Prediction of In Vivo Intestinal Absorption Enhancement on P-Glycoprotein Inhibition, from Rat In Situ Permeability MANTHENA V.S. VARMA, RAMESH PANCHAGNULA Department of Pharmaceutics, National Institute of Pharmaceutical Education and Research (NIPER), Phase X, SAS. Nagar, Punjab 160062, India
Received 30 August 2004; revised 16 October 2004; accepted 8 December 2004 Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jps.20309
ABSTRACT: The purpose of this study is to determine the functional role of Pglycoprotein (P-gp) in intestinal absorption of drugs and to quantitatively predict the in vivo absorption enhancement on P-gp inhibition. In situ single-pass rat ileum permeability and aqueous solubility were measured for a set of 16 compounds. Permeability studies were also carried out in the presence of P-gp inhibitor to estimate the permeability enhancement on P-gp inhibition. A significant correlation was obtained between rat ileum permeability and the literature human intestinal absorption (HIA), Fa,human (r ¼ 0.891; p < 0.01). Compounds with permeability >0.2 104 cm/s are completely absorbed; however, few practically insoluble compounds were overestimated with this relationship. Inhibition of P-gp increased the permeability ( p < 0.05) of three moderately and three highly permeable compounds. Efflux inhibition ratio (EIR), the ratio of permeability due to P-gp-mediated efflux activity and passive permeability only, for these compounds was in the order of digoxin > paclitaxel > fexofenadine > quinidine > verapamil > cyclosporine. Integration of EIR with permeability versus Fa,human predicted that modulation of P-gp has no significant effect on the absorption of highly permeable compounds (quinidine, verapamil, and cyclosporine A), while for moderately permeable compounds (digoxin, paclitaxel, and fexofenadine), P-gp profoundly influences the intestinal permeability. The in situ permeability in rat ileum may be used to predict the in vivo P-gp function and its quantitative contribution to intestinal drug absorption. Integration of the functional activity of P-gp with the characteristics of BCS may explain drug interactions and explore the possible pharmacokinetic advantage on P-gp inhibition. ß 2005 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci 94:1694–1704, 2005
Keywords: P-glycoprotein; prediction; solubility; permeability; biopharmaceutic classification system
INTRODUCTION There is increasing evidence that P-glycoprotein (P-gp) have a much broader substrate specificity and a number of drugs, including HIV protease inhibitors like indinavir, saquinavir, ritonavir, and anticancer drugs like paclitaxel and vinblas-
Correspondence to: Ramesh Panchagnula (Telephone: 91172-2214682; Fax: 91-172-2214692; E-mail:
[email protected]) Journal of Pharmaceutical Sciences, Vol. 94, 1694–1704 (2005) ß 2005 Wiley-Liss, Inc. and the American Pharmacists Association
1694
tin have been reported to be substrates for P-gp using in vitro methodologies.1,2 The significant expression of P-gp in normal intestine raises the questions of its functional role and of whether it can influence the drug absorption processes. In vivo studies confirmed that P-gp significantly limits oral bioavailability of several drugs where intestinal permeability showed dose dependence with increased permeability as lumen concentration increases, or absorption enhancement on selective P-gp inhibition.3 However, on the other hand, literature also indicated no influence of P-gp on bioavailability of a number of P-gp
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
INTESTINAL ABSORPTION ENHANCEMENT ON P-GLYCOPROTEIN INHIBITION
substrates.4,5 For example, drugs like verapamil, quinidine, and digoxin have instant absorption (>80%), suggesting that for some P-gp substrates, intestinal permeability studies can overemphasize the importance of secretory transport.6 Further, intestinal transport of many P-gp substrates is practically not polarized despite drug efflux, which can be ascribed to a lesser control by P-gp caused by higher passive transmembrane movement rates and/or to per se lower activity of P-gp. It should be appreciated that both passive permeability and P-gp-mediated efflux processes operating in mutually opposite directions, contribute to overall drug permeability and thus influencing intestinal absorption. In vivo studies permit determination of absolute or relative bioavailability and provide actual picture on drug interactions and pharmacokinetic alterations. However, for obvious reasons high throughput and simple in situ animal models are more reliable when intestinal permeability is ratelimiting to absorption.7,8 Much of the information on drug interaction with P-gp has come from use of cultured epithelial models such as Caco-2 and MDRI-MDCKII.9,10 The relevance of these findings to P-gp effects on drug permeability in normal intestine is uncertain, given that these cells are derived from colonic tumor (Caco-2), overexpressing efflux transporters and exhibit permeability characteristics different from normal epithelium. In situ permeability in rats provide more meaningful estimates, as the rat model is a good predictor to human intestinal absorption.11,12 The specific objectives of the present study were to (1) develop a relationship between rat in situ ileum permeability to the fraction absorbed in humans (Fa,human), (2) determine the functional role of P-gp in limiting intestinal absorption of drugs, and (3) quantitatively predict the absorption enhancement on P-gp modulation by integrating the P-gp functional activity to the relationship between permeability and Fa,human.
MATERIALS AND METHODS Chemicals Paclitaxel and cyclosporine were a gift from Dabur India Ltd. (New Delhi, India). Fexofenadine HCl and hydrochlorthiazide were received from Aristo Pharmaceuticals Ltd. (Daman, India). Indinavir sulphate and lovastatin were generously provided by Matrix Lab. Ltd. (Hyderabad,
1695
India). Propranolol HCl was from Sun Pharmaceutical Industries Ltd. (Mumbai, India). Digoxin was gifted by Burroughs Wellcome India Ltd. (Mumbai, India). Frusemide and ranitidine were from Dr. Reddys’ Lab. (Hyderabad, India). LPhenylalanine was purchased from Sisco Lab. (Mumbai, India). Other compounds quinidine, verapamil, imipramine, sulindac, and D-glucose were purchased from Sigma Chemicals Co. (St. Louis, MO). Solvents used for quantification were of HPLC grade (JT Baker, Mexico), and all other chemicals and reagents were of analytical grade.
