7 Westferry Circus ● Canary Wharf ● London E14 4HB ● United Kingdom
Telephone +44 (0 20 7418 8400 Facsimile +44 (0)20 7418 8416
E-mail info@ema.europa.eu Websi e ww .ema.europa.eu An agency of the European Union
© European Medicines Agency, 2012. Reproduction is authorised provided the source is acknowledged.
24 May 2012
EMA/CHMP/BMWP/86289/2010
Committee for Medicinal Products for Human Use (CHMP)
Guideline on immunogenicity assessment of monoclonal
antibodies intended for in vivo clinical use.
Draft agreed by Similar Biological Medicinal Products Working Party October 2010
Adoption by CHMP for release for consultation November 2010
End of consultation (deadline for comments) May 2011
Final agreed by BMWP March 2012
Adoption by CHMP 24 May 2012
Date for coming into effect 1 December 2012
Disclaimer: This guideline is intended as an addendum to Guideline on Immunogenicity assessment of
biotechnology-derived therapeutic proteins EMEA/CHMP/BMWP/14327/2006 and should be read in
conjunction.
Keywords Immunogenicity, monoclonal antibodies, similar biological medicinal
products, clinical use, assay strategy.
2/10
Guideline on immunogenicity assessment of monoclonal
antibodies intended for in vivo clinical use.
Table of contents
Executive summary ..................................................................................... 3
1. Introduction (background)...................................................................... 3
2. Scope....................................................................................................... 3
3. Legal basis .............................................................................................. 3
4. Problems experienced with screening and confirmatory assays used in
assessing immunogenicity of mAbs............................................................. 4
4.1. Assays for antibody detection ................................................................................4
4.2. Presence of mAb product in samples for analysis......................................................5
4.3. Confirmatory Assays.............................................................................................5
4.4. Controls ..............................................................................................................5
5. Assessing the neutralising capacity of antibodies induced against mAbs 6
6. Immunogenicity risk management of mAbs ............................................ 6
6.1. Risk identification .................................................................................................6
6.2. Risk assessment...................................................................................................8
6.3. Risk monitoring and mitigation...............................................................................9
3/10
Executive summary
This guideline addresses issues relating to the unwanted immunogenicity of monoclonal antibodies
(mAbs) intended for clinical use. These include factors impacting on immunogenicity of mAbs, the
clinical consequences of immunogenicity, assay related problems, assessing neutralizing antibodies
induced by mAbs and consideration of a risk-based approach for the evaluation of immunogenicity of
mAbs.
1. Introduction (background)
Immunogenicity can be a significant problem in the treatment of patients with therapeutic biologicals
and this is addressed in the ‘Guideline on Immunogenicity Assessment of Biotechnology-Derived
Therapeutic Proteins’ by the CHMP (adopted April 2008, referred to henceforth as ‘the general
guideline’), which in principle is applicable to monoclonal antibodies (mAbs). While many aspects of
immunogenicity of mAbs are not different from those for other therapeutic proteins, there are several
aspects that require more specific considerations. MAbs are not expected to induce antibodies that
cross-react and neutralize an endogenous counterpart (as can occur with EPO) and are not used as
replacement therapies. Often mAbs are used as therapeutic or diagnostic agents where alternative
treatments or diagnostics may exist. However, some specific aspects of immunogenicity are exclusively
or primarily relevant for mAbs or novel mAb derivatives (e.g. Fab fragments, scFv, nanobodies,
minibodies) and these are addressed in this guideline.
MAbs comprise a large important class of therapeutic biologicals. The range of clinical indications with
potential for treatment with mAbs is very wide. Many mAb products are known to be associated with
unwanted immunogenicity and in some cases the immunogenicity causes impaired clinical responses or
rare serious adverse reactions which require clinical intervention. The wide range of mAbs in
development, and approved for different clinical indications precludes specific guidelines that are
pertinent to all situations.
2. Scope
The general principles adopted and explained in this document apply to the development and
systematic evaluation of an unwanted immune response against a therapeutic or in vivo diagnostic
mAb in recipients. The guideline applies to mAbs, their derivatives, and products of which they are
components, e.g., conjugates, Fc linked fusion proteins.
This guideline considers the major quality and clinical aspects that are important for adequately
addressing the problems with detection of and risk related to the development of an unwanted immune
response to the particular mAb in the particular clinical indication sought.
