1 WHO/mAb_DRAFT/1 Mar 2016 2
ENGLISH ONLY 3
4
Guidelines on evaluation of monoclonal antibodies as 5
similar biotherapeutic products (SBPs) 6
7 NOTE: 8
This document has been prepared for the purpose of inviting comments and suggestions on the 9
proposals contained therein, which will then be considered by the Expert Committee on 10
Biological Standardization (ECBS). Publication of this early draft is to provide information about 11
the proposed Guidelines on evaluation of monoclonal antibodies as similar biotherapeutic 12
products (SBPs) to a broad audience and to improve transparency of the consultation process. 13
14
The text in its present form does not necessarily represent an agreed formulation of the 15
Expert Committee. Written comments proposing modifications to this text MUST be 16
received by 5 April 2016 in the Comment Form available separately and should be addressed 17
to the World Health Organization, 1211 Geneva 27, Switzerland, attention: Department of 18
Essential Medicines and Health Products (EMP). Comments may also be submitted electronically 19
to the Responsible Officer: Dr Kai Gao at email: [email protected]. 20
21
The outcome of the deliberations of the Expert Committee will be published in the WHO 22
Technical Report Series. The final agreed formulation of the document will be edited to be in 23
conformity with the "WHO style guide" (WHO/IMD/PUB/04.1). 24
25
© World Health Organization 2016 26 27
All rights reserved. Publications of the World Health Organization can be obtained from WHO Press, World Health Organization, 28 20 Avenue Appia, 1211 Geneva 27, Switzerland (tel.: +41 22 791 3264; fax: +41 22 791 4857; e-mail: [email protected]). 29 Requests for permission to reproduce or translate WHO publications – whether for sale or for non-commercial distribution – 30 should be addressed to WHO Press, at the above address (fax: +41 22 791 4806; e-mail: [email protected]). 31 32 The designations employed and the presentation of the material in this publication do not imply the expression of any opinion 33 whatsoever on the part of the World Health Organization concerning the legal status of any country, territory, city or area or of its 34 authorities, or concerning the delimitation of its frontiers or boundaries. Dotted lines on maps represent approximate border lines 35 for which there may not yet be full agreement. 36 37 The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed or recommended 38 by the World Health Organization in preference to others of a similar nature that are not mentioned. Errors and omissions 39 excepted, the names of proprietary products are distinguished by initial capital letters. 40 41 All reasonable precautions have been taken by the World Health Organization to verify the information contained in this 42 publication. However, the published material is being distributed without warranty of any kind, either expressed or implied. The 43 responsibility for the interpretation and use of the material lies with the reader. In no event shall the World Health Organization be 44 liable for damages arising from its use. 45
46 The named authors [or editors as appropriate] alone are responsible for the views expressed in this publication. 47
WHO/mAb_DRAFT/1 Mar 2016
Page 2
Contents 1
1. Introduction ............................................................................................................................... 3 2
2. Aim .............................................................................................................................................. 3 3
3. Scope ........................................................................................................................................... 4 4
4. Special considerations for characterization and quality assessment .................................... 4 5
5. Special considerations for nonclinial evaluation of monoclonal antibodies......................... 7 6
6. Special considerations for clinical evaluation of monoclonal antibodies ........................... 11 7
Authors and acknowledgements ................................................................................................ 24 8
References .................................................................................................................................... 26 9
10
WHO/mAb_DRAFT/1 Mar 2016
Page 3
1. Introduction 1
Monoclonal antibodies (mAbs) are a major class of recombinant deoxyribonucleic acid (rDNA) 2
technology derived biotherapeutic products and as such have achieved outstanding success in 3
treating many life-threatening and chronic diseases. Some of these targeted therapy products are 4
ranked in the top ten lists of annual global pharmaceutical revenue successes. As patents and data 5
protection measures on monoclonal antibody products have expired, or are nearing expiry, 6
considerable attention has turned towards producing copies of the monoclonal antibody innovator 7
products with a view to making more affordable products that may improve global access to these 8
so called block buster products. 9
10
Structurally, monoclonal antibodies are highly complex biological macromolecules with size and 11
charge variants, various post translational modifications, including different glycosylation 12
patterns, N and C terminal heterogeneity, long half-lives and the potential for inducing 13
immunogenicity. In addition, and importantly, they have several possible functional domains 14
within a single molecule, depending on such as the antigen-binding region, the complement –15
binding region and the constant part interacting with Fc receptors. Each individual monoclonal 16
antibody may therefore present a unique profile, a characteristic which needs to be taken into 17
consideration during the evaluation of these products as biosimilars. Therefore, comparability 18
studies of candidate biosimilar monoclonal antibodies and a biological reference product 19
monoclonal antibody are quite demanding and challenging for developers of these products as 20
well as for regulators. 21
22
WHO guidelines on evaluation of similar biotherapeutic products (SBPs) were adopted by the 23
Expert Committee on Biological Standardization (ECBS) in 2009 (1). This document provided a 24
set of globally acceptable principles regarding the regulatory evaluation of SBPs and served well 25
as a basis for setting national requirements for SBPs in general. However, in 2014 WHO was 26
requested to update these SBPs guidelines to take account of technological advances in the 27
characterization of rDNA derived products. In response, WHO organized an informal 28
consultation on the possible amendment of the 2009 SBP guidelines with an additional focus on 29
similar biotherapeutic products containing monoclonal antibodies in 2015. The outcome of these 30
discussions was that there was no need to revise the main body of the existing WHO SBPs 31
guideline. All participants, including national regulatory agencies and industry, recognized and 32
agreed that the evaluation principles described in WHO SBP guidelines are still relevant and 33
applicable. However, it was also agreed that there was a need for additional guidance for the 34
evaluation of biosimilar mAbs. 35
36
2. Aim 37
This product specific document is intended to provide special considerations for the evaluation of 38
monoclonal antibodies developed as similar biotherapeutic products and to complement the 39
WHO/mAb_DRAFT/1 Mar 2016
Page 4
existing WHO Guidelines. This document should therefore be read in conjunction with the 1
existing guidelines on biotherapeutics published by the WHO(1,2). Guidance on various aspects 2
of rDNA-derived medicines, SBPs and mAbs are also available from several other bodies. These 3
WHO Guidelines are not intended to conflict with, but rather to complement existing documents. 4
It takes the form of WHO guidelines rather than recommendations since mAbs represent a 5
heterogeneous class of products and the quality attributes of mAbs can vary from product to 6
product. Furthermore, one mAb product may have multiple indications, therefore comparability 7
study design of mAb SBPs will need to be adapted to suit the product in question, and follow 8
case by case manner. 9
10
3. Scope 11
The scope of products covered by this WHO guidance includes all biologically active mAbs 12
products used in the treatment of human diseases and which are prepared by rDNA technology. 13
14
4. Special considerations for characterization and quality assessment 15
Monoclonal antibodies for therapeutic use are preparations of an immunoglobulin or a fragment 16
of an immunoglobulin with specificity for a target ligand and derived from a single clone of cells. 17
Each molecule of monoclonal consists of two heavy and two light polypeptide chains which 18
linked by di-sulfide bonds and form a Y-shaped protein molecule. The defined specificity of a 19
monoclonal antibody is based on the binding region for an antigen that is located in the Fab part 20
of the molecule. Within their Fc-region they have the ability to bind to specific receptors leading 21
to immune effector functions such as antibody-dependent cellular cytotoxicity (ADCC) and 22
complement-dependent cytotoxicity (CDC), etc. Monoclonal antibodies are glycoproteins with 23
glycosylation sites in the Fc-portion of the heavy chains, with further glycosylation sites possible 24
dependant on the molecule. 25
26
From a regulatory perspective, mAb production and quality assessment is based on the same 27
principles as used for the evaluation of other rDNA-derived biotherapeutic proteins. Therefore, 28
the ‘WHO Guidelines on the quality, safety, and efficacy of biotherapeutic protein products 29
prepared by recombinant DNA technology’ (1) clearly apply applies to mAb SBPs. Biosimlar 30
mAb products should also comply with WHO ‘Guidelines on evaluation of similar biotherapeutic 31
products (SBPs)’ (2). A stepwise approach to the evaluation of mAb products in comparison with 32
a reference biological product is indicated, comprising first of a thorough head to head 33
comparability assessment of quality attributes, followed by head to head comparative nonclinical 34
and clinical assessment. The demonstration of biosimilarity of a candidate biosimilar mAb 35
