Movable double paddle

Abstract

IThe present application concerns a testing apparatus, including a central shaft, and a movable assembly defining a central lumen for receiving the central shaft, a disk and at least one blade, the movable assembly being axially translatable relative to the central shaft.

Description

ConfidentialMERCK-0039PCTMOVABLE DOUBLE PADDLECROSS-REFERENCE OF RELATED APPLICATIONS

[0001] This application claims the benefit of the filing date of U.S. Provisional Application No. 63 / 736,657, filed December 20, 2024, the disclosure of which is hereby incorporated herein by reference.FIELD OF THE INVENTION

[0002] The present disclosure generally relates to dissolution experiments. More specifically, the present disclosure relates to devices and method for carrying out and improved dissolution experiments or for mixing materials or phases of materials.BACKGROUND OF THE INVENTION

[0003] It is preferable for medication dosage forms to permit a range of release characteristics, such as immediate, delayed, controlled, or sustained release. For instance, dosage forms can be made that enable the drug's active ingredient to be released in a customized manner over a number of hours, a day, a week, or longer. Some of these release characteristics may allow a particular medicine to be taken less frequently.

[0004] Many medications are designed to dissolve over time. These dosage forms could contain a range of excipients that are intended to dissolve, degrade, or disintegrate over a predetermined amount of time. It is also possible to design a release profile to dissolve at a specific pH, such as lower in the stomach or possibly higher in the intestines later on. Some dose forms, including tablets, capsules, suppositories, or sublingual capsules, may be made to release the medication immediately or gradually, over a comparatively short or lengthy period of time. Drug dosage forms may be formulated with a variety of excipients or matrix components that influence the release of the drug active from the matrix. These excipients may include hydrophilic polymers, hydrophobic polymers, surfactants, disintegrants, waxes or other components.

[0005] In vivo studies may be costly. Conversely, in vitro testing may be used to ascertain a drug's release over time using precisely crafted equipment in regulated environments. These tests can then be associated with in vivo testing to produce commercial formulations. Usually, these tests are created for medication formulations or dosage forms that release the drug through a dissolvingConfidentialMERCK-0039PCT or disintegration mechanism. For oral formulations or dosage forms, the rate at which the drugs are released and / or dissolve in gastrointestinal fluids is important in the design and use of orally administered formulations and dosage forms. Particularly, the drug must be released or dissolved before it can be absorbed by the body. The rate at which the drug enters into solution is known in the art as the dissolution rate, and the determination of the dissolution rate in vitro is known as dissolution testing.

[0006] Dissolution testing can help determine how much of a medication is accessible at a specific absorption site at different times. Furthermore, the creation of tailored delivery methods is aided by the correlation that can be established between the systemic blood levels of a drug and its dose form and availability at certain absorption sites. It would be useful to further improve these dissolution testing devices and methods to obtain more accurate information relating to the dissolution rate of certain drugs.

[0007] Thus, there exists a need for devices that improve upon and advance the methods of carrying out dissolution testing.SUMMARY OF THE INVENTION

[0008] In some examples, a testing apparatus, includes a central shaft, and a movable assembly defining a central lumen for receiving the central shaft, a disk and at least one blade, the movable assembly being axially translatable relative to the central shaft.

[0009] In some examples, a method of dissolution testing includes providing a testing apparatus including a central shaft, and a movable assembly defining a central lumen for receiving the central shaft, a disk and at least one blade, placing the testing apparatus within a vessel, and moving the assembly relative to the central shaft.BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Various embodiments of the presently disclosed dissolution testing apparatuses are disclosed herein with reference to the drawings, wherein:

[0011] FIG. 1 is a perspective view of an apparatus for operating and controlling a plurality of release testing vessels;

[0012] FIG. 2 is an exploded view of a first embodiment of a paddle and a testing vessel;ConfidentialMERCK-0039PCT

[0013] FIGS. 3A-3C are schematic illustrations showing the use of a testing apparatus as a vessel fills with solution.

[0014] FIGS. 4A-4B show schematic perspective illustrations of a testing apparatus having a movable assembly.

[0015] FIG. 4C illustrates potential cross-sectional shapes for a central lumen of the movable assembly of FIG. 4A.

[0016] FIGS. 5A-5C are schematic illustrations showing the use of a testing apparatus with a movable assembly as a vessel fills with solution.

[0017] FIGS. 6A-7B illustrate some variations in the disks and / or blades of a movable double paddle apparatus.

[0018] FIGS. 9-12 illustrate and summarize experimental data relating to biphasic dissolution in a gastrointestinal simulator to forecast in vivo performance of certain drugs.

[0019] Various embodiments are described below with reference to the appended drawings. It is to be appreciated that these drawings depict only some embodiments of the disclosure and are therefore not to be considered limiting of its scope.DETAILED DESCRIPTION OF THE INVENTION

[0020] Despite the various improvements that have been made to dissolution testing, conventional devices and methods suffer from some shortcomings. Therefore, there is a need for further improvements to the devices and methods used to improve upon and advance the methods of dissolution testing. Among other advantages, the present disclosure may address one or more of these needs.

