Interpreting Survival Observations Using Phenomenological Models

It is generally accepted that the dose-rate dependence of the mammalian-cell survival curve reflects the modification of initial damage by cellular biochemical processes. However, these processes have not been adequately described to allow predictions of the effects at low doses or of radiations having higher stopping powers. The kinetics of damage induction and removal have been studied extensively, even though specific mechanisms have yet to be demonstrated. For example, it can be shown that one component of cell killing depends on the square of the concentration of radiation damage, a dependence that can be explained by sublethal damage models as well as by models featuring misrepair of potentially lethal damage. We have demonstrated two components of radiation-damage repair in plateau-phase Chinese hamster ovary cells. Analysis of repair rates shows that one component has a mean repair time of less than 1 hour, and another has a time of 18 hours. Repair rates, and fractions of damage repaired, appear independent of the initial amounts of damage produced. We have compared our data with the predictions of several different phenomenological models. These observations do not appear compatible with simple models that assume either the saturation of a rapid repair process or the interaction of pairs of sublesions. Rather, they support more-complex models that consider combinations of both sublethal and potentially lethal damage or multiple-step processes. Our observations are also consistent with models that consider the accessibility of damage to the repair enzymes.