Evaluation of the Structural Determinants of Polymeric Precipitation

Subscriber access provided by Libraries of the | University of Hawaii at Manoa

Article

Evaluation of the structural determinants of polymeric precipitation inhibitors using solvent shift methods and principle component analysis Dallas B. Warren, Christel A.S. Bergström, Hassan Benameur, Christopher J.H. Porter, and Colin W. Pouton Mol. Pharmaceutics, Just Accepted Manuscript • DOI: 10.1021/mp300576u • Publication Date (Web): 30 Apr 2013 Downloaded from http://pubs.acs.org on May 4, 2013

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Molecular Pharmaceutics is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 68

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

EVALUATION OF THE STRUCTURAL DETERMINANTS OF POLYMERIC PRECIPITATION INHIBITORS USING SOLVENT SHIFT METHODS AND PRINCIPLE COMPONENT ANALYSIS Dallas B. Warrena, Christel A.S. Bergströmb,c, Hassan Benameurd, Christopher J.H. Porterc* and Colin W. Poutona* a

Drug Discovery Biology, Monash Institute of Pharmaceutical Sciences, Monash University, Parkville, Victoria, Australia

b

Uppsala University Drug Optimization and Pharmaceutical Profiling Platform, Department of Pharmacy, Uppsala University, Uppsala, Sweden c

Drug Delivery, Disposition and Dynamics, Monash Institute of Pharmaceutical Sciences, Monash University, Parkville, Victoria, Australia and

d

Pharmaceutical Sciences Capsugel R&D, Illkirch Graffentstaden, France

* Corresponding authors: C.J.H. Porter Email: [email protected] Tel: +61 3 9903 9649, C.W. Pouton Email: [email protected] Tel: +61 3 9903 9562

1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract The presence of polymers within solid dose forms, such as solid dispersions, or liquid or semisolid formulations, such as lipid based formulations, can promote the maintenance of drug supersaturation after dissolution or dispersion/digestion of the vehicle in the gastrointestinal tract. Transiently stable supersaturation delays precipitation, increases thermodynamic activity and may enhance bioavailability and reduce variability in exposure. In the current study a diverse range of 42 different classes of polymers, with a total of 78 polymers across all classes, grades and molecular weights were examined, to varying degrees, as potential polymeric precipitation inhibitors (PPIs) using a solvent shift method to initiate supersaturation. To provide a deeper understanding of the molecular determinants of polymer utility the data was also analysed, along with a range of physicochemical descriptors of the polymers employed, using principle component analysis (PCA). Polymers were selectively tested for their ability to stabilise supersaturation for 9 poorly water soluble model drugs, representing a range of nonelectrolytes, weak acids and weak bases. In general, the cellulose based polymers (and in particular hydroxypropylmethyl cellulose, HPMC, and its derivatives) provided robust precipitation inhibition across most of the drugs tested. Subsequent PCA indicate that there is consistent PPI behaviour of a given polymer for a given drug type, with clear clustering of the performance of polymers with each of the nonelectrolytes, weak bases and weak acids. However, there are some exceptions to this, with some specific drug type – polymer interactions also occurring. Polymers containing primary amine functional groups should be avoided as they are prone to enhancing precipitation rates. An inverse relationship was also documented for the number of amide, carboxylic acid and hydroxyl functional groups, therefore for general good PPI performance the number of these contained within the polymer should be minimised. Molecular weight is a poor predictor of performance, having only a minor influence, and in some cases higher molecular weight enhances the precipitation process. The importance of ionic interactions to the ability of a PPI to stablise the supersaturated state was demonstrated by the advantage of choosing a polymer with an opposite charge with respect to the drug. Additionally, when the polymer charge is the same as the supersaturated drug, precipitation is likely to be enhanced. A PCA model based on polymer molecular properties is presented, which has a central oval region where the polymer will general perform well across all three drug types. If the polymer is located outside of this region, then they either show compound specific inhibition or enhance precipitation. Incomplete separation of the PPI performance based on the molecular properties on the polymers indicates that there are some further molecular properties that might improve the correlation.


