Você está na página 1de 10

ADMET&DMPK1(3)(2013)1928;doi:10.5599/admet.1.3.

OpenAccess:ISSN:18487718

Perspective

WhatADMEtestsshouldbeconductedforpreclinicalstudies?
HongWan
ShanghaiHengruiPharmaceuticalCO.,LTD.
Dept.ofDMPK/Tox,Shanghai,P.R.China
Email:wanh@shhrp.com
Tel:+86(0)21547590661313;Fax:+86(0)2154759072;Mobile:+8618036618308

Received:June06th,2013;Revised:July01st,2013;Published:July03rd,2013

Abstract
Thepharmaceutical industry has beenevolving in recentyears, whichmade numerous CROsand biotech companies
conductingdrugdiscoveryanddevelopmentprogramsandservicesinChina.Asaverycomprehensiveandimportant
technology platform bridging efficacy and safety, the DMPK has become a mature discovery function to optimise
ADMEpropertiesindrugdesignandscreening,anddramaticallymitgateattritionratesduringthelastdecades.Inthis
article,theauthoraddressesseveralfrequentquestionsassociatedwithADME/DMPKstudies,e.g.,whatADMEtests
should be conducted for preclinical studies? What should a typical investigational new drug (IND) enabling package
cover?WhichADME/DMPKstudiesrequiregoodlaboratorypractice(GLP)ornonGLP?WhatdoesagoodPKprofile
looklike? Theauthorpresentsastraightforwardoverviewofthesebasicquestionsfromhismanyyearsexperience
inbothpharmaceuticalresearchandCROinsupportingdrugdiscoveryprojectsandINDfiling.

Keywords
DMPK,PK,toxicity,safety,screeningcascade,metabolite,drugdruginteraction,investigationalnewdrug

1.DMPK/ADME:definitionandstudypurpose
DMPK,orDrugMetabolismandPharmacokinetics,isanimportantpartofstudiesoftenreferredtoas
ADME(Absorption,Distribution,Metabolism,andElimination).
-

Absorption(howmuchandhowfast,oftenreferredtoastheabsorbedfractionorbioavailability)
Distribution(wherethedrugisdistributed,howfastandhowextensive)
Metabolism(howfast,whatmechanism/route,whatmetaboliteisformed,andwhethertheyare
activeortoxic)
Elimination(howfast,whichroute).

In the drug discovery process, early in vitro ADME screening and in vivo PK profiling provide a basis for
choosingnewmolecularentities(NMEs)andleadcompoundsthathavedesirabledrugmetabolism,PKor
safety profiles, necessary for drug candidate selection (CS) and latestage preclinical and clinical
development. The ADME properties of a drug allow the drug developer to understand the safety and
efficacydatarequiredforregulatoryapproval.

doi:10.5599/admet.1.3.9

19

Wan

ADMET&DMPK1(3)(2013)1928

Today,DMPKstudiesperformedinvitroandinvivohavebecomemorestandardisedproceduresacrossthe
pharmaceutical industry. A couple of typical examples include the most commonly used in vitro ADME
studies,suchaslivermicrosomesandthewholehepatocytemodelsforinvitrometabolism.Bothmodels
contain major metabolism enzymes, such as CYP450 for phase I metabolism and
UDPglucuronosyltransferase(UGT)forphaseIImetabolism.Theserelevantinvitromodelsarealsoapplied
to CYP inhibition/induction studies as well as drug metabolite identification and profiling across species.
Other key in vitro assays, such as Caco2 or MDCK cellbased models are often used for intestinal
permeabilityevaluation.ForinvivoPKstudies,AssociationforAssessmentandAccreditationofLaboratory
AnimalCare(AAALAC)accreditedanimals,suchasmice,rats,dogs,andnonhumanprimatesareemployed
togenerateinvivoPKdatalikedrugclearance(CL),bioavailability(F%),exposure(AUC),halflife(t1/2),and
distributionvolume(L).Currently,almostallDMPKrelatedassaysarecarriedoutbyavailableautomated
technology platforms combined with highthroughput liquid chromatographymass spectrometry
(LC/MS/MS)bioanalysis,whichhasconsiderablyspeededuptheADMEdatagenerationfordecisionmaking
duringthedrugdiscoveryanddevelopmentprocess.
2.WhatADMEtestsshouldbeconductedforpreclinicalstudies?