Animals and Legal Prerequisites Sprague-Dawley rats (270–350 g) were used to perform in situ single-pass perfusion. Anesthesia, surgical, and perfusion procedures were justified in detail, and were approved by the Institutional Animal Ethics Committee (IAEC, NIPER). The study complied with local and federal requirements for animal studies.
Solubility Studies Equilibrium solubility of drugs was determined by the shake-flask method (n ¼ 3). An excess amount of the drug was added in pH 7.4 10 mM phosphate buffer and equilibrated at 37 0.28C with vigorous shaking in a shaker water bath (Julabo, Germany) for 24 h. Samples were filtered using a 0.22-mm filter (Millipore, USA). Aliquots of each filtrate were diluted appropriately and analyzed by UV spectrophotometry (DU 640i Beckman, USA).
Permeability Studies: In situ Perfusion Experiments The surgical procedure and the in situ single-pass perfusion experiments were performed according to the methods described previously.12–14 Rats were fasted for 16–18 h, prior to study, with tap water ad libidum. After anesthesia via intraperitoneal administration of thiopental sodium (50 mg/kg), rats were placed on a heating pad to maintain body temperature at 378C. The intestine of the rats was exposed by a midline abdominal incision, and a 12–15 cm ileum segment (5 cm above the ileoceceal junction) was isolated and cannulated at both ends with glass tubing. The segment was rinsed with phosphate buffer saline JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
1696
VARMA AND PANCHAGNULA
(10 mL) and the perfusion solution maintained at 378C was pumped at a flow rate of 0.1 mL/min using a syringe pump (Harvard Apparatus PHD 2000 pump, MA, USA). The perfusion solution (pH 7.4) consisted of NaCl 48 mM, KCl 5.4 mM, Na2HPO4 2.8 mM, NaH2PO4 4 mM and D-glucose 1 g/L; and contained drugs with or without quinidine (200 mM) as a P-gp inhibitor. The concentrations of compounds in the perfusion solution were low enough to avoid precipitation in the lumen during the perfusion studies (lovastatin, 20 mM; digoxin, 20 mM; paclitaxel, 10 mM; fexofenadine, 100 mM; indinavir, 100 mM; ranitidine, 100 mM; frusemide, 50 mM; sulindac, 50 mM; hydrochlorthiazide, 100 mM; cyclosporine, 20 mM; verapamil, 200 mM; quinidine, 100 mM; imipramine, 100 mM; propranolol, 100 mM; D-glucose, 1 g/ L and L-phenylalanine, 100 mM). Insoluble compounds (lovastatin, digoxin, paclitaxel, and cyclosporine) were added as stock solutions of DMSO, giving a final solvent concentration of <0.2%. Higher concentrations were used for highly soluble high-dose compounds considering the in vivo availability of the compound in the solution at the absorption site. A two-step perfusion procedure was followed to determine the permeability of compounds with and without the P-gp inhibitor.13 This included sampling every 5 min for a 30-min perfusion period with perfusion solution containing test compound only, after 20 min equilibrium, and then switching to perfusion solution containing both the test compound and inhibitor, and similarly sampling every 5 min for a 30- min perfusion period after a 20-min equilibrium. Permeability was determined in four individual rats (n ¼ 4) for each compound. Out of four rats (n ¼ 4) used for each compound, two rats were first perfused with only test compound solution and then switched to perfusion solution containing both the test compound and P-gp inhibitor, while the other two rats were first perfused with solution containing both the test compound and P-gp inhibitor and then switched to perfusion solution containing only compound. Equilibrium of 20 min prior to sampling was found to be sufficient for both washout and to reach an initial steady state.13 Water flux was quantified by weight and volume measurements.13,15 To further keep a check on the intraand interindividual variability and presence of steady state, permeability of propranolol (passive, highly permeable), frusemide (passive, low permeable), and D-glucose (carrier-mediated, highly permeable) were monitored, by coperfusing along JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
with each individual compounds without and with P-gp inhibitor.
Bioanalysis by RP-HPLC All quantifications were performed by RP-HPLC with dual wavelength UV detector; except that D-glucose was estimated using a GOD-PODbased assay kit (Autozyme, ACCURex Biomedical Pvt. Ltd., Mumbai, India). HPLC was conducted on Waters equipment (Waters Corp., MA, USA) equipped with 600 controller pumps, Waters 2487 UV-Vis dual l absorbance detector, and configured to Millenium32 software. The perfusion sample was loaded onto the column by means of a 717plus autosampler (Waters Corp., MA, USA). Chromatography was performed on a 5-mm, 4.6 250-mm Symmetry1 C18 column (Waters Corp., MA, USA) attached with 5 mm, 3.9 20 mm Symmetry1 C18 guard column (Waters Corp., MA, USA). Chromatographic conditions for indinavir, lovastatin, and cyclosporine were previously published from our laboratory.16,17 Other methods used were optimized and validated for estimating individual compounds or for simultaneous analysis. For chromatography of digoxin, sulindac, imipramine, propranolol, and frusemide, the mobile phase consisted of 25% methanol, 25% acetonitrile, and 50% pH 3.0 (20 mM) acetate buffer was pumped on the isocratic mode at a flow rate of 0.5 mL/min with UV monitoring at 220 and 275 nm.18 For paclitaxel, quinidine, propranolol, and frusemide, the mobile phase composed of 70% methanol, 26% pH 5.0, 20 mM acetate buffer, and 4% isopropyl alcohol was pumped at a flow rate of 0.6 mL/min and chromatograms were recorded at 220 and 230 nm. In the case of L-phenylalanine, hydrochlorthiazide, ranitidine, and fexofenadine, the mobile phase consisted of 52% methanol, 41.7% pH 5.0 (20 mM) acetate buffer, and 6.3% isopropyl alcohol was pumped at a flow rate of 0.6 mL/min and chromatograms were recorded at 230 and 275 nm. Validation of the methods was evaluated by accuracy and precision in the working range of 50–110% of the drug concentration used in perfusion solution for individual compounds.
