Accelerating Development for Challenging Compounds -Assessing Multiple Technologies in Parallel

A special article collection companion to the webinar broadcast 20th January 2016. To view the webinar, click here

Image courtesy of Catalent Pharma Solutions © 2016 Catalent, Inc. All rights reserved

Sponsored byhttp://www.catalent.com/

Table of Contents

IntroductionAccelerating development for challenging compounds by assessingMultiple technologies in parallel when trying to improve bioavailability Transcript of the webinar broadcast 20th January 2016 pages 1-3

A control strategy for bioavailability enhancement by size reduction: Effectof micronization conditions on the bulk, surface and blending characteristicsof an active pharmaceutical ingredientPowder Technology, Volume 258, May 2014, Pages 222-233Dolapo Olusanmi, Dimuthu Jayawickrama, Dongsheng Bu, Gary McGeorge,Helen Sailes, Joanne Kelleher, John F. Gamble, Umang V. Shah, Mike Tobyn pages 4-15

Continuous manufacturing of solid lipid nanoparticles by hot melt extrusionInternational Journal of Pharmaceutics, Volume 471, Issues 1–2, 25 August 2014, Pages 153-156 Hemlata Patil, Vijay Kulkarni, Soumyajit Majumdar, Michael A. Repka pages 16-19

Formulation and development of pH-independent/dependent sustained release matrix tablets of ondansetron HCl by a continuous twin-screw melt granulation process International Journal of Pharmaceutics, Volume 496, Issue 1, 30 December 2015, Pages 33-41Hemlata Patil, Roshan V. Tiwari, Sampada B. Upadhye, Ronald S. Vladyka, Michael A. Repka pages 21-29

TOCV4.indd 1 19/04/2016 09:58:12

IntroductionRestrictions in bioavailability, stability and the delivery profile leads to developmental challenges. Leveraging advanced technologies can result in achieving these criteria and help attain the target product profile. The questions that arise are:

How to assess the risk and the opportunities for these techniques? Why is it better to use them in parallel? How can we balance the risk and opportunities? What should we look for in terms of costs and timelines?

Specifically at mid-stage drug development, high attrition rates and high proportion of poorly soluble candidates in drug development portfolio mandates the need for sophisticated drug development strategies. In phase I, there is approximately 40% attrition of all molecules going into phase II; from phase II going into phase III the attrition is approximately 65%. This then translates to only 2 out of 10 phase I molecules actually reaching phase III. Furthermore, only 1 of those 2 from phase III will actually reach the patient later on. The industry in general acknowledges this attrition curve and agrees with executing a lean development paradigm to overcome it.

The trend towards more lipophilic drug candidates in early stage portfolio is driven by varieties of factors. For instance, the use of high throughput screening for hit generation and hit-to-lead optimization in the medical chemistry process and the very nature of the therapeutic targets. Approximately 70% of the molecules which arise in the developmental pipeline have low solubility and high permeability. In other words they are BCS class II compounds.

0

2

4

8

10

12

preclinical phase I phase II phase III submission patient

10

2 1

High Attrition Rate Demands Lean Development

6

Num

ber o

f Mol

ecul

es

1

The higher proportion of candidates with pronounced low solubility increases a number of downstream risks. For example, low bioavailability, high intra- or inter-individual variability and higher risk for a significant food effect. Advanced formulations technologies and strategies during mid stage can minimize these risks.

Parallel approach in early development can create valuable data and enable sound decision making during subsequent mid-stage development. It is advisable to utilize multiple platforms in parallel early on during developability and pre-formulation stage and phase I when considering formulation and the PK profile in humans a little bit closer. In the middle of phase I, there is a need to make a prominent decision, namely, which of the previously considered formulation platforms might be most suitable for phase II and beyond? In phase II, scalability topics such as formulation space, design of the pivotal formulation and clinical trial material supply chains become very important.

This eBook features two key technologies to enhance the bioavailability namely, solid dispersion and particle size reduction.

The first article by Dolapo Olusanmi et al demonstrates the effect of varying micronization conditions on an API for which micronization is deemed necessary to ensure consistent drug delivery after human administration. Material micronized to different extents are confirmed as different by surface area, surface energy, particle size analysis, bulk density and surface adhesion measurements.

Developability & Preformulation

Formulation & PK profile

Scalability

Manufacturability

Formulation and PK profile

suitable first-in-man formulations

bioavailability

Cmax, Cmin, t0.5

Scalability

formulation space

design of pivotal formulation

unit operations & process design

clinical trial material supply

shelf-life

Phase I Phase II

2

The second article by Hemlata Patil et al proposes and demonstrates the formulation of SLN for pharmaceutical applications by combining two processes: hot melt extrusion (HME) technology for melt-emulsification and high-pressure homogenization (HPH) for size reduction.

The third article by Hemlata Patil et al highlights the challenge in the development of pH-independent/dependent sustained release tablets of ondansetron HCl dihydrate (OND) given its its pH-dependent solubility and relatively short elimination half-life. The article demonstrates that the continuous melt granulation technique within a twin-screw extruder is a viable method to develop a sustained release tablet of OND.

Note: The introduction includes excerpts from the recent webinar on “Novel Technologies to Deliver Oral Oncology Therapies”. Click to view the webinar.

3

Powder Technology 258 (2014) 222–233

Contents lists available at ScienceDirect

Powder Technology

j ourna l homepage: www.e lsev ie r .com/ locate /powtec

A control strategy for bioavailability enhancement by size reduction:Effect of micronization conditions on the bulk, surface and blendingcharacteristics of an active pharmaceutical ingredient

Dolapo Olusanmi a,⁎, Dimuthu Jayawickrama a, Dongsheng Bu a, Gary McGeorge a, Helen Sailes c,Joanne Kelleher c, John F. Gamble b, Umang V. Shah d, Mike Tobyn b

a Bristol-Myers Squibb Pharmaceuticals, 1 Squibb Drive, New Brunswick, NJ 08903, USAb Bristol-Myers Squibb Pharmaceuticals, Reeds Lane, Moreton, Wirral CH46 1QW, UKc Bristol-Myers Squibb Pharmaceuticals, Watery Lane, Swords, Dublin, Irelandd Imperial College London, South Kensington, London SW7 2AZ, UK

⁎ Corresponding author. Tel.: +1 732 227 6131.E-mail address: dolapo.olusanmi@bms.com (D. Olusan

http://dx.doi.org/10.1016/j.powtec.2014.03.0320032-5910/© 2014 Elsevier B.V. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 13 September 2013Received in revised form 7 March 2014Accepted 8 March 2014Available online 14 March 2014

Keywords:MicronizationMillingQbDBlendingSurface energySurface areaCohesion

In a Quality by Design (QbD) development environment the effect of early process parameters on downstreammanufacturing parameters, and the ultimate effect on drug product quality, need to be understood. For poorlysoluble drugs, size reduction is frequently employed to obtain consistent in-vivo exposures. As a result,micronization is a key early stage processing step for many active pharmaceutical ingredients (APIs).This paper demonstrates the effect of varying micronization conditions on an API for which micronization isdeemed necessary to ensure consistent drug delivery after human administration. Material micronized to differ-ent extents are confirmed as different by surface area, surface energy, particle size analysis, bulk density and sur-face adhesion measurements.Thesematerial characteristics can be correlatedwith the outcomes from a key processing step, blending. The evo-lution of the blending process is followed using PAT techniques, so that an overall understanding of the relation-ship between particle properties and blend uniformity can be demonstrated.Execution of such a study during drug development can enable selection of the appropriate control strategy toensure production of API in the desired rangewhere consistent optimal bioavailability and downstream process-ability are achieved.

© 2014 Elsevier B.V. All rights reserved.

1. Introduction

In the pharmaceutical industry a significant percentage of marketeddrugs [1] and new chemical entities [2] are classified as BCS Class II,where bioavailability is controlled by the drug release from the productmatrix due to the inherent low solubility of the drug substance. In orderto improve the absorption of such drug substanceswhichmay befirst inline in terms of their therapeutic benefits pharmaceutical scientistsoften employ alternative approaches to designing the active pharma-ceutical ingredient (API) and also the finished product. These includebut are not limited to amorphous forms, salt formation of the API, co-crystal formation, spray dried intermediates, deposition on inorganicmesoporous materials and particle size control. The formation of saltsof ionisable drug entities, often weak salts [3], while increasing solubil-ity will also alter properties such as hygroscopicity, electrostatics, melt-ing point, polymorphism and mechanical properties; all of which canimpact chemical stability [4]. Additionally, not all drug entities are

mi).

ionisable and in fact developing salt forms of APIs is not beneficial inevery case [5]. In some cases crystal size modification can be employed.This can be achieved either through bottom-up (crystallization) or top-down (size reduction) approaches to obtain the desired dissolution rateenhancement.

Furthermore, for low dose formulations particle size control is need-ed to ensure content uniformity (CU), a regulated quality standard oforal solid dosage forms. A key aspect of the drug development processis crystallization, and a primary challenge of crystallization is associatedwith difficulties in controlling the size and shape distribution of crystals,especially when complex crystallization processes are involved [6,7],and with increasing complexity of molecules [8]. In cases where bio-availability of drug entities can be improved by size and shapemodifica-tion the ideal scenario would be to “dial a particle size and shapedistribution” using robust bottom-up approaches from crystallizationthat can be scaled up to commercial scale, while maintaining requiredpolymorphic form, crystallinity and stability. The use of supercriticalfluid technology to produce crystalline API of narrow size distributionin the sub-micron range has been developed and discussed [9,10], how-ever, instances where the process has been scaled up for commercial

4

223D. Olusanmi et al. / Powder Technology 258 (2014) 222–233

scale manufacture have not been found, therefore as yet it is not a com-mercially viable option.

Rohrs et al. [11] developed a method to estimate the particle sizelimits needed to ensure that content uniformity criteria for a solid dosageform are met based on median particle size, particle size distributionspread and dose. The method assumes uniform mixing of the blend,and a log-normal particle size distribution for the API, with key input fac-tors such as d[0.5], ratio of the d[0.9]/d[0.5], dose andAPI particle density.For a given dose themaximumvalue of d[0.5] that can be accommodatedwithout the risk of failing blend content uniformity is affected by the ratioof the d[0.9]/d[0.5]. As the latter increases, the d[0.5] value has to be re-duced in order to pass content uniformity. This is particularly applicableto low dose formulations where the presence of large particles wouldbe expected to have a greater impact on CU. Further reduction of thedose at a specified tablet weight would increase the risk of CU failure.

As a consequence of the aforementioned factors when particle sizereduction is required to enhance bioavailability and/or ensure contentuniformitymilling, particularlymicronization, is oftenwidely employed.Themechanical activation duringmilling may inadvertently cause vary-ing and uncontrolled degrees of disruption to the particle surfaces [9], aswell as exposure of non-habit surfaces due to plastic deformation andfracture. Therefore the principles of Quality by Design (QbD) should beapplied to this process.

The implementation of QbD, a systematic approach to drug develop-ment, in the pharmaceutical industry is advocated by the regulatoryagencies, and the principles are given in the ICH Q8 guideline [12].This practice involves determining the functional relationships thatlink material attributes and processing parameters to product ‘criticalquality attributes’ (CQAs). The use of Process Analytical Technology(PAT) during development enhances process understanding thus lead-ing to improved control of the manufacturing process [13]. In additionto drug product CQAs that impact the desired quality, safety and effica-cy, CQAs can additionally include those properties that affect down-stream processability [12]. A substantial effort is made to ensure theformer part because without it a product cannot make it onto the mar-ket, however, the latter is also important. The design space, i.e. the link-age between the process inputs and CQAs, should be determined foreach unit operation that forms part of the manufacturing process. Formicronization the optimal design space is achieving the powder proper-ties required for the biopharmaceutics requirements, while ensuringthe best possible downstream processability. Downstream processabil-ity can be defined in terms of ease of bulk powder handling, processingtime to ensure quality etc.; essentially all these can be summed up intoprocessing efficiency. Without establishing an optimal design space forthe micronization operation a wide range of materials can be producedwhich meet the specifications for particle size, e.g. d[0.9] b 40 μm, butwhich may differ widely in terms of processability. As discussed byOlusanmi et al. [14] the 3-dimensional long-range order associatedwith crystalline organic materials can result in asymmetrical propertiesof the various crystal facets which can consequently impact their me-chanical properties. The surface chemistry on each crystal facet mayvary as the various crystallographic planes dissect the unit cell (the fun-damental building block of crystals repeated in space to form the crystalstructure) at different planar orientations and angles. Thus milling mayresult in the exposure of differing chemical groups, and differing surfaceconcentration of those groups [15]. As a consequence, the surface ener-getics of a micronized sample may be significantly different to the orig-inal un-milled material.

The surface energetics of pharmaceutical powders are often charac-terized in terms of their dispersive surface energy, with a higher disper-sive surface energy often suggested to represent a more “reactive”surface [16]. The total surface free energy of a solid can be split into dis-persive (non polar) and specific (polar) components. The former arenon-specific interactions which are due to long-range London disper-sive forces, while the latter are specific short-range directional chemicalinteractions which involve charge redistribution and sharing [17].

Techniques used to characterize the surface energetics of pharma-ceutical powders include IGC (Inverse Gas Chromatography) [9,18–22], sessile drop contact angle methods on compacted powders,and contact angle methods on macroscopic single crystals [15,23,24].There are several drawbacks associated with the sessile drop contactangle methods and these have been discussed extensively by a numberof authors [25–27]. Furthermore dispersive surface energy measure-ments by liquid wetting angle techniques are difficult to implement re-producibly on free-flowing powders [28]. For sessile drop contact anglemeasurements on individual crystal faces large specially grown crystalsare required, but these give better indications of dispersive surface ener-gy anisotropy resulting from surface chemistry differences, rather thanan averaged value obtained frompowder samples [15,24]. Surface ener-gy values determined bymeasurements at infinite dilution, as describedby Thielmann [13], aremostly biased towards the highest energy sites ofthe powder sample [29] because the amount of probe injected onlycovers a small percentage, b0.1%, of the powder sample [17]. This hasthe benefit of distinguishing subtle differences in the surface properties,even in the case of nominally similar batches of the same material [9].

In some cases the distribution of active sitesmay bemore relevant tothepractical properties of amaterial than thehigh energy sites [16]. Thisis particularly pertinent in cases where the effect of milling on disper-sive surface energy is of interest. In the case of materials with cleavageplanes, following exposure of such faces during processes such as mill-ing or agitated drying, the properties of the cleavage plane may domi-nate the surface energetics of the resulting product. Heng et al. [18]observed an increase in the dispersive surface energy, measured byIGC, of paracetamol form I powder with milling, and this was attributedto an increase in the proportion of the (010) facet, the cleavage plane,which exhibited the highest dispersive surface energy compared tothe other facets as determined by sessile drop contact angle measure-ments on individual crystal faces of macroscopic crystals. Exposure ofnon-habit surfaces due to fracture may subsequently introduce furthercomplexities in the interaction of organic particulates with differentchemical entities compared to unmilled material. Overall, processeswhich involve size reductionmay further compound the dispersive sur-face energy heterogeneity of the product due to exposure of new sur-faces coupled with the possible disordering of crystalline surfacescaused by mechanical stress. Therefore when micronization is requiredit is essential that during development, the effect of micronization ex-tent within the acceptable range for the biopharmaceutical require-ments on powder properties (both surface and bulk) and downstreamprocessability is understood.

Work has been reported in literature on the effect of milling on thesurface properties of APIs and particulate materials in general. Whilesome authors have reported decreases in surface energy with milling[19,21,30], others have observed milling to increase the surface energyof powders [18,31]. Where surface energy increases withmilling extenthas been reported it has often been attributed to the exposure of higherenergy crystal facets and/or the creation of higher energetic surfaces onthe particles. Typically a number of these publications report changes tothe degree of crystallinity of the milled material either due to localizedlattice disorder and/or reduced crystallinity [20,30,32], changes to thesurface chemistry of milled material [18,31]. An important question,which has not been addressed, is what is the effect of micronizationand associated change in surface energetics on the bulk handling andprocessing behavior of pharmaceutical API during development andmanufacture? Publications where the effect of these milling-inducedchanges are demonstrated are limited e.g. Vippagunta et al. [33]. To en-able better implementation of quality by design andmore efficient drugdevelopment it is important to understand the impact of such changesto other downstream processing operations, and this subject would bethe focus of this present work.

