feature
EU decision-making for marketing authorization of advanced therapy medicinal products: a case study
SofiekedeWilde1,z,DelphiG.M.Coppens2,z,JarnoHoekman2,3,MarieL.deBruin2,4, HubertG.M.Leufkens2,Henk-JanGuchelaar1andPaulineMeij1,p.meij@lumc.nl
A comparative analysis of assessment procedures for authorization of all European Union (EU) applications for advanced therapy medicinal products (ATMPs) shows that negative opinions were associated with a lack of clinical efficacy and identified severe safety risks. Unmet medical need was often considered in positive opinions and outweighed scientific uncertainties. Numerous quality issues illustrate the difficulties in this domain for ATMP development. Altogether, it suggests that setting appropriate standards for ATMP authorization in Europe, similar to elsewhere, is a learning experience.
The experimental characteristics of authorized ATMPs urge regulators, industry, and clinical practice to pay accurate attention to post-marketing risk management to limit patient risk. Methodologies for ATMP development and regulatory evaluations need to be continuously evaluated for the field to flourish.
Introduction
Overthepastdecade,therehasbeenincreased interestinthedevelopmentofATMPstowards marketingauthorization.In2009,RegulationEC No.1394/2007cameintoforceasthefirstspe- cificregulatoryframeworkforapprovalofthis potentiallynewclassofmedicinalproductsin theEU[1,2].ByAugust2017,thenumberof ATMPregulatoryproceduresformarketingau- thorizationwas16,anumberthathasbeen coinedasrelativelylowgiventherecentim- pressiveadvancesinbasicmolecularandclinical scienceinthefieldofATMPs[3–5].
ItiswellknownthatATMPdevelopersface variousscientificandtechnologicalchallenges, frommanufacturingandqualityissues[6]to
preclinicalandclinicalefficacyandsafetyissues [1].Moreover,additionalhurdlesinthetrajec- torytowardsapprovalareexperiencedbyaca- demicdevelopers,suchasalackofregulatory knowledge,insufficientfinancialsupport,and clinicaltrial-relatedproblems,suchasrecruit- ment[7].AlthoughRegulationECNo.1394/2007 includeshigh-levelrequirementsforapproval, becausethefieldisrapidlyevolving,standard- izationofregulatoryrequirementsforapproval isdifficultandperhapsundesirable.Conse- quently,duringthedecision-makingprocess, regulatorsneedtodealwithnovelissuesthat havenotbeenpreviouslydiscussedinother regulatoryprocedures[8].Giventhesedevel- opmentalandregulatorycomplexities,scientific
uncertaintiesduringbenefit–riskassessments areprevalent.
Inthisstudy,weprovideinsightintodecision- makingforapprovalofATMPsinEuropebe- tween1January2009and1July2017by characterizingregulatoryassessmentproce- duresformarketingauthorization,andanalyz- ingidentifiedmajorissuesandconsiderations forbenefit–riskoutcomes(seeAppendix1inthe supplementalinformationonline[9–13]).
Cohortanalysisofassessmentprocedures Fromthe14ATMPsincludedinourstudy,five werestandardapprovals,threewereapproved viaanexpeditedpathway(definedascondi- tionalapprovalorapprovalunderexceptional FeaturesPERSPECTIVE
1328 1359-6446/ã2018TheAuthors.PublishedbyElsevierLtd.ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).
circumstancesforthisstudy),andsixwere nonapproved(Table1).Theproductprofilesof allassessedATMPsareshowninTable2.Char- acteristics,suchasATMPsubtype,startingma- terial,administrationroute,andstorage conditions,werediverseforthedifferentsub- mittedproducts.Orphandrugdesignationwas assignedtoallexpeditedapprovedproducts, whereasonlyone(outoffive)standardap- provedproductsandhalf(threeoutofsix)ofthe nonapprovedproductsweredesignatedorphan drugs.Fortheexpeditedapprovedproducts,no alternativetreatmentwasavailable,whereasthis appliedonlytooneoutoffivestandardap- provedproductsandtwo(outofsix)nonap- provedproducts.
Allstandardapprovalsweretestedaccording tostandardsonsterility,purity,andviabilityupon release.However,fortheexpeditedapprovalsand nonapprovals,thesereleasetestswerenotalways discussedintheEuropeanpublicassessment report(EPAR).Remarkablewastheunspecified shelf-lifeandstorageconditionsfornonapproved products(fouroutofsix).