product with respect to the RBP in terms of quality is a pre-requisite for moving forward into 36
comparative nonclinical and clinical studies. 37
38
Although general principles for quality assessment of biosimilar mAbs are already described in 39
existing WHO rDNA and SBP Guidelines (1,2), characterization of the biological activity of 40
mAbs is not considered in detail. However, mAb products are complex and variable 41
WHO/mAb_DRAFT/1 Mar 2016
Page 5
glycoproteins which have diverse and variable effector functions. This implies that potency 1
assessment for mAb biosimilars is particularly important and has some unique characteristics. 2
Also, the expression system used for production of mAbs can, in some cases, considerably affect 3
the structure of the mAb product. Therefore, the quality part of this document will focus on 4
specific considerations for assessment of mAb potency and on the impact of expression system 5
selected for production. 6
7
4.1 Strategy for mAb biological activity assessment 8
Potency/biological activity of mAb products is an important parameter and needs to be measured 9
appropriately by in vitro, in vivo, biochemical (including immunochemical assays) and/or 10
physicochemical assays as appropriate. Since changes of higher-order structure in a mAb could 11
alter its bioactivity, the analysis of bioactivity is very useful as a means of evaluating the 12
comparability of higher-order structure. 13
14
An understanding of the mode of action of the mAb is important for considering the strategy for 15
biological activity assessment both in the characterization and the comparability study. MAbs 16
exert their action by various mechanisms ranging from simple binding to antigen, which alone 17
mediates the clinical effect, to binding antigen and mediating one or more immunobiological 18
mechanisms which combine to produce the overall clinical response. These immunobiological 19
properties may play a role in the mechanism of action and/or have an impact on product safety 20
and efficacy. Therefore, a detailed analysis of the biological activity of the mAb, demonstrating 21
the mechanism of action (e.g. ligand neutralization, ADCC, CDC, apoptosis), ability for binding 22
to Fc gamma and neonatal Fc receptors, as well as complement binding– should be provided as 23
appropriate. For example, Rituximab which is a chimeric mAb specific for CD20 requires Fc 24
function and especially ADCC for its clinical efficacy in all its clinical indications. Assessment 25
of Fc function (including ADCC) is therefore paramount for this mAb and biosimilar candidate 26
products should have comparable ADCC to the reference product. However Infliximab (a tumor 27
necrosis factor alfa (TNF alfa) antagonist) seems not to require Fc function for its clinical 28
efficacy in rheumatoid arthritis and psoriasis although this may be needed for effective treatment 29
of Crohn’s disease and inflammatory bowel disease (IBD). 30
31
Assays for measuring Fc function can be technically demanding. For example ADCC assays 32
require appropriately responsive target and efficient effector cells and identifying or producing 33
these can be difficult and arduous. The use of continuously growing cell lines may overcome this 34
is some cases. It is often necessary to engineer cells to obtain appropriate responses. Different 35
data may be generated with different assay formats and different cells/cell combinations. 36
37
If it can be shown that binding alone is the major mode of action of the mAb and 38
immunobiological activities are consistent, ligand binding assays may be used as a surrogate for 39
WHO/mAb_DRAFT/1 Mar 2016
Page 6
potency estimation for mAbs (for routine control) rather than classic bioassays. In some cases 1
these non-cell based potency assays may show technical advantages compared to cell-based 2
assays. 3
4
Although simple binding may seem to be the only mechanism operating to achieve clinical 5
efficacy, other effects may also play a role in this. In some cases multiple functions of the mAb 6
may be involved in an additive or synergistic manner to produce an overall combined clinical 7
effect and this may be hard to dissect experimentally to allow a clear understanding of how the 8
mAb mediates its clinical potency. Therefore, if intact mAbs are used, care must be taken not to 9
assume that the Fc mediated immunobiological effects of the product are not involved in clinical 10
efficacy, even when simple antigen binding is considered to be the primary mode of action. In 11
this regard, use of a properly design cell based assay for measuring potency has an advantage. 12
13
4.2 Consideration of production cell line changes and resultant glycoform 14
differences 15
The WHO SBP guideline allows for the use of different expression systems for production of the 16
biosimilar and the reference product, if the manufacturer can demonstrate convincingly that the 17
structure of the molecule is not affected or that the clinical profile of the product will not change. 18
However, the use of a different host cell type may not be advisable for biosimilar glycoproteins 19
because glycosylation patterns can vary significantly between different host cell types. Although 20
mAbs are not highly glycosylated products, their molecular complexity and the importance of 21
some post translational modifications requires as close as possible production processes 22
(including the cells used to produce product) for the biosimilar and reference product to achieve 23
acceptable biosimilarity. For instance, fucose bound by a α1-6 linkage to the trimannosyl core 24
portion of N-linked carbohydrate chains or a non-reduced terminal galactose regulates the 25
cytotoxic activity of antibodies. Therefore, the similarity of carbohydrate chains should be 26
examined and compared for then SBP and reference biotherapeutic product (RBP). Also, it is 27
useful to assess the relationship between the carbohydrate chain structures of mAbs (SBP and 28
RBP) and their pharmacokinetics as glycosylation is known to potentially affect this. 29
30
In view of the above, the use of different cell lines can result in differences in glycoforms present 31
on products, which may or not have clinical consequences. For example, production cells based 32
on mouse cell lines, such as SP2/0 and NS0 secrete mAbs with the carbohydrate structure gal-33
alpha-1, 3-gal present on the carbohydrate moiety. Humans cannot produce the gal-alpha-1, 3-gal 34
structure as they lack the necessary enzyme for its synthesis; however many humans produce 35
antibodies against this. In a proportion of these individuals the antibodies are of the IgE class and 36
this sensitization can result in anaphylactic reactions (often serious) if they are treated with 37
mouse cell line derived, gal-alpha-1, 3-gal containing mAbs. This pre-existing antibody problem 38
has been particularly evident for the mAb product Erbitux (Cetuximab; anti epidermal growth 39
factor receptor receptor (EGFR)). The problem can be avoided by using human or some clones of 40
Chinese hamster ovary (CHO) cells for production of mAb products as these cells cannot 41
WHO/mAb_DRAFT/1 Mar 2016
Page 7
synthesise the gal-alpha-1, 3 -gal antigen. However this type of phenomenon can have important 1
implications for biosimilar mAb development. For example, producing a copy of Erbitux in 2
mouse cells could result in a biosimilar, but this would probably show the same gal-alpha-1, 3 -3
gal related anaphylaxis problems as the reference product. However production of the mAb in 4
CHO cells may avoid the anaphylaxis problem, as the gal-alpha-1, 3 -gal structure would not be 5
present on the mAb, but the differences in glycosylation, and possibly other post-translational 6
modifications may disqualify the new mAb as a biosimilar, even though it has advantages over 7
the Erbitux reference product. 8
9
Therefore choice of expression system for biosimilar mAbs needs careful consideration as 10
various potential issues need to be assessed. Development of mAb products as biosimilars rather 11
than as stand-alone products also needs thoughtful assessment, including a consideration of the 12
clinical effects (beneficial and adverse) of the potential reference product(s). 13
14
4.3 Standards 15
International standards and reference preparations have been established for a wide range of 16
biological substances prepared by rDNA technology. These standards and materials are used 17
either to calibrate assays directly or to calibrate secondary standards or manufacturers’ working 18
standards. A list of such materials is available on the WHO website1. Each
standard or reference 19
preparation is held by one of the WHO custodian laboratories (e.g. the National Institute for 20
Biological Standards and Control, Potters Bar, England). 21
Biological assays used to assess the biological activity of mAbs should be validated and traced to 22
WHO IS or WHO reference reagents when available. The different uses of Reference Products 23
and International Standards in biosimilar product development and evaluation are described 24
elsewhere (1,3). 25 1
See: http://www.who.int 26
27
5. Special considerations for nonclinical evaluation of monoclonal antibodies 28
As regards nonclinical development, as for all SBPs, a step-wise approach should be applied to 29
evaluate the similarity of biosimilar and reference mAb. Nonclinical studies should be performed 30
before initiating clinical trials. In vitro studies should be conducted first and a decision then made 31
as to the extent of what, if any, in vivo work will be required. 32
The following approach may be considered and should be tailored to the similar biotherapeutic 33
product concerned on a case-by-case basis, also need to be fully justified in the nonclinical 34
overview. 35
36
5.1 In vitro studies 37
SBP - general aspects 38
WHO/mAb_DRAFT/1 Mar 2016
Page 8
In order to assess any difference in biological activity between the similar and the reference 1
biotherapeutic product, data from a number of in vitro studies, some of which may already be 2
available from quality-related assays, should be provided. 3
As for all SBPs, the following general principles apply to similar biotherapeutic mAbs: 4