This guideline is aimed at products at final development stage (e.g. marketing authorization application
stage). However, many of the principles are relevant to earlier phases of development.
3. Legal basis
Directive 2001/83/EC, as amended in particular in Directive 2001/83/EC Art 10(4) and Part II of the
Annex I of Directive 2001/83/EC, as amended. This guideline should be read in conjunction with other
relevant guidelines, e.g.:
 Guideline on similar biological medicinal products (CHMP/437/04).
4/10
 Guideline on similar biological medicinal products containing biotechnology-derived proteins as
active substance: non-clinical and clinical issues (EMEA/CPMP/42832/2005).
 Guideline on production and quality control of monoclonal antibodies and related substances
(EMEA/CHMP/BWP/157653/2007).
 ICH guideline S 6 (R1) Preclinical safety evaluation of biotechnology-derived pharmaceuticals
(EMA/CHMP/ICH/731268/1998).
 Guideline on Immunogenicity Assessment of Biotechnology-derived Therapeutic Proteins
(EMEA/CHMP/BMWP/14327/2006).
 European Pharmacopeia monograph on monoclonal antibodies.
 Guidelines on comparability of biotechnology-derived medicinal products after a change in the
manufacturing process: non-clinical and clinical issues (EMEA/CHMP/BMWP/101695/2006).
 Guideline on bioanalytical method validation (EMEA/CHMP/EWP/192217/2009).
 Eudralex Volume 9A of the Rules governing Medicinal Products in the European Union (Part I,
Chapter 3: Requirements for Risk Management Systems) to be replaced by Good Vigilance
Practice (Volume on Risk Management Systems) after July 2012.
4. Problems experienced with screening and confirmatory
assays used in assessing immunogenicity of mAbs
4.1. Assays for antibody detection
In principle, any immunoassay format can be used to measure antibodies against mAbs. However,
assays used to detect antibodies against mAbs are often more problematic, difficult and can be
technically challenging. Many standard assay formats involve the use of anti-immunoglobulin reagents
such as antibodies against immunoglobulins, protein A or protein G, but these are inappropriate for use
in detecting antibodies against mAbs as they very often bind to the product itself. Thus, for example
simple ELISAs and radio-immunoprecipitation assays are not usually suitable for use with mAbs unless
they are adapted to overcome this problem. Therefore, different assay approaches have to be
developed for mAbs. A common approach is to use the ‘bridging’ format e.g. for ELISAs or
electrochemiluminescence (ECL) assays which does not require anti-immunoglobulin reagents and so
can be directly applied to studies with mAbs. In some cases, this procedure may be less sensitive than
other immunoassay methods and may require significant development effort to produce a suitable
assay. It also will not efficiently detect the IgG4 antibodies which can be produced in some cases.
Another approach is to use a Surface Plasmon Resonance (SPR) procedure. This does not require antiimmunoglobulin
reagents for detecting antibodies against mAbs. It is a real-time procedure and is
therefore fast and also detects rapidly dissociating antibodies which can be missed by other methods.
However, as SPR simply detects protein binding to the coated chip it needs to be confirmed that the
signal is caused by antibodies. It may be less sensitive than other methods for detecting high affinity
antibodies and, in the absence of automated sampling systems may have a low throughput.
Samples (normally serum or plasma) may contain substances that interfere with the assays i.e. matrix
effects which produce false positive or negative results and/or incorrect assessment of antibody
content. Well known examples of such interfering substances are complement components, mannose
binding protein, Fc receptors, soluble target molecules, complement receptor 1 and rheumatoid factors,
but other substances including the product itself can also cause problems. Assays often need to be
‘tailored’ to reduce artefacts and achieve acceptable background signal levels, sensitivity and specificity.
Applicants need to justify the suitability of the chosen approach, taking into full consideration the
limitations of the respective methods.
5/10
4.2. Presence of mAb product in samples for analysis
Intact mAb products have relatively long half lives and persist in circulation for long periods. Even
fragments can persist in blood for several days. This can cause significant problems in detection of
antibody responses due to the presence of mAb product in samples collected for antibody assessment.
