Particle therapy is distinguished by its highly localized energy deposition pattern, which significantly differs from conventional X-ray radiotherapy. To address these differences, particle therapy treatment plans are typically optimized based on the biological dose, defined as the product of the physical dose and its biological effects. However, the biological effect is often derived from in-vitro data, overlooking more clinically relevant endpoints such as tumor control probability (TCP) and normal tissue complication probability (NTCP). We have developed a mechanistic microdosimetric-based mathematical model (MS-GSM2) Battestini et al., Radiother. Oncol. (2025), Battestini et al., Front. Phys. (2023), Cordoni et al., Phys. Rev. E (2021), Cordoni et al., Rad. Res. (2022) that accounts for the stochastic nature of energy deposition at the single-cell level, considering the impact of radiation chemistry. This model simulates subsequent DNA damage formation, DNA damage repair, cell repopulation, and cell division. Extending the Relative Seriality Model Kallman et al., Int. J. Radiat. Biol. (1992), we have created a mechanistic approach to describe clinically relevant endpoints such as TCP and NTCP. The model's single-cell resolution allows it to account for energy deposition and tissue heterogeneity, considering different organ volume effects and oxygen gradients. Additionally, the model can compute damage repair for each cell over time, enabling the implementation of various fractionation schemes. Our model consistently reproduces NTCP experimental results across various particles, including protons, helium, and carbon ions Saager et al., Radiother. Oncol. (2018), Hintz et al., Radiother. Oncol. (2022), Karger et al., Int. J. Radiat. Oncol. Biol. Phys. (2006), Sørensen et al., Radiother. Oncol. (2022), for both conventional (Figure 1A-C) and Ultra-High Dose Rate (UHDR) (Figure 1D) irradiations. It accurately models different fractionation schemes for the rat spinal cord tolerance and mouse skin injury. Furthermore, we have investigated the role of single-cell energy deposition stochasticity, oxygenation, and irradiated volume on the predicted endpoints. Comparison of the MS-GSM2-driven NTCP results with in-vivo experimental data, for different radiation qualities and biological endpoints, under Conv and UHDR regime. (A, B, C) Conv irradiation of rat spinal cord with 13.0 keV/µm carbon ions, 2.9 keV/µm helium ions, and 3.9 keV/µm protons, for different fractionation schemes. We calibrate the biological pathway parameters of MS-GSM2 on carbon ion data, while we predict the results for helium ion and proton irradiation. (D) Conv and UHDR irradiation of mouse skin with protons. We calibrate the biological pathway parameters of MS-GSM2 on Conv data, while we predict the results for UHDR irradiation. Decades of research have demonstrated that microdosimetry has significant potential to assess the effects of radiation. It is highly effective in predicting macroscopic clinically relevant endpoints, showcasing its potential for clinical application.
Battestini et al. (Thu,) studied this question.