The studies should be comparative in nature and should not just assess the response per se. To 5
obtain unambiguous results, the methods used should be scientifically valid and suitable for 6
their purpose. The studies should be sensitive, specific and sufficiently discriminatory to 7
provide evidence that observed differences in quality attributes are clinically not relevant. 8
Functional studies should be designed to be sensitive enough to detect differences in the 9
concentration–activity relationship between the SBP and the RBP. 10
They should be performed with an appropriate number of batches of the reference product and 11
of the similar biotherapeutic product representative of the material intended for clinical use. 12
Assay and batch-to-batch variability will affect the number needed. The number tested should 13
be sufficient to draw meaningful conclusions on the variability of a given parameter for both 14
the similar biotherapeutic and the reference product and on the similarity of both products. 15
Together, these assays should cover the whole spectrum of pharmacological/ toxicological 16
aspects known to be of clinical relevance for the reference product and for the product class. 17
The manufacture should discuss to what degree the in vitro assays used are representative/ 18
predictive for the clinical situation according to current scientific knowledge. 19
Since in vitro assays may often be more specific and sensitive to detect differences between the 20
biosimilar and the reference product than studies in animals, these assays can be considered as 21
paramount for the nonclinical biosimilar comparability exercise. 22
23
Similar biotherapeutic mAbs - specific aspects 24
For similar biotherapeutic mAbs, the nonclinical in vitro program should usually include relevant 25
assays for the following specific topics: 26
Binding studies: 27
– binding to target antigen(s); 28
– binding to representative isoforms of the relevant three Fc gamma receptors (FcγRI, FcγRII 29
and FcγRIII), FcRn and complement (C1q); 30
Functional studies/Biological Activities: 31
– Fab-associated functions (e.g. neutralization of a soluble ligand, receptor activation or 32
blockade); 33
– Fc-associated functions (e.g. antibody-dependent cell-mediated cytotoxicity, ADCC; 34
complement-dependent cytotoxicity, CDC; complement activation); 35
Together these assays should broadly cover the functional aspects of the mAb even though some 36
may not be considered essential for the therapeutic mode of action. However, an evaluation of 37
ADCC and CDC is generally not needed for mAbs directed against non-membrane bound targets. 38
39
Additional notes 40
WHO/mAb_DRAFT/1 Mar 2016
Page 9
As indicated in the ICH S6 (R1) guideline(4), tissue cross-reactivity studies with mAbs are not 1
suitable to detect subtle changes in critical quality attributes and are thus not recommended for 2
assessing comparability. 3
4
As for all SBPs, if the comparability exercise using the above strategy indicates that the test mAb 5
and the reference mAb cannot be considered biosimilar, it may be more appropriate to consider 6
developing the product as a stand-alone. 7
8
5.2 In vivo studies 9
Determination of the need for in vivo studies 10
As for SBPs in general, based on the totality of quality and nonclinical in vitro data available 11
and on the extent to which there is residual uncertainty about the similarity of the proposed 12
biotherapeutic mAb and the reference mAb, it is at the discretion of the National Regulatory 13
Authorities (NRAs) to waive/not to waive the request for nonclinical in vivo studies. If the 14
quality comparability exercise and the nonclinical in vitro studies are considered satisfactory 15
and no issues are identified which would block direct entrance into humans, an in vivo animal 16
study may not be considered necessary. 17
However, some mAbs may mediate effects that cannot be fully elucidated by in vitro studies. 18
Therefore, in such cases nonclinical in vivo studies may be necessary to provide 19
complementary information. 20
21
General aspects to be considered for all SBPs including similar biotherapeutic mAbs 22
If there is a need for additional in vivo information, the availability of a relevant animal 23
species or other relevant models (e.g. transgenic animals, transplant models) should be 24
considered. 25
If a relevant in vivo animal model is not available the manufacture may choose to proceed to 26
human studies taking into account principles to mitigate any potential risk. 27
Factors to be considered when the need for additional in vivo nonclinical studies is evaluated, 28
include but are not restricted to: 29
– presence of potentially relevant quality attributes that have not been detected in the 30
reference product (e.g. new post-translational modification structures); 31
– presence of potentially relevant quantitative differences in quality attributes between the 32
similar biotherapeutic product and the reference product; 33
– relevant differences in formulation, e.g. use of excipients not widely used for mAbs. 34
Although each of the factors mentioned here do not necessarily warrant in vivo testing, 35
these issues should be considered together to assess the level of concern and whether there 36
is a need for in vivo testing. 37
If product-inherent factors that impact PK and/or bio-distribution, like extensive glycosylation, 38
cannot sufficiently be characterized on a quality and in vitro level, in vivo studies may be 39
WHO/mAb_DRAFT/1 Mar 2016
Page 10
necessary. The manufacture should then carefully consider if these should be performed in 1
animals or as part of the clinical testing, e.g. in healthy volunteers. 2
3
Performance of in vivo studies 4
The following explanations apply to all SBPs, including similar biotherapeutic mAbs. 5
6
General aspects 7
If an in vivo evaluation is deemed necessary, the focus of the study/studies (Pharmacokinetic, PK 8
and/or Pharmacodynamics, PD and/or safety) depends on the need for additional information. 9
10
Animal studies should be designed to maximize the information obtained. The duration of the 11
study (including observation period) should be justified, taking into consideration the PK 12
behaviour of the reference mAb and its clinical use. 13
14
Effects of SBPs are often species-specific. In accordance with ICH S6(R1) and WHO 15
“Guidelines on the quality, safety and efficacy of biotherapeutic protein products prepared by 16
recombinant DNA technology (2,4) ,in vivo studies should only be performed in relevant species, 17
i.e. (a) species which is/are pharmacologically and/or toxicologically responsive to the SBP. 18
19
PK and/or PD studies 20
When the model allows, the PK and PD of the SBP and the RBP should be quantitatively 21
compared, including, if feasible, a dose-response assessment including the intended exposure in 22
humans. 23
24
In vivo assays may include the use of animal models of disease to evaluate functional effects on 25
PD markers or efficacy measures. 26
27
Safety studies 28
For safety studies a flexible approach should be considered, in particular if non-human primates 29
are the only relevant species. The conduct of standard repeated dose toxicity studies in non-30
human primates is usually not recommended. If appropriately justified, a repeated dose toxicity 31
study with refined design (e.g. using just one dose level of biosimilar and reference product 32
and/or just one gender and/or no recovery animals) and/or an in-life evaluation of safety 33
parameters (such as clinical signs, body weight and vital functions) may be considered. 34
Depending on the endpoints needed, it may not be necessary to sacrifice the animals at the end of 35
the study. 36
37
For repeated dose-toxicity studies where only one dose is evaluated, this would usually be 38
selected at the high end of the dosing range and should be justified on the basis of the expected 39
toxicity of the RBP. 40
41
WHO/mAb_DRAFT/1 Mar 2016
Page 11
The conduct of toxicity studies in non-relevant species (i.e. to assess unspecific toxicity only, 1
based on impurities) is not recommended. Due to the different production processes used by the 2
biosimilar and reference product manufacturers, qualitative differences of process related 3
impurities will occur (e.g. host cell proteins). Such impurity level should be kept to a minimum to 4
minimize any associated risk. 5
6
Immunogenicity studies 7
Qualitative or quantitative difference(s) of product-related variants (e.g. glycosylation patterns, 8
charge variants) may have an effect on immunogenic potential and potential to cause 9
hypersensitivity. In general, these effects are difficult to predict from animal studies and should 10
be further assessed in clinical studies. 11
12
However, while immunogenicity assessment in animals is generally not predictive for 13
immunogenicity in humans, it may be needed for PK/toxicokinetics (TK) interpretation of in vivo 14
animal studies. Therefore, adequate blood samples should be taken and stored for future 15
evaluations if then needed. 16
17
Local tolerance studies 18
Studies on local tolerance are usually not required. If excipients are introduced for which there is 19
no or little experience with the intended clinical route, local tolerance may need to be evaluated. 20
If other in vivo studies are performed, evaluation of local tolerance may be part of the design of 21
that study instead of the performance of separate local tolerance studies. 22
23
Other studies 24
Studies regarding safety pharmacology and reproduction toxicology are usually not required for 25
nonclinical testing of SBPs. 26
27
In accordance with ICH S6 (R1) and WHO “Guidelines on the quality, safety and efficacy of 28
biotherapeutic protein products prepared by recombinant DNA technology” (2,4), genotoxicity 29
and (rodent) cancerogenicity studies are not required for (similar) biotherapeutic products. 30
31
6. Special considerations for clinical evaluation of monoclonal antibodies 32
6.1 PK 33
6.1.1 Aim of comparative PK studies 34
Comparative PK studies should be used to further confirm the comparability of mAb SBPs that is 35
established through comparative structural, functional and nonclinical studies. For instance, a 36