Hit-to-Lead Lead-optimisation Candidate

Potency <100 nM
Enzyme/cell based assays
LogD7.4
(1.5-4.5)

Solubility 7.4

Metabolic stability
CLint (h, r, d, m)
Hepatocytes
(r, h)

Dose =

Ce(ss) CL
F
MTD
(rodent)

In vivo PK (m, r, d, m)
F>20%, CL<50% BF,
dose<100mg

PPB (r, h, d, m)
fu> 0.1-1%

PK/PD
Dose response

hERG
IC50

Blood/plasma ratio
(r, h, d, m)

Met ID, CYP ID


Metabolism route

Non-GLP tox
(rodent, dog)

MDCK
PAMAP

Caco-2 permeability
> 0.3x 1E-6cm/s, efflux

CYP inhibition (IC50)


Major CYP>1-10 M
Reactive metabolite
(GSH)

Solubility 6.5
>10 M (FaSSIF)

> 1-10 M (PBS buffer)

CYP
Induction
Transporter, P-gp etc
Substrate/inhibition

CS (candidate
selection)

Clinical
Candidate
(GLP)
Human PK
dose, PK/PD

IND enabling
(Candidate)

GLP tox
(safety package)

Figure1:TypicalADME/PKscreeningcascade

Figure1highlightsatypicalADME/DMPKscreeningcascadetosupportdrugdiscoveryprogramsfromearly
hittolead, to lead optimisation and candidate evaluations. As depicted in Figure 1, a number of DMPK
studies are required before CS, followed by nonGLP and GLP toxicology studies for the INDenabling
package that is necessary for clinical candidate development. Various studies and assays should be
conducted at different stages to address ADME issues and meet specific criteria to enable the project to
20

ADMET&DMPK1(3)(2013)1928

ADMETestsforPreclinicalStudies

moveforward.Itshouldbenotedthatthesecriteria,asexemplifiedinthiscascade,arenotfixednumbers
or filters, and depend on other parameters as well as the specific therapeutic targets of different drug
discovery programs. This cascade just demonstrates a traditional downstream of ADME/PK tests for a
typicaldrugdiscoveryprogramsbeforetheINDenablingstage.Often,attheearlystageofhittolead,key
in vitro assays and studies, such as solubility, liver microsomal stability, CYP inhibition, and permeability
should be conducted to provide key data for chemists to select the more metabolically stable and
active/potentcompoundsforfurtherinvivoPKstudies.SolubilityandLogDdatacanserveasabasistorun
variousinvitroassaysaswellasfordrugstabilityassessmentandselectionofsuitableformulations.Also,it
is well known that Log D/Log P a crucial factor governing passive membrane partitioning, influencing
permeabilityoppositetoitseffectonsolubility.Moreover,iftheinvitrometabolicstabilitydatafromphase
I metabolism cannot predict in vivo hepatic plasma clearance, the phase II in vitro metabolism in
hepatocytesandevenhepatictransporterssuchasOATP1B1arethennecessarytofindtherelevantinvitro
modelsforinvivometabolismpredictiontodrivechemiststodesignandmakemorepotentandselective
compounds with improved PK profiles. Even at the very early stage of hittolead, selective in vivo PK
studiesarevaluabletoconfirmwhethertheappliedinvitroassays(invitrometabolismandabsorption)can
serve as good predictive models for in vivo PK in terms of plasma clearance and bioavailability. In some
circumstances, if both in vitro liver microsomes and hepatocyte models might not predict well in vivo
metabolism,moreextensiveinvivoPKstudieshavetobeperformedasaresultofpoorinvitroandinvivo
correlationsbasedonevaluationsbyasimplifiedwellstirredmodel.Aclearstructureactivityrelationship
(SAR)/ADMEcertainlyfacilitatesthedesignofnewcompounds.Forthedrugabsorptionevaluations,there
arethreecommonlyusedmodels,suchasPAMPA,andCaco2andMDCKpermeability.Fromthecostpoint
of view, the PAMPA membrane permeability assay is the most costeffective among these three in vitro
models, with a fast turnaround, while the Caco2 and MDCKMDR1 models can offer more accurate
absorption prediction as well as allow efflux and transporter studies for substrate and inhibition
identification.

