Nasal powders of thalidomide for local treatment of nose bleeding in persons affected by hereditary hemorrhagic telangiectasia

Nasal

powders

of

thalidomide

for

local

treatment

of

nose

bleeding

in

persons

affected

by

hereditary

hemorrhagic

telangiectasia

G.

Colombo

a

,

F.

Bortolotti

a

,

V.

Chiapponi

b

,

F.

Buttini

b

,

F.

Sonvico

b

,

R.

Invernizzi

c

,

F.

Quaglia

c

,

C.

Danesino

d

,

F.

Pagella

e

,

P.

Russo

f

,

R.

Bettini

b

,

P.

Colombo

g

,

A.

Rossi

b,

*

a

DepartmentofLifeSciencesandBiotechnology,UniversityofFerrara,ViaFossatodiMortara17/19,44121Ferrara,Italy

b

DepartmentofPharmacy,UniversityofParma,ParcoAreadelleScienze27/A,43124Parma,Italy

c

DepartmentofInternalMedicine,UniversityofPavia,IRCCSFondazionePoliclinicoSanMatteo,PiazzaleGolgi19,27100Pavia,Italy

d

DepartmentofMolecularMedicine,GeneralBiologyandMedicalGeneticsUnit,UniversityofPavia,ViaForlanini6,27100Pavia,Italy

e_{DepartmentofOtorhinolaryngology,IRCCSFondazionePoliclinicoSanMatteo,PiazzaleGolgi19,27100Pavia,Italy} f_{DepartmentofPharmacy,UniversityofSalerno,viaGiovanniPaoloII132,84084Fisciano,SA,Italy}

g

PlumeStarssrl,StradaG.Inzani1,43125Parma,Italy

ARTICLE INFO Articlehistory: Received10May2016

Receivedinrevisedform30June2016 Accepted1July2016

Chemicalcompoundsstudiedinthisarticle: Thalidomide(PubChemCID:5426)

b-cyclodextrin(PubChemCID:444041) Hydroxypropyl-b-cyclodextrin(PubChem CID:44134771)

Sulphobutylether-b-cyclodextrin (PubChemCID:91886200) Lecithin(PubChemCID:57369748) Keywords: Thalidomide Telangiectasia Nasalpowders Cyclodextrins Mucoadhesion Nasaldevices ABSTRACT

Inthisworknasalpowderformulationsofthalidomideweredesignedandstudiedtobeusedbypersons affectedbyhereditaryhemorrhagictelangiectasiaasacomplementaryanti-epistaxistherapy,withthe goalofsustainingtheeffectobtained withthalidomideoraltreatmentafter itsdiscontinuationfor adverseeffects.Threenasalpowderswerepreparedusingascarriersb-CDoritsmorehydrophilic derivatives such as hydropropyl-b-CD and sulphobutylether-b-CD and tested with respect to technological andbiopharmaceutical featuresafter emissionwithactive andpassivenasalpowder devices.Forallformulatedpowders,improveddissolutionratewasfoundcomparedtothatoftheraw material,makingthalidomidepromptlyavailableinthenasalenvironmentataconcentrationfavouring anaccumulationinthemucosa.Theverylimitedtransmucosaltransportmeasuredinvitrosuggestsalow likelihoodofsignificantsystemicabsorption.Thetopicalactiononbleedingcouldbenefitfromthepoor absorptionandfromthefactthatabout2–3%ofthethalidomideappliedonthenasalmucosawas accumulatedwithinthetissue,particularlywiththeb-CDnasalpowder.

ã2016TheAuthors.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense (http://creativecommons.org/licenses/by/4.0/).

1.Introduction

Hereditaryhemorrhagictelangiectasia(HHT)isararegenetic diseasecharacterizedbyvascularmalformationswithaprevalence ofabout1–2in10,000individuals(GovaniandShovlin,2009).HHT is caused by mutations in genes encoding proteins mostly expressedinvascularendothelialcells.Theseproteinsareinvolved in transforming the growth factor-beta superfamily signalling

pathway.Clinicalmanifestationsincludearteriovenous malforma-tions (AVMs), which can be classi_fied as small telangiectases (regionsofcapillarydilation)andlargeAVMs,wherearteriolesand venulesaredirectlyconnectedwithnocapillariesinbetween.In particular,smalltelangiectasesaretypicalofthenasalseptum,oral mucosa and gastrointestinal tract, whereas large AVMs mainly occurinthelung,brainandliver.Astheselesionsareduenotonly toanexpansionofthevascularlumen,butalsotothinningofthe vascular wall, theycause chronic nasal and/or gastrointestinal bleeding and severeanaemia.Indeed, recurrentepistaxisis the mostcommonmanifestationofnasalHHTandbeginsbytheageof 10–20yearsinmanypatients,increasinginseveritywithageing. Epistaxismaybesosevereastorequirebloodtransfusionsandoral ironsupplementation(Guttmacheretal.,1995).

Abbreviations:AVMs,arteriovenousmalformations;HHT,hereditary hemor-rhagictelangiectasia;SNES,simulatednasalelectrolytesolution;THAL, thalido-mide.

* Correspondingauthor.

E-mailaddress:[email protected](A. Rossi).

http://dx.doi.org/10.1016/j.ijpharm.2016.07.002

0378-5173/ã2016TheAuthors.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/).

ContentslistsavailableatScienceDirect

International

Journal

of

Pharmaceutics

Treatmentsforepistaxisincludesurgicalreplacementofnasal epitheliumwithskin,laser ablation,topicalapplicationof anti-inflammatory drugs or manipulation of the coagulation and fibrinolyticpathways (Guttmacheret al.,1995).However,these optionsdo not permanentlyresolve bleeding and are not well accepted by patients. Recent in vivo studies showed that oral administrationofthalidomidereducedthefrequencyandduration ofnosebleedsinHHTpatients(Invernizzietal.,2015;Lebrinetal., 2010). The anti-angiogenic and anti-hemorrhagic activities of thalidomideareduenotonlytoadirectinhibitionofendothelial cellproliferationandmigration,butalsotoanincreaseinmural cellcoverage,whichstabilizesbloodvesselsandpreventsbleeding. However, the observed anti-epistaxis effect of thalidomide is reversibleand nosebleeding recursafterdiscontinuationofthe oraltreatment.Thus,systemicadministrationmustberesumed,at theriskofcausingadverseeffects,suchasreversibleneuropathy, constipationandsedation(vonMoosetal.,2003).

WhenHHTinvolvesthenasalmucosa,thetopical administra-tion of thalidomide inside the nasal cavity for a maintenance treatment, complementing the systemic administration, is a therapeutic hypothesis. Thalidomide is practically insoluble in water and undergoes rapid and spontaneous hydrolysis in an aqueousenvironment(Goosenetal.,2002; Pereiraetal.,2013). Therefore,thalidomideshouldbeformulatedasdry powderfor insufflation, to increase physico-chemical and microbiological stability(VidgrenandKublik,1998)aswellasdrugconcentration onandcontacttimewiththenasalmucosa(Colomboetal.,2011; Jiangetal.,2010).

