Comparative study of relative biological effectiveness of different microdosimetric kinetic models in heavy-ion pencil beam scanning for hepatocellular carcinoma

Methods Based on the Siemens IONTRIS heavy-ion pencil beam scanning system at the Heidelberg Ion Beam Therapy Center (HIT), a detailed beamline geometry model for carbon and oxygen ion beams was constructed using the TOPAS MC platform. The macroscopic physical dose distributions in water phantoms were obtained, and the corresponding microdosimetric quantities were subsequently calculated. These microdosimetric parameters were then incorporated into three different microdosimetric kinetic models implemented in the MONAS tool, enabling the calculation of cell survival outcomes and realizing a multiscale simulation framework spanning from energy deposition to cellular biological response.By comparison with the survival data of HCC cell lines (Hep3B, HepG2, HUH7, and PLC) after carbon and oxygen ion irradiation obtained from the PIDE database and reported by Habermehl et al., the cell survival curves and RBE10 values calculated using three modified microdosimetric kinetic models were evaluated to assess the computational accuracy and clinical applicability of the different models.Results The combined TOPAS–MONAS multiscale Monte Carlo simulation framework for carbon and oxygen ion pencil beam scanning radiotherapy demonstrated high accuracy. The simulated spread-out Bragg peaks (SOBP) of carbon ions (119~134 MeV/u) and oxygen ions (140~162 MeV/u) showed good agreement with published experimental data, with relative errors in SOBP width below 3% and mean energy deviations less than 2%. The calculated LET values were 114 keV/μm for carbon ions and 138 keV/μm for oxygen ions, consistent with experimental measurements.Cell survival curves revealed pronounced differences in radiosensitivity among the investigated HCC cell lines, with Hep3B and HUH7 being the most radiosensitive, whereas PLC exhibited the highest radioresistance. The calculated RBE10 values ranged from 2.1 to 3.4 for carbon ions and from 1.7 to 3.4 for oxygen ions, showing overall good agreement with experimental observations. Among the three evaluated microdosimetric kinetic models, the mSMKM model achieved the highest predictive accuracy, with mean absolute error (MAE) ≤ 0.07 and root mean square error (RMSE) ≤ 0.08.Conclusions This study validated the feasibility of the coupled TOPAS–MONAS multiscale simulation framework for heavy-ion radiotherapy modeling of hepatocellular carcinoma. Noticeable discrepancies were observed among different microdosimetric kinetic models in predicting RBE in high-LET regions, whereas the mSMKM model demonstrated superior stability and robustness. These findings provide quantitative evidence for the selection of RBE models and the optimization of TPS parameters, and offer methodological support for the realization of individualized and precision heavy-ion radiotherapy.Objectice To compare and investigate the differences in the calculation of biological effects of hepatocellular carcinoma (HCC) cells among different microdosimetric kinetic models (MKM-z*, DSMKM, and mSMKM) under heavy-ion pencil beam scanning radiotherapy, and to evaluate their applicability in clinical treatment planning systems (TPS).