Applying quality by design to vaccine development - BioProcess ...

Application of Quality byDesign to Vaccine DevelopmentCambridge Healthtech Institute’s Seventh AnnualThe Immunotherapeutics and Vaccine SummitAugust 13-16, 2012Cambridge, MASheila G. Magil, Ph.D.BioProcess Technology Consultantswww.bptc.com

Presentation Outline‣ What is QbD, briefly• Overview― History― Goals• Advantages and Disadvantages‣ Why apply QbD principles to vaccine development• Vaccines are not monoclonal antibodies• QbD application at each development stage‣ Example2 From Clone to Clinic®

ICH QbD Guidance‣ Q8 Pharmaceutical Development‣ http://www.ich.org/fileadmin/Public_Web_Site/ICH_Products/Guidelines/Quality/Q8_R1/Step4/Q8_R2_Guideline.pdf‣ Q9 Quality Risk Management‣ http://www.ich.org/fileadmin/Public_Web_Site/ICH_Products/Guidelines/Quality/Q9/Step4/Q9_Guideline.pdf‣ Q10 Pharmaceutical Quality System‣ http://www.ich.org/fileadmin/Public_Web_Site/ICH_Products/Guidelines/Quality/Q10/Step4/Q10_Guideline.pdf‣ Q11 Development and Manufacture of Drug Substances‣ http://www.ich.org/fileadmin/Public_Web_Site/ICH_Products/Guidelines/Quality/Q11/Step_2/Q11_Step_2.pdf4 From Clone to Clinic®

Mab vs VaccinesDifferences Effecting Development StrategyMAb Vaccines ImplicationsUsually well‐characterized Often difficult to characterize Less definitive analyticalcomparabilityMore difficult to monitorproduct quality mid‐ processClear Link to MOAMay have biomarker surrogatefor clinical performanceConsistent process andproductTherapeutic patientpopulationWell‐understood process gooddetectability for test methodsDifficult to establish clinicalpotency surrogateMore complex, lesspredictable process/productProphylactic patientpopulationLess understood process,difficult to measure attributechangesChallenges to improve processpost‐licensureVariability overproduct/process life cycle“Process is Product”philosophyEmpirical process models forlinking parameter inputs toquality outputsMore stringent threshold forreporting manufacturingchanges5 From Clone to Clinic®

QbD for Vaccines‣ A‐VAX: Applying Quality by Design to Vaccines‣ http://www.pda.org/Home‐Page‐Content/CMC‐VWG‐A‐VAX.aspx‣ Produced by the CMC‐Vaccines Working Group• GSK, MedImmune, Merck, Pfizer, sanofi pasteur‣ Placed in the public domain by PDA‣ Use the A‐Vax study product and process description for general discussion ofQbD approach for development of vaccines6 From Clone to Clinic®

What is QbD?‣ Quality by Design (QbD) is a concept first outlined by well‐known quality expertJoseph M. Juran in various publications, most notably Juran on Quality by Design.Juran believed that quality could be planned, and that most quality crises andproblems relate to the way in which quality was planned in the first place.‣ QbD identifies characteristics that are critical to quality from the perspective ofpatients, translates them into the attributes that the drug product should possess,and establishes how the critical process parameters can be varied to consistentlyproduce a drug product with the desired characteristics.‣ Systematic process to build quality into a product from the inception to finaloutput.‣ Quality by Design is a systematic approach to development that begins withpredefined objectives and emphasizes product and process understanding andprocess control, based on sound science and quality risk management.7 From Clone to Clinic®

What Does That Mean?‣ Systematic approach to development• Identify the customer• Use tools to prioritize activities• Iterative process building on knowledge gained‣ Predefined objectives• CQAs• QTPPs‣ Emphasizes product and process understanding and process control• Identification of critical process parameters linked to CQAs• Prior knowledge‣ Based on sound science• First principles understanding‣ Quality risk management• Use risk assessment tools consistently• Revise risk assessments as knowledge is gained8 From Clone to Clinic®

