METHOD OF TESTING CRYSTALLINITY IN AMORPHOUS PHARMACEUTICAL COMPOSITIONS

Abstract

Methods of testing pharmaceutical compositions for the presence or absence of active pharmaceutical ingredient (API) crystallinity in an amorphous solid dispersion or solid-state solution using UV/vis spectrometry is provided. Testing may be performed standalone or during manufacturing of a pharmaceutical composition. A predictive model provides for quantitative analysis of the amount of crystalline API based on UV/vis data of corresponding reference samples. Also provided is an apparatus for manufacturing a pharmaceutical composition.

Claims

1. A method of generating a predictive model for determining the amount of crystallinity of an active pharmaceutical ingredient (API) in an amorphous solid dispersion or solid-state solution comprising the steps of; (i) subjecting a plurality of reference samples of dispersions or solutions spanning a range of API crystallinity to UV/vis spectroscopy, (ii) measuring a reflectance and/or transmission spectrum of each reference sample, and (iii) processing the spectra gathered in step (ii) to generate a predictive crystallinity model.

2. A method according to claim 1 wherein the number of reference samples is 5 or more.

3. A method according to any one of the preceding claims, wherein the spectra are pre-processed before step (iii) to normalise and/or smooth the spectra

4. A method according to any one of the preceding claims wherein the spectra of the reference samples are processed to derive a feature that correlates with crystallinity across at least a portion of said range of crystallinity.

5. A method according to claim 4, wherein said feature comprises: at least a first principle component derived from principle components analysis (PCA) of the spectra, optionally wherein the variance of said first principle component by crystallinity is substantially linear across the range of crystallinity of said plurality of dispersions or solutions; or a lightness value L* of CIELAB colour space derived from the spectra.

6. A method of testing a pharmaceutical composition comprising an API in an amorphous solid dispersion or solid-state solution for crystallinity of the API comprising the steps of; (i) subjecting the dispersion or solution to UV/vis spectroscopy, (ii) measuring a reflectance and/or transmittance spectrum, and (iii) determining the presence or absence of crystallinity of the API by comparing measured reflectance and/or transmittance spectrum to that expected for a completely amorphous sample.

7. A method according to claim 6, further comprising the step of (iv) when crystalline API is found to be present, determining the amount crystallinity in the dispersion or solution by comparing the observed spectrum to a predictive model.

8. A method according to claim 7, wherein the predictive model is generated according to any one of claims 1 to 5.

9. A method according to claim 7 or claim 8, wherein comparing the observed spectrum to the predictive model comprises processing the observed spectrum in the same way as the spectra of said plurality of reference samples according to any one of claims 1 to 5.

10. A method according to any one of claims 6 to 9, wherein the amount of crystallinity in the dispersion or solution is measured at 50 wt % and below.

11. A method according to any one of claims 6 to 10, wherein the amount of crystallinity in the dispersion or solution is measured at 1 wt % and above.

12. A method of manufacturing a pharmaceutical composition comprising the steps of; (i) forming an API into an amorphous solid dispersion or solid state solution, (ii) testing the dispersion or solution for crystallinity of the API one or more times by the method of any one of claims 6 to 10, and (iii) where the dispersion or solution has an amount of crystallinity within an acceptable range, processing the composition into a finished pharmaceutical product.

13. A method according to claim 12 wherein: the forming of the API into an dispersion or solution is performed by extrusion, ball-milling or spray drying; and/or the testing of the dispersion or solution for the presence or absence of crystallinity one or more times is performed in-line.

14. A method according to claim 13 wherein the forming is performed by extrusion and the testing is performed at one or more of the point of API input, upstream of the point of extrusion and at the point of extrusion.

15. A method according to claim 13 wherein the forming is performed by ball-milling in a ball mill having one or more transparent points wherein the testing is performed through the one or more transparent points substantially perpendicular to the axis of motion.

16. A method according to claim 13 wherein the forming is performed by spray drying and the testing is performed at one or more of the point of API input, at the point of atomisation in a drying chamber or at the point of settling after atomisation.

17. A method according to any one of claims 12 to 16 wherein the acceptable range for crystallinity is 1 wt % or less.

18. A method according to any one of claims 13 to 17 wherein the amount of API input to the screw extruder, ball mill or spray drier is automatically adjusted, when required, to ensure crystallinity is within the acceptable range.

19. A method according to any one of claims 13 to 18 wherein the speed of the screw extruder, ball-mill or spray drier is automatically adjusted, when required, to ensure crystallinity is within the acceptable range.

