2285 - Sensitivity Analysis of Microdosimetric Lineal Energy to Parameter Variations in the Microdosimetric Kinetic Model for Carbon Ion Radiotherapy

Purpose/Objective(s): The modified microdosimetric kinetic model (mMKM) is widely used to calculate the relative biological effectiveness (RBE) in carbon ion radiation therapy. The lineal energy y represents energy deposition per unit distance on a microscopic scale, while y* is its saturation-corrected counterpart and serves as a key component in RBE calculations. This study aims to investigate the sensitivity of y* to uncertainties in other parameters within the mMKM.

Materials/Methods: A treatment plan was created for a spread-out Bragg peak (SOBP) in water, with a width of 60 mm and an energy of 350 MeV/u. The output file from the treatment planning system of a clinical case was converted into an input file and processed using PHITS (Particle and Heavy Ion Transport code System), a Monte Carlo simulation tool. The spectrum of y along the beam direction was obtained from the simulation, and y* was calculated based on the y spectrum, the quadratic coefficient in LQ model (ß), domain radius (r_d), and the nucleus radius (R_n). According to Inaniwa et al, the standard model parameters used in Japan, based on human salivary gland (HSG) cells, are as follows: ß = 0.0615, r_d = 0.32 m, and R_n = 3.9 m. Comparisons were made between y* at the center of the SOBP and y* values obtained by individually varying the standard parameters by up to ± 50%.

Results: The y* value increases when R_n increases. A +50% and -50% variation in R_n resulted in the largest change in y*, with an increase of 15.6% and a decrease of 27.8%, respectively. The y* value decreases with increasing ß, though the effect is relatively small, with a -7.9%/+13.3% change in y* when the variation of ß is +50%/-50%. Variations in r_d cause fluctuations in y* without a consistent directional trend. However, the overall impact is minimal, with the largest change being 6.4%.

Conclusion: The study indicates that y* is largely insensitive to variation in ß and r_d. Although R_n has a greater impact on y*, its variation is predictable and confined in a narrow range given the limited extent of R_n variation (typically 2.5 – 5 µm in human cells). Further studies will assess the potential of y as a universal metric for radiation quality for carbon ion therapy and other particle therapy, given its relative insensitivity to other mMKM parameters, and its consistency across different tumor and normal tissues.