Setting Specifications How and Where to Control Karin Sewerin on behalf of EBE MedImmune September 9, 2011
Specifications – Presentation outline • Monitoring of Critical Quality Attributes (CQAs) • Testing Types of testing Types of acceptance criteria • Handling of CQAs and non-CQAs 2
Rationale for Setting Specifications Molecular Molecular structure structure Methods Methods Impurities Impurities validation validation Process Process Stability of Stability of Specifications Specifications evaluation evaluation product product Manufacturing Manufacturing Regulatory Regulatory process and process and requirements requirements control control Batch to batch Batch to batch variation variation 3
Starting point - for setting specifications •The ideal case o No Critical Quality Attributes have been identified o Stable and robust process has been established Drug Product Specifications Attribute Test Identity Suitable method Potency Relative activity Content A 280 Purity Endotoxins Safety Sterility • The real case o Add each CQA identified o Select a suitable method 4
Risk Assessment - for control of CQAs Detectability Detectability - capability of - capability of analytical methods analytical methods Probability Probability Severity - relationship of Severity - relationship of Define CQAs product knowledge Define CQAs product knowledge – impact on S&E and process – impact on S&E and process capability capability Design of Control Strategy 5
Control Strategy – for each step and the process Procedural controls Procedural controls Output I nput Control Control Process step Examples: Examples: •RM Spec •IPC on eluate •IPC on eluate from this step from previous •IP test step •DS/ DP Specs •IP test Routine Monitoring 6
Testing - for routine monitoring of QAs 1. In Process Controls (IPC) Confirm QAs on each batch Control input for next step Real Time Release Testing (RTRT) 2. End Product Testing Drug Substance and Drug Product Skip lot testing 3. Life cycle management Real time stability monitoring Accelerated stability testing Comparability testing 7
Example – Real Time Release Testing • Mab – Oligosacharide profile Criticality assessment Impact on: Biological activity / Immunogenicity / Clearance? No - attribute is NOT a CQA – no routine monitoring required Yes - attribute is a CQA - should be monitored • Prior knowledge Glycosylation is formed during cell culture production and doesn’t change during down stream processing or storage • In-Process testing in lieu of End Product testing CQA testing is performed on the harvested bulk Report result In process Control in Batch records On Certificate of Analysis with reference place of testing 8
End Product Testing – DS vs DP • Drug Substance vs Drug Product testing No DS release testing required Direct filling of Purified DS DP Certificate of Analysis includes testing of all CQAs DS release is required if Storage of DS before fill Pooling – release testing to confirm product quality before mixing of several batches 9
Life cycle management – testing strategy Specification Testing • Stability testing Shelf life specifications Routine monitoring on a periodic basis • Comparability testing Assessment of impact after process changes Investigational Testing • Forced degradation study / Accelerated stability Initial assessment of the molecular capability and understand the degradation pathway Selection of stability indicating methods Identify critical condition in the process Formulation development 10
Types of Acceptance Criteria Examples • Quantitative Protein Concentration: Protein Concentration: 10 ± 1 mg/mL 10 ± 1 mg/mL Range < X < Host Cell Protein: “Open ended” Host Cell Protein: < 10 ppm < 10 ppm Not More Than Not Less Than Identity: Identity: Profile comparable to Profile comparable to • Qualitative Reference Material* Reference Material* Compare to reference material * To a well characterized Reference Material 11
Quantitative Acceptance Criteria Target / Set point Out of trend Out of trend Alert Lim its for Regulatory monitoring Acceptance Lim it of process performance 12
Quantitative Acceptance Criteria Target / Set point Out of trend Out of trend Alert Lim its Acceptance Acceptance Action Lim its -- Action An action - An action - to confirm that the Examples: to confirm that the •Corrective action quality of the quality of the •Additional testing product product •End product testing 13
Quantitative Acceptance Criteria Target / Set point Out of trend Out of trend Alert Lim its An action to An action to confirm the confirm the quality of the quality of the Action Lim its -- product product Acceptance Lim its Out Of Out Of OOS investigation Specifications Specifications Acceptance / Rejection 14
Example - alert limits x x x x x x x x x x x Acceptance x x limits x x x x x 15
Example - alert limits x x x x x x x x x x x Acceptance x Alert x limits x x Limits x x x 16
Example - action limits Control of HCP Process Validation In Process Control Action if > 500 000 ppm Affinity Affinity 5 – 10 000 ppm > 10 000ppm < 10 000 ppm IEX IEX 1-200 ppm IEX IEX 5-8 ppm End Product Testing < 10 ppm < 1 0 ppm DS Bulk 17
Process vs Clinical Experience • Experience vary between companies Example 1 One company has a more consistent process than another company Example 2 Clinical Experience Clinical Experience Clinical experience is broader than Process Process Consistency batches or Process Process Consistency batches are broader than clinical experience Design Space Design Space Design space different What are the right specifications? 18
Handling of QA:s - monitoring 1. Consistency for Non-Critical QAs • Process performance – internal monitoring 2. Clinical limits for Critical QAs • Within regulatory limits Internal Acceptance limits limits for trending Batch # Registration 19
Summary • A strategy was presented for selection critical quality attributes (CQA) and for selection of type of test to use for a specific attribute (IPC or end product testing) • Acceptance criteria for regulatory specifications should be based on clinical experience and process capability • The process for performance should be monitored using internal limits without regulatory commitment 20
Acknowledgements • Mark Schenerman, MedImmune • Alistair Kippen, MedImmune • Ronald Imhoff, Janssen 21
Recommend
More recommend