Thedesignofpivotalclinicaltrialswasmore robustforstandardversusexpeditedapproved andnonapprovedproducts.Formost(fourout offive)ofthestandardapprovals,arandomized controlledPhase3clinicaltrialwasperformed.
Bycontrast,thiswasthecaseforonlytwo(outof six)nonapprovedandfornoneoftheexpedited approvedproducts.Thenumberofpatients recruitedwashigherforthestandardapproved products(mean:244patients,range:12–341) comparedwithnonapprovedproducts(mean:
120patients,range:26–241)andexpedited approvedproducts(mean:57patients,range:
14–106).Thedefinedprimaryendpointswere consideredclinicallyrelevantforallstandard approvedproducts,forsomeexpeditedap-
provedATMPs(twooutofthree)andforhalf (threeoutofsix)ofthenonapprovedproducts.
Asignificanteffectontheprimaryendpoint wasdemonstratedforallstandardapproved products.Bycontrast,significanteffectswere notdemonstratedintwo(outofthree)expe- ditedapprovedproductsandinfive(outofsix) nonapprovedproducts.Noaddedclinicalben- efitwasdemonstratedformostofthestandard approved(fouroutoffive)andforallthenon- approvedproducts.Addedclinicalbenefitwas demonstratedforallexpeditedapprovedpro- ductsbecauseofthelackofalternativethera- pies.
Analysisofmajorissues
Majorissueswereevaluatedacrossassessment procedures,regardlessoffinalregulatoryopin- ion(Table3;fordetaileddescriptionsseeTable S1inAppendix2inthesupplementalinfor- mationonline).
Forquality,majorissueswerenotedforall products;forexample,thevector(expedited approvaloneoutofthree,nonapproved:two outofsix)andspecificreleasetests(standard approved:oneoutoffive,expeditedapproved:
threeoutofthree,nonapproved:fiveoutofsix).
Whereasdevelopersoftheapprovedproducts wereabletoresolvetheobjectionsbeforefinal regulatorydecision-making,developersofthe nonapprovedproductswereunabletoresolve thesemajorissues,whichweremostlyraised earlyduringtheassessmentprocedure,and decidedtowithdrawtheirproduct.
Mostofthemajorissuesrelatedtopreclinical studieswereraisedfornonapprovedproducts, concerninganimalmodels(oneoutofsix), toxicology(fouroutofsix)andefficacystudies (oneoutofsix).Bycontrast,nomajorissues werenotedfortheapprovedproducts,except
forone(outofthree)expeditedapproved product,whichconcernedtoxicologyandwas unresolveduponfinaldecision-making.Inad- dition,majorissuesindicatedfornonapproved productswerestillunresolvedatthetimeof finaldecision-making.
Forclinicaltrialdesign,mostmajorissues werealsoraisedfornonapprovedproducts.
Theseissuesconcernedmethodologicalissues orinvalidclinicaltrialdesign(fiveoutofsix)and changeofendpointsoruncertainclinicalrele- vanceofanendpoint(twooutofsix).Achange ofendpointswasalsonotedasamajorissuefor onestandardandoneexpeditedapproved product.Fortheapprovedproducts,themajor concernswereconsideredresolved,whereasall majorissuesaroundclinicaltrialdesignforthe nonapprovedproductswereunresolvedupon finaldecision-making.
Majorissuesrelatedtoclinicaloutcomeswere raisedforallnonapprovedproductsandfor Glybera1,oneoftheapprovedproducts.Alack offavorableclinicaloutcomesfornonapproved productsrelatedtobothefficacy(sixoutofsix) andsafety(fiveoutofsix).Furthermore,good clinicalpractice(GCP)wasanissueinthree(out ofsix)dossiersandpharmacodynamicsdata weretoolimitedintwo(outofsix)nonapproved products.
