- Email: [email protected]
Kinetic assessment of luminal degradation of orally effective prodrugs for rational drug development
Kinetic Assessment of Luminal Degradation of Orally Effective Prodrugs for Rational Drug Development TAKASHI MIZUMA Department of Drug Absorption and Pharmacokinetics, School of Pharmacy, Tokyo University of Pharmacy and Life Sciences, 1432-1 Horinouchi, Hachioji, Tokyo 192-0392, Japan
Received 9 March 2009; revised 24 May 2009; accepted 29 May 2009 Published online 21 July 2009 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jps.21867
ABSTRACT: Although prodrugging (prodrug derivatization) is a powerful technique for improving the pharmacokinetic characteristics of drugs, the intestinal pharmacokinetics of prodrugs has yet to be elucidated fully. A previous article reported the kinetic requirement of prodrugs to overcome membrane barriers. In the present article, the luminal degradation of prodrugs was kinetically assessed to understand crucial factors in the intestinal absorption of prodrugs and to show a rational development procedure. A kinetic model equation involving luminal degradation clearance (CLdeg) was derived, and CLdeg was estimated according to the equation with in vitro and in vivo reported data of two kinds of ampicillin prodrugs (lenampicillin and pivampicillin) and one acyclovir prodrug (valacyclovir). For lenampicillin ((2,2-dimethyl-1-oxopropoxy)methyl ester derivative), CLdeg was approximately 1.7 times as large as absorption clearance (CLabs), whereas for pivampicillin ((5-methyl-2-oxo-1,3-dioxol-4-yl)methyl ester derivative), CLdeg was approximately one tenth of CLabs. For valacyclovir (acyclovir prodrug), CLdeg was negligible. These results indicate that not only membrane permeability but also luminal stability should be assessed for the rational development of orally effective prodrugs, and that luminal stabilization can improve the intestinal absorption of prodrugs. A procedure was proposed to develop orally effective prodrugs considered for luminal degradation as well as membrane permeability. ß 2009 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci 99:1078–1086, 2010
Keywords:
prodrugs; absorption; kinetics; drug design; oral drug delivery
INTRODUCTION Orally active prodrugs have been developed to overcome their poor intestinal absorption and are on the market1; nevertheless, no kinetic strategy for rationally designing prodrugs has been established because the intestinal pharmacokinetics of orally active prodrugs has hardly been clarified. Recently, the author reported a kinetic strategy to develop orally effective prodrugs overcoming membrane permeability.2 Moreover, even an
Correspondence to: Takashi Mizuma (Telephone: þ81-42676-3181; Fax: þ81-426-76-3142; E-mail: [email protected]) Journal of Pharmaceutical Sciences, Vol. 99, 1078–1086 (2010) ß 2009 Wiley-Liss, Inc. and the American Pharmacists Association
1078
orally active prodrug, lenampicillin, was enzymatically and/or nonenzymatically degraded to a large extent in the luminal tract, and ampicillin formed from lenampicillin in the luminal tract mostly contributed to the fraction of orally administered lenampicillin absorbed.3 This indicates that nonenzymatic and/or enzymatic degradation of prodrugs in the luminal tract should be assessed, but in the previous report, luminal degradation was assessed from the aspect of the extent but not the rate; therefore, in the present article, the objectives were to assess the luminal degradation of lenampicillin, pivampicillin, and acyclovir to help understand the factors involved in intestinal absorption, and to propose a procedure for the rational development of prodrugs that takes into consideration luminal
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 2, FEBUARY 2010
LUMINAL DEGRADATION OF ORALLY EFFECTIVE PRODRUG
1079
degradation of the prodrug, as well as membrane permeability.
THEORY AND METHODS Fraction Absorbed (Fa) of Prodrug and Drug A kinetic model of orally administered prodrug is shown in Figure 1. Based on this model, simultaneous differential equations are as follows: dXp;l ¼ ðkdeg þ ka;p Þ Xp;l dt
(1)
dXp;b ¼ ka;p Xp;l dt
(2)
dXd;l ¼ kdeg Xp;l ka;d Xd;l dt
(3)
dXd;b ¼ ka;d Xd;l dt
(4)
Xp,l and Xp,b are the amounts of prodrug in the luminal tract and blood (body), respectively. ka,p and kdeg are the rate constants for the absorption of prodrug and the luminal degradation of prodrug to drug, respectively. Xd,l and Xd,b are the amounts of drug in the luminal tract and the blood (body), respectively. ka,d is the rate constant for drug absorption.
Figure 1. Kinetic model for intestinal pharmacokinetics of orally administered prodrug. Xp,l and Xp,b are the amounts of prodrug in the luminal tract and blood (body), respectively. ka,p and kdeg are the rate constants for the absorption of prodrug and the luminal degradation of the prodrug to drug, respectively. Xd,l and Xd,b are the amounts of drug in the luminal tract and the blood (body), respectively. ka,d is the rate constant for drug absorption.
The amount of drug absorbed (Xa,d), which formed from prodrug in the luminal tract, across the intestine membrane to the body during time T is
Xa;d ¼ Dose kdeg ka;d ðkdeg þ ka;p Þ expðka;d TÞ ka;d expððkdeg þ ka;p Þ TÞ 1 þ ka;d ðkdeg þ ka;p Þ ka;d þ ðkdeg þ ka;p Þðka;d kdeg ka;p Þ The amount of prodrug absorbed (Xa,p) across the intestinal membrane to the body during time T is Xa;p ¼
Dose ka;p ð1 expððkdeg þ ka;p Þ TÞÞ kdeg þ ka;p (5)
Accordingly, the sum of the amount of prodrug and drug absorbed during time T (Xa) following oral administration of the prodrug is Xa ¼ Xa;p þ Xa;d
DOI 10.1002/jps
(7)
Therefore, the Fa is Fa ¼
¼
(6)
Xa Dose
ka;p ð1 expððkdeg þ ka;p Þ TÞÞ kdeg þ kdeg þ ka;p kdeg þ ka;p kdeg ðkdeg þ ka;p Þ expðka;d TÞ ka;d expððkdeg þ ka;p Þ TÞ þ ðkdeg þ ka;p Þðka;d kdeg ka;p Þ
(8)
(9)
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 2, FEBUARY 2010
1080
MIZUMA
Under the following conditions: ka;p ¼ a Ppro
(10)
ka;d ¼ a Pdru
(11)
A¼aT
(12)
Then, from Eqs. (16) and (17), general expressions (Eqs. 18 and 19, respectively) are obtained for stable compounds (drug or prodrug) in the luminal tract. P and CL are the permeation coefficient ( P) and absorption clearance (CLabs). A or B is calculated from Pdru or CLabs,d
Fa is expressed by
Fa ¼
Fa ¼ 1 expðA PÞ
Ppro ð1 expðA ðPdeg þ Ppro ÞÞ Pdeg þ Pdeg þ Ppro Pdeg þ Ppro Pdeg ðPdeg þ Ppro Þ expðA Pdru Þ Pdru expðA ðPdeg þ Ppro ÞÞ þ ðPdeg þ Ppro ÞðPdru Pdeg Ppro Þ
where Ppro and Pdru are the membrane permeabilities of the prodrug and drug, respectively, which can be obtained using Caco-2 cell monolayers,4 an artificial membrane,5–7 and so on. a is a scaling constant, including several factors such as the membrane surface area and Pdeg ¼ kdeg/a. Since membrane permeability ( P) area of the membrane (S) ¼ clearance (CL), Fa is also expressed by
Fa ¼
(18)
(13)
or Fa ¼ 1 expðB CLabs Þ
(19)
Eq. (18) corresponds to an equation in drug studies (not prodrugs) reported by several researchers.8–10 Here they have assumed that drugs are stable in the luminal tract.
