Purely in Silico BCS Classification: Science Based Quality Standards

Permeability classification with the different in silico methods was correct for 69–72.4% ...... New scientific approaches to international regulato...
0 downloads 0 Views 2MB Size
Subscriber access provided by BRANDEIS UNIV

Article

Purely in-silico BCS classification: Science based quality standards for the world's drugs Arik Dahan, Omri Wolk, Young Hoon Kim, Chandrasekharan Ramachandran, Gordon M. Crippen, Toshihide Takagi, Marival Bermejo, and Gordon L Amidon Mol. Pharmaceutics, Just Accepted Manuscript • DOI: 10.1021/mp400485k • Publication Date (Web): 04 Oct 2013 Downloaded from http://pubs.acs.org on October 5, 2013

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Molecular Pharmaceutics is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

Purely in-silico BCS classification: Science based quality standards for the world's drugs

Arik Dahan1,*, Omri Wolk1, Young Hoon Kim2,3, Chandrasekharan Ramachandran2, Gordon M. Crippen2, Toshihide Takagi2, Marival Bermejo4, and Gordon L. Amidon2

1.

Department of Clinical Pharmacology, School of Pharmacy, Faculty of Health Sciences, Ben-Gurion University of the Negev, Beer-Sheva 84105, Israel

2.

Department of Pharmaceutical Sciences, College of Pharmacy, University of Michigan, Ann Arbor, Michigan 48109.

3.

Korea Food and Drug Administration, Seoul, South Korea

4.

Department of Engineering, Pharmacy Section, Miguel Hernandez University, Alicante, Spain

*

Corresponding Author: Department of Clinical Pharmacology,

School of Pharmacy, Faculty of Health Sciences, Ben-Gurion University of the Negev, P.O. Box 653, Beer-Sheva 84105, Israel. E-mail: [email protected]. Tel: +972-8-6479483. Fax: +972-8-6479303.

1 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ABSTRACT BCS classification is a vital tool in the development of both generic and innovative drug products. The purpose of this work was to provisionally classify the world's top selling oral drugs according to the BCS, using in-silico methods. Three different in-silico methods were examined: the wellestablished group contribution (CLogP) and atom contribution (ALogP) methods, and a new method based solely on the molecular formula and element contribution (KLogP). Metoprolol was used as the benchmark for the low/high permeability class boundary. Solubility was estimated in-silico using a thermodynamic equation that relies on the partition coefficient and melting point. The validity of each method was affirmed by comparison to reference data and literature. We then used each method to provisionally classify the orally administered, IR drug products found in the WHO Model list of Essential Medicines, and the top-selling oral drug products in the United States (US), Great Britain (GB), Spain (ES), Israel (IL), Japan (JP), and South Korea (KR). A combined list of 363 drugs was compiled from the various lists, and 257 drugs were classified using the different in-silico permeability methods and literature solubility data, as well as BDDCS classification. Lastly, we calculated the solubility values for 185 drugs from the combined set using insilico approach. Permeability classification with the different in silico methods was correct for 69-72.4% of the 29 reference drugs with known human jejunal permeability, and for 84.6–92.9% of the 14 FDA reference drugs in the set. The correlations (r2) between experimental log P values of 154 drugs and their CLogP, ALogP and KLogP were 0.97, 0.82 and 0.71, respectively. The different in-silico permeability methods produced comparable results; 30-34% 2 ACS Paragon Plus Environment

Page 2 of 54

Page 3 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

of the US,GB,ES and IL top selling drugs were Class 1, 27-36.4% were Class 2, 22-25.5% were Class 3, and 5.46-14% were Class 4 drugs, while ~8% could not be classified. The WHO list included significantly less Class 1 and more Class 3 drugs in comparison to the countries lists, probably due to differences in commonly used drugs in developing vs industrial countries. BDDCS classified more drugs as class 1 compared to in-silico BCS, likely due to the more lax benchmark for metabolism (70%), in comparison to the strict permeability benchmark (metoprolol). For 185 out of the 363 drugs, in-silico solubility values were calculated, and successfully matched the literature solubility data. In conclusion, relatively simple in-silico methods can be used to estimate both permeability and solubility. While CLogP produced the best correlation to experimental values, even KLogP, the most simplified in-silico method that is based on molecular formula with no knowledge of molecular structure, produced comparable BCS classification to the sophisticated methods. This KLogP, when combined with a mean melting point and estimated dose, can be used to provisionally classify potential drugs from just molecular formula, even before synthesis. 49-59% of the world's top-selling drugs are highly soluble (Class 1 and Class 3), and are therefore candidates for waivers of in-vivo bioequivalence studies. For these drugs, the replacement of expensive human studies with affordable in-vitro dissolution tests would ensure their bioequivalence, and encourage the development and availability of generic drug products in both industrial and developing countries. KEYWORDS:

BCS, BDDCS, In silico, NME, Partition coefficient,

Permeability, Solubility. 3 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

INTRODUCTION

Since its publication in 1995 1, the biopharmaceutics classification system (BCS) has become an increasingly important tool in the development and regulation of both generic and innovative drug products worldwide 2-4. The BCS has revealed that the fundamental parameters controlling the rate and extent of oral absorption are the intestinal permeability and the aqueous solubility/dissolution of the drug dose. The BCS has presented a new paradigm in bioequivalence (BE); historically, BE standard was essentially empirical, based on relative in-vivo bioavailability (BA) studies, i.e. plasma levels, AUC and Cmax. By revealing the fundamental parameters determining the in-vivo oral drug absorption process, the BCS has enabled BE assurance through mechanistic tools, rather than empirical observation; if two drug products that contain the same active pharmaceutical ingredient (API) have a similar gastrointestinal (GI) concentration-time profile (i.e., in-vivo GI release) under all luminal conditions, then a similar rate and extent of absorption is ensured for these products, and they will be necessarily bioequivalent 5. Therefore, BE can be assured based on dissolution tests that provide the mechanistic proof for similar bioavailability, rather than empirical in-vivo human studies 6, 7. This is the scientific and mechanistic rationale provided by the BCS, for the regulatory waiver of in-vivo BE 8-12. Likewise, the BCS greatly influenced the decision making in the discovery and early development of new molecular entities (NMEs). Early determination of the rate-limiting step in the oral absorption

4 ACS Paragon Plus Environment

Page 4 of 54

Page 5 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

process, and preliminary BCS classification of pipeline compounds, may greatly facilitate intelligent lead compound selection, prioritization of salt, polymorph, and other formulation strategies 4, 13-15. In the past decade, the validity and broad applicability of the BCS have been the subject of extensive research and discussion, including an effort to draw a BCS classification of many drug products. The majority of the data is based on secondary aqueous solubility references and permeability estimations based on correlations with Log P and CLogP. As such, the classifications are provisional and can be revised as more experimental data become available. Kasim et al 16 provisionally classified the WHO Essential Medicines List, using standard solubility references (from e.g. Merck Index, USP, etc) that together with the maximum dose strengths allowed to estimate the dose number (D0), and permeability estimations were based on correlation of the n-octanol/water partition coefficients using both Log P and CLogP of the uncharged form of the drug molecule 16. This provisional classification was later extended to include the top 200 drugs on the United States, Great Britain, Spain, and Japan lists 17, as well as common herbs used in Western and Chinese medicine 18, 19. A subsequent classification of the WHO list of Essential Medicines that was based primarily on human fraction absorbed (Fabs) literature data for the permeability assignment, produced similar results 20

