- Email: [email protected]
Prediction of Absorbability of Poorly Water-soluble Drugs Based on Permeability Obtained through Modified In Vitro Chamber Method
Drug Metab. Pharmacokinet. 20 (6): 428–434 (2005).
Regular Article Prediction of Absorbability of Poorly Water-soluble Drugs Based on Permeability Obtained through Modiˆed In Vitro Chamber Method Etsushi WATANABE1, Noriyuki KINOSHITA1, Masayuki TAKAHASHI2 and Masahiro HAYASHI1 1Faculty 2Pharmaceutical
of Pharmaceutical Sciences, Science University of Tokyo, Japan Formulation Research Laboratory, Daiichi Pharmaceutical Co., LTD., Tokyo, Japan
Full text of this paper is available at http://www.jstage.jst.go.jp/browse/dmpk
Summary: Permeability is an underlying parameter to control drug absorption. For highly water-soluble drugs, the high correlation between their permeability and fraction absorbed in humans is reported. In the present study, to predict the absorbability of poorly water-soluble drugs in humans, a new experimental method of the permeation study was proposed and subjected to examination. Firstly, using the in vitro chamber method modiˆed to contain 5z (ˆnal concentration) dimethyl sulufoxide (DMSO) in both compartments of the chamber (DMSO-MS), the eŠect of DMSO on membrane integrity was evaluated. Secondly, the correlation between the apparent permeability coe‹cients (Papp) obtained through DMSO-M or DMSO-MS and fractions absorbed in humans were investigated using 7 poorly water-soluble drugs. Membrane integrity of the rat intestinal tissues was maintained after using DMSO-MS, as with that after using the conventional in vitro chamber method. Papp of two paracellular markers obtained through DMSO-MS was not diŠerent from that obtained through the conventional chamber method. In the permeation study of the P-glycoprotein substrate, Papp from both mucosal to serosal and serosal to mucosal sides obtained through DMSO-MS was not signiˆcantly diŠerent from that obtained through the conventional chamber method. The correlation between Papp obtained through DMSO-MS and Fa which was expressed by the equation of Fa＝1－exp (－Papp×1.38×105) (r2＝0.980), was more favorable than the correlation between Papp obtained through DMSO-M and Fa which was expressed by the equation of Fa＝ 1－exp (－Papp×2.12×105) (r2＝0.875). These results showed that DMSO-MS was a useful method for predicting the absorbability of poorly water-soluble drugs.
Key words: poorly water-soluble drugs; permeability; dimethyl sulfoxide; intestinal tissue; fraction absorbed chamber experiment using isolated rat intestinal tissues and Fa in humans.3) Based on above results, the permeability of water-soluble drugs was an underlying parameter for estimating their absorbability in the development of a new chemical entity. In contrast, for poorly water-soluble drugs, prediction of their absorbability requires their dissolution rates as well as their permeability, because of their slow and incomplete dissolution in the gastrointestinal tract.1) However, many trials have been conducted to improve the solubility of poorly water-soluble drugs by using the techniques and materials such as solid dispersion, melt extrusion, cyclodextrin and sealing liquids into hard
Introduction According to a theory of the macroscopic mass balance model proposed by Amidon et al., the intestinal absorbability of a water-soluble drug which is rapidly and completely dissolved in the gastrointestinal tract can be determined by its permeability.1) Arturrson et al. report a good correlation between the apparent permeability coe‹cient (Papp) of water-soluble drugs obtained in the in vitro chamber experiment using Caco-2 cell monolayers, and fraction absorbed (Fa) in humans.2) Additionally, we had previously shown a good correlation between Papp obtained in the in vitro
Received; July 6, 2005, Accepted; September 16, 2005 To whom correspondence should be addressed : Etsushi WATANABE, Technology Administration Department, Daiichi Pharmaceutical Co., LTD., 1-8 Nihonbashi-Koamicho, Chuo-ku, Tokyo, 103-8541, JAPAN. Tel. ＋81-3-5640-0655; Fax. ＋81-3-5640-1744, E-mail: watanqsf＠ daiichipharm.co.jp
428
Prediction of Absorbability of Poorly Water-soluble Drugs
gelatin capsules.4–6) Using these techniques, the drug with high permeability is considered to have the high absorbability regardless of its intrinsic solubility. Therefore, it is necessary for the recent development of a new chemical entity to estimate its permeability regardless of its solubility. We had previously predicted Fa values of 5 poorly water-soluble drugs based on their Papp obtained through the modiˆed chamber method with an addition of 5z dimethyl sulufoxide (DMSO) in the mucosal side (DMSO-M).3,7) However, predicted Fa values of griseofulvin, nifedipine and indomethacin were not favorably comparable with their Fa values in humans.3) In the present study, a new experimental method in which the chamber method was modiˆed with an addition of 5z DMSO to both sides of chamber (DMSO-MS) was proposed, and subjected to examination. Firstly, the eŠects of the DMSO on the membrane integrity were evaluated by the measurement of released enzymes and membrane conductance. Secondly, the eŠects of DMSO on the membrane permeability were investigated using Papp of ‰uorescein isothiocyanatelabeled dextran 4000, sulfasalazine, antipyrine and rhodamine-123. Finally, the relationship between Papp values of 7 poorly water-soluble drugs obtained through DMSO-MS and their Fa values in humans8–14) was investigated. Methods
Materials: Fluorescein isothiocyanate-labeled dextran 4000 (FD4), sulfasalazine (SFZ), antipyrine (APY), rhodamine-123 (Rho123), indomethacin (IDM), triamterene (TAT), nifedipine (NFD), phenytoin (PHT), griseofulvin (GSF), azathioprine (AZT) and desipramine (DSP) were purchased from Sigma Chemical Co., Ltd. (St. Louis, MO). LDH CII (#271–54101) and ALP enzyme kit (#274–04401) were purchased from Wako Pure Chem. (Osaka, Japan). Other chemicals used were of the analytical grade. Animals: Male Wistar rats at ages of 7 weeks or older (body weight: about 200–250 g) were used. The rats were fasted for about 24 h with free access to water before the experiment. Necessary approvals for the experimental protocol of animals were obtained from the Ethical Committee of Science University of Tokyo (Tokyo, Japan). Experiment using modiˆed in vitro chamber method (DMSO-MS): Rats were anesthetized with urethane, and the abdominal wall was cut open along the midline. A 2 to 3 cm jejunum segment was excised and placed in warm Ringer's solution (pH of 7.4) in an atmosphere of 5z CO2 gas mixture. The epithelium was a 95z O2 W then exposed via a longitudinal incision along the mesenterium, and the underlying muscle layer was removed. Finally the epithelium was mounted onto an Us-