QbD = Enhanced Product Quality‣ The product is designed to meet patientneeds and performance requirements‣ The process is designed to consistently meetproduct critical quality attributes‣ The impact of starting raw materials andprocess parameters on product quality is wellunderstood‣ Critical sources of variability are identifiedand controlled through appropriate controlstrategiesDesign Space: “The multidimensional combinationand interaction of input variables (e.g., materialattributes) and process parameters that have beendemonstrated to provide assurance of quality”Ref: AS Rathore & H. Winkle, Nature Biotech 2009H. Winkle, BPI Conference 2010From Clone to Clinic®

QbD Based on Scientific & Risk Based ApproachRef: D. Low, BPI Conference 2010From Clone to Clinic®

Critical Quality Attributes“…those physical, chemical, biological, or microbiological properties or characteristicsthat should be within an appropriate limit, range, or distribution to ensure the desiredproduct quality…” ‐ ICH Q8Risk (Criticality) α Impact x UncertaintyRisk that an attribute impactssafety or efficacyImpact attribute has on safetyor efficacy. Determined byavailable knowledgeUncertainty in assigning impact– determined by relevance ofknowledgeManufacturabilityStability Immunogenicity= Safety ?11From Clone to Clinic®

What are Critical Process Parameters (CPPs)‣ Independent process parametersmost likely to affect CQAs, including• Purity• Qualitative and quantitativeimpurities• Physical characteristics• Microbial quality‣ Determined by sound scientificjudgment and based on research,scale‐up or manufacturing experience‣ Controlled and monitored to maintainCQAs12From Clone to Clinic®

Critical Process Parameters Drive Critical Quality AttributesRange of Raw MaterialAndFacility AttributesProcess Designed toLimit Product Variability‣ Define Drug Substance in terms of CQAs‣ Identify CPP that affect these CQAs‣ Determine range of each CPP that produces acceptableproduct to establish the Design Space

Knowledge SpaceRef: B. Davis, FIP Conference 2007

Design Space‣ Design space is proposed by the applicant and is subject to regulatory assessmentand approval• Any process that runs within previously defined and approved design space isnot considered a change• Operation outside of previously approved design space requires a regulatorypost‐approval change processRepeat UnitsYieldRM 1RM1From Clone to Clinic®

Creating a Design Space: Putting it All Together‣ Summarize results ofmultiple experiments todefine response surface‣ Validate the process todemonstrate consistentproduction of productwith desiredcharacteristics% LysisFeed RateDesign SpaceRaw Material 1RepeateUnitsFrom Clone to Clinic®

Downstream Design Space:Ion Exchange Column Load Conditions‣ Purification of Ps by anion exchange chromatography‣ Column equilibrated with 10 mM TRIS‐Phosphate buffer‣ Variation of load solution pH will impact product yield and purityProduct peak(s)From Clone to Clinic®

Design Space: Anion Exchange Column Yield and Purity‣ Best yield at pH 7.0‣ However, additionalcontaminant present inpool that is not seen athigher pHProductContaminant‣ Can subsequentprocess steps removethis contaminant?From Clone to Clinic®

Overall Design Space: Final Product PuritySubsequent process step removes process contaminantregardless of pH at which Anion Exchange column is runA –Anion exchange column pool, pH 7.0B–Anion exchange column pool, pH 8.6C–Column 2 pool following loading with “A”D –Column 2 pool following loading with “B”From Clone to Clinic®

QbD Work Flow for Development of a VaccinePotential CQAsInterim CQAsFinal CQAControl StrategyPlatform KnowledgeTox Study RangesNon‐clinical studiesDose rangingProcessValidationFew Tests andNarrow AcceptanceRangesMultiple Tests withWide Acceptance RangesBLA SubmissionPhase 1 Phase 2 Phase 3 Commercial21 From Clone to Clinic®

