Introduction to Bioequivalence - BEBAC • Consultancy Services for ...

Introduction to BioequivalenceBioequivalence StudiesREALITY…REALITY…REALITY…REALITY…Reference product (MR / IR)Reference product (MR / IR)Defining study objectivesAdhering to guidelinesAdhering to guidelinesEthical considerationsEthical considerationsProtocol developmentdevelopmentFasting / fedAssessing Assessing clinical clinical and safety safetylaboratory laboratory facilities facilitiesSelecting CROsSelecting CROsSingle dose / multiple multiple doseSelecting Selecting subjectsMoscow, 23 May 20124 • 30

Introduction to BioequivalenceOverviewBioequivalenceSurrogate of clinical equivalence orMeasure of pharmaceutical quality?Types of studiesPharmacokinetic (PK)Pharmacodynamic (PD)Clinical (equivalence and/or safety/efficacy)Moscow, 23 May 2012 5 • 30

Introduction to BioequivalenceOverviewTypes of studies (cont’d)Healthy SubjectsPatientsSingle doseMultiple doseCross-over, replicateParallelReference product (MR, IR, solution)Moscow, 23 May 2012 6 • 30

Introduction to BioequivalenceOverviewTypes of studies (cont’d)Food effectPK interactionDesign IssuesDose regimenFasted / fed stateType of standard mealsBioanalytics (not GLP!)Parent drug / metabolite(s) / enantiomers / pro-drugsValidation / routine applicationMoscow, 23 May 2012 7 • 30

Introduction to BioequivalenceOverviewEthics (GCP!)Dose levels / number of administered dosesNumber / volume of blood samplesDrug and/or adverse effectsClinical performance (GCP!)CRO selectionResponsibilities of sponsor / investigatorAudits / monitoringMoscow, 23 May 2012 8 • 30

Introduction to BioequivalenceOverviewNCA / PK (PD)Sampling scheduleMetrics (AUC, C max ; AUEC, Ae max ,…)Design, methods, evaluationSample sizeEstimation from previous and/or pilot studies,literatureHighly variable drugsBiostatisticsModels & assumptionsProtocol, evaluation, reportMoscow, 23 May 2012 9 • 30

Introduction to BioequivalenceOverview‘What if’-scenariosCommon pitfallsBlind review‘Failed’ studiesDeficiency lettersMoscow, 23 May 2012 10 • 30

Introduction to BioequivalenceAssumptionsWorld ‘Truth’αβModel ‘Data’H0HATheory ‘Reality’Moscow, 23 May 2012 11 • 30

Introduction to BioequivalenceTerminologyBioavailabilityrelative BAComparative BAabsolute BABioequivalenceFood effectPK interactionPilot studyMoscow, 23 May 2012 12 • 30

Introduction to BioequivalenceDefinitionsEMEA Guideline on BE (2010)A bioequivalence study is basically a comparative bioavailabilitystudy designed to establish equivalencebetween test and reference products. Comparative BA, designed to demonstrate BE, reference = innovator’s product.Russian BE Guideline (2008)Two drug preparations are considered to be bioequivalentif bioavailability of drug substance is the same.EMEA Human Medicines Evaluation Unit / CPMPGuideline on Investigation of Bioequivalence(2010)http://bebac.at/Guidelines.htm - EUMinistry of Health and Social Development Russian FederationDrugs Bioequivalence Evaluation (2008)http://bebac.at/Guidelines.htm - RUMoscow, 23 May 2012 13 • 30

Introduction to BioequivalenceBioequivalence…Comparative BAtrue experiment; no bibliographic comparisonDesigned to demonstrate BEvariability,deviation of test from reference,drop-out rate,…to be able (statistical power!) to demonstrate BEReference = Innovator’s product#1: BE [90%–125%]#2: BE [80%–110%]#3: not BE [76%–103%]; (but ‘BE’ to #2)Moscow, 23 May 2012 14 • 30

