Oral peptide formulations and methods of manufacture and use thereof

Abstract

The invention generally relates to oral peptide formulations and methods of manufacture and use thereof. In certain aspects, the invention provides an oral peptide pharmaceutical formulation comprising: a peptide as an active agent; a permeation enhancer; and a gel forming polymer.

Description

[0001] Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION

[0002] ORAL PEPTIDE FORMULATIONS AND METHODS OF MANUFACTURE AND USE THEREOF

[0003] Related Application

[0004] The present application claims the benefit of and priority to U.S. provisional patent application serial number 63 / 709,052, filed October 18, 2024, the content of which is incorporated by reference herein in its entirety.

[0005] Field of the Invention

[0006] The invention generally relates to oral peptide formulations and methods of manufacture and use thereof.

[0007] Background

[0008] With the recent approval of several oral peptide products, notably Rybelsus (semaglutide) in 2019, Mycapssa (octreotide) in 2020, Lupkynis (voclosporin) in 2021, there is a resurgence of interest in oral formulation strategies. This interest is also driven by the large number of peptides currently in clinical development. All of the aforementioned recently approved peptide products are based on enabling formulation technologies, addressing sub-optimal physicochemical properties that lead to the poor oral absorption often observed for beyond-rule-of-five peptides.

[0009] For semaglutide and octreotide, the peptides are poorly permeable due to their high molecular weight and hydrophilicity, and the formulations contain a permeation enhancer (PE). Semaglutide contains the PE, salcaprozate sodium (SNAC), as a tablet formulation, while octreotide is formulated as an enteric coated capsule with sodium caprylate (C8) as the PE, together with surfactants and lipids. Voclosporin is a modified form of cyclosporine and, like cyclosporine, is formulated as a self-emulsifying drug delivery system (SEDDS) by dissolving the drug in a mixture of ethanol, surfactants and lipids. In the case of voclosporin, the SEDDS formulation is used to counteract the low aqueous solubility of the drug. These enabling formulations achieve different extents of bioavailability. Even with permeation enhancers, semaglutide and octreotide achieve low oral bioavailability. The oral bioavailability of semaglutide when dosed with SNAC is 0.4-1%, while octreotide has a reported oral bioavailability of 0.7%. The bioavailability of voclosporin is estimated as approximately 50%, Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION consistent with its much higher permeability, and successful mitigation of its low solubility through the SEDDS formulation.

[0010] Summary

[0011] The invention generally relates to oral peptide formulations including permeation enhancers and methods of manufacture and use thereof. Permeation enhancers have been studied for many years, and comprise a group of chemically diverse compounds, with different proposed mechanisms of action. These include membrane fluidization and disruption of tight junctions. Ideal permeation enhancers effectively increase peptide permeability by rapidly and reversibly impacting gastrointestinal membrane properties. Reversibility is important to conserve the protective function of the gastrointestinal membrane. As demonstrated herein, high local concentrations of PEs are required to increase the membrane transport of hydrophilic peptides. This necessitates strategies that can deliver the PE to the absorption site, synchronously with the peptide. For a semaglutide formulation for example, it has been demonstrated that absorption occurs chiefly in the stomach, from an eroding formulation that exerts local effects on the gastric mucosa and fluids including buffering the pH to higher values. In contrast, an octreotide formulation may be enterically coated, to bypass the gastric environment and deliver the peptide to the small intestine.

[0012] Regardless of the site of absorption, co-localization of peptide and PE, as well as high local PE concentrations are considered important for successful improvement of permeability. Consequently, a challenge for oral drug delivery of peptides and PEs is the co-localization of peptide and PE at the membrane surface whereby the PE forms a sufficiently concentrated solution to perturb the membrane. This is challenging to achieve with a solid oral delivery system, especially when targeting the small intestine.

[0013] The invention recognizes that congruence (i.e. synchronous or near synchronous release) can be achieved with copovidone (PVPVA) formulations, whereby the polymer controls the release rate, and drug is released at a similar rate to the release rate of the neat polymer. Given the anticipated need for advanced peptide formulations of both hydrophilic and hydrophobic peptides, combined with PEs, the invention provides in certain aspects herein release properties of model peptides and PEs, alone and in combination with polymers (e.g., copovidone) to determine if congruent release of peptide and PE can be achieved for these systems. Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION

[0014] Cyclosporine and octreotide were selected as model hydrophobic and hydrophilic peptides respectively. SD and SNAC were identified as PEs of contemporary interest. Surface normalized dissolution rate measurements were performed using Wood’s dissolution apparatus.

[0015] In certain aspects, the invention provides an oral peptide pharmaceutical formulation comprising: a peptide as an active agent; a permeation enhancer; and a gel forming polymer. In other aspects, the invention provides methods for manufacturing oral peptide pharmaceutical formulation, the method comprising: combining a peptide as an active agent, a permeation enhancer, and a gel forming polymer to form a mixture; and formulating the mixture into an oral peptide pharmaceutical formulation.

[0016] In certain embodiments of the formulations and methods herein, the permeation enhancer is at least one selected from the group consisting of sodium decanoate, sodium octanoate, salcaprozate sodium, Labrasol, and a combination thereof. In certain embodiments of the formulations and methods herein, the the gel forming polymer is at least one selected from the group consisting of copovidone, polyvinylpyrrolidone, hydroxypropylmethylcellulose, polyvinylalcohol, polyacrylic acid, and a combination thereof. In certain embodiments of the formulations and methods herein, the peptide is a hydrophobic peptide. In certain embodiments of the formulations and methods herein, the peptide is a hydrophilic peptide. In certain embodiments of the formulations and methods herein, the peptide is cyclosporine or octreotide.

[0017] In certain embodiments of the formulations and methods herein, the oral pharmaceutical formulation is a solid dispersion. In certain embodiments of the formulations and methods herein, the peptide is 60% or less drug loading of the oral pharmaceutical formulation. In certain embodiments of the formulations and methods herein, the peptide is between 1-2% drug loading of the oral pharmaceutical formulation. In certain embodiments of the formulations and methods herein, a ratio of the permeation enhancer; and the gel forming polymer is 50:50 w / w.

[0018] Brief Description of the Drawings

[0019] FIG. 1 panels A-F show the chemical structures of (Panel A) octreotide acetate (OCA), (Panel B) cyclosporine (CYC), (Panel C) decanoic acid (DA), (Panel D) sodium decanoate (SD), (Panel E) Salcaprozate sodium (SNAC), and (Panel F) copovidone (PVPVA).

[0020] FIG. 2 panels A-F show percent release Vs. time profiles and normalized release rates for surface normalized dissolution of (a,b) neat components, (c,d) PVPVA from a SD: PVPVA Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION 50:50 dispersion, (e,f) OCA: SNAC: PVPVA 1 .6% DL in 50 mM pH 6.8 phosphate buffer. Error bars represent standard deviations, n = 3.

[0021] FIG. 3 panels A-D showpercent release Vs. time profdes and normalized release rates for surface normalized dissolution of (Panels A-B) PVPVA in SD:PVPVA 50:50, (Panels C-D) OCA:SD:PVPVA 1.6% DL in 50 mM pH 6.8 phosphate buffer. Error bars represent standard deviations, n = 3.

[0022] FIG. 4 panels A-E Percent release vs. time profdes for surface normalized dissolution of (Panel A) CYC:SD: PVPVA 1.6% DL, (Panel B) CYC:SD: PVPVA 20% DL, (Panel C) CYC: SNAC: PVPVA 1.6% DL, (Panel D) CYC: SNAC: PVPVA 20% DL, (Panel E) CYC: PVPVA 20% DL in 50 mM pH 6.8 phosphate buffer. Error bars represent standard deviations, n = 3.

[0023] FIG. 5 panels A-E show normalized release rates calculated after surface normalized dissolution of (Panel A) OCA:SD: PVPVA 1.6% DL, (Panel B) OCA: SNAC: PVPVA 1.6% DL, (Panel C) CYC:SD: PVPVA 1.6% and 20% DL, (Panel D) CYC: SNAC: PVPVA 1.6% and 20% DL, (Panel E) neat components. Error bars represent standard deviations, n = 3.

[0024] FIG. 6 shows surface normalized dissolution of CYC:SD: PVPVA 1.6% and 20% DL in pH 1.6 and two-stage pH shift experiment from pH 1.6 to pH 6.8 for CYC:SD: PVPVA 20% DL. Error bars represent standard deviations, n = 3.

[0025] FIG. 7 panels A-E show powder x-ray diffractograms of (Panel A) OCA:SD: PVPVA 1.6% DL, (Panel B) OCA: SNAC: PVPVA 20% DL, (Panel C) CYC:SD: PVPVA 1.6% and 20% DL, (Panel D) CYC: SNAC: PVPVA 1.6% and 20% DL, (Panel E) CYC: PVPVA 20% DL.

[0026] FIG. 8 panels A-B show particle size analysis using zetasizer at various time points during surface normalized dissolution of (Panel A) OCA:SD: PVPVA 1.6% DL, (Panel B) OCA: SNAC: PVPVA 1.6% DL.

[0027] FIG. 9 A panels A-D and FIG. 9B panels E-G show fluorescence spectra of (Panel A) various concentrations of OCA, (Panel B) OCA at 25 ug / mL in various media, (Panel C) various concentrations of SNAC, (Panel D) samples from surface normalized dissolution of OCA:SD:PVPVA in pH 6.8 phosphate buffer, (Panel E) samples from surface normalized dissolution of OCA: SNAC :PVPVA in pH 6.8 phosphate buffer, (Panel F) samples from surface normalized dissolution of OCA:SNAC:PVPVA in pH 1.6 media, (Panel G) samples from surface normalized dissolution of OCA:SD:PVPVA in pH 6.8 phosphate buffer. Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION

[0028] FIG. 10 panel A shows CMC of SD, and FIG. 10 panel B shows CMC of SD in presence of PVPVA in 50 mM pH 6.8 phosphate buffer.

[0029] FIG. 11 shows two common transient permeation enhancers.

[0030] FIG. 12 shows sodium decanoate phase behavior.

[0031] FIG. 13 shows improvement of peptide permeability by permeation enhancers.

[0032] FIG. 14 is a table showing that PEs may be rapidly absorbed, and therefore the relative release rates of PE’s and peptides is important.