mAb SBP may differ in its affinity for FcRn receptors from its RBP which may lead to could 37
have a shorter or longer half-life. As a consequence of shorter half-life, drug exposure would be 38
reduced which may lead to a lower of efficacy (5). Comparative PK studies may also be useful in 39
WHO/mAb_DRAFT/1 Mar 2016
Page 12
monitoring the impact of anti-drug antibody formation, and establishing evidence in support of 1
indication extrapolation. It is not necessary to study the mAb SBP in every indication that is 2
being sought. However, separate comparative PK studies may be required to bridge multiple 3
distinct conditions that have been authorized for the reference mAb. The number of studies 4
required depends both on the degree of similarity between the mAb SBP and the reference 5
(ascertained from the data of chemical and manufacture control (CMC)) and on the indications 6
for which the mAb SBP is proposed. The design of the comparative PK studies depends on 7
various factors, including clinical context, safety profiles, and PK characteristics of the reference 8
(target-mediated disposition, linear or non-linear PK, time-dependency, half-life, etc.). 9
10
6.1.2 Study design and population 11
A single-dose PK study in healthy volunteers is generally recommended, as it can be considered 12
as a sensitive and homogenous population (6). A parallel group design is usually required for 13
mAbs since a single-dose cross-over design is rarely appropriate due to the long half-lives of 14
mAbs and the potential influence of immunogenicity on the PK profile. Regarding the use of 15
healthy volunteers to study the PK of mAbs, there are some key considerations that should be 16
taken into account. First, healthy subjects may exhibit a greater immunogenic response than a 17
population with disease due to less target-mediated clearance and lack of concomitant medication, 18
or background standard of care (e.g. immunosuppressive therapies) and other population factors 19
(7). 20
21
Second, healthy subjects may not be appropriate if the disease relevant to the indication, or its 22
treatment, is known to alter the PK of the reference product. For instance, receptor density, 23
concomitant use of other therapeutic agents, disease type, degree of systemic inflammation, FcγR 24
polymorphisms, the maximum concentration achieved in relation to antigenic mass are some of 25
the situations where healthy subjects would not be appropriate. Third, a clinically relevant dose 26
of anti-cancer mAb may not be considered as ethical in healthy volunteers due to safety concerns. 27
Thus, a lower therapeutic dose or a dose below the therapeutic dose may be required. 28
29
It may be necessary to perform the PK study for mAb SBPs in a sensitive patient population 30
rather than in healthy volunteers, if the likely risk/benefit ratio would be unfavorable in healthy 31
volunteers as in the case of Rituximab. Unnecessary exposure to risk (safety, medical reasons) 32
would be viewed as unethical. It is preferable that the chosen patient population be different from 33
that selected to demonstrate the absence of clinically meaningful differences in safety and 34
efficacy. 35
36
The choice of a particular population for PK analysis also depends on the range of therapeutic 37
indications of the mAb under development. For example, if a reference mAb is authorized both as 38
an anti-inflammatory agent and as an anticancer antibody (e.g. as with Rituximab) PK data in one 39
therapeutic area may complement clinical data obtained in the other therapeutic area and thus can 40
also strengthen the evidence for indication extrapolation. In addition, the most sensitive 41
WHO/mAb_DRAFT/1 Mar 2016
Page 13
population in whom PK characteristics is compared may not be the same as the population 1
chosen for comparison of efficacy. To aid in selecting a sensitive study population, patient 2
characteristics, such as the level of antigen/receptor and severity of disease, should be considered. 3
Factors such as these may have an impact on the assessment of PK comparability. 4
5
6.1.3 Regimen 6
MAbs are often indicated for both monotherapy and as a part of combination regimens that 7
incorporate immune-suppressants or chemotherapeutics. It may be sensible to study the 8
comparative PK in the monotherapy setting in order to minimize the sources for variability. 9
10
6.1.4 PK characteristics of the reference mAb 11
The PK of the mAb may be affected by factors such as the antigen/ receptor level (e.g. related to 12
tumor burden in oncology), the existence of target-mediated clearance and/or receptor shedding. 13
These factors may be indication specific and should be considered when selecting the dosage 14
regimen and the population in whom to compare the PK of the SBP to the reference. 15
16
6.1.5 Doses 17
A dose should be selected that will enable the detection of potential PK differences between the 18
mAb SBP and the reference mAb. MAbs generally possess a high degree of target selectivity, 19
with many exhibiting nonlinear distribution and elimination, influenced by binding to their target. 20
In general, it is recommended that the PK be compared using the lowest recommended 21
therapeutic dose. For mAbs that are eliminated by target-mediated disposition, a low dose (i.e. 22
one at which target-mediated disposition (TMD) is not saturated) may be particularly useful to 23
detect differences in PK (8). 24
25
6.1.6 Routes of administration 26
Administration via a route that requires an absorption step is preferred. Where the route of 27
administration requires an absorption step, such as the subcutaneous route (SC), standard 28
comparisons of Cmax and AUCT may be used to assess PK comparability. It is also expected that 29
the SC route is more immunogenic than the intravenous route (IV), which makes the SC route 30
more appropriate for the investigation of potential immunogenic differences between the mAb 31
SBP and the reference. However, many mAbs are only administered via IV injection or infusion, 32
which may necessitate a more comprehensive comparative assessment of PK parameters that 33
reflect clearance including half-life (t1/2), volume of distribution (Vd), clearance (CL)and the 34
elimination rate constant (Kel). 35
36
If the application is envisaged with the use of auto-injector, it would be necessary to provide 37
clinical data on that (e.g. third arm in the clinical study). 38
39
WHO/mAb_DRAFT/1 Mar 2016
Page 14
6.1.7 Sampling times 1
Primary PK comparability studies should include sufficient sampling to accurately characterize 2
Cmax and should adequately characterize the late elimination phase such that the terminal 3
disposition rate constant can be reliably estimated. In single-dose studies, optimal sampling 4
should continue past the expected last quantifiable concentration (AUCT), and the concentration-5
time curve should cover at least 80% of AUC0-inf. 6
7
In a multiple dose study, the primary parameters should be the truncated AUC after the first 8
administration until the second administration (AUC0-t) and AUC over a dosage interval at steady 9
state. The proposed duration of the study typically corresponds to 5 half-lives of the mAb. 10
11
6.1.8 Specific assays for serum drug concentration 12
It is preferable to have a single, validated bio-analytical assay to detect both the mAb SBP and 13
the reference mAb. The bio-analytical assay should be appropriate for the detection and 14
quantification of mAbs and should be demonstrated to be bio-analytically comparable with 15
respect to its ability to precisely and accurately quantify both the mAb SBP and the reference (9). 16
17
6.1.9 Parameters in PK 18
The PK endpoints necessary to demonstrate PK comparability may differ depending on the route 19
of administration and the study design. In general, single-dose studies with an absorption step 20
should compare AUC-t and Cmax between the mAb SBP and the reference mAb. If a multiple 21
dose study is performed in patients, sampling at first dose and at steady state should be performed. 22
PK parameters that should be evaluated include AUC, Cmax, and Ctrough, clearance and half-life. 23
For mAbs that are administered only intravenously, the aforementioned parameters should be 24
compared as well as parameters that reflect the clearance of the product. 25
26
Anti-drug antibodies (ADA) should be measured in parallel of PK assessment using the most 27
appropriate sampling time points and a subgroup analysis by ADA status should be performed. If 28
ADA is measured, characterization (neutralizing capacity) over time should be determined since 29
this may add valuable information on observed PK parameters. The PK analysis on the ADA 30
negative samples is of particular interest, providing the clearest picture on PK similarity. 31
32
6.2 PD 33
The overall objective of the development program is to establish comparability. Therefore, the 34
selection of the primary patient population is driven by the need for homogeneity and sensitivity. 35
36
6.2.1 PD markers and PD assay 37
The effect of the mAb SBP/ reference on exploratory PD markers (biomarkers) may be included 38
as additional support for comparability. Formal inferential analysis is not recommended since 39
reliable PD markers (i.e. those that are predictive of clinical outcomes) have not been identified 40
WHO/mAb_DRAFT/1 Mar 2016
Page 15
for most mAbs. If PD markers are considered relevant, manufacturers should pay attention to the 1
following factors (6): 2
sensitive enough to detect relevant differences, and if they can be measured with sufficient 3
precision; 4
the use of multiple PD markers, if they exist, is recommended; 5
to study dose-concentration-response relationships or time-response relationships provided 6
that the selected doses are within the linear part of the established dose-response curve of RBP; 7
a clear dose-response relationship is shown; 8
at least one PD marker is an acceptable surrogate marker and is related to patient outcome; 9
the PD assay should at least be relevant to a pharmacological effect of the biological product 10
(PD assay is highly dependent on the pharmacological activity of the product, the approach for 11