. More recently, a larger dataset of drugs was provisionally classified using

dose number calculated from reference solubility data, and permeability across Caco-2 cell monolayers 21. At the early stage of drug discovery and development, very little amount of the API is available for thorough evaluation of BCS classification. 5 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Hence, a reliable BCS classification based solely on in-silico approaches can be extremely valuable. Certainly, the underlying assumptions and methods used in any computational approach should be carefully evaluated; however, the continuous progress, convenience, and feasibility of in-silico methods attract increasing interest. The purpose of this work was to provisionally classify the world's top selling oral drugs according to the BCS, using in-silico methods. For the permeability estimations, three different in-silico methods were examined: the well-established group contribution (CLogP) and atom contribution (ALogP) methods, and a new method based solely on the molecular formula and element contribution (KLogP), even without any knowledge of the molecular structure. Solubility was estimated in-silico using the thermodynamic equation developed by Amidon and Williams 22 that allows for the deduction of solubility from the partition coefficient and the melting point, which together with the highest dose strengths, allowed us to calculate the dose number. The validity of each method was affirmed by comparison to reference data and literature. We then used each method to provisionally classify the orally administered, immediate-release (IR) drug products found in the WHO Model list of Essential Medicines, and the top-selling oral drug products in the United States (US), Great Britain (GB), Spain (ES), Israel (IL), Japan (JP), and South Korea (KR). This BCS classification was also compared to previously published BDDCS classification 8, 23, 24. This work shows that adequate accuracy may be achieved when classifying drugs using relatively simplified in-silico methods, which may be extremely useful in the discovery and early development of NMEs. The results of the provisional classification suggest 6 ACS Paragon Plus Environment

Page 6 of 54

Page 7 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

that more than 50% of the top world's drug products are highly soluble and are hence candidates for a biowaiver.

7 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

METHODS

Drug lists Lists of the top 200 US, GB, ES and JP drugs were obtained from IMS Health (Fairfield, CT). A list of the top 165 IL drugs was obtained from the Israeli Ministry of Health. The list of the top 300 KR drugs was obtained from the Korea Food and Drug Administration (KFDA). The WHO Model List of Essential Medicines is available at the WHO website (www.who.int). Only immediate-release solid oral dosage forms were selected for the provisional BCS classification, and the selection was based on information obtained from the Electronic Orange Book for the drugs in the US list (www.fda.gov), www.medicines.org.uk for those in the GB list, www.portalfarma.com and information provided by the Spanish Agency of Medicines for those in the ES list, the Israeli Ministry of Health (www.health.gov.il) for those in the IL list, the Japanese Orange Book (www.jpora452.rsjp.net) for those in the JP list, the KFDA ezDrug Information (www.ezdrug.mfds.go.kr) and KIMS Online (www.kimsonline.co.kr) for those in the KR list. From the WHO list, only drugs that were specified as having oral IR formulation were included. The number of oral IR drugs on the US, GB, ES, IL, JP, KR, and WHO lists were 113, 102, 106, 97, 113, 126, and 133, respectively. A combined list was compiled by selecting oral IR drug products from the six lists that (a) appeared at least once on one or more lists; and (b) that was at the highest dose strength on the lists. A total of 363 drugs were thus on the combined list.

8 ACS Paragon Plus Environment

Page 8 of 54

Page 9 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

Partition coefficients n-octanol/water partition coefficients of the uncharged drug were calculated using several different methods. Experimentally measured (MLogP) values were obtained from BioLoom 5.0 (BioByte Corp., Claremont, CA). Calculations of Log P based on the contributions of functional groups and fragments attached to a base molecule (CLogP) were generated with algorithms based on theoretical treatments developed by Leo 25, and chemical structures of the drug as depicted in the Merck Index using BioLoom 5.0 and ChemDraw Ultra 8.0 software (CambridgeSoft Corp., Cambridge, MA). Log P values were also calculated using Molecular Operating Environment (MOE version 2004.03) based on atomic contribution to lipophilicity developed by Crippen and Wildman (ALogP) 26, 27. Lastly, a simplified approach based on element contribution, also used to calculate Log P. This parameter, hereinafter designated KLogP was derived solely from the molecular formula of the drug; 154 drugs from the common list for which the experimental MLogP values were available were classified using 9 parameters. Each parameter represents one of the elements that are commonly present in organic compounds, i.e, C, H, N, O, F, P, S, Cl and Br. The number of repetitions of each element was extracted for every drug from its molecular formula. The resulting list of elements types and numbers in each drug was then fitted against the experimental MLogP data by linear regression and least squares fit using Microsoft Office Excel 2010 software, producing the following equation:

9 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

KLog P = (nH × H) + (nC × C) + (nN × N) + (nO × O) + (nF × F) + (nP × P) + (nS × S) + (nCl × Cl) + (nBr × Br) Where ni = number of type i atoms, and the corresponding contribution coefficients are the output of Excel "LINEST" function: H = -0.0060, C = 0.2463, N = -0.3163, O = -0.4515, F = -0.1033, P = -0.7753, S = 0.0450, Cl = 0.6534, and Br = -0.2781.

Ionization constants and distribution coefficients Log D, the pH-dependent distribution coefficient for singly ionized species, was calculated from the estimated CLogP and the ionization constant (pKa) using the following equations 28: For acids, Log D = CLogP – Log (1+10 (pH - pKa)) For bases, Log D = CLogP – Log (1+10 (pKa - pH)) CLogP and literature pKa values were obtained from BioLoom 5.0 (BioByte Corp., Claremont, CA). The acid or base characteristic of drug dissociation was validated with ADMET Predictor 1.2.3 (SimulationPlus Inc., Lancaster, CA). Log D was calculated for pH=6.5 and pH=7.4, as these are the physiological pH levels at the upper and lower small intestine, respectively.

Correlation of human intestinal permeability with partition and distribution coefficients 10 ACS Paragon Plus Environment

Page 10 of 54

Page 11 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

The permeability classification is based on correlations of the experimentally determined human intestinal permeability for 29 reference drugs 17 with the aforementioned partition (CLogP, ALogP or KLogP) and distribution (LogD6.5, LogD7.4) coefficients. Metoprolol was chosen as the reference compound for permeability since 95% of the drug is known to be absorbed from the gastrointestinal tract 29-31. Thus, drugs that exhibit higher/equal or lower partition coefficients, distribution coefficients and human intestinal permeability than the corresponding value for metoprolol are considered high-permeability and low-permeability drugs, respectively. Drugs that exhibit human intestinal permeability greater than metoprolol but with partition or distribution coefficients values lower than that of metoprolol are termed false negatives. False positives, on the other hand, are drugs with partition or distribution coefficients values higher than that of metoprolol but with corresponding experimental human intestinal permeability that is lower than that of metoprolol.

Solubility

Solubility from reference literature The aqueous drug solubility values (mg/mL) were obtained from the Merck Index, the USP, the Japanese prescription information, and the KFDA ezDrug Information. For cases where the reported solubility was significantly different in the above three sources, the lowest listed value was used in dose

11 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

number calculations. Whenever specific solubility values were not available, the lower limit of the solubility range defined in the USP was chosen as a conservative estimate (Table S1). For drugs that were listed as practically insoluble (pi), a more conservative value of 0.01 mg/mL (rather than 0.1 mg/mL in the USP definition) was used in dose number calculations.

In-silico drug solubility The intestinal permeability of a drug is often dependent on its hydrophobicity, which in turn is inversely correlated to the drug’s solubility in the aqueous GI milieu. For this reason, octanol/water partition coefficients, commonly used to assess the permeability of drugs, can also be utilized to assess their solubility. This method was used to calculate In silico solubility values based on the thermodynamic equation developed by Amidon and Williams 22: Log Cs = 1.05 – 0.0099MP – PC Where Cs is the molar aqueous solubility of the drug at 25 °C, MP is the melting point (in °C) for the uncharged drug molecule, and PC is one of the aforementioned in-silico partition coefficient estimations (CLogP, ALogP or KLogP). In-silico solubility could be calculated for a total of 185 drugs on the combined list for which experimentally determined melting points were available.