This normally results in an artefactually low estimate of antibody content of affected samples and can
be so pronounced as to cause false negative results. Several approaches have been proposed to
overcome this problem. One possibility is to delay sampling until levels of mAb product have declined
sufficiently to no longer cause problems. This has been claimed to resolve the problem with some mAb
products, but requires careful assessment as it has the potential to fail to detect immunogenicity, as
induced antibodies may have declined to undetectable levels by the time the samples are taken.
Another approach is to use methodology which is least affected by the problem. Some ECL based
immunoassays seem much less affected by residual product in samples than other methods, including
conventional bridging ELISAs. A commonly described procedure for dealing with the problem is to
include a preliminary antigen-antibody dissociation step in the assay design so that any complexes
present are disrupted before antibody is detected. Various versions of assays which include acid
incubations, sometimes coupled with affinity separation of product have been described for this but
need to be carefully evaluated to show that the additional steps do not invalidate the assay. A final
possibility is to dilute samples so that residual product present is insufficient to interfere with the
assay. This approach needs care as it may result in a false negative assessment of immunogenicity if
the assay is not sufficiently sensitive to detect antibodies in the diluted samples. In some cases it may
be necessary to assay samples for the amount of residual mAb. In many cases, anti-mAb method
development, validation and testing utilizes a combination of all three approaches to reduce product
interference.
4.3. Confirmatory Assays
Confirmatory assays can suffer from the same problems as screening assays. It is important to select
an appropriate confirmatory assay taking account of the characteristics of the screening assay. Use of
Protein A and Protein G may be appropriate in confirmatory assays to demonstrate that the positive
response is due to an immunoglobulin; however other approaches can be used for this.
4.4. Controls
Generation of positive control sera is in general a critical issue for immunogenicity studies for mAbs.
The chosen positive control serum or purified antibody is important for monitoring assay sensitivity and
specificity. If human sera are not available (as is possible during early phases of product development)
then use of animal sera is the only option. Choice of species for this has important consequences. Nonhuman
primates produce significant anti-CDR and anti-framework responses against human or
humanized mAbs, which may closely mimic human responses and may be an appropriate positive
control. However, non-primate species usually produce antibodies primarily against the constant
regions of the mAb, which is unlike human responses. Use of an anti-idiotypic antiserum or mAb can,
in some cases, provide a useful positive control. Selection of appropriate negative controls is
important. For confirmatory assays, spiking samples with an irrelevant mAb can be used to confirm
specificity.
6/10
5. Assessing the neutralising capacity of antibodies induced
against mAbs
MAbs exert their action by various mechanisms ranging from simple binding to antigen, which alone
mediates the clinical effect, to binding antigen and mediating one or more immunobiological
mechanisms which combine to produce the overall clinical response. Therefore, although simple
binding may seem to be the only mechanism operating to achieve clinical efficacy, other effects may
also play a role in this. In some cases multiple functions of the mAb may be involved in an additive or
synergistic manner to produce an overall combined clinical effect and this may be hard to dissect
experimentally to allow a clear understanding of how the mAb mediates its clinical potency. Therefore,
if intact mAbs are used, care must be taken not to assume that the Fc mediated immunobiological
effects of the product are not involved in clinical efficacy, even when simple antigen binding is
considered to be the primary mode of action. In this regard, use of a cell based assay for measuring
neutralization has an advantage. In such cases a thorough biological characterization of the mAb must
be undertaken, using appropriate biological and immunological assays. Following this, the properties of
the mAb need to be assessed to allow selection of an appropriate neutralizing assay strategy.
Antibodies which neutralize the biological activity of biological products may diminish clinical efficacy of
the product. It is normally expected that the neutralizing capacity of any antibodies induced will be
measured. Any deviations from this need to be justified. For most biological products, the most
appropriate neutralizing antibody assay is a bioassay which measures the neutralization of the
bioactivity of the product by antibodies. However, the nature of the clinical mode of action of mAbs
implies that induced antibodies which block mAb binding to target are those which are mostly
associated with reduced clinical efficacy. Therefore, competitive ligand binding assays may be the
neutralizing assays of choice for mAbs rather than classic bioassays. This distinguishes mAbs from
other classes of biologicals with regard to immunogenicity assessment.