CLabs;p ð1 expðB ðCLdeg þ CLabs;p ÞÞ CLdeg þ CLdeg þ CLabs;p CLdeg þ CLabs;p CLdeg ðCLdeg þ CLabs;p Þ expðB CLabs;d Þ CLabs;d expðB ðCLdeg þ CLabs;p ÞÞ þ ðCLdeg þ CLabs;p ÞðCLabs;d CLdeg CLabs;p Þ
B¼
A S
(15)
where CLa,p, CLa,d, and CLdeg are the absorption clearance of the prodrug, absorption clearance of the drug, and degradation clearance of the prodrug to drug in the luminal tract, respectively. CLa,p and CLa,d can be obtained using Caco-2 cell monolayers,4 an artificial membrane,5–7 and so on. When a prodrug is stable in the luminal tract, Pdeg and CLdeg are null. Accordingly, Fa of the prodrug is expressed by Fa ¼ 1 expðA Ppro Þ
(16)
Fa ¼ 1 expðB CLabs;p Þ
(17)
or
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 2, FEBUARY 2010
(14)
Once A or B is calculated according to Eq. (18) or (19) from Fa and P or CLabs of a stable compound, for example, a drug in the present study, Fa of the prodrug can be predicted under the assumption that the prodrug is stable in the luminal tract. When the predicted Fa of the prodrug is lower than the observed (reported) Fa, this suggests that the prodrug may be unstable in the luminal tract, and that Pdeg and CLdeg can be estimated by Eqs. (13) and (14), respectively.
Data Collection and Estimation of Luminal Degradation Clearance (CLdeg) CLdeg is estimated using Eq. (14) from B, Fa, CLabs,p, and CLabs,d. B is obtained according to DOI 10.1002/jps
LUMINAL DEGRADATION OF ORALLY EFFECTIVE PRODRUG
Eq. (19) from Fa and CLabs of the drug described above. Fa, CLabs,d, and CLabs,p were retrieved from the literature. Fa were from the literature on lenampicillin,11 ampicillcin,12 pivampicillin,13 valacyclovir,14 and acyclovir.15 CLabs,p of lenampicillin and CLabs,d of ampicillin were from the previous report.3 When studies (pivampicillin and its drug study16 and valacyclovir and its drug study17) with Caco-2 cell monolayers reported the permeation coefficient ( P), CLabs of the prodrug (CLabs,p) and drug (CLabs,d) were calculated from the reported P of the prodrug and drug, respectively, using the following equation: CLabs ðnL=min=cm2 Þ ¼ 60 106 P ðcm=sÞ (20)
Simulation of Impact of Luminal Degradation Clearance on Absorbed Fraction Impact of luminal degradation of the prodrug on the fraction absorbed of prodrug was simulated based on Eq. (14) with the following values of ampicillin3 as an example: CLabs;d ðCLabs of drugÞ ¼ 3:8 nL=min=cm2 B ¼ 0:06203
intra-tissue metabolism data (Class E) CLp;mi fp;muc CLd;mi fd;muc OE CLd;is CLp;is CLp;im CLd;im
CLp;mi fp;muc > CLd;mi fd;muc OE CLd;is CLp;is < CLp;im CLd;im
CLd;is CLp;is CLp;im CLd;im
ðCLp;is CLd;im þ CLp;is CLd;is þ CLmet;int CLd;is Þ CLd;mi fd;muc OE ðCLp;im CLd;is þ CLp;is CLd;is þ CLmet;int CLd;is Þ CLp;mi fp;muc Oral Prodrug Classification System (OPCS) Required minimum Fa ( Fa,min), which is required Fa for orally effective prodrug, is expressed by the following equation: Fa;min ¼ 1 expðB CLabs;d OEÞ
(21)
CLabs;p;min ¼ CLabs;d OE CLabs,p,min is the absorption clearance required for an orally effective prodrug. Orally effective prodrugs are classified based on two factors, the uptake ratio and efflux ratio, of the prodrug and drug (Fig. 2), which is based on the previous report.2 When the prodrug satisfies both Eqs. (22) and (23), the prodrug is effective, irrespective of DOI 10.1002/jps
(22) (23)
where CLp,m-i and CLd,m-i are uptake clearances of the prodrug and drug, respectively. fp,muc and fd,muc are unbound fractions of the prodrug and drug, respectively, on the mucosal side. CLp,i-s and CLp,i-m are efflux clearances of the prodrug from inside the intestinal tissue to the serosal side and from the inside of intestinal tissue to the mucosal side, respectively. CLd,i-s and CLd,i-m are efflux clearances of the drug from inside the intestinal tissue to the serosal side and from the inside of intestinal tissue to the mucosal side, respectively. When the prodrug satisfies Eqs. (24) and (25) or Eqs. (26) and (27), the prodrug is effective only when satisfying the conditional Eq. (28) (Class M)
CLp;mi fp;muc < CLd;mi fd;muc OE
Accordingly, Fa is expressed by a function of CLdeg and CLabs,p.