12 ACS Paragon Plus Environment

Page 12 of 54

Page 13 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

Dose Number Calculations The FDA guidelines specify that a drug substance is considered highly soluble when the highest dose strength is soluble in 250 ml or less of aqueous media over the pH range of 1-7.5. The volume estimate of 250 ml is derived from typical BE study protocols that prescribe administration of a drug product to fasting human volunteers with a glass (about 8 ounces) of water. The dimensionless dose-normalized solubility parameter, named dose number (D0), is calculated by the following equation 32:

Do =

(M

/ V 0) Cs 0

Where M0 is the maximum dose strength (milligrams, obtained from the electronic databases specified under "Drug Lists"), Cs is either the reference or calculated solubility (milligrams per milliliter), and V0 = 250 mL. Drugs with D0 ≤ 1 were classified as high-solubility drugs and drugs with D0 > 1 were assigned as low solubility drugs.

13 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

RESULTS

Correlation of measured and in silico partition coefficients The experimentally measured octanol/water partition coefficients (MLogP) available from the BioLoom database for 154 drugs along with in silico partition coefficients obtained using different algorithms are listed in Table S2. Linear correlation plots of the different in silico partition coefficients vs. experimentally determined octanol-water partition coefficients are shown in Figure 1. CLogP values estimated using either BioLoom 5.0 or ChemDraw 8.0 exhibited excellent correlations (r2 = 0.97) with measured values. The correlation of measured partition coefficients with in-silico ALogP or KLogP (based solely on molecular formula) was lower yet adequate (r2 = 0.82 and 0.71, respectively). The interrelation between the various in silico estimations mirrored the correlations observed with the measured values. Thus, the following linear correlations were evident (plots not shown): CLogP (BioLoom 5.0) vs. CLogP (ChemDraw 8.0) r2 = 0.97; CLogP (ChemDraw 8.0) vs. ALogP (MOE 2004.03) r2 = 0.89; ALogP (MOE 2004.03) vs. KLogP (Molecular Formula) r2 = 0.81; CLogP (BioLoom 5.0) vs. ALogP (MOE 2004.03) r2 = 0.78; CLogP (BioLoom 5.0) vs. KLogP (Molecular Formula) r2 = 0.68; CLogP (ChemDraw 8.0) vs. KLogP (Molecular Formula) r2 = 0.64.

14 ACS Paragon Plus Environment

Page 14 of 54

Page 15 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

Correlation of human jejunal permeability with partition coefficients The experimentally determined human jejunal permeability for 29 drugs 30, 33-35

is listed in Table 1 along with in silico partition coefficients of the

uncharged molecular form. Plots of experimentally determined human jejunal permeability against MLogP, CLogP, Log P, and KLog P, are shown in Figure 2, panels a-d, respectively. The classification of permeability based on metoprolol as a reference compound was correct for 20 (69.0%) and 19 (65.5%) drugs with CLogP using BioLoom 5.0 and ChemDraw 8.0, respectively. The permeability classification with ALogP and KLogP was similar and correct for 21 of the 29 drugs (72.4%). In general, polar molecules such as cephalexin 36, 37, enalapril 38, 39, levodopa 40, 41, L-leucine 40, 42, phenylalanine 43, 44, and valacyclovir 45-47 that are known to be transported by carrier-mediated mechanisms were classified as low permeability drugs (false negatives) on the plots. Conversely, in a few cases, drugs such as losartan 48, 49

and furosemide 50-52 that are substrates for efflux transporters were

classified as high permeability drugs (false positives). A breakdown of possible reasons for misclassification of drug intestinal permeability using the various partition coefficients is provided in Table 2.

Correlation of human intestinal permeability with distribution coefficients The calculated Log D values at pH 6.5 and pH 7.4 for the 29 reference drugs are also listed in Table 1. A plot of the experimentally determined human jejunal permeability against calculated Log D at pH 6.5 and pH 7.4 15 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

are shown in Figure 2, panels e and f, respectively. The classification of permeability based on metoprolol as a reference compound was correct for 20 of the 29 drugs (69.0%). The 4 drugs, cimetidine 53-55, furosemide 50, 51, ranitidine 53, 56, 57, and losartan 48, 49 that were classified as false positives are substrates for P-glycoprotein mediated efflux (Table 2). Similarly, the prediction of human jejunal permeability with Log D at pH 7.4 was correct for 20 of the 29 drugs (69.0%) (Table 2). The 3 drugs that were false positives, cimetidine, ranitidine and losartan are again substrates for P-glycoprotein. Furthermore, permeability classification with the different in silico methods was correct for 84.6–92.9% of the 14 FDA reference drugs in the 29 drugs set.

Solubility, permeability, and provisional BCS classification of the US, GB, ES, IL, JP, KR and WHO lists

Classification of drug permeability Drugs were classified as high or low permeability by comparing their partition coefficients values to the corresponding values of the reference drug metoprolol. Thus, drugs with CLogP, ALogP or KLogP values greater than or equal to 1.49, 1.61, and 1.87, respectively, were classified as highpermeability drugs. A total of 64-74.4%, 64-71.2%, 63.3-70.7%, 56.7-66%, 61.2-71.7% and 58.5-64.3% of the drugs on the US, GB, ES, IL, JP, KR list were classified as high-permeability drugs (Figure 3). The WHO list presented

16 ACS Paragon Plus Environment

Page 16 of 54

Page 17 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

significantly less high-permeability drugs (a 13%-25% difference) compared to the various countries lists. In general, the four different in-silico methods produced similar permeability classification in each list. However, ChemDraw 8.0, the older software used for the calculation of CLogP, was unable to classify roughly 4-15.6% of the drugs on the various lists. In addition, KLogP significantly underestimated high-permeability drugs in the JP list (~10% difference).

Classification of drug solubility by reference solubility data Dose numbers were calculated using solubility values from literature and maximum dose strength of the oral IR drug product on the six lists. A total of 49-59% of the drugs on each list were classified as high-solubility drugs using the highest maximum dose strengths on the US, GB, ES, JP, KR, and WHO lists (Figure 4). The KR list had a relatively high percentage of drugs for which there were no available data on solubility (~10%).

Provisional BCS classification of the various lists by in-silico permeability and reference solubility. Drug solubility values from reference literature along with in silico partition coefficients were used to provisionally classify the oral IR drugs on the various lists (Table S3 and Figure 5). An examination of Figure 5 reveals that the provisional BCS classifications of Western countries lists were quite similar. Thus, 30-34% of the US, GB, ES and IL top selling drugs were BCS

17 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Class 1, 27-36.4% were Class 2, 22-25.5% were Class 3, and 5.46-14% were Class 4 drugs, while up to 8% could not be classified, depending on the method used to calculate partition coefficients. The JP list presented a slightly higher percentage of Class 1 drugs (about 37% with all methods excluding KLogP). In contrast, the KR lists had significantly less Class 1 drugs (23.826.3%), as well as a high percentage of drugs that could not be classified (10.2%-14.6%). The WHO list presented significantly less Class 1 (17.421.2%) drugs and more Class 3 (29.5-34.1%) drugs when compared to the countries lists.

BCS and BDDCS The provisional BCS classification based on reference solubility values and in silico partition coefficients was compared to previously published BDDCS classification 58. From a total of 363 drugs on the combined list, 257 drugs were also classified by Benet et al. and were included in our analysis (Figure 6). BDDCS classification resulted in a significantly higher percentage of Class 1 drugs when compared to BCS in silico classification methods (~10% difference), likely due to the more lax benchmark for metabolism (70%), in comparison to the strict permeability benchmark (metoprolol).

Solubility classification using in-silico methods The solubility classification based on in silico solubility values was possible for a total of 185 drugs on the combined list for which melting point

18 ACS Paragon Plus Environment

Page 18 of 54

Page 19 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

data were available. Table S4 lists the solubility classification for the 185 drugs based on the different in silico methods, as well as reference literature solubility. Table S5 lists the maximum dose strengths, melting points, reference solubility, CLogP (BioLoom 5.0), KLogP, and the resulted in silico dose number (D0) calculated with CLogP (BioLoom 5.0) and KLogP for these 185 drugs. The average melting point of these 185 drugs (Table S5) was 162.7°C. The solubility classification based on estimation with various in-silico partition coefficients and experimental melting points along with the classification based on reference literature values is shown in Figure 7. An excellent agreement was obtained between the in-silico based classification and reference solubility data; 49.7% were classified as high-solubility drugs using reference literature solubility values, vs. 48.6-55.1% by the different in silico methods when experimental melting points were used. Notably, in-silico solubility classification with CLogP (BioLoom 5.0) and KLogP with experimental melting points vs. average melting point (162.7°C) produced similar results, with only ~2% difference (Figure 7).

