Solubility enhancement of carvedilol using drug–drug cocrystallization with hydrochlorothiazide

Increasing hydrophilicity of poorly water-soluble drugs is a major challenge in drug discovery and development. Cocrystallization is one of the techniques to enhance the hydrophilicity of such drugs. Carvedilol (CAR), a nonselective beta/alpha1 blocker, used in the treatment of mild to moderate congestive heart failure and hypertension, is classified under BCS class II with poor aqueous solubility and high permeability. Present work is an attempt to improve the solubility of CAR by preparing cocrystals using hydrochlorothiazide (HCT), a diuretic drug, as coformer. CAR-HCT (2:0.5) cocrystals were prepared by slurry conversion method and were characterized by DSC, PXRD, FTIR, Raman, and SEM analysis. The solubility, stability, and dissolution (in vitro) studies were conducted for the cocrystals. The formation of CAR-HCT cocrystals was confirmed based on melting point, DSC thermograms, PXRD data, FTIR and Raman spectra, and finally by SEM micrographs. The solubility of the prepared cocrystals was significantly enhanced (7.3 times), and the dissolution (in vitro) was improved by 2.7 times as compared to pure drug CAR. Further, these cocrystals were also found to be stable for 3 months (90 days). It may be inferred that the drug–drug (CAR-HCT) cocrystallization enhances the solubility and dissolution rate of carvedilol significantly. Further, by combining HCT as coformer could well be beneficial pharmacologically too.


Background
Pharmaceutical formulation is a complex field by itself as it needs to adapt itself to comply with the ever-growing demands of new therapeutic requirements. Many APIs exhibit poor water solubility and dissolution rate which may correlate to their suboptimal bioavailability [1]. More particularly, it is important in case of formulating solid dosage forms of drugs by oral route as it is the most preferred and patient-friendly mode of administration. There are several ways to improve oral bioavailability, like formulations using surfactants [2], liposomes [3], nanoemulsions [4], carbon nanotube [5], polymer micelles [6], and microemulsions [7] as well as nanoparticle-based methods [8]. Many of these formulations may face problems during formulations due to tedious procedures, and some of them suffer from solubility and stability issues also. Structural modification through crystal engineering is a challenging scientific way to improve biopharmaceutical properties [9]. Crystal engineering provides several possibilities to develop single-or multi-component alterations of an API, including the synthesis of pharmaceutical cocrystals. Cocrystals are crystalline complexes of active/ neutral compounds that form a unique crystalline lattice through non-covalent bonds, particularly hydrogen bonds. A distinct advantage with cocrystallization is preservation of the intrinsic pharmacological properties of APIs while altering their physicochemical profile (melting point, solubility, dissolution, etc.) [10].
Cocrystals are solid, crystalline materials that are being used increasingly for the development of APIs with improved physicochemical properties [11][12][13][14] such as solubility, stability, and bioavailability [15,16]. Cocrystallization is a crystal engineering technique which utilizes the noncovalent interactions such as hydrogen bonding, van der Waals forces, and π-π stacking and halogen bonding [17,18]. Among them, hydrogen bonding is the most notable because of its strength and directionality [19,20]. Pharmaceutical cocrystals can be conventionally classified into two types based on coformer used: API-coformer cocrystals (coformer from FDA-approved GRAS list) [21] and drugdrug cocrystals (coformer is replaced by another API). The current popularity of cocrystals need not be overemphasized as some cocrystals entered the market [Suglat® (ipragliflozin L-proline) in Japan and Entresto® (sacubitril and valsartan) in USA] and several more in late stages of development [22,23].
Drug-drug cocrystals evoked a lot of interest in crystal engineering because of tailoring the physicochemical properties of many new and existing APIs [16,24] and protection of intellectual property in terms of their novelty, utility, and non-obviousness [25,26]. It represents a promising area of research especially in the light of the fact that combined therapies are frequently prescribed for effective treatment options in a large number of clinical conditions. Cocrystals with multiple APIs might overcome the drawbacks of conventional combination therapy. The drugdrug cocrystals may not only lead to the modification of the fundamental properties but also provide new opportunities for the development of combined drug therapies which could offer potential advantages like synergistic and/ or additive effects [27], reduction of the dose, cost-effective for treatment [28,29] besides offering improved patient compliance [30,31]. A good number of such attempts involving drug-drug cocrystals that are recorded in literature indicate their emerging popularity. A consolidated list of drug-drug cocrystals, their literature citation/preclinical development phase has been given chronologically in Table  1. Number of traditional techniques have been employed to address the poor solubility issues of CAR. Those include cyclodextrins [43], solid lipid nanoparticles, [44] and selfemulsifying drug delivery systems [45].
In the present study, an attempt has been made to produce drug-drug cocrystals of CAR with HCT as coformer. It is well-known that the therapeutic efficacy of CAR can be increased when used in combination with HCT as adjuvant therapy [46]. This antihypertensive drug (CAR) is free from raising metabolic concerns when combined with HCT [47]. Based on these obvious clinical advantages, and both have functional groups that are required to form hydrogen bond between CAR and HCT, this combination has been chosen for the present study. Simple combination therapy can give an additive or synergistic effect, but when we apply crystal engineering approaches as cocrystallization, it can potentially bring about certain distinct advantages like improved pharmaceutical profiles of drugs and more importantly an improved therapeutic benefit [16]. Further, CAR produces more beneficial effects than the other beta blockers like atenolol/metoprolol by improving glucose, lipid metabolism, and lipid peroxidation in hypertensive and diabetic patients [48][49][50]. But poor aqueous solubility of CAR is a major obstacle in formulation development, which leads to poor bioavailability. We aimed to prepare drug-drug cocrystals of CAR ( Fig. 1) with HCT ( Fig. 2) as coformer using cocrystallization technique and study the solubility and consequently bioavailability, and the results are described in this communication.

