Radiobiological Modelling of Non-Small-Cell Lung Cancer: A Tumour Control Probability Perspective
No Thumbnail Available
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Radiobiology has developed a methodology to evaluate the radiation response of a wide spectrum of cell lines, which can be used to validate a radiation response using mathematical models. The fundamental motivation for this thesis was to examine the ability of a mechanistic TCP model to predict treatment outcomes for a wide range of treatment strategies for NSCLC, such as hypofractionation, standard fractionation, and hyperfractionation. A fully heterogeneous population TCP model based on the linear-quadratic (LQ) cell survival concept combined with the Poisson statistic was established to predict local tumour control after one, two and three years. The TCP model was created using data from 25 publications. The TCP model was fitted to the clinical outcome data using optimised radiosensitivity values produced by the Nelder–Mead simplex algorithm. The statistical analysis for early-stage NSCLC resulted in R2 values of 0.96, 0.96 and 0.97 and wRMSE values of 3.9%, 5.2% and 5.9% for one-, two- and three-year local tumour control rates, respectively. Based on variations in the TCP with the clonogenic density model, it was estimated that 60% of the dose was sufficient to maintain the TCP after two years for areas with lower clonogenic cell density. The predicted TCP for the T1a vs T1b group after applying 2 Gy in 30 fractions was 48% ± 1.09% vs 43% ± 1.05% according to the 7th edition, and 51.5% ± 0.80% vs 47.8% ± 0.85% according to the 8th edition. Furthermore, the TCP model outcomes for the radiochemotherapy arm exhibited superior TCPs than the radiotherapy alone arm. For the radiochemotherapy arm, TCPs were as high as 75.5%, 50.6% and 41.4% at one-, two- and three-year local tumour control rates, respectively. In contrast to the radiotherapy alone arm results, the model yielded low TCPs values: 51.8%, 38.1% and 33.2% at one-, two- and three-year local tumour control rates, respectively. An important finding of this research is that the T distribution schema of the 7th edition was slightly inferior to that of the 8th edition in terms of specifying the adjacent T subcategories. For instance, the statistical analysis for 3D-CRT resulted in R2 values of 0.89, 0.95 and 0.93 for the 7th edition and R2 values of 0.93, 0.95 and 0.94 for the 8th edition. This thesis concludes that the TCP model is appropriate for the analysis and evaluation of radiotherapy treatment plans as well as radiochemotherapy plans.