6. Immunogenicity risk management of mAbs
6.1. Risk identification
The immunogenicity of mAbs is complex and there are a number of often poorly understood factors
which makes it difficult to predict with any certainty whether a therapeutic or diagnostic monoclonal
antibody is likely to provoke a clinically relevant immune response. In vitro non-clinical approaches
aimed at identifying T-cell epitopes have been developed but these have limited capacity to predict
immunogenicity of a therapeutic in humans. However, such procedures can be useful for selecting
candidate molecules for further development.
Standard aspects of immunogenicity as described in the general guideline should be addressed for
every new therapeutic mAb, taking into account its characteristics, the nature of the intended use and
the therapeutic indication.
Preliminary immunogenicity data from early clinical studies can provide information which may be of
use for planning later studies, e.g. exploring the performance of bioanalytical assays, detection of preexisting
antibodies or other factors that could confound recognition of treatment-emergent antibodies
against mAbs. Based on a risk identification and assessment strategy as further described below, the
standard immunogenicity testing programme may be reduced with thorough justification, or may need
to be intensified, depending on the level of risk identified. The applicant should always present a
thorough risk identification that takes full account of the nature of the product along with to its
intended use.
7/10
a) Prior knowledge
Available knowledge or lack of knowledge concerning other similar mAbs (e.g., from the same target
class, expressed in the same expression system) is an important consideration. The risk perception
may be higher if the methodology to either detect antibodies against mAbs or to detect clinical
consequences (e.g. mAb trough concentration, PD parameters and response to mAb treatment) of
antibodies against mAbs is not sufficiently sensitive. In such cases, more extensive monitoring of the
anti-mAb response dynamics relative to therapeutic outcome may have to be considered.
b) Mab structure
In principle, antibodies can be produced against various epitopes present on different parts of the mAb
molecule e.g. variable or constant regions. For heterologous e.g. rodent sequence or chimaeric mAbs,
recognition of the antibody as being foreign is the primary basis for antibody mediated immunity and
antibodies can be elicited against any part of the molecule. With humanised or human sequence mAbs
the immune response is predominantly anti-idiotypic (as the complementarity determining regions are
hypervariable in sequence ), which clearly can compromise clinical responses to the mAb. However, in
some cases, antibodies can be induced against the constant region of human or humanised mAbs and
this can affect the effector functions of the mAb with potential consequences on the clinical response.
There is less clinical experience with emerging mAb based constructs and this may add to the
perception of risk. Special consideration should be given to next generation products, for example,
bispecific mAbs or mAb fragments and their capacity to expose hidden antigenic determinants.
Altered glycosylation patterns may decrease or increase the immunogenic properties of the molecule
(e.g. change in shielding of the protein backbone). Non-typical glycosylation patterns, e.g. as
encountered when adopting entirely novel expression systems, may introduce a higher immunogenicity
risk as compared to more commonly used expression systems.
Other factors that influence immunogenicity include impurities arising from the production process and
other quality attributes. Therefore, the analytical and clinical approach to assess, characterize and
potentially mitigate such potential risks may have to be more extensive, and the risk related to the
product quality should be appropriately identified.
For example, a mAb against a target where substantial previous experience exists, but that is
produced using a novel expression system, may have a perceived low risk as regards its mechanism of
action, but a higher risk as regards potential impact of impurities where little experience exists
regarding their effects on safety.
c) Mechanism of action:
The mode of action of the mAb (e.g., cytolytic, apoptotic), and especially the nature of the target
molecule (e.g., immunosilencing, immunostimulating), needs to be adequately characterized and
comprehensively investigated. Antibody responses against mAbs that target the idiotype of a mAb
usually result in diminished efficacy. Likewise, the impact of antibodies against mAbs that recognize
allotypic or other regions should be considered carefully since immune complex formation may result in
undesirable effects in recipients.
Indirect effects of the antibodies induced by mAbs may also be important, e.g. it is possible that mAbs
that target molecules involved in signalling cascades may induce antibodies which cross-link the target
molecules in an agonistic manner, potentially leading to enhanced activation of the immune system
and possibly cytokine release syndromes. This may be difficult to predict at the individual patient level.
For agonistic mAbs or for mAbs where cross-linking could on theoretical considerations lead to
immunoactivation, applicants should consider careful observation of patients in early clinical trials to
see if such events occur.