The objective of this milling study is to establish a control strategyfor producing micronized API with consistent powder properties inthe desired target product profile for dissolution enhancement.

5

224 D. Olusanmi et al. / Powder Technology 258 (2014) 222–233

Furthermore the effect of milling extent of a commercially available ac-tive pharmaceutical ingredient, compound X, on its powder propertiesand subsequent downstream processing behavior is determined. Boththe surface and bulk properties of themilledmaterial are characterized,with the former conducted using B.E.T. surface area and dispersive sur-face energy heterogeneity analysis. Subsequently the effect of the deter-mined powder properties of the API batches during the blending of aformulation of low (b5% w/w) drug loading is assessed. Batches witha wide range of milling parameters and particle sizes were producedto determine a suitable design space that would meet all CQArequirements.

2. Materials and methods

2.1. Materials

The API used in this study is manufactured by Bristol–Myers Squibb(BMS). Excipients typically used in a dry granulation process were uti-lized in the study.

The probe liquids for contact angle measurements, ethylene glycol(N99%), dimethyl sulfoxide (N99%), and glycerol (N99%) were obtainedfrom Sigma-Aldrich. Formamide (N99.5%), and diiodomethane (N99%)were obtained from Acros Organics.

2.2. Milling

0.8 kg of a batch of the crystallized API, Batch A, was milled in 4separate lots of 200 g to different extents, runs 1 to 4, using a 4-inch jet mill. Different milling conditions were used with the aimof producing materials with distinguishable bulk and surface prop-erties to enable evaluation on the CQAs, and thus aid implementa-tion of the appropriate control strategy. An additional batch, BatchB, of the same API was milled to three extents as part of a millingcross trial analysis i.e. low pressure/high feed rate, high pressure/low feed rate and mid-point between the two, with no limits seton particle size distribution. In total 8 batches milled to different ex-tents were assessed in this study.

2.3. Content uniformity (CU) modeling

A proprietary implementation of the Rohrs [11] CUmodel was usedto assess the potential effect of particle size on CU for a low drug load,b5%, and low unit dose b10 mg.

2.4. Surface area

The surface area of the milled samples was obtained using a Gemini2390a surface area analyser supplied byMicromeritics, USA,with Nitro-gen as the adsorbate. About 0.5–1 g of sample was analyzed for surfacearea measurements. To ensure accurate surface area determination,sampleswere outgassed prior to analysis for 24h at 50 °C usingnitrogengas to remove residualmoisture adsorbed on the surface of the particleswhichmight affect the accuracy of the results. During analysis, the sam-ples were equilibrated for 10 s followed by an evacuation rate of66.6 kPa/min. Multi-point B.E.T. measurements were taken in therange 0.05–0.3 p/p0. Two samples were analyzed for each batch andthe average determined.

2.5. Surface energy analysis

The dispersive and specific surface energy of the unmilled andmilled samples was characterized using a Surface Energy Analyser(SEA), supplied by Surface Measurement Systems (Alperton, UK). TheSEA, a next generation Inverse Gas Chromatography (IGC) system, hasbeen used to successfully characterize the surface energy of pharmaceu-tical materials at finite dilution, and has been previously described in an

earlier publication [31]. An advantage that the SEA has over the tradi-tional IGC is that the former is designed to determine the surface en-ergy at targeted surface coverages; increasing the concentration ofprobe molecules will increase the number of lower energy (“lessactive”) sites that are involved in the interaction with the probemolecules as high energy sites are interacted with preferentially[16,34].

Prior to analysis each milled sample was filled into silanised glasscapillary columns of 3 mm inside diameter and 300 mm length by ver-tical tapping, with both ends of the columns stoppered with silanisedglass wool to prevent the movement of the sample bed during analysis.

The dispersive surface energy analysis method involved the injec-tion of a series of dispersive n-alkane probes, namely decane, nonane,octane and heptane, the volumes of which were controlled in order toobtain specific surface coverages in the range of 0.15 up to 20%. For spe-cific surface energy analysis twomono-polar probes, chloroform (Lewisacid) and ethyl acetate (Lewis base), were also injected at equivalentsurface coverages as used in the dispersive surface energy method. Sur-face adsorbedwater and other impuritieswere removed through condi-tioning of the columns using dry helium for 2 h at 30 °C/0% RH. For oneof the samples duplicate analysis of the same sample column was con-ducted to ensure that these degassing conditions were sufficient toclean the samples without impacting a physical change during analysis.Themeasured results remained unchanged for the duplicate analyses. Acolumn gas flow rate of 10 cm3/min was used for all samples, and thecolumnsweremaintained at 30 °C and 0% RH. To determine the columndead time, methane gas was injected at a volume of 0.208 cm3. This ac-counts for any differences in the packing behavior of the powder sam-ples due to their morphology.

The dispersive surface energy data were analyzed using the Dorrisand Gray [35] approach at peakmax. Voekel et al. [36] reported that dis-persive surface energy values calculated from the Dorris and Gray [35]method were comparable to those obtained using the Schultz [37]method.

For the Dorris and Gray approach used here the dispersive compo-nent of the surface energy is obtained from a plot of [RT ln VN] versusthe carbon number of the alkane probes, where R is the universal gasconstant (J/mol K), T is the temperature (K), and VN is the net retentionvolume, a fundamental surface thermodynamic property of the solid–vapor interaction, described in Eq. (1).

VN ¼ jm

� F � tR−t0ð Þ � T273:15

� �ð1Þ

where j is the James–Martin correction, m is the mass of the sample inthe column (g), F is the carrier gas flow rate (cm3/min), tR is the reten-tion time of the probe (minutes), t0 is the mobile phase hold-up time(minutes) and T is the column temperature (K). The dispersive surfaceenergy can be calculated using the peak max and the peak center ofmass, the latter of which accounts for non-normal peaks in the elutiondata. In this study tailing in the elution peaks was not observed there-fore the peak max approach was used.

Determination of the specific component of the surface energy wasconducted using the approach described by Dong et al., [38] from aplot of [RT ln VN] versus the molar deformation polarization of theprobes, PD, as described in Eq. (2).

PD ¼MW � r2−1

� �ρl � r2 þ 2

� �0@

1A ð2Þ

where MW, r, and ρl are the molar mass, reflective index, and liquiddensity, respectively, of the probe.

6

Table 1Properties of liquid probes used for sessile drop contact angle measurements.

Probe liquids Density (kg/m3) γLV (mJ/m2) γLd (mJ/m2) γL

p (mJ/m2)

Diiodomethane 3325 50.8 50.8 0.0Formamide 998 58.3 32.3 26Ethylene glycol 1109 48.0 29.0 19.0Dimethyl sulfoxide 1100 43.6 34.9 8.7

225D. Olusanmi et al. / Powder Technology 258 (2014) 222–233

The acid (Lewis acceptor), γs+ and base (Lewis donor), γs

− numbersof the solid based on the van Oss [39] concept were obtained from thespecific free energy values using Eq. (3).

−ΔG ¼ NA � a � 2 γ−l � γþ

s

� �1=2 þ γþl � γ−

s

� �1=2

� ð3Þ

where ΔG is the specific free energy of adsorption, NA is Avogadro'snumber, a is the cross-sectional area of the probemolecule, and γl

+ andγl− represent the electron acceptor and donor parameters, respectively,

of the probe molecule. The specific surface energy was calculated usingthe geometric mean of γs

+ and γs− as described in Eq. (4).

γSPS ¼ 2 � γ−

s � γþs

� �1=2 ð4Þ

where γsSP is the specific surface energy of the sample. The total surface

energy is the sum of the dispersive and specific surface energy values.

2.6. Particle size analysis

Particle size characterization was carried out using two techniques,laser light scattering and image based analysis.

2.6.1. Laser light scattering (LLS)Laser light scatteringwas carried out on aMalvernMastersizer 2000

(Malvern Instruments, Malvern, UK), equipped with a liquid dispersionsystem. The dispersion medium was an aqueous solution of Tween 80saturated with the API and filtered. For particle size measurements50 mg of API were dispersed into 5 ml of the dispersion media andvortexed for 15 s. Measurements were taken over 10 s, at obscurationlevels between 10 and 25%. Ultrasonication was applied during themeasurements at 25% level over 10 s.

2.6.2. Image based analysisImage based particle size analysis was conducted using a Malvern

Morphologi G3 particle characterisation system (Malvern Instruments,Malvern, UK). Samples were dry dispersed using the systems automat-ed sample dispersion unit onto a glass plate situated on an automatedsample stage. Particle imaging was conducted using a 20× magnifica-tion lens with vertical z-stacking enabled, taking an additional 5 planesabove the focal plane (equivalent to 27.2 μm in total), to account for 3-dimensionality within the samples. Morphological filtering of the rawdata was conducted, in order to remove partially imaged/overlappingparticles, on a sample by sample basis using a combination of convexity,solidity and particle width filters.

2.7. Bulk density and wall friction angle characterization

Bulk density was characterized using the FT4 powder rheometer(Freeman Technology, Worcestershire, UK). The API sample is loadedinto afixed volume vessel and “conditioned” before any analysis. Duringthe conditioning cycle the FT4 rheometer blade traverses downwardsinto the powder bed and then upwards at conditions which effect slic-ing rather than compacting of the powder bed. This conditioning cyclegently displaces the powder in order to loosen and slightly aerate it,and therefore makes the result independent of operator's method ofloading powder into the sample vessel or excess air [40]. Subsequentlyexcess powder is removedwith the aid of a vessel ‘split’ so that the pow-der is level with the top of the vessel which has a fixed volume of 25ml.Bulk density is determined as the mass of the powder in this 25 mmbore, 25 ml vessel. Duplicate analysis of each milled batch wasconducted.

The wall friction angle (WFA) of the powders was also evaluatedusing the FT4 Powder Rheometer and 304 stainless steel discs of

estimated 1.2 μm and 0.05 μm surface roughness as specified by themanufacturer. The WFA test is proposed to determine the frictional ad-hesion of powder samples towards a specific surface e.g. stainless steel.However the authors of this work are also interested to assess whetheror not the set-up can be used to determine differences in adhesion, ifany, due to increased “reactivity” of the powder sample duringparticle-surface interactions as a result of milling propensity. The testswere conducted in a 25 mm bore, 10 ml borosilicate glass vessel. Thisset up has the advantage that a smaller amount of material is requiredcompared to the bulk density test described above. However theamount of sample used is still greater than desired, particularly in theearly stages of drug development where only small amounts of APIare available. Following a conditioning cycle (as described above) andconsolidation of the powder bed using a vented piston a step sequenceof predetermined normal stress of 15 kPa was applied while the afore-mentioned wall friction disc was rotated and the frictional torque mea-sured over a range of stresses between 1 and 18 kPa. It is assumed thatonly the surface of the powder bed is sheared by the disc. The shearingstress is calculated from themeasured frictional torque and thewall co-efficient of friction is determined from the slope of the shear stress ver-sus normal stress data. This described test method was developed byFreeman Technology, and the wall friction coefficient is defined belowin Eq. (5);

Wall friction coefficient; μ ¼ tan Kð Þ ¼ shear stressnormal stress

ð5Þ

where Ø is the friction angle.The coupons were gently but thoroughly cleaned with acetone be-

fore each analysis, and all tests on the FT4 rheometer were carried outin environmentally controlled laboratory conditions.

2.8. Sessile drop contact angle measurements on stainless steel

A Krüss Drop Shape Analyser DSA 10 (Krüss GmbH, Hamburg,Germany) was used for the advancing contact angle measurements.Ethylene glycol, formamide, diiodomethane and dimethyl sulfoxidewere used as probe liquids, properties of which are reported inTable 1. Contact angle data using water could not be obtained due tothe liquid completely wetting the stainless steel surface such that anangle could not be measured. Measurements were obtained in openair at ambient conditions. Probe liquid droplet was monitored usingCCD camera and the shape of the droplet was fitted with a tangentmethod to obtain contact angle data using the Drop Shape Analysis soft-ware (DSA version 1.0) (KrüssGmbH, Hamburg, Germany). Aminimumof 4 droplets on 3 different stainless steel coupons were measured. Be-fore analysis with each probe liquid the coupons were placed in 50 mlof acetone and sonicated for 1 min at 30 °C. The contact angle datawas analyzed using the method reported by Owens and Wendt [41] todetermine surface energy.

2.9. Blending

To determine the effect of the milling on the processing behavior ofthe material blending of the milled batches was conducted.

7

(a) Two NIR blending profiles. A: fast and B: slow

(b) Ln(RSDe) vs revolutions from Figure 1(a) with linear fit.

0

20

40

60

80

100

120

0 100 200 300 400

NIR

pre

dic

ted

po

ten

cy

Rotations

A

B

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

0 20 40 60 80 100

Ln

RS

De

Revolutions

A

B

Linear (A)

Linear (B)

Fig. 1. (a): Two NIR blending profiles. A: fast and B: slow. (b): Ln(RSDe) vs revolutionsfrom Fig. 1 (a) with linear fit.

Table 2Particle size distribution of batches obtained from jet-milling process and milling param-eters.

Batch number Milling parameters Particle size (μm)

Feedrate (Kg/h) Pressure (Bar) d[0.1] d[0.5] d[0.9]

Batch-A Run 1 9.6 1.5 2.17 20.5 83.5Batch-A Run 2 9.6 2.0 1.94 15.9 59.9Batch-A Run 3 8.4 3.0 1.29 7.92 33.6Batch-A Run 4 7.2 3.0 1.02 5.40 19.6Batch-B Unmilled 51.0 127.0 350.4Batch-B Cross trial 1 7.2 5.0 1.13 6.31 19.9Batch-B Cross trial 2 3.6 9.0 0.69 2.43 5.97Batch-B Cross trial 3 5.5 7.0 1.09 4.50 11.7

226 D. Olusanmi et al. / Powder Technology 258 (2014) 222–233

2.9.1. Sample preparationThe API and excipients were charged into the bin prior to the blend

operation in an order such that the API was sandwiched in the powderbed. The batch size for each run was ~1.4 kg.

2.9.2. Blending and NIR monitoringA 6 L size tote bin (A&M process equipment Ltd, Ontario,

Canada) was used for blending using a portable laboratory tumbleblender (Model ATS050LP, A&M process equipment Ltd, Ontario,Canada). A 3-inch bin lid was custom modified with sapphire windowto house a non contact online NIR instrument. The NIR spectra of blend-ing powder were acquired with a Brimrose 5030 near infrared spectro-photometer (Brimrose Inc., Sparks, MD). The blender was operated at15 RPM for all the blending steps. The total number of revolutions forblending was set to 400. At each revolution NIR spectra were acquiredwhen the NIR is closest to the ground (i.e. at the bottom of the bin).This position ensured complete coverage of the sapphire window onthe bin lid by the powder. The general acquisition parameters were:wavelength range 1100–2300 nm, number of scans 15, and resolution2 nm. All the data were acquired in reflection mode. A custom-madesoftware based on MatLab (Matlab Inc., Natick, MA, USA) andPLS_Toolbox (Eigenvector Research Inc., Wenatchee,WA, USA) was de-veloped in house to determine real time API potency during blending.

2.9.3. Calibration modelAmultivariate calibration model based on Partial Least Square (PLS)

was developed and validated to quantitatively determine the amountAPI in the blend. The calibration model was developed using offlinesamples encompassing the API potency range. The gravimetric amountpresent in the calibration standard was used as the reference value forcalibration. Absorbance spectrawere first pretreatedwith standard nor-mal variate (SNV), followed by second derivative (Svaitzky_Golay, 11smoothing and 2nd order polynomial). API peak region of the secondderivative spectra were used for the calibration model. A calibrationmodel based on Partial Least Square (PLS) was developed with one fac-tor. The cross validation error for the model was ~3%.

2.9.4. Quantitative measurement of blending profilesIn this study it was necessary to develop a quantitative parameter in

order to compare different blending profiles. The kinetics and homoge-neity of blending can be described using the relative standard deviation(RSD) of the content of a component in a mixture. A novel approach toquantify blending profiles has been described by Shi et.al. [42] using aterm called the RSDE as defined in Eq. (6) and its relationship to firstorder mixing kinetics.