However,theperformanceofapowderformulationfornasal deliverywillbehighlydependentondrugdissolutionrateinthe liquidliningthenasalmucosa.Complexationwith

b

-cyclodextrins (

b

-CDs) hasbeen described as increasingthalidomide aqueous solubility(Kratzetal.,2012;Krennetal.,1992).

b

-CDhasprovento beanexcellentsolubilityandabsorptionenhancerinnasaldrug delivery(Challaetal.,2005;DeAscentiisetal.,1996;Merkusetal., 1999),beingthemostusedinpharmaceuticalapplicationssinceits cavitysizeissuitableformanydrugs.Some

b

-CDderivativesused intopical drugformulations, suchashydroxypropyl-

b

-CD

(HP-b

-CD)andsulphobutylether-

b

-CD(SBE-

b

-CD),havehigherwater solubilityandabettertoxicologicalprofilethan

b

-CD(Badr-Eldin etal.,2008;deAraujoetal.,2008;Fukudaetal.,2008;Loftssonand Masson,2001;Marttinetal.,1998).

Hence, in this work novel nasal powder formulations of thalidomidewerestudiedtoenablethalidomideadministration inthenoseandprovideHHTpatientstreatedbyoralthalidomide

withacomplementaryanti-epistaxistherapy.Theultimategoalis tosustain theeffect obtainedwith theoral treatment after its interruption.Threepowderswerepreparedusingascarriers

b

-CD orthemorehydrophilicHP-

b

-CDandSBE-

b

-CD,aimingtoidentify the bestperforming thalidomidenasal powderwith respectto technological and biopharmaceutical features. In particular, thalidomidedissolutionandtransportacrossrabbitnasalmucosa andpowdermucoadhesionwereinvestigated.Then,asnasaldrug administration requires a device for dosedelivery, the powder formulationswerecombinedwithdifferentdevicestodetermine theplumeemissionanddepositionprofileinamodelcastofthe humannose.Studyingtheeffectofthalidomidenasal administra-tioninHHTinhumanswasoutofourscopehereandwillbethe subjectofafuturework.

2.Materialsandmethods 2.1.Materials

Thalidomide was purchased from Olon S.p.A. (Milan, Italy). PhenacetinwasobtainedfromCarloErba(Milan,Italy).

b

-cyclo-dextrin and hydroxypropyl-

b

-cyclodextrin were supplied by Roquette (Lestrem, France). Sulphobutylether-

b

-cyclodextrin was kindly donated by CYDEX Pharmaceuticals (Lawrence, KS, USA).Lecithin(LIPOIDS45)wassuppliedbyLipoidAG (Steinhau-sen,Switzerland).HPLC-gradeacetonitrilewasobtainedfromVWR InternationalS.r.l. (Milan,Italy).Allotherreagentsand solvents usedwereanalyticalgrade.

2.2.Methods 2.2.1.HPLCassay

Thalidomide quantificationwas performed by reversephase highperformance liquid chromatography(HPLC)using thetwo isocraticmethodsreportedinTable1.Fordrugquantificationin powdersandinphasesolubilityanddissolutionratesamples,the methodcomplied with theUSP 34 Thalidomide Official Mono-graph.SystemsuitabilitywasassessedaccordingtoUSP34.

Themethodfordeterminingthalidomideconcentrationinthe transport and mucoadhesion experiments was developed in-house.Thismethodwas validatedfor repeatability(RSD0.053% and0.063%,forareaandretentiontimevaluesrespectively;n=6 injections)andlimitsofquantificationanddetection(LOQ0.65ng/ ml;LOD0.22ng/ml;n=3injections).

Table1

ConditionsoftheHPLCmethodsused. Conditions Method1 (USP34)

Method2 (in-house) Apparatus ChromatopacLC-10ASpump,ESAtemperature-controlledautosampler

andSPD-10UVdetector (Shimadzu,Kyoto,Japan)

Agilent1100serieswithauto-samplerandUV–visdetector(Agilent, SantaClara,CA,USA)

Stationaryphase C18SynergiHydro4mmcolumn(150
3.9mm)a

C18Luna3mmcolumn (150
4.6mm)a

Mobilephase Water:acetonitrile:H3PO4

85:15:0.1(v/v)

50mMNaH2PO4:Acetonitrile

60:40(v/v);pH3 Internalstandard(IS) Phenacetin –

Flowrate(ml/min) 1 0.6

Detectionwavelength (nm) 237 220 Temperature(_{C) 25}_{C Ambient} Injectionvolume(ml) 50 40 Thalidomideretention time(min) 18(24forIS) 4.8 a

2.2.2.Phasesolubilitystudies

Phase solubility studies of thalidomide in the presence of cyclodextrinswerecarriedoutinphosphatebuffersolution0.01M pH5.0.Excessamountsofthalidomide(about50mg)wereadded to25mlofbuffercontainingincreasingconcentrationsof cyclo-dextrins.Dispersionsweremagneticallystirredfor5h,at room temperature,thenclearedthrougha0.45

m

mmembranefilterand assayedforthalidomideconcentrationwiththeHPLCMethod1 (Table1).

2.2.3.Nasalpowdermanufacturing

Thalidomide-cyclodextrinpowderswereproducedbya knead-ingtechnique.2.5gofthalidomideand7.0gofcyclodextrinraw materialswere blended in a Turbula1 mixer (Type T2A; WAB, Muttenz,Switzerland).Theblendwastransferredtoamortarand kneadedwithasolutionof0.5glecithinin15mlofethanol.The wetmassinthemortarwasovendriedat40_{Cuptocomplete} solvent evaporation. The dried product was forced through a 0.4mmmeshsieve.Thisprocedureallowedforthepreparationof threenasalpowderscontainingthalidomide,lecithinandoneof thethreecyclodextrins(

b

-CD,HP-

b

-CDorSBE-

b

-CD)intheratio of25:5:70(w/w).

2.2.4.Powderphysico-chemicalcharacterization

DifferentialScanning Calorimetry (DSC)measurements were performedonanindiumcalibratedDSC821e_{calorimeter(Mettler}

Toledo S.p.A., Milan, Italy). Accurately weighed samples (about 5mg) were placed in aluminium pans, crimped and pierced. Sampleswereheatedata rate of 10C/min,from30 to300C, undera100ml/minflowof drynitrogen.Theresulting thermo-gramswereanalyzedwithSTAReSoftwaretoassessdrug/excipient interactions.

