See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/23982577

Quality-control issues and approaches in vaccine development

Article in Expert Review of Vaccines · March 2009

DOI: 10.1586/14760584.8.2.227 · Source: PubMed

CITATIONS READS

45 11,874

8 authors, including:

Johan van der Gun Leo A van der Pol

Bilthoven Biologicals Intravacc
12 PUBLICATIONS 176 CITATIONS 86 PUBLICATIONS 2,048 CITATIONS

SEE PROFILE SEE PROFILE

Gideon F A Kersten
Intravacc
138 PUBLICATIONS 5,556 CITATIONS

SEE PROFILE

All content following this page was uploaded by Gideon F A Kersten on 19 February 2015.

The user has requested enhancement of the downloaded file.

 Review
For reprint orders, please contact reprints@expert-reviews.com

Quality-control issues and

approaches in vaccine
development
Expert Rev. Vaccines 8(2), 227–238 (2009)

Bernard Metz†, The successful development of vaccines is a lengthy process, requiring input from different
Germie van den expertises, such as research and development, quality control, quality assurance, production,
Dobbelsteen, regulatory affairs, and marketing and sales. A cornerstone in vaccine development is the
availability of a panel of high-quality assays that can reduce the risk of failure in clinical trials
Cécile van Els,
and licensure. This review highlights the quality-control issues and approaches in vaccine
Johan van der Gun, development from different viewpoints: discovery, process development, assay development,
Lonneke Levels, clinical development and the postlicensing phase.
Leo van der Pol,
Nynke Rots and Keywords : animal model • clinical trial • correlates of protection • physicochemical and immunochemical
technique • process development • quality control • vaccine development
Gideon Kersten
†
Author for correspondence
Research and Development
In this review, the problems and opportunities to link a certain immunological effect to specific
Unit, Netherlands Vaccine
Institute, Bilthoven,
related to quality control (QC) during vaccine product characteristics. This, in turn, presents
The Netherlands development are discussed. Vaccines can be a challenge to develop appropriate assays to
bernard.metz@nvi-vaccin.nl considered an extraordinary group of biologics. monitor the product quality for a vaccine. In
They are heterogeneous with respect to compo- addition, the large diversity of pathogens also
sition (e.g., live-attenuated, polysaccharide and generates considerable variation in the produc-
many others) and, in most cases, they are not tion technology of vaccines [4] . As a result, each
‘well defined’. Other biologics, such as hormones vaccine needs a tailor-made set of tests in order
and therapeutic antibodies, can often be char- to establish its quality. The current regulatory
acterized by relatively straightforward methods requirements for a biopharmaceutical product
to establish quality. These assays include isoelec- are well summarized in three International
tric focusing, electrophoresis, light scattering to Conference on Harmonization (ICH) guide-
determine aggregates, tryptic digest ana­lysis and lines: Q7, Q8 and Q9 [5–7] . In ICH guideline
mass spectrometry, among others. An in vitro Q7 on current good manufacturing practice
cell-based assay is often used to measure potency. (cGMP), it is indicated that the manufacturing
Although these techniques are also used to char- process must be defined by preset criteria for the
acterize vaccines, the results are less or not at all critical process parameters, ensuring a safe and
predictive for the quality of the vaccine. effective product in humans. In ICH guideline
Vaccine development is a complex challenge Q8 on pharmaceutical development, it is sug-
because: gested that a design space for the production
• There is often a lack of knowledge on how process should be explored and defined, which
vaccines exert a protective response in humans; will provide a high-quality product. Design
space is described as the multidimensional
• Vaccines are a heterogeneous product group as combination and interaction of input variables
a result of enormous pathogen diversity; and process parameters that have been demon-
strated to provide assurance of quality [7] . In
• Extensive clinical testing is required for market ICH guideline Q9 on risk management, it is
approval [1–3] . suggested that risk assessment be performed
For the development of a vaccine, there is during product development to determine areas
often insufficient scientific understanding on its with suboptimal process control and how to deal
exact protective effect, which makes it difficult with this to reach an overall acceptable level of

www.expert-reviews.com 10.1586/14760584.8.2.227 © 2009 Expert Reviews Ltd ISSN 1476-0584 227

