1. What You Should Know When You Make Manufacturing Changes to Biotechnology Products May 16-18, 2011 | Beijing, China Mark Rosolowsky, Ph.D. Vice President, Global Regulatory Sciences-CMC Bristol-Myers Squibb

2. Disclaimers –The information within this presentation represents the views of the presenter and is based on the presenter’s expertise and experience 2

3. 3 Overview –Comparability - key points for consideration –Discuss changes to biotechnology products: During development Post-marketing

4. 4 Why are Changes to Biotechnology Products so Complex? Small Molecules –Usually synthetic, organic compounds having well defined structures and chemical characteristics –Typically produced through chemical synthesis –Usually micromolecules having molecular weights of less than 500 Daltons –Generally very stable, and not extremely sensitive to heat Biotechnology Products –Usually a protein- or carbohydrate-based product with complex structure –Either composed of / or extracted from a living organism or produced via cell culture –Macromolecular by nature, and usually have a molecular weight greater than 500 Daltons –Tend to be rather labile, and are usually very heat- and sheer- sensitive –Tend to be immunogenic

5. What is Comparability? ICHQ5E Definitions Comparable: –A conclusion that products have highly similar quality attributes before and after manufacturing process changes and that no adverse impact on the safety or efficacy, including immunogenicity, of the drug product occurred. This conclusion can be based on an analysis of product quality attributes. In some cases, nonclinical or clinical data might contribute to the conclusion. Comparability Exercise: –The activities, including study design, conduct of studies, and evaluation of data, that are designed to investigate whether the products are comparable. 5

6. Comparability Exercise Considerations –No one right answer when deciding a comparability package –Decision must be made with consideration of: Product clinical development plan Complexity of the product Stage of development of the product Robustness of analytical methods Existence of relevant animal models Previous health authority interactions 6

7. What is “Highly Similar”? –Comparability does not require quality attributes of pre-change and post-change product to be identical –“Highly similar” depends upon whether: Existing knowledge can adequately support that differences in quality attributes have no adverse impact upon safety or efficacy –Side-by-side analysis of “post-change” vs. “pre- change” product is useful for contemporaneous evaluation 7

8. 8 Comparability is a Sequential Process Comparability is a Sequential Process Non-clinical Clinical and/or Pharmacovigilance Quality Confirmatory clinical testing may be necessary Non-clinical studies may be necessary, if non-clinical studies cannot discern relevant differences, then… If the analytical procedures used are not sufficient to discern relevant differences, then… OVERALL GOAL: Assess potential impact to safety and efficacy of the product

9. Parameters to Consider –Production step where changes are introduced. –Potential impact of changes to: purity, physicochemical and biological properties considering complexity and degree of knowledge (e.g., impurities, product related substances). –Availability of suitable analytical techniques to detect potential modifications –Understanding of relationship between quality attributes and safety and efficacy, based on overall nonclinical and clinical experience. –Relevant physicochemical and biological characterization data regarding quality attributes; –Need for stability data, including accelerated or stress conditions, to provide insight into potential product differences in the degradation pathways 9

10. Parameters to Consider (continued) –Batches used for demonstration of manufacturing consistency; –Historical data that provide insight into potential “drift” of quality attributes –Critical control points in the manufacturing process that affect product characteristics, –Impact of the process change on the quality of in-process materials & ability of downstream steps to accommodate material from a changed cell culture process; –Adequacy of the in-process controls (critical control points & in- process testing: In-process controls for post-change process should be confirmed, modified, or created, as appropriate, to maintain product quality –Nonclinical or clinical characteristics of the drug product and its therapeutic indications 10

11. Risk Assessment –Utilize prior knowledge and development studies –Categorize risk change of process parameters based upon potential to impact product quality –Examples: High: Change to Master Cell Bank (MCB) Moderate: Media composition change using established raw materials, <50% output Low: Step optimization (e.g. flow rate, wash volumes, elution collection criteria) –Provides an effective tool for internal discussion of change & subsequent communications to regulators regarding the assessment 11

12. Risk Assessment Model

13. Example Risk Assessment Tool 13  Each risk factor category has built-in drop-down boxes with potential values  Higher values indicate greater risk  Team discussion is critical to document thought process that drove the scores (“no one right answer”) Total risk score generated by multiplying individual risk factors

14. Evaluating Changes During Development 14

15. Phase-based Approach to Changes Early Development: –Before nonclinical studies: Comparability is not generally not a concern Subsequent nonclinical and clinical studies using the post-change product as part of the development process support change –Early phases of nonclinical and clinical studies: Comparability testing is generally not as extensive as for an approved product –As knowledge and information accumulate, the comparability exercise will generally become more comprehensive

16. Phase-based Approach to Changes During Pivotal Clinical Studies: –Changes are discouraged –Sponsor should seek scientific advice from the relevant health authorities If process changes are introduced in late (post- pivotal studies) stages of development: –Thorough comparability exercise is generally required: Physicochemical and biological in vitro studies Clinical pharmacokinetic and / or pharmacodynamic comparability studies may also be required –If comparability exercise cannot rule out impact to the efficacy and safety profile: Additional clinical studies may be required 16

