coverage and OAR constraints. In general, there were slight dose differences between the LETd-based and standard dose-based plans, as shown by the dose volume histograms of Figure 1. Regarding the D1%, dose differences between the dose-based and LETd-based plans, averaged on the entire patient group, were the following: (0.1±0.3, max 0.3), (-0.1±0.7, max 0.7), (0.4±0.7, max 0.7), (0.0±0.3, max 0.3) Gy(RBE) for brainstem, left optical nerve, right optical nerve and chiasm, respectively. For brainstem D0.01cc, the difference was (0.0±0.3, max 0.5) Gy(RBE). If RBE models other than 1.1-fixed RBE were used, higher differences were found. For D1%, dose differences between the dose-based and LETd-based plans recalculated with McMahon model (k= 0.055 um/keV) were the following: (0.8±0.8, max 3.1), (0.4±0.8, max 2.1), (1.0±0.9, max 2.9), (0.1±0.4, max 0.8) Gy(RBE) for brainstem, left optical nerve, right optical nerve and chiasm, respectively. For brainstem D0.01cc the difference was (0.7±1.0, max 3.0) Gy(RBE). When plans were recalculated with McNamara model (a/b = 2Gy), D1% dose differences were the following: (0.8±0.8, max 3.1), (0.4±0.7, max 2.1), (0.9±0.8, max 2.9), (0.1±0.4, max 0.8) Gy(RBE) for brainstem, left optical nerve, right optical nerve and chiasm, respectively. For brainstem D0.01cc the difference was (0.7±0.9, max 2.9) Gy(RBE). Values are summarized in Table 1.

Figure 1 Upper part: LETd Volume Histograms for a LETd based plan (full lines) and a dose based plan (dashed lines) for the investigated OARs. Lower part: Dose Volume Histograms for the aforementioned plans, calculated with RBE 1.1 model (left) and McMahon model (right)