 Review Metz, van den Dobbelsteen, van Els et al.

control. Because a high product quality requires the identification for the development of a candidate vaccine, suitable tests are
of critical process parameters and definition of the design space, needed, whereas, for the assay-development reference, materi-
it will be clear that identifying the product quality attributes als of high quality are mandatory (Figure 3) . This results in an
and the development of assays that provide this information are iterative process where the proof of concept evolves into a final
essential for process development of biopharmaceuticals. vaccine product, and an experimental test becomes a validated
To meet the complex challenge for developing a new vaccine, it assay (see also the section on QC and process development).
seems best to follow an integrated approach by combining input The values of critical parameters are known and defined in the
from different expertises, such as R&D, QC, quality assur- product specification (Figure 4) . These specifications also include
ance (QA), production, regulatory affairs, and marketing and alarm limits besides the set points and acceptance limits of the
sales, into a project plan with milestone achievements with criteria critical parameters.
reflecting the best compromise between expertises. This review is
written from different viewpoints, using the main activities and Toolbox
phases in vaccine development as starting points: discovery, pro- In the past, the QC of vaccines depended mainly on methods to
cess development, assay development, clinical development and ensure that the products were safe and potent [8] . QC was often
the postlicensing phase (Figure 1) . The review provides an overview based on final lot testing in animals. Regulatory guidelines are
of the main QC issues and approaches in vaccine development, usually close behind state-of-the-art with regard to analytics.
supported with typical examples. This can result in additional requirements in the areas of iden-
tity, stability, purity, structure of the antigen and consistency in
Assay development production, in order to obtain regulatory approval. How these
The availability of a panel of high-quality assays in the develop- regulatory expectations can be practically applied is reviewed in
ment stage of a vaccine can reduce the risk of failure during clini- detail by Lebron et al. [9] . A toolbox containing a large variety
cal trials. High quality not only means that the assays generate of analytical techniques can be applied for the assessment of
valid data but also that the data are meaningful (i.e., information vaccine quality.
about product quality is provided). Whether an assay is relevant In the last few decades, the number of techniques used for
is not always immediately obvious. Ideally, a vaccine is a well- the characterization of antigens has increased substantially. A
defined product – that is, its structure and composition are known huge variety of analytical techniques is available for the charac-
in detail. Often this is impossible to achieve because of technical terization of vaccines. An overview of the techniques is given in
reasons (it cannot be done) or economical reasons (it is too costly Table 1. Techniques can be divided into two categories that will
and time consuming). However, a vaccine developer must provide now be discussed:
convincing data to proceed to toxicity and stability studies, to
begin clinical studies and to achieve registration. The assays that Techniques for measuring the concentration &
are needed to demonstrate this depend to a large extent on the integrity of antigens
type of vaccine. For live-attenuated viral vaccine assays, measur- This category includes colorimetric assays and analytical separa-
ing the number of infectious particles and the level of impurities tion techniques used for the quantification of intermediate and
is important. In the case of a recombinant protein, assays may be final products and for the detection of impurities in the vaccine.
crucial to address chemical and conformational stability, aggrega- In general, colorimetric assays are easy to perform. They are
tion and the presence of post-translational modifications, at least applied to measure protein concentrations, polysaccharide con-
during the development phase. When the product is licensed, the tent, number of primary amine groups and host cell DNA, among
number of tests is reduced again (Figure 2) . This can be carried out others. Separation techniques, such as liquid and gas chromato­
because the production process is validated and consistent produc- graphy and gel electrophoresis, are often used to determine the
tion has been demonstrated. A selection of release tests can now stability and purity of protein and polysaccharide vaccines. Mass
ensure the quality, potency and safety of the vaccine. spectrometry is increasingly used to characterize antigens and
The development of assays often occurs in parallel with the impurities. Identification, fragmentation and other structural deg-
vaccine development itself, because suitable assays do not initially radation (oxidation and deamidation) can be measured relatively
exist. Vaccine and assay development are mutually dependent: easily and quickly.

Discovery Small-scale process development Early clinical Scale-up and validation Late clinical Market

Animal model Assay development Assay validation

Time

Figure 1. Highly schematic overview of a vaccine-development sequence. The subject of this paper is discussed from five
viewpoints: assay development, animal models/correlates of protection, process development, clinical development and postlicensure.

