Mathematical Model Studies on the Optimal Scheduling of the Treatment of Systemic Malignant Disease by Radiation

[摘要] The work reported in this thesis deals with mathematical model studies on the optimal scheduling of treatment of systemic malignant disease by radiation. To provide the necessary background to the original aspects of the work reviews of several fields are required. Chapter 1 is a general review of normal tissue radiobiology. Chapter 2 is a review of human tumour radiobiology. Chapter 3 focusses on one particular isoeffect model, the linear-quadratic or LQ model, which is employed throughout this thesis to describe the effects of radiation on normal tissues. These basic radiobiological principles are applied to the clinical modalities of total body irradiation (TBI) and biologically targeted radiotherapy (BTR). Chapter 4 reviews the principles of TBI. Chapter 5 is a review of published data on the in-vitro radiosensitivities of human leukaemia/lymphoma and neuroblastoma, two conditions which require a systemic approach to treatment. Chapter 8 is a review of the principles of BTR. The original work is contained in the appendix to chapter 3, which examines the correspondence between the LQ model and CRE models for continuous radiation exposures with constant and exponentially decaying dose-rates; chapter 6, Which examines the question of whether fractionated or low dose-rate TBI is the superior method of treatment; chapter 7, where the optimal scheduling of fractionated TBI is investigated; chapter 9, where the LQ isoeffect model and a dosimetric approach is used for the evaluation of alternative therapeutic strategies for the treatment of widespread micrornetastatic disease by BTR. Finally, in chapter 10 a simple model is used to investigate optimal scheduling of BTR, TBI and marrow rescue. CONCLUSIONS 1/ Comparison of the LQ model and the CRE model for continuous radiation exposures: for constant dose-rates it is found that, when late-effect parameter values are used in the LQ model, there is a correspondence between the models' predictions. There is no correspondence between models when acute-effect parameter values are used in the LQ model. In the case of exponentially decaying dose-rates the predictions of the CRE and LQ models appear more divergent, although again the use of late rather than acute-effect parameter values in the LQ model gives a closer match to the CRE. 2/ Fractionated TBI is predicted to be preferable to low dose-rate TBI treatment. Although theoretically the methods can be equivalent, low dose-rate treatments would have to be over impractically long treatment times. 3/ In the case of external beam TBI, fractionated low dose-rate treatments do not appear to offer a significant improvement over fractionated high dose-rate treatments. This is because in order to achieve a significant increase in dose or reduction in toxicity impractically long exposure times are required. It is expected that this finding will be true in general for external beam radiotherapy, not just in the case of TBI. 4/ Optimal fractionation schedules for the treatment of leukaemia/lymphoma and neuroblastoma by TBI are predicted to be accelerated and hyperfractionated. It is suggested that a two fraction per day schedule of 10 fractions of 1.3-1.5 Gy is a suitable candidate for clinical evaluation. 5/ It is concluded that knowledge of ratios for tumours and normal tissues is, by itself, insufficient information to enable prediction of optimal schedules. 6/ In the case of BTR, dose-rate effects are predicted to be important for late-responding tissues. Tolerance doses may be greater or less than those for fractionated radiotherapy depending on the effective radionuclide half-life. 7/ When injected activities of targeted radionuclide are restricted by haemopoietic tolerance, curative therapy is unlikely. 131-I appears to be a better radionuclide warhead for therapy of micrometastases than 90-Y. 8/ The use of bone marrow rescue in conjunction with BTR seems to offer curative potential, however reasons are presented why a combined strategy using BTR, TBI and marrow rescue is likely to be preferable. 9/ For optimal scheduling of BTR, TBI and marrow rescue, the main characteristics of BTR which determine curative potential are its specificity and sensitivity. Specificity is defined here as the ratio of initial dose-rate at the tumour cells to that in the dose-limiting tissue. Sensitivity is inversely related to the proportion of tumour cells which escape targeting. Where biological targeting is highly specific but some tumour cells escape, a phenomenon of "overkill" will largely determine the optimal schedules. It is predicted that these are likely to consist of combinations of BTR and external beam TBI with the TBI component being the greatest in terms of radiation dose to the whole body.

[发布机构] University:University of Glasgow

[关键词] Nuclear physics and radiation, Oncology