Poggesi I (1), Simeoni M (2), Germani M(1), De Nicolao G(2), Rocchetti M(1)
(1)Pharmacia Italia S.p.A., Nerviano, Italy; (2)University of Pavia, Italy
Introduction: We developed a minimal pharmacokinetic-pharmacodynamic model describing the effect of an anticancer agent to the tumor growth in animals (PAGE 2004, Simeoni et al.). In this model, the growth of tumors in non-treated animals (unperturbed growth) is described by an exponential growth followed by a linear growth phase. Three parameters are estimated: the initial tumor mass, the rate constant of the exponential growth and the rate of the linear growth.
Aim: A substantial inter-experiment variability was observed in the tumor growth of non-treated animals. Since this was considered a possible limitation for the prospective use of the PK-PD model, a population analysis was implemented for studying the different sources of variability.
Methods: the weights of tumors in control animals obtained from different experiments were modeled using NONMEM (v. V). The frequency distributions of the various parameters were evaluated. The variability contributions describing the inter-individual and the inter-experiment components were considered.
Results: The model fitted well the experimental data. The distribution of the parameters indicated that, whilst the specific rate of exponential growth was reasonably constant across all experiments, the other parameters were more variable. The inter-individual variability of these parameters was well described using a lognormal distribution.
Conclusion: This population approach allowed a better comprehension of the meaning of the model parameters. The specific rate of exponential growth was more constant across experiments, suggesting that this parameter is related to the growth characteristics of the inoculated cell lines. The initial burden of tumor cells was affected by a higher inter-experiment variability, likely due to possible differences in the experimental procedure of inoculation. Also the rate of tumor growth in the linear phase was variable, possibly due to different immunological reactions among the animals.
Reference: PAGE 13 (2004) Abstr 535 [www.page-meeting.org/?abstract=535]
Poster: poster