Dinesh de Alwis and Leon Aarons
Department of Pharmacy, University of Manchester M17 9PL, UK
Ondansetron is a potent and selective 5-HT3 antagonist that is used to prevent and treat chemotherapy and radiotherapy induced nausea and emesis in children and adult cancer patients [1].
In order to develop a population model one needs to clearly describe the relationship between pharmacokinetic structural parameters and covariates such as age, weight and gender. This can be a tedious and time consuming task as the number of covariates and the various combinations are numerous.
Maitre et al [2] suggested a three step approach to develop a population model. This method is dependent on entering covariates in a stepwise manner into a nonlinear mixed effects modelling program (NONMEM), [3], which is a time consuming operation. Mandema et al [4] suggested using a generalized additive model (GAM) to allow nonlinear covariate-parameter relationships to be studied. This method showed that a population or covariate model could be developed outside NONMEM.
In the present study, stepwise regression, was used alongside the three step approach in order to develop covariate models more efficiently outside NONMEM. This was done using ondansetron plasma concentration-time data obtained from four clinical studies: two phase I and two phase II.
The data analysis was first carried out on the full data set (rich data). The full data set was then randomly reduced by 50% and then by 75%. A test data set consisting of one phase I and two phase II studies were used for prediction purposes. Both methods showed good agreement with the full data set but perhaps due to the quality of empirical Bayes estimates, there was greater disparity between the methods with the reduced data sets.
1) Markham A and Sorkin E.M Ondansetron: An update of its therapeutic use in chemotherapy-induced and postoperative nausea and vomiting. Drugs.46:931-957. 1993.
2) Maitre P.O et al. A three-step approach combining bayesian regression and NONMEM population analysis: Application to midazolam. J.Pharmacokin.Biopharm.19:377-384. 1991.
3) Beal S.L and Sheiner L.B NONMEM users guides parts 1-VI, NONMEM project group, University of California, San francisco. 1989.
4) Mandema J.W et al. Building population pharmacokinetic- pharmacodynamic models. Pharmacokin.Biopharm. 20:511-527. 1992.
Reference: PAGE 5 () Abstr 550 [www.page-meeting.org/?abstract=550]
Poster: poster