Rapid release pharmaceutical formulations containing algal cellulose

Abstract

There is provided a pharmaceutical composition comprising cellulose obtained from algae, or a derivative of said cellulose, and an active pharmaceutical ingredient (e.g. from Type 2 and 4 BCS class), wherein the active pharmaceutical ingredient is in a predominantly amorphous form. Compositions of the invention find particularly utility as formulations comprising BCS Type 2 and 4 drugs, including NSAIDs or other drugs, that may be employed in the treatment of migraine or dysmenorrhea, as well as formulations comprising other poorly soluble active ingredients where rapid release in vivo is advantageous.

Claims

1. A pharmaceutical composition comprising cellulose obtained from algae, or a derivative of said cellulose, and an active pharmaceutical ingredient, wherein the active pharmaceutical ingredient is: at least 90% by weight amorphous; solid under ambient conditions; a Type 2 or 4 BCS active pharmaceutical ingredient; and a molecule which contains at least one aromatic ring or polycondensed cyclic structure, and the cellulose or derivative thereof is obtained from algae of the Cladophorales or Siphonocladales order.

2. The pharmaceutical composition according to claim 1, wherein the cellulose or derivative thereof is obtained from algae of the genus Cladophora.

3. The pharmaceutical composition according to claim 1, wherein the cellulose or derivative thereof is at least 80% crystalline.

4. The pharmaceutical composition according to claim 1, wherein the cellulose or derivative thereof is at least 90% crystalline.

5. The pharmaceutical composition according to claim 1, wherein the composition is formed by a process involving heat-assisted extrusion of a mixture of the cellulose or cellulose derivative and the active pharmaceutical ingredient.

6. The pharmaceutical composition according to claim 1, wherein the weight ratio of the active pharmaceutical ingredient to the cellulose is at most 1:3.

7. The pharmaceutical composition according to claim 1, wherein the active pharmaceutical ingredient is a non-steroidal anti-inflammatory drug, a steroid or a cholate.

8. The pharmaceutical composition according to claim 7, wherein the non-steroidal anti-inflammatory drug, steroid or cholate is selected from the group consisting of ibuprofen, ketoprofen, flurbiprofen, naproxen, aspirin, ethenzamide, mefenamic acid, flufenamic acid, tolfenamic acid, indomethacin, sulindac, pyroxicam, progesterone, estradiol, progestin, estrogen, cholic acid, deoxycholic acid and ursodeoxycholic acid.

9. A pharmaceutical composition comprising cellulose obtained from algae, or a derivative of said cellulose, and an active pharmaceutical ingredient as defined in claim 1, wherein the composition is obtained by a process involving heating a mixture of the active pharmaceutical ingredient and cellulose or cellulose derivative to a temperature close to or above the glass transition temperature of the active pharmaceutical ingredient, wherein the cellulose or derivative thereof is obtained from algae of the Cladophorales or Siphonocladales order.

10. The pharmaceutical composition according to claim 1, wherein the molecule also contains at least one hydrogen bond donor or hydrogen bond acceptor.

11. The pharmaceutical composition according to of claim 1, wherein the cellulose or derivative thereof is obtained from algae of the genus Cladophora, and wherein the weight ratio of the active pharmaceutical ingredient to the cellulose is from 1:3 to 1:9.

12. The pharmaceutical composition according to claim 1, wherein the active pharmaceutical agent is a steroid.

13. A method of preparing a pharmaceutical composition comprising mixing together an active pharmaceutical ingredient as defined in claim 1 and cellulose obtained from algae of the Cladophorales or Siphonocladales order, or a derivative of said cellulose.

14. The method of claim 13, further comprising processing the mixture by: heat-assisted-extrusion; static heat sealing; heat-assisted intensive mixing, mixing under reduced pressure, heating; mild grinding which does not adversely affect the pore structure of excipient; and/or co-spray drying, or rotary evaporation at reduced pressure, with a solvent, preferably wherein the solvent is a mixture of water and lower alkyl alcohol.

15. A method of treating dysmenorrhea or migraine, said method comprising administering a pharmaceutical composition as defined in claim 1 to a subject suffering from dysmenorrhea or migraine, wherein the active pharmaceutical ingredient is a non-steroidal anti-inflammatory drug.

16. The method of claim 11, wherein the weight ratio of the active pharmaceutical ingredient to the cellulose is at most 1:3.

Description

(1) FIG. 1 shows DSC profiles of ibuprofen (IBU) and cellulose. FIG. 1(a) shows the DSC profiles of pure crystalline IBU, pure MCC and Cladophora cellulose samples. The DSC profile for crystalline IBU is characterized by sharp melting endotherm at 78° C. The DSC profiles for cellulose were characterized by a water evaporation broad endotherm. In a mixture the two endothermic events will overlap due to the relatively low melting temperature of IBU. FIG. 1(b) shows the DSC profile of both physical and heated mixtures of IBU and MCC (10% IBU). Both in the physical and heated mixtures of IBU with MCC the endothermic event corresponding to melting of IBU is clearly visible, although the melting temperature is shifted slightly to lower temperatures, i.e. 73° C. (heated) vs. 76° C. (physical). FIG. 1(c) shows the DSC profile of both physical and heated mixtures of IBU and cladophora-derived cellulose (CLAD) (10% IBU). The endothermic event corresponding to melting of IBU is clearly visible in the physical mixture but it is completely absent in the heated mixture;

(2) FIG. 2 represents the XRD profile of IBU and cellulose. FIGS. 2(a) and 2(b) show the XRD profiles of IBU-MCC 10% physical and heated mixtures, respectively. The dotted line represents the XRD profile of crystalline XRD. The sharp peaks of crystalline IBU overlaid on MCC background are clearly visible in both physical and heated mixtures of IBU-MCC. FIGS. 2(c) and 2(d) show the XRD profiles of IBU-CLAD 10% physical and heated mixtures, respectively. The characteristic peaks for crystalline IBU, otherwise seen in the physical mixture of MCC, are either substantially suppressed (physical mixture) or completely disappear in the heated mixture of Cladophora;

