Implementation of Quality by Design at Merck Serono
Transcription
Implementation of Quality by Design at Merck Serono
44th International SFSTP Conference – June 6 and 7 2012 FROM SCIENTIFIC UNDERSTANDING TOWARDS PRODUCT QUALITY (by Design) Implementation of Quality by Design at Merck Serono Lessons learnt from the FDA QbD Pilot Program Dr Hervé Broly – V.-P. Biotech Process Sciences Dr Hervé BROLY - MERCK SERONO 44th International SFSTP Conference June 6-7 2012 Contents Brief Introduction to QbD FDA Pilot Program Outcomes Next Steps and Activities Dr Hervé BROLY - MERCK SERONO 44th International SFSTP Conference June 6-7 2012 Low Biotech Industry Efficiency and High Regulator Workload Triggered New Initiatives The Issue: Observation of disturbing industry trends (Dr. Janet Woodcock, FDA Nov. 2001) Regulatory Agency perception Industry perception Insufficient Product / process understanding Recalls, withdrawals, Approval delays Persistent cGMP violations More inspections, observations More severe enforcement Excessive cGMPs / Regulatory pressure Cost / political pressures Change is difficult, innovation is a risk Quality by Design (QbD) related guidelines: ICH Q8, 9, 10, 11 Dr Hervé BROLY - MERCK SERONO 44th International SFSTP Conference June 6-7 2012 The Response: FDA's "GMPs for the 21st Century“ and the PAT initiative Knowing functional links between clinical safety/efficacy, CQAs and CPPs is key Link Link Clinical safety and efficacy Attributes by design Variants Impurities CQA: a Quality Attribute which, when outside a controlled range, induces a change of efficacy and/or safety Dr Hervé BROLY - MERCK SERONO CPPs: CQAs: Process parameters Raw materials CPP: a Process Parameter which, when set outside of the Design Space, induces a change of a Critical Quality Attribute outside of its acceptable range 44th International SFSTP Conference June 6-7 2012 Quality by Design is an enhanced approach to drug substance and drug product development A systematic approach to pharmaceutical development Development starts with clear pre-defined objectives In-depth scientific understanding of the molecule, the product and the process CQA-driven process and control strategy Natural, “seamless” transition to manufacturing operations Quality risk management throughout development and production Dr Hervé BROLY - MERCK SERONO 44th International SFSTP Conference June 6-7 2012 Poor product/process knowledge causes rigid control strategies and tighter regulatory oversight When the CQAs do not drive the development of the manufacturing process: Process-capability defines target quality Quality of product not “fit for purpose” Limit opportunities for process improvements Inadequate or overly rigid control strategy End-product testing main criterion for Manufacturing operations Tight acceptance criteria Little possibility to adapt process to variability of materials Tighter regulatory oversight Dr Hervé BROLY - MERCK SERONO 44th International SFSTP Conference June 6-7 2012 QbD implies several loops of continuing activities and iterative progression TPP Target molecular properties Product knowledge Q-TSP Large scale Manuf acturing in vitro, non-clinical clinical experience Q-TPP Prior knowledge and literature Process development Process Qualif ication at small scale Process Characterization DoE Critical Quality Attributes Quality Risk Management Critical Process parameters Control strategy Raw Material qualif ication Process knowledge Dr Hervé BROLY - MERCK SERONO Continuous verif ication Def inition of design space Robustness Validation Expanded Change Protocol Quality systems 44th International SFSTP Conference June 6-7 2012 Contents Brief Introduction to QbD FDA Pilot Program Outcomes Next Steps and Activities Dr Hervé BROLY - MERCK SERONO 44th International SFSTP Conference June 6-7 2012 FDA and Sponsors need to gain experience with QbD submissions for biotechnology products FDA QbD Pilot Program aims at: Submitting quality data for biotechnology products Quality by Design principles and risk management Control strategy Expanded change protocol Gain information and facilitate Agency review of quality-by-design dossiers Several meeting with applicants Development of new guidance for industry in partnership with selected biotech companies Dr Hervé BROLY - MERCK SERONO 44th International SFSTP Conference June 6-7 2012 The development of atacicept BDS process using a QbD approach has been discussed with FDA Three face-to-face meetings with the FDA were held between Aug. 2010 and Oct. 2011 + one additional meeting planned in 2012 August 2010: General concepts July 2011: Selection of CQAs and CPPs October 2011: Control Strategy