RSDE ¼

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiXxi−Eð Þ2n−1ð ÞE

vuuut0BBBB@

1CCCCA ð6Þ

− dRSDE

dr¼ kRSDEð Þ ð7Þ

−LnRSDE ¼ krþ cð Þ ð8Þ

where xi is an individual spectrums determined concentration for onerevolution, E is the expected target mean concentration and n is thenumber of measurements. Consequently, the RSDE term is a measureof the relative deviation for the target composition, as opposed to thetraditional RSD metric that looks at the difference from the average ofthemetric. A linear relationship (Eq. (8)) betweenRSDE and thenumber

of revolutions (r) can be derived assuming a first order kinetics(Eq. (7)). Eq. (8) enables one to determine a rate, or blending, coeffi-cient (k) to quantify the blending process. As an example Fig. 1(a) illus-trates two blending profiles, one fast (A) and another slow (B)generated with online NIR and multivariate calibration, and the deter-mined LnRSDE profiles are shown in Fig. 1(b). These plots show two dis-tinct linear relationships between LnRSDE and revolutions characterizedby their respective rate constants (slope of the linear regression plot).The use of quantitative rate constants allows one to compare blendingprofiles using a single parameter, hence minimizing the subjectivity ofthe analysis.

8

Fig. 2. Geometric (volume based) particle size distribution profiles of milled batches, measured using Malvern Morphologi G3.

227D. Olusanmi et al. / Powder Technology 258 (2014) 222–233

3. Results and discussion

3.1. Particle size distribution and content uniformity

The Malven LLS particle size distributions attained for the milledsamples and the mill settings used to achieve them are shown inTable 2.

The attained d[0.9] and d[0.5] values cover a wide range; an overlayof the Morphologi G3 determined PSD profiles of all the batches shownin Fig. 2. It is noted thatwhile the general trends are similar in both casesthe median sizes obtained from the Morphologi are generally higherthan those obtained from the wet dispersed LLS method, possibly dueto differences in the lower particle size resolution range of the two sys-tems and different sample preparation approaches. From Fig. 2 it can beseen that with increased milling extent the population of coarse parti-cles gradually decreases while the population of fine particles increases,with some batches having skewed distributions.

A graph, based on the method developed by Rohrs et al. [11], thatgives an estimate of the effect of d[0.9]/d[0.5] ratio and d[0.5] on content

Fig. 3. Graphical estimation of effect of particl

uniformity for the relevant drug load and unit dose is shown in Fig. 3.For b5% drug loading and b10 mg unit dose, application of the method,which assumes uniform blending, indicates that provided that the d[0.9]/d[0.5] ratio is not greater than 2, d[0.5] values up to 90 μm are ac-ceptable to ensure acceptable content uniformity. Further decreasingthe dose would of course increase the chances of failing CU with theseparticle size specifications. The actual obtained d[0.9]/d[0.5] values arein the range of 2.5 to 4. This ratio is greater than the desired target fora narrow size distribution, according to the method of Rohrs et al.[11], however, all the batches are predicted to pass tablet content uni-formity since the d[0.5] values are considerably smaller than 90 μm.As expected by increasing the d[0.9]/d[0.5] ratio from 2 to 4 there is asignificant reduction in the predicted maximum acceptable d[0.9]from 90 to 25 μm.

An analysis of the mill settings against the PSD and surface areavalues showed that for themill used the feed rate provides the best cor-relation to the surface area of the milled product, with an R2 value of0.94, compared to R2 of 0.77 for mill pressure. Lower feed rate giveshigher surface area, while the reverse is the case for mill pressure.

e size distribution on content uniformity.

9

Table 3Summary of surface energy results, arranged in order of surface area.

Batch number Average B.E.T. SSA m2/g n = 2 Dispersive surface energy at 5%surface coverage mJ/m2

Free energy of absorption at5% surface coveragekJ/mol

Specific surface energy 5%surface coverage mJ/m2

Total surface energy 5%surface coverage mJ/m2

Ethyl acetate Chloroform

Batch-B-unmilled 0.117 35.1 6.52 7.74 53.6 88.6Batch-A Run 2 0.954 43.9 7.03 8.90 68.6 112.5Batch-A Run 1 1.08 48.1 7.24 9.22 72.8 120.9Batch-A Run 3 1.89 50.2 7.37 9.33 76.5 126.7Batch-B Cross trial 1 2.34 48.6 7.39 9.20 75.2 123.7Batch-B Cross trial 1 2.34 48.7 7.27 9.36 75.2 123.8Batch-A Run 4 3.42 53.2 7.54 9.58 79.3 132.5Batch-B Cross trial 3 3.98 48.8 7.25 9.29 74.7 123.6Batch-B Cross trial 2 5.00 50.8 7.54 9.51 78.3 129.2

228 D. Olusanmi et al. / Powder Technology 258 (2014) 222–233

3.2. Surface area and surface energy

The surface area and surface energy of all the samples are shown inTable 3. The surface area results are the mean of two measurements,with % RSDgenerally 3%or less. Surface area provides a good descriptionof powder samples compared to d[0.9] alone, as the former takes intoaccount surface contributions from all the particles, both coarse andfine, present in the sample.

For the surface energy measurements, one of the samples, Batch-BCross trial 1 highlighted in bold in Table 3, was analyzed in duplicateusing two separate samples to determine reproducibility. As the repro-ducibility was observed to be acceptable, and with the analysis time forthemeasurement up to three days per sample, replicate analysis for theremaining samples was not conducted.

The reproducibility of the two samples analyzed from Batch-B Crosstrial 1 iswithin 1mJ/m2 across themeasured and calculated parameters.Therefore for comparison between the batches any differences in thedispersive surface energy values above 1 mJ/m2 can be said to be real.

Cross trial samples 1–3 were milled using Batch-B, while runs 1 to 4were from Batch-A. When these sets of samples are grouped accordingto batch history the dispersive surface energies at 5% coverage are ob-served to increase with increasing surface area and milling extent. Theincrease in specific surface energy of the batches withmilling is concur-rent with that of the dispersive surface energy (shown in Appendix A,Figs. 1A & 2A, respectively). A combination of these two factors leadsto an overall increase in the total surface energy with milling extent,as shown in Fig. 4. However, despite Batch-B Cross trial 2 having thehighest milling extent and surface area of all the batches, its dispersivesurface energy is lower than Batch-A Run 4 which has a lower millingextent and surface area.

Comparison of the surface energy heterogeneity profiles of Batch-BCross trial 2 and Batch-A Run 4, shown in Fig. 5, indicates that Batch-ARun 4 has a higher dispersive surface energy at closer to 0% surface

80

90

100

110

120

130

140

0.00 1.00 2.00 3.00 4.00 5.00 6.00

Tota

l Su

rfac

e E

ner

gy

5 %

cov

erag

e m

J/m

2

B.E.T. SSA (m2/g)

Runs 1-4

Cross-trials

Fig. 4. Variation of total surface energy of milled batches as a function of surface area.

coverage, which represents the surface energy of the “highest” energysites. The reason for this observation is unclear andmay be partly attrib-uted to differences in batch history; unfortunately, a sample of the un-milled API thatwas used to obtain Batch-Awas not available for analysisin order to confirm or refute this hypothesis.

It is of interest to understand the reason for the observed surface en-ergy increase with milling extent. Milling extent clearly and expectedlyincreases the surface area of the product. Comparison of the surface en-ergy heterogeneity profiles of runs 1 to 4 regressed to ~0% surface cov-erage (Fig. 6), i.e. similar to infinite dilution conditions where highsensitivity to “higher energy” surfaces is expected, shows that not onlyare the surface energy values at 0% greater than at 5%, but also thatthe extent of increase from runs 1 to 4 across the heterogeneity profilesare similar. Therefore the data suggests that milling has created higherenergetic sites. It is possible that the process of milling, during impart-ment ofmechanical energy, could lead to the creation of more energeticsurfaces; suchmechanical energymay result in the creation of localizedlattice disorder on the surface of the particles in the sample. In fact thisphenomenon has been reported in literature. With increased millingenergy the extent of creation of localized lattice disordermight increase.

In such a case, “higher” energetic sites would be created thus leadingto higher measured surface energy values at ~0% surface coverage, es-sentially infinite dilution region. This hypothesis was explored by pow-der x-ray diffraction (PXRD) analysis of some of the milled batches todetermine the presence of peak broadening, which can be caused by lat-tice disorder. It is important to note that PXRD analysis of the sampleswas conducted months after SEA analyses. Aging time post-millinghas been observed to affect the surface energetics and flowability ofMetformin HCl due to re-ordering of mechanically induced surface de-fects [33]. Nevertheless no significant differences in peak widths, norpresence of amorphous material, were observed for the batches.

45

47

49

51

53

55

57

59

61

0 0.02 0.04 0.06 0.08 0.1Dis

per

sive

su

rfac

e en

erg

y m

J/m

2

Surface coverage n/nm

Batch-B Cross-trial 2

Batch-A Run 4

Fig. 5. Surface energy comparison for batches Batch-B Cross trial 2 and Batch-A Run 4. Theprofiles of both samples are well described by an exponential fit with R2 values of greaterthan 0.99 in both cases.

10

35

40

45

50

55

60

65

0 0.02 0.04 0.06 0.08 0.1

Su

rfac

e en

erg

y (m

J/m

2 )

Surface coverage (n/nm)

Batch A run 1

Batch A run 2

Batch A run 3

Batch A run 4

Fig. 6. Surface energy heterogeneity profiles of runs 1 to 4.

Table 4Raw data from advancing contact angle measurements on stainless steel coupons.

Stainless steel coupon 1

Probe 1 2 3 4 Average Stdev

Ethylene glycol 47.0 48.6 44.8 48.9 47.3 1.88Formamide 47.6 48.3 48.7 51.5 49.0 1.71Diiodomethane 42.2 43.9 44.8 43.8 43.7 1.08Dimethyl sulfoxide 45.1 42.5 45.2 44.7 44.4 1.27R2 for Owens–Wendt [41] plot = 0.75Calculated surface energy values: dispersive = 34.0 mJ/m2, specific = 5.1 mJ/m2

Stainless steel coupon 2

Probe 1 2 3 4 Average Stdev

Ethylene glycol 47.3 45.5 48.5 45.1 46.6 1.59Formamide 54.5 56.1 55.2 52.4 54.6 1.58Diiodomethane 45.6 43.6 42.6 42.8 43.7 1.37Dimethyl sulfoxide 42.3 41.9 43.3 41.3 42.2 0.84R2 for Owens–Wendt [41] plot = 0.75Calculated surface energy values: dispersive = 33.5 mJ/m2, specific = 4.4 mJ/m2

Stainless steel coupon 3

Probe 1 2 3 4 Average Stdev

Ethylene glycol 50.0 50.1 48.3 42.4 47.7 3.63Formamide 49.4 53.5 53.6 52.1 52.2 1.96Diiodomethane 44.5 45.1 43.0 42.1 43.7 1.37Dimethyl sulfoxide 42.2 46.7 41.6 43.3 43.5 2.28R2 for Owens–Wendt [41] plot = 0.80Calculated surface energy values: dispersive = 35.2 mJ/m2, specific = 4.1 mJ/m2

50

100

150

200

250

300

0.00 1.00 2.00 3.00 4.00 5.00 6.00

Wo

rk o

f co

hes

ion

an

d a

dh

esio

n (

mJ)

Surface area (m2/g)

Work of cohesion

Work of adhesion

Fig. 7. Variation of work of adhesion and cohesion with surface area.

229D. Olusanmi et al. / Powder Technology 258 (2014) 222–233

Therefore the available data suggests that although milling creates“higher energetic” sites (based on surface energy data) these cannotbe attributed to lattice disorder. The consequences of the formation ofthese higher energetic sites by milling are explored further in the fol-lowing section.

Using the surface energy values reported here the work of cohesionand adhesion (mJ/m2), were calculated using Eqs. (9) and (10), respec-tively.

WCohesion ¼ 2ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiγDA � γD

A

� �qþ 2

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiγSPA � γSP

A

� �qð9Þ

WAdhesion ¼ 2ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiγDA � γD

SS

� �qþ 2

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiγSPA � γSP

SS

� �qð10Þ

In Eqs. (9) and (10), the subscripts A and SS refer to API and stainlesssteel, respectively, while the superscripts D and SP refer to dispersiveand specific surface energy values, respectively.

Work of adhesion towards stainless steel was estimated usingdispersive and specific surface energy values of 34.2 ± 0.9 mJ/m2

and 4.6 ± 0.5 mJ/m2, respectively, determined from advancingcontact angle measurements on stainless steel coupons described inSection 2.8. The raw data for the contact angle measurements and theR2 values of the Owens–Wendt plots are provided in Table 4. The resultsgenerally show good reproducibility for four measurements on eachstainless coupon, and across the three coupons. The slightly higher stan-dard deviation for SS coupon 3 and the generally low R2 values of 0.75–0.80 for themeasurements could be due to residual organic surface con-tamination. Mantel and Wightman [43] observed that the measuredsurface energy values, particularly the polar surface energy, of stainlesssteel surfaces determined by contact anglemeasurements using organicprobe molecules are highly dependent on the pre-measurementcleaning treatment ranging fromwater to O2 plasma cleaning.Measure-ments obtained from surfaces cleaned with water and acetone gave thelowest values,while those thatwereO2 plasma cleaned gave thehighestvalues of both dispersive and polar surface energy.

The coupon cleaning treatment used in this study is sonication in ac-etone (described in Section 2.8), therefore it is expected that there areresidual organic contaminants that affect the accuracy of the results.Nevertheless routine cleaning of tablet press tooling and equipmentsused for manufacture is generally done using organic solvents, for ex-ample IPA or acetone, sometimes with sonication for tablet presstoolings. Therefore while the dispersive and specific surface energyvalues reported here may not be accurate for a contaminant-free stain-less steel surface they represent a realistic state of stainless steel tooling/equipment surfaces that the API would come into contact with duringroutine drug product manufacture. Furthermore the stainless steel cou-pons used for the wall friction angle measurements on the FT4 rheom-eter, described in Section 2.7, were also cleaned with acetone prior to

each use therefore the comparisons made between the theoreticalwork of adhesion to stainless steel from surface energy measurementsand the adhesion experiments on the FT4 are valid and adequate forthe purpose of this manuscript.

The variation of work of adhesion and cohesion with surface area isshown in Fig. 7. While work of adhesion does not change significantly,the work of cohesion shows a more marked increase with milling ex-tent. The latter effect is contributed to by the fact that both the specificand dispersive surface energies of the samples increase with millingextent.

3.3. FT4 rheometer characterization

The first set of wall friction angle (WFA) valueswas obtained using astainless steel coupon of 1.20 μm surface roughness. TheWFA°, which isdetermined from the slope of the shear stress versus applied normalstress as shown in Fig. 8 for Batch-A Run 3, measures the angle of wallfriction between the powder surface and the rotating disc used. Repro-ducibility analysis of the test using Batch-A Run 3 showed values to bereproducible to within ±0.3°. From the values shown in Table 5 it canbe observed that for the 1.2 μm surface roughness disc, hereafter

11

0.0

2.0

4.0

6.0

8.0

10.0

12.0

14.0

0.0 5.0 10.0 15.0 20.0

Mea

sure

d s

hea

r st

ress

, kP

a

Applied normal stress, kPa

Batch-A Run 31.20 microncoupon run 1

Batch-A Run 31.20 microncoupon run 2

Fig. 8. Slope of shear vs. applied normal stress. Reproducibility of WFA at 1.2 μm usingbatch-A Run 3.

25

27

29

31

33

35

37

39

41

80 90 100 110 120 130 140Wal

l Fri

ctio

n A

ng

le °

(1.2

SR

C)

Total surface energy (mJ/m2)

Runs 1 - 4

Cross-trials

Fig. 9. Correlation between WFA° and total surface energy at 5% coverage which shows alinear incremental relationship.

230 D. Olusanmi et al. / Powder Technology 258 (2014) 222–233

referred to as 1.2SRC, the adhesion against the stainless steel couponsurface increases with milling extent. The differences in the numericalvalues are small but greater than the ±0.3° expected variability basedon the reproducibility analysis.