Analysisofbenefit–riskassessment ForstandardapprovedATMPs,benefit–risk balancesweremainlybasedonclinicalefficacy results(Table4).Thebeneficialefficacyout- comesandafavorablesafetyprofileresultedin apositiveopinionforMACI1.Thebeneficial efficacytrendforChondrocelect1and Imlygic1combinedwithsatisfactorysafety profilesresultedinstandardapproval,despite ampleregulatorydiscussionabouttheclinical TABLE1
Productsusedintheanalysisa
Product ATMPsubtype Startingmaterial Approvaltype Dateoffinaloutcome
Chondrocelect1 TEP Autologous Standardapproval October2009
Imlygic1 GTMP–invivo N/A Standardapproval October2015
MACI1 TEP Autologous Standardapproval April2013
Provenge1 CTMP Autologous Standardapproval June2013
Strimvelis1 GTMP–exvivo Autologous Standardapproval April2016
Holoclar1 TEP Autologous Conditionalapproval December2014
Zalmoxis1 CTMP Allogeneic Conditionalapproval June2016
Glybera1 GTMP–invivo N/A Underexceptionalcircumstances October2012
Advexin GTMP–invivo N/A Nonapproval(withdrawn) December2008
CLG GTMP–invivo N/A Nonapproval(withdrawn) June2009
Cerepro GTMP–invivo N/A Nonapproval(withdrawn) April2007
Heparesc CTMP Allogeneic Nonapproval October2015
Hyalograft TEP Autologous Nonapproval(withdrawn) January2013
OraNera TEP Autologous Nonapproval(withdrawn) March2013
aAbbreviations:CLG,ContusugeneLadenovecGendux;CTMP,celltherapymedicinalproduct;GTMP,genetherapymedicinalproduct;TEP,tissueengineeringproduct.
FeaturesPERSPECTIVE
trialdesign.Significantandclinicallyrelevant efficacyofProvenge1combinedwiththeac- knowledgedunmetmedicalneedforthetarget indication(oncology),outweighedtherisksand uncertaintiesrelatedtothesafetyprofile.
CompellingefficacyoutcomesforStrimvelis1,
withtheacknowledgedunmetmedicalneed, outweighedrisksanduncertaintiessurround- inglatentsevereadverseevents[14].Despite thesefavorableregulatoryopinions,divergent positionsweresubmittedfortwoapproved products(Imlygic1:N=1;Provenge1:N=13).
Asaprerequisiteforconditionalapproval pathways,thebodyofevidencewasoverallless robustandassociatedwithmoreuncertainty comparedwithstandardapprovedATMPs (Table4).Uncertaintyaboutsignificantclinical benefitsforHoloclar1wasrecognizedbecause TABLE2
Elementswithvariablesscoredpermarketingapprovaltypea,b
Element Variable SA(N=5) CA(N=2) UEC(N=1) NA(N=6)
Productprofile
Producttype GTMP 2 0 1 3
CTMP 1 1 0 1
TEP 1 1 0 2
Combined 1 0 0 0
Startingmaterial Autologous 4 1 0 2
Allogeneic 0 1 0 1
Notapplicable 1 0 1 3
Endproduct Refrigerated 2 0 0 0
Roomtemperature 2 1 0 0
Nitrogen-cryopreserved 0 1 0 1
Other-cryopreserved 1 0 1 1
Unspecified 0 0 0 4
Previousapprovedinotherjurisdictions Yes 3 0 0 0
Indicationarea Cancer 2 0 0 2
Congenital,hereditary,neonataldiseases 1 0 1 2
Eyediseases 0 1 0 1
Immunesystemdiseases 0 1 0 0
Musculoskeletaldiseases 2 0 0 1
Lackalternativetreatment Yes 1 2 1 2
Orphandrugdesignation Yes 1 2 1 3
Scientificevidence
Quality Potencyassay 5 2 1 6
Release:sterility 5 1 0 4
Release:purity 5 1 0 4
Release:viability 5 2 0 2
Release:activity 3 1 1 3
Preclinical Toxicity 4 2 1 6
Efficacy 5 1 1 6
Dose 3 1 1 6
Pivotaltrialdesign RCT 4 0 0 2
ClinicalprimaryEP 5 1 0 4
ClinicalrelevanceprimaryEP 5 2 0 3
Significantoutcome 5 1 0 1
Clinicaloutcome SignificantprimaryEP 5 1 0 1
Beneficialeffect 1 2 1 0
Regulatoryprocess
Scientificadvice 5 2 1 4
Restrictedlabeling 5 1 1 0
aPerelement,variablesarescoredforeach(non-)approvaltypeofATMP.
bAbbreviations:CA,conditionalapproved;CTMP,celltherapymedicinalproduct;EP,endpoint;GTMP,genetherapymedicinalproduct;NA,non-approved;RCT,randomizedcontrolled trial;SA,standardapproved;TEP,tissueengineeringproduct.