429
sing-type chamber having a tissue surface area of 0.64 cm2. Ringer's solution (5 mL) was added to both the donor (mucosal) and receiver (serosal) chambers. After 10-min preincubation, drug solution (5 mL) solubilized by Ringer's solution and Ringer's solution (5 mL), both of which included 5z (ˆnal concentration) DMSO, were added to the mucosal and serosal sides, respectively. The initial concentrations of drug solutions were 0.2 mM, except for Rho123 solution (0.026 mM). After the permeation study was started, samples (100 mL) were withdrawn from both sides at regular intervals (20 min.) up to 100 min. In permeation study, Papp from mucosal to serosal sides was basically evaluated. On the other hand, in the study of Rho123, Papp from serosal to mucosal sides as well as that from mucosal to serosal sides was evaluated. In control group, the conventional chamber experiment was performed to obtain the permeation of FD4, SDZ, APY and Rho123 in Ringer's solution without DMSO. For Rho123, we examined the transport from the serosal to mucosal sides as well as that from mucosal to serosal sides. Details of DMSO-M had been previously reported.3,7) In this study, experiments were conducted in 3 groups of control, DMSO-M and DMSO-MS groups using the control method, DMSO-M and DMSO-MS, respectively. Estimation of membrane integrity: Enzymes (LDH and ALP) released from the intestinal tissues were evaluated in the mucosal side buŠer 100 min after the start of permeation study. ALP activity was determined using the ALP enzyme kit by the Bessey-Lowry method. LDH activity was determined using the LDH CII kit (i.e., by spectrophotometeric determination (l＝560 nm) of diformazan formation). Membrane electrical parameters were evaluated by the membrane conductance (DGt) measured by the methods of Yamashita et al. and Sawada et al.15,16) Assay: The concentrations of FD4 and Rho123 were determined using a ‰uorescent microplate reader operating at excitation wavelength (EX) of 480 nm and emission wavelength (EM) of 520 nm, and EX of 486 nm and EM of 546 nm, respectively. The concentrations of other drugs were determined by HPLC (Alliance 2690, Waters Corporation, MA, USA) using a Waters 2487 Dual Absorbance Detector (Waters). Details of the HPLC conditions had been previously reported.3,7,12) Data Analysis: The apparent permeability coe‹cient per unit membrane surface area (Papp (cm W sec)) was calculated according to the following equation: Papp＝dM W dt×1 W AC0 where dM W dt is the steady-state appearance rate of drugs to the serosal side ( mmol W sec), C0 is the initial drug concentration on the mucosal side (mM), and A is the surface area of the membrane (cm2).7,17,18) Fa values
430
Etsushi WATANABE et al.
of 10 water-soluble drugs and 7 poorly water-soluble drugs were quoted from the literature.2,8–14,19–21) When the quoted Fa values had the certain range, median value was used. The relationship between in vitro permeability of water-soluble drugs in rats and in vivo absorption in humans was investigated using a complete radial mixing model (CRM).19,22) The numerical expression, which indicated relationship of Papp (cm W sec) of 7 poorlywater-soluble drugs as well as 10 water-soluble drugs with fraction absorbed in humans (Fa), was described using the following equation: Fa＝1－exp (－Papp×f ) where f was the lump constant including the factors of eŠective surface area of the intestine, transit time and scale-up factor in species diŠerence between rat and human. The relationship between Papp and Fa expressed as f was determined by non-linear regression analysis (Levenberg-Marquardt least squares algorithm; Delta Graph 4.0(r) by SPSS Inc., USA). Statistical Analysis: Results are expressed as means ±S.D. For multiple group analysis of variance, one-
Fig. 1. Release of alkaline phosphatase (ALP) and lactate dehydrogenase (LDH) from the intestinal tissue in control, DMSO-M and DMSO-MS groups. Data represent the means±S.D. of 3 experiments.
factor ANOVA was used. A p value of less than 0.05 was considered to be statistically signiˆcant. Results EŠects of DMSO addition on membrane integrity: Figure 1 shows the results of enzyme release measurement on the mucosal side 100 min. after the start of experiment in control, DMSO-M and DMSO-MS groups. There was no signiˆcant diŠerence in the release of ALP or LDH among the three groups. No signiˆcant changes in DGt were induced by DMSO-M or DMSOMS (Fig. 2). EŠects of DMSO addition on Papp of FD4, SFZ and APY: Figures 3 and 4 show Papp of FD4, SFZ and APY obtained in control, DMSO-M and DMSO-MS groups. No signiˆcant diŠerence in Papp of FD4, SFZ (Fig. 3) or APY (Fig. 4) was observed among the three groups. EŠects of DMSO addition on Papp of Rho123: Figure 5 shows Papp of Rho123 from mucosal to serosal sides and that from serosal to mucosal sides in control, DMSO-M and DMSO-MS groups. No signiˆcant diŠerence in Papp of Rho123 transported in either direction was observed among the three groups.
Fig. 2. Changes of membrane conductance (DGt) in control, DMSO-M and DMSO-MS groups. Data represent the means±S.D. of 3 experiments.
Fig. 3. Apparent permeability coe‹cients of paracellular permeant compounds of ‰uorescein isothiocyanate-labeled dextran 4000 (FD4) and sulfasalazine (SFZ), in control, DMSO-M and DMSO-MS groups. Data represent the means±S.D. of 3 experiments.
431
Prediction of Absorbability of Poorly Water-soluble Drugs
Papp and Fa of poorly water-soluble drugs: Papp values of 7 poorly water-soluble drugs were obtained through DMSO-M and DMSO-MS. Table 1 shows Papp values, predicted Fa values and actual Fa in humans. Papp
values of IDM and NFD obtained through DMSO-MS were signiˆcantly diŠerent from those obtained through DMSO-M. Estimation of relationship between Papp of 7 poorly water-soluble drugs and their Fa in humans: The correlation curve (solid line) between Papp of 7 poorly water-soluble drugs obtained through DMSO-M or DMSO-MS and their Fa in human is shown in Fig. 6. The correlation curve (dashed line) of 10 water-soluble drugs reported previously (Fa＝1－exp (－Papp×1.51× 105)) is overlaid with Fig. 6.3) The best ˆt for the data obtained through DMSO-M and DMSO-MS was observed when f was 2.12×105 (r2＝0.875) and 1.38× 105 (r2＝0.980), respectively. Discussion
Fig. 4. Apparent permeability coe‹cients of a transcellular permeant compound of antipyrine (APY) in control, DMSO-M and DMSO-MS groups. Data represent the means±S.D. of 3 experiments.
