- Email: [email protected]
In vitro dissolution absorption system (IDAS2): Use for the prediction of food viscosity effects on drug dissolution and absorption from oral solid dosage forms
European Journal of Pharmaceutical Sciences 143 (2020) 105164
Contents lists available at ScienceDirect
European Journal of Pharmaceutical Sciences journal homepage: www.elsevier.com/locate/ejps
In vitro dissolution absorption system (IDAS2): Use for the prediction of food viscosity effects on drug dissolution and absorption from oral solid dosage forms
T
⁎
Svitlana Silchenko , Nourdine Nessah, Jibin Li, Li-Bin Li, Yuehua Huang, Albert J. Owen, Ismael J. Hidalgo Absorption Systems LLC, Exton, PA 19341, USA
A R T I C LE I N FO
A B S T R A C T
Keywords: viscosity of medium oral solid dosage products dissolution permeation IDAS2 Caco2 cell monolayer
Existing in vitro dissolution or permeation models to predict food effect are mainly based on Pharmacopeias’ compendial media, which specify such variables as pH, bile salts, lipolytic enzymes, and phospholipids content. However, the viscosity of food in the gastrointestinal (GI) tract is not taken into account, although it can affect both the dissolution of the oral solid dosage form and absorption of the released drug. Here, a new in vitro dissolution absorption system (IDAS2) is utilized, which comprises a dissolution apparatus USP2 (DISTEK) equipped with specially constructed permeability chambers containing Caco-2 monolayers, thereby allowing dissolution and transepithelial absorption to be ascertained simultaneously. The IDAS2 was used to evaluate the effect of medium viscosity on both the dissolution of oral solid dosage forms and absorption of released drugs. Such information, which is not ordinarily determined in dissolution and permeation studies, will be helpful to the formulators developing robust oral dosage forms. Commercially available solid dosage forms of ten model drugs from across all BCS classifications were used in this evaluation: metoprolol, minoxidil, and propranolol from BCS class 1; carbamazepine, ketoprofen, and simvastatin from BCS class 2; atenolol and ranitidine from BCS class 3; and acetazolamide and saquinavir from BCS class 4. The study revealed the applicability of IDAS2 as a tool for in vitro screening of dissolution and absorption of intact oral solid products to predict food viscosity effect. The most profound viscosity effect on dissolution and absorption was observed of solid dosage forms for the BCS class 2 compounds carbamazepine and simvastatin. A higher medium viscosity significantly slowed down the dissolution rate of tested BSC class 4 compounds acetazolamide and saquinavir, without significant effect on their absorption. The solid dosage forms least affected by the viscosity of the medium tested were the BCS class 1 compounds minoxidil and propranolol.
1. Introduction The effect of food on rate and extent of drug absorption is included as part of bioavailability and bioequivalence studies in the FDA Guidance for Industry (U.S. FDA Guidance for Industry, 2002). In vivo, food inhibits gastric emptying and modifies intestinal transit time (Winstanley and Orme, 1989). The GI pH value, bile acid concentrations and fluid composition are different in the fasted versus the fed state (Lentz, 2008; Perez de la Cruz Moreno et al., 2006; Riethorst et al., 2016). These factors can affect the wettability, disintegration, and dissolution of oral solid dosage forms, as well as diffusion and absorption of the released drug. Several existing in vitro dissolution/absorption models for solid oral dosage forms have been used to predict in
⁎
vivo drug absorption for the fasted state (Kataoka et al., 2011; Kourentas et al., 2016; Mudie et al., 2010). The modified Noyes-Whitney equation describes the drug dissolution process, i.e. the relationship of the mass of drug dissolved versus time and the variables which have an impact on dissolution. In this equation, the rate of dissolution is dependent on the drug concentration in the boundary layer (solubility), drug bulk concentration, drug diffusivity (diffusion coefficient), and empirical thickness of the hydrodynamic boundary layer surrounding the solid dosage form. Dissolution is also affected by intrinsic properties of the drug or of the solid dosage form, as well as the physiological environment (in vivo) or dissolution medium (in vitro). The in vivo GI fluid composition, changes along the GI tract, as well as with food consumption and digestion. These changes
Corresponding author at: Department of Physical Chemistry and Formulation, Absorption Systems, LLC, 436 Creamery Way, Suite 600, Exton, PA 19341, USA. E-mail address: [email protected] (S. Silchenko).
https://doi.org/10.1016/j.ejps.2019.105164 Received 17 July 2019; Received in revised form 14 November 2019; Accepted 20 November 2019 Available online 21 November 2019 0928-0987/ © 2019 Published by Elsevier B.V.
European Journal of Pharmaceutical Sciences 143 (2020) 105164
S. Silchenko, et al.
Fig. 1. Schematic illustration of in vitro dissolution absorption systems IDAS1 (A) and IDAS2 (B).
because the passive diffusion, para-cellular, and active drug transporters are involved in the drug permeation in parallel (Berben et al., 2018). Correctly adjusted biorelevant medium (Yamashita et al., 2000) has shown good in vitro-in vivo correlation under fed and fasted conditions (Kataoka et al., 2011; Markopoulos et al., 2014). Despite mimicking various in vivo intestinal conditions (e.g. medium pH, bile acid concentration) on the apical side of the cells, Caco-2 cell monolayers might be sensitive to food components and excipients. The first integrated dissolution/Caco-2 permeation system was reported by Ginski et al. (1999). The integrated system was tested on several commercial solid oral dosage forms and proved to be a valuable instrument for formulators in formulation development. Another in vitro tool with simultaneous dissolution/permeation (D/P) measurement across a Caco-2 cell monolayer permeability barrier was introduced by Kataoka et al. (2003). This D/P system was successfully used for predicting formulation effects on the oral absorption of drugs with low aqueous solubility (Kataoka et al., 2012) in evaluation of food effects and dose strength on drug dissolution and permeation with IVIVC projection (Buch et al., 2009; Kataoka et al., 2011, 2006). Absorption Systems LLC adapted this D/P system, and termed it “In-vitro Dissolution Absorption System 1″ (IDAS1). The IDAS1 tool using fasted state simulated small intestinal fluid (FaSSIF) as dissolution media along with conventional Ussing chambers with mounted Caco-2 cells using modified FaSSIFmod were found to accurately predict the formulation effect on fenofibrate in vivo performance (Forner et al., 2017). However, the existing in vitro models of translating dissolution/ permeation data into in vivo performance have limited applicability when it comes to pharmaceutical testing. For instance, the dissolution chamber in the IDAS1 is too small to accommodate an intact drug tablet. In contrast, more recently developed tools such as the in vitro dissolution absorption system 2 (IDAS2), where dissolution of intact solid oral product is integrated with the permeation of released drug through Caco-2 monolayer membrane, are more capable of implementing current formulation development trends. The IDAS2 was designed at Absorption Systems LLC (Li and Hidalgo, 2017) consists of a commercially available dissolution apparatus (USP1 or USP2), fitted with a specially designed lid with two attached permeability chambers (Fig. 1, B). The IDAS2 has successfully been tested on commercial solid oral dosage forms of differently sized indomethacin formulations (Li et al., 2019).
include differences in pH and differences in the concentrations of bile salts, lipase, and pepsin. Additionally, the viscosity of the fluid might change with food consumption as well, further affecting the drug dissolution and bioavailability (Reppas et al., 1998). To predict in vivo drug dissolution, various biorelevant media have been suggested for use with in vitro dissolution studies. The recent most comprehensive publication (Markopoulos et al., 2015) suggests four levels of biorelevant media simulating various gastrointestinal conditions for in-vitro dissolution studies. The publication mentions the importance of postprandial medium viscosity on drug in vitro dissolution. The viscosity of the medium itself may also have an impact on the wettability, disintegration, and dissolution of a drug product, as well as the diffusion of the released drug (Parojcic et al., 2008; Radwan et al., 2012). For example, a negative food effect on trospium chloride bioavailability has previously been observed and explained as an intrinsic property of the drug. However, it has been shown that a higher medium viscosity prolonged the disintegration time and decreased the dissolution rates of trospium chloride tablets, which might also have contributed to the observed negative food effect on trospium chloride bioavailability (Heinen et al., 2013). Even though the comprehensive media mentioned above have been useful in dissolution studies for predicting in-vitro/in-vivo correlation, formulators developing oral solid dosage forms are in need of more diverse methods and systems to predict drug release and absorption. To predict drug absorption using drug permeability screening, various cellfree permeation systems with two classes of barriers, biomimetic and dialysis membranes, have been used (Berben et al., 2018). Some advantages of such cell-free permeation tools is robustness across a broad pH range and stability toward solubilizing excipients used in advanced solubility-enhancing formulations of poorly soluble drugs (Flaten et al., 2006). The in vitro setup where dissolution testing is combined with a cell-free sink permeation bag has been used to investigate food effect on nanosized and microsized fenofibrate (Hens et al., 2015). The authors were able to demonstrate usefulness of the combined dissolution and permeation tool for prediction of food effect on drug absorption. However, a major disadvantage of cell-free permeation systems is that they are only applicable for predicting passive drug transport. The Caco-2 cell monolayer based models, routinely used as in vitro models for drug permeability screening in BCS classification studies (Amidon et al., 1995) allow for generation of biorelevant results 2
European Journal of Pharmaceutical Sciences 143 (2020) 105164
S. Silchenko, et al.
Transwell and Snapwell plates to grow cell monolayers for the permeability studies. The culture medium was changed every other day until use (20 to 28 days after seeding).
The purpose of the current study was to show the applicability of the IDAS2 to simultaneous dissolution and absorption screening of oral solid dosage during development, with regards to food viscosity effect in particular.
2.4. IDAS1 and IDAS2 2. Materials and methods Diagrams of both IDAS1 and IDAS2 are presented in Fig.1 (Part A and Part B, respectively). Both IDAS1 and IDAS2 are comprised of a dissolution chamber/vessel and one or two permeation chambers separated by human Caco-2 cell monolayers. Both systems also enable simultaneous measurements of drug dissolution and permeation in vitro. However, only IDAS2 permits evaluation of intact oral solid dosage forms (tablets or capsules).
2.1. Materials and dosage forms Propranolol, metoprolol, minoxidil, carbamazepine, simvastatin, ketoprofen, atenolol, ranitidine, acetazolamide, saquinavir mesylate, Dglucose, 2-(N-morpholino)ethanesulfonic acid (MES), and bovine serum albumin (BSA) were obtained from Sigma-Aldrich (St. Louis, MO). Hanks’ balanced salt solution (HBSS), Dulbecco's modified Eagle's medium (DMEM), Dulbecco's phosphate-buffered saline (DPBS), fetal bovine serum (FBS), penicillin-streptomycin mixture, non-essential amino acids (NEAA), sodium pyruvate, and trypsin were obtained from Life Technologies (Grand Island, NY). N-(2-Hydroxyethyl)piperazine-N(2-ethanesulfonic acid) (HEPES, 1 M solution) was purchased from Thermo Fisher Scientific (Waltham, MA). Costar Transwell® (6-well format, 4.67 cm2 insert area, 0.4 µm pore size) and Snapwell® plates (6well format, 1.13 cm2 insert area, 0.4 µm pore size) were purchased from Corning Life Sciences (Corning, NY). Hydroxypropyl methylcellulose (HPMC, Methocel K100 Premium LV) was kindly provided by DOW Chemical. All chemicals were stored properly at all times and used before their expiration dates. The commercial solid oral dosage forms used in the study are presented in Table 1.
2.4.1. Viscosity effect on permeation profile with IDAS1 In this first study, the effect of viscosity on active pharmaceutical ingredient (API) permeation was evaluated with IDAS1. Four test drugs were selected, one from each of the four BCS classes: propranolol (Class 1), carbamazepine (Class 2), atenolol (Class 3), and acetazolamide (Class 4)). The media, 8 ml of HBSSg buffer pH 6.5 with or without 1.4% HPMC and 5.5 ml of HBSSg buffer pH 7.4 with 4% BSA, were loaded into the donor side and receiver side, respectively, which were separated by a Caco-2 monolayer. The system was pre-warmed to 37°C, and each chamber was independently stirred at 200 rpm. Each API was pre-dissolved in DMSO and introduced to the mucosal side at time 0. The high permeability compounds propranolol and carbamazepine were dosed at a concentration of 10 µM, while the low permeability compounds atenolol and acetazolamide were dosed at a concentration of 100 µM. Stock solutions were prepared in DMSO at concentrations of 10 mM for propranolol and carbamazepine, 100 mM for acetazolamide, and 60 mM for atenolol, such that the final concentration of DMSO on the donor side was no more than 0.1%. Receiver sample aliquots of 1 ml were routinely taken from the receiver side at 5, 10, 20, and 30 min for highly permeable API (propranolol and carbamazepine) and at 30, 60, 90, and 120 min for API with low permeability (atenolol and acetazolamide). All samples were analyzed by LC–MS/MS.
