Biosimilars in the EU – Regulatory update with focus on CMC CMC Strategy Forum Japan 2015 Tokyo Marriott Hotel, Tokyo, Japan November 9-10, 2015 Niklas Ekman, Ph.D., Senior Researcher Finnish Medicine Agency (FIMEA), Helsinki, Finland Disclaimer: The views expressed are my personal views and should not be understood or quoted as being made on behalf of or reflecting the position of the Finnish Medicines Agency, the European Medicines Agency or one of its committees or working parties.

What is a biosimilar? Current EU regulatory definition of biosimilars A biosimilar is a biological medicinal product that contains a version of the active substance of an already authorised original biological medicinal product (reference medicinal product). A biosimilar demonstrates similarity to the reference medicinal product in terms of quality characteristics, biological activity, safety and efficacy based on a comprehensive comparability exercise  The scientific principles of a biosimilar comparability exercise are based on those applied for evaluation of the impact on changes in the manufacturing process of a biological medicinal product (as outlined in ICH Q5E)

Lääkealan turvallisuus- ja kehittämiskeskus

Nov 9, 2015

CMC JPN 2015; [email protected]

2

Available guidance in the EU Definitions and main principles

Overarching Guideline CHMP/473/04 Rev. 1 “Guideline on Similar Biological Medicinal Products” Adopted 2005, revised 2014

General guidelines

Nonclinical and Clinical EMEA/CHMP/BMWP/42832/2005 Rev. 1

Adopted 2006, revised 2015

Class/ product specific guidance

Quality EMA/CHMP/BWP/247713/2012 Adopted 2006, revised 2014

ADOPTED

UNDER REVISION

Somatropin (2006) Epoetin (2006, revised 2010) IFNα (2009) Monoclonal Abs (2012) rFSH (2013) IFNβ (2013) Insulin (2006, revised 2015)

G-CSF (2006, CP 2015) LMWH (2009, dGL 2013)

Lääkealan turvallisuus- ja kehittämiskeskus

Nov 9, 2015

Immunogenicity assessment of biotech proteins (2008, CP 2014)

CMC JPN 2015; [email protected]

3

Biosimilar product experience in EU • 29 Marketing Authorisation (MA) applications for biosimilars reviewed by the CHMP • 21 positive (2 withdrawn by the MA holders post-approval), 7 withdrawn during evaluation, 1 negative

• 19 biosimilar medicinal products currently holding a valid marketing authoriation • 1 somatropin (authorised in 2006), 5 epoetin (first authorised in 2007), 8 filgrastim (2008), 2 infliximab (2013), 2 follitropin alfa (2013), 1 insulin glargine (2014)

• 4 biosimilar MA applications currently under review (Sep 2015) • Enoxaparin sodium, etanercept, infliximab, insulin human

Lääkealan turvallisuus- ja kehittämiskeskus

Nov 9, 2015

CMC JPN 2015; [email protected]

4

EMA biosimilar scientific advices 2004-2014 % mAb/ mAb-like proteins:

40

2010

2011

2012

2013

2014

69%

65%

63%

55%

76%

H1/2015 ~70%

35

Number of advices

30 25 20

15 10 5

0 2004

2005

2006

2007

2008

2009

First advice Lääkealan turvallisuus- ja kehittämiskeskus

Nov 9, 2015

2010

2011

2012

2013

2014

Follow-up advice

CMC JPN 2015; [email protected]

5

General principles for biosimilar development • In principle, the concept of a biosimilar is applicable to any biological medicinal product • The success of developing a biosimilar depends on; • The ability to manufacture a close copy of the reference medicinal product in a consistent manner • The ability to perform thorough physicochemical and biological characterization and to understand the clinical relevance of any differences detected • The ability to demonstrate bioequivalence • The availability of suitable clinical models; sensitive endpoints, possibility to identify relevant comparability margins

Lääkealan turvallisuus- ja kehittämiskeskus

Nov 9, 2015

CMC JPN 2015; [email protected]

6

General principles for biosimilar development • Similar physicochemical and biological properties is the basis for successful biosimilar development • Composition, physical properties, primary and higher order structures, purity, product-related isoforms and impurities, and biological activity • Note, no ”scaling” of similarity in EU legislation and guidelines → “similar”, ”highly similar” and ”highly similar with finger-print like similarity” = ”similar” • It may be challenging to claim biosimilarity if relevant quality differences are confirmed for which absence of an impact on safety or efficacy will be difficult to justify • Clinical data cannot be used to justify substantial differences in quality attributes

Lääkealan turvallisuus- ja kehittämiskeskus

Nov 9, 2015

CMC JPN 2015; [email protected]

7

Manufacturing process development • The development and documentation for biosimilars should cover two distinct aspects • Performance and consistency of the manufacturing process of the biosimilar on its own • Quality attributes of the target product profile should be comparable to the reference

• Quality Target Product Profile (QTPP) • Based on data collected on the chosen reference medicinal product, including publicly available information and data obtained from extensive characterisation of the reference medicinal product • Form the basis for the development of the biosimilar product and its manufacturing process • The manufacturing system needs to be properly designed to achieve the QTPP (e.g. with regard to the expression system) Lääkealan turvallisuus- ja kehittämiskeskus

Nov 9, 2015

CMC JPN 2015; [email protected]

8

The analytical similarity exercise Picture modified from Nat Biotechnol. 2008 Sep;26(9):985-90.

