Natural and radiation carcinogenesis in man I. Theory of initiation phase

Abstract A review is given of the evidence and arguments supporting the hypothesis that carcinogenesis is a multi-stage process. In addition to the broad distinction between ‘initiation’and‘promotion’ phases, it is highly probable that initiation itself is multi-stage in character. Previous models of the initiation phase are briefly described; they do not resolve the question as to how many stages are involved. A factor common to all these models, and to the new one described here, is that, other things being equal, cancer incidence should be related to the number of cells at risk. This prediction receives some support from human epidemiological statistics, and from murine experimental evidence. The new model is based on the traditional assumption that each stage of the initiation process is marked by some form of somatic mutation most probably of nuclear genes. It is constructed to permit the analysis of events throughout the life span and not merely during adult life. Equations are derived in which the average somatic mutation frequency appears as a function of the number of somatic mutations needed to initiate any particular stage, at any given age from fertilization. Familial and other evidence shows that certain individuals, comprising a subpopulation, may have a predisposition (usually genetically-based) to a particular malignancy or group of malignancies. The effect of this population heterogeneity on mortality statistics is analysed mathematically.