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

(log Kd — 2.5; cf. Table 7.4), are in second rank ordering. We took it as a challenge to explain these anomalies in our optimized in vitro GIT model.

As Fig. 7.58 indicates, our task was to explain the ordering of the eight probe molecules in the human in vivo target, but subjecting the eight probe molecules to each of the 50 PAMPA lipid models. For each PAMPA model, the regression correlation coefficient, r2, was used to assess the appropriateness of the model.

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

Since the widely accepted in vitro permeability model in the pharmaceutical industry is based on the use of cultured cells, such as Caco-2 or MDCK, it was appropriate to analyze the regression correlation coefficients based on the comparisons of Caco-2 log Pe and the log Pe values based on the human jejunal measurements [56].

Figure 7.59 shows a plot of log PHJP (human jejunal permeabilities) vs. log pCaco-2 taken 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.

Figure 7.60 Human jejunal permeabilities compared to Caco-2 permeabilities from Artursson's group.

log Pe Caco"2

Figure 7.60 Human jejunal permeabilities compared to Caco-2 permeabilities from Artursson's group.

7.8.3 How Well Do PAMPA Measurements Predict the Human Jejunal Permeabilities?

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 (r2 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.

Introducing negative-charge phospholipids to the 2% DOPC at pH 7.4 improved the correlations significantly (models 5.0, 6.0, 7.0, 8.0, 9.0, 10.0). Sink conditions only marginally improved the correlations for the dodecylcarboxylic acid (1.1% DA) and phosphatidic acid (0.6% PA) models (models 6.1 and 7.1). The phospha-tidylglycerol (PG) models (models 9.1 and 10.1) did not correlate well under sink conditions. The modified Chugai model at pH 7.4 performed well (r2 0.60), but was unstable under sink conditions.

Several egg lecithin models were tested at pH 7.4. The Avanti egg lecithin behaved differently from the Sigma-Aldrich egg lecithin, and was unstable under sink conditions when cholesterol was added. The correlation coefficients were

TABLE 7.23 Correlation (r2) between Human Jejunal and PAMPA Permeabilities





No Sink

With Sink

0 0

Post a comment