Data Treatment and Statistics Permeabilities (without and with P-gp inhibitor) are calculated after correcting the outlet concentration for water flux on the basis of the ratio of
INTESTINAL ABSORPTION ENHANCEMENT ON P-GLYCOPROTEIN INHIBITION
volume of perfusion solution collected and infused for each sampling point (5 min). Cin P eff;Control ðorÞ Peff ;Inh ¼ Q 1 2prl Cout ð1Þ where Q is the flow rate, Cin and Cout are the respective inlet and outlet concentration, r is radius of intestine (0.21 cm), and l is length of intestine measured after completion of perfusion.12,19 To evaluate the quantitative contribution of Pgp to limit the intestinal absorption, we have determined the EIR for all the drugs in the dataset (Figs. 2 and 3). The permeability obtained by in situ perfusion without and with P-gp inhibitor can be described by eqs. 2 and 3, respectively.20,21 Peff;Control ¼ PPD PPgp
ð2Þ
Peff;Inh ¼ PPD
ð3Þ
Considering one-site Michaelis-Menten saturation kinetics for P-gp substrates,1,9,22 EIR can be obtained by EIR ¼
Peff;Control PPgp ¼1 PPD Peff ;Inh
ð4Þ
EIR quantifies the passive drug transport attenuation by P-gp across the intestinal enterocytes. EIR provide information on the extent to which P-gp influences intestinal absorption of P-gp substrates across the enterocytes. EIR ¼ 0.75 (value lies between 0 and 1) indicates that P-gpmediated efflux transport (PP-gp) attenuates the passive permeability (PPD) by 75%. Values were indicated as mean SD for permeability in four independent rats. Statistical difference between the permeabilities of the drugs without and with P-gp inhibitor was evaluated by one-way ANOVA (SigmaStat version 2.03, SPSS Inc., IL, USA) at p < 0.05 and p < 0.01.
RESULTS Solubility and Permeability Determinations Permeability (Peff,Control) was determined in rat ileum using in situ single-pass perfusion technique. The Peff,Control values ranged from 0.009 0.006 104 cm/s to 1.30 0.24 104 cm/s with lovastatin showing least and l-phenylalanine, the maximum permeability (Table 1). Equilibrium solubility determined by shake-flask method or
1697
taken from the literature, and/or in situ permeability data was used to estimate maximum absorbable dose (MAD), an indicative of the maximum amount of drug that can be absorbed on oral administration, and dose number, an indicative of the maximum amount of drug that is soluble in 250 mL of water. For digoxin, paclitaxel, fexofenadine, indinavir, frusemide, and cyclosporine, MAD is less than the therapeutic dose, indicating that drug absorption for these compounds is limited by their solubility and/or permeability. Least-square nonlinear regression showed a sigmoidal relationship between Peff,Control and Fa,human (Fig. 1). The correlation was found to be highly significant (r ¼ 0.891; p < 0.01) when paclitaxel and cyclosporine were excluded. Paclitaxel and cyclosporine were found to be practically insoluble in water with a dose number >> 1 (Table 1). Thus, the reason for paclitaxel and cyclosporine having less Fa,human (0.04 and 0.30, respectively) than that expected from their Peff,Control can be ascribed to the solubility limitation. Overall, Peff,Control versus Fa,human relationship indicated that compounds with Peff,Control >0.2 104 cm/s are completely absorbed. In the present dataset seven compounds showed characteristics of high permeability while nine showed poor or moderate permeability. Functional Activity of P-gp and Human Intestinal Absorption Complete inhibition of P-gp by quinidine (200 mM) or cyclosporine (20 mM) significantly increased the permeability (Peff,Inh) of three moderately permeable compounds (digoxin, paclitaxel, and fexofenadine) and three highly permeable compounds (quinidine, verapamil, and cyclosporine) (Figs. 2 and 3). Other compounds that also showed increased permeability on P-gp inhibition included indinavir and sulindac. Further to check the intra- and interindividual variability, permeabilities of propranolol, frusemide, and D-glucose was monitored along with the individual compounds. These high (propranolol and D-glucose) and low (frusemide) permeability markers were shown to have no interaction with P-gp.25 The permeability data of individual compounds was considered only when the variability in permeability was within 10% of mean for propranolol and frusemide and 20% of mean for D-glucose. To confirm the complete efflux inhibition by quinidine (200 mM), we examined the inhibitor concentration-dependent effect on JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
1698
VARMA AND PANCHAGNULA
Table 1. Solubility, Permeability and Intestinal Absorption Parameters of Selected Compounds
No.
