Vangelis Karalis, Aristides Dokoumetzidis, Panos Macheras
Laboratory of Biopharmaceutics & Pharmacokinetics, National & Kapodistrian University of Athens, Athens, Greece
Objectives: To develop a criterion using the dC/dt versus C plot for testing whether the slope of the terminal (logC vs. t) data points, represents the real exponential terminal phase of the data points.
Methods: For classical first-order elimination kinetics in one or a multi-compartment system, the relationship between dC/dt and C for the terminal phase data points is: dC/dt = – Kz*C (Eq.1) where Kz refers either to the elimination rate constant for the one-compartment system or to the smallest of the rate constants associated with the multi-compartment system. According to Eq.1, the dC/dt versus C plot of the terminal segment data points is a straight line with slope matching Kz and zero y-intercept. Due to the errors associated with the variables dC/dt, C, simulation studies were performed using errorless and contaminated with error C, t data. Several scenarios were investigated such as pharmacokinetic scenarios (e.g., one- and two-compartment models), sampling schemes, approximation method for the dC/dt estimate etc. In all cases, the slope and the intercept of the regression lines were calculated using Eq.1.
Results: For errorless data and sampling schemes ensuring real terminal phase data points following exponential elimination, the intercept values were found to be very close (non statistically significant) to zero. This small deviation is associated with the calculation errors of dC/dt. Negative y-intercepts statistically significant than zero were calculated for experimental data reported in the literature obeying power law elimination. For C, t error data generated from linear one and two compartment models, the y- intercept estimate is related to the sampling scheme and the standard error of the Kz estimate. In this case, a relationship between the y-intercept estimate, the standard error of the Kz estimate and the relative ratio of AUC of the area up to the last quantifiable concentration over area extrapolated to infinity was developed.
Conclusions: A computational method was explored for the validation of the terminal slope estimate relying on the analysis of terminal data points. Significant components for this criterion were found to be the relative ratio of AUC and the standard error of the Kz estimate.
Reference: PAGE 24 (2015) Abstr 3557 [www.page-meeting.org/?abstract=3557]
Poster: Methodology - Estimation Methods