S1730
Physics - Dose prediction/calculation, optimisation and applications for particle therapy planning
ESTRO 2026
Purpose/Objective: Proton RBE depends on LET, dose, fractionation and tissue type. LET distributions are heterogeneous since LET increases near the end of the proton beam Bragg peak, posing risks for OARs but offering opportunities for LET painting in the CTV. We assessed how optimizing LET in the primary CTV leads to improved CTV-RBE and TCP in IMPT for lung cancer patients. Material/Methods: 10 lung cancer patients, previously treated at our center to 55-66 GyE in 20-30 fractions were included. Clinical ITV-based treatment plans consisted of 3-5 patient-specific beams and were optimized on the average CT with 7mm/3% setup/range robustness settings. Dose was prescribed to the voxel-wise minimum D98%,ITV=0.95Dpres across 28 scenarios, averaged over four breathing phases of the 4D planning CT. For each patient, three additional primary CTV-based plans were made, assuming elimination of breathing motion via breath-hold or gating, and including four opposing beams to facilitate LET optimization. The plans were: (i) robust dose- optimized plans based on clinical robustness goals (VWmin D98%,CTV=0.95Dpres), (ii) LET optimized plans, with enforced pencil-beam placement by splitting the CTV into four beam-specific sub-volumes and (iii) robust LET-optimized plans, combining (ii) with P(D99.8%,CTV=0.95Dpres)=90% in a probabilistic evaluation of 10,000 fully fractionated treatments with polynomial chaos expansion [1]. In (iii), probabilistic dose inhomogeneity up to P(D2%,CTV=1.4Dpres)=95% was accepted. Critical OAR constraints were never exceeded. Plans were evaluated in terms of dose (RBE=1.1), dose-averaged LET (LETd), voxel-wise LETd × dose, and McNamara RBE ( α / β = 10 Gy) for which, the median, near-minimum (D98%), and near-maximum (D2%) values in the CTV were calculated in the patient population. TCPs for dose, RBE=1.1 dose and McNamara RBE-weighted dose were calculated based on the equivalent uniform dose [2]. Results: LETd, dose, LETd x dose and McNamara RBE in the CTV were all significantly (p<0.001) higher for the LET optimized plans (ii, iii) compared to the robust dose opt plans (i) (Fig. 1). TCP was significantly (p<0.001) increased for the LET optimized plans (ii, iii) compared to the robust dose opt plans (i) (Fig. 2). No significant differences (p>0.05) in TCP were observed between the RBE=1.1 dose and McNamara RBE-weighted dose in the robust LET optimized plans (iii).
Conclusion: Creating non-inferior FLASH-ready treatment plansis patient and strategy dependent. When feasible, sparing of OARs can be further enhanced through a stronger FLASH effect when using hybrid BPTPB. Keywords: FLASH, proton therapy, OAR sparing
Digital Poster 2377
Enhanced CTV-RBE and TCP through LET painting and probabilistic evaluation guided planning in IMPT lung cancer patients Max O.M. Koch 1,2 , Jenneke I de Jong 2,1 , koen Crama 3,1 , Mischa S Hoogeman 2,1 , Steven J.M. Habraken 3,1 1 Medical Physics and Informatics, HollandPTC, Delft, Netherlands. 2 Radiotherapy, Erasmus MC Cancer Institute, Rotterdam, Netherlands. 3 Radiation Oncology, Leiden University Medical Center, Leiden, Netherlands
Made with FlippingBook - Share PDF online