Page 2 of 68

Page 3 of 68

Abstract graphic PPIs with similar precipitation inhibition performance (colour scale green-red) have similar molecular properties

8000

No Polymer 0.0001% 0.0003% 0.001% 0.003% 0.01% 0.03% 0.1% 0.3%

6000

Signal

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


4000

Increasing polymer concentration decreases precipiation rate

2000

0 0

1000

2000

3000

4000

5000

Time (sec)

Keywords supersaturation, lipid formulation, crystallisation, polymer, inhibition, metastable



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1 INTRODUCTION The successful delivery and subsequent absorption of poorly water soluble drugs after oral administration is challenging, and the bioavailability of many Biopharmaceutical Classification System (BCS) Class II drugs is limited by slow or incomplete dissolution. Strategies to address the issue of low water solubility include techniques such as modulation of crystal form, particle size reduction, salt formation, and the development of enabling formulations such as solid dispersions or lipid based formulations 1. Each of these approaches aims to increase the apparent solubility of a poorly water soluble drug within the gastrointestinal tract or to increase the rate of dissolution. Recently, however, it has become apparent that the requirement for effective solublisation within the gastrointestinal tract need not require significant change to equilibrium solubility 2-4. Instead, transient periods of supersaturation may be sufficient to maintain solublisation and to provide the driving force for absorption. Indeed supersaturation may provide advantages for absorption over and above that of sub-saturated solutions by virtue of an increase in thermodynamic activity. The generation of a supersatured state and subsequent inhibition of precipitation has been described, by analogy, as a “spring and parachute” 5. The “spring” to create supersaturation may occur, for example, via dissolution of solid dose forms containing amorphous drug, by dissolution of pharmaceutical salts or by the dispersion or digestion of lipid based delivery systems 4. Approaches to reduce the rate of precipitation from supersaturated solutions, i.e. potential “parachutes”, include the addition of polymers, surfactants and cyclodextrins 6. Polymeric precipitation inhibitors (PPIs) have been shown to stabilise supersaturation and therefore enhance oral absorption in both solid dose forms such as solid dispersions 7 and liquid and semi-solid formulations such as lipid based formulations 3, 4, 8-11. For example, Gao et al. 3 found that the addition of 5% w/w hydroxypropylmethyl cellulose (HPMC) to a self-emulsifying drug delivery system (SEDDS) of paclitaxel resulted in oral bioavailability of 9.5% in male Sprague-Dawley rats compared to 0.9 % in the absence of the polymer. PPIs aim to maintain drug in a kinetically stabilised (but thermodynamically unstable) supersaturated state for sufficient time to allow absorption. In doing so, PPIs slow down drug precipitation/crystallisation via inhibition of nucleation, crystal growth or both. In contrast, PPIs typically have little impact on drug solubility directly, i.e. via cosolvency, when incorporated into formulations at relatively low proportions 12. The ability of PPIs to kinetically stabilise the supersaturated state is thought to result from intermolecular interactions between the drug and polymer in solution (through hydrogen bonding, hydrophobic and ionic interactions), the ability of the polymer to sterically hinder the crystallisation process, a decrease in the drug selfdiffusion coefficient or combination thereof 13-15. The current study presents the results of an extensive screen of the potential utility of a range of PPIs with a series of model poorly water soluble drugs and represents a significant expansion of previously published studies 13. Data from the literature suggests that common attributes of a “good” PPI exists. However, the current scarcity of data and complexity of the gastrointestinal lumen are hindering their identification beyond specific examples. In order to obtain the 4 ACS Paragon Plus Environment