Traditionally,theCYPinhibitionstudyisconductedatalatestage,e.g.,fromleadoptimisationtoCS.This
has been now moved to earlier stages, even hittolead. Table 1 shows a general decision tree on CYP
inhibitionbasedondrugdruginteraction(DDI)assessment,withaninterpretationofpotentialforinvivo
inhibition(derivedfromFDAGuidanceforIndustry:DrugInteractionStudiesStudyDesign,DataAnalysis,
ImplicationsforDosing,andLabelingRecommendations)[1].Anestimated[I]/Kiratioofgreaterthan0.1is
considered positive and a followup in vivo evaluation is recommended. The likelihood of an in vivo
interactionisprojectedbasedonthe[I]/Kiratiowhere[I]representsthemeansteadystateCmaxvaluefor
totaldrug(boundplusunbound)followingadministrationofthehighestproposedclinicaldose.Astheratio
increases,thelikelihoodofaninteractionincreases.Also,forCYP3Ainhibition,twostructurallyunrelated
substratesshouldbeevaluatedinvitro.Ifoneofthetwoevaluationssuggestsapotentialinteraction(i.e.,
[I]/Kiofmorethan0.1),aninvivoevaluationshouldbecarriedout.Attheearlierscreeningstage,IC50bya
singlepointscreeningassay,whichisamorecosteffectiveapproachwithacceptablygoodcorrelationwith
full curve IC50 assay, can be used for initial DDI evaluation. A full curve IC50 or Ki determination can be
utilisedforDDIassessmentatalatephaseforcandidatessuchasNMEs.Althoughquantitativepredictions
ofinvivoDDIfrominvitrostudiesarenotpossible,rankorder acrossthedifferentCYPenzymesfor the
samedrugmayhelpprioritiseinvivoDDIassessment.

doi:10.5599/admet.1.3.9

21

Wan

ADMET&DMPK1(3)(2013)1928

Table1:PredictionofclinicalrelevanceofcompetitiveCYPinhibition

Likelihood
Likely
Possible
Unlikely

IC50(M)
<1
1<IC50 <10
>10

[I]/Ki [1].
>1
0.1<[I]/Ki<1
<0.1

[I]istheconcentrationofinhibitorexposedtotheactivesiteoftheenzymeinvivo, Kiistheinhibition
constant,whichcanbeobtainedbyafullcurveIC50assay,orroughlyestimatedfromtheIC50,i.e.,
KiIC50/2

SimilarstudiesassociatedwithtransportersareneededtoidentifywhethertheNMEscouldbeapotential
substrateorinhibitorforPgporotherimportanttransportersforDDIasearlyaspossible.Morecomplex
decisionmakingtreesandrecommendedmodelsfortransportermediatedDDIassessmentareoutlinedin
theFDAsnewguidanceondruginteractionstudies.ItshouldbepointedoutthattheFDAhasextended
DDIguidancefromthe2006draft,with52pages,tothecurrent75pages(issuedin2012),reflectingthe
importanceandcomplexityofdruginteractionstudiesamongtheDMPK[1,2].Thisisattributedtothehigh
percentageofdrugfailurescausedbyDDI.
Anotherimportantaspectincludesthedrugmetaboliteidentificationandprofilingaswellasmetabolism
pathwaysinconnectionwithCYPmetabolismenzymeidentification(thesocalledCYPID).Nowadays,the
rapidmetaboliteidentification(MetID)ataveryearlystagehasalsobecomerelativelyeasy,suchaswith
Metabolynxassisted structure identification software, which is especially necessary for metabolically
unstablecompounds,whosebiotransformationcanhelpchemistsunderstandthesiteofmetabolicliability,
forfurtherstructuremodificationforadesirablemetabolismprofile.WiththeissuanceoftheFDAsfinal
guidanceforindustryonsafetytestingofdrugmetabolites(MIST),thereisincreasedconcernforobtaining
metabolite data as early as possible in preclinical studies. The discovery of disproportionate drug
metaboliteslateindrugdevelopmentcanpotentiallycausedevelopmentandmarketdelays.Asguidedin
MISTpublishedin2008aswellasin theICH M3 (R2) publishedin2009 [3,4],ifametaboliteisa unique
human metabolite, or more commonly, if a circulating metabolite is present at disproportionately higher
levels in humans than in the animal species used in toxicology studies, additional nonclinical safety
assessmentstudiesmayberequired.AsexemplifiedinFigure2,anymetabolitesfoundtobe>10%(based
onsystemicexposureatsteadystate)oftotaldrugrelatedmaterialshouldthenbemonitoredinrepeating
dose animal toxicology studies to determine if they are present at higher levels in other species as well.
Exceptions might be possibleforlowerriskmetabolites,suchasmostphaseIIconjugated metabolites. If
the major metabolite exposure in at least one species is greater than in the human, then no additional
safetystudiesofthemetaboliteareneeded.Itshouldbenotedthatthemetabolitedecisiontreeshownin
Figure 2 is only a suggestion or recommendation, with an approximate 10% cutoff, which does not
necessarily mean a metabolite testing safety guarantee. For instance, if a drug shows a low predicted
efficacydose(10mgorlower),ametabolitefoundtobeover10%oftotaldrugrelatedmaterialmightnot
requirefurthersafetytests.In contrast,forahighlydosedcompound,even ifametaboliteislowerthan
10%oftotaldrugrelatedmaterial,itcouldbetoxic,especiallywithaccumulationofthemetabolite.Also,a
reactivemetabolitemightbetoxicataverylowconcentration.Overall,ametabolitesafetytestingdecision
shouldconsidermanyfactors,suchasinvitroinvivocorrelationsandespeciallyinvivoanimaltoxicology
findings,e.g.,abnormalitiesorevenanimaldeathswithrepeat dosesonanimalstudies.Inthis case,itis
essentialtoidentifythebiotransformedmetabolitestounderstandwhethertheobservedtoxicityisdueto
theaccumulationofparentexposureortoxicmetabolites,asevenmetabolitesatalevellowerthan10%