We had previously reported that Papp of 5 poorly water-soluble drugs could be obtained through a new experimental condition by adding 5z DMSO to the mucosal side of the in vitro chamber mounted on the
Fig. 5. Apparent permeability coe‹cients of a P-gp substrate of rhodamine 123 (Rho123) in control, DMSO-M and DMSO-MS groups. In each group, the transport of Rho123 both from mucosal to serosal (M to S) sides and from serosal to mucosal (S to M) sides was examined. Data represent the means±S.D. of 3 experiments.
Table 1. Apparent permeability coe‹cients (Papp) of poorly water-soluble drugs obtained through DMSO-M and DMSO-MS, and predicted fractions absorbed (Fa) in humans DMSO-M Poorly Water-soluble Drugs
Triamterene Desipramine Azathioprine Indomethacin Nifedipine Phenytoin Griseofulvin a
DMSO-MS
a
Papp, mean±S.D. [10－ 6 cm W sec]
Predicted Fab
Papp, mean±S.D. [10－ 6 cm W sec]
Predicted Fab
5.11±1.91 5.24±1.06 5.53±0.38 9.04±1.35 15.12±3.19 22.01±2.26 22.12±2.32
0.54 0.55 0.57 0.74 0.90 0.96 0.96
6.32±0.33 6.03±0.96 6.63±1.42 26.93±3.14** 43.17±5.07** 27.47±2.81 32.13±5.48
0.61 0.60 0.63 0.98 1.00 0.98 0.99
Solubilized by 5z (ˆnal concentration) dimethylsulfoxide Predicted based on the relationship between Papp and Fa of water-soluble drugs. c Quoted from references shown in parentheses. d Fa value of GSF was calculated based on the results reported by Bates et al. (13 and 14). **pº0.01 compared with DMSO-M b
Previously Reported Fa in Humansc
0.40–0.70(10) 0.50(12) 0.60(12) 1.00(8) 0.90–1.00(28) 1.00(9) 1.00(13,14)d
432
Etsushi WATANABE et al.
Fig. 6. The theoretical line calculated using the data of the fraction absorbed in humans versus the apparent permeability coe‹cients (Papp) of 7 poorly water-soluble drugs obtained through DMSO-M and DMSO-MS (solid line). The fraction absorbed was calculated based on the data in references of 8, 9, 10, 11, 13 and 14. The dashed line was the theoretical line calculated using the data of the 10 water-soluble drugs by non-linear regression as described in a previously report (3). Papp values represent the mean of the data from more than three experiments.
isolated rat intestinal tissue.3) Additionally, we predicted Fa values of these poorly water-soluble drugs based on the correlation between permeability of 10 water-soluble drugs and their fraction absorbed in humans. Compared with Fa in humans, predicted Fa may be overestimated for GSF and underestimated for NFD and IDM.3) In our previous report, the human Fa of GSF was quoted from clinical data obtained after oral administration of 500 mg tablets.3) This Fa value was not suitable for the measurement of absorbability of GSF, since GSF in this dosage form might be incompletely dissolved in the gastrointestinal tract.1) In the present study, Fa of GSF in humans was quoted from clinical data obtained after oral administration of GSF solution. Bates et al. report that the relative bioavailability of GSF dissolved in mono chloroform and encapsulated in gelatin is 150z, while the absolute bioavailability of pulverized GSF encapsulated in gelatin is 70z.13,14) Therefore, Fa value of GSF when GSF was completely dissolved in the gastro-intestinal tract was 1.0. As previously reported, log D of TAT, IDM, NFD, PHT and GSF was 1.6, 1.0, 3.2, 2.3 and 2.2, respectively.3) In addition, log P of AZT and DSP was reported to be 0.1 and 3.0, respectively.23,24) Though log D of NFD was very high,3,19) that of IDM was very low (1.01) because of almost complete ionization in neutral pH.3,25) On the other hand, Tashiro et al. reported that a high log P of nonionic type IDM was 4.4.25) Yamashita et al. suggested that the permeation of highly lipophilic drugs such as IDM and NFD from the cells to the serosal side through the intestinal tissue membrane appeared to be limited.26) Therefore, to estimate the permeability of poorly water-soluble drugs with high lipophilicity, it is necessary to improve their solubility and the lipophilicity in the serosal side medium. In the present study, we proposed a method in which 5z DMSO was added to the serosal side as well as the mucosal side.
Release of ALP and LDH was considered to represent the toxicity to the brush border and epithelial membranes, respectively.27) In DMSO-MS, the release of ALP and LDH was not aŠected (Fig. 1). DGt, an indicator of opening of a tight junction, was not signiˆcantly changed (Fig. 2). These results indicated that the membrane integrity was not aŠected by using DMSO-MS. To investigate the eŠect on permeability, we estimated Papp of paracellular permeant compounds such as FD4 (high molecule) and SFZ (low molecule) (Fig. 3), as well as that of APY (Fig. 4) which was a representative transcellular permeant compound. Papp of either compound obtained through DMSO-MS was not signiˆcantly diŠerent from that obtained through the control method or DMSO-M. Additionally, we estimated Papp of Rho123, a P-gp substrate (Fig. 5). Regardless of the transport directions, there was no signiˆcant diŠerence in Papp of Rho123 between DMSO-MS and control groups. Based on these results, we concluded that DMSO-MS could be applied to the transport study of poorly water-soluble drugs. During the experiment, either poorly water-soluble drug did not precipitate out in the transport medium. Therefore, the membrane permeability of poorly water-soluble drugs could be obtained in the same manner as that of water-soluble drugs. Papp of highly lipophilic drugs such as IDM and NFD in DMSO-MS group was higher than that in DMSO-M group (Table 1). This might be due to increased lipophilicity of receiver solution in the serosal side. The correlation between Papp obtained through DMSO-MS and Fa, which was expressed by the equation of Fa＝1－exp (－Papp×1.38×105) (r2＝0.980), was more favorable than that between Papp obtained through DMSO-M and Fa, which was expressed by the equation of Fa＝1－exp (－Papp×2.12×105) (r2＝0.875) (Table 1, Fig. 6). In conclusion, we established a new method for
Prediction of Absorbability of Poorly Water-soluble Drugs
predicting the drug permeability of poorly water-soluble drugs. This method was easy to be operated and likely to give reproducible data, and therefore could be widely applied to the studies of drug permeability. Our results suggested that DMSO-MS was useful method for predicting Fa value of poorly water-soluble drugs including those with high lipophilicity.