[0033] FIG. 15 illustrates synchronous versus asynchronous release.

[0034] FIG. 16 shows certain model peptides: insulin, octreotide, and cyclosporine.

[0035] FIG. 17 shows an exemplary composition for insulin minitablets.

[0036] FIG. 18 shows release of insulin in different media.

[0037] FIG. 19 shows insulin solubility as a function of pH.

[0038] FIG. 20 shows impact of pes on release of insulin: biorelevant media.

[0039] FIG. 21 shows sodium decanoate solubility.

[0040] FIG. 22 shows local pH during dissolution.

[0041] FIG. 23 shows release profdes: SNAC.

[0042] FIG. 24 shows release profiles: CIO versus insulin.

[0043] FIG. 25 shows octreotide intrinsic dissolution rate and surface area normalized release from formulations.

[0044] FIG. 26 shows copovidone (PVPVA) used as the gel-forming polymer.

[0045] FIG. 27 shows intrinsic dissolution rates of octreotide and SNAC.

[0046] FIG. 28 shows comparison of intrinsic dissolution rates of octreotide, SNAC and

[0047] PVPVA.

[0048] FIG. 29 shows binary and ternary dispersions containing SNAC and PVPVPA.

[0049] FIG. 30 shows rate comparison between systems.

[0050] FIG. 31 shows ternary dispersions containing octreotide, sodium decanoate (SD) and

[0051] PVPVPA.

[0052] FIG. 32 shows details regarding the lipophilic peptide, cyclosporine.

[0053] FIG. 33 is a table providing evidence that sodium decanoate can improve permeability of cyclosporine.

[0054] FIG. 34 shows improved release by inclusion of pes in dispersions. Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION

[0055] FIG. 35 shows comparison of component release rates for SNAC.

[0056] FIG. 36 shows comparison of component release rates for sodium decanoate (SD).

[0057] Detailed Description

[0058] The invention generally relates to oral peptide formulations and methods of manufacture and use thereof.

[0059] Materials

[0060] Octreotide acetate (OCA) was purchased from APIChem Technology (Zhejiang, China). Sodium decanoate (SD) was supplied by Sigma Aldrich (MO, USA). Salcaprozate sodium (SNAC) was ordered from MedchemExpress (NJ, USA). Decanoic acid (DA) was supplied by Sigma Aldrich (MO, USA). Cyclosporine was purchased from Watson Laboratories, Inc (IL, USA). Poly vinylpyrrolidone-co-vinyl acetate (PVPVA) was supplied by BASF (Ludwigshafen, Germany) under the brand name Kollidon VA64. Ethanol (EtOH), methanol (MeOH), hydrochloric acid (HC1), trifluoroacetic acid (TFA), acetonitrile (ACN), sodium hydroxide (NaOH), sodium phosphate dibasic anhydrous (Na2HPO4), sodium phosphate monobasic monohydrate (NaFbPC .EbO) were purchased from Fischer Chemicals (Fair Lawn, NJ, USA).

[0061] Methods

[0062] Analysis of Peptide, Permeation Enhancers and Polymer

[0063] For OCA and SNAC, an Ascentis Express (Sigma-Aldrich, St. Louis, MO) 90 A C 18 column with dimensions of 15 cm x 4.6 mm and particle size of 5 pm was used for high- performance liquid chromatography (HPLC) analysis. For CYC, a Zorbax Eclipse Plus (Agilent, Santa Clara, CA) C18 column with dimensions of 4.6 mm x 250 mm and a particle size of 5 pm was employed, while for copovidone, an A2500, Aqueous GPC / SEC column (300 x 8.0 mm) (P / N CLM3016, Malvern Panalytical, Worcestershire, UK) column was used. For OCA, 30:70 v / v 0.1% TFA in water and methanol was used as a mobile phase at 0.5 mL / min with an injection volume of 30 DL with UV detection at 210 nm. For CYC, 20:80 v / v water and acetonitrile was used as a mobile phase at 2.0 mL / min with injection volume of 100 DL with UV detection at 210 nm and the column was maintained at 75 DC. For SNAC, 20:80 v / v 0.1% TFA in water and methanol was used as a mobile phase at 0.5 mL / min with an injection volume of 20 Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION uL and UV detection at 230 nm. For PVPVA, 80:20 v / v 50 mM pH 7.4 phosphate buffered saline (PBS) and methanol was used as a mobile phase at 0.5 mL / min with an injection volume of 50 uL and UV detection at 205 nm.

[0064] Preparation of Solid Dispersions

[0065] Solid dispersions were prepared with SD: PVPVA, SNAC: PVPVA and DA: PVPVA at various w / w ratios. 50:50 w / w ratios. OCA:SD: PVPVA and OCA: SNAC: PVPVA at 1.6% drug loading (DL) were prepared by keeping the permeation enhancer to polymer ratio at 50:50 w / w. CYC: SD: PVPVA and CYC: SNAC: PVPVA ASDs were prepared at 1.6%, 5%, 10% and 20% DL with a 50:50 w / w ratio of permeation enhancer to polymer. For all dispersions, the components were added to a round bottom flask, dissolved in methanol and the solvent was evaporated using a rotary evaporator (Hei-VAP Core rotary evaporator, Heidolph Instruments, Schwabach, Germany) equipped with an Ecody st EcoChyll S cooler (Ecodyst, Apex, NC, USA) with a water bath maintained at 45 degrees C. Dispersions were kept under vacuum overnight to remove excess solvent. They were then triturated using a mortar and pestle and stored in desiccator.

[0066] Powder X-Ray Diffraction (PXRD)

[0067] PXRD was used to check the crystallinity of the dispersion powders. The diffractograms of neat OCA, CYC, SD, DA and SNAC as well their PVPVA-based dispersions were collected using a Rigaku SmartLab diffractometer (Rigaku corporation, Tokyo, Japan) with data collected at 0.5 deg / min over a 2 degrees range of 4-40 degrees, a step size of 0.02° 2 degrees at 44 mA current and 40 kV voltage. Powder x-ray diffraction laboratory (PXDL) software was used to analyze the data.

[0068] Surface Normalized Dissolution Rate Experiments

[0069] Surface normalized dissolution rate experiments were performed using an intrinsic dissolution rate (IDR) Wood’s apparatus (Agilent Technologies, Santa Clara, CA). Briefly 100 mg of dispersion was added to a die with a surface area of 0.5 cm2 and compressed using a Carver press (Carver, Wabash, IN) at 1500 psi with the pressure held for 1 minute. The die was then attached to the rotating spindle at 100 rpm. Only one surface of the die was exposed to the Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION buffer where experiments were performed in a jacketed beaker maintained at 37 degrees C. Samples were collected at regular intervals every 5 minutes for first 30 minutes and then every 10 minutes until 60 minutes. Samples were analyzed for various components using the HPLC methods described above. The surface area normalized release rate was calculated using equation 1. where k is the slope of the regression line, V is the volume of dissolution medium (100 mL), S is the surface area of the die exposed to the dissolution medium (0.5 cm2) and x is the weight fraction of each component.

[0070] IDR experiments were performed on select dispersions in pH 1.6 acidic media. A two- stage dissolution experiment was also performed by carrying out surface normalized dissolution in 100 mL of pH 1.6 acid media for one hour, and then adding 15 mL of concentrated pH 7.3 phosphate buffer to increase the pH to pH 6.8 and performing surface normalized dissolution for another hour.

[0071] Particle Size Analysis Using Dynamic Light Scattering

[0072] The particle size of any colloidal species generated during surface normalized dissolution experiments at 37 degrees C was measured using a Zetasizer ULTRA by Malvern Instruments (Westborough, MA, USA) equipped with an Avalanche Photo Diode detector. Samples were taken at fixed time intervals during dissolution experiments with various dispersions in pH 1.6 acidic media and pH 6.8 phosphate buffer. The measurements were performed on filtered (0.2 degrees nylon filter) and unfiltered samples in disposable polystyrene plastic cuvettes of 10 cm pathlength at 37 degrees C to simulate the dissolution bath temperature.

[0073] Fluorescence spectroscopy

[0074] Fluorescence spectra were obtained to probe potential interactions between various components after dissolution given that OCA and SNAC are both autofluorescent. Calibration curves of emission intensity as a function of concentration for OCA and SNAC were collected using a Shimadzu RF-5301pc spectrofluorometer (Kyoto, Japan) in the concentration range of 1- Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION

[0075] 100 ug / mL. The excitation wavelength was 296 nm and emission in the range of 250 to 550 nm was detected. The excitation slit width was 10 nm and the emission slit width of 3 nm. Spectra of neat components at a fixed concentration were compared with the solution resulting from a dispersion dissolved in buffer to achieve the same concentration. Samples were also collected following release of various dispersions and spectra were obtained from these solutions.

[0076] Self-Association ofPEs in the Presence and Absence ofPVPVA

[0077] To measure the self-association of SD, a stock solution of SD was prepared at 50 mg / mL and serial dilutions were prepared in the range of 0.1- 40 mg / mL in 50 mM pH 6.8 phosphate buffer. Pyrene was used as an environment-sensitive fluorescent probe which interacts with micelles / vesicles with a change in fluorescence spectrum. The final pyrene concentration added to each SD solution was 0.4 ug / mL and the ratio of fluorescence at Ii (372 nm) / h (383 nm) was plotted against the concentration of SD. All vials were equilibrated at 37 degrees C for 30 min prior to measurements taken with a Shimadzu RF-5301pc spectrofluorometer (Kyoto, Japan). The SD concentration where a decrease in the I1 / I3 ratio was observed was taken as the critical vesicle concentration (CVC). CVC in the presence of the polymer was evaluated in a similar way by adding 1 mg / mL ofPVPVA to each sample and evaluating the pyrene fluorescence spectrum for SD concentration over the range of 0.1- 40 mg / mL, again in 50 mM pH 6.8 phosphate buffer.

[0078] Results

[0079] Surface Area Normalized Dissolution Rate Experiments

[0080] FIG. 2 panels A-F show the percent release time profiles and surface area normalized dissolution rates in 50 mM pH 6.8 phosphate buffer of neat components, as well as various OCA formulations with SNAC, while FIG. 3 panels A-D show the corresponding data for formulations with SD. Surface normalized release rates were calculated using equation 1 applied to the linear portion of the curve. Of note, neat OCA has a rapid release rate (FIG. 2 panel A) of 4.5 mg.min- '.cm'2, presumably driven by its high solubility and amorphous nature. Its intrinsic dissolution rate (IDR) is higher than that of neat PVPVA (~3 mg.min'1. cm-2). Neat SNAC also released quite fast (~2 mg. min-1. cm-2), but with a lower IDR than the peptide or polymer.