Compounds
Solubility (mg/mL)
Peff,Control (104 cm/s)
K ab (min1)
MADc (mg)
1 2 3 4 5 6 7 8 9 10 11 12
Lovastatin Digoxin Paclitaxel Fexofenadine Indinavir Sulphate Ranitidine HCl Frusemide Sulindac Hydrochlorthiazide Cyclosporin Verapamil HCl Quinidine 2H2OHCl Imipramine HCl Propranolol HCl D-glucose L-Phenylalanine
0.000a 0.010 (insoluble)a 0.001 0.000 0.807 0.025 0.426 0.076 100.0 (freely soluble)a 0.095 0.013 3.0a 0.01a 0.005 0.002 0.545 0.125 1.290 0.025
0.009 0.006 0.040 0.010 0.052 0.020 0.055 0.012 0.070 0.008 0.073 0.007 0.100 0.015 0.154 0.040 0.160 0.040 0.360 0.030 0.545 0.022 0.589 0.023
0.0037 0.0097 0.0121 0.0127 0.0156 0.0161 0.0214 0.0320 0.0331 0.0722 0.1083 0.1169
0.0 80 8000.00 4.4 0.25 0.10 0.7 284 854.17 459.6 180 0.89 298.7 400 3.76 72461.5 300 0.01 92.1 40 1.69 4317.0 200 0.27 14.9 25 10.00 16.2 600 480.00 2655.9 80 0.59 6784.9 200 0.62
169.84 25.64 179.93 31.36 909.0a 29.6a
0.712 0.054 0.890 0.039 0.894 0.153 1.300 0.240
0.1409 0.1756 0.1764 0.2257
13 14 15 16
1076780.0 1421628.0 7216088.7 340531.1
TDd (mg)
25 40 — —
Dose No.e
0.00 0.00 — —
BCSClassf IV III IV III IV III IV III IV II I I I I — —
a
Taken from Ref. 23 Absorption rate constant, Ka ¼ Peff,human A/V, where Peff,human ¼ Peff,control 3.253 þ 0.032 104 and intestinal area to volume ratio (A/V) was taken as 10cm1.24 c MAD ¼ CsxKa SIWV SITT, where Cs is the equilibrium solubility, SIWV is small intestinal water volume (250 mL), SITT is small intestinal transit time (180 min). d Therapeutic dose taken from WHO model list of essential drugs, IInd Ed. (Nov. 1999, http://www.who.int/medicines); and Maximum recommended therapeutic dose database, USFDA, CDER (http://www.fda.gov/cder). e Dose No. ¼ TD/Cs/250. f Compounds with Dose No. < 1 are considered as highly soluble and compounds with Peff,Control > 0.2 104 cm/s are considered as highly permeable. b
the transport of digoxin (Fig. 4). Quinidine at 50 mM concentration did not significantly enhance the permeability of digoxin ( p > 0.05), while at 100 and 200 mM, permeability of digoxin significantly increased ( p < 0.01). It is also obvious from the Figure 4 that no significant change in permeability was observed between 100 and 200 mM of quinidine, indicating 100 mM and above concentration completely inhibits the Pgp-mediated efflux transport.
Figure 1. Correlation between in situ rat ileum permeability and fraction absorbed in human after oral administration, for compounds listed in Table 1. Compounds are identified by their corresponding number from Table 1. The sigmoidal relationship (eq. 5) was obtained by least-square nonlinear regression (r ¼ 0.891, p < 0.01). JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
Figure 2. Passive permeability (Peff,Inh or PPD; Black bars), P-gp-mediated efflux (PP-gp; gray bars) and EIR (open triangles) of low and moderately permeable compounds. **p < 0.01, statistically significant difference with reference to Peff,Control of the same compound (values in Table 1).
INTESTINAL ABSORPTION ENHANCEMENT ON P-GLYCOPROTEIN INHIBITION
1699
Figure 3. Passive permeability (Peff,Inh or PPD; Black bars), P-gp-mediated efflux (PP-gp; gray bars) and EIR (open triangles) of highly permeable compounds. *p < 0.05 and **p < 0.01, statistically significant difference with reference to Peff,Control of the same compound (values in Table 1).
To evaluate the quantitative functional role of P-gp, the intestinal efflux inhibition ratio (EIR), the ratio of permeability due to P-gp-mediated efflux activity (PP-gp), and passive permeability only (PPD), was calculated. EIR of moderately permeable compounds (with EIR >0.2) from the present dataset was in the order of digoxin > paclitaxel > fexofenadine > sulindac > indinavir with a mean ratio of 0.526 (Fig. 2). High permeability compounds (quinidine > verapamil > cyclosporine) showed mean EIR 0.27 (Fig. 3). This indicate that drugs with moderate permeability (i.e., 0.01 104
Figure 5. Prediction of role of P-gp in intestinal drug absorption. Effect of complete P-gp inhibition on the fraction absorbed in human as a function of Peff,control, for P-gp substrates with different EIR. Profiles were generated using eqs. 4 and 5. Inset shows the relationship between HIA enhancement ratio and Peff,control of P-gp substrates with different EIR.
Human intestinal absorption (HIA) enhancement ratio, the ratio of Fa,human in the complete absence and presence of P-gp-mediated efflux transport, was found to be more than 1.5-fold for P-gp substrates with EIR ¼ 0.75 and Peff,Control between 0.002 and 0.1 104 cm/s. However, P-gp substrates with Peff,Control > 0.28 104 cm/s are not affected by P-gp induction or inhibition. P-gp substrates with minimum EIR (e.g., <0.2) are least affected by P-gp modulation. Digoxin, paclitaxel, and fexofenadine with high EIR and moderate Peff,Control show maximum enhancement in Fa,human on P-gp inhibition (Table 2). Although highly permeable P-gp substrates quinidine, verapamil, and cyclosporine demonstrated significant efflux transport, they showed no significant improvement in Fa,human on P-gp inhibition.
DISCUSSION
Figure 4. Effect of quinidine concentration on the ileum permeability of digoxin (20 mM). The data points and error bars represent the mean SD of four replicates.
