1. 1 The Value of Applying Quality by Design - Not Just Monoclonals, But Across Products and Systems Tony Mire-Sluis, Executive Director, Global Product Quality and Quality Sciences

2. There Are Four Basic Principals for QbD Design Develop Monitor Improve 2

3. There Are Four Basic Principals for Developing Quality Systems 3 Plan Execute Monitor Improve

4. The Principles of QbD Are Not Unique to a Product or Process! 4

5. We Don’t Need to ‘Invent’ Something New… Dr. Deming Figured It Out In ~1950

6. The Principles of QbD are not ‘a Regulatory Strategy’ but Provide Business and Quality Benefits QbD principles can provide: Improved success rates for process and product development through increased understanding Reducing risks To patients, process failures, method invalids, NCs etc. Improving product quality More robust and consistent processes Less manufacturing failures Reduced complaints Continuous improvement 6

7. 7 How the Basic Principles Can Be Applied Regardless of Product or Process Establish Design Targets Establish Design Targets Develop Process to Meet Targets Understand Process Effective and Efficient Control Strategies Post approval Lifecycle Management Molecular Assessments And Engineering Design spaceRisk based control strategy Fit for purpose processes Efficient Continuous Improvement Integrated Controls Operational Testing Raw materials Continuous verification QbD filing/ECP QMS DesignDevelop and ExecuteMonitorImprove

8. 8 Molecular Design Can Be Applied Across Any Product Class – Our Understanding of CQAs goes Beyond MAbs Sequence & hot-spot analysis Engineering to impact CQA’s Fit to FIH platform High productivity Optimal signal peptide usage Downstream process performance Compatibility with FIH platform formulation pH-jump study Serum based pH jump Particulation propensity Develop predictive particulation assays Degradation specific screens Develop degradation-rate database and Arrhenius models to enable decisions based on predicted rate at 2-8  C Concentration & viscosity screen Engineer in low-viscosity attributes by incorporating principles of QbD

9. 9 Examples of Assays Used During Molecular Design to Select Best Lead Candidates Sequence binning used to identify antibodies with diversity and desirable attributes t=24h pH jump study performed to eliminate candidates with potential issues in vivo Viscosity testing for ease of SC delivery at high concentration Expression Assessment to achieve high productivity

10. Design Principles and their Benefits do Not Just Apply to Product or Process Product Manufacturing Process Equipment Facility Utilities Raw Materials Containers Transport 10

11. Risk Assessment is a Critical Element of the Development of Many Systems Risk Assessments can be applied across many different programs: Critical Quality Attributes Equipment selection and maintenance Non conformances Facility design Packaging Supply Finance Overall Risk Assessment RPN Impact Severity Capability Occurrence Testing Strategy Detection =X X 11

12. Appropriate Monitoring is Also Neither Product or Process Dependant  Monitoring and analysis of data should be applied regardless of the product, process or system being used  Incoming Inspections  During the process  Lot Release  Stability  Non conformances  Complaints  APR  Frequent cross functional assessments Raw Data (Plant Level) Deviation Diagnosis (Variable Level) Abnorma l Batch 12

13. 13 Applying QbD Principles to Raw Materials – A Case Study of Vials

14. Raw Materials are an Essential part of both Product and Process Raw materials and their controls are vital at all stages of manufacture In fact, they can make up the product itself beyond just the protein – water, buffering components, tonicity agents, polysorbate, primary packaging etc. and QbD principles apply to RMs at all stages of manufacture Application of the principles is carried out on a risk based approach depending where they are in the manufacturing process (downstream or upstream) or how much they impact the quality of the product (e.g. media components are downstream away from the final product but can have a tremendous impact on protein CQAs during fermentation) 14

15. 15 It is Essential to Understand a Raw Material - Design Initial Characterization Raw material itself Design appropriate Analytical Methods Lot to lot variability Set Specifications Interaction with process Impact to Critical Process Parameters Affect on Product Impact on Critical Quality Attributes How much is required depends on where in the process the RM is used Managing Change over time Supplier need to communicate when changes are proposed / made The company needs to evaluate if changes matter to their processes… ideally before they are implemented

16. 16 What do you Need to Characterize in a Raw Material? – What is There Relying on the manufacturer’s CoA is often not be enough One needs to understand all the components of the raw material as the manufacturer may not even measure the parameters you require control of: Product Heterogeneity Additives Preservatives Degradation products Contaminants

17. 17 What do you Need to Characterize in a Raw Material? – What Might Appear There must be a thorough understanding of the degradation pathway of the raw material – relying on an expiration date from the manufacturer may not be enough The impact of degradants on the product or process is needed There needs to be an understanding of the use of raw materials over time and the impact to the expiration date of the product The impact of handling the raw material over time must be assessed (aliquoting, light, temperature, oxygen etc.)