228 Expert Rev. Vaccines 8(2), (2009)

 Quality-control issues & approaches in vaccine development Review

A selection of physicochemical and

immunochemical techniques is often used
Start clinical
Development to demonstrate consistency of production
Process validated
(in-process controls) in comparability stud-
Number of assays

Licensing ies after a process change or for assuring the

quality of final product (release testing).
Characterization assays Another hurdle is the quality assessment
of individual antigens in combination vac-
cines. The efficacy of all antigens must be
Release assays and IPC proven by a correlate of protection or in
an efficacy study. The quality assurance of
a combination vaccine is mainly based on
animal studies and immunochemical and
Time serological analyses. Many physico­chemical
assays are not useful because they cannot
discriminate between two or more anti-
Figure 2. Assay effort as a function of time. The number of assays stabilizes before gens. Challenges in the development, QC
clinical trial materials are prepared. A large number of assays is generally needed to
complete the Investigational Medicinal Product Dossier preceding clinical studies. After
and licensure of combination vaccines are
process validation and licensing, it is possible to reduce the number of tests when it has reviewed in three publications [13–15] .
been shown that the production process is robust and the product can be prepared
reproducibly. What is left are batch-release assays and IPCs. Evaluation of safety & potency
IPC: In-process control. The safety and potency assessment of a can-
didate vaccine is a major target during its
Techniques for measuring the biological activity, active development. Vaccine development or life-cycle management of
concentration or conformation of antigens existing vaccines benefit from the availability of fast and accurate
This group comprises animal tests, cell-based assays (see the in vitro alternatives for in vivo safety tests. Examples are the use of
section entitled ‘Animal models & other tests’) and physico- a PCR method to measure virulent poliovirus in oral polio vac-
chemical and immunochemical techniques. Physicochemical cine preventing neurovirulence testings in monkeys [16,17] , and a
and immunochemical assays are more and more important in very sensitive in vitro Vero cell assay for the detection of residual
the quality assessment of vaccines. These techniques can reveal diphtheria toxin described in the Ph Eur. Monograph [18] . Vaccine
conformational differences in the antigen structure between dis- safety is guaranteed further by developing consistent production
tinct batches or the accessibility of immunodominant epitopes processes and postmarketing surveillance.
for monoclonal antibodies. The techniques belonging to this In most cases, potency assays are performed in animals. The
category can assure the integrity of the antigenic structure current potency tests can be divided into four different types:
that is often important for the potency of vaccines, although • The immunization–challenge test
a direct link between antigen structure and potency or efficacy
is often not known. Furthermore, physicochemical techniques • Serological analyses (e.g., for measuring neutralizing antibodies)
have been demonstrated to be very valu-
able for the assessment of the quality of
the adsorbed final product. Many vac- Vaccine
cines contain aluminium phosphate or candidate
aluminium hydroxide, to which the anti-
gen is adsorbed. Furthermore, the antigen
concentration in the final lot is relatively Reference Analyses on
materials product
low. This has hampered the characteriza-
tion of final lot material. In order to study
these antigens, they must be desorbed or
Assay
the adjuvant has to be dissolved. This pre-
treatment may have an effect on antigen
conformation. Usually, only the potency
and the degree of adsorption are measured.
Recently, advances in spectroscopy, such as Figure 3. Process development and assay development are interdependent.
increased sensitivity, improved optics and Assays are needed to characterize the vaccine as it is being developed and the reference
better software, make it possible to study product is needed to develop and qualify assays. Product and assay specifications must
antigens in the adsorbed form [10–12] . be determined early in the process to avoid too many iterations.

www.expert-reviews.com 229
 Review Metz, van den Dobbelsteen, van Els et al.