13)
14)
References 1)
2)
3)
4)
5)
6)
7)
8)
9)
10)
11)
12)
Amidon, G. L., Lennernas, H., Shah, V. P. and Crison, J. R.: A theoretical basis for a biopharmaceutic drug classiˆcation: the correlation of in vitro drug product dissolution and in vivo bioavailability. Pharm. Res., 12: 413–420 (1995). Artursson, P. and Karlsson, J.: Correlation between oral drug absorption in humans and apparent drug permeability coe‹cients in human intestinal epithelial (Caco-2) cells. Biochem. Biophys. Res. Commun., 175: 880–885 (1991). Watanabe, E., Takahashi, M. and Hayashi, M.: A possibility to predict the absorbability of poorly watersoluble drugs in humans based on rat intestinal permeability assessed by an in vitro chamber method. Eur. J. Pharm. Biopharm., 58: 659–665 (2004). Hulsmann, S., Backensfeld, T., Keitel, S. and Bodmeier, R.: Melt extrusion–an alternative method for enhancing the dissolution rate of 17beta-estradiol hemihydrate. Eur. J. Pharm. Biopharm., 49: 237–242 (2000). Nagarsenker, M. S., Meshram, R. N. and Ramprakash, G.: Solid dispersion of hydroxypropyl beta-cyclodextrin and ketorolac: enhancement of in-vitro dissolution rates, improvement in anti-in‰ammatory activity and reduction in ulcerogenicity in rats. J. Pharm. Pharmacol., 52: 949–956 (2000). Tanino, T., Ogiso, T. and Iwaki, M.: EŠect of sugarmodiˆed beta-cyclodextrins on dissolution and absorption characteristics of phenytoin. Biol. Pharm. Bull., 22: 298–304 (1999). Watanabe, E., Sudo, R., Takahashi, M. and Hayashi, M.: Evaluation of absorbability of poorly water-soluble drugs: Validity of the use of additives. Biol. Pharm. Bull., 23: 838–843 (2000). Yazdanian, M., Glynn, S. L., Wright, J. L. and Hawi, A.: Correlating partitioning and caco-2 cell permeability of structurally diverse small molecular weight compounds. Pharm. Res., 15: 1490–1494 (1998). Stewart, B. H., Chan, O. H., Lu, R. H., Reyner, E. L., Schmid, H. L., Hamilton, H. W., Steinbaugh, B. A. and Taylor, M. D.: Comparison of intestinal permeabilities determined in multiple in vitro and in situ models: relationship to absorption in humans. Pharm. Res., 12: 693–699 (1995). Pruitt, A. W., Winkel, J. S. and Dayton, P. G.: Variations in the fate of triameterene. Clin. Pharmacol. Ther., 21: 610–619 (1977). Grundy, J. S., Eliot, L. A. and Foster, R. T.: Extrahepatic ˆrst-pass metabolism of nifedipine in the rat. Biopharm. Drug. Dispos., 18: 509–522 (1997). Matsumoto, N., Masuhara, K. and Ogata, H.:
15)
16)
17)
18)
19)
20)
21)
22)
23)
24)
25)
26)
433
``Rinshou Yakubutsu Doutaigaku'' [Clinical Pharmacokinetics]. Maruzen CO., LTD., Tokyo, 2000, pp. 173–174. Bates, T. R., Fung, H. L., Lee, H. and Tembo, A. V.: Comparative bioavailability of anhydrous griseofulvin and its chloroform solvate in man. Res. Commun. Chem. Pathol. Pharmacol., 11: 233–243 (1975). Bates, T. R., Pieniaszek, H. J. Jr, Sequeira, J. A. and Rasmussen, J. E.: Gastrointestinal absorption of griseofulvin from corn oil-in-water emulsions: eŠect of amount of corn oil ingested in man. Arch. Dermatol., 113: 302–306 (1977). Tanaka, Y., Taki, Y., Sakane, T., Nadai, T., Sezaki, H. and Yamashita, S.: Characterization of drug transport through tight-junctional pathway in Caco-2 monolayer: comparison with isolated rat jejunum and colon. Pharm. Res., 12: 523–528 (1995). Sawada, T., Ogawa, T., Tomita, M., Hayashi, M. and Awazu, S.: Role of paracellular pathway in nonelectrolyte permeation across rat colon epithelium enhanced by sodium caprate and sodium caprylate. Pharm. Res., 8: 1365–1371 (1991). Lennernas, H., Nylander, S. and Ungell, A. L.: Jejunal permeability: a comparison between the ussing chamber technique and the single-pass perfusion in humans. Pharm. Res., 14: 667–671 (1997). Ungell, A. L., Nylander, S., Bergstrand, S., Sjoberg, A. and Lennernas, H.: Membrane transport of drugs in diŠerent regions of the intestinal tract of the rat. J. Pharm. Sci., 87: 360–366 (1998). Fagerholm, U., Johansson, M. and Lennernas, H.: Comparison between permeability coe‹cients in rat and human jejunum. Pharm. Res., 13: 1336–1342 (1996). Hendeles, L., Weinberger, M. and Bighley, L.: Absolute bioavailability of oral theophylline. Am. J. Hosp. Pharm., 34: 525–527 (1977). Balon, K., Riebesehl, B. U. and Muller, B. W.: Drug liposome partitioning as a tool for the prediction of human passive intestinal absorption. Pharm. Res., 16: 882–888 (1999). Sinko, P. J., Leesman, G. D. and Amidon, G. L.: Predicting fraction dose absorbed in humans using a macroscopic mass balance approach. Pharm. Res., 8: 979–988 (1991). Lucangioli, S. E., Kenndler, E., Carlucci, A., Tripodi, V. P., Scioscia, S. L. and Carducci, C. N.: Relation between retention factors of immunosuppressive drugs in microemulsion electrokinetic chromatography with biosurfactants and octanol-water partition coe‹cients. J. Pharm. Biomed. Anal., 33: 871–878 (2003). Bhattacharjee, A. K., Kyle, D. E., Vennerstrom, J. L. and Milhous, W. K.: A 3D QSAR pharmacophore model and quantum chemical structure–activity analysis of chloroquine (CQ)-resistance reversal. J. Chem. Inf. Comput. Sci., 42: 1212–1220 (2002). Tashiro, Y., Shichibe, S., Kato, Y., Hayakawa, E. and Itoh, K.: EŠect of lipophilicity on in vivo iontophoretic delivery. I. NSAIDs. Biol. Pharm. Bull., 24: 278–283. (2001). Yamashita, S., Furubayashi, T., Kataoka, M., Sakane,
434
27)
Etsushi WATANABE et al.
T., Sezaki, H. and Tokuda, H.: Optimized conditions for prediction of intestinal drug permeability using Caco-2 cells Eur. J. Pharm. Sci., 10: 195–204 (2000). Eichman, J. D. and Robinson, J. R.: Mechanistic studies on eŠervescent-induced permeability enhancement.
28)
Pharm. Res., 15: 925–930 (1998). Grundy, J. S., Eliot, L. A., Kulmatycki, K. M. and Foster, R. T.: Grapefruit juice and orange juice eŠects on the bioavailability of nifedipine in the rat. Biopharm. Drug. Dispos., 19: 175–183 (1998).