[0021] As used herein, the term “proximal,” when used in connection with a component of an apparatus, refers to the end of the component closer to the drive unit; whereas the term “distal,” when used in connection with a component of an apparatus, refers to the end of the component farthest from the drive unit. Likewise, the terms “trailing” and “leading” are to be taken as relative to the drive unit. “Trailing” is to be understood as relatively close to the drive unit, and “leading” is to be understood as relatively farther away from the drive unit. Moreover, as used herein, the terms “medicament,” “medication,” and “drug” are used generically interchangeably and it will be understood that the drugs described may include biologies, therapeutics, medicaments, topical ointments, and the like.ConfidentialMERCK-0039PCT

[0022] A generic testing apparatus for testing the rate of drug release resulting from diffusion is depicted in FIG. 1. As shown, a multi- vessel testing apparatus 10 includes, in this embodiment, several mixing vessels 11, each with a fitted cover 12 and a shaft 13 for rotating a stirring paddle. Testing apparatus 10 is coupled to controls 15, 16 to monitor the performance of the various parts, keep the temperature in each vessel constant, rotate the shaft at a constant speed, and / or periodically sample the medium in each vessel. Additional details of each mixing vessel are shown in FIG. 2, which shows vessel 11 to be stirred by paddle stirrer 14 which is affixed to, and orthogonally disposed with respect to, the distal end of shaft 13. Stirrer 14 may be rotated in direction “R” (e.g., clockwise or counterclockwise) by a motor or other mechanism in the test apparatus 10 (not shown).

[0023] Turning to Figs. 3A-3B, a testing apparatus 10 is shown including vessel 11 and paddle stirrer 14. Vessel 11 may be used to mix a material of two layers. For example, vessel 11 may be used to mix an organic phase 20 and an aqueous phase 30. Initially, paddle stirrer 14 is disposed near the bottom of vessel 11, and a distance DI is defined between stirrer 14 and the boundary “B” of organic phase 20 and aqueous phase 30 as shown in Fig. 3 A. A nozzle 19 may continue to introduce solution during the dissolution test. This introduction of aqueous solution 30 may result in an increase in the aqueous phase and rising of the organic phase, and this may increase the distance between stirrer 14 and the boundary “B” to distance D2 in Fig. 3B, and to distance D3 in Fig. 3C due to the stirrer being disposed at a fixed axial position. Without being bound by any particular theory, it is believed that a large distance between stirrer 14 and the boundary “B” may result in suboptimal stirring and / or poor results of dissolution testing used to predict bioperformance of oral test dosage forms.

[0024] Fig. 4A illustrates one example of a movable double paddle apparatus 100 according to the present disclosure. As shown, apparatus 100 includes a paddle stirrer 114 similar to stirrer 14. In this example, an axially movable assembly 120 is coupled to a central shaft 113. Assembly 120 may include a disk 121, an upper blade 122 and a lower blade 124, the upper and lower blades 122,124 being connected by a stem 125. Stem 125 may be sized as desired so that the distance between the two blades is sufficient to allow them to straddle a mixing boundary as will be discussed in greater detail below. Two blades 122,124 are shown, although it will be understood that a single blade may be used, or that more than two blades (e.g., three, four, or five blades) are possible. Assembly 120 may include a central lumen 127 that extends from the assembly’sConfidentialMERCK-0039PCT proximal end 130 to its distal end 132 to accept a shaft or column, such as central shaft 113. In some examples, central lumen 127 may be non-cylindrical (i.e., central lumen 127 may have a square, triangular, cruciform, diamond, or other similar uniform or irregular cross-section, etc. as shown in Fig. 4C) so that rotation of central shaft 113 also rotates assembly 120, and with it blades 122,124. In some examples, central shaft 113 and central lumen 127 have complementary shapes. Central lumen 127 may be slightly oversized with respect to central shaft 113 so that assembly 120 can slide or translate up and / or down along central shaft 113, as needed. In some examples, a first length, LI may be defined between the stirrer 114 and lower blade 124, and the first length may be between 1 cm and 3 cm. In some examples, this length LI may be modified based on the volume of organic phase 20 and / or aqueous phase 30.

[0025] Thus, assembly 120 may be mounted as a sleeve on central shaft 113 and may be axially translatable relative to central shaft 113 and / or stirrer 114, but not rotatable relative to central shaft 113 and / or stirrer 114. Fig. 4B illustrates the same apparatus where assembly 120 has translated upward along central shaft 127. In some examples, a second length, L2 may be defined between the stirrer 114 and lower blade 124, and the second length may be between 1 cm and 15 cm. In some examples, assembly 120 may translate over a length of between 1 cm and 15 cm. Assembly 120 may be formed of relatively low-density material (e.g., less than 1 g / cm3) (e.g., powder nylon, wood, cork, bamboo, styrofoam, certain plastics, such as PLA and the like, foam, rubber, fiberglass, aluminum, titanium, etc.) that is capable of floating in an aqueous solution. In some examples, the density of the material is in the range of 1010-1190 kg / m3.