assay validation and the characteristics of the assay performance may differ depending on the 12
specific PD assay). 13
14
6.3 Primary clinical study 15
The purpose of the efficacy trials is to confirm comparable clinical performance of the SBP and 16
the reference product. The comparative assessment of efficacy and safety is a key component of 17
the clinical assessment for mAb SBPs. The extent and nature of the nonclinical in vivo studies 18
and clinical studies to be performed depend on the level of evidence obtained in the previous 19
step(s) including the robustness of the physicochemical, biological and nonclinical in vitro data. 20
It is important to note that clinical data cannot be used to justify substantial differences in quality 21
attributes. If dose comparative and sensitive PD studies cannot be performed to convincingly 22
demonstrate comparability in a clinically relevant manner for mAbs, evidence of comparability 23
needs to be obtained by other studies. If there is residual uncertainty that remains based on 24
comparative structural and functional characterization, animal testing and human PK/PD data, 25
comparative clinical trials should be carried out to demonstrate clinical comparability between 26
the proposed mAb SBP and the reference mAb. These clinical trials should be randomized, 27
adequately powered, and preferably double blinded. 28
29
The primary clinical study is preferred to be conducted in a population that is different from the 30
population for the PK/PD study. Manufacturer of a biosimilar product should perform a thorough 31
analysis of the available clinical data in the public domain for the reference product in order to 32
determine the population-endpoint combination that is likely to provide a relevant and sensitive 33
model for detecting clinically meaningful differences in efficacy and safety. However, the choice 34
of setting for the primary clinical study is dependent on many intrinsic and extrinsic factors, 35
which may not always lead to the selection of the most sensitive/relevant population. The type 36
and number of comparative clinical trials required for the proposed mAb SBP could be affected 37
by several factors including a) the nature and complexity of the mAbs; b) the limitations of 38
studies comparing structural and functional characteristics; c) the extent to which differences in 39
WHO/mAb_DRAFT/1 Mar 2016
Page 16
structure, function, nonclinical data and human PK/PD can predict clinical outcomes; d) the 1
degree of understanding of mechanism(s) of action of mAb and disease pathology; and e) the 2
extent of clinical experience with the reference mAb with respect to safety, efficacy, and 3
immunogenicity. The clinical data obtained from one or more clinical studies in the most 4
sensitive populations can also be used to support extrapolation to other indications in which the 5
proposed mAb SBP has not been tested. 6
7
6.3.1 Clinical trial design 8
Clinical trial design and statistical analysis of equivalence and non-inferiority trials that are 9
already addressed in the WHO “Guidelines on evaluation of similar biotherapeutic products 10
(SBPs)” clearly apply to biosimilar mAbs (1). The guidelines stress the importance of clearly 11
stating the specific design selected for a given study, and includes details on the determination of 12
the comparability margin, sample size determination, and statistical analyses. For biosimilar 13
mAbs, extrapolation to other indications is especially important, and additional considerations are 14
required in order to design a meaningful trial to support additional indications. 15
16
Although equivalence or non-inferiority studies would be acceptable for the comparative clinical 17
studies of the biosimilar mAb to the RBP, equivalence trials are generally preferred. Detailed 18
explanations regarding the advantages and disadvantages of equivalence/ non-inferiority design 19
for similar biotherapeutic products are provided in WHO Guidelines and also available from 20
several other bodies on SBPs (1,10,11,12). Special considerations for clinical trial design of 21
mAbs developed as biosimilars are provided below. 22
23
A demonstration of equivalence, as opposed to non-inferiority, is especially important given that 24
extrapolation to other indications is one of the goals of the development program for the 25
biosimilar mAb. Non-inferiority trials are one sided and hence do not exclude the possibility that 26
the biosimilar mAb could be found to be superior to the RBP. Such a finding would create 27
challenges in providing a justification for extrapolation to other indications of the RBP. 28
29
An important consideration during the planning and design of the equivalence trial for the 30
biosimilar mAb is that it should be conducted with a sensitive and well established clinical model 31
regarding both the study population and the study endpoints. Assay sensitivity is important to 32
provide some confidence that the trial, as planned and designed, will have the ability to detect 33
differences between the biosimilar mAb and the RBP if such differences exist (13). A trial that 34
lacks sensitivity could lead to the erroneous conclusion of equivalence of the biosimilar mAb to 35
the RBP. The selected study population should not only be representative of the approved 36
therapeutic indications of the RBP, but should also be sensitive to detect potential differences 37
between the biosimilar mAb and the RBP. Hence, historical scientific evidence should be 38
provided which shows that appropriately designed and conducted trials with the RBP against 39
placebo for the approved indication have reliably demonstrated the superiority of the RBP over 40
placebo. For situations in which relevant scientific advances make it necessary for the study 41
WHO/mAb_DRAFT/1 Mar 2016
Page 17
population or study endpoints to deviate from those leading to approval of the RBP for the 1
specific indication, the proposed deviations should be discussed with the relevant regulatory 2
authorities during the planning and design stage of the trial, and should include details on how 3
the comparability margin(s) will be determined. 4
5
The selected study endpoints are also important determinants of sensitivity. Endpoints that are 6
sensitive to detecting clinically meaningful differences between the biosimilar mAb and the RBP 7
should be used. Although endpoints such as PD measures are expected to be more sensitive than 8
clinical endpoints, there are few established PD endpoints for mAbs and clinical endpoints will 9
likely have to be utilized. Effort should be made to use the same primary endpoint(s) as those that 10
were used for establishing efficacy of the RBP. However, if that is not feasible, historical 11
evidence to support sensitivity of the selected endpoint(s) as well as the determination of the 12
equivalence margin(s) should be provided. 13
14
Since the efficacy of the RBP compared to placebo has already been demonstrated previously and 15
it is also considered clinically important to ensure that the biosimilar mAb retains a substantial 16
fraction of the effect of the RBP, an equivalence margin that preserves a fraction of the smallest 17
effect size that the RBP can be expected to have relative to a placebo control is the most suitable 18
margin. The fraction of the effect size of the RBP that must be retained by the biosimilar mAb 19
should be clearly justified in each case, and should take into account the smallest clinically 20
important difference in a given setting. A commonly used value is the 50 % rule in which 50 % 21
of the effect size of the RBP is preserved, and its wide use appears to be historical. However, the 22
50 % rule is not expected to be justifiable in all cases as it could result in a margin that is larger 23
than what is considered the smallest clinically important difference. Once the margin has been 24
selected, the determination of the required sample size should be based on methods specifically 25
designed for equivalence trials. 26
27
Statistical analysis of data from equivalence trials is typically based on the indirect confidence 28
interval comparison which requires specification of the equivalence limits (14). Equivalence is 29
demonstrated when the confidence interval for the selected metric of the treatment effect falls 30
entirely within the lower and upper equivalence limits. If a p-value approach is used, then the p-31
values should be computed based on the Two-One Sided Test (TOST) procedure testing 32
simultaneously the null hypotheses of inferiority and superiority. In using the TOST procedure, 33
equivalence is demonstrated when the p-values obtained are less than the significance level used. 34
35
6.3.2 Study population 36
In order to detect differences between the mAb SBP and the reference mAb, clinical trials for 37
mAb SBPs should be carried out in an appropriately sensitive patient population using endpoints 38
that can accurately demonstrate both the similarity of the biosimilar to the reference product, and 39
WHO/mAb_DRAFT/1 Mar 2016
Page 18
its efficacy and safety in the indication (see Section 6.3.3). Rationale for the selected study 1
population should be provided. In general, a homogeneous population of patients would 2
minimize inter-patient variability and thus increase the likelihood of detecting differences 3
between the mAb SBP and the reference mAb. Patients who have not received previous treatment, 4
e.g. first-line therapy, are considered to be more homogeneous than patients who have previously 5
received several or different lines of therapy. Ideally, the observed clinical effects should be 6
triggered by the direct action of the mAb SBP/reference mAb without interference of other 7
medications, as concomitant medications may affect or mask differences in PK/PD, efficacy, 8
safety and/or immunogenicity of the tested products. To validate the effect of the reference mAb 9
and the sensitivity of the study in the chosen study population, historical data should be used to 10
justify the selection of the study population and equivalence margin. This could generally be 11
done through meta-analysis or systematic review. 12
13
Monoclonal antibodies can function through various mechanisms of action, such as through 14
