PKQuest: capillary permeability limitation and plasma protein binding - application to human inulin, dicloxacillin and ceftriaxone pharmacokinetics

Abstract

Background: It is generally assumed that the tissue exchange of antibiotics is flow limited (complete equilibration between the capillary and the tissue water). This assumption may not be valid if there is a large amount of plasma protein binding because the effective capillary permeability depends on the product of the intrinsic capillary permeability (PS) and the fraction of solute that is free in the blood (fwB). PKQuest, a new generic physiologically based pharmacokinetic software routine (PBPK), provides a novel approach to modeling capillary permeability in which the only adjustable parameter is the PS of muscle.

Methods: All the results were obtained by applying PKQuest to previously published human pharmacokinetic data.

Results: The PKQuest analysis suggests that the highly protein bound antibiotics dicloxacillin and ceftriaxone have a significant capillary permeability limitation. The human muscle capillary PS of inulin, dicloxacillin and ceftriaxone was 0.6, 13 and 6 ml/min/100 gm, respectively. The ceftriaxone protein binding is non-linear, saturating at high plasma concentrations. The experimental ceftriaxone data over a wide range of intravenous inputs (0.15 to 3 gms) was well described by PKQuest. PKQuest is the first PBPK that includes both permeability limitation and non-linear binding.

Conclusions: Because of their high degree of plasma protein binding, dicloxacillin and ceftriaxone appear to have a diffusion limited exchange rate between the blood and tissue and are not flow limited as had been previously assumed. PKQuest and all the examples are freely available at http:\\www.pkquest.com.

Figure 1

Comparison of the model prediction (solid line) and the experimental data (squares) for inulin venous plasma concentration.

Figure 2

Comparison of the model prediction (solid line) and the experimental data (squares) for dicloxacillin venous plasma concentration. Left panel: Capillary permeability limitation (fclear_muscle 0.26). Right panel: Flow limited – no permeability limitation (fclear_muscle = 1.0).

Figure 3

Comparison of the capillary permeability limited model prediction (solid line) and the experimental data (squares) for ceftriaxone venous plasma concentrations. The four panels correspond to inputs of 0.15, 0.5, 1.5 and 3.0 grams given as a 5 minute constant IV infusion. Because the protein binding saturates, the effective capillary permeability limitation (fclear) decreases as the concentration increases.

Figure 4

Comparison of the flow limited model prediction (solid line) and the experimental data (squares) for ceftriaxone venous plasma concentrations. Left panal: 0.15 gm IV input; Right panel: 3.0 gm input.

Figure 5

Free plasma (red) and tissue (black) water ceftriaxone concentration. The dashed lines correspond to the flow limited model results and the solid lines are for the case of a capillary permeability limitation. The PBPK parameters and input conditions are identical to those used in figs. 3 and 4.