Key Planning Step: Identify the Customer‣ Overall patients, physicians and Regulators‣ Each process step may have different customers‣ Upstream customer is downstream‣ Downstream customer is formulation,‣ Do not forget that at each step the customer is also the manufacturing, quality andregulatory groups22 From Clone to Clinic®

Next Step: Target Product Profile (TPP) from A-VAX Case StudyParameterMechanism of ActionIndicationPrimary EndpointsKey ClaimsSecondary EndpointsFormulation/DosingApprovals andRecommendationsDescriptionA‐VAX is a pentavalent vaccines containing the capsularpolysaccharide (Ps) of 5 serotypes, each linked to a recombinant noninfectiousvirus‐like particle (VLP) and adjuvanted with Al salt.Expected to produced enhanced cellular and humoral antigen specificprotective immunityActive immunization of 2‐60‐month old infants and children forprevention of disease related illnesses due to causative agent70% reduction of confirmed disease within 1 year after dosing in thetarget population, Safe and tolerableFavorable risk‐benefit ration, can be dosed with other pediatricvaccines, achieves WHO stability requirementsReduction of disease specific hospitalizations and ER visits, reductionof causitive agent specific disease, duration of protection > 1 yearAntigen/Adjuvant in pre‐filled syringe or single dose vial, containerspre‐mixed prior to injection, 3 doses administered 2 months apart –preferably on standard pediatric vaccine scheduleExpecting Advisory Committee on Immunization Practices and otheruniversal recommendations23 From Clone to Clinic®

Quality Target Product ProfileParameterKey ClaimsFormulation/DosingDescriptionEasy to administer, 0.5 mL subcutaneous delivery in an outpatient settingusing a 1 mL syringe (27g x ½ inch needle)Stable 2 years at RT or 4 years at 2‐8°CStability following reconstitution:24 hours physical and chemical (2‐8°C) or 8 hours at RTNo animal or human derived material used in manufactureSterile product, DS (Ps‐VLP) can be sterile filtered3 doses (containing 5g each of 4 Ps and 50g of fifth Ps adsorbed to300g Al adjuvant) administered 2 months apartLyophilized and reconstituted with standard dilutents contaiing adjuvant,rapid reconstitution profile with suitable viscosityComposition: sugar, surfactant, buffer (isotonic and normal pH) and Ps‐VLPLabel volume 0.5 mLPrimary packaging: Single dose vial, clear Type I glass, latex‐free stopperand seal with flip‐off plastic buttonSecondary packaging and shipping: allowed shipping excursiontemperature 2‐40°C for 3 days (10 vials/carton)24 From Clone to Clinic®

Control Strategy Vocabulary 1TerminologyQuality AttributeCritical Quality Attribute(CQA)Less Critical QualityAttributeAcceptance CriteriaPerformance AttributeDefinitionA physical, chemical, biological or microbiological property orcharacteristic of the product whose variability might have potentialimpact on the safety and efficacy of the product. At early stages ofdevelopment some of these are liley to be “potential CQAs”A physical, chemical, biological or microbiological property orcharacteristic of the product that should be within an appropriatelimit, range or distribution to ensure the desired product quality (ICHQ8 (R2))A quality attribute determined through risk analysis to be less criticalto assurance of desired product quality efficacy and safetyNumerical limits, ranges or other suitable measures for acceptancewhich the DS or DP or intermediate materials should meet toconform with the specification of the results of analytical proceduresA physical, chemical, biological or microbiological property orcharacteristic of the product whose variability might have a potentialimpact on process performance25 From Clone to Clinic®