2.2. Media preparation and viscosity measurement Buffer solution, HBSS supplemented with 10 mM D-glucose (HBSSg) at pH 6.5, was used on the donor side for both IDAS1 and IDAS2 to simulate the low viscosity of fasting condition stomach contents. To simulate the high viscosity fed condition, HPMC (Methocel K100 Premium LV) was dissolved in HBSSg buffer at 1.4% (w/v) concentration and used on the donor side. The medium viscosity was measured with a micro-viscometer (Model: µVISC, RheoSense, Inc.). The measured viscosity of 1.4% HPMC (Methocel K100 Premium LV) in HBSSg buffer was 110 mP.s. In all experiments, HBSSg buffer pH 7.4 containing 4.5% BSA (w/v) was used on the receiver side.
2.4.2. Viscosity effect on dissolution and permeation profile with IDAS2 For the second study, IDAS2 was used to evaluate the effect of viscosity on a wider range of compounds. Ten test drugs were selected from across all four BCS classes: metoprolol, minoxidil, and propranolol from BCS class 1; carbamazepine, ketoprofen, and simvastatin from BCS class 2; atenolol and ranitidine from BCS class 3; and acetazolamide and saquinavir from BCS class 4. The dissolution system vessels, DISTEK 2100B, were equipped with either rotating paddles for tablets or baskets for capsules. They were loaded with 500 ml of HBSSg buffer pH 6.5 (500 ml) for the low viscosity experiment, and HBSSg buffer pH 6.5 with 1.4% HPMC for the high viscosity experiment. Two permeation chambers (receiver side) were submerged in each vessel for both the low and high viscosity experiments, each containing 8 ml of HBSSg buffer pH 7.4 with 4.5% BSA, with Caco-2 monolayers separating the
2.3. Preparation of Caco-2 monolayers Caco-2 cells, clone C2bbe1, were maintained in DMEM containing 10% FBS, 1% NEAA, 4 mM L-glutamine, 1 mM sodium pyruvate, 100 IU/ml penicillin, and 100 µg/ml streptomycin in a humidified incubator (37 °C, 5% CO2). The culture medium was changed three times weekly, and cell growth was observed by light microscopy. When the stock cultures were approximately 80% confluent, the cells were harvested with trypsin-EDTA and seeded at a density of 60,000 cells/cm2 onto collagen-coated polycarbonate membrane filters (0.4 µm pore size) in Table 1 Commercial solid oral dosage forms tested. Drug Tablets
Strength (mg/unit)
MW
Manufacturer
NDC
Dose conc. (µg/ml)
Metoprolol tartrate Minoxidil Propranolol HCl Carbamazepine Ketoprofen (capsule) Simvastatin Atenolol Ranitidine HCl Acetazolamide Saquinavir
100 10 80 200 50 20 100 300 125 500
684.6 209.3 295.8 236.3 254.3 418.6 266.3 350.9 222.2 670.9
Mylan Pharmaceuticals Watson Pharm Mylan Pharmaceuticals Novartis Pharmaceuticals Teva Pharmaceuticals Lupin Pharmaceuticals IPR Pharmaceuticals GSK Taro Pharmaceuticals Genentech
0378-0047-01 0591-5643-01 0378-0185-01 0078-0509-05 009303193-01 68180-479-01 52427-431-90 0173-0393-40 51672-4022-1 0004-0244-51
400 40 320 800 200 80 400 1200 500 2000
3
European Journal of Pharmaceutical Sciences 143 (2020) 105164
S. Silchenko, et al.
dissolution and permeation chambers. The BCS classification of drug solubility is based on dose per administrated volume of 250 ml. To keep the dissolution part close to physiological conditions, dose/250 ml, the drug tablets or capsules (two per vessel) were dosed to the dissolution vessel (donor side) and the paddles/baskets rotated at 50 revolutions per minute (RPM). Atenolol was co-dosed with all drugs at 100 µM concentration, as a marker to monitor the Caco-2 cell monolayer integrity. Sample aliquots of 2 ml were withdrawn without replacement from the dissolution vessels at 5, 10, 15, 30, 60, 90, and 120 min. The samples were filtered through nylon syringe filters (0.45 µm pore size), discarding the first 1 ml of filtrate to saturate the filter membrane. Sample aliquots of 1 mL were withdrawn from the permeation chambers at 15, 30, 60, 90, and 120 min, and replaced with the same volume of blank media after each sampling. All samples were analyzed by either UV–Vis absorption or LC–MS/MS.
Table 3 Precursor-to-product transition of compounds in LC-MS/MS.
2.5. Sample analysis Donor samples were analyzed with CLARIOstar® microplate reader (BMG Labtech, Baden-Württemberg, Germany) UV/vis absorption. When necessary, samples were diluted with corresponding dissolution medium (HBSSg at pH 7.4), and the drug concentrations were determined by comparing measured absorbance against a standard curve obtained in the matching matrix. The wavelength details are provided in Table 2. For receiver samples, 100 µl of sample was spiked with 300 µl of 100% acetonitrile containing deuterated internal standard in a 96-well plate. The plate was shaken at 850 RPM for 5 min and then centrifuged at 3000 RPM for 10 min. After that, 100 µl of supernatant was transferred into a new 96-well plate, prefilled with 400 µl water in each well. The new mixture was shaken at 850 RPM for 5 min and then centrifuged at 3000 RPM for 10 min. Finally, 10 µl of supernatant was injected for LC–MS/MS analysis. Sample quantification was conducted by coupling a triple quadruple mass spectrometer (API4000, Applied Biosystem, Foster City, CA) to an Agilent 1260 system (Santa Clara, CA). The LC column was a Thermo Hypersil BDS C18 30 × 2.1 mm i.d., 3 µm, with guard column. (Thermo Fisher Scientific, Waltham, MA). Mobile phase buffer was 25 mM ammonium format pH 3.5. The aqueous mobile phase was 90% water and 10% buffer, and the organic mobile phase consisted of 90% acetonitrile and 10% buffer. A gradient was used to achieve chromatographic separation, starting from 100% aqueous phase and linearly progressing to 100% organic phase over a period of 1.5 min, maintaining the 100% organic phase for 0.5 min. At 2.1 min the gradient returned to 100% aqueous, and the column re-equilibrated for 1.5 min, for a total run time of 3.6 min. A sample volume of 10 μl was injected onto the LC column, and the flow rate was 0.25 ml/min. Multiple reactions monitoring mode was used for compound quantification in the API 4000, with Turbo Ionspray. The precursor-to-product transitions for all measured compounds are listed in the Table 3.
UV Wavelength (nm)
Acetazolamide Atenolol Carbamazepine Ketoprofen Metoprolol Minoxidil Propranolol Ranitidine Saquinavir Simvastatin
308 275 284 261 222 285 222 313 239 243
Q1/Q3
DP
CE
CXP
IS
Temperature
Acetazolamide Acetazolamide-D3 Atenolol Atenolol-D7 Carbamazepine Carbamazepine-D8 Metoprolol Metoprolol-D7 Minoxidil Minoxidil-D10 Ketoprofen Ketoprofen-D3 Propranolol Propranolol-D7 Ranitidine Ranitidine-D6 Saquinavir Saquinavir-D10 Simvastatin Simvastatin-D6
223/181.2 226/182 267.2/145.1 274.2/145.1 237/194.2 245.5 /202.2 268.2/116.1 275.5/123.2 210/193.4 220.4/203.3 255.2/208.9 258.2/212.2 260.2/116.1 267.2/123.2 315/176.2 321.2/176.3 671.4/571 681/572 419.4/199.3 425.6/285.6
35 35 50 50 50 90 76 85 80 80 60 100 70 70 60 60 120 135 50 80
20 20 40 40 30 27 26 28 23 22 20 22 30 30 25 30 45 46 30 15
11 11 15 15 5 12 10 7 13 12 13 13 5 5 15 10 10 11 5 10
5000 5000 5000 5000 5000 5000 5000 5000 5000 5000 5000 5000 5000 5000 5000 5000 5000 5000 5000 5000
500 500 500 500 500 500 500 500 500 500 500 500 500 500 500 500 500 500 500 500
2.6. Data analysis and statistical analysis The data are presented as mean concentration ± standard deviation. All IDAS1 experiments were performed with six replicates (n = =6). The IDAS2 dissolution was performed in triplicate (n = =3), with six replicates (n = =6) for permeation. This is due to the structure of the IDAS2, with one dissolution chamber for every two permeation chambers. To determine whether or not observed differences were statistically significant (p < 0.05), unpaired t-tests were performed using Microsoft® Office Excel 2010. 3. Results 3.1. Assessment of viscosity effect on drug permeability (IDAS1) When the permeability of pre-dissolved drugs was measured using the IDAS1, the viscosity of the media did not significantly impact the expected permeability of any of the four test compounds (propranolol, carbamazepine, atenolol, and acetazolamide). In all cases, the rate of permeation was similar between the low and high viscosity media. Fig. 2 shows the in vitro permeation time profiles of all compounds in both the low and high viscosity media. As expected, the rate of permeation for the highly permeable compounds (2A) was found to be at least 10-fold faster compared to the low permeability compounds (2B). To compare the effect of viscosity on drug permeated versus time, the area under the percent permeated-time curve (AUCperm) for each tested drug was calculated from the permeation profile in high viscosity LV medium and low viscosity medium, and the ratio of AUCHV perm/ AUCperm was estimated using the approach described by Higashino et al. (2014). . HV The unpaired t-test was performed on AUCLV perm (% min) vs AUCperm (%.min). The permeation was found to be not significantly different between low viscosity and high viscosity media for all tested drugs (Table 4).
Table 2 Detection wavelength. Drug
Compounds
3.2. Assessment of viscosity effect on dissolution of solid dosage form and drug permeability (IDAS2) Dissolution of ten oral solid dosage forms and absorption of released drugs at low viscosity vs high viscosity was performed using IDAS2. Drugs were chosen from all four BCS classes (metoprolol, minoxidil, and propranolol from BCS class 1; carbamazepine, ketoprofen, and simvastatin from BCS class 2; atenolol and ranitidine from BCS class 3; and acetazolamide and saquinavir from BCS class 4) in order to provide 4
European Journal of Pharmaceutical Sciences 143 (2020) 105164
S. Silchenko, et al.
a notably slower rate of drug permeation (Fig. 3A-2, 3B-2, and 3C-2). In order to evaluate if the rates of dissolution (followed by a difference in permeability and absorbance) in high viscosity medium and low viscosity medium were significantly different from each other, the area under the curve for percent of drug dissolved, AUCdiss, (%.min) and present of drug permeated, AUCperm, (%.min) for LV and HV media were calculated and used for statistical analysis and comparison. To estimate the effect of viscosity on AUCdiss and AUCperm, the ratio of each AUC was calculated as follows.
Propranolol Carbamazepine Atenolol Acetazolamide
1 2 3 4
AUCperm (%⋅min) LV
HV
22.1 71.1 14.1 12.3
22.6 60.8 12.7 13.2
± ± ± ±
0.72 6.69 2.10 3.59
AUCperm ratio HV/LV ± ± ± ±
1.8 4.73 0.78 2.10
AUCperm(HV/LV) ratio = AUCperm (HV)/ AUCperm (LV)
(2)
3.2.2. Effect of viscosity on dissolution and permeation of BCS 2 drug products (carbamazepine, simvastatin, ketoprofen) Due to dissolution-rate limited adsorption, the adsorption of compounds classified as BCS2 (low solubility/high permeability), tend to be sensitive to the presence of food (Martinez and Amidon, 2002). Solid oral products of carbamazepine, simvastatin, and ketoprofen (Table 1) were tested with the IDAS2 for dissolution and permeation in both low viscosity and high viscosity dissolution medium. The dissolved drug and permeated drug for each product are presented in Fig. 4, as a percent of the dose dissolved or permeated vs time. The viscosity of the dissolution medium significantly delayed the drug release for all three tested products of BCS 2 drugs (Fig. 4A-1, 4B-1, and 4C-1). However, the measured equilibrium solubility of tested drugs in the HV medium was similar to solubility in the LV medium (data not shown). Based on these observations, it is probable that the wetting and disintegration of solid product in the HV medium caused a significantly different dissolution profile compared to the LV medium for the tested solid dosage forms. The delay in drug release from each solid dosage form in the HV medium was reflected in a notably slower rate of drug permeation (Fig. 4A-2, 4B-2, and 4C-2). The AUCdiss (%.min) of drug dissolved and AUCperm (%.min) of drug permeated for carbamazepine, simvastatin, and ketoprofen were calculated and used for statistical analysis. The AUCdiss, AUCperm, and AUC ratio values for the compounds in high and low viscosity media are presented in Table 5 and Fig. 7. A Student's t-test was used to calculate statistical difference, and the observed dissolution and permeation for all three drugs, carbamazepine, simvastatin, ketoprofen, were significantly different (p < 0.05) between the HV and LV media (Table 4). The effect of viscosity on drugs from the BCS 2 group was more apparent than in the BCS 1 group (Table 5). The AUCdiss(HV/LV) ratio was calculated to be 0.42, 0.76 and 0.34 for carbamazepine, ketoprofen, and simvastatin respectively, showing different sensitivity of each solid dosage form drug toward the viscosity of the dissolution medium. The negative effect of viscosity on carbamazepine, ketoprofen, and simvastatin permeability was apparent, with AUCperm(HV/LV) ratio values of 0.28, 0.47, and 0.55 for each drug, respectively (Table 5).
