Discrepancies in pharmacokinetic analysis results obtained by using two standard population pharmacokinetics software programs
Abstract
Multiple standard software packages for population pharmacokinetics (PK) modeling are currently available. These programs may significantly vary in the algorithms used for modeling plasma concentrations as a function of time course. We compared the population PK parameters obtained by using two standard software packages, p-pharm and saam ii, for analysis of a similar data set of serum samples of doxorubicin obtained from 11 infants and children with malignant diseases. Plasma drug concentrations were fitted to time by a two-compartment intra-vascular PK model by saam ii and p-pharm programs. The population parameters obtained from the analysis by the two software programs were substantially different. For example, Vd was almost five times larger when using saam ii compared with p-pharm (9.6 L/kg vs. 2.0 L/kg, respectively), whereas t½β was about 30 times larger in the latter (7.7 h vs. 206.9 h, respectively). When considering the results reported from a population PK analysis, validation of the results by different software should be considered, especially when extreme, unexpected values are obtained.
