Proposing a Clinical Model for RBE Based on Proton Track-End Counts.

Abstract

In proton therapy, the clinical application of linear energy transfer (LET) optimization remains contentious, in part because of challenges associated with the definition and calculation of LET and its exact relationship with relative biological effectiveness (RBE) because of large variation in experimental in vitro data. This has raised interest in other metrics with favorable properties for biological optimization, such as the number of proton track ends in a voxel. In this work, we propose a novel model for clinical calculations of RBE, based on proton track end counts. We developed an effective dose concept to translate between the total proton track-end count per unit mass in a voxel and a proton RBE value. Dose, track end, and dose-averaged LET (LET_d) distributions were simulated using Monte Carlo models for a series of water phantoms, in vitro radiobiological studies, and patient treatment plans. We evaluated the correlation between track ends and regions of elevated biological effectiveness in comparison to LET_d-based models of RBE. Track ends were found to correlate with biological effects in in vitro experiments with an accuracy comparable to LET_d. In patient simulations, our track end model identified the same biological hotspots as predicted by LET_d-based radiobiological models of proton RBE. These results suggest that, for clinical optimization and evaluation, an RBE model based on proton track end counts may match LET_d-based models in terms of information provided while also offering superior statistical properties.