(3) FIG. 3 shows XRD profiles of heated IBU-CLAD samples in various ratios, (i.e. 10, 20 and 30% by weight IBU relative to the IBU-CLAD mixture) and stored at 40% relative humidity (RH) for 1 (FIG. 3(a)), 2 (FIG. 3(b)), and 6 months (FIG. 3(c)). It is seen from the graph that in heated IBU-CLAD samples below 20% IBU is essentially amorphous for up to a 6-month period;

(4) FIG. 4 shows the FTIR profiles of pure ibuprofen, pure cellulose (MCC or CLAD) and IBU-cellulose (MCC or CLAD) in the regions corresponding to the stretch C═O bond (FIG. 4(a)/(c)/(e)) and C—H vibrations in the aromatic ring (FIG. 4(b)/(d)/(f)). FIG. 4(a) shows the FTIR profile corresponding to the stretch C═O bond (ca. 1720 cm.sup.−1) for pure ibuprofen. FIG. 4(b) shows the FTIR profile corresponding to the C—H vibrations in the aromatic ring (ca. 900 to 750 cm.sup.−1) for pure ibuprofen. FIG. 4(c) shows the FTIR profile corresponding to the stretch C═O bond (ca. 1720 cm.sup.−1) for IBU-MCC in physical and heated mixtures. No shift in the position of C═O peak is observed. FIG. 4(d) shows the FTIR profile corresponding to the C—H vibrations in the aromatic ring (ca. 900 to 750 cm.sup.−1) for IBU-MCC physical and heated mixture, which appear similar. FIG. 4(e) shows the FTIR profile corresponding to the stretch C═O bond (ca. 1720 cm.sup.−1) for IBU-CLAD in the physical and heated mixtures. A slight shift to the right side is observed in the heated mixture suggesting significant molecular rearrangement. FIG. 4(f) shows the FTIR profile corresponding to the C—H vibrations in the aromatic ring (ca. 900 to 750 cm.sup.−1) for IBU-CLAD physical and heated mixture. The FTIR profile for heated mixture is significantly distorted suggesting molecular interaction between cellulose and IBU involving the aromatic ring;

(5) FIG. 5 shows the in vitro dissolution profile of IBU in mixtures with cellulose (10% IBU by weight relative to the mixture). FIG. 5(a) shows the dissolution of IBU from physical and heated mixtures of IBU-MCC. The dotted line represents the dissolution of pure crystalline IBU as a benchmark. It is seen that the dissolution of IBU is slightly improved in the heated MCC mixture compared to physical mixture and pure IBU. FIG. 5(b) shows the dissolution of IBU in CLAD physical and heated mixtures. It is seen in FIG. 5(b) that the dissolution of IBU is slightly improved in the physical mixture compared to pure IBU. The most dramatic improvement in dissolution of IBU is observed in the heated IBU-CLAD mixture;

(6) FIG. 6 shows the plasma concentration of IBU following administration of a physical mixture of IBU-MCC, and physical and heated mixtures of IBU-CLAD. In each case, the starting mixture contained (10% IBU by weight relative to the mixture). It is seen in the graph that CLAD mixtures exhibit improved bioavailability compared to the IBU-MCC mixture. The fastest absorption was observed for the heated IBU-CLAD mixture;

(7) FIG. 7 shows the DSC profiles for flufenamic acid (FFA) mixtures with cellulose. The DSC profile for pure FFA (not shown) showed a distinct endotherm at 135° C. corresponding to the melting temperature of crystalline FFA. The DSC profile of MCC in mixture with FFA (FIG. 7(a)) displayed several distinguishable peaks close to the melting region of FFA, typically below the specific melting temperature. The absence of the melting peak of FFA (at 135° C.) in the DSC profiles in heated FFA-CLAD samples (FIG. 7(b)) is indicative of an amorphous structure of FFA in the heated samples, since fully amorphous materials do not exhibit a melting endotherm;

(8) FIG. 8 shows the XRD profiles of FFA with celluloses (MCC—Figs (a) and (b); Cladophora—Figs. (c) and (d)). In the physical mixtures (FIGS. 8(b) and 8(d)), the sharp peaks of crystalline FFA are visible. In the heated samples (FIGS. 8(a) and 8(c)) the peaks are significantly suppressed and in the heated CLAD sample the peaks are absent; the shift in the characteristic sharp diffraction peaks of crystalline FFA in mixture compared to pure FFA suggest molecular rearrangement and formation of polymorphs, while the absence of sharp diffraction peaks suggests amorphous FFA. As many as 8 different FFA polymorphs are known, see Lopez-Mejia et al. J. Am. Chem. Soc. 2012, 134, 9872-9875;

(9) FIG. 9 shows the FTIR results of FFA with celluloses in the regions corresponding to the stretch C═O bond (FIGS. 9(a)/(c)) and C—H vibrations in the aromatic ring (FIGS. 9(b)/(d)). FIG. 9(a) shows the FTIR profile corresponding to the stretch C═O bond (ca. 1650 cm.sup.−1) for FFA-MCC in physical and heated mixtures. No significant shift in the position of C═O peak is observed. FIG. 9(b) shows the FTIR profile corresponding to the C—H vibrations in the aromatic ring (ca. 1000 to 600 cm.sup.−1) for FFA-MCC physical and heated mixtures, which appear similar. FIG. 9(c) shows the FTIR profile corresponding to the stretch C═O bond (ca. 1650 cm.sup.−1) for FFA-CLAD in physical and heated mixtures. The region of 1000 to 600 cm.sup.−1 and 1800 to 1400 cm.sup.−1 in FTIR spectra are the most informative, since characteristic bands corresponding to the functional groups of FFA are distinctively present in these two regions. Detailed analysis of FTIR spectra is provided in S. Jabeen, T. J. Dines, S. A. Leharne and B. Z. Chowdhry, “Raman and IR spectroscopic studies of fenamates—Conformational differences in polymorphs of flufenamic acid, mefenamic acid and tolfenamic acid,” Spectrochim. Acta Part A Mol. Biomol. Spectrosc., vol. 96, pp. 972-985, 2012. The region of 1000 to 600 cm.sup.−1 is predominantly related to aromatic out-of-plane C—H deformations and benzene ring deformations. In addition to the bands corresponding to C—H vibrations in the aromatic rings, the vibrations in CF.sub.3 group appear in this region of 1000 to 600 cm.sup.−1 as well. More specifically, the characteristic bands at 889, 787 and 760 cm.sup.−1 appear due to aromatic ring deformations, while the bands at 659 and 652 cm.sup.−1 are associated with vibrations of the CF.sub.3 group. The spectral range between 1800 and 1400 cm.sup.−1 involves signals of mixed character. The bands at 1423, 1454, and 1493 cm.sup.−1 are due to in-plane aromatic C—H deformation, benzene ring stretching and C—N stretching. The bands at 1519 and 1578 cm.sup.−1 are of mixed origin and arise due to benzene ring C—N stretching as well as in-plane N—H deformation. In particular, the band at 1655 cm.sup.−1 arises due to carbonyl C═O group stretching.