Q4 2012: Continuous verification, Expanded Change Protocol, Engineering Design Space Dr Hervé BROLY - MERCK SERONO 44th International SFSTP Conference June 6-7 2012 The atacicept QbD program encompassed a first meeting to discuss the general concepts Identification of cQAs through risk ranking • Identification of CPPs • Qualification of scaledown models • Process characterization by DOE Implementation of adaptive control strategies Clarified cell culture harvest HMW ≤ 26% 26% ≤ HMW ≤ 33% Capture elution at 3.75 ≤ pH ≤ 4.10 DSP anywhere within Design Space MOR Definition and validation of a global multi-step design space A: Load aggregates 10.9 Cell culture process 9.4 HMW: 2.8 7.9 MOR 6.5 230 5.0 6.0 6.3 6.5 6.8 7.0 C: elution N aC l 210 B: Elution pH Protein A 190 170 MOR 150 130 7.00 4 .5 0 7.25 7.50 7.75 8.00 8.25 8.50 B: elution pH Cation exchange B: Load conductivity Process flow Definition of MOR 3 .8 8 3 .2 5 MOR 2 .6 3 1 0 .9 2 .0 0 7 .0 0 7.5 0 8 .0 0 8 .5 0 A: Load pH Anion exchange A: Load aggregates 6 .5 0 9 .4 7 .9 MOR 6 .5 5 .0 6.0 6 .3 6 .5 B: Elution pH 6 .8 7 .0 Hydroxyapatite Dr Hervé BROLY - MERCK SERONO DS 44th International SFSTP Conference June 6-7 2012 33% < HMW ≤ 41% Capture elution at 3.75 ≤ pH ≤ 4.10 Yes HMW ≥ 41% HMW Capture elution at 3.90 pH 4.10 HMW 15% No DSP steps at centerpoint or BC Reject Step characterizations were used to define a Global Multi-Step Design Space MOR Single step design space determination (MOR) • guarantees drug substance quality for all CQAs • overlap of models of all CQAs defines MOR Cell culture process A: Load aggregates 33.10 28.63 HMW forms: 18 MOR 24.15 Multi-Step Design Space determination: • Start from drug substance • Move up to the cell culture process • Link between steps defined by most limiting CQAs 19.67 230 15.20 3.72 3.85 3.97 C: Elution pH Protein A 4.10 210 C: elution NaCl 3.60 190 MOR 170 150 4.5 0 130 7.00 7.25 7.50 7.75 8.00 B: elution pH 8.25 8.50 B: Load conductivity Process flow Definition of MOR Cation exchange 3.8 8 3.2 5 MOR 2.6 3 1 0 .9 Link between steps: The most limiting CQAs are both input Anion exchange variables (CPP: quality of load) and process outputs (quality of eluate) 2.0 0 7 .00 7 .5 0 A: Load pH 8.0 0 8 .50 A: Load aggregates 6.5 0 9 .4 7 .9 MOR 6 .5 5 .0 6 .0 6 .3 6 .5 B: Elution pH 6.8 7 .0 Hydroxyapatite Dr Hervé BROLY - MERCK SERONO 44th International SFSTP Conference June 6-7 2012 DS An adaptive manufacturing strategy for aggregates delivers DS with QTPP target quality Process flexibility: Control strategy of individual steps adapted to incoming material quality Clarified cell culture harvest Yes No HMW 26% Yes No 26 HMW 33% Yes Capture elution at pH 3.75 Capture elution at pH 3.80 Yes DSP anywhere within Design Space Dr Hervé BROLY - MERCK SERONO Capture elution at pH 3.75 HMW 15% Capture elution at 3.90 pH 4.10 HMW 41% No No DSP steps at centerpoint or BC 44th International SFSTP Conference June 6-7 2012 Reject General QbD development approach seemed to be in good agreement with FDA expectations The general QbD approach is consistent with ICH guidelines The concept of Global Multi-Step Design Space and Design Space confirmation in robustness studies appear adequate Adaptive process control strategies are acceptable But some aspects deserved further discussions Dr Hervé BROLY - MERCK SERONO 44th International SFSTP Conference June 6-7 2012 Identification of CQAs is expected for all molecules – regardless of development approach Start from the complete list of molecule quality attributes Avoid early elimination of CQAs – a comeback is difficult Extracellular Extracellular domain domain of TACI of TACIreceptor receptor • 12 Cys form 6 intrachain • 12 Cys form 6 intrachain disulphide disulphidebridges bridges • • Cleavage Cleavagesites: sites:positions positions32, 32, 49, 81 49, 81 • • Met Metoxidation: oxidation:positions positions2,2,19 19 Fc Fcdomain domainof of human human immunoglobulin immunoglobulin • • 66Cys Cysform form22intra intra/ /22interchain interchain disulphide disulphidebridges bridges • • Potential Potentialcleavage cleavagesites: sites: positions positions137, 137,144, 144,173, 173,176 176 • • Met oxidation: positions Met oxidation: positions118, 118, 294 294 • • Potential Potentialdeamidation deamidationsites: sites: positions positions181 181and and250 250 • • N-glycosylation: N-glycosylation:position position163 163 • • Heterogeneity HeterogeneityofofC-terminal C-terminal Lys: position 313 Lys: position 313 Product properties – – – – Conformation Potency / Product concentration Post-translational modifications Physico-chemical modifications Other quality attributes Product-related