PowderX-raydiffractometry (PXRD)patternswere recorded on a PW1050 diffractometer (Philips Analytical, Almelo, The Netherlands),equippedwithacurvedgraphitecrystal,usingCuK

a

radiation.Thescanningrateemployedwas0.5min1overa2

u

rangeof2–50_.

Morphology of powders was analyzed byscanning electron microscopy(SEM)(SigmaHD,CarlZeiss,Oberkochen,Germany)at EHT 1.00kV. A double-sided adhesive tape was placed on an aluminiumstubandaminimalamountofpowderwasdispersed by gently tapping on the edge of the stub. The samples were analyzedatdifferentmagnificationsafter30minof depressuriza-tion.

Particle size distribution of powders was measured using laserlightdiffractionapparatus(Spraytec,MalvernInstruments Ltd., Malvern, UK), by suspending powder aliquots in a 0.1% solution (w/v) of Span 85 in cyclohexane. The analysis was performed in a 100ml cell under magnetic stirring, using a 300mm lens.

Bulk and tapped volumes were determined with a 25ml graduatedcylinderwith0.5mlgraduationmarksaccordingtoPh. Eur.,onpowdersamplesweighingabout8–8.5g.Thebulkvolume wasreadafteragentlemanualtappingofthecylinder.Thetapped volume was measured with a Tapped Density Tester (Erweka GmbH,Heusenstamm,Germany)aftertappingsequencesof1250 taps.Frombulkand tappedvolumes, theCompressibilityIndex (or Carr Index, CI) was calculated according to the following formula:

CI¼100ð

n

0

n

fÞ=

n

0

wherevoisthebulkvolumeandvfisthetappedvolume.

The static angle of repose was determined using the glass funneltechniqueaccordingtothePh.Eur.8thEd.

2.2.5.Biopharmaceuticalcharacterization

2.2.5.1.Invitrodissolutionrate. Thalidomidedissolutionratewas measured using a USP Apparatus II dissolution tester (Erweka DT6R; Erweka GmbH, Heusenstamm, Germany). Samples of thalidomide-cyclodextrin powders, equivalent to about 5–6mg thalidomide,werespreadonthesurfaceofthedissolutionmedium (900mlofPBS0.01MpH5.0)at370.5Candstirredat100rpm. At fixed time points, samples filtered through a 0.45

m

m membrane filter were analyzed for thalidomide concentration byHPLC.

2.2.5.2.Invitrotransportacrossexcisednasalmucosa. Thalidomide transport across excised rabbit nasal mucosa from the nasal powderswasstudiedinvitrousingverticalFranz-typediffusion cells(diffusionarea:0.58cm2_{;Vetrotecnica,Padova,Italy).Rabbit}

nasalmucosawasextractedfromrabbitheadssuppliedbyalocal slaughterhouse(PolaS.r.l.,FinaleEmilia,MO,Italy)onthedayof theexperimentandtransportedtothelabinarefrigeratedbox.The extraction procedure, completed within 2h from the animal’s death,hasbeendescribedelsewhere(Balduccietal.,2013).Franz cells were assembled withthetissue’s mucosalside facingthe donor.Thereceptorcompartmentwasfilledwith5mlofasolution composed of acetate buffer pH5.0 and PEG 400(50:50v/v) to assuresinkconditions.Theassembledcellwaslefttoequilibrate for15minat37Cbeforeintroducingtheformulation.

Asaturatedsolutionofthalidomidein50mMsodiumacetate bufferpH5.0wasusedasreference.Onemlofthissolutionor3mg of nasalpowder (corresponding to750

m

g of thalidomide)was introducedinthedonor.Withthepowder,100

m

lofacetatebuffer pH5.0wasaddedtothedonorforpowderdissolution.Thecellwas protected from light during the experiment, which lasted 4h (Bortolottietal.,2009).Thisexperimentdurationwasconsistent withviabilitydataintheliteratureforanimalexcisednasaltissue inthecourseofinvitrostudies:3–4hforbovinenasalmucosaand upto10hoursforisolatedrabbitnasalmucosawithoxygenation (Bechgaardetal.,1992;Schmidtetal.,2000).Thereceptorsolution wassampledonlyattheendoftheexperiment.

The residual thalidomide in the donor compartment was quantitatively recovered with aliquots of mobile phase for subsequent HPLC analysis (Method 2 in Table 1). Thalidomide accumulatingwithinthetissuethicknesswasthenrecoveredby comminutingthemucosawithasurgicalbladeandhomogenizing in5mlofwaterwithUltra-Turrax1IKA(T10basicmodel,IKA1 -WerkeGmbH&Co.KG,Staufen,Germany)for3min.Twomilliliters ofmethanolwerethenaddedandhomogenizationcontinuedfora further 30s to disrupt the cell membranes and dissolve the thalidomide.Thehomogenatewascentrifugedandthe superna-tant injected as such. Thalidomide amounts recovered from receptor and donor compartments and the amount extracted from the mucosa allowed for calculating the mass balance (expressed as % of the thalidomide loaded in the donor). The numberofreplicatesforthesestudieswasbetween6and10. 2.2.5.3.Invitromucoadhesionstudieswithrabbitnasalmucosa. The biological material for these experiments was supplied to and handledatthelabasdescribedbefore.Thenasalseptumwiththe mucosalayeringonthecartilagewasextractedfromtherabbit’s headandusedasasurfaceforpowderdeposition.Theseptumwith the lining mucosawas flat and rectangular (area 4.90.8cm2; meanS.D.,n=7).

TheseptuminaPetridishwasintroducedintoaglassdesiccator atarelativehumidityof75–76%atroomtemperature(Wexlerand Hasegawa, 1954). About 20mg of thalidomide-cyclodextrin powder formulation were manually sprinkled on the whole available mucosa surface.The dish wascovered and leftin the

desiccatorfor 20min.ThreeFalconTM_{50mlconical tubes were}

each filled with 15ml of simulated nasal electrolyte solution (SNES)atroomtemperature.SNEScontained8.77mg/mlsodium chloride,2.98mg/mlpotassiumchlorideand0.59mg/mlcalcium chloride dehydrate (Castile et al., 2013). 20min after powder deposition,thenasalseptumwas pinchedwithtweezersonits shorter side. Holding it firmly, the septum was continuously submergedandwithdrawnfromtheliquidinthetubefor1min with a frequency of 1s in and 1s out. The procedure was repeatedwiththesecondandthirdtubeforatotal“washingtime” of3min.

Thesuspensionsinthethreetubeswerecentrifugedtoseparate the solid (detached powder) from the supernatant (8000rpm, 40min).Thesedimentsweretreatedwithmethanoltoquantify thalidomideinthepowderphysicallydetachedfromthemucosa bythewashingprocedure,i.e., thenonadheredthalidomideas solid. The supernatantswere alsoanalyzed by HPLCMethod 2 (Table1)toquantifythefractionofthalidomidesolubilizedbythe SNES,i.e., thenonadheredthalidomideas solution.Finally,the mucosa with the residual powder attached was treated with methanol to dissolve thalidomide for quantification (adhered thalidomide).