(10) The bands associated with stretching and deformations in the FFA molecule were studied to reveal potential interactions between the drug and the excipient. The FTIR spectra for physical blends of Cladophora cellulose-FFA display an apparent difference of band characteristics in both selected regions, i.e. the characteristic bands appeared to shift to higher wavenumbers, become broader and generally decrease in intensity. This is observed particularly for bands corresponding to CF.sub.3 vibration at 659 and 652 cm.sup.−1, aromatic out-of-plane C—H deformations at 889 cm.sup.−1 and in-plane aromatic C—H deformation at 1422 cm.sup.−1. The characteristic FFA bands in physical mixture with MCC demonstrated no particular deviation from relevant band positions.

(11) The FTIR spectra for heated samples revealed even more remarkable deviations from the characteristic band positions of FFA in reference spectra. In the selected spectral regions, a shift in band positions is observed at nearly all bands of pure FFA for each heated cellulose-FFA mixture. The characteristic bands of FFA shifted in their wavenumber position and were further accompanied by a broader appearance and lower intensity in spectra for heated MCC-FFA. The heated Cladophora cellulose-FFA mixture expressed the most obvious shift to higher wavenumbers for characteristic FFA bands in the region of 1800-1500 cm.sup.−1. Furthermore, the appearance of these peaks was much more prominent for Cladophora cellulose than MCC. The shift in the band position at 1655 cm.sup.−1 was prominent in spectra for heated samples of Cladophora cellulose-FFA where the band position shifted to higher wavenumber in the heated sample of Cladophora cellulose-FFA. Since the band associated with carbonyl stretching at 1655 cm.sup.−1 is especially sensitive to changes in the electrostatic environment of the molecule, a shift in the wavenumber position at this band is particularly indicative of interaction involving carboxylic group of FFA and cellulose. When the strength of the carbonyl bond is weakened by an intramolecular interaction, a shift to lower wavelengths is observed. The observed shift in the band position at 1655 cm.sup.−1 in the heated mixture of MCC-FFA appeared inferior when compared to FFA in formulation with Cladophora cellulose. While some peaks for heated MCC-FFA shifted in the band position, other peaks such as 760, 1519 and 1578 cm.sup.−1 were diffuse and of remarkably low intensity, which could be due to interference from water in these samples. Overall, the results from FTIR analysis suggest that a potential interaction is present between FFA and the different celluloses, particularly for FFA in formulation with Cladophora cellulose in the heated samples;

(12) FIG. 10 shows the in vitro dissolution profiles of FFA with celluloses in simulated intenstinal fluid. The heated mixture of FFA and CLAD (FIG. 10(b)) shows an immediate and rapid increase of the FFA concentration in the solvent. This rapidity of the increase was much greater than for the physical mixture of FFA and CLAD. For heated and physical mixtures of FFA and MCC (FIG. 10(a)), the release profiles were broadly similar;

(13) FIG. 11 shows the DSC results for mixtures of Cladophora cellulose with various drugs: ketoprofen (Fig. (a)), flurbiprofen (Fig. (b)), naproxen (Fig. (c)), indomethacin (Fig. (d)), sulindac (Fig. (e)), piroxicam (Fig. (f)), flufenamic acid (Fig. (g); N.B, flufenamic acid here was mixed using Turbula Mixer and heated to 138° C. for 3 hours as opposed to Example 2), and mefenamic acid (Fig. (h)). In each case, the endothermic event visible corresponding to melting of the drug in the physical mixture is completely absent in the heated mixture;

(14) FIG. 12 shows the in vitro dissolution profile of NAP alone and in mixtures with cellulose (10% NAP by weight relative to the mixture);

(15) FIG. 13 shows the in vitro dissolution profile of PRO alone and in mixtures with cellulose (10% PRO by weight relative to the mixture);

(16) FIG. 14 shows the DSC results for normal and heated mixtures of Cladophora cellulose with β-estradiol in various proportions. At lower drug concentrations (10 wt %; FIG. 14(a)), no endothermic event is visible for the melting of the drug for the normal and heated mixtures. At higher drug concentrations (30 wt %; FIG. 14(b)), an endothermic event corresponding to melting of the drug in the physical mixture is visible but is completely absent in the heated mixture;

(17) FIG. 15 shows the XRD profile of a heated mixture of FFA and CLAD cellulose following storage for 4 months at 50° C. and 75% relative humidity; and

(18) FIG. 16 shows the TGA results for a heated mixture of FFA and CLAD cellulose before and after storage for 4 months at 50° C. and 75% relative humidity.