20. A method according to any one of claims 13 to 19, wherein the temperature of the extruder barrel, ball-miller or spray drier is automatically adjusted, when required, to ensure crystallinity is within the acceptable range.

21. A method according to any one of claims 12 to 20 wherein UV/vis reflectance and/or transmittance are measured continuously.

22. A method according to any one of the preceding claims, wherein the reflectance and/or transmittance spectrum is measured at wavelengths of 210 to 800 nm.

23. A method according to any one of the preceding claims wherein only the reflectance spectrum is measured.

24. A method according to any one of claims 13 to 22 wherein only the transmission spectrum is measured.

25. A method according to any one of the preceding claims, wherein the particle size of the dispersion or solution is measured prior to being subjected to UV/vis spectroscopy.

26. A method according to claim 25, wherein the dispersion or solution is sized to match the size of one or more reference samples of known crystallinity, wherein the reference samples match the API and carrier of the dispersion or solution.

27. A method according to any one of the preceding claims, wherein the API is a compound of 2000 g/mol or less molecular weight.

28. A method according to any one of the preceding claims, wherein the API is at least 95% pure prior to being incorporated into said dispersion or solution.

29. A method according to any one of the preceding claims, wherein the API comprises 5 wt % or less water.

30. A method according to any one of the preceding claims, wherein the API is a non-nucleoside reverse transcriptase inhibitor (NNRTI), preferably etravirine (ETR), or is a non-steroidal anti-inflammatory drug (NSAID), preferably piroxicam (PRX).

31. A method according to any one of the preceding claims, wherein the carrier of the dispersion or solution comprises an amorphous polymer.

32. A method according to claim 31, wherein the amorphous polymer is one or more of a cellulose polymer, a vinyl polymer, a polymethacrylate polymer and a polyalkylene glycol polymer.

33. A method according to claim 32 wherein the amorphous polymer is one or more of ethyl cellulose, hydroxyethyl cellulose, hydroxypropyl cellulose, hydroxypropyl methylcellulose (HPMC), HPMC acetate succinate (HPMCAS); polyvinyl acetate phthalate; polymethacrylates; polyalkylene glycols such as polyethylene glycol (PEG), copolymers of PEG, polypropylene glycol (PPG), poloxamers (triblock polymers with a central hydrophobic PPG block flanked by two hydrophilic PEG blocks), soluplus (a PEG-polyvinyl acetate-polyvinyl caprolactam graft copolymer (PVAc-PVCap-PEG)); polyvinylpyrrolidone (povidone), copovidone, vinylpyrrolidone-vinyl acetate copolymers and a linear random copolymer (e.g. 60:40) of N-vinyl-2-pyrrolidone and vinyl acetate.

34. A method according to any one of the preceding claims, wherein the carrier of the dispersion or solution comprises 5 wt % or less water.

35. An apparatus for manufacturing a pharmaceutical composition as an amorphous solid dispersion or solid-state solution comprising a quality control system for testing the amount of crystallinity of an API in the composition by means of one or more UV/vis spectrometers configured to measure reflectance and/or transmittance intensity.

36. An apparatus according to claim 35 comprising a screw extruder, a ball mill or a spray drier, wherein the or each UV/vis spectrometer is positioned at an in-line measurement point to measure reflectance and/or transmittance of the dispersion or solution.

37. An apparatus according to claim 36 comprising a screw extruder wherein the or at least one of the UV/vis spectrometers is configured to measure reflectance and/or intensity at one or more of the point of API input, upstream of the point of extrusion and the point of extrusion.

38. An apparatus according to claim 36 comprising a ball mill having one or more transparent points wherein the or at least one of the UV/vis spectrometers is configured to measure reflectance and/or transmittance through the one or more transparent points and substantially perpendicular to the axis of motion.

39. An apparatus according to claim 36 comprising a spray drier wherein the or at least one of the UV/vis spectrometers is configured to measure reflectance and/or transmittance at one or more of the point of API input, the point of atomisation in a drying chamber and the point of settling after atomisation.

40. An apparatus according to any one of claims 35 to 39 wherein the or each UV/vis spectrometer is photonically connected to the or each measurement point by optical fibre.

41. An apparatus according to any one of claims 36 to 40 wherein the apparatus is configured to automatically adjust the amount of API input to the screw extruder, ball mill or spray drier, when required, in response to the in-line measurements of the or each UV/vis spectrometer.