substances or impurities Process-related impurities Contaminants Any cQA should be considered critical unless proven otherwise Dr Hervé BROLY - MERCK SERONO 44th International SFSTP Conference June 6-7 2012 Suitability of the Design Space defined at small-scale must be confirmed at large scale Option 1: Large-scale process outside center points ? pH Option 2: Engineering Design Space The definition of Design Space requires to include consideration of all input variables Dr Hervé BROLY - MERCK SERONO 44th International SFSTP Conference June 6-7 2012 Ba s e …the second meeting covered Critical Quality Attributes and Critical Process Parameters TPP Product knowledge Q-TSP Critical Quality Attributes Target molecular properties Q-TPP in vitro, non-clinical clinical experience Prior knowledge and literature Large scale Manuf acturing Process development Process Qualif ication at small scale Process Characterization DoE Critical Process parameters Quality Risk Management COMBINATIONS Binding site Glycation Raw Material qualif ication Cleavage Deamidation Oxidation Mutations FcRn interaction FcγRs interaction Process knowledge Dr Hervé BROLY - MERCK SERONO Def inition of design space Robustness Validation 44th International SFSTP Conference June 6-7 2012 Prot A interaction C1q binding Linkage between CQAs and CPPs is of key importance Develop the approach to establish linkages in great detail Justify the design attributes (e.g. CDR, Absence of Fc – effector functions…) Consider interactions with impurities, input materials, other CQAs Clinical Safety and Efficacy Linkage Linkage cQAs cPPs Attempting to minimise the number of CQAs, CPPs or release tests to be performed may be risky: justify selection based on sound scientific grounds Dr Hervé BROLY - MERCK SERONO 44th International SFSTP Conference June 6-7 2012 Any QA will remain critical until proven otherwise through quality, nonclinical or clinical data Induce variant Molecular modelling Isolate variant Bioassay no Bioactivity? – Model to explain variants CQA yes no Target binding? Biacore CQA yes no In vitro metabolism? Metabolism in serum CQA Fill knowledge gaps yes no PK/PD in animals Rodent study CQA yes no PK/PD data convincing? yes Non critical QA Dr Hervé BROLY - MERCK SERONO CQA – Cleavage products should be evaluated in the CQA assessment – Confirm non-criticality of C-term lysine on PK/PD profile for subcutaneous route 44th International SFSTP Conference June 6-7 2012 Risk assessment needs to build on each other and covers all manufacturing steps Quantitative rankings for impact and severity Dr Hervé BROLY - MERCK SERONO 44th International SFSTP Conference June 6-7 2012 …and the third meeting focused on the control strategy …and overall the Control Strategy ensures that the product is produced to meet all quality targets Find the Critical Control Points Run next QA no Was this the last QA to evaluate? The Control Strategy is a planned set of controls that assures performance and product quality yes End Product knowledge Start Hazard Analysis and Critical Control Points Run next step Is the QA identified as a CQA? Not a Critical Control Point no Critical Quality Attributes yes yes Assign the right testing strategy Does the step increase CQA levels? Does the step decrease the CQA no yes Raw material or equipment qualification no Is this the last step? no no Control Strategy yes Is this the last step where the CQA increases? no Continuous verification yes Process parameter and procedural controls Release and / or stability testing Are the follow steps decreasing the CQA level? yes Critical Process parameters Is CQA below target of LoQ/LOD? no yes yes Is testing at this step necessary for safety or input to next step? Can input material be tested? no yes yes Process targets for critical quality attributes Test input material Establish verification and monitoring process Is testing of the CQA at this step possible? yes no Develop analytical procedure and implement Critical Control Point Design space Trending Is the CQA defined by the process alone, e.g., by a design space? Periodic quality review yes Apply design space Critical Process Parameters (Process parameter control, procedural procedure) no Is the CQA to be defined by analytical procedure ? yes Apply testing controls (IPC, release and/or stability testing) no Is the CQA to be defined by Raw material or equipment qualitication? Quality Risk Management Testing targets for critical quality attributes yes Apply testing controls (IPC, release and/or stability testing) Process knowledge Quality Risk Management Input material controls Raw Material qualification Control strategy Raw Material