22

ADMET&DMPK1(3)(2013)1928

ADMETestsforPreclinicalStudies

can accumulate. Such additional toxicity testing of the major metabolite can be informative to confirm
whetherthecauseoftoxicityisthemetaboliteornot.Inthiscase,forrodentoracuterattoxicology,the
identification and profiling of drug metabolites using toxicokinetic plasma samples can be used. In short,
relevantstudiesmustbecarriedouttodetermineifhumanmetabolitesareadequatelyevaluatedduring
nonclinicalsafetystudies.

DisproportionateDrugMetabolite

10%parent

>10%parent

systemicexposure
AUC

systemicexposure
AUC
Formedinany
animaltest

No further
needed

testing
No

Yes
Howmuch?

Exposureinanimal
studies
does
not
approach

Exposureinanimal
studies
does
approach

humanexposure

humanexposure

Nonclinicaltestingof
thedrugmetabolite

Nofurthertesting
needed

Figure2:FDAsuggestedmetabolitedecisiontreeflowdiagram

InadditiontotheCYPinhibition/inductionandtransporterstudiesassociatedwithDDI,othersafetyrelated
studies, such as reactive metabolism and hERG should be performed as early as possible. The hERG
inhibition test for cardiotoxicity has been shown to predict increased risk of arrhythmia, which can be
initiallyevaluatedunder nonGLPconditionbyinvitroapproachesthatarecosteffectiveandhaveafast
turnaround,andlaterconfirmedusingQTintervalprolongationmeasurementsaccordingtoFDAguidelines
[5]. Another importantinvivostudyisbasicsafetyteststhatare mostcommonlyperformedonratsand
dogs,e.g.,nonGLPacutetoxicology(7dayor14dayratanddogrepeatdose)fororaldrugs,whichusually
includebodyweight,clinicalchemistries,haematology,andhistopathology(highdoseandcontrolonlyifno
doi:10.5599/admet.1.3.9

