Parameters of Radiosensitivity

Abstract

Dr. B. Fertil, Dr. E. P. Malaise, and co-workers have shown (1-5) that despite wide variability in the shapes of published in vitro survival curves for cells of human origin, there are, on average, systematic differences in the shapes of survival curves in the low-dose region among cells of distinct histological types. Moreover, tumor cells whose survival curves have steep initial regions tend to be more radiocurable than those whose average survival curves are more shallow at low doses (2, 4). These results are quite interesting since they support the hypothesis that human cell lines differ in intrinsic radiosensitivity and that such differences contribute to differences in clinical response. In the course of their studies, Dr. Fertil and Dr. Malaise have suggested a number of ways to characterize cell radiosensitivity in terms of a single parameter. Based on data in the literature, they have shown (4) that hulman cell lines can be distinguished by the average values of each of three different parameters: the initial slope of the survival curve (a), the surviving fraction at a small dose such as 2 Gy (SF2), and the mean inactivation dose (D), which is equal to the area under the survival curve in linear coordinates. Of these, they advocate D most strongly, since they find it to provide the best discrimination among cell types (4). Strictly speaking, the radiosensitivity of a cell population cannot be completely characterized by the value of a single parameter, since a minimum of two parameters (e.g., a and (3) is needed to describe the curved shape of a typical cell survival curve over a range of doses. This fact provided the motivation for a recent publication (6) in which I investigated the relationship between the value of D and the shape of the underlying survival curve. In that paper, I showed that it is theoretically possible for curves with different D values to have precisely the same surviving fraction at 2 Gy and, conversely, that there are survival curves with equal D values but which differ in radiosensitivity over the entire low-dose region (<4 Gy). In response, Fertil and Malaise argue (7) that pairs of survival curves with discrepant values of D and SF2, such as those that I presented, are merely theoretical and