Purely in-silico BCS classification Table S6 lists the provisional BCS classification of the 185 drugs based on in-silico and reference solubility classification, and permeability classification using CLogP, ALogP, and KLogP with experimental melting points. The corresponding in-silico provisional BCS classification based on CLogP (BioLoom 5.0) and KLogP and an average melting point (162.7 °C) can also be found on Table S6. A comparison plot of the BCS classification of

19 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

the 185 oral IR drugs using in silico and reference literature solubility approaches is shown in Figure 8. For a given solubility classification approach, it is evident that the BCS classification was quite similar when different in-silico partition coefficient methods were used. In contrast, the classification by the two solubility approaches exhibited some systematic differences. It can be seen that the in-silico solubility approaches underestimated Class 1 and overestimated Class 2 drugs by an identical average of 4.3 ± 1.0%. Similarly, overestimation of Class 3 and underestimation of Class 4 drugs by an identical average of 7.3 ± 0.7% was obtained with the in-silico solubility approaches compared to the reference literature data.

20 ACS Paragon Plus Environment

Page 20 of 54

Page 21 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

DISCUSSION

In this work we show that simplified in-silico methods can be used to estimate permeability, solubility, and BCS classification. While CLogP produced the best correlation to experimental values, even KLogP, the most simplified in-silico method that is based on molecular formula with no knowledge of molecular structure, produced comparable BCS classification to the more sophisticated methods. This KLogP, when combined with a mean melting point and estimated dose, can be used to provisionally classify potential drugs from just molecular formula, even before synthesis. This may become extremely valuable in the early stages of drug discovery and development, when very little (or no) amount of the API is available. The data presented in this work demonstrate that 49-59% of the world's top-selling drugs are highly soluble (BCS Class 1 or 3), and are therefore candidates for waivers of in-vivo bioequivalence studies. For these drugs, the replacement of expensive human studies with affordable in-vitro dissolution tests would ensure their bioequivalence, and encourage the development and availability of generic drug products in industrial, and even more important, in developing countries.

In silico partition coefficients In this work, we used three different in silico methods to estimate the octanol/water partition coefficient. Among these methods, the functional group

21 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

contributions (CLogP) is the oldest and most established one 25. Indeed, CLogP has produced an excellent correlation (r2 = 0.97) with MLogP (Figure 1, panels a and b). However, this method requires detailed knowledge of many intra-molecular factors, including various electronic and steric effects and hydrogen bonding Interactions 25. An in-silico method based on atom type contributions (ALogP) also showed good correlation with MLogP (r2 = 0.82) (Figure 1, panel c). This calculation allows for some intra-molecular factors to be ignored and thus simplifies the model 27, although this method still requires prior knowledge of the complete chemical structure of the compound. A third in-silico method, based on the atomic contribution approach, was also used to estimate the octanol-water partition coefficient; this revolutionary method (dubbed KLogP) relies strictly on element type contributions, with no prior knowledge of molecular structure. The reasonable correlation of KLogP with MLogP (r2 = 0.71) (Figure 1 panel d) indicates that the octanol/water partition coefficient can be adequately deduced merely from the molecular formula. 29 drugs for which experimentally determined human jejunal permeability values are available were used to validate the in-silico partition coefficient estimations. The permeability classification was correct for 69.072.4% of this set of 29 drugs using the different in-silico partition coefficient approaches (Figure 2, panels a-d), and for 84.6–92.9% of the 14 FDA reference drugs in the set. Majority of the misclassified drugs in the above set are known to be transported by carrier-mediated mechanisms (Table 2). The disadvantage of inaccurate prediction of permeability for carrier mediated substrates was expected and plausible, since carrier-mediated transport is dependent on specific molecule-transport interactions rather than on 22 ACS Paragon Plus Environment

Page 22 of 54

Page 23 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

lipophilicity considerations 59. Thus, the effects of carriers on overall permeability must be modeled separately. The comparable permeability classification of the 29 reference drugs using the different in-silico methods (69.0–72.4%) suggests that simple element type contributions allow similar results to that based on more sophisticated in-silico methods. This may become extremely valuable in the early stages of drug discovery and development, as very little (or no) amount of the API is available.

In-silico distribution coefficients Since ionized molecules do not readily cross the cellular membranes of the epithelium, the intestinal permeability of ionizable drugs may be pH dependent 9, 29, 60, 61. We estimated Log D values at pH 6.5 and pH 7.4 for the 29 reference drugs, and compared them to the respective human intestinal permeability values, to evaluate this pH-dependency. The results did not show any improvement of accuracy when Log D was used (Figure 2, panels e and f). Furthermore, Log D is often difficult to be determined or calculated, especially for drugs with multiple ionizable groups. For these reasons, we chose to proceed with Log P for the subsequent provisional BCS classification.

Permeability classifications The provisional permeability classification of drugs on the individual US, GB, ES, IL, JP, KR and WHO lists illustrated in Figure 3 indicates good

23 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

agreement between the different in-silico methods. However, a few trends can be observed. First, the differences in classification between the two CLogP methods may be atributable in part to the failure of the ChemDraw 8.0 software to estimate the permeability for some drugs. On average, 7% of the drugs on all lists are classified differently by the two methods; this is also the average percent of drugs with unavailable CLogP (ChemDraw 8.0) values. Second, the differences in classification between the CLogP (BioLoom 5.0) and KLogP methods are insignificant for drugs on the US, GB, ES, IL and WHO lists (average difference of 1.9%). However, the differences between the two methods for drugs on the JP, and KR lists are larger (average difference of 7.9%). Although it is not clear why this systematic difference occurs, it is possible that a greater percentage of drugs on the JP and KR lists may be structurally and inherently more complex than those on the western markets and requires more refined contribution methods to estimate partition coefficients.

Solubility classification based on reference and literature data The percentage of high-solubility drugs using reference and literature solubility data was similar for drugs on the US, GB, ES, and WHO lists (55.1 ± 1.1). The slightly higher percentage of high-solubility drugs on the JP list (59.3%) may reflect the significantly lower average maximum dose strength of the drugs on this list noted earlier 17. On the other hand, the lower percentage of high-solubility drugs on the KR list (49%) may be due to the higher percentage of drugs for which solubility values were unavailable (Figure 4).

24 ACS Paragon Plus Environment

Page 24 of 54

Page 25 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

For generic products of these drugs, bioequivalence can be ensured by well validated in-vitro dissolution studies. Low solubility drugs are more complex, and would require the development of more advanced dissolution assays which could better simulate the physiological dissolution process.

Provisional BCS classification by in-silico permeability and reference solubility The provisional BCS classification of drugs on the various lists based on reference solubility dose number and the different in-silico methods (Figure 5) exhibits some distinct trends. In general, the classification of drugs on the US, GB, ES an IL lists were quite similar regardless of the in-silico method being used. Distinct differences are evident with the JP, KR, and WHO lists compared to the US, GB, ES and IL lists. The percentage of Class 1 drugs on the JP list was slightly higher than those on the other lists (~35.0%). In contrast, the percentage of Class 1 drugs on the KR list was lower (~24. %) which may be explained by the relatively high percentage of unclassifiable drugs in this list. Also, the percentage of Class 1 drugs on the WHO list were lower (~19%) while the percentage of Class 3 drugs was higher than those on the US, GB, ES and IL lists (~32%), suggesting that this list is inherently different then the countries lists, as it contains drugs that are commonly used in developing but not industrial countries. The low percentage of Class 4 drugs on all lists indicates that the unfavorable biopharmaceutical characteristics of Class 4 compounds severely hamper their development as therapeutic agents.