EXPERIMENTAL

(19) A typical process for obtaining Cladophora-derived cellulose from a suitable source (e.g. Cladophora green algae) is disclosed in Mihranyan et al. Int. J. Pharm. 2004; 269 (2), 433-442. In such a method, algae is bleached (e.g. with NaClO.sub.2 adjusted with acetic acid or another suitable buffer to pH 4-5) under appropriate conditions (e.g. at about 60° C. for about 3 hours). The solution is then cooled, filtered, washed until conductivity of the wash solution is less not more than 75 μS/cm, and dried. Typically the drying is achieved through spray-drying using an outlet temperature of not less than 95° C. The resulting product is then washed until neutrality and filtered. The filtered product may be further washed with a basic solution (e.g. 0.5M NaOH or 17.5% w/v solution) before being dried and ground, e.g. hummer-type mill, e.g. Fitz Mill type D, UK. The ground material is treated using acidic hydrolysis (e.g. by adding it to a 5% HCl solution and then heating to boiling). The solution is then cooled, filtered, washed until the conductivity of wash water is not more than 75 μS/cm and dried, e.g. spray-dried using an outlet temperature not lower than 95° C.

(20) One skilled in the art will understand that variations of this manufacturing process may be employed e.g. with respect to the bleaching procedure the use of H.sub.2O.sub.2 or an alkaline metal percarbonate salt could be employed to substitute halogen atom containing bleaches (e.g. chlorites and hypochlorites) due to environmental and safety concerns (hazardous chlorine gas evolution). Further, the hydrochloric acid may be substituted by another suitable mineral acid such as sulphuric acid or phosphoric acid in the acidic hydrolysis step.

(21) All active pharmaceutical ingredients were purchased from Sigma Aldrich with a purity of no less than 98-99%.

Example 1—Ibuprofen (IBU)

(22) Product Preparation

(23) Cladophora-derived cellulose (CLAD) was obtained from Cladophora green algae using the method disclosed in Mihranyan et al. Int. J. Pharm. 2004; 269 (2), 433-442.

(24) Physical mixtures of drug (Ibuprofen) and either microcrystalline cellulose (MCC) or cladophora-derived cellulose (CLAD) were prepared by blending the drug substance with the cellulose. Unless otherwise stated, the weight ratio between the drug and cellulose was 1:9. The surface area of the cladophora-derived cellulose was found to be 98.79 m.sup.2/g (as measured by N.sub.2 gas adsorption technique according to the Brunauer Emmett Teller (BET) method). Typically, in a glass vial 5 mg of the drug was mixed with 45 g of the cellulose using a Turbula mixer (Switzerland) for 15 minutes.

(25) Heated mixtures of drug and cellulose (e.g. MCC or CLAD) were prepared by obtaining a physical by the process above and placing it is a sealed vial. The mixture was then heated to around 10° C. higher than the corresponding melting temperature of the drug for 3 hours. All samples were used after 24 hrs from the time of preparation after cooling to room temperature.

(26) Differential Scanning Calorimetry (DSC)

(27) The DSC measurements were performed with Q 2000 TA instrument (USA). The samples were first cooled from room temperature to −40° C. and then heated to around 10° C. higher than the melting temperature at 10° C./min heating rate. Typically, 10 mg of 1:9 drug-cellulose mixture was used per measurement. For pure substances, 1 mg of drug and 10 mg of cellulose were used per measurement. The pan containing the sample was punctured to avoid overpressure. Results are shown in FIG. 1.

(28) Fourier Transform Infrared Spectroscopy (FTIR)

(29) FTIR spectrum was used to follow the interactions between the model drug and cellulose. The range for measurements was set from 4000 to 400 cm.sup.−1. The most informative regions relevant for this study included the area from 1800 to 1600 cm.sup.−1, corresponding to stretch of C═O bonds, and the area from 1000 to 600 cm.sup.−1, corresponding to C—H vibrations in the aromatic ring. The measurements were performed with Bruker Tensor 27 FT-IR according to the pellets technique using potassium bromide (KBr). The amount of model drug substance in the 1:9 drug cellulose mixture was about 2 mg. The amount of KBr used was around 200 mg. Results are shown in FIG. 4.

(30) Powder X-Ray Diffraction (PXRD)

(31) An X-ray diffractometer (D8 Twin-Twin, Bruker) with Bragg-Brentano geometry (CuKα radiation; λ=1.54 Å) was used. Results are shown in FIG. 2.

(32) Storage Stability Study

(33) Heated IBU-CLAD samples were prepared as described above containing various proportions of cellulose and drug: 10, 20 and 30% by weight IBU relative to the IBU-CLAD mixture. Each sample was stored at 40% relative humidity (RH) for up to 6 months. Results of XRD analyses at 1, 2, and 6 months are shown in FIG. 3.

(34) In Vitro Drug Release in Water

(35) The IBU release from cellulose mixtures was studied using UV-spectroscopy (U1700 Shimadzu, Japan). The concentration of IBU was monitored at λ=222 nm. During each measurement, samples were drawn every 2 minutes during 38 minutes. An IBU-MCC and IBU-CLAD mixtures of about 50 mg with 1:9 drug-to-cellulose weight ratio were used. The drug release was performed in 250 ml of deionised water. Results are shown in FIG. 5.

(36) Oral Pharmacokinetic Study in Rats

(37) Naive SPF Wistar rats (HanTac:WH strain; 5-6 weeks; 100 g) were used for pharmacokinetic studies performed by oral gavage as the administration route. Samples of ibuprofen (3 mg) with cellulose in 1:9 weight ratio were used. The dose administered to each animal was 30 mg/kg. The formulations were administered by flashing the contents of the powder vial with a total of 2 ml of purified water by oral gavage. On Day 1, blood samples were collected at the following time points in relation to dosing: 0 (pre-treatment), 15, 30, 45, 60, and 120 min after dosing. At each time point 3 animals were sampled (1 animal per sample). The collected blood samples were frozen at −20° C. until analysed.

(38) Analytical Procedure

(39) The liquid chromatography system used was a LC-10AD pump with a SIL-HTc autosampler (Shimadzu, Kyoto, Japan) and a HyPurity C18 column (3 μm particle size, 50×4.6 mm from Thermo Scientific, MA, USA) with a guard column (HyPurity C18 column, 3 μm particle size 10×4.0 mm. from Thermo Scientific, MA, USA). For detection a Quattro Ultima [Waters, Milford, Mass., USA operated in selected reaction monitoring (SRM) mode with negative electrospray ionization was used. Data analysis was performed using Masslynx 4.1 software (Micromass, Manchester, UK)].

