Establishing Comparability with Process and Manufacturing Changes for Recombinant Vaccine: A case study of HPV Vaccine

Dicky Abraham

Global Vaccine and Biologics Commercialization Merck & Co., Inc WCBP, Jan 2015

1

Outline • • • • • •

HPV and HPV Vaccine Importance of characterization Characterization of Gardasil® Comparability matrix Comparability Assessment for facility scale-up Conclusions 2

Rationale For An HPV Vaccine • Human Papillomavirus (HPV) infection causes cervical and anal cancers, and genital warts. • Pap smear has reduced cervical cancer rates, but it has significant drawbacks: – repeat testing required – limited sensitivity & specificity – psychosocial impact of a “pre-cancer” diagnosis – high costs – not feasible in the developing world

• A vaccine to prevent infection with common pathogenic HPV types is a major advance in anogenital cancer control. 3

Immunologic Basis For Candidate HPV Vaccines • Vaccine candidates have been evaluated in animal models of papillomavirus infection. • L1 VLP Vaccines – L1 HPV major capsid protein self-assembles into virus-like particles (VLPs) – In preclinical studies using species-specific VLPs: • vaccination protection against infection / disease • efficacy associated with development of neutralizing antibodies • transfer of serum from vaccinated to unvaccinated animals transfers protection

4

HPV Coat (Capsid) Proteins Expressed in Yeast HPV Virion

infectious HPV 6, 11 - genital warts HPV 16, 18 - cervical cancer

HPV Virus Like Particle

non-infectious vaccine

Clone L1 Gene for the Capsid Protein for Each HPV Type Intracellular expression in S. cerevisiae

Key expression features: - tightly-regulated promoter - common host strain - common vector

5

GARDASIL Merck’s Quadrivalent HPV Vaccine • Quadrivalent HPV (Types 6, 11, 16, 18)

L1 virus-like particle

(VLP) vaccine-approved in 2006 • VLPs manufactured in yeast • Highly purified VLPs • Well Characterized • Approved world-wide

6

Structural Model of HPV VLP Virus-Like Particle (~20,000 kDa)

L1 Capsomere L1 protein

(~280 kDa)

(Atomic force microscopy image of a single VLP)

(55 or 57 kDa) 72 Capsomere

5 x L1 (Crystal structure coordinates courtesy of Prof. S. C. Harrison, Harvard University)

~ 3 nm

~ 10 nm

~ 60 nm 7

Why Characterize a Vaccine? • Historically for complex vaccines and biologics “The Process Defines the Product” • Modern vaccines are amenable to modern biochemical characterization techniques • Requirement to demonstrate that we understand our product (CTD Sections S3.1, S3.2, S4 & P5). • Release, Process Characterization & Comparability • Rational bridge to clinical data • Functional potency test 8

HPV VLP is Very Complex HPV Virus Like Particle IgG Immunoglobin (Therapeutic mAb)

~150X

~150X

Zocor®

Aspirin

9

Method – Structure Link Amino Acid Sequence

α-Helix

Assembly

Adsorption

Disassembly

β-Sheet

Aluminum Adjuvant

1° Structure Mass Spec, Purity, Integrity, Deamidation, & Free Thiols in Denatured Form

2° Structure CD & FT-IR

3° and 4° Structure DSC, DLS, IC50, Relative Antigenicity, Free Thiols in Native Form & L1Oligomer

MBAP DLS, DSC, IVRP & IC50

10

Primary Structure • • • •

Mass spectrometry SDS –PAGE Deamidation Free Thiols

11

Secondary Structure • Circular Dichroism • Fourier Transform IR Spectroscopy

12

Tertiary and Quaternary Structure • • • • • •

Differential Scanning Calorimetry Dynamic Light Scattering Relative antigenicity Atomic Force Microscopy Cryo EM Potency (in Vitro Relative Potency)

13

Summary Product Characterization • Meets Regulatory Expectations • Supports modern in vitro potency tests • Focused testing (rather than testing to infinity) is desired. • Provides data base to observe changes in product over time • Supports development • Supports post licensure process changes • Analytical Matrix Approach to comparability based on identifiable & relevant changes. 14

ICH Quidelines (Q5E) • Comparabillity goal is to ensure the quality, safety and efficacy of drug product produced by a changed manufacturing process: – Achieved through collection and evaluation of the relevant data – Affirm there is NO adverse impact on the drug product due to the manufacturing process changes.

• Comparability does not necessarily mean that the quality attributes of the pre-change and post-change product are identical, – But highly similar –

Existing knowledge is sufficiently predictive to ensure that any differences in quality attributes have no adverse impact upon safety or efficacy of the drug product.

• Comparability can be based on a combination of analytical testing, biological assays, and, in some cases, nonclinical and clinical data. – Aassurance of comparability through analytical studies alone, • Non-clinical or clinical studies are not warranted-where the relationship between specific quality attributes and safety and efficacy has been established.

15

Comparability for a Well Characterized Vaccine • Distinguishes difference in final vaccine structure • Sensitive to process • Distinguishes intermediate from final product • Essential data base for future process changes • Comparability Plans and Protocols 16

Comparability Plan • Used for “specified” or “well-characterized” biologic products. – “Historical” Comparability – Therapeutic / Clinical Equivalence

• Fundamental limitation: can’t always state if a change matters. – Focused testing (rather than testing to infinity) is desired. – (Analytical) Matrix Approach to comparability based on identifiable & relevant changes (more case-by-case).