that play a role in the attachment of patho-

Tested range
gens to host cells is an important step when
attempting to develop an animal model.
Specification Using genetic methods, transgenic mice
expressing human target receptors, such
Alarm limits as CD46 or CD66, have been developed.
These transgenic mice have been used
as an animal model for measles virus [21]
or Neisseria infections [22] . Alternatively,
‘humanized immunodeficient mice’, in
which human tissue that retains immuno­
Set point logical functions is transplanted into mice,
can be used. In mice transplanted with
Parameter value
human umbilical cord blood CD34 + cells,
human immune functions could be recon-
Figure 4. Vaccine quality requires the identification and specification of critical
structed and studied in the presence of
product and process parameters. Each parameter must be tested in a range to
demonstrate its effect on product quality. For a desired product quality, each parameter HIV infections [23] .
value has to be measured within the specification range. The ranges of all parameters However, the immune responses induced
must be assessed and summarized in the product specification before the realization of by infection may not always be the same as
clinical trials. those providing protection after vaccina-
• Cell-based assays (e.g., for detecting cellular immune responses) tion [2] . To further complicate the search for
correlates of protection, the immune system is redundant and the
• Titration assays different types of responses to vaccines act synergistically.
For live-attenuated vaccines, potency is often indirectly
determined by measurement of the number of viable particles, Serological correlates of protection
determined either by colony counting or by virus titration. In Many assays have been developed that measure antibody levels
some cases, it appears feasible to use quantitative PCR (qPCR) or some other aspects of the adaptive immune response, such as
instead of virus titration [19] . The advantage is that qPCR is faster, antibody concentration, antibody avidity (a surrogate marker
involves less hands-on time and is probably more accurate. In the for memory), functional antibody assays (i.e., pathogen or toxin
following section, the first three types of assays are discussed in neutralizing) and cellular immune responses (T-cell prolifera-
more detail. tion and cytokine quantification). Although mucosal and cellular
immune responses clearly contribute to protection induced by
Animal models & other functional tests some vaccines, most vaccines licensed today depend on the induc-
A correlate of protection is defined as a specific immune response tion of serum antibodies for their efficacy. For various pathogens,
to a vaccine that is closely related to protection against infection, particular levels of antibodies have been identified or suggested
disease or other defined end points [2] . Correlates of protection to that confer protection (see the ‘Clinical development’ section) [2] .
bacterial and viral pathogens are desirable for identifying protec- Antibody responses are easy to measure in ELISA or in newly
tive antigens, demonstrating the immunogenicity of a candidate developed multiplex assays analyzing titers for several antigens
vaccine and its potential efficacy and permitting optimization of simultaneously in very small serum volumes [24–27] . Ideally,
the dose, delivery vehicle, adjuvant and schedule of immuniza- these assessments should be supplemented with functional anti-
tion. The identification of an immune response as a correlate for body assays, such as complement-mediated bactericidal activity,
protection against a certain disease would greatly facilitate the neutralizing or opsonic activity, or passive protection in animal
rational development of an effective vaccine against this disease. models of disease [2] . The induction of autoimmune or blocking
However, finding such a correlate can be difficult. antibodies, such as the capsular polysaccharide of N. meningi-
tidis type B [28] , the outer surface protein of Borrelia burgdorferi
Animal models [29] , wrongly formulated N. meningitidis pore protein A [30] or
Vaccination and challenge studies in animals provide an ideal disease-enhancing antibodies, as seen in the 1960s with inac-
platform to compare induction of immune response parameters tivated paramyxovirus vaccines [31] , should be prevented by the
and assess the correlation of these parameters with protection. It use of appropriate animal models and immunoassays and the
is only in animal models that interactions of the organism with characterization of any crossreaction with human tissues.
the innate, humoral and cellular immune system can be assessed. Recent experience with waning immunity in children after con-
However, it turned out to be very difficult to develop animal jugate vaccination during infancy [32] indicates that knowledge of
models for strictly human pathogens, such as Neisseria menin- the precise duration of direct protection conferred by vaccines and
gitidis, due to the specificity of a range of surface proteins that long-term protection by the induction of immunologic memory
interact with the host receptors [20] . Unravelling the molecules should be included in preclinical and clinical studies.

230 Expert Rev. Vaccines 8(2), (2009)

 Quality-control issues & approaches in vaccine development Review

Table 1. Physicochemical and immunochemical techniques for vaccine development.