23

Wan

ADMET&DMPK1(3)(2013)1928

pathology is seen at the high dose). This study should be designed to establish a dose that induces a
minimaltoxiceffect,oralternatively,toestablishasafetymargin(e.g.,10100xdependingonthedifferent
therapeutic targets). The selection of dose levels for repeat oral dose in toxicology studies is generally
based on the minimum efficacy dose (MED) and the maximum tolerated dose (MTD). Based on the
measured AUC (exposures) and toxicokinetic data, an approximate safety margin and compound
accumulationcharacteristicscan beknown,whichareparamountfor thedecisionmaking onwhetherto
followupaGLPtoxdesignornot.SuchnonGLPstudieshelptoidentifymoreaccuratedoseleveldesignfor
thesubsequentGLPstudies.
In summary, a comprehensive nonclinical ADME/PK package, including key toxicity data, should be
generallycompletebythetimeofCS.TheultimategoalsofallADME/PK/toxstudiesbeforetheCSstageare
tobetterunderstandthecompoundsmetabolitemediatedtoxicityandsafetyprofiletomakeaconcrete
decisionforthepurposeofenablingIND,whichisacrucialpartofthedrugapprovalprocess.
3.INDenablingpackageforGLPornonGLP?
What preclinical studies should be conducted to enable an IND? Should all studies be performed in GLP
laboratoriesforINDfiling?SomeresearchersaremiscommunicatingthatatINDstage,allstudiesshouldbe
conductedinGLPlaboratories.Actually,theFDArequiresGLPdocumentationonlyforsafetyrelatedinvivo
studieslikeGLPtox.Inotherwords,invitroADMEandinvivoPKstudiesdonotneedtobeperformedin
GLPlabs.Infact,fromabioanalyticaldataqualitypointofview,thereisnoapparentdiscrepancybetween
nonGLP and GLP laboratories regarding bioanalytical procedures, because most nonGLP bioanalysis
methodsareperformedaccordingtoFDAbioanalysisguidance[6].Themajordifferencecanbeinthestudy
design,e.g.,atypicalnonGLPtoxstudyutilisesasmallgroupofanimals(threeorfourpergroup),whilea
GLPtox study usually utilises more than 10 animals per group with half males and half females in one
study.Inaddition,morestrictdocumentationmustberecordedforthepurposeofregulatorycompliance.
Obviously,theGLPstudyhasmuchhighercoststhanasimilarnonGLPstudy,whichisnotrecommended
beforeCSfortheINDenablingpackage.KeycomponentsoftheINDdatapackageincludepharmacology,
toxicologyandsafetypharmacology,ADMEandchemistry,andthemanufacturingandcontrolsectionsof
the submission. A typical INDenabling ADME package contains data from the following studies:
bioanalytical method validation in one rodent and one or more nonrodent species; single and
multipledose PK; dose proportionality and absolute bioavailability in one rodent and one or more
nonrodent species; and in vitro CYP inhibition/induction in human liver microsomes for DDI assessment
(includinglikelytransporterstudies).Otherdata,suchasmassbalanceandroutesofexcretioninrodents,
includingbilaryexcretioninrats,metaboliteprofilingandidentificationinrodentsandnonrodents,tissue
distributionusingradiolabelledcompoundbyquantitativewholebodyautoradiography(QWBA)formpart
of the surrogate endpoints, although these studies are not required for submission. Overall, complete
ADMEandtoxicologydatacanmakethesubmissionpackagemorecompellingtomoveacompoundinto
phaseIclinicalstudies.Knowingtheobjectives,expectations,andprocessesofassemblingandfilinganIND
isthekeyfactornotonlyforasuccessfulfiling,butitcanalsoaccelerateapromisingclinicaldevelopment
path forward. More detailed information on the general requirements and strategies for successful IND
filingscanbefoundinFDAsguidance[7],whichisbeyondthescopeofthecurrentmaintopic.Inorderto
reducethepotentialdevelopmentcost,exploratoryIND(eIND)preclinicalstudiesarerecommendedbythe
FDA [8]. Such eIND studies are conducted prior to the traditional dose escalation, safety, and tolerance
studiesthatordinarilyinitiateaclinicaldevelopmentprogram,whichislessextensivethanfortraditional
IND studies. Overall, eIND studies provide the opportunity to already obtain human data in the drug
discoveryphase,suchasimportantinformationonPK,PK/PD,andbasictoxicologydataforakeydecision

24

ADMET&DMPK1(3)(2013)1928

ADMETestsforPreclinicalStudies

for selecting the most promising lead product. Even a microdose approach (e.g., <100 g, or <1/100 of
MED) in a human study, using an advanced technology such as accelerator mass spectrometry (AMS)
combinedwithradiolabelledcompoundcanbeverybeneficialtogainhumanPKinformationasearlyas
possibleforINDenabling[9].
4.WhatisagoodPKprofile?
AnotherfrequentquestionaskedbydrugdiscoveryscientistsisIsthiscompoundsPKgood?Ideally,the
desirablePKprofilesofapreclinicalcandidatethatisbeingconsideredfortakingintodevelopmentshould:

haveacceptablesolubilityfordevelopment;
becompletelyabsorbed,preferablyviapassiveabsorption;
havehighbioavailability(e.g.,F>50%)fororaldrug;
have a low plasma clearance CL (e.g, <30% blood flow), long halflife (t1/2) (e.g., >6 hrs), and
acceptabledistributionvolume;
havealinearkinetics,i.e.,exposureproportionaltodoseandaclearPK/PDcorrelation;
beeliminatedbyseveralpathways,i.e.,renalexcretionandhepaticmetabolism,alsometabolised
preferablybymorethanoneenzymeforderiskingDDI;
haveasimplemetaboliteprofile,withnoreactivemetabolite;
havenoobviousCYPandmajortransporterslikePgpinhibitionorinductionorlowDDIpotential;
and
haveasufficientoratleastacceptablesafetymargin(safetymargin>10x,dependingondifferent
therapeutictargets).

