A tumour control probability based approach to the development of plan acceptance criteria for planning target volume in intensity modulated radiation therapy for non-small-cell lung cancer

Abstract

The probability of achieving local tumour control is function both of the Prescribed Dose (PD) and the distribution of that dose within the Planning Target Volume (PTV). Conventionally 3-Dimensional Conformal Radiation Therapy (3DCRT) has complied with the homogeneity constraints recommended by ICRU 50. With the implementation of Intensity Modulated Radiation Therapy (IMRT) these homogeneity constraints are less easily achieved, suggesting the need for new approaches to the evaluation of the dose volume relationship within the PTV. The aim of this study is to establish new plan acceptance criteria for the PTV, based on individualised control probability using a TCP model, in the setting of a dose per fraction dose escalation in Non-Small-Cell Lung Cancer (NSCLC). A cuboid PTV of 100 cc was assumed, divided into 1,000 voxels of equal dimension. Dose values from within a pre-defined range were allocated to each voxel using a random number generator. Virtual Dose Volume Histograms (DVH) were produced and TCP values calculated using the formalism of Nahum and Tait. Using model parameters obtained from the literature for NSCLC (α/β = 10, αmean = 0.38 Gy-1, K = 104, σα = 0.088, ρ = 107 cc-1) TCP values were calculated for uniform, homogenous (ICRU 50) and inhomogeneous (non ICRU 50) irradiation. Three levels of dose escalation were assessed i.e. 72 Gy/24 daily fractions (fr), 81 Gy/27 fr and 90 Gy/30 fr. For the dose range assessed the TCP values for uniform irradiation was equivalent to those obtained with homogeneous irradiation. In the case of inhomogeneous irradiation of the PTV, underdosage (<95%) was found to have a more significant impact on TCP than overdosage (>107%). The impact of underdosage on the TCP value was a function of the volume receiving less than 95% of the dose, the range of dose <95% and the prescribed dose level. ICRU equivalent TCP values were found for inhomogeneous irradiation which tolerated underdosage of the PTV of up to 5% for the 72 Gy scheme, 10% for 81 Gy and 17% for 90 Gy. For each dose level, physical acceptance criteria were determined in terms of volume of PTV receiving <95% of the dose and dose range. e.g. for 72 Gy / 24 fr, an acceptable TCP value was obtained with 5% of volume between 90 and 95% and 1% of the volume between 85 and 90%. The TCP model suggests that the dose homogeneity in relation to PTV as recommended by ICRU 50 may be relaxed. Although there is some uncertainty regarding the model parameters, TCP based physical acceptance criteria take into account tumour radiosensitivity, dose fractionation, and biological cell parameters and may provide a clinically more relevant definition of optimum dose distribution within the PTV