42. An apparatus according to any one of claims 36 to 41 wherein the apparatus is configured to automatically adjust the speed of the screw extruder, ball mill or spray drier, when required, in response to the in-line measurements of the or each UV/vis spectrometer.

43. An apparatus according to any one of claims 36 to 42 wherein the apparatus is configured to automatically adjust the temperature of the screw extruder, ball mill or spray drier, when required, in response to the in-line measurements of the or each UV/vis spectrometer.

44. An apparatus according to any one of claims 35 to 43, wherein the or each UV/vis spectrometer is configured to measure reflectance and/or transmittance intensity over the region of 210 to 800 nm.

45. An apparatus according to any one of claims 35 to 44 wherein the or each UV/vis spectrometer is configured to measure reflectance and/or transmittance continuously.

46. An apparatus according to any one of claims 35 to 45 for manufacturing a pharmaceutical composition as an amorphous solid dispersion or solid-state solution according to any one of claims 12 to 32.

Description

SUMMARY OF THE FIGURES

[0071] So that the invention may be understood, and so that further aspects and features thereof may be appreciated, embodiments illustrating the principles of the invention will now be discussed in further detail with reference to the accompanying figures, in which:

[0072] FIG. 1 shows five UV/vis transmission spectra of 15, 18, 20, 23 and 25 wt % PRX samples at 230 to 810 nm. An overall increase in absorbance is seen with increasing amounts of PRX with a particularly large increase in overall absorbance is observed between 23% and 25% PRX.

[0073] FIG. 2 shows five XRD spectra obtained on a Bruker instrument of pure solid PRX, Kollidon VA64, and 20, 25 and 30 wt % PRX mixtures. Traces of the XRD fingerprint of crystalline PRX are first observed in the 30% PRX spectra.

[0074] FIG. 3 shows overlaid UV/vis spectra for polymeric mixtures of HMPC-E5 (VIVAPHARM® hydroxypropyl methylcellulose) and sodium stearoyl fumarate (SSF) having 20, 10, 5, 2.5, 1.25 and 0.625 wt % crystalline etravirine.

[0075] FIG. 4 shows overlaid UV/vis reflectance spectra for ETR dispersed in hydroxypropyl methyl cellulose (HMPC-E5) wherein the total amount of ETR is 20 or 25 wt %.

[0076] FIG. 5 shows a good correlation between L* value and crystallinity of thirteen unmilled ETR samples that each have different crystallinities as determined beforehand by Raman spectroscopy obtained on a Kaiser instrument. The total amount of ETR in each sample, including both amorphous and crystalline content, was 25 wt %. An unmilled sample is the direct solid monolithic form that results upon solidification after extrusion.

[0077] FIG. 6 shows the reflectance UV/vis transmission spectra at 220 to 800 nm for the same thirteen samples of ETR as FIG. 5.

[0078] FIG. 7 shows the XRD spectra obtained on a Rigaku instrument for the same thirteen samples of ETR as FIG. 5.

[0079] FIG. 8 shows the UV reflectance spectra for five of the thirteen ETR samples of FIG. 5 that have substantially no crystallinity (<1% as determined by Raman spectroscopy). Good reliability and reproducibility of the overall spectrum shape and reflectance amount between these very similar samples is observed.

[0080] FIG. 9 shows a plot of loading contributing to Factor-1 (y-axis), the first principle component following PCA on the reflectance spectra ETR-containing samples as described in Method 3, against wavelength in nm from 290 to 400 (x-axis). As shown, Factor-1 provides the first eigenvector that explains the variance in the dataset.

[0081] FIG. 10 shows a plot of variance (y-axis) against Factors in the PCA (x-axis), showing that Factor-1 and Factor-2 explain around 98% and 2% of the variance of the dataset, respectively.

[0082] FIG. 11 shows a scatter plot of Factor-2 (y-axis) against Factor-1 (x-axis) derived from the PCA analysis in method 3. As can be seen the samples cluster according to crystallinity (i.e. 2.5% crystallinity samples (squares), 5% crystallinity samples (circles), 10% crystallinity samples (triangles) and 20% crystallinity samples (diamonds)).

[0083] FIG. 12 shows a plot of the predicted crystallinity by Factor-1 (y-axis) against the known (reference) levels of crystallinity of the samples (x-axis). The inset shows the linearity of the correlation.

DETAILED DESCRIPTION

[0084] The features disclosed in the foregoing description, or in the following claims, or in the accompanying drawings, expressed in their specific forms or in terms of a means for performing the disclosed function, or a method or process for obtaining the disclosed results, as appropriate, may, separately, or in any combination of such features, be utilised for realising the invention in diverse forms thereof.

