International Journal of Pharmaceutics 461 (2014) 367379Contents lists available at ScienceDirectInternationalJournalofPharmaceuticsj our nal homepage: www. el sevi er . com/ l ocat e/ i j phar mPharmaceuticalNanotechnologyRedispersiblefastdissolvingnanocompositemicroparticlesofpoorlywater-solubledrugsAnaghaBhakay,MohammadAzad,EcevitBilgili,RajeshDaveOtto H. York Department of Chemical, Biological and Pharmaceutical Engineering, New Jersey Institute of Technology, Newark, NJ, USAart icleinfoArticle history:Received 3 July 2013Received in revised form9 October 2013Accepted 30 November 2013Available online 11 December 2013Keywords:NanoparticlesWetmedia millingFluidized bed coatingSpray dryingRedispersionFast dissolutionabstractEnhancedrecovery/dissolutionoftwowetmedia-milled,poorlywater-solubledrugs,Griseofulvin(GF)and Azodicarbonamide(AZD),incorporatedintonanocompositemicroparticles(NCMPs)viauidizedbeddrying(FBD)andspray-drying(SD)wasinvestigated.Theeffectsofdryingmethod,drugloading,drugaqueoussolubility/wettabilityas wellas synergisticstabilizationofthemilledsuspensionsonnanoparticlerecovery/dissolutionwereexamined.DrugnanoparticlerecoveryfromFBDandSDpro-duced NCMPshavinghighdrugloadingswasevaluatedupongentleredispersionviaopticalmicroscopyandlaserdiffraction.Duringwet-milling,hydroxypropylcellulose(HPC)alonestabilizedmorewettabledrug (AZD)nanoparticleswithslightaggregation,butcouldnotpreventaggregationoftheGF nanopar-ticles.Incontrast,well-dispersed,stablenanosuspensionsofbothdrugswereproducedwhensodiumdodecylsulfate(SDS)andHPCwerecombined.TheFBDandSDNCMPswithoutSDSexhibitedincompletenanoparticlerecovery,causingslowerdissolutionforGF,butnotforAZD,likelyduetohigheraqueoussolubility/wettabilityofAZD.ForhighactiveloadedNCMPs(FBD50wt%,SD80wt%)ofeitherdrug,HPCSDStogetherowingtotheirsynergisticstabilizationledtofastredispersibility/dissolution,cor-roboratedviaopticalmicroscopyandparticlesizing.Thesepositiveattributescanhelpdevelopmentofsmaller,highdrug-loadeddosageformshavingenhancedbioavailabilityandbetterpatientcompliance. 2013 Elsevier B.V. All rights reserved.1. IntroductionPoor aqueous solubility has been a major issue in achieving ade-quate oral bioavailabilityfor a large percentage of newlydiscovereddrug compounds (Kesisoglou et al., 2007). Size reduction of drugcrystals increases the specic surface area, which can increase thedissolution rate of drugs according to the NoyesWhitney equation(Noyes and Whitney, 1897). Wetstirred media milling is a robustprocess and has been commonly used to produce drug nanopar-ticles with desired drug loadings (Bruno et al., 1996; Muller andPeters, 1998; Bilgili andAfolabi, 2012). Nanoparticles showa strongtendency to aggregate due to enhanced Brownian motion, rela-tively high surface energy, and large specic surface area (Guptaand Kompella, 2006). Therefore, steric, electrostatic or a combina-tion of both stabilizing mechanisms can be used during wet mediamilling to prevent aggregation and to stabilize the nanoparticlesduring storage (Kim, 2004). Combined use of polymers and anionicsurfactants is known to have synergistic stabilization effects onthe nanosuspensions (Ryde and Ruddy, 2002; Gupta and Kompella,2006). In spite of the use of stabilizers, nanoparticle aggregationand particle size growth by Ostwald ripening maystill take placeCorresponding author. Tel.: +1 973 596 5860; fax: +1 973 642 7088.E-mail address: dave@njit.edu (R. Dave).gradually upon prolonged storage. To mitigate these issues andto meet high patient/clinical demand for solid dosage forms,nanosuspensions are usually converted into dry nanocompositemicroparticles (NCMPs), which can be subsequently compressedinto tablets or lled into capsules.Spray drying (SD), freeze drying, and granulation/coating in u-idized bed (Ryde and Ruddy, 2002; Lee, 2003; Abdelwahed et al.,2006; Van Eerdenbrugh et al., 2008; Niwa et al., 2011; Bhakayet al., 2013; Cerdeira et al., 2013) have been widely used to convertnanosuspensions into nanocomposite microparticles and nallyinto various solid dosage forms. However, any form of drying ofthe nanosuspensions generally causes the nanoparticles to aggre-gate due to liquid removal leading to loss of the large surface areaof nanoparticles (Lee, 2003). Lee (2003) and Bhakay et al. (2013)observed that due to aggregation during the drying processes, thedrug nanoparticles formulated with a polymer, hydroxypropyl cel-lulose (HPC) as a dispersant, were not recovered fully and quicklyfrom the NCMPs after redispersion of the powders in water. Toovercome this issue, a signicant amount of water-soluble disper-sants/cryoprotectants such as sugars (lactose, trehalose, sucrose),sugar alcohols (Mannitol, xylitol), or cyclodextrins are added tothe suspensions before drying for effective recovery of nanopar-ticles from the NCMPs (Schwarz and Mehnert, 1997; Konan et al.,2002; RydeandRuddy, 2002; VanEerdenbrughet al., 2008; KimandLee, 2010). Unfortunately, the increase in dispersant loading has a0378-5173/$ see front matter 2013 Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.ijpharm.2013.11.059368 A. Bhakay et al. / International Journal of Pharmaceutics 461 (2014) 367379negative effect on the drug loading; hence, dispersant type/loadingmust be selected in such a way to ensure fast redispersibility/drugdissolution while still maintaining high drug loading in the NCMPs.Recent studies have shown that drug nanoparticles were com-pletely recovered from nanocomposite microparticles (NCMPs)prepared via uidized bed-drying (Basa et al., 2008; Bhakay et al.,2013) or spray-drying (Niwa et al., 2011; Cerdeira et al., 2013)when a combination of polymer such as HPC, polyvinylpyrrolidone(PVP), and hydroxypropylmethyl cellulose (HPMC) with an anionicsurfactant, sodium dodecyl sulfate (SDS), was used. Niwa et al.(2011) studied the effects of using a combination of PVP with SDSon the redispersion/dissolution of phenytoin nanoparticles fromspray-dried NCMPs with/without Mannitol; however, they did notconsider formulations with PVP alone or emphasize the effects ofdrugloadinginNCMPs whichchangedwith/without the additionofMannitol. Basa et al. (2008) reported fast redispersion/dissolutionof ketoconazole nanoparticles from uidized bed granules thatwere formulated with a combination of polymer with SDS, butdid not present a comparative analysis of the synergistic effects ofthe polymer with SDS. Bhakay et al. (2013) investigated the recov-ery of the GF nanoparticles fromNCMPs, formulated with/withoutsurfactants, and examined the synergy between HPC and SDS byconsidering the importance of redispersion in water as well asredispersion conditions. Their work compared various degrees ofagitation/shear during redispersion tests in order to identify mostsuitable method that can easily discriminate between different for-mulations in terms of ability to recover nanoparticles from dryNCMPs. However, they did not assess the effects of drug load-ing, aqueous solubility/wettability of drugs, and the drying methodeither on the redispersion or on the dissolution. The effects of drugloading were addressed by Bose et al. (2012) by increasing the drugloading from 10% to 20% in uidized bed granules, which slowedthe dissolutionrate of the drug fromNCMPs formulatedwithHPMCand TPGS (non-ionic surfactant). It is noted that all of the abovestudies used a single drying method.Besides formulation/drug parameters mentioned above, differ-ent drying processes can affect the morphology of the NCMPsand their redispersibility. Most studies on the recovery of drugnanoparticles from spray-dried (Lee, 2003; Niwa et al., 2011), u-idizedbed-dried(Basa et al., 2008; Bhakayet al., 2013), freeze-dried(Abdelwahed et al., 2006) and spray-granulated (Bose et al., 2012)dry powders used a single drying method. On the other hand,Kim and Lee (2010) compared the effects of three drying meth-ods namely vacuum, convective and freeze drying, on the recoveryof ve poorly water-soluble drug nanoparticles formulated withHPCandpolymeric dispersants suchas carrageenan, gelatin, alginicacid after drying. However, they did not examine the subsequenteffects on the dissolution proles of these powders. Bourezg et al.(2012) compared the effects of uidized bed drying, spray dry-ing, and freeze drying on the redispersibility of lipid nanoparticlescontaining the drug, spirinolactone, and their dissolution. How-ever, the drug loading in the dried NCMPs was low (2.3%, w/w)which was limited by the drug solubility in the lipid, and also drugwas not in the formof nanocrystals. Overall, these studies providemany useful insights by examining one or more important aspects,but also point to a need for comprehensive assessment of thecombinedeffects of drug loading, drying method, aqueous drug sol-ubility/wettability, and synergistic effects of polymer with anionicsurfactant on the redispersion/dissolution of drug nanoparticles,which is the subject of the current paper. To be specic, based onthe results presented in the literature discussed above, it wouldnot be possible to answer questions such as (a) would the syner-gistic stabilization work on any given drug; (b) what would be theconcentration of the polymer/surfactant that will be required forenhanced redispersion/dissolution as various other conditions arechanged such as the drug aqueous solubility, drying method, etc.;(c) how would the drying methods and use of additional disper-sants affect the synergistic stabilization; and, (d) would the changein drug loading require a different polymer/surfactant concentra-tion for improved drug nanoparticle recovery. The present papermakes an attempt to answer some