Table 4 In-vitro Area under the concentration-time curve of permeated (AUCperm) model drugs in low viscosity (LV) and high viscosity (HV) media using the IDAS1. BCS
(1)
The AUCdiss, AUCperm, and AUC ratio values for the compounds in high and low viscosity media are presented in Table 5 and Fig. 7. A Student's t-test was used to calculate statistical difference, and the observed dissolution and permeation for all three drugs, propranolol, metoprolol and minoxidil, were found to be significantly different (p < 0.05) between the HV and LV media (Table 5). At the same time, the negative effect of viscosity on drug dissolution was very similar, with an AUCdiss(HV/LV) ratio value of approximately 0.9 for all three tested BCS1 solid dosage forms (Table 5). Notably, the negative effect of viscosity on drug absorption and AUCperm(HV/LV) ratio displayed some variability, with ratios of 0.83 > 0.69 > 0.53 for minoxidil, propranolol, and metoprolol, respectively (Table 4). The negative effect of viscosity on drug permeability in the presence of PEG400 on the apical side of the D/P system was mentioned by Kataoka et al. (2012). One of the possible explanations for decreased rate of permeability for tested drug 3H-dexamethasone was viscosity effect of PEG400.
Fig. 2. In vitro permeation time profiles of model drugs in low (___) and high (——) viscosity media measured using the IDAS1: Panel A for highly permeable drugs propranolol (■/ÿ) and carbamazepine (●/o/), Panel B for low permeable drugs atenolol (♦/à) and acetazolamide (▼/r), respectively.
Drug
AUCdiss(HV/LV) ratio = AUCdiss (HV)/ AUCdiss (LV)
1.03 0.86 0.90 1.07
data from both low and high permeability and solubility compounds. 3.2.1. Effect of viscosity on dissolution and permeation of BCS 1 drug products (metoprolol, minoxidil, and propranolol) Three solid oral products (Table 1) of metoprolol, minoxidil, and propranolol, drugs classified as BCS1 (high solubility/high permeability), were tested with the IDAS2 for dissolution and permeation in both low viscosity and high viscosity dissolution medium. The dissolved drug and permeated drug for each product are presented in Fig. 3, as a percent of the dose dissolved or permeated vs time. As can be seen from the dissolution profiles (Figs. 3A-1, 3B-1, and 3C1), the viscosity of dissolution medium delayed the drug release for all three tested products within first 60 min. This effect was especially pronounced for propranolol within the first 10 min, where the wetting and disintegration of propranolol tablets in high viscosity medium was notably slower compared to low viscosity medium. Although this phenomenon was not observed to such an extent in the dissolution profile of minoxidil tablets, there was still some delay in release from the tablets in high viscosity compared to low viscosity. The dissolution profile of metoprolol tablets in low viscosity vs high viscosity also demonstrated this effect within first 60 min of the experiment, and the dissolution profiles for LV and HV were similar after. The delay in drug release from each solid dosage form in the HV medium was reflected in 5
European Journal of Pharmaceutical Sciences 143 (2020) 105164
S. Silchenko, et al.
Fig. 3. In vitro dissolution and permeation time profiles of BCS1 model drugs in low viscosity (●) and high viscosity (ο) dissolution media. Individual graphs represent metoprolol dissolution (3A-1) and permeation (3A-2), minoxidil dissolution (3B-1) and permeation (3B-2), and propranolol dissolution (3C-1) and permeation (3C-2).
Table 5 In-vitro area under the concentration-time curve of dissolved (AUCdiss) and permeated (AUCperm) model oral dosage forms in low viscosity (LV) and high viscosity (HV) media using the IDAS2. Drug Tables
Metoprolol Minoxidil Propranolol Carbamazepine Ketoprofen Simvastatin Atenolol Ranitidine Acetazolamide Saquinavir ⁎
BCS
1 1 1 2 2 2 3 3 4 4
AUCdiss (%⋅min) LV
HV
AUCperm ratio HV/LV
AUCdiss ratio LV
HV
AUCperm (%⋅min) HV/LV
8003 ± 42.4 11,757 ± 695 11,212 ± 572 2751 ± 71.9 9437 ± 317 4216 ± 475 12,285 ± 273 12,216 ± 357 4836 ± 539 651 ± 31.9
7190 ± 364* 10,688 ± 150 10,092 ± 367* 1169 ± 126* 7151 ± 107* 1417 ± 116* 10,202 ± 197* 10,692 ± 487* 3407 ± 244* 495 ± 27.1*
0.90 0.91 0.90 0.42 0.76 0.34 0.83 0.88 0.70 0.76
84.3 ± 7.62 66.5 ± 3.56 180 ± 13.7 202 ± 23.3 1574 ± 195 73.2 ± 14.0 2.55 ± 0.10 2.64 ± 1.45 30.6 ± 6.62 4.44 ± 1.07
44.6 ± 5.89* 55.3 ± 6.17* 125 ± 16.8* 56.6 ± 14.9* 743 ± 69.2* 40.1 ± 12.2* 1.12 ± 0.20* 1.98 ± 0.37 29.2 ± 3.59 3.94 ± 0.60
0.53 0.83 0.69 0.28 0.47 0.55 0.44 0.75 0.95 0.89
p < 0.05, significantly different from LV group.
6
European Journal of Pharmaceutical Sciences 143 (2020) 105164
S. Silchenko, et al.
Fig. 4. In vitro dissolution and permeation time profiles of BCS2 model drugs in low viscosity (●) and high viscosity (ο) dissolution media. Individual graphs represent carbamazepine dissolution (4A-1) and permeation (4A-2), ketoprofen dissolution (4B-1) and permeation (4B-2), and simvastatin dissolution (4C-1) and permeation (4C-2).
dissolution in the HV medium after that point. This is most likely due to a slower diffusivity of ranitidine in HV medium as compared to the LV medium. For atenolol, the delay in drug release from the solid product was propagated throughout the test, resulting in a slower rate of drug permeation (Fig. 5A-2). However, the permeation rate of ranitidine seemed to be unaffected by the difference in dissolution rate (Fig. 5B-2). It could be explained by low concentration of ranitidine on a receiver side and high standard deviation in measured ranitidine concentration. The AUCdiss (%.min) of drug dissolved and AUCperm (%.min) of drug permeated for atenolol and ranitidine were calculated and used for statistical analysis. The AUCdiss, AUCperm, and AUC ratio values for the compounds in high and low viscosity media are presented in Table 5 and Fig. 7. A Student's t-test was used to calculate statistical difference, and the observed dissolution and permeation profiles for both atenolol and ranitidine were significantly different (p < 0.05) between the HV
3.2.3. Effect of viscosity on dissolution and permeation of BCS 3 drug products (atenolol, ranitidine) Two solid oral products (Table 1) of atenolol and ranitidine, drugs classified as BCS 3 (high solubility/low permeability), were tested with the IDAS2 for dissolution and permeation in both low viscosity and high viscosity dissolution medium. The dissolved drug and permeated drug for each product are presented in Fig. 5, as a percent of the dose dissolved or permeated vs time. The dissolution profiles of both atenolol and ranitidine in low and high viscosity medium (Fig. 5A-1 and 5B-1) was similar to those of the drugs in the BCS 1 group. The dissolution of atenolol tablets was delayed for the first 60 min, most likely due to slower wetting or disintegration of the solid product. The dissolution profile for ranitidine tablets was similar between the low and high viscosity media at the early stage of dissolution (i.e. the first 15 min time point), but exhibited some delay of 7
European Journal of Pharmaceutical Sciences 143 (2020) 105164
S. Silchenko, et al.
Fig. 5. In vitro dissolution and permeation time profiles of BCS3 model drugs in low viscosity (●) and high viscosity (ο) dissolution media. Individual graphs represent atenolol dissolution (5A-1) and permeation (5A-2), and ranitidine dissolution (5B-1) and permeation (5B-2).
unaffected by the difference in dissolution rate. The AUCdiss (%.min) of drug dissolved and AUCperm (%.min) of drug permeated for acetazolamide and saquinavir were calculated and used for statistical analysis. The AUCdiss, AUCperm, and AUC ratio values for the compounds in high and low viscosity media are presented in Table 5 and Fig. 7. A Student's t-test was used to calculate statistical difference, and the observed dissolution profiles for both acetazolamide and saquinavir were significantly different (p < 0.05) between the HV and LV media (Table 5). The permeation profiles were not found to be significantly different from each other (p ≥ 0.05), indicating that there was no difference in the permeated amount of both released drugs in two media. The effect of viscosity on dissolution for BCS4 drugs was very similar to BCS2 group (Table 5). The AUCdissHV/LV ratios were calculated to be 0.70 and 0.76 for acetazolamide and saquinavir, respectively. The AUCpermHV/LV ratios were calculated to be 0.95 and 0.89 for acetazolamide and saquinavir, respectively, which was not similar to the ratios obtained from any other BCS group (Table 5).
and LV media (Table 5). The effect of viscosity on drugs from the BCS 3 group was very similar to the BCS 1 group. The AUCdissHV/LV ratios were calculated to be 0.83 and 0.88 and the AUCperm HV/LV ratios were calculated to be 0.44 and 0.75 for atenolol and ranitidine, respectively (Table 5). 3.2.4. Effect of viscosity on dissolution and permeation of BCS 4 drug products (acetazolamide, saquinavir) Two solid oral products (Table 1) of acetazolamide and saquinavir, drugs classified as BCS 4 (low solubility/low permeability), were tested with the IDAS2 for dissolution and permeation in both low viscosity and high viscosity dissolution medium. The dissolved drug and permeated drug for each product are presented in Fig. 6, as a percent of the dose dissolved or permeated vs time. The dissolution profiles of acetazolamide and saquinavir tablets showed significant difference at low viscosity compared to high viscosity, while the permeation profiles did not show any difference for either drug. The dissolution profiles of acetazolamide and saquinavir in low and high viscosity media were similar to those of the drugs in the BCS 2 group in the sense that the percentage of drug dissolved over 120 min for both acetazolamide and saquinavir was below 100% (Fig. 6A-1and 6B-1). Based on the shape of the dissolution profile, the low diffusivity of acetazolamide in high viscosity medium might be the reason for dissolution differences in two media, while limited wettability and disintegration of the tablet was responsible for the decreased the dissolution rate of saquinavir within first 60 min. For both of the BCS 4 drugs, it was apparent that the delay in drug release from each solid product did not affect the rate of drug permeation between the two media (Fig. 6A-2 and 6B-2). The rate of permeation for both acetazolamide and saquinavir seemed to be
4. Discussion The effect of food on drug absorption in vivo depends on meal composition, viscosity level, and/or specific window of drug adsorption (Welling, 1996). Using Echo planar MRI, the viscosity of gastric medium after 12 min of ingestion of model meal with increased viscosity in healthy humans has been reported to be from 30 to 4000 mPa.s (Marciani et al., 2000). On the other hand, the luminal viscosity significantly changes with meal composition and specific location along GI tract. Depending on level of luminal viscosity, property of viscosity enhancing agents, the drug absorption pattern (local or not) the food viscosity may affect the drug dissolution and absorption of all BCS 8
European Journal of Pharmaceutical Sciences 143 (2020) 105164
S. Silchenko, et al.
Fig. 6. In vitro dissolution and permeation time profiles of BCS4 model drugs in low viscosity (●) and high viscosity (ο) dissolution media. Individual graphs represent acetazolamide dissolution (6A-1) and permeation (6A-2), and saquinavir dissolution (6B-1) and permeation (6B-2).
The negative viscosity effect on drug dissolved, AUCdiss, was observed for all oral products tested, suppressing the total amount dissolved within 120 min. For all products tested with the IDAS2, the area under the curve (%.min) values for drug dissolved (AUCdiss) and drug permeated (AUCperm) are presented in Fig. 7. The observations provided here are in agreement with the previously reported data, though in some cases to lesser extent. The dissolution study of metoprolol tartrate (BCS class 1) tablets in the medium with viscosity enhancing agent showed decreased metoprolol release (Cvijic et al., 2014). Much more pronounced negative viscosity effect of HPMC on atenolol (BCS class 3)
classes 1, 2, 3, 4 (Reppas et al., 1998). The consumption of meal high on water-soluble fibers has shown significant effect on adsorption of some drugs with no effect on another. The significant negative effect of food viscosity on drug bioavailability has been reported for Crixivan® (indinavir, BCS class 2) which was codosed in humans with 500 ml of 2% HPMC in water (Carver et al., 1999). At the same time the effect of meal viscosity might depend on the specific location of drug adsorption. If the drug absorption is not site specific then the luminal viscosity will have minimal effect on total amount absorbed (Martinez and Amidon, 2002).