In reality, we can readily rank which compounds PK is better than others from the same series of
compoundsduringthescreening.However,weshouldnotruleoutacompoundbyonlylookingatitsPK
data,especiallyforthosewithsuperpotentNMEs.Aslongastheirpotenciesandmechanismareattractive,
they should be tested in in vivo efficacy models despite poor PK profiles. Indeed, a number of marketed
drugshaveshownpoorPKprofilesbuthavestillbecomethetopeversellingdrugLipitor(Atorvastatin),
withabsolutebioavailabilityof5%and14%inratandhuman,respectively[10,11].AsillustratedinFigure3,
inananalysisof600marketdrugs[12],morethan30%ofthemhaverelativelylowbioavailability(F<30%),
and 22% of the drugs have even lower bioavailability (F <10 %). Nearly 50% of the drugs have
moderatetohighclearance,and17%showhighclearance,orextremelyhighclearance(evenoverblood
flow).
As mentioned above, some of these top selling drugs have many undesirable properties in terms of
physicochemical properties, such as solubility, Log D, in vitro microsomal stability (Clint), and
permeability/effluxdata,aswellashighinvivoclearanceorlowbioavailability.IfweuseonlyPKdatato
screen the compounds, some potential druglike compounds might be discarded. In other words, the
AtorvastatinlikecompoundswouldneverhavebecomegooddrugsfromthesoleviewofPKprofiles.These
examples suggest that a better understanding of in vitroin vivo correlation and frontloading PK/PD is
crucialforcompoundselectionevenatanearlystage.Obviously,asuperPDcansaveapoorPKaslongas
thePDeffectisnotable.

doi:10.5599/admet.1.3.9

25

Wan

ADMET&DMPK1(3)(2013)1928

Figure3:Distributionofhumanclearanceandbioavailabilityofmarketdrugs

5.Concludingremarks
Drugdiscoveryanddevelopmentareahighrisk,highcostbuthighlyrewardingbusiness.Manyresources
areinvested,and thuswastedon, candidate productsthataresubsequently foundto haveunacceptable
profileswhenevaluatedinhumanonlylessthan10%ofINDsfromNMEsprogressbeyondthesubmission
ofnewdrugapplication(NDA)accordingtotheFDAsreport[8].Asaverycomprehensiveandimportant
science and technology platform bridging efficacy and safety, DMPK has become a discovery function to
optimiseADMEpropertiesindrugdesign,toreduceattritionrates.Recentstudieshavedemonstratedthat
it is viable to implement highthroughput cuttingedge technologies and costeffective ADME/DMPK
screening assays, and design relevant studies to address various ADME issues at different stages for
deriskingtheinvestmentandavoidingdevelopmentmistakesearly.

Apparently, at drug discovery stage, a better understanding of SAR and in vitroin vivo correlations, and
moreinputondrugdesigncanbeanticipated.Moreover,frontloadingPK/PDforabetterunderstandingof
the efficacyexposure relationship and dose prediction as early as possible can benefit projects. The
integrationofinvitroandinvivodrugmetabolismdatawithphysicochemicalpropertiesaswellasPKand
PD data is highly recommended for safety and toxicity liability evaluation, e.g., are there any drug
accumulations?Arethereanypotentialtoxicmetabolites?ArethereanylikelyDDIs?Areestimatedsafety
margins enough? These safetyrelated questions have to be answered as early and clearly as possible to
deriskthedrugdiscoveryprojects.AnotherconcernisriskassessmentassociatedwithhERG,CYPinhibition
and induction, metabolite profiling and safety, transporter substrate and inhibition, safety margin, etc.,
which is of ultimate importance for CS and decisionmaking for the INDenabling package. Future ADME
studieswillcontinuetofocusonP450enzymesandtransportersimpactonsafedrugdeliveryaswellas
accuratepredictionofhumanPK,PK/PDmodelingandoptimumclinicaldosage.