[0026] In use, testing apparatus 100 may be used in a manner similar to testing apparatus 10 by being placed within a vessel 11 with paddle stirrer 114 being disposed near the bottom of vessel 11. In this example, assembly 120 may be disposed over, and coupled to, central shaft 113. Assembly 120 may be formed of a low-density material that floats so that disk 121 always remains at least partially above organic phase 20. Thus, in Fig. 5A, disk 121 is disposed at least partially above disk 121, and upper and lower blades 122,124 are spaced to straddle the boundary “B” of organic phase 20 and aqueous phase 30. As nozzle 19 continues to introduce solution during the dissolution test, there is an increase in the aqueous phase and rising of the organic phase. However, assembly 120 may translate relative to central shaft 113 and float upward so that disk 121 remains at least partially above organic phase 20 (Fig. 5B) with upper and lower blades 122,124 continuing to straddle the boundary “B” of organic phase 20 and aqueous phase 30. This continues even afterConfidentialMERCK-0039PCT the full introduction of solution from nozzle 19 (Fig. 5C). Thus, by having a floating assembly 120, the position of upper and lower blades 122,124 relative to the boundary “B” may be maintained, and sufficient mixing or agitation of organic phase 20 and aqueous phase 30 may be maintained. A passive system has been described where the movable assembly 120 floats in response to the contents of the vessel 11 without any action needed from the user or the motor. Alternatively, more complex configurations are contemplated where the assembly 120 may be actively moved up and / or down relative to the shaft by an actuation mechanism (e.g., gears, rack- and-pinion, a pullwire, etc.) in response to the filling of vessel 11. In any case, movement of assembly 120 may maintain a constant stirring rate at the boundary even as additional solution is introduced.

[0027] Variations are possible. For examples. Figs. 6A and 6B illustrate two axially movable assemblies 620A and 620B. Fig. 6A illustrates a small paddle assembly 620A and Fig. 6B illustrates a relatively larger paddle assembly 620B. Notably, disk 621B and blades 622B,624B of paddle assembly 620B are wider than disk 621 A and blades 622A.624A of paddle assembly 620A. Additionally, paddle stirrers 614A,614B may have different shapes and / or sizes as shown. Fig. 6C illustrates a top view of axially movable assemblies 620A and some relative sizes of central shaft 613. Additionally, a disk may include features to help it float. For example, Fig. 7A illustrates a disk 721 A with a recessed upper surface 722, and Fig. 7B illustrates another disk 72 IB with a hollowed inner chamber 724.

[0028] Experimental Data

[0029] Experiments were carried out using the axially movable assemblies disclosed herein. The gastrointestinal simulator alpha (GIS-a) is an in vivo predictive transfer dissolution method that mimics the pH changes and peristalsis in the gastrointestinal tract that are necessary in the biorelevant dissolution of drugs under the Biopharmaceutics Classification System (BCS) class II and IV. It can be used to provide increased understanding to the dissolution, precipitation, and supersaturation of various low- solubility drugs, but lacks insights on absorption. Conducting experiments in the GIS-a with an added biphasic format in the jejunal compartment not only improves the overall observed dissolution but also adds an absorptive sink that can be used to simulate the absorption that happens in the intestinal walls. In this study, the objective was to evaluate the dissolution of four representative BCS class II drugs using the GIS-a with the biphasic format. A customized paddle according to the present disclosure was also used in the jejunalConfidentialMERCK-0039PCT chamber. This paddle floats in the organic layer as the aqueous volume increases and maintains the hydrodynamics in both the aqueous and organic phases. The combination of the biphasic GIS- a and the moving paddle resulted to improved dissolution profiles of fenofibrate, danazol, and celecoxib while not affecting that of ritonavir. Incorporating these dissolution profiles in a PBPK model using GastroPlus® also resulted to pharmacokinetic parameters that correlate well with observed clinical values. Overall, this methodology considers both dissolution and absorption and proves to be a useful tool in predicting the in vivo performance of low- solubility drugs.

[0030] 1. Introduction

[0031] Dissolution testing is one of the key methods in the biopharmaceutics toolkit as it allows for prediction of clinical release profiles of oral formulations without relying too heavily on animal testing. The compendial dissolution tests approved by the USP, such as the USP I and II. are considered as the gold standard in dissolution testing. However, they are usually used as a quality control test for oral dosage forms and thus, are most meaningful in a commercial environment of a finished product. Dissolution testing of newly discovered APIs is still a challenge to formulation scientists. Throughout the years, dissolution tests have evolved, and various improvements have been incorporated into models so they could better predict in vivo outcome. However, these tests usually fail to replicate the complex in vivo physiology which limits in vivo predictability. Add to this the increasing complexity of new molecules being discovered as a result of advances in the field of combinatorial chemistry and high-throughput screening. Around 70% of these new compounds belong to the BCS class II and another 20% in BCS class IV. Both classes have very low aqueous solubility, but BCS class II compounds have higher permeability than the BCS class IV compounds. These properties dictate the behavior and bioavailability of an oral drug.

[0032] The dissolution method must be able to maintain sink conditions. For low solubility compounds, this necessitates the use of very large dissolution media volumes, which is very hard to handle and impractical. Other strategies that have been employed include adding different types of surfactants, using co-solvents like alcohols and propylene glycol to enhance solubility, and changing the pH for pH-sensitive compounds. The caveat is that these conditions are no longer representative of physiological conditions and would not be satisfactory in establishing in vitro-in vivo correlations. The biphasic model of dissolution was introduced as a means to mimic the redissolution and absorption process that happens to dissolved substances in the GI tract. This system includes the use of an organic solvent that is not miscible with water. Upon dissolution of theConfidentialMERCK-0039PCT lipophilic drug, the dissolved particles could freely partition into the organic layer, thus preventing saturation, allowing re-dissolution of precipitated particles, and effectively maintaining sink conditions

[0033] In this study, the multi-compartment transfer dissolution system, gastrointestinal simulator alpha (GIS-a) was equipped with a 1-decanol layer to serve as the absorptive phase in the jejunal chamber. The GIS-a is able to model important factors of the GI tract such as the variable pH in different chambers, the transfer kinetics, the rate of gastric emptying, and the luminal fluid volume in the different chambers. Past work indicates that in vivo plasma profiles of BCS II compounds can be predicted better by the GIS-a than the USP II. Building up on that information, a biphasic capability was added in order to simulate drug absorption, making it closer to physiological conditions. The dissolution profiles of four BCS class II compounds generated in this biphasic format will then be used as input in the physiologically based pharmacokinetic (PBPK) model using GastroPlus®. Results showed that there is an improvement in the observed release profiles of fenofibrate, danazol, and celecoxib while the dissolution profile of ritonavir remained similar. Consequently, this improvement led to better in vivo prediction in GastroPlus®. This methodology could be applied to other low solubility compounds especially if they are not performing well in compendial systems.

[0034] 2. Materials and Methods

[0035] 2.1. Materials. Fenofibrate (134 mg) and danazol (100 mg) capsules were obtained from Glenmark Pharmaceuticals USA Inc. (Mahwah, NJ) and Teva Pharmaceuticals USA, Inc. (Fairfield, NJ), respectively. Celecoxib capsule (400 mg) and ritonavir tablet (100 mg) were obtained from Aurobindo Pharma USA, Inc. (East Windsor, NJ). Acetonitrile, trifluoroacetic acid, phosphoric acid, hydrochloric acid, sodium dihydrogen phosphate monohydrate, sodium chloride, 1-decanol, and methanol were purchased from Fisher Scientific Inc. (Pittsburgh, PA). All chemicals were of analytical grade or HPLC grade, unless otherwise specified.

[0036] 2.2. Preparation of Dissolution Media. Simulated gastric fluid (SGF) was prepared by making a 0.01 M HC1 solution from a 0.1 M stock solution of HC1. The two times concentrated simulated intestinal fluid (SIF) consisted of 100 mM Na2HPC>4 at pH 6.8 with 30 mM NaCl. The components were weighed out and dissolved in appropriate volume of Milli-Q® purified water to yield the desired concentration and the pH was adjusted to 6.8 using 10N sodium hydroxide solution (Fisher Sicentific, Somerset, NJ). Stock solutions were used within 1 week of preparation.ConfidentialMERCK-0039PCT

[0037] 2.3. Dissolution using the GIS-a. Briefly, three compartments representing the stomach, duodenum, and jejunum were used. The gastric chamber was filled with 50 mL of SGF and 250 mL of water as the dose volume, representing the standard clinical protocol of administering an 8 oz glass of water along with the dose. The duodenal chamber was filled with 50 mL of SIF, and the volume was kept constant throughout the experiment. The jejunal chamber collected the output coming from the duodenal chamber. During the experiment, SGF and 2x SIF were kept at separate compartments and pumped into the gastric and duodenal chambers, respectively, at a rate of 1 mL / min, to maintain their pH environment. The solution transfer rate was programmed with the first-order like kinetics as reported previously with the gastric half-emptying time set to 8 min. The paddle speed was set at 100 rpm in the stomach but was kept at 50 rpm in the duodenal and jejunal chamber. The whole set up was kept at 37° C using a water bath. To initiate the experiment, the dosage form was dropped into the stomach chamber. A spiral sinker was used for dosage forms in capsule form to prevent them from floating. A 750 pL sample was taken from all three chambers at 5, 10, 20, 30, 60, and 90 min. The samples were then spun at 16400 x g for 1 min. Afterwards, 500 pL of the supernatant was taken and mixed with 500 pL of methanol to prevent precipitation. The samples were then analyzed using HPLC.

[0038] 2.4. Biphasic Dissolution using the GIS-a. The conditions were kept the same as indicated above with a few modifications in the jejunal compartment. Initially, the jejunal compartment was filled with 100 mL water and 150 mL of 1 -decanol. This initial condition was designed to prevent the premature mixing of aqueous components coming from the duodenum with the organic phase and helps to maintain the interface between the immiscible phases. A custom paddle similar to those of Figs. 7A-7B was also used. This paddle was designed to have a moving paddle that follows the interface between water and 1 -decanol as the water layer collects from the duodenal chamber. This ensures proper mixing and hydrodynamics is maintained on both immiscible layers. The whole paddle was 3D printed using powdered Nylon which held well in both water and 1 -decanol during our tests. The speed of the paddle was kept at 50 rpm to avoid turbulent mixing in both phases. Samples were taken at the same time intervals as stated above from all three compartments plus the organic layer. The aqueous samples were then spun at 16400 x g for 1 min and the supernatant was diluted 1:1 with methanol prior to HPLC analysis. The 1 -decanol samples of fenofibrate, danazol, and celecoxib were diluted 20x with a 1:1 mixture of methanol and waterConfidentialMERCK-0039PCT prior to HPLC analysis. The 1 -decanol samples of ritonavir were directly analyzed at 240 nm using a UV-Vis Spectrophotometer.

[0039] 2.5. HPLC Analysis. A Waters HPLC system (Waters Corporation, Milford, MA) composed of Alliance 2696 and Waters 2487 UV detector controlled by Empower 3® software was used to analyze all samples. HPLC grade water with 0.1% phosphoric acid (Solvent A) and HPLC grade acetonitrile with 0.1% trifluoroacetic acid (Solvent B) were used as the mobile phase. Ten microliters of sample were injected and resolved in an Atlantis T3 (Waters Corporation, Milford, MA) 5 pm, 4.6 x 50 mm C18 analytical column. Gradient elution was employed from 40% to 90% solvent B for a period of 2 min then constant at 90% solvent B for another 2 min before going back down to 40% at a constant flow rate of 1 mL / min. The UV detector was set at 254 nm for all compounds. Standard curves for all drugs were also generated and used to quantify the area under the peaks for each compound. Each concentration was obtained from the area under the peak. The dissolved amount was determined based on the concentration and the volume in the vessel at the specific time. The percent drug dissolved was calculated based on the dissolved amount divided by the dose then multiplied by 100.

[0040] 2.6. UV-Vis Spectrometry. A spectrofluorometer plate reader (Spectramax Gemini, Molecular Devices) was used to analyze ritonavir samples in 1 -decanol. Briefly, the 1 -decanol samples obtained from each time point from the GIS-a were analyzed directly without further dilution. A spectral scan from 200 to 300 nm was made, and the maximum absorbance at 240 nm was obtained. The concentration was then obtained by fitting into a calibration curve from ritonavir standards, and the percent drug dissolved was calculated similar to the method described above.

[0041] 2.7. PBPK modeling using GastroPlus®. To compare the obtained dissolution profiles with the conventional GIS and the Biphasic format, GastroPlus® was utilized to predict the in vivo plasma concentrations. Briefly, the computational software, GastroPlus® 9.8.3 (SimulationPlus, Inc., Lancaster, CA), was ran using a Lenovo ThinkPad computer with Intel Core i5 processors. Known physicochemical and pharmacokinetic values for all four drugs are listed in Fig. 8 and was used in GastroPlus® alongside the experimental dissolution data. This is a combination of physicochemical properties of each drug available in the literature and pharmacokinetic parameters calculated using the PKPlus module within GastroPlus®. A standard physiological condition model within the software was used: Human PhysiologicaLFasted alongside the Opt LogD Model SA / V 6.1. The PBPK models are built with the PKPlus module within GastroPlus® using clinicalConfidentialMERCK-0039PCT i.v. data for celecoxib and p.o. data for danazol, fenofibrate, and ritonavir, that was available in the literature. Figure 8 shows physicochemical and pharmacological properties of each drug for GastroPlus® simulation.

[0042] 3. Results

[0043] Biphasic dissolution is usually carried out using compendial dissolution systems such as the USP II apparatus, with a few innovative approaches reported in the literature. This results to a constant volume ratio between the aqueous and organic phases. The GIS-a has been shown to be biopredictive, but it could still be improved to increase its in vivo predictive capability. One way to do this is to incorporate an organic layer to serve as an absorptive sink in the intestinal compartment. This would improve the modeling of not only the release and distribution kinetics but also the absorption kinetics of a drug. The biphasic dissolution profiles obtained for the model drugs, fenofibrate, danazol, celecoxib, and ritonavir are shown in the next section.

[0044] 3.1. In Vitro Biphasic Dissolution

[0045] The GIS-biphasic dissolution data for fenofibrate, danazol. celecoxib. and ritonavir obtained for this study are shown in Figure 9. The normal GIS-a dissolution were taken from a previously published study and used as a baseline reference. The solvent compatibility of the 3D- printed dynamic paddle was assessed by immersing the paddle in the aqueous buffer and 1 -decanol for a period of 2 hours with stirring. No relevant changes in the material were observed. Long term stability of the material was not conducted because it is outside the scope of this study. Only one paddle was used for all drug samples and trials and there was no significant contamination and / or difference in dissolution profiles between trials were observed.

[0046] Fenofibrate is a neutral drug belonging to the BSC class II due to its low solubility but high permeability. Because it is a neutral compound, the variability in pH in the GI tract has little to no influence on the overall dissolution of the drug. Fenofibrate is used to reduce elevated low-density lipoprotein cholesterol (LDL-C). total cholesterol, triglycerides, and apolipoprotein B, and also increases high-density lipoprotein cholesterol (HDL-C). Its active form is fibric acid and it is highly bound to plasma proteins (>99%). The dissolution profile of fenofibrate capsules in the GIS-a is shown in Figure 9A. Results show that most of fenofibrate got transferred in the 1-decanol layer upon reaching the jejunal chamber. A zero-order release profile could be observed but only ~3% was dissolved and partitioned in the 1-decanol layer after 90 mins of dissolution. The aqueous layers had <1% dissolved in them regardless of the pH.ConfidentialMERCK-0039PCT

[0047] Similarly, danazol is a neutral compound under BCS class TI. As a neutral compound, its solubility is not dictated by the pH in the environment, but rather it is influenced by its dissolution rate and the amount of time spent in the GI tract. Danazol has been used to treat endometriosis, fibrocystic breast disease, breast cancer, and hemophilia. The absolute bioavailability of danazol under fasted conditions has been determined to be 11 ± 5.2%. This low bioavailability can be explained partially by the slow dissolution rate of the drug as it passes through the GI tract. Figure 9B shows the biphasic dissolution profile of danazol obtained from the GIS-a. Up to -15% partitioned into the 1 -decanol layer, and less than 1% is in the aqueous media in all three compartments.

[0048] The benzene sulfonamide drug, celecoxib, is known to work by selectively inhibiting cyclooxygenase-2, which is one of the key enzymes responsible for prostaglandin synthesis. Celecoxib is a weak acid with a physiologically relevant pKa value of 10.7. As such, the degree of ionization will be affected by the pH of its environment, and it will have higher solubility at higher pH values. The higher amount of celecoxib dissolved in the intestinal compartments than in the stomach confirms this. Celecoxib has a very poor water solubility in acidic media, and the high pKa value is not reached in physiological conditions. The absorption is thus severely limited by its solubility resulting in variable bioavailability. The absolute bioavailability of celecoxib in humans has not been investigated but was reported as 22-40% when given as capsules to beagle dogs. Results show that celecoxib readily partitions to the organic layer with -7% dissolution (Figure 9C).

[0049] Ritonavir is an antiviral drug originally intended for treating HIV infections. More recently, its use has been extended to serve as P-glycoprotein and CYP3A4 inhibitor, as a combination therapy with anti-virals for treatment of hepatitis C, and is an ingredient in the COVID-19 treatment Paxlovid. Ritonavir is a weak base, therefore, its solubility is expected to be greater in the stomach chamber where pH is high. This can be observed with the obtained dissolution profile for ritonavir tablets. The addition of the 1-decanol layer did not improve the overall dissolution of ritonavir. The majority of the drug was dissolved in the stomach chamber, as shown in Figure 9D, and some precipitation happened when it entered the duodenum chamber. This largely contributed to the amount that transferred to the jejunum chamber where the 1-decanol is located.

[0050] 3.2. Comparison with Monophasic GIS-aConfidentialMERCK-0039PCT

[0051] Figure 10 shows the total percentage dissolved in the duodenal and jejunal compartments for the monophasic GIS-a and includes the percent dissolved in the 1-decanol layer for the biphasic GIS-a. The monophasic GIS-a experiments were done separately from the biphasic in order to effectively assess the advantage of using the biphasic format in the dissolution of BCS class II compounds. For fenofibrate, danazol, and celecoxib, the introduction of an organic layer to serve as an absorptive sink greatly benefitted their dissolution. The dissolved amount for fenofibrate increased from 0.2 to 3.2% which is a 16x improvement to its dissolution (Figure 10A). Danazol and celecoxib changed from 0.4 to 17.2% (43x) and 0.4 to 7.2% ( 18x), respectively (Figures 10B and 10C). On the other hand, there was no improvement observed for the weakly basic drug, ritonavir. Its dissolution decreased 1.5x from 15.2% in the monophasic to 10.3% in the biphasic format (Figure 10D). This could be explained by the observed precipitation of the drug due to the change in pH as it passed through the intestinal compartments. This limited the amount of drug that partitioned into the organic layer.

