A stochastic clonal growth model for describing quantitative changes in size and number of putative preneoplastic lesions was modified to analyze the time-course information of cell proliferation and glutathione S-transferase pi (GST-P) foci within a medium-term bioassay. The study used F344 rats and a single initiating event using diethylnitrosamine (200 mg/kg ip) at Week 0. After a 2-week recovery period, chemical treatment began by gavage administration of pentachlorobenzene (PeCB; 100 micromol/kg/day, 7 days/week) in a corn oil vehicle and continued for 6 weeks. One week after beginning gavage dosing, a two-thirds partial hepatectomy was performed and the animals were serially euthanized at 48, 120, 168, 624, and 840 h postsurgery, which corresponds to 216, 288, 336, 792, and 1008 h following the beginning of PeCB treatment, respectively. For analysis, two types of models were evaluated for describing the time-course changes in GST-P foci. First, a sequential model describing the transformation of normal cells into a homogenous initiated cell population (i.e., one-cell model). Second, a two-cell model that describes a heterogeneous foci population by splitting the initiated cell population into two distinct types. In our study, the one-cell model was unable to adequately represent the time-course data for changes in both size and number of foci. In contrast, the two-cell model, which was parameterized to describe a negative selection mechanism, produced adequate simulations of both the size and number of foci. This model-based analysis suggested that the differences between PeCB-treated and untreated animals were primarily in parameters involving the rates of cell death.
Copyright 2000 Academic Press.