Materials
Pure APIs (I.P. grade) CAR (m.r. 114-117°C) and HCT (m.r. 269-272°C) were purchased from Yarrow Chem Products Mumbai, India, and were used without further purification. Ethanol (99.9%, AR grade) and hydrochloric acid (HCl) were procured from Himedia, Mumbai, India, and FINAR Chemicals, Ahmedabad, Gujarat, India, respectively. All other reagents of analytical grade were purchased from commercial sources and were used directly. Melting points were measured in open capillaries on a melting point apparatus (MEL-TEMP, Barnstead, International 1001D, USA) and are uncorrected. Water filtered through a double-distilled water purification system (EROSE 1077, India) was used in all solubility experiments.

Slurry conversion method (SC) [35]
Cocrystals of CAR (MW 406.5 g/mole) and HCT (MW 297.74 g/mole) were prepared by the SC method with 2:0.5 molar ratio (i.e., 2mM of CAR and 0.5mM of HCT were taken in a flat bottom flask, and 5-8mL of ethanol was added). Magnetic stirrer (REMI 1ML, Q-19, India) was used for stirring the suspension mixture at room temperature for 24 h. The resultant slurry was dried, and the solid obtained was stored in a desiccator for their characterization.
Solvent evaporation method (SE) [51,52] CAR-HCT (2:0.5) cocrystals, prepared by SC method were taken in glass beaker and added 5-8 ml of ethanol to dissolve. Then, the beaker was slightly heated and stirred with glass rod until complete dissolving of the solute. To this solution, a pinch of charcoal was added and filtered. The resulted solution was kept for evaporation at room temperature until it grows a single crystal. After 5 days, colorless rectangular-shaped single crystals were formed.

Melting point determination
The m.p./m.r. is a characteristic physical property of solids (the temperature at which solid phase exists in equilibrium with the liquid phase). It was determined using melting point apparatus, in open capillaries and is uncorrected.