[0081] For SNAC dispersions, release of each of the components could be monitored. The surface area normalized release rate of SNAC from a binary dispersion with PVPVA is shown in Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION

[0082] FIG. 2 panel B, and it is apparent that the PE and polymer released at the same normalized release rates. The release rate of PVPVA was approximately doubled in the presence of SNAC, while the normalized release rate of SNAC increased by approximately 4-fold. Thus, complete and simultaneous release was observed within 25-30 minutes for each component from 50:50 SNAC: PVPVA dispersions. The impact of different SNAC:PVPVA ratios was also evaluated (Figure SI). While the 10:90 SNAC: PVPVA dispersion showed a slightly slower release rate for each component (complete release within 40 min), dispersions of 20:80 SNAC: PVPVA or those with higher amounts of SNAC, all showed the same rapid and congruent release of both components within 30 min. For a ternary dispersion of OCA: SNAC: PVPVA 1.6% DL, (FIG. 2 panel C), it was observed that release of each component was more rapid than that of the release rate of the individual components, and that each component released simultaneously. Thus, the release of the peptide and the PE could be controlled in the ternary system to be simultaneous. The normalized release rates of the components from the OCA dispersion with SNAC: PVPVA 50:50 were around 8 mg. min’1. cm’2.

[0083] SD release rate could not be measured due to analytical constraints, but visual observation indicated that complete dissolution from the IDR die had occurred within 7-9 minutes, indicating that it had a faster intrinsic dissolution rate than SNAC. A binary dispersion of PVPVA and SD led to rapid release of the polymer (FIG. 3 panels A-D). For the neat polymer, approximately 60 min was required for complete release (FIG. 2 panel A), while, somewhat surprisingly, this was reduced to only 10 min in the presence of SD. To further explore this effect, dispersions with different ratios of SD:PVPVA were prepared, and the polymer release rate was observed. It was found that polymer release was complete within 10 min when there was a minimum ratio of SD:PVPVA of 20:80. However, for the 10:90 SD: PVPVA dispersion, polymer release was complete over 25 min, approximately twice as fast as for neat polymer. These results suggest that the amount of SD in the formulation could be varied to control the polymer release rate.

[0084] Ternary dispersions of OCA:SD: PVPVA, containing 1.6% drug, also showed rapid release of OCA and PVPVA (and presumably SD as well, FIG. 3 panels A-D). OCA and PVPVA released at the same normalized release rate. Thus, combining the SD with PVPVA led to a highly synergistic improvement in the release of each component. The normalized release rates of components from OCA dispersions with SD: PVPVA 50:50 were around 25 mg. min’ Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION hern'2Thus, even the release of the highly hydrophilic OCA was enhanced relative to that of neat OCA when added to the SD-polymer blend.

[0085] The release profiles for various dispersions of CYC with and without a permeation enhancer are shown in FIG. 4 panels A-E, while the corresponding normalized release rates are shown in FIG. 5 panels A-E together with the IDR profile for neat amorphous drug. In contrast to OCA, the release relate of neat amorphous CYC was extremely slow, with less than 0.1% release after 6 h. At a 1.6% DL CYC: SNAC: PVPVA dispersions maintained congruent release of all components, where the release rates were similar to those seen for the components from binary dispersion of PVPVA and SNAC, as well as from the ternary dispersion with OCA. CYC:SD: PVPVA 1.6 % DL dispersions also showed rapid release of drug and polymer, at similar rate as observed for the corresponding OCA dispersion, where peptide and polymer release was complete within around 10 minutes (FIG. 4 panel B).

[0086] When the drug loading was increased to 20%, very little CYC release was observed from the binary CYC: PVPVA ASD over the experimental time period (FIG. 4 panels A and E), while PVPVA release was severely impeded. Addition of a permeation enhancer to PVPVA at a 50:50 w / w ratio improved the release of CYC to about 15% for both SD and SNAC dispersions. However, for the 20% DL, there was a loss of congruency for the CYC:SD: PVPVA dispersion (FIG. 4 panels C and E) where PVPVA continued to release rapidly, but drug did not. For CYC: SNAC: PVPVA dispersions, both polymer and SNAC continued to release to a small extent after drug release reached at plateau. XRPD patterns of the remaining dispersion after 60 min of dissolution testing showed that the remaining solid material was amorphous. Therefore, the plateau in CYC release was not due to crystallization of the peptide.

[0087] Release experiments were also conducted in conditions that mimic the gastric compartment. FIG. 5 panels A-E show that complete release of drug from CYC:SD: PVPVA 1.6% DL occurred in pH 1.6 media, while about 10% drug was released from the 20% DL. SD is expected to convert to decanoic acid at pH 1.6 and may not dissolve, hence, a two-stage pH shift dissolution experiment was performed to see if additional CYC was released at higher pH. A small increase in the amount of CYC release was observed following media transfer. However, overall, for the 20% DL dispersion, the extent of CYC release was similar for the two stage experiment, as for the single stage experiments shown in FIG. 4 panels A-E. Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION

[0088] Powder X-Ray Diffraction (PXRD)

[0089] PXRD studies were performed to evaluate the crystalline / amorphous nature of dispersions and neat components (FIG. 7 panels A-E). Neat OCA and PVPVA were amorphous whereas both as-received permeation enhancers SD, and SNAC, as well as CYC were crystalline. The dispersion of CYC: PVPVA 20% DL without any PE was amorphous whereas all other dispersions had crystallinity arising from the PE, while OCA and CYC remained amorphous in the dispersions. SNAC crystallinity appeared reduced in the dispersions based on peak broadening while SD crystallinity appeared largely unchanged by the presence of the polymer.

[0090] Size Analysis Using Dynamic Light Scattering

[0091] The sizes of colloidal species formed during surface normalized dissolution of OCA:SD: PVPVA 1.6% DL and OCA: SNAC: PVPVA 1.6% DL in pH 6.8 phosphate buffer at various time points for up to one hour are shown in FIG. 8 panels A-B. The average colloid size for the experimental duration was around 10 ± 5 nm for filtered samples. Unfiltered samples showed poor quality data with interference from large species in the samples. For release studies in pH 1.6, the average colloid size was around 10 ± 5 nm for filtered samples and around 230 ± 10 nm for unfiltered samples of OCA:SD: PVPVA for a one hour experiment.

[0092] Fluorescence Spectroscopy

[0093] Both OCA and SNAC are autofluorescent compounds (FIG. 9A panels A-D and FIG. 9B panels E-G). Neat OCA and neat SNAC fluorescence spectra at various concentrations are shown in FIG. 9A panels A and C. The fluorescence spectrum of SNAC is sensitive to pH, where the peak undergoes a red-shift at lower pH. FIG. 9A panel B shows that the fluorescence spectrum of OCA does not change between pH 6.8 and pH 1.6. The fluorescence spectrum of OCA from dissolved OCA:SD: PVPVA or OCA: PVPVA overlaps the neat OCA spectra at an equivalent concentration. This may suggest an absence of interactions between OCA with either SD or PVPVA at the concentrations tested. For OCA: SNAC: PVPVA dissolved in pH 6.8, the OCA fluorescence intensity was notably quenched relative to that of OCA alone. Similar observations were found for samples taken from surface normalized dissolution experiments at Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION various time points. Fluorescence quenching of OCA in the presence of SNAC was not observed at pH 1.6.

[0094] Self-association of PEs

[0095] FIG. 10 panel A shows the pyrene I1 / I3 ratio as a function of SD concentration. A decrease in the ratio was observed at around 2.7 mg / mL, corresponding to the CMC of SD in 50 mM pH 6.8 phosphate buffer. FIG. 10 panel B shows the corresponding data in the presence of 1 mg / mL of PVPVA, where a decrease in pyrene I1 / I3 was observed at a slightly higher value of 2.9 mg / mL.

[0096] Discussion

[0097] Effective implementation of PEs requires co-administration of the PE and drug at the absorption site. Furthermore, extensive studies have shown that the effectiveness of a PE is highly dependent on its concentration. Studies demonstrating the link between PE concentration and absorption enhancement have been performed under highly controlled conditions. These typically involve rodent intestinal perfusion studies using solutions of known PE concentration, or intra-intestinal administration to administer the formulation to a specific location.

[0098] During oral dosing, the fate of the dosage form is much less controlled, and hence achieving the optimum concentrations of PE and drug at the absorption site is much more challenging. Some of these challenges include the harsh gastric environment which may destroy the peptide, or lead to peptide but not PE solubilization, as well as dilution by secretions and variations in gastrointestinal motility which impact mixing extent and distribution of components. If the peptide and PE have different solubilities, then they will dissolve at different rates, potentially impacting their ability to co-localize at the absorption site. Importantly, PEs such as SD are rapidly absorbed (within 15-30 minutes), creating additional challenges for achieving favorable conditions for enhanced drug absorption and further emphasizing the need for synchronous drug and PE delivery at a rate sufficient to achieve an effective local PE concentration. For SD, various studies have suggested that a concentration of at least 10 mM or higher at the membrane surface is required.

[0099] Much effort has been directed to exploring new PEs as well as elucidating the mechanism of action of established PEs. However, there has been relatively little work studying the release Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION properties of components from solid oral dosage forms containing PEs. In a seminal study, Rubenstein and coworkers prepared eroding matrix as well as immediate release tablets containing sulpiride, a poorly absorbed drug, and SD. They demonstrated that enhanced drug absorption was afforded following rat intrajejunal administration using eroding tablets where drug and PE had synchronized release, relative to tablets where the PE released faster than the drug. Additionally, more rapid synchronized release yielded higher absorption extents than slower synchronized release. Their study thus demonstrated the importance of simultaneous release of drug and PE, as well as the timeframe of the release process.

[0100] When considering conventional immediate release formulations, the relative dissolution rates of the peptide and PE will depend on their solubility in surrounding media, their relative surface areas, and their diffusion coefficients. From the data shown in Figure 4e, it is apparent that OCA dissolves approximately three times as fast as SNAC under the experimental conditions employed. OCA is a highly hydrophilic peptide, with a rapid dissolution rate.