[0085] While the invention has been described in conjunction with the exemplary embodiments described above, many equivalent modifications and variations will be apparent to those skilled in the art when given this disclosure. Accordingly, the exemplary embodiments of the invention set forth above are considered to be illustrative and not limiting. Various changes to the described embodiments may be made without departing from the scope of the invention.

[0086] For the avoidance of any doubt, any theoretical explanations provided herein are provided for the purposes of improving the understanding of a reader. The inventors do not wish to be bound by any of these theoretical explanations.

[0087] Any section headings used herein are for organizational purposes only and are not to be construed as limiting the subject matter described.

[0088] Throughout this specification, including the claims which follow, unless the context requires otherwise, the words “have”, “comprise”, and “include”, and variations such as “having”, “comprises”, “comprising”, and “including” will be understood to imply the inclusion of a stated integer or step or group of integers or steps but not the exclusion of any other integer or step or group of integers or steps.

[0089] It must be noted that, as used in the specification and the appended claims, the singular forms “a,” “an,” and “the” include plural referents unless the context clearly dictates otherwise. Ranges may be expressed herein as from “about” one particular value, and/or to “about” another particular value. When such a range is expressed, another embodiment includes from the one particular value and/or to the other particular value. Similarly, when values are expressed as approximations, by the use of the antecedent “about,” it will be understood that the particular value forms another embodiment. The term “about” in relation to a numerical value is optional and means, for example, +/−10%.

[0090] The words “preferred” and “preferably” are used herein refer to embodiments of the invention that may provide certain benefits under some circumstances. It is to be appreciated, however, that other embodiments may also be preferred under the same or different circumstances. The recitation of one or more preferred embodiments therefore does not mean or imply that other embodiments are not useful, and is not intended to exclude other embodiments from the scope of the disclosure, or from the scope of the claims.

Examples

[0091] Method 1—In-Line Crystallinity Detection of Piroxicam in Kollidon VA64

[0092] A twin-screw extruder was used to assess the solubility maximum for a mixture of the non-steroidal anti-inflammatory drug (NSAID) piroxicam (PRX) and the carrier polymer Kollidon VA64. The mixture was continuously analysed while traversing the extruder screw using a system of in-line UV/vis spectrometers and the input of the PRX was adjusted until crystallinity was first detected.

##STR00001##

Formula 1. 4-hydroxy-2-methyl-1,1-dioxo-N-pyridin-2-yl-1$1{circumflex over ( )}{6},2-benzothiazine-3-carboxamide (PRX)

[0093] The samples were analysed by a UV/vis spectrometer in transmission mode (FIG. 1) and an XRD spectrum of the same samples was also obtained for comparison and verification (FIG. 2). The XRD samples required milling and analysis offline whereas UV/vis has the advantages that it can be performed rapidly in-line without milling.

[0094] UV/vis transmission spectra were obtained for 15, 18, 20, 23 and 25 wt % PRX samples. There is a clear overall absorbance increase between the 23 and 25 wt % PRX samples. This correlates with the observation that mixtures appear amorphous, homogeneous and transparent below 25 wt % PRX loading. However, at 25 wt % PRX, the first microcrystalline domains are observed, leading to increased scattering and a decrease in transmitted light (i.e. an increase in absorbance).

[0095] XRD spectra were obtained on a Bruker instrument for each of pure solid PRX, Kollidon VA64, and 20, 25 and 30 wt % PRX mixtures separately. By comparison, XRD does not detect the presence of crystallinity in the 25 wt % PRX mixture. Only at 30 wt % PRX loading, well above the maximum solubility when there is substantially higher crystallinity, does the XRD spectra first show peaks that correlate with the pure XRD spectra of PRX. The Kollidon VA64 spectra shows that the polymer component of the mixture is amorphous.

[0096] This experiment demonstrates that UV/vis can have a greater sensitivity to low amounts of crystallinity compared to XRD.

Method 2—UV/Vis Calibration Using a Milled Crystalline ETR Dilution Series

[0097] Etravirine (0.82271 g), sodium stearoyl fumarate (0.03291 g) and HMPC-E5 (2.444 g) were added to a small sealable plastic bag to provide a 25 wt % ETR mixture. The ETR was 100% crystalline. The HMPC-E5 was cryomilled for 20 minutes and dried before use.