Introduction to BioequivalenceBioequivalence…EMA GL on BE (2010)Two medicinal products containing the sameactive substance are considered bioequivalentif they are pharmaceutically equivalent or pharmaceuticalalternatives and their bioavailabilities(rate and extent) after administration in thesame molar dose lie within acceptable predefinedlimits. These limits are set toensure comparable in vivo performance,i.e. similarity in terms ofsafety and efficacy.Moscow, 23 May 2012 15 • 30

Introduction to BioequivalenceBioequivalence…Russian GL on BE (2008)Two drug preparations are considered to bebioequivalent if bioavailability of drug substanceis the same. Bioavailability – percentageamount of the drug substance entered systemicblood flow (extent of absorption) and the rate ofthis process (rate of absorption).Moscow, 23 May 2012 16 • 30

Introduction to BioequivalenceGlobal Harmonization?In almost all regulations two PK metrics are necessaryto demonstrate BE, namelyextent (AUC tor AUC ∞) andrate (C max) of exposure.One exception: US-FDA (where AUC t and AUC ∞ mustdemonstrate extent of exposure)Although stated in the GL, such a requirementis statistically flawed. Multiplicity issues (what is the patient’s risk?) Impossible α-adjustment (interdependence)There can be only one!Moscow, 23 May 2012 17 • 30

Introduction to BioequivalenceHistory of BEBioequivalenceProblems first noticed with NTIDs (NarrowTherapeutic Index Drugs) in the late 1970sIntoxications (and even some fatallities!) werereported (warfarin, digoxin, phenytoin)Warfarin, digoxin: Patients switched betweenformulations which were got approval solely basedon in vitro data (innovator ↔generic)Phenytoin: The innovator’s API was changed from amicrocrystalline to an amorphous form resulting in10times higher plasma concentrations in steady stateMoscow, 23 May 2012 18 • 30

Introduction to BioequivalenceHistory of BEBioequivalenceSurrogate of clinical equivalence (1980+)Studies in steady state in order to reduce variabilityStudies based on active metaboliteWider acceptance range if clinical justifiable(not FDA!)Measure of pharmaceutical quality (2000+)Single dose studies preferredGenerally parent drugWidening of acceptance range exceptional(except FDA HVDs and EMA C maxof HVDs)Moscow, 23 May 2012 19 • 30

Introduction to BioequivalenceEarly 1980sFirst methodFDA’s 75/75 RuleBE, if 75% of subjectsshow ratios of 75%-125%.Not a statistic, variableformulations may pass bychance…BE CabanaAssessment of 75/75 Rule: FDA ViewpointJ Pharm Sci 72, 98-99 (1983)JD HaynesFDA 75/75 Rule: A ResponseJ Pharm Sci 72, 99-100 (1983)T R T/R 75%-125%1 71 81 87.7% yes2 61 65 93.8% yes3 80 94 85.1% yes4 66 74 89.2% yes5 94 54 174.1% no6 97 63 154.0% no7 70 85 82.4% yes8 76 90 84.4% yes9 54 53 101.9% yes10 99 56 176.8% no11 83 90 92.2% yes12 51 68 75.0% yes75.0%Moscow, 23 May 2012 20 • 30

Introduction to BioequivalenceMid 1980s IEarly methodTesting for a significantdifference (t-test) at α 0.05Problem: High variability in differences→ formulation will pass (p ≥ 0.05) Low variability in differences→ formulation will fail (p < 0.05) This is counterintuitive andthe opposite of what we actuallywant!T R T–R1 71 81 -102 61 65 -43 80 94 -144 66 74 -85 94 54 +406 97 63 +347 70 85 -158 76 90 -149 54 53 +110 99 56 +4311 83 90 -712 51 68 -17mean 75 73 +2SD 16 15 23CV% 21.4% 20.6% 940%t -table 2.2010t -calc 0.3687n.s.Moscow, 23 May 2012 21 • 30

Introduction to BioequivalenceExample2520Epanutin (Acid, Parke Davis): ReferencePhenhydan (Acid, Desitin): F=151% (p>0.05)Epilan-D (Na-salt, Gerot): F=139% (p>0.05)Difhydan (Ca-salt, Leo): F=22% (p

