Analysis of intestinal perfusion data for highly permeable drugs using a numerical aqueous resistance--nonlinear regression method.

PURPOSE: To develop, validate and apply a method for analyzing the intestinal perfusion data of highly permeable compounds using the Numerical Aqueous Resistance (NAR) theory and nonlinear regression (NAR-NLR) and to compare the results with the well-established Modified Boundary Layer (MBL) Analysis. METHODS: The NAR-NLR method was validated and the results were compared to the MBL analysis results using previously reported cephradine jejunal perfusion data. Using the Single Pass Intestinal Perfusion (SPIP) method, the concentration dependence of intestinal permeability was investigated for formycin B, proline, and thymidine, three compounds reported to be absorbed by carrier-mediated transport processes. The MBL and NAR-NLR analyses were then applied to the three sets of SPIP data. RESULTS: The results demonstrate that the intrinsic MBL transport parameters were highly variable and, in one case, the analyses failed to give a statistically significant Michaelis constant. The MBL mean dimensionless wall permeabilities (P*w) were greater than the NAR-NLR P*w and were also highly variable. In all cases, the NAR-NLR variability was significantly lower than the MBL variability. The extreme variability in the MBL-calculated P*w is due to the sensitivity of P*w when the fraction of unabsorbed drug (Cm/Co) is low or, alternatively, when P*w approached the aqueous permeability, P*aq. CONCLUSIONS: The NAR-NLR method facilitates the analysis of intestinal perfusion data for highly permeable compounds such as those absorbed by carrier-mediated processes at concentrations below their Km. The method also allows for the use of a wider range of flow conditions than the MBL analysis resulting in more reliable and less variable estimates of intestinal transport parameters as well as intestinal wall permeabilities.