[0101] In contrast, CYC is a much more hydrophobic peptide, and the IDR data shown herein illustrates a very slow release rate. Therefore, the relative dissolution rates of peptides versus PEs from simple formulations will be highly dependent on their individual characteristics and are unlikely to be similar. For the OCA: SNAC system, for comparable particle sizes, it would be anticipated that the peptide would faster than the PE when combined in an immediate release tablet, whereas the converse would be true for CYC: SNAC. In both instances, this difference in release rates is likely to impact the effectiveness of the PE at improving peptide membrane permeability.

[0102] In particular, slower release of the peptide relative to the PE is likely to be an issue. Remarkably, forming a dispersion with PVPVA eliminates the anticipated differences in release rates, whereby each of the three components releases at the same, rapid normalized rate. Furthermore, when the peptide is present at a very low weight percent, as would be typical for a highly potent peptide, little difference is seen in release rate for OCA versus CYC, eliminating the impact of peptide solubility. When the percentage of the lipophilic peptide increases, the release rate and extent from the SNAC: PVPVA dispersion is, however, reduced although release is still rapid, albeit incomplete. Similar, rapid release rates were observed for dispersions containing the two peptides when formulated with SD and PVPVA, suggesting that the use of Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION PVPVA to achieve synchronous release of peptide and PE can be applied to different types of PEs.

[0103] Interestingly, PVPVA was able to synchronize release of peptide and PE, even though the PE was in crystalline form in the dispersion. This can likely be attributed to the gel forming properties of PVPVA. Extensive recent studies have shown that this polymer forms a gel layer of >200pm depth when immersed in water. If the PE and peptide are trapped in the gel layer, then gel layer erosion will control the release of all components. This is likely the main mechanism behind the observed synchronous release of components at low peptide loadings.

[0104] Further, coformulation of the polymer and PE led to synergistic increases in release rate for all components, that could be tailored to some extent by varying the PE:polymer ratio. The mechanism of the enhanced release rate of the PEs in the presence of PVPVA is likely due to enhanced water sorption in the binary mixture. This would increase the rate of gel erosion, and hence the rate of release of all of the components. This was confirmed for PVPVA and SNAC mixtures. In certain embodiments, it is possible to use a liquid filled capsule, such as the commercial OCA formulation. For solid oral dosage forms, they postulate that the same bolus effect as that achieved by a liquid formulation could be afforded by a solid dosage form that dissolved within 5-15 minutes. Remarkably, this is readily achieved with the SD: PVPVA dispersion, where complete dissolution from the IDR apparatus is achieved within 10 min, despite the limited surface area exposed to the release medium. The dispersion approach provides the advantages of a solid dosage form, coupled an extremely rapid solubilization process.

[0105] The observations herein are extremely promising in terms of modulating and synchronizing the release of PE and peptides from oral formulations. While we have employed PVPVA as a gel forming polymer with relatively rapid release rate as a neat polymer, there are many other pharmaceutically acceptable gel forming polymers that could also be used for this strategy. These polymers could afford faster or slower release, depending on the desired in vivo performance. This opens up many potential formulation avenues to achieve optimized release of peptides and PEs, ultimately achieving improved peptide absorption.

[0106] Amorphous Solid Dispersions (ASPs) Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION

[0107] In certain embodiments, the pharmaceutical formulation is an Amorphous solid dispersions (ASDs). Amorphous solid dispersions (ASDs) are a drug formulation technology that disperses drug molecules in an amorphous (non-crystalline) form within a polymer matrix to improve the drug's solubility and bioavailability. By stabilizing the drug in a metastable amorphous state, ASDs can achieve significantly higher drug dissolution rates compared to their crystalline counterparts. Common preparation methods include hot-melt extrusion and spray drying, utilizing polymers such as copovidone and HPMCAS to enhance solubility and stabilize the drug against crystallization. ASD’s are further described in Bhujbal et al., “Pharmaceutical amorphous solid dispersion: A review of manufacturing strategies”, Acta Pharmaceutica Sinica B, Volume 11, Issue 8, August 2021, Pages 2505-2536, the content of which is incorporated by reference herein in its entirety.

[0108] ASDs can be manufactured by several methods but the underlying principle for their formation is the same. Firstly, the lattice structure of a crystalline drug is broken and converted into a liquid state by applying heat or dissolving it in a solvent. The system is then rapidly cooled (if using heat) or dried (if using solvent), causing it to fall out of the equilibrium at the Tg. This results in the generation of a solid drug in an amorphous state. To generate an amorphous state, the liquefied drug should be cooled or dried at a sufficiently fast rate. ASD manufacturing methods can be broadly classified into solvent-based methods and melting or fusion methods. Solvent evaporation-based methods include spray drying (SD), electrospraying, and rotary evaporation, wherein the drug and polymer are dissolved in a solvent which is then evaporated to form an ASD7,80. These are suitable methods for thermolabile drugs. In melting methods, the physical mixture of the drug and the polymer is melted and solidified rapidly to form the ASD81. Although some of the methods to produce ASDs have been well-established, researchers have made consistent efforts over the past decade to further improve and understand them. In addition, novel manufacturing techniques are constantly emerging. Therefore, the purpose of this review is to provide an updated overview of manufacturing techniques for ASDs.

[0109] Different manufacturing processes will generate ASD products with different physical and functional properties. Therefore, an adequate understanding of manufacturing processes and their impact on product properties is crucial for obtaining a successful ASD product. To aid the development of robust ASDs, we will discuss the impact of formulation, equipment, and process variables together with downstream processing on the critical physical stability of ASDs for each Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION method. In addition, the advantages and limitations of each of the processes have been evaluated. Lastly, in consideration of the several variables mentioned above, selection strategies have been proposed to identify suitable manufacturing methods.

[0110] Solvent evaporation-based methods

[0111] The solvent evaporation processes for preparing ASDs essentially involve drug-polymer dissolution in organic solvent systems and their subsequent evaporation. Aqueous solvents can also be used in conjunction with organic systems to enhance polymer solubility, and / or reduce the extent of organic solvent usage. Amorphous drug-polymer dispersion is generated by rapidly evaporating the solvent from the drug-polymer solution. Because organic solvent evaporation is usually performed at temperatures well below the drug melting point, solvent evaporation is particularly suitable for thermolabile formulation systems83.

[0112] An important consideration when developing an ASD using a solvent evaporation process is the choice of a solvent system. The most challenging aspect of this method is obtaining a solvent system that can solubilize the drug-polymer system and be compatible with the formulation84, and has a low residue in the product. Poor or partial solubility of the constituents may lead to longer processing times and non-homogenous ASDs. In order to obtain the desired solvent parameters, often a combination of solvents is used. In such cases, azeotrope forming solvents such as water with ethanol (95.5%, w / w) or isopropanol (87.7%, w / w) are preferred. This is because binary solvents with different evaporation rates can cause a variable degree of supersaturation that can result in rapid precipitation of selective components at some point in the evaporation process. Such an event generates a strong potential for phase separation85. Similarly, significant differences in the solubility of components can result in a faster rate of drying and selective precipitation of the component with lower solubility on the droplet surface. This in turn could further affect ASD stability. The solvent should not affect the physical or chemical stability of the formulation constituents during the manufacturing process before being evaporated from the system. The amount of residual solvent(s) in the final ASD products must be within the acceptable values of the International Council for Harmonization Q3C (R6) guideline. This guideline defines three different classes of solvents: Classes I, II, and III. Of these, the Class I solvents are to be avoided and Class III are the most preferable. However, Class II solvents can also be used to a limited extent if Class III solvents fail to Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745

[0113] PATENT APPLICATION provide adequate manufacturing conditions. Adequate removal of the residual solvent from the final product is particularly crucial when toxic organic solvents have been employed90,91. Therefore, ASD preparation by solvent evaporation-based method is usually followed by secondary drying. Other important considerations for a solvent-evaporation based method are operator exposure to harmful organic solvents and the environmental impact of solvent waste.

[0114] Additional important solvent properties include drying efficiency, combustibility, viscosity, and toxicity. The drying efficiency is governed by the extent of heat and mass transfer which in turn depends on heat supply and solvent vapor removal. Thus, in order to obtain adequate drying efficiency, solvent parameters affecting its evaporation rate such as vapor pressure, boiling point, specific heat, the heat of vaporization, and viscosity need to be assessed84. A high feed solution viscosity can also cause sticking of the solution to the processing equipment, which can result in low product yields97. Typical solvents used for solvent evaporation methods are water98, alcohols (methanol, ethanol or isopropanol) or other organic solvents such as dichloromethane, acetone, ethyl acetate, and methyl ethyl ketone (Table 1).

[0115] Table 1. Commonly used solvents for ASD preparation.

[0116] Solvent Boiling point (°C) Solubility in water (g / 100 g) Density (at 25 °C, g / mL) Viscosity (at 25 °C, cP) Dielectric constant ICH Class (limit ppm)

[0117] Acetone 56.2 Miscible 1.049 0.295 20.7 Class 3

[0118] Butanone 79.6 29 0.805 0.4 18.51 Class 3

[0119] Butyl acetate 126.1 0.68 0.882 0.685 5.07 Class 3

[0120] Chloroform 61.7 0.795 1.498 0.536 4.81 Class 2 (60)

[0121] Dichloromethane 39.6 1.32 1.326 0.413 9.08 Class 2 (600)

[0122] Dimethyl acetamide 165 Miscible 0.937 0.92 37.78 Class 2 (1090)

[0123] Dimethyl formamide 153 Miscible 0.944 0.97 36.7 Class 2 (880)

[0124] Dimethyl sulfoxide 189 25.3 1.092 1.987 47 Class 3

[0125] Ethanol 78.5 Miscible 0.789 1.04 24.6 Class 3

[0126] Ethyl acetate 77 8.7 0.895 0.428 6 Class 3

[0127] Glycerin 290 Miscible 1.261 954 42.5 -

[0128] Isopropanol 82.6 Miscible 0.786 1.96 18.2 Class 3 Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745

[0129] PATENT APPLICATION

[0130] Methanol 64.6 Miscible 0.791 0.543 32.6 Class 2 (3000)

[0131] Tetrahydrofuran 66 Miscible 0.889 0.48 7.52 Class 2 (720)

[0132] Water 100 0.998 1 78.5 -

[0133] -, not applicable.