Possible reasons for the increase inWFA with milling extent include(1) greater surface energy of the powder particles with milling extentcausing them to be more “reactive” towards the steel surface and/or(2) increased surface area with milling extent causing a larger contactarea for “binding”, particularly in the grooves on the surface of the cou-pon. This greater contact area would have the effect of increasing fric-tion between the powder particles and the metal surface. Todetermine the contribution of increased “reactivity” of powder samplestowards stainless steel due to increased surface energywithmilling, thewall friction tests were conducted on fresh samples of selected batchesusing a coupon of 0.05 μm surface roughness (0.05SRC). The 0.05SRChas amuch smoother surface finish, and therefore is expected to exhibitless frictional adhesion to the powder particles. In this case the differ-ences in the WFA of the batches are much less pronounced, and thetrend observed for 1.2 μm surface roughness no longer holds. This ob-served surface roughness effect is in agreementwith thework of Bunkeret al. [44] who found that surface roughness played a significant role inthe observed adhesion between lactose particles and punch surfaces;for the same lactose batch punches with rougher surfaces exhibitedgreater adhesion during compression. In fact the batches tested withthe 0.05SRC here exhibit similar WFA° under these conditions ~35 ±0.6°. Therefore option 1 of the proposed mechanisms above can beruled out. If the increase in adhesion of the samples is due to increased“reactivity” onewould expect this effect to hold regardless of the surfaceroughness of the coupon used. This corroborates the SEA analysis in theprevious section where it was deduced that for this material althoughincreased milling extent creates higher energetic sites it does not leadto a proportional increase in the work of adhesion towards the stainlesssteel coupons used.

Interestingly there is a linear relationship between theWFA° valuesat 1.2SRC and the total surface energy values measured by SEA (Fig. 9).

Table 5FT4 measured data.

Batch number B.E.T. surface area (m2/g)(average of n = 2)

Batch-A Run 2 0.95Batch-A Run 1 1.08Batch-A Run 3 1.89Batch-A Run 4 3.42Batch-B (unmilled) 0.117Batch-B Cross trial 1 2.34Batch-B Cross trial 2 5.00Batch-B Cross trial 3 3.98

The wall friction values increase with surface energy and this trend isgrouped according to batch history. The observed correlation between1.2SRC and surface energy may be solely attributed to a surface area ef-fect; the frictional adhesion increases with surface area due to increas-ing contact area between powder bed and steel surface, while thesurface energy also increaseswith surface area as a consequence ofmill-ing extent for reasons already discussed.

Another outcome of the increased milling extent is a reduction in thebulk density of the batches. The reduction in bulk density is proportionalto surface area increaseswithmilling extent, irrespective of the batch his-tory.With reducing particlemass it is expected that the impact of electro-static forces will increase in competition with gravity forces. Particulatematerials with small particle size and low bulk density provide ideal con-ditions for tribo-charging [45]; the electrification of particulate materialsin pharmaceutical systemswhich can arise due to sliding, rolling and im-pact during processing and handling operations [46]. Such charging,which has been shown to affect adhesion and agglomeration of particu-late materials [47], can potentially have detrimental effects on productquality, and result in loss of powder through deposition, segregationwithin mixed formulations etc. [45]. Furthermore bulk density is one ofthe key factors which affect the flowability of bulk solids [48].

3.4. Effect of milling propensity and powder properties on blending

In this section the effect of milling degree and resulting powderproperties on the blending unit operation is described. This portion ofthe study was carried out about ~4 months after milling these batchestherefore the API batches would have undergone a significant restingtime. This reflects a realistic manufacturing scenario where milledbatches may undergo resting times of weeks, and sometimes months,during transport and storage before further processing. Fig. 10 illus-trates blendingprofiles and the corresponding LnRSDE for some selectedblends. These profiles include blendsmanufactured with API of variablesize (d[0.9] PSD of 83.49 μm, 33.6 μmand 5.97 μm). Selecting these sizesof particles provides the ability to evaluate themixing behavior of large,

Bulk density g/ml(average of n = 2)

Wall friction angle (°) using a SS coupon ofestimated 1.2 μm surface roughness

0.462 31.60.438 32.30.367 35.50.281 37.80.546 26.00.293 36.40.240 38.80.273 37.7

12

0.5

1.5

2.5

3.5

4.5

-150

-100

-50

0

50

100

150

0 50 100 150 200 250 300 350 400L

nR

SD

E

AP

I Po

ten

cy in

ble

nd

Number of rotations

D90 83.49

D90 33.6

D90 5.97

D90 5.97 LnRSDE

D90 33.6 LnRSDE

D90 83.49 LnRSDE

* ****

Fig. 10. Online NIR blending profiles of pre-blends.

R² = 0.87

-0.030

-0.025

-0.020

-0.015

-0.010

-0.005

0.000

0.0 1.0 2.0 3.0 4.0 5.0 6.0

Slo

pe

of

Ln

RS

DE

(ble

nd

ing

kin

etic

s)

B.E.T. Surface area (m2/g)

Fig. 11. Blending rate as a function of surface area.

231D. Olusanmi et al. / Powder Technology 258 (2014) 222–233

medium and small sized API in a low drug load formulation. The profileswith large API particles (d[0.9] values of 83.49 μm and 33.6 μm) exhibitrelatively faster kinetics to reach blend uniformity. For both profiles themacro mixing (convection type) is completed by ~100 revolutions anddiffusion type mixing is completed by ~175 revolutions. In comparisonto this, the blend manufactured with API with a d[0.9] of 5.97 μm dem-onstrated completion ofmacromixingwithin ~75 revolutions but diffu-sion type of mixing is still not achieved after 400 revolutions. It seemsthis slow but steady micro mixing (i.e. diffusion type) is happeningthroughout the blending time. A similar behavior can easily be visual-ized with corresponding LnRSDE plots. The slopes of the LnRSDE vs. rev-olutions are clearly different for the three batches as shown in thegraph. The largest and smallest slopes correspond to the largest particlesize (d[0.9] values of 83.49 μm) and smallest particle size (d[0.9] valuesof 5.97 μm) blends respectively.

Notably, the blending profile of 5.97 μmblend has high intensity APIpeaks as denoted by asterisks. The presence of a significant number of“API concentration spikes” is the result of areas of concentrated APIpassing the window and being sampled during the measurement. Thespikes occur either because of the presence of agglomerates or aggre-gates. When an agglomerate appears it carries more weight in contrib-uting to the spectrum and results in intense peaks from time to timeduring the blending profile. In general it is not surprising to observeAPI concentration spikes at the beginning of blending profiles, however,the continuous presence of such concentration spikes throughout theblending profile is indicative of presence of aggregates/agglomeratesthat are difficult to break with the shear applied during tumble blend-ing. The observation of larger particles exhibiting faster blending kinet-ics than smaller particles of the same chemical entity is in corroborationwith that of Bellamy et al. [49].

The theoretical work of cohesion from surface energy values,displayed in Fig. 7, shows a general increase with surface area. Howevera closer look at thedata shows that the sampleswith d[0.9] of 33 μmand5.97 μm have similar theoretical cohesion values despite the consider-able differences in surface area, 1.89 m2/g and 5 m2/g, respectively.The work of cohesion data is contrary to the clear differences in theblending kinetics of these two batches that have been elucidated inthis current section.

Collectively the data presented in this study indicates that surfaceenergeticsmay not be themain factors governing the blending of partic-ulatematerials evaluated here. The surface area of the d[0.9]= 5.97 μmbatch is over 2.5 times that of the d[0.9] = 33.6 μm batch, and the bulkdensity of the former is 50% lower. These differences in surface areaappear to have a greater impact on the cohesivity and “blendability”observed here. Fig. 11 shows a linear correlation between the surfacearea of the milled batches and rate constant obtained from the LnRSDE

data obtained using the procedure described in Fig. 1. As shown in thisstudy the API batch smallest particle size distribution also has thehighest surface area. From the PSD data shown in Table 2, large particleswere not present in batch B cross trial 2 immediately post-milling. For-mation of agglomerates, particularly of small particles, can potentiallyoccur during periods of storage and shipping, and exacerbated in thepresence of compaction forces. During blending it is expected that themacroscopic bulk lumps break up during convective mixing, but as the“chunks” break up into smaller elements, it is possible that the areasof the intrinsic powder that have higher cohesivity will stay intact forlonger. Increased electrostatic effect that is expected for lower bulk den-sity materials (not measured in this study) may also be a contributingfactor. The concentration spikes observed during blending are likelydue to the inability of the process to quickly and completely break upcohesive agglomerates within the blend as a consequence of the smallparticle size and associated cohesivity of this material. As shown inthis example the presence of such small API particles clearly results ina slower blending profile, even accounting for the presence of the APIspikes. It is expected that small particles will take longer to blend thanthe samemass of large particles because numerically it is difficult to dis-perse large number of smaller particles compared to larger particles. Inadditionwhen small particles are present as agglomerates it takes addi-tional shear energy to break them down and as a result these API attri-butes can lead to incomplete blend with reduced blend homogeneity.Fig. 11, which shows that blending time to reach uniformity increaseswith surface area of the API, clearly demonstrates the influence of pow-der properties on the manufacturing process. The appropriatemicronization control strategy is one that gives optimal bioavailabilityenhancement without detrimental effects to downstream processingefficiency.

4. Conclusions

The effect of the degree of milling on the surface properties of an APIhas been demonstrated. Furthermore the consequence of such millinginduced changes has been evaluated in a unit operation that is com-monly utilized during pharmaceutical drug manufacture.

Increased milling extent causes changes to the surface as well asbulk properties of thismaterial, and this has a potential, if notmitigated,to lead to a direct negative impact on handling and downstreamprocessing.

In a QbDmanufacturing paradigm the effect of bulk properties on allaspects of the manufacturing process needs to be understood. Here it isshown that milling parameters produce materials with different bulkand surface properties, and that these parameters can influence a keyunit process, blending during drug product manufacture. Such work,carried out on a small scale using appropriate tools, can provide aguide for setting lower limit of particle size specifications. Specificationson the upper limit of particle size is obviously important forbiopharmaceutics reasons, while lower limit specifications would en-sure that powder handling and downstream processing are not mademore challenging than necessary.

13

232 D. Olusanmi et al. / Powder Technology 258 (2014) 222–233

Akey learning point from this study is thatwhenmilling is necessaryit is important to conduct it appropriately in order to utilize the benefi-cial effects for biopharmaceutics requirements while minimizing thedeleterious ones. Milling, one unit operation in a series of drug productprocessing steps, beyond what is required can lead to additional chal-lenges during bulk handling and even downstream processing.

Acknowledgments

The authors would like to thank Dr Denette Murphy (Bristol–MyersSquibb U.S.A.) for PXRD analyses, and Drs Jatin Patel and Omar Sprockel(both Bristol–Myers Squibb U.S.A.) for their review and feedback on thepublication.

Appendix A

50

55

60

65

70

75

80

85

0 1 2 3 4 5 6

Sp

ecif

ic S

urf

ace

En

erg

y5

% c

over

age

mJ/

m2

B.E.T. SSA (m2/g)

Runs 1-4

Cross-trials

Fig. 1A. Variation of specific surface energy at 5% coverage with surface area.

30

35

40

45

50

55

0 1 2 3 4 5 6

Runs 1-4

Cross-trials

Dis

per

sive

Su

rfac

e E

ner

gy

5 %

cov

erag

e m

J/m

2

B.E.T. SSA (m2/g)

Fig. 2A. Variation of dispersive surface energy at 5% coverage with surface area.

References

[1] B. Griffin, Advances in lipid-based formulations: overcoming the challenges of lowbioavailability for poorly water soluble drug compounds, Am. Pharm. Rev. (March2012) 41–47.

[2] C. Giliyar, D.T. Fikstad, S. Tyavanagimatt, Challenges and opportunities in oral deliv-ery of poorly water-soluble drugs, Drug Deliv. Technol. 6 (2006) 57–63.

[3] M.B. Maurin, D.J.W. Grant, P.H. Stahl, The physicochemical background: fundamen-tals of ionic equilibria, in: P.H. Stahl, C.G. Wermuth (Eds.), Handbook of Pharmaceu-tical Salts: Properties, Selection and Use, Wiley-VCH, New York, 2002, p. 12.

[4] J.M. Merritt, S.K. Viswanath, G.A. Stephenson, Implementing quality by design inpharmaceutical salt selection: a modeling approach to understanding dispropor-tionation, Pharm. Res. 30 (2013) 203–217.

[5] G.A. Stephenson, A. Aburub, T.A. Woods, Physical stability of salts of weak bases inthe solid-state, J. Pharm. Sci. 100 (2011) 1607–1617.

[6] R.D. Braatz, Advanced control of crystallization processes, Annu. Rev. Control. 26(2002) 87–89.

[7] N. Rodriguez-Hornedo, D. Murphy, Significance of controlling crystallization mech-anisms and kinetics in pharmaceutical systems, J. Pharm. Sci. 88 (1999) 651–660.

[8] S. Desikan, S.R. Anderson, P.A. Meenan, P.H. Toma, Crystallisation challenges in drugdevelopment: scale-up from laboratory to pilot plant and beyond, Curr. Opin. DrugDiscov. Dev. 3 (2000) 723–733.

[9] J.C. Feeley, P. York, B.S. Sumby, H. Dicks, Determination of surface properties andflow characteristics of salbutamol sulphate, before and after micronization, Int. J.Pharm. 172 (1998) 89–96.

[10] B.Y. Shekunov, P. York, Crystallisation processes in pharmaceutical technology anddrug delivery design, J. Cryst. Growth 211 (2000) 122–136.

[11] B.R. Rohrs, G.E. Amidon, R.H. Meury, P.J. Secrest, H.M. King, C.J. Skoug, Particle sizelimits to meet USP content uniformity criteria for tablets and capsules, J. Pharm.Sci. 95 (2006) 1049–1059.

[12] International conference on harmonization of technical requirements for registra-tion of pharmaceuticals for human use, Q8 (R1), Pharmaceutical development revi-sion 1.

[13] R.W. Bondi Jr., J.K. Drennen III, Quality by design and the importance of PAT in QbD,Sep. Sci. Technol. 10 (2001) 195–224.

[14] D. Olusanmi, K. Roberts, M. Ghadiri, Y. Ding, The breakage behaviour of Aspirinunder quasi-static and single particle impact loading: effect of crystallographic an-isotropy, Int. J. Pharm. 411 (2011) 49–63.

[15] J.Y.Y. Heng, A. Bismarck, A. Lee, K. Wilson, D.R. Williams, Anisotropic surface chem-istry of aspirin crystals, J. Pharm. Sci. 96 (2007) 2134–2144.

[16] F. Thielmann, Introduction into characterisation of porous materials by inverse gaschromatography, J. Chromatogr. A 1037 (2004) 115–123.

[17] P.P. Ylä-Mäihäniemi, J.Y.Y. Heng, F. Thielmann, D.R. Williams, Inverse gas chromato-graphic method for measuring the dispersive surface energy distribution for partic-ulates, Langmuir 24 (2008) 9551–9557.

[18] J.Y.Y. Heng, F. Thielmann, D.R. Williams, The effects of milling on the surface proper-ties of form I paracetamol crystals, Pharm. Res. 23 (2006) 1918–1927.

[19] P. York, M.D. Ticehurst, J.C. Osborn, R.J. Roberts, R.C. Rowe, Characterisation of thesurface energetics of milled dl-propranolol hydrochloride using inverse gas chroma-tography and molecular modeling, Int. J. Pharm. 174 (1998) 179–186.

[20] H.E. Newell, G. Buckton, D.A. Butler, F. Thielmann, D.R. Williams, The use of inversephase gas chromatrography to measure the surface energy of crystalline, amor-phous, and recently milled lactose, Pharm. Res. 18 (2001) 662–666.

[21] M. Ohta, G. Buckton, Determination of the changes in surface energetics ofcefditoren pivoxil as a consequence of processing induced disorder and equilibra-tion to different relative humidity levels, Int. J. Pharm. 269 (2004) 81–88.

[22] F. Thielmann, D.J. Burnett, J.Y.Y. Heng, Determination of the surface energy distribu-tions of different processed lactose, Drug Dev. Ind. Pharm. 33 (2007) 1240–1253.