(40) Quantitation was performed using multiple reaction monitoring (MRM) mode to monitor product ion (m/z) transitions. Ibuprofen and ibuprofen-D3 SRM transitions were m/z 204.9.160.9 and m/z 207.9.163.9, respectively. The source dependent parameters maintained for ibuprofen and ibuprofen-D3: 3.8 kV; source temperature: 125° C.; desolvation temperature: 450° C.; cone gas flow; 35 L/h and desolvation gas flow: 1000 L/h. Cone voltage (V) and collision energy (eV) were 35 and H, respectively for both ibuprofen and ibuprofen-D3.

(41) Several mobile phase A candidates varying in pH were tested to achieve the most optimal reverse phased HPLC separation such as 0.1% formic acid, 0.005% formic acid and 5 mM ammonium acetate. The latter was selected. HPLC separation was performed using 5 mM ammonium acetate as a mobile phase A (MPA) and 5 mM ammonium acetate in 90:10 (v:v) acetonitrile:water as mobile phase B (MPB). The flow rate was 0.80 mL/min and the column temperature was RT ° C. Isocratic elution with 45% MPB was used. The autosampler temperature was 4° C. and the injection volume was 10 μL. The retention time for ibuprofen and ibuprofen-03 was 2.07 min and the total run time was 4 min. A basic autosampler wash, 50:50 (v:v) water:methanol, was used to reduce carryover.

(42) Acetonitrile, methanol, ammonium acetate and formic acid were purchased from Merck (Darmstadt, Germany). The water was purified using a Milli-Q system (Millipore, Bedford, Mass.).

(43) Calibration Standards

(44) Ibuprofen (MW206.28 g/mol) stock solution corrected for purity and salt form was prepared in duplicate (IBU #Weight #1 9.71 mg in 5 ml DMSO, 9.41 mM and IBU #Weight #2 5.26 mg in 5 ml DMSO, 5.0999 mM) in dimethyl sulfoxide. Ibuprofen-D3 (MW209.28 g/mol) stock solution (5.14 mg in 5 ml DMSO, 4.912 mM) was also prepared in dimethyl sulfoxide. All stock solutions were stored at −20° C. Intermediate stock solutions in acetonitrile were kept at 4° C.

(45) Calibration standards were prepared by spiking blank plasma from three male Sprague-Dawley rats with ibuprofen. Calibration standard concentrations were initially selected as 5.1, 10.2, 51, 102, 510, 1020, 1530, 1785, 2040 and 5100 nM. Due to unexpectedly high concentration of ibuprofen in several samples additional calibration standards were prepared as 10200, 20400, 15300, 51000, 76500 and 100000 nM Calibration-standards were stored at −20° C. Quadratic regression analysis with 1/y weighing was performed to quantify the concentration of the standards. The determination coefficient (R2) was greater than or equal to 0.99.

(46) Analytical Sample Preparation

(47) Prior to analysis, all frozen pre- and post-treatment samples and calibration standard samples were thawed and allowed to equilibrate at room temperature. To an aliquot of 50 μL of plasma sample 100 μL of ice-cold 0.1% formic acid in acetonitrile spiked with 200 nM ibuprofen-D3 was added. Further samples were vortexed for 20 s and centrifuged at 10.000 g for 3 min at room temperature. One hundred μL of the supernatant was mixed with 100 μL mobile phase A (5 mM ammonium acetate followed by vigorous vortexing and centrifugation at 10.000 g for 1 min. Ten μL were injected into the column.

(48) Data Review

(49) All chromatograms were reviewed to ensure that chromatographic peak shape and peak integration was satisfactory. Run acceptance criteria were set prior the analysis based on the results from calibration standards.

(50) Results

(51) Table 4 summarizes the results of the pharmacokinetic study in rats based on the results in FIG. 6. The data support the improved bioavailability and shortened peak plasma concentration time for the heated IBU-CLAD mixture.

(52) TABLE-US-00005 TABLE 4 IBU 30 mg/kg in rats PK parameters. The results are average of 3 measurements with standard error. AUC.sub.0-t AUC.sub.0-∞ MRT, Group min (μg ml.sup.−1) Min (μg ml.sup.−1) min T.sub.1/2, min IBU-MCC 197.0 ± 20.8  419.9 ± 328.3 192 ± 202  133 ± 140 (physical mixture) IBU-CLAD 1912.8 ± 136.3 2479.1 ± 165.4 87 ± 11 60 ± 7 (physical mixture) IBU-CLAD 2323.7 ± 170.1 3026.0 ± 186.4 85 ± 11 59 ± 8 (heated mixture) MRT: mean residence time

Example 2—Flufenamic Acid (FFA)

(53) Product Preparation

(54) Physical mixtures of FFA and cellulose (MCC or CLAD) were prepared by blending the drug substance with the cellulose. Unless otherwise stated, the weight ratio between the FFA and cellulose was 1:9. The surface area of the cellulose was found to be 98.79 m.sup.2/g (as measured by N.sub.2 gas adsorption technique according to the Brunauer Emmett Teller (BET) method). Typically, in a glass vial 5 mg of FFA was mixed with 45 g of the cellulose using a Vortex mixer for 15 minutes.

(55) Heated mixtures of FFA and cellulose (e.g. MCC or CLAD) were prepared by heating the physical mixture above in a sealed vial at 120° C. for 2 hours. All samples were used after 24 hrs from the time of preparation after cooling to room temperature.

(56) DSC

(57) Thermal analysis was performed with TA instrument (Model Q-2000) on both cellulose-drug mixtures and pure substances. Samples were placed inside hermetically sealed aluminium crucibles with punctured lids, in order to avoid overpressure caused by water evaporation. An empty pan was used as a reference. The analysis was conducted in the temperature range from −40° C. to 150° C. with a heating rate of 10° C. min.sup.−1. N.sub.2 gas, at a flow of 50 mL min.sup.−1, was applied during analysis. Initially the samples were cooled from room temperature to −40° C., then heated to 150° C. and finally cooled to 25° C. again. All of the samples were stored at ambient conditions for 24 hours prior to DSC measurements. For the heated cellulose-FFA mixtures the heating conditions were 120° C. for 2 hours.

