1

REGULATORY PERSPECTIVE ON THE EVALUATION OF CLONALITY OF MAMMALIAN CELL BANKS

Rachel Novak

CDER/OPQ/OBP/DBRRI

January 23, 2017

www.fda.gov

Disclaimer

This presentation should not be used in place of

regulations, published FDA guidances or discussions

with the Agency.

3

Introduction

• This talk will focus on a regulatory perspective on: – Expectations of clonality

– Mutual understanding between regulators and industry

– How reviewers assess adequate assurance of clonality

– Management of cell lines that have low assurance of clonality

– Management of non-clonal cell banks

4

Basis for the Expectation of Clonality

• Scientific considerations: – To minimize the heterogeneity within the master cell bank (MCB) to

allow for a well-controlled processes that is capable of consistent manufacture of a product.

– When cell banks are non-clonal, every potential change made to the upstream process (raw materials, process parameters, manufacturing site, etc.) presents the potential to put selective pressure on the cultures, which may result in changes to the manufacturing process or the final product.

• Regulatory considerations– No single regulatory document explicitly states that cell banks should be

monoclonal; however, it is expected that clonal cell lines are developed as referenced by ICH Q5D and EMA/CHMP.

www.fda.gov

5

INDUSTRY PERSPECTIVEClonality

6

March 2016

7

Mutual Understanding

• To get safe and effective products (not cells) to patients as soon as possible

• We understand the nature of the cells used to make biologics– They are complex (product and the cells)

– Cultures can become heterogeneous

– Genetic and phenotypic drift cannot be directly controlled

– The same genetic plasticity that allows for these cells to be used to make biologics also makes the cell vulnerable to selective pressure and can potentially put the product and process at risk when changes are made.

• Assurance of clonality is part of the overall control strategy– Consistency of the process

– Product quality

8

It’s all about minimizing risk

• Starting with a cell line that has a high probability of being monoclonal reduces (but doesn’t eliminate) residual uncertainty when it comes to making changes to the manufacturing process.

• The small scale studies and process validation studies done to support licensure for cell culture are a snapshot of the process at that time and there is no way to predict how/what changes could shift the clonal balance.

• We understand that its time consuming and costly to front load development; however, starting with a cell bank that has a high assurance of monoclonality could result in less work down the road.

9

It can happen..

• Problems with cell culture, due to non-clonality, can (and have) lead to serious disruptions in manufacturing

– Won’t find things you are not looking for

– Quality System may not be sensitive enough to pick up changes

– This could lead to potential drug shortages.

10

THE REVIEWERS APPROACHClonality

11

Cell Line Development

• Generally, INDs are submitted to the Agency after MCB has already been developed, although it may not be fully characterized.

• Information submitted at the time of IND should include:

– Cell bank testing for adventitious agents per ICH Q5D

– Information on the parental cell line history

– Adaptation to serum-free conditions (if applicable)

– How the cloning process was performed (Limiting Dilution Cloning, FACS, ClonePix, Clone Select, etc.)

– How the final clone was expanded, assessed and selected

• Adaptation to serum-free conditions should be performed prior to cloning step. If adaptation occurs post-cloning, additional cloning steps are needed.

– Serum added as part of expanding single-cell clones will not require additional rounds of cloning.

www.fda.gov

12

Reviewer Considerations for Clonality at the IND stage

• At the IND stage, reviewers will do a initial assessment of the information provided about the clonality of the MCB. If significant deficiencies are noted, then the appropriate comments will be communicated.

• Lack of assurance of clonality is not necessarily a hold issue.

www.fda.gov

13

Probability and Assurance

• The use of “assurance” and “probability” in relation to clonality are described as:

– Probability refers to a numerical calculation provided by the sponsor.

– Assurance refers to an assessment of all the information provided, including probability calculations and supplemental data/information.

www.fda.gov

14

How to assess the adequacy of clonality?

• 3 Aspects to be considered:

– Probability (Numerical)

– Assurance (Probability and Additional Information/Data)

– Final Control Strategy (CS)

www.fda.gov

15

“Acceptable” Probability

• There is not a firm numerical threshold for probability.

• The probability of clonality must be calculated prospectively, no back-calculating.

• Typically, two rounds of LD at the appropriate dilution level is acceptable.

• Other techniques including FACS and ClonePix can provide high probability of clonality when performed using the correct procedures and parameters.

• Imaging techniques could be used to supplement data from LD, FACS or ClonePix

www.fda.gov

16

“Acceptable” Assurance

• If there is “acceptable” probability of clonality then additional assurance is not typically necessary.

• The additional data provided for assurance will not definitively prove clonality but rather provide supporting evidence.

• The type of data used to support clonality is dependent on the way the cells were cloned. – FISH would not be appropriate for cell lines that we cloned using site-

specific integration.

www.fda.gov

17www.fda.gov

Acceptability is dependent on adequacy final control strategy:

– High probability Acceptable (no additional CS modification)

– Low probability + High Assurance Acceptable (no additional CS modification)

– Low probability + Little Assurance + Augmented CS Acceptable

18

Considerations at the BLA stage

• Adequate assurance of clonality should be provided at the time of the BLA submission.

• Having low assurance of clonality of the MCB at the time of licensure does not necessarily preclude approvability of the application.

• Augmentation of the control strategy could be an acceptable approach to managing a non-clonal MCB for licensure.

www.fda.gov

19

• Changes made to the control strategy for managing a potential non-clonal cell line will be product and application dependent .

• Some strategies that have been implemented: – Adding additional specifications (LC-MS/MS for Sequence Variants,

Glycosylation despite not impacting MOA, etc.)

– Tighter limits on the limit of in vitro cell age

– Establishing additional critical process parameters (growth parameters escalated to CPP)

– Trending and Statistical Process Control

– Additional risk assessment for changes in critical raw materials (media, components, etc.)

– Tighter controls for re-qualification of a new WCB

Augmentation of the Control Strategy

20

Non-clonal Lines Post-Licensure

• Sponsors should consult the Agency as soon as possible if evidence of a non-clonal MCB becomes apparent.

• Sponsors should establish a plan to demonstrate the adequacy of the control strategy including a sub-clone analysis assessing the affect of different clonal populations on CQAs and process attributes.– Accumulating this type of data is time consuming.

• In the interim, the sponsor should develop a plan to keep the process well-controlled. This could include:

• the use of statistical process controls to evaluate drift in the product

• establishing additional critical process parameters

• additional characterization for each lot of drug substance manufactured

• limiting future manufacturing until the issue is addressed.

21

Recloning the MCB

• Recloning of the MCB can introduce substantial risk and this risk increases as development proceeds.

• It might not be advisable to reclone the MCB late in development; introduction of a more robust control strategy may be more suitable.– May become difficult to link pre-clinical lots with clinical/commercial lots

– Issues with comparability

22

Summary

• The expectation is that MCBs are monoclonal.• Clonality is part of the overall control strategy for product

development. • Lack of assurance of clonality does not necessarily result in

INDs being placed on hold or preclude approvability of BLA submissions.

• Reviewers will look at the totality of the evidence to determine whether sufficient information has provided adequate assurance of clonality and provide the appropriate communication.

• To address a low probability and assurance of clonality, the control strategy can be supplemented to manage a non-clonal MCBs to support licensure.

23

Acknowledgments

• Sarah Kennett

• Joel Welch

• Chana Fuchs

• Juhong Liu

• Kathleen Clouse