18. 18 Supplier Management is Necessary for Ensuring Raw Material Quality Expectations must be clear We need to understand the manufacturing processes for raw materials so we know how they could impact product quality, allow us to set appropriate specifications and help us during investigations – Design and Monitoring Notification of change - Monitoring Thorough investigation of defects – Continuous Improvement Supplier site visits are key Audit of Quality system - Plan Technical visits to understand processes, process capability and identify indirect product contact materials - Design Technical visits to evaluate changes and ensure that they are managed properly - Monitor

19. LowMediumHigh Preferred Audits every 4 year Audits every 3 year Standard Audits every 4 yearAudits every 3 yearAudits every 2 year Marginal Audits every 3 yearAudits every 2 yearAnnual Audits Material and Supplier Criticality Assessment Supplier Performance Rating One can Use a Risk-Based Approach to Segment the Supplier Base Quality Assessment = 40% Technical Assessment = 40% Business Assessment = 20% Quality 50% Delivery 30% Service 20% CAPAs manage ‘up or out’ plans for marginal suppliers

20. 20 Vial FMEAs were conducted by a Global team - Design Vials are an essential part of a product in which they are filled, critical to product quality and close to the patient FMEAs were conducted at fill finish lines across internal and contractor manufacturing sites Risks were identified, mitigation plans developed, and CAPAs documented completion Identifying what were the important attributes for vials was essential to develop appropriate specifications: Extractables/Leachables Interactions with equipment Handling and Storage Possible glass defects

21. 21 An Example of Specifications Required for Vials – Design/Monitor 1.Finish Inner Diameter (ID) 2.Finish Outer Diameter (OD) 3.Lip/Flange Height 4.“H” Dimension 5.Vial Length (Height) 6.Vial Outer Diameter 1 2 3 5 6 4

22. 22 Incoming Inspections of Raw Materials to Assure Specifications are Met IQA review received docs (CoA, Receiving Inspection Report) Sample as per SOP IQA inspect samples as per SOP IQA perform batch evaluation and disposition of material IQA Receipt of Material Inspection Evaluate & Disposition Sampling

23. 23 Inspection for Tubular Glass Vials Includes over 50 Criteria Defect Classification: Critical Visual Examples: Wrong component Split on vial interior with nonremovable, marked deposit Internal Airline – Elongated gaseous inclusion that appears as a vertical line. Chipped (Broken Finish) – (if the seal is compromised) A finish that has actual piece of glass broken out of it. Contamination – Foreign substance (particle, stain, dirt) deposited on the internal surface Crack – Fracture that penetrates completely through the glass wall Malformed – Finish is grossly distorted or deformed, if seal is compromised (lip malformed).

24. 24 Automated Inspection of Vials Prior to Lot Release - Monitor Neck Inspection Camera C1/C6 Body Inspection Camera C2/C7 Base Inspection Camera C3/C8 Cap Inspection Camera C4/C9

25. 25 A Global Primary Container Team was established to Monitor Vial Performance The Primary Container Team: Comprises of all Amgen site, quality and technical groups Tracks key glass metrics from the glass process monitoring, NCs, and Product Complaints Identifies and implements Glass Handling best practices Alignment to new equipment requirements Identifies Quality/Process improvements and container standardization

26. 26 Embedding the Quality System - Improve Primary packaging issues are monitored globally and lessons learned applied A governance body is covering vial platform projects and tasks A global team is ensures Amgen wide scope and communication of tasks and work streams The team monitors Amgen production and CMC activities The creation of a global team allows for continuous monitoring and improvement Primary Packaging Network Product Complaints Glass Process Monitoring Δ Control NC Glass Handling Project Risk Assessment

27. 27 In Summary The basic principals of QbD can (and should) be applied across product types as well as multiple processes and systems We don’t have to label it QbD and treat it like something special, it just makes good business sense and we have been doing it to a certain extent anyway and it can be applied to all products and across multiple systems/processes Using the principals of QbD in many cases does not require a great deal of cost – mostly good planning, increased understanding and more successful execution A risk based approach can be used to help appropriately focus QbD efforts