FeaturesPERSPECTIVE
oftheretrospective,nonrandomized,uncon- trolledobservationalstudydesign.Yet,thiswas outweighedbythemanageablerisksandac- knowledgedunmetmedicalneed.Unmet medicalneedoutweighednonconfirmatory clinicalbenefitandsafetybecauseofuncer- taintyinclinicaltrialdesignforZalmoxis1.A
divergentpositionwasundersignedbythree membersoftheCommitteeformedicinal productsforhumanuse(CHMP).
Glybera1wasapprovedunderexceptional circumstances(EUC)aftera longandexten- siveassessment procedure,involvingmany re-evaluationsbytheCommitteeofadvanced
therapies(CAT)andCHMP [15].Many uncertaintiesaboutquality,efficacy, and safetyledto unfavorablerecommendations for approvaltwice. Beforethefinalre-exam- ination,alackofrobust efficacyoutcomes wasconsideredamajorconcern.Yet,apost- hocanalysisrevealedabeneficialeffectwith TABLE3
Majorissuesmentionedintheassessmentreportsformarketingauthorizationa,b
Drugcategory Authorization Quality Preclinical Clinicaltrialdesign Clinicaloutcome
Approved(N=8) Standard(N=5) Inprocesscontrol(1) Endpoint(1)
Releasespecification(1) Specificreleasetest(1) Conditional(N=2) Specificreleasetest(2)
UEC(N=1) Vector(1) Toxicology(1) Endpoint(1) Efficacy(1)
Specificreleasetest(1) Safety(1)
Non-approved(N=6) Vector(2) Toxicology(4) Design(5) PD(2)
GMPfacility(3) Animalmodel(1) Endpoint(2) GCP(3)
Inprocesscontrol(2) Efficacy(1) Efficacy(6)
GMOtest(1) Safety(5)
Startingmaterial(1) Specificreleasetest(1) Specificreleasetest(4)
aPercategory(quality,preclinical,clinicaltrialdesign,andclinicaloutcome)themajorissuesincludingthenumberofproductsforwhichthatmajorobjectionwasraisedismentioned:
Green,resolvedattimeoffinaldecision;orange,acceptableattimeoffinaldecision;red,unresolvedattimeoffinaldecision.
bAbbreviations:GCP,goodclinicalpractice;GMO,geneticallymodifiedorganism;PD,pharmacodynamics.
TABLE4
Benefit–riskassessmentpercategorya,b
Drug Quality Preclinical Design Efficacy Safety Unmetmedicalneed Benefit–Risk
Standardapproval
Chondrocelect1 ++ +/ + ++ ++
Imlygic1 + + +
MACI1 ++ ++ ++ ++ ++
Provenge1 +/ ++ ~ +
Strimvelis1 +/ ++ ~ ++
Conditionalapproval
Holoclar1 +/ + + ~ ++
Zalmoxis1 +/ +/ ~ +
Approvalunderexceptionalcircumstances
Glybera1 +/ +/ +/ ~ +
Nonapproval
Advexin
CLG
Cerepro + +
Heparesc +/
Hyalograft +/
OraNera
a,unsatisfactory,unresolvedmajorobjections;,uncertainty,concerns,andrisks,trendtowardsunsatisfactory;+/–,neutral,mentionedbutnoclearjudgement;+,uncertainty, trendtowardssatisfactory;++,satisfactory;emptybox,notmentioned;~,unmetmedicalneedconsideredinbenefit/risk-assessment.
bAbbreviation:CLG,ContusugeneLadenovecGendux.
FeaturesPERSPECTIVE
Glybera1forasubgroupofpatients (N=5).
The unmetmedical needfor thissubgroup was crucialto reachapprovalfrom UEC, takingthe ultra-orphanstatusintoconsider- ation. Consequently,thelabelwasrestricted tothispatientgroup.ThefinalCHMPopinion was notsupportedby16 members,who undersigned adivergentposition.
