Two-stage and Weibull models for carcinogenesis applied to the ED01 discontinued dosing data.

Abstract

The two-stage clonal expansion model for a single, less-than-lifetime period of dosing is formulated and applied to the liver and bladder tumor data from the EDMo study. The model successfully predicts liver tumor incidence for time points beyond termination of dosing with 2-acetylaminofluorene, but it is unsuccessful for bladder tumor incidence. A discontinued dosing version of the Weibull model is proposed and is shown to predict successfully both liver and bladder tumor incidences for time points after termination of dosing.

The Variation in Slope of Concentration–Effect Relationships

“…Sigmoid curves are obtained when e!ect is plotted against the logarithm of the test concentration (Finney, 1971). The lognormal, log-logistic, and Weibull distributions are frequently used to describe the concentration}e!ect curve (Finney, 1971;Christensen, 1984;Kodell and Felton, 1991;Wagner and L+kke, 1991). Although using a distribution as descriptor for a concentration}e!ect curve is not a realistic method (0 and 100% e!ect levels are the distribution's asymptotes and, therefore, will theoretically never be reached), it is simple, frequently used, and applicable to a relative comparison of the slopes of many concentration}e!ect curves.…”

Corrections to: ‘‘Age distribution of cancer in mice’’

“…The ED01 study appears to be the only research available with data on carcinogenesis in a substantial population of elderly mice with known age. Others have estimated incidence from the ED01 database, and fit it to two-stage models, with variable results (Cohen and Ellwein, 1990;Kodell and Felton, 1991;Lensing and Kodell, 1995;Travis et al, 1996). However, our work here appears to be the only research available on age-specific cancer mortality in a substantial population of older mice.…”

Relationship of DNA Adducts Derived from 2-Acetylaminofluorene to Cell Proliferation and the Induction of Rodent Liver and Bladder Tumors

“…It is by utilizing information of this type in producing estimates of model parameters that are biologically consistent across the various experimental doses and across different experimental regimens that a realistic appraisal of the carcinogenic process can be estimated. With traditional statistical approaches, a unified explanation of the tumor incidence in liver and bladder and in the continuous and discontinuous experiments is not possible (13).…”