1081
(24) (25) (26) (27)
(28)
where CLmet,int is the intrinsic clearance for intratissue metabolism. When the prodrug satisfies Eqs. (29) and (30) or Eqs. (31) and (32), the prodrug is ineffective, irrespective of intra-tissue metabolism data (Class I) CLp;mi fp;muc ¼ CLd;mi fd;muc OE CLd;is CLp;is < CLp;im CLd;im CLp;mi fp;muc < CLd;mi fd;muc OE CLd;is CLp;is CLp;im CLd;im
(29) (30) (31) (32)
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 2, FEBUARY 2010
1082
MIZUMA
Figure 2. Oral prodrug classification system (OPCS) showing kinetic classification and criteria for orally effective prodrugs. CLp,m-i and CLd,m-i are uptake clearances of the prodrug and drug, respectively. fp,muc and fd,muc are unbound fractions of the prodrug and drug, respectively, on the mucosal side. CLp,i-s and CLp,i-m are efflux clearances of the prodrug from the inside of the intestinal tissue to the serosal side and from the inside of the intestinal tissue to the mucosal side, respectively. CLd,i-s and CLd,i-m are efflux clearances of the drug from the inside of intestinal tissue to the serosal side and from the inside of intestinal tissue to the mucosal side, respectively. CLmet,int is intrinsic clearance of intra-tissue metabolism.
RESULTS AND DISCUSSION Reported Fa Versus Predicted Fa of Prodrugs on the Assumption of Stability in the Luminal Tract The relationships between Fa and CLabs are shown in Figure 3. The predicted Fa ( Fa,pred) of prodrugs was calculated according to Eq. (19) from B and CLabs on the assumption that prodrugs are stable in the luminal tract. Table 1 summarizes the calculated values (B, Fa,pred, and CLdeg) and the reported data (CLabs,d, CLabs,p, and Fa) that were used for the calculation. The reported Fa of JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 2, FEBUARY 2010
lenampicillin was approximately 50%, whereas the predicted Fa ( Fa,pred) of lenampicillin was almost 100% (Fig. 3a). This clearly indicates that the reported Fa of lenampicillin was lower than the predicted Fa. Also, the reported Fa of pivampicillin is slightly lower than the predicted Fa (Fig. 3b). In contrast, the reported Fa of valacyclovir is comparable to the predicted Fa (Fig. 3c). These results suggest that lenampicillin and pivampicillin are unstable in the luminal tract, whereas valacyclovir may be stable in the luminal tract. The absorption clearance or permeation coefficient of lenampicillin and its drug, pivampicillin DOI 10.1002/jps
LUMINAL DEGRADATION OF ORALLY EFFECTIVE PRODRUG
1083
Figure 4. Impact of luminal degradation clearance (CLdeg) on fraction absorbed ( Fa). Condition: Equation (14), B ¼ 0.06203, CLabs,d ¼ 3.8 nL/min/cm2.
and its drug, and valacyclovir and its drug was from the reports of Mizuma,3 He et al.,16 and Katragadda et al.,17 respectively, giving each B value due to a difference in experiment conditions conducted by research groups. CLdeg of Lenampicillin, Pivampicillin, and Valacyclovir CLdeg of lenampicillin, pivampicillin, and valacyclovir are summarized in Table 1. CLdeg of lenampicillin was 1107 nL/min/cm2, approximately 1.7 times as large as CLabs,p. This indicates that luminal degradation impacts on Fa, and the stabilization of lenampicillin in the luminal tract can improve Fa, whereas CLdeg of pivampicillin was 162 nL/min/cm2, and was approximately onetenth of CLabs,p. This indicates that pivampicillin degrades but its CLdeg is much lower than its CLabs,p. The results of these ampicillin prodrugs, lenampicillin and pivampicillin, suggest that (5-methyl-2-oxo-1,3-dioxol-4-yl)methyl ester derivative (pivampicillin) may be more stable than (2,2-dimethyl-1-oxopropoxy)methyl ester derivative (lenampicillin) in the luminal tract when compared with the respective membrane permeability.
Figure 3. Relationship between fraction absorbed ( Fa) and absorption clearance (CLabs) of drugs (closed circles) and their prodrugs (open circles). Curves represent predicted Fa calculated by the equation Fa ¼ 1exp(B CLabs), with CLabs of drug (CLabs,d). (a) lenampicillin (ampicillin prodrug), (b) pivampicillin (ampicillin prodrug), (c) valacyclovir (acyclovir prodrug).
Table 1. Kinetic Characteristics of Orally Effective Prodrugs B
Drug
Prodrug (Drug) Lenampicillin (ampicillin) Pivampicillin (ampicillin) Valacyclovir (acyclovir) DOI 10.1002/jps
CLabs,d (nL/min/cm2)
Fa
3.80 120 90.0
0.210 0.210 0.225
Prodrug
CLabs,p 1/(nL/min/cm2) (nL/min/cm2) Fa,pred 0.06203 0.001964 0.002832
660.0 1932 265.8
Fa
1.00 0.504 0.978 0.920 0.529 0.542
CLdeg CLdeg/ (nL/min/cm2) CLabs,p 1107 162.0 —
1.68 0.084 —
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 2, FEBUARY 2010
1084
MIZUMA
Scheme 1. Procedures for rational development of orally effective prodrugs.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 2, FEBUARY 2010
DOI 10.1002/jps
LUMINAL DEGRADATION OF ORALLY EFFECTIVE PRODRUG
In contrast, the predicted Fa of valacyclovir was comparable to the reported Fa (Tab. 1), indicating that the luminal degradation of valacyclovir is negligible. At the same time, because PEPT1mediated absorption of valacyclovir has been reported,18 it is another possibility that valacyclovir in the luminal tract was degraded but Caco-2 cells with lower PEPT1 activity gave lower CLabs,p. Thus, lower Fa due to luminal degradation and underestimation of in vitro CLabs,p resulted in similar values of Fa and the predicted Fa. The kinetic analysis in the present study was triggered by the difference between the predicted Fa of prodrug on the assumption of stability in the luminal tract and the observed Fa. This difference may have resulted from other causes, but the predicted Fa was lower but not higher than the observed Fa for all prodrugs studied in this report. Therefore, luminal degradation of the prodrug is considered to be the most appropriate cause for the difference at present. B value is a scaling factor used for absorption clearance, and is similar to A value, which is a scaling factor used for permeation coefficient. Several researchers have used an equation equivalent to Eq. (18) involving A value, which has several factors. Thus, this analysis was based on the assumption that A and B values are constant, as several researchers have done for drugs.