The extent of interaction with P-gp, the passive permeability, and the overall effective permeability of drugs determines the quantitative functional role of P-gp in limiting intestinal absorption in vivo. In the present study, we have quantitatively estimated the enhancement in HIA when P-gp-mediated efflux transport is inhibited, by using in situ intestinal perfusion model in Sprague-Dawley rats. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
1700
VARMA AND PANCHAGNULA
Table 2. Predicted Fraction Absorbed in Humans from Rat Ileum Permeability in the Presence and Absence of P-gp-Mediated Efflux Fa,human
Compounds Lovastatin Digoxin Paclitaxel Fexofenadine Indinavir Sulphate Ranitidine HCl Frusemide Sulindac Hydrochlorthiazide Cyclosporine Verapamil HCl Quinidine 2H2OHCl Imipramine HCl Propranolol HCl D-glucose L-Phenylalanine
Literature (Reference)
Predicted from Peff,Control (with P-gp activity)
Predicted from Peff,Inh (without P-gp activity)
HIA Enhancement ratioa
0.1026 0.8127 0.0528 0.3029 0.6330 0.5031 0.6132 0.9033 0.9031 0.3034 1.0035 1.0036 1.0037 0.9532 1.00 1.00
0.19 0.05 0.43 0.06 0.50 0.12 0.52 0.07 0.59 0.04 0.60 0.03 0.7 0.05 0.84 0.09 0.85 0.05 0.98 0.01 1.00 0.00 1.00 0.00 1.00 0.00 1.00 0.00 1.00 0.00 1.00 0.00
0.21 0.05 0.92 0.03 0.86 0.07 0.79 0.08 0.70 0.10 0.65 0.03 0.69 0.06 0.92 0.08 0.85 0.08 0.99 0.01 1.00 0.00 1.00 0.00 1.00 0.00 1.00 0.00 1.00 0.00 1.00 0.00
1.09 2.15 1.71 1.54 1.17 1.07 0.98 1.09 0.99 1.01 1.00 1.00 1.00 1.00 1.00 1.00
a
HIA enhancement ratio is the ratio of Fa,human predicted from permeabilities in the absence and presence of P-gp activity.
The selected compounds (excluding cyclosporine and paclitaxel) demonstrated close agreement in correlation between in situ rat ileum permeability and Fa,human with a sigmoidal relationship (Fig. 1). Fa;human ¼ 1 e½2ðPeff;Control 3:253þ0:03210
4
Þtres =rf
ð5Þ where tres is the intestinal transit time considered to be 180 min, r is the radius of human intestine (1.75 cm), and f is correction factor considered as 2.8.12 This relationship very closely matches to the previously reported correlation between jejunum permeability in rats and Fa,human.12 Similar type of relationships were also demonstrated with Caco-2 cell monolayers and other permeability models.38,39 Using this relationship, drugs can be categorized as low permeable (poorly absorbed, Fa,human < 20%), moderately permeable (incomplete absorption; 20% < Fa,human < 80%) and highly permeable (completely absorbed; Fa,human > 80%). Drugs with Peff,Control < 0.01 104 cm/s can be considered as low permeable, and those with Peff,Control > 0.2 104 cm/s as highly permeable. Based on the dose number and the permeability JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
limits, all the compounds of the dataset were classified according to BCS.21 It is obvious from Table 1 that all the compounds that showed BCS class I characteristics were reported to be completely absorbed in humans. Except for digoxin, sulindac, and hydrochlorthiazide, which featured class III, other compounds falling into class II–IV showed incomplete absorption. However, the outliers to the Peff,Control versus Fa,human relationship include paclitaxel and cyclosporine, which showed considerable difference in Fa,human than that expected from their permeabilities. The less Fa,human for paclitaxel and cyclosporine can be attributed to the solubility limited absorption. The practically insoluble nature of these compounds resulted in a MAD therapeutic dose (Table 1). Normalization of dose to MAD indicated complete absorption for cyclosporine.24 These results are in concordance to the human jejunum permeability studies, which indicated cyclosporine as a low solubilityhigh permeability BCS class II drug, and its incomplete and highly variable absorption (30%) from Sandimmune oral formulation is the result of poor dissolution.19 Similarly, lower than expected Fa,human (0.05 vs 0.50) of paclitaxel may be partially explained by its solubility limitation. Pharmacokinetic analysis of oral paclitaxel (Taxol1, BMS) revealed that dose-escalation of oral pacli-
INTESTINAL ABSORPTION ENHANCEMENT ON P-GLYCOPROTEIN INHIBITION
taxel from 60–300 mg/m2 resulted in significant increases in the AUC of paclitaxel; however, these increases were moderate and not proportional with increases in dose due to the poor aqueous solubility of paclitaxel and consecutive limited dissolution in the GIT.28 Of the moderately permeable compounds, digoxin, paclitaxel, and fexofenadine showed a statistically significant increase in the permeability in the presence of the P-gp inhibitor, while highly permeable drugs quinidine, verapamil, and cyclosporine showed increased permeability in the presence of the P-gp inhibitor. EIR for the drugs, which showed statistical significant difference (p < 0.05) between Peff,Control and Peff,inh, are in the order of digoxin > paclitaxel > fexofenadine > quinidine > verapamil > cyclosporine. In the case of moderately permeable compounds, EIR was high compared to that of highly permeable compounds (Figs. 2 and 3). Admittedly, the dataset is small to make a general conclusion, but this trend could be due to (1) high P-gp activity towards these compounds and/or (2) slow flux/passive diffusion of these compounds through the membrane, leading to more accessibility for P-gp at subsaturation levels. The increased permeability in the presence of the P-gp inhibitor quinidine (200 mM) is expected due to the complete inhibition of P-gp-mediated efflux.40 Digoxin showed quinidine concentrationdependent permeability up to 100 mM, with no change in permeability in the presence of quinidine (100 mM) and quinidine (200 mM) (Fig. 4). This data indicates that quinidine (200 mM) completely
1701
reverses the transport attenuation, and therefore, the EIR calculated represents the complete P-gpmediated efflux. Integration of EIR with a Peff,Control versus Fa,human relationship provided the basis of predicting the HIA enhancement of drugs on P-gp inhibition (Fig. 5). Peff,Inh at various EIR was calculated with eq. 4 and the obtained Peff,Inh was used to predict the Fa,human on complete P-gp inhibition using eq. 5. This model has the advantage of combining the quantitative functional activity of P-gp with the in vivo relevance of permeability to intestinal absorption. Complete inhibition of P-gp has a profound effect on the Fa,human of low and moderately permeable drugs, quantitatively dictated by the P-gp activity towards the compound (Table 2). Based on these principles, we propose a classification system that differentiates P-gp substrates with intestinal absorption relevance in vivo from those having no significance (Table 3). Although, the classification is based on a rather limited set of compounds, it provides valuable insight on the differential absorption behavior and the actual in vivo relevance of P-gp in limiting intestinal absorption of compounds. Completely absorbed drugs (except cyclosporine, which belong to BCS class II) showed the features of BCS class I (see Table 1), where drug absorption is not limited by the solubility and/ or permeability. Therefore, even compounds like quinidine and verapamil, whose permeability is significantly attenuated by P-gp-mediated efflux activity, are not influenced by P-gp induction or