Fig. 7. In vitro area under the concentration-time curve of dissolved (AUCdiss, 7A) and permeated (AUCperm, 7B) model drugs in low viscosity (LV) and high viscosity (HV) dissolution media. *All drugs in tablet form except ketoprofen in capsule form. 9
European Journal of Pharmaceutical Sciences 143 (2020) 105164
S. Silchenko, et al.
values as compared to low permeability drugs (atenolol, ranitidine, BCS 3, and acetazolamide and saquinavir, BCS 4). The viscosity of the medium made an impact on the dissolution rate and amount dissolved of all tested oral products. However, the impact on the final amount permeated was not as uniform. In order to visualize the viscosity effect on amount absorbed, the AUCdiss HV/LV ratio and AUCperm HV/LV ratio were plotted next to each other (Fig. 8). The ratio of AUCperm HV/LV to AUCdiss HV/LV was also calculated, and is presented in Table 6. If decreased dissolution was the only driving force behind the decreased amount absorbed, then the ratio of AUCperm HV/LV to AUCdiss HV/LV was close to one, for example minoxidil or ranitidine. If there was some other driving force behind the decreased amount absorbed, such as diffusivity or permeability, then the ratio of AUCperm HV/LV to AUCdiss HV/LV was less than one, as with metoprolol, carbamazepine, ketoprofen, and atenolol. In instances where the negative effect of viscosity on dissolution was not reflected in the amount absorbed, the AUCperm HV/LV to AUCdiss HV/LV ratio was greater than one, as with simvastatin, acetazolamide, and saquinavir. In general, the dissolution of compounds with low solubility (BCS class 2 and 4) were most impacted by the high viscosity medium tested, while the high solubility compounds (BCS class 1 and 3) displayed a definite but less noticeable decrease in dissolution. For all compounds except minoxidil, the difference between the AUCdiss in HV and LV media was considered statistically significant. .
Fig. 8. Ratios of AUC in high viscosity over AUC in low viscosity. *All drugs in tablet forms except ketoprofen in capsule form. Table 6 Viscosity effect on amount dissolved and amount permeated. Drug tables
BCS
AUCdiss ratio
AUCperm ratio
5. Conclusion
AUCpermHV/LV/ AUCdissHV/
Current in vitro models for food effect prediction are mainly constructed on Pharmacopeias compendial media, with such variables as pH, bile salts, lipolytic enzymes, and phospholipids content. From this study, it is clear that the viscosity of the stomach and intestinal contents is also an important variable for predicting food effect. The viscosity of the media may change both the dissolution and permeation profiles of a consumed oral product and lead to an underestimation or overestimation of drug absorption. The effect of viscosity of the dissolution medium on dissolution of ten intact oral solid products and absorbance of released drugs was investigated with the new in vitro tool, IDAS2. The study revealed the applicability of the IDAS2 as a model for in vitro screening of dissolution and absorption of intact oral solid products to predict food viscosity effect. The proposed experimental approach exhibited the potential to differentiate between oral solid products and their sensitivity toward the viscosity of the medium they are being dosed in. The IDAS2 might be very helpful for formulators developing new oral dosage forms or re-formulating the existing oral drugs. The authors are considering in future exploring IDAS2 applicability for drug dissolution and permeation from intact solid dosage form using fed media with viscosity enhancing agent present.
LV
Metoprolol Minoxidil Propranolol Carbamazepine Ketoprofen Simvastatin Atenolol Ranitidine Acetazolamide Saquinavir
1 1 1 2 2 2 3 3 4 4
HV/LV
HV/LV
0.90 0.91 0.90 0.42 0.76 0.34 0.83 0.88 0.70 0.76
0.53 0.83 0.69 0.28 0.47 0.55 0.44 0.75 0.95 0.89
0.59 0.91 0.77 0.66 0.62 1.63 0.53 0.86 1.36 1.17
release rate from the tablets tested compared to our results was reported elsewhere (Cvijic et al., 2014). The Cmax of paracetamol (BCS class 1) significantly decreased in dog plasma when co-dosed with 4% of guar gum, water-soluble polysaccharide (Reppas et al., 1998). The effect of viscosity on dissolution and absorption of drug mentioned above was much more pronounced compared to our data. The difference in the level of viscosity effect can be attributed to much higher viscosity levels in the reference, 3550 mP.s versus 110 mP.s used in this study. The high solubility drugs (metoprolol, minoxidil, and propranolol, BCS 1, and atenolol and ranitidine, BCS 3) exhibit the highest AUCdiss compared to low solubility drugs (carbamazepine, simvastatin, BCS 2, and acetazolamide and atenolol, BCS 4) with exception of ketoprofen. Ketoprofen amount dissolved, AUCdiss, was comparable to highly soluble drugs. This profile is most likely due to the fact that ketoprofen is a weak acid with pKa around 4.45 (Drug Bank database), with solubility increasing at pH above pKa. At the pH of 6.5 used in the donor compartment of the IDAS2 ketoprofen behaved similarly to a BCS 1 compound rather than BCS 2. However, from the experiments, reported here it is apparent that the viscosity of the dissolution medium was a rate limiting step on drug permeation/absorption mainly for those drugs classified as BCS 4, exhibiting both low permeability and solubility. As anticipated based on their classifications, the highly permeable drugs (metoprolol, minoxidil, and propranolol, BCS 1, and carbamazepine, ketoprofen, simvastatin, BCS 2) exhibited higher AUCperm
Acknowledgments The authors acknowledge help of a scientific writer Katherine Pin during preparation of the manuscript. References Amidon, G.L., Lennernas, H., Shah, V.P., Crison, J.R., 1995. A theoretical basis for a biopharmaceutic drug classification: the correlation of in vitro drug product dissolution and in vivo bioavailability. Pharm. Res. 12, 413–420. Berben, P., Bauer-Brandl, A., Brandl, M., Faller, B., Flaten, G.E., Jacobsen, A.C., Brouwers, J., Augustijns, P., 2018. Drug permeability profiling using cell-free permeation tools: overview and applications. Eur. J. Pharm. Sci. 119, 219–233. Buch, P., Langguth, P., Kataoka, M., Yamashita, S., 2009. IVIVC in oral absorption for fenofibrate immediate release tablets using a dissolution/permeation system. J. Pharm. Sci. 98, 2001–2009. Carver, P.L., Fleisher, D., Zhou, S.Y., Kaul, D., Kazanjian, P., Li, C., 1999. Meal composition effects on the oral bioavailability of indinavir in HIV-infected patients. Pharm. Res. 16, 718–724.
10
European Journal of Pharmaceutical Sciences 143 (2020) 105164
S. Silchenko, et al.
Cvijic, S., Parojcic, J., Langguth, P., 2014. Viscosity-mediated negative food effect on oral absorption of poorly-permeable drugs with an absorption window in the proximal intestine: in vitro experimental simulation and computational verification. Eur. J. Pharm. Sci. 61, 40–53. Drug Bank database. Drug card for Ketoprofen. https://www.drugbank.ca/drugs/ DB01009(accessed 10 October 2019). Flaten, G.E., Bunjes, H., Luthman, K., Brandl, M., 2006. Drug permeability across a phospholipid vesicle-based barrier 2. Characterization of barrier structure, storage stability and stability towards pH changes. Eur. J. Pharm. Sci. 28, 336–343. Forner, K., Hidalgo, I., Lin, J., Ackermann, M., Langguth, P., 2017. Dissolution/permeation: the importance of the experimental setup for the prediction of formulation effects on fenofibrate in vivo performance. Pharmazie 72, 581–586. Ginski, M.J., Taneja, R., Polli, J.E., 1999. Prediction of dissolution-absorption relationships from a continuous dissolution/Caco-2 system. AAPS PharmSci 1, E3. Heinen, C.A., Reuss, S., Amidon, G.L., Langguth, P., 2013. Ion pairing with bile salts modulates intestinal permeability and contributes to food-drug interaction of bcs class iii compound trospium chloride. Mol. Pharm. 10, 3989–3996. Hens, B., Brouwers, J., Corsetti, M., Augustijns, P., 2015. Gastrointestinal behavior of nano- and microsized fenofibrate: in vivo evaluation in man and in vitro simulation by assessment of the permeation potential. Eur. J. Pharm. Sci. 77, 40–47. Higashino, H., Hasegawa, T., Yamamoto, M., Matsui, R., Masaoka, Y., Kataoka, M., Sakuma, S., Yamashita, S., 2014. In vitro-in vivo correlation of the effect of supersaturation on the intestinal absorption of BCS Class 2 drugs. Mol. Pharm. 11, 746–754. Kataoka, M., Itsubata, S., Masaoka, Y., Sakuma, S., Yamashita, S., 2011. In vitro dissolution/permeation system to predict the oral absorption of poorly water-soluble drugs: effect of food and dose strength on it. Biol. Pharm. Bull. 34, 401–407. Kataoka, M., Masaoka, Y., Sakuma, S., Yamashita, S., 2006. Effect of food intake on the oral absorption of poorly water-soluble drugs: in vitro assessment of drug dissolution and permeation assay system. J. Pharm. Sci. 95, 2051–2061. Kataoka, M., Masaoka, Y., Yamazaki, Y., Sakane, T., Sezaki, H., Yamashita, S., 2003. In vitro system to evaluate oral absorption of poorly water-soluble drugs: simultaneous analysis on dissolution and permeation of drugs. Pharm. Res. 20, 1674–1680. Kataoka, M., Sugano, K., da Costa Mathews, C., Wong, J.W., Jones, K.L., Masaoka, Y., Sakuma, S., Yamashita, S., 2012. Application of dissolution/permeation system for evaluation of formulation effect on oral absorption of poorly water-soluble drugs in drug development. Pharm. Res. 29, 1485–1494. Kourentas, A., Vertzoni, M., Stavrinoudakis, N., Symillidis, A., Brouwers, J., Augustijns, P., Reppas, C., Symillides, M., 2016. An in vitro biorelevant gastrointestinal transfer (BioGIT) system for forecasting concentrations in the fasted upper small intestine: design, implementation, and evaluation. Eur. J. Pharm. Sci. 82, 106–114. Lentz, K.A., 2008. Current methods for predicting human food effect. AAPS J. 10, 282–288. Li, J., Hidalgo, I.J., 2017. System for the concomitant assessment of drug dissolution,
absorption and permeation and methods of using the same. U.S. Patent9,546–991. Li, J., Li, L.B., Nessah, N., Huang, Y., Hidalgo, C., Owen, A., Hidalgo, I.J., 2019. Simultaneous analysis of dissolution and permeation profiles of nanosized and microsized formulations of indomethacin using the in vitro dissolution absorption system 2. J. Pharm. Sci. 108, 2334–2340. Marciani, L., Gowland, P.A., Spiller, R.C., Manoj, P., Moore, R.J., Young, P., Al-Sahab, S., Bush, D., Wright, J., Fillery-Travis, A.J., 2000. Gastric response to increased meal viscosity assessed by echo-planar magnetic resonance imaging in humans. J. Nutr. 130, 122–127. Markopoulos, C., Thoenen, F., Preisig, D., Symillides, M., Vertzoni, M., Parrott, N., Reppas, C., Imanidis, G., 2014. Biorelevant media for transport experiments in the Caco-2 model to evaluate drug absorption in the fasted and the fed state and their usefulness. Eur. J. Pharm. Biopharm. 86, 438–448. Markopoulos, C., Andreas, C.J., Vertzoni, M., Dressman, J., Reppas, C., 2015. In-vitro simulation of luminal conditions for evaluation of performance of oral drug products: choosing the appropriate test media. Eur. J. Pharm. Biopharm. 93, 173–182. Martinez, M.N., Amidon, G.L., 2002. A mechanistic approach to understanding the factors affecting drug absorption: a review of fundamentals. J. Clin. Pharmacol. 42, 620–643. Mudie, D.M., Amidon, G.L., Amidon, G.E., 2010. Physiological parameters for oral delivery and in vitro testing. Mol. Pharm. 7, 1388–1405. Parojcic, J., Vasiljevic, D., Ibric, S., Djuric, Z., 2008. Tablet disintegration and drug dissolution in viscous media: paracetamol IR tablets. Int. J. Pharm. 355, 93–99. Perez de la Cruz Moreno, M., Oth, M., Deferme, S., Lammert, F., Tack, J., Dressman, J., Augustijns, P., 2006. Characterization of fasted-state human intestinal fluids collected from duodenum and jejunum. J. Pharm. Pharmacol. 58, 1079–1089. Radwan, A., Amidon, G.L., Langguth, P., 2012. Mechanistic investigation of food effect on disintegration and dissolution of BCS class III compound solid formulations: the importance of viscosity. Biopharm. Drug Dispos. 33, 403–416. Reppas, C., Eleftheriou, G., Macheras, P., Symillides, M., Dressman, J.B., 1998. Effect of elevated viscosity in the upper gastrointestinal tract on drug absorption in dogs. Eur. J. Pharm. Sci. 6, 131–139. Riethorst, D., Mols, R., Duchateau, G., Tack, J., Brouwers, J., Augustijns, P., 2016. Characterization of human duodenal fluids in fasted and fed state conditions. J. Pharm. Sci. 105, 673–681. U.S. FDA, 2002. Guidance for industry food-effect bioavailability and Fed bioequivalence studies. https://www.fda.gov/files/drugs/published/Food-Effect-Bioavailabilityand-Fed-Bioequivalence-Studies.pdf(accessed 11 July 2019). Welling, P.G., 1996. Effects of food on drug absorption. Annu. Rev. Nutr. 16, 383–415. Winstanley, P.A., Orme, M.L., 1989. The effects of food on drug bioavailability. Br. J. Clin. Pharmacol. 28, 621–628. 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.
