Joint BWP / QWP workshop with stakeholders in relation to prior knowledge and its use in regulatory applications Application of Prior Knowledge for Process Parameter Definition Bob Kuhn, Ph.D., Director CMC Lifecycle Management, Amgen Inc. London, Nov. 23, 2017 1
Process parameter (PP) definition • PP definition requires – Establishment of acceptable ranges in which relevant quality criteria are met – Assignment of c riticality based on potential to impact CQAs • For platform processes and unit operations, there can be strong commonality between PPs and their impact • Effective PP definition requires an effective risk and an inclusive knowledge based framework
Process parameter characterization sorting tool assesses potential criticality, risks and knowledge requirements • Assess risk related to process excursions for each PP and CQA: – Severity (S) of the impact of a PP excursion – Occurrence (O) frequency of an excursion outside acceptable performance – S x O = Relative Risk (RR) Potential CPP, In-depth knowledge required Higher severity of to assess criticality and justify range impact to CQA Score S & O Higher RR Lower severity of Non-CPP, does not impact to CQA Lower RR require additional studies (document justification) Prior knowledge is an essential input to enable focus on high risk parameters 3
Prior Knowledge Assessments (PKA’s) can be applied to systematically analyze platform process data Can be view as “experiments”, addressing specific question(s)… Except using historical data as the “laboratory”
PKAs process borrows from the principles used for Systematic Reviews Frame the Question: (i.e. “Does unit process parameter X control product quality Y in step Z”?) Materials: Identification of prior knowledge sources • Relevance requirements are based on the question • Reliability requirements are based on how the PKA is to be used Methods: Develop processes for data consolidation and analysis. Review: Compile and consolidate and analyze information from sources. Documentation: Conclusions, recommendations.Does the data meet a burden of proof?
Example - Process Impact Rating (PIR) applied to identify the most impactful operating parameters 4 4 4 3 Large Effect Effect Magnitude 4 3 3 3 Small CQA Effect No 2 2 1 1 Effect <1X NOR >1X - <2X NOR 2X – 3X NOR > 3X NOR Perturbation Magnitude Process Parameter Normalizes quantitative impact across products and processes to assess relative impact and importance
Number Report 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 Example - assessment of process parameter impact for James E. Seely, Roger A. Hart, Prior Knowledge Assessments, BioProcess International, 10, 9, 2012 Product M M M M M K G G D C B B L J H F E E A I + + + + + + + + Operating parameter 1 + Operating parameter 2 - - chromatography step for one CQA Operating parameter 3 - Higher risk operating Operating parameter 4 parameters Operating parameter 5 Operating parameter 6 Operating parameter 7 7 Operating parameter 8 Operating parameter 8 Operating parameter 9 + + + + + Operating parameter 10 + + + + + + Operating parameter 11 Operating parameter 12 Operating parameter 13 Operating parameter 14 - Operating parameter 15 + + Operating parameter 16 Operating parameter 17
Case study – prior knowledge assessment for cation exchange chromatography for platform MAb process • Chromatography step option for platform MAb processes • Operated in bind and elute mode • Primary purpose is clearance of impurities Systematically evaluated process design and characterization data from 14 MAb products, as well as extensive manufacturing data. Methodology for this analysis described in Seely and Hart, Prior knowledge Assessments - Leveraging Platform Process Experience to Develop Targeted Process Characterization Strategies, Bioprocess International, October 2012
Extensive platform data clearly identify high risk parameters (radial plots of normalized impact) Impurity 2 Impurity 1 Load Rate (g/L resin) Resin lot / ligand density 0.8 Load pH Load Rate (g/L resin) Bed Height Load cond 0.5 Bed Height Load pH 0.6 0.4 Temperature Load cond Temperature Load treatment 0.4 0.3 Flow rate Load treatment 0.2 0.2 Flow rate Equil pH 0 0.1 Gradient length Equil pH 0 Gradient length Equil Conductivity / concentration -0.2 -0.1 Equil Conductivity / Stop Collect concentration Stop Collect Equil volume Start Collect Equil volume Start Collect Wash pH Elution salt Elution salt concentration/cond Wash cond./ concentration Wash pH concentration/cond Elution buffer pH Wash volume Elution buffer pH Wash cond./ concentration Elution buffer Concentration Elution buffer Wash volume Concentration Same process parameters impact impurities 1 and 2
Extensive manufacturing data across multiple processes indicate the load impurity levels markedly impact impurity 1 and 2 clearance 1 Impurity 1 1 Impurity 1 Product A Product B 0.1 0.1 0.01 0.01 Load Pool Load Pool Impurity 2 1 1 Product A Impurity 2 Product B 0.5 0.5 0 0 Load Pool Load Pool Example plots – observed for multiple products
Prior knowledge assessment resulted in informed, focused, and effective process characterization PKA Findings PC Strategy • • High risk parameters clearly Focus PC on small number of identified potential critical parameters • • Parameter interactions not Perform feed challenge/spiking practically significant studies to: • Assess clearance capability • No impact of raw materials • Establish performance (including resin) requirements for prior step(s) • Feed stream quality impacts step • Inform control strategy testing performance for impurities 1 and 2 requirements • Significant excess clearance capacity for impurities 3 and 4
Recommend
More recommend
