Article pubs.acs.org/crystal
Nanocrystal Recovery by Use of Carrier Particles Shahzeb Khan,† Marcel de Matas,‡ Smitha Plakkot,§ and Jamshed Anwar*,∥ †
Computational Biophysics Laboratory, Institute of Life Sciences Research, University of Bradford, Bradford, West Yorkshire BD7 1DP, United Kingdom ‡ Product Development, AstraZeneca, Redesmere Building, Silk Road Business Park, Charter Way, Macclesfield SK10 2NA, United Kingdom § Lena Nanoceutics Limited, Bradford, West Yorkshire BD7 1DP, United Kingdom ∥ Chemical Theory and Computation, Department of Chemistry, Lancaster University, Faraday Building, Lancaster LA1 4YB, United Kingdom ABSTRACT: Aqueous dispersions of nanoparticles (nanosuspensions) of poorly soluble drugs are an effective option for addressing issues of low and erratic bioavailability. They are, however, not attractive as dosage forms due to their predisposition to physical and chemical instability. Here we describe an effective method for isolating nanocrystals in solid form from a suspension, which opens up the possibility of formulating nanocrystals in solid dosage forms such as tablets. The method involves the use of carrier particles to adsorb and recover nanocrystals from a liquid suspension. The method is illustrated by using carrier particles of dibasic calcium phosphate to recover nanocrystals of ibuprofen and glibenclamide produced by both size reduction and crystallization. Respective recoveries of the nanocrystals were in excess of 90%. Powders of carrier−nanocrystal particles yielded dissolution rates similar to those of the native nanocrystals and substantially faster than marketed tablets and micronized suspensions of the drugs, confirming that the high surface area of the nanocrystals is retained during the adsorption process.
■
INTRODUCTION The problem of low aqueous solubility is a significant issue for many marketed drugs and potential drug candidates that can result in poor and erratic bioavailability. Indeed, up to 40% of molecules in company development pipelines exhibit markedly low aqueous solubility, as do upward of 60% of molecules derived from high-throughput screens.1,2 For these molecules, consistent and enhanced bioavailability can be achieved if the dissolution rate can be sufficiently enhanced or if the drug substance can be solubilized in the gastrointestinal milieu. Approaches used to enhance dissolution rate include particle size reduction,3 enhancement of solubility through production of high-energy solid forms (e.g., solid dispersions),4 and presentation of the drugs in the form of nanocrystals.5 Besides their immense surface area, the literature suggests that nanocrystals also offer increased saturation solubility alongside a concomitant decrease in the diffusional pathway adjacent to the nanocrystal surface, which can all converge to substantially increase the rate of dissolution.6 Although the technology for producing nanocrystal-based formulations is maturing, there are still important issues that have limited its adoption and application in the pharmaceutical industry.7−9 A particular issue is the difficulty experienced in the recovery and isolation of nanocrystals in the dry state for incorporation into solid dosage forms such as tablets. Presenting the drug in the © 2014 American Chemical Society
solid state has the potential to markedly enhance the physical and chemical stability of nanocrystal formulations. Current methods for isolating nanocrystals in the dry state include spray drying,10 freeze-drying,11,12 and nanosuspensions being used as granulating fluids for wet granulation followed by drying.13 Each of these methods has its own limitations. Spray drying and wet granulation accompanied by drying can result in chemical instability due to the high energy input involved in solvent vaporization.14 Freeze-drying, in contrast, is an expensive, timeconsuming process that can adversely affect the resulting particle size distribution,12 while the use of nanosuspensions as granulating fluid can result in irreversible aggregation of primary particles with the potential for decreased dissolution rate.15 We report here a simple, effective approach for recovering nanocrystals from a suspension by use of inert carrier particles. While the concept of using carrier particles is not new (consider its widespread application in dry-powder inhaler formulations16,17 and in ordered mixing18), to our knowledge this is the first open literature application of this approach for isolating nanocrystals from suspension. The isolation of Received: September 25, 2013 Revised: December 14, 2013 Published: January 13, 2014 1003
dx.doi.org/10.1021/cg401432m | Cryst. Growth Des. 2014, 14, 1003−1009
Crystal Growth & Design
Article