Table 3. Classification Scheme of the Drugs Based on the Relevance of P-gp-Mediated Efflux Transport In Vivo Class
Subclass
Examples from the dataset
Moderate Permeability High Respondersa
Low Respondersb
Digoxin Paclitaxel Fexofenadine
Indinavir Sulindac Ranitidine Lovastatin
Nonrespondersc Frusemide Hydrochlorthiazide
High Permeability Cyclosporine Verapamil Quinidine Imipramine Propranolol D-glucose L-phenylalanine
a
Highly influenced by the level of expression and induction/inhibition of P-gp. Criteria ) EIR > 0.5 and HIA enhancement ratio > 1.5. b Reasonably influenced by the level of expression and induction/inhibition of P-gp. Criteria ) 0.1 < EIR < 0.5 and 1.1 < HIA enhancement ratio < 1.5. c Not influenced by P-gp. Criteria ) EIR < 0.1 and HIA enhancement ratio < 1.1. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
1702
VARMA AND PANCHAGNULA
inhibition in vivo.6 However, these compounds have the strong affinity to P-gp and inhibit P-gpmediated efflux, and may show clinically significant increase in intestinal absorption of various P-gp substrates.41,42 P-gp substrates belonging to class II (cyclosporine) are fairly permeable, and may be well absorbed into systemic circulation. However, low dissolution rate limits the concentration of the drug at the site of absorption, thereby leading to less passive diffusion and further preventing saturation of efflux transporter.21 Thus, saturation of P-gp by providing high drug concentration at the site of absorption or inhibiting P-gp transport may significantly improve absorption of drugs like cyclosporine. Compounds with low and moderate permeability show a profound effect on Fa,human with small changes in permeability, and are further classified as high-, low-, and nonresponders with reference to P-gp activity. Intestinal absorption of high responders may be markedly affected by interindividual difference in the expression level of P-gp and/or simultaneous administration of P-gp substrates/inhibitors, whereas low responders may only be minimally affected by an alteration in P-gp function.43 High responders digoxin and fexofenadine belong to BCS class III, while paclitaxel belongs to BCS class IV. These drugs have less intrinsic permeability, which to a significant extent is attributable to the P-gp-mediated efflux. Concomitant oral administration of fexofenadine and ketoconazole showed an increase in Cmax and area under plasma concentration–time curve, which is consistent with inhibition of P-gpmediated efflux, because fexofenadine is not subjected to any metabolism.44 Similarly, a significant increase in the bioavailability of digoxin and paclitaxel when administered with P-gp inhibitors were reported.45,46 In vivo absorption of low responders indinavir, sulindac, and ranitidine have minimal efflux activity due to substrate recognition, while incomplete absorption of nonresponders is attributable to their intrinsic solubility and permeability limitations and P-gp has no significance. Overall, P-gp activity plays a significant role in limiting intestinal permeability of BCS class II–IV P-gp substrates based on their EIR, but not for BCS class I compounds irrespective of P-gp activity.47 The jejunum has a relatively shorter transit time (40 min) favoring absorption of compounds with high membrane permeability. Such compounds may partition into the membrane at such a fast rate that P-gp-mediated efflux is overruled by passive influx, unaffecting net abJOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
sorption. However, because of the solubility and/or permeability limitations the absorption site for drugs belonging to BCS class II–IV is shifted more towards the ileum, which has a transit time of 140 min.48 The high P-gp expression levels in the lower intestine make the moderately absorbed Pgp substrates more susceptible to P-gp-mediated efflux.49,50 The rat is the best Fa,human predictor animal model for passive permeability, and further, there exist a similar level of P-gp expression51 and overlapped substrate specificity with quantitatively the same affinity for a large number of P-gp substrates in rat mdr1a and human MDR1.52 Thus, permeability studies in rat ileum provides a more meaningful forecast on in vivo absorption of P-gp substrates.
CONCLUSIONS Equilibrium solubility and in situ permeability in rats were determined for a set of 16 compounds. The in situ permeability showed a significant correlation to Fa,human. Compounds can be classified according to BCS based on the limits of 0.2 104 cm/s for high permeability and the dose number <1 for high solubility. Incompletely absorbed compounds from the dataset were found to be solubility and/or permeability limiting. Inhibition of P-gp has moved compounds of BCS class III and IV (digoxin, fexofenadine, and paclitaxel) to class I and II, respectively, by significantly increasing the permeability. Due to solubility and/or permeability limitations, a major part of the administered dose of class II–IV compounds is absorbed from the lower intestine, where the P-gp expression is reportedly more. Thus, permeability data in the ileum will be useful in developing a predictive models for in vivo relevance of P-gp-mediated efflux. EIR quantifies the functional activity of P-gp-mediated efflux transport and integration of EIR with the permeability versus Fa,human relationship can predict the P-gp effect on the intestinal absorption in vivo. In general, these predictions indicated that no pharmacokinetic advantage can be obtained on P-gp inhibition for BCS class I compounds, irrespective of their P-gp activity. However, these compounds by virtue of interaction with P-gp, are involved in drug interactions with other P-gp substrates, as exemplified with quinidine and verapamil. P-gp substrates showing the features of BCS class II–IV may be profoundly influenced by P-gp modulation, based on their P-gp activity.