17. Guidelines for Acceptance Criteria Setting for Analytical Comparability a Orthogonal support for release test or independent attribute assessment for structure-function relationship b If a stability indicating attribute is identified under recommended storage conditions c If a forced degradation or stressed condition degradation product attribute is identified and believed relevant to structure-function

18. Case Study - Development 18

19. Case Study Process Changes Overview of Changes –Drug Substance New MCB (higher producing subclone of current MCB) New DS manufacturing site (Site “A”  Site “B”) New DS manufacturing process (cell culture and purification) –Drug Product New DP manufacturing site (Site “X”  Site “Y”) Minor change to sterile filtration (0.1 µm  0.2 µm) Reason for Changes –Increase drug substance yield ~4X –Manufacturing control –COG

20. Comparability Exercise Goal - to ensure the quality, safety and efficacy of drug product produced by a changed manufacturing process How - through collection and evaluation of the relevant data based on process and product knowledge –Analytical assays –Biological assays –Nonclinical data –Clinical data

21. Analytical Comparability Established at multiple points –In-process –Release Release tests Extended characterization tests –Stability profile Recommended storage condition Accelerated/Stress storage conditions –Downstream Drug substance changes may only be seen in the drug product (release, stability profile )

22. Analytical Acceptance Criteria In-process –Comparable process/product related impurity/adventitious agent clearance Release tests –Current (pre-change) specification Additional tests may be needed Extended characterization –May need to evaluate additional pre-change batches to establish appropriate acceptance criteria Evaluation against historical data (i.e. clinical experience) Side-by-side comparison of pre- and post-change product by various analytical characterization techniques

23. Binding Kinetics by Surface Plasmon Resonance Association Dissociation

24. Electron Spray Ionization Mass Spectrometry Electron Spray Ionization Mass Spectrometry Reference material (pre-change) Post-change product Slight increase in one subtype observed

25. Tryptic Peptide Maps Overlay of pre- and post-change materials

26. Comparison of Carbohydrate Profiles by HPAEC-PAD 26 Pre-change material Post-change material

27. Isoelectrically Focused Isoelectrically Focused Lane 1: Reference material (pre-change) Lanes 2 through 5: Post-change material

28. Cation Exchange Profiles Gray line: Pre-change material Black line: Post-change material

29. Evaluating Changes for Marketed Products 29

30. Changes to Marketed Products Changes to Marketed Products Changes during life-cycle are inevitable: –Yield increases necessary to meet market demand / address cost of goods issues –Quality improvements are necessary to adhere to current GMPs –Unexpected events require corrective action, such as process parameter changes –Vendor / supply issues necessitate use of alternate materials 30

31. Hypothetical Process Evolution Hypothetical Process Evolution Process Step Phase IIPhase III & Commercial Process CProcess DProcess E Cell Bank XYZ-01XYZ-01.1Same Media DECD-CHO, eRDF & Yeastolate Same, Plus Additional Minerals Same Production Conditions No temp. shift Two phase temp. shift Same Single temp. shift Downstream Sequence Six columns Five columns SameChange of one resin Other Process Changes N/A Centrifugation & filtration parameters Filter changesN/A

32. Case Study – Marketed Product 32

33. Case Study: Change in Media –Quality: Media component change resulted in minor differences to quality attributes –Nonclinical: Study conducted using non- human primate model with previously well- established concordance to human PK for the product Model had demonstrated sensitivity to changes in: –Minor glycosylation alterations –Moderate sialic acid profile shifts 33

34. Non-Human Primate Results Red line = pre-change Blue line = post-change Time in Hours Serum Concentration [ug/mL]

35. FDA Feedback –Quality attributes evaluated against historic data alone deemed insufficient “…the Agency strongly encourages the use of side-by-side analysis as the most rigorous assessment of comparability…” –Side-by-side comparison required: “Since the product approved for marketing authorization was made using the…media..., this product should be directly compared to the product produced from the proposed process…”

36. Regulatory Experiences with Comparability Regulatory Experiences with Comparability –Comparability exercises can be successfully used to support changes –Most changes can be supported on Quality attributes alone Specifications alone are generally NOT sufficient to support comparability –Additional characterization required –Side-by-side analysis preferred by FDA –Additional non-clinical & clinical data may be required –Safety & efficacy data are generally not required 36

37. In Summary –Comparability is a key issue for biotech products –There is no single, correct strategy to demonstrate comparability –Decision must be made with consideration of multiple factors: Complexity of the product Stage of development of the product Process knowledge & robustness of analytical methods Existence of relevant animal models Previous health authority interactions –Comparability should be approached stepwise: Quality Non-clinical Clinical 37

38. Acknowledgements Cheryl Watson Reb Russell Dave Peck Charlene Craig Gary Lazarus 38