11
Contents lists available at ScienceDirect
European Journal of Pharmaceutical Sciences journal homepage: www.elsevier.com/locate/ejps
In vitro dissolution absorption system (IDAS2): Use for the prediction of food viscosity effects on drug dissolution and absorption from oral solid dosage forms
T
⁎
Svitlana Silchenko , Nourdine Nessah, Jibin Li, Li-Bin Li, Yuehua Huang, Albert J. Owen, Ismael J. Hidalgo Absorption Systems LLC, Exton, PA 19341, USA
A R T I C LE I N FO
A B S T R A C T
Keywords: viscosity of medium oral solid dosage products dissolution permeation IDAS2 Caco2 cell monolayer
Existing in vitro dissolution or permeation models to predict food effect are mainly based on Pharmacopeias’ compendial media, which specify such variables as pH, bile salts, lipolytic enzymes, and phospholipids content. However, the viscosity of food in the gastrointestinal (GI) tract is not taken into account, although it can affect both the dissolution of the oral solid dosage form and absorption of the released drug. Here, a new in vitro dissolution absorption system (IDAS2) is utilized, which comprises a dissolution apparatus USP2 (DISTEK) equipped with specially constructed permeability chambers containing Caco-2 monolayers, thereby allowing dissolution and transepithelial absorption to be ascertained simultaneously. The IDAS2 was used to evaluate the effect of medium viscosity on both the dissolution of oral solid dosage forms and absorption of released drugs. Such information, which is not ordinarily determined in dissolution and permeation studies, will be helpful to the formulators developing robust oral dosage forms. Commercially available solid dosage forms of ten model drugs from across all BCS classifications were used in this evaluation: metoprolol, minoxidil, and propranolol from BCS class 1; carbamazepine, ketoprofen, and simvastatin from BCS class 2; atenolol and ranitidine from BCS class 3; and acetazolamide and saquinavir from BCS class 4. The study revealed the applicability of IDAS2 as a tool for in vitro screening of dissolution and absorption of intact oral solid products to predict food viscosity effect. The most profound viscosity effect on dissolution and absorption was observed of solid dosage forms for the BCS class 2 compounds carbamazepine and simvastatin. A higher medium viscosity significantly slowed down the dissolution rate of tested BSC class 4 compounds acetazolamide and saquinavir, without significant effect on their absorption. The solid dosage forms least affected by the viscosity of the medium tested were the BCS class 1 compounds minoxidil and propranolol.
1. Introduction The effect of food on rate and extent of drug absorption is included as part of bioavailability and bioequivalence studies in the FDA Guidance for Industry (U.S. FDA Guidance for Industry, 2002). In vivo, food inhibits gastric emptying and modifies intestinal transit time (Winstanley and Orme, 1989). The GI pH value, bile acid concentrations and fluid composition are different in the fasted versus the fed state (Lentz, 2008; Perez de la Cruz Moreno et al., 2006; Riethorst et al., 2016). These factors can affect the wettability, disintegration, and dissolution of oral solid dosage forms, as well as diffusion and absorption of the released drug. Several existing in vitro dissolution/absorption models for solid oral dosage forms have been used to predict in
⁎
vivo drug absorption for the fasted state (Kataoka et al., 2011; Kourentas et al., 2016; Mudie et al., 2010). The modified Noyes-Whitney equation describes the drug dissolution process, i.e. the relationship of the mass of drug dissolved versus time and the variables which have an impact on dissolution. In this equation, the rate of dissolution is dependent on the drug concentration in the boundary layer (solubility), drug bulk concentration, drug diffusivity (diffusion coefficient), and empirical thickness of the hydrodynamic boundary layer surrounding the solid dosage form. Dissolution is also affected by intrinsic properties of the drug or of the solid dosage form, as well as the physiological environment (in vivo) or dissolution medium (in vitro). The in vivo GI fluid composition, changes along the GI tract, as well as with food consumption and digestion. These changes
Corresponding author at: Department of Physical Chemistry and Formulation, Absorption Systems, LLC, 436 Creamery Way, Suite 600, Exton, PA 19341, USA. E-mail address: [email protected] (S. Silchenko).
https://doi.org/10.1016/j.ejps.2019.105164 Received 17 July 2019; Received in revised form 14 November 2019; Accepted 20 November 2019 Available online 21 November 2019 0928-0987/ © 2019 Published by Elsevier B.V.
European Journal of Pharmaceutical Sciences 143 (2020) 105164
S. Silchenko, et al.
Fig. 1. Schematic illustration of in vitro dissolution absorption systems IDAS1 (A) and IDAS2 (B).
because the passive diffusion, para-cellular, and active drug transporters are involved in the drug permeation in parallel (Berben et al., 2018). Correctly adjusted biorelevant medium (Yamashita et al., 2000) has shown good in vitro-in vivo correlation under fed and fasted conditions (Kataoka et al., 2011; Markopoulos et al., 2014). Despite mimicking various in vivo intestinal conditions (e.g. medium pH, bile acid concentration) on the apical side of the cells, Caco-2 cell monolayers might be sensitive to food components and excipients. The first integrated dissolution/Caco-2 permeation system was reported by Ginski et al. (1999). The integrated system was tested on several commercial solid oral dosage forms and proved to be a valuable instrument for formulators in formulation development. Another in vitro tool with simultaneous dissolution/permeation (D/P) measurement across a Caco-2 cell monolayer permeability barrier was introduced by Kataoka et al. (2003). This D/P system was successfully used for predicting formulation effects on the oral absorption of drugs with low aqueous solubility (Kataoka et al., 2012) in evaluation of food effects and dose strength on drug dissolution and permeation with IVIVC projection (Buch et al., 2009; Kataoka et al., 2011, 2006). Absorption Systems LLC adapted this D/P system, and termed it “In-vitro Dissolution Absorption System 1″ (IDAS1). The IDAS1 tool using fasted state simulated small intestinal fluid (FaSSIF) as dissolution media along with conventional Ussing chambers with mounted Caco-2 cells using modified FaSSIFmod were found to accurately predict the formulation effect on fenofibrate in vivo performance (Forner et al., 2017). However, the existing in vitro models of translating dissolution/ permeation data into in vivo performance have limited applicability when it comes to pharmaceutical testing. For instance, the dissolution chamber in the IDAS1 is too small to accommodate an intact drug tablet. In contrast, more recently developed tools such as the in vitro dissolution absorption system 2 (IDAS2), where dissolution of intact solid oral product is integrated with the permeation of released drug through Caco-2 monolayer membrane, are more capable of implementing current formulation development trends. The IDAS2 was designed at Absorption Systems LLC (Li and Hidalgo, 2017) consists of a commercially available dissolution apparatus (USP1 or USP2), fitted with a specially designed lid with two attached permeability chambers (Fig. 1, B). The IDAS2 has successfully been tested on commercial solid oral dosage forms of differently sized indomethacin formulations (Li et al., 2019).
include differences in pH and differences in the concentrations of bile salts, lipase, and pepsin. Additionally, the viscosity of the fluid might change with food consumption as well, further affecting the drug dissolution and bioavailability (Reppas et al., 1998). To predict in vivo drug dissolution, various biorelevant media have been suggested for use with in vitro dissolution studies. The recent most comprehensive publication (Markopoulos et al., 2015) suggests four levels of biorelevant media simulating various gastrointestinal conditions for in-vitro dissolution studies. The publication mentions the importance of postprandial medium viscosity on drug in vitro dissolution. The viscosity of the medium itself may also have an impact on the wettability, disintegration, and dissolution of a drug product, as well as the diffusion of the released drug (Parojcic et al., 2008; Radwan et al., 2012). For example, a negative food effect on trospium chloride bioavailability has previously been observed and explained as an intrinsic property of the drug. However, it has been shown that a higher medium viscosity prolonged the disintegration time and decreased the dissolution rates of trospium chloride tablets, which might also have contributed to the observed negative food effect on trospium chloride bioavailability (Heinen et al., 2013). Even though the comprehensive media mentioned above have been useful in dissolution studies for predicting in-vitro/in-vivo correlation, formulators developing oral solid dosage forms are in need of more diverse methods and systems to predict drug release and absorption. To predict drug absorption using drug permeability screening, various cellfree permeation systems with two classes of barriers, biomimetic and dialysis membranes, have been used (Berben et al., 2018). Some advantages of such cell-free permeation tools is robustness across a broad pH range and stability toward solubilizing excipients used in advanced solubility-enhancing formulations of poorly soluble drugs (Flaten et al., 2006). The in vitro setup where dissolution testing is combined with a cell-free sink permeation bag has been used to investigate food effect on nanosized and microsized fenofibrate (Hens et al., 2015). The authors were able to demonstrate usefulness of the combined dissolution and permeation tool for prediction of food effect on drug absorption. However, a major disadvantage of cell-free permeation systems is that they are only applicable for predicting passive drug transport. The Caco-2 cell monolayer based models, routinely used as in vitro models for drug permeability screening in BCS classification studies (Amidon et al., 1995) allow for generation of biorelevant results 2
European Journal of Pharmaceutical Sciences 143 (2020) 105164
S. Silchenko, et al.
Transwell and Snapwell plates to grow cell monolayers for the permeability studies. The culture medium was changed every other day until use (20 to 28 days after seeding).
The purpose of the current study was to show the applicability of the IDAS2 to simultaneous dissolution and absorption screening of oral solid dosage during development, with regards to food viscosity effect in particular.
2.4. IDAS1 and IDAS2 2. Materials and methods Diagrams of both IDAS1 and IDAS2 are presented in Fig.1 (Part A and Part B, respectively). Both IDAS1 and IDAS2 are comprised of a dissolution chamber/vessel and one or two permeation chambers separated by human Caco-2 cell monolayers. Both systems also enable simultaneous measurements of drug dissolution and permeation in vitro. However, only IDAS2 permits evaluation of intact oral solid dosage forms (tablets or capsules).
2.1. Materials and dosage forms Propranolol, metoprolol, minoxidil, carbamazepine, simvastatin, ketoprofen, atenolol, ranitidine, acetazolamide, saquinavir mesylate, Dglucose, 2-(N-morpholino)ethanesulfonic acid (MES), and bovine serum albumin (BSA) were obtained from Sigma-Aldrich (St. Louis, MO). Hanks’ balanced salt solution (HBSS), Dulbecco's modified Eagle's medium (DMEM), Dulbecco's phosphate-buffered saline (DPBS), fetal bovine serum (FBS), penicillin-streptomycin mixture, non-essential amino acids (NEAA), sodium pyruvate, and trypsin were obtained from Life Technologies (Grand Island, NY). N-(2-Hydroxyethyl)piperazine-N(2-ethanesulfonic acid) (HEPES, 1 M solution) was purchased from Thermo Fisher Scientific (Waltham, MA). Costar Transwell® (6-well format, 4.67 cm2 insert area, 0.4 µm pore size) and Snapwell® plates (6well format, 1.13 cm2 insert area, 0.4 µm pore size) were purchased from Corning Life Sciences (Corning, NY). Hydroxypropyl methylcellulose (HPMC, Methocel K100 Premium LV) was kindly provided by DOW Chemical. All chemicals were stored properly at all times and used before their expiration dates. The commercial solid oral dosage forms used in the study are presented in Table 1.
2.4.1. Viscosity effect on permeation profile with IDAS1 In this first study, the effect of viscosity on active pharmaceutical ingredient (API) permeation was evaluated with IDAS1. Four test drugs were selected, one from each of the four BCS classes: propranolol (Class 1), carbamazepine (Class 2), atenolol (Class 3), and acetazolamide (Class 4)). The media, 8 ml of HBSSg buffer pH 6.5 with or without 1.4% HPMC and 5.5 ml of HBSSg buffer pH 7.4 with 4% BSA, were loaded into the donor side and receiver side, respectively, which were separated by a Caco-2 monolayer. The system was pre-warmed to 37°C, and each chamber was independently stirred at 200 rpm. Each API was pre-dissolved in DMSO and introduced to the mucosal side at time 0. The high permeability compounds propranolol and carbamazepine were dosed at a concentration of 10 µM, while the low permeability compounds atenolol and acetazolamide were dosed at a concentration of 100 µM. Stock solutions were prepared in DMSO at concentrations of 10 mM for propranolol and carbamazepine, 100 mM for acetazolamide, and 60 mM for atenolol, such that the final concentration of DMSO on the donor side was no more than 0.1%. Receiver sample aliquots of 1 ml were routinely taken from the receiver side at 5, 10, 20, and 30 min for highly permeable API (propranolol and carbamazepine) and at 30, 60, 90, and 120 min for API with low permeability (atenolol and acetazolamide). All samples were analyzed by LC–MS/MS.
2.2. Media preparation and viscosity measurement Buffer solution, HBSS supplemented with 10 mM D-glucose (HBSSg) at pH 6.5, was used on the donor side for both IDAS1 and IDAS2 to simulate the low viscosity of fasting condition stomach contents. To simulate the high viscosity fed condition, HPMC (Methocel K100 Premium LV) was dissolved in HBSSg buffer at 1.4% (w/v) concentration and used on the donor side. The medium viscosity was measured with a micro-viscometer (Model: µVISC, RheoSense, Inc.). The measured viscosity of 1.4% HPMC (Methocel K100 Premium LV) in HBSSg buffer was 110 mP.s. In all experiments, HBSSg buffer pH 7.4 containing 4.5% BSA (w/v) was used on the receiver side.