[23] R. Ho, D. Wilson, J.Y.Y. Heng, Crystal habits and the variation in surface energy het-erogeneity, Cryst. Growth Des. 9 (2009) 4907–4911.

[24] J.Y.Y. Heng, A. Bismarck, D.R. Williams, Anisotropic surface chemistry of crystallinepharmaceutical solids, AAPS Pharm. Sci. Technol. 7 (2006) E1–E9.

[25] G. Buckton, Assessment of the wettability of pharmaceutical powders, J. Adhes. Sci.Technol. 7 (1993) 205–219.

[26] G. Buckton, J.M. Newton, Assessment of the wettability of powders by use of com-pressed powder discs, Powder Technol. 46 (1986) 201–208.

[27] J.Y.Y. Heng, A. Bismarck, A.F. Lee, K. Wilson, D.R. Williams, Anisotropic surface ener-getic and wettability of macroscopic form I paracetamol crystals, Langmuir 22(2006) 2760–2769.

[28] C. Levoguer, D. Butler, F. Thielmann, D. Williams, Characterising the physico-chemical properties of pharmaceutical materials, Pharm. Technol. Eur. 12 (2000)36–42.

[29] J. Domingue, D. Burnett, F. Thielmann, Using inverse gas chromatography (IGC) toinvestigate process-related changes in surface and bulk properties of pharmaceuti-cal materials, Am. Lab. 35 (2003) 32–37.

[30] A. Otte, T. Carvajal, Assessment of milling-induced disorder of two pharmaceuticalcompounds, J. Pharm. Sci. 100 (2011) 1793–1804.

[31] J.F. Gamble, M. Leane, D. Olusanmi, M. Tobyn, E. Supuk, J. Khoo, M. Naderi, Surfaceenergy analysis as a tool to probe the surface energy characteristics of micronizedmaterials — a comparison with inverse gas chromatography, Int. J. Pharm. 422(2012) 238–244.

[32] P. Begat, D.A.V. Morton, J.N. Staniforth, R. Price, The cohesive–adhesive balances indry powder inhaler formulations I: direct quantification by atomic forcemicroscopy,Pharm. Res. 21 (2004) 1591–1597.

[33] R.R. Vippagunta, R. LoBrutto, C. Pan, J.P. Lakshman, Investigation of metformin HCllot-to-lot variation on flowability differences exhibited during drug product pro-cessing, J. Pharm. Sci. 99 (2010) 5030–5039.

[34] J.F. Gamble, R.N. Davé, S. Kiang, M.M. Leane, M. Tobyn, S.S.Y. Wang, Investigating theapplicability of inverse gas chromatography to binary powdered systems: an appli-cation of surface heterogeneity profiles to understanding preferential probe-surfaceinteractions, Int. J. Pharm. 445 (2013) 39–46.

[35] G.M. Dorris, D.G. Gray, Adsorption of n-alkanes at zero surface coverage on cellulosepaper and wood fibres, J. Colloid Interface Sci. 77 (1980) 353–362.

[36] A. Voekel, K. Batko, K. Adamska, B. Strezemiecka, Determination of Hansen solubilityparameters by means of gas–solid inverse gas chromatography, Adsorpt. Sci.Technol. 26 (2008) 93–102.

[37] J. Schultz, L. Lavielle, C. Martin, The role of the interface in carbon fibre-epoxy com-posites, J. Adhes. 23 (1987) 45–60.

[38] S. Dong, M. Brendlé, J.B. Donnet, Study of solid surface polarity by inverse gas chro-matography at infinite dilution, Chromatographia 28 (1989) 469–472.

[39] C.J. van Oss, Interfacial Forces in Aqueous Media, Marcel Dekker, New York, 1994.[40] R.E. Freeman, J.R. Cooke, L.C.R. Schneider, Measuring shear properties and normal

stresses generated within a rotational shear cell for consolidated and non-consolidated powders, Powder Technol. 190 (2009) 65–69.

[41] D.K. Owens, R.C. Wendt, Estimation of the surface free energy of polymers, J. Appl.Polym. Sci. 13 (1969) 1741–1747.

[42] W. Shi, D. Hajedemos, O. Sprockel, An energetic approach of powder mixing by ma-nipulating cohesive interaction in a magnetic field, Powder Technol. 235 (2013)400–404.

[43] M. Mantel, J.P. Wightman, Influence of the surface chemistry on the wettability ofstainless steel, Surf. Interface Anal. 21 (1994) 595–605.

14

233D. Olusanmi et al. / Powder Technology 258 (2014) 222–233

[44] M. Bunker, J. Zhang, R. Blanchard, C.J. Roberts, Characterising the surface adhesivebehaviour of tablet tooling components by atomic force microscopy, Drug Dev.Ind. Pharm. 37 (2011) 875–885.

[45] E. Supuk, A. Zarrebini, J.P. Reddy, H. Hughes, M.M. Leane, M.J. Tobyn, P. Timmins, M.Ghadiri, Tribo-electrification of active pharmaceutical ingredients and excipients,Powder Technol. 217 (2012) 427–434.

[46] G. Rowley, Quantifying electrostatic interactions in pharmaceutical solid systems,Int. J. Pharm. 227 (2001) 47–55.

[47] K.R. LaMarche, F.J. Muzzio, T. Shinbrot, B.J. Glasser, Granular flow anddielectrophoresis: the effect of electrostatic forces on adhesion and flow of dielectricgranular materials, Powder Technol. 199 (2010) 180–188.

[48] A.J. Tamara, C. Yap, M.A. Mannan, Gravity pipe flow of polymeric bulk solids inpneumatic conveying system, Chem. Eng. Sci. 61 (2006) 7836–7849.

[49] L.J. Bellamy, A. Nordon, D. Littlejohn, Effects of particle size and cohesive propertieson mixing studied by non-contact NIR, Int. J. Pharm. 361 (2008) 87–91.

15

Rapid communication

Continuous manufacturing of solid lipid nanoparticles by hot meltextrusion

Hemlata Patil, Vijay Kulkarni, Soumyajit Majumdar, Michael A. Repka *Department of Pharmaceutics & Drug Delivery, Pii Center for Pharmaceutical Technology, School of Pharmacy, The University of Mississippi, MS 38677, USA

A R T I C L E I N F O

Article history:Received 10 March 2014Received in revised form 11 May 2014Accepted 16 May 2014Available online 20 May 2014

Keywords:Solid lipid nanoparticlesHot melt extrusionHigh pressure homogenizer

A B S T R A C T

Solid lipid nanoparticles (SLN) can either be produced by hot homogenization of melted lipids at highertemperatures or by a cold homogenization process. This paper proposes and demonstrates theformulation of SLN for pharmaceutical applications by combining two processes: hot melt extrusion(HME) technology for melt-emulsification and high-pressure homogenization (HPH) for size reduction.This work aimed at developing continuous and scalable processes for SLN by mixing a lipid and aqueousphase containing an emulsifier in the extruder barrel at temperatures above the melting point of the lipidand further reducing the particle size of emulsion by HPH linked to HME in a sequence. The developednovel platform demonstrated better process control and size reduction compared to the conventionalprocess of hot homogenization (batch process). Varying the process parameters enabled the productionof SLN below 200 nm (for 60 mg/ml lipid solution at a flow rate of 100 ml/min). Among the severalprocess parameters investigated, the lipid concentration, residence time and screw design played majorroles in influencing the size of the SLN. This new process demonstrates the potential use of hot meltextrusion technology for continuous and large-scale production of SLN.

ã 2014 Elsevier B.V. All rights reserved.

1. Introduction

Solid lipid nanoparticles (SLN) is a potential drug deliverysystem that has attracted increasing attention in the recent yearsas a carrier system for cosmetic ingredients, nutraceuticals andpharmaceutical drugs (Dingler and Gohla, 2002). SLN differs fromnanoemulsions based on the lipid nature, wherein SLN replaces theliquid lipid to high melting point glycerides or waxes. SLN has beenreported for controlled drug delivery, bioavailability enhancementby modification of dissolution rate and/or improvement of tissuedistribution and targeting of drugs by using various applicationroutes. SLN are mainly formulated by non-solvent or solvent-basedtechniques. The solvent-based techniques utilize organic solventsto dissolve the solid lipids followed by evaporation of the solvent(s)from the emulsion to obtain SLN. Non-solvent techniques liquefythe solid lipid above its melting point and subsequently areconverted to a nanoemulsion by cooling, which results in SLN.

High-pressure homogenization (HPH), microemulsions, high-speed stirring or ultra-sonication and membrane emulsificationare some of the non-solvent based approaches. Solvent evapora-tion by precipitation in o/w emulsion processes for formulation of

SLN entails disadvantages such as the very use of organic solventsas well as the requirement of large amounts of surfactants(Hou et al., 2003). HPH (hot and cold homogenization) has beenexplored for its feasibility in scaling-up. R. H. Muller in 2010conducted the scale up studies for production of stavudine-loadedSLN from laboratory scale (40 g) to medium scale (10 kg) and largescale (20/60 kg). This scale up study is very significant; however, itis only related to one part of the SLN production process, which isthe homogenization step for size reduction (Muller et al., 2011).However, these methods for SLN preparation involve multistepprocesses (melting of lipids, dispersion or dissolution of the drug inmelted lipids, preparation of aqueous dispersions and finally sizereduction), hence, rendering it a batch process. In the pharmaceu-tical industry, a continuous process is almost always preferred overbatch processes as continuous processing decreases the cost ofproduction by decreased space requirements, labor and resources.A continuous process can provide higher efficiency and improvedproduct quality attributes, whereas risks of batch-to-batchvariation require careful and complex procedures and controlsthat can lead to variable product outcomes such as differentparticle size, polydispersion indices, zeta potential and variationsin entrapment efficiency of the drug into the SLN. New approachesare therefore needed, both to increase the quality of the productand to reduce the production time. Hot melt extrusion is aninnovative technology for the production of a variety of dosage

* Corresponding author. Tel.: +1 662 915 1155.E-mail address: marepka@olemiss.edu (M.A. Repka).

http://dx.doi.org/10.1016/j.ijpharm.2014.05.0240378-5173/ã 2014 Elsevier B.V. All rights reserved.

International Journal of Pharmaceutics 471 (2014) 153–156

Contents lists available at ScienceDirect

International Journal of Pharmaceutics

journal homepage: www.elsev ier .com/locate / i jpharm

16

forms, offering several advantages over traditional processingtechniques and is cost-effective (Khinast et al., 2013; Roblegg et al.,2011; Repka et al., 2007; Crowley et al., 2007).

Hot melt extrusion (HME) technology is a continuous process ofpumping raw materials at high temperatures and pressuresresulting in a product of uniform shape and density (Breitenbach,2002; Maniruzzaman et al., 2012). To the best of our knowledge,HME has not been reported in the literature for the manufacturingof SLN. The purpose of this current work was to develop acontinuous process for SLN production using hot melt extrusion forpreparing a pre-emulsion and high-pressure homogenization forfurther size reduction. With this process, HME and HPH areconnected in a sequence through an insulated connection. Processparameters, such as the zone of liquid addition (ZA), barrel zonetemperature (ZT), screw speed (SS), liquid temperature (LT), screwdesign (SD), and lipid concentration (LC) were optimized for theformulation of SLN by this hot melt extrusion–high pressurehomogenization method. First, three parameters that is ZA, ZT andSS play a very important role to ensure the lipid is completely in amolten state and the drug is completely melted and dissolved inthe lipid before coming in contact with the surfactant solution. The

HME barrel zone temperature for all zones between the feedingzone to the liquid addition zone and screw speed should besufficient to melt the lipid completely when it contacts thesurfactant aqueous solution and, hence, due to high sheargeneration within the extruder, the two phases mix with eachother to form an emulsion. The scheme for continuousmanufacturing of SLN is summarized in Fig. 1. SLN were preparedby the figure-illustrated hot melt extrusion and high-pressurehomogenization method.

Glyceryl behenate (Compritol1 888 ATO, Gattefosse, France) wasfed into the co-rotating twin-screw extruder (11 mm Process 11,ThermoFisher Scientific, Karlsruhe, Germany) using a gravimetricfeeder. A 1.5% w/w Tween 80 (Sigma–Aldrich, USA) aqueous solutionheated to the equivalent to the extrusion temperature was injectedinto the extruder barrel through an injection port using a peristalticpump. The melt-extrusion was performed by varying the formula-tion parameters, process parameters and screw configuration asdescribed inTable 1. The barrel temperature for zone 2, 3 and the restof the zones (4–8) along with the die were varied based on the screwspeed. The screw configuration with three mixing zones was usedeither as manufacturer’s standard screw configuration or a modified

Table 1Experimental design matrix and response variable values obtained (data represent mean � SD, n = 6).

Batch Screw speed Zone of liquidaddition

Barrel temp.a

(�C)Screw design Lipid conc.

(% w/w)Z-average(nm)

PDI Zeta potential(mV)

F1 240 3 150–100–83 Std. 6 85.1 � 7.12 0.353 � 0.025 �31.8 � 0.11F2 240 4 150–100–83 Std. 6 286.9 � 5.12 0.312 � 0.012 �28.6 � 0.09F3 240 3 150–100–83 Modi. 6 245.5 � 6.91 0.408 � 0.041 �33.3 � 0.18F4 240 3 120–85–75 Std. 6 303.9 � 7.64 0.295 � 0.028 �29.2 � 0.08F5 160 3 150–100–83 Std. 6 266.3 � 4.71 0.264 � 0.015 �30.6 � 0.15F6 160 3 150–100–83 Modi. 6 194.8 � 5.82 0.391 � 0.009 �34.3 � 0.04F7 160 3 150–100–83 Modi. 12 846.7 � 21.52 0.625 � 0.085 �15.7 � 0.48F8 Conventional process for SLN preparation 6 248.2 � 13.42 0.412 � 0.038 �30.6 � 0.12

a The barrel temperature for zone no. 2–3–4 through the die (F6 optimized batch barrel temp. in zone 2 – 150 �C, zone 3 – 100 �C and zone 4 through the die – 83 �C).

Fig. 1. Schematic representation of continuous preparation of SLNs using hot melt extrusion connected to a high pressure homogenizer.

154 H. Patil et al. / International Journal of Pharmaceutics 471 (2014) 153–156

17

screwconfiguration (Fig. 2). The hot pre-emulsion resulting from thehot melt extrusion process was passed through insulated tubingconnectedtotheHMEdieand thesample holderof thehigh-pressurehomogenizer (Avestin Emulsiflex C5, Canada) for size reduction. Thehigh-pressure homogenization was performed at 75 �C and 1000 barpressure to reduce the particle size of the melt extruded pre-emulsion. The HPH parameters were constant for all of the batches.Further, the size-reduced emulsion was cooled at room temperatureto obtain SLN. The mean particle size of SLN was determined byphoton correlation spectroscopy (Zetasizer-Nano-ZS, Malvern, UK)at 25 �C. The effect of process variables on particle size and zetapotential are described inTable 1. The SLN prepared using HME–HPHwere compared to SLN prepared by a conventional method. Theconventional SLN were prepared as described by Wang et al., usingthe same lipid and surfactant concentration (Wang et al., 2012). Theproduced pre-emulsion was then passed through a high pressurehomogenizer. The selected variables including the screw speed,screw configuration, processing temperature and lipid concentra-tion, significantly influenced the particle size of the SLN (Table 1). InF1andF2, the fed lipidwascompletely meltedbefore reachingzone 3and therefore, additionof surfactant solutioneitheratzone 3 or4 hadno significant effect on quality of emulsion formulation at theemployed processing conditions. Modifying the screw configuration(F3) by including more 90� elements at the second mixing zone(Fig. 2) decreased the particle size of the SLN. The mixing elementsgeometry allows the increase of the radial mixing of material insidethe barrel by maintaining the flow channels in contact with eachother. In addition, a higher shear rate inside the kneading elementsresults in the reduction in particle size. It was observed that higherscrew speeds and lower barrel temperatures for zone 2 and 3 (F4)resultedin incomplete meltingof the lipid reachingthe zone of liquidaddition hence, resulting in inadequate formation of the pre-emulsion and consequently higher particle size of SLN. Screw speedcan be converted in terms of residence time of material inside theextruder barrel. At lower screw speeds (F5), a decrease in the particlesize of SLN was observed due to the increase in contact time betweenthe melted lipid and the aqueous surfactant solution resulting in animproved pre-emulsion formation prior to high-pressure homoge-nization. Based on these results, processing of SLN at lower screwspeeds and modified screw configurations produced SLN with

particle sizes below 200 nm and narrow PDI. These data may beexplained due to the synergistic effect of increased residence timeand high shear mixing. Furthermore, using the same processingparameters and increasing the lipid concentration from 6% to 12% w/w, demonstrated a significant increase in particle size. At 12% lipid,surfactant concentration was insufficient to lower the surfacetension between the oil and water phase. SLN processed with HME(F6) exhibited lower particle size of SLN compared to theconventional process utilized for SLN preparation.