Impact of Luminal Degradation on Fraction Absorbed The impact of luminal degradation on the fraction absorbed is shown in Figure 4. Ampicillin, the CLabs (CLabs,d) of which is 3.8 nL/min/cm2, is exemplified. When 50% of minimum Fa is required and CLdeg is null, only 11.9 nL/min/cm2 of CLabs,p is needed, whereas when 50% of minimum Fa is required and CLdeg is 1000 nL/min/cm2, CLabs,p must be higher than 577.5 nL/min/cm2. In other words, when 50% of minimum Fa is required and CLdeg is 16.0 nL/min/cm2, only 16.0 nL/min/ cm2 of CLabs,p equal to CLdeg is needed. This indicates that CLdeg impacts on Fa of the prodrug, and that even if membrane permeability is improved by prodrugging, luminal degradation must be studied to rationally develop orally effective prodrugs. A future study is to experimentally estimate the luminal degradation kinetics. DOI 10.1002/jps
1085
Procedure for Rational Development of Orally Effective Prodrugs Because luminal degradation of the prodrug or prodrug candidate should be considered as described above, a procedure for developing orally effective prodrugs, which is considered for luminal degradation of the prodrug, is proposed (Scheme 1) based on the previous report.3 The first decision condition (decision condition 1) is based on the predicted Fa, which is calculated by Eq. (1) with CLabs of a prodrug candidate determined by the in vitro experiment. When the predicted Fa is higher than or equal to the minimum required Fa ( Fa Fa,min), the prodrug candidate is promising. The second decision condition (decision condition 2) is based on OPCS (Fig. 2), which is based on the previously reported decision tree for the development procedure of an orally effective prodrug (kinetic classification and criteria for effective membrane-permeable prodrugs).2 When a prodrug candidate is classified into Class E indicating ‘‘Effective,’’ the prodrug candidate is promising. When a prodrug candidate is classified into Class I (Ineffective), another prodrug candidate should be designed with consideration of the results of OPCS. When a prodrug candidate satisfies Eq. (1) (decision condition 1) and is classified into Class E (decision condition 2), the prodrug candidate may be forwarded to ‘‘Go’’ in the in vivo study. Fa is calculated and CLdeg is assessed according to Eq. (3) with data from the in vivo study. When Fa is higher than or equal to a required minimum Fa, the prodrug candidate is judged successful. When CLdeg is larger than CLabs,p, for example, CLdeg/CLabs,p is 2 or higher, the luminal stability of the prodrug candidate should be improved. Since the present study focused on prodrugs with an ester bond, the next target is an orally active thiamine prodrug, fursultiamine,19 which has a disulfide bond. In summary, luminal degradation of the oral prodrugs available on the market was assessed using in vitro and in vivo data in humans. The luminal degradation clearance of lenampicillin was higher than the absorption clearance, indicating that not only membrane permeability but also luminal stability must be studied for the development of orally effective prodrugs. This article is the first report assessing the luminal degradation kinetics of oral prodrugs in humans. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 2, FEBUARY 2010
1086
MIZUMA
ACKNOWLEDGMENTS This study was supported in part by a grant from The Research Foundation for Pharmaceutical Sciences.
REFERENCES 1. Ettmayer P, Amidon GL, Clement B, Testa B. 2004. Lessons learned from marketed and investigational prodrugs. J Med Chem 47:2393–2404. 2. Mizuma T. 2008. Pharmacokinetic strategy for designing orally effective prodrugs overcoming biological membrane barriers: Proposals of kinetic classification and criteria for membrane-permeable prodrug-likeness. Chem-Bio Inform J 8:25–32. www. cbi.or.jp/cbi/CBIj/vol8/8_25-E.pdf (accessed June 9, 2008). 3. Mizuma T. 2008. Integrated pharmacokinetic assessment and strategy for orally effective prodrugs overcoming luminal degradation and biological membrane barriers. Chem-Bio Inform J 8:58–68. www.cbi.or.jp/cbi/CBIj/vol8/8_58-E.pdf (accessed November 18, 2008). 4. Artursson P, Karlsson J. 1991. Correlation between oral drug absorption in humans and apparent drug permeability coefficients in human intestinal epithelial (Caco-2) cells. Biochem Biophys Res Commun 175:880–885. 5. Kansy M, Senner F, Gubernator K. 1998. Physicochemical high throughput screening: Parallel artificial membrane permeation assay in the description of passive absorption processes. J Med Chem 41:1007–1010. 6. Kerns EH. 2001. High throughput physicochemical profiling for drug discovery. J Pharm Sci 90:1838– 1858. 7. Sugano K, Nabuchi Y, Machida M, Aso Y. 2003. Prediction of human intestinal permeability using artificial membrane permeability. Int J Pharm 257: 245–251. 8. Amidon GL, Sinko PJ, Fleisher D. 1988. Estimating human oral fraction dose absorbed: A correlation using rat intestinal membrane permeability for passive and carrier-mediated compounds. Pharm Res 5:651–654.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 2, FEBUARY 2010
9. Yamashita S, Furubayashi T, Kataoka M, Sakane T, Sezaki H, Tokuda H. 2000. Optimized conditions for prediction of intestinal drug permeability using Caco-2 cells. Eur J Pharm Sci 10:195–204. 10. Zhu C, Jiang L, Chen TM, Hwang KK. 2001. A comparative study of artificial membrane permeability assay for high throughput profiling of drug absorption potential. Eur J Med Chem 37:399– 407. 11. Schering-Plough. 2008. Package insert of varacillin tablets. Tokyo, Japan: Schering-Plough. 12. Meiji Seika. 2005. Package insert of viccillin capsules. Tokyo, Japan: Meiji Seika. 13. Ehrnebo M, Nilsson SO, Bore´ us LO. 1979. Pharmacokinetics of ampicillin and its prodrugs bacampicillin and pivampicillin in man. J Pharmacokinet Biopharm 7:429–451. 14. Soul-Lawton J, Seaber E, On N, Wootton R, Rolan P, Posner J. 1995. Absolute bioavailability and metabolic disposition of valacyclovir, the L-valyl ester of acyclovir, following oral administration to humans. Antimicrob Agents Chemother 39:2759– 2764. 15. Benet LZ, Sheiner LB. 1985. Design and optimization of dosage regimens: Pharmacokinetic data. In: Gilman AG, Goodman LS, Rall TW, Murad F, editors. The pharmacological basis of therapeutics. 7th edition. New York: Macmillan Publishing Co. pp. 1663–1733. 16. He X, Sugawara M, Takekuma Y, Miyazaki K. 2004. Absorption of ester prodrugs in Caco-2 and rat intestine models. Antimicrob Agents Chemother 48:2604–2609. 17. Katragadda S, Jain R, Kwatra D, Hariharan S, Mitra AK. 2008. Pharmacokinetics of amino acid ester prodrugs of acyclovir after oral administration: Interaction with the transporters on Caco-2 cells. Int J Pharm 362:93–101. 18. Han H, de Vrueh RL, Rhie JK, Covitz KM, Smith PL, Lee CP, Oh DM, Sade´ e W, Amidon GL. 1998. 5’-Amino acid esters of antiviral nucleosides, acyclovir, and AZT are absorbed by the intestinal PEPT1 peptide transporter. Pharm Res 15:1154– 1159. 19. Kitamori N, Itokawa Y. 1993. Pharmacokinetics of thiamin after oral administration of thiamin tetrahydrofurfuryl disulfide to humans. J Nutr Sci Vitaminol 39:465–472.
