Some problems arising in analysis of large-scale animal irradiation experiments

This paper describes some of the problems that occur in experiments designed particularly to study the effects of irradiation. One of the purposes of such experiments is to provide data that allow an informed judgment on estimates of the risks for man after exposure to radiation.It is no trivial matter to establish whether exposure to agents, such as radiation, increases the probability of a tumor arising de novo or whether the effect is an advancement of the time of appearance of naturally occurring tumors. In many murine tumors, the problem of establishing which of these two possibilities is the case is increased by the high natural incidence. In order to interpret the data from carcinogenesis experiments, a knowledge of the natural history of the tumor of interest is necessary. In this paper we present data that help to elucidate the biology of lung tumors in hybrid B6CF₁/An1 mice.The rising incidence of many tumors late in life raises the question of whether a common systemic factor is involved. In experiments with female mice, relatively low doses of radiation may sterilize the ovary with subsequent hormonal imbalances that in turn may influence the appearance of tumors in a number of hormone-influenced tissues. We have found that radiation may increase not only the age-specific rates of tumors but also the probability of the number of different tumor types.The possibility of a lack of independence of the occurrences of different tumors complicates the data analysis. To answer some of the questions we have raised, it would be of considerable help to have data on the prevalence of tumors; however, such data usually involve serial killing which is costly. We have examined the possibility of obtaining prevalence data for lung tumors by the determination of the incidence of lung tumors in mice dying from causes other than lung tumors. There appears to be no difference in the prevalence of lung tumors determined by using either this approach or serial killing.