2. Conclusion

In conclusion, SLN were successfully prepared by the newlyproposed method using hot melt extrusion and high-pressurehomogenization technology. Continuous production of SLN can beachieved by this novel method. High shear generation inside theHME barrel results in the formation of a very good quality pre-emulsion. As shown in Table 2, the pre-emulsion obtained fromHME is much improved as compared to the pre-emulsion preparedby a conventional method. Therefore, compared to other knownprocesses for producing solid lipid nanoparticles, such as preparingthe pre-emulsion with ultra-turrex in a batch process, the HMEprocess produces SLN with a decreased particle size, decreasedpolydispersion indices, and decreased zeta potential. The quality ofthe pre-emulsion affects the quality of the final product attributesto a large extent and it is desirable to obtain droplets in the sizerange of a few micrometers. The developed technique resulted inpromising results for continuous processing of pharmaceuticalsemploying minimal manual steps, providing a cost effectiveprocess, resulting in shorter processing times, exhibiting less

Table 2Comparison between pre-emulsion quality obtained by HME and a conventionalmethod.

No. Method Particle size(nm)

PDI Zeta potential(mV)

1 HME 799.1 0.754 �22.52 Conventional method 1812.3 1.000 �20.2

Fig. 2. Screw configuration and mixing zones used in the preparation of SLNs by HME.

H. Patil et al. / International Journal of Pharmaceutics 471 (2014) 153–156 155

18

formulation variation and providing dependable quality productattributes. This technology may be used in preparation of SLN fororal as well as for parenteral use. Future studies will includeincorporation of drug and optimization of the processingparameters to enhance the drug loading efficiency and stabilityof the SLN.

References

Breitenbach, J., 2002. Melt extrusion: from process to drug delivery technology. Eur.J. Pharm. Biopharm. 54, 107–117.

Crowley, M.M., Zhang, F., Repka, M.A., Thumma, S., Upadhye, S.B., Battu, S.K.,McGinity, J.W., Martin, C., 2007. Drug Dev. Ind. Pharm. 33, 909–926.

Dingler, A., Gohla, S., 2002. Production of solid lipid nanoparticles (SLN): scaling upfeasibilities. J. Microencapsul. 19, 11–16.

Hou, D.Z., Xie, C.S., Huang, K.J., Zhu, C.H., 2003. The production and characteristics ofsolid lipid nanoparticles (SLNs). Biomaterials 24, 1781–1785.

Khinast, J., Baumgartner, R., Roblegg, E., 2013. Nano-extrusion: a one-step processfor manufacturing of solid nanoparticle formulations directly from the liquidphase. AAPS PharmSciTech 14, 601–604.

Maniruzzaman, M., Boateng, J.S., Snowde, M.J., Douroumis, D., 2012. A review of hot-melt extrusion: process technology to pharmaceutical products. ISRN Pharm.1–9.

Muller, R.H., Shegokar, R., Singh, K.K., 2011. Production & stability of stavudine solidlipid nanoparticles from lab to industrial scale. Int. J. Pharm. 416, 461–470.

Repka, M.A., Battu, S.K., Upadhye, S.B., Thumma, S., Crowley, M.M., Zhang, F., Martin,C., McGinity, J.W., 2007. Pharmaceutical applications of hot-melt extrusion: PartII. Drug Dev. Ind. Pharm. 33, 1043–1057.

Roblegg, E., Jäger, E., Hodzic, A., Koscher, G., Mohr, S., Zimmer, A., et al., 2011.Development of sustained-release lipophilic calcium stearate pellets via hotmelt extrusion. Eur. J. Pharm. Biopharm. 79, 635–645.

Wang, S., Chen, T., Chen, R., Hu, Y., Chen, M., Wang, Y., 2012. Emodin loaded solidlipid nanoparticles: preparation, characterization and antitumor activitystudies. Int. J. Pharm. 430, 238–246.

156 H. Patil et al. / International Journal of Pharmaceutics 471 (2014) 153–156

19

˝ 2

016

Cat

alen

t Ph

arm

a So

luti

ons.

All

righ

ts r

eser

ved.

Learn more at catalent.com/optiform us + 1 888 SOLUTION eu + 800 8855 6178 solutions@catalent.com

easierIntegrated solution at one simple price with minimal material needed.

simplerOptimal recommendations based on real data from a dedicated scientific advisor.

fasterAccelerated parallel process with superior technologies allowing for optimized animal PK materials in 12 weeks!

Rigorous science. Superior technologies. From molecule to dose form.

new optiform‰ solution suite

enhanced bioavailability in 12 weeks!

International Journal of Pharmaceutics 496 (2015) 33–41

Formulation and development of pH-independent/dependentsustained release matrix tablets of ondansetron HCl by a continuoustwin-screw melt granulation process

Hemlata Patil a, Roshan V. Tiwari a, Sampada B. Upadhye b, Ronald S. Vladyka b,Michael A. Repka a,c,*aDepartment of Pharmaceutics and Drug Delivery, School of Pharmacy, The University of Mississippi, University, MS 38677, USAbCatalent Pharma Solutions, 14 School House Road, Somerset, NJ 08873, USAcCenter for Pharmaceutical Technology, School of Pharmacy, The University of Mississippi, Oxford, MS 38677, USA

A R T I C L E I N F O

Article history:Received 16 February 2015Received in revised form 28 March 2015Accepted 6 April 2015Available online 8 April 2015

Keywords:OndansetronMelt-granulationSustain releaseHot melt extrusionpH-dependent/independent drug release

A B S T R A C T

The objective of the present study was to develop pH-independent/dependent sustained release (SR)tablets of ondansetron HCl dihydrate (OND), a selective 5-HT3 receptor antagonist that is used forprevention of nausea and vomiting caused by chemotherapy, radiotherapy and postoperative treatment.The challenge with the OND API is its pH-dependent solubility and relatively short elimination half-life.Therefore, investigations were made to solve these problems in the current study. Formulations wereprepared using stearic acid as a binding agent via a melt granulation process in a twin-screw extruder.The micro-environmental pH of the tablet was manipulated by the addition of fumaric acid to enhancethe solubility and release of OND from the tablet. The in vitro release study demonstrated sustainedrelease for 24 h with 90% of drug release in formulations using stearic acid in combination with ethylcellulose, whereas 100% drug release in 8 h for stearic acid-hydroxypropylcellulose matrices. Theformulation release kinetics was correlated to the Higuchi diffusion model and a non-Fickian drug releasemechanism. The results of the present study demonstrated for the first time the pH dependent releasefrom hydrophilic-lipid matrices as well as pH independent release from hydrophobic-lipid matrices forOND SR tablets manufactured by means of a continuous melt granulation technique utilizing a twin-screw extruder.

ã 2015 Elsevier B.V. All rights reserved.

Contents lists available at ScienceDirect

International Journal of Pharmaceutics

journal homepage: www.elsev ier .com/locate / i jpharm

1. Introduction

Several approaches have been used to develop sustain release(SR) matrix formulations and among them the most interestingapproach is based on melt granulation techniques. Melt granula-tion is a process by which pharmaceutical powders are efficientlyagglomerated by the use of a low melting point binding material,which after melting act as a binding liquid (Perissutti et al., 2003).The advantage of this process over other techniques is that it is asolvent free process, and thus there is no need for a drying step,which consumes time and energy. Also, the percent of finesproduced by the melt granulation process is less as compared tothe wet and dry granulation process (Tan et al., 2014). Several

* Corresponding author at: Department of Pharmaceutics and Drug Delivery, PiiCenter for Pharmaceutical Technology School of Pharmacy, The University ofMississippi, University, MS 38677, USA.

E-mail address: marepka@olemiss.edu (M.A. Repka).

http://dx.doi.org/10.1016/j.ijpharm.2015.04.0090378-5173/ã 2015 Elsevier B.V. All rights reserved.

researchers have used melt granulation processes using differentkinds of low melting point excipients as binders (polyethyleneglycols 3000, 6000 and 8000, various types of waxes and lipids)(Voinovich et al., 1999, 2001; Yang et al., 2007). Over the last fewyears several techniques have been investigated for melt granula-tion processes such as high-shear granulation, fluidized bedprocessing and others. (Aoki et al., 2015; Campisi et al., 1999). Hotmelt extrusion (HME) is another thermal processing technique thathas attracted interest as a novel technique for melt granulation (Liuet al., 2001; Yang et al., 2007). HME is a continuous, simple, easy toscale up and efficient process (Maniruzzaman et al., 2012; Patilet al., 2014; Repka et al., 2007).

Nausea and vomiting are the common problems in cancerpatients undergoing chemotherapy, radiation therapy, and post-operative treatment or due to the advancement in cancer itself.Serotonin (5-hydroxytryptamine) subtype-3 (5-HT3) receptorplays an important role in emetogenic pathways in relation tomassive release of serotonin from damaged enterochromaffin cellsin the gastrointestinal tract following chemotherapy (Harris, 2010;

21

Table 1Compositions of the investigated granules (all quantities given in percentile).

Formulation OND EC HPC SA FA Magnesium stearate MCC

F1 20 37 – 12 – 0.2 30.5F2 20 – 37 12 – 0.2 30.5F3 20 37 – 12 2.5 0.2 28F4 20 37 – 15 2.5 0.2 25F5 20 40 – 12 2.5 0.2 25F6 20 40 – 15 2.5 0.2 22F7 20 – 37 12 2.5 0.2 28F8 20 – 37 15 2.5 0.2 25F9 20 – 40 12 2.5 0.2 25F10 20 – 40 15 2.5 0.2 22

Magnesium stearate (0.3%) added in all batches extra-granularly before compression.

34 H. Patil et al. / International Journal of Pharmaceutics 496 (2015) 33–41

McManis and Talley, 1997). Serotonin antagonists exert theiranti-emetic action via 5-HT3 receptors located centrally (chemo-receptor trigger zone of the area postrema and nucleus tractussolitaries) as well as peripherally (enterochromaffin cells of theenteric nervous system) (Glare et al., 2008; Hesketh and Gandara,1991). 5-HT3 receptor can be antagonized by drugs such asondansetron, tropisetron, granisetron, dolasetron, palonosetronand ramosetron as well as by certain drugs having prokinetic(increasing gut motility) action such as renzapride, cisapride andmetoploramide (McManis and Talley, 1997; Warr, 2008).

Ondansetron, a carbazole derivative, is a potent, highly selectivecompetitive 5-HT3 receptor antagonist (Ye et al., 2001). Intrave-nous and oral dosage forms of OND are commercially available, forexample: oral solutions, orally disintegrating tablets and film-coated tablets. The recommended oral dose regimen of OND is8 mg, three times a day. Following oral administration OND is wellabsorbed from the GIT and its bioavailability is approximately 67%and the elimination half-life is relatively short, approximately3–5 h (Gungor et al., 2010; Rojanasthien et al., 1999).

The present study has investigated HME techniques to prepareOND granules. Stearic acid is commonly used as a tablet lubricantin oral formulations. Also it has been extensively explored as abinder. In the current study, stearic acid is used as a melt binder inthe formulations. Two different polymers were used as matrixformers such as the water-insoluble and almost un-swellable ethylcellulose (Siepe et al., 2006), and the water-soluble and swellablehydroxypropyl cellulose (HPC) in combination with stearic acid.OND exhibits pH dependent solubility and therefore, it is freelysoluble in gastric fluid’s low pH (Venkatesh et al., 2009), butpractically insoluble at pH > 6. Several approaches have been usedin the past to overcome pH-dependent solubility of weakly basicdrugs (Streubel et al., 2000). The commonly used approach is theaddition of acid pH modifiers to the matrix tablets (Gabr, 1992).Therefore, based on preliminary studies, these investigators haveused fumaric acid to create a suitable micro-environmental pH,which increased drug solubility at higher pH.

Currently a commercially available orodispersible tablet as wellas conventional OND tablets are meant for immediate release andare given 3–4 times a day due to its short half-life (3–5 h).Therefore, there is a need to develop a sustained releaseformulation for OND, which will exhibit a pH-independent releaseprofile. The aim of this novel study was to: (1) develop aninnovative sustained release OND tablet by continuous twin-screwmelt granulation processing. (2) To investigate the influence offormulation parameters on the physical properties of the hot-meltextruded granules and compressed tablets containing stearic acidas a release-retarding agent. (3) Study the release kinetics for ONDtablets prepared by either hydrophilic or hydrophobic polymers incombination with stearic acid. (4) To study the effect of fumaricacid on drug release behavior from the tablet matrix.

2. Materials and methods

2.1. Materials

Ondansetron HCl dihydrate was purchased from ChemsceneLLC (New Jersey, USA). Hydroxypropyl cellulose (Klucel1 EF) waskindly gifted by Ashland Specialty Ingredients. (Wilmington, DE).Ethyl cellulose (Ethocel1 Standard 10) was gifted from Dowchemical company. Fumaric acid and Magnesium stearate waspurchased from Spectrum Laboratory Products Inc. (Gardena, CA).Microcrystalline cellulose (Avicel1 PH 101) was gifted from FMCBiopolymer (Philadelphia, PA). Stearic acid was purchased fromEMD Millipore (Billerica, MA). All the other reagents such asmethanol (impurities <0.1%) used in this study were of theanalytical grade.

2.2. Compatibility of OND with different excipients

To study the compatibility of OND with the polymers and otherexcipients, physical mixtures were prepared by mixing the drugwith each formulation excipients in the ratio of 1:1. Pure drug(OND) and two physical mixtures were characterized by differen-tial scanning calorimetry (Diamond DSC, PerkinElmer) using Pyrismanager software (PerkinElmer Life and Analytical Sciences, 719Bridgeport Ave., CT, USA) and Fourier transform infrared spectros-copy (Agilent Technologies Cary 660, Santa Clara, CA). For the DSCstudy, samples were prepared by sealing 3–5 mg of pure API andphysical mixture in hermetically sealed aluminum pans and thethermal analysis was performed under an inert nitrogen atmo-sphere at a heating rate of 10 �C/min over a temperature range of40–250 �C. FTIR studies were conducted in the range of 4000–400 cm�1. The bench was equipped with an ATR (Pike TechnologiesMIRacle ATR, Madison, WI), which was fitted with a single bouncediamond coated ZnSe internal reflection element.

2.3. Hot melt granulation

Ondansetron HCl dihydrate (20% w/w) was blended with otherexcipients as shown in Table 1 using a V-shell blender (Globe-Pharma, MaxiblendTM New Brunswick, NJ) for 20 min at 25 rpm,after passing through US# 35 mesh screen to remove anyaggregates that may have formed. Hot melt granulations wereprocessed without an extrusion die in a fully intermeshingco-rotating twin-screw extruder (11 mm Process 11TM ThermoFisher Scientific) with modified screw design (Fig. 1). This extruderbarrel consists of a total of 8 zones. The first zone (no heating), alsoknown as feeding zone, is where the physical mixture is fed. Basedon the preliminary studies zone 2 and zone 3 were set at 110 �C,zone 4 was set at 70 �C and the remaining subsequent zones (zone5–zone 8) were heated to 50 �C. The set temperatures of all zoneswere below the melting point of OND (189 �C). The modified co-rotating screw configuration consists of mixing elements at twoplaces, one at zone 3 and second before the discharge elements.The extruder was allowed to equilibrate at set temperature for30 min before starting the trials. The physical mixtures wereloaded into the volumetric feeder, fed from the extruder feedingzone and further processed through extruder barrel at the settemperatures. Powder feed rate (7.2 g/min) and screw speed(100 rpm) were selected from the preliminary data and keptconstant for all batches. Granules were collected from the open endof the extruder after steady state was reached for a trial condition.At the end of the granulation process the collected granules werestored in a foil-lined sealed bag.

