FW: Duration of Absorption Time From Depot as Covariate

Dear Paul, Masoud, Leonid and Jeroen, I think your comments and solutions were all excellent. Wanted to add two comments from my experience with around 10 IVIVC studies on sustained-release formulations. At least for these drugs, it was 10 of 10 times possible to establish a link between the release profile in vitro and the absorption profiles in humans. As Leonid wrote, it is often not possible to assume that time-in-vitro and time in vivo correlate 1:1. After applying a nonlinear time transformation, this conversion of in-vitro-time to in-vivo-time held up for various formulations of the same drug. Therefore, I would not use the fraction released in vitro directly as input-function of the amount of drug absorbed in vivo. We developed a fairly flexible and simple absorption model [1] which can accommodate a large range of absorption profiles including sharp peaks, "flat peaks", and dual peaks for cefuroxime axetil (see Fig. 2). If one does not have the full data or time available to implement the elegant approaches from Kiyohiko, Martin, and Masoud, this model may be an alternative. The basic idea of the cefuroxime model is that gastric release can be described by a Michaelis-Menten process and that the associated Vmax can change over time (for example using time past dose or time past last meal as the reference point). We have successfully adapted this model to multiple dose data over various dose levels with or without between occasion variability and the model was robust in NONMEM, SADAPT, & NPAG. Best wishes Juergen Jurgen Bulitta, Ph.D., Senior Scientist Ordway Research Institute, Albany, NY [1] New semiphysiological absorption model to assess the pharmacodynamic profile of cefuroxime axetil using nonparametric and parametric population pharmacokinetics. Bulitta JB, Landersdorfer CB, Kinzig M, Holzgrabe U, Sorgel F. Antimicrob Agents Chemother. 2009 Aug;53(8):3462-71