Differential scanning calorimetry (DSC)
Thermal data were obtained by differential scanning calorimeter (Perkin Elmer 4000, USA). About 3-5mg of samples (CAR and HCT) were accurately weighed using electronic analytical balance (KERRO P7, BL-2204, India) were taken into DSC sample pans (Perkin Elmer, 20 μL) and were crimped hermetically. The samples were scanned in the temperature range of 30 to 300°C at a heating rate of 10°C/min under a continuously purged dry nitrogen flow rate 50 ml/min. An empty pan was used as a reference. Data analysis was performed using the Pyris software.

Powder X-ray diffraction (PXRD)
The PXRD measurements were carried out with X-ray powder diffractometer (PANalytical XPERT-PRO, PW3040/ 60, the Netherlands) operating at 40 kV, 40 mA, using Cu-Kα radiation (λ = 1.5418 Å) with X' Celerator RTMS (real-time multiple strips) detectors. The powder samples (100-200 mg) were placed in a sample holder (2-mm-thick aluminum plate with 20-mm square thick center hole), and the samples were scanned from 6 to 40°, 2θ at a step size of 0.0167°a nd step time of 0.5 s. PXRD data and diffractograms were analyzed using the PANalytical X-pert High score plus Data viewer software version 1.4. Fourier transform-infrared spectroscopy (FTIR) FTIR spectra were recorded individually by a FTIR spectrophotometer (Bruker, Spectrum RXI, USA). One milligram (1 mg) of sample was mixed with 60 mg of potassium bromide (KBr) and pressed to form a pellet. Scans were recorded in the range of 4000-400 cm −1 at a spectral resolution of 4 cm −1 using extended scanning mode of the instrument. The data were developed and analyzed using Omnic 6.2.

Raman spectral studies (RS)
The Raman spectra were recorded using Bruker RFS 27: Stand-alone FT-Raman spectrometer (Bruker Optics, Ettlingen, Germany) equipped with a germanium detector. Spectra were acquired using an excitation wavelength of 1064 nm of Nd:YAG laser radiation (power 250 mW). The data presented using Opus 5.5 version.

Scanning electron microscope (SEM)
The morphology of the samples was investigated by high resolution scanning electron microscope (JEOL 5400, Japan). The small amount of sample was dispersed on to a carbon tube (double adhesive carbon-coated tape) adhered to an aluminum stub. These sample stubs were coated with a thin layer of gold by employing POLARAN-E3000 sputter coater. SEM analysis was carried out using an accelerating voltage of 20 kV after gold sputtering.

Accelerated stability studies (AST)
Accelerated stability studies (AST) were performed (for 90 days) as per ICH guidelines. The samples were kept in a photostability chamber (BLS 30, 12 CFT, Samiksha Industries Co, India) at 40 ± 2°C, 75% RH ± 5%. The samples of cocrystals (n = 3) were taken out at an interval of 0, 1, 2, and 3 months and were subjected to PXRD analysis (PANalytical X-pert High score plus Data viewer software version 1.4.) to ascertain the crystallinity of the prepared cocrystals.
Solubility studies [53] The solubility of pure APIs and prepared cocrystals were performed in distilled water, hydrochloric acid solution (0.1 N HCl), and phosphate buffer solutions pH 6.8 and pH 7.4. Excessive amounts of pure APIs and cocrystals were added to 10 mL of the above-mentioned dissolution media separately in a vial and then agitated continuously at room temperature for 24 h using a water bath mechanical shaker (REMI RSB-12, India). After equilibrium is attained, the solutions were filtered through Whatman filter paper (pore size 0.45 μm, GE Health Care Life Sciences, USA); filtrates were suitably diluted and analyzed on UV-visible double beam spectrophotometer (ELICO SL 210, India) at wavelength (λ) 248 nm (for CAR), and no interference was observed by HCT at this wavelength. The results were analyzed (Graphpad prism, version 5 software USA) and presented as the mean of three replicates.
Phase transformation studies [13] The prepared cocrystals were subjected to assess their stability in aqueous medium, and the study was conducted by agitating the cocrystals for 24 h using a magnetic stirrer (REMI IML, Q-19, India) at 300 rpm. The resultant slurry was filtered, dried, and was subjected to PXRD studies.
Dissolution studies (in vitro) [13] The dissolution studies were conducted in 900 mL of dissolution medium (0.1 N HCl) at 50 rpm maintained at 37 ± 0.5°C in a dissolution apparatus (USP-certified Electrolab TDT-08 L Mumbai, India) using the paddle apparatus (Apparatus 2, USP). The drug (CAR, 100 mg) or its equivalent cocrystals were added to the dissolution medium, and the samples were withdrawn at regular time intervals (0, 5, 10, 15, 20, 25, 30, 45 min) for up to 1 hour. The volume of the dissolution medium was adjusted to 900 mL by replacing it each time the sample was drawn with fresh medium in order to maintain the sink conditions. The samples were immediately filtered through a 0.45-μm membrane filter (E. Merck Pvt Ltd, India), suitably diluted, and analyzed spectrophotometrically (ELICO SL 210 Double beam UV-Visible Spectrophotometer) at 248 nm.