2.4. Micromeritic properties of granules

Bulk density was calculated by measuring the volume of 5 ggranules in a 10 ml graduated cylinder. The cylinder was tapped

22

Fig. 1. Modified screw configuration.

H. Patil et al. / International Journal of Pharmaceutics 496 (2015) 33–41 35

100 times until no further reduction in the volume of the granuleswas observed. Tapped density was calculated using the volume ofthe granules after tapping. Flow properties of the granules such asangle of repose, Carr’s index, and Hausner’s ratio were alsocalculated. The angle of repose was determined by the funnelmethod. The accurately weighed powder was placed in a funnel.The height of the funnel through which the powder passes wasadjusted in such a way that the tip of the funnel just touched the tipof the cone of the powder. The powder was allowed to flow throughthe funnel freely onto the fixed base. The diameter of the powdercone and height of the cone was measured. The angle of repose (u)was calculated using the following equation:

TanðuÞ ¼ hr

where ‘h’ and ‘r’ are the height and radius of the powder conerespectively.

Carr’s (compressibility) index (Corti et al., 2008) and Hausner’sratio (HR) was determined according to the following formula:

CI ¼ ðTapped density � Bulk densityÞTapped density

� 100

HR ¼ Tapped densityBulk density

2.5. Particle size distribution

Particle size distribution was analyzed by the sieve analysismethod. Two standard sieves (U.S.A. Standard Test Sieve) withmesh sizes of 500 mm and 1400 mm were used to conduct the sieveanalysis. Three fractions were collected, that is, granule size morethan 1400 mm, granule size in-between 1400 and 500 mm andgranule size less than 500 mm which are considered as fines andthe percentile weight distribution was determined for all threefractions.

2.6. Tablet compression

Prior to direct tablet compression, granules were mixed with0.3% magnesium stearate. Tablets were manually prepared bydirect compression on a single punch tablet press (MCTMI,GlobePharma Inc., New Brunswick, NJ), by using an 8 mm flatround punch at a compression force of 130 kg/cm.

2.7. Evaluation of tablet properties

Compressed tablets were evaluated for hardness, thickness,friability and drug content uniformity. Ten tablets were randomlyselected and tested for their hardness (Hardness tester,

Schleuniger). An additional ten tablets were randomly selectedand tested for their thickness using a digital Vernier caliper(Montana). The friability was determined as percentage weightloss of tablets (weighing 6.5 g) using a dual scooping projectionVander Kamp friabilator (Vankel Industries Inc., Chatham, NJ) for4 min at 25 rpm.

2.8. Evaluation of gel layer pH

To study the effect of fumaric acid, the pH of the tablet surfacewas measured within 10 h by an Oakton pH meter (pH Spear, FisherScientific) equipped with a contact electrode. The dissolutionmedia used to conduct this test was 0.1N HCl (pH 1.2) for the first2 h, followed by pH 6.8 buffer for the remaining 8 h. The tabletswere removed at an interval of 1 h to check the surface pH up to10 h (Dvorackova et al., 2013). Briefly, the tested tablets wereremoved from the dissolution vessel at set time interval andbathed with purified water to remove residual buffer from thetablet surface. Then the contact electrode was slightly pressed intothe gel layer of the tablet to measure the pH.

2.9. In vitro drug release

In vitro drug release was measured using USP dissolutionapparatus II (Hanson SR8) set at 50 rpm and equipped with UV–visprobes (Rainbow Dissolution Monitor, PION) collecting spectraevery 2 min for the first 2 h and then for every 25 min until 24 h at305 nm. The test dissolution media were 700 ml of 0.1N HCl (pH1.2) with 1% SLS for the first 2 h, then 200 ml of 0.2 M tribasicsodium phosphate (pH 12.5) with 1% SLS to provide a final pH of 6.8for 24 h (media were maintained at 37 � 0.5 �C) to simulate thetablet transit from stomach (pH 1.2) to the intestine (pH 6.8) (Cortiet al., 2008; He et al., 2014). The drug release studies wereconducted in triplicates and the mean values were plotted versustime.

2.10. Similarity and dissimilarity factor analysis

The similarity (f2) and dissimilarity (f1) factor was used toevaluate pH-independent release patterns of OND from theoptimized tablets in the release media pH 1.2 and pH 6.8.Similarity and dissimilarity factors are calculated by using thefollowing equation (Moore and Flanner, 1996):

f 2 ¼ 50log1ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

1 þ 1=pð ÞSPi¼1ðRt � TtÞ2

q � 100

0B@

1CA

f 1 ¼ f½St¼1n jRt � Ttj�g � 100

here, Rt and Tt are the cumulative percent of drug dissolved frommatrices for the references and test samples at time t and n is the

23

36 H. Patil et al. / International Journal of Pharmaceutics 496 (2015) 33–41

number of time points. The similarity factor values ranges between0 and 100. The similarity between two profiles increases when f2value approaches 100, whereas dissimilarity occurs with adecrease of the f2 value (less than 50) (Pillay and Fassihi, 1998).

2.11. Drug release kinetics and mechanism

In order to study the mechanism of drug release from thematrix tablets, the experimental data was evaluated kinetically bythe following equations (Bose et al., 2013; Dvorackova et al., 2013;Higuchi, 1963; Korsmeyer et al., 1983),

Zeroorderequation :Qt ¼ Q0 þ K0t

Firstorderequation : InQt ¼InQ0 þ K1t

2:303

Fig. 2. (a) DSC and (b) FTIR spectra of pure OND, phys

Higuchiequation : Qt ¼ KHt1=2

Korsmeyer � Peppas equation :Qt

Q1¼ Kkpt

n

where, Qt is the amount of drug released in time t, Q0 is the initialamount of the drug in the solution, Q1 is the amount of drugreleased after infinite time, K0 is the zero order release rateconstant, K1 is the first order release rate constant, KH is theHiguchi diffusion rate constant, Kkp is the release constantcomprised of structural and geometrical characteristics of thetablets and ‘n’ is the release exponent indicating the mechanism ofthe drug release.

ical mixture for EC-SA matrix and HPC-SA matrix.

24

Fig. 3. Particle size distributions of the OND granules (F1–F10) (determined bysieve analysis).

H. Patil et al. / International Journal of Pharmaceutics 496 (2015) 33–41 37

2.12. Statistical data analysis

The differences between batches were analyzed by one-wayanalysis of variance (ANOVA) followed by Student’s t-test. Adifference of p < 0.05 was considered statistical significant. Allvalues were reported as the mean of three recordings.

3. Results and discussion

3.1. Compatibility of OND with different excipients

Drug-excipients compatibility studies were performed utilizingDSC and FTIR. Any abrupt or drastic changes in the thermalbehavior of either the drug or polymer may indicate a possibleinteraction between drug and polymer. DSC of drug-polymerphysical mixture for both EC-SA and HPC-SA matrices showed awell recognizable endothermic peak of OND at 181.74 �C, thetemperature slightly shifted to the lower temperature thancorresponding to the melting point of the pure drug (188.58 �C).This result indicates that the drug is compatible with the polymerbut due to the presence of stearic acid which is acting as aplasticizer results in lowering the melting temperature of OND.There was no change in the glass transition temperature andmelting endotherms of other excipients used in the formulationconfirming the absence of any drug-polymer interaction. Also,there were no considerable changes in the IR peaks of OND whenmixed with other excipients further confirming the absence ofdrug-excipient interaction. The FTIR spectra of the pure drug andphysical mixture were compared and the characteristic peak forthe physical mixture was found to be superimposable to that of thepure drug. This result indicates that there was no drug polymerinteraction. A broad band of bonded ��OH of OND was observedfrom 3481 cm�1 to 3245.97 cm�1, and a peak of ��CH stretchingwas found at 2900 cm�1 indicating the presence of a methyl group,and at 1680 cm�1 (��C¼O stretching) indicating a keto group in allof the formulations’ physical mixtures (Fig. 2).

3.2. Micromeritic properties of granules

Granule flowability and compactability affects the die fillingand tablet mechanical characteristics; therefore it is essential tostudy the granule flow properties. Various tests have beenperformed on all formulations such as bulk density, tappeddensity, angle of repose, Carr’s index and Hausner’s ratio.According to the flow property classification of USP, the angle ofrepose in a range of 25–30 indicates excellent flow properties. Allthe prepared formulations showed excellent flowability(25 < angle of repose < 30). As shown in Table 2, it is clear thatas the amount of the meltable binder and polymer increased, theangle of repose also increased. Carr’s index and Hausner’s ratio arealso the measurement of flow properties and good flowing

Table 2Granule properties of the different formulations of ondansetron hydrochloride (OND).

Formulation Angle of repose (�) BD (gm/cm3)

F1 26 � 1.2 19 � 0.75F2 27 � 0.78 19.5 � 0.62F3 26 � 0.94 20 � 0.71F4 27 � 1.02 20 � 0.55F5 27.5 � 1.42 21 � 0.94F6 28.5 � 1.08 17.5 � 0.29F7 27 � 0.97 19 � 0.84F8 28 � 1.05 17 � 0.53F9 28 � 1.09 15.5 � 0.97F10 28.5 � 1.16 18 � 1.01

BD – Bulk Density.TD – Tapped Density.

granules should have a Carr’s index between 10.0 and 18.0%,whereas Hausner’s ratio should be less than 1.25. When the Carr’sindex and Hausner’s ratio are adequate, the powder flows atminimum bulk density and consolidates to maximum densityinside the die, prior to compression (Wells, 1988). Carr’s index andHausner’s ratio in Table 2 are between 10.0 and 18.0% and less than1.25 respectively, confirming the good flow of granules obtained inall the formulations.

3.3. Particle size distribution

The results of sieve analysis are shown in Fig. 3. These resultswere useful to understand the effect of the concentration of stearicacid, EC and HPC on the granule size fractions (F), defined as fines(F < 500 mm), fraction of interest for tableting (500 mm < F < 1400mm) being the granulation yield, and oversized granules(F > 1400 mm). Large size granules are obtained with the increasein the amount of binder stearic acid or polymer, as a result ofincrease in the cross-linking inside the matrix granules, whichleads to the increase in the matrix density of the granules.

3.4. Evaluation of tablets

The formulated tablets were evaluated for hardness, friability,weight variation, and thickness. These parameters were found tobe within the acceptable limits as shown in Table 3. OND contentuniformity in the prepared tablets was determined. Briefly, 20tablets were selected randomly and triturated with the help of amortar and pestle. The amount equivalent to weight of one tabletwas weighed and OND content was determined using a UV–visspectrophotometer at lmax 305 nm against blank. All of theformulations possessed exceptional drug content (>98%) as well as

TD (gm/cm3) Carr’s index (%) Hausner’s ratio

16 � 0.26 15.79 � 1.95 1.19 � 0.02716 � 0.46 17.95 � 0.25 1.22 � 0.003

17.5 � 0.42 12.5 � 1.00 1.14 � 0.01316 � 0.95 20 � 2.55 1.25 � 0.04018 � 0.61 14.28 � 0.93 1.17 � 0.012

14.5 � 0.38 17.14 � 0.79 1.25 � 0.01116 � 0.97 15.79 � 1.38 1.19 � 0.01914 � 0.82 17.65 � 2.25 1.21 � 0.03313 � 0.29 16.13 � 3.39 1.19 � 0.048

15.5 � 0.67 13.89 � 1.11 1.16 � 0.014

25

Fig. 4. Change of gel layer pH within the dissolution test of the matrix tablets (2 h in1.2 and following 2 h pH media changes to 6.8).

Fig. 5. In vitro drug release profile for formulations F3–F10 by two-step dissolutionmedia method.

38 H. Patil et al. / International Journal of Pharmaceutics 496 (2015) 33–41

content uniformities (<2% RSD) after the melt granulation process.This is indicative of a robust formulation and process.

3.5. Tablet surface pH measurement using contact electrode

To determine the effect of fumaric acid on OND solubility andrelease characteristics, the tablet surface pH was measured everyhour during 10 h of the study using a pH Spear, Oakton pH meter. Itwas expected that the pH of the tablet surface for all of the sampleswould be similar to the pH value of the dissolution media. For thefirst 2 h when 0.1N HCl (pH 1.2) dissolution media was used thetablet surface measured a pH of 1.2–1.4, whereas after 2 h whenphosphate buffer was added, the dissolution media pH waschanged to 6.8, at that time the differences in tablet surface pHwere observed between samples during different time intervals.Formulation F1 and F2 (without fumaric acid) indicated pH valuesthe same as pH of the dissolution media (for first 2 h in 0.1N HCl thetablet pH was 1.2–1.4, and in HCl-phosphate buffer pH was6.8–7.0). Formulation F3 and F7 (with fumaric acid) measured pH1.2–1.4 for the first 2 h in 0.1N HCl and after 2 h pH of the tabletsurface was changed to 3.5–4.0 in the HCl-phosphate bufferdissolution media of pH 6.8 (Fig. 4). This indicates that fumaric acidis dissolved inside the tablet microenvironment and thusmaintaining the acidity inside the tablet, which assisted in thesolubilization of OND in dissolution media with higher pH.

3.6. Effect of pH of dissolution media, fumaric acid and differentpolymer matrices on drug release

In vitro dissolution studies for all formulations were performedusing the two-step dissolution method by changing pH, whichbetter corresponds to the real conditions in the gastrointestinaltract (2 h in pH 1.2 and following 22 h in pH 6.8) (Dvorackova et al.,2013) (Fig. 5). Also, in vitro dissolution studies were performed inboth 0.1N HCl and pH 6.8 buffer individually to see the effect ofchange in the pH of the dissolution media on drug release. Asexpected, the pH of the dissolution medium was found tosignificantly affect (p < 0.05) the release rate of OND from thematrix system. Fig. 6 shows a lower percentage of OND release atpH 6.8 as compared with the high percentage of OND release in0.1N HCl at 24 h, which is due to the very low aqueous solubility ofOND at basic pH.

To study the effect of fumaric acid on drug release, in vitrodissolution studies for formulations containing fumaric acid(F3/F7) and formulations without fumaric acid (F1/F2) wereconducted in both pH 1.2 and pH 6.8 buffer media. Formulation F3and F4 composition contained fumaric acid to create a constantacidic micro-environment inside the tablets. Therefore, irrespec-tive of the pH of the surrounding dissolution medium, fumaric acidassisted in the solubilization and release of OND (weakly basicdrug) in high pH dissolution medium corresponding to theintestinal pH. These data contribute to solving the pH-dependent

Table 3Comparisons of the physical properties of the matrix tablets containing Ondansetron h

Formulation Hardness (Kp) Thickness (mm)

F1 6.3 � 0.07 2.40F2 6.2 � 0.05 2.38F3 6.0 � 0.01 2.37F4 6.7 � 0.06 2.39F5 6.2 � 0.02 2.40F6 6.8 � 0.04 2.37F7 5.9 � 0.03 2.39F8 6.6 � 0.08 2.41F9 6.1 � 0.04 2.40F10 6.9 � 0.03 2.38

solubility problem of OND. Ideally, fumaric acid should dissolveslowly so that it will remain inside the tablet during the entireperiod of drug release. The dissolution results clearly indicated thatthe presence of fumaric acid in the matrices increased thedissolution of OND in phosphate buffer pH 6.8. Approximately90% of drug was released within 24 h from formulation F3containing fumaric acid in both pH 1.2 and pH 6.8 buffer. Themaintenance of a constant and low acidic microenvironment byfumaric acid created the most favorable conditions for ONDrelease. However, formulation F1 and F2 which do not containfumaric acid demonstrated slower drug release in pH 6.8 ascompared to that of pH 1.2 media. These observations indicate thatthe presence of fumaric acid inside the tablet may affect initialporosity of the tablet as well as assisting in maintaining the acidicpH inside the tablet, which resulted in the increased drug release.

ydrochloride dihydrate.

Weight (mg) Friability (%) Content uniformity (%)

150.2 0.32 100.50 � 1.25150.3 0.35 100.20 � 1.43151.2 0.31 98.46 � 1.26149.8 0.30 101.30 � 0.75150.3 0.32 101.70 � 0.61150.7 0.35 103.40 � 0.45151.4 0.31 102.60 � 1.06150.9 0.35 99.80 � 1.43151.3 0.32 102.40 � 0.79151.5 0.36 103.80 � 1.02

26

Fig. 6. Effect of pH of surrounding media on OND release from formulations F1 andF2.

