POPULATION-KINETICS AND CONDITIONAL ASSESSMENT OF THE OPTIMAL DOSAGE REGIMEN USING THE P-PHARM SOFTWARE PACKAGE

The adjustment of individual dosage regimen is an adaptive control pro cess based upon an individual response to a pharmacokinetic model. To attain this objective, if is very helpful to know the characteristics of the population to which the subject belongs, in terms of mean param eters and interindividual variability. Usually the available informati on consists of incomplete and sparse data. For this reason it is essen tial to employ a computational methodology based on non-linear mixed-e ffect procedures in order to obtain a population parameter estimate. A Bayesian methodology can then be applied from the population paramete rs to the specific data for the individual requiring a dosage adjustme nt (such data includes drug concentration(s) of the active drug, demog raphic data, etc). The result of the Bayesian calculation supplies the required individual pharmacokinetic parameters. An optimal dosage reg imen can be defined on the basis of therapeutical criteria (concentrat ion ranges) as well as practical constraints such as: the size of avai lable unitary drug dosages, feasible drug intake times, penalties asso ciated with expected concentrations falling outside the therapeutic co ncentration ranges. In this paper we present the methodology and resul ts obtained using the P-Pharm software tool. P-Pharm implements a non- linear mixed-effect population parameter estimation algorithm based on the EM algorithm. This method allows the inclusion of explicit variab les into the calculations, it implements an individual Bayesian parame ter estimation procedure and also an algorithm for the conditional ass essment of the optimal dosage regimen given a list of practical constr aints.