Control Strategy Vocabulary 2TerminologyKey PerformanceAttribute (KPA)Critical ProcessParameter (CPP)Key ProcessParameter (KPP)Design SpaceFormalExperimental DesignDefinitionA parameter that when controlled ensures optimal process performanceA process parameter whose variability has an impact on a CQA andtherefore should be monitored or controlled to ensure the processproduces the desired qualityAn adjustable process parameter of the process that when maintainedwithin a narrow range ensures optimum process performance. A KPPdoes not meaningfully affect CQAs. Ranges for KPPs are establishedduring process development and changes to these ranges will bemanaged within the Quality SystemThe multidimensional combination and interaction of input variables andprocess parameters that have been demonstrated to provide assuranceof qualityA structured, organized method for determining the relationship betweenfactors affecting a process and the output of that process (DOE)26 From Clone to Clinic®

QbD for Vaccines27From Clone to Clinic®

Wide Range of Technologies Available for Vaccine ProductionToxoid(Bacterial)Live,attenuatedVirus‐likeparticleInactivated Subunit DNAKilled,Empirical Recombinant metabolically Purified RecombinantactiveVectorConjugate(e.g., proteinpolysaccharide)From Clone to Commercial®

Example:Application of QbD to Influenza VaccinesProcesses adapt to Seasonal Strain VariationsFrom Clone to Clinic®

Influenza Virus Types‣ Influenza Type A• Infects humans, birds, pigs, horses• Most prevalent, occurs in several Sub‐Types• Most likely to demonstrate Antigenic Drift Shift• Major Antigenic Shift may lead to pandemic‣ InfluenzaType B• Infects humans, seals• Less volatile, no sub‐types• Antigenic drift, slower than Type A‣ Influenza Type C• Infects humans• Comparatively rare –not included in standard vaccinesFrom Clone to Clinic®

Diagram of Influenza VirusHemagglutinin HANeuraminidase NAFrom Clone to Clinic®

Identification of New Influenza Virus Strains‣ New ‘A’ viruses arise by recombination or re‐assortment (dual infections) inanimals, birds‣ Most often in the Far East• People living in close proximity to pigs, ducks, etc.‣ Chain of WHO Influenza Diagnostic Laboratories• Constantly checking new isolates for antigenic “Shift” versus “Drift”‣ Requirement for change in vaccine strains when significant drift or a major shiftoccurs, MOST OFTEN IN TYPE A.‣ Notified by WHO in January ‐ March for Western World.‣ CDC, Atlanta GA issues new strains for manufacturers to use‣ These may need additional adaptation to eggs or cell cultureFrom Clone to Clinic®

A Wide Variety of Influenza A Strains ExistH1N1 H1N2H2N2H3N1 H3N2 H3N8H5N1 H5N2 H5N3 H5N8 H5N9H7N1 H7N2 H7N3 H7N7H9N2H9N7•H1N1 was the 1918 “Spanish Flu” type: 20% mortality world‐wide•H1N1 variant was the 2009 Pandemic flu•H2N2 caused the 1957 “Asian Flu”•H3N2 caused 1968 “Hong Kong Flu”•H5N1 is the type associated with “Avian Flu”

Influenza Vaccine Production: Viral PropagationRecombinant Seed Virus (New Isolate X Well‐growing strain)PrepareSubstrateInfect &IncubateRemove Cells,Purify VirusInactivate VirusEgg based…IncubateEmbryonated Eggs(1,000s needed)Cell based…Scale up cellculture from WCBInoculateAllantoic SacInfect inBioreactorHarvest AllantoicFluid, (10‐12ml/egg)Clarify, Ultra‐CentrifugeCentrifuge, TFF,ChromatographyTreat withformaldehydeSub‐unit vaccines arethen solvent‐extractedAttenuated live vaccinestreated to limitinfectivityFrom Clone to Clinic®

Recombinant Influenza VaccineIdentifyTarget StrainObtain sequence of antigeniccomponents (HA +/‐NA)GeneraterecombinantexpressionconstructGenerally requires no more than two weeksProduceRecombinantprotein or VLP onLarge ScaleHarvest, purify(protein or VLP)Fill/finish similar to standardbiopharmaceuticalsFrom Clone to Clinic®