Mucoadhesionwas expressedaspercentof adhered thalido-mide versusthe total thalidomide recovered. For each powder formulation,atleast3replicateswerecarriedout.

2.2.6.Powderanddevicecombination(deliveryanddepositionin nasalcast)

Theamountofnasalpowderemittedupondeviceactuationwas measured and pictures of the delivery sequence were taken (Buttini et al., 2012; Russo et al., 2004). Three devices were employed:oneactivesingledose(MIAT,Milan,Italy),oneactive multi-dose(TeijinLtd.,Tokyo,Japan)andapassivebi-dose(Aptar, LouveciennesCedex,France)device.MIATandAptardevicesare designedfordeliveringadoseofpowderpre-meteredinasize3 capsuleandinabi-doseblisterreservoirrespectively.Bothwere loadedwithapproximately20mgofpowder,correspondingtoa 5-mgthalidomidedose.TheTeijindevicehasaninternalreservoirfor thepowderbulkandameteringchamber.MIATandTeijindevices haveasqueezablerubberbulbmanuallyactuatedbyanoperatorto generatetheairflow.Since theAptardevice isbreath-actuated, 15L/minairflowwasdrawnthroughitfor2susingavacuumpump toextractthepowderfromthereservoir.Quantitativedeliveryof thepowderfromtheinsufflatorwasdeterminedbyweighingthe insufflatorbeforeandaftereachactuation.Sequencesofimagesof thepowderplume duringthe actuationwererecordedusing a videocamera(PanasonicHDC-TM700,PanasonicCorp.,Kadoma, Japan).

Plumedepositioninahumannasalmodelwascarriedoutusing asiliconcast(TeijinPharmaLtd.Tokyo,Japan)(Bettinietal.,1999). Insuf_flationwiththesamethree devices wasperformedin one

nostriloftheassemblednasalcast.Thenasalcavitieswerethen separatedandpicturesweretaken.

2.2.7.Powderstabilitystudy

Thestabilityofthalidomide-cyclodextrinpowderswasstudied intheshortterm,namelyfor3monthsofstorageat25C/60%RH andfor1monthat40C/75%RH.About20mgofeachformulation, correspondingtoa5-mgthalidomidedose,werestoredinsize3 HPMC capsulesand testedatfixed time pointsfor thalidomide content,invitrodissolutionrateandpowderinsufflationusingthe single-doseMIATdevice.

2.2.8.Statisticalanalysis

Statistical analysis of data was performed by applying an unpairedtwo-tailedStudent’st-test.Significancewasacceptedatp values<0.05.

3.Resultsanddiscussion

Nasal powders of thalidomide have to induce a local anti-angiogenicaction onthemicrovasculature ofthenasal mucosa compromised by HHT. To exert a local effect or to have drug absorption,theformulationmustenabledrugdissolutioninthe small volume of liquid lining the mucosa. On site powder dissolutionleadstolocaldrugsaturationconcentration. Thalido-mideisveryslightlysolubleinwaterandtherawmaterialdoesnot weteasily.Thus,itwasdecidedtointroduceacyclodextrininthe formulationasasolubilityenhancer(Kaleetal.,2008;Kratzetal., 2012;Krennetal.,1992).

Inordertoestablishtheoptimalthalidomideto

b

-CD,HP-

b

-CD orSBE-

b

-CDratiointhenasalproduct,phasesolubilitystudies werecarriedoutinphosphatebufferpH5.0at25C.Thisbuffer wasusedassolventbecausethalidomideismorestableatacidic pH.Allthree

b

-CDsincreasedthalidomidesolubility,whichwas about42

m

g/mlintheabsenceofcyclodextrin.Thephasesolubility diagramsshowthatthalidomidesolubilityincreasedlinearlyasa function ofcyclodextrin concentration(Fig.1).The relationship betweencyclodextrinandthalidomideconcentrationsledtoanAL

-type diagram, which is typical of 1:1 complexes. Calculated stabilityconstantswere313M1,105M1and201M1for

b

-CD, HP-

b

-CDandSBE-

b

-CDrespectively.Thecomplexationefficiency of the three cyclodextrins, whose inverse value expresses the numberofCDmoleculesinsolutionthatforma1:1complexwith thalidomide,was0.05for

b

-CD,0.017forHP-

b

-CDand0.032for SBE-

b

-CD(Loftssonetal.,2005).

3.1.Manufacturingandphysico-chemicalcharacterizationof thalidomide/cyclodextrinpowders

b

-CD,HP-

b

-CDandSBE-

b

-CDwereusedfortheformulationof thenasalpowderbecausetheyallenhancedthalidomidesolubility

Fig.1.Phasesolubilitydiagramsofthalidomideinpresenceof(a)b-CD,(b)HP-b-CDand(c)SBE-b-CD(meanvaluesS.D.;n=3).Linearfitting:(a)y=0.16626+0.04830x (R2

=0.99),(b)y=0.19209+0.01669x(R2

=0.99)and(c)y=0.19313+0.03153x(R2

inaqueousmediumtodifferentextents.Thethreenasalpowders wereproducedbykneadingandcontained25%thalidomide(w/w) and70%cyclodextrin(w/w)withtheobjectiveofhaving5mgof drugin20mgofpowdertoinsufflate.Thismeantthatthemolar ratio between thalidomide and the CD was 1:0.64, 1:0.53 and 1:0.51 respectively for

b

-CD, HP-

b

-CD and SBE-

b

-CD. The formulationalsocontained5%(w/w)lecithin, whichwasadded duringkneadingasethanolsolutiontoimprovethewettabilityof thalidomide.

Inordertoverifywhetheracomplexwas formedduringthe kneading process, the powders were studied by differential scanningcalorimetryincomparisonwithcrystallinethalidomide (THAL)rawmaterial(Fig.2).TheTHALrawmaterialthermogram showedasharpendothermicpeakat276Ccorrespondingtothe drug melting point. The thermograms of the thalidomide/CD powders showed a broad endothermic band around 100C, correspondingtothereleaseofwaterfromthecyclodextrincavity. Themeltingpeakofthalidomidewasstillpresent,althoughmuch

smaller and shifted to lower temperatures, likely due to an interaction of the two compounds in the mixture. In fact, the enthalpyoffusionforthepuredrugwas158Jg1,whereasitwas 44.32Jg1,21.92Jg1and133.94Jg1respectivelyforTHAL/

b

-CD,THAL/HP-

b

-CDandTHAL/SBE-

b

-CD.Thislastvalueshould not be considered as fully reliable, being affected by the degradationpeak of SBE-

b

-cyclodextrin that superposedtothe THALmeltingpeak.