Ontheotherhand,fromthepointofviewofregulatoryreview,theFDAhasbeenbecomingincreasingly
conservativethanever,exceptforoncologyandlifethreateningdiseaseswhereariskbenefitassessment
canbeconsidered.Ifadrugistoxicatclinicallyrelevantdoses,theINDmaynevergoforward.Therefore,a
completestudy/datapackageincludingasmuchdetailedtoxicityinformationaspossiblecanfacilitateIND
filing and increase the clinical trial success rate, which should cover pharmacology (efficacy and animal
26

ADMET&DMPK1(3)(2013)1928

ADMETestsforPreclinicalStudies

proof of principle) studies, a rodent general toxicology study of at least a 14day duration, a nonrodent
general toxicology study of at least a 14day duration, genotoxicity studies, safety pharmacology studies,
ADME/PK as well as bioanalytical validations, etc. In the end, we must clearly answer these two key
questions:
(1)Isthepreclinicalcandidatesafeenoughtotakeintohumans?
(2)Isthepreclinicalcandidateeffectiveinpatients?
Acknowledgements:TheauthorwouldliketothankDr.KinTamforproofreadingthisarticlewithvaluable
comments.
Declaration: This article reflects the authors personal opinions. The author declares that he has no
competingfinancialinterests.

References
[1] FDA:DraftGuidanceforIndustry:DrugInteractionStudiesStudyDesign,DataAnalysis,
Implicationsfordosing,andLabelingRecommendation,(Sept.2006)
[2]

FDA:GuidanceforIndustry:DrugInteractionStudiesStudyDesign,DataAnalysis,Implicationsfor
dosing,andLabelingRecommendation(February2012)

[3]

FDA:GuidanceforIndustry:Safetytestingofdrugmetabolites(February2008)

[4]

ICHM3(R2):Guidanceonnonclinicalsafetystudiesfortheconductofhumanclinicaltrialsand
marketingauthorizationforpharmaceuticals(June2009)

[5]

FDA:GuidanceforIndustryS7BNonclinicalEvaluationofthePotentialforDelayedVentricular
Repolarization(QTIntervalProlongation)byHumanPharmaceuticals(2005)

[6]

FDA:GuidanceforIndustryBioanalyticalMethodValidation(2001)

[7]

FDA:GuidanceforIndustryContentandFormatofInvestigationalNewDrugApplications(INDs)for
Phase1StudiesofDrugs,IncludingWellCharacterized,Therapeutic,Biotechnologyderived
Products(1997)

[8]

FDA:GuidanceforIndustry,Investigators,andReviewersExploratoryINDStudies(Jan2006)

[9]

G.Lappin,M.Seymour,Bioanalysis2(2010)13151324.

[10]

Y.Y.Lau,H.Okochi,Y.HuangandL.Z.Benet,DrugMetabolismandDisposition34(2006)11751181.

[11]

H.Lennernas,ClinicalPharmacokinetics42(2003)11411160.

[12]

L.S.Goodman,A.Gilman,ThePharmacologicalBasisofTherapeutics,McGrawHillPublishers,New
York,USA,2006.

Abbreviations
ADME

AbsorptionDistributionMetabolismExcretion

CRO

ContactResearchOrganization

DMPK

DrugMetabolismandPharmacokinetics

PK

Pharmacokinetics

doi:10.5599/admet.1.3.9

27

Wan

ADMET&DMPK1(3)(2013)1928

CYP450

CytochromeP450

MDCK

MadinDarbycaninekidneycells

PPB

PlasmaProteinBinding

hERG

humanEtheragogoRelatedGene

MetID

MetaboliteIdentification

PD

Pharmacodynamics

Pgp

Pglycoprotein

MTD

MaximumToleratedDose

IND

InvestigationalNewDrug

GLP

GoodLabPractice

CL

Clearance

Absolutebioavailability

Ce(ss)

Effectiveplasmaconcentrationatsteadystage

Doseinterval

CS

CandidateSelection

PMAPA

ParallelArtificialMembranePermeabilityAssay

NMEs

NewMolecularEntities

2013bytheauthors;licenseeIAPC,Zagreb,Croatia.Thisarticleisanopenaccessarticledistributedunderthetermsand
conditionsoftheCreativeCommonsAttributionlicense(http://creativecommons.org/licenses/by/3.0/)

28