and turbulence generated within this narrow gap provide the potential for rupture and shearing of particles, leading to an ultrafine product in the submicrometer size range.21 The suspension produced during processing is continually recycled through a stainless steel screen, which retains the grinding media and prevents contamination of product. The drug material was presented for size reduction in the form of a stabilized aqueous suspension, with the dispersion medium (150 mL) being composed of sodium lauryl sulfate (0.1% w/w), PVPK30 (0.5% w/w), and 6 cP grade HPMC (0.5% w/w). The drug substance was mixed with the stabilizer solution to give a 250 mL suspension containing 2.6% (w/w) of the drug material. The resultant suspension was then placed into the feedstock hopper of the size reduction system. The suspension was processed for 60 min by recycling through the size reduction chamber. In-process samples were taken at intervals of 5, 10, 15, 30, 45, and 60 min, and the particle size was measured by dynamic light scattering (DLS) on a Zetasizer Nano instrument (Malvern Instruments Ltd., U.K.). For glibenclamide, nanocrystals were also prepared by controlled crystallization by infusing 10 mL of drug solution (5 mg/mL) dissolved in polyethyene glycol 400 (PEG-400) at a rate of approximately 100 mL/min into 90 mL of a stabilizer solution composed of HPMC 15 cP (1% w/v) and PVP K-30 (0.5% w/w). Details of the method have been presented earlier.22 The solution was stirred in a beaker by use of a magnetic stirrer at 1200 rpm, with the sample temperature maintained at 25 °C. Five batches were prepared to ascertain the variability in the process. The effect of stirring rate was also examined with the stirring rate being set to approximately 400, 800 and 1200 rpm. The particle size stability of glibenclamide and ibuprofen nanosuspensions was monitored regularly by dynamic light scattering (DLS) over a period of 1 month, with the nanosuspensions being stored at three different temperatures (4, 25, and 40 °C). These studies enabled an assessment of the degree of aggregation and Ostwald ripening in the nanosuspensions. Nanocrystal Recovery by Carrier Particles. The carrier particle material used was dibasic calcium phosphate anhydrous (DCP). Particle size analysis, carried out on a Sympatec HELOS and RODOS laser diffraction particle size analyzer (Sympatec Instruments, U.K.), revealed D90 = 44.85 μm, D50 = 33.50 μm, and D10 = 20 μm. The adsorption efficiency of the carrier particles was investigated in a range of carrier particle concentrations: 30, 60, 90, 120, 150, and 180 mg/ mL. The carrier particles were added to 10 mL of each nanosuspension with the resultant sample then stirred for 5 min by use of a magnetic stirrer at 400 rpm at a temperature of 25 °C. The samples were then filtered through 0.4 μm filter paper. The filtered extracts were dried at room temperature and subjected to analysis of active agent content by HPLC (see below). To enable an objective comparison of adsorption efficiency, the glibenclamide nanosuspensions prepared by comminution were diluted to a concentration equivalent to the nanosuspensions prepared by controlled crystallization, giving a final drug concentration of 0.49 mg/mL for both process variants. The drug content in the isolated nanocrystal−carrier powders was determined by HPLC by weighing a known quantity of the dried adsorbate and diluting with the mobile phase [55/45 (v/v) acetonitrile and water for ibuprofen; 55/45 (v/v) ammonium phosphate (0.02 M) and acetonitrile for glibenclamide] to a volume of 100 mL to give a nominal concentration of 20 μg/mL, if it was assumed that complete adsorption of drug had occurred. The resultant solution was sonicated and then centrifuged at 14 700 rpm for 30 min to allow the dissolution of available drug substance in the mobile phase. The supernatant layer was then analyzed in triplicate for drug content by HPLC. The injection volume was 25 μL for ibuprofen solutions and 50 μL for samples containing glibenclamide. The HPLC system comprised a Waters 2695 Model connected to a UV detector. For glibenclamide, the UV detector wavelength was set at 254 nm. The mobile phase solvent system comprised monobasic ammonium phosphate (0.02 M) and acetonitrile at a ratio of 45/55 (v/v). The flow rate of the mobile phase was set at 1.5 mL/min. An Ultra II TM C18 5 μm 250 × 4.6 mm column was used, which was
nanocrystal−carrier particles in powder form presents the possibility of developing solid dosage forms such as tablets and capsules for oral administration with a marked enhancement in dissolution rate. A key benefit is the complete elimination of the problems of aggregation and Ostwald’s ripening that plague nanocrystals in a liquid environment. The nanocrystal−carrier particles are shown to consistently reproduce the high dissolution rate of the original nanocrystal suspension. The methodology is robust and from the industry’s perspective relatively low-cost. Given that adsorption of the nanocrystals onto the carrier particle must depend on the nature of the drug, crystal particle size, and surface characteristics of the stabilized nanocrystals, we evaluated the adsorption efficiency for nanocrystals of two model drug compounds, ibuprofen and glibenclamide (Figure 1), produced by comminution, and in
Figure 1. Molecular structures of (a) glibenclamide and (b) ibuprofen.