DOI 10.1002/jps
Received 9 March 2009; revised 24 May 2009; accepted 29 May 2009 Published online 21 July 2009 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jps.21867
ABSTRACT: Although prodrugging (prodrug derivatization) is a powerful technique for improving the pharmacokinetic characteristics of drugs, the intestinal pharmacokinetics of prodrugs has yet to be elucidated fully. A previous article reported the kinetic requirement of prodrugs to overcome membrane barriers. In the present article, the luminal degradation of prodrugs was kinetically assessed to understand crucial factors in the intestinal absorption of prodrugs and to show a rational development procedure. A kinetic model equation involving luminal degradation clearance (CLdeg) was derived, and CLdeg was estimated according to the equation with in vitro and in vivo reported data of two kinds of ampicillin prodrugs (lenampicillin and pivampicillin) and one acyclovir prodrug (valacyclovir). For lenampicillin ((2,2-dimethyl-1-oxopropoxy)methyl ester derivative), CLdeg was approximately 1.7 times as large as absorption clearance (CLabs), whereas for pivampicillin ((5-methyl-2-oxo-1,3-dioxol-4-yl)methyl ester derivative), CLdeg was approximately one tenth of CLabs. For valacyclovir (acyclovir prodrug), CLdeg was negligible. These results indicate that not only membrane permeability but also luminal stability should be assessed for the rational development of orally effective prodrugs, and that luminal stabilization can improve the intestinal absorption of prodrugs. A procedure was proposed to develop orally effective prodrugs considered for luminal degradation as well as membrane permeability. ß 2009 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci 99:1078–1086, 2010
Keywords:
prodrugs; absorption; kinetics; drug design; oral drug delivery
INTRODUCTION Orally active prodrugs have been developed to overcome their poor intestinal absorption and are on the market1; nevertheless, no kinetic strategy for rationally designing prodrugs has been established because the intestinal pharmacokinetics of orally active prodrugs has hardly been clarified. Recently, the author reported a kinetic strategy to develop orally effective prodrugs overcoming membrane permeability.2 Moreover, even an
Correspondence to: Takashi Mizuma (Telephone: þ81-42676-3181; Fax: þ81-426-76-3142; E-mail: [email protected]) Journal of Pharmaceutical Sciences, Vol. 99, 1078–1086 (2010) ß 2009 Wiley-Liss, Inc. and the American Pharmacists Association
1078
orally active prodrug, lenampicillin, was enzymatically and/or nonenzymatically degraded to a large extent in the luminal tract, and ampicillin formed from lenampicillin in the luminal tract mostly contributed to the fraction of orally administered lenampicillin absorbed.3 This indicates that nonenzymatic and/or enzymatic degradation of prodrugs in the luminal tract should be assessed, but in the previous report, luminal degradation was assessed from the aspect of the extent but not the rate; therefore, in the present article, the objectives were to assess the luminal degradation of lenampicillin, pivampicillin, and acyclovir to help understand the factors involved in intestinal absorption, and to propose a procedure for the rational development of prodrugs that takes into consideration luminal
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 2, FEBUARY 2010
LUMINAL DEGRADATION OF ORALLY EFFECTIVE PRODRUG
1079
degradation of the prodrug, as well as membrane permeability.
THEORY AND METHODS Fraction Absorbed (Fa) of Prodrug and Drug A kinetic model of orally administered prodrug is shown in Figure 1. Based on this model, simultaneous differential equations are as follows: dXp;l ¼ ðkdeg þ ka;p Þ Xp;l dt
(1)
dXp;b ¼ ka;p Xp;l dt
(2)
dXd;l ¼ kdeg Xp;l ka;d Xd;l dt
(3)
dXd;b ¼ ka;d Xd;l dt
(4)
Xp,l and Xp,b are the amounts of prodrug in the luminal tract and blood (body), respectively. ka,p and kdeg are the rate constants for the absorption of prodrug and the luminal degradation of prodrug to drug, respectively. Xd,l and Xd,b are the amounts of drug in the luminal tract and the blood (body), respectively. ka,d is the rate constant for drug absorption.
Figure 1. Kinetic model for intestinal pharmacokinetics of orally administered prodrug. Xp,l and Xp,b are the amounts of prodrug in the luminal tract and blood (body), respectively. ka,p and kdeg are the rate constants for the absorption of prodrug and the luminal degradation of the prodrug to drug, respectively. Xd,l and Xd,b are the amounts of drug in the luminal tract and the blood (body), respectively. ka,d is the rate constant for drug absorption.
The amount of drug absorbed (Xa,d), which formed from prodrug in the luminal tract, across the intestine membrane to the body during time T is
Xa;d ¼ Dose kdeg ka;d ðkdeg þ ka;p Þ expðka;d TÞ ka;d expððkdeg þ ka;p Þ TÞ 1 þ ka;d ðkdeg þ ka;p Þ ka;d þ ðkdeg þ ka;p Þðka;d kdeg ka;p Þ The amount of prodrug absorbed (Xa,p) across the intestinal membrane to the body during time T is Xa;p ¼
Dose ka;p ð1 expððkdeg þ ka;p Þ TÞÞ kdeg þ ka;p (5)
Accordingly, the sum of the amount of prodrug and drug absorbed during time T (Xa) following oral administration of the prodrug is Xa ¼ Xa;p þ Xa;d
DOI 10.1002/jps
(7)
Therefore, the Fa is Fa ¼
¼
(6)
Xa Dose
ka;p ð1 expððkdeg þ ka;p Þ TÞÞ kdeg þ kdeg þ ka;p kdeg þ ka;p kdeg ðkdeg þ ka;p Þ expðka;d TÞ ka;d expððkdeg þ ka;p Þ TÞ þ ðkdeg þ ka;p Þðka;d kdeg ka;p Þ
(8)
(9)
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 2, FEBUARY 2010
1080
MIZUMA
Under the following conditions: ka;p ¼ a Ppro
(10)
ka;d ¼ a Pdru
(11)
A¼aT
(12)
Then, from Eqs. (16) and (17), general expressions (Eqs. 18 and 19, respectively) are obtained for stable compounds (drug or prodrug) in the luminal tract. P and CL are the permeation coefficient ( P) and absorption clearance (CLabs). A or B is calculated from Pdru or CLabs,d
Fa is expressed by
Fa ¼
Fa ¼ 1 expðA PÞ
Ppro ð1 expðA ðPdeg þ Ppro ÞÞ Pdeg þ Pdeg þ Ppro Pdeg þ Ppro Pdeg ðPdeg þ Ppro Þ expðA Pdru Þ Pdru expðA ðPdeg þ Ppro ÞÞ þ ðPdeg þ Ppro ÞðPdru Pdeg Ppro Þ
where Ppro and Pdru are the membrane permeabilities of the prodrug and drug, respectively, which can be obtained using Caco-2 cell monolayers,4 an artificial membrane,5–7 and so on. a is a scaling constant, including several factors such as the membrane surface area and Pdeg ¼ kdeg/a. Since membrane permeability ( P) area of the membrane (S) ¼ clearance (CL), Fa is also expressed by
Fa ¼
(18)
(13)
or Fa ¼ 1 expðB CLabs Þ
(19)
Eq. (18) corresponds to an equation in drug studies (not prodrugs) reported by several researchers.8–10 Here they have assumed that drugs are stable in the luminal tract.