Regular Article Prediction of Absorbability of Poorly Water-soluble Drugs Based on Permeability Obtained through Modiˆed In Vitro Chamber Method Etsushi WATANABE1, Noriyuki KINOSHITA1, Masayuki TAKAHASHI2 and Masahiro HAYASHI1 1Faculty 2Pharmaceutical
of Pharmaceutical Sciences, Science University of Tokyo, Japan Formulation Research Laboratory, Daiichi Pharmaceutical Co., LTD., Tokyo, Japan
Full text of this paper is available at http://www.jstage.jst.go.jp/browse/dmpk
Summary: Permeability is an underlying parameter to control drug absorption. For highly water-soluble drugs, the high correlation between their permeability and fraction absorbed in humans is reported. In the present study, to predict the absorbability of poorly water-soluble drugs in humans, a new experimental method of the permeation study was proposed and subjected to examination. Firstly, using the in vitro chamber method modiˆed to contain 5z (ˆnal concentration) dimethyl sulufoxide (DMSO) in both compartments of the chamber (DMSO-MS), the eŠect of DMSO on membrane integrity was evaluated. Secondly, the correlation between the apparent permeability coe‹cients (Papp) obtained through DMSO-M or DMSO-MS and fractions absorbed in humans were investigated using 7 poorly water-soluble drugs. Membrane integrity of the rat intestinal tissues was maintained after using DMSO-MS, as with that after using the conventional in vitro chamber method. Papp of two paracellular markers obtained through DMSO-MS was not diŠerent from that obtained through the conventional chamber method. In the permeation study of the P-glycoprotein substrate, Papp from both mucosal to serosal and serosal to mucosal sides obtained through DMSO-MS was not signiˆcantly diŠerent from that obtained through the conventional chamber method. The correlation between Papp obtained through DMSO-MS and Fa which was expressed by the equation of Fa＝1－exp (－Papp×1.38×105) (r2＝0.980), was more favorable than the correlation between Papp obtained through DMSO-M and Fa which was expressed by the equation of Fa＝ 1－exp (－Papp×2.12×105) (r2＝0.875). These results showed that DMSO-MS was a useful method for predicting the absorbability of poorly water-soluble drugs.
Key words: poorly water-soluble drugs; permeability; dimethyl sulfoxide; intestinal tissue; fraction absorbed chamber experiment using isolated rat intestinal tissues and Fa in humans.3) Based on above results, the permeability of water-soluble drugs was an underlying parameter for estimating their absorbability in the development of a new chemical entity. In contrast, for poorly water-soluble drugs, prediction of their absorbability requires their dissolution rates as well as their permeability, because of their slow and incomplete dissolution in the gastrointestinal tract.1) However, many trials have been conducted to improve the solubility of poorly water-soluble drugs by using the techniques and materials such as solid dispersion, melt extrusion, cyclodextrin and sealing liquids into hard
Introduction According to a theory of the macroscopic mass balance model proposed by Amidon et al., the intestinal absorbability of a water-soluble drug which is rapidly and completely dissolved in the gastrointestinal tract can be determined by its permeability.1) Arturrson et al. report a good correlation between the apparent permeability coe‹cient (Papp) of water-soluble drugs obtained in the in vitro chamber experiment using Caco-2 cell monolayers, and fraction absorbed (Fa) in humans.2) Additionally, we had previously shown a good correlation between Papp obtained in the in vitro
Received; July 6, 2005, Accepted; September 16, 2005 To whom correspondence should be addressed : Etsushi WATANABE, Technology Administration Department, Daiichi Pharmaceutical Co., LTD., 1-8 Nihonbashi-Koamicho, Chuo-ku, Tokyo, 103-8541, JAPAN. Tel. ＋81-3-5640-0655; Fax. ＋81-3-5640-1744, E-mail: watanqsf＠ daiichipharm.co.jp
428
Prediction of Absorbability of Poorly Water-soluble Drugs
gelatin capsules.4–6) Using these techniques, the drug with high permeability is considered to have the high absorbability regardless of its intrinsic solubility. Therefore, it is necessary for the recent development of a new chemical entity to estimate its permeability regardless of its solubility. We had previously predicted Fa values of 5 poorly water-soluble drugs based on their Papp obtained through the modiˆed chamber method with an addition of 5z dimethyl sulufoxide (DMSO) in the mucosal side (DMSO-M).3,7) However, predicted Fa values of griseofulvin, nifedipine and indomethacin were not favorably comparable with their Fa values in humans.3) In the present study, a new experimental method in which the chamber method was modiˆed with an addition of 5z DMSO to both sides of chamber (DMSO-MS) was proposed, and subjected to examination. Firstly, the eŠects of the DMSO on the membrane integrity were evaluated by the measurement of released enzymes and membrane conductance. Secondly, the eŠects of DMSO on the membrane permeability were investigated using Papp of ‰uorescein isothiocyanatelabeled dextran 4000, sulfasalazine, antipyrine and rhodamine-123. Finally, the relationship between Papp values of 7 poorly water-soluble drugs obtained through DMSO-MS and their Fa values in humans8–14) was investigated. Methods
Materials: Fluorescein isothiocyanate-labeled dextran 4000 (FD4), sulfasalazine (SFZ), antipyrine (APY), rhodamine-123 (Rho123), indomethacin (IDM), triamterene (TAT), nifedipine (NFD), phenytoin (PHT), griseofulvin (GSF), azathioprine (AZT) and desipramine (DSP) were purchased from Sigma Chemical Co., Ltd. (St. Louis, MO). LDH CII (#271–54101) and ALP enzyme kit (#274–04401) were purchased from Wako Pure Chem. (Osaka, Japan). Other chemicals used were of the analytical grade. Animals: Male Wistar rats at ages of 7 weeks or older (body weight: about 200–250 g) were used. The rats were fasted for about 24 h with free access to water before the experiment. Necessary approvals for the experimental protocol of animals were obtained from the Ethical Committee of Science University of Tokyo (Tokyo, Japan). Experiment using modiˆed in vitro chamber method (DMSO-MS): Rats were anesthetized with urethane, and the abdominal wall was cut open along the midline. A 2 to 3 cm jejunum segment was excised and placed in warm Ringer's solution (pH of 7.4) in an atmosphere of 5z CO2 gas mixture. The epithelium was a 95z O2 W then exposed via a longitudinal incision along the mesenterium, and the underlying muscle layer was removed. Finally the epithelium was mounted onto an Us-
429