8/10
d) Clinical factors
Immunogenicity is significantly influenced by clinical factors. Immunogenicity for mAbs can be age
related e.g. protein turnover is different in children compared to adults and this can result in
differences in observed immunogenicity, e.g. for antibodies used in treatment of juvenile arthritis
compared to rheumatoid arthritis at comparable doses. Previous exposure to similar or related
monoclonal antibodies can also influence immunogenicity. Therapeutic antibodies used with
intermittent dosing schemes have a higher likelihood of inducing immunogenicity than when used in a
scheduled and repeated dosing scheme.
Whether antibodies against a mAb have clinically significant effects depends on the binding site of the
antibody, the affinity of the antibody for the mAb and the titre of the antibodies that develop.
Antibodies against mAbs can occur transiently and then disappear during treatment or alternatively
persist throughout treatment or for longer. For some mAb therapies, the development of antibodies
has no apparent adverse clinical consequences but for others it reduces efficacy or is associated with
therapy related adverse events.
6.2. Risk assessment
Numerous factors contribute to the generation of an immune response against a mAb and these need
consideration as part of risk assessment.
Factors impacting the incidence and severity of an immune response against a mAb (product-,
process- and patient-/disease-related risk factors), can form the basis of an approach where such risk
factors are weighed against availability and feasibility of risk assessment (or identification) and
mitigation strategies.
Risk identification, based on the factors discussed above, leads to an assessment that integrates the
individual risks in clinical context and an appropriately designed immunogenicity programme as part of
clinical development. The assessment of risk requires a multidisciplinary approach taking into account
all risks identified (e.g. related to product quality control strategy, including product formulation,
justification of acceptance limits for product-related variants and process-related impurities). This also
implies that the overall risk assessment should be linked to any comparability exercise potentially
performed during development, in case changes to the mAb used occur at different stages of product
development.
A pivotal aspect of risk assessment is, therefore, the evaluation of the rate of occurrence and the
clinical consequences of an unwanted immune response, and if these consequences can be prevented,
appropriately measured, and/or treated medically. Depending on the risk identified and available
measures to monitor and mitigate that risk, the testing programme for immunogenicity may be less or
more than that described in the general guideline. In any case, applicants should justify and discuss
their approach.
Depending on the class and subclass of the mAb (which affects immunobiological functions e.g. binding
to Fc receptors) or the mechanism of action, individual mAb products may not all have the same
clinical consequences associated with an unwanted immune response. For example, mAbs can be
neutralized by antibodies resulting in a reduced efficacy, or result in adverse events such as infusion
reactions and/or immune complex formation. Such infusion reactions can be severe, but can (unlike
allergic hypersensitivity reactions) be potentially handled by appropriate clinical measures such as the
use of pre-medication. Likewise, in case of loss of efficacy, the availability of other mAbs or related
therapeutic proteins as alternative treatment options may be an important factor for a risk mitigation
strategy. As a general guiding principle, sufficient data to estimate the severity, rate of occurrence and
detectability of the risks should be presented at the time of applying for a marketing authorisation.
9/10
These risks may have to be further substantiated by post-marketing surveillance and monitoring as
required.
As a starting point, the following considerations may be helpful for risk assessment and mitigation:
 A risk stratification based on risk identification principles as discussed in the previous section,
integrated with product-related factors like for example identification of intrinsic immunogenic
motifs, physico-chemical profile including aggregates or other product-related or processrelated
variants, information from formulation development for example solubility at
physiologic pH, location of the target antigen etc.
 Assay performance details as discussed elsewhere in this guideline, especially the extent to
which the sensitivity of the selected mAb antibody assay format is compromised by residual
circulating product.
 In case of unavoidable assay shortcomings, the availability of measures that can complement
mAb antibody monitoring, e.g. PD measurements or PK parameters.
 The availability of assays that detect early immune responses (e.g. early measurement of
binding mAbs, measurement of IgM to detect early immune responses).
 The vulnerability of the patient population; therapeutic index; auto-immune status, use of
immuno-suppressant co-medication etc.
 In the oncology setting, loss of response may be more difficult to detect than in other clinical
conditions, since it may be difficult to relate tumour progression to the development of
antibodies; progression of the disease and consequential loss of response to therapy is usually
observed in virtually all patients after a certain time, and it may be challenging to distinguish
this from effects mediated by antibodies. Therefore, more intensive measurement may have to
take place during the clinical trials to estimate what to expect in a post-authorisation setting,
especially when therapeutic alternatives are available.