NonapprovalofATMPswasassociatedwith numerousscientificdeficiencies(Table4).Halfof thenonapprovedproductshadanunsatisfactory profileforallscientificevidenceelements.Forall nonapprovedproducts,theclinicaltrialdesign wasregardedasunsatisfactory,whichhindered regulatorsfromevaluatingtheclinicaldata.Pos- itiveresultsrelatedtoqualityandpreclinical studiesweredemonstratedforCerepro.However, anunsatisfactoryclinicaltrialdesignandclinical outcomesresultedinnonapproval.ForHeparesc, theclinicalsafetyprofilewasacceptable,butthe clinicaltrialdesignandclinicalefficacywere judgedtobeunsatisfactory.ForHyalograftonly clinicalefficacywasacceptable,butotheraspects wereunsatisfactory.Forfour(outofsix)nonap- provedproducts,unmetmedicalneedwasac- knowledged,butdidnotoutweighscientific deficiencies.Duringtheapplicationprocedure, fiveoutofsixnonapprovedproductswere withdrawnbythecompanybeforeafinaldecision wasmadebytheregulators.
Pharmaceuticalquality
Thenumerousscientificissuesrelatedtophar- maceuticalqualitydemonstratethatthisdo- mainremainsproblematicintheATMPfield[6].
Amainpharmaceuticalqualityissueinthe submittedapplicationsconcernedthelevelof validationofreleasetestingqualitycontrol(QC) fordifferentclinicaltrialstagesandforapproval.
EUGMPrequirementsappeartobemore stringentcomparedwithotherjurisdictions(e.
g.,USAorJapan)andmightimposedevelop- menthurdles.Inthiscontext,boththerevised first-in-humanclinicaltrialsEUGuidelineandthe EUGMPguidelineforATMPsgivehintsofquality aspects,suchaspotencytestinganduseof biomarkers,althoughtheproofofthatexpec- tationwill‘beintheeating’[16–18].
Potencyalsofrequentlyraisedmajorobjections forbothapprovedandnonapprovedATMPs.ATMP developersexperiencedifficultiesinproperpo- tencytestingbecauseofthelackofsuitableanimal models,withlittleorevennoknowledgeaboutthe mechanismofaction,and,therefore,alsolack validatedbiomarkers.Developerscouldprevent failureduringlate-stagedevelopmentthrough early investment in potency evaluation[19]. Vector- relatedproblemsbelongtothefundamentalde-
velopmentaspectsofsuchproductsandshould havebeenresolvedbeforesubmissionforap- proval.Thisalsoaccountsfornondefinedend- product storage conditionsand shelf-life, which are allassociatedwithnegativeopinionsforapproval.
IncontrasttotheearlydaysofATMPregulation, itisnowpossibletoconditionallyreleasea productbyusingarapid-releasetest.Ourfindings demonstratethatalackofafinalreleasetestwas oftenresolvedbythedevelopmentofarapid- releasetestforapprovedATMPs.Inthisstudy,we analyzedthequalityaspectsthatwerementioned and,thus,discussedintheEPARs.Althoughwe comparethedifferentapprovals,wedonotthink thatthequalityrequirementsdependonthe approvalpathway.However,theobjectionsthat werediscussedintheEPARscouldhaveinflu- encedtheapprovaltype.Furthermore,incom- parabilityofthecommercialproductandclinical trialproductraisedmajorobjections.Thisshould andcouldbeavoidedbyconsideringfuture aspectsofdevelopmentandproperclinicaltrial designduringtheearlystagesofATMPdevel- opment[7,20]topreventwithdrawalsatDay120 forthosedeveloperswhomightnothavethe resourcestotackleresolvablemajorissues.
Clinicaldevelopment
Theobservedsuboptimalclinicaltrialdesigns thatcreateuncertaintyaroundclinicaloutcomes areinlinewithearlierreportsofdevelopment hurdlesexperiencedinthefield[5,21].However, halfofcurrentlyapprovedATMPstargetorphan diseases,forwhichrobustclinicaltrialdesignis notalwayspossibleasaresultofsmallpatient populationsoralackofalternativetreatment [5,22,23].Therefore,ourobservationsofsub- optimalstudydesignsunderexpeditedapproval ofATMPs,suchaslowernumbersofrecruited patients,shouldbeinterpretedwithinthe contextoforphandrugs.Yet,observationsof suboptimalstudydesign,suchasnonrandom- izedtrialdesignwithoutacomparator,areinline withfindingsforconditionallyapprovednon- orphandrugsintheEU[13].
Somemajorconcernsrelatedtoclinicaltrial design,suchasachangeofprimaryendpoint, werealsoraisedforstandardapprovedATMPs.
