Abstract
The present approach enables a noncompartmental assessment of log-concave plasma concentration-time profiles following oral drug administration. Observed log-concavity corresponds to a nonparametric class of residence time distributions with the following properties: (1) The fractional rate of elimination kB(t) (failure rate of the distribution) increases monolonically until reaching the terminal exponential coefficient kB,Z.(2) The relative dispersion of body residence times CV B 2 (ratio of variance to the squared mean , VBRT/MBRT2,)acts as a shape parameter of the curve. The role of the input process in determining the shape of the concentration profile is discussed. In this connection evidence is provided for the importance of log-concave percent undissolved versus time plots, introducing the general concept of a time-varying fractional rate of dissolution. The governing factor for the appearance of log-concavity is the ratio of mean absorption time to mean disposition residence time (MAT/MDRT);this factor exceeds a particular threshold value which depends on the distributional properties of the drug. Generalizing previous approaches which are valid for first-order input processes, the “flipflop” phenomenon and the problem of “vanishing of exponential terms” are explained using fewer assumptions. Upper bounds for the elimination time (more than 90% eliminated) and the cutoff error in AUCdetermination are presented. The concept of logconcavity reveals general features of the pharmacokinetic behavior of oral dosage forms exhibiting a dominating influence of the absorption/dissolution process.

This publication has 14 references indexed in Scilit: