Pharmaceutical preparations

Abstract

Formulating the SGLT2 inhibitor, bexagliflozin, as a sustained release tablet can improve its pharmacokinetic profile. These tablets can allow for lower peak plasma concentrations C max while maintaining plasma concentrations at therapeutic levels for the desired period of time. For example, a lower dose can be administered while providing the same pharmacological effect.

Description

[0001] Cross-reference to related applications

[0002] This application claims priority to International Application No. PCT / CN2020 / 119816, filed on October 5, 2020, which is incorporated herein by reference in its entirety for all purposes. Technical Field

[0003] This invention provides pharmaceutical formulations of SGLT2 inhibitors for the treatment of diabetes and other diseases, particularly oral formulations of bexagliflozin with improved pharmacokinetic properties. Background of the Invention

[0005] Bexagliflozin (EGT0001442, EGT1442, THR1442, THR0001442) is SGLT2 (renal sodium) + Inhibitors of glucose transporter (GLP) can be used to treat and manage a variety of diseases, including diabetes (see: Zhang et al. (2011) Pharmacol Research 63(4): 284-93; Allegretti et al. (2019) American Journal of Kidney Disease 74: 328 doi: 10.1053 / j.ajkd.2019.03.417; Zhang et al. (2019) Xenobiotic doi: 10.1080 / 00498254.2019.1654634). Its oral solid dosage form (e.g., see NCT01377844 or NCT01029704) and oral solution dosage form have been clinically tested in humans and have been shown to be well-tolerated, providing sustained and clinically meaningful improvements in glycemic control, as well as reducing weight and blood pressure in adult patients with diabetes (Halvorsen et al. (2019) Diabetes Obes Metab doi:10.1111 / dom.13833, Halvorsen et al. (2019) Diabetes Obes Metab 21:2248 doi:10.1111 / dom.13801). Invention Overview

[0007] Human studies of besagliflozin administered in oral capsule or oral solution form have shown that the peak-to-trough ratio (C0.05) of plasma concentration in subjects receiving besagliflozin is significantly higher. max Compared to C min and C max Compared to C 24hThe peak plasma concentration (Cp) is relatively high, decreasing sharply in phase α. The inventors discovered that formulating besagliflozin into sustained-release tablets yields better pharmacokinetic characteristics. Compared to standard immediate-release formulations, these tablets are administered at lower doses but still provide the same pharmacological effects (peak plasma concentration Cp). max Lower (while maintaining therapeutic plasma concentrations) and reduced potential side effects for a given dose. Adverse drug reactions are well known in the art, especially unpredictable idiosyncratic reactions (which are often not detected during pre-approval trials but, if discovered, can lead to restrictions or withdrawal of the approved drug) are more likely to occur with drugs that must be administered in high doses, and the likelihood of adverse reactions usually increases with C. max The amount increases with the amount of [something]. Therefore, in a first aspect, the present invention provides a sustained-release tablet of besagliflozin.

[0008] In fasting subjects, the preferred tablets of the first aspect released besagliflozin in vivo, resulting in plasma C1 per milliliter of besagliflozin. max This is at least 125,000 times lower than the total content of besaggliflozin per tablet. Therefore, for example, a tablet containing 20 mg of besaggliflozin on an empty stomach has a lower C... max ≤160 ng / mL. Ideally, plasma C max It contains at least 135,000 times less than besagliflozin (i.e., the C content of a tablet containing 20 mg of besagliflozin is lower). max ≤148 ng / mL), or even at least 145,000 times lower (i.e., C of a 20 mg tablet). max ≤138ng / mL).

[0009] According to a second aspect, the present invention provides a sustained-release tablet containing 10 mg to 20 mg of bexaggliflozin, and in a fasting subject, plasma C... max ≤160 ng / mL. Ideally, C max ≤133 ng / mL. In one embodiment, the tablet contains 10 mg of besagliflozin, with a plasma C... max ≤80 ng / mL; in another embodiment, the tablet contains 20 mg of besagliflozin, with a plasma C... max ≤160ng / mL.

[0010] According to a third aspect, the present invention provides a sustained-release tablet containing 30 mg to 60 mg of bexaggliflozin, and in a fasting subject, plasma C... max <400 ng / mL. In one embodiment, the tablet contains 40 mg of besagliflozin, with a plasma C... max <320 ng / mL; in another embodiment, the tablet contains 50 mg besagliflozin, whose plasma C max<400ng / mL.

[0011] For the first and second aspects, the preferred tablets contain 20 mg of bexaggliflozin, and the AUC in fasting subjects is [not specified]. 0-t At 600 and 1200 ng h mL -1 Between. Similarly, tablets containing 20 mg besaggliflozin are preferred, and the AUC in fasting subjects is [missing information]. 0-∞ At 675 and 1275 ng h mL -1 between.

[0012] For the first and second aspects, the preferred tablets contain 20 mg of bexaggliflozin, and in fasting subjects, plasma C... max Between 80 and 150 ng / mL.

[0013] For the first, second, and third aspects, it is preferred that the tablets be administered 24 hours after drug administration (i.e., C). 24h Besaglitazone plasma concentration ≥5 ng / mL, ideally ≥10 ng / mL.

[0014] For the first, second, and third aspects, the preferred time (T0) for the tablet to reach the maximum bexagliflozin plasma concentration in fasting subjects is... max The time is 2 to 6 hours, and ideally 2 to 4.5 hours.

[0015] As explained in detail below, the properties defined for tablets are typically measured after administration of a representative batch of sample tablets, and the appropriate average of the results (e.g., geometric mean) is calculated. With this in mind, the geometric mean plasma Ca of bexagliflozin released from a batch of tablets in a fasting subject is considered in the first aspect. max The total content of besagliflozin per tablet was at least 125,000 times / mL lower. Similarly, the geometric mean plasma Cg of a batch of tablets in the second aspect was lower in fasting subjects. max ≤160 ng / mL (e.g., geometric mean C max ≤133ng / mL); for example, a tablet containing 10mg bexaggliflozin has a geometric mean C max ≤80 ng / mL, or tablets containing 20 mg besagliflozin, with a geometric mean C max ≤160 ng / mL. Similarly, the geometric mean plasma C in a batch of tablets from the third aspect in fasting subjects. max <400 ng / mL; for example, a tablet containing 40 mg bexagliflozin has a geometric mean plasma C0.05. max <320 ng / mL, or tablets containing 50 mg bexagliflozin, with a geometric mean plasma C0.05 max<400 ng / mL. Furthermore, in subjects on an empty stomach, a batch of tablets containing 20 mg bexagliflozin from both the first and second aspects showed (i) geometric mean AUC. 0-t At 600 and 1200 ng h mL -1 Between, and / or (ii) geometric mean AUC 0-∞ At 675 and 1275 ng h mL -1 Between, and / or (iii) geometric mean plasma C max Between 80 and 150 ng / mL. Similarly, for the first, second, and third aspects, the geometric mean C of a batch of tablets... 24h ≥5ng / mL, ideally ≥10ng / mL.

[0016] For a batch of tablets according to the first, second and third aspects, bexagliflozin plasma C max and C min The ratio of the median values ​​can be less than 10, for example, between 5 and 10, between 6 and 8, or between 7 and 8. Therefore, the peak / trough ratio found in existing technologies can be avoided. Preferred C min The median was at least 10 ng / mL. As shown below, these pharmacokinetic parameters represent robust statistical estimates based on over 800 subjects worldwide taking various extended-release bexagliflozin tablets.

[0017] The inventors also observed that besagliflozin is a substrate of P-gp, and that the amount of besagliflozin absorbed in the large intestine is minimal. Since P-gp expression increases with increasing distance from the small intestine, absorption in the duodenum may be greater than in the ileum, and [ 14 [C] Quantitative mass balance studies of besagliflozin showed minimal colonic absorption (Zhang et al. Xenobiotica. 2019 Aug 27: 1-11. doi: 10.1080 / 00498254.2019.1654634). Since sustained-release tablets aim for high release of most of the drug in the small intestine, potential practical incompatibility may exist. The tablets of this invention advantageously incorporate an adaptation that helps retain the drug in the stomach. Therefore, the majority of the sustained release of besagliflozin can occur in the stomach, allowing absorption of the drug to occur in the ideal location in the small intestine, thus providing favorable pharmacokinetic characteristics. Even though besagliflozin is unstable in a long-term acidic environment and readily degraded by acid, tablets with gastric retention adaptation show good function in vivo.

[0018] Various modifiers can be used to help retain the tablets of the present invention in the stomach, including but not limited to: (i) adding an effervescent excipient that provides buoyancy during the release of gas by gastric acid; (ii) rapidly dispersing the tablet into multiple particles or spheres within the stomach to prevent the complete tablet from being expelled from the stomach in a single event; (iii) using a low-density excipient to provide buoyancy or a floating tablet; and / or (iv) adding a mucosal adhesive to the tablet. These four methods can be used alone or in combination to provide a favorable tablet for the delivery of bexagliflozin.

[0019] According to a fourth aspect, the present invention provides a sustained-release tablet containing besagliflozin and a mucosal adhesive. Ideally, such a tablet has a density lower than that of the gastric contents. It may also be effervescent (especially upon contact with gastric acid) and / or dispersed into multiple particles or spheres upon contact with gastric acid.

[0020] According to a fifth aspect, the present invention provides a solid oral dosage form, typically a sustained-release tablet, containing besagliflozin, and in an in vitro dissolution test in simulated gastric fluid (see below), releasing ≤17% besagliflozin after 1 hour and ≥80% after 8 hours. In one embodiment, it releases 20-45% (inclusive) of besagliflozin after 3 hours and / or 45-75% (inclusive) of besagliflozin after 5 hours. The tablet may be from a production batch that has passed formal dissolution acceptance criteria, as discussed below.

[0021] According to a sixth aspect, the present invention provides a solid oral dosage form, typically a sustained-release tablet, containing besagliflozin, and having an f2 value > 50 in an in vitro dissolution test in simulated gastric juice (see below) compared to a reference tablet, wherein f2 is proportional to 1 plus the decimal logarithm of the mean square error:

[0022]

[0023] In the formula: n is the number of time points for dissolution measurement; R i T is the dissolution percentage of the reference tablet at time point i; i It represents the dissolution percentage of the solid oral dosage form at time point i.

[0024] The reference tablet is a sustained-release tablet containing besagliflozin, which, in an in vitro dissolution test simulating gastric fluid, releases ≤17% besagliflozin at 1 hour, ≥80% at 8 hours, and optionally, 20-45% (inclusive) besagliflozin at 3 hours and / or 45-75% (inclusive) besagliflozin at 5 hours. Three suitable reference tablets, such as reference tablets (a) to (c), will be disclosed in more detail below, wherein tablet (c) is the preferred tablet. The preferred value of n is at least 3, for example, between 4 and 8.

[0025] According to a seventh aspect, the present invention provides a batch of besagliflozin extended-release tablets, wherein, when administered to healthy, fasting subjects, one scenario is that tablets from the first representative sample group of that batch produce C... max First logarithm and AUC 0-t The first average logarithm, another different case is the C produced by the second representative sample tablet from that batch. max The second logarithm of the mean and AUC 0-t The second logarithm of the mean, and C max The difference between the first and second logarithms and AUC 0-t The difference between the first and second logarithms of the mean both show 90% confidence intervals, with endpoints between -0.22314 and +0.22314. Detailed assessment of these parameters is given in the “Bioequivalence” section below, including, for example, the use of randomized crossover studies in appropriate trial populations. Ideally, each tablet in this batch contains 5 mg, 10 mg, or 20 mg of besagliflozin.

[0026] According to an eighth aspect, the present invention provides a batch of besagliflozin extended-release tablets. In one scenario, when administered to healthy subjects, a single tablet representing the first tablet sample group is provided to fasting subjects, and in another scenario, a single tablet representing the second tablet sample group is provided to subjects in a postprandial state (e.g., as described in the "Bioequivalence" section below and its references, 30 minutes after a standard high-fat, high-calorie meal), ln(C max ) and ln(AUC 0-t The average difference (from C in the postprandial state) max Logarithm and AUC 0-t Subtracting the fasting state from the logarithm max Logarithm and AUC 0-t The logarithmic values ​​all showed 90% confidence intervals, with endpoints ranging from -0.22314 to +0.58779. Ideally, each tablet in this batch contains 5 mg, 10 mg, or 20 mg of besagliflozin.

[0027] According to a ninth aspect, the present invention provides a batch of besagliflozin extended-release tablets. In one scenario, when administered to healthy, fasting subjects, each subject is provided with a single tablet from a first representative tablet sample group without prior administration of a parenteral GLP-1 receptor agonist; in another scenario, a single tablet from a second representative tablet sample group is provided to the subject 30 minutes after administration of an approved dose of a parenteral GLP-1 receptor agonist, ln(C max ) and ln(AUC 0-tThe average difference (from C of the second sample group) max Logarithm and AUC 0-t Subtract C from the first sample group in the logarithm max Logarithm and AUC 0-t The logarithmic values ​​all showed 90% confidence intervals, with upper limits less than 0.69315. Ideally, each tablet in this batch contains 5 mg, 10 mg, or 20 mg of besagliflozin.

[0028] According to a tenth aspect, the present invention provides a batch of besagliflozin extended-release tablets, wherein, in administration to healthy subjects, one scenario is that a single tablet from a first representative tablet sample group is administered on an empty stomach; another scenario is that a single tablet from a second representative tablet sample group is administered on a postprandial stomach (e.g., 30 minutes after a standard high-fat, high-calorie meal, as described in the "Bioequivalence" section below), with a T0 in the postprandial state. max Subtract T in fasting state max The median difference is less than or equal to 3.5 hours. The median difference is the difference between 50% of subjects having a median above the median and 50% of subjects having a median below the median. For example, in an ordered list of differences, for an odd number of subjects (e.g., 2n+1 subjects), the median is the difference at the midpoint of the list (subject n+1), and for an even number of subjects (e.g., 2n subjects), the median is the arithmetic mean of the differences between the two subjects on either side of the midpoint (subjects n and n+1). Ideally, each tablet in this batch contains 5 mg, 10 mg, or 20 mg of besagliflozin.

[0029] According to the eleventh aspect, the present invention provides a solid oral dosage form, typically a sustained-release tablet, containing besagliflozin, which produces first plasma C10 in fasting subjects. max The first AUC 0-t And the first T max A second plasma C was produced in subjects after a meal. max The second AUC 0-t And the second T max , where (i) the second C max The ratio divided by the first is between 0.8 and 1.8; (ii) the second AUC 0-t The ratio divided by the first is between 0.8 and 1.8; or (iii) the second T max The ratio divided by the first is between 0.8 and 3.0.

[0030] As described below, the properties defined for tablets are typically measured after administration of a representative batch of samples modeled after the tablets. Therefore, the eleventh batch of tablets provides the first geometric mean plasma C in fasting subjects. maxThe first geometric mean AUC 0-t And the first T max Median, and provides a second geometric mean plasma C in postprandial subjects. max The second geometric mean AUC 0-t And the second T max Median, where (i) the second geometric mean C max (ii) The ratio of the first to the second geometric mean AUC is between 0.8 and 1.8; 0-t The ratio divided by the first is between 0.8 and 1.8; or (iii) the second T max The ratio of the median to the first is between 0.8 and 3.0.

[0031] According to a twelfth aspect, the present invention provides a solid oral dosage form, typically a sustained-release tablet, containing besagliflozin, which generates first plasma C1 in subjects who have not previously received parenteral GLP-1 receptor agonists. max The first AUC 0-t And the first T max A second plasma C was generated in subjects who received parenteral GLP-1 receptor agonists. max The second AUC 0-t And the second T max , where (i) the second C max The ratio of the first to the second AUC is between 0.8 and 2.0; 0-t The ratio divided by the first is between 0.8 and 2.0; or (iii) the second T max The ratio divided by the first is between 0.8 and 3.0.

[0032] As described below, the properties defined for tablets are typically measured after administration of a representative batch of samples, modeled after tablets. Therefore, the first geometric mean plasma C1 in a batch of tablets from aspect 12 was obtained in subjects who had not received prior parenteral GLP-1 receptor agonist administration. max The first geometric mean AUC 0-t and the first median T max Furthermore, it produced a second geometric mean plasma C in subjects who received prior parenteral GLP-1 receptor agonists. max The second geometric mean AUC 0-t And the second T max Median, where (i) the second geometric mean C max (ii) The ratio divided by the first is between 0.8 and 2.0; the second geometric mean AUC 0-t The ratio divided by the first is between 0.8 and 2.0; or (iii) the second T maxThe ratio of the median to the first is between 0.8 and 3.0.

[0033] The present invention also provides a method for treating patients, which is discussed in more detail below. Attached Figure Description

[0034] Figure 1 The geometric mean plasma concentration (ng / mL) of besagliflozin in fasting subjects is shown as a function of time after administration (hours). Closed circles (●) represent data for 20 mg capsules, while other symbols represent 15 mg tablets (XR5), respectively. XR8 (△) or XR11 (○).

[0035] Figure 2 The dosage shown is 10 mg (●), 15 mg (○), or 30 mg. Geometric mean plasma concentration (ng / mL) of besagliflozin in fasting subjects taking the tablets.

[0036] Figure 3 shows the percentage of besagliflozin dissolved (vertical axis) after 1 hour (◆), 3 hours (■), 5 hours (▲), or 8 hours (X) in in vitro dissolution tests. Tablets were dissolved at 25°C (…). Figure 3A ) or 30℃ Figure 3B The storage time under these conditions is up to 60 months (x-axis). The figure shows the mean measured using the regression line method. Detailed Implementation

[0037] The present invention provides a sustained-release tablet, which improves the pharmacokinetic properties of besagliflozin compared to a capsule formulation.

[0038] Bexagliflozin

[0039] Besaglitazone is a C-aryl glucosinolate SGLT2 inhibitor with formula (I):

[0040]

[0041] Its IUPAC name is (2S,3R,4R,5S,6R)-2v(4-chlorov3-(4-(2-cyclopropoxyethoxy)benzyl)phenyl)-6v(hydroxymethyl)tetrahydro-2H-pyran-3,4,5-triol. Its CAS registration number is 1118567-05-7.

[0042] The tablets of the present invention contain besagliflozin, typically in crystalline solid form (e.g., see WO2011 / 153953). In some embodiments, besagliflozin may be present in the form of an ester (mono, di, tri, or tetra), but typically besagliflozin is used in the form of a tetraol of formula (I) as shown above. Furthermore, in some embodiments, besagliflozin may be present in a cocrystal form, such as a cocrystal with proline, as disclosed in WO2010 / 022313, “THR1474” (molar ratio of besagliflozin to proline is 1:2). These forms of besagliflozin may optionally be present in the tablets of the present invention as a solvate. The present invention includes all such forms of besagliflozin.

[0043] The amount of besagliflozin in the tablets of this invention is typically from 1 mg to 100 mg, preferably in the range of 5 mg to 50 mg (e.g., 10-20 mg in the second aspect of this invention). Tablets containing 5 mg, 10 mg, or 20 mg are particularly preferred. These values ​​are expressed as tetraols of formula (I). These sustained-release tablets of these strengths (especially 20 mg) have good therapeutic effects.

[0044] The specific content of besagliflozin in tablets mentioned should be understood within the general context of pharmaceutical formulations. Therefore, the content can be determined according to, for example, the United States Pharmacopeia General Principles. <905> The dosage uniformity is specified in the European Pharmacopoeia 2.9.40 or the Japanese Pharmacopoeia 6.02. Measurement. If the tablets are approved for medicinal use in a particular region, the relevant licenses, sales permits, prescribing information, product characteristics summary, product information, patient literature, etc., will specifically mention the content of besagliflozin, for example, in tablet strengths of 5 mg, 10 mg, 20 mg, 40 mg, or 50 mg.

[0045] The tablets of this invention may include impurities and / or degradation products related to besagliflozin. If so, the content of these should be ≤1.0% of the total mass of the besagliflozin tablets, and the content of any particular impurity or degradation product should be ≤0.20% of the total mass of besagliflozin.

[0046] General Rules

[0047] The term "comprising" includes both "including" and "consisting." For example, a composition that "comprising" X may consist entirely of X, or may include other things, such as X+Y.

[0048] The term "approximately" related to the numerical value x is optional; for example, it could refer to x ± 10%.

[0049] The word “substantially” does not exclude “completely”; for example, a composition that is “substantially free” of Y can be completely free of Y. The word “substantially” may be omitted when defining the invention if necessary.

[0050] The term “between” refers to two values, such as 10 mg and 20 mg. The range “between” includes 10, 15 and 20 mg.

[0051] Statistical pharmacokinetic parameters obtained from non-compartmental models are the most commonly used in this field, and are summarized as follows:

[0052] “T max "This refers to the time when the maximum plasma concentration was observed, and when referring to a population, unless otherwise specified, it is given as the population median."

[0053] “C max "This is the highest observed plasma concentration."

[0054] “C min "C" represents the lowest observed plasma concentration, a value typically obtained before repeated dosing in a regularly administered regimen. For example, in daily dosing, C min Results are usually obtained 24 hours after the previous dose.

[0055] "AUC" is the area under the curve of plasma concentration over time, obtained by the linear trapezoidal rule. According to this rule, AUC can be obtained by multiplying the sum of the arithmetic mean of the concentrations at two adjacent sampling points by the time difference between these sampling points: (C(t)) i )+C(t i+1 ))(t i+1 -t i ) / 2.

[0056] AUC 0-t "" represents the AUC from time 0 (e.g., intake time) to the final quantifiable concentration.

[0057] AUC 0-∞ "" represents the AUC from time 0 to infinity, obtained by extrapolation from a simple (single-phase) exponential decay. AUC 0-∞ =AUC 0-t +C last / k el C last It is the final quantifiable concentration, k. el It is the terminal elimination rate constant.

[0058] “t 1 / 2 "t" refers to the terminal half-life, also known as the elimination half-life. If the empirically determined terminal elimination kinetics are not first-order in time, then t cannot be defined. 1 / 2 T1 / 2 =-ln(2) / k el ≈0.693 / k el .

[0059] The terms “d(0.1)”, “d(0.5)”, and “d(0.9)” are the particle threshold diameters that fall within the minimum 10%, 50%, and 90% of the total volume of all particles, respectively. Therefore, in d(0.9), 90% of the sample volume can be found in particles with a diameter smaller than d(0.9).

[0060] The "logarithm" used here defaults to the natural logarithm. The function of parameter x is usually written as ln(x), where, for clarity, x = e^x. ln(x) If the base of the logarithm is 10, then the logarithm is called the decimal logarithm, and the function of the parameter y is written as logy. 10 (y), where, to avoid any ambiguity, y = 10 log10(y) .

[0061] The term "solid oral dosage form" in this article refers to any solid (or semi-solid) dosage form that can be administered orally. It can take the form of tablets, solid pills, capsules, capsules, encapsulated gels or encapsulated liquids, or combinations or combinations thereof, such as layers or subcomponents of various shapes and properties that may be embedded in a matrix or contained in capsules or capsules.

[0062] The size of a tablet “batch” can range from 100 tablets to a complete production batch (e.g., all tablets made from the same quantity of initial materials and subjected to a series of identical manufacturing operations, or any total number of tablets that have undergone similar manufacturing operations and are collected for testing or distribution). The definition of a “production batch” includes that provided by Section 201.3 of Title 21 of the United States Code, which defines a “batch of a specified quantity of a drug or other material that is made to have uniform properties and quality within specified limits and is produced within the same production cycle according to a single production order.”

[0063] When the term "representative" is applied to a unit or sample in a batch, it refers to a unit or sample that was not pre-selected for any specific characteristics (such as weight, density, hardness, or coating hue), has no manufacturing defects, and is essentially randomly drawn from the batch.

[0064] The phrase "substantially random" means completely random, such as when each unit in a batch has an equal probability of being selected, or when the selection is carried out through a process designed to achieve a practically balanced representation of the sampled batch. For example, representative units can be drawn at regular intervals during production or coating to avoid sample imbalance in units with slightly different properties, such as units produced at the beginning or end of a run being overrepresented. Such units can be said to be drawn substantially randomly from the batch.

[0065] As used herein, a “sample set” refers to a collection of units or samples that can be analyzed individually or collectively to assess the characteristics of the entire batch or population. When used for in vitro or in vivo testing of tablet properties, a sample set refers to a collection of individual tests from which the overall characteristics of the batch of tablets are assessed.

[0066] A characteristic defined for any particular unit (e.g., tablet) should be understood as the characteristic of a representative unit drawn from the production batch, whose members transmit or demonstrate a reference characteristic in appropriate tests, typically requiring the consumption of multiple units from the production batch. Therefore, when referring to a unit producing a specific pharmacokinetic parameter, it should be understood that this parameter is typically measured after administration of a representative sample from the production batch canonicalized by that unit, and the appropriate statistical characteristics of the result are calculated. Parameters based on bexagliflozin plasma concentrations (e.g., C...) max The AUC and T are usually expressed as geometric mean, while T max It is typically expressed as the population median. Furthermore, when a pharmacokinetic parameter is defined as having a range of values, it should be understood that administration of a representative sample of the production batch modeled by that unit to a suitably composed test cohort will produce a characteristic parameter (e.g., geometric mean or median) falling within that range of values.

[0067] For example, when referring to the statistical values ​​produced by tablets (e.g., geometric mean C) max When a value falls within a certain range, it should be understood that, within an appropriately composed cohort, administering a representative sample of production batches exemplified by that tablet will produce statistical values ​​(e.g., geometric mean C) falling within that range. max ).

[0068] A “suitably composed cohort” refers to a set of test subjects, typically consisting of a sample size of healthy men and women with sufficient capacity to assess the required pharmacokinetic parameters. A sample size providing sufficient capacity can be calculated as follows. In routine practice, for example, to demonstrate bioequivalence for regulatory purposes, twelve or more subjects of different sexes are typically selected; if sex imbalanced, the total sample size is 24 subjects. Such testing usually requires participants to abstain from alcohol and avoid foods known to significantly affect drug metabolism. Although not, for example, a regulatory requirement, it should be understood that, in order to determine whether the sample group represents the tablets of the present invention, the test cohort should generally consist of individuals close to the midpoint of a young, healthy adult population. Therefore, for example, the cohort should not include a large number of individuals who are overweight or underweight, abnormally thin or obese, elderly, or have special dietary habits, or are taking medications, herbal preparations, or supplements, as these could confound the test results.

[0069] The “sample size providing adequate power” used to determine pharmacokinetic parameters is the number of individuals in the cohort required to achieve a specific degree of differentiation between two experimental condition groups (e.g., taking tablets from one source or tablets from another). Methods for calculating statistical power are well-known in the art. In its simplest form, statistical power describes the probability of obtaining a statistically significant result in a study when a predicted difference actually exists between two populations. Power tests typically determine a minimum sample size to detect true between-group differences, which, due to randomness, may have a specific tendency to fail. For example, a 90% power means that in 9 out of 10 studies, a statistically significant result will occur, but in 1 study, even if a difference exists, it will be meaningless. Therefore, 100% minus the probability of a false negative. In the testing of pharmacokinetic parameters, a typical power value is 90% or greater; for clarity, “adequate power” is defined here as 95% or greater. To calculate power, the variability of the measured indicator, usually expressed as standard deviation, and the difference to be measured (the difference between the values ​​of the two groups of indicators being measured) must be entered. If there is significant uncertainty in the standard deviation of the indicator in the population, it can be determined empirically. When used to determine non-inferiority, efficacy calculations are used to determine whether the difference between the two groups is less than a certain required sample size. For example, bioequivalence studies are two-sided non-inferiority trials designed to demonstrate that the difference between two formulations is within a certain range.

[0070] When a meal status (e.g., fasting or postprandial) is specified, the fasting status is achieved by each subject refraining from consuming any food or beverage other than water for at least ten hours prior to tablet intake; the postprandial status is achieved by each subject consuming a standard high-fat, high-calorie diet as provided by regulatory guidelines (e.g., FDA Guidance for Industry: Bioavailability and Bioequivalence Studies Submitted in NDAs or INDs - General Considerations, March 2014), and taking the tablet 30 minutes after the meal. Further information on how to achieve these specific meal statuses is provided in the Bioequivalence section below.

[0071] Extended-release tablets

[0072] Besagliflozin has been used in human subjects in various dosage forms. In a radioactive tracer material balance study, healthy male volunteers were administered an aqueous solution containing 50 mg of besagliflozin, and the C... max 692 ng / mL -1 AUC 0-t 2523 ng h mL -1 AUC 0-∞ 2604 ng h mL -1 ;T max For 0.5h, t 1 / 2 The C10 time was 5.6 h (Zhang et al. (2019), ibid.). After dose standardization, the C10 time per mg of besagliflozin was 5.6 h (Zhang et al. (2019), ibid.). max 13.84 ng / mL -1 .

[0073] Oral capsule formulations of besagliflozin were well tolerated in healthy and diabetic subjects at single and repeated doses up to 100 mg. The capsules provide relatively rapid release of besagliflozin in vivo, but subsequent plasma concentrations showed a high peak / trough ratio. On a fasting state, capsules containing 6.7, 16.7, and 34 mg of besagliflozin showed high peak-to-trough ratios after dose standardization. max The concentrations were 12.6, 11.3, and 11.5 ng / mL, respectively. -1 mg -1 T max The median times were 1, 2, and 1 hour, respectively. Based on these values, for example, the C values ​​for capsules containing 20 mg besagliflozin in fasting subjects... max The concentration was between 226 and 252 ng / mL, and occurred approximately 1 to 2 hours after administration (i.e., T). max(Approximately 1 to 2 hours). Oral solutions have the fastest absorption rate and exhibit the lowest Tg. max and maximum dose standardized C max C contains capsules with 34mg of besaggliflozin. 24h Greater than 10 ng / mL. Plasma concentration drops sharply in the α phase (i.e., the distribution phase of the standard two-compartment model).

[0074] Compared to immediate-release capsules, the inventors have found that formulating besagliflozin into sustained-release tablets improves its pharmacokinetic characteristics. These tablets provide a lower C60 concentration. max (e.g., 8 ng / mL / mg bexaggliflozin or less), while a 20 mg tablet still maintains a C10 level of around 10 ng / mL. 24h Lower C max The risk of side effects is reduced, but the drug remains effective because it is effective for 24 hours after administration (i.e., C14 hours after administration). 24h In formulations exhibiting plasma concentrations of 10 ng / mL or higher, urinary glucose excretion is close to its maximum.

[0075] Therefore, a first aspect of the present invention provides a sustained-release tablet of besagliflozin.

[0076] Sustained-release (also known as long-acting or sustained-release) tablets release their contents in the body over a period of time after ingestion. Ideally, release should begin immediately after ingestion (e.g., once the tablet enters the stomach) and should not be delayed. Therefore, the tablets of this invention typically do not have an enteric coating, as this would result in delayed release.

[0077] The tablets of this invention should provide a single-peak plasma concentration of besagliflozin as a function of time (in most subjects). Therefore, after administration of a single tablet to a subject, the plasma concentration of besagliflozin should show only one peak (e.g., see...). Figure 1 and 2 ).

[0078] The tablets of this invention release bexagliflozin in vitro with essentially zero-order kinetics.

[0079] The plasma concentration of besagliflozin reaches C max It can then be reduced in a biphasic manner.

[0080] As mentioned above, in fasting subjects, plasma C... max Approximately 226 to 252 ng / mL, or C per milliliter of plasma. max The total content of besagliflozin in the capsules is 80,000 to 90,000 times lower. However, in the preferred tablets of the present invention, the C in fasting subjects... maxThe content of bexagliflozin should be at least 125,000 times lower than that in tablets. Therefore, the C of 20mg tablets... max ≤160 ng / mL. Ideally, C max The ratio of bexaglitazone to net content is even higher than 125,000 times, for example ≥135,000 times or ≥145,000 times.

[0081] Therefore, the present invention specifically provides a sustained-release tablet having a geometrically mean plasma Cg in healthy subjects in a fasting state (e.g., in a cohort of no fewer than 6 fasting subjects weighing more than 60 kg). max ≤8 ng / mL / mg besagliflozin (ideally ≤6 ng / mL / mg). In one embodiment, the tablet contains 10 mg besagliflozin, and C max ≤80 ng / mL; in another embodiment, the tablet contains 20 mg of bexaggliflozin, and C max ≤160ng / mL.

[0082] When the C of the tablets of the present invention max When the concentration is ≤160 ng / mL, it is preferably ≤150 ng / mL, and ideally between 80-150 ng / mL (especially for 20 mg bexagliflozin). For 20 mg tablets, C is preferred. max It is between 85-145 ng / mL, with C being more preferred. max Between 95-140 ng / mL.

[0083] The sustained-release tablets of the present invention showed plasma C-values ​​of bexagliflozin in fasting subjects. 2a ≥3 ng / mL. As mentioned above, plasma C 24h At a concentration of ≥10 ng / mL, urinary glucose excretion is close to its maximum; therefore, the preferred tablets of this invention can provide ≥10 ng / mL of plasma glucose. 24h For example, in the range of 10-25 ng / mL. In one embodiment, the tablet contains 10 mg of bexaggliflozin, and C 24h ≥3 ng / mL; in another embodiment, the tablet contains 20 mg of bexagglitazone, and C 24h ≥6ng / mL.

[0084] As stated above, the plasma Tg of besagliflozin capsule formulation in fasting subjects max Approximately 1 hour. Conversely, the preferred tablets of the present invention show a T in fasting subjects. max Typically, the time to release is between 2 and 6 hours. Therefore, compared to immediate-release capsules, the tablets of this invention can delay the T-release of besagliflozin. max .

[0085] In fasting subjects, the plasma AUC of the preferred tablet of the present invention is per milligram of bexagliflozin. 0-t 15-60 ng hmL -1 Between. In one embodiment, the AUC of a tablet containing 10 mg bexaggliflozin. 0-t 150-600 ng h mL -1 Between, for example, 350-450 ng h mL -1 Between; in another embodiment, the AUC of a tablet containing 20 mg besaggliflozin 0-t 600-1200 ng h mL -1 Between, for example, 650-1150 ng h mL -1 between.

[0086] In fasting subjects, the plasma AUC of the preferred tablet of the present invention is per milligram of bexagliflozin. 0-∞ In 17.5-65 mgh mL -1 Between. In one embodiment, the AUC of a tablet containing 10 mg bexaggliflozin. 0-∞ 410-510 ng h mL -1 Between; in another embodiment, the AUC of a tablet containing 20 mg besaggliflozin 0-∞ 675-1275 ng h mL -1 Between, for example, 750-1200 ng h mL -1 between.

[0087] In fasting subjects, the preferred tablets of the present invention have a t 1 / 2z The terminal elimination half-life is between 7 and 14 hours, for example, between 8 and 13 hours.

[0088] C max T max t 1 / 2z C 24h AUC 0-t AUC 0-∞These are standard pharmacokinetic parameters. They can be calculated manually or using modeling software well-known in the art (e.g., the Phoenix WinNonlin software package using a non-compartmental model). The basic knowledge for calculating these parameters is well-known (see, for example, Rowland & Tozer (2019), Clinical Pharmacokinetics and Pharmacodynamics: Concepts and Applications, ISBN 978-1496385048, or Jambhekar & Breen (2012), Basic Pharmacokinetics, ISBN 978-0853699804). Typically, the parameters are derived from the mean (e.g., geometric mean) of a group of healthy adults, with a group of at least 12 (and typically between 24 and 36). The parameters should be measured according to standards and practices acceptable to drug regulatory agencies such as the FDA, EMEA, MHLW, or WHO. These values ​​can be measured at appropriate intervals after tablet ingestion (e.g., every hour) or at increasingly sparse sampling intervals (e.g., 1, 3, 5, 7, 9, 11, 13, 15, 20, and 24 hours after ingestion).

[0089] The above pharmacokinetic parameters are defined for fasting subjects, i.e., subjects who have fasted overnight for at least 10 hours. The parameters for besagliflozin differ between fasting and postprandial subjects; if the tablet is administered after a meal (e.g., 30 minutes after a meal), C... max C 24h AUC 0-t and AUC 0-∞ Generally, the values ​​are high. Fasting subjects assessed and tested for the parameters defined in this paper were healthy, fasting adults (male and female) of Caucasian descent (i.e., not diabetic and not on medication for other diseases) weighing between 60 and 100 kg (e.g., approximately 75 kg). The same behavior may occur in other subjects (e.g., Asian subjects, or patients with lower body weight), but the population assessed for the parameters should meet these criteria. Testing in a cohort of at least six subjects is typical.

[0090] With the required C max T max C 24h AUC 0-t and / or AUC 0-∞Sustained-release tablets can be prepared according to the following guidelines, combined with common knowledge of sustained-release tablet preparation, for example, as in Chapter 32 of Collett & Moreton (2007), *Pharmaceutics: The Science of Dosage Form Design* (3rd ed.), Lordi (1986), *Theory and Practice of Industrial Pharmacy* (3rd ed.), Timmins et al. (2014), "Hydrophilic Matrix Tablets for Oral Controlled Release" (ISBN 978-1493915187), Sushma et al. (2014), "Matrix Tablets: An Approach Towards Sustained-Release Drug Delivery" (ISBN 978-3659579110), and Rasul et al. (2011), "Sustained-Release Tablets..." The following documents describe how to synthesize glipizide extended-release tablets using hydroxyethyl and hydroxypropyl cellulose: ISBN 978-3844323719, "Design and Manufacture of Pharmaceutical Tablets" (ISBN 978-0128021828) by Eyjolfsson (2014) and "Extended Release Tablet of Antidiabetic Drug: Development, Optimization and Evaluation" (ISBN 978-3659448140).

[0091] Therefore, the manufacturing principle of sustained-release tablets is well known in the art. Compared to immediate-release capsules at any given dose, the use of sustained-release tablet technology reduces the C60 of besagliflozin. max This is consistent with the pharmacokinetic characteristics desired by this invention. Consistent with known design principles, modifying the properties of sustained-release tablets allows for control over the degree of reduction.

[0092] There are three main methods for achieving sustained release from tablets: (i) using a monolithic matrix, where drug particles are dispersed in a soluble or insoluble matrix; (ii) a reservoir or membrane control system; or (iii) an osmotic pump system. Soluble matrix-based tablets, including compressed mixtures of besagliflozin and water-swellable hydrophilic polymers, begin to dissolve and release besagliflozin over a prolonged period once they enter the gastrointestinal tract. Insoluble matrix-based tablets, including mixtures of besagliflozin and waxes or water-insoluble substances (such as fats or polymers), allow water to diffuse and dissolve the besagliflozin, releasing it. The water diffusion pathway can be part of the tablet after ingestion or may occur after ingestion as a channeling agent leaches from the tablet. A reservoir-based tablet includes a membrane through which besagliflozin must diffuse, and this diffusion is facilitated by the hydration of the membrane. The membrane is typically made of polymers that remain intact during release, such as acrylic copolymers, ethyl cellulose, shellac, and zein. The osmotic pump system is similar to a reservoir system, but the hydration of the tablet core generates hydrostatic pressure, forcing the dissolved besagliflozin through the pores in the semi-permeable coating of the tablet core. Details of suitable release-regulating agents for these tablets are provided below.

[0093] These conventional methods are well-known, and anyone skilled in tablet formulation technology can use any of these methods to prepare and test tablets and adjust them according to the desired pharmacokinetic characteristics. Tablet properties can be modified based on the characteristics of the formulation method used. For example: for soluble matrices, release can be controlled by selecting the chemical, physical properties, and quantity of the water-swellable hydrophilic polymer; for water-insoluble (e.g., wax) matrices, release can be controlled by selecting the quantity of water-insoluble substances and the properties and quantity of channeling agents; for insoluble polymer matrices, the pore structure of the matrix is ​​a key parameter, and matrices with high rigidity and few pores generally result in slower release; for reservoir systems, membrane selection is critical, especially the selection and quantity of membrane plasticizers, but adding water-soluble components to the membrane can also increase the release rate; for osmotic pump systems, the rate at which water can enter the tablet core and the rate at which besagliflozin can exit the coating pores control the tablet release characteristics. Therefore, the composition and design principles for controlling tablet release characteristics while providing physically stable tablets are well known, and those skilled in tablet formulation technology can utilize any of these methods to prepare and test (in vitro and in vivo) tablets to obtain the release characteristics of the product, providing the required C for any given quantity of bexagliflozin. max T max C 24h AUC 0-t and AUC 0-∞ .

[0094] Preferred tablets of the present invention comprise besagliflozin dispersed in a water-insoluble (e.g., wax) matrix (e.g., based on disorbate diglyceride, as described below).

[0095] In addition to employing these technologies to delay tablet release, we also hope to adapt the tablets to gastric retention (as described below) to increase their sustained-release ratio in the duodenum, thereby delaying the passage of besagliflozin through the small intestine to further refine its pharmacokinetic behavior in vivo.

[0096] In vivo trials are being conducted to determine the levels of C in the human body. max T max C 24h AUC 0-t and / or AUC 0-∞ Prior to this, in vitro dissolution tests on tablets may be useful to provide some preliminary predictions and facilitate design modifications. These in vitro tests are used in regulatory settings to ensure that tablets can effectively and safely deliver the required therapeutic amount of drug into the bloodstream, and involve formal dissolution acceptance tests on tablets in manufacturing batches intended for human patients. Such formal acceptance tests can ensure that the required amount of besagliflozin is delivered into the body within the required time intervals.

[0097] Therefore, the present invention provides a solid oral dosage form (most typically a sustained-release tablet) containing besagliflozin, and in an in vitro dissolution test simulating gastric fluid (see below), releases ≤17% of besagliflozin after 1 hour and ≥80% after 8 hours. Thus, at least 83% of besagliflozin remains in the dosage form after 1 hour of the in vitro dissolution test, but at least 80% is released after 8 hours (this includes examples of 100% release within 8 hours). The tablet releases less besagliflozin after 1 hour of the dissolution test than an immediate-release capsule containing the same amount of besagliflozin. In one embodiment, the dosage form releases 20-45% (inclusive) of besagliflozin after 3 hours and 45-75% (inclusive) of besagliflozin after 5 hours.

[0098] In an embodiment of the invention, a dosage form (e.g., a sustained-release tablet) releases 20-45% besagliflozin after 3 hours in an in vitro dissolution test, and the dosage form may be prepared to release 23-43% besagliflozin after 3 hours.

[0099] In embodiments of the invention, a dosage form (e.g., a sustained-release tablet) releases 45-75% besagliflozin after 5 hours in an in vitro dissolution test. The dosage form can be prepared to (a) release 45-72% besagliflozin after 5 hours, (b) release 50-70% besagliflozin after 5 hours, (c) release 49-69% besagliflozin after 5 hours, or (d) release 48-68% besagliflozin after 5 hours. More generally, the dosage form can release xy% besagliflozin after 5 hours, wherein: x is selected from 45, 47, 48, 49, or 50; and y is selected from 68, 69, 70, 72, or 75.

[0100] In one embodiment, the sustained-release tablets can release 23-43% besagliflozin after 3 hours and (2) 45-72%, 50-70%, 49-69%, or 48-68% besagliflozin after 5 hours in an in vitro dissolution test. Therefore, these percentages can be used as standards for use at 3 hours and 5 hours in the in vitro dissolution tests disclosed herein.

[0101] Because determining these release characteristics is necessarily destructive, these parameters do not need to be determined directly for a specific tablet of interest, but rather using tablets prepared with the same ingredients and manufacturing process. Therefore, a batch of tablets can be prepared by a specific method, and an in vitro dissolution test can be performed on a sample group of representative tablets from that batch. If the test results meet the above requirements, then the tablets prepared by this production method are the tablets of the present invention. Therefore, the present invention also provides tablets from any such production batch.

[0102] The in vitro dissolution test used for these determinations is one of the standards in the art, especially for sustained-release tablets, for example, see USP. <711> Dissolution rate or Ph Eur. 2.9.3. More details are below.

[0103] Studies of specific tablet formulations can provide IVIVC (in vitro-in vivo correlation), which describes the in vitro characteristics of the tablet (e.g., the rate or extent of drug release) and the associated in vivo response (e.g., C). max or AUC 0-t The relationship between the two is discussed. This type of model helps in the rational development, evaluation, and modification of the sustained-release tablets of the present invention.

[0104] Ideally, the release of a sustained-release formulation should be independent of dietary status. However, if such an effect is unavoidable (e.g., if the sustained-release mechanism depends on the mechanism by which the contents are released from the stomach, as in several embodiments of the present invention), it is desirable that the consequences of prior food consumption be predictable and limited, and that there is no risk of adverse side effects or insufficient therapeutic effect on the patient under any circumstances. The tablets of the present invention meet these criteria.

[0105] Various drugs are known to affect gastrointestinal motility as side effects or therapeutic mechanisms. Among drugs that affect gastric emptying and are often co-delivered with oral antidiabetic drugs are glucagon-like peptide-1 (GLP-1) receptor agonists. GLP-1 receptor agonists inhibit gastric emptying and may mimic a postprandial state; therefore, administration prior to besagliflozin may increase besagliflozin exposure. Currently, most GLP-1 receptor agonists are delivered subcutaneously, but an oral delivery formulation of semaglutide has recently been approved, and more such or synthetic agonists may be approved in the future. As with mealtimes, it is hoped that the consequences of GLP-1 receptor agonist administration will be predictable and limited. The tablets of the present invention meet these criteria.

[0106] Gastric retention

[0107] By following the above guidelines and combining them with common knowledge of sustained-release tablet preparation, tablets with the required C can be prepared. max T max C 24h AUC 0-t and / or AUC 0-∞ Extended-release tablets with specific properties. A further approach to modifying tablets to achieve desired parameters is to incorporate gastric retention adaptations into the tablets, particularly one or more of the four adaptations discussed below. The overall goal of the gastric retention adaptations discussed herein is to delay the passage of besagliflozin through the small intestine, thereby promoting the majority of besagliflozin's extended release occurring in the stomach or small intestine (see Hou et al. (2003) Crit Rev Ther DrugCarrier Syst 20:459-97). These adaptations have all been shown to reduce plasma C10 levels compared to immediate-release capsules. max At the same time, it can still provide effective C 24h And T max AUC 0-t and AUC 0-∞ All are within the required range.

[0108] One approach to achieving the desired in vivo pharmacokinetic behavior is to incorporate an effervescent excipient into the tablet, particularly one that effervesces upon contact with gastric acid, such as carbonates or bicarbonates, or hydrogen carbonate salts, such as sodium bicarbonate. Due to gas release, the tablet tends to float during effervescence, thus delaying its progression toward the pyloric sphincter at the fundus of the stomach (e.g., see Wei et al. (2001) *Drug Dev Ind Pharm* 27:469-74, Ray & Prusty (2010) *International Journal of Applied Pharmaceutics* 2:12-16). A tablet matrix containing bicarbonate provides an additional advantage in protecting besaglitazone from acid degradation. As shown in the examples, the addition of an effervescent excipient reduces C... max This helps to obtain the desired pharmacokinetic characteristics.

[0109] A second approach to achieving the desired in vivo pharmacokinetic behavior is to construct a tablet that disperses into numerous particles or microspheres upon contact with gastric contents, thus creating a sustained release. Typically, the stomach takes longer to expel multiple small particles / microspheres compared to a large tablet. A similar approach was disclosed in Pharm Dev Technol 16(4):316-30 by Aburahma & Hamza Yel (2011), who used rapidly disintegrating components to compress sustained-release beads.

[0110] A third approach to achieving the desired in vivo pharmacokinetic behavior is to use low-density excipients, thereby providing buoyancy or floating tablets. By using sufficient amounts of low-density excipients, tablets with a density below the total density of gastric contents can be prepared, allowing them to float in the stomach and delaying their transport to the pyloric sphincter without requiring effervescence (e.g., Srikanth Meka et al. (2014) Acta Pharm Sinica 64: 485-494). As illustrated in the example, this approach can effectively reduce C... max The density of stomach contents is approximately 1.004-1.010 g / cm³. 3 Therefore, the density of the tablet should be lower than this density, so that it can float under ideal conditions.

[0111] Buoyancy, and the length of time the tablet retains buoyancy during degradation, can be assessed in vitro in simulated gastric fluid at 37°C. In some embodiments, the tablets of the present invention retain buoyancy (i.e., remain on the surface) until 90% of besagliflozin is released. In some embodiments, the tablets of the present invention can retain buoyancy for 5 hours or longer, for example, 8 hours or longer. In practice, the tablets can be studied using the same techniques as the in vitro dissolution tests discussed below, for example in a device containing 900 mL of 0.1 N HCl (simulated gastric fluid) at 37 ± 0.5°C. The density of the tablets can be determined by displacement method using analytical grade benzene as the displacement medium.

[0112] A fourth approach to achieving the desired in vivo pharmacokinetic behavior is to incorporate a mucosal adhesive into the tablet. Mucosal adhesives allow the tablet to interact with the gastrointestinal mucosal surface (e.g., the stomach wall), thereby delaying tablet progression. This approach is discussed, for example, in Jha & Nanda (2013), *Asian Journal of Biomed Pharm Sci* 3:44-49. Various mucosal adhesives suitable for incorporation into tablets are known in the art; they are typically hydrophilic polymers. Generally, good mucosal adhesives possess strong hydrogen-bonded groups (-OH, -COOH), strong anionic charges, sufficient flexibility to penetrate cytosolic polysaccharide-protein complexes to extend the glycan network, surface tension properties suitable for wetting mucus / mucosal tissue surfaces, and / or high molecular weight (see Yadav et al. (2010), *Journal of Chemical Pharm Research* 2:418-32). Examples of mucosal adhesives are as follows. Some mucosal adhesives are known to provide sustained-release properties to tablets (e.g., HPMC, polyethylene oxide), and therefore both effects can be effectively utilized in the tablets of the present invention. In the tablets of the present invention, the useful amount of the mucosal adhesive can be 10-25% of the total tablet weight.

[0113] Therefore, a fourth aspect of the invention provides a sustained-release tablet containing besagliflozin and a mucosal adhesive. Compared to an equivalent tablet with the same composition except for the absence of a mucosal adhesive, the amount of mucosal adhesive added to the tablet can delay the passage of the tablet through the stomach and / or duodenum in the body. Preferred mucosal adhesives for the tablets of the present invention are nonionic polyethylene oxide polymers, particularly those with an average molecular weight of 800,000 or higher, for example, 900,000-5,000,000. These hydrophilic polymer powders are available in pharmacopoeia-grade products under the trade name POLYOX. TM These are derived from Dow Chemical and have a molecular weight of 100,000-7,000,000. They act as both mucosal adhesives and provide sustained-release properties, thus enabling them to effectively achieve both functions in the tablets of this invention. Appropriate amounts of mucosal adhesives have been discussed above.

[0114] The four methods discussed above can be used individually to deliver sustained-release tablets of besagliflozin with the desired pharmacokinetic parameters. In particular, each method can reduce C60 levels when compared to immediate-release formulations. max The degree of reduction can be controlled to some extent, particularly by increasing specific adaptability, thereby providing the required C for any given amount of besagliflozin tablets. max For example, increasing the amount of effervescent excipients or increasing the number of individual particles / microcapsules can increase gastric retention to some extent, thereby correspondingly reducing C. max Similarly, increased buoyancy can also increase gastric retention, although the extent to which buoyancy can be increased is actually limited. Finally, increasing the content of mucosal adhesive, or using a stronger mucosal adhesive, will increase gastric retention, although, again, there are practical limitations to the ability of tablet mucosal adhesives. However, in general, those skilled in tablet formulation technology in the art can use these methods to prepare and test tablets and adjust them according to desired pharmacokinetic properties.

[0115] Although these four methods can be used individually, combining them is more advantageous.

[0116] The inventors have discovered that the first method itself can effectively reduce C. max However, these tablets exhibit significant inter-patient variability (especially T). max In cases where one does not wish to be bound by theory, this behavior may occur if, in some patients, the tablet leaves the stomach earlier than expected, and thereafter it no longer undergoes acid-driven effervescent disintegration, thus reducing drug release and bioavailability. To mitigate this problem, the first and second approaches can be combined, for example, by compressing multiple effervescent particles into a single tablet, so that the individual effervescent particles are released as the tablet disperses in the stomach.

[0117] The second method is technically difficult to implement sustainably, although it reduces C max However, the effect is not as noticeable (e.g., not as noticeable as the first method). Furthermore, the commercial shelf life of the granules is relatively short, so the second method is not preferred, whether used alone or in combination with any other method.

[0118] When using more than one method to improve pharmacokinetic behavior, one option is to combine a third and fourth method to obtain a low-density tablet containing a mucosal binder. As shown in the examples, this combination of methods provides tablets with advantageous properties for delivering besagliflozin in the human body. Therefore, the present invention provides a sustained-release tablet containing besagliflozin and a mucosal binder, wherein the density of the tablet is lower than that of human gastric acid. Suitable mucosal binders and their amounts, as well as further details on suitable densities, have been discussed above.

[0119] Gastric retention can be measured by adding a radionuclide to the formulation and directly recording the proportion of the formulation remaining in the stomach as a function of time after administration using a suitable scintillation camera. While this method offers relatively high accuracy, it has two main drawbacks: (i) the radionuclide is typically not found in commercial products, thus deviating the formulation's composition from the intended commercial form; and (ii) conducting such experiments is both difficult and expensive, exposing participants to additional risks of exposure to radioactive materials. Therefore, the determination of gastric retention can be inferred from other characteristics of the formulation, such as the Tg of the formulation. max T with immediate-release formulations max Compare, or compare T in a fasting state max and T in the post-meal state max Comparison. As mentioned above, using [ 14 Studies of besagliflozin [C] have shown that colonic absorption is minimal, with the majority of absorption occurring in the small intestine. The effect of meal status is consistent with this description. For example, in a fasting state, T... max The absorption time is 1 to 2 hours, but 5 hours in the postprandial state, which is explainable if gastric contents need to be released to achieve the maximum absorption rate. The T-day of besagliflozin extended-release tablets in the U20 formulation (see below) in the fasting state... max The retention time is 3.5 hours, and 5 hours in the postprandial state, which is consistent with the view that they remain in the stomach for a longer time than sustained-release formulations.

[0120] Tablet ingredients

[0121] As described above, the tablets of the present invention, in addition to besagliflozin, typically or optionally contain: one or more controlled-release agents (e.g., components for forming a matrix or film); one or more matrix or film modifiers (e.g., channeling agents or wicking agents); one or more solubilizers; one or more flow aids, lubricants, and / or flow phase aids; one or more disintegrants; one or more fillers; one or more binders; one or more density modifiers and / or effervescent components; one or more colorants; one or more flavoring agents; one or more antioxidants; and / or one or more mucosal adhesives. These components are typically present in the tablet as a mixture, but may also be present in different proportions in layers or discrete geometries, such as particles or spheres of one composition embedded in another composition, or in sheets or blocks of materials of different volumes of compositions.

[0122] A typical tablet consists of a core of one component surrounded by a coating or outer layer of another component. The tablets of this invention typically include a coating.

[0123] Examples of controlled-release agents used to form the matrix include, but are not limited to, water-swellable hydrophilic polymers (e.g., hydroxypropyl cellulose or methyl cellulose, sodium carboxymethyl cellulose, alginate, alginic acid, gelatin, xanthan gum (with or without locust bean gum), carbomer, polyethylene oxide, galactomann, etc.), waxes (e.g., hydrogenated vegetable oils, microcrystalline waxes, carnauba wax, etc.) and insoluble polymers (e.g., ethyl cellulose). These components may constitute 15%–40% of the tablet weight.

[0124] A particularly useful controlled-release agent for forming the tablet matrix of the present invention is diglyceride behenate, because it has advantages in terms of lower density than gastric juice and resistance to lipases. Diglyceride behenate is a known sustained-release agent (e.g., as described by Opta et al. in (2013) *Int J Pharm Tech Res* 5:622-8). Preferred tablets contain 30-35% diglyceride by weight. "Diglyceride behenate" is a preferred pharmaceutical description of commercially available mixtures of glycerides (including mono, di, and tribenzyl behenates) that are currently primarily in the form of diesters. Monobenzyl behenate has two regioisomers, and diglyceride behenate has two regioisomers. Previously, the term "glyceride behenate" has been used to describe commercially available ester mixtures, but the disadvantage of this term is that it implies the composition is primarily in the form of a monobenzyl behenate, which is inaccurate. Commercial formulations of diglyceride behenate contain 40-60% diester by weight. Any reference to "di-behenate glycerol" in this article should be understood as referring to products containing a mixture of behenate glycerides, rather than the amount of di-behenate glycerol contained therein.

[0125] Commercial formulations of diglyceride behenate are formulated to improve their performance in pharmaceutical manufacturing processes, such as improving blending or flow properties. The inventors have discovered that micronized or atomized formulations (e.g., Compritol 888ATO) can improve these properties. TM It has good properties for preparing the tablets of the present invention.

[0126] Examples of controlled-release agents used to form films include, but are not limited to, ethyl cellulose, acrylic polymers (e.g., Eudragit RL&RS). TM Shellac and zein. These can be combined with plasticizers such as dibutyl phthalate, diethyl phthalate, dibutyl sebate, or citrate. Plasticizers typically comprise about 10-25% of the membrane polymer by weight, sufficient to allow the membrane to fully coalesce into a thin film without making it too elastic, plastic, soft, or permeable.

[0127] Examples of matrix modifiers include, but are not limited to, sugars, polyols, and soluble salts. These can alter the diffusion properties of the matrix, as well as the rate and extent of its hydration, thereby modulating the release of besagliflozin. Channel modifiers include sodium chloride, sugars, and polyols (e.g., lactose), comprising 10-30% of the tablet weight.

[0128] Examples of solubilizers include, but are not limited to, surfactants (including ionic and nonionic surfactants), such as sodium lauryl sulfate, cetyltrimethylammonium bromide, polysorbates (e.g., polysorbate 20 or 80), poloxamers (e.g., poloxamer 188 or 207), and polyethylene glycol. Preferred tablets include poloxamers, ideally micronized, for example in micronized form (EP-A-1661558). Poloxamers with an average particle size between 10 and 200 μm are useful. The most preferred poloxamer is micronized poloxamer 188. The preferred amount of poloxamer 188 in the tablets of this invention is 10-12% of the tablet weight. Higher contents of poloxamer facilitate faster tablet release.

[0129] In some embodiments, the surfactant may be combined with amorphous besagliflozin in a manner disclosed in WO2018 / 167589 with the aim of obtaining tablets with durable stability and good bioavailability (and optionally, bioequivalent to the reference tablets disclosed herein). Useful surfactants for such embodiments are available by trade name SEPITRAP. TM 80 and Dubcare TM GPE810 obtained. SEPITRAP TM 80 is a powdered form of polysorbate 80 microcapsules, in which polysorbate 80 is adsorbed onto a porous magnesium aluminosilicate support. (Dubcare) TM GPE810 is a mixture of PEG-8 caprylic / capric glycerides.

[0130] Examples of lubricants, flow aids, and flow promoters include, but are not limited to, magnesium stearate, calcium stearate, stearic acid, hydrogenated vegetable oil, glyceryl palmitate, dibenzyl behenate, sodium stearate fumarate, colloidal silica, and talc. The amount of lubricant in the tablet is typically 1-5% of the tablet weight. Preferred tablets of the present invention comprise magnesium stearate and / or colloidal silica (e.g., in an amorphous, anhydrous form). Preferred tablets contain 1.5-2.5% magnesium stearate by weight and / or 1.0-1.5% colloidal silica by weight.

[0131] Examples of disintegrants include, but are not limited to, starch, cellulose, cross-linked PVP, sodium glycolate starch, and sodium cross-linked carboxymethyl cellulose.

[0132] Examples of fillers (also known as leavening agents or diluents) include, but are not limited to, starch, maltodextrin, polyols (e.g., lactose), and cellulose. Preferred tablets of the present invention comprise lactose and / or microcrystalline cellulose (e.g., Avicel series products; see Doelker et al., Drug Development and Industrial Pharmacy (1995) 21: 643-61). Lactose can be used in anhydrous or hydrated form (e.g., monohydrate) and is typically prepared by spray drying, fluidized bed granulation, or roller drying. The particle size of the microcrystalline cellulose is preferably about 150-200 μm. The tablets preferably contain 11-13% lactose by weight and / or 18-20% microcrystalline cellulose by weight. Lactose monohydrate obtained by spray drying is preferred.

[0133] Examples of adhesives include, but are not limited to, cross-linked PVP, HPMC, microcrystalline cellulose, sucrose, starch, etc.

[0134] Examples of effervescent components include, but are not limited to, carbonates or acid carbonates (bicarbonates), such as sodium bicarbonate.

[0135] Examples of antioxidants include, but are not limited to, tert-butyl-p-hydroxyanisole, butylated hydroxytoluene, sodium metabisulfite, propyl gallate, and cysteine. Preferred tablets contain butylated hydroxytoluene as an antioxidant.

[0136] Examples of mucosal adhesives include, but are not limited to: carbomer (a polymer crosslinked with acrylic acid and polyolefin ether or diethylene glycol), crosslinked carboxyl-polymethylene, carboxymethyl cellulose (such as sodium carboxymethyl cellulose), hydroxyethyl cellulose, hydroxypropyl methyl cellulose, polycarboferric sulfate, astragalus gum, poly(acrylic acid / divinylbenzene), alginate (e.g., sodium alginate), guilarin, and polyoxyethylene (also known as polyethylene oxide or polyethylene glycol). As described above, in the tablets of the present invention, the useful amount of the mucosal adhesive can be 10-25% of the total tablet weight. Preferred adhesive components included in the tablets of the present invention are nonionic polyethylene oxide polymers, particularly those with an average molecular weight (e.g., number average) of not less than 800,000 (based on rheological measurements). Preferred tablets comprise 16-20% by weight of polyethylene oxide.

[0137] While uncoated tablets can be used, coated tablets are more commonly used, in which case conventional non-enteric coatings can be used. The coating can be white or colored, such as blue. Suitable coatings include, but are not limited to, polymer film coatings, such as those containing polyvinyl alcohol, such as 'Opadry II'. TM (This includes partially hydrolyzed PVA, titanium dioxide, polyethylene glycol 3350, and talc, with optional colorants such as indigo carmine, iron oxide yellow, or FD&C yellow #6). The coating amount is typically 2.5–3.5% of the tablet core weight.

[0138] Some components can play multiple roles in tableting. For example, diglyceride behenate can be used as a controlled-release agent in the tablet matrix, or as a gastric retention excipient (due to its density), or as a lubricant, while polyethylene oxide can be used as a controlled-release agent or as a mucosal adhesive. Therefore, a component can play multiple roles in a tablet, but usually a component is used for only one purpose, and thus its quantity and location (in the tablet and / or in the manufacturing process) are selected accordingly.

[0139] The tablets of this invention typically have a hardness between 20 and 100 N, and more typically between 20-60 N, 30-40 N, or 60-90 N. Hardness can be determined according to USP. <1217> The tablets can be conveniently evaluated using the Dr. Schleuniger Pharmatron tester, which drives the anvil to compress the tablets at a constant rate until they break.

[0140] The tablets of this invention typically have a brittleness of ≤1% by weight. Brittleness can be determined according to USP. <1216> An assessment will be conducted.

[0141] The tablets of this invention typically have a moisture content of ≤5% by weight. The moisture content can be determined according to USP. <921> An assessment will be conducted.

[0142] The tablets of the present invention can be simply prepared by direct compression (with subsequent coating if necessary).

[0143] Preferred tablets

[0144] The preferred tablets of the present invention comprise: besagliflozin; glyceryl disorbate; polyethylene oxide; lactose (anhydrous or preferably monohydrate); poloxamer 188 (preferably micronized); microcrystalline cellulose; colloidal silica; and magnesium stearate; optionally, a coating composed of polyvinyl alcohol.

[0145] Each of the tablets described herein comprises the following components: besagliflozin, between 3 and 60 mg; disorbate diglyceride, between 100 and 140 mg; polyethylene oxide, between 50 and 75 mg; lactose, between 40 and 50 mg; poloxamer 188, between 40 and 45 mg; microcrystalline cellulose, between 60 and 80 mg; colloidal silica, between 4 and 5 mg; magnesium stearate, between 6 and 9 mg; and optionally, a coating of 10-12 mg composed of polyvinyl alcohol.

[0146] The three preferred tablets of the present invention comprise one of the following cores, further details of which are well known and can also be found in the Handbook of Pharmaceutical Excipients (edited by Sheskey, Cook & Cable; 8th edition 2016):

[0147] Bexagliflozin 5mg 10mg 20mg Polyethylene oxide, average molecular weight 900,000 65mg 65mg 65mg Diglyceride behenate 120mg 120mg 120mg Lactose (anhydrous or monohydrate, e.g., spray-dried) 45mg 45mg 45mg Polosham 188 42mg 42mg 42mg microcrystalline cellulose 70mg 70mg 70mg Colloidal silica 4.5mg 4.5mg 4.5mg magnesium stearate 7.5mg 7.5mg 7.5mg

[0148] The core hardness is preferably 40-60N or 60-90N, and the brittleness is <1% by weight.

[0149] The present invention also provides a tablet whose core coating is composed of one of three polymer films: polyvinyl alcohol, titanium dioxide, and polyethylene glycol 3350. The amount of coating may be 3% of the weight of the core.

[0150] In these preferred tablets: poloxamer 188 should be micronized; lactose may be anhydrous, but is preferably monohydrate; optionally, the coating may consist of polyvinyl alcohol, titanium dioxide, polyethylene glycol 3350, talc, brilliant blue FCF and indigo carmine, such as Opadry II blue products.

[0151] The present invention also provides an oral dosage form (particularly a solid oral dosage form, such as tablets) whose geometric mean C in a cohort of healthy subjects max and geometric mean AUC 0-t The C of its logarithmic transformation max AUC of the logarithmic transformation0-t The 90% confidence interval, after exponentiation, falls exactly on the geometric mean C of the reference tablets (see also tablets U5, U10, and U20 below) in the same cohort, which have one of the following components. max and geometric mean AUC 0-t Between 80.00% and 125.00%.

[0152] (a) A tablet comprising: a core consisting of a mixture of 5 mg besagliflozin, 65 mg nonionic polyethylene oxide with an average molecular weight of about 900,000, 120 mg glyceryl disorbate powder, 45 mg spray-dried lactose monohydrate, 42 mg micronized poloxamer 188, 70 mg microcrystalline cellulose, 4.5 mg amorphous anhydrous colloidal silica, and 7.5 mg magnesium stearate; and a film coating consisting of 10.77 mg of a mixture comprising PVA, titanium dioxide, polyethylene glycol 3350, talc, brilliant blue FCF, and indigo carmine (such as Opadry II). TM Blue 85F99153); the tablet core is formed by compression using a 14.8×6.5mm tablet punch, with a hardness between 40-60N.

[0153] (b) A tablet comprising: a core consisting of a mixture of 10 mg besagliflozin, 65 mg of nonionic polyethylene oxide with an average molecular weight of about 900,000, 120 mg of glyceryl disorbate powder, 45 mg of spray-dried lactose monohydrate, 42 mg of micronized poloxamer 188, 70 mg of microcrystalline cellulose, 4.5 mg of amorphous anhydrous colloidal silica, and 7.5 mg of magnesium stearate; and a film coating consisting of 10.92 mg of a mixture comprising PVA, titanium dioxide, polyethylene glycol 3350, talc, brilliant blue FCF, and indigo carmine (such as Opadry II). TM Blue 85F99153); the tablet core is formed by compression using a 14.8×6.5mm tablet punch, with a hardness between 40-60N.

[0154] (c) A tablet comprising: a core consisting of a mixture of 20 mg besagliflozin, 65 mg of nonionic polyethylene oxide with an average molecular weight of approximately 900,000, 120 mg of glyceryl disorbate powder, 45 mg of spray-dried lactose monohydrate, 42 mg of micronized poloxamer 188, 70 mg of microcrystalline cellulose, 4.5 mg of amorphous anhydrous colloidal silica, and 7.5 mg of magnesium stearate; and a film coating consisting of 11.22 mg of a mixture comprising PVA, titanium dioxide, polyethylene glycol 3350, talc, brilliant blue FCF, and indigo carmine (such as Opadry II). TMBlue 85F99153); the tablet core is formed by compression using a 14.8×6.5mm tablet punch, with a hardness between 40-60N.

[0155] These reference tablets (a), (b), and (c) can be manufactured by: (i) mixing besagliflozin, colloidal silica, and 80% of MCC, and then sieving the mixture; (ii) adding the remaining MCC to obtain mixture “A”; (iii) sieving polyethylene oxide, glyceryl disorbate, and lactose to obtain mixture “B”; (iv) mixing mixtures “A” and “B” together; (v) adding sieved magnesium stearate, and then mixing further; (vi) compressing the material into a tablet core, for example using a 14.8 × 6.5 mm tablet punch and a suitable die; (vii) dusting; and (viii) coating, for example using a 12% or 18% w / w suspension of coating material to obtain a coating that increases the tablet weight by about 3%. The besagliflozin formulation used to manufacture these reference tablets should have the solid crystalline form disclosed in WO2011 / 153953. Preferred embodiments of this type of formulation have a particle size distribution of d(0.9)≤700μm.

[0156] Evaluation of C of logarithmic transformation max and AUC 0-t Further details regarding whether the 90% confidence interval of the value falls within the 80.00–125.00% range of the value achieved by the reference tablet will be given in the next section, for example, using randomized crossover studies in appropriate test populations.

[0157] Bioequivalence

[0158] Therefore, the present invention provides oral dosage forms that are bioequivalent to the reference tablets (a) to (c). The oral dosage forms contain the same molar amount of besagliflozin as the relevant reference tablets, i.e., the same amount as 5 mg, 10 mg or 20 mg of besagliflozin of formula (I).

[0159] In the fields of bioavailability and bioequivalence, it is well known how to determine whether any particular tablet meets regulatory requirements for bioavailability and pharmacokinetic bioequivalence, for example: Niazi (2014), *Handbook of Bioequivalence Testing*, 2nd Edition, ISBN 978-1482226379; (FDA Guidance for Industry: Bioequivalence Studies with Pharmacokinetic Endpoints for Drugs Submitted Under an ANDA), December 2013; (FDA Guidance for Industry: Bioavailability and Bioequivalence Studies Submitted in NDAs or INDs - General Considerations), March 2014; *FDA Guidance for Industry: Bioavailability and Bioequivalence Studies Submitted in NDAs or INDs - General Considerations*, March 2014; *FDA Guidance for Industry: Validation of Bioanalytical Methods*. Guidance for Industrp: Bioanalytical Method Validation, May 2018; Guideline on The Investigation of Bioequivalence, EMA, January 2010 (CPMP / EWP / QWP / 1401 / 98Rev.1 / Corr**); and Guideline on the pharmacokinetic and clinical evaluation of modified release dosage forms, EMA, November 2014 (EMA / CPMP / EWP / 280 / 96Corr1).

[0160] Many individual-specific factors can affect drug concentrations in plasma. Therefore, the quality of the subject, whether the medication was taken on an empty stomach or after a meal, the degree of liver and / or kidney impairment, any accompanying medications, food, alcohol, or tobacco use, and the influence of sex, race, genetics, and culture are all important considerations. Consequently, even under optimal control conditions, drug concentrations can vary from person to person. The characteristics of extended-release formulations, such as the percentage of dissolution as a function of time, can be most accurately determined by referencing the characteristics of in vitro tests (see other sections of this document). When referencing in vivo test characteristics, it is appropriate to adjust or normalize the effects to the expected behavior in well-characterized prototype subjects.

[0161] However, from a practical standpoint, even the characteristics of prototype subjects cannot record all inter-individual variations. For this reason, comparisons between formulations are typically conducted by administering each formulation to be compared to the same individual—for example, one day using the reference formulation and another day using the comparison formulation, and vice versa. A considerably long period is usually allowed (at least 10 times the half-life of the previous formulation) so that pre-administration of one formulation is unlikely to affect subsequent administrations. Because significant inter-individual variability is almost always present, comparisons are usually made between groups of individuals, typically no fewer than 12. Bioequivalence is indicated when comparisons of pharmacokinetic assays between subjects taking the two formulations meet certain criteria.

[0162] In principle, there are many methods to define bioequivalence between formulations, but this article adopts a general standard for regulatory purposes: two formulations are considered bioequivalent if the lower limit of the 90% confidence interval of the logarithm of the geometric mean of the tested formulation parameter is ≥80.00% of the geometric mean of the same parameter in the reference formulation after exponentiation, and if the upper limit of the 90% confidence interval of the logarithm of the geometric mean of the tested formulation parameter is ≤125.00% of the geometric mean of the same parameter in the reference formulation after exponentiation. The maximum achievable drug concentration (C0) is also considered. max The area under the curve (AUC) of drug concentration over time from the start of administration to the final detectable time. 0-t The area under the curve (AUC) of drug concentration over time extrapolated from the start of dosing to infinity. 0-∞ These are typical parameters that must be met for this test. Geometric mean and logarithm are used in these calculations because most physiological variables, including drug plasma concentrations, typically follow a log-normal distribution on repeated sampling of the same individual and on sampling of different individuals in a population.

[0163] Therefore, the present invention provides a sustained-release tablet containing besagliflozin, wherein C max and AUC 0-tBioequivalent to any of the reference tablets (a) to (c).

[0164] To ensure statistical power, C will be measured in multiple subjects. max and AUC 0-t For example, in a group containing at least 12 (usually between 24 and 36) healthy adults.

[0165] To confirm bioequivalence, two-period, two-sequence, two-treatment, single-dose, crossover, single-dose parallel, or replicate designs can be used. The preferred design is a two-period, two-sequence, two-treatment, single-dose, crossover design using healthy subjects. Each subject should be randomized to receive either treatment (experimental and reference drugs). The most accurate, sensitive, and reproducible method should be used to measure plasma drug concentrations. For besagliflozin, the preferred method is separation using validated high-performance or ultra-high-performance liquid chromatography (HPLC) followed by tandem mass spectrometry (MS / MS). For extended-release besagliflozin tablets, both fasting and postprandial bioequivalence studies should be conducted. The highest dose of the formulation should be tested in each case. Multiple-dose (e.g., steady-state) studies are not recommended.

[0166] Typically, at least 12 subjects with evaluable data are required to support the determination of bioequivalence. For studies conducted in a fasting state, participants must fast for at least 10 hours prior to administration and should refrain from drinking water for 1 hour before and 1 hour after administration. No food should be provided for at least 4 hours after administration. 240 mL of water may be provided with the investigational product.

[0167] For studies conducted in a postprandial state, fasting for at least 10 hours prior to the event is required, followed by a standard high-fat, high-calorie meal of 800 to 1000 kcal, of which approximately 150, 250, and 500–600 kcal are derived from protein, carbohydrates, and fat, respectively (see *Guidance for Industry: Bioequivalence Studies with Pharmacokinetic Endpoints for Drugs Submitted Under an ANDA*, (2013), and *Guideline on the pharmacokinetic and clinical evaluation of modified release dosage forms* (EMA / CPMP / EWP / 280 / 96Corr1), section 5.1.4.1). The meal should be taken within 30 minutes or less, and the medication should be administered within 30 minutes of the start of the meal. No additional food should be provided for at least 4 hours.

[0168] To test at any mealtime, venous blood samples should be drawn at appropriate time intervals, typically totaling 12 to 18 samples, covering at least three final elimination half-lives of the drug. It is recommended to test at the expected T... max Dense sampling nearby to obtain the most accurate C max .

[0169] Because C is determined max and AUC 0-t Each test requires the consumption of one tablet, and even if the tablets are identical in all respects and administered to the same subjects, differences exist between each test. However, the pharmacokinetic parameters are based on the C-values ​​of a representative sample group of subjects administered from one production batch of tablets. max It is determined by the mean of AUC. The mean is the geometric mean, not the arithmetic mean. (Using C...) max For example, in a cohort with 6 subjects, the geometric mean C max There are 6 subjects C max The sixth root of the product. If C... max Exponentializing the logarithmic arithmetic mean yields the same result. For each subject C max The logarithm of each will together form the subject C max The logarithmic distribution.

[0170] To compare the second batch of tablets with the first batch, the measurement process can be repeated by administering the second batch of tablets to the same subjects. (In practice, the dosing order for each subject is randomized, so some people are given the second batch of tablets first, while others are given the first batch of tablets first.) The measurement is then performed by analyzing the C-values ​​of the second batch of tablets. max Subtract the C of the first batch of tablets from the logarithm max The difference is calculated logarithmically. The exponent of this difference is the C of the second tablet. max With the first tablet of C max The proportion, if the difference is 0 (e 0 =1), then the proportion is 1. Using the common statistical method for analyzing the difference between two groups of values ​​(analysis of variance), determine the endpoints of the 90% confidence interval for the logarithmic difference. For the two distributions to be considered biologically equivalent, the endpoints of the 90% confidence interval for the logarithmic difference must fall between -0.22314 and +0.22314. If these two values ​​are indexed, they are 80.00% and 125.00% respectively (e.g., e...). -0.22314 =0.8000).

[0171] While it is considered advantageous to administer tablets from each production batch to each subject to minimize the variation between measurements, a similar approach can be used if two production batches of tablets are evaluated using different subject cohorts, where the mean difference between the logarithms of the two cohorts is calculated and a 90% confidence interval for the logarithmic difference is constructed.

[0172] This type of test can be used to determine whether the tablets in question are the tablets defined herein. A batch of tablets produced using an unknown manufacturing process was compared with reference C using the methods described above. max and AUC 0-t Comparing a batch of tablets of the present invention as defined herein, two batches of C max and AUC 0-t If the endpoints of the 90% confidence interval for the logarithmic difference fall between -0.22314 and +0.22314, then the tablets produced by the unknown manufacturing process meet the relevant C. max and AUC 0-t The required tablets.

[0173] The above inference is that if a group of subjects takes the same batch of tablets of the present invention twice, and refers to C... max and AUC 0-t By definition, the first and second C max and AUC 0-t The endpoints of the 90% confidence interval for the logarithmic difference between the two values ​​will fall between -0.22314 and +0.22314.

[0174] More formally, two representative sample groups from the same batch produced C in a cohort of healthy subjects. max Logarithm and AUC 0-t The logarithmic inter-group mean difference, with the endpoints of the 90% confidence interval falling between -0.22314 and +0.22314. The difference from the previous paragraph is that the testing order of the two sample groups can be randomly assigned among the subjects in this cohort, as recommended in bioequivalence testing regulatory guidance documents.

[0175] In vitro dissolution test

[0176] Methods for detecting sustained-release solid oral dosage forms are well known in the art, including those outlined in USP. <711> It specifies the types of devices and methods for detecting immediate-release and sustained-release solid oral dosage forms.

[0177] The testing of besagliflozin extended-release tablets was conducted in a USP device 1 (e.g., a basket device with a nominal capacity of 1 liter), with 900 mL of 0.1 N HCl (i.e., simulated gastric fluid) added and stirred at 50 rpm while maintaining the temperature at 37 ± 0.5 °C. Individual tablets were placed in the device, and 10 mL of liquid was drawn without replacement at specified time points (e.g., 1, 3, 5, and 8 hours) for sampling. The concentration of besagliflozin in the liquid sample at each time point was determined (e.g., by a validated HPLC method) to calculate the tablet release. If this method involves filtering the drawn liquid prior to HPLC analysis, to avoid variations caused by possible interactions between besagliflozin and the filter (e.g., with PVDF material), the first portion of the liquid (e.g., 3.5 mL of the 10 mL sample) can be filtered, and subsequent portions (e.g., the remaining 6.5 mL of the 10 mL sample) can then be analyzed.

[0178] The test can be divided into up to three stages, called levels. In the first stage (Level 1 test), 6 tablets are analyzed. Success is recorded if no single value exceeds each specified range and no single value is below the specified value at the final test time. If this standard is not met, another 6 tablets are analyzed (Level 2 test). Level 2 testing is performed if the average of all 12 units is within each specified range (i.e., 1, 3, 5, and 8 hours), and the final test time is not lower than the specified amount, and tablets outside each specified range do not exceed 10% of the labeled amount (i.e., 2 mg for a 20 mg tablet), and tablets below the specified amount are not lower than 10% of the labeled amount at the final test time. If Level 2 standards are not met, Level 3 testing must be performed. Another 12 tablets are tested. The average of the 24 tablets must be within each specified range and the final test time must not be lower than the specified amount. Of the 24 tablets, no more than two tablets exceeded the labeled amount by 10% outside each specified range, and no more than two tablets exceeded the labeled amount by 10% below the specified amount at the final test time; no more than 20% of the labeled amount (i.e., 4 mg for a 20 mg tablet) of any tablet outside the specified range, or no more than 20% of the labeled amount of any tablet below the specified amount at the final test time.

[0179] A production batch of bexagliflozin extended-release tablets is considered to have passed formal dissolution acceptance testing if it meets the success criteria for at least one of the three tests. A representative tablet from the production batch needs to meet these criteria, such as those specified in the USP. <711> As specified in Acceptance Form 2. In fact, once successful, the test terminates. No further testing should be performed; for example, if it fails at Level 3, the test should be restarted from Level 1.

[0180] Therefore, the present invention provides a sustained-release tablet containing besagliflozin, wherein a production batch of said tablet has components, testing, or manufacturing methods within the formally acceptable range of variations in process, testing, or composition for formulations U5, U10, U20, or U40 (see below). Of these four formulations, U20 is the most preferred formulation for the treatment of diabetes.

[0181] The present invention also provides a sustained-release tablet containing besagliflozin, wherein the tablet is derived from a production batch having a formulation having U5, U10, U20 or U40 components (see below).

[0182] Similarly, the present invention provides a solid oral dosage form (particularly tablets, such as sustained-release tablets) containing besagliflozin, and in an in vitro dissolution test simulating gastric juice, having an f2 value > 50 compared to one of the reference tablets (a), (b), or (c) as defined above, where f2 is the decimal logarithmic reciprocal square root transformation value of the sum of squared errors: Where: n is the number of time points for dissolution measurement; Ri This is the percentage of dissolution of the reference tablet at time point i; T i It represents the dissolution percentage of the solid oral dosage form at time point i.

[0183] This invention provides a sustained-release tablet containing besagliflozin, which, in an in vitro dissolution test simulating gastric fluid, releases ≤17% besagliflozin after 1 hour and ≥80% besagliflozin after 8 hours. Preferably, the tablet releases 20-45% besagliflozin after 3 hours and / or 45-75% besagliflozin after 5 hours. As described above, within the 45-75% release range after 5 hours, the tablet may release (a) 45-72% besagliflozin, (b) 50-70% besagliflozin, (c) 49-69% besagliflozin, or (d) 48-68% besagliflozin. Furthermore, within the 20-45% release range after 3 hours, the tablet may release 23-43% besagliflozin.

[0184] This invention also provides a solid oral dosage form, typically a sustained-release tablet, containing besagliflozin, and passing a formal dissolution acceptance test simulating gastric juice (see above), with the standard being the release of ≤17% besagliflozin after 1 hour and ≥80% besagliflozin after 8 hours. Preferably, the standard for the dissolution acceptance test is the release of 20-45% besagliflozin (e.g., between 23-43%) after 3 hours and / or the release of 45-75% besagliflozin (e.g., 45-72%, 50-70%, 49-69%, or 48-68% as described above) after 5 hours. In the formal dissolution acceptance test, these dosage forms at least pass the USP test. <711> One of the formal Level 3 test plans specified in Acceptance Form 2.

[0185] Treatment

[0186] The tablets of this invention can be used to treat diabetes and its symptoms, particularly type 2 diabetes. More specifically, the tablets of this invention can be used as an adjunct to diet and exercise to improve glycemic control in adults with type 2 diabetes.

[0187] This invention provides a method for treating a subject suffering from diabetes or its symptoms. The method comprises administering the tablets of this invention to the patient, typically involving repeated administration (e.g., once daily) indefinitely or until a desired therapeutic effect is achieved. Typically, besagliflozin is administered at doses of 5 mg, 10 mg, 20 mg, or 40 mg once daily.

[0188] Similarly, the present invention provides tablets of the present invention for use in this treatment method.

[0189] This invention also provides the use of besagliflozin and at least one pharmaceutically acceptable excipient in the preparation of a medicament for treating diabetes, wherein said medicament is a tablet of the invention as described herein. Pharmaceutically acceptable excipients(s) may be selected as described herein to prepare the sustained-release tablets of the invention.

[0190] As described above, the single tablets of the present invention preferably contain 5 mg, 10 mg, 20 mg, or 40 mg of besagliflozin. Therefore, the methods and uses of the present invention generally involve administering 5 mg, 10 mg, 20 mg, or 40 mg (or integer multiples thereof) of besagliflozin to a subject, for example, 5 mg, 10 mg, 20 mg, or 40 mg once daily.

[0191] These treatments and uses are also applicable to diabetic patients receiving other diabetes therapies, such as GLP-1 receptor agonists (e.g., exenatide, lixisenatide, duraglutide, liraglutide, abiglutide, or semaglutide). As discussed elsewhere herein, the tablets of this invention can be safely used in patients without requiring changes to the prescription regimen.

[0192] Existing GLP-1 receptor agonists include exenatide, lixinatide, liraglutide, abiglutide, duraglutide, and semaglutide (Gentilella et al., 2019, Diabetes Metabolism Research and Review, 35: e3070, doi: 10.1002 / dmrr.3070). The first two are exendin-4 analogs, a peptide isolated from the saliva of venomous lizards that induces severe hypoglycemia in its prey, thus aiding the lizard in hunting. The latter four are modified human GLP-1 analogs that can prolong plasma half-life. The approved dosages for these agonists are as follows: exenatide, 5 pg or 10 μg subcutaneously twice daily, or a sustained-release stock formulation once weekly; lixisenatide, 20 μg subcutaneously once daily; for maintenance therapy, 1.2 or 1.8 mg liraglutide subcutaneously once daily; others are subcutaneously once weekly, with abiglutide at a dose of 30 or 50 mg, duraglutide at a dose of 0.75 or 1.5 mg, and semaglutide at a dose of 0.5 or 1.0 mg.

[0193] Embodiments of the present invention

[0194] Example 1 - Effervescent Tablets

[0195] Effervescent tablets containing 10, 15, or 20 mg of besagliflozin were developed. Early tablets consisted of hydroxypropyl methylcellulose (HPMC; low and medium viscosity), lactose monohydrate, sodium bicarbonate, and magnesium stearate, formed by direct compression. In stability studies, these excipients were shown for the first time to be compatible with besagliflozin (e.g., decomposition was observed when citric acid monohydrate was tested as an effervescent agent). Besagliflozin and lactose monohydrate (diluent) were mixed and sieved, then HPMC, sodium bicarbonate, and silica were added in a mixer. Finally, magnesium stearate was added as a lubricant, and the tablets were formed.

[0196] Two target release profiles were initially proposed, yielding ≥80% release of besagliflozin at 12 or 18 hours, as evaluated in an in vitro dissolution test of the tablets in 900 ml of 0.1 N HCl (USP apparatus 2, 37 ± 0.5 °C, 50 rpm, with slot). The tablet composition is as follows:

[0197]

[0198] Tablets containing a mixture of HPMCs released 68% of their contents at 10 hours and 82% at 14 hours. In contrast, tablets containing a single HPMC released 62% at 10 hours, 75% at 12 hours, and 89% at 16 hours.

[0199] Two other batches were prepared:

[0200] Bexagliflozin 10 6.67％ 10 6.67％ HPMC (Low) 40.5 27％ 28.125 18.75％ HPMC (Medium) - - 9.375 6.25％ lactose monohydrate 83 55.33％ 86 57.33％ Sodium bicarbonate 15 10％ 15 10％ magnesium stearate 1.5 1％ 1.5 1％ total 150 100％ 150 100％

[0201] Both tablets had similar release profiles (75%) up to 12 hours, but the release was slightly faster after using the HPMC mixture (91% vs. 87% at 18 hours).

[0202] Various tablets were further prepared, and the final tablet composition was as follows:

[0203] Bexagliflozin 10 6.67％ HPMC (Low) 37.5 25％ lactose monohydrate 86 57.33％ Sodium bicarbonate 15 10％ magnesium stearate 1.5 1％ total 150 100％

[0204] First, different formulations of 20mg tablets were tested:

[0205] Bexagliflozin 20 13.33％ 20 13.33％ HPMC (Low) 45 30％ 30 20％ HPMC (Medium) - - 15 10％ lactose monohydrate 67.75 45.17％ 67.75 45.17％ Sodium bicarbonate 15 10％ 15 10％ Colloidal silica 0.75 0.5％ 0.75 0.5％ magnesium stearate 1.5 1％ 1.5 1％ total 150 100％ 150 100％

[0206] The release profiles of these tablets were slower than expected (less than 75% release at 12 hours in both cases), so modifications were made. The final selected 15mg and 20mg tablet compositions are as follows:

[0207] Bexagliflozin 20 13.33％ 15 10％ HPMC (Low) 37.5 25％ 37.5 25％ lactose monohydrate 76 50.67％ 81 54％ Sodium bicarbonate 15 10％ 15 10％ Colloidal silica 0.75 0.5％ 0.75 0.5％ magnesium stearate 0.75 0.5％ 0.75 0.5％ total 150 100％ 150 100％

[0208] Further in vitro studies have shown that low-viscosity HPMC (19-24% methoxy, 7-12% hydroxypropyl, apparent viscosity of 2% aqueous solution at 20°C approximately 3000 mPa·s) can be used as the sole release-controlled polymer to obtain the desired release profile. Adhesion can be avoided by using 1% magnesium stearate. Therefore, the final batch used in clinical studies has the following composition (mass in mg) and release profile:

[0209]

[0210] First, lactose monohydrate and besagliflozin are mixed, then HPMC, sodium bicarbonate, and silicon dioxide are added, and finally magnesium stearate is added to produce tablets containing these three ingredients. The mixture is compressed directly using a 7mm punch. The tablets are stable for one month at 40°C and 75% relative humidity.

[0211] These three extended-release (XR) tablets and the 20 mg immediate-release (IR) tablet were administered together in human clinical trials to evaluate pharmacokinetics and pharmacodynamics. Dosage was administered once daily for 5 days, either on an empty stomach (days 1 and 2) or after a meal (day 3). The mean p-value ± SD on an empty stomach was:

[0212] <![CDATA[C max (ng / mL)]]> 238±85.1 54.6±22.9 75.9±23.1 99.9±77.9 <![CDATA[T max (h)]]> 1.0 3.0 5.0 4.0 <![CDATA[AUC 0-24h (of h mL) -1 )]]> 961±252 341±123 525±169 632±334 <![CDATA[AUC 0-∞ (of h mL) -1 )]]> 1024±263 391±133 615±170 746±321 <![CDATA[t 1 / 2z (High) 7.14±3.88 8.15±2.30 8.17±2.85 9.42±3.45

[0213] Therefore, compared to 20mg immediate-release tablets, 20mg extended-release tablets have a higher C... max It has approximately 40% bioavailability and a longer half-life, but its bioavailability is significantly reduced by about 30%. The absorption and elimination of the three sustained-release doses are consistent, C max The AUC value increases with increasing dose.

[0214] In the 20 mg IR formulation, the lower C max and longer T max This indicates that food reduces the amount of besagliflozin and delays its absorption. Although food reduces the absorption rate of the 20 mg IR formulation, its impact on overall bioavailability is minimal.

[0215] In the 10mg XR formulation, in addition to shortening the average T... max Apart from this, food appeared to have little effect on PK. However, PK test results showed that the median T value was higher in both fasting and postprandial states. max They are the same.

[0216] In the 15mg and 20mg XR formulations, food reduced T... max However, the mean C values ​​for both doses were higher in both fasting and postprandial states. max and average AUC 0-∞Similar results. These test results indicate that after taking 15mg and 20mg of the XR formulation, food may accelerate but does not increase the absorption of besagliflozin.

[0217] In terms of pharmacodynamics, all tablets were associated with significant dose-dependent glucuria in healthy subjects. Compared to the IR formulation, glucose excretion occurred later in the XR formulation, but the total daily glucose excretion was comparable. Generally, urinary glucose excretion was highest in the first 12 hours and on day 2 after administration in both fasting and postprandial states for all tablets. Beyond 24 hours, the effect of food on glucose excretion appeared to be minimal in all tablets, as postprandial excretion fell within the range of fasting excretion.

[0218] Although these XR formulations successfully lowered the C50 of besagliflozin max However, bioavailability and pharmacokinetics are more unstable than expected. In particular, T... max The variation is unacceptable and may be due to a failure to retain the tablet in the stomach. Premature expulsion from the stomach could also explain sporadic reductions in bioavailability, partly due to the disruptive stress caused by acid-driven effervescence. Therefore, further development of XR formulations is needed to reduce this variability.

[0219] Example 2 - Granule-releasing capsules

[0220] Dispersing the capsule into numerous microspheres or small particles in the stomach reduces the likelihood of the total amount of besagliflozin being expelled from the stomach in a single event. Therefore, two methods relying on the release of multiple besagliflozin granule capsules have been proposed. The first involves releasing low-density particles that float in gastric acid; the second involves releasing coated particles.

[0221] Five floating particulate formulations (“float” capsules) were prepared and evaluated in vitro in 0.1N HCl as described previously. The content (mg / capsule) of these capsules and the percentage of besagliflozin released after 12 hours are as follows:

[0222]

[0223] *Poly(ethyl acrylate-co-methyl methacrylate-co-trimethylaminoethyl methacrylate chloride) 1∶2∶0.1

[0224] Twenty-one formulations with coated pellets were tested, seven of which had three different coatings. The composition and 12-hour dissolution percentage are as follows:

[0225]

[0226] *Poly(ethyl acrylate-co-methyl methacrylate-co-trimethylaminoethyl methacrylate chloride) 1∶2∶0.2

[0227]

[0228] Based on in vitro dissolution tests, the following formulation was selected:

[0229]

[0230] To prepare the granules: besagliflozin, disorbate diglyceride (sustained-release and flotation agent), ethyl acrylate / methyl methacrylate copolymer (Eudragit RS PO; matrix material), microcrystalline cellulose (MCC; filler), and polyvinylpyrrolidone (binder and disintegrant) were mixed; then water was added to form wet granules. The wet granules were extruded and balled to obtain wet granules, which were then dried to form floating granules, which were then filled into capsules.

[0231] To prepare the coated granules, besagliflozin, microcrystalline cellulose (filler), poloxamer 188 (solvent), and polyvinylpyrrolidone (binder and disintegrant) were mixed, and then water was added to form wet granules. The wet granules were obtained by extrusion and balling, and then dried. Talc (lubricant), TEC (plasticizer), and water were mixed to form a coating composition, which was then mixed with two Eudragit copolymer components (sustained-release coating) to obtain a suspension. This was used to coat the dry granules, and the coated granules were then filled into capsules.

[0232] Accelerated stability studies showed that the compositions were stable for 8 weeks at 40°C and 75% RH, but their dissolution profiles changed significantly (slower for coated particles and faster for floating particles). Therefore, these formulations successfully altered the pharmacokinetic characteristics of besagliflozin, but their shelf life is not optimal for commercial purposes.

[0233] Example 3 - Floating Tablets

[0234] The tablets floating in the stomach contents can delay their transport from the stomach, thereby avoiding rapid and premature expulsion from the stomach as described in Example 1 above.

[0235] Two prototype formulations were prepared, with the following compositions (mg / tablet):

[0236] Bexagliflozin 15 15 Nonionic polyethylene oxide 105 60 Diglyceride behenate 100 120 Lactose (filler) - 25 Microcrystalline cellulose (MCC) (filler) 77 77 Colloidal silica 1.5 1.5 magnesium stearate 1.5 1.5 total 300 300

[0237] These tablets were compressed to a hardness of 40N or 50N and then subjected to in vitro dissolution tests as described in Examples 1 and 2. The percentage of besagliflozin released at 8 and 12 hours was as follows:

[0238] 8 hours 59.8 67.9 82.7 51.4 12 hours 95.6 95.2 100 76.9

[0239] Based on these results, the final tablet formulation selected is as follows:

[0240] Bexagliflozin 15 5 Active ingredients Polyethylene oxide (PEO) 105 35 Mucosal adhesion matrix Diglyceride behenate 100 33.3 Flame retardants and flotation agents microcrystalline cellulose 77 25.7 filler Colloidal silica 1.5 0.5 Flow aid magnesium stearate 1.5 0.5 lubricant total 300 100％

[0241] These tablets are formed by combining (a) besagliflozin, MCC, disorbate diglyceride and PEO, and (b) silica and magnesium stearate, and then directly compressing the combination of (a) and (b) into tablets.

[0242] Accelerated stability studies showed that the tablets were stable for 8 weeks at 40°C and 75% RH, with minimal differences in dissolution profiles.

[0243] Example 4 - Faster-Release Tablets

[0244] Further work was conducted to achieve faster release (aiming for complete release within 4-6 hours) from the mucosal adhesion tablets of Example 3, while maintaining a similar tablet composition and direct compression manufacturing technique. Therefore, the tablet composition was modified, resulting in two formulations after further investigation:

[0245] Bexagliflozin 15 4.3 15 4.3 Polyethylene oxide 65 18.8 50 14.3 Diglyceride behenate 120 34.7 120 34.4 lactose 45 13.0 - - Polosham 188 42 12.1 87 24.9 Microcrystalline cellulose (MCC) 50 14.5 50 14.3 Polyvinylpyrrolidone - - 15 4.3 Colloidal silica 4.5 1-3 6 1.7 magnesium stearate 4.5 1-3 6 1.7 total 346mg 100％ 349mg 100％

[0246] These tablets are prepared in the same manner as in Example 3, that is, all components except lubricant and gliding agent are mixed, then they are mixed with a combined lubricant / gliding agent, and directly compressed into tablets with a hardness of 30N.

[0247] The tablets dissolved as follows over 8 hours:

[0248] J％ 0.0 10.7 23.6 38.3 59.5 85.7 94.7 94.2 K％ 0.0 7.4 20.8 46.8 83.9 95.4 94.2 92.0

[0249] The tablets are stable for at least 8 weeks at 40°C and 75% RH. After storage, the difference in the dissolution profiles of tablet J is negligible, but the release profile of tablet K is slightly faster. Furthermore, both tablets become slightly harder after storage.

[0250] Therefore, a faster release than in Example 3 was successfully achieved.

[0251] Example 5 - Lactose-free sustained-release tablets

[0252] Tablet J of Example 4 comprises lactose. This is an animal-derived material, so alternative fillers were tested to achieve a similar release profile for the tablet. Specifically, mannitol, sorbitol, xylitol, and maltodextrin (45 mg each) were tested as alternatives.

[0253] First, it was demonstrated that all four components are compatible with besaglitazone.

[0254] Tablets obtained by replacing lactose with mannitol exhibit similar release behavior, with both types of tablets showing a release percentage >90% within 5 hours in vitro. Attempting to achieve higher tablet hardness (45-55 N) resulted in shorter floating times and therefore slightly faster dissolution.

[0255] Tablets using maltodextrin, sorbitol, and xylitol release slightly faster than those using lactose and mannitol, possibly because of their higher solubility.

[0256] Overall, it is feasible to achieve similar solubility by replacing lactose with alternative excipients.

[0257] Example 6 - Sustained-Release Tablets for Clinical Trials

[0258] Five floating mucosal adhesion tablets were prepared for clinical trials, including the final formulation from Example 3 and tablet J&K from Example 4. Their composition and properties are as follows:

[0259] Bexagliflozin 15 15 15 15 15 Polyethylene oxide 105 85 65 65 50 Diglyceride behenate 100 100 120 120 120 Anhydrous lactose - 45 45 45 - Poloxamer 188, micronized - - 42 42 87 MCC 77 77 50 50 50 PVPP - - - - 15 Colloidal silica 1.5 1.5 1.5 4.5 6 magnesium stearate 1.5 1.5 1.5 4.5 6 total 300 325 340 346 349 Hardness range 40N 40-50N 20-30N 30N 30N

[0260] Typically, these are prepared by combining (a) a mixture of besagliflozin and MCC, (b) a mixture of lubricant and flow aid, and (c) a mixture of the remaining components. The mixture is then compressed to the desired hardness using a rotary compressor with a 14×6mm sheet punch. Brittleness does not exceed 1% w / w.

[0261] Some adhesion was observed when using formulation N. To address this issue, the amount of magnesium stearate was increased to 4.5 mg. The amount of silica was further increased to prepare formulation O.

[0262] Formulations L, M, and O exhibited the best overall properties in terms of solubility and stability. These three tablet types were selected for further investigation of the effect of dissolution time: for tablet L, the release changed from 80% to 90% between 10 and 12 hours; for tablet M, this occurred between 8 and 10 hours; and for tablet O, this occurred between 5 and 6 hours. Therefore, these tablets were named XR11, XR8, and XR5 to reflect their dissolution rates and were used for clinical trial testing.

[0263] Example 7 - Alternative Dosage in Clinical Trial Tablets

[0264] Based on the results of XR5 in Example 6 (Tablet O), floating mucosal adhesion tablets were further prepared using the same method, but containing 10 mg or 30 mg of besagliflozin. Furthermore, the film coating of these tablets was made from Opdary II white. The final tablets had the following composition (mg / tablet):

[0265]

[0266] The coated tablets were cured at 50°C for 24 hours to investigate the effect on hardness. The dissolution-release profile and hardness of the tablets were not affected by curing; therefore, this treatment method was not further investigated.

[0267] Accelerated stability studies have shown that it has no effect on the dissolution performance of tablets.

[0268] In in vitro dissolution tests (as described above), the release of besagliflozin was as follows:

[0269] 1 9％ 9％ 8％ 3 47％ 44％ 44％ 5 85％ 82％ 80％ 8 96％ 96％ 94％

[0270] These tablets exhibited the expected stability and release characteristics, and therefore, along with the XR5, XR8, and XR11 tablets from Example 6, they were included in human clinical trials.

[0271] Example 8 - Particle Size Distribution

[0272] In an in vitro dissolution study of XR5 tablets containing 20 mg or 30 mg besagliflozin, the effect of the particle size distribution of crystalline besagliflozin on tablet dissolution was evaluated. Various particle size distributions were tested, with d(0.9) ranging from approximately 10 μm to 700 μm (i.e., the diameter of crystalline besagliflozin particles comprising 90% of the total volume not exceeding 10 μm to 700 μm), for example, d(0.9) of 220 μm or 325 μm. The dissolution profiles of these tablets with different d(0.9) did not show significant changes; therefore, the particle size distribution of crystalline besagliflozin is not considered an important parameter for tablet dissolution.

[0273] Example 9 - Clinical Trial

[0274] A two-phase, open-label, phase 1 clinical trial was conducted to evaluate the pharmacokinetics of multiple oral administrations of floating tablets in healthy male subjects. Part 1 evaluated the pharmacokinetics (PK) in XR5, XR8, or XR11 tablets (Example 6). Part 2 evaluated the 5-hour release profiles of three doses (10 mg, 15 mg, and 30 mg) of tablets (Example 7). Secondary objectives were to evaluate the safety and tolerability of besagliflozin and to assess the effect of food on PK parameters.

[0275] Part 1 used a crossover design. Twenty subjects received one of three 15 mg tablets or 20 mg capsules (size 2, white, opaque gelatin capsules containing 20 mg besagliflozin and silanized microcrystalline cellulose). There were four dosing periods with no washout. The first dosing period consisted of two days of administration on an empty stomach, followed by one day of administration after a meal. The second through fourth dosing periods consisted of one day of administration on an empty stomach followed by one day of administration after a meal. Subjects were randomly assigned to receive one of the four formulations in a 4-period crossover study, or one of 24 formulations, with the single constraint that each formulation was included five times in the first dosing period.

[0276] Part 2 was a parallel study in 30 subjects. The medication was administered once daily for two days on an empty stomach and once daily after a meal.

[0277] Subjects, in an upright position, swallowed the tablet (or capsule) with approximately 200 mL of water without chewing. Dosing on an empty stomach occurred after at least 10 hours of overnight fasting. For the fasting dose group, breakfast was provided 1 hour after administration. Dosing after a meal occurred 30 minutes after the start of a standard meal. Plasma concentrations of besagliflozin were determined using a validated HPLC-MS / MS method in K2EDTA-anticoagulated whole venous blood samples (see below).

[0278] Figure 1 The geometric mean plasma concentration of besagliflozin in fasting subjects during Part 1 of the trial is shown. The capsules have high C... max However, the use of XR5, XR8, or XR11 tablets successfully reduced C. max This provides an extended absorption phase for all subjects taking these three tablets on an empty stomach, with a T... max The median time was 3 hours (compared to 1 hour for capsules). Considering their lower dosage (15 mg vs 20 mg), the standardized C of tablets... max The concentration decreased to <5 ng / mL / mg, while that in capsules was 10.2 ng / mL / mg. Postprandial C max It also decreased, but to a lesser extent; the XR11 saw the largest decrease. At C... max Subsequently, the plasma concentrations of both tablets and capsules decreased in a biphasic manner. Overall, the specific pharmacokinetic parameters are as follows:

[0279]

[0280] Figure 2 The geometric mean plasma concentrations of besagliflozin in fasting subjects are shown in Part 2 of the trial. All doses (10, 15, and 30 mg) in fasting subjects showed a prolonged absorption phase, T0.max The median time is 3 hours. (The sentence fragment about reaching C appears unrelated and likely refers to a different topic.) max Subsequently, the plasma concentrations of all three tablets decreased in a biphasic manner. Within the dose range of 10-30 mg, the exposure (AUC)... 0-24h and C max Increases in α are generally dose-proportional, but clearance and volume of distribution are dose-independent. Overall, the specific pharmacokinetic parameters are as follows:

[0281]

[0282] Summary of pharmacokinetic studies: Compared to capsule formulations, XR11, XR8, and XR5 tablets had longer absorption times. Under both dietary conditions, the mean exposure after administration of the 20 mg capsule formulation was higher than that after administration of the 15 mg extended-release formulation. Among the extended-release formulations, XR5 showed the highest exposure. Compared to administration on an empty stomach, administration of the extended-release formulation after a meal resulted in a lower AUC. 0-24 and C max Exposure increased from 27% to 49% and from 71% to 97%, respectively. Administration of 10, 15, and 30 mg XR5 formulations resulted in dose-proportional increases in exposure.

[0283] Analytical methods for human plasma samples

[0284] As described above, the concentration of besagliflozin in human plasma samples was determined using a validated HPLC MS / MS method. An example of a suitable method is provided below.

[0285] The internal label 'IS' indicates besagliflozin, in which 6 hexose carbons are... 13 C-substituted. Other internal standards may also be used, such as tolbutamide, but isotopically labeled internal standards are preferred.

[0286] Each run included both "blank + IS" and "blank + drug" samples to monitor the effect of IS on the analyte, and vice versa. All standards and reconstitution solvents were methanol. The matrix was human plasma anticoagulated with K2EDTA.

[0287] The analytical procedure was performed as follows: Thaw the standards, QCs, blank matrix, and study samples (if applicable), vortex for approximately 3 minutes, and then pipette; add 100 μL of blank plasma to the blank, blank + IS, blank + drug, test, and calibration standards; add 5 μL of besagliflozin spiking solution containing 16000 ng / mL to the blank + drug; perform a spiking test with 5 μL of 80 ng / mL spiking solution; add 5 μL of spiking solution to each concentration of calibration standard; add 100 μL of the appropriate concentration and multiple repetitions of QC sample to the QC tube; if applicable, add 100 μL of each study sample to the appropriate test tube; add 5 μL of MeOH to the blank, blank + IS, QCs, and study sample tubes, if applicable; add 50 μL of IS to the test, blank + IS, calibration standards, and QC (and study samples, if applicable); add 50 μL of MeOH to the blank and blank + drug tubes; vortex at high speed for approximately 2 minutes.

[0288] The protein precipitation extraction procedure was as follows: Add 500 μL of acetonitrile (ACN) to all tubes; vortex at high speed for about 3 minutes, then centrifuge at 3000 rpm for 10 minutes; transfer the supernatant to 16 x 100 mm marked tubes; evaporate to dryness under a nitrogen stream at 40 °C for about 10 minutes; add 200 μL of MeOH to each tube to reconstruct all samples, vortex at high speed for about 1 minute; transfer to autosampler vials for LC-MS / MS analysis; centrifuge the vials at 3000 rpm for about 5 minutes.

[0289] The equipment used was: a Shimadzu DGU 14A vacuum degasser; solvent delivery systems, LC-10ADvp and SCL-10Avp, from Shimadzu; an auto-injector, HTC PAL, for CTC analysis; and a 35°C column heater, TS-130, from Phenomenex. TM Mass spectrometer, triple quadrupole MS (API 4000), Sciex.

[0290]

[0291]

[0292]

[0293] Example 10 - Additional Tablet Strength

[0294] To supplement Example 9, 3 mg and 90 mg besagliflozin tablets were prepared. The 3 mg tablets were similar to those of Example 9, but the excipients were removed from the 90 mg tablets, thus eliminating their flocculation properties. Placebo tablets were prepared to observe flocculation properties. All tablets retained the mucosal adhesive. The new tablet composition is as follows:

[0295]

[0296] The lack of MCC in tablet S affects compressibility and leads to severe lamination. Therefore, 90mg tablets with 25mg or 50mg MCC, or a combination containing 20mg lactose and 25mg MCC, are further prepared. Additionally, lubricants and glidants are co-sieved with besagliflozin to reduce lamination. Based on observed dissolution and flocculation, the following tablets are prepared for clinical use:

[0297]

[0298]

[0299] As previously described, the mixture of (a) besagliflozin and MCC co-sieve and (b) polyethylene oxide, poloxamer, lactose and glyceryl disorbate was then added to a mixture of (c) magnesium stearate and silica. The material was compressed to the desired hardness in a 14×6 mm sheet punch, and then tablets were coated.

[0300] Through the USP as described above <711> The release of besagliflozin was evaluated using an in vitro dissolution test (USP apparatus 1, 900 M 10.1 N HCl added, stirred at 50 rpm at 37 °C, sample taken without replacement). The table below shows the appropriate chromatographic conditions for detecting besagliflozin in 0.1 N HCl. 10 mL of sample from apparatus 1 was passed through a 10 μm PVDF filter, and 50 μL was injected into the column.

[0301]

[0302] The results of the in vitro dissolution test are as follows:

[0303] 1 8％ 8％ 6％ 5％ 3 35％ 31％ 29％ 24％ 5 62％ 57％ 51％ 44％ 8 90％ 84％ 78％ 72％

[0304] In clinical studies of patients after meals and on an empty stomach, the specific pharmacokinetic parameters of T10 and T30 tablets are as follows:

[0305]

[0306] Example 11 - Further Mucosal Adhesion Clinical Tablets

[0307] Based on the above embodiments, tablets for clinical research were prepared as follows:

[0308]

[0309]

[0310] The tablets are prepared as follows: (i) besagliflozin, colloidal silica, and 80% of MCC are co-sieved using a vibrating screen with a #20 sieve; (ii) the sieved materials are mixed in a container drum at 14 rpm (U5) or 18 rpm (U10 and U20) for 6 minutes; optionally (iii) the materials and the remaining MCC are sieved at 1000 rpm in a conical mill with an 813 μm sieve to obtain mixture 'A'; (iv) a vibrating screen with a #20 sieve is used. (v) The mixtures 'A' and 'B' are sieved through a sieve to obtain a mixture 'B'; (vi) The mixtures 'A' and 'B' are mixed in a container drum at 14 rpm; (vi) Magnesium stearate sieved through a #30 sieve is added and mixed in a container drum at 14 rpm; (vii) This material is compressed into tablet cores using a Korsch XL100 tablet press with 10 punch sets, a 20-50 rpm feeder and a 55-70 rpm turntable, or a Killian T-300 tablet press with 32 punch sets and a minimum feeder, using a 14.8 × 6.5 mm angled sheet punch and a suitable die; (viii) Dust is removed; and (ix) Coating is performed in a 600 mm (U5) or 800 mm (U10 and U20) coating pan using a suspension of 18% w / w coating material.

[0311] In an in vitro dissolution test using USP apparatus 1, with 900 mL of 0.1 N HCl, at a temperature of 37 ± 0.5 °C and a stirring speed of 50 rpm, the release of besagliflozin was as follows:

[0312] 1 10％ 9％ 6％ 3 40％ 34％ 27％ 5 66％ 58％ 48％ 8 93％ 88％ 80％ 10 95％ 96％ 94％

[0313] The tablets have been proven to be stable. U20 tablets were selected for use in clinical trials requiring 20 mg of bexaggliflozin.

[0314] Other batches of tablets were prepared in a similar but slightly different manner. For example, step (vii) was modified to use a Killian T-200 press with 19 punches. Furthermore, the concentration of the coating material in step (ix) was reduced from 18% to 12%. The tablets prepared using the modified process exhibited the desired properties.

[0315] In vitro dissolution tests of the reference batch for preparing U20 tablets showed that the release of besagliflozin was 7%, 27%, 50%, and 86% after 1, 3, 5, and 8 hours, respectively. Nine other production batches were tested (all at 1, 5, and 8 hours; five were also tested at 3 hours), and the f2 values ​​ranged from 54 to 94 compared to the reference tablets.

[0316] Example 12 - Stability Test

[0317] U20 tablets can be stored for up to 5 years at 25°C / 60% RH or 30°C / 75% RH, and according to USP <711> The dissolution characteristics of tablets at different time points (3, 6, 9, 12, 18, 24, 36, 48 and 60 months) were detected in an in vitro dissolution test simulating gastric fluid.

[0318] Figure 3 shows the average percentage of besagliflozin released from six representative preserved tablets after 1, 3, 5, and 8 hours in simulated gastric fluid. After a full 5 years of storage under both conditions, dissolution tests showed that the percentage of release remained below 17% after 1 hour, remained in the range of 20-45% (even between 23-43%) after 3 hours, remained in the range of 45-75% (even between 48-68%) after 5 hours, and remained above 80% after 8 hours.

[0319] For samples stored at 25°C, linear regression showed a very slight positive slope for the average percentage of release after 1 and 8 hours, and a very slight negative slope for the average percentage of release after 3 and 5 hours. For samples stored at 30°C, linear regression showed a very slight positive slope for the percentage of release after 8 hours, and a very slight negative slope for the average percentage of release after 1, 3, and 5 hours. However, under both storage conditions, the 95% confidence upper and lower limits for all four dissolution time points were greater than and less than zero, respectively, indicating that the slopes were not significantly different from zero. Furthermore, the small changes over time are consistent with the interpretation that the tablet release profile did not change significantly after 5 years of storage.

[0320] Example 13 - Efficacy of the U20 formulation in a randomized controlled trial

[0321] To support later-stage clinical development, seven batches of U20 tablets were prepared, five of which contained approximately 800,000 tablets. The trials were conducted as follows, each involving 200-1700 participants:

[0322]

[0323] Example 14 - Clinical pharmacology and food effects study of U20

[0324] In vivo studies provided further characterization of the U20 formulation during five clinical pharmacology studies that investigated the effects of food consumption on the pharmacokinetics of besagliflozin delivery, as well as the effects of co-administration with other drugs. The compilation below presents results only for those groups (i.e., control groups) in the studies described later that were not co-administered with other drugs.

[0325] Offer U20 tablets after fasting for at least 10 hours overnight, and do not provide food or nutrients for 4 hours after administration. Take the tablets with 240 ml of water, but do not provide water one hour before or one hour after administration. Do not use in combination with other medications.

[0326] The geometric mean of a specified number of subjects (n) is as follows:

[0327] (kg) <![CDATA[ng mL -1 ]]> <![CDATA[ng h mL -1 ]]> <![CDATA[ng hmL -1 ]]> h A 77.1(18) 125(18) 1101(18) 1154(18) 10.3(18) B 77.3(18) 117(18) 958(18) 1012(17*) 12.6(17*) C 72.4(16) 98(16) 698(16) 761(16) 12.2(16) D 77.1(20) 96(20) 703(20) 776(17*) 12.4(17*) E 72.6(24) 134(24) 1074(24) 1149(24) 11.7(24) total 75.2(96) 114(96) 900(96) 972(92) 11.8(92)

[0328] *Some subjects were unable to accurately estimate the final elimination stage.

[0329] These data demonstrate the general expected variation in pharmacokinetic parameters derived from in vivo analyses of the formulation in the subject cohort. These data also illustrate the importance of testing in a crossover design, so that each individual can use themselves as a control. The dose-standardized C of besagliflozin. max The average value was 5.7 ng / mL. -1 mg -1 In a fasting state, the dose-standardized C of immediate-release capsules containing 6.7, 16.7, and 34 mg bexaggliflozin is... max The corresponding values ​​were 12.6, 11.3, and 11.5 ng / mL, respectively. -1 mg -1 The standardized dose C for administering a 50 mg oral solution on an empty stomach. max 13.8 ng / mL -1 mg -1 AUC of 50mg oral solution 0-t 2523 ng h mL -1 This is equivalent to 1009 ng h ml of a 20 mg dose intensity. -1 Therefore, compared to rapidly absorbed oral solutions, the dose-standardized C of the U20 formulation is... max Significantly reduced, while dose-normalized AUC 0-t It decreased only slightly.

[0330] In multiple studies, the effects of eating food beforehand have been largely consistent. In a dedicated study of food effects on randomly assigned subjects, 133.7 ng / mL was observed compared to fasting. -1 C max In comparison, the geometric mean C in the postprandial state max 175.7 ng / mL -1 Or the geometric mean C under fasting conditions max 131.4%. In the postprandial state, AUC 0-t and AUC 0-∞There was also an increase, but the proportion was small, at 13.9% and 11.1% respectively. T after administration in an empty stomach max The median time to administration (T) was 3.5 hours after administration in a postprandial state. max The median duration was 5 hours. Other studies compared the pharmacokinetics after administration under different dietary conditions, with the fasting time (T0.05) being [missing information]. max The median is typically 3 hours, while the postprandial T... max The median duration is typically 5 hours. Therefore, the advantage of the formulation of this invention is that the effects of consuming a high-fat, high-calorie meal in advance are relatively moderate, and the pharmacokinetic parameters do not change significantly after administration under both dietary conditions.

[0331] In some clinical pharmacology studies, subjects took the drug in a postprandial state, defined as a fasting period of at least 10 hours followed by the consumption of a high-calorie, high-fat diet within 30 minutes of the start of the meal. They took tablets of U20 30 minutes after the start of the meal and then abstained from other food for at least 4 hours. The table below shows the results of such studies, expressed as geometric mean.

[0332] kg <![CDATA[ng mL -1 ]]> <![CDATA[ng h mL -1 ]]> <![CDATA[ng hmL -1 ]]> h F 76.1(18) 159(17) 1142(18) 1205(17*) 8.0(17*) G 79.9(16) 162(15) 1056(16) 1047(15*) 12.2(15*) H 77.1(16) 159(16) 969(16) 1035(16) 10.5(16) I 71.7(25) 176(23) 1223(23) 1276(23*) 10.9(23*) total 75.6(75) 165(71) 1106(71) 1165(71) 10.1(71)

[0333] *Some subjects were unable to accurately estimate the final elimination stage.

[0334] In a drug-drug interaction clinical pharmacology study, the effect of the GLP-1 receptor agonist exenatide on the pharmacokinetics of besaglitazone was investigated in a randomized crossover study. GLP-1 receptor agonists are known to delay gastric emptying, and the besaglitazone formulation has a gastric retention mechanism; therefore, the potential for this delay to adversely affect besaglitazone delivery was considered a concern (e.g., see Section 5.1.4.2 of the Guideline on the pharmacokinetic and clinical evaluation of modified releasedosage forms (EMA / CPMP / EWP / 280 / 96Corr1)). In this study, participants were assigned to receive either besaglitazone alone or a combination of exenatide and besaglitazone. Each group alternated between the two treatments in a crossover design (two-phase, two-treatment crossover design), with a 7-day washout period between the two treatment phases. Compared with besagliflozin alone, the systemic exposure, e.g., AUC, was lower when besagliflozin was administered 30 minutes after subcutaneous delivery of 10 pg exenatide. 0-t AUC 0-∞ and C maxThe AUC of besaglitazone combined with exenatide increased by approximately 48%, 38%, and 25%, respectively. 0-t AUC 0-∞ and C max The geometric least squares mean values ​​for besaglitazone and exenatide alone [90% confidence intervals] were 147.50% [130.23%, 167.07%], 137.56% [122.28%, 154.75%], and 125.27% [104.45%, 150.24%], respectively. Although the endpoints of the confidence intervals fell outside the 80–125% range, indicating that the interaction between exenatide and besaglitazone resulted in changes in exposure, the effect of exenatide on the pharmacokinetics of besaglitazone was not significant enough to jeopardize patient safety or warrant a change in prescribing patterns. The primary PK parameter AUC of besaglitazone in combination with exenatide was significantly lower in the same subjects compared to besaglitazone alone. 0-t and AUC 0-∞ Change <22%, C max The change was approximately 32%. When bexagliflozin was administered 30 minutes after exenatide injection, absorption was delayed, and the post-administration T... max The median was 5.00 hours, while the T after administration of besaglitazone alone was... max The median is 2.00 hours.

[0335] Example 15 - Population Pharmacokinetic Modeling

[0336] Sparse sampling of plasma drug concentrations from a large number of diverse populations, combined with pharmacokinetic models (population PK models), is a tool for exploring potential influencing factors (covariates) of drug pharmacokinetics. The population PK analysis of besagliflozin used samples from healthy volunteers or diabetic subjects participating in pharmacokinetic assessment studies, diabetic subjects participating in a sparse sampling program to obtain samples from a multicenter, international clinical trial, subjects with moderate hepatic impairment, and hypertensive subjects participating in the open-label introductory phase (the phase where all subjects received besagliflozin). Participants were recruited from North America, Europe, and East Asia. The database used for analysis contained 884 subjects and 6247 concentration records. The analysis included participants exposed to T3, T10, T30, and T90 formulations, as well as U5, U10, and U20 formulations. The majority of subjects were exposed to the U20 formulation. Subjects who consented to participate in the sparse sampling program provided three blood samples at different hours after administration (typically 6 to 8 weeks after the start of administration). The study data included administration history (dose intensity, date and time of administration), plasma concentrations and corresponding sample collection dates and frequencies, demographic descriptions, general laboratory usage, and records of concomitant medications. The model initially included the following terms: dietary status, age, weight, body mass index (BMI), body surface area, albumin, alanine aminotransferase (ALT), aspartate aminotransferase (AST), bilirubin, creatinine clearance, dose, sex, race, disease status, country, and concomitant medications.

[0337] The data were well fitted using a transport compartment model for the absorption phase and a typical two-compartment model (central and peripheral compartments) for the elimination phase. The changes in the absorption rate constant, elimination rate, and central compartment volume among individuals all followed a log-normal distribution, although the actual distribution had a hypertrophic tail. Overall, the final PPK model accurately described the observed data. In the PPK model, body weight, creatinine clearance, dietary status, and Asian ethnicity were important. Larger patients had lower exposure, while decreased creatinine clearance was associated with higher exposure. Postprandial status C max Decreased, but AUC and C min Similar to the fasting state. Population PK estimates of food effects contradict those of authoritative food effect studies, and item-by-item analysis of population PK study data suggests that food effect data appear to be biased against overall population data. Asian races and higher C max It is related to the clearance rate.

[0338] In a group PK simulation of a reference population composed of healthy Caucasians, C max The median was 112 ng / mL. -1 C min The median was 14 ng / mL. -1 C max With Cmin The ratio was 7.67, and the median AUC over 24 hours remained stable at 1023 ng h mL. -1 The median simulated value for diabetes in Caucasian individuals is about 10% lower. max With C min The ratio is 7.66. The C of the latter group... min The first and third quartiles were 10.6 and 20.2 ng / mL, respectively. -1 Higher than 10 ng / mL -1 Target concentration (approximately in vitro IC50) s0 10 times (C) min ≥10ng mL -1 and C max With C min A ratio of less than 10 is the design goal for long-acting sustained-release formulation development projects.

[0339] Example 16 - Clinically Acceptable Solid Dosage Forms

[0340] The inventors have provided tablet compositions and methods for their preparation, and ensured that the sustained-release formulations of this invention perform according to stringent and recognized standards in in vitro dissolution testing. However, not all in vivo behaviors of a formulation can be captured by in vitro testing. If different formulations are designed to have similar properties through different material components or different principles(s) of sustained release, then formal bioequivalence testing can confirm that the in vivo properties are similar. Such testing ensures that the absorption rate and extent of the new composition are not significantly or unacceptably altered.

[0341] The U5, U10, and U20 formulations of gastric retention tablets have been tested and demonstrated statistically significant therapeutic effects in large-scale randomized controlled trials in human patients with diabetes. To ensure that further formulations can provide similar therapeutic benefits and are clinically acceptable, each batch of tablets administered to humans should undergo a formal dissolution test using the aforementioned three-stage process (i.e., as per USP). <711> As described in Acceptance Form 2, in the test method based on USP device 1, 900 mL of 0.1 N HCl is added first, and the temperature is maintained at 37 ± 0.5 °C. The mixture is stirred at 50 rpm. During the test, the release of besagliflozin should not exceed 17% after 1 hour, 23% to 43% after 3 hours, 45% to 75% after 5 hours, and not less than 80% after 8 hours.

[0342] If any substantial changes are made to the formulation, in addition to passing these formal dissolution acceptance tests, it must be demonstrated that the tablets are at least in C0.05 equivalent to a clinically acceptable reference batch of tablets. max and AUC 0-t It exhibits in vivo bioequivalence in terms of parameters.

[0343] A formulation is clinically acceptable if (i) it has been proven effective in treating a disease or condition and is manufactured in a well-controlled and pre-defined manner within acceptable limits of its ingredients and preparation process, and passes a formal dissolution acceptance test; or (ii) it deviates from the original manufacturing range of its ingredients and / or preparation process but passes a formal dissolution acceptance test and demonstrates bioequivalence to the original formulation. This invention includes all such clinically acceptable oral solid dosage forms.

[0344] The following standards (from the FDA's March 2014 "Guidance for Industry: CMC Postapproval Manufacturing Changes To Be Documented in Annual Reports," Appendix B) describe the extent to which formulation modifications typically do not require bioequivalence documentation. Additionally, Appendix A of this guidance outlines certain other potentially acceptable changes.

[0345] 1. Any changes made to comply with the official guidelines for U20 formulations, once specified, except for relaxation of acceptance criteria or deletion of trials.

[0346] 2. Completely or partially remove ingredients that only affect the color, flavor, or odor of the formulation without changing other approved specifications.

[0347] 3. Variations in non-controlled-release excipients, expressed as a percentage (w / w) of the total formulation approved in the original application, are less than or equal to the following percentage ranges: filler (lactose monohydrate, MCC) ±5%, lubricant (magnesium stearate) ±0.25%, flow aid (colloidal silica) ±0.1%, and film coating (Opadry II blue) ±1%.

[0348] 4. If the technical grade and specifications of the excipient remain unchanged, but the supplier of the excipient changes.

[0349] 5. Changes in controlled-release excipients (polyethylene oxide, poloxamer 188, glyceryl disorbate) are less than or equal to 5%, expressed as a percentage (w / w) of the total amount of controlled-release excipients in U20. Following the change, the total weight of the dosage form and its strength should remain the same as U20.

[0350] It should be understood that the present invention has been described above by way of example only, and modifications can be made to the above examples within the scope and spirit of the present invention.

Claims

1. A tablet comprising: (a) 3 to 60 mg bexagliflozin; (b) 100 to 140 mg disorbate; (c) 50 to 75 mg of polyethylene oxide with an average molecular weight of 900,000; (d) 40 to 50 mg of lactose monohydrate; (e) 40 to 45 mg poloxamer 188; (f) 60 to 80 mg of microcrystalline cellulose; (g) 4 to 5 mg colloidal silica; and (h) 6 to 9 mg magnesium stearate.

2. The tablet as described in claim 1, characterized in that, The tablets showed plasma besagliflozin Cmax ≤ 8 ng / mL per mg of besagliflozin in fasting subjects weighing more than 60 kg.

3. The tablet as described in claim 2, characterized in that, (i) the tablet contains 10 mg besagliflozin and Cmax ≤ 80 ng / mL, (ii) the tablet contains 20 mg besagliflozin and Cmax ≤ 160 ng / mL, or (iii) the tablet contains 40 mg besagliflozin and Cmax ≤ 320 ng / mL.

4. The tablet of claim 1, wherein the T-cell release of bexaggliflozin in a fasting human subject is... max Between 2 and 6 hours.

5. The tablet of claim 1, containing 20 mg of bexaggliflozin, and the AUC of fasting human subjects. 0-t Between 600 and 1200 ng h mL -1 between.

6. The tablet of claim 1, containing 20 mg of besagliflozin, and the AUC of fasting human subjects... 0-∞ 675-1275 ng h mL -1 between.

7. The tablet of claim 1, comprising 20 mg besagliflozin, and wherein plasma C in fasting human subjects is [not specified]. max Between 80-150 ng / mL.

8. The tablet as claimed in claim 1, characterized in that, The plasma concentration C of bexagliflozin in the tablets max Median and C min The proportion of the median is less than 10.

9. The tablet as claimed in claim 1, characterized in that, It is a sustained-release tablet containing 20 mg besagliflozin, and in an in vitro dissolution test conducted using a United States Pharmacopeia (USP) apparatus at 1,900 mL 0.1 N HCl, 37 ± 0.5 °C, and 50 rpm, it released ≤ 17% besagliflozin after 1 hour.

10. The tablet as claimed in claim 1, characterized in that, It is a sustained-release tablet containing 20 mg besagliflozin, and in an in vitro dissolution test conducted using a United States Pharmacopeia (USP) apparatus at 1,900 mL 0.1 N HCl, 37 ± 0.5 °C, and 50 rpm, it released 20-45% of besagliflozin after 3 hours.

11. The tablet as claimed in claim 1, characterized in that, It is a sustained-release tablet containing 20 mg besagliflozin, and in an in vitro dissolution test conducted using a United States Pharmacopeia (USP) apparatus at 1,900 mL 0.1 N HCl, 37 ± 0.5 °C, and 50 rpm, it released 45-75% of besagliflozin after 5 hours.

12. The tablet as claimed in claim 1, characterized in that, It is a sustained-release tablet containing 20 mg besagliflozin, and in an in vitro dissolution test conducted using a United States Pharmacopeia (USP) apparatus at 1,900 mL 0.1 N HCl, 37 ± 0.5 °C, and 50 rpm, it released ≥ 80% of besagliflozin after 8 hours.

13. The tablet as described in claim 1, characterized in that, Bexagliflozin is a crystalline solid.

14. The tablet as described in claim 1, characterized in that, The tablet has a coating surrounding the tablet core.

15. The tablet as claimed in claim 14, characterized in that, The coating (i) contains polyvinyl alcohol, and (ii) accounts for 2.5-3.5% of the weight of the core.

16. The tablet as described in claim 1, characterized in that, The tablet has a hardness between 20 and 100 N and / or a brittleness ≤ 1% by weight.

17. A tablet comprising: (a) 20 mg bexagliflozin; (b) 65 mg of polyethylene oxide with an average molecular weight of 900,000; (c) 120 mg disorbate; (d) 45 mg lactose monohydrate; (e) 42 mg poloxamer 188; (f) 70 mg microcrystalline cellulose; (g) 4.5 mg colloidal silica; and (h) 7.5 mg magnesium stearate.

18. Use of the tablet according to any one of claims 1-17, characterized in that, Used to prepare drugs for treating diabetes or its symptoms.

Images