of these questions.This study deals with the preparation of redispersible, fast-dissolving nanocomposite microparticles with high drug loadingvia dryingof wet-milledsuspensions of poorlywater-solubledrugs.Two drying methods: uidized bed drying (FBD) and spray dry-ing (SD) are compared and their effects are examined on thestructure/morphology/size of the NCMPs with various loadingsof two model drugs, Azodicarbonamide (AZD) and Griseofulvin(GF). The twodrying methods differ signicantly with regards todrying time scales of a drop of suspension: spray drying occur-ring within few seconds while uidized bed coating taking fewminutes (Masters, 1985; Bilgili et al., 2011). Additionally, eachdrying method produces different morphologies and hence dif-ferent microstructure upon drying that could affect the recoveryand dissolution from dried drug nanosuspensions, in particularconsidering that FBD employs water-soluble carrier particles. Thetwo drugs were selected because they differ in aqueous solubil-ity and wettability (see Table 1), which is unlike most previouslyreported studies. HPC or a combination of HPC with SDS wasusedfor imparting physical stability to the wet media-milled AZD andGF suspensions. HPC also acts as a dispersant or matrix former dur-ing drying. In addition to HPC and SDS, Mannitol wasused as anadditional dispersant in one of the spray-dried formulations, whichhelps lower the drug loading if requiredandfor the purpose of com-paring with the nanocomposites formed using the uid-bed dryingmethod. The prepared suspensions were dried via spray drying oruidbeddrying, inthe latter, wet-milledsuspensions were sprayedonto Pharmatosecarrier particles. The NCMPs were dispersed inwater employing paddle stirring (Bhakay et al., 2013) to evaluatethe extent of drug nanoparticle recovery which closely resemblesthe hydrodynamics for drug dissolution in a USP II (paddle) appara-tus. As will be shown, this study demonstrates that synergistic useof HPCSDS combination as dispersants leads to enhanced redis-persibility and drug dissolution for twodrugs, irrespective of themethod of drying and resultant particle size of the NCMPs, evenat high drug loadings in the NCMPs. Hence, such NCMPs mayhelpformulators to develop smaller tablets/capsules with higher load-ing of poorly water-soluble drugs for better patient compliance andenhanced efcacy of the drug.2. Material and methods2.1. MaterialsGF and AZD were used as model poorly water-soluble drugs inthis study (see Table 1 for details). GF and AZD were purchasedfromLetco Medicals (Decatur, Alabama, USA) and Pfaltz and BauerInc. (Waterbury, CT, USA), respectively. HPC (SL grade) and SDSTable 1Physicochemical properties of the poorly water-soluble drugs.Drugs Molecular weight(g/mol)logP Solubility in water(g/ml)Contact angle withwater ()Initial particle size(m) D10, D50, D90Griseofulvin (GF) 352.7 3.5 8.9 57 1.0 5.1, 19.9, 54.1Azodicarbonamide (AZD) 116.1 1.7 60 44 1.7 2.2, 5.1, 11.2A. Bhakay et al. / International Journal of Pharmaceutics 461 (2014) 367379 369Table2Formulation of the drug suspensions and drug content in the NCMPs.Run no. Drug Loading of drug(% w/w)aHPC (% w/w)aSDS (% w/w)aOther additives(% w/w)aDrying method Drug content in NCMPsFBDbSDb1 GF 10 0 0 FBD and SD 8.2 6.7 91.4 2.92GF 10 2.5 0 FBD and SD 12.4 5.4 81.0 3.93GF 10 2.5 0.5 FBD and SD 12.4 2.4 76.9 5.54GF 10 2.5 0.5 FBD 48.8 3.95GF 10 2.5 0.5 10 (Mannitol) SD 41.2 4.36AZD 10 2.5 0 FBD and SD 10.7 2.2 80.2 1.97AZD 10 2.5 0.5 FBD and SD 11.9 2.6 74.0 3.28AZD 10 2.5 0.5 FBD 40.4 3.49AZD 10 2.5 0.5 10 (Mannitol) SD 41.4 4.9a% (w/w) is with respect to weight of suspensions for suspensions and weight of NCMPs after drying.bFluidized bed dried (FBD) NCMPs and spray dried (SD) NCMPs.purchased fromNisso America Inc. (NewYork, NY, USA) and SigmaAldrich(Milwaukee, WI,USA), respectively, wereusedas stabilizersfor the wet media milling. Mannitol was used as an additional dis-persant inthe NCMPs andpurchasedfromPure Bulk Inc. (Roseburg,OR, USA). PharmatoseDCL 11, a grade of lactose monohydrateprovided by DMVInternational (Netherlands), was used as thewater-soluble carrier (core) for FBD.2.2. Preparation of drug suspensions by wet media millingThe formulations usedinthis study are presentedinTable 2. Thewet-milled suspensions and the NCMPs prepared are labeled basedon the precursor suspension formulation/composition through-out the paper. All percentages refer to weight by weight (w/w)with respect to (wrt) the weight of suspension, unless otherwiseindicated. As can be seen from Table 1, GF and AZD have differ-ent aqueous solubility, hydrophobicity/wettability, andmechanicalproperties. As a result of these differences in the properties of thetwo drugs, different milling conditions had to be used to preparenanosuspensions (suspensions with particle size