2.4.2. Viscosity effect on dissolution and permeation profile with IDAS2 For the second study, IDAS2 was used to evaluate the effect of viscosity on a wider range of compounds. Ten test drugs were selected from across all four BCS classes: metoprolol, minoxidil, and propranolol from BCS class 1; carbamazepine, ketoprofen, and simvastatin from BCS class 2; atenolol and ranitidine from BCS class 3; and acetazolamide and saquinavir from BCS class 4. The dissolution system vessels, DISTEK 2100B, were equipped with either rotating paddles for tablets or baskets for capsules. They were loaded with 500 ml of HBSSg buffer pH 6.5 (500 ml) for the low viscosity experiment, and HBSSg buffer pH 6.5 with 1.4% HPMC for the high viscosity experiment. Two permeation chambers (receiver side) were submerged in each vessel for both the low and high viscosity experiments, each containing 8 ml of HBSSg buffer pH 7.4 with 4.5% BSA, with Caco-2 monolayers separating the
2.3. Preparation of Caco-2 monolayers Caco-2 cells, clone C2bbe1, were maintained in DMEM containing 10% FBS, 1% NEAA, 4 mM L-glutamine, 1 mM sodium pyruvate, 100 IU/ml penicillin, and 100 µg/ml streptomycin in a humidified incubator (37 °C, 5% CO2). The culture medium was changed three times weekly, and cell growth was observed by light microscopy. When the stock cultures were approximately 80% confluent, the cells were harvested with trypsin-EDTA and seeded at a density of 60,000 cells/cm2 onto collagen-coated polycarbonate membrane filters (0.4 µm pore size) in Table 1 Commercial solid oral dosage forms tested. Drug Tablets
Strength (mg/unit)
MW
Manufacturer
NDC
Dose conc. (µg/ml)
Metoprolol tartrate Minoxidil Propranolol HCl Carbamazepine Ketoprofen (capsule) Simvastatin Atenolol Ranitidine HCl Acetazolamide Saquinavir
100 10 80 200 50 20 100 300 125 500
684.6 209.3 295.8 236.3 254.3 418.6 266.3 350.9 222.2 670.9
Mylan Pharmaceuticals Watson Pharm Mylan Pharmaceuticals Novartis Pharmaceuticals Teva Pharmaceuticals Lupin Pharmaceuticals IPR Pharmaceuticals GSK Taro Pharmaceuticals Genentech
0378-0047-01 0591-5643-01 0378-0185-01 0078-0509-05 009303193-01 68180-479-01 52427-431-90 0173-0393-40 51672-4022-1 0004-0244-51
400 40 320 800 200 80 400 1200 500 2000
3
European Journal of Pharmaceutical Sciences 143 (2020) 105164
S. Silchenko, et al.
dissolution and permeation chambers. The BCS classification of drug solubility is based on dose per administrated volume of 250 ml. To keep the dissolution part close to physiological conditions, dose/250 ml, the drug tablets or capsules (two per vessel) were dosed to the dissolution vessel (donor side) and the paddles/baskets rotated at 50 revolutions per minute (RPM). Atenolol was co-dosed with all drugs at 100 µM concentration, as a marker to monitor the Caco-2 cell monolayer integrity. Sample aliquots of 2 ml were withdrawn without replacement from the dissolution vessels at 5, 10, 15, 30, 60, 90, and 120 min. The samples were filtered through nylon syringe filters (0.45 µm pore size), discarding the first 1 ml of filtrate to saturate the filter membrane. Sample aliquots of 1 mL were withdrawn from the permeation chambers at 15, 30, 60, 90, and 120 min, and replaced with the same volume of blank media after each sampling. All samples were analyzed by either UV–Vis absorption or LC–MS/MS.
Table 3 Precursor-to-product transition of compounds in LC-MS/MS.
2.5. Sample analysis Donor samples were analyzed with CLARIOstar® microplate reader (BMG Labtech, Baden-Württemberg, Germany) UV/vis absorption. When necessary, samples were diluted with corresponding dissolution medium (HBSSg at pH 7.4), and the drug concentrations were determined by comparing measured absorbance against a standard curve obtained in the matching matrix. The wavelength details are provided in Table 2. For receiver samples, 100 µl of sample was spiked with 300 µl of 100% acetonitrile containing deuterated internal standard in a 96-well plate. The plate was shaken at 850 RPM for 5 min and then centrifuged at 3000 RPM for 10 min. After that, 100 µl of supernatant was transferred into a new 96-well plate, prefilled with 400 µl water in each well. The new mixture was shaken at 850 RPM for 5 min and then centrifuged at 3000 RPM for 10 min. Finally, 10 µl of supernatant was injected for LC–MS/MS analysis. Sample quantification was conducted by coupling a triple quadruple mass spectrometer (API4000, Applied Biosystem, Foster City, CA) to an Agilent 1260 system (Santa Clara, CA). The LC column was a Thermo Hypersil BDS C18 30 × 2.1 mm i.d., 3 µm, with guard column. (Thermo Fisher Scientific, Waltham, MA). Mobile phase buffer was 25 mM ammonium format pH 3.5. The aqueous mobile phase was 90% water and 10% buffer, and the organic mobile phase consisted of 90% acetonitrile and 10% buffer. A gradient was used to achieve chromatographic separation, starting from 100% aqueous phase and linearly progressing to 100% organic phase over a period of 1.5 min, maintaining the 100% organic phase for 0.5 min. At 2.1 min the gradient returned to 100% aqueous, and the column re-equilibrated for 1.5 min, for a total run time of 3.6 min. A sample volume of 10 μl was injected onto the LC column, and the flow rate was 0.25 ml/min. Multiple reactions monitoring mode was used for compound quantification in the API 4000, with Turbo Ionspray. The precursor-to-product transitions for all measured compounds are listed in the Table 3.
UV Wavelength (nm)
Acetazolamide Atenolol Carbamazepine Ketoprofen Metoprolol Minoxidil Propranolol Ranitidine Saquinavir Simvastatin
308 275 284 261 222 285 222 313 239 243
Q1/Q3
DP
CE
CXP
IS
Temperature
Acetazolamide Acetazolamide-D3 Atenolol Atenolol-D7 Carbamazepine Carbamazepine-D8 Metoprolol Metoprolol-D7 Minoxidil Minoxidil-D10 Ketoprofen Ketoprofen-D3 Propranolol Propranolol-D7 Ranitidine Ranitidine-D6 Saquinavir Saquinavir-D10 Simvastatin Simvastatin-D6
223/181.2 226/182 267.2/145.1 274.2/145.1 237/194.2 245.5 /202.2 268.2/116.1 275.5/123.2 210/193.4 220.4/203.3 255.2/208.9 258.2/212.2 260.2/116.1 267.2/123.2 315/176.2 321.2/176.3 671.4/571 681/572 419.4/199.3 425.6/285.6
35 35 50 50 50 90 76 85 80 80 60 100 70 70 60 60 120 135 50 80
20 20 40 40 30 27 26 28 23 22 20 22 30 30 25 30 45 46 30 15
11 11 15 15 5 12 10 7 13 12 13 13 5 5 15 10 10 11 5 10
5000 5000 5000 5000 5000 5000 5000 5000 5000 5000 5000 5000 5000 5000 5000 5000 5000 5000 5000 5000
500 500 500 500 500 500 500 500 500 500 500 500 500 500 500 500 500 500 500 500
2.6. Data analysis and statistical analysis The data are presented as mean concentration ± standard deviation. All IDAS1 experiments were performed with six replicates (n = =6). The IDAS2 dissolution was performed in triplicate (n = =3), with six replicates (n = =6) for permeation. This is due to the structure of the IDAS2, with one dissolution chamber for every two permeation chambers. To determine whether or not observed differences were statistically significant (p < 0.05), unpaired t-tests were performed using Microsoft® Office Excel 2010. 3. Results 3.1. Assessment of viscosity effect on drug permeability (IDAS1) When the permeability of pre-dissolved drugs was measured using the IDAS1, the viscosity of the media did not significantly impact the expected permeability of any of the four test compounds (propranolol, carbamazepine, atenolol, and acetazolamide). In all cases, the rate of permeation was similar between the low and high viscosity media. Fig. 2 shows the in vitro permeation time profiles of all compounds in both the low and high viscosity media. As expected, the rate of permeation for the highly permeable compounds (2A) was found to be at least 10-fold faster compared to the low permeability compounds (2B). To compare the effect of viscosity on drug permeated versus time, the area under the percent permeated-time curve (AUCperm) for each tested drug was calculated from the permeation profile in high viscosity LV medium and low viscosity medium, and the ratio of AUCHV perm/ AUCperm was estimated using the approach described by Higashino et al. (2014). . HV The unpaired t-test was performed on AUCLV perm (% min) vs AUCperm (%.min). The permeation was found to be not significantly different between low viscosity and high viscosity media for all tested drugs (Table 4).
Table 2 Detection wavelength. Drug
Compounds
3.2. Assessment of viscosity effect on dissolution of solid dosage form and drug permeability (IDAS2) Dissolution of ten oral solid dosage forms and absorption of released drugs at low viscosity vs high viscosity was performed using IDAS2. Drugs were chosen from all four BCS classes (metoprolol, minoxidil, and propranolol from BCS class 1; carbamazepine, ketoprofen, and simvastatin from BCS class 2; atenolol and ranitidine from BCS class 3; and acetazolamide and saquinavir from BCS class 4) in order to provide 4
European Journal of Pharmaceutical Sciences 143 (2020) 105164
S. Silchenko, et al.
a notably slower rate of drug permeation (Fig. 3A-2, 3B-2, and 3C-2). In order to evaluate if the rates of dissolution (followed by a difference in permeability and absorbance) in high viscosity medium and low viscosity medium were significantly different from each other, the area under the curve for percent of drug dissolved, AUCdiss, (%.min) and present of drug permeated, AUCperm, (%.min) for LV and HV media were calculated and used for statistical analysis and comparison. To estimate the effect of viscosity on AUCdiss and AUCperm, the ratio of each AUC was calculated as follows.
Propranolol Carbamazepine Atenolol Acetazolamide
1 2 3 4
AUCperm (%⋅min) LV
HV
22.1 71.1 14.1 12.3
22.6 60.8 12.7 13.2
± ± ± ±
0.72 6.69 2.10 3.59
AUCperm ratio HV/LV ± ± ± ±
1.8 4.73 0.78 2.10
AUCperm(HV/LV) ratio = AUCperm (HV)/ AUCperm (LV)
(2)
3.2.2. Effect of viscosity on dissolution and permeation of BCS 2 drug products (carbamazepine, simvastatin, ketoprofen) Due to dissolution-rate limited adsorption, the adsorption of compounds classified as BCS2 (low solubility/high permeability), tend to be sensitive to the presence of food (Martinez and Amidon, 2002). Solid oral products of carbamazepine, simvastatin, and ketoprofen (Table 1) were tested with the IDAS2 for dissolution and permeation in both low viscosity and high viscosity dissolution medium. The dissolved drug and permeated drug for each product are presented in Fig. 4, as a percent of the dose dissolved or permeated vs time. The viscosity of the dissolution medium significantly delayed the drug release for all three tested products of BCS 2 drugs (Fig. 4A-1, 4B-1, and 4C-1). However, the measured equilibrium solubility of tested drugs in the HV medium was similar to solubility in the LV medium (data not shown). Based on these observations, it is probable that the wetting and disintegration of solid product in the HV medium caused a significantly different dissolution profile compared to the LV medium for the tested solid dosage forms. The delay in drug release from each solid dosage form in the HV medium was reflected in a notably slower rate of drug permeation (Fig. 4A-2, 4B-2, and 4C-2). The AUCdiss (%.min) of drug dissolved and AUCperm (%.min) of drug permeated for carbamazepine, simvastatin, and ketoprofen were calculated and used for statistical analysis. The AUCdiss, AUCperm, and AUC ratio values for the compounds in high and low viscosity media are presented in Table 5 and Fig. 7. A Student's t-test was used to calculate statistical difference, and the observed dissolution and permeation for all three drugs, carbamazepine, simvastatin, ketoprofen, were significantly different (p < 0.05) between the HV and LV media (Table 4). The effect of viscosity on drugs from the BCS 2 group was more apparent than in the BCS 1 group (Table 5). The AUCdiss(HV/LV) ratio was calculated to be 0.42, 0.76 and 0.34 for carbamazepine, ketoprofen, and simvastatin respectively, showing different sensitivity of each solid dosage form drug toward the viscosity of the dissolution medium. The negative effect of viscosity on carbamazepine, ketoprofen, and simvastatin permeability was apparent, with AUCperm(HV/LV) ratio values of 0.28, 0.47, and 0.55 for each drug, respectively (Table 5).