sing-type chamber having a tissue surface area of 0.64 cm2. Ringer's solution (5 mL) was added to both the donor (mucosal) and receiver (serosal) chambers. After 10-min preincubation, drug solution (5 mL) solubilized by Ringer's solution and Ringer's solution (5 mL), both of which included 5z (ˆnal concentration) DMSO, were added to the mucosal and serosal sides, respectively. The initial concentrations of drug solutions were 0.2 mM, except for Rho123 solution (0.026 mM). After the permeation study was started, samples (100 mL) were withdrawn from both sides at regular intervals (20 min.) up to 100 min. In permeation study, Papp from mucosal to serosal sides was basically evaluated. On the other hand, in the study of Rho123, Papp from serosal to mucosal sides as well as that from mucosal to serosal sides was evaluated. In control group, the conventional chamber experiment was performed to obtain the permeation of FD4, SDZ, APY and Rho123 in Ringer's solution without DMSO. For Rho123, we examined the transport from the serosal to mucosal sides as well as that from mucosal to serosal sides. Details of DMSO-M had been previously reported.3,7) In this study, experiments were conducted in 3 groups of control, DMSO-M and DMSO-MS groups using the control method, DMSO-M and DMSO-MS, respectively. Estimation of membrane integrity: Enzymes (LDH and ALP) released from the intestinal tissues were evaluated in the mucosal side buŠer 100 min after the start of permeation study. ALP activity was determined using the ALP enzyme kit by the Bessey-Lowry method. LDH activity was determined using the LDH CII kit (i.e., by spectrophotometeric determination (l＝560 nm) of diformazan formation). Membrane electrical parameters were evaluated by the membrane conductance (DGt) measured by the methods of Yamashita et al. and Sawada et al.15,16) Assay: The concentrations of FD4 and Rho123 were determined using a ‰uorescent microplate reader operating at excitation wavelength (EX) of 480 nm and emission wavelength (EM) of 520 nm, and EX of 486 nm and EM of 546 nm, respectively. The concentrations of other drugs were determined by HPLC (Alliance 2690, Waters Corporation, MA, USA) using a Waters 2487 Dual Absorbance Detector (Waters). Details of the HPLC conditions had been previously reported.3,7,12) Data Analysis: The apparent permeability coe‹cient per unit membrane surface area (Papp (cm W sec)) was calculated according to the following equation: Papp＝dM W dt×1 W AC0 where dM W dt is the steady-state appearance rate of drugs to the serosal side ( mmol W sec), C0 is the initial drug concentration on the mucosal side (mM), and A is the surface area of the membrane (cm2).7,17,18) Fa values
430
Etsushi WATANABE et al.
of 10 water-soluble drugs and 7 poorly water-soluble drugs were quoted from the literature.2,8–14,19–21) When the quoted Fa values had the certain range, median value was used. The relationship between in vitro permeability of water-soluble drugs in rats and in vivo absorption in humans was investigated using a complete radial mixing model (CRM).19,22) The numerical expression, which indicated relationship of Papp (cm W sec) of 7 poorlywater-soluble drugs as well as 10 water-soluble drugs with fraction absorbed in humans (Fa), was described using the following equation: Fa＝1－exp (－Papp×f ) where f was the lump constant including the factors of eŠective surface area of the intestine, transit time and scale-up factor in species diŠerence between rat and human. The relationship between Papp and Fa expressed as f was determined by non-linear regression analysis (Levenberg-Marquardt least squares algorithm; Delta Graph 4.0(r) by SPSS Inc., USA). Statistical Analysis: Results are expressed as means ±S.D. For multiple group analysis of variance, one-
Fig. 1. Release of alkaline phosphatase (ALP) and lactate dehydrogenase (LDH) from the intestinal tissue in control, DMSO-M and DMSO-MS groups. Data represent the means±S.D. of 3 experiments.
factor ANOVA was used. A p value of less than 0.05 was considered to be statistically signiˆcant. Results EŠects of DMSO addition on membrane integrity: Figure 1 shows the results of enzyme release measurement on the mucosal side 100 min. after the start of experiment in control, DMSO-M and DMSO-MS groups. There was no signiˆcant diŠerence in the release of ALP or LDH among the three groups. No signiˆcant changes in DGt were induced by DMSO-M or DMSOMS (Fig. 2). EŠects of DMSO addition on Papp of FD4, SFZ and APY: Figures 3 and 4 show Papp of FD4, SFZ and APY obtained in control, DMSO-M and DMSO-MS groups. No signiˆcant diŠerence in Papp of FD4, SFZ (Fig. 3) or APY (Fig. 4) was observed among the three groups. EŠects of DMSO addition on Papp of Rho123: Figure 5 shows Papp of Rho123 from mucosal to serosal sides and that from serosal to mucosal sides in control, DMSO-M and DMSO-MS groups. No signiˆcant diŠerence in Papp of Rho123 transported in either direction was observed among the three groups.
Fig. 2. Changes of membrane conductance (DGt) in control, DMSO-M and DMSO-MS groups. Data represent the means±S.D. of 3 experiments.
Fig. 3. Apparent permeability coe‹cients of paracellular permeant compounds of ‰uorescein isothiocyanate-labeled dextran 4000 (FD4) and sulfasalazine (SFZ), in control, DMSO-M and DMSO-MS groups. Data represent the means±S.D. of 3 experiments.
431
Prediction of Absorbability of Poorly Water-soluble Drugs
Papp and Fa of poorly water-soluble drugs: Papp values of 7 poorly water-soluble drugs were obtained through DMSO-M and DMSO-MS. Table 1 shows Papp values, predicted Fa values and actual Fa in humans. Papp
values of IDM and NFD obtained through DMSO-MS were signiˆcantly diŠerent from those obtained through DMSO-M. Estimation of relationship between Papp of 7 poorly water-soluble drugs and their Fa in humans: The correlation curve (solid line) between Papp of 7 poorly water-soluble drugs obtained through DMSO-M or DMSO-MS and their Fa in human is shown in Fig. 6. The correlation curve (dashed line) of 10 water-soluble drugs reported previously (Fa＝1－exp (－Papp×1.51× 105)) is overlaid with Fig. 6.3) The best ˆt for the data obtained through DMSO-M and DMSO-MS was observed when f was 2.12×105 (r2＝0.875) and 1.38× 105 (r2＝0.980), respectively. Discussion
Fig. 4. Apparent permeability coe‹cients of a transcellular permeant compound of antipyrine (APY) in control, DMSO-M and DMSO-MS groups. Data represent the means±S.D. of 3 experiments.
We had previously reported that Papp of 5 poorly water-soluble drugs could be obtained through a new experimental condition by adding 5z DMSO to the mucosal side of the in vitro chamber mounted on the
Fig. 5. Apparent permeability coe‹cients of a P-gp substrate of rhodamine 123 (Rho123) in control, DMSO-M and DMSO-MS groups. In each group, the transport of Rho123 both from mucosal to serosal (M to S) sides and from serosal to mucosal (S to M) sides was examined. Data represent the means±S.D. of 3 experiments.