25 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

BCS and BDDCS Wu and Benet 62 have noticed that the high-permeability characteristics of BCS class 1 and 2 drugs allow ready access to metabolizing enzymes within hepatocytes and suggested that there is a good correlation between the extent of drug metabolism and the permeability as defined in the BCS. This is the BDDCS, claiming that if the major route of elimination of a given drug is metabolism, then the drug is high-permeable and if the major route of elimination is renal and biliary excretion of unchanged drug, then that drug should be classified as low-permeability 23. The cutoff was originally set at ≥50% metabolism but later changed to 70% or 90% of an oral dose in human. Additional implications of the BDDCS, e.g., food effect and significance of transporter/enzyme interplay in drug interactions, were suggested as well 24, 63

. The key questions, to what extent metabolism can be used as a

surrogate for intestinal permeability and under what circumstances drug metabolism may not be viable for permeability predictions, were investigated. Takagi et al 17 compared the BCS and BDDCS classification of 168 drugs, and revealed excellent agreement between BDDCS and BCS for class 2 and 4 drugs, but not for class 1 and 3. They have shown that the differences could be reduced depending on the choice of permeability (fraction absorbed) or percent metabolism borderline for the high/low classification. More recently, Chen and Yu 64 evaluated the extent of metabolism of 51 high-permeability drugs. By using a cutoff of 50% metabolism, 37 drugs (73%) were classified

26 ACS Paragon Plus Environment

Page 26 of 54

Page 27 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

as extensively metabolized, pointing out that high permeability as defined by the BCS does not necessarily dictate extensive metabolism. In accordance with the observations of Takagi et al 17, Figure 6 illustrates that more drugs are classified as Class 1 according to the BDDCS 58

vs the current BCS classification. Further analysis of the data reveals that

these differences were mainly between Class 1 and Class 3; form the total of 106 BDDCS Class 1 drugs, 7 were classified as BCS Class 2 drugs, 18 BCS Class 3 drugs and 8 BCS Class 4 drugs. This trend is likely due to the more lax benchmark for metabolism (70%) 58, in comparison to the strict benchmark (metoprolol) we used for the permeability classification. Examination of the human intestinal absorption of metoprolol reveals that it is in fact completely (100%) absorbed, as evidenced by radiolabel mass-balance studies in five human subjects 31, thus making metoprolol an overly conservative reference drug for the low/high permeability class boundary. Indeed, the angiotensin II blocker losartan has been shown to have a lower human permeability than that of metoprolol yet a high fraction dose absorbed based on mass-balance. Likewise, isotretinoin was reported to have 90% fraction dose absorbed yet human permeability lower than that of metoprolol 3. Similarly, when Takagi et al used alternative reference drugs as the permeability benchmarks, such as cimetidine, ranitidine, or atenolol, the agreement between BDDCS and BCS was greatly improved 17. The original choice of metoprolol as the reference drug was intentionally conservative in the initial guidance, to minimize the chances for exceptions to the dissolution standard that was going to be recommended for biowaivers.

27 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

While classification based solely upon metabolism may fail to correctly classify drugs that are highly absorbed but are excreted unchanged into urine and bile (e.g., amoxicillin, trimethoprim, lomefloxacin, zalcitabine, and chloroquine), lipophilicity considerations alone would not be able to predict active carrier-mediated transport of drugs. Despite these differences, the two approaches indicate that granting a waiver from in-vivo BE studies is justified for the majority of drugs 2, 5, 8.

In-silico solubility classification Figure 7 clearly demonstrates that in-silico methods may be a reliable alternative to estimate the biopharmaceutical properties of NMEs, even before synthesis, by employing the thermodynamic equation developed by Amidon and Williams 22. The overall similarity of the solubility classification using insilico solubility estimates and various in-silico partition coefficients with that obtained with reference literature solubility (Figure 7) suggests that in-silico octanol-water partition coefficients can be useful in the in-silico calculation of aqueous solubility. Solubility BCS classification depends on the maximal dose strength as well as the intrinsic solubility of the drug. This puts a theoretical constraint on in-silico based BCS classification of NMEs, as the highest dose strength of a drug can only be determined in the clinical stages of R&D. However, the recent work of Pham-The et al 21 and Broccatelli et al 24 suggests that the dose has a smaller impact on the dose number when compared to the intrinsic solubility. Thus, computerized models may assign a probabilistic 28 ACS Paragon Plus Environment

Page 28 of 54

Page 29 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

solubility classification for a drug based on a range of hypothetical doses with no major loss of accuracy. Another obstacle to complete in-silico prediction of solubility for NMEs is the need for melting point value that can only be determined by empirical experiments. Therefore, we examined whether the average melting point may serve as a surrogate to experimentally determined melting point values. the in-silico solubility classifications using average melting point and CLogP (BioLoom 5.0) or KLogP, were remarkably similar to that obtained using experimental melting points. Taken together, these results suggest that the solubility and permeability classifications of NMEs can be successfully predicted from just molecular formulas and dose assumptions and average melting point.

29 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

CONCLUSIONS

In conclusion, relatively simplified in-silico methods can reliably predict permeability, solubility, and BCS classification. The most simplified in-silico method (KLogP) that is based on molecular formula with no knowledge of molecular structure, generally produced comparable BCS classification to the more sophisticated methods. This KLogP, when combined with a mean melting point and estimated dose, can be used to provisionally classify potential drugs from just molecular formula, even before synthesis. This may become extremely valuable in the early stages of drug discovery and development, when very little (or no) amount of the API is available. We have also demonstrated that more than half of the world's topselling drugs are highly soluble, and are therefore candidates for waivers of in-vivo BE studies. For these drugs, the replacement of expensive human studies with affordable in-vitro dissolution tests would ensure their bioequivalence, and encourage the development and availability of generic drug products in industrial, and even more important, in developing countries.

Acknowledgements The manuscript represents the personal opinions of the authors and does not necessarily represent the views or policies of the Korea Food and Drug Administration.

30 ACS Paragon Plus Environment

Page 30 of 54

Page 31 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

REFERENCES

1.

Amidon, G. L.; Lennernas, H.; Shah, V. P.; Crison, J. R. A theoretical

basis for a biopharmaceutic drug classification: the correlation of in vitro drug product dissolution and in vivo bioavailability. Pharm Res 1995, 12, (3), 41320. 2.

Amidon, K. S.; Langguth, P.; Lennernas, H.; Yu, L.; Amidon, G. L.

Bioequivalence of oral products and the biopharmaceutics classification system: Science, regulation, and public policy. Clinical Pharmacology & Therapeutics 2011, 90, (3), 467-470. 3.

Dahan, A.; Lennernäs, H.; Amidon, G. L. The fraction dose absorbed,

in uumans, and high jejunal human permeability relationship. Molecular Pharmaceutics 2012, 9, (6), 1847-1851. 4.

Lennernäs, H.; Abrahamsson, B. The use of biopharmaceutic

classification of drugs in drug discovery and development: current status and future extension. Journal of Pharmacy and Pharmacology 2005, 57, (3), 273285. 5.

Dahan, A.; Miller, J. M.; Amidon, G. L. Prediction of solubility and

permeability class membership: provisional BCS classification of the world's top oral drugs. AAPS J 2009, 11, (4), 740-6. 6.

Cook, J. A.; Davit, B. M.; Polli, J. E. Impact of biopharmaceutics

classification system-based biowaivers. Molecular Pharmaceutics 2010, 7, (5), 1539-1544.

31 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

7.

Polli, J. In vitro studies are sometimes better than conventional human

pharmacokinetic in vivo studies in assessing bioequivalence of immediaterelease solid oral dosage forms. AAPS J 2008, 10, (2), 289-299. 8.