Page 4 of 68

Page 5 of 68

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


fundamental information to address this gap in current understanding, this study of PPI using a wide variety of polymers classes and model drugs was initiated. A simplified approach was adopted to remove other phenomenological effects. This method does not model the processes that may occur within a physiological system; however it does allow isolation of the influence of the PPI on supersaturated drug precipitation and the relationship between the drugs and the polymers inhibiting precipitation. The drugs employed include non-electrolytes – carbamazepine, danazol and ethinylestradiol, weak bases – amiodarone, halofantrine and itraconazole, and weak acids – meclofenamic acid, mefenamic acid and tolfenamic acid. The properties and structures of these drugs are shown in Table 1. The PPIs cover a diverse range of 42 different classes of polymers, including cellulose derivatives (such as hydroxypropylmethyl cellulose, hydroxypropylmethyl cellulose acetate succinate and methyl cellulose), poly methacrylates, poloxamers, polyvinyl alcohol, polystyrene sulfonate and polyvinyl pyrrolidone. A total of 78 polymers across all classes, grades and molecular weights were tested to varying degrees. The molecular structures of the well defined polymers are shown for the cellulose derivatives in Table 2 and all other polymers in Figure 2. A number of the polymers tested have poorly characterised molecular structures and are omitted from the table and figure. Principle component analysis (PCA) was used to identify specific and generalized PPIs and drug molecules that cluster together with similar variable profiles, to determine the link between PPI performance and polymer molecular properties, to determine whether PPI performance is a function of differences in polymer and drug charge, and validate this model. The data has allowed the identification of the key properties of PPIs that dictate optimal utility when coadministered with poorly water soluble drugs.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

2 MATERIALS AND METHODS 2.1 MATERIALS The following polymers were obtained from Sigma-Aldrich Pty Ltd, Australia; alginic acid (G Alginic), gum arabic (G Arabic), locust bean gum (G Locust), xanthan gum (G Xanthan), hydroxylethyl cellulose ethoxylate quaternized (HECEQ), hydroxypropylmethyl cellulose (HPMC), methyl 2-hydroxyethyl cellulose (MHEC), poly(acrylic acid) (PAA 1.8, PAA 1,250 and PAA 3,000), poly allylaimine hydrogen chloride (PAAH), poly(2-acrylamido-2-methyl-1propanesulfonic acid) (PAAMPS), polyacrylamide (PAC), poly(acrylamide-co-acrylic acid) (PACAA 200 and PACAA 5,000), poly(acrylamide-co-diallyldimethylammonium chloride) (PACDDA), polyanetholesulfonic acid (PAESA), poly(bis(2-chloroethyl) ether-alt-1,3-bis(3(dimethylamino)propyl)urea) (PCEDPU), poly diallyldimethylammonium chloride (PDDA), poly(dimethylamino-co-epichloorohydrin-co-ethylenediamine) (PDEE), polyethyleneimine (PEI), poly(ethylene glycol)-block-poly(propylene glycol)-block-poly(ethylene glycol) (PEPE 1.1 PEPE 14.6) poly(2-ethyl 2-oxazoline) (PEOX 50, PEOX 200 and PEOX 500), polyepoxysuccinic acid (PESA), poly(glycidyl methacrylate) (PGMA), poly(methyl vinyl ether-alt-maleic anhydride) (PMVEM), poly(propylene glycol) bis(2-aminopropyl ether) (PPGAE), poly(sodium 4-styrene sulfonate) (PSSS 70 and PSSS 1,000), poly(4-styrenesulfonic acid-co-maleic acid) (PSAMA), polyvinyl alcohol (PVA 50 and PVA 94), polyvinylpyrrolidone (PVP 10, PVP 40, and PVP 360) and cross-linked PVP (PPVP), and poly(1-vinylpyrrolidone-co-2-dimethylaminoethyl methacrylate) (PVPDAM). HPMC E5, HPMC E4M(1), HPMC E4M(2), HPMC E10M and HPMC F4M, and methylcellulose (MC A15 and MC A4M) was supplied by The Dow Chemical Company, c/o Colorcon Asia Pacific Pty Ltd, Australia. Hercules Chemical Company Inc, c/o APS Healthcare, Nuplex Industries Pty Ltd Australia supplied the following polymers; ethyl cellulose N100 (EC), hydroxyethyl cellulose (HEC 250GF), hydroxypropyl cellulose HXF (HPC), HPMC K200M and K4M, MC A4C and sodium carboxylmethyl cellulose (SCMC 7H, SCMC 9M and SCMC 12M). HEC 30000 was supplied by Shandong Head Co Ltd and Carbomers (C 340 and C 934) by Suichang Tinci Materials Technology Co Ltd, all c/o Ceechem Pty Ltd, Australia. The Eudragit polymers (E E100, E L100, E L10055, E RL100 and E S100) were supplied by Evonik Degussa Australia Pty Ltd. Pluronic polymers (P F68, P L62, P L64, P L92, P P103 and P P105) were obtained from BASF Australia Ltd. HPMC 606 and HPMC 904, hydroxypropylmethyl cellulose acetate succinate (HPMCAS LF, HPMCAS MF and HPMCAS HF), hydroxypropylmethyl cellulose phthalate 55S (HPMCP) and MC SM4 samples were obtained from ShinEtsu Chemical Co. Ltd, c/o ANZChem Pty Ltd, Australia. A complete alphabetical listing of the abbreviations used to identify the polymers can also be found in the Supporting Information. Danazol was supplied by Sterling Pharmaceuticals, Australia, halofantrine HCl by GlaxoSmithKline Pharmaceuticals (India), itraconazole by Lee Pharma Ltd, India, amiodarone HCl, carbamazepine, ethinylestradiol, meclofenamic acid sodium salt, mefenamic acid and 6 ACS Paragon Plus Environment