Yet,regulatorsevaluatedscientificevidenceas sufficientforstandardapproval.Inaddition,un- metmedicalneedwasacknowledgedandtaken intoaccountfordecision-making.Bycontrast,a robustclinicaltrialdesignandclinicaloutcomes aremandatoryforstandardapprovalofconven- tionalproducts[12].ThissuggeststhatEUreg- ulatorsareexploringanappropriateregulatory standardforATMPs,whereconventionalproducts couldbeusedasausefulreference.
Considerationsforbenefit–riskanalysis Here,orphandesignationamongtheapproved ATMPsskewedthelevelofscientificevidencetoa nonconfirmatorynature.Thereisampleconcern that,inthefieldofnotonlyorphandrugs,butalso targetedoncologyproducts,thenatureofevi- dencebecomeslessconfirmatorywiththeuseof nonrandomizeddataandsurrogateendpoints [24].Therelativelyhighnumberoforphandesig- nationsinthefieldofATMPswillimpactthereg- ulatoryconsiderationsformarketingapprovalin thefuture[25].Unmetmedicalneedhasanim- portantroleindecision-makingfortheapprovalof orphanATMPs,providedthatthedatashouldat leastshowsomebeneficialtrendsofefficacyora favorablesafetyprofiletoreceiveapproval.This featureisalsoseeninthefieldofregulatingorphan drugs[11].Yet,considerationsofunmetmedical needdidnotleadtoahigherrateofpositive opinionsonorphandrugapprovalcomparedwith treatmentswithoutunmetmedicalneed[12].This apparentdissimilaritybetweenorphanATMPsand orphannewentitiesneedstobeexploredfurther.
Surprisingly,conditionalapprovalandapproval UECfororphanATMPsarenotprimarilyinitiatedby thedevelopers,butbytheregulators.Inlinewith previouswork,thesefindingssuggestthatcondi- tionalapprovalisfrequentlyusedasarescue optionforapproval[13].For(ultra-)orphanindi- cations,developersshouldtakeconditionalap- provalandapprovalUECintotheirstrategic considerationsformarketingauthorizationinstead ofleavingthistotheregulatorstopropose.
Critically,observationsofalackofclinical efficacyfornonapprovalofATMPsareinline withargumentationfornegativebenefit–risk opinionsonconventionalmedicinalproducts.
Earlierresearchonconventionalmedicinal productsshowedthatbeneficial,clinicallyrele- vantefficacyoutcomesaredeterminantsfor approval[11].Furthermore,ourfindingsindicate thattheprocessofdecision-makingleadingto nonapprovalissimilarbetweenATMPsand conventionalmedicinalproducts.Earlierre- searchshowedthatmajorissuesthatwereun- resolvedatthetimeofthefinaldecisionoften ledtowithdrawalbytheapplicant[21].Strik- ingly,theunresolvedmajorissuesofnonap- provedATMPsunderlinethechallengestothe developmentofATMPs[6,19,26].Glybera1isthe onlyapprovedproductthatappearstobean exceptiontotheruletobeapproveddespiteits uncertainbenefit–riskprofile;itwasapproved afteralongregulatoryprocesswitharestricted labelandmanyuncertainties[27].Currently,the marketingauthorizationholderhasdecidednot toextendthemarketingauthorizationofthe product.
FeaturesPERSPECTIVE
Futureimplications
ThecurrentcentralizedsystemforATMPs,in- cludingCATexpertsandarangeofadvantagesfor ATMPdevelopers,createsopportunitytolearn andgainexperiencewiththeseinnovativepro- ductsaswellastheunderlyingscienceand technology[28].Asthefielddevelops,itisim- portantthatregulatorystandards(incrementally) coevolvetotailorproceduresanddecision-mak- ingfortheseATMPs.Ourobservationsindicate thatEUregulatorsareinclinedtobeadaptive[29]
andtoendorseATMPsforapproval,without compromisingnecessaryevidentiarysupportfor positivebenefit/riskopinions.Therearealsonu- merousregulatoryadaptationsthataretobe implementedsoon(e.g.thenewClinicalTrial Regulation)intheEU.ThesewillalsoaffectATMP development[30].Othershavebeenrecently implemented,suchasthenewregulatorypath- wayforprioritymedicines(PRIME).Manyofthe investigationalmedicinesthatwereincludedin thePRIMEschemeareATMPs[31].Development effortsarealsorapidlyevolving.TheATMPsdis- cussedherereflectastartofahugeclinical developmentpipeline[3–5,30],forwhichappli- cationsforapprovalwillbefiledinduecourse.
