Practical challenges in the CMC development of biosimilars...Nature Reviews Drug Discovery. February 2011. CASSS Bay Area Discussion Group ... Analytical similarity methodology: selection

Practical challenges in the CMC

development of biosimilars

Simon Hotchin

Executive Director Regulatory Affairs

Amgen Inc.

CASSS Bay Area Discussion Group

12 Nov 2015 2

Overview

• Introduction

• Statistical methodologies in the assessment of analytical

similarity

• Selection of analytical methods for similarity assessment

• Reference product bridging assessments

• Discussion points

Introduction


12 Nov 2015 4

Analytical similarity remains the

cornerstone of biosimilar development

•Analytical data form the

cornerstone of the similarity

assessment

•Products must be ‘highly similar’

to be eligible for approval as

biosimilars

•Increasing the extent and

discriminatory power of the

analytical assessment may permit

a more targeted and selective

approach to nonclinical and

clinical studies.

Clinical

Clinical Pharmacology

Nonclinical

Analytical


12 Nov 2015 5

Approval requires a careful assessment

of the totality of evidence

• Biosimilar or Biosimilarity means:

• that the biological product is highly similar to the

reference product notwithstanding minor

differences in clinically inactive components; and

• there are no clinically meaningful differences

between the biological product and the reference

product in terms of the safety, purity, and potency of

the product.

Analytical and Functional

Data

Clinical Data


12 Nov 2015 6

The definition of biosimilarity poses

some practical questions

• How similar is similar and how can a level playing field

be assured?

• How does analytical method capability impact the

assessment of similarity, and what methods should be

used?

• How to conduct efficient global biosimilar development

when the approval pathway has an inherently national

aspect?

Analytical similarity standards:

statistical methodologies


12 Nov 2015 8

FDA has recommended a tiered attribute

assessment for demonstrating similarity

• Tier 1 is based on a test of

equivalence

– Analytically similar if 90%

confidence interval of the true

mean diff. is within equivalence

margins (δ1, δ2)

– The equivalence margin is

1.5sR where sR is the true

standard deviation of the RP

• Tier 2 is based on majority of

results falling within a quality

range of +/- X SD

(δ1δ2)

90% CI

Fail

Fail

Yi Tsong, FDA, DIA/FDA statistics Forum 2015


12 Nov 2015 9

There are potentially some practical

challenges with the proposed approach

• Existing knowledge of attribute criticality may not be

sufficient to clearly delineate the appropriate testing tier

• Reliable estimates of the true mean and variance of the

reference product can be difficult to establish during

development, and both can change over the

development time frame

• A statistical approach based on data from to-be-

commercial lots may not be the most effective means to

assure a robust process to deliver similar product


12 Nov 2015 10

The appropriate tier assignment for an

attribute may be less than obvious

• Product labels and literature may

include information suggesting a

role for e.g. effector function in the

MoA of a product

• However, understanding regarding

the quantitative relationship

between attributes and

safety/efficacy in patients is

typically incomplete

• Scientific judgment on tiering

assignment must be made based

on the available evidence and

application of risk management

principles

• Differing conclusions may be

reached

• FDA vs Sponsor

• Sponsor vs Sponsor

Reproduced from: Jiang et al. Advances in the assessment and control of

the effector functions of therapeutic antibodies. Nature Reviews Drug

Discovery. February 2011.


12 Nov 2015 11

Cell Line and Process

Development

Nonclinand

Clinical Dev.

Submission and

Approval


DevelopmentNonclinical Clinical

Submission and Approval

Post Approval

Estimating the true mean of the

reference product can be challenging

Reference product

lifecycle

Biosimilar product

lifecycle

Simulated data for illustrative purposes


12 Nov 2015 12

Identifying Tier 1 attributes early is

critical to demonstration of similarity

▲Probability to pass Tier 1▲

Adequate power

Means match

Some results slightly outside

RP range

▼Probability to pass Tier 1▼

Inadequate power

Means do not match

All results inside RP range

• Early cell line and process development

focused on matching mean for attribute

• Similarity testing protocol captured sufficient

results for attribute

• Early cell line and process development

targeted matching RP range

• RP variability greater than anticipated

• Similarity testing protocol assumed reduced

testing sufficient



12 Nov 2015 13


Development

Nonclinand

Clinical Dev.

Submission and

Approval

Pre-Approval Post-Approval

Actual shifts in the reference product

may also be apparent

Reference product

lifecycle

Biosimilar product

lifecycle



12 Nov 2015 14

Data from at-scale lots may not be the

sole means to establish similarity

• The Tier 1 statistical approach rewards increasing sample size

– Minimum of 6 lots recommended for equivalence testing

– 10 or more required to achieve power levels that sponsors may deem

appropriate when multiple attributes are included in Tier 1

– May drive sponsors to manufacture additional lots for analytical similarity

purposes

• Is this form of ‘extended process validation’ the most appropriate

means to ensure similarity of future lots?

– Could data from a validated small scale model and/or generated by

mathematical process modelling be included in the statistical analysis?

– Should data from at-scale production be considered as confirmatory

evidence?