Baculovirus Expression of Influenza Antigens• Influenza antigens can be expressed by cloning into Baculoviruses and growing ininsect cell culture– Baculovirus has limited host range (No infection of mammalian cells)– Rapid scale‐up and production of new vaccine strains possibleIsolate Target GeneCreate RecombinantBaculovirus StrainGrow Up a Batch ofInsect CellsInfect & IncubateHarvestInfected CultureClarify HarvestOrNovavax: Purify VLPCollect CellsProtein Sciences: Extract & Purify rHAFrom Clone to Clinic®

expresSF+ Engineered for Vaccine Manufacturing‣ Cell line developed from Sf9 Cells• Selective pressure in serum‐free media withadded insulin (0.4 mM)• unique phenotypic and genotypic properties‣ Protein Sciences (Meriden, CT, USA) has developed FluBlok®using their proprietary BEVS baculovirus expression system• Contains three recombinant HA antigens― Dose of each antigen is 3X egg‐produced dose• Over 50,000 doses delivered with minimal adverseevents― Highly effective protective immunity generated• BLA filed, awaiting approvalUninfected SF+Infected SF+From Clone to Clinic®

Protein Sciences FluBlok ® ExpressionBaculovirus Expression Vector System (BEVS)• Engineer baculovirus with thegene of interest (e.g.Hemagglutinin)• Baculoviruses highly specific toinsect cells• Powerful promoter generateshigh yield of protein of interest• Culture expression ofinsect cells in afermenter• Infect cells withengineered virus• Incubate infection for ~48 ‐72 hours• Protein forms rosettes• Purify protein to > 90%into final product• Formulate with PBS intovaccineFrom Clone to Clinic®

QbD in Influenza Vaccine Case Study: Flublok ®‣ Flublok must be changed each year to produce the current strains• In 2011 the strains were the same for US as in 2010• Impact of strain changes on antigen biochemistry can be conservative orsignificant—Driven by epidemiology• Process must cope with annual potential change of input antigen whileoperating within predefined design space—SOP in place at Protein Sciences to drive process adaptation to newstrains

FluBlok ® Production – Downstream ProcessStep Method Purposeupstreamdownstreamfermentation bioreactor rHA expressionharvest centrifugation cell recoveryextraction non‐ionic detergent rHA solubilizationclarification depth filtration Turbidity reductionIEXCation exchange (SP orCM)rHA capture, purificationHIC Phenyl sepharose rHA purificationQ membrane Mustang Q DNA removalUltrafiltration TFF Buffer exchangeFrom Clone to Clinic®

Process Steps have Differing Sensitivity to ChangesEvaluation of Risks in Process StepsupstreamdownstreamstepfermentationharvestIsotypevariability(H1, H3, B)Seasonalstrain variabilityuniversaluniversalVariableparametersextraction yes infrequent pH; [detergent]clarification yes no DF Filter typeIEXHICQ membraneTFFyes often pH; [salt]; columntype; [detergent]yes some pH; [salt];[detergent]universaluniversalFrom Clone to Clinic®

Qbd: Flexible Response and Process Consistency‣ How to respond quickly to seasonal strain changes while maintaining a stable,validated downstream production process?Seasonal strain changes:Necessitate rapid processadjustmentsProcess Validation:FDA requires process consistencyunder the FluBlok BLA.Commercial‐scale production:Requires a stable process platformProcess Development Group:requires guidance in order to workrapidlyFrom Clone to Clinic®

PSC SOP RG0006: a Framework for FluBlok ®Seasonal Strain-change Development‣ Applied to the process development for each new FluBlok component strain‣ Developed with FDA guidance to define acceptable modifications in downstreamprocess parameters within the scope of a validated process.‣ Assures FDA that:• any process changes are necessary and minimal.• process performance remains consistent season to season‣ Provides PSC Process Development group with guidance so that:• development procedures and acceptance criteria are clearly defined.• PD can work rapidly, knowing that process changes within RG0006 do notrequire regulatory modification.‣ Process changes outside of RG0006 require regulatory modification (ChangeControl)SOP RG0006 guides the work flow anddefines the downstream process design spaceFrom Clone to Clinic®