 For administration of the mAb at home or in the clinic: Treatment with mAbs in the clinic may
offer advantages for immediate mitigation of infusion reactions or anaphylaxis should they
occur, but home use for subcutaneously administered mAbs may offer advantages for patient
care. Thus, applicants will have to weigh the risk of unwanted immune responses and their
consequences against the intended clinical use. For example, mAbs with a high incidence of
reactions following subcutaneous use may be less suitable for home use.
 Availability of alternative treatments or diagnostic procedures in case of loss of efficacy or
induction of infusion reactions or anaphylaxis.
6.3. Risk monitoring and mitigation
Following such an approach of risk identification and assessment, applicants should carefully plan this
concept early in product development, and revisit and update it regularly during product development
and the entire lifecycle of the mAb product as new data become available. At the beginning of clinical
development, applicants may, for example, have to assign a higher risk level to the mAb, although the
mechanism of action per se may not necessarily suggest a higher risk, if other factors make this
necessary. The risk level may, depending on the results of larger clinical trials, need to be reconsidered
following the trials. At the time of marketing authorisation application, applicants should
thoroughly justify and discuss their overall concept for the design and extent of immunogenicity testing
used during their development programme.
10/10
For products that are claimed to exhibit a particular advantage as regards immunogenic potential, (e.g.
a claim in the Summary of Product Characteristics) appropriate data to support this is usually required.
Depending on the outcome of the risk assessment, in some cases more detailed and additional testing
may be necessary during clinical trials. For example, in some instances, IgE testing prior to clinical
administration needs to be considered for patients if the mAb contains non-human carbohydrate
structures, e.g. Galactose-α-1,3-Galactose, in order to prevent severe anaphylaxis. Another instance
where development of IgE testing should be considered is where the incidence of allergic reactions was
high on first administration during early clinical development of the product. While measuring levels of
IgG subclasses, or other Ig classes like IgA, are normally not a standard requirement for
immunogenicity assessment of mAbs, they may become necessary as part of such an approach when
certain risks have been identified (e.g. nasal application). However, the determination of neutralizing
capacity and of transient versus persistent mAbs by repeated sampling is usually required.
The frequency and timing of sampling and analysis may vary depending on the identified relative risk
level. For lower risk mAbs it may be possible to reduce sampling frequency in later stages of
development, provided that no adverse events or reduced efficacy has been observed. Nevertheless,
banking of samples should be undertaken on a routine basis over the whole development programme.
Such an approach needs to be duly justified. For mAbs where a higher risk is determined, sampling
may have to be more frequent during the whole clinical development. In this situation it may be
advisable to analyze samples in real time. It may be necessary during the clinical development to
measure antibody levels, PK, PD markers, efficacy and safety simultaneously and over a period of
repeated treatments. This allows assessment of the clinical significance of antibody development, and
also whether the antibody effect changes over time, which could occur as a result of increase in titre
and/or isotype switching/affinity maturation of the antibody response. Non-neutralizing anti-mAb
antibodies may indirectly impact efficacy by binding to the mAb product and impacting on its
pharmacokinetic properties. Therefore, measurement of PK parameters may be a potential help when
planning how to measure an anti-mAb antibody response.
Information derived from assays can be used for risk management. For example, where risk
identification and assessment has concluded that early identification of an immune response is required
and allows for the possibility of discontinuing treatment with the mAb, development of low affinity IgM
antibodies can be an indicator of an early immune response, and measuring IgM responses could help
in early identification of patients mounting an immune response. Likewise, the detection of binding,
non-neutralizing antibodies may be an early indication of later development of neutralizing antibodies.
Risk mitigation strategies could, for example, include studying how to effectively handle patients that
show an immune response, e.g. if dosing schedules can be intensified without compromising patient
safety etc. However, feasibility considerations will have to be taken into account as part of this.
When filing a Marketing Authorisation Application, it is recommended that applicants present an
integrated summary of their strategy on risk identification, characterisation, monitoring, minimization
and mitigation. This risk-based approach should also have been taken into account for the Risk
Management Plan (RMP), discussing how risks identified by data from the development programme,
and potential risks, or missing information should be handled in a post-authorisation setting.