(58) The measurements were performed in triplicate, and the estimated amount of drug in the mixtures was 10 wt %. Results are shown in FIG. 7.

(59) XRD

(60) The characteristic X-ray diffraction patterns were generated using an (D8 Twin-Twin, Bruker) instrument with Bragg-Brentano geometry for both FFA, as a pure drug, and FFA in a blend with different celluloses. The samples were scanned at room temperature (25° C.), CuKα radiation was utilized (λ=1.54 Å) with 2θ angle set between 10 and 60°. Pure FFA (5 mg) was used as the reference and 50 mg of both normal and heated samples of cellulose-FFA blends (approx. 10 wt % FFA in each sample) were scanned once during the analysis. Results are shown in FIG. 8.

(61) FTIR

(62) FTIR analysis was conducted on FFA, as a pure substrate, and FFA in blend with different celluloses. Cellulose-FFA blends from both heated and normal, i.e. unheated, samples were analysed. The FTIR spectra were obtained on a Bruker Tensor 27 (Germany) with KBr pellets. A background scan on air was subtracted from all spectra using the instrument software (Opus 7.0, Bruker, Germany). The approximate sample content in 200 mg KBr pellets was 10 wt % (i.e. 1 wt % drug). The collected data was normalized with respect to C—H stretching vibration at 2897 cm.sup.−1. The FTIR spectra was collected with the following parameters: 64 scans at a spectrum resolution of 4 cm.sup.−1 over a range from 4000 to 400 cm.sup.−1. Results are shown in FIG. 9.

(63) FFA Release

(64) Calibration Standard

(65) Stock solution containing 10 μg mL.sup.−1 FFA was prepared by dissolving FFA in simulated intestinal fluid (SIF). Various amounts of stock solution between 0.1 mL and 1 mL were transferred to plastic vials and frozen at −27° C. The vials containing stock solution were freeze-dried overnight using a Scanvac CoolSafe 55-4 (LaboGene ApS, Lynge, Denmark). A total of seven working standard solutions with FFA concentration range between 0.1 and 10 μg mL.sup.−1 were prepared by dissolving the vials containing freeze-dried FFA with 1 mL of a polar solvent, consisting of Acetonitrile-DMSO (4:1, vol/vol). Fluorescence spectral measurements were performed on an Infinite M200 microplate reader by Tecan Gmbh (Austria) equipped with two monochromators (excitation and emission). Black 96-well round-bottom (Corning 96 Round Bottom, Polystyrol) microplates were used. The maximum emission intensity of the drug in the working standards was measured spectrophotometrically at λ.sub.ex=289 with a full band scan from 400 nm to 500 nm.

(66) Dissolution measurements were performed with the rotating paddle technique at 37.0±0.5° C. and 50 rpm with SOTAX (AT7 Smart, Switzerland) dissolution apparatus. Simulated intestinal fluid (SIF, enzyme-free, from Sigma) was selected as the dissolution medium and prepared by diluting 20 mL of concentrated SIF with 480 mL deionized water. Normal and heated mixtures were poured to dissolution vessels with 500 mL dissolution medium and samples of 1 ml were extracted at various time points between 15 minutes and 5 hours (15 min, 30 min, 1 h, 2 h, 3 h, 4 h and 5 h). A total of seven samples with a volume of 1 mL was collected for each cellulose-FFA formulation and passed through a syringe filter into 2 ml plastic vials. The vials were frozen at −27° C. and further freeze-dried. The vials containing the freeze-dried FFA were filled with 1 mL Acetonitrile-DMSO (4:1, vol/vol) solvent and manually shaken until the collected FFA was dissolved. Two parallel measurements were performed for each formulation. Spectrofluorometric analysis together with previously described regression analysis was used to estimate the concentration of released FFA at different time points.

(67) Results

(68) Table 5 shows the enthalpies of FFA mixtures based on the analysis of results presented in FIG. 10. It is seen from Table 5 that while the degree of crystallinity of FFA is suppressed in all mixtures with both celluloses it is only for the heated FFA-CLAD sample that FFA is fully amorphous.

(69) TABLE-US-00006 TABLE 5 Melting enthalpies of FFA in pure form and in mixtures with different celluloses. Results are presented as averages with standard deviation (n = 3). ΔH.sub.melt (J/g.sub.mix) T.sub.onset (° C.) T.sub.melt (° C.) Crl.sub.FFA.% FFA 95.1 ± 3.1  133.9 ± 0.1 135.0 ± 0.3 100 FFA-MCC (physical mixture) 8.8 ± 3.5 134.1 ± 0.0 138.0 ± 1.0 58 FFA-MCC (heated mixture).sup.a 0.5 ± 0.2 120.7 ± 1.4 123.9 ± 0.7 3 3.3 ± 0.3 130.8 ± 0.7 133.2 ± 0.2 19 FFA-CLAD (physical mixture) 4.6 ± 2.3 134.1 ± 0.0 136.1 ± 0.5 29 FFA-CLAD (heated mixture) 0 0 0 0 .sup.aTwo peaks could be detected at near the melting temperature of FFA in each sample

Example 3—Other Active Pharmaceutical Ingredients

(70) The invention is further illustrated for other NSAIDs such as ketoprofen, flurbiprofen, naproxen, mefenamic acid, indomethacin, pyroxicam, sulindac. Mixtures of drug and cellulose were prepared in the same manner as described above in respect of Example 1, i.e. for each respective active substance the mixture was heated statically to melting temperature or slightly above for 3 hours. The results of DCS measurements are shown in FIG. 11.

(71) In all these cases a similar trend was observed during solid state characterization of CLAD mixtures: in DSC profiles, no melting endotherm was observed for heated drug-CLAD mixtures as compared to the physical drug-CLAD mixture and pure crystalline drug; in XRD profiles, the sharp peaks corresponding to crystalline drug disappear or are essentially depressed for the heated drug-CLAD mixture as compared to the physical drug-CLAD mixture and pure crystalline drug; in FTIR profiles, shifts in the position for C═O group as well as significant distortion of peaks for aromatic vibrations were observed in the heated drug-CLAD mixture as compared to the physical drug-CLAD mixture and pure crystalline drug.