[0134] There are several methods based on the principle of solvent evaporation like spray drying, electrospraying, fluidized bed drying, supercritical fluid methods, spray freeze-drying, as well as many laboratory-scale methods. Each method has unique processing procedures and parameters, which may impact ASD product properties. Small changes in the processing conditions can lead to substantial differences in product characteristics and performances 6. Therefore, a fundamental understanding of different processes is essential for selecting the most appropriate manufacturing method.

[0135] Spray drying

[0136] Spray drying is one of the widely used processes for manufacturing ASDs because it is a continuous and commercially scalable drying process. The spray drying process constitutes several steps. First, the feed solution / suspension containing the drug and the polymer (and possibly other additives) is pumped into the drying chamber through a spray-nozzle. The different types of commonly used nozzles are known in the art. The two-fluid nozzle has been the most commonly used nozzle for preparing spray-dried ASDs, particularly on a laboratoryscale. The energy required to atomize the liquid is primarily provided by a gas. Liquid fed into the nozzle under low pressure can be mixed either internally or externally with the gas. Another widely used nozzle in the pharmaceutical industry is the pressurized nozzle, which solely uses the feed liquid pressure for atomization. The potential energy of the liquid is converted into kinetic energy within the pressure-swirl nozzle. Due to internal instability as well as instability arising from the interaction with the surrounding air, the annular liquid lamella disintegrates 109. The pressure-swirl nozzle does not atomize highly viscous liquids effectively. A higher solution viscosity decreases the swirl intensity and leads to a higher liquid throughput as the cross- sectional area of the liquid increases. Pressurized nozzles provide the ability to produce larger particles with better flow properties and are easy to scale-up. This is particularly advantageous for downstream processing, as it can improve powder flow, die filling, compression, and tablet uniformity. Sildenafil was spray-dried with poly (lactide-co-glycolide) using a pressurized Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION nozzle to form microparticulates (4-8 pm) of ASDs. No significant changes in physicochemical properties or in vitro drug release were observed during a scaled-up manufacturing process using the same procedure.

[0137] The choice of the feed pump depends on the feed material viscosity, the type of atomization nozzle, and the drying capacity. The droplets atomized by the nozzle come in contact with the heated gas, which causes evaporation of the solvent in the drying chamber. The duration of particle residence in the drying chamber will vary depending on the equipment and process parameters, however, it is usually in the range of a few milliseconds. Industrial spray dryers are equipped to have a gas flow rate as high as 5000 kg / h that can result in a solvent evaporation capacity of up to 400 kg / h. The dried material is carried to the cyclone separator, where the heavier particles are separated from the drying gas and collected. The finer particles are removed with the exhaust gases which are collected via a filter. Particles can deposit at the bottom of the drying chamber in some cases and may be scraped. Scraping can be done with the aid of vibratory devices, and / or compressed air. Although mechanical brushes can also be used, they might result in additional stresses. One of the concerns in using spray drying is the amount of residual solvent. Therefore, spray drying is usually followed by secondary drying.

[0138] ASD product characteristics and performance can vary significantly by fine-tuning the formulation and process parameters. Relevant manufacturing parameters are known in the art. Of these, two of the crucial processing variables are the inlet temperature and the feed rate. Optimization of these factors is essential to obtain a homogenous amorphous dispersion. Selection of the inlet temperature is dictated by the physical and chemical stability of formulation constituents and the boiling point of the solvent(s). During rapid solvent evaporation from the atomized droplet, if the surface film formed is permeable, a porous particle is formed. A hollow particle with a thicker shell is formed if the initial film is impermeable. Slower rates of evaporation can provide adequate time for molecular rearrangement. This can cause phase separation or even crystallization. The extent of phase separation / crystallization is dependent upon the strength of the drug-polymer interactions. Studies have shown that the extent of drug- polymer miscibility varies depending on the location in the spray dryer from which it has been collectedl27. Naproxen-PVP-VA ASDs collected from the cyclone of the (Pro-C-epT Micro) spray dryer showed the narrow glass transition width, indicating a higher degree of drug-polymer miscibility relative to the ASDs sampled from the collectorl27. Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION

[0139] Recently, solvent composition during spray drying has also been shown to have a significant impact on ASDs despite the complete solubilization of drug and polymer in the selected solvent system. In their work, Li et al. found that water addition to the solvent system can lead to phase separation of ASDs during spray drying even for an initially one-phase feed solution. For the investigated ritonavir-PVPVA system, phase separation appeared to be subtle at a 25% drug loading with the co-solvents of water and methanol (10:90 ratio). However, a significant reduction in drug release rate was noted for this batch. Both experimental and modeling results indicated that the extent of phase separation increased when a higher amount of water (60:40 water-methanol ratio) was added to the spray solvent. When comparing systems prepared from the same solvent composition but with varying drug-loads, higher-drug loaded ASDs were more prone to phase separation than the low-drug-loading system. However, the impact of phase separation on drug release rates of high-drug loaded ASDs was minimal, likely due to the already compromised release often seen at high drug loadings85. A change in the cosolvent ratio has also been shown to alter the surface composition of spray-dried ASDs, likely due to the varying evaporation and diffusion kineticsl28,129. These observations are particularly significant, considering a higher surface ratio of drug to polymer is known to result in an increased tendency of the amorphous drug to recrystallize 130. For an in-depth reading regarding the fundamentals of ASD particle engineering by spray drying, readers are referred to some previously published reviews.

[0140] The solubility of the drug and excipients in the feed solvent limits the output of a traditional spray drying process. If the solubility of the solute and excipient in the solvent is very different, spray-dried ASDs obtained from such solutions are often not homogeneous. Recently, modified spray drying techniques have been developed for ASD production in which the aqueous, organic, or combination feed solution is heated by a heat exchanger before being atomized and spray-dried. This process generally leads to spray-dried ASDs that are more homogeneousl36. The increase in temperature of the feed solution increases the solubility of the drug and other excipients. Such temperature increase of feed solution can reduce its viscosity and enable improved uniformity of atomizationl36. In addition, this modification allows for the rapid evaporation of the solvent and shorter times of particle solidification than conventional spray drying. However, operation safety and formulation stability concerns must be carefully evaluated before employing this method for preparing ASDs. Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION

[0141] Spray drying has some limitations. A major concern in spray-dried ASDs is solvent residue. The low bulk density ASD powder prepared by this method often needs densification to improve its flow properties for further processing into the final dosage forms. Additionally, adhesion of the material to the equipment walls reduces product yield which can be a limiting factor early in development where the developers have a limited amount of the active ingredient, especially for those expensive drugs. Adding silicon dioxide into the feed has been shown to increase bulk density, and therefore the recovery of ASD product. Nevertheless, spray drying remains one of the most popular methods due to its applicability to a wide variety of compounds and its ability to obtain a product with a high drug load and the desired particle properties by fine-tuning multiple processing variables. For example, albendazole was found to be degrading up to 97.4% during hot-melt extrusion, despite the precautions of earlier forced chemical and thermal degradation tests. However, no degradation was seen for spray-dried ASDs. The capability of spray drying to prepare high drug-loaded indomethacin-PVP ASDs was compared with co-milling and supercritical anti-solvent process in another report. The spray drying method could prepare stable formulations up to a higher drug load (80%, w / w) relative to the other processes (60%, w / w). In a very short timeframe, spray drying can generate particles with a size range from nano-to micro-meter scales. Nanoparticles of celecoxib-phospholipoid E80 and trehalose were prepared via spray drying. Microspheres (3-10 pm) of caffeine ASDs were generated when it was spray-dried with poly (lactic-co-glycolic acid) and polylactic acid. It is also a process that can be scaled up from laboratory to industrial manufacturing. For example, Sawicki et al. showed that spray drying is more suited to scale-up than freeze-drying. For a Phase one clinical trial of docetaxel or paclitaxel, spray drying was a method of choice over freeze- drying since the values of both saturation solubility and precipitation onset time of spray-dried ASDs were either similar or better to the freeze-dried ASD.

[0142] Electrospraying

[0143] In electrospraying, electrical forces atomize feed solution (containing drug and other additives) into small droplets in the range of a few nanometers or micrometers. Similar to spray drying, the rapid rate of solvent evaporation contributes to the formation of the amorphous drug state within the ASD. With atomization of the drug-containing solvent and rapid drying, the method is somewhat similar to that of spray drying. However, one of the key advantages of electrospraying over spray drying and many other techniques is its capability to produce small Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION particles, with a narrow particle size distribution. The ‘free-fall’ of the droplets with subsequent rapid solvent evaporation allows for minimal to no agglomeration of the electrosprayed particlesl48. This technique additionally offers adaptability with basic equipment designs to generate particles with the desired size, shape, and morphology.

[0144] The standard electrospraying set-up is comprised of four significant parts: a siphoning system (usually a pump), a spray nozzle set-up with a variable high voltage, and a grounded substrate. The electrically conductive feed is pumped gradually into the spray nozzle, which has an applied electrical potential difference. When the feed solution is ejected from the nozzle, at an adequately high applied voltage, the free charges on the solution surface create an electrical pressure. This results in the generation of the ‘Taylor cone’, where the meniscus at the nozzle tip is shaped like a cone. The solution at the tip of the cone has a high free charge and is pulled away rapidly towards the collector, forming a highly charged solution jet. During the flight to the collector, solvent evaporation on the primary droplet surfaces causes them to shrink. This increases charge concentration causing the droplet to undergo Coulomb fission and break into even smaller droplets. These nano- or micro-sized droplets allow for instant solvent evaporation so that only solidified particles reach the grounded substrate. Several studies have also used an additional assembly of a corona neutralizer. The neutralizer is placed opposite to or concentrically around the nozzle and is used to prevent further Coulomb fission and disruption of charged droplets to obtain monodispersed particles. Besides the Taylor cone-jet mode, many other electrospraying modes can be achieved by adjusting the electrical potential. However, the Taylor-cone mode, which emits a steady stream of microscopic jet and breaks up periodically into uniformly sized droplets is the preferred mode for the generation of monodisperse particles.