Chen, M.-L.; Amidon, G.; Benet, L.; Lennernas, H.; Yu, L. The BCS,

BDDCS, and regulatory guidances. Pharm Res 2011, 28, (7), 1774-1778. 9.

Dahan, A.; Miller, J. M.; Hilfinger, J. M.; Yamashita, S.; Yu, L. X.;

Lennerna s, H.; Amidon, G. L. High-permeability criterion for BCS classification: Segmental/pH dependent permeability considerations. Molecular Pharmaceutics 2010, 7, (5), 1827-1834. 10.

Löbenberg, R.; Amidon, G. L. Modern bioavailability, bioequivalence

and biopharmaceutics classification system. New scientific approaches to international regulatory standards. European Journal of Pharmaceutics and Biopharmaceutics 2000, 50, (1), 3-12. 11.

Martinez, M. N.; Amidon, G. L. A mechanistic approach to

understanding the factors affecting drug absorption: A review of fundamentals. The Journal of Clinical Pharmacology 2002, 42, (6), 620-643. 12.

Yu, L.; Amidon, G.; Polli, J.; Zhao, H.; Mehta, M.; Conner, D.; Shah, V.;

Lesko, L.; Chen, M.-L.; Lee, V. L.; Hussain, A. Biopharmaceutics classification system: The scientific basis for biowaiver extensions. Pharm Res 2002, 19, (7), 921-925. 13.

Cook, J.; Addicks, W.; Wu, Y. Application of the biopharmaceutical

classification system in clinical drug development - An industrial view. AAPS J 2008, 10, (2), 306-310. 14.

Ku, M. S. Use of the biopharmaceutical classification system in early

drug development. AAPS J 2008, 10, (1), 208-212.

32 ACS Paragon Plus Environment

Page 32 of 54

Page 33 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

15.

van de Waterbeemd, H.; Gifford, E. ADMET in silico modelling:

towards prediction paradise? Nature Reviews Drug Discovery 2003, 2, (3), 192-204. 16.

Kasim, N. A.; Whitehouse, M.; Ramachandran, C.; Bermejo, M.;

Lennernas, H.; Hussain, A. S.; Junginger, H. E.; Stavchansky, S. A.; Midha, K. K.; Shah, V. P.; Amidon, G. L. Molecular properties of WHO essential drugs and provisional biopharmaceutical classification. Molecular Pharmaceutics 2004, 1, (1), 85-96. 17.

Takagi, T.; Ramachandran, C.; Bermejo, M.; Yamashita, S.; Yu, L. X.;

Amidon, G. L. A provisional biopharmaceutical classification of the top 200 oral drug products in the United States, Great Britain, Spain, and Japan. Molecular Pharmaceutics 2006, 3, (6), 631-43. 18.

Fong, S. Y. K.; Liu, M.; Wei, H.; Löbenberg, R.; Kanfer, I.; Lee, V. H. L.;

Amidon, G. L.; Zuo, Z. Establishing the pharmaceutical quality of Chinese herbal medicine: A provisional BCS classification. Molecular Pharmaceutics 2013, 10, (5), 1623-1643. 19.

Waldmann, S.; Almukainzi, M.; Bou-Chacra, N. A.; Amidon, G. L.; Lee,

B.-J.; Feng, J.; Kanfer, I.; Zuo, J. Z.; Wei, H.; Bolger, M. B.; Löbenberg, R. Provisional biopharmaceutical classification of some common herbs used in western medicine. Molecular Pharmaceutics 2012, 9, (4), 815-822. 20.

Lindenberg, M.; Kopp, S.; Dressman, J. B. Classification of orally

administered drugs on the World Health Organization Model list of Essential Medicines according to the biopharmaceutics classification system. European Journal of Pharmaceutics and Biopharmaceutics 2004, 58, (2), 265-78.

33 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

21.

Pham-The, H.; Garrigues, T.; Bermejo, M.; González-Álvarez, I.;

Monteagudo, M. C.; Cabrera-Pérez, M. Á. Provisional classification and in silico study of biopharmaceutical system based on caco-2 cell permeability and dose number. Molecular Pharmaceutics 2013, 10, (6), 2445-2461. 22.

Amidon, G. L.; Williams, N. A. A solubility equation for non-electrolytes

in water. International Journal of Pharmaceutics 1982, 11, 249-256. 23.

Benet, L.; Amidon, G.; Barends, D.; Lennernäs, H.; Polli, J.; Shah, V.;

Stavchansky, S.; Yu, L. The use of BDDCS in classifying the permeability of marketed drugs. Pharm Res 2008, 25, (3), 483-488. 24.

Broccatelli, F.; Cruciani, G.; Benet, L. Z.; Oprea, T. I. BDDCS class

prediction for new molecular entities. Molecular Pharmaceutics 2012, 9, (3), 570-80. 25.

Leo, A. J. Calculating log Poct from Structures. Chemical Reviews

1993, 93, (4), 1281-1306. 26.

Ghose, A. K.; Crippen, G. M. Atomic physicochemical parameters for

three-dimensional-structure-directed quantitative structure-activity relationships. 2. Modeling dispersive and hydrophobic interactions. Journal of Chemical Information and Computer Sciences 1987, 27, (1), 21-35. 27.

Scott, A. W.; Crippen, G. M. Prediction of physicochemical parameters

by atomic contributions. Journal of Chemical Information and Modeling 1999, 39, 868-873. 28.

Hansch, C.; Leo, A. J., Substituent Constants for Correlation Analysis

in Chemistry and Biology. ed.; Willey: New York, 1979; 'Vol.' p.

34 ACS Paragon Plus Environment

Page 34 of 54

Page 35 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

29.

Fairstein, M.; Swissa, R.; Dahan, A. Regional-dependent intestinal

permeability and BCS classification: Elucidation of pH-related complexity in rats using pseudoephedrine. AAPS J 2013, 15, (2), 589-597. 30.

Lennernas, H. Human intestinal permeability. Journal of

Pharmaceutical Sciences 1998, 87, (4), 403-10. 31.

Regardh, C.; Borg, K.; Johansson, R.; Johnsson, G.; Palmer, L.

Pharmacokinetic studies on the selective beta1-receptor antagonist metoprolol in man. Journal of Pharmacokinetics and Biopharmaceutics 1974, 2, (4), 347-64. 32.

Oh, D. M.; Curl, R. L.; Amidon, G. L. Estimating the fraction dose

absorbed from suspensions of poorly soluble compounds in humans: a mathematical model. Pharm Res 1993, 10, (2), 264-70. 33.

Lennernas, H. Intestinal permeability and its relevance for absorption

and elimination. Xenobiotica 2007, 37, (10-11), 1015-51. 34.

Lennernäs, H. Animal data: The contributions of the Ussing chamber

and perfusion systems to predicting human oral drug delivery in vivo. Advanced Drug Delivery Reviews 2007, 59, (11), 1103-1120. 35.

Lennernas, H. Human jejunal effective permeability and its correlation

with preclinical drug absorption models. Journal of Pharmay and Pharmacology 1997, 49, (7), 627-38. 36.

Chu, X.-Y.; Sánchez-Castaño, G. P.; Higaki, K.; Oh, D.-M.; Hsu, C.-P.;

Amidon, G. L. Correlation between epithelial cell permeability of cephalexin and expression of intestinal oligopeptide transporter. Journal of Pharmacology and Experimental Therapeutics 2001, 299, (2), 575-582.

35 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

37.

Ganapathy, M. E.; Brandsch, M.; Prasad, P. D.; Ganapathy, V.;

Leibach, F. H. Differential recognition of beta -lactam antibiotics by intestinal and renal peptide transporters, PEPT 1 and PEPT 2. Journal of Biological Chemistry 1995, 270, (43), 25672-7. 38.

Han, H. K.; Rhie, J. K.; Oh, D. M.; Saito, G.; Hsu, C. P.; Stewart, B. H.;

Amidon, G. L. CHO/hPEPT1 cells overexpressing the human peptide transporter (hPEPT1) as an alternative in vitro model for peptidomimetic drugs. Journal of Pharmaceutical Sciences 1999, 88, (3), 347-50. 39.