(72) Heated mixtures of these isopropionic acid derivatives, enolic acid derivatives, isopropionic acid derivatives, or anthranilic acid derivatives with nanocellulose each showed a significant degree of molecular rearrangement and with the drug becoming amorphous.

Example 4—Naproxen, progesterone and β-estradiol

(73) Naproxen, progesterone, and β-estradiol were used as supplied by Sigma Aldrich. Cladophora cellulose was used as supplied by FMC Corp.

(74) Product Preparation

(75) Typically, 100 mg blends containing approximately 10% drug (either Naproxen, progesterone or β-estradiol) were prepared by mixing 10 mg drug with 90 mg of cellulose powder in 1 mL glass vials. Additionally, mixtures containing 20% (by weight) drug were made for progesterone, and mixtures containing 20% and 30% (by weight) drug were made for estradiol. The vials were sealed with plastic screw caps and vortexed for 30 seconds. Each cellulose-drug blend was analysed in both heated and unheated (normal) form. In order to form the heated cellulose-drug blends, vials containing the cellulose-drug blends were placed in a preheated oil bath for 2 hours at the same temperature as the melting point of the drug.

(76) Fourier Transform Infrared Spectroscopy (FTIR)

(77) FTIR spectra were obtained for naproxen, progesterone and β-estradiol (each mixed with cellulose) using the processes described in Example 1. The approximate sample content in 200 mg KBr pellets was 10% (i.e. 1% drug). The collected data was normalised with respect to C—H stretching vibration at 2897 cm.sup.−1. The FTIR spectra were collected with the following parameters: 32 scans at a spectrum resolution of 4 cm.sup.−1 over a range from 4000 to 400 cm.sup.−1.

(78) Thermogravimetric Analysis (TGA)

(79) TGA was conducted on a Mettler TG50 apparatus (10 K min.sup.−1; 35° C. to 20° C. above the melting point of the drug) on 10-20 mg samples in aluminium crucibles under a nitrogen atmosphere (60 mL min.sup.−1). Both normal and heated 10% mixtures were analysed and compared to the pure crystalline substance of corresponding drug content, i.e. 1-2 mg. The temperature in the furnace was continuously monitored, and the heat flow curves were collected both for heating and cooling phase to record melting or re-crystallization events.

(80) The weight normalized enthalpies of melting observed in cellulose-drug mixtures were calculated using the STARe Excellence software (Mettler Toledo) and compared to the weight normalized enthalpy for the pure crystalline drug, to give a rough estimate of the degree of the drug crystallinity (“Crl”) as follows.

(81) $\begin{array}{cc}C=\frac{\Delta H_m}{\Delta H_d}\times 100& \left(1\right)\end{array}$
where ΔH.sub.melt is weight normalised melting enthalpy of the drug in a specific cellulose-drug blend [in Joules/g] and ΔH.sub.drug is the melting enthalpy of a pure crystalline drug [in Joules/g].
X-Ray Diffraction (XRD)

(82) X-ray diffraction patterns were generated using the apparatus described in Example 1. The samples were scanned at room temperature (25° C.), CuKα radiation monochromatized with a graphite crystal was utilized (λ=1.54 Å) with 2θ angle set between (10 and 45°). Pure drug (about 2 mg) was used as the reference and 10-20 mg of both normal and heated samples of blends (approx. 10% drug in each sample) were scanned once during XRD analysis.

(83) Drug Dissolution Studies

(84) In order to study the release kinetics of naproxen and progesterone in formulations with Cladophora cellulose, dissolution measurements were made on normal and heated mixtures and compared to the dissolution of the pure crystalline drug of corresponding drug content. Dissolution rate was determined by the standardized USP paddle method. Release profiles for the drug were created by spectrofluorometric analysis on samples at various time points.

(85) Standard Drug Solutions and Fluorometric Calibration

(86) Stock solutions containing 16 μg mL.sup.−1 naproxen or 15 μg mL.sup.−1 progesterone were prepared by dissolving the drug in phosphate buffered saline (pH=7.4 at 25° C.). Various amounts of stock solution (between 0.1 and 1 mL) were transferred to plastic vials. These vials were dried overnight. For naproxen (NAP), a total of six working standard solutions with NAP concentration range between 2 and 12 μg mL.sup.−1 were prepared by dissolving the contents of the vials with 1 mL of pure acetonitrile. For progesterone (PRO), the same method was used but with seven working solutions with PRO concentration range between 1.5 and 9 μg mL.sup.−1.

(87) Fluorescence spectral measurements were performed on Infinite M200 Tecan (Austria) microplate reader equipped with two monochromators (excitation and emission) and UV Xenon light source. Black 96-well round-bottom microplates (Corning 96 Round Bottom, polystyrol) were used. The excitation wavelength was set at λ.sub.ex=230 nm. The emission intensity of the drug was scanned between 280 and 400 nm. The maximum emission intensity for naproxen was at λ.sub.em=350 nm and for progesterone at λ.sub.em=300 nm.

(88) Drug Release

(89) Dissolution measurements were performed using apparatus as described in Example 2. Phosphate buffer was selected as the dissolution medium. Five hundred mL of the prepared solution were used per beaker. Normal and heated mixtures were poured into the beakers containing the dissolution medium. Samples of 1 mL were extracted at various time points between 5 minutes and 6 hours. A total of twelve samples was taken per run. The contents of each sample were evaporated until all of the water had gone. The solid residue was re-dissolved in 1 mL using acetonitrile. Two parallel measurements were performed for each formulation. The dissolution rate was estimated from the average intensities of two measurements using the calibration curve as described above.

