Comprehensive study on regional human intestinal permeability and prediction of fraction absorbed of drugs using the Ussing chamber technique

The purpose of this study was to evaluate the use of human intestinal tissue in Ussing chamber to predict oral and colonic drug absorption and intestinal metabolism. Data on viability, correlation between apparent permeability coefficients (P(app)) and fraction absorbed (f(a)) after oral and colonic administration, regional permeability, active uptake and efflux of drugs as well as intestinal metabolism were compiled from experiments using 159 human donors. Permeability coefficients for up to 28 drugs were determined using one or several of four intestinal regions: duodenum, jejunum, ileum and colon and 10 drugs were studied bidirectionally. Viability was monitored simultaneously with transport experiments by recording potential difference (PD), short-circuit current (SCC) and the resistance (TER). Intestinal metabolism was studied using testosterone and midazolam as probe substrates. There was a steep sigmoidal correlation between P(app) in the Ussing chamber, using jejunal segments, and oral f(a) in humans, for a set of 25 drugs (R(2): 0.85, p<0.01). A clear sigmoidal relationship was also obtained between P(app) in colonic segments and f(a) after colonic administration in humans for a set of 10 drugs (R(2): 0.93, p<0.05). Regional permeability data showed a tendency for highly permeable compounds to have higher or similar P(app) in colon as in the small intestinal segments, while the colonic regions showed a lower P(app) for more polar compounds as well as for d-glucose and l-leucine. Bidirectional transport (mucosa to serosa and serosa to mucosa direction) in jejunum showed well functioning efflux- and uptake asymmetry. Intestinal metabolic extraction during transport across jejunum segments was found for both testosterone and midazolam. In conclusion, viable excised human intestine mounted in the Ussing chamber, is a powerful technique for predicting regional fraction absorbed (f(a)), transporter-mediated uptake or efflux as well as intestinal metabolism of drug candidates in man. Furthermore, a sigmoidal relationship of P(app) vs. f(a) was obtained when permeability data from the present study were merged with data from 2 other independent laboratories (R(2): 0.83, p<0.01). The correlation curve reported can be used by any laboratory for predictions of human permeability and f(a)(.) In addition, for the first time a correlation curve between colonic P(app) and human colonic f(a) is reported, which demonstrates the usefulness of this methodology in early assessment of the colonic absorption potential of extended release formulation candidates.