Additional evidence of an interaction occurring between thalidomide and cyclodextrins was obtained by PXRD analysis (SupplementaryMaterial).Thalidomiderawmaterialpeakswere sharpand intense,con_firming itscrystallinenature.Incontrast, thalidomidekneadedwiththecyclodextrinsshowedlessintense peaks.

Powderparticlesize,packingandflow arekeyproperties in view of nasal dosage form manufacturing, delivered dose uniformity, aerodynamicbehaviour during insufflation, dissolu-tionrateandinteractionwiththewetmucosa.Particlesizeanalysis by laser diffraction showed a mean volume diameter (Dv0.5)

between13and22

m

mforallpowders(Table2).Thissizewasalso confirmedbySEManalysis(Fig.3).Theparticleshapeofkneaded productsappearsasagglomerateswithsharpedges,resultingfrom the sieve comminution after drying. Nevertheless, this non sphericalmorphologydidnotnegativelyaffectpowderflowand packing. Bulk density and repose angle values of these nasal powdersallowedforclassifyingtheformulationscontaining

b

-CD and HP-

b

-CD as free-flowing according to the Pharmacopoeia (Table2).TheSBE-

b

-CDnasalpowderhadinferiorflowproperties andlooserpacking.

3.2.Nasalpowderstability

Forthalidomide-cyclodextrinpowdersstoredinsize3HPMC capsulesat25C/60%RHfor3monthsandat40C/75%RHfor1 month,

b

-CDpowderandHP-

b

-CDpowderremainedstableinall conditions with respecttothalidomide content.Similarly,

SBE-b

-CDpowderdidnotshowasignificantdecreaseinthalidomide

Fig.2.DSCthermogramsof(a)thalidomiderawmaterial,(b)b-CDnasalpowder, (c)HP-b-CDnasalpowderand(d)SBE-b-CDnasalpowder.

content,despiteitscolourturningtopaleyellowafterstorageat 40C. Stability was also confirmed with respect to in vitro dissolutionrateandpowderdelivery.

3.3.Biopharmaceuticalcharacterization

3.3.1.Invitrothalidomidenasalpowderdissolutionrate

Dissolutionrateofthalidomidenasalpowderswasmeasuredin vitroinanaqueousmediumatpH5.0over2h(Goosenetal.,2002; Pereiraetal.,2013;Schumacheretal.,1965).Theseexperiments werecarriedoutwiththepurposeofcomparingthedissolution behaviourofthenasalpowdersinsinkconditionsforthequality assessmentofsolidpreparations.Thalidomiderawmaterial,i.e., thesolid material before the formulation process,was used as reference.

In these conditions only 46.00.1% of thalidomide raw materialdissolvedin1h,reaching72.00.9%attheendofthe secondhour(Fig.4).Itwasobservedthatduringthefirstminute drugpowderfloatedonthedissolutionmediumsurface,owingto itsdifficultwettability.Incontrast,kneadedpowderscontaining cyclodextrins together with lecithin dissolved more rapidly, reaching841%,992%and932%respectivelyofthalidomide insolutionafter60min.forthepowdercontaining

b

-CD,HP-

b

-CD andSBE-

b

-CD.Thisimprovementindrugdissolutionratehadtwo main explanations. Firstly, solid state analysis of the powders showedthat thalidomideand cyclodextrins partiallyinteracted. Thisinteractionatthemolecularlevelincreasedthedrugsolubility inaqueousenvironment(asdemonstratedinthephasesolubility studies).Thesecondpositivecontributiontothalidomide dissolu-tionratecamefromlecithin,owingtoitssurface-activeproperties. Thepositiveeffectoflecithinonwettabilityincreasedthepowder surfaceinteractingwiththesolvent.

3.3.2.Invitrothalidomidetransportacrossrabbitnasalmucosa InvitroTHALtransportacrossexcisedrabbitnasalmucosaina 4-hexperimentwasnegligiblefromallpowders,aswellasfrom thethalidomidesaturatedsolution.Havingobservedthat trans-mucosal diffusion was very low, if not absent, the THAL concentrationinthereceptor compartmentwas measuredonly attheendoftheexperiment.THALconcentrationwasaboveLOQ onlyintheexperimentswiththethalidomidesaturatedsolution and thenasalpowdercontaining

b

-CD(0.090.10

m

g/5mland 0.130.09

m

g/5ml respectively). Independently of the formula-tion,morethan80%oftheloadedthalidomideremainedinthe donor.Thethalidomidepresentinthemucosaattheendofthe permeationexperimentwasextractedfromthetissuesampleused asbarrier.Theamountsrecoveredareshowninthebargraphof

Fig.5.

Anypossibledrugdegradationoccurringduringthepermeation wasassessedbydeterminingthethalidomidemassbalanceafter all transport experiments (Table 3). The mass balance for the powder formulations was higher than 90%, indicating that the experimentalvariabilityanddrugstabilitywereundercontrol.The massbalancebelow90%forthesaturatedsolutionhastotakeinto accountthattheamountofthalidomideloadedinthedonorwas muchlowerthanwiththepowders(65vs750

m

grespectively).

Insummary,thetransportexperimentsshowedthat thalido-midewaspoorlydiffusedacrossthenasalepithelium,remainingto acertainextentinthetissue.Thenegligibletransportacrossthe mucosa is a pivotal result with respect to topical application, becauseitreducesthelikelihoodofthalidomidebeingsignificantly absorbedfromthenasalcavityintothesystemiccirculation.Infull awareness that these in vitro data cannot substitute the bioavailabilitystudyfollowingnasaladministration,in anycase theyindicatethatinvivosystemicabsorptionmaynotbeclinically relevant. In fact, these datawere collectedafter 4hof contact betweenformulationandmucosa,whereasresidencetimeinside thenasalcavity,evenforasolidformulation,maynotbeaslong.

Table2

Particlesizeandpowderflowofthalidomide/cyclodextrinpowders(n=3). NasalPowder Dv0.1 (mm) Dv0.5 (mm) Dv0.9 (mm) BulkDensity (g/ml) TappedDensity (g/ml) Compressibility Index Angleofrepose b-CD 3.10.1 19.20.5 58.52.9 0.550.01 0.700.01 20.0 28.51.2 HP-b-CD 3.20.2 18.12.1 61.65.7 0.600.01 0.800.01 20.0 21.82.6 SBE-b-CD 3.50.3 21.23.5 61.110.3 0.480.01 0.600.01 23.3 34.60.7

Fig.4.Dissolutionprofilesof ()thalidomide raw material,(&)b-CDnasal powder,(4)HP-b-CDnasalpowderand(^)SBE-b-CDnasalpowder(meanS.D.; n=3).