Table 1. Apparent permeability coe‹cients (Papp) of poorly water-soluble drugs obtained through DMSO-M and DMSO-MS, and predicted fractions absorbed (Fa) in humans DMSO-M Poorly Water-soluble Drugs
Triamterene Desipramine Azathioprine Indomethacin Nifedipine Phenytoin Griseofulvin a
DMSO-MS
a
Papp, mean±S.D. [10－ 6 cm W sec]
Predicted Fab
Papp, mean±S.D. [10－ 6 cm W sec]
Predicted Fab
5.11±1.91 5.24±1.06 5.53±0.38 9.04±1.35 15.12±3.19 22.01±2.26 22.12±2.32
0.54 0.55 0.57 0.74 0.90 0.96 0.96
6.32±0.33 6.03±0.96 6.63±1.42 26.93±3.14** 43.17±5.07** 27.47±2.81 32.13±5.48
0.61 0.60 0.63 0.98 1.00 0.98 0.99
Solubilized by 5z (ˆnal concentration) dimethylsulfoxide Predicted based on the relationship between Papp and Fa of water-soluble drugs. c Quoted from references shown in parentheses. d Fa value of GSF was calculated based on the results reported by Bates et al. (13 and 14). **pº0.01 compared with DMSO-M b
Previously Reported Fa in Humansc
0.40–0.70(10) 0.50(12) 0.60(12) 1.00(8) 0.90–1.00(28) 1.00(9) 1.00(13,14)d
432
Etsushi WATANABE et al.
Fig. 6. The theoretical line calculated using the data of the fraction absorbed in humans versus the apparent permeability coe‹cients (Papp) of 7 poorly water-soluble drugs obtained through DMSO-M and DMSO-MS (solid line). The fraction absorbed was calculated based on the data in references of 8, 9, 10, 11, 13 and 14. The dashed line was the theoretical line calculated using the data of the 10 water-soluble drugs by non-linear regression as described in a previously report (3). Papp values represent the mean of the data from more than three experiments.
isolated rat intestinal tissue.3) Additionally, we predicted Fa values of these poorly water-soluble drugs based on the correlation between permeability of 10 water-soluble drugs and their fraction absorbed in humans. Compared with Fa in humans, predicted Fa may be overestimated for GSF and underestimated for NFD and IDM.3) In our previous report, the human Fa of GSF was quoted from clinical data obtained after oral administration of 500 mg tablets.3) This Fa value was not suitable for the measurement of absorbability of GSF, since GSF in this dosage form might be incompletely dissolved in the gastrointestinal tract.1) In the present study, Fa of GSF in humans was quoted from clinical data obtained after oral administration of GSF solution. Bates et al. report that the relative bioavailability of GSF dissolved in mono chloroform and encapsulated in gelatin is 150z, while the absolute bioavailability of pulverized GSF encapsulated in gelatin is 70z.13,14) Therefore, Fa value of GSF when GSF was completely dissolved in the gastro-intestinal tract was 1.0. As previously reported, log D of TAT, IDM, NFD, PHT and GSF was 1.6, 1.0, 3.2, 2.3 and 2.2, respectively.3) In addition, log P of AZT and DSP was reported to be 0.1 and 3.0, respectively.23,24) Though log D of NFD was very high,3,19) that of IDM was very low (1.01) because of almost complete ionization in neutral pH.3,25) On the other hand, Tashiro et al. reported that a high log P of nonionic type IDM was 4.4.25) Yamashita et al. suggested that the permeation of highly lipophilic drugs such as IDM and NFD from the cells to the serosal side through the intestinal tissue membrane appeared to be limited.26) Therefore, to estimate the permeability of poorly water-soluble drugs with high lipophilicity, it is necessary to improve their solubility and the lipophilicity in the serosal side medium. In the present study, we proposed a method in which 5z DMSO was added to the serosal side as well as the mucosal side.
Release of ALP and LDH was considered to represent the toxicity to the brush border and epithelial membranes, respectively.27) In DMSO-MS, the release of ALP and LDH was not aŠected (Fig. 1). DGt, an indicator of opening of a tight junction, was not signiˆcantly changed (Fig. 2). These results indicated that the membrane integrity was not aŠected by using DMSO-MS. To investigate the eŠect on permeability, we estimated Papp of paracellular permeant compounds such as FD4 (high molecule) and SFZ (low molecule) (Fig. 3), as well as that of APY (Fig. 4) which was a representative transcellular permeant compound. Papp of either compound obtained through DMSO-MS was not signiˆcantly diŠerent from that obtained through the control method or DMSO-M. Additionally, we estimated Papp of Rho123, a P-gp substrate (Fig. 5). Regardless of the transport directions, there was no signiˆcant diŠerence in Papp of Rho123 between DMSO-MS and control groups. Based on these results, we concluded that DMSO-MS could be applied to the transport study of poorly water-soluble drugs. During the experiment, either poorly water-soluble drug did not precipitate out in the transport medium. Therefore, the membrane permeability of poorly water-soluble drugs could be obtained in the same manner as that of water-soluble drugs. Papp of highly lipophilic drugs such as IDM and NFD in DMSO-MS group was higher than that in DMSO-M group (Table 1). This might be due to increased lipophilicity of receiver solution in the serosal side. The correlation between Papp obtained through DMSO-MS and Fa, which was expressed by the equation of Fa＝1－exp (－Papp×1.38×105) (r2＝0.980), was more favorable than that between Papp obtained through DMSO-M and Fa, which was expressed by the equation of Fa＝1－exp (－Papp×2.12×105) (r2＝0.875) (Table 1, Fig. 6). In conclusion, we established a new method for
Prediction of Absorbability of Poorly Water-soluble Drugs
predicting the drug permeability of poorly water-soluble drugs. This method was easy to be operated and likely to give reproducible data, and therefore could be widely applied to the studies of drug permeability. Our results suggested that DMSO-MS was useful method for predicting Fa value of poorly water-soluble drugs including those with high lipophilicity.