[0052] 3.3. In Vivo Predictions

[0053] The in vivo predictive capability of GIS-a was assessed by taking the dissolution profiles and using them as input to the PBPK modeling software, GastroPlus®. The release profile was assumed as a controlled release dissolution profile as previously done. The dissolution of an undissolved poorly soluble drug can be assumed to be equivalent to the drug release from a controlled release formulation. The PBPK model was also built using physicochemical properties available in the literature or predicted by the ADMET predictor (Fig. 8). In addition, i.v. data for danazol was fed into the PKPlus module within GastroPlus® to predict pharmacokinetic values. For fenofibrate, celecoxib, and ritonavir, only oral data are available in the literature and were used for the PKPlus module. This modeling approach is done not to accurately predict the observed plasma profiles but to estimate how the biphasic format could affect the model. Results of the GastroPlus® modeling are shown in Figure 11.

[0054] The biphasic experimental format improved the in vivo predictions for fenofibrate, danazol and celecoxib. This is likely due to the increased rate in dissolution from the incorporation of the absorptive phase. On the other hand, the plasma concentration vs time profile of ritonavir did not improve. The simulated curve for the biphasic GIS-a is similar to the curve generated from the monophasic GIS-a. The predicted pharmacokinetic parameters are listed in the table shown in Fig. 12 along with the reported clinical data. For fenofibrate, the Cmax value was slightlyConfidentialMERCK-0039PCT underpredicted at 2.80 pg / mL compared to 3.21 pg / mL reported in the literature. The model for danazol accurately predicted the Fa% but slightly underestimated the Cmax at 29.5 pg / mL. Nonetheless, both models for neutral compounds correlate well with the observed values. For the weakly acidic drug celecoxib, the overall distribution and elimination curve agreed well with the clinical data; however, all PK parameters were underestimated. Lastly, the dissolution of ritonavir remained the same despite the addition of the organic layer and the predicted PK parameters were all underestimated as well.

[0055] 4. Discussion

[0056] Dissolution apparatuses, like the GIS-a, remain to be one of the key toolkits in biopharmaceutics that allows for formulation optimization of a drug product before it enters animal testing. In the biphasic format, key aspects that need to be considered include the type of organic solvent used, volume ratio between the organic and aqueous (phases? — feels like a noun is missing), absorption surface area, and hydrodynamics. Several organic solvents have been used in the past like hexane, cyclohexane, 1 -octanol, 1 -nonanol, and 1 -decanol, to simulate the drug absorption through the gut wall. In this study, 1 -decanol was chosen as the organic solvent because of its lower aqueous solubility and vapor pressure compared to the more commonly used 1-octanol. 1 -Decanol is also more manageable to use because it has less of a noticeable bad smell than 1- octanol. A constant organic volume of 150 mL was maintained throughout the experiment making the volume ratio (VR) change from 1.5 at the start to 0.3 after 90 min. Previous biphasic systems reported in the literature have volume ratios ranging from 0.2 to 1.0. Generally, the amount of drug dissolved in the aqueous phase remains the same regardless of volume ratio used, however, more precipitation can be observed at lower volume ratios which could affect the amount that partitions into the organic phase.

[0057] The physiological human absorption surface area per 100 cm3can be estimated to be between 1.9 cm’1and 2.3 cm4, considering tubular compression. This is hard to achieve using in vitro dissolution vessels that are rigid and limited in size. The surface area to volume ratio in the jejunal chamber used in this study ranges from 0.31 cm4to 0.12 cm’1. Despite this flaw in the design, there is improvement in the dissolution profiles obtained using the biphasic format in the GIS-a apart from the weak base, ritonavir (Figure 10). The 3D-printed moving paddle was designed to improve the hydrodynamics in the jejunal chamber and to help account for the significantly lower contact surface area in the organic phase. The improved mixing with theConfidentialMERCK-0039PCT movable paddle ensures sufficient mass transport between the aqueous and organic interface. This is evident with the results obtained for fenofibrate, danazol, and celecoxib.

[0058] As an overall assessment of this biphasic-GIS-a format, GastroPlus® was utilized in order to predict in vivo performance, using biphasic dissolution data as input. Results are shown in Figure 11 and were also compared to existing clinical data for further validation. Without the biphasic format, only the dissolution data for ritonavir was able to come close to what was observed in vivo. The other three drugs quickly reached saturation point, and less than 5% was dissolved. The addition of the organic layer allowed fenofibrate, danazol, and celecoxib to solubilize past their saturation concentrations. On the other hand, the dissolution of ritonavir remained the same in the two formats. This lack of improvement could relate to the pH dependence of the solubility of ritonavir; ritonavir was more soluble in the acidic environment of the stomach chamber but precipitated out in the higher pH environment in the duodenal chamber. The decreased drug concentration by ritonavir precipitation in the duodenal chamber reduced the drug solution to the jejunal chamber and resulted in no difference in the drug amount in the aqueous and the organic phases. Overall, the combined predictive power of the biphasic GIS-a and the PBPK modeling in GastroPlus® were able to predict the pharmacokinetic parameters, like Cmax and AUC, that correlate well with observed clinical values.