Fig.5. Amountsofthalidomideextractedfromthemucosaattheendofthe4-h diffusionexperiment(thalidomideloadedattimezero:65mgforthesaturated solution,750mgforthepowders;meanS.D.,n=6–10).

Indeed,itmustbeborneinmindthatthedamaged microvascula-tureofthenasalmucosainHHTisthetargetofthelocalapplication ofthalidomide.Thus,thalidomideaccumulationwithinthetissue isfavourabletoitslocalaction.

Comparingthenasalpowderformulations,the

b

-CDpowder led to the greatest THAL accumulation within the tissue.

b

-cyclodextrinisknownasanabsorptionenhancerinnasaldrug delivery (Marttin et al.,1998). The mechanism underlying this actionhasbeenrelatedtotheinteractionwiththenasalepithelial membranes and transient opening of tight junctions. Several studiesindicatedifferentenhancingpropertiesof

b

-cyclodextrin andderivativesdependingonthesubstitutingmoiety.Ingeneral, methylated derivatives appeared more effective compared to hydroxypropyl-

b

-cyclodextrinwithrespecttonasalabsorptionof drugs(Merkusetal.,1999).Thehigherabsorptionenhancingeffect of

b

-cyclodextrincomparedtotheotherderivativeswasalsoseen fromthetracesofthalidomidedetectedinthereceptor compart-ment.

3.3.3.Mucoadhesion

Thalidomideinteractionwithcyclodextrinsimprovedpowder mucoadhesioncomparedtothedrugrawmaterialaspowder.In this work the mucoadhesionwas measuredby a methodology consistinginwashing,incontrolledconditionswitha simulated nasalfluid(SNES),asampleofnasalmucosaonwhichthenasal powderhadbeenpreviouslydeposited.Theresultsareexpressed asmeanvaluesstandarddeviationofreplicatescarriedoutby twooperators.Theobtainedvariabilityof theindividualdatais consistentwithareproducibleprocedure.

For the three formulations,the average percentage of THAL retainedonthemucosasurfaceafterthewashingprocedurewith thesimulatednasalfluid(ADHERED)andthedrugamountsnot retainedandrecovered intheremovedsolid ordissolved inthe fluid(NONADHERED)areshowninFig.6.Torevealapossibletime dependencyofmucoadhesion,theproductwasremovedfromthe

mucosabywashingthreetimesat1minintervals.Itwasfoundthat thelowerthethalidomidedissolutionrateinwater,thegreaterthe mucoadhesion.Indeed,50%ofthe

b

-cyclodextrinpowderwasstill retainedonthemucosasurfaceafterthewashing.Thiswasthe same powder enabling the highest thalidomide accumulation withinthemucosaltissue(seeFig.5).Asthemorehydrophilic

HP-b

-CDandSBE-

b

-CDincreasedwettabilityanddissolutionratein aqueousmedia,thecorrespondingnasalpowdersmoreeasilylost thecontactwiththemucosaandwereremoved. Therespective adhered amountswerenot significantlydifferent(p=0.16). The solid material physically detached from the mucosa remained dispersedordissolvedinthewashingliquidphase.Infact,between 4% (

b

-CD powder) and 16% (HP-

b

-CD powder) of removed thalidomide was found in solution. Thus, dissolution by the washingfluidwasmostlyrelevantforthemostsolubleHP-

b

-CD powder(p<0.05vsSBE-

b

-CDpowder).

Froma kinetic pointof view, mostpowderdetachment and dissolution occurred during the first minute of washing, in particularfortheHP-

b

-CDandSBE-

b

-CDpowders(805%and 828% respectively; p=0.23). Then, between 3% and 6% of thalidomide was washed away during the second and third minute, without significant differences betweentheabove two powders.Incontrast,atime-dependencyofmucoadhesionwasnot foundforthe

b

-CDpowder,whosetotalnonadheredfractionwas mainly recovered as dispersed solid in SNES (Supplementary material).

3.4.Powderandinsufflationdevicecombination

A nasal drug product is a typical combination product, comprisingformulationanddeviceforitsdelivery.Thechoiceof aproperdeviceisdrivenbythetherapyrequirementsandtarget depositionsite,aswellasthecharacteristicsoftheproducttobe administered. For the local action desired in HHT,thalidomide depositionshould coverthe maximummucosal surfacefor the longest possible time. Moreover, the device should effectively deliverthepowder,whilelimitingitsdispersioninthe environ-ment.Inthisregard,inthecaseofthalidomideapassivedevice maybepreferable.

Inordertoproposeanasaldevicetailoredtotheinvestigated formulation, powder delivery and depositionwere tested with three commerciallyavailable devices withdifferentinsufflation mechanisms,namelyanactivesingle-dose(MIAT),anactive multi-dose(Teijin) and a disposablepassivebi-dose (Aptar)one. The

Fig.6.Percentageofthalidomideretainedon(ADHERED)andremovedfromthemucosasurface(NONADHERED)after3minwashingin45mlofsimulatednasalfluid.The nonadheredamountswererecoveredasdispersedsolidorinsolutioninthewashingfluid;meanS.D.,n=3–4.*p<0.05accordingtounpairedtwo-tailedStudent’st-test. Table3

Massbalance frominvitrotransportexperiments inFranzcells (meanS.D.; numberofreplicatesinparentheses).

Formulation Massbalance(%) Saturatedsolution(10) 837

b-CD(6) 966

HP-b-CD(6) 912 SBE-b-CD(7) 962

threedevicesoperatedatmarkedlydifferentinsufflationairflows, i.e., 4, 0.6 and 15L/min respectively. Wewanted to investigate whetherandinwhatwaytheactuationmodeandairflowaffected powderdosedeliveryanddepositionsiteinsidethenosecast.

Table4presentsthepowderamountsdelivereduponactuation ofeachoftheconsidereddevices.Thereportedvaluesarethemean oftenshots.CapsulesorblistersusedinMIATandAptarcontained 20mgofpowder,correspondingtoathalidomidedoseof5mg.The delivereddoseforthemulti-dosedevice(Teijin)wastheaverageof tenconsecutivemeteredactuations.

Powderdeliverydependedonbothformulationtechnological propertiesanddeviceairflowduringinsufflation.WithMIATand Teijin devices the delivered dose was in line withthe powder reposeanglevalues(HP-

b

-CD>

b

-CD>SBE-

b

-CD;seeTable2).The differencesin delivered amount comingfrom thepowder _flow characteristicswerelevelleddownwiththeAptardevice.Thiswas reasonably due to the higher operating airflow. The powder formulatedwithHP-

b

-CDwastheonlyonequantitativelyemitted withoutsignificantdifferencesamongthedevices.Thevalueswere withinthelimits(15%ofthelabelclaim)oftheguidelineonthe pharmaceutical qualityof inhalation and nasal products (EMA, 2006).