Process Development Under SOP RG0006‣ What does RG0006 tell the Protein Sciences Process Development group?• Organizes work flow between small‐scale and large‐scale.• Specifies scale‐down models.• Specifies conditions which may be altered.• Defines acceptance criteria for each step• Provides a small‐scale validation of process performance prior to transfer.• Specifies reporting activities for process transfer.From Clone to Clinic®

RG0006 Governs the Variable Processing StepsupstreamdownstreamProcessstepIsotypevariability(H1, H3, B)SeasonalstrainvariabilityVariableparametersModifiableunder RG0006?fermentation universal noharvest universal noextraction yes infrequent pH; [detergent] yesclarification yes noIEX yes oftenHIC yes someDepth filtertypepH; [salt];column type;[detergent]pH; [salt];[detergent]noyesyesQ membrane universal noTFF universal noOnly process steps with seasonal variabilityare modifiable under RG0006From Clone to Clinic®

Process Development under RG0006:Road Map to a New Seasonal ProcessTry the process for the previous year’sisotypic strain at small scaleWorks?Doesn ’t work?Evaluate at full scale(single cGMP run)Small‐scale development& optimizationWorksDoes not workSmall‐scale validation(limits for pH, etc.)cGMP full‐scale evaluationFinal full‐scaleSeasonal ProcessFrom Clone to Clinic®

Key Changes to H1 DSP Between 2009 & 2010‣ H1 Process modifications were successfully evaluated at small‐scale in PD prior totransfer to manufacturingProcess step H1 2009 H1 2010 RationaleExtraction 0 mM NaCl 50 mM NaCl Improved yieldIEX –wash 2 0 mM NaCl 100 mM NaCl Increased purityIEX ‐ elution 100 mM NaCl 250 mM NaCl Increased step‐yieldHIC ‐ equilibration 0 % CHAPS 0.05% CHAPS Improved purityFrom Clone to Clinic®

FluBlok Summary‣ FluBlok production requires evaluation and modification in response to seasonalinfluenza strain changes.‣ SOP RG0006 provides a PD & regulatory framework for adaptation within avalidated process.‣ RG0006 provides multiple routes to development of a full‐scale process.‣ Process changes outside of RG0006 may be addressed through Change Control(BLA modification)FluBlok and PanBlok ® development supported byHHS contract HHS0100200900106CFrom Clone to Clinic®

QbD and Influenza Vaccine-An Existing Paradigm‣ Influenza vaccines have always been manufactured within a design space• Annual strain changes impact egg production as much as recombinantproduction‣ Provides an operational case study for FDA review and implementation to otherproduct classes‣ Enables regulatory flexibility and adaptation to emergency or rapid responserequirements‣ Quality by design as a regulatory shift will, in the long run, facilitate manufacturingdecisions and product approvals for vaccines and other productsFrom Clone to Clinic®

Product Control Through Process ControlCell lineProductConsistencyMedia and FeedsPotencyCulture TimeQbDStructureColumn LoadingIntermediate HoldTimePATStabilitySafetyFormulation BufferEfficacyFrom Clone to Clinic®

Acknowledgements‣ BioProcess Technology Consultants• Howard Levine, Ph.D.• Alex Kanarek, Ph.D.• Susan Dana Jones, Ph.D.‣ Protein Sciences• Robert Boulanger, Ph.D.• Al Price, Ph.D.• Dan Adams, Ph.D.From Clone to Clinic®

Thank You!Sheila G. Magilsmagil@bptc.com1-781-281-2710BioProcess Technology Consultants, Inc.12 Gill Street, Suite 5450Woburn, MA 01801

Follow us…www.bptc.comwww.bioprocessblog.combptcGlobalwww.linkedin.com/company/bioprocess-technology-consultants-inc.