Thus,thecurrentanalysisreflectsdecision-mak- ingforasmallsampleoffirst-generationATMPs, makingitdifficulttodrawgeneralizableconclu- sionsforthefuture.Itispossiblethatsome observationsaredrivenbyproductspecificity and/ordiseasecharacteristicsinsteadofbyreg- ulatoryapprovalpathways.Therefore,itiscrucial tocontinuetomonitorregulatoryoutcomesand evaluatetheATMPregulatoryframework.
Concludingremarks
EUregulatorsaremakingimportantstepsinthe fieldofATMPsbybalancingevidentiarysupport andmedicalneedswithcriticalscientificuncer- taintiesthatcouldhampermarketingapproval.
Thedevelopment,regulation,andclinicaluseof mostATMPsarestillcoevolving.Inthiscontext, definingappropriateregulatorystandardstaking intoaccountthecomplexitiesinherittothese productsiscritical.Ourobservationsconcurnot onlywithcurrentdefinedstandardsforATMPs, butalsowiththeavailablespacethatregulations allowforfacilitatedpathways.Aslongastherisks areacceptable,thisappearstobethewayfor- ward.Yet,becauseofthenoveltyandlackof clinicalexperienceinthisfield,regulators,and thoseinindustryandclinicalpracticeneedtopay accurateattentiontopostmarketingsurveillance andrisk-minimizationmeasures,inparticularfor thoseproductswithahighdegreeofscientific uncertaintyuponpointofapproval.Forthefield toflourish,developersandregulatorsneedto
collaboratetocontinuouslymonitorandevolve methodologiesandregulationsforATMPs.
AppendixA. Supplementarydata Supplementarydataassociatedwiththisarticle canbefound,intheonlineversion,athttps://
doi.org/10.1016/j.drudis.2018.03.008.
References
1 Schneider,C.K.etal.(2010)Challengeswithadvanced therapymedicinalproductsandhowtomeetthem.
Nat.Rev.DrugDiscov.9,195–201
2 EU(2004)RegulationEC.1394/2007onadvancedtherapy medicinalproductsandamendingDirective2001/83.EC andRegulation(EC)No/726.Off.J.EU324,121–137 3 Maciulaitis,R.etal.(2012)Clinicaldevelopmentof
advancedtherapymedicinalproductsinEurope:
evidencethatregulatorsmustbeproactive.Mol.Ther.
20,479–482
4 deWilde,S.etal.(2016)Clinicaldevelopmentofgene- andcell-basedtherapies:overviewoftheEuropean landscape.Mol.Ther.3,16073
5 Hanna,E.etal.(2016)Advancedtherapymedicinal products:currentandfutureperspectives.J.Mark.
AccessHealthPolicy4,31036
6 Salmikangas,P.etal.(2015)Manufacturing, characterizationandcontrolofcell-basedmedicinal products:challengingparadigmstowardcommercial use.Regen.Med.10,65–78
7 deWilde,S.etal.(2016)Hurdlesinclinical implementationofacademicadvancedtherapy medicinalproducts:anationalevaluation.Cytotherapy 18,797–805
8Coppens,D.G.M.etal.(2017)Globalregulatorydifferences forgeneandcellbasedtherapies:consequencesand implicationsforpatient accessandtherapeuticinnovation.
Clin.Pharmacol.Ther.103,120–127
9 EMA(2018)EuropeanPublicAssessmentReports.EMA [10] EMAInspections,HumanMedicines
PharmacovigilanceandCommitteesDivision(2017) CATMonthlyReportofApplicationProcedures, GuidelinesandRelatedDocumentsonAdvanced Therapies—July2017.EMA
11 Putzeist,M.etal.(2012)Determinantsforsuccessful marketingauthorisationoforphanmedicinalproducts intheEU.DrugDiscov.Today17,352–358
12 Putzeist,M.etal.(2013)EUmarketingauthorization reviewoforphanandnon-orphandrugsdoesnot differ.DrugDiscov.Today18,1001–1006 13 Hoekman,J.etal.(2015)Useoftheconditional
marketingauthorizationpathwayforoncology medicinesinEurope.Clin.Pharmacol.Ther.98,534–541 14 Monaco,L.andFaccio,L.(2017)Patient-drivensearch forrarediseasetherapies:theFondazioneTelethon successstoryandthestrategyleadingtoStrimvelis.
