Lifetime-independent risk of a second primary cancer in artificial populations subjected to frailty and cell senescence effects

Abstract

In the past, the age-specific incidence of most forms of cancer was widely thought to increase monotonically with age. However, cancer registry data show that the specific incidence for many forms of cancer increases with age, reaches a maximum, and then decreases. At least two hypotheses have been proposed to explain this decrease: the cell senescence hypothesis and the frailty hypothesis. The objective of our work was to formulate a stochastic model of cancer incidence to estimate the lifetime-independent odds ratio (LIOR) measuring the risk of developing a second primary cancer, conditioned on a first cancer diagnosis, relative to the risk of developing a first primary cancer, in two artificial populations: one where cancer susceptibility is universal and one where only a small proportion of individuals are born susceptible, or frail, to developing one or more cancers. The predicted LIOR values were significantly greater than 1, only 180 Luis Francisco Soto-Ortiz and James P. Brody when a frailty effect was introduced in the model. This result suggests that if a limited proportion of the United States population were innately susceptible to developing cancer, this would be reflected in a LIOR value significantly greater than 1. In addition, the model predicts that the smaller the pool of susceptibles in a population, the larger the computed LIOR value would be. This one-to-one mapping between a LIOR value and the corresponding proportion of susceptibles indicates that a LIOR value could be used to indirectly estimate the proportion of individuals born predisposed to developing a particular type of cancer. Moreover, this result raises the possibility that a LIOR value could be used as a metric to assess how the relative risk of developing a second primary cancer, varies by cancer type in a given geographical region.