Table 4 In-vitro Area under the concentration-time curve of permeated (AUCperm) model drugs in low viscosity (LV) and high viscosity (HV) media using the IDAS1. BCS
(1)
The AUCdiss, AUCperm, and AUC ratio values for the compounds in high and low viscosity media are presented in Table 5 and Fig. 7. A Student's t-test was used to calculate statistical difference, and the observed dissolution and permeation for all three drugs, propranolol, metoprolol and minoxidil, were found to be significantly different (p < 0.05) between the HV and LV media (Table 5). At the same time, the negative effect of viscosity on drug dissolution was very similar, with an AUCdiss(HV/LV) ratio value of approximately 0.9 for all three tested BCS1 solid dosage forms (Table 5). Notably, the negative effect of viscosity on drug absorption and AUCperm(HV/LV) ratio displayed some variability, with ratios of 0.83 > 0.69 > 0.53 for minoxidil, propranolol, and metoprolol, respectively (Table 4). The negative effect of viscosity on drug permeability in the presence of PEG400 on the apical side of the D/P system was mentioned by Kataoka et al. (2012). One of the possible explanations for decreased rate of permeability for tested drug 3H-dexamethasone was viscosity effect of PEG400.
Fig. 2. In vitro permeation time profiles of model drugs in low (___) and high (——) viscosity media measured using the IDAS1: Panel A for highly permeable drugs propranolol (■/ÿ) and carbamazepine (●/o/), Panel B for low permeable drugs atenolol (♦/à) and acetazolamide (▼/r), respectively.
Drug
AUCdiss(HV/LV) ratio = AUCdiss (HV)/ AUCdiss (LV)
1.03 0.86 0.90 1.07
data from both low and high permeability and solubility compounds. 3.2.1. Effect of viscosity on dissolution and permeation of BCS 1 drug products (metoprolol, minoxidil, and propranolol) Three solid oral products (Table 1) of metoprolol, minoxidil, and propranolol, drugs classified as BCS1 (high solubility/high permeability), were tested with the IDAS2 for dissolution and permeation in both low viscosity and high viscosity dissolution medium. The dissolved drug and permeated drug for each product are presented in Fig. 3, as a percent of the dose dissolved or permeated vs time. As can be seen from the dissolution profiles (Figs. 3A-1, 3B-1, and 3C1), the viscosity of dissolution medium delayed the drug release for all three tested products within first 60 min. This effect was especially pronounced for propranolol within the first 10 min, where the wetting and disintegration of propranolol tablets in high viscosity medium was notably slower compared to low viscosity medium. Although this phenomenon was not observed to such an extent in the dissolution profile of minoxidil tablets, there was still some delay in release from the tablets in high viscosity compared to low viscosity. The dissolution profile of metoprolol tablets in low viscosity vs high viscosity also demonstrated this effect within first 60 min of the experiment, and the dissolution profiles for LV and HV were similar after. The delay in drug release from each solid dosage form in the HV medium was reflected in 5
European Journal of Pharmaceutical Sciences 143 (2020) 105164
S. Silchenko, et al.
Fig. 3. In vitro dissolution and permeation time profiles of BCS1 model drugs in low viscosity (●) and high viscosity (ο) dissolution media. Individual graphs represent metoprolol dissolution (3A-1) and permeation (3A-2), minoxidil dissolution (3B-1) and permeation (3B-2), and propranolol dissolution (3C-1) and permeation (3C-2).
Table 5 In-vitro area under the concentration-time curve of dissolved (AUCdiss) and permeated (AUCperm) model oral dosage forms in low viscosity (LV) and high viscosity (HV) media using the IDAS2. Drug Tables
Metoprolol Minoxidil Propranolol Carbamazepine Ketoprofen Simvastatin Atenolol Ranitidine Acetazolamide Saquinavir ⁎
BCS
1 1 1 2 2 2 3 3 4 4
AUCdiss (%⋅min) LV
HV
AUCperm ratio HV/LV
AUCdiss ratio LV
HV
AUCperm (%⋅min) HV/LV
8003 ± 42.4 11,757 ± 695 11,212 ± 572 2751 ± 71.9 9437 ± 317 4216 ± 475 12,285 ± 273 12,216 ± 357 4836 ± 539 651 ± 31.9
7190 ± 364* 10,688 ± 150 10,092 ± 367* 1169 ± 126* 7151 ± 107* 1417 ± 116* 10,202 ± 197* 10,692 ± 487* 3407 ± 244* 495 ± 27.1*
0.90 0.91 0.90 0.42 0.76 0.34 0.83 0.88 0.70 0.76
84.3 ± 7.62 66.5 ± 3.56 180 ± 13.7 202 ± 23.3 1574 ± 195 73.2 ± 14.0 2.55 ± 0.10 2.64 ± 1.45 30.6 ± 6.62 4.44 ± 1.07
44.6 ± 5.89* 55.3 ± 6.17* 125 ± 16.8* 56.6 ± 14.9* 743 ± 69.2* 40.1 ± 12.2* 1.12 ± 0.20* 1.98 ± 0.37 29.2 ± 3.59 3.94 ± 0.60
0.53 0.83 0.69 0.28 0.47 0.55 0.44 0.75 0.95 0.89
p < 0.05, significantly different from LV group.
6
European Journal of Pharmaceutical Sciences 143 (2020) 105164
S. Silchenko, et al.
Fig. 4. In vitro dissolution and permeation time profiles of BCS2 model drugs in low viscosity (●) and high viscosity (ο) dissolution media. Individual graphs represent carbamazepine dissolution (4A-1) and permeation (4A-2), ketoprofen dissolution (4B-1) and permeation (4B-2), and simvastatin dissolution (4C-1) and permeation (4C-2).
dissolution in the HV medium after that point. This is most likely due to a slower diffusivity of ranitidine in HV medium as compared to the LV medium. For atenolol, the delay in drug release from the solid product was propagated throughout the test, resulting in a slower rate of drug permeation (Fig. 5A-2). However, the permeation rate of ranitidine seemed to be unaffected by the difference in dissolution rate (Fig. 5B-2). It could be explained by low concentration of ranitidine on a receiver side and high standard deviation in measured ranitidine concentration. The AUCdiss (%.min) of drug dissolved and AUCperm (%.min) of drug permeated for atenolol and ranitidine were calculated and used for statistical analysis. The AUCdiss, AUCperm, and AUC ratio values for the compounds in high and low viscosity media are presented in Table 5 and Fig. 7. A Student's t-test was used to calculate statistical difference, and the observed dissolution and permeation profiles for both atenolol and ranitidine were significantly different (p < 0.05) between the HV
3.2.3. Effect of viscosity on dissolution and permeation of BCS 3 drug products (atenolol, ranitidine) Two solid oral products (Table 1) of atenolol and ranitidine, drugs classified as BCS 3 (high solubility/low permeability), were tested with the IDAS2 for dissolution and permeation in both low viscosity and high viscosity dissolution medium. The dissolved drug and permeated drug for each product are presented in Fig. 5, as a percent of the dose dissolved or permeated vs time. The dissolution profiles of both atenolol and ranitidine in low and high viscosity medium (Fig. 5A-1 and 5B-1) was similar to those of the drugs in the BCS 1 group. The dissolution of atenolol tablets was delayed for the first 60 min, most likely due to slower wetting or disintegration of the solid product. The dissolution profile for ranitidine tablets was similar between the low and high viscosity media at the early stage of dissolution (i.e. the first 15 min time point), but exhibited some delay of 7
European Journal of Pharmaceutical Sciences 143 (2020) 105164
S. Silchenko, et al.
Fig. 5. In vitro dissolution and permeation time profiles of BCS3 model drugs in low viscosity (●) and high viscosity (ο) dissolution media. Individual graphs represent atenolol dissolution (5A-1) and permeation (5A-2), and ranitidine dissolution (5B-1) and permeation (5B-2).
unaffected by the difference in dissolution rate. The AUCdiss (%.min) of drug dissolved and AUCperm (%.min) of drug permeated for acetazolamide and saquinavir were calculated and used for statistical analysis. The AUCdiss, AUCperm, and AUC ratio values for the compounds in high and low viscosity media are presented in Table 5 and Fig. 7. A Student's t-test was used to calculate statistical difference, and the observed dissolution profiles for both acetazolamide and saquinavir were significantly different (p < 0.05) between the HV and LV media (Table 5). The permeation profiles were not found to be significantly different from each other (p ≥ 0.05), indicating that there was no difference in the permeated amount of both released drugs in two media. The effect of viscosity on dissolution for BCS4 drugs was very similar to BCS2 group (Table 5). The AUCdissHV/LV ratios were calculated to be 0.70 and 0.76 for acetazolamide and saquinavir, respectively. The AUCpermHV/LV ratios were calculated to be 0.95 and 0.89 for acetazolamide and saquinavir, respectively, which was not similar to the ratios obtained from any other BCS group (Table 5).
and LV media (Table 5). The effect of viscosity on drugs from the BCS 3 group was very similar to the BCS 1 group. The AUCdissHV/LV ratios were calculated to be 0.83 and 0.88 and the AUCperm HV/LV ratios were calculated to be 0.44 and 0.75 for atenolol and ranitidine, respectively (Table 5). 3.2.4. Effect of viscosity on dissolution and permeation of BCS 4 drug products (acetazolamide, saquinavir) Two solid oral products (Table 1) of acetazolamide and saquinavir, drugs classified as BCS 4 (low solubility/low permeability), were tested with the IDAS2 for dissolution and permeation in both low viscosity and high viscosity dissolution medium. The dissolved drug and permeated drug for each product are presented in Fig. 6, as a percent of the dose dissolved or permeated vs time. The dissolution profiles of acetazolamide and saquinavir tablets showed significant difference at low viscosity compared to high viscosity, while the permeation profiles did not show any difference for either drug. The dissolution profiles of acetazolamide and saquinavir in low and high viscosity media were similar to those of the drugs in the BCS 2 group in the sense that the percentage of drug dissolved over 120 min for both acetazolamide and saquinavir was below 100% (Fig. 6A-1and 6B-1). Based on the shape of the dissolution profile, the low diffusivity of acetazolamide in high viscosity medium might be the reason for dissolution differences in two media, while limited wettability and disintegration of the tablet was responsible for the decreased the dissolution rate of saquinavir within first 60 min. For both of the BCS 4 drugs, it was apparent that the delay in drug release from each solid product did not affect the rate of drug permeation between the two media (Fig. 6A-2 and 6B-2). The rate of permeation for both acetazolamide and saquinavir seemed to be
4. Discussion The effect of food on drug absorption in vivo depends on meal composition, viscosity level, and/or specific window of drug adsorption (Welling, 1996). Using Echo planar MRI, the viscosity of gastric medium after 12 min of ingestion of model meal with increased viscosity in healthy humans has been reported to be from 30 to 4000 mPa.s (Marciani et al., 2000). On the other hand, the luminal viscosity significantly changes with meal composition and specific location along GI tract. Depending on level of luminal viscosity, property of viscosity enhancing agents, the drug absorption pattern (local or not) the food viscosity may affect the drug dissolution and absorption of all BCS 8
European Journal of Pharmaceutical Sciences 143 (2020) 105164
S. Silchenko, et al.
Fig. 6. In vitro dissolution and permeation time profiles of BCS4 model drugs in low viscosity (●) and high viscosity (ο) dissolution media. Individual graphs represent acetazolamide dissolution (6A-1) and permeation (6A-2), and saquinavir dissolution (6B-1) and permeation (6B-2).
The negative viscosity effect on drug dissolved, AUCdiss, was observed for all oral products tested, suppressing the total amount dissolved within 120 min. For all products tested with the IDAS2, the area under the curve (%.min) values for drug dissolved (AUCdiss) and drug permeated (AUCperm) are presented in Fig. 7. The observations provided here are in agreement with the previously reported data, though in some cases to lesser extent. The dissolution study of metoprolol tartrate (BCS class 1) tablets in the medium with viscosity enhancing agent showed decreased metoprolol release (Cvijic et al., 2014). Much more pronounced negative viscosity effect of HPMC on atenolol (BCS class 3)
classes 1, 2, 3, 4 (Reppas et al., 1998). The consumption of meal high on water-soluble fibers has shown significant effect on adsorption of some drugs with no effect on another. The significant negative effect of food viscosity on drug bioavailability has been reported for Crixivan® (indinavir, BCS class 2) which was codosed in humans with 500 ml of 2% HPMC in water (Carver et al., 1999). At the same time the effect of meal viscosity might depend on the specific location of drug adsorption. If the drug absorption is not site specific then the luminal viscosity will have minimal effect on total amount absorbed (Martinez and Amidon, 2002).