Introduction to BioequivalenceMid 1980s IILater methodFDA’s 80/20 ruleAt least 80% power to be ableto demonstrate a 20%difference (t-test) at α 0.05 Essentially the 75/75 rule inmore statistical terms. Power 71.5% < 80! (not BE) In any study (even at ‘true’ T=R)with variabilitys 2 n > 6.44it is impossible to show BE!T R T–R1 71 81 -102 61 65 -43 80 94 -144 66 74 -85 94 54 +406 97 63 +347 70 85 -158 76 90 -149 54 53 +110 99 56 +4311 83 90 -712 51 68 -17mean 75 73 +2SD 16 15 23t -table 2.2010t -calc 0.3687n.s.power 71.59%Moscow, 23 May 2012 23 • 30

Introduction to BioequivalenceLate 1980sTOST (Two One-Sided Tests)First formulation of the problembased on equivalencerather than a difference Two One-Sided t-tests Bioequivalent ifp(120%) ≤0.05 Equivalent to a 90% confidenceinterval within an acceptancerange of 80% – 120%DA SchuirmannA Comparison of the Two One-Sided Tests Procedure and thePower Approach for Assessing the Equivalence ofAverage BioavailabilityJ Pharmacokin Biopharm 15, 657–680 (1987)T R T–R1 71 81 -102 61 65 -43 80 94 -144 66 74 -85 94 54 +406 97 63 +347 70 85 -158 76 90 -149 54 53 +110 99 56 +4311 83 90 -712 51 68 -17p(120%) 0.0344p(total) 0.0414T/R 103.32%90% CI (lo) 88.35%90% CI (hi) 118.30%Moscow, 23 May 2012 24 • 30

Introduction to BioequivalenceHuman Guineapigs IBE studies as a surrogate for clinical efficacy /safety (‘essential similarity’)We want to get unbiased estimates, i.e., the pointestimate from the study sample …XˆTestPE =XˆReference… should be representative for the population ofpatients.µTestFPop=µReferenceMoscow, 23 May 2012 25 • 30

Introduction to BioequivalenceHuman Guineapigs IIBE studies as a special case of documentedpharmaceutical qualityThe in vivo release in the biostudy …XˆTestPE =XˆReference… should be representative for the in vitroperformance.f2⎡⎢⎢= 50 ⋅ log ⎢⎢⎢⎢⎢⎣1+t=n∑t=1100⎡⎣R( t) − T ( t)⎤⎦n2⎤⎥⎥⎥⎥⎥⎥⎥⎦Moscow, 23 May 2012 26 • 30

Introduction to BioequivalenceScience → RegulationsWe can’t compare bioavailabilities in theentire population of patientsScientific Reductionism (based on assumptions)‘Similar’ concentrations in healthy subjects willlead to ‘similar’ effects in patients.Equal doses and inter-occasion clearances!DT ⋅ FT DR ⋅ FRAUCT= , AUCR=CLCL[ D = ɶ D , CL = ɶ CL ]FrelT R T R( BA)FFTT= =RɶAUCAUCTRMoscow, 23 May 2012 27 • 30R

Introduction to BioequivalenceModels vs. RealityMoscow, 23 May 2012 28 • 30

Introduction to BioequivalenceA ReminderRoseis a roseis a roseis a rose. Gertrude Stein (1913)Guidelinesare guidelinesare guidelines.Henrike Potthast (ca. 2004)In advanced engineering, you expected failure; you learnedas much from failures as from successes – indeed if younever suffered a failure you probably weren’t t pushing theenvelope ambitiously enough.Stephen Baxter; Transcendent, Chapter 36 (2006)Moscow, 23 May 2012 29 • 30

Introduction to BioequivalenceThank You!Introduction toBioequivalenceOpen Questions?Helmut SchützBEBACConsultancy Services forBioequivalence and Bioavailability Studies1070 Vienna, Austriahelmut.schuetz@bebac.atMoscow, 23 May 201230 • 30