T4056 Development of Correlation for Long Acting ... · Development of In Vitro-In Vivo Correlation...

Post on 03-Jun-2020

5 views 0 download

Transcript of T4056 Development of Correlation for Long Acting ... · Development of In Vitro-In Vivo Correlation...

DEVELOPINGSCIENCE.IMPACTINGHEALTH.

Development of In Vitro-In Vivo Correlation for Long Acting Injectable Microsphere Formulations A. Shahraz, J. Mullin, J. Spires, V. Lukacova, M. B. Bolger, W. Woltosz Author Affiliation: Simulations Plus, Inc QR

CodePURPOSEThe concept of in vitro-in vivo correla3ons (IVIVCs) for long-ac3ng injectable (LAI)microsphere formula3ons has gainedmore significance in the past decade. However,this problem is challenging due to the mul3phase release characteris3cs ofcomposi3onally equivalent formula3ons, the lack of a mechanis3c deconvolu3onmethod,andthelackofcompendialinvitroreleasetes3ngmethods.Thisstudyaimsto(1) determine whether an IVIVC can be established for long-ac3ng injectablemicrosphere formula3ons with different release profiles; (2) assess the role ofmechanis3cdeconvolu3ons indetermining the invivo releaseprofiles;and (3)explorethe poten3al for using a triphasic Weibull func3on on improving results of thedeconvolu3onprocess.

CONCLUSION(S)ThepossibilityofestablishingIVIVCsforlong-ac3nginjectablemicrosphereformula3onswas inves3gated. The results show promise, but the desired success is yet to beachieved.Severalimportantaspectshavebeendiscovered,including:•  The significanceofusinga triple-Weibull func3on in themechanis3cdeconvolu3on

and describing the more complex in vivo release of some long-ac3ng injectablemicrospheres

•  The role of the selected op3miza3on objec3ve func3on in the mechanis3cdeconvolu3on,specificallyinthecaseoflackingsufficientinvivosamplingpoints

•  Theeffectofusing interpolated invitrodata inestablishingan IVIVCandpredic3ngthecorrectshapeoftheCp-3meprofile

We aim to con3nue our study on the pharmacokine3cs of LAImicrosphereswith thegoalofbeUercharacterizingthe invivoenvironmentalparametersthataffectpolymerdegrada3on,microspheredispersion,andAPIpharmacokine3csaWerinvivoinjec3onoftheseformula3ons.

RESULT(S)AlevelAIVIVCwasestablishedforeachofthetestcompoundsandformula3onsandseveralimportantaspectswerediscoveredintheprocess.

(1)  Complex in vivo Profile: The in vivo release profiles for these long-ac3ng injectablemicrospheresmay be complex and cannot be always accurately described by a single- ordouble-Weibullfunc3on.Atriple-Weibullfunc3onwasrequiredtoaccuratelydescribetheinvivoreleaseprofilesfororn3deformula3ons.

(2)Op9miza9onTargetCriteria:WhenfiXng the invivo releaseprofileagainst theen3reobservedCp-3meprofile, theerroronCmax isoWenhigher thanallowedby the IVIVCcriteriadue to the number of other concentra3on points outweighing the contribu3on of the singleCmax value. This issue can be addressed by including addi3onal weight on Cmax during thedeconvolu3on/op3miza3onprocess.

(3)InsufficientinvitroSampling:Thedensityofinvitrosamplingpointsisalsoimportantforpredic3onofthecorrectshapeoftheCp-3meprofile.InthecaseofhuperzineA,theIVIVCwasable to predict the Cmax and AUC for new formula3ons, but a smaller ini3al peak was notcapturedduetothelackof invitrodatapointsrepresen3ngtheini3alreleaseofthedrug.Thislimita3onwasaddressedbyusinginterpolateddatapointsintheinvitrodissolu3onprofile.

METHOD(S)WeaUempted toestablishvalidatedLevelA IVIVCsusingGastroPlus™9.6 (Simula3onsPlus,Inc.) for several drugs formulatedas long-ac3ng injectablemicrospheres. Literaturedata forseveral poly (d,l- lac3de-co-glycolide) (PLGA) or poly(d,l-lac3de) (PLA) formula3onsadministered subcutaneously in Sprague-Dawley rats for olanzapine [1], intramuscularly inbeagledogsforhuperzineA[2],andsubcutaneously inrats fororn3de[3]wereused inthisstudy. Foreachcasestudy,thePKwasestablishedbasedonplasmaconcentra3on-3me(Cp-3me) profiles aWer intravenous (IV) (huperzine A), intraperitoneal (IP) (olanzapine), orsubcutaneous (SC) (orn3de) injec3on. The PK model was subsequently linked to theintramuscular or subcutaneous controlled release model in the Addi3onal Dosage RoutesModule(ADRM) inGastroPlus,andthecompletemodelwasusedtodeconvolutethe invivoreleaseprofileforeachformula3on. Finally,levelAIVIVCsbetweenthedeconvolutedinvivorelease and in vitro dissolu3on profiles were established and evaluated across differentformula3onsforeachtestcompound.

OBJECTIVE(S)TheObjec3vesofthisstudyaretoinves3gatethefeasibilityofestablishingIVIVCforlongac3nginjectablemicrosphereformula3onsandtodiscoverimportantaspectsofthisprocess.

FUNDING/GRANTS/ENCORE/REFERENCEOROTHERUSETheworkwasdoneunderfundingfromtheFDA(grant1U01FD005463-02)References[1]D’Souza,S.,Faraj,J.A.,Giovagnoli,S.,DeLuca,P.P.,Interna3onalJournalofBiomaterials2014;2014Ar3cleID407065[2]Chu,D.,Fub,X.,Liu,W.Liu,K.,Li,Y.,,Interna3onalJournalofPharmaceu3cs2006;325:116–123[3]Kostanski,J.W.,DaniB.A.,Reynolds,G.,Bowers,C.Y.,DeLuca,P.P.,,AAPSPharmSciTech2000;1(4).

CONTACTINFORMATION:azar@simula3ons-plus.com

T4056

(a) (b)

SimulatedCp-@meprofilesforanaltrexoneLAImicrosphereformula@onconsidering(a)targetobservedCp-@meprofileand(b)addi@onalweightaddedtoCmaxintheop@miza@onfunc@on.

(a) (b)

SimulatedCp-@meprofilesaKerSCinjec@onofoneorn@deLAIformula@oninrat.TheinvivoreleaseprofilewasfiMedas(a)double-Weibulland(b)triple-Weibullfunc@on.

PredictedCp-@meprofilesforahuperzineALAImicrosphereformula@onbasedonIVIVCbuilt(a)withand(b)withoutusingtheinterpolatedinvitrodata

Unabletopredictini3alburst

𝑦↓𝑖𝑛 𝑣𝑖𝑣𝑜  =0.949𝑥↓𝑖𝑛 𝑣𝑖𝑡𝑟𝑜 ↑0.885 

(b)

abletopredictini3alburst

(a)

𝑦↓𝑖𝑛 𝑣𝑖𝑣𝑜  =−0.801𝑥↓𝑖𝑛 𝑣𝑖𝑡𝑟𝑜 ↑3 + 0.982𝑥↓𝑖𝑛 𝑣𝑖𝑡𝑟𝑜 ↑2 + 0.731𝑥↓𝑖𝑛 𝑣𝑖𝑡𝑟𝑜 +0.016