receptor blockade or agonist activity (e.g. vascular endothelial growth factor (VEGF) and EGFR), 15
induction of apoptosis, or delivery of a drug or cytotoxic agent, and immune-mediated cell killing 16
mechanisms, e.g. CDC, ADCC and regulation of T cell function. Because the mechanisms 17
involved in one disease may differ from those involved in another, extensive consideration 18
should be given to the setting in which clinical comparability is to be tested, especially where it is 19
known that extrapolation to other indications and uses will be sought. 20
21
In some jurisdictions, clinical studies in an unauthorized population (e.g. new indication, line of 22
therapy, combined therapy, disease severity) may be acceptable to demonstrate “no clinically 23
meaningful differences” for biosimilars. Manufacturers of biosimilar products should consult 24
relevant regulatory authorities prior to conducting such studies. 25
26
6.3.3 Primary study endpoint 27
Clinically relevant and sensitive study endpoints within a sensitive population should be selected 28
to improve the detection of potential differences between the mAb SBP and the reference product. 29
In general, clinical outcomes, surrogate outcomes or a combination of both can be used as 30
primary endpoints in mAb SBP trials. The same study endpoints for the innovator products may 31
be used, as a large body of historical data is generally available in the public domain for setting 32
the equivalence margin and calculating the sample size. The study endpoints may be different 33
from those traditionally used or study guideline recommended endpoints for the innovator 34
products, as these endpoints may not be considered as the most sensitive endpoints to detect 35
clinically meaningful differences in an equivalent trial setting. A surrogate endpoint can be used 36
as the primary endpoint, when the surrogacy to the clinical outcome is well-established or 37
generally accepted, as seen with pCR in neoadjuvant treatment of breast cancer. The choice of 38
the study endpoint should always be scientifically justified. A more sensitive clinical endpoint 39
could be ones used as secondary endpoints for the innovator products, primary or secondary 40
endpoints for the innovator products at different time points of analysis, and/or new surrogates. 41
WHO/mAb_DRAFT/1 Mar 2016
Page 19
For example, overall response rate (ORR) or complete response (CR) rate as endpoints for 1
clinical efficacy studies of mAb SBPs in oncology trials are recommended, as these endpoints 2
may be sensitive and are not time related. However, if progression-free survival (which is one 3
endpoints recommended for clinical efficacy testing for innovator products) is more sensitive 4
than ORR then this may be the preferred option. Likewise, in rheumatoid arthritis (RA), both 5
continuous outcomes, e.g. changes in disease activity score of 28 joints (DAS28) over time, and 6
dichotomous outcomes, e.g. American College of Rheumatology score 20 (ACR20), are 7
considered acceptable in rheumatoid arthritis (RA) for determining clinical comparability (15). 8
9
When the primary efficacy endpoints that were used for the RBP cannot be used for SBP, it is 10
advisable to include some common endpoints as secondary endpoints to facilitate comparisons 11
between the mAb SBP and the mAb RBP. The role of these secondary endpoints in the overall 12
interpretation of the study results should be clearly defined, particularly whether the secondary 13
endpoints are used to support or to confirm equivalency or similarity. 14
15
National Regulatory Authorities (NRAs) may not always agree upon the choice of study 16
endpoints. For a biosimilar manufacturer with a global development program that is guided or 17
required by various NRAs to fulfil local regulatory or clinical practice requirements, it may be 18
possible to pre-specify different primary study endpoints with the statistical power in the same 19
trial to comply with various regulatory requirements. 20
21
6.3.4 Safety 22
6.3.4.1 General consideration 23
Comparative safety data should normally be collected pre-authorization, their amount depending 24
on the type and severity of safety issues known for the reference product. Study population 25
should inform safety events of interest and evaluate for overt new mAb SBP-specific safety 26
findings. Care should be given to compare the nature, severity and frequency of the adverse 27
reactions between the mAb SBP and the reference product in clinical trials that enrolled a 28
sufficient number of patients treated for an acceptable period of time. Clinical safety issues 29
should be captured throughout the clinical development during initial PK and/or PD evaluations 30
and also in the primary clinical study establishing comparability. It is also useful to compare the 31
safety profile of the mAb SBP with that of the reference product throughout the life-cycle of the 32
reference mAb to ensure that there is no new safety signal elicited by the mAb SBP. 33
34
6.3.4.2 Immunogenicity 35
Recombinant DNA-derived biotherapeutics, including monoclonal antibodies, are recognized by 36
the human immune system. This immune recognition may lead to an immune response that may 37
have an impact on efficacy and safety. From the regulatory point of view, animal data are usually 38
not sufficiently predictive of the human immune response. Thus, immunogenicity needs to be 39
WHO/mAb_DRAFT/1 Mar 2016
Page 20
investigated as part of the clinical trial program of a monoclonal antibody SBP. The analysis of 1
the immunogenicity of DNA-derived biotherapeutics is outlined in the WHO Guidelines on the 2
quality, safety, and efficacy of biotherapeutic protein products prepared by recombinant DNA 3
technology and Guidelines on evaluation of similar biotherapeutic products (SBPs). This general 4
guidance should be taken into account in reviewing biosimilar mAbs. In addition, details 5
regarding the advantages and disadvantages of particular assays as well as some considerations 6
for interpretation of the results and decision making process are provided in several review 7
articles (16,17,18,19). 8
9
Immunogenicity of a SBP monoclonal antibody 10
The basic data package contains the incidence, titer, neutralization capacity, and persistence of 11
product antibodies determined by appropriate assays, as well as their pharmacokinetic and 12
clinical correlations. The immunogenicity program needs to be tailored for each product. Thus, 13
the manufacturer should present a summary of the immunogenicity program that justifies the 14
selected approach. This summary should address the following topics as appropriate: 15
16
Risk assessment 17
previous knowledge of the immunogenicity of the reference product, such as the presence of 18
immunogenic structures in the active substance as well as the incidence, type, persistence and 19
clinical correlations of the antibodies; 20
findings of the physicochemical and structural comparisons between the similar mAb and its 21
reference product, including process-related impurities and aggregates; 22
differences in formulation and packaging, e.g. potential impurities and leachables; 23
route and/or the mode of administration of the product; 24
patient related and disease-related factors. Such as state of the immune system, concomitant 25
immune-modulatory therapy, and potential pre-existing immunity. 26
27
The risk-based immunogenicity program 28
The manufacturer should present a risk-based immunogenicity assessment program: 29
The basis of the immunogenicity assessment is the testing of patient samples pre-treatment, 30
during treatment and post-treatment in an appropriate set of assays. The measurement of 31
antibodies to monoclonal antibodies is methodologically challenging and as standard assay 32
formats involving anti-immunoglobulin reagents are inappropriate for this product class, 33
alternative methods should be used. Like other biotherapeutic products, a multi-tiered 34
approach for assessment is needed. The developer has to validate assays for screening, 35
confirmation, and neutralization capacity. Special attention should be paid to the choice of the 36
control sera, determination of cut points and to the interference caused by the presence of 37
matrix components and of the residual drug in the sample. To mitigate the potential 38
interference, corrective measures should be implemented. For example, drug interference may 39
be overcome by allowing time for clearance of the drug from the circulation prior to sampling 40
WHO/mAb_DRAFT/1 Mar 2016
Page 21
or by dissociating immune complexes or by removal of the drug. Inclusion of any of these 1
measures must not compromise the detection of antibodies or patient treatment. 2
Integration of the product antibody testing in the comparative clinical trials: sampling schedule 3
and duration of the follow up. It is particularly important to synchronise the sampling for 4
product antibody determination and pharmacokinetic measurements as well for assessments of 5
safety and efficacy. 6
Special emphasis should be put on the potential association of product antibodies with loss of 7
efficacy, infusion reactions as well as to acute and delayed hypersensitivity. The manufacturer 8
should systematically use terminology and definitions to characterize potentially immune-9
mediated symptoms according to relevant publications (20). 10
The manufacturer should take the dose, dosing schedule, intermittent of dosing into account. 11
The vulnerability of the patient population(s) and the expected risks of immunogenicity should 12
be taken into account in planning for the intensity of monitoring. 13
The manufacturer should give a description and analysis of the use of pre-medication or de-14
immunization measures to mitigate acute infusion/injection-related reactions and other, 15
possibly immune-mediated reactions. 16
Product antibodies should be assayed post-treatment not earlier than 12 weeks in order to 17
demonstrate the reduction of antibody titres or appearance of circulating antibodies possibly 18
suppressed by the product. 19
20
Comparative immunogenicity 21
The lack of standardization and rapid evolution of the assay methodology makes it difficult to 22