(90) Results

(91) For naproxen, the following observations were made: in DSC profiles, a much smaller melting endotherm was observed for heated and normal drug-CLAD mixtures as compared to the pure crystalline drug; in XRD profiles, the sharp peaks corresponding to crystalline drug disappear or are essentially depressed for the heated and normal drug-CLAD mixtures as compared to the pure crystalline drug. There is also a slight reduction in the sharpness of peaks corresponding to crystalline drug in the heated drug-CLAD mixture as compared to the normal drug-CLAD mixture; in the FTIR profile for naproxen, shifts in the positions for the C═O group were observed in the heated drug-CLAD mixture as compared to the physical drug-CLAD mixture and pure crystalline drug.

(92) The drug release measurements are shown in FIG. 12. The results show substantially accelerated release and dissolution of drug from the drug-cellulose mixtures as compared to dissolution of the pure drug. Dissolution was fastest for the drug-CLAD samples.

(93) Table 6 shows the enthalpies of naproxen (NAP) mixtures based on the analysis of results obtained. It is seen from Table 6 that the degree of crystallinity of NAP is suppressed in both mixtures with cellulose. The heated NAP-CLAD sample shows the greatest degree of amorphicity for NAP.

(94) TABLE-US-00007 TABLE 6 Melting enthalpies of NAP in pure form and in mixtures with CLAD cellulose. ΔH.sub.melt (J/g) T.sub.onset (° C.) T.sub.peak (° C.) Crl.sub.NAP (%) NAP melt. −193.6 153.2 156.1 100 NAP recryst. 147.5 123.7 124.6 76 NAP-CLAD-N 1:9 −2.52 142.7 156.8 1.30 NAP-CLAD-H 1:9 −0 140.9 154.7 0

(95) For progesterone, the following observations were made: in DSC profiles, a much smaller melting endotherm was observed for the heated and normal drug-CLAD mixtures as compared to the pure crystalline drug; in XRD profiles, the sharp peaks corresponding to crystalline drug disappear or are essentially depressed for the heated and normal drug-CLAD mixtures as compared to the pure crystalline drug; in the FTIR profile for progesterone, a shift in the position for the C═O group as well as significant distortion of peaks for aromatic vibrations were observed in the heated drug-CLAD mixture as compared to the pure crystalline drug. The shift in position for the C═O group was greatly reduced for the physical drug-CLAD mixture.

(96) The drug release measurements are shown in FIG. 13. The results show accelerated release and dissolution of drug from the drug-cellulose mixtures as compared to dissolution of the pure drug. Dissolution was fastest for the drug-CLAD samples. The results showing accelerated drug release and dissolution for progesterone are interesting as this is evidence that such effects are potentially observable across a broad range of drug molecules (for example, progesterone does not contain a carboxylic acid group unlike most NSAIDs).

(97) Table 7 shows the enthalpies of progesterone (PRO) mixtures (both 10% and 20% by weight) based on the analysis of results obtained. It is seen from Table 7 that the degree of crystallinity of PRO is suppressed in all mixtures with cellulose. The two heated PRO-CLAD samples show the greatest degree of amorphicity for the drug.

(98) TABLE-US-00008 TABLE 7 Melting enthalpies of PRO in pure form and in mixtures with CLAD cellulose. ΔH.sub.melt (J/g) T.sub.onset (° C.) T.sub.peak (° C.) Crl.sub.PRO (%) PRO −108.3 127.6 130.3 100 PRO-CLAD-N 1:9 −19.1 126.3 136.4 12.9 PRO-CLAD-N 2:8 −30.2 129.0 134.3 27.8 PRO-CLAD-H 1:9 −15.2 125.8 136.4 14.0 PRO-CLAD-H 2:8 −11.5 128.9 134.0 10.6

(99) For β-estradiol, the following observations were made: in DSC profiles, no melting endotherm was observed for heated or normal drug-CLAD mixtures (10 wt % drug) as compared to the pure crystalline drug. In mixtures containing 30 wt % drug, no melting endotherm was observed for the heated mixture, whereas a melting endotherm was evident for the normal mixture; in XRD profiles, the sharp peaks corresponding to crystalline drug disappear or are essentially depressed for the heated drug-CLAD mixture as compared to the normal drug-CLAD mixture; in the FTIR profile for progesterone mixtures, alteration of peak positions for C—H vibrations was observed in the heated drug-CLAD mixture as compared to the physical drug-CLAD mixture and pure crystalline drug.

(100) Table 8 shows the enthalpies of β-estradiol (EST) mixtures (10, 20 and 30 wt % drug) based on the analysis of results obtained. It is seen from Table 8 that the degree of crystallinity of EST is suppressed in all mixtures with cellulose. For any given drug:cellulose ratio, the heated EST-CLAD sample showed the greater degree of amorphicity for EST.

(101) TABLE-US-00009 TABLE 8 Melting enthalpies of EST in pure form and in mixtures with CLAD cellulose. ΔH.sub.melt (J/g) T.sub.onset (° C.) T.sub.peak (° C.) Crl.sub.EST (%) EST −136 176.6 178.1 100 EST-CLAD-N 1:9 −10.1 169.7 177.2 7.4 EST-CLAD-N 2:8 −1.5 175.2 185.9 1.1 EST-CLAD-N 3:7 −23.5 170.4 178.3 17.2 EST-CLAD-H 1:9 0 169.2 177.0 0 EST-CLAD-H 2:8 −11.7 173.6 183.7 8.5 EST-CLAD-H 3:7 −0.9 172.3 178.4 0.7 The TGA measurements for β-estradiol are shown in FIG. 14.

Example 5—Flufenamic Acid (FFA) Storage Stability Study

(102) Heated mixtures of FFA and CLAD cellulose were prepared according to the process described in Example 2 containing 10% by weight FFA relative to the FFA-CLAD mixture. Samples were stored in gelatin capsules at 50° C. and 75% relative humidity (RH) for up to 4 months. The samples were analysed by XRD and TGA at various time points according to the methods described in Example 2. The XRD pattern obtained at 4 months is shown in FIG. 15, and the TGA results for before and after storage are shown in FIG. 16.

(103) TGA analysis showed that the degree of crystallinity of FFA was essentially 0% at all time points from 0 to 4 months, and thus remained significantly suppressed throughout the duration of the study.

(104) The XRD analysis also showed minimal evidence of an increase in the crystallinity of the FFA during storage.