CLabs;p ð1 expðB ðCLdeg þ CLabs;p ÞÞ CLdeg þ CLdeg þ CLabs;p CLdeg þ CLabs;p CLdeg ðCLdeg þ CLabs;p Þ expðB CLabs;d Þ CLabs;d expðB ðCLdeg þ CLabs;p ÞÞ þ ðCLdeg þ CLabs;p ÞðCLabs;d CLdeg CLabs;p Þ
B¼
A S
(15)
where CLa,p, CLa,d, and CLdeg are the absorption clearance of the prodrug, absorption clearance of the drug, and degradation clearance of the prodrug to drug in the luminal tract, respectively. CLa,p and CLa,d can be obtained using Caco-2 cell monolayers,4 an artificial membrane,5–7 and so on. When a prodrug is stable in the luminal tract, Pdeg and CLdeg are null. Accordingly, Fa of the prodrug is expressed by Fa ¼ 1 expðA Ppro Þ
(16)
Fa ¼ 1 expðB CLabs;p Þ
(17)
or
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 2, FEBUARY 2010
(14)
Once A or B is calculated according to Eq. (18) or (19) from Fa and P or CLabs of a stable compound, for example, a drug in the present study, Fa of the prodrug can be predicted under the assumption that the prodrug is stable in the luminal tract. When the predicted Fa of the prodrug is lower than the observed (reported) Fa, this suggests that the prodrug may be unstable in the luminal tract, and that Pdeg and CLdeg can be estimated by Eqs. (13) and (14), respectively.
Data Collection and Estimation of Luminal Degradation Clearance (CLdeg) CLdeg is estimated using Eq. (14) from B, Fa, CLabs,p, and CLabs,d. B is obtained according to DOI 10.1002/jps
LUMINAL DEGRADATION OF ORALLY EFFECTIVE PRODRUG
Eq. (19) from Fa and CLabs of the drug described above. Fa, CLabs,d, and CLabs,p were retrieved from the literature. Fa were from the literature on lenampicillin,11 ampicillcin,12 pivampicillin,13 valacyclovir,14 and acyclovir.15 CLabs,p of lenampicillin and CLabs,d of ampicillin were from the previous report.3 When studies (pivampicillin and its drug study16 and valacyclovir and its drug study17) with Caco-2 cell monolayers reported the permeation coefficient ( P), CLabs of the prodrug (CLabs,p) and drug (CLabs,d) were calculated from the reported P of the prodrug and drug, respectively, using the following equation: CLabs ðnL=min=cm2 Þ ¼ 60 106 P ðcm=sÞ (20)
Simulation of Impact of Luminal Degradation Clearance on Absorbed Fraction Impact of luminal degradation of the prodrug on the fraction absorbed of prodrug was simulated based on Eq. (14) with the following values of ampicillin3 as an example: CLabs;d ðCLabs of drugÞ ¼ 3:8 nL=min=cm2 B ¼ 0:06203
intra-tissue metabolism data (Class E) CLp;mi fp;muc CLd;mi fd;muc OE CLd;is CLp;is CLp;im CLd;im
CLp;mi fp;muc > CLd;mi fd;muc OE CLd;is CLp;is < CLp;im CLd;im
CLd;is CLp;is CLp;im CLd;im
ðCLp;is CLd;im þ CLp;is CLd;is þ CLmet;int CLd;is Þ CLd;mi fd;muc OE ðCLp;im CLd;is þ CLp;is CLd;is þ CLmet;int CLd;is Þ CLp;mi fp;muc Oral Prodrug Classification System (OPCS) Required minimum Fa ( Fa,min), which is required Fa for orally effective prodrug, is expressed by the following equation: Fa;min ¼ 1 expðB CLabs;d OEÞ
(21)
CLabs;p;min ¼ CLabs;d OE CLabs,p,min is the absorption clearance required for an orally effective prodrug. Orally effective prodrugs are classified based on two factors, the uptake ratio and efflux ratio, of the prodrug and drug (Fig. 2), which is based on the previous report.2 When the prodrug satisfies both Eqs. (22) and (23), the prodrug is effective, irrespective of DOI 10.1002/jps
(22) (23)
where CLp,m-i and CLd,m-i are uptake clearances of the prodrug and drug, respectively. fp,muc and fd,muc are unbound fractions of the prodrug and drug, respectively, on the mucosal side. CLp,i-s and CLp,i-m are efflux clearances of the prodrug from inside the intestinal tissue to the serosal side and from the inside of intestinal tissue to the mucosal side, respectively. CLd,i-s and CLd,i-m are efflux clearances of the drug from inside the intestinal tissue to the serosal side and from the inside of intestinal tissue to the mucosal side, respectively. When the prodrug satisfies Eqs. (24) and (25) or Eqs. (26) and (27), the prodrug is effective only when satisfying the conditional Eq. (28) (Class M)
CLp;mi fp;muc < CLd;mi fd;muc OE
Accordingly, Fa is expressed by a function of CLdeg and CLabs,p.
1081
(24) (25) (26) (27)
(28)
where CLmet,int is the intrinsic clearance for intratissue metabolism. When the prodrug satisfies Eqs. (29) and (30) or Eqs. (31) and (32), the prodrug is ineffective, irrespective of intra-tissue metabolism data (Class I) CLp;mi fp;muc ¼ CLd;mi fd;muc OE CLd;is CLp;is < CLp;im CLd;im CLp;mi fp;muc < CLd;mi fd;muc OE CLd;is CLp;is CLp;im CLd;im
(29) (30) (31) (32)
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 2, FEBUARY 2010
1082
MIZUMA
Figure 2. Oral prodrug classification system (OPCS) showing kinetic classification and criteria for orally effective prodrugs. CLp,m-i and CLd,m-i are uptake clearances of the prodrug and drug, respectively. fp,muc and fd,muc are unbound fractions of the prodrug and drug, respectively, on the mucosal side. CLp,i-s and CLp,i-m are efflux clearances of the prodrug from the inside of the intestinal tissue to the serosal side and from the inside of the intestinal tissue to the mucosal side, respectively. CLd,i-s and CLd,i-m are efflux clearances of the drug from the inside of intestinal tissue to the serosal side and from the inside of intestinal tissue to the mucosal side, respectively. CLmet,int is intrinsic clearance of intra-tissue metabolism.