compare immunogenicity studies. Therefore, pre-licensing comparative immunogenicity studies 23
are always needed in the development of SPBs (1,12). Immunogenicity testing of the SBP and 24
the reference product should be conducted within the biosimilar comparability exercise by using 25
the same assay format and sampling schedule. Parallel group design is recommended because of 26
the long half-life of antibodies and because it may be difficult to interpret immunogenicity after a 27
switch. 28
29
Assays and mAb characterization 30
Antibody assays should ideally be capable of detecting all antibodies against both the biosimilar 31
and the reference molecule. Thus, assays can be performed with both the reference and biosimilar 32
molecule as the antigen/capture agent in parallel in order to measure the immune response against 33
the product that was received by each patient. The challenge of this assay is to develop two 34
assays with similar sensitivity and specificity. Cross-testing all serum samples by both tests is 35
useful to confirm the similar assay performance as well as similar antigen epitopes. The use of a 36
single assay with the active substance of the biosimilar as the antigen/capturing agent for 37
evaluation of all samples (including those from reference product treated patients) will be able, in 38
WHO/mAb_DRAFT/1 Mar 2016
Page 22
principle, to detect all antibodies developed against the biosimilar molecule, i.e. it is, in principle, 1
conservative. However, this assumption needs to be justified by the manufacturer. 2
3
Following identification of confirmed antibody positive samples, characterization of the 4
antibodies is required. Determination of their neutralizing potential is essential and deviation 5
from this requires justification. Although a cell-based bioassay or a non-cell-based competitive 6
ligand binding (CLB) assay can be used, the latter should only be used if relevant to the 7
mechanism of action (MoA) of the product. For example, a CLB assay is appropriate in a 8
scenario where a therapeutic mAb acts by binding to a soluble ligand thereby blocking it from 9
interacting with its receptor thus inhibiting the biological action of the ligand. Since the assay 10
procedure measures binding to the target and inhibition of the binding activity if neutralizing 11
antibodies are present, it is reflective of the therapeutic's MoA. For intact mAbs where effector 12
functions are likely to contribute to the clinical effect, cell-based bioassays are recommended as 13
the mechanism of action cannot be adequately reflected in a non-cell-based CLB assay. 14
15
Cell-based assays are preferred for the assay of neutralizing mAb since they represent the more 16
physiological situation. However, bioassays may not be sufficiently sensitive or measure 17
neutralization effectively. In such cases, alternative functional methods including non-cell-based 18
CLB assays could be explored to assess neutralizing antibodies. It such cases, it is recommended 19
to seek advice from regulators. 20
21
Additional studies beyond the standard data package, such as immunoglobulin class and IgG 22
subclass, will be necessary in special situations (e.g. occurrence of anaphylaxis). It may also be 23
necessary to locate the antigenic sites (e.g. antigen-binding region vs constant region of antibody 24
molecule). Banking of patient samples is necessary in order to have the possibility for re-testing 25
in case of technical problems in the original assay. 26
27
Clinical immunogenicity study 28
The selected patient population should be sensitive for differences in immunogenicity. It is also 29
important that the pivotal immunogenicity study includes PK measurements. Ideally, the repeat 30
dose clinical trials should include sampling for immunogenicity and PK (trough levels) as well as 31
simultaneous efficacy and safety assessments to establish the clinical impact of immunogenicity. 32
If the study will include patients previously treated with the reference mAb a subgroup analysis 33
of previously treated patients should be performed. 34
35
The sampling schedule should be optimized for the demonstration of similar onset and 36
persistency of products antibodies to the test and the reference product. 37
38
The duration of the follow up of immunogenicity should be sufficient to demonstrate similar 39
persistence and clinical impact of the product antibodies. In chronic administration, the minimum 40
follow up is six months. 41
WHO/mAb_DRAFT/1 Mar 2016
Page 23
Immunogenicity should be followed after licensing by monitoring possible immune-mediated 1
adverse effect. Special immunogenicity studies may be necessary in high risk situations, e.g. 2
when the reference product is known to have serious but rare immune-mediated effects, such as 3
anaphylaxis. 4
5
Evaluation of immunogenicity includes the incidence, titer, neutralization capacity and 6
persistency as well as correlations to exposure, safety and efficacy. For time being, there is no 7
generally accepted statistical methodology that could be used to define the limits of comparable 8
immunogenicity. In general, excess immunogenicity of a SBP is not compatible with 9
comparability unless the sponsor can convincingly show that the product antibodies have no 10
clinical relevance and that the underlying difference between the SBP and the reference product 11
does not signal an otherwise important problem. 12
13
Less immunogenicity may be acceptable if the manufacturer can justify that the clinical data of 14
the reference product is still representative for the SBP. A subgroup analysis of antibody-negative 15
patients should be performed to exclude the impact of immunogenicity. 16
17
6.4 Indication extrapolation 18
Indication extrapolation is the leveraging of efficacy and safety data from clinical studies in one 19
indication to support the authorization of other indications in which the biosimilar has not been 20
studied, but for which the reference product is authorized and well characterized. As part of 21
guidance for biosimilar development, the WHO guidance document has provided 22
recommendations regarding extrapolation of clinical data across indications. The principles 23
recommended for indication extrapolation for biosimilars stated in the WHO guidance apply to 24
mAb SBPs. 25
26
In addition to these principles, some special attentions should be paid to mAbs. Monoclonal 27
antibodies have both Fab and Fc effector functions and may exert their clinical effect through a 28
variety of mechanisms. These include, for example, ligand blockade, receptor blockade, receptor 29
down-regulation, cell depletion (via ADCC, CDC, apoptosis), and signaling induction. A 30
particular mAb may act through one or a combination of these or other mechanisms. Where a 31
therapeutic mAb is indicated for a variety of diseases, various mechanisms of action may be 32
important depending on the indication in question. In order to support extrapolation, the 33
mechanisms that contribute to the efficacy of the mAb in each indication should ideally be well 34
understood and clearly defined. It may not be possible to extrapolate when mechanistic 35
differences are observed between the biosimilar and the reference product during the 36
comparability exercise (e.g. in in vitro or ex vivo biological assays). This is especially true if the 37
affected mechanism is not considered active in the studied indication since there would be no 38
clinical data available to assess the impact of the difference on the indication to be authorized via 39
WHO/mAb_DRAFT/1 Mar 2016
Page 24
extrapolation. Due to their ability to exert effects through various mechanisms of action, 1
extrapolation can be difficult when a mAb is indicated for a variety of diseases in which the 2
important mechanisms of action are not well understood for each indication (21). Disease 3
pathophysiology is another important determinant in the extrapolation assessment. Some mAbs 4
hold indications for the treatment of diseases that bear little resemblance to each other. For 5
example, if a reference mAb is authorized both as an anti-inflammatory agent and as an 6
anticancer antibody, the scientific justification as regards extrapolation between the two (or more) 7
indications is more challenging. Since the pharmacokinetic characteristics of a mAb (e.g. linear 8
vs. non-linear PK) may differ between various diseases, additional clinical data may be required 9
to support extrapolation from one to the other. Also, posology may differ between indications 10
(e.g. body surface area adjusted dosing vs. fixed dosing). In situations, such as the one described, 11
manufacturers should consider performing clinical trials in the two unrelated settings if they seek 12
to have a mAb SBP authorized for all indications held by the reference mAb. 13
14
The possibility of extrapolating safety including immunogenicity data requires careful 15
consideration, and may have to involve more specific studies including PK/PD or 16
immunogenicity studies. Immunogenicity is related to multiple factors including the route of 17
administration (e.g. SC vs IV), treatment regimen (e.g. continuous vs intermittent), and patient-, 18
disease-, and treatment-related factors (e.g. immune status). Therefore, extrapolation of 19
immunogenicity data is not self-evident and always requires convincing justification (22). 20
21
6.5 Pharmacovigilance and post approval consideration 22
Pharmacovigilance measures should be in place once a mAb SBP is approved in order to ensure 23
its long-term safety and efficacy. The general requirements for pharmacovigilance are the same 24
as for any approved new drug. As described in WHO documents, including recording of product 25
name, batch number, manufacturers’ name, and INN where exits are further detailed in the WHO 26
guidelines (1,2). Clinically important adverse events-induced by mAb SBPs occur at a relatively 27
low frequency and the probability of them occurring during the time frame of the clinical trial is 28
also low. Additionally, due to relatively small sample size, mAb SBP clinical trials may only 29
have the statistical power to detect common adverse events. Thus, a targeted pharmacovigilance 30