[0145] For electrospraying, the processing conditions are optimized such that the solution kinetic energy and surface tension are overcome by the electrostatic repulsion, allowing the jet to break into smaller droplets. However, if the kinetic energy in the Taylor cone and the surface tension exceeds the electrostatic repulsion, usually due to the presence of high molecular weight polymers, the charged solution jet will not break into droplets. This forms fine polymeric fibers (instead of particles) with their diameter ranging from a couple of nanometers to a few micrometers 163. This process is known as electrospinning. Although ASDs can be prepared using electrospinning, electrospraying is preferred due to its ability to generate spherical and monodisperse particles with better flow characteristics relative to electrospun fibers. Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION

[0146] Bohr et al. formulated celecoxib-poly (lactic-co-gly colic acid) (PLGA) microspheres of ASDs with electrospraying. Not only was the drug release from ASDs faster than that of a pure crystalline drug, but optimization of electrospraying allowed for varying ASD properties like particle size and porosity which helped alter drug release profiles. Yu et al. prepared ketoprofen- PVP ASD nanoparticles using electrospraying with drug-polymer ratios of 1 :10, 2: 10, and 5: 10. Almost complete drug release was achieved for all the ratios within 1 min at a 50 mg drug dose in phosphate buffer solution. Similar rapid release (<1 min) was exhibited for acetaminophen / PVP K25 prepared by electrospraying. Electrospraying can also be useful for preparing ASDs of drugs that are poorly soluble in common solvents or have a high melting point, which makes using traditional heat or solvent-based methods challenging. For instance, quercetin is one such drug with a high melting point of 326 °C. Li et al. formulated a quercetin- PVP ASD by electrospraying their dimethylacetamide and ethanol-based solution, which exhibited rapid release (<10 s) and 10-fold higher permeation rates across porcine sublingual mucosae than crystalline quercetin.

[0147] Although there are studies indicating successful ASD formation by electrospraying, there are also reports highlighting the complexity of electrospraying in preparing ASDs. When clarithromycin ASDs were prepared by electrospraying, Mohammadi et al. observed an incomplete amorphization of the drug. Similarly, even though the -EUDRAGIT (polymethacrylate-based copolymers) LI 00 ASDs prepared by Zhang et al. had a significantly enhanced drug release, incomplete amorphization (<5% crystallization) was observed in the formulation. Besides the nature of the drug, ambient pressure is one of the factors that results in incomplete amorphization during electrospraying. For example, Ny strom et al. reported the varying effect of ambient pressure on drug amorphization during electrospraying. Electrospraying the solutions of budesonide and piroxicam at low pressure (0.3 bar) led to powders exhibiting higher degree of amorphization compared to solutions electrosprayed at atmospheric pressure. On the other hand, indomethacin solutions electrosprayed at lowered pressure (0.3 bar) prompted the formation of more crystalline drug than that produced at atmospheric pressure.

[0148] An important parameter impacting the crystallinity of electrosprayed solids is the electrical field, especially for dipolar compounds. Increased interaction of the compound dipole moments with a strong electrical field, can cause molecular rearrangements that promote Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION crystallization. Also, it is important for the feed solution to have sufficient conductivity and low viscosity so that the solution jet can be broken into smaller droplets.

[0149] Since even a small trace of crystalline phase can induce subsequent crystallization of the ASD, residual crystallinity is a parameter that needs to be monitored while manufacturing ASDs, especially via electrospraying. Although there is some advancement in scale-up approaches of electrospraying like multi-spray nozzle, nozzle-free, and high-speed electrospinning, the use of electrospinning for ASD preparation is restricted due to its low production rates and complex process design. For in-depth reading regarding the fundamentals of electrospraying, readers can refer to some previously published reviews.

[0150] Fluidized bed technology

[0151] Fluid bed technology is used for various pharmaceutical unit operations including granulation (fluidized bed granulator), coating (fluidized bed coater), drying (fluidized bed dryer), and cooling. Along similar lines, fluidized bed coaters and granulators are also used to manufacture ASDs. For ASD preparation, the drug-polymer solutions are sprayed onto inert excipient cores, with the solvent evaporation and ASDs layering occurring simultaneously. The organic solvent can be recovered and recycled. This method has been used to formulate both controlled- and immediate-release solid dispersions. Direct formation of ASD granules by this method reduces additional downstream processing steps, which aids in avoiding potential stability issues during these processes. In addition, such granules allow an additional coating of suitable excipients which can control the release profdes or enhance the ASD stability. For example, indomethacin-PVP ASDs prepared by coating onto sugar spheres were further coated with various polymers to achieve the desired drug release and diffusion rates.

[0152] Based on the nozzle location, there are four types of setups available for fluidization. In both the bottom and tangential spray process, the drying air and feed solution are introduced in the same direction. This presents a more controlled particle movement and allows for a uniform ASD coating. Therefore, these two configurations are usually the preferred fluid bed technology methods to prepare ASDs. Dipyridamole controlled release ASD pellets were formulated by Beten et al.192 using the bottom spray fluidized bed process. When Ho et al. prepared nifedipine-HPMC ASDs by layering them on sugar spheres using a fluidized bed coater, they observed varying rates of drug release with varying drug-polymer ratios. Similarly, Sun et al. Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION employed a fluidized bed coating method to prepare silymarin-PVPP ASDs. Additionally, Zhang et al. demonstrated the success of this process to prepare ASDs at low air temperatures of <30 °C for thermolabile drugs. This technique was also utilized to produce wax-based (floating) ASD pellets which gave a sustained release for 12 h. The wax-based core was coated with protocatechuic acid and ethylcellulose solution in a single-step fluidized bed coating method. Fluid bed layering has also been used to prepare an amorphous complex of drug and cyclodextrins to enhance the drug dissolution profile. For ASD preparation, fluid bed layering can be challenging on a larger scale due to stickiness of the material and over-agglomeration; though there are still a few marketed products manufactured using the fluid bed process, probably because the fluidized bed offers a one-step granulation-coating-drying approach with the possibility of relatively lower operating temperatures than spray drying.

[0153] Supercritical fluids

[0154] Supercritical fluids (SCFs) have also been used to produce ASDs. SCFs are simply gases that simultaneously present gaseous and liquid state properties under specific pressure and temperature conditions. The liquid property of SCFs is useful for drug-polymer solubilization whereas the gaseous property aids in solid diffusion and solvent evaporation. Although theoretically nearly all gases can be SCFs, practically only a few are used due to the limitation of attainable temperature and pressure conditions. In fact, >98% of all SCF applications have been developed with carbon dioxide (CO2). This is because CO2 has a low critical temperature (31 °C) and pressure (7.4 MPa), making it easier to achieve adequate conditions for an SCF process. In addition, it is non-flammable, reusable, non-toxic, and inexpensive. Another gas used as an SCF in the pharmaceutical industry is trifluoromethane. The major advantage of SCF-based methods is that they are relatively greener in nature and have lower production costs compared to other solvent-evaporation processes. The solvent evaporation process can also be controlled more by adjusting the temperature and pressure conditions. The low viscosity of SCFs result in a high diffusivity and rapid solvent evaporation with faster and higher yields. Drawbacks of using SCF-based methods for ASD preparation include the difficulty in removing residual organic solvents (if used), and the high capital investment.

[0155] SCF-based processes can be divided into three groups. The first includes processes that use SCFs as solvents. An example of such a process is the rapid expansion of a supercritical Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION solution (RESS). In RESS the drug and polymer / s are dissolved in an SCF and the SCF is then rapidly expanded by sudden decompression. This is usually done by passing the SCF solution through an orifice at low pressure. RESS is an advantageous single-step process requiring minimal to no organic solvent. However, a typical issue with this process is particle agglomeration. Many poorly water-soluble compounds have inadequate solubility in SCFs under moderate conditions of temperature (<60 °C) and pressure (<300 bar). Cosolvents such as methanol can be added to CO2 to aid drug solubility. However, this necessitates an additional step for removing the residual solvent which increases process complexity and expense.

[0156] The second group consists of processes that use SCFs as antisolvents. In these processes, the drug-polymer system is solubilized in an organic solvent and then mixed with an SCF which acts as an antisolvent. The solubilization capability of the organic solvent reduces as SCF is being added to the organic solvent. Thus, the new liquid mixture now becomes supersaturated with respect to the drug (and the polymer) causing them to precipitate as ASDs. To successfully produce an ASD by this method, the drug and polymer should possess high solubility in the selected organic solvent and limited / no solubility in the SCF. Also, the selected organic solvent should be miscible with the selected SCF (antisolvent). The precipitated particles are subsequently filtered. There are different techniques that have the same underlying principle but differ in the mixing method of the solvent and antisolvent. The examples include precipitation with compressed fluid antisolvent (PCA), supercritical antisolvent (SAS) precipitation, aerosol solvent extraction system (ASES), gas antisolvent (GAS) precipitation, and solution-enhanced dispersion by SCF (SEDS). For the PCA method, the organic solvent solution is introduced via capillaries into the controlled chamber containing the SCF214, whereas in SAS the solvent solution is introduced via a nozzle. These methods have been applied to provide ASDs on an industrial-scale. Atorvastatin, megestrol acetate, and valsartan are ASDs produced by the SAS approach and have exhibited improved solubility and bioavailability. Indomethacin, cefdinir, and glycyrrhizic acid ASDs produced by this process have shown enhanced solubility whereby the formulated powder also exhibited uniform particle size. Glibenclamide ASDs prepared by the SAS were shown to have similar solubility as those prepared by solvent evaporation using a rotary evaporator. A recent study further demonstrated improved permeability of zidovudine- poly (1-lactic acid) solid dispersions relative to the pure crystalline drug when tested in an ex vivo everted rat intestinal sac model. The observed effect was attributed to the enhanced poly (1- Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION lactic acid) (polymer) plasticization which increased the extent of drug diffusion in the polymer matrix. In the ASES method, the organic solvent solution and SCF are sprayed at the same time (using different nozzles) into the chamber. Itraconazole-HPMC ASD particles (100-500 nm) produced by ASES showed >609-fold increase in the amount of drug released during dissolution compared to the pure crystalline drug.