Fig. 7. In vitro area under the concentration-time curve of dissolved (AUCdiss, 7A) and permeated (AUCperm, 7B) model drugs in low viscosity (LV) and high viscosity (HV) dissolution media. *All drugs in tablet form except ketoprofen in capsule form. 9
European Journal of Pharmaceutical Sciences 143 (2020) 105164
S. Silchenko, et al.
values as compared to low permeability drugs (atenolol, ranitidine, BCS 3, and acetazolamide and saquinavir, BCS 4). The viscosity of the medium made an impact on the dissolution rate and amount dissolved of all tested oral products. However, the impact on the final amount permeated was not as uniform. In order to visualize the viscosity effect on amount absorbed, the AUCdiss HV/LV ratio and AUCperm HV/LV ratio were plotted next to each other (Fig. 8). The ratio of AUCperm HV/LV to AUCdiss HV/LV was also calculated, and is presented in Table 6. If decreased dissolution was the only driving force behind the decreased amount absorbed, then the ratio of AUCperm HV/LV to AUCdiss HV/LV was close to one, for example minoxidil or ranitidine. If there was some other driving force behind the decreased amount absorbed, such as diffusivity or permeability, then the ratio of AUCperm HV/LV to AUCdiss HV/LV was less than one, as with metoprolol, carbamazepine, ketoprofen, and atenolol. In instances where the negative effect of viscosity on dissolution was not reflected in the amount absorbed, the AUCperm HV/LV to AUCdiss HV/LV ratio was greater than one, as with simvastatin, acetazolamide, and saquinavir. In general, the dissolution of compounds with low solubility (BCS class 2 and 4) were most impacted by the high viscosity medium tested, while the high solubility compounds (BCS class 1 and 3) displayed a definite but less noticeable decrease in dissolution. For all compounds except minoxidil, the difference between the AUCdiss in HV and LV media was considered statistically significant. .
Fig. 8. Ratios of AUC in high viscosity over AUC in low viscosity. *All drugs in tablet forms except ketoprofen in capsule form. Table 6 Viscosity effect on amount dissolved and amount permeated. Drug tables
BCS
AUCdiss ratio
AUCperm ratio
5. Conclusion
AUCpermHV/LV/ AUCdissHV/
Current in vitro models for food effect prediction are mainly constructed on Pharmacopeias compendial media, with such variables as pH, bile salts, lipolytic enzymes, and phospholipids content. From this study, it is clear that the viscosity of the stomach and intestinal contents is also an important variable for predicting food effect. The viscosity of the media may change both the dissolution and permeation profiles of a consumed oral product and lead to an underestimation or overestimation of drug absorption. The effect of viscosity of the dissolution medium on dissolution of ten intact oral solid products and absorbance of released drugs was investigated with the new in vitro tool, IDAS2. The study revealed the applicability of the IDAS2 as a model for in vitro screening of dissolution and absorption of intact oral solid products to predict food viscosity effect. The proposed experimental approach exhibited the potential to differentiate between oral solid products and their sensitivity toward the viscosity of the medium they are being dosed in. The IDAS2 might be very helpful for formulators developing new oral dosage forms or re-formulating the existing oral drugs. The authors are considering in future exploring IDAS2 applicability for drug dissolution and permeation from intact solid dosage form using fed media with viscosity enhancing agent present.
LV
Metoprolol Minoxidil Propranolol Carbamazepine Ketoprofen Simvastatin Atenolol Ranitidine Acetazolamide Saquinavir
1 1 1 2 2 2 3 3 4 4
HV/LV
HV/LV
0.90 0.91 0.90 0.42 0.76 0.34 0.83 0.88 0.70 0.76
0.53 0.83 0.69 0.28 0.47 0.55 0.44 0.75 0.95 0.89
0.59 0.91 0.77 0.66 0.62 1.63 0.53 0.86 1.36 1.17
release rate from the tablets tested compared to our results was reported elsewhere (Cvijic et al., 2014). The Cmax of paracetamol (BCS class 1) significantly decreased in dog plasma when co-dosed with 4% of guar gum, water-soluble polysaccharide (Reppas et al., 1998). The effect of viscosity on dissolution and absorption of drug mentioned above was much more pronounced compared to our data. The difference in the level of viscosity effect can be attributed to much higher viscosity levels in the reference, 3550 mP.s versus 110 mP.s used in this study. The high solubility drugs (metoprolol, minoxidil, and propranolol, BCS 1, and atenolol and ranitidine, BCS 3) exhibit the highest AUCdiss compared to low solubility drugs (carbamazepine, simvastatin, BCS 2, and acetazolamide and atenolol, BCS 4) with exception of ketoprofen. Ketoprofen amount dissolved, AUCdiss, was comparable to highly soluble drugs. This profile is most likely due to the fact that ketoprofen is a weak acid with pKa around 4.45 (Drug Bank database), with solubility increasing at pH above pKa. At the pH of 6.5 used in the donor compartment of the IDAS2 ketoprofen behaved similarly to a BCS 1 compound rather than BCS 2. However, from the experiments, reported here it is apparent that the viscosity of the dissolution medium was a rate limiting step on drug permeation/absorption mainly for those drugs classified as BCS 4, exhibiting both low permeability and solubility. As anticipated based on their classifications, the highly permeable drugs (metoprolol, minoxidil, and propranolol, BCS 1, and carbamazepine, ketoprofen, simvastatin, BCS 2) exhibited higher AUCperm
Acknowledgments The authors acknowledge help of a scientific writer Katherine Pin during preparation of the manuscript. References Amidon, G.L., Lennernas, H., Shah, V.P., Crison, J.R., 1995. A theoretical basis for a biopharmaceutic drug classification: the correlation of in vitro drug product dissolution and in vivo bioavailability. Pharm. Res. 12, 413–420. Berben, P., Bauer-Brandl, A., Brandl, M., Faller, B., Flaten, G.E., Jacobsen, A.C., Brouwers, J., Augustijns, P., 2018. Drug permeability profiling using cell-free permeation tools: overview and applications. Eur. J. Pharm. Sci. 119, 219–233. Buch, P., Langguth, P., Kataoka, M., Yamashita, S., 2009. IVIVC in oral absorption for fenofibrate immediate release tablets using a dissolution/permeation system. J. Pharm. Sci. 98, 2001–2009. Carver, P.L., Fleisher, D., Zhou, S.Y., Kaul, D., Kazanjian, P., Li, C., 1999. Meal composition effects on the oral bioavailability of indinavir in HIV-infected patients. Pharm. Res. 16, 718–724.
10
European Journal of Pharmaceutical Sciences 143 (2020) 105164
S. Silchenko, et al.
Cvijic, S., Parojcic, J., Langguth, P., 2014. Viscosity-mediated negative food effect on oral absorption of poorly-permeable drugs with an absorption window in the proximal intestine: in vitro experimental simulation and computational verification. Eur. J. Pharm. Sci. 61, 40–53. Drug Bank database. Drug card for Ketoprofen. https://www.drugbank.ca/drugs/ DB01009(accessed 10 October 2019). Flaten, G.E., Bunjes, H., Luthman, K., Brandl, M., 2006. Drug permeability across a phospholipid vesicle-based barrier 2. Characterization of barrier structure, storage stability and stability towards pH changes. Eur. J. Pharm. Sci. 28, 336–343. Forner, K., Hidalgo, I., Lin, J., Ackermann, M., Langguth, P., 2017. Dissolution/permeation: the importance of the experimental setup for the prediction of formulation effects on fenofibrate in vivo performance. Pharmazie 72, 581–586. Ginski, M.J., Taneja, R., Polli, J.E., 1999. Prediction of dissolution-absorption relationships from a continuous dissolution/Caco-2 system. AAPS PharmSci 1, E3. Heinen, C.A., Reuss, S., Amidon, G.L., Langguth, P., 2013. Ion pairing with bile salts modulates intestinal permeability and contributes to food-drug interaction of bcs class iii compound trospium chloride. Mol. Pharm. 10, 3989–3996. Hens, B., Brouwers, J., Corsetti, M., Augustijns, P., 2015. Gastrointestinal behavior of nano- and microsized fenofibrate: in vivo evaluation in man and in vitro simulation by assessment of the permeation potential. Eur. J. Pharm. Sci. 77, 40–47. Higashino, H., Hasegawa, T., Yamamoto, M., Matsui, R., Masaoka, Y., Kataoka, M., Sakuma, S., Yamashita, S., 2014. In vitro-in vivo correlation of the effect of supersaturation on the intestinal absorption of BCS Class 2 drugs. Mol. Pharm. 11, 746–754. Kataoka, M., Itsubata, S., Masaoka, Y., Sakuma, S., Yamashita, S., 2011. In vitro dissolution/permeation system to predict the oral absorption of poorly water-soluble drugs: effect of food and dose strength on it. Biol. Pharm. Bull. 34, 401–407. Kataoka, M., Masaoka, Y., Sakuma, S., Yamashita, S., 2006. Effect of food intake on the oral absorption of poorly water-soluble drugs: in vitro assessment of drug dissolution and permeation assay system. J. Pharm. Sci. 95, 2051–2061. Kataoka, M., Masaoka, Y., Yamazaki, Y., Sakane, T., Sezaki, H., Yamashita, S., 2003. In vitro system to evaluate oral absorption of poorly water-soluble drugs: simultaneous analysis on dissolution and permeation of drugs. Pharm. Res. 20, 1674–1680. Kataoka, M., Sugano, K., da Costa Mathews, C., Wong, J.W., Jones, K.L., Masaoka, Y., Sakuma, S., Yamashita, S., 2012. Application of dissolution/permeation system for evaluation of formulation effect on oral absorption of poorly water-soluble drugs in drug development. Pharm. Res. 29, 1485–1494. Kourentas, A., Vertzoni, M., Stavrinoudakis, N., Symillidis, A., Brouwers, J., Augustijns, P., Reppas, C., Symillides, M., 2016. An in vitro biorelevant gastrointestinal transfer (BioGIT) system for forecasting concentrations in the fasted upper small intestine: design, implementation, and evaluation. Eur. J. Pharm. Sci. 82, 106–114. Lentz, K.A., 2008. Current methods for predicting human food effect. AAPS J. 10, 282–288. Li, J., Hidalgo, I.J., 2017. System for the concomitant assessment of drug dissolution,
absorption and permeation and methods of using the same. U.S. Patent9,546–991. Li, J., Li, L.B., Nessah, N., Huang, Y., Hidalgo, C., Owen, A., Hidalgo, I.J., 2019. Simultaneous analysis of dissolution and permeation profiles of nanosized and microsized formulations of indomethacin using the in vitro dissolution absorption system 2. J. Pharm. Sci. 108, 2334–2340. Marciani, L., Gowland, P.A., Spiller, R.C., Manoj, P., Moore, R.J., Young, P., Al-Sahab, S., Bush, D., Wright, J., Fillery-Travis, A.J., 2000. Gastric response to increased meal viscosity assessed by echo-planar magnetic resonance imaging in humans. J. Nutr. 130, 122–127. Markopoulos, C., Thoenen, F., Preisig, D., Symillides, M., Vertzoni, M., Parrott, N., Reppas, C., Imanidis, G., 2014. Biorelevant media for transport experiments in the Caco-2 model to evaluate drug absorption in the fasted and the fed state and their usefulness. Eur. J. Pharm. Biopharm. 86, 438–448. Markopoulos, C., Andreas, C.J., Vertzoni, M., Dressman, J., Reppas, C., 2015. In-vitro simulation of luminal conditions for evaluation of performance of oral drug products: choosing the appropriate test media. Eur. J. Pharm. Biopharm. 93, 173–182. Martinez, M.N., Amidon, G.L., 2002. A mechanistic approach to understanding the factors affecting drug absorption: a review of fundamentals. J. Clin. Pharmacol. 42, 620–643. Mudie, D.M., Amidon, G.L., Amidon, G.E., 2010. Physiological parameters for oral delivery and in vitro testing. Mol. Pharm. 7, 1388–1405. Parojcic, J., Vasiljevic, D., Ibric, S., Djuric, Z., 2008. Tablet disintegration and drug dissolution in viscous media: paracetamol IR tablets. Int. J. Pharm. 355, 93–99. Perez de la Cruz Moreno, M., Oth, M., Deferme, S., Lammert, F., Tack, J., Dressman, J., Augustijns, P., 2006. Characterization of fasted-state human intestinal fluids collected from duodenum and jejunum. J. Pharm. Pharmacol. 58, 1079–1089. Radwan, A., Amidon, G.L., Langguth, P., 2012. Mechanistic investigation of food effect on disintegration and dissolution of BCS class III compound solid formulations: the importance of viscosity. Biopharm. Drug Dispos. 33, 403–416. Reppas, C., Eleftheriou, G., Macheras, P., Symillides, M., Dressman, J.B., 1998. Effect of elevated viscosity in the upper gastrointestinal tract on drug absorption in dogs. Eur. J. Pharm. Sci. 6, 131–139. Riethorst, D., Mols, R., Duchateau, G., Tack, J., Brouwers, J., Augustijns, P., 2016. Characterization of human duodenal fluids in fasted and fed state conditions. J. Pharm. Sci. 105, 673–681. U.S. FDA, 2002. Guidance for industry food-effect bioavailability and Fed bioequivalence studies. https://www.fda.gov/files/drugs/published/Food-Effect-Bioavailabilityand-Fed-Bioequivalence-Studies.pdf(accessed 11 July 2019). Welling, P.G., 1996. Effects of food on drug absorption. Annu. Rev. Nutr. 16, 383–415. Winstanley, P.A., Orme, M.L., 1989. The effects of food on drug bioavailability. Br. J. Clin. Pharmacol. 28, 621–628. 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.
11