• An extensive, side-by-side comparability exercise is required. • Analytical methods should be appropriately qualified and applied with sufficient sensitivity to detect relevant differences • Orthogonal methods should be used whenever possible • Quantitative comparability ranges should be established • Ranges should be based primarily on the measured quality attribute ranges of reference product and should not be wider than the range of variability of the representative reference product batches

Lääkealan turvallisuus- ja kehittämiskeskus

Nov 9, 2015

CMC JPN 2015; [email protected]

9

The use of statistics in the analytical similarity exercise • Currently, the CHMP does not recommend any specific statistical method for assessment of biosimilarity • EMA/CHMP/BWP/247713/201 “A descriptive statistical approach to establish ranges for quality attributes could be used”

• Proposals for statistical analysis are welcomed but should be justified in the dossier • QA criticality ranking (link to the overall statistical approach), assay considerations, level of the QA in the product etc. • Appropriateness and limitations of the statistical analysis

• Various factors can influence the outcome of statistical tests • Sample size (unit of observation), origin (sampling strategy, sources of variability) and distribution (normality) • Risk of false positive conclusion • Acceptance/ equivalence ranges chosen Lääkealan turvallisuus- ja kehittämiskeskus

Nov 9, 2015

CMC JPN 2015; [email protected]

10

The use of statistics in the analytical similarity exercise • Similarity ranges established based on statistical intervals can lead to wide ranges with little (or no) clinical relevance • Depending on the chosen statistical tool (e.g. TI, PI, SD) and the data set available, the use of statistical intervals may result in an inability to detect relevant difference

• If inferential statistics is used, testing for equivalence (two-sided) should generally be applied • One-sided test could be acceptable for certain QAs (e.g. impurities)

• The final conclusion on analytical biosimilarity can not be drawn only based on statistical analyses • Raw data should always be provided in a suitable format • A statistically insignificant difference could in principle be clinically significant! Lääkealan turvallisuus- ja kehittämiskeskus

Nov 9, 2015

CMC JPN 2015; [email protected]

11

Experience from addressing the impact of glycosylation differences observed for a biosimilar monoclonal antibody Marketing Authorisation Application for Remsima/ Inflectra1 1 European

public assessment report (EPAR) available at www.ema.europa.eu

Lääkealan turvallisuus- ja kehittämiskeskus

Nov 9, 2015

CMC JPN 2015; [email protected]

12

Glycosylation and effector function analyses • Recombinant mAbs contain a large amount of different glycan types • As some of the glycan structures may affect activity, clearance and/or immunogenicity, detailed comparison of glycan structures is usually necessary • • • •

N-linked glycan analysis; site, structure and occupancy Carbohydrate content; neutral and amino sugars Sialic acid content (NANA and NGNA) Antennary profile, high-mannose variants

• Binding to Fc receptors (FcRn, FcγRI, FcγRII, FcγRIII) and complement (C1q) • Fc associated functional studies; ADCC and CDC • Especially if the MoA is known to involve Fc functions Lääkealan turvallisuus- ja kehittämiskeskus

Nov 9, 2015

CMC JPN 2015; [email protected]

13

Remsima/ Inflectra; Qaulity aspects • High similarity between the biosimilar and the reference demonstrated for • Primary, secondary and tertiary structures • In vitro TNFa neutralisation, binding affinity (soluble and transmembrane TNFa, TNFb, FcγRIa, FcγRIIa, FcRn, C1q), in vitro functional tests (apoptosis, CDC, ADCC using PBMNC effector cells from healthy volunteers)

• Minor differences reported for • C-terminal lysine content, aggregates, intact IgG level, charged molecular variants, glycosylation pattern • Binding to FcγRIIIa • Ex vivo binding to NK cells of healthy donors or CD patients (genotype dependent, only seen for V/V and V/F genotypes). No difference in the presence of diluted CD patient serum (to mimic the in vivo environment) Lääkealan turvallisuus- ja kehittämiskeskus

Nov 9, 2015

CMC JPN 2015; [email protected]

14

Potential mechanisms of action for anti-TNFs Primary MoA

Additional potential MoAs, especially in inflammatory bowel diseases (IBD)

Neutralisation of soluble TNFa

Modified from Thalayasingam N. et al., Best Pract Res Clin Rheumatol. 2011 Aug;25(4):549-67 Lääkealan turvallisuus- ja kehittämiskeskus

Nov 9, 2015

CMC JPN 2015; [email protected]