RESULTS AND DISCUSSION Reported Fa Versus Predicted Fa of Prodrugs on the Assumption of Stability in the Luminal Tract The relationships between Fa and CLabs are shown in Figure 3. The predicted Fa ( Fa,pred) of prodrugs was calculated according to Eq. (19) from B and CLabs on the assumption that prodrugs are stable in the luminal tract. Table 1 summarizes the calculated values (B, Fa,pred, and CLdeg) and the reported data (CLabs,d, CLabs,p, and Fa) that were used for the calculation. The reported Fa of JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 2, FEBUARY 2010
lenampicillin was approximately 50%, whereas the predicted Fa ( Fa,pred) of lenampicillin was almost 100% (Fig. 3a). This clearly indicates that the reported Fa of lenampicillin was lower than the predicted Fa. Also, the reported Fa of pivampicillin is slightly lower than the predicted Fa (Fig. 3b). In contrast, the reported Fa of valacyclovir is comparable to the predicted Fa (Fig. 3c). These results suggest that lenampicillin and pivampicillin are unstable in the luminal tract, whereas valacyclovir may be stable in the luminal tract. The absorption clearance or permeation coefficient of lenampicillin and its drug, pivampicillin DOI 10.1002/jps
LUMINAL DEGRADATION OF ORALLY EFFECTIVE PRODRUG
1083
Figure 4. Impact of luminal degradation clearance (CLdeg) on fraction absorbed ( Fa). Condition: Equation (14), B ¼ 0.06203, CLabs,d ¼ 3.8 nL/min/cm2.
and its drug, and valacyclovir and its drug was from the reports of Mizuma,3 He et al.,16 and Katragadda et al.,17 respectively, giving each B value due to a difference in experiment conditions conducted by research groups. CLdeg of Lenampicillin, Pivampicillin, and Valacyclovir CLdeg of lenampicillin, pivampicillin, and valacyclovir are summarized in Table 1. CLdeg of lenampicillin was 1107 nL/min/cm2, approximately 1.7 times as large as CLabs,p. This indicates that luminal degradation impacts on Fa, and the stabilization of lenampicillin in the luminal tract can improve Fa, whereas CLdeg of pivampicillin was 162 nL/min/cm2, and was approximately onetenth of CLabs,p. This indicates that pivampicillin degrades but its CLdeg is much lower than its CLabs,p. The results of these ampicillin prodrugs, lenampicillin and pivampicillin, suggest that (5-methyl-2-oxo-1,3-dioxol-4-yl)methyl ester derivative (pivampicillin) may be more stable than (2,2-dimethyl-1-oxopropoxy)methyl ester derivative (lenampicillin) in the luminal tract when compared with the respective membrane permeability.
Figure 3. Relationship between fraction absorbed ( Fa) and absorption clearance (CLabs) of drugs (closed circles) and their prodrugs (open circles). Curves represent predicted Fa calculated by the equation Fa ¼ 1exp(B CLabs), with CLabs of drug (CLabs,d). (a) lenampicillin (ampicillin prodrug), (b) pivampicillin (ampicillin prodrug), (c) valacyclovir (acyclovir prodrug).
Table 1. Kinetic Characteristics of Orally Effective Prodrugs B
Drug
Prodrug (Drug) Lenampicillin (ampicillin) Pivampicillin (ampicillin) Valacyclovir (acyclovir) DOI 10.1002/jps
CLabs,d (nL/min/cm2)
Fa
3.80 120 90.0
0.210 0.210 0.225
Prodrug
CLabs,p 1/(nL/min/cm2) (nL/min/cm2) Fa,pred 0.06203 0.001964 0.002832
660.0 1932 265.8
Fa
1.00 0.504 0.978 0.920 0.529 0.542
CLdeg CLdeg/ (nL/min/cm2) CLabs,p 1107 162.0 —
1.68 0.084 —
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 2, FEBUARY 2010
1084
MIZUMA
Scheme 1. Procedures for rational development of orally effective prodrugs.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 2, FEBUARY 2010
DOI 10.1002/jps
LUMINAL DEGRADATION OF ORALLY EFFECTIVE PRODRUG
In contrast, the predicted Fa of valacyclovir was comparable to the reported Fa (Tab. 1), indicating that the luminal degradation of valacyclovir is negligible. At the same time, because PEPT1mediated absorption of valacyclovir has been reported,18 it is another possibility that valacyclovir in the luminal tract was degraded but Caco-2 cells with lower PEPT1 activity gave lower CLabs,p. Thus, lower Fa due to luminal degradation and underestimation of in vitro CLabs,p resulted in similar values of Fa and the predicted Fa. The kinetic analysis in the present study was triggered by the difference between the predicted Fa of prodrug on the assumption of stability in the luminal tract and the observed Fa. This difference may have resulted from other causes, but the predicted Fa was lower but not higher than the observed Fa for all prodrugs studied in this report. Therefore, luminal degradation of the prodrug is considered to be the most appropriate cause for the difference at present. B value is a scaling factor used for absorption clearance, and is similar to A value, which is a scaling factor used for permeation coefficient. Several researchers have used an equation equivalent to Eq. (18) involving A value, which has several factors. Thus, this analysis was based on the assumption that A and B values are constant, as several researchers have done for drugs.