13)
14)
References 1)
2)
3)
4)
5)
6)
7)
8)
9)
10)
11)
12)
Amidon, G. L., Lennernas, H., Shah, V. P. and Crison, J. R.: A theoretical basis for a biopharmaceutic drug classiˆcation: the correlation of in vitro drug product dissolution and in vivo bioavailability. Pharm. Res., 12: 413–420 (1995). Artursson, P. and Karlsson, J.: Correlation between oral drug absorption in humans and apparent drug permeability coe‹cients in human intestinal epithelial (Caco-2) cells. Biochem. Biophys. Res. Commun., 175: 880–885 (1991). Watanabe, E., Takahashi, M. and Hayashi, M.: A possibility to predict the absorbability of poorly watersoluble drugs in humans based on rat intestinal permeability assessed by an in vitro chamber method. Eur. J. Pharm. Biopharm., 58: 659–665 (2004). Hulsmann, S., Backensfeld, T., Keitel, S. and Bodmeier, R.: Melt extrusion–an alternative method for enhancing the dissolution rate of 17beta-estradiol hemihydrate. Eur. J. Pharm. Biopharm., 49: 237–242 (2000). Nagarsenker, M. S., Meshram, R. N. and Ramprakash, G.: Solid dispersion of hydroxypropyl beta-cyclodextrin and ketorolac: enhancement of in-vitro dissolution rates, improvement in anti-in‰ammatory activity and reduction in ulcerogenicity in rats. J. Pharm. Pharmacol., 52: 949–956 (2000). Tanino, T., Ogiso, T. and Iwaki, M.: EŠect of sugarmodiˆed beta-cyclodextrins on dissolution and absorption characteristics of phenytoin. Biol. Pharm. Bull., 22: 298–304 (1999). Watanabe, E., Sudo, R., Takahashi, M. and Hayashi, M.: Evaluation of absorbability of poorly water-soluble drugs: Validity of the use of additives. Biol. Pharm. Bull., 23: 838–843 (2000). Yazdanian, M., Glynn, S. L., Wright, J. L. and Hawi, A.: Correlating partitioning and caco-2 cell permeability of structurally diverse small molecular weight compounds. Pharm. Res., 15: 1490–1494 (1998). Stewart, B. H., Chan, O. H., Lu, R. H., Reyner, E. L., Schmid, H. L., Hamilton, H. W., Steinbaugh, B. A. and Taylor, M. D.: Comparison of intestinal permeabilities determined in multiple in vitro and in situ models: relationship to absorption in humans. Pharm. Res., 12: 693–699 (1995). Pruitt, A. W., Winkel, J. S. and Dayton, P. G.: Variations in the fate of triameterene. Clin. Pharmacol. Ther., 21: 610–619 (1977). Grundy, J. S., Eliot, L. A. and Foster, R. T.: Extrahepatic ˆrst-pass metabolism of nifedipine in the rat. Biopharm. Drug. Dispos., 18: 509–522 (1997). Matsumoto, N., Masuhara, K. and Ogata, H.:
15)
16)
17)
18)
19)
20)
21)
22)
23)
24)
25)
26)
433
``Rinshou Yakubutsu Doutaigaku'' [Clinical Pharmacokinetics]. Maruzen CO., LTD., Tokyo, 2000, pp. 173–174. Bates, T. R., Fung, H. L., Lee, H. and Tembo, A. V.: Comparative bioavailability of anhydrous griseofulvin and its chloroform solvate in man. Res. Commun. Chem. Pathol. Pharmacol., 11: 233–243 (1975). Bates, T. R., Pieniaszek, H. J. Jr, Sequeira, J. A. and Rasmussen, J. E.: Gastrointestinal absorption of griseofulvin from corn oil-in-water emulsions: eŠect of amount of corn oil ingested in man. Arch. Dermatol., 113: 302–306 (1977). Tanaka, Y., Taki, Y., Sakane, T., Nadai, T., Sezaki, H. and Yamashita, S.: Characterization of drug transport through tight-junctional pathway in Caco-2 monolayer: comparison with isolated rat jejunum and colon. Pharm. Res., 12: 523–528 (1995). Sawada, T., Ogawa, T., Tomita, M., Hayashi, M. and Awazu, S.: Role of paracellular pathway in nonelectrolyte permeation across rat colon epithelium enhanced by sodium caprate and sodium caprylate. Pharm. Res., 8: 1365–1371 (1991). Lennernas, H., Nylander, S. and Ungell, A. L.: Jejunal permeability: a comparison between the ussing chamber technique and the single-pass perfusion in humans. Pharm. Res., 14: 667–671 (1997). Ungell, A. L., Nylander, S., Bergstrand, S., Sjoberg, A. and Lennernas, H.: Membrane transport of drugs in diŠerent regions of the intestinal tract of the rat. J. Pharm. Sci., 87: 360–366 (1998). Fagerholm, U., Johansson, M. and Lennernas, H.: Comparison between permeability coe‹cients in rat and human jejunum. Pharm. Res., 13: 1336–1342 (1996). Hendeles, L., Weinberger, M. and Bighley, L.: Absolute bioavailability of oral theophylline. Am. J. Hosp. Pharm., 34: 525–527 (1977). Balon, K., Riebesehl, B. U. and Muller, B. W.: Drug liposome partitioning as a tool for the prediction of human passive intestinal absorption. Pharm. Res., 16: 882–888 (1999). Sinko, P. J., Leesman, G. D. and Amidon, G. L.: Predicting fraction dose absorbed in humans using a macroscopic mass balance approach. Pharm. Res., 8: 979–988 (1991). Lucangioli, S. E., Kenndler, E., Carlucci, A., Tripodi, V. P., Scioscia, S. L. and Carducci, C. N.: Relation between retention factors of immunosuppressive drugs in microemulsion electrokinetic chromatography with biosurfactants and octanol-water partition coe‹cients. J. Pharm. Biomed. Anal., 33: 871–878 (2003). Bhattacharjee, A. K., Kyle, D. E., Vennerstrom, J. L. and Milhous, W. K.: A 3D QSAR pharmacophore model and quantum chemical structure–activity analysis of chloroquine (CQ)-resistance reversal. J. Chem. Inf. Comput. Sci., 42: 1212–1220 (2002). Tashiro, Y., Shichibe, S., Kato, Y., Hayakawa, E. and Itoh, K.: EŠect of lipophilicity on in vivo iontophoretic delivery. I. NSAIDs. Biol. Pharm. Bull., 24: 278–283. (2001). Yamashita, S., Furubayashi, T., Kataoka, M., Sakane,
434
27)
Etsushi WATANABE et al.
T., Sezaki, H. and Tokuda, H.: Optimized conditions for prediction of intestinal drug permeability using Caco-2 cells Eur. J. Pharm. Sci., 10: 195–204 (2000). Eichman, J. D. and Robinson, J. R.: Mechanistic studies on eŠervescent-induced permeability enhancement.
28)
Pharm. Res., 15: 925–930 (1998). Grundy, J. S., Eliot, L. A., Kulmatycki, K. M. and Foster, R. T.: Grapefruit juice and orange juice eŠects on the bioavailability of nifedipine in the rat. Biopharm. Drug. Dispos., 19: 175–183 (1998).