15

Can efficacy and safety data be extrapolated to all Remsima/Inflectra indications applied for? • Remsima was shown to contain a lower level of afucosylated glycans than Remicade, resulting in a lower binding affinity to FcγRIIIa and FcγRIIIb • Directive 2001/83, part II, Annex 1 In case the originally authorised medicinal product has more than one indication, the efficacy and safety of the medicinal product claimed to be similar has to be justified or, if necessary, demonstrated separately for each of the claimed indications

 As Fc-mediated functions could potentially be involved in the mode of action of infliximab, the Company was asked to provide further experimental data, in combination with an overall discussion, confirming that the differences detected do not affect efficacy or safety in any of the applied indications Lääkealan turvallisuus- ja kehittämiskeskus

Nov 9, 2015

CMC JPN 2015; [email protected]

16

• No difference in reverse signaling through tmhTNFa • Inhibition of apoptosis • Blockade of pro-inflammatory cytokine production

• No difference in blocking soluble hTNFa in an in vitro IBD model • Suppression of proinflammatory cytokine (IL-6 and IL-8) secretion from costimulated epithelial cell line • Suppression of epithelial cell line apoptosis

• No difference in Complementdependent cytotoxicity (CDC) activation Only for illustration, complete list available in the EPAR Lääkealan turvallisuus- ja kehittämiskeskus

Nov 9, 2015

CMC JPN 2015; [email protected]

17

Clinical impact of the difference in FcγRIIIa binding? • No difference in Regulatory Macrophage function (regMø) • Quantity of induced regulatory macrophages, suppression of T cell proliferation, in vitro wound healing

• No difference in Antibody-dependent cell-mediated cytotoxicity (ADCC) using • tmhTNFα-Jurkat cells as target cells and PBMCs (from healthy donors or CD patients), NK cells (from healthy donors) whole blood (from healthy donors) as effector cells • LPS-stimulated monocytes (from healthy donors or CD patients) as target cells and PBMC as effector cells

 Difference in ADCC functional assay detected only with • tmhTNFα-Jurkat cells as target cells and NK cells from CD patient donors (158V/V or 158V/F genotypes, but not 158F/F) as effector cells Lääkealan turvallisuus- ja kehittämiskeskus

Nov 9, 2015

CMC JPN 2015; [email protected]

18

Conclusion from the Remsima assessment • Functional difference was seen only in an ADCC assay employing artificially high tmTNFα expressing Jurkat target cells in combination with highly purified NK effector cells • No differences in experimental models regarded as more relevant to the pathophysiological conditions in CD patients

• No published reports describing the induction of ADCC by TNF antagonists in CD patients • No firm evidence that the FcγRIIIa polymorphism has an impact on the clinical course of CD  The CHMP concluded that the differences detected were not clinically meaningful, a positive opinion issued was issued

Lääkealan turvallisuus- ja kehittämiskeskus

Nov 9, 2015

CMC JPN 2015; [email protected]

19

Prerequisites for extrapolation • Similar physicochemical and structural characteristics, similar biological function in in vitro models • Similar human pharmacokinetics (exposure), pharmacodynamics, efficacy, safety, and immunogenicity at least in one therapeutic indication1 • Sound scientific justification • Clinical experience and available literature data from the reference product • Mechanism of action of the active substance in each indications • Evidence that the lead indication is representative for the other therapeutic indications, both with regard to safety and efficacy

Extrapolation of indication(s) is always a case-by-case decision and will depend on the totality of evidence presented 1) For

simple biologics, safety and efficacy studies may not always be necessary

Lääkealan turvallisuus- ja kehittämiskeskus

Nov 9, 2015

CMC JPN 2015; [email protected]

20

Summary – Value of Physiochemical and Biological Data in Demonstrating Biosimilarity  Analytical and biological comparability is the cornerstone for successful biosimilar development and forms the basis for the non-clinical and clinical comparability programs  As demonstrated by the Remsima case, thorough physicochemical and biological characterisation is also foremost important for successful extrapolation of indications  For more simple biologics, certain (non-) clinical studies could be waived by the CHMP based on comprehensive physicochemical and biological comparability data • Only appropriate if biosimilarity can be convincingly concluded based on physicochemical and functional characterisation (in combination with PK and PD profiles) • Could at some point be possible for more complex biologics as well? Lääkealan turvallisuus- ja kehittämiskeskus

Nov 9, 2015

CMC JPN 2015; [email protected]

21

Thank you for your attention! EMA Website http://www.ema.europa.eu/ema/

Biosimilar guidelines http://www.ema.europa.eu/ema/index.jsp?curl=pages/regulation/general/general_cont ent_000408.jsp&mid=WC0b01ac058002958c

Remsima EPAR (European public assessment report) http://www.ema.europa.eu/ema/index.jsp?curl=pages/medicines/human/medicines/00 2576/human_med_001682.jsp&mid=WC0b01ac058001d124 Niklas Ekman, Ph.D.

Senior Researcher, Quality/CMC Assessor Finnish Medicine Agency (FIMEA), Helsinki, Finland [email protected] Lääkealan turvallisuus- ja kehittämiskeskus

Nov 9, 2015

CMC JPN 2015; [email protected]

22