Impact of Luminal Degradation on Fraction Absorbed The impact of luminal degradation on the fraction absorbed is shown in Figure 4. Ampicillin, the CLabs (CLabs,d) of which is 3.8 nL/min/cm2, is exemplified. When 50% of minimum Fa is required and CLdeg is null, only 11.9 nL/min/cm2 of CLabs,p is needed, whereas when 50% of minimum Fa is required and CLdeg is 1000 nL/min/cm2, CLabs,p must be higher than 577.5 nL/min/cm2. In other words, when 50% of minimum Fa is required and CLdeg is 16.0 nL/min/cm2, only 16.0 nL/min/ cm2 of CLabs,p equal to CLdeg is needed. This indicates that CLdeg impacts on Fa of the prodrug, and that even if membrane permeability is improved by prodrugging, luminal degradation must be studied to rationally develop orally effective prodrugs. A future study is to experimentally estimate the luminal degradation kinetics. DOI 10.1002/jps
1085
Procedure for Rational Development of Orally Effective Prodrugs Because luminal degradation of the prodrug or prodrug candidate should be considered as described above, a procedure for developing orally effective prodrugs, which is considered for luminal degradation of the prodrug, is proposed (Scheme 1) based on the previous report.3 The first decision condition (decision condition 1) is based on the predicted Fa, which is calculated by Eq. (1) with CLabs of a prodrug candidate determined by the in vitro experiment. When the predicted Fa is higher than or equal to the minimum required Fa ( Fa Fa,min), the prodrug candidate is promising. The second decision condition (decision condition 2) is based on OPCS (Fig. 2), which is based on the previously reported decision tree for the development procedure of an orally effective prodrug (kinetic classification and criteria for effective membrane-permeable prodrugs).2 When a prodrug candidate is classified into Class E indicating ‘‘Effective,’’ the prodrug candidate is promising. When a prodrug candidate is classified into Class I (Ineffective), another prodrug candidate should be designed with consideration of the results of OPCS. When a prodrug candidate satisfies Eq. (1) (decision condition 1) and is classified into Class E (decision condition 2), the prodrug candidate may be forwarded to ‘‘Go’’ in the in vivo study. Fa is calculated and CLdeg is assessed according to Eq. (3) with data from the in vivo study. When Fa is higher than or equal to a required minimum Fa, the prodrug candidate is judged successful. When CLdeg is larger than CLabs,p, for example, CLdeg/CLabs,p is 2 or higher, the luminal stability of the prodrug candidate should be improved. Since the present study focused on prodrugs with an ester bond, the next target is an orally active thiamine prodrug, fursultiamine,19 which has a disulfide bond. In summary, luminal degradation of the oral prodrugs available on the market was assessed using in vitro and in vivo data in humans. The luminal degradation clearance of lenampicillin was higher than the absorption clearance, indicating that not only membrane permeability but also luminal stability must be studied for the development of orally effective prodrugs. This article is the first report assessing the luminal degradation kinetics of oral prodrugs in humans. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 2, FEBUARY 2010
1086
MIZUMA
ACKNOWLEDGMENTS This study was supported in part by a grant from The Research Foundation for Pharmaceutical Sciences.
REFERENCES 1. Ettmayer P, Amidon GL, Clement B, Testa B. 2004. Lessons learned from marketed and investigational prodrugs. J Med Chem 47:2393–2404. 2. Mizuma T. 2008. Pharmacokinetic strategy for designing orally effective prodrugs overcoming biological membrane barriers: Proposals of kinetic classification and criteria for membrane-permeable prodrug-likeness. Chem-Bio Inform J 8:25–32. www. cbi.or.jp/cbi/CBIj/vol8/8_25-E.pdf (accessed June 9, 2008). 3. Mizuma T. 2008. Integrated pharmacokinetic assessment and strategy for orally effective prodrugs overcoming luminal degradation and biological membrane barriers. Chem-Bio Inform J 8:58–68. www.cbi.or.jp/cbi/CBIj/vol8/8_58-E.pdf (accessed November 18, 2008). 4. Artursson P, Karlsson J. 1991. Correlation between oral drug absorption in humans and apparent drug permeability coefficients in human intestinal epithelial (Caco-2) cells. Biochem Biophys Res Commun 175:880–885. 5. Kansy M, Senner F, Gubernator K. 1998. Physicochemical high throughput screening: Parallel artificial membrane permeation assay in the description of passive absorption processes. J Med Chem 41:1007–1010. 6. Kerns EH. 2001. High throughput physicochemical profiling for drug discovery. J Pharm Sci 90:1838– 1858. 7. Sugano K, Nabuchi Y, Machida M, Aso Y. 2003. Prediction of human intestinal permeability using artificial membrane permeability. Int J Pharm 257: 245–251. 8. Amidon GL, Sinko PJ, Fleisher D. 1988. Estimating human oral fraction dose absorbed: A correlation using rat intestinal membrane permeability for passive and carrier-mediated compounds. Pharm Res 5:651–654.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 99, NO. 2, FEBUARY 2010
9. Yamashita S, Furubayashi T, Kataoka M, Sakane T, Sezaki H, Tokuda H. 2000. Optimized conditions for prediction of intestinal drug permeability using Caco-2 cells. Eur J Pharm Sci 10:195–204. 10. Zhu C, Jiang L, Chen TM, Hwang KK. 2001. A comparative study of artificial membrane permeability assay for high throughput profiling of drug absorption potential. Eur J Med Chem 37:399– 407. 11. Schering-Plough. 2008. Package insert of varacillin tablets. Tokyo, Japan: Schering-Plough. 12. Meiji Seika. 2005. Package insert of viccillin capsules. Tokyo, Japan: Meiji Seika. 13. Ehrnebo M, Nilsson SO, Bore´ us LO. 1979. Pharmacokinetics of ampicillin and its prodrugs bacampicillin and pivampicillin in man. J Pharmacokinet Biopharm 7:429–451. 14. Soul-Lawton J, Seaber E, On N, Wootton R, Rolan P, Posner J. 1995. Absolute bioavailability and metabolic disposition of valacyclovir, the L-valyl ester of acyclovir, following oral administration to humans. Antimicrob Agents Chemother 39:2759– 2764. 15. Benet LZ, Sheiner LB. 1985. Design and optimization of dosage regimens: Pharmacokinetic data. In: Gilman AG, Goodman LS, Rall TW, Murad F, editors. The pharmacological basis of therapeutics. 7th edition. New York: Macmillan Publishing Co. pp. 1663–1733. 16. He X, Sugawara M, Takekuma Y, Miyazaki K. 2004. Absorption of ester prodrugs in Caco-2 and rat intestine models. Antimicrob Agents Chemother 48:2604–2609. 17. Katragadda S, Jain R, Kwatra D, Hariharan S, Mitra AK. 2008. Pharmacokinetics of amino acid ester prodrugs of acyclovir after oral administration: Interaction with the transporters on Caco-2 cells. Int J Pharm 362:93–101. 18. Han H, de Vrueh RL, Rhie JK, Covitz KM, Smith PL, Lee CP, Oh DM, Sade´ e W, Amidon GL. 1998. 5’-Amino acid esters of antiviral nucleosides, acyclovir, and AZT are absorbed by the intestinal PEPT1 peptide transporter. Pharm Res 15:1154– 1159. 19. Kitamori N, Itokawa Y. 1993. Pharmacokinetics of thiamin after oral administration of thiamin tetrahydrofurfuryl disulfide to humans. J Nutr Sci Vitaminol 39:465–472.
DOI 10.1002/jps