the case of glibenclamide also by a simple crystallization process. For the carrier particle, we have selected dibasic calcium phosphate, as it has a low aqueous solubility and has previously been reported as good inert carrier for adsorption of microparticles.19
■
MATERIALS AND METHODS
Materials. The chosen model drugs glibenclamide (batch PPC/08/ GLB/057) and ibuprofen (batch 7050-1077) were purchased from Anzen Exports, Kolkata, India, and Albemarle Corp., U.S., respectively. Hydroxypropylmethylcellulose (HPMC) of viscosity grades 6 cP (batch 8028213) and 15 cP (batch 7068037) were kindly provided by Shin-Etsu, Japan. Polyvinylpyrrolidone K-30 (PVP), batch 08297047GO, was provided gratis by BASF, Germany. Sodium lauryl sulfate (SLS, batch 08421LE), sodium hydroxide (NaOH, batch S47417-479), and monobasic potassium phosphate (MPP, batch SZE90330), were purchased from Sigma Aldrich, UK. Poly(ethylene glycol) 400 (PEG-400, batch 0917861) and acetonitrile (batch 0809411) were purchased from Fisher Scientific, U.K. Calcium phosphate dibasic (CaHPO4 anhydrous, batch A0280641) was purchased from Acros Organics, Belgium. Glibenclamide 5-mg tablets (batch 9A06HL) were obtained from Teva Ltd., U.K., and ibuprofen 200-mg tablets (batch LL11845) from Wockhard Ltd., U.K. The tablets were used in comparative dissolution studies. Preparation of Nanocrystals. Nanocrystals of glibenclamide and ibuprofen were prepared by use of the DENA DM100 size reduction system.20 The DENA DM100 system comprises a fast-rotating conical rotor (1500 rpm) constructed from a soft polymer, which sits inside a conical polymeric sleeve. Grinding media (0.2 mm yttrium-reinforced zirconium beads) are housed inside indentations within the rotor that form a narrow gap between the outer sleeve and rotor. The high shear 1004
dx.doi.org/10.1021/cg401432m | Cryst. Growth Des. 2014, 14, 1003−1009
Crystal Growth & Design
Article
maintained at a temperature of 30 °C. For ibuprofen content, the wavelength of the UV detector was set at 214 nm and a Vydac 202TP C18 5 μm, 4.6 × 250 mm column was employed with the temperature maintained at 30 °C. The mobile phase consisted of a binary 50/50 (v/v) mixture of water and acetonitrile. The pH of the mobile phase was 2.8, and the flow rate of the mobile phase was set to 1 mL/min. Scanning Electron Microscopy. Scanning electron microscopy (SEM) images were taken of the raw dibasic calcium phosphate carrier material, along with that of the composite samples (nanocrystals of the drug adsorbed onto the carrier particles) by use of a Quanta 400 SEM (FEI Company, Cambridge, U.K.). These images were taken to confirm surface adsorption of the nanocrystals. The sample preparation involved fixing the powder samples on to a metal stub with the aid of double-sided adhesive tape. The SEM was calibrated by use of a gold grid supplied with the instrument. Dissolution Testing. Dissolution tests were carried out in triplicate for nanosuspensions of glibenclamide and ibuprofen, composite powder samples of adsorbed nanocrystals on carrier particles, and microcrystalline suspensions of both drugs. These tests were carried out to ascertain whether the dissolution rate of the nanocrystals was compromised by adsorption onto carrier particles. Each test employed a unit dose equivalent to 5 mg of glibenclamide and 200 mg of ibuprofen. The actual sample masses were 1.8 and 2.8 g for the glibenclamide nanocrystal−carrier powder (for nanocrystals produced by comminution and controlled crystallization, respectively) and 81.63 g for ibuprofen nanocrystal−carrier powder. These rather substantial sample sizes reflect the low loading that appears to be a characteristic of the developed carrier particle approach to nanocrystal recovery. The USP dissolution apparatus II (USP, 2008) was employed with 900 mL of the dissolution medium (buffer at pH 7.2 for ibuprofen and pH 7.5 for glibenclamide) at 37 °C with a paddle speed of 75 rpm. Aliquots of 5 mL of the dissolution medium were sampled by use of a syringe fitted with a filter (0.2 μm) at 0, 2, 6, 10, 15, 30, 45, and 60 min, and the sample volume was replaced with an equivalent volume of fresh medium. All the samples were analyzed by the HPLC method given above to quantify drug concentration. The percentage of the nominal dose of each drug released for each of the samples at specific time intervals was then calculated.