EMBOMol.Med.9,289–292
15 Melchiorri,D.etal.(2013)Regulatoryevaluationof GlyberainEurope—twocommittees,onemission.
Nat.Rev.DrugDiscov.12,719
16 USDepartmentofHealthandHumanServicesetal.
(2011)GuidanceforIndustryPotencyTestsforCellular andGeneTherapyProducts.OCOD
17 EMA(2017)GuidelineonStrategiestoIdentifyand MitigateRisksforFirst-in-HumanandEarlyClinicalTrials withInvestigationalMedicinalProducts.EMA 18 EC(2017)EudraLexTheRulesGoverningMedicinal
ProductsintheEuropeanUnionVolume4Good
ManufacturingPractice—GuidelinesonGood ManufacturingPracticespecifictoAdvancedTherapy MedicinalProducts.EC
19 Bravery,C.A.etal.(2013)Potencyassaydevelopment forcellulartherapyproducts:anISCTreviewofthe requirementsandexperiencesintheindustry.
Cytotherapy15,9–19
20 deWilde,S.etal.(2017)Understandingclinical developmentofchimericantigenreceptorTcell therapies.Cytotherapy19,703–709
21 Pignatti,F.etal.(2002)Thereviewofdrugapplications submittedtotheEuropeanMedicinesEvaluation Agency:frequentlyraisedobjections,andoutcome.
Eur.J.Clin.Pharmacol.58,573–580
22 Fischbach,M.A.etal.(2013)Cell-basedtherapeutics:
thenextpillarofmedicine.Sci.Transl.Med.5 179ps7 23 Wirth,T.etal.(2013)Historyofgenetherapy.Gene525,
162–169
24 Kesselheim,A.S.etal.(2011)Characteristicsofclinical trialstosupportapprovaloforphanvsnonorphan drugsforcancer.JAMA305,2320–2326
25 Farkas,A.M.etal.(2017)Advancedtherapymedicinal productsforrarediseases:stateofplayofincentives supportingdevelopmentinEurope.Front.Med.4,53 26 CommitteeforAdvancedTherapiesetal.(2010)
Challengeswithadvancedtherapymedicinalproducts andhowtomeetthem.Nat.Rev.DrugDiscov.9,195–
201
27 Watanabe,N.etal.(2015)Re-examinationofregulatory opinionsinEurope:possiblecontributionforthe approvalofthefirstgenetherapyproductGlybera.
Mol.Ther.2,14066
28 Salmikangas,P.etal.(2015)Marketingregulatory oversightofadvancedtherapymedicinalproducts (ATMPs)inEurope:theEMA/CATperspective.Adv.Exp.
Med.Biol.871,103
29 Liberti,L.etal.(2015)Adaptivelicensingandfacilitated regulatorypathways:asurveyofstakeholder perceptions.Clin.Pharmacol.Ther.98,477–479 30 Borán,T.etal.(2017)Clinicaldevelopmentand
commercializationofadvancedtherapymedicinal productsintheEuropeanUnion:howaretheproduct pipelineandregulatoryframeworkevolving?Human GeneTher.Clin.Dev.28,126–135
31 EMA(2018)PRIME:PriorityMedicines.EMA
SofiekedeWilde1,z DelphiG.M.Coppens2,z JarnoHoekman2,3 MarieL.deBruin2,4 HubertG.M.Leufkens2 Henk-JanGuchelaar1 PaulineMeij1,*
1DepartmentofClinicalPharmacy& Toxicology, Leiden University Medical Center, Leiden, The Netherlands
2Utrecht Institute for Pharmaceutical Sciences UtrechtUniversity,Utrecht,TheNetherlands
3InnovationStudiesGroup,FacultyofGeosciences, UtrechtUniversity,Utrecht,TheNetherlands
4CopenhagenCentreforRegulatoryScience,Uni- versityofCopenhagen,Copenhagen,Denmark
*Correspondingauthor.
zTheseauthorscontributedequallytothiswork.
FeaturesPERSPECTIVE