[0059] 5. Conclusion

[0060] In the pharmaceutical industry, it is desirable to choose the most suitable tool to provide reliable in vivo predictions in humans, not only to cut costs but also to increase efficiency in the lab. Various tools are available in the literature but knowledge of which one to use is convoluted as there is no single tool and methodology that can be applied to everything. The same is true for the GIS-a. Despite its capabilities and it being more biorelevant than compendial apparatuses, it still fails to accurately predict in vivo performance of low- solubility drugs under the BCS class II and IV on its own. There is still a need to innovate and optimize the parameters used in the experiment. Nonetheless, this study shows that the addition of an organic layer to act as an absorptive phase, improved the observed dissolution of fenofibrate, danazol, and celecoxib while not affecting that of ritonavir. The 3D-printed moving paddle was able to provide the desired hydrodynamics in the jejunal chamber, both in the aqueous phase and the organic phase. This paddle has the advantage of adapting to the change in fluid volume in jejunal chamber as more fluid gets transferred from the duodenal chamber. The use of 1 -decanol also proved to be effectiveConfidentialMERCK-0039PCT in acting as an absorptive sink with fewer drawbacks than the commonly used 1 -octanol. For the neutral drugs, fenofibrate and danazol, there was a 16- and 43-fold increase in the dissolution, respectively. On the other hand, there was an 18-fold increase in the dissolution of the weakly acidic drug, celecoxib and about a 1.5-fold decrease for the weakly basic drug, ritonavir. The predictions using GastroPlus® and the biphasic dissolution data closely replicated what was reported in the clinic. The bioavailability of danazol was accurately predicted using this method. Overall, the biphasic GIS-a format using the movable paddle can provide more accurate in vivo prediction of BCS class II drugs, especially for neutral and weakly acidic drugs. Given that both dissolution and absorption are key parameters that drive the bioavailability of a drug, it is important to keep developing and optimizing bio-predictive tools like the GIS-a.

[0061] It is to be understood that the embodiments described herein are merely illustrative of the principles and applications of the present disclosure. For example, the number and / or sizes of the blade may be varied. Moreover, certain components or steps of a method of using the device are optional, and the disclosure contemplates various configurations and combinations of the steps disclosed herein. Additionally, as used herein, the term “coupleable” refers to two or more components that cooperate, join or engage one another. It will be understood that where two or more components are said to be “coupled” or “coupleable” that they may also be unitarily or integrally formed. It is therefore to be understood that numerous modifications may be made to the illustrative embodiments and that other arrangements may be devised without departing from the spirit and scope of the present disclosure as defined by the appended claims.

[0062] It will be appreciated that the various dependent claims and the features set forth therein can be combined in different ways than presented in the initial claims. It will also be appreciated that the features described in connection with individual embodiments may be shared with others of the described embodiments.

Claims

ConfidentialMERCK-0039PCTCLAIMS1. A testing apparatus, comprising: a central shaft; and a movable assembly defining a central lumen for receiving the central shaft, a disk, and at least one blade, the movable assembly being axially translatable relative to the central shaft.

2. The testing apparatus of claim 1, wherein the movable assembly comprises a material that includes polyamide(nylon).

3. The testing apparatus of claim 1, wherein the movable assembly comprises a material that has a density of less than 1190 kg / m3.

4. The testing apparatus of claim 1, wherein the at least one blade comprises an upper blade and a lower blade.

5. The testing apparatus of claim 1, wherein the upper blade and the lower blade are spaced apart by a stem defining a first distance.

6. The testing apparatus of claim 1, wherein the central shaft terminates in a stirrer.

7. The testing apparatus of claim 1, wherein the central lumen is non-cylindrical.

8. The testing apparatus of claim 7, wherein the central lumen has a square- shaped cross-section.

9. The testing apparatus of claim 1, wherein the central lumen and the central shaft have complementary shapes.

10. The testing apparatus of claim 1, wherein the central lumen is rotationally fixed relative to the central shaft.

11. A method of dissolution testing comprising: providing a testing apparatus including a central shaft, and a movable assembly defining a central lumen for receiving the central shaft, a disk, and at least one blade; placing the testing apparatus within a vessel; and moving the assembly relative to the central shaft.

12. The method of claim 11, wherein moving the assembly relative to the central shaft comprises passively moving the assembly by allowing it to float.ConfidentialMERCK-0039PCT13. The method of claim 1 1 , wherein the movable assembly comprises a material having a density of less than 1190 kg / m3.

14. The method of claim 11, further comprising introducing a first material and a second material to the vessel, the first material and the second material defining a boundary therebetween.

15. The method of claim 11, further comprising introducing an aqueous phase and an organic phase to the vessel.

16. The method of claim 14, wherein providing at least one blade comprises providing an upper blade and a lower blade.

17. The method of claim 16, further comprising the step of positioning the upper blade above the boundary and positioning the lower blade below the boundary.

18. The method of claim 11, further comprising the step of rotating the central shaft and the at least one blade with a motor.

Images