■
narrow size distribution.23 Nanocrystals of glibenclamide prepared by controlled crystallization had an average particle size of approximately 300 ± 3 nm with low polydispersity 0.2 ± 0.02; the particle size here was slightly lower than that attained by comminution, 342 nm. Physical stability of the nanocrystals is an important issue, and nanocrystals in suspensions are prone to aggregation and Ostwald’s ripening with increased potential for sedimentation and caking. Differences in stability between formulations might be an indicator of unique particle characteristics relating to the methods of particle generation and the associated formulation components. These differences could go on to induce unpredictable variation in the adsorption results. Furthermore, if there were significant changes taking place in the size of nanoparticles during storage, the extent of adsorption could in principle vary, depending on the days lapsed after the nanocrystals were prepared. In view of these considerations, the particle size stability of the nanoparticles as a function of time was assessed at three temperatures, 4, 25, and 40 °C. For both drugs produced by either of the two distinct methods (size reduction and crystallization), the particle size was found to be relatively stable for periods up to 30 days. Only a slight increase in the average particle size was observed during the first 10 days (Figure 2), with a stable size distribution observed thereafter. The adsorption studies were carried out after the 10-day “equilibration period” had elapsed. The efficiency of nanocrystal recovery (percentage mass of nanocrystals adsorbed relative to that in the initial nanosuspension) as a function of carrier particle concentration is shown in Figure 3. The results reveal an increasing percent recovery of nanocrystals with increasing carrier particle
RESULTS AND DISCUSSION
The systems being studied, nanosuspensions and nanocrystals adsorbed onto carrier particles, are clearly heterogeneous in nature. Therefore, one would expect variability in results, which can make it hard to generalize findings. In view of this consideration, it is essential that we characterize the systems thoroughly, to control known factors and cover the effect of broader variables. We have followed this philosophy here and have considered two distinct molecules for the study and also utilized the two main (and disparate) approaches for preparing nanocrystals, namely, top-down comminution and bottom-up crystallization. The results clearly reveal that the carrier particle approach to recovering nanocrystals from suspension is effective and robust. It is possible to recover/isolate in general beyond 90% by mass for ibuprofen, and beyond 95% for glibenclamide from aqueous suspensions, with the resultant powders giving rapid dissolution rates similar to those observed for the original aqueous nanosuspensions. The effectiveness of the comminution approach using the DENA DM100 size reduction system has been detailed earlier.20,21 Monitoring the particle size as a function of processing time revealed that 45 min was sufficient to induce the maximum attainable reduction in particle size, after which further size reduction was limited. After 60 min the respective particle size for glibenclamide and ibuprofen was 342 ± 3 and 440 ± 5 nm, respectively. Average polydispersity index (PDI) values for both drug materials were
