ESTRO 2024 - Abstract Book

S4727

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

ESTR0 2024

Cooperation Unit Radiation Oncology, Heidelberg, Germany. 12 National Center of Oncological Hadrontherapy (CNAO), Medical Physics, Pavia, Italy

Purpose/Objective:

With the ongoing establishment of new carbon ion centers in the world, a larger number of patients will be treated with this modality, increasing the therapeutic arsenal in the US as well. Biological modeling in carbon ions represents the weakest point in the current clinical approaches, which could harm the potential benefit of this modality. A clear example of these shortcomings is the clinical data of IPI (Ion Prostate Irradiation, ClinicalTrial.gov under NCT01641185), a prospective randomized phase 2 trial comparing proton and carbon ion treatments of prostate cancer using 66 Gy (RBE) in 20 fractions employing standard clinical RBE modeling (RBE = 1.1 for protons and RBE LEM based for carbon ions). Biochemical relapse-free survival data are 85% (1H) and 50% (12C).[1] We thus created a treatment planning and tumour control prediction framework, which could improve the comparability of treatment arms.

Material/Methods:

The original treatment plans of IPI's 91 prostate tumor patients irradiated with proton (46) and carbon ions (45) were re-calculated using mMKM ((α/β)x = 3.1 Gy). Based solely on the response data of photon-irradiated patient groups, which were stratified according to risk and usage of ADT, we derived parameters for an mMKM-based Poisson tumor control probability (TCP) model. With the mMKM-based model in mind, we created new carbon and helium ion plans that meet the desired biological dose criteria. These newly optimized plans were then compared against existing proton plans as well as the clinical experience of Japanese centers using the same fractionation scheme.

Results:

mMKM predictions suggested a large deviation in biological dose between the 1H and 12C arms. Patients irradiated with protons received in average 3.25 ± 0.08 GyRBEmMKM/Fx, while the carbon ions irradiated patients received 2.51 ± 0.05 GyRBEmMKM/Fx. Our prediction of TCP was 86 ± 3% for 1H and 52 ± 4% for 12C, matching the clinical results of 85% (1H) and 50% (12C). Newly optimized carbon ion plans with the mMKM/TCP model could reproduce clinical data at Japanese centers employing the same fractionation scheme [2]. Application of the TCP model to calculated helium ion plans, showed similar dose coverage as corresponding carbon ion and existing proton plans, while sparing lateral OARs.

Conclusion:

Our thorough analysis of long-term observed IPI clinical data displayed the effect of an applied RBE model on the comparability of the two treatment arms. It also presents a TCP model derived from published photon data that predicts the observed lack of iso-effectiveness of the IPI trial based on modern biological modeling (mMKM). The developed framework has been used to predict tumor control found by other carbon ion centers and to prepare the way for upcoming prostate treatment with helium beams at our facility.

Keywords: prostate, radiobiological effectiveness (RBE)