Human jejune permeability

Winiwarter, S., Bonham, N. M., Ax, F., Hallberg, A., Lennemas, H., Karlen, A. Correlation of human jejunal permeability (in vivo) of drugs with experimentally and theoretically derived parameters. A multivariate data analysis approach. J. Med. Chem. 1998, 41, 4939-4949. 

The strategy for the development of the oral absorption model at pION is illustrated in Fig. 7.58. The human jejunal permeabilities reported by Winiwarter et al. [56] were selected as the in vivo target to simulate by the in vitro model. In particular, three acids, three bases and two nonionized molecules studied by the University of Uppsala group were selected as probes, as shown in Fig. 7.58. They are listed in the descending order of permeabilities in Fig. 7.58. Most peculiar in the ordering is that naproxen, ketoprofen, and piroxicam are at the top of the list, yet these three acids are ionized under in vivo pH conditions and have lipophilicity (log Kj) values near or below zero. The most lipophilic molecules tested, verapamil and carbamazepine... 

Figure 7.59 Human jejunal permeabilities compared to Caco-2 permeabilities from several groups.

How Well Do Caco-2 Permeability Measurements Predict Human Jejunal Permeabilities ... 

Figure 7.59 shows a plot of logP JP (human jejunal permeabilities) vs. log pCaco-2 takgjj from the literature, based on the work of more than 11 laboratories. The r2 for the correlation is 0.62. It is clear from the plot that some laboratories better predicted the HJP than other laboratories. Figure 7.60 shows the plot of the results published by Artursson s group [506,512,603], where r2 was calculated as 0.95, the most impressive value of all the comparisons. It is noteworthy that naproxen, ketoprofen, and piroxicam were not available for the comparison in the Fig. 7.60 plot. 

Table 7.23 shows the results for 47 specific PAMPA models tested at pION, according the the scheme in Fig. 7.58. The two columns on the right are the r2 values in the comparisons. The neutral-lipid models (1.0, 1A.0, 2.0, 3.0, and 4.0) at pH 7.4 do not explain the permeability trend indicated in the human jejunal permeabilities [56]. Octanol was least effective, with r2 0.01. This should not be too surprising, since we did note that the appearance of naproxen, ketoprofen, and piroxicam at the top of the HJP ordering was unexpected. Our expectations were based on the octanol-water lipophilicity scale, which clearly does not correlate with the HJP trend. Adding a sink condition to the 2% DOPC model (model 1.1) improves correlation (r1 increases from 0.33 to 0.53). The addition of cholesterol to the 2% DOPC/dodecane system made the model unstable to the surfactant-created sink condition. 

Figure 7.61 Correlation between human jejunal permeabilities and soy lecithin models (in dodecane) at pH 7.4.

Figure 7.62 Correlation between human jejunal permeabilities [vs. PAMPA (double-sink)] and soy lecithin models under gradient pH conditions.

Figure 7.63 Human jejunal permeabilities compared to pION s double-sink sum-Pe PAM PA GIT model.

The pION double-sink GIT model, with donor pH 5, predicts the human jejunal permeabilities as well as the best reported Caco-2 model (Artursson s), and a lot better than the rest of the reported Caco-2 models, as shown in Fig. 7.63. 

Figure 7.67 Human intestinal absorption compared to (a) pION s double-sink sum-P, PAMPA GIT model and (b) human jejunal permeabilities [56],...

The lengthy permeability chapter (Chapter 7) recounts the study of many different artificial membrane formulations, comparing transport results of each to human jejunal permeabilities. A very promising in vitro screening system was described the double-sink sum-Pe PAMPA GIT model. It is most applicable to molecules that are classified as soluble in the BCS scheme. 

Predictions of in vivo Human Jejunal Permeabilities using the Improved 20% Soy Lecithin with Surfactant in vitro PAMPA Technique... 

The studies of various compositions revealed that the 20% soy lecithin-dodecane membrane with 35 mM in the acceptor wells has substantially improved predictive value compared with the 2% DOPC model. Fine-tuning of the model components may be guided by the in vitro-in vivo (IV-IV) correlations, comparing the improved PAMPA model permeabilities to the human jejunal permeabilities measured by Winiwarter et al. [36] (Table 3.6). Table 3.7 lists the results of comparisons of various models. The best correlations were realized with the 20% soy lecithin-dodecane system, employing 35 mM SLS in the acceptor compartment, but a better... 

Tab. 3.7. in vitro-in vivo correlations, PAMPA versus human jejunal permeabilities. 

We have developed a new set of rules, called J-Alert , that contains cut-offs for human jejunal permeability, water solubility, the number of oxygen-based H-bond acceptors, partial charge sums on the oxygen-based H-bond donors and acceptors, and a low level cut-off for the Moriguchi log P [28]. The J-Alert rules and a two-letter abbreviation are listed as follows ... 

QMPRPlus was used to generate in silico estimates of log P, aqueous solubility, and human jejunal permeability from 3D molecular structures. The predictive... 

Winiwarter S, Bonham NM, Ax F, Hallberg A, Lennernas H and Karlen A (1998) Correlation of Human Jejunal Permeability (in Vivo) of Drugs With Experimentally and Theoretically Derived Parameters A Multivariate Data Analysis Approach. J Med Chem 41 pp 4939 1949. 

Human Jejune Permeability. Lennernas has reported data on 38 molecules, some of which were collected from the literature. A larger set of 51 molecules (45 drugs and six amino acids) was used by SimulationsPlus to develop their commercially available model from the previous publication of the same author and other literature sources. 

Figure 17.1 Correlation of rank order for ADMET Risk and human intestinal absorption (H IA%). The ADMET Risk score ranged from 0 to 5, with 5 being compounds with the greatest risk of having poor ADMET properties. Within a single ADMET Risk number, the compounds were ranked according to ascending estimated human jejunal permeability (ADMET Predictor, Simulations Plus, Inc.). Spearman rank correlation coefficient was 0.7 (pcO.OOl).

We have developed a two-step procedure for the in silico screening of compound libraries based on biopharmaceutical property estimation linked to a mechanistic simulation of GI absorption. The first step involves biopharmaceutical property estimation by application of machine learning procedures to empirical data modeled with a set of molecular descriptors derived from 2D and 3D molecular structures. In silico methods were used to estimate such biopharmaceutical properties as effective human jejunal permeability, cell culture permeability, aqueous solubility, and molecular diffusivity. In the second step, differential equations for the advanced compartmental absorption and transit model were numerically integrated to determine the rate, extent, and approximate GI location of drug liberation (for controlled release), dissolution, and absorption. Figure 17.3 shows the schematic diagram of the ACAT model in which each one of the arrows represents an ordinary differential equation (ODE). 

---