Analytical similarity methodology:

selection of analytical procedures


12 Nov 2015 16

Stability

Primary structure

Biological function

Particles and

aggregates

Process-related

impurities

General properties

and excipients

Higher order

structureProduct-related

substances and

impurities

Attributes related to the amino acid

sequence and all post-translational

modifications, including glycansBiological and functional

activities, including receptor binding

and immunochemical properties

Quantitative levels of

product variants and

their identities

Degradation profiles

denoting stability

Subvisible and submicron particles

and aggregates of various sizes

Integrity of the secondary, tertiary,

and quaternary structure

Properties of the finished

drug product, including

strength and formulation Impurities from host cells

and downstream process

Comprehensive analytical similarity assessment reduces the degree of uncertainty


12 Nov 2015 17

The choice and performance of analytical

tools matters

Slide from Kozlowski, S (CDER) presentation at 2014 Biomanufacturing Technology Summit, Rockville, MD, June 13, 2014

Reliability - Methods are qualified and fit-for-use for intended purposes

Relevance - Methods provide meaningful information and may predict clinical performance

Resolution - Methods are sensitive and capable of resolving differences which are critical attributes


12 Nov 2015 18

Different methods provide differing levels of

information relative to criticality and similarity

Critical

glycan

attribute

Glycan

Map

profile

Monosaccharide

analysis

Rel

evan

ce

Reliability

0.0

10.0

20.0

30.0

4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 22.0 24.0 26.0

min

A2

G0

F

A1

G0

F

M5

A2G

0

A1

G1

F

A2

G1

F(a

)A

2G

1F

(b)

M6

A2

G2

F

M7

Monosaccharide analysis

Glycan mapConfidence in similarity


12 Nov 2015 19

Rel

evan

ce

Reliability

NK92

ADCC

FcgRIIIa

binding

PBMC

ADCC

Glycan

map

Confidence in manufacturing

Knowing the correlation of orthogonal methods

helps develop a robust control strategy

High resolution and reliable methods suited for process control can guide the process and product development.

Orthogonal assays with increasing biological relevance to discern clinically meaningful differences and confirm functional similarity

Response (LU))

0.0

10.0

20.0

30.0

4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 22.0 24.0 26.0

min

A2G

0F

A1G

0F

M5

A2G

0

A1G

1F

A2G

1F

(a)

A2G

1F

(b)

M6 A2G

2F

M7

Chung, mAbs, 326-340, May 2012

ADCC vs. Afucosylated glycan

Analytical similarity requirements:

reference product considerations


12 Nov 2015 21

Background

• Biosimilar regulations continue to roll-out & evolve around the world

• The majority of the regulations necessarily require demonstration of

similarity to a local reference product approved on that market

• Use of analytical, nonclinical and clinical data generated with a

foreign reference product can be negotiated, by providing evidence

to demonstrate the equivalence of the local and foreign reference

products.

• The level of evidence required for bridging varies.


12 Nov 2015 22

Why bridging?

• Is a foreign comparator truly representative of the local reference product

with respect to quality, safety and efficacy?

• Some regulators may require analytical (± PK) bridging to a locally procured

reference product

• Possible factors causing geographic divergence in originator product quality

attributes

– Supply chain partitioning between geographies (eg, local manufacturing)

– Licensing arrangements without shared manufacturing

– Asynchronous implementation of major process variations

Hypothetical concerns? Consider epoetin alfa, supplied by 3 distinct companies using independent DS and DP sources. Only one originator

product was associated with a specific quality related safety issue (eg, PRCA cluster in Europe, 1998-2002).

Casadevall et al. NEJM 346:469-475 (2002).


12 Nov 2015 23

Problem Statement: How to go from

this…

EU US

Pivotal Trials

Analytical Sim

Nonclinical

Analytical Sim

Bridging

Analytical

Bridging PK

CORE

X Originator product lots sourced from country X

Map of comparisons made between product lots(originator to biosimilar or originator sourced in EU to originator sourced in US)


12 Nov 2015 24

To this…

EU US

JP

Pivotal Trials

Analytical Sim

Nonclinical

Analytical Sim

Bridging

Analytical

Bridging PK

AUS

CN

RSA

TURBRMEX

RUS

KR

CAN

??

?

?

? ? ?

CORE

??

?

X Originator product lots sourced from country X


12 Nov 2015 25

Various bridging options are possible.

Some may represent a higher challenge.

• Minimal similarity risks and costsNo additional justification

• Minimal similarity risks and costsPaper based justification

• Similarity risks and costsLimited analytical bridge

• Similarity risks and costsComprehensive analytical

bridge

• Significant similarity risks and costs

Analytical plus BE

• High similarity risks and costsStand-alone development

Lacking other evidence of a

harmonized originator product

supply would these approaches

protect patients from geographic

“divergence”?


12 Nov 2015 26

Conclusions

• High standards will help to build confidence and drive biosimilar adoption.

Statistics can play a role in assessment of similarity, but approaches based

on comparisons of means may be practically challenging to implement.

• Ensuring early transparency to the identity of the highest risk attributes for a

given molecule would increase predictability in biosimilar development.

• The application of methods that have been demonstrated to be capable of

discriminating differences in critical attributes will increase confidence in

similarity.

• Flexibility in the application of regional bridging requirements will ensure

global access to high quality biosimilar medicines


12 Nov 2015 27

Acknowledgements

• Jennifer Liu

• Margaret Karow

• Rick Burdick

• Gino Grampp

• Richard Markus

Documents

Practical challenges in the CMC development of biosimilars...Nature Reviews Drug Discovery. February 2011. CASSS Bay Area Discussion Group ... Analytical similarity methodology: selection