Encrypted login | home

Program Information

An LET-Based Model of Proton RBE Consistent with New High Accuracy RBE Data Measured Near Distal Fall-Off

no image available
O Vassiliev

O Vassiliev*, F Guan , L Bronk , D Grosshans , R Mohan , UT MD Anderson Cancer Center, Houston, TX


SU-I-GPD-T-103 (Sunday, July 30, 2017) 3:00 PM - 6:00 PM Room: Exhibit Hall

Purpose: A recent study reported on high spatial resolution measurements of proton RBE as a function of depth and linear energy transfer (LET). Near the distal edge, measured RBEs were higher than previous models predict. The purpose of this study was to develop a model consistent with newer measurements at all depths yet simple enough to be used for clinical RBE calculations.

Methods: Cell survival data for each of two cell lines (H460 and H1437) were analyzed separately. For each cell line, a global fit of all survival curves measured at different depths was performed using the linear quadratic model. This produced alpha as a function of depth, whereas beta was assumed to be the same for all curves. Fluence energy spectra was calculated for all depths using Monte Carlo. Spectra near the distal edge were relatively broad. Characterizing the beam by a single quantity, such as the dose- or track-averaged LET, would be an oversimplification. Therefore, in the proposed model, the alpha for any given depth is represented as a convolution of fluence spectrum and alphas for monoenergetic protons. The latter quantities were found by fitting the model to the alpha vs depth curve.

Results: The model developed can predict cell survival and RBE for any given proton spectrum including locations near the distal edge. The model has a total of five parameters with four needed to calculate alpha. The difference between predicted and measured alphas exceeded two sigmas only for ~12% of points (the sigma includes only uncertainty of the fit). To implement the model for patient RBE calculations, for all voxels in the volume of interest the dose and the fluence spectra are required.

Conclusion: To predict high RBE consistent with observations of recent high resolution experiments, fluence spectra needs to be taken into account.

Contact Email: