Further Radiobiologic Modeling of Palliative Radiotherapy: Use of Virtual Trials

Abstract

To study duration of response in palliative radiotherapy in a population of tumors. Models of dynamic changes in cell number with time were used to develop a function for the remission time (T(rem)) after palliative radiotherapy: [See Equation], where BED is the biologically effective dose, t(1) the duration of symptoms (i.e., the time between the onset of symptoms and the initiation of radiotherapy), K the daily BED repopulation equivalent, alpha the linear radiosensitivity parameter in the linear-quadratic model, and z the tumor regression rate. Simulations of clinical trials show marked variations in remission statistics depending on the tumor characteristics and are highly compatible with the results of clinical trials. Dose escalation produces both a higher proportion and extended duration of remissions, especially in tumors with high alpha/beta ratios and K values, but the predicted dose responses of acute and late side effects show that caution is necessary. The prospect of using particle beam therapy to reduce normal tissue radiation exposures or using hypoxic sensitizers to improve the tumor cell kill might significantly improve the results of palliative radiotherapy in carefully selected patients and could also be used for safer palliative re-treatments in patients with the potential for prolonged survival. The effect of tumor heterogeneity in determining palliative responses probably exceeds that in radical radiotherapy; as few as 100 patients in each treatment arm produce statistically unreliable results. Virtual trials of palliative radiotherapy can be useful to test the effects of competing schedules and better determine future strategies, including improved design of clinical trials as well as combinations of radiotherapy with other anticancer modalities.