Abstract

PurposeTo develop an empirical model to predict radiosensitivity and relative biological effectiveness (RBE) after helium (He) and carbon (C) ion irradiation with or without DNA repair inhibitors. MethodsWe characterized survival in eight human cancer cell lines exposed to 6 MV photons and to He- and C-ions with linear energy transfer (LET) values of 2.2-60.5 keV/m to verify that the radiosensitivity parameters (D5%, D10%, D20%, D37%, D50% and SF2Gy) correlate linearly between photon and ion radiation with or without DNA-PKcs or ATR inhibitors. Then, we parameterized the LET response of the parameters governing these linear correlations up to LET values of 225 keV/m using the data in the Particle Irradiation Data Ensemble (PIDE) v3.2 database, creating a model that predicts a cells ion radiosensitivity, RBE and ion survival curve for a given LET on the basis of the cells photon radiosensitivity. We then trained this model using the PIDE database as a training dataset, and validated it by predicting the radiosensitivity of the cell lines we exposed to He- and C- ions with LET ranging from 2.2-60.5 keV/m. ResultsRadiosensitivity to ions depended linearly with radiosensitivity of photons in the range of investigated LET values and the slopes and intercepts of these linear relationships within the PIDE database vary exponentially and linearly, respectively. Our model predicted ion radiosensitivity (e.g., D10%) within 5.1-21.3%, RBED10% within 5.0-17.1%, and ion mean inactivation dose within 6.7-25.1% for He- and C-ion LET ranging from 2.2-60.5 keV/m. ConclusionsRadiosensitivity to He- and C-ions depend linearly with radiosensitivity to photons and can be used to predict ion radiosensitivity, RBE and cell survival curves for clinically relevant LET values from 2.2-60.5 keV/m, with or without drug treatment. SUMMARYWe present a new empirical model capable of predicting clonogenic cell survival of cell lines exposed to helium and carbon ion beams. Our model is based on an observed linear correlation between radiosensitivity to ions and photons across a wide range of LET values. This linear correlation can be used to predict ion RBE, radiosensitivity, and the cell survival curve for a given LET all based on a cells photon survival curve.