Picture sequences of emitted plumes show that powder deliveryfromthedeviceswascompletedinlessthanonesecond (Supplementary Material). The plume development withAptar wasdifficulttovisualizeowingtotheexperimentalneedtoconfine itin anairdrawingtubeconnected withthenoseadapter.The combinedeffectofformulationanddevicevariableswasre_flected intheplumeshape.Themoreflowablepowderscontaining

b

-CD andHP-

b

-CDledtocloselypackedplumes.Incontrast,themore cohesiveSBE-

b

-CDpowdergeneratedlessdenseclouds,con firm-ingalessefficientpowderdeliveryforthisformulation.

Takingintoaccountthedisease,thetargetdepositionsiteand desiredintranasaldistributionofthedrug,afinalstudywascarried out generating the powder plume with the three insufflators directlyinto a silicone humannasal cast.The cast allowed for visualizingdifferencesamongthedistributionofthepowdersina humannasalcavity.OnlytheHP-

b

-CDpowderwasconsideredfor this study becausethe delivered amount was the same for all devices (see photographsof depositionpatterns in the Supple-mentaryMaterial).Eachdevicegaverisetoadifferentdegreeof coverageof thevarious regions of thenasal cavity. During the experiment,wequalitativelyobservedthat theAptarand Teijin devices deposited the powder more homogeneously in the vestibular, pre-turbinate and turbinate regions, whereas in the MIAT device larger particles were deposited in the turbinate region.

4.Conclusions

Thenasalpowderformulationofthalidomidewithlecithinand

b

-cyclodextrinoritsderivativeshydroxypropyl-

b

-cyclodextrinor sulphobutylether-

b

-cyclodextringaverisetostableproductswith suitable technological properties for nasal insufflation. From a biopharmaceuticalpoint ofview, theimproved dissolution rate comparedtothatoftherawmaterialmadethalidomidepromptly available in the aqueousnasal environment at a concentration

giving rise to an accumulation in the mucosa, with limited transmucosaltransport.Thisinvitroobservationsuggestsalow likelihoodofsignificantsystemicabsorption.Thetopicalactionon bleedingcouldbenefitfromthepoorabsorptionandfromthefact thatabout2–3%ofthethalidomideappliedonthemucosainvitro was accumulated within thetissue, particularly with the

b

-CD nasalpowder.Furthermore,residencetimeinthenasalcavitymay beprolongedbythemucoadhesiveactionprovidedbythisnasal powder.Consideringthenasalcavitycoverage,depositionof

b

-CD nasalpowderwaspredictedasmoreefficientwhenusingapassive device,i.e.,adevicedeliveringthepowderonlywhenthepatient inhalesthroughit.Anadditionaladvantageofthisactuationmode couldbethelimitedpowderdispersionintheenvironmentaround the patient. This is relevant with a formulation containing thalidomide. Thus, in nose telangiectasia, nasal powders could enabletopicalthalidomidetreatmentasacomplementtosystemic oraladministrationformaintenancetherapy,thuscontributingto asaferre-positioningofaneffectiveactivedrug.

Authordisclosurestatement

Theauthorsdeclarenoconflictofinterest. Acknowledgements

TheauthorswouldliketothankTelethonFoundation(Grant number GGP13036) for research fundingand Prof. Luca Nicola Ferraro (University of Ferrara) for his valuable contribution. Authors alsoacknowledgethe greathelp fromformer students Dr.GretaAdorni(UniversityofParma),Dr.BenedettaMarchesano (University of Ferrara) and Dr. Maria Teresa Stigliani Ph.D. (UniversityofSalerno).

AppendixA.Supplementarydata

Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j. ijpharm.2016.07.002.

References

Badr-Eldin,S.M.,Elkheshen,S.A.,Ghorab,M.M.,2008.Inclusioncomplexesof tadalafilwithnaturalandchemicallymodifiedbeta-cyclodextrins.I: preparationandin-vitroevaluation.Eur.J.Pharm.Biopharm.70,819–827.

Balducci,A.G.,Ferraro,L.,Bortolotti,F.,Nastruzzi,C.,Colombo,P.,Sonvico,F.,Russo, P.,Colombo,G.,2013.Antidiureticeffectofdesmopressinchimeraagglomerates bynasaladministrationinrats.Int.J.Pharm.440,154–160.

Bechgaard,E.,Gizurarson,S.,Jørgensen,L.,Larsen,R.,1992.Theviabilityofisolated rabbitnasalmucosaintheUssingchamber,andthepermeabilityofinsulin acrossthemembrane.Int.J.Pharm.87,125–132.

Bettini,R.,D’Andria,C.,Catellani,P.L.,Santi,P.,Massimo,G.,Cocconi,D.,Colombo,P., 1999.Nicotinenasalpowder:designandcharacterization.S.T.PPharm.Sci.9, 457–462.

Bortolotti,F.,Balducci,A.G.,Sonvico,F.,Russo,P.,Colombo,G.,2009.Invitro permeationofdesmopressinacrossrabbitnasalmucosafromliquidnasal sprays:theenhancingeffectofpotassiumsorbate.Eur.J.Pharm.Sci.37,36–42.

Buttini,F.,Colombo,P.,Rossi,A.,Sonvico,F.,Colombo,G.,2012.Particlesand powders:toolsofinnovationfornon-invasivedrugadministration.J.Control. Release161,693–702.

Castile,J.,Cheng,Y.H.,Simmons,B.,Perelman,M.,Smith,A.,Watts,P.,2013. DevelopmentofinvitromodelstodemonstratetheabilityofPecSys1_,aninsitu

nasalgellingtechnology,toreducenasalrun-offanddrip.DrugDev.Ind.Pharm. 39,816–824.

Challa,R.,Ahuja,A.,Ali,J.,Khar,R.K.,2005.Cyclodextrinsindrugdelivery:an updatedreview.AAPSPharmSciTech6,E329–E357.

Colombo,G.,Lorenzini,L.,Zironi,E.,Galligioni,V.,Sonvico,F.,Balducci,A.G., Pagliuca,G.,Giuliani,A.,Calzà,L.,Scagliarini,A.,2011.Braindistributionof ribavirinafterintranasaladministration.AntiviralRes.92,408–414.

deAraujo,M.V.,Vieira,E.K.,SilvaLazaro,G.,Conegero,L.S.,Almeida,L.E.,Barreto,L. S.,daCostaJr.,N.B.,Gimenez,I.F.,2008. Sulfadiazine/hydroxypropyl-beta-cyclodextrinhost-guestsystem:characterization,phase-solubilityand molecularmodeling.Bioorg.Med.Chem.16,5788–5794.

Table4

Mean delivered powder amount (mg) from Aptar, MIAT and Teijin devices (meanS.D.,n=10).