Results
Designing of cocrystal is the primary aspect of the cocrystallization technique. Both CAR and HCT contain hydrogen bond donor and acceptor groups. Based on hydrogenbonding principles, we predicted that the possible intermolecular interactions between the CAR and HCT may be amino-amino interactions and same as depicted in Fig. 3 and was verified using spectral analysis.

Meting point determination
The m.p. of the CAR-HCT cocrystal was found to be 125.81°C. The pure CAR and HCT m.p. were found to be at 116.52°C and HCT 272.30°C respectively.

DSC studies
DSC thermograms of the pure CAR, HCT, and CAR-HCT cocrystals were shown single melting endothermic peaks with onset temperatures of 117.12°C, 276.03°C, and 126.31°C respectively (Fig. 4).

RS studies
The RS analysis was carried out to investigate the intermolecular interactions between the participating APIs. The spectrum for pure CAR has a strong intensity signal at 3344.18 cm −1 , 2996.23 cm −1 , and 1629.55 cm −1 corresponding to NH, CH, and C=C stretching, and the spectrum of HCT shows a strong intensity signal at 3361.71 cm −1 corresponding to NH stretching, 2948.87 cm −1 , 1375.89 cm −1 , and 709.05 cm −1 corresponding to CH, S=O, and C-Cl stretching respectively (Fig. 7) ( Table 3). The spectrum of CAR-HCT cocrystals shows NH bands shifted from 3344.18 cm −1 , 3361.71 cm −1 , to 3270.07 cm −1 , CH signal shifted to 2941.44 cm −1 , and C= C signal shifted to 1590.91.

Scanning electron microscopy (SEM)
The SEM image of the CAR appeared to be irregular in shape and of HCT was small irregular rod-shaped, and the CAR-HCT cocrystal image appeared markedly different from parent APIs (Fig. 8).

AST studies
The AST studies (performed according to the ICH tripartite guidelines) were based on PXRD data, and it was found that CAR-HCT cocrystals were fully analogous before and after 30, 60, and 90-day period (Fig. 9).

Solubility studies
As the objective is to examine the likely enhancement in solubility of CAR using cocrystallization, saturated solubility studies of CAR alone and the prepared cocrystals were performed in an aqueous medium, 0.1 N HCl, and phosphate buffers pH = 7.4 and 6.8 by subjecting the samples to shaking in a water bath mechanical shaker for 24 h at room temperature and were analyzed using double beam UV spectrophotometer. The pure CAR showed low solubility in all the dissolution media performed (aqueous medium = 0.248 mg/mL, 0.1 N HCL = 0.675 mg/mL, phosphate buffers pH 7.4 = 0.508 mg/mL and pH 6.8 = 0.553 mg/mL), whereas the CAR-HCT showed high solubility in 0.1 N HCl with 4.94 mg/mL, in pH 6.8 phosphate buffer with 2.208 mg/mL, in pH 7.4 phosphate buffer with 1.505 mg/mL, and in aqueous medium with 0.475 mg/mL (Fig. 10).

