this approach to examine the kinetics of repair of carcinogenic injury. Studies have been conducted examining repair kinetics associated with skin tumorigenesis following localized irradiation of rat skin by Burns and coworkers (1975). In the mouse, repair kinetics were determined by examining tumor development in the mouse ovary and mouse lung following whole-body irradiation (Yuhas 1974; Ullrich 1984; Ullrich and others 1987). The experimental design for these studies has been to compare tumorigenic effects following a single acute exposure with the effects after a similar total dose split into two equal fractions separated in time by hours or days. When there is interaction between the two doses the tumorigenic effectiveness would be predicted to approximate that for the single acute exposure, while if there is recovery from carcinogenic injury, the effectiveness of the split doses would be lower. A simple approach to determining whether cellular-based or tissue-based factors play a limiting role in radiation tumorigenesis is to compare a 24 h fractionation scheme with that in which the time between fractions is much longer and more compatible with tissue kinetics. A convenient time to use has been 30 d between fractions. Not surprisingly, considering the role of cell killing in its pathogenesis, studies examining radiation-induced ovarian tumorigenesis have indicated a recovery time between fractions of 24 h or less (Yuhas 1974). Likewise data for skin tumorigenesis in the rat, for which cell-killing effects appear to play a role in neoplastic development, a recovery time of approximately 4 h has been reported (Burns and others 1973, 1975).
More interesting are data for the induction of lung adenocarcinomas in Balb/c mice (Ullrich and others 1987). Cell killing has not been seen to play a major role in the pathogenesis of this tumor, and the doses used in the fractionation studies are not in the range where cell killing would be likely to produce significant tissue damage (Meyer and others 1980; Meyer and Ullrich 1981). A comparison of the tumorigenic effects of two 1 Gy fractions separated by either 24 h or 30 d with that for a single dose of 2 Gy indicated full recovery by 24 h with no further reduction in tumorigenic effectiveness when the time between fractions was increased to 30 d. Additional studies compared the lung tumorigenic effects produced at a total dose of 2 Gy delivered as a single acute exposure to those of multiple 100 mGy fractions separated by 24 h as well as to continuous low-intensity exposures delivered at a dose rate of 4 mGy/h. The observation of a similar reduction in lung adenocarcinomas following both the low-dose-rate and the fractionated exposure regimens also provides support for recovery kinetics in the range of 24 h or less.
While fractionation studies suggest that tissues can recover from radiation-induced carcinogenic injury and that this recovery is likely based on kinetics associated with repair of DNA and chromosomal-type damage, another important question is the persistence of radiation-initiated cells once the initial damage has been produced. Two studies using different experimental systems have addressed this issue. Hoshino and Tanooka (1975) examined the persistence of latent carcinogenic damage in irradiated mouse skin. In this study they gave a dose of irradiation that by itself would not result in the development of skin tumors and followed this with promotion using 4NQO over intervals from 11 to 400 days after irradiation. Importantly, they found that radiation-initiated cells could persist as latent carcinogenic damage for up to 400 d. Yokoro and his coworkers (1977), in studies examining the interaction of radiation and hormones in breast cancer development, found that latent radiation carcinogenic damage could be produced in rat mammary glands by a single low dose of radiation and that the expression of this damage could be enhanced by subsequent stimulation with prolactin. As in the Hoshino and Tanooka study, the latent radiation-initiated cells were found to persist for a substantial portion of the rat’s lifetime.
It has been known for decades that radiation reduces the life span of animals, and studies in mice and dogs have been conducted using life-span shortening as a means to quantify radiation effects (NCRP 1980; Storer and others 1982; Carnes and Fritz 1991; Carnes and others 2002, 2003). The rationale for such studies has been that life shortening, although a complex end point, can serve as an integrated measure of the deleterious effects of radiation. The degree of life shortening from a specific radiation dose can vary as a function of strain, species, gender, and physiological status of the animals (Storer and others 1979, 1982; Korshurnikova and Shilnikova 1996). This variation is largely a function of the spectrum of spontaneous and induced disease and the age distribution of disease occurrence. For example, a great degree of life shortening is observed in animals susceptible to the induction of thymic lymphoma or myelogenous leukemia, both of which occur relatively early following exposure to ionizing radiation (Storer and others 1979, 1982; Storer and Ullrich 1983).
In contrast, in animals that are not susceptible to such early developing neoplasms, but rather develop late-occurring solid tumors following radiation exposure, substantially less life shortening is observed at the same radiation dose. Regardless of the degree of life shortening observed however, analyses of experimental studies indicate that at low doses of radiation and for radiation delivered at low dose rates, such life shortening is due almost entirely to radiation-induced cancer (Storer and others 1979, 1982; Carnes and others 2002, 2003).
Single acute doses in the range of 500 mGy and higher increased life shortening attributable to nonneoplastic effects, but at lower doses and for a wide range of doses deliv-