qualification Quality systems 44th International SFSTP Conference June 6-7 2012 Expanded Change Protocol Process parameter controls Procedural controls Testing Characterisation Stability Testing Large scale Manufacturing Dr Hervé BROLY - MERCK SERONO Process controls Process monitoring Release Testing In Process Testing GMP PPs are in a continuum – a PP not critical inside a given range may be critical outside CPPs determine the desired quality Systematically evaluate all process parameters Carefully control the PP continuum Procedural / process parameter controls ensure they remain in the predetermined range Limits of the PPs Control strategy preserves status at Design Space definition Dr Hervé BROLY - MERCK SERONO 44th International SFSTP Conference June 6-7 2012 Immunogenicity will remain an area of remaining risk The overall risk assessment for immunogenicity should include use of concomitant medications patient immune status assay performance potential for anti-drug antibodies to cross-react with an endogenous non-redundant protein, links between CQAs The evaluation of the criticality of aggregates on immunogenicity is independent of the amount present in the Drug Substance or Drug Product Combine in silico, in-vitro, in-vivo, clinical data to evaluate the impact of product-related substances and impurities on immunogenicity Dr Hervé BROLY - MERCK SERONO 44th International SFSTP Conference June 6-7 2012 Much focus is put on input materials where column resins life time prediction will be a challenge Develop the raw materials qualification concept Impurities derived from non-bioactive raw materials should be considered for potential interaction with the product or other impurities Chromatographic resins have a particular issue : repeated use Stability of performance needs to be ensured over their life-cycle How can this stability be predicted upon arrival of a new batch? Ideas: Online composite quality measure Establishment of quality parameters linking loss of performance to column age for each used resin and performance test at receipt Surrogate marker or proxy, e.g., the most limiting CQA for the step Dr Hervé BROLY - MERCK SERONO 44th International SFSTP Conference June 6-7 2012 The control strategy must cover all relevant QAs and PPs and include a convincing monitoring plan Set acceptance criteria for CQAs based on effects on safety and efficacy Do not base on process capability only Justify experiments performed, trying to find the edge of failure is a good concept Control Strategy - a complex network of controls permitting consistent achievement of desired quality Process targets Testing targets Knowledge management and Process monitoring, e.g., using a SDMS Need to consider developing a robust testing strategy that will enhance process understanding and build confidence in the Design Space Dr Hervé BROLY - MERCK SERONO 44th International SFSTP Conference June 6-7 2012 The quality is in focus, not in process performance or manufacturing costs Avoid to refer to process performance – the Agency is interested in quality Reduce failures / Introduce innovation Closely adhere to concepts and definitions as found in ICH guidances Pre-submission meetings to outline the dossier contents will be a “must” Change management vs Reg. Authorities “What they have always seen before” -syndrome Example : charge variants. Dr Hervé BROLY - MERCK SERONO 44th International SFSTP Conference June 6-7 2012 Contents Brief Introduction to QbD FDA Pilot Program Outcomes Next Steps and Activities Dr Hervé BROLY - MERCK SERONO 44th International SFSTP Conference June 6-7 2012 The main areas for improvement are the input material and variant qualification Develop continuous verification and expanded change protocol concepts Complete the activities in the identified areas for improvement Input material qualification Variant qualification Immunogenicity Scale qualification Dr Hervé BROLY - MERCK SERONO 44th International SFSTP Conference June 6-7 2012 In summary QbD QbDimproves improves product productand and process process understanding understanding Dr Hervé BROLY - MERCK SERONO MercK MercK Serono Serono has hasvalidated validatedits its approach approachto toQbD QbD with withthe theFDA FDA 44th International SFSTP Conference June 6-7 2012 Merck MerckSerono Serono needs needsto todeepen deepen product productvariant variant and andinput input material material knowledge knowledge Starting medicinal product development with the end in mind by applying Quality by Design principles has in-depth product and process knowledge as the principal outcome Dr Hervé BROLY - MERCK SERONO 44th International SFSTP Conference June 6-7 2012