[0157] As an alternative to the above methods, the SCF can be added to the organic solution in the GAS technique. A key unique feature of this method is that the SCF can be used as an antisolvent in the supercritical condition for the entire time of processing. The pressure applied varies continually from one Bar to the final pressure. With an increase in pressure, the concentration of the gas employed increases, causing the ASD to precipitate. Compared to other SCF methods, GAS is a slow process that can allow for molecular rearrangement and is therefore not ideal for ASDs. SEDS, another process of the second group uses a unique patented nozzle that allows SCF to function both as an antisolvent and a dispersing agent for the organic solution. Thus, the organic solution and SCF are atomized simultaneously228. There are two types of a nozzle in the SEDS: one with two channels (one each for SCF and organic solution) and another with three channels. The nozzle with three channels provides more choices in operating variables. For example, one channel can be for the organic solution with a drug, a second for polymer in aqueous solution, and a third for SCF. Puerarin microparticles produced by SEDS were amorphous whereas the ones prepared by GAS were crystalline, likely due to the slow process of GAS.

[0158] The third group includes methods that use SCFs as solutes, including precipitation from gas-saturated solutions (PGSS). It utilizes the ability of supercritical CO2 (scCCh) to diffuse into organic compounds like polymers. When scCCh diffuses into the polymer, it lowers polymer Tg and viscosity. In the PGSS process, the drug-polymer physical mixtures are first introduced into SCF. Elevated pressure and SCF cause the mixture to melt. This non-viscous solution is sprayed into a receiving chamber with controlled pressure. Due to rapid decompression, SCF escapes the solid matrix, and ASDs are formed. This method is particularly suitable for materials like PVP and PLGA which easily absorb SCF. PGSS is an advantageous SCF-based process since it does not require organic solvents and usually operates at low pressures with lower gas consumption relative to other processes such as RESS. Composite solid dispersions of fenofibrate and stearoyl macrogol-32 glycerides have been prepared using PGSS by Pestieau et al. Biphasic dissolution Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION of the formulation indicated a significantly higher concentration of the drug in the aqueous (0.1 mol / L HC1) as well as organic (octanol) layer relative to the physical mixture. In another study, progesterone solid dispersions were prepared with PGSS232. Amongst the tested variables, a yield of 94.7% and the highest extent of progesterone dissolution after 20 min (85.6%) was observed for lower progesterone-to-excipient ratio (1: 10) and process values of higher pressure (186 bar), higher temperature (60 °C), and a longer processing time (30 min).

[0159] SCFs have also been used as processing aids in combination with melt extrusion (usually in RESS or PGSS mode). Among SCF-based processes, SAS-based processes are generally favored for ASD preparation due to easier scale-up and more tunable variables. Besides the formulation composition (discussed in previous methods), the key variables that affect product attributes in the SCF-based methods are pre-expansion conditions (temperature and pressure of vessel), nozzle type (atomization, dimension) and angle of impact of the jet stream, feed rate of solution, flow rate of the SCF, and final drying / extraction time.

[0160] Spray-freeze-drying

[0161] Spray-freeze-drying is one of the cryogenic technology for preparing ASD powders. During the spray-freeze-drying process, the feed solution or suspension comprising of the drug, polymer, and possibly other excipients are atomized and sprayed directly into a cryogenic liquid. The frozen particles are then transferred to a freeze-dryer to generate a flowable ASD powder. This process can be further categorized depending on the type of injection devices (capillary, rotary, pneumatic, ultrasonic, two-fluid / three fluid nozzle), location of the nozzle, and the composition of the cryogenic liquid (liquid nitrogen, liquid argon, compressed CO2). However, generally, spray-freeze-dried powders are amorphous and porous, which results in a high surface area and a high dissolution rate. Also, relative to spray drying, spray-freeze-drying offers a higher control over particle size, compatibility with more excipients, less thermal stress, and a higher yield. The spray -freeze-drying process has been successfully used to enhance the dissolution and bioavailability of poorly soluble drugs like carbamazepine, and danazol. He et al. demonstrated that this technology could be employed for producing stable and free-flowing baicalein powers that could not be obtained by a conventional solvent evaporation method (i.e., rotavapor). Pluronic F68 was used as a carrier and inhibited baicalein from crystallizing. In comparison with the ASDs prepared by rotary evaporation, the spray-freeze-dried ASDs Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION exhibited significantly enhanced baicalein dissolution rate and improved oral bioavailability in rats, likely as a result of their porous structure and higher specific surface area.

[0162] Since spray-freeze-drying is a combination of spraying and freeze-drying processes, the manufacturing considerations for preparing stable ASD systems using this method are a combination of those two as well (discussed in their respective sections). Nevertheless, an important distinction is the solvents used for preparing the feed solution. Spray drying solvents may have a high ratio of organic components. However, for spray-freeze-drying, it is preferable that the feed solution have a low organic component since freeze-drying usually does not permit high organic solvent content. Another concern with spray -freeze-drying is that the porous and low-density structures of the ASDs may make them fragile and difficult for secondary processing.

[0163] Other solvent-based methods

[0164] Solvent casting is a commonly used laboratory process for preparing ASDs. A drug- polymer solution (mostly organic) is spread onto a substrate and then the solvent is evaporated at room temperature under normal pressure. The solvent evaporation can also be sped up using a hot plateor by placing it in an oven (at low temperatures and pressure) and followed by cooling. The films that form are usually milled to obtain powders. However, this method is limited to drugs that can be solubilized in solvents with low boiling points like ethanol, chloroform, dichloromethane, or their mixtures. Also, it might be challenging to completely evaporate the residual solvent.

[0165] Solvent evaporation using a rotary evaporator is another frequently used process for a small-scale ASD preparation. In this process, the drug-polymer solution, typically using an organic solvent, is evaporated under vacuum and at slightly elevated temperatures. Simultaneous application of vacuum and heat increases the rate of solvent evaporation and allows the use of a solvent with higher boiling points if required. In the rotary evaporator process, solvents like tetrahydrofuran or dimethylformamide that could not be used in a solvent casting process can be used. The final product is collected from the flask and, if necessary, can be further milled.

[0166] Sandhu et al. reported rotating jet-spinning as an alternative to electrospinning for ASD production with particles in the nano- or micro-meter size range. The typical set-up consists of a rotating reservoir containing drug-polymer solutions attached to a motor. The reservoir can either Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION be perforated, or equipped with a side nozzle, or possibly placed over metal plates with perforations for ejecting the solution. The drug-polymer solution is placed into a preheated or room temperature rotating metal container (spinning reservoir) which rotates at high speed (generally in the range of 2000-13,000 rpm). When the reservoir is rotated (about its axis) at a rate that overcomes the capillary and centrifugal forces, the solution jet is ejected from a side nozzle / perforations on the reservoir or a gap between metal plates holding the reservoir. This jet is propelled along a long spiral trajectory extended by the centrifugal forces. This generates a higher surface area. Solvent evaporation also occurs during this step. The solvent evaporation rate is dependent upon the solvent diffusion coefficient in the polymer261. Relative to solvent evaporation methods like spray drying or electrospinning, challenges posed by this method include its lower ability to remove the residual organic solvent(s) to a satisfactory level, and the necessity of a batch mode. Further study on the scale-up capability of this method for producing ASDs is warranted for the feasibility of industrial manufacturing.

[0167] Melting-based methods

[0168] In melt-based methods, the formulation components are heated to form dispersions, followed by cooling. Solvent avoidance is a significant benefit of the melting techniques. However, a major drawback of these methods is that the high temperatures may induce drug degradation84. Melting methods also require sufficient solubility / miscibility of drugs in the polymer melt, which can be very difficult for certain molecules to achieve.

[0169] An efficient ASD production by melting depends on the operational as well as compositional variables. The use of a polymer in the production of an ASD is of course to stabilize the amorphous form, but it is also important for processing because the polymer provides a molten medium for drug solubilization or dispersion. Therefore, the polymers used in melting processes are usually polymers or waxes with a low melting point or Tg. Commonly used carriers include PVP, PVPVA, cellulose esters and acrylates, and polymethacrylate derivatives. There are some commercial polymers specifically designed for melting processes such as HPMCAS, SOLUPLUS (amphiphilic polyvinyl caprolactam-poly vinyl acetatepolyethylene glycol graft copolymer from BASF), and AFFINISOL (HPMC HME is a water- soluble polymer) HPMC HME (modified HPMC). These carriers can also be used in combination to achieve improved amorphization, stability, dissolution, and bioavailability. Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION

[0170] Sufficient plasticization is essential to form an ASD during the melting process. Although the drug itself can provide some plasticizing effect, additional plasticizers can be added to aid the mixing of drug and polymer. Melting-based ASD preparation processes are also often characterized by high shear stress, which can be reduced by the incorporation of a plasticizer. Plasticizers reduce the viscosity of the mixture and therefore can lower the processing temperature. Some examples of frequently used plasticizers are d-a-tocopheryl, poloxamers, low molecular-weight poly (ethylene glycol)s (PEGs), and surfactants. Plasticizer selection is dependent on its intended formulation functionality, such as lowering the processing temperature or lowering the melt viscosity. Added plasticizers that remain in the product, can affect its properties and performance (physicochemical stability, dissolution, Tg, hygroscopicity, or appearance). In addition, conventional plasticizers are used in a concentration range of 5%-30%, w / w which increases the total weight of the formulation and may result in large, unacceptable dosage forms. Therefore, an ideal plasticizer should be capable of providing the desired plasticizing effect and then be removed from the formulation to mitigate its possible negative effects before final processing. For this reason, scCCh, low boiling solvents, or reagents that can evaporate or sublime are being used.

[0171] Hot-melt extrusion (HME)

[0172] For commercial-scale production, only two types of melting processes are available, hot- melt extrusion (HME) and melt agglomeration (discussed later in the ‘Granulation’ subsection). HME, and especially the twin-screw melt extrusion, is one of the most widely used techniques for producing ASDs. The drug and the polymer are mixed, melted, dispersed, and then extruded under specific processing conditions. The HME process, which can be divided theoretically into five steps, namely: feeding, melting and plasticizing, mixing / kneading and conveying, discharging, and cooling for further downstream processing. Important parts of a hot-melt extruder include a feed hopper, barrel, extrusion screws, torque sensors, heating-cooling system and dies.