Powder Aptar MIAT Teijin

b-CD 20.41.2 19.00.7 18.50.6 HP-b-CD 21.01.7 21.00.7 21.71.2 SBE-b-CD 20.41.2 17.61.9 13.90.8

DeAscentiis,A.,Bettini,R.,Caponetti,G.,Catellani,P.L.,Peracchia,M.T.,Santi,P., Colombo,P.,1996.Deliveryofnasalpowdersofbeta-cyclodextrinby insufflation.Pharm.Res.13,734–738.

EuropeanMedicinesAgency:EMEA/CHMP/QWP/49313/2005CorrGUIDELINEON THEPHARMACEUTICALQUALITYOFINHALATIONANDNASALPRODUCTS2006.

http://www.ema.europa.eu/docs/en_GB/document_library/ Scientific_guideline/2009/09/WC500003568.pdf,May2016.

Fukuda,M.,Miller,D.A.,Peppas,N.A.,McGinity,J.W.,2008.Influenceofsulfobutyl etherbeta-cyclodextrin(captisol)onthedissolutionpropertiesofapoorly solubledrugfromextrudatespreparedbyhot-meltextrusion.Int.J.Pharm.350, 188–196.

Goosen,C.,Laing,T.J.,du,P.J.,Goosen,T.C.,Flynn,G.L.,2002.Physicochemical characterizationandsolubilityanalysisofthalidomideanditsN-alkylanalogs. Pharm.Res.19,13–19.

Govani,F.S.,Shovlin,C.L.,2009.Hereditaryhemorrhagictelangiectasia:aclinical andscientificreview.Eur.J.Hum.Genet.17,860–871.

Guttmacher,A.E.,Marchuk,D.A.,WhiteJr.,R.I.,1995.Hereditaryhemorrhagic telangiectasia.N.Engl.J.Med.333,918–924.

Invernizzi,R.,Quaglia,F.,Klersy,C.,Pagella,F.,Ornati,F.,Chu,F.,Matti,E.,Spinozzi, G.,Plumitallo,S.,Grignani,P.,Olivieri,C.,Bastia,R.,Bellistri,F.,Danesino,C., Benazzo,M.,Balduini,C.L.,2015.Efficacyandsafetyofthalidomideforthe treatmentofsevererecurrentepistaxisinhereditaryhemorrhagic

telangiectasia:resultsofanon-randomisedsingle-centre,phase2study.Lancet Haematol.2,e465–e473.

Jiang,L.,Gao,L.,Wang,X.,Tang,L.,Ma,J.,2010.Theapplicationofmucoadhesive polymersinnasaldrugdelivery.DrugDev.Ind.Pharm.36,323–336.

Kale,R.,Tayade,P.,Saraf,M.,Juvekar,A.,2008.Molecularencapsulationof thalidomidewithsulfobutylether-7beta-cyclodextrinforimmediaterelease property:enhancedinvivoantitumorandantiangiogenesisefficacyinmice. DrugDev.Ind.Pharm.34,149–156.

Kratz,J.M.,Teixeira,M.R.,Ferronato,K.,Teixeira,H.F.,Koester,L.S.,Simoes,C.M.O., 2012.Preparation,characterization,andinvitrointestinalpermeability evaluationofthalidomide-hydroxypropyl-beta-cyclodextrincomplexes.AAPS PharmSciTech.13,118–124.

Krenn,M.,Gamcsik,M.P.,Vogelsang,G.B.,Colvin,O.M.,Leong,K.W.,1992. Improvementsinsolubilityandstabilityofthalidomideuponcomplexation withhydroxypropyl-beta-cyclodextrin.J.Pharm.Sci.81,685–689.

Lebrin,F.,Srun,S.,Raymond,K.,Martin,S.,vandenBrink,S.,Freitas,C.,Bréant,C., Mathivet,T.,Larrivée,B.,Thomas,J.L.,Arthur,H.M.,Westermann,C.J.,Disch,F., Mager,J.J.,Snijder,R.J.,Eichmann,A.,Mummery,C.L.,2010.Thalidomide stimulatesvesselmaturationandreducesepistaxisinindividualswith hereditaryhemorrhagictelangiectasia.Nat.Med.16,420–428.

Loftsson,T.,Masson,M.,2001.Cyclodextrinsintopicaldrugformulations:theory andpractice.Int.J.Pharm.225,15–30.

Loftsson,T.,Hreinsdóttir,D.,Másson,M.,2005.Evaluationofcyclodextrin solubilizationofdrugs.Int.J.Pharm.302,18–28.

Marttin,E.,Verhoef,J.C.,Merkus,F.W.,1998.Efficacy,safetyandmechanismof cyclodextrinsasabsorptionenhancersinnasaldeliveryofpeptideandprotein drugs.J.DrugTarget.6,17–36.

Merkus,F.W.,Verhoef,J.C.,Marttin,E.,Romeijn,S.G.,vanderKuy,P.H.,Hermens,W. A.,Schipper,N.G.,1999.Cyclodextrinsinnasaldrugdelivery.Adv.DrugDeliv. Rev.36,41–57.

Pereira,B.G.,LigorioFialho,S.,MariadeSouza,C.,DantasCassali,G.,Silva-Cunha,A., 2013.Evaluationoftheeffectsofthalidomide-loadedbiodegradabledevicesin solidEhrlichtumor.Biomed.Pharmacother.67,129–132.

Russo,P.,Buttini,F.,Sonvico,F.,Bettini,R.,Massimo,G.,Sacchetti,C.,Colombo,P., Santi,P.,2004.Chimeralagglomeratesofmicroparticlesfortheadministration ofcaffeinenasalpowders.J.DrugDeliv.Sci.Technol.14,449–454.

Schmidt,M.C.,Simmen,D.,Hilbe,M.,Boderke,P.,Ditzinger,G.,Sandow,J.,Lang,S., Rubas,W.,Merkle,H.P.,2000.Validationofexcisedbovinenasalmucosaasin vitromodeltostudydrugtransportandmetabolicpathwaysinnasal epithelium.J.Pharm.Sci.89,396–407.

Schumacher,H.,Smith,R.L.,Williams,R.T.,1965.Themetabolismofthalidomide: thespontaneoushydrolysisofthalidomideinsolution.Br.J.Pharmacol. Chemother.25,324–337.

Vidgren,M.T.,Kublik,H.,1998.Nasaldeliverysystemsandtheireffectondeposition andabsorption.Adv.DrugDeliv.Rev.29,157–177.

vonMoos,R.,Stolz,R.,Cerny,T.,Gillessen,S.,2003.Thalidomide:fromtragedyto promise.SwissMed.Wkly.133,77–87.

Wexler,A.,Hasegawa,S.,1954.Relativehumidity-temperaturerelationshipsof somesaturatedsaltsolutionsinthetemperaturerange0_to50_{C.J.Res.Natl.}