Yuasa, H.; Fleisher, D.; Amidon, G. L. Noncompetitive inhibition of

cephradine uptake by enalapril in rabbit intestinal brush-border membrane vesicles: an enalapril specific inhibitory binding site on the peptide carrier. Journal of Pharmacology and Experimental Therapeutics 1994, 269, (3), 1107-1111. 40.

Kanai, Y.; Endou, H. Functional properties of multispecific amino acid

transporters and their implications to transporter-mediated toxicity. Journal of Toxicological Sciences 2003, 28, (1), 1-17. 41.

Tsuji, A.; Tamai, I.; Nakanishi, M.; Amidon, G. Mechanism of

absorption of the dipeptide alpha-methyldopa-phe in intestinal brush-border membrane vesicles. Pharm Res 1990, 7, (3), 308-9. 42.

Friedman, D.; Amidon, G. Oral absorption of peptides: influence of pH

and inhibitors on the intestinal hydrolysis of leu-enkephalin and analogues. Pharm Res 1991, 8, (1), 93-96. 43.

Amidon, G.; Chang, M.; Fleisher, D.; Allen, R. Intestinal absorption of

amino acid derivatives: importance of the free alpha-amino group. Journal of Pharmaceutical Sciences 1982, 71, (10), 1138-41.

36 ACS Paragon Plus Environment

Page 36 of 54

Page 37 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

44.

Dahan, A.; Khamis, M.; Agbaria, R.; Karaman, R. Targeted prodrugs in

oral drug delivery: the modern molecular biopharmaceutical approach. Expert Opinion on Drug Delivery 2012, 9, (8), 1001-13. 45.

Han, H. K.; Oh, D. M.; Amidon, G. L. Cellular uptake mechanism of

amino acid ester prodrugs in Caco-2/hPEPT1 cells overexpressing a human peptide transporter. Pharm Res 1998, 15, (9), 1382-6. 46.

Landowski, C. P.; Sun, D.; Foster, D. R.; Menon, S. S.; Barnett, J. L.;

Welage, L. S.; Ramachandran, C.; Amidon, G. L. Gene expression in the human intestine and correlation with oral valacyclovir pharmacokinetic parameters. Journal of Pharmacology and Experimental Therapeutics 2003, 306, (2), 778-786. 47.

Sun, J.; Dahan, A.; Amidon, G. L. Enhancing the intestinal absorption

of molecules containing the polar guanidino functionality: A double-targeted prodrug approach. Journal of Medicinal Chemistry 2010, 53, (2), 624-632. 48.

Soldner, A.; Benet, L. Z.; Mutschler, E.; Christians, U. Active transport

of the angiotensin-II antagonist losartan and its main metabolite EXP 3174 across MDCK-MDR1 and caco-2 cell monolayers. British journal of pharmacology 2000, 129, (6), 1235-43. 49.

Weiss, J.; Sauer, A.; Divac, N.; Herzog, M.; Schwedhelm, E.; Böger, R.

H.; Haefeli, W. E.; Benndorf, R. A. Interaction of angiotensin receptor type 1 blockers with ATP-binding cassette transporters. Biopharmaceutics & Drug Disposition 2010, 31, (2-3), 150-161. 50.

Flanagan, S. D.; Benet, L. Z. Net secretion of furosemide is subject to

indomethacin inhibition, as observed in Caco-2 monolayers and excised rat jejunum. Pharm Res 1999, 16, (2), 221-24.

37 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

51.

Flanagan, S. D.; Cummins, C. L.; Susanto, M.; Liu, X.; Takahashi, L.

H.; Benet, L. Z. Comparison of furosemide and vinblastine secretion from cell lines overexpressing multidrug resistance protein (P-glycoprotein) and multidrug resistance-associated proteins (MRP1 and MRP2). Pharmacology 2002, 64, (3), 126-34. 52.

Kim, J.-S.; Mitchell, S.; Kijek, P.; Tsume, Y.; Hilfinger, J.; Amidon, G. L.

The suitability of an in situ perfusion model for permeability determinations:  Utility for BCS class I biowaiver requests. Molecular Pharmaceutics 2006, 3, (6), 686-694. 53.

Collett, A.; Higgs, N. B.; Sims, E.; Rowland, M.; Warhurst, G.

Modulation of the permeability of H2 receptor antagonists cimetidine and ranitidine by P-glycoprotein in rat intestine and the human colonic cell line Caco-2. Journal of Pharmacology and Experimental Therapeutics 1999, 288, (1), 171-8. 54.

Dahan, A.; Amidon, G. L. Segmental dependent transport of low

permeability compounds along the small intestine due to P-glycoprotein: The role of efflux transport in the oral absorption of BCS class III drugs. Molecular Pharmaceutics 2009, 6, (1), 19-28. 55.

Dahan, A.; West, B. T.; Amidon, G. L. Segmental-dependent

membrane permeability along the intestine following oral drug administration: Evaluation of a triple single-pass intestinal perfusion (TSPIP) approach in the rat. European Journal of Pharmaceutical Sciences 2009, 36, (2–3), 320-329. 56.

Bourdet, D.; Pollack, G.; Thakker, D. Intestinal absorptive transport of

the hydrophilic cation ranitidine: a kinetic modeling approach to elucidate the

38 ACS Paragon Plus Environment

Page 38 of 54

Page 39 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

role of uptake and efflux transporters and paracellular vs. transcellular transport in Caco-2 cells. Pharm Res 2006, 23, (6), 1178-1187. 57.

Lentz, K.; Polli, J.; Wring, S.; Humphreys, J.; Polli, J. Influence of

passive permeability on apparent P-glycoprotein kinetics. Pharm Res 2000, 17, (12), 1456-60. 58.

Benet, L. Z.; Broccatelli, F.; Oprea, T. I. BDDCS applied to over 900

drugs. AAPS J 2011, 13, (4), 519-47. 59.

Giacomini, K. M.; Huang, S. M.; Tweedie, D. J.; Benet, L. Z.; Brouwer,

K. L.; Chu, X.; Dahlin, A.; Evers, R.; Fischer, V.; Hillgren, K. M.; Hoffmaster, K. A.; Ishikawa, T.; Keppler, D.; Kim, R. B.; Lee, C. A.; Niemi, M.; Polli, J. W.; Sugiyama, Y.; Swaan, P. W.; Ware, J. A.; Wright, S. H.; Yee, S. W.; ZamekGliszczynski, M. J.; Zhang, L. Membrane transporters in drug development. Nature Reviews Drug Discovery 2010, 9, (3), 215-36. 60.

Neuhoff, S.; Ungell, A.-L.; Zamora, I.; Artursson, P. pH-Dependent

passive and active transport of acidic drugs across Caco-2 cell monolayers. European Journal of Pharmaceutical Sciences 2005, 25, (2–3), 211-220. 61.

Varma, M. V.; Gardner, I.; Steyn, S. J.; Nkansah, P.; Rotter, C. J.;

Whitney-Pickett, C.; Zhang, H.; Di, L.; Cram, M.; Fenner, K. S.; El-Kattan, A. F. pH-dependent solubility and permeability criteria for provisional biopharmaceutics classification (BCS and BDDCS) in early drug discovery. Molecular Pharmaceutics 2012, 9, (5), 1199-1212. 62.

Wu, C. Y.; Benet, L. Z. Predicting drug disposition via application of

BCS: transport/absorption/ elimination interplay and development of a biopharmaceutics drug disposition classification system. Pharm Res 2005, 22, (1), 11-23.

39 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

63.

Custodio, J. M.; Wu, C.-Y.; Benet, L. Z. Predicting drug disposition,

absorption/elimination/transporter interplay and the role of food on drug absorption. Advanced Drug Delivery Reviews 2008, 60, (6), 717-733. 64.