Page 6 of 68

Page 7 of 68

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


tolfenamic acid by Sigma-Aldrich Pty Ltd, Australia. Analytical grade sodium dihydrogen phosphate, disodium hydrogen phosphate and sodium chloride were used for the aqueous phase buffer. Propylene glycol was obtained from Merck Pty Ltd, Australia.

2.2 SUPERSATURATION CHALLENGE USING SOLVENT SHIFT The solvent shift method was used to generate a supersaturated solution of the studied poorly water soluble drugs in solution containing different polymers at varying concentrations. Drugs were initially dissolved in the cosolvent (propylene glycol) at close to maximum concentration (90% of saturated solubility), with the following concentrations (mg cm-3); 1.0 amiodarone, 10.0 carbamazepine, 1.0 danazol, 1.0 ethinylestradiol, 1.0 halofantrine, 1.0 itraconazole, 8.0 meclofenamic acid, 2.0 mefenamic acid and 3.0 tolfenamic acid. For a particular drug, this maximised the precipitation “challenge” and therefore provided conditions under which the relative performance of the polymers could be most readily probed. This approach led to variability in the ultimate extent of supersaturation across the different drugs. However, in the current studies, the ability of the polymers to inhibit precipitation was the focus of the work conducted, rather than an exploration of intrinsic precipitation patterns and this was not highly dependent on the supersaturation ratio generated (see Supplementary Information, Influence of supersaturation ratio on precipitation inhibition performance). Propylene glycol was selected as the cosolvent due to more similarities in its physical properties to lipid formulation components than other cosolvents previously used to generate a supersaturated solution, i.e., dimethylformamide 16, 17, dimethylacetamide 18, 19 and 1,3-dioxolane 20. Propylene glycol is also an acceptable component for oral administration. The cosolvent phase of propylene glycol (300 µl) was then added into 3,000 µl of buffered aqueous phase (simulated endogenous intestinal fluid, pH 6.5, 18 mM NaH2PO4, 12 mM Na2HPO4, 98 mM NaCl, ionic strength 0.152 M 21) within a 20 ml glass vial (Econo Glass Vial, PerkinElmer) containing dissolved polymer as required, thereby generating the supersaturated solution. These final dispersed solution conditions were maintained for all experiments. The phosphate buffer system was utilised due to its simplicity and common use. Nonetheless, aqueous phase conditions can have an effect on the degree of supersaturation22 and future studies might usefully explore the impact of differing, e.g. biorelevant, media on PPI performance. The resulting aqueous dispersed system (90.6% w/w water and 9.4% w/w propylene glycol) generated the following approximate drug supersaturation ratios (based on aqueous solubilities from Table 1); 1 amiodarone, 3 carbamazepine, 145 danazol, 11 ethinylestradiol, 545 halofantrine, 1090 itraconazole, 13 meclofenamic acid, 8 mefenamic acid and 628 tolfenamic acid. The principal polymer aqueous phase concentrations tested were 0.001 and 0.1% w/v. This range was selected based on the realization that an approximate concentration of 0.02% w/v would be obtained when a 1 g capsule containing 5% w/w polymer is dispersed within stomach contents of 250 cm3. After cosolvent addition, four x 300 µl samples of the dispersion were taken and immediately placed into wells of a 96-well microplate and the plate introduced into the nephelometer. Blanks containing propylene glycol (without dissolved drug) dispersed into the appropriate aqueous phase were run in parallel. The delay, 30 to 60 s, between mixing of solutions and the first nephelometer turbidity reading was recorded. All measurements were performed at room temperature, 20oC. 7 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Solution turbidity was monitored for a period of 1.5 h (5400 s) using a NEPHELOstar Galaxy (BMG Labtechnologies) microplate nephelometer (λ = 635nm laser), which records the turbidity as a function of back scattered light, not light adsorption. The following program settings were used; gain = 70, cycle time = 30 s, measurement time per well = 0.30 s, position delay 0.5 s, orbital shaking with 2 mm width for 5 s at end of each cycle. 96-well microplates, comprising flat bottomed wells made from polystyrene (NUNC) were used.