INTESTINAL ABSORPTION ENHANCEMENT ON P-GLYCOPROTEIN INHIBITION
REFERENCES 1. Varma MVS, Ashokraj Y, Dey CS, Panchagnula R. 2003. P-glycoprotein inhibitors and their screening: A perspective from bioavailability enhancement. Pharmacol Res 48:347–359. 2. Huisman MT, Smit JW, Schinkel AH. 2000. Significance of P-glycoprotein for the pharmacology and clinical use of HIV protease inhibitors. Aids 14: 237–242. 3. Williams WC, Sinko PJ. 1999. Oral absorption of the HIV protease inhibitors: A current update. Adv Drug Deliv Rev 39:211–238. 4. Chiou WL, Chung SM, Wu TC. 2000. Apparent lack of effect of P-glycoprotein on the gastrointestinal absorption of a substrate, tacrolimus, in normal mice. Pharm Res 17:205–208. 5. Chiou WL, Chung SM, Wu TC, Ma C. 2001. A comprehensive account on the role of efflux transporters in the gastrointestinal absorption of 13 commonly used substrate drugs in humans. Int J Clin Pharmacol Ther 39:93–101. 6. Sandstrom R, Karlsson A, Knutson L, Lennernas H. 1998. Jejunal absorption and metabolism of R/Sverapamil in humans. Pharm Res 15:856–862. 7. Stewart BH, Helen Chan O, Jeayk N, Fleisher D. 1997. Discrimination between drug candidates using models for evaluation of intestinal absorption. Adv Drug Deliv Rev 23:27–45. 8. Zhang Y, Bachmeier C, Miller DW. 2003. In vitro and in vivo models for assessing drug efflux transporter activity. Adv Drug Deliv Rev 55:31–51. 9. Gao J, Murase O, Schowen RL, Aube J, Borchardt RT. 2001. A functional assay for quantitation of the apparent affinities of ligands of P-glycoprotein in Caco-2 cells. Pharm Res 18:171–176. 10. Mahar Doan KM, Humphreys JE, Webster LO, Wring SA, Shampine LJ, Serabjit-Singh CJ, Adkison KK, Polli JW. 2002. Passive permeability and P-glycoprotein-mediated efflux differentiate central nervous system (CNS) and non-CNS marketed drugs. J Pharmacol Exp Ther 303:1029–1037. 11. Chiou WL, Barve A. 1998. Linear correlation of the fraction of oral dose absorbed of 64 drugs between humans and rats. Pharm Res 15:1792–1795. 12. Fagerholm U, Johansson M, Lennernas H. 1996. Comparison between permeability coefficients in rat and human jejunum. Pharm Res 13:1336–1342. 13. Hanafy A, Langguth P, Spahn-Langguth H. 2001. Pretreatment with potent P-glycoprotein ligands may increase intestinal secretion in rats. Eur J Pharm Sci 12:405–415. 14. Fagerholm U, Lindahl A, Lennernas H. 1997. Regional intestinal permeability in rats of compounds with different physicochemical properties and transport mechanisms. J Pharm Pharmacol 49:687–690. 15. Issa C, Gupta P, Bansal AK. 2003. Implications of density correction in gravimetric method for water
16.
17.
18.
19.
20.
21.
22.
23. 24.
25.
26.
27.
1703
flux determination using rat single-pass intestinal perfusion technique: A technical note. AAPS PharmSciTech 4:133–138. Panchagnula R, Bansal T, Varma MVS, Kaul CL. 2004. Reversed-phase liquid chromatography with ultraviolet detection for simultaneous quantitation of indinavir and propranolol from ex-vivo rat intestinal permeability studies. J Chromatogr B Anal Technol Biomed Life Sci 806:277–282. Sharma P, Chawla H, Panchagnula R. 2002. Analytical method for monitoring concentrations of cyclosporin and lovastatin in vitro in an everted rat intestinal sac absorption model. J Chromatogr B Anal Technol Biomed Life Sci 768:349–359. Varma MVS, Kapoor N, Sarkar M, Panchagnula R. 2004. Simultaneous determination of digoxin and permeability markers in rat in situ intestinal perfusion samples by RP-HPLC. J Chromatogr B 813:347–352. Chiu YY, Higaki K, Neudeck BL, Barnett JL, Welage LS, Amidon GL. 2003. Human jejunal permeability of cyclosporin A: Influence of surfactants on P-glycoprotein efflux in Caco-2 cells. Pharm Res 20:749–756. Troutman MD, Thakker DR. 2003. Novel experimental parameters to quantify the modulation of absorptive and secretory transport of compounds by P-glycoprotein in cell culture models of intestinal epithelium. Pharm Res 20:1210–1224. Varma MVS, Khandavilli S, Ashokraj Y, Jain A, Dhanikula A, Sood A, Thomas NS, Pillai OP, Sharma P, Gandhi R, Agrawal S, Nair V, Panchagnula R. 2004. Biopharmaceutic classification system: A scientific framework for pharmacokinetic optimization in drug research. Curr Drug Metab 5:375–388. Jang SH, Wientjes MG, Au JL. 2001. Kinetics of Pglycoprotein-mediated efflux of paclitaxel. J Pharmacol Exp Ther 298:1236–1242. The Merck Index. 2001. 13 ed. Rathway, NJ: Merck Research laboratories. Hilgers AR, Smith DP, Biermacher JJ, Day JS, Jensen JL, Sims SM, Adams WJ, Friis JM, Palandra J, Hosley JD, Shobe EM, Burton PS. 2003. Predicting oral absorption of drugs: A case study with a novel class of antimicrobial agents. Pharm Res 20:1149–1155. Polli JW, Wring SA, Humphreys JE, Huang L, Morgan JB, Webster LO, Serabjit-Singh CS. 2001. Rational use of in vitro P-glycoprotein assays in drug discovery. J Pharmacol Exp Ther 299:620–628. Lennernas H, Fager G. 1997. Pharmacodynamics and pharmacokinetics of the HMG-CoA reductase inhibitors. Similarities and differences. Clin Pharmacokinet 32:403–425. Hinderling PH, Hartmann D. 1998. Pharmacokintics of digoxin and main metabolite derivatives in healthy humans. Ther Drug Monit 13:381–401.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