Phase transformation studies
The phase transformation studies of CAR-HCT cocrystals were conducted in line with the solubility studies. The solids obtained after the equilibrium time were dried and subjected to PXRD analysis (Fig. 11).

In vitro dissolution studies
The in vitro dissolution studies were performed in 0.1 N HCl dissolution medium. The initial release of pure CAR was 14.41% for the first 20 min, and the same was 35.91% at 1 h. In case of CAR-HCT, cocrystal initial release was 61.66% for the first 20 min, and at 1 h, the release was 98.94% (Fig. 12).

Single crystal analysis
Our attempts to generate single crystals of the CAR-HCT cocrystals were not successful. Although the formed crystal morphologically appeared as single crystal, but it was a twin crystal (Fig. 13). Perhaps it is due to the uneven molar ratio of the two APIs involved. However, efforts are in progress to get a single crystal of the cocrystal, and the results of those studies will be part of our future communications.

Discussion
Cocrystallization is currently recognized as a successful method to improve the physicochemical properties of solid APIs leading to requisite therapeutic benefit; cocrystals offer an effective optimized physicochemical (i.e., solubility, stability) and mechanical properties of the active drugs but they also offer the possibility of combined therapies that come along with intellectual property opportunities [14]. In addition, multidrug cocrystals may also offer beneficial synergistic or  additive pharmacological outcomes. The design of possible cocrystals involves the evaluation of potential non-covalent and non-ionic intermolecular interactions between two or more active ingredients; in general, cocrystal formers bear functional groups that can form hydrogen bonds [54].
In designing potential cocrystals, a key point is the possibility of intermolecular interactions between components, in particular, hydrogen bonding, that may trigger the molecular assembly and determine the supramolecular architecture [55]. The formation of a cocrystal depends upon the existence of a coformer that dramatically influences the cocrystal properties, mainly solubility, and dissolution [56]. To date, the selection of a suitable coformer is based on several predictive criteria [57]; however, the formation of a cocrystal can be validated only through experimental techniques. The most frequently used coformers are compounds bearing functional groups such as carboxylic, amide, hydroxylic, or amino groups [58]. Earlier, Gasper et al. reported CAR cocrystals [59] with markedly higher improvement in solubility and dissolution rate. The same group further formulated CAR cocrystals for gastroretentive drug delivery systems [60]. We assume that this new compound CAR+HCT can be linked together by aminoamino interactions through hydrogen bonds.
The m.p. of CAR-HCT cocrystal was markedly different and lower than the pure APIs. This is one of the typical characteristic features of cocrystal. In DSC analysis, the melting temperature of CAR-HCT cocrystal was distinct from pure CAR and HCT, which indicated the formation of a new crystalline phase. From the PXRD

Conclusion
Cocrystallization is a simple and promising approach to improve the physicochemical and pharmacokinetic behavior of poorly water-soluble drug substances. Taking the advantages of drug-drug cocrystals in overcoming the problems associated with single drug therapy, reduction of dosage, CAR-HCT cocrystals were prepared (molar ratio 2:0.5). They were characterized by thermal (DSC), diffractometric methods (PXRD) and were further confirmed by spectral analysis (FT-IR and Raman). The SEM analysis results determined that the morphological behavior of CAR-HCT was different from individual CAR and HCT. The prepared CAR-HCT cocrystals were stable up to 3 months (90 days). Their solubility was significantly improved 7.3 times in 0.1 N HCl than the pure CAR alone. The in vitro dissolution rate of cocrystals was enhanced by 2.7 times than pure CAR which may lead to enhanced bioavailability. Enhancement in solubility and subsequent dissolution rate need to be further demonstrated in biological (in vivo) models. Formulating the cocrystals in suitable mode of clinical settings is being planned to take these encouraging observations to logical conclusion.