is essential for the identification and assessment of potential post-market risk for mAb SBPs. 31
32
Authors and acknowledgements 33
The first draft of these guidelines was prepared by the following authors for the parts indicated: 34
Dr K. Gao, World Health Organization (WHO), Geneva, Switzerland; Dr E. Griffiths, consultant, 35
United Kingdom; Dr H.K. Heim, Federal Institute for Drugs and Medical Devices (BfArM), 36
Germany; Dr I. Knezevic, World Health Organization (WHO), Geneva, Switzerland; Dr P. Kurki, 37
Finnish Medicines Agency, Finland; Dr E. Niklas, Finnish Medicines Agency, Finland; Dr C. 38
Njue, Health Canada, Canada; Dr R. Thorpe, consultant, United Kingdom; Dr M. Wadhwa, 39
National Institute for Biological Standards and Control (NIBSC), United Kingdom; Dr J. Wang, 40
Health Canada, Canada; Dr E. Wolff-Holz, Paul Ehrlich Institute, Germany (Clinical).The 41
WHO/mAb_DRAFT/1 Mar 2016
Page 25
drafting group took into consideration and proposals from the WHO informal consultation on 1
monoclonal antibodies as similar biotherapeutic products (SBPs) held in Geneva in April 2015 2
with following participants: 3
4
Temporary Advisers: Dr M.C. Bielsky, Medicines and Healthcare products Regulatory 5
Agency(MHRA), United Kingdom; Dr E. Griffiths, Consultant, United Kingdom; Dr H.K. Heim, 6
Federal Institute for Drugs and Medical Devices (BfArM), Germany; Dr J. Joung, Ministry of 7
Food and Drug Safety, Republic of Korea; Dr C. Njue, Health Canada, Canada; Dr R. Thorpe, 8
Consultant, United Kingdom; Dr M. Wadhwa, National Institute for Biological Standards and 9
Control(NIBSC), United Kingdom; Dr J. Wang, Health Canada, Canada; Dr J. Wang, National 10
Institutes for Food and Drug Control (NIFDC), People's Republic of China; Dr T. Yamaguchi, 11
Pharmaceuticals and Medical Devices Agency (PMDA), Japan. 12
13
Representatives of National Regulatory Authorities: Mrs A. Abas, Ministry of Health Malaysia, 14
Malaysia; Dr A. Abdelaziz, Jordan Food and Drug Administration (JFDA), Jordan; Dr 15
K.Bangarurajan, Central Drug Standards Control Organization (CDSCO), India; Dr B. 16
Boonyapiwat, Ministry of Public Health, Thailand; Ms J. Dahlan, National Agency of Drug and 17
Food Control, Indonesia; Mrs D. Darko, Food and Drug Authority, Ghana; Mr D. Goryachev, 18
Ministry of Health, Russia; Dr C. Ilonze, National Agency For Food and Drug Administration 19
and Control (NAFDAC), Nigeria; Mrs W. Jariyapan, Ministry of Public Health, Thailand; Dr D. 20
Khokal, Health Sciences Authority, Singapore; Dr P. Kurki (on behalf of EMA BMWP), Finnish 21
Medicines Agency, Finland; Ms I. Lyadova, Ministry of Health, Russia; Mrs S. Marlina; National 22
Agency of Drug and Food Control, Indonesia; Dr J. Meriakol; Ministry of Health, Kenya; Mr K. 23
Nam, Ministry of Food and Drug Safety, Republic of Korea; Mrs Y.H. Nunez, Centro para el 24
Control Estatal de la Calidad de los Medicamentos (CECMED), Cuba; Dr C. Gonzalez, Ministry 25
of Health, Colombia; Mrs B. Valente, National Health Surveillance Agency (ANVISA), Brazil; 26
Ms N. Vergel, National Institute of Drugs and Food, Colombia; Dr L. Wang, National Institutes 27
for Food and Drug Control (NIFDC), People's Republic of China; Dr E. Wolff-Holz, Paul Ehrlich 28
Institute, Germany; Dr S. Xie, China Food and Drug Administration, People's Republic of China. 29
Representatives of the Latin American Association of Pharmaceutical Industries (ALIFAR): Dr E. 30
Spitzer, Argentina. 31
32
Representatives of the Developing Countries Vaccine Manufacturers’ Network (DVCMN): Dr A. 33
Qu, Shanghai Institute of Biological Products Company Limited, People's Republic of China; Ms 34
L. S, United States of Americanti, Bio Farma, Indonesia; Dr A. Zhang, China National Biotec 35
Group Company Limited, People's Republic of China. 36
37
Representatives of the Generic Pharmaceutical Alliance (IGPA)/ European Generic Medicines 38
Association (EGA): Dr Stan Bukofzer, Hospira, IL, United States of America; Mrs Suzette Kox, 39
WHO/mAb_DRAFT/1 Mar 2016
Page 26
European Generic Medicines Association, Brussels, Belgium; Dr Martin Schiestl, Sandoz GmbH 1
Biochemiestrasse, Kundl, Austria; Dr Christopher Webster, Baxter, Cambridge, MA, United 2
States of America. 3
4
Representatives of the International Federation of Pharmaceutical Manufacturers and 5
Associations (IFPMA): Mrs J. Bernat, International Federation of Pharmaceutical Manufacturers 6
& Associations, Geneva, Switzerland; Dr K. Heidenreich, Novartis, Basel, Switzerland; Dr J. 7
Macdonald, Pfizer, Tadworth, United Kingdom; Dr S. Ramanan, Amgen, CA, United States of 8
America; Dr T. Schreitmueller, F. Hoffmann – La Roche Ltd., Basel, Switzerland; Dr J. Siegel, 9
Johnson & Johnson, Washington DC, United States of America; Dr K. Watson, AbbVie Ltd, 10
Maidenhead, United Kingdom. 11
12
Representatives of the Individual Manufacturers: Mr R.Ivanov, Biocad, Saint-Petesburg, Russia; 13
Dr A. Kudrin, Celltrion, Incheon, Republic of Korea; Mr B. Kim, Celltrion, Incheon, Republic of 14
Korea. 15
16
Representatives of WHO Regional Office: Ms H.B. Pedersen, EU/HTP Health Technologies and 17
Pharmaceuticals (EU/RGO/DSP/HTP); Dr M.L. Pombo, Pan American Health Organization 18
(PAHO), Washington DC, United States of America. 19
20
WHO Secretariat: Dr K. Gao, Dr H. Kang, Dr I. Knezevic and Dr David Wood, Technologies 21
Standards and Norms (TSN) Team, Essential Medicines and Health Products (EMP) Department, 22
Health Systems and Innovation (HIS) Cluster, World Health Organization (WHO), Geneva, 23
Switzerland. 24
25
References 26
1. Guidelines on evaluation of similar biotherapeutic products (SBPs). In: WHO Expert 27
Committee on Biological Standardization: sixtieth report. Geneva: World Health Organization; 28
2013(WHO Technical Report Series, No. 977), Annex 2. 29
30
2. Guidelines on the quality, safety and efficacy of biotherapeutic protein products prepared by 31
recombinant DNA technology. In: WHO Expert Committee on Biological Standardization: 32
sixty-fourth report. Geneva: World Health Organization; 2014 (WHO Technical Report Series, 33
No. 987), Annex 4. 34
35
3. Thorpe R, et al. Intended use of reference products and WHO International 36
Standards/Reference Reagents in the development of similar biological products (biosimilars). 37
Biologicals, 2011, 39(5): 262-265. 38
39
WHO/mAb_DRAFT/1 Mar 2016
Page 27
4. Preclinical safety evaluation of biotechnology-derived pharmaceuticals. ICH Guideline S6 1
(R1). Geneva, International Conference on Harmonisation of Technical Requirements for 2
Registration of Pharmaceuticals for Human Use, 2011. 3
4
5. Wang W, et al. Monoclonal antibodies with identical fc sequences can bind to FcRn 5
differentially with pharmacokinetic consequences. Drug Metab Dispos. 2011; 39(9):1469–6
1477. 7
8
6. Committee for Medicinal Products for Human Use. Guideline on similar biological medicinal 9
products containing monoclonal antibodies – nonclinical and clinical issues. London, 10
European Medicines Agency, 2012(EMA/CHMP/BMWP/403543/2010). 11
12
7. Guidance for industry. Clinical pharmacology data to support a demonstration of 13
biosimilarity to a reference product. Rockville, MD, Food and Drug Administration, 2014. 14
15
8. Pippig, S. D, et al. Biosimilar Monoclonal Antibodies. In: Stefan Dübel, Janice M. Reichert, 16
editors. Handbook of therapeutic antibodies. Wiley-VCH Verlag GmbH and Co. KGaA, 17
Weinheim, Germany. doi: 10.1002/9783527682423.ch24. 2014:681-704. 18
19
9. Marini JC, et al. Systemic Verification of Bioanalytical Similarity between a biosimilar and a 20
reference biotherapuetic: Committee recommendations for the development of a single ligand 21
binding assay to support PK assessments. The AAPS Journal, 2014, 16(6):1149-1158. 22
23
10. Guidance for sponsors: Information and submission requirements for subsequent entry 24
biologics (SEBs). Ottawa, Health Canada, 2010. 25
26
11. Guidance for Industry: Scientific Considerations in Demonstrating Biosimilarity to a 27
Reference Product. Silver Spring, MD, Food and Drug Administration, 2015. 28
29
12. Committee for Medicinal Products for Human Use. Guideline on similar biological medicinal 30
products containing biotechnology-derived proteins as active substance: Nonclinical and 31
clinical issues EMA nonclinical and clinical. London, European Medicines Agency, 2015. 32
33
13. Choice of control group and related issues in clinical trials. ICH Guideline E10. Geneva, 34
International Conference on Harmonisation of Technical Requirements for Registration of 35
Pharmaceuticals for Human Use, 2000. 36
37
14. Shein-Chung C, Jen Pei L, authors.Design and Analysis of Clinical Trials: Concepts and 38
Methodologies. John Wiley & Sons, New Jersey; 2004. 39
WHO/mAb_DRAFT/1 Mar 2016
Page 28
15. Kudrin A, et al. Case studies on clinical evaluation of biosimilar monoclonal antibody: 1
Scientific considerations for regulatory approval. Biologicals, 2015, 43(1):1 -10. 2
3
16. Wadhwa M, et al. Immunogenicity assessment of biotherapeutic products: An overview of 4
assays and their utility. Biologicals, 2015, 43(5):298-306. 5
6
17. Knezevic I, et al. Immunogenicity assessment of monoclonal antibody products: A simulated 7
case study correlating antibody induction with clinical outcomes. Biologicals, 2015 8
43(5):307-317. 9
10
18. Gupta S, et al. Recommendations for the validation of cell-based assays used for the detection 11
of neutralizing antibody immune responses elicited against biological therapeutics. J Pharm 12
Biomed Anal, 2011, 55(5):878-888. 13
14
19. Shankar G, et al .Recommendations for the validation of immunoassays used for detection of 15
host antibodies against biotechnology products. J Pharm Biomed Anal, 2008, 48(5):1267-16
1281. 17
18
20. Kang SP, et al. Infusion-Related and Hypersensitivity Reactions of Monoclonal Antibodies 19
Used to Treat Colorectal Cancer—Identification, Prevention, and Management. Journal of 20
Supportive Oncology, 2007, 5(9):451–457. 21
22
21. Scott BJ, et al. Biosimilar Monoclonal Antibodies: A Canadian Regulatory Perspective on the 23
Assessment of Clinically Relevant Differences and Indication Extrapolation. J Clin Phamacol, 24
2015, 55(S3) S123 –S132. 25
26
22. Weise M, et al. Biosimilars: the science of extrapolation. Blood, 2014,124 (22):3191-3196. 27
Top Related