Technique Information regarding Identity Purity Stability Structure*‡ Ref.
Chromatography
Reversed phase Purity, stability, degradation and +* + + 1° [60–62]
protein modifications
Ion exchange Protein modifications, degradation, aggregation 0 + 0 1° [60,63]
and purity
Size exclusion Hydrodynamic size, aggregation, purity and 0 + + 1°, 4° [64,65]
oligomeric repartition
Immunochemical techniques
Biosensor ana­lysis Antigen concentration, epitope integrity and + 0 + 3° § [45,66]
binding kinetics
ELISA Antigen concentration + 0 + 3° § [25,27,67]

Immunoblotting Size, protein modifications, degradation + 0 + 1°, 4° [64,67]

and aggregation
Spectroscopy
Circular dichroism Secondary and tertiary structure - - 0 2°, 3° [45,68–71]

Electron microscopy Imaging of supramolecular structures 0 0 + 4° [68,72,73]

and integrity
Fluorescence spectroscopy Tertiary structure of proteins, protein unfolding - - 0 3° [12,45,74]
and refolding
Infrared spectroscopy Excipients and protein structure - 0 - 2° [11,75–77]

Light scattering Aggregation - - + 4° [71,75]

Nuclear magnetic resonance Excipients and polysaccharide structures - 0 - - [63]

UV absorbance spectroscopy Tertiary structure of proteins, protein unfolding - - + 3° [71]

and refolding, and aggregation
Electrophoresis
Capillary electrophoresis Primary structure of proteins and 0 0 + 1° [78,79]
protein degradation
IEF pI, protein modifications, degradation, 0 + + 1° [65,80]
aggregation and purity
PAGE (SDS or native) Size of proteins, protein modifications, 0 + + 1°, 4° [64,81]
degradation, aggregation and purity
2D electrophoresis (IEF plus Size, pI, protein modifications, degradation, +* + + 1° [82]
SDS-PAGE) aggregation and profile of protein impurities
Mass spectrometry
Primary structure of proteins, protein + 0 + 1° [60–
62,83,84]
modifications and degradation
Differential scanning calorimetry
Thermodynamics of protein unfolding - - + 3° [12,85,86]

+: Yes; 0: Neutral/possibly; -: No/unfavorable.

*
In combination with mass spectrometry or N-terminal sequencing by Edman degradation.
‡
1°: Pimary structure; 2°: Secondary structure; 3°: Tertiary structure; 4°: Quaternary structure of proteins.
§
If monoclonal antibodies are used that recognize conformational epitopes.
IEF: Isoelectric focusing; PAGE: Polyacrylamide gel electrophoresis; SDS: Sodium dodecyl sulfate

Memory B cells are important for long-term humoral immu- an independent parameter of antigen-specific immune memory.
nity after vaccination [33,34] . For most antigens, it is unclear Several B-cell ELISPOT assays to quantitate antigen-specific
whether memory B-cell and serum antibody levels correlate well, memory B cells in human blood have been developed [35,36] . These
and it is therefore important to be able to track memory B cells as assays utilize a 5- or 6‑day polyclonal stimulation procedure of

www.expert-reviews.com 231
 Review Metz, van den Dobbelsteen, van Els et al.

peripheral blood mononuclear cells (PBMCs) followed by an released by antigen-stimulated innate immune cells have been
antigen-specific ELISPOT for the detection of memory B cells shown to be critical for the differentiation of the CD4 + T-cell
that have differentiated into antibody-secreting cells in vitro. response into either a Th1-, Th2- or Th17-type response. These
Thus, while immediate protection induced by vaccines requires are distinguished by the production of particular sets of T-cell
the development of specific plasma B cells for antibody produc- cytokines associated with different effector mechanisms. A study
tion, sustained protection is conferred by a memory B-cell pool. of profiles of cytokines is, therefore, more informative than a
How important that is for vaccine efficacy needs to be investi- study of individual cytokines. In addition to cell-based assays,
gated. For some pathogens, the replenishment of plasma cells the ana­lysis and quantification of T-cell cytokines (and chemo­
by memory B cells must be continuous, as a certain minimum kines) secreted in biological fluids and tissue culture supernatant
level of serum antibodies is needed for protection (e.g., N. men- has become widely used in vaccine research. A major step in this
ingitidis). In other cases, the maintenance of a memory B-cell field is the