##STR00002##

Formula 2. 4-[6-amino-5-bromo-2-(4-cyanoanilino)pyrimidin-4-yl]oxy-3,5-dimethylbenzonitrile (ETR)

[0098] The particle size distributions of the components were measured using a Mastersizer and are shown in the table below.

TABLE-US-00001 TABLE 1 Particle size distributions of the components. Component D10 D50 D90 crystalline etravirine (ETR) 8 μm 26 μm 87 μm HPMC-E5 (cryomilled for 20 3 μm 17 μm 60 μm minutes)

[0099] The sealed bag was alternatively kneaded and shaken for 2 minutes in total. The resulting powder was used in a serial dilution series to provide mixtures having 20, 10, 5, 2.5, 1.25, 0.625 and 0.3125 wt % crystalline ETR. The diluent was a completely amorphous solid dispersion having the same ratios of ETF, SSF and HPMC-E5 to maintain the ETR loading at 25%.

[0100] Multiple UV/vis reflectance spectra were obtained for each sample of the serial dilution. A spectra for mixtures having 20, 10, 5, 2.5, 1.25 and 0.625 wt % crystalline ETR is shown in FIG. 3.

[0101] The overall reflectance increases with crystallinity at all measured wavelengths. There is a clear correlation between the amount of crystalline ETR (20 to 0.6 wt %) and the reflectance observed in each UV/vis spectrum. The amount of crystalline ETR was independently verified for the same samples using off-line XRD and Raman spectroscopy.

[0102] Furthermore, FIG. 4 shows overlaid UV/vis spectra of another series of ETR preparations that differ in crystallinity. The same clear correlation between the amount of crystalline ETR (0 to 9.8 wt %) and the reflectance observed in each UV/vis spectrum.

Method 3—Predictive Modelling of Crystallinity

[0103] To prepare a predictive model to determine crystallinity, each reflectance spectra for each amount of crystalline ETR in Method 2 was first converted to its second derivative and then processed by standard normal variate (SNV) analysis, and Savitzky Golay analysis was performed to generate a moving average across all data points in each spectrum. Following these data processing steps, which normalise and smooth the spectral data, the data were subjected to principle component analysis (PCA) to provide a first principle component (Factor-1) that was found to explain around 98% of the variance of the spectra (see FIGS. 9 and 10). The plot of Factor-1 (y-axis) against wavelength in the range 290-400 nm (x-axis) is shown in FIG. 9 and represents a simplified “spectrum” that best fits the whole dataset of the samples across the examined range of crystallinity.

[0104] Factor-2 largely explains the remaining 2% variation in the dataset (FIG. 10). A scatter plot of Factor-2 (y-axis) against Factor-1 (x-axis) is shown in FIG. 11. The samples were found to cluster according to crystallinity (i.e. 2.5% crystallinity samples (squares), 5% crystallinity samples (circles), 10% crystallinity samples (triangles) and 20% crystallinity samples (diamonds)). This shows that processing the UV/vis spectra data to form a PCA model successfully resolves the spectra, distinguishing the samples by crystallinity in the range 2.5-20%.

[0105] In FIG. 12, the predicted crystallinity by Factor-1 (y-axis) is plotted against the known (reference) levels of crystallinity of the samples (x-axis). A high degree of correlation is seen (R-squared values above 0.99). This example shows that processing the UV/vis spectra, optionally after data normalisation and/or smoothing, to generate a PCA model facilitates determination of crystallinity of one or more samples via the methods of the present invention.

[0106] Importantly, without wishing to be bound by any particular theory, the present inventors believe that many other API/carrier combinations will be amenable to such PCA modelling because PCA is an adaptable analysis that can be used to reduce the complexity or dimensionality of the spectral data without the need to understand what causes a particular spectral shape or pattern.

[0107] In some cases, the data may be further processed by partial least squares regression (PLSR) to conveniently compare and predict the crystallinity of sample(s) based on their UV/vis spectra, particularly with reference to a “standard curve” formed of values obtained from a plurality of reference samples of known crystallinity and which have the same API and carrier as the subject sample(s).

[0108] Because total reflectance and total absorbance of a sample vary with crystallinity, another way to model the UV/vis spectral data is to generate and plot the L* values of reference samples having a known and differing crystallinity.

[0109] L* is the lightness value component of L*a*b* according to ‘CIELAB colour space’ as defined by the International Commission on Illumination (CIE) where the darkest black is L*=0 and the brightest white is L*=100. Such a plot of L* value against crystallinity is shown in FIG. 5. A clear and linear correlation between crystallinity and L* value, particularly above 1% crystallinity, is observed.