A study of biologically guided dose painting radiation treatment (BGDP-RT) framework in routine clinic

Abstract

A radiation treatment regimen which incorporates the heterogeneous tumour biological information into the dose optimisation and delivery holds a potential to individualise the radiation treatment that may lead to optimum treatment outcome for a patient. This study explored a framework of biologically guided dose painting radiation treatment (BGDP-RT) for external beam treatment using the volumetric modulated arc therapy modality and the Poisson tumour control probability (TCP) model. The intra-tumour biological heterogeneity (ITBH) was modelled with the radiosensitivity α of varying degrees of heterogeneity that was evaluated in a classification problem employing a basic convolutional neural network model. The effect of ITBH on the Poisson TCP and the D_37 and γ_37 TCP parameters was studied under the uniform and the biologically optimum dose distributions. The biologically based treatment planning (BBTP) was tested within a treatment planning system (TPS) according to the recommendations of the American Association of Physicists in Medicine Task Group Report 166. The relevant normal tissue toxicity of the BBTP plans was evaluated with respect to the Quantitative Analyses of Normal Tissue Effects in the Clinic information. The deliverability of BBTP plans was tested on a C-arm linear accelerator by 2-D detector array fluence measurement. The feasibility of BGDP-RT dose optimisation was explored within a research TPS for the dose painting by contours and the dose painting by numbers techniques. The uncertainty in BGDP-RT was studied in terms of multi-leaf collimator leaf positional error (MLCPE) estimated from the detector output measurement and the related changes in the generalised equivalent uniform dose to targets and the dose to organs at risk reported in the literature. The BBTP and dose delivery may be feasible with an application of a validated dose response model and the precision radiation treatment based on patient and tumour specific biological information.