Estimating Human Jejunal Effective Permeability Using Numerical Deconvolution of Oral Solution Concentration Data

Accurate determination of human intestinal effective permeability (P_eff) typically relies on technically challenging human intestinal perfusion experiments. This study aimed to develop a numerical deconvolution method to estimate jejunal permeability using concentration data obtained from oral solution and intravenous administration.

Twenty-seven drugs (4 acidic, 11 basic, 7 amphophilic, and 5 neutral) were included using predefined criteria. Numerical deconvolution of intravenous and oral solution data across five time windows (0.25-0.75, 0.5-1, 0.75-1.25, 1-1.5, and 1.5-2 h) was applied to estimate regional jejunal permeability. Estimates were generated under two conditions: uncorrected (P_{eff, uncorr}) and corrected for first-pass metabolism (P_{eff, FP_corr}). The resulting permeability values were compared with observed human perfusion data (P_{eff, obs}), as well as with the reported three-parameter average model and the site-specific data-based deconvolution method.

Correlations of jejunal permeability were performed for 26 drugs (lisinopril excluded due to unreliable P_{eff, obs}). The 0.25-0.75 h window showed the best agreement with observed values. Within this window, correlations with P_{eff, obs} were moderate for P_{eff, uncorr} (R² = 0.47) and strong for P_{eff, FP_corr} (R² = 0.83). The proposed method outperformed the three-parameter average model (R²: 0.83 vs 0.57) and was comparable to the site-specific deconvolution method (R²: 0.94 vs 0.96) using overlapping compounds.

Our oral solution-based numerical deconvolution method, accounting for first-pass metabolism, enables accurate estimation of jejunal permeability.