2.3 DATA CLASSIFICATION AND ANALYSIS The signal versus time data provided by the nephelometer provides a measure of the amount of precipitate present within the sample well. Details of how this assertion was validated are given below. In general, precipitation leads to the generation of particulates that in turn result in scatter and, ultimately, this manifests as an increase in turbidity. Due to the complex nature of the signal versus time curves observed, a classification system of three characteristic precipitation profiles has been developed 13, see Figure 1. Precipitation curves of Type C, see Figure 1c, have a high initial rate of turbidity increase (i.e. precipitation) that slows with time, reaching an asymptotic maximum value when complete. This is further divided into two subtypes; C1 (typical of profiles obtained in the absence of a PPI) and C2, which exhibits a second phase in which the signal subsequently fluctuates and then decreases due to precipitate flocculation. Type B (of which there are three subtypes, see Figure 1b) is the more complex of the three types, with 2 or 3 distinct precipitation rates that all reach an asymptotic maximum value by the end of the experiment. B1 has a higher initial rate which slows dramatically and after a further delay accelerates again, B2 has a moderate initial rate which also accelerates after a delay, and B3 has a slow initial rate that also accelerates after a delay time. The final type is Type A, see Figure 1a, where precipitation is inhibited for the entire period of the experiment; A1 has precipitation almost completely inhibited within detection limits. A2 shows precipitation occurring at a very slow and decreasing rate. PPIs that show Type A1 behaviour are the most desirable, Types A2, B1 and B2 may be suitable depending on the period required to support absorption. Systems showing C1/C2 profiles are the least effective at precipitation inhibition, with some even increasing the precipitation rate. For each combination of drug and polymer, the precipitation curve was classified and precipitation rate(s) determined using linear regression of the appropriate sections of the signal versus time curve, see Figure 1d. The lower limit of detection for the precipitation rate using this technique was 0.0010 units s-1, or 5.4 units over the time period of the precipitation experiments. (Note that in an attempt to simplify classification, labelling of the regions has changed relative to that originally proposed by Warren et al. (2010) 13; with Region III and Region I being interchanged ). The performance of the PPIs were quantified by calculating the relative rate of precipitation (RR), defined as the logarithm of the ratio of the precipitation rate (R) in the presence of polymer to the rate with no polymer present (Rno polymer), see Equation 1. A logarithmic scale was required since the precipitation rates observed ranged from 0.001 to 100 units s-1, a variation of over five orders of magnitude. For example, a RR of -2 indicates that the 8 ACS Paragon Plus Environment

Page 8 of 68

Page 9 of 68

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


precipitation rate was decreased by a factor of 100. RI, RII and RIII denote the relative rate of precipitation in Region I, II and III (see Figure 1d), respectively. The RR was used to classify the performance of a given PPI using the following performance ranking system; superior (RR

Evaluation of the Structural Determinants of Polymeric Precipitation

Recommend Documents