H. Patil et al. / International Journal of Pharmaceutics 496 (2015) 33–41 39

To clarify the important role of fumaric acid in thepH-independent/pH-dependent release of OND in both gastricand intestinal fluid, dissolution profiles of both EC-SA and HPC-SAmatrix OND tablets containing fumaric acid (F3 and F7) arecompared in Fig. 7a and b.

This research group also studied the effect of different types ofpolymers in combination with stearic acid on the drug releaseprofile. It was observed that the release of the drug depends uponthe type of polymer used in the formulation. Ethyl cellulose(hydrophobic polymer) along with stearic acid sustained the drugrelease over the period of 24 h (90% drug release). However,hydroxypropyl cellulose (hydrophilic polymer) in combinationwith stearic acid sustained drug release over a period of 7–9 h(100% drug release) (Fig. 8). This effect is contributed to the natureof polymer in use. The formulations with EC-SA matricesdemonstrated a retarded release of the drug compared to themore hydrophilic matrices. This may be due to the hydrophobicnature of the EC polymer, which prevents the penetration of thedissolution medium into the tablet matrix leading to slowerdissolution and diffusion of the drug molecules from the matrix

Fig. 7. Effect of fumaric acid on drug release from (a) EC-SA matrix (b) HPC-SAmatrix.

system. These properties sustain the drug release from the matrixof the tablet for a longer period of time as compared to HPC-SAmatrix. In formulations with HPC-SA matrices, it was observed thatthe release of the drug was faster. HPC due to its water-solublenature dissolves faster and acts as a channeling agent. In ourcurrent study, we have used HPC EF, which has low viscosity inwater and therefore, it dissolves faster in aqueous media. In theHPC-SA matrix systems, two drug release mechanisms arepossible. One mechanism is that after coming in contact withaqueous media HPC EF are hydrated and swelled resulting information of a hydrogel through which the dissolved drug diffusesand transfers into the dissolution media. Secondly, dissolution ofHPC in dissolution media results in a further increase in drugrelease. Also, due to the hydroxyl ion (from pH 6.8 media) fumaricacid is dissolved resulting in more pore formation in the HPCmatrix as compared to EC-SA matrices. Drug release occurs bymixed mechanisms of diffusion and erosion of the matrix(Espinoza et al., 2000; Siepe et al., 2006). Thus, the HPC-SAsystem retarded drug release (<10 h) as compared to EC-SAmatrices.

The in vitro study also revealed that the release of drug wasretarded with the proportional increase in the polymer and binderconcentration such as in the case of the EC-SA matrix (Fig. 9a). Invitro drug release decreased from 90% to 64% and 71.9% (F3 to F4and F5) with the increase in polymer or lipid concentrationrespectively. In formulation F6, where the percentage of both ECand SA in the tablet composition was increased, the drug releasefurther reduced to 58% at the end of 24 h of dissolution study. Onthe other hand, the increase in the percentage of HPC and SA informulation F10 resulted in faster drug release. Complete drugrelease was obtained within 6.5 h for formulation F10, whereas itwas 9.5 h for formulation F7 (Fig. 9b). The drug release mechanismseen here was the combined effect of diffusion and erosionprocesses due to the dissolution of HPC and fumaric acid.

3.7. Similarity and dissimilarity

Similarity factor (f2) and dissimilarity factor (f1) were calculatedusing the release profiles of the formulations in pH 1.2 and 6.8 tosee the effect of fumaric acid, pH of surrounding media, and type ofpolymer used in the formulations.

As shown in Table 4, f2 value for formulations F1 and F2 in pH 1.2and 6.8 media were 40 and 29 and, the f1 values were 36 and 49,respectively. These results indicate that the release profile offormulation F1 and F2 was not similar as the f1 value is less than 50and f2 value is more than 15. These data indicate that formulationF1 and F2 exhibited pH-dependent release profiles and thereforeshows different profiles as change in the pH of the dissolutionmedia. This research group also studied the effect of addition of

Fig. 8. Effect of type of polymers (ethyl cellulose and hydroxypropyl cellulose) usedin the formulations on drug release profile by two step dissolution method.

27

Fig. 9. Effect of concentration of polymers and binder (a) ethyl cellulose (37% and40%) (b) hydroxypropyl cellulose (37% and 40%) used in the formulation on releaseof OND by two step dissolution method.

Table 4Similarity and dissimilarity factor analysis.

Formulations Similarity factor Dissimilarity factor

F1 (pH 1.2 and 6.8) 40 36F2 (pH 1.2 and 6.8) 29 49F3 (pH 1.2 and 6.8) 74 6F7 (pH 1.2 and 6.8) 44 25F3/F7 41 36

40 H. Patil et al. / International Journal of Pharmaceutics 496 (2015) 33–41

fumaric acid on the similarity and dissimilarity factor. f2 values forformulations containing fumaric acid such as F3 and F7 in pH 1.2and 6.8 were 74 and 44, whereas f1 values were 6 and 25,respectively, indicating similarity in the dissolution profile forformulation F3, whereas slight dissimilarity in F7. The dissolutionprofiles and f1/f2 values for formulation F3 confirms that it exhibitspH-independent solubility and pH changes do not influencedissolution of this formulation.

Table 5Mathematical modeling and release kinetics of OND from the prepared formulations (

Formulation Zero order plot correlation coeff.(R2)

First order plot correlation coef(R2)

F1 0.9533 0.1383F2 0.9249 0.2162F3 0.9386 0.6328F4 0.9776 0.2201F5 0.9617 0.3551F6 0.9721 0.1789F7 0.7938 0.5678F8 0.765 0.6615F9 0.7243 0.3957F10 0.7221 0.4745

Use of different polymers in the formulation compositionillustrated different dissolution profiles. As shown in Table 4, f2value for formulations F3/F7 was 41 and f1 value was 36 confirmingthe differences in the dissolution profiles.

3.8. Determination of drug release kinetics

In order to describe the release behavior of OND from differentformulations, the dissolution profiles were analyzed according tokinetic equations such as zero-order, first-order, Higuchi andKorsmeyer–Peppas. The regression coefficient values of differentrelease kinetic equations for all developed formulations werecompared as shown in Table 5. All batches containing EC incombination with stearic acid (F3–F6), showed a very goodcorrelation to the Higuchi equation (r2 = 0.9902–0.9985). TheHiguchi plot showed high linearity in comparison to other releasekinetic equations, which indicates that the drug release is heavilygoverned by the diffusion process. To confirm the diffusionalmechanism, the data were fitted into Korsmeyer–Peppas equation.As shown in Table 4, release exponent (n) found for batches F3–F6was less than 0.5. Release exponent n < 5 is considered consistentwith a diffusion-controlled release, whereas values of n between0.5 and 1 indicates non-Fickian (anomalous) release mechanisms.

The mechanism of drug release through matrices containingwater soluble and swellable polymers in combination withlipophilic binders is a more complex process. Formulationscontaining HPC in combination with stearic acid (F7–F10)demonstrated a good fit to the Korsmeyer–Peppas equation,indicating combined effects of diffusion and erosion mechanismsfor drug release. Moreover, the release exponent n was within therange of 0.50–0.89, indicating a non-Fickian diffusion mechanismand that drug release was governed by both diffusion and matrixerosion. Thus, drug release was controlled by more than oneprocess in the case of the HPC-SA matrices.

4. Conclusion

It can be concluded from this study that the continuous meltgranulation technique within a twin-screw extruder is a viablemethod to develop a sustained release tablet of OND. The granulesprepared with the melt granulation binder (stearic acid) exhibitedgood flowability as well as good compressibility with a feweramount of fines. Stearic acid in combination with EC (F3)demonstrated prolonged release of OND for 24 h with 90% drugrelease, whereas stearic acid in combination with HPC (F7) showed100% drug release over a period of 9 h. The incorporation of fumaricacid in the formulation led to the improvement of OND releasefrom both EC-SA and HPC-SA systems. Thus, this work formulated a

F1–F10).

f. Higuchi’s plot correlation coeff.(R2)

Korsmeyer–Peppas plot

correlation coeff.(R2)

Diffusional exponent(n)

0.9897 0.9889 0.420.9939 0.9929 0.430.9973 0.9985 0.460.9914 0.9908 0.390.9992 0.9959 0.450.9964 0.9902 0.400.9215 0.9989 0.650.9029 0.9991 0.630.8808 0.9997 0.560.8792 0.9987 0.57

28

H. Patil et al. / International Journal of Pharmaceutics 496 (2015) 33–41 41

sustained release OND matrix tablet that was not influenced by pHin dissolution testing. In conclusion, the novel aspect of the currentstudy is its demonstration of the feasibility to continuouslymanufacture ready-to-compress melt granules of OND by atwin-screw extruder that can be processed into tablets to providepH-dependent/pH-independent sustained release of the drug.

References

Aoki, H., Iwao, Y., Uchimoto, T., Noguchi, S., Kajihara, R., Takahashi, K., Ishida, M.,Terada, Y., Suzuki, Y., Itai, S., 2015. Fine granules showing sustained drug releaseprepared by high-shear melt granulation using triglycerin full behenate andmilled microcrystalline cellulose. Int. J. Pharm. 478, 530–539.

Bose, A., Wong, T.W., Singh, N., 2013. Formulation development and optimization ofsustained release matrix tablet of itopride HCl by response surfacemethodology and its evaluation of release kinetics. Saudi Pharm. J. 21, 201–213.

Campisi, B., Vojnovic, D., Chicco, D., Phan-Tan-Luu, R., 1999. Melt granulation in ahigh shear mixer: optimization of mixture and process variables using acombined experimental design. Chemom. Intell. Lab. Syst. 48, 59–70.

Corti, G., Cirri, M., Maestrelli, F., Mennini, N., Mura, P., 2008. Sustained-releasematrix tablets of metformin hydrochloride in combination with triacetyl-beta-cyclodextrin. Eur. J. Pharm. Biopharm. 68, 303–309.

Dvorackova, K., Dolezel, P., Maskova, E., Muselik, J., Kejdusova, M., Vetchy, D., 2013.The effect of acid pH modifiers on the release characteristics of weakly basicdrug from hydrophlilic–lipophilic matrices. AAPS PharmSciTech 14, 1341–1348.

Espinoza, R., Hong, E., Villafuerte, L., 2000. Influence of admixed citric acid on therelease profile of pelanserin hydrochloride from HPMC matrix tablets. Int. J.Pharm. 201, 165–173.

Gabr, K., 1992. Effect of organic acids on the release patterns of weakly basic drugsfrom inert sustained release matrix tablets. Eur. J. Pharm. Biopharm. 38, 199–202.

Glare, P.A., Dunwoodie, D., Clark, K., Ward, A., Yates, P., Ryan, S., Hardy, J.R., 2008.Treatment of nausea and vomiting in terminally ill cancer patients. Drugs 68,2575–2590.

Gungor, S., Okyar, A., Erturk-Toker, S., Baktir, G., Ozsoy, Y., 2010. Ondansetron-loadedbiodegradable microspheres as a nasal sustained delivery system: in vitro/invivo studies. Pharm. Dev. Technol. 15, 258–265.

Harris, D.G., 2010. Nausea and vomiting in advanced cancer. Br. Med. Bull. 96, 175–185.

He, W., Wu, M., Huang, S., Yin, L., 2014. Matrix tablets for sustained release ofrepaglinide: preparation, pharmacokinetics and hypoglycemic activity inbeagle dogs. Int. J. Pharm. 478, 297–307.

Hesketh, P.J., Gandara, D.R., 1991. Serotonin antagonists: a new class of antiemeticagents. J. Natl. Cancer Inst. 83, 613–620.

Higuchi, T., 1963. Mechanism of sustained-action medication. Theoretical analysisof rate of release of solid drugs dispersed in solid matrices. J. Pharm. Sci. 52,1145–1149.

Korsmeyer, R.W., Gurny, R., Doelker, E., Buri, P., Peppas, N.A., 1983. Mechanisms ofsolute release from porous hydrophilic polymers. Int. J. Pharm. 15, 25–35.

Liu, J., Zhang, F., McGinity, J.W., 2001. Properties of lipophilic matrix tabletscontaining phenylpropanolamine hydrochloride prepared by hot-meltextrusion. Eur. J. Pharm. 52, 181–190.

Maniruzzaman, M., Boateng, J.S., Snowden, M.J., Douroumis, D., 2012. A review ofhot-melt extrusion: process technology to pharmaceutical products. ISRNPharm. 2012, 436763.

McManis, P.G., Talley, N.J., 1997. Nausea and vomiting associated with selectiveserotonin reuptake inhibitors. CNS Drugs 8, 394–401.

Moore, J.W., Flanner, H.H., 1996. Mathematical comparison of dissolution profiles.Pharm. Technol. 20, 64–74.

Patil, H., Kulkarni, V., Majumdar, S., Repka, M.A., 2014. Continuous manufacturing ofsolid lipid nanoparticles by hot melt extrusion. Int. J. Pharm. 471, 153–156.

Perissutti, B., Rubessa, F., Moneghini, M., Voinovich, D., 2003. Formulation design ofcarbamazepine fast-release tablets prepared by melt granulation technique. Int.J. Pharm. 256, 53–63.

Pillay, V., Fassihi, R., 1998. Evaluation and comparison of dissolution data derivedfrom different modified release dosage forms: an alternative method. J.Controlled Release 55, 45–55.

Repka, M.A., Battu, S.K., Upadhye, S.B., Thumma, S., Crowley, M.M., Zhang, F., Martin,C., McGinity, J.W., 2007. Pharmaceutical applications of hot-melt extrusion: partII. Drug Dev. Ind. Pharm. 33, 1043–1057.

Rojanasthien, N., Manorot, M., Kumsorn, B., Nawoot, S., Teekachunhatean, S.,Sangdee, C., Apisariyakul, A., 1999. Pharmacokinetics and bioavailability studiesof generic ondansetron and the innovator preparation, in healthy Thai malevolunteers. J. Med. Assoc. Thai. 82, 713–720.

Siepe, S., Herrmann, W., Borchert, H.H., Lueckel, B., Kramer, A., Ries, A., Gurny, R.,2006. Microenvironmental pH and microviscosity inside pH-controlled matrixtablets: an EPR imaging study. J. Controlled Release 112, 72–78.

Streubel, A., Siepmann, J., Dashevsky, A., Bodmeier, R., 2000. pH-independentrelease of a weakly basic drug from water-insoluble and -soluble matrix tablets.J. Controlled Release 67, 101–110.

Tan, D.C., Chin, W.W., Tan, E.H., Hong, S., Gu, W., Gokhale, R., 2014. Effect of binderson the release rates of direct molded verapamil tablets using twin-screwextruder in melt granulation. Int. J. Pharm. 463, 89–97.

Venkatesh, G., Lai J.W., Vyas, N.H., Purohit, V., 2009. Drug delivery systemscomprising weakly basic drugs and organic acids. Google Patents.

Voinovich, D., Campisi, B., Moneghini, M., Vincenzi, C., Phan-Tan-Luu, R., 1999.Screening of high shear mixer melt granulation process variables using anasymmetrical factorial design. Int. J. Pharm. 190, 73–81.

Voinovich, D., Moneghini, M., Perissutti, B., Franceschinis, E., 2001. Meltpelletization in high shear mixer using a hydrophobic melt binder: influence ofsome apparatus and process variables. Eur. J. Pharm. Biopharm. 52, 305–313.

Warr, D.G., 2008. Chemotherapy-and cancer-related nausea and vomiting. Curr.Oncol. 15, S4–9.

Wells, J.I., 1988. Pharmaceutical Preformulation: The Physicochemical Properties ofDrug Substances. Ellis Horwood, Chichester.

Yang, D., Kulkarni, R., Behme, R.J., Kotiyan, P.N., 2007. Effect of the melt granulationtechnique on the dissolution characteristics of griseofulvin. Int. J. Pharm. 329,72–80.

Ye, J.H., Ponnudurai, R., Schaefer, R., 2001. Ondansetron: a selective 5-HT(3) receptorantagonist and its applications in CNS-related disorders. CNS Drug Rev. 7, 199–213.

29