1704
VARMA AND PANCHAGNULA
28. Malingre MM, Meerum Terwogt JM, Beijnen JH, Rosing H, Koopman FJ, van Tellingen O, ten Bokkel Huinink WW, Swart M, Duchin K, Schellens JHM. 2000. Phase I and pharmacokinetic study of oral paclitaxel. J Clin Oncol 18:2468–2475. 29. Simons FE, Bergman JN, Watson WT, Simons KJ. 1996. The clinical pharmacology of fexofenadine in children. J Allergy Clin Immunol 98:1062–1064. 30. Lin JH. 1997. Human immunodeficiency virus protease inhibitors, from drug design to clinical studies. Adv Drug Deliv Rev 27:215–237. 31. Yazdanian M, Glynn SL, Wright JL, Hawi A. 1998. Correlating partitioning and caco-2 cell permeability of structurally diverse small molecular weight compounds. Pharm Res 15:1490–1494. 32. Winiwarter S, Bonham NM, Ax F, Hallberg A, Lennernas H, 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 41:4939–4949. 33. Davies NM, Watson MS. 1997. Clinical pharmacokinetics of sulindac. A dynamic old drug. Clin Pharmacokinet 32:437–459. 34. Lindholm A, Henricsson S, Lind M, Dahlqvist R. 1988. Intraindividual variability in the relative systemic availability of cyclosporin after oral dosing. Eur J Clin Pharmacol 34:461–464. 35. McTavish D, Sorkin EM. 1989. Verapamil. An updated review of its pharmacodynamic and pharmacokinetic properties, and therapeutic use in hypertension. Drugs 38:19–76. 36. Physicians’ Desk Reference. 2003. 57 ed. Montvale, NJ: Thomson, PDR. 37. Sallee FR, Pollock BG. 1990. Clinical pharmacokinetics of imipramine and desipramine. Clin Pharmacokinet 18:346–364. 38. Artursson P, Karlsson J. 1991. Correlation between oral drug absorption in humans and apparent drug permability coefficients in human epithelial (Caco2) cells. Biochem Biophys Res Commun 175:880– 885. 39. Irvine JD, Takahashi L, Lockhark K, Cheong J, Tolan JW, Selick HE, Grove JR. 1999. MDCK (Madin-Darby canine kidney) cells: A tool for membrane permeability screening. J Pharm Sci 88: 28–33. 40. Stephens RH, Tanianis-Hughes J, Higgs NB, Humphrey M, Warhurst G. 2002. Region-dependent modulation of intestinal permeability by drug efflux transporters: In vitro studies in mdr1a(/) mouse intestine. J Pharmacol Exp Ther 303:1095– 1101. 41. Kharasch ED, Hoffer C, Altuntas TG, Whittington D. 2004. Quinidine as a probe for the role of pglycoprotein in the intestinal absorption and clinical effects of fentanyl. J Clin Pharmacol 44: 224–233.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 8, AUGUST 2005
42. Kharasch ED, Hoffer C, Whittington D. 2004. The effect of quinidine, used as a probe for the involvement of P-glycoprotein, on the intestinal absorption and pharmacodynamics of methadone. Br J Clin Pharmacol 57:600–610. 43. Hoffmeyer S, Burk O, von Richter O, Arnold HP, Brockmoller J, Johne A, Cascorbi I, Gerloff T, Roots I, Eichelbaum M, Brinkmann U. 2000. Functional polymorphisms of the human multidrug-resistance gene: Multiple sequence variations and correlation of one allele with P-glycoprotein expression and activity in vivo. Proc Natl Acad Sci USA 97:3473– 3478. 44. Cvetkovic M, Leake B, Fromm MF, Wilkinson GR, Kim RB. 1999. OATP and P-glycoprotein transporters mediate the cellular uptake and excretion of fexofenadine. Drug Metab Dispos 27:866–871. 45. Westphal K, Weinbrenner A, Giessmann T, Stuhr M, Franke G, Zschiesche M, Oertel R, Terhaag B, Kroemer HK, Siegmund W. 2000. Oral bioavailability of digoxin is enhanced by talinolol: Evidence for involvement of intestinal P-glycoprotein. Clin Pharmacol Ther 68:6–12. 46. Meerum Terwogt JM, Beijnen JH, ten Bokkel Huinink WW, Rosing H, Schellens JHM. 1998. Co-administration of cyclosporin enables oral therapy with paclitaxel. Lancet 352:285. 47. Varma MVS, Sateesh K, Panchagnula R. 2004. Functional role of P-glycoprotein in limiting intestinal absorption of drugs: Contribution of passive permeability to P-glycoprotein-mediated efflux transport. Mol Pharm 2:12–21. 48. Kaus LC, Gillespie WR, Hussain AS, Amidon GL. 1999. The effect of in vivo dissolution, gastric emptying rate, and intestinal transit time on the peak concentration and area-under-the-curve of drugs with different gastrointestinal permeabilities. Pharm Res 16:272–280. 49. Mouly S, Paine MF. 2003. P-glycoprotein increases from proximal to distal regions of human small intestine. Pharm Res 20:1595–1599. 50. Siegmund W, Ludwig K, Engel G, Zschiesche M, Franke G, Hoffmann A, Terhaag B, Weitschies W. 2003. Variability of intestinal expression of Pglycoprotein in healthy volunteers as described by absorption of talinolol from four bioequivalent tablets. J Pharm Sci 92:604–610. 51. Stephens RH, O’Neill CA, Warhurst A, Carlson GL, Rowland M, Warhurst G. 2001. Kinetic profiling of P-glycoprotein-mediated drug efflux in rat and human intestinal epithelia. J Pharmacol Exp Ther 296:584–591. 52. Yamazaki M, Neway WE, Ohe T, Chen I-W, Rowe JF, Hochman JH, Chiba M, Lin JH. 2001. In vitro substrate identification studies for P-glycoproteinmediated transport: Species difference and permeability of in vivo results. J Pharmacol Exp Ther 296:723–735.