• “Characterization” vs “Comparability” Testing

17

Analytical Matrix for Comparability Method Release - Potency Release – Purity & Integrity IC50 Rel. Antigen. DSC DLS Free Thiols Oligomer Deamidation • Monitor clinically-relevant structural features that are desirable

properties of the product. • Emphasis on release assays & epitope structure (conformation) during manufacture & storage.

18

HPV VLP Manufacturing Process This image cannot currently be displayed.

Produce in recombinant yeast

Release intracellular HPV VLPs

Digest DNA/RNA

Clarify yeast lysate

Remove bulk yeast impurities Reduce nucleic acid and aggregates Remove residual impurities and buffer exchange.

Jansen et al. (Merck & Co.) US Patent 5,888,516, 1999; Cook et al., Protein Exp. Purif., 17, 477, 1999

19

Manufacturing Change: Facility Scale-Up • A small scale purification facility was used to produce clinical, process validation and launch drug substance lots • Scale up facility constructed to meet projected market demand • Process for new facility was scaled up – Required targeted changes to manage larger production scale

20

Process Comparability • Limited process changes with new facility • In-process inputs & outputs to demonstrate new facility process is comparable to launch facility  Critical Process Parameters (CPPs) and Critical Quality Attributes (CQAs)  Key Operating Parameters (KOPs) and Key Process Attributes (KPAs)

• Process characteristics to demonstrate comparability to launch facility process characteristics  Qualitative comparisons of chromatography profiles, pH profiles, and impurity clearance profiles 21

Process Peformance Process Step

Attribute

Launch Facility Range

Lot 1

Lot 2

Lot 3

VLP Release

Parameter 1

95 – 98

100

95

95

VLP Capture and Purification

Parameter 2

69 - 74

87

80

78

Parameter 3

1.5 – 2.9

2.9

2.9

2.9

Parameter 4

96 - 98

98

96

97

Parameter 5

0.5 – 1.7

1.4

1.3

1.3

VLP Disassembly

Parameter 6

98 – 99

99

99

99

VLP Reassembly Cross linking

Parameter 7

99 – 100%

100

100

100

Parameter 8

11 – 13

11

13

11

Parameter 9

81 – 98

98

86

93

Parameter 10

717 - 928

856

777

875

Parameter 11

106 - 114

107

106

103

Sterile Filtration

Parameter 12

> 100

97

102

108

Alum Adsorption

Parameter 13

300 - 322

338

317

320

Polishing Purification

Buffer Exchange

• •

For each parameter performance was evaluated against acceptance limit and historical performance (launch facility ranges) Excursions were investigated

22

Example: Impurity clearance

•

Comparable clearance profile for all impurities and process residuals 23

Summary: Process Perfomance • In-process outputs, Critical Quality Attributes, Key Process Attributes demonstrated comparability to launch facility • Qualitative comparison demonstrated process is comparable to launch facility  Chromatography profiles  pH profiles  Impurity and process residual clearance profiles 24

Product Comparability Overview Analytical Measures Mfg Database

Expected Results

• Product Characterization • Impurities • Stability • Represent manufacturing diversity • Analytical variability

• Acceptance Criteria (Primary Limit) • Alert Limit (Secondary Limit)

Process

Comparable Meets acceptance and alert limits

Change

Fails acceptance criteria

Not

Meets acceptance Limits but outside Alert limits

Investigate Investigation includes: Weighting of parameter Evaluate data from all lots Clinical Experience Results from other parameters

Comparable

25

Developing Limits from our database

Limits prospectively established to define potential deviations from comparability  Provides rigor and credibility

Two types of Limits

 Acceptance Limits = deviations generally means failure to demonstrate comparability

 Alert levels = deviations results in investigation, but not deemed a failure a priori  Limits/Levels must be sufficiently rigorous but not so tight as to cause “nuisance alarms” 26

Analytical Comparability: Drug Substance

27

Investigation of Excursions •

T6 Potency (IVRP) Investigation: Alert Level excursion Lot #3 •

Assay: No atypicals found in assay run

•

IC50 & Relative Antigenicity assays (orthogonal immunoassays) had expected results

•

Stability data showed expected IVRP

 Conclusion: assay variability, lot is comparable •

T16 Free Thiols: Alert Level excursion Lot #3 •

Assay measures non-crosslinked SH molecules, result can suggest particle may be malformed and/or unstable

•

Stability evaluations showed no change in stability profile

•

Particle size and level of monomer consistent with other lots

 Conclusion: excursion does not have an impact on product quality or stability, lot is comparable •

Differential Scanning Calorimetry (DSC): all lots •

Limited number of runs used to set alert limit

•

Run to run variability not represented in alert limits

•

Samples from Phase III lots run concurrently with new facility lots showed similar alert limit excursion

 Conclusion: lots are comparable

28

Conclusion •

A modestly scaled purification facility was used to supply Phase III and launch material for GARDASIL®

•

A larger facility was required to meet the expected demand for the product

•

A key component of the licensure for the new facility was the demonstration of comparability of drug substance and drug product made in the two facilities

•

The comparability assessment shows the product made in the new facility is comparable to that from the launch facility

•

Merck leveraged comparability to accelerate licensure of larger-scale facility, providing world-wide supply ASAP

29

Acknowledgments A. Acosta R. Capen J. Bendas J. Dashnau P. DePhillips P. Desai M. Kosinski M. MacNeil W. Muthoga D. Opalka

B. Oswald R. Peluso C. Potter (Scripps) R. Sitrin J. Robinson R. Rustandi S. Sagar C. Schultz A. Sturgess

G. Takle V. Towne Y. Wang M. Washabaugh D. Wohlpart Q. Zhao

30