[0173] The HME equipment has a flexible design, which enables processes to be tailored to achieve the desired outcomes and to accommodate varying raw materials, by adjusting the modular design elements, namely the screws and barrels. It is also possible to apply HME technology to drugs liable to oxidation and hydrolysis by excluding oxygen and moisture from Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION the mixture. The barrels can be flanged together or linked via internal tie rods. The most critical component of the extruder is the screw, which determines the quality and quantity (per unit time) of the extruded material. Based on the screw configuration, the process may be categorized into single-screw or twin-screw extrusion. A twin-screw provides many benefits over the single screw model, with decreased drug residence time in the extruder and facilitated continuous mass flow with better mixing. It can also be designed with two separate hoppers, both of which can vary in temperature over a wide range and are equipped with the self-cleaning function. In addition, twin-screw extruders can limit thermal stress on the materials by reducing the ‘nonmotion’ and preventing localized overheating. It also presents easier feeding of materials and less propensity to overheat288.

[0174] Depending on the desired shear level and operating speed, two screws can be designed in different configurations of co-rotating (rotating in the same direction) or counter-rotating (rotating in the opposite direction), depending on the desired mixing intensity. Co-rotating screws are usually used in pharmaceutical manufacturing because they generate relatively lower shear forces than the counter-rotating. Nakamichi et al. demonstrated that the physicochemical properties of the extruded material were substantially impacted by the machine operating conditions. The study reported that since the kneading screws kept the material in the machine for a prolonged period under shear, stable ASDs exhibiting super-saturation upon dissolution could be prepared irrespective of changes in operating conditions such as the rate of revolution of the screws (20-100 rpm) and the amount of water (0%-50%) introduced in the feed. When the kneading screws were detached from the screws and only the feed screw elements were used for ASD preparation, the extent of amorphization and dissolution profiles of the extruded material was substantially impacted by the machine operating conditions. Although partial crystallinity was observed in all the batches, the extent of crystallinity decreased with a reduction in rotation rate and amount of water. Slow screw rotation and the addition of a sufficient amount of water to the mixture increased the rate of dissolution of the drug, although no super-saturation occurred in any of the batches tested.

[0175] The kneading paddles play a vital role in drug amorphization. Verhoeven et al. reported that at least one kneading zone was necessary for the homogeneous distribution of metoprolol tartrate in ethyl cellulose matrices, even though the homogeneity of distribution and drug release rate were not significantly altered by the number of kneading zones or their location along the Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION extruder barrel. The impact of screw configuration on the dissolution behavior was studied by Liu et al. They observed that indomethacin dissolution into the polymer melt was accelerated by the kneading blocks. Adequate shear stress, shear rate and mean residence time (which is linked to lower screw speed and feeding rate), are required to ensure complete drug amorphization. However, excessively high shear stress, shear rate, and residence times pose the risk of material degradation.

[0176] The molten mass can be conveyed to the dies of a variety of shapes and sizes depending on the desired product properties. For example, the production of films and patches use flat dies, whereas pelletization and spheronization use circular dies. Injection molding can also be utilized to fill the molten drug-polymer mass into molds. Tablets, granules, capsules, or ear inserts can be produced in these molds. After cooling at room temperature, the resulting product is collected and can be further milled to obtain the required particle sizes. Apart from these important variables, other parameters such as temperature, die geometry, barrel design, and screw speed can also impact the final product properties. Recent work has shown that the degassing process can remove volatile substances and subsequent air bubbles, which facilitates consistent production of extrudates with improved cross-sectional uniformity. In order to reduce torque during screw rotation, certain minimum temperatures capable of plasticizing the material are required in HME. Material flow properties are also critical to ensure a consistent feed rate from the hopper. For poorly flowing material, flow aids such as spray-dried mannitol, maltodextrin, and colloidal silica can be used. As some of these flow aids are crystalline, the miscibility of the drug in the polymer may be affected, further complicating the solid-state and chemical analysis.

[0177] HME is continuous, single-step, solvent-free, and capable of scaling up. The HME method can be equipped with on-line and in-line quality-control analysis such as near-infrared, Raman, and dielectric spectroscopies, which facilitate quality by design and continuous manufacturing. Some drawbacks include the higher energy usage and exclusion of thermolabile compounds. Changes in the design of the equipment (e.g., presence of kneading elements), as well as the addition of plasticizers, can lead to a reduction in processing temperatures and / or residence time, and thus minimize the potential of thermal degradation of drugs during the processing.

[0178] Incorporation by Reference Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION

[0179] References and citations to other documents, such as patents, patent applications, patent publications, journals, books, papers, web contents, have been made throughout this disclosure, including to the Supplementary. The Supplementary, and all other such documents are hereby incorporated herein by reference in their entirety for all purposes.

[0180] Equivalents

[0181] The invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The foregoing embodiments are therefore to be considered in all respects illustrative rather than limiting on the invention described herein.

[0182] EXAMPLES

[0183] The data herein show that it hs been found that release of PEs and peptides from solid oral dosage forms is complex. Release is likely to be highly pH dependent for PE salts of weak acids. Addition of a gelling polymer can be helpful to synchronize release of PE and peptide, as well as modify the release rate.

[0184] Example 1 : Release Rate of Peptides From Formulations Containing Permeation Enhancers

[0185] Two common transient permeation enhancers are shown in FIG. 11. Sodium decanoate phase behavior is shown in FIG. 12. Improvement of peptide permeability by permeation enhancers is shown in FIG. 13.

[0186] From the work herein, it appears that there are certain formulation / physiological factors limiting PE Effectiveness, which include:

[0187] • PE dissolution rate may be too slow to reach threshold concentration (ideally ~ 25mM);

[0188] • Slow PE dissolution rate might lead to greater dilution by GI fluids;

[0189] • Peptide and PE dissolution rates may be different, impacting co-localization - PE and peptide both need to be present at epithelium; and

[0190] • Rapid PE absorption will decrease its concentration.

[0191] It appears that PEs may be rapidly absorbed (see FIG. 14), and therefore the relative release rates of PE’s and peptides is important. FIG. 15 illustrates synchronous versus asynchronous release. Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745

[0192] PATENT APPLICATION

[0193] FIG. 16 shows certain model peptides: insulin, octreotide, and cyclosporine. FIG. 17 shows an exemplary composition for insulin minitablets. FIG. 18 shows release of insulin in different media. FIG. 19 shows insulin solubility as a function of pH. FIG. 20 shows impact of pes on release of insulin: biorelevant media. FIG. 21 shows sodium decanoate solubility. FIG. 22 shows local pH during dissolution. FIG. 23 shows release profdes: SNAC. FIG. 24 shows release profiles: CIO versus insulin.

[0194] Example 3 : Octreotide Intrinsic Dissolution Rate and Surface Area Normalized Release from Formulations

[0195] FIG. 25 shows octreotide intrinsic dissolution rate and surface area normalized release from formulations. FIG. 26 shows copovidone (PVPVA) used as the gel-forming polymer. FIG. 27 shows intrinsic dissolution rates of octreotide and SNAC. FIG. 28 shows comparison of intrinsic dissolution rates of octreotide, SNAC and PVPVA. FIG. 29 shows binary and ternary dispersions containing SNAC and PVPVPA. FIG. 30 shows rate comparison between systems. FIG. 31 shows ternary dispersions containing octreotide, sodium decanoate (SD) and PVPVPA.

[0196] FIG. 32 shows details regarding the lipophilic peptide, cyclosporine. FIG. 33 is a table providing evidence that sodium decanoate can improve permeability of cyclosporine. FIG. 34 shows improved release by inclusion of pes in dispersions. FIG. 35 shows comparison of component release rates for SNAC. FIG. 36 shows comparison of component release rates for sodium decanoate (SD).

Claims

Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATIONWhat is claimed is:

1. An oral peptide pharmaceutical formulation comprising: a peptide as an active agent; a permeation enhancer; and a gel forming polymer.

2. The oral peptide pharmaceutical formulation of claim 1, wherein the permeation enhancer is at least one selected from the group consisting of sodium decanoate, sodium octanoate, salcaprozate sodium, Labrasol, and a combination thereof.

3. The oral peptide pharmaceutical formulation of claim 2, wherein the the gel forming polymer is at least one selected from the group consisting of copovidone, polyvinylpyrrolidone, hydroxypropylmethylcellulose, polyvinylalcohol, polyacrylic acid, and a combination thereof.

4. The oral peptide pharmaceutical formulation of claim 3, wherein the peptide is a hydrophobic peptide.

5. The oral peptide pharmaceutical formulation of claim 3, wherein the peptide is a hydrophilic peptide.

6. The oral peptide pharmaceutical formulation of claim 3, wherein the peptide is cyclosporine or octreotide.

7. The oral peptide pharmaceutical formulation of claim 3, wherein the oral pharmaceutical formulation is a solid dispersion.

8. The oral peptide pharmaceutical formulation of claim 7, wherein the peptide is 60% or less drug loading of the oral pharmaceutical formulation.Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION9. The oral peptide pharmaceutical formulation of claim 8, wherein the peptide is between 1-2% drug loading of the oral pharmaceutical formulatio.

10. The oral peptide pharmaceutical formulation of claim 7, wherein a ratio of the permeation enhancer; and the gel forming polymer is 50:50 w / w.

11. A method for manufacturing oral peptide pharmaceutical formulation, the method comprising: combining a peptide as an active agent, a permeation enhancer, and a gel forming polymer to form a mixture; and formulating the mixture into an oral peptide pharmaceutical formulation.

12. The method of claim 11, wherein the permeation enhancer is at least one selected from the group consisting of sodium decanoate, sodium octanoate, salcaprozate sodium, Labrasol, and a combination thereof.

13. The method of claim 12, wherein the the gel forming polymer is at least one selected from the group consisting of copovidone, polyvinylpyrrolidone, hydroxypropylmethylcellulose, polyvinylalcohol, polyacrylic acid, and a combination thereof.

14. The method of claim 13, wherein the peptide is a hydrophobic peptide.

15. The method of claim 3, wherein the peptide is a hydrophilic peptide.

16. The method of claim 13, wherein the peptide is cyclosporine or octreotide.

17. The method of claim 13, wherein the oral pharmaceutical formulation is a solid dispersion.

18. The method of claim 17, wherein the peptide is 60% or less drug loading of the oral pharmaceutical formulation.Attorney Docket No.: PURD- 159 / 01 WO 28593 / 745 PATENT APPLICATION19. The method of claim 18, wherein the peptide is between 1-2% drug loading of the oral pharmaceutical formulatio.

20. The method of claim 17, wherein a ratio of the permeation enhancer; and the gel forming polymer is 50:50 w / w.

Images