Chen, M.-L.; Yu, L. The use of drug metabolism for prediction of

intestinal permeability. Molecular Pharmaceutics 2009, 6, (1), 74-81.

40 ACS Paragon Plus Environment

Page 40 of 54

Page 41 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

Legend for Figures

Figure 1

Correlation plots of in-silico vs experimental partition

coefficients. (a) CLogP (BioLoom 5.0); (b) CLogP (ChemDraw 8.0); (c) ALogP (MOE 2004.03); and (d) KlogP (Molecular Formula). Figure 2

Correlation plots of human jejunal permeability with partition and

distribution coefficients values for 29 reference drugs. (a) CLogP (BioLoom 5.0); (b) CLogP (ChemDraw 8.0); (c) ALogP (MOE 2004.03); (d) KLogP (Molecular Formula); (e) LogD (pH 6.5); and (f) LogD (pH 7.5). Red dashed lines, the Peff value and the relevant coefficient value of the reference drug metoprolol. Yellow point, metoprolol; Blue points, correctly classified; Red points, false negatives; Green points, false positives. Figure 3

Comparison of the permeability classification of oral IR drug

products on the US, GB, ES, IL, JP, KR, and WHO lists using various in-silico partition coefficient approaches. Figure 4

Comparison of the solubility classification of oral IR drug

products on the US, GB, ES, IL, JP, KR, and WHO lists using reference literature solubility values. Figure 5

Comparison of the provisional BCS classification of oral IR drug

productss on the US, GB, ES, IL, JP, KR, and WHO lists using solubility classification based on reference literature solubility and permeability classification based on the various in-silico partition coefficient approaches.

41 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 6

Comparison of the classification of 257 drugs according to the

BCS using the four different in-silico permeability methods, and literature BDDCS classification. Figure 7

Comparison of the solubility classification of 185 drugs using in-

silico estimations with the different in-silico approaches, and reference literature solubility values. Figure 8

Comparison of the provisional BCS classification of 185 drugs

using solubility classification based on in-silico estimations or reference literature solubility values, and permeability classification using the various insilico partition coefficient approaches.

42 ACS Paragon Plus Environment

Page 42 of 54

Page 43 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

Molecular Pharmaceutics

Table 1

Human jejunal permeability and in-silico parameters for 29 reference drugs Permeability Classification based on: Drug

pKa

CLogP

CLogP

ALogP

KLog P

LogD

LogD

Human

CLogP

ALogP

KLog P

LogD

LogD

(BioLoom

(BioLoom

(ChemDraw

(MOE

(Molecular

(pH 6.5)

(pH 7.4)

permeability

(BioLoom

(MOE

(Molecular

(pH

(pH

4

5.0)

5.0)

8.0)

2004.03)

Formula)

(x10 cm/s)

5.0)

2004.03)

Formula)

6.5)

7.4)

α-methyldopa

9.12

-2.26

-2.26

0.44

0.26

-4.88

-3.99

0.1

C

C

C

C

C

amoxicillin

2.64

-1.87

-1.87

0.02

0.67

-5.73

-6.63

0.3

C

C

C

C

C

antipyrine

1.44

0.2

0.2

1.78

1.55

0.2

0.2

5.6

Fn

C

Fn

C

C

atenolol

9.6

-0.11

-0.11

0.45

1.33

-3.21

-2.31

0.2

C

C

C

C

C

carbamazepine

NI

2.38

2.38

3.39

2.54

2.38

2.38

4.3

C

C

C

C

C

cephalexin

2.67

-1.84

-1.84

0.44

1.13

-5.67

-6.57

1.56

Fn

Fn

Fn

Fn

Fn

cimetidine

6.9

0.38

0.35

0.6

0.51

-0.17

0.26

0.26

C

C

C

Fp

Fp

creatinine

4.8

-1.76

-1.77

-1.01

-0.46

-3.47

-4.36

0.29

C

C

C

C

C

desipramine

10.65

4.47

4.47

3.53

3.67

0.32

1.22

4.5

C

C

C

C

C

enalapril

3.04

0.67

0.67

1.6

1.87

-2.79

-3.69

1.57

Fn

C

C

Fn

Fn

enalaprilat

2.2

0.75

0.88

1.13

1.4

-3.55

-4.45

0.2

C

C

C

C

C

fluvastatin

4.39

4.05

4.05

4.63

3.53

4.05

4.05

2.4

C

C

C

C

C

furosemide

3.34

1.9

1.9

1.89

0.7

-1.26

-2.16

0.05

Fp

Fp

C

Fp

C

Hydrochlorothiazide

7.9

-0.36

-0.37

-0.35

-0.34

-1.78

-0.98

0.04

C

C

C

C

C

inogatran

1.3

-0.14

-

0.37

1.24

-5.34

-6.24

0.03

C

C

C

C

C

ketoprofen

4.5

2.76

2.76

3.11

2.5

0.76

-0.14

8.7

C

C

C

C

C

43

ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

Page 44 of 54

levodopa

2.32

-2.82

-2.82

0.05

0.03

-7

-7.9

3.4

Fn

Fn

Fn

Fn

Fn

l-leucine

2.33

-1.67

-1.67

0.44

0.18

-5.84

-6.74

6.2

Fn

Fn

Fn

Fn

Fn

lisinopril

1.68

-1.82

-1.69

1.24

1.78

-6.64

-7.54

0.33

C

C

C

C

C

losartan

4.7

3.85

4.11

4.27

3.59

3.84

3.85

1.15

Fp

Fp

Fp

Fp

Fp

metoprolol

9.56

1.49

1.49

1.61

1.87

-1.57

-0.67

1.34

ref

ref

ref

ref

ref

naproxen

4.15

2.82

2.82

3.04

2.01

0.47

-0.43

8.5

C

C

C

C

C

phenylalanine

1.83

-1.56

-1.56

0.64

0.93

-6.23

-7.13

4.08

Fn

Fn

Fn

Fn

Fn

piroxicam

5.1

1.89

1.89

1.58

0.91

0.47

-0.41

6.65

C

Fn

Fn

C

C

propranolol

9.47

2.75

2.75

2.58

2.6

-0.22

0.68

2.91

C

C

C

C

C

ranitidine

8.45

0.67

0.63

1.05

0.49

-1.28

-0.42

0.27

C

C

C

Fp

Fp

terbutaline

8.72

0.48

0.48

1.52

1.17

-1.74

-0.86

0.3

C

C

C

C

C

valacyclovir

1.9

-1.22

-1.22

-0.93

-0.62

-1.22

-1.22

1.66

Fn

Fn

Fn

C

Fn

verapamil

8.73

4.47

4.47

5.09

3.98

2.24

3.12

6.8

C

C

C

C

C

C, correctly classified; Fn, false negative; Fp- false positive.

44

ACS Paragon Plus Environment

Page 45 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

Table 2

Possible reasons for misclassification of permeability of reference drugs

Drug

Evidence for Transport

False negatives antipyrine

no evidence of carrier-mediated transport

cephalexin

substrate for peptide transporters

enalapril

substrate for peptide transporters

levodopa

substrate for amino acid transporters

L-leucine

substrate for amino acid transporters

phenylalanine

substrate for amino acid transporters

piroxicam

no evidence for carrier-mediated transport

valacyclovir

substrate for peptide transporters

False positives cimetidine

secretion by P-glycoprotein

furosemide

secretion in rat jejunum and Caco-2 cells

losartan

secretion by P-glycoprotein

ranitidine

secretion by P-glycoprotein

45 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

190x142mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 46 of 54

Page 47 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

190x142mm (300 x 300 DPI)

ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

300x209mm (260 x 260 DPI)

ACS Paragon Plus Environment

Page 48 of 54

Page 49 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

300x209mm (260 x 260 DPI)

ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

190x142mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 50 of 54

Page 51 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

300x209mm (260 x 260 DPI)

ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

190x142mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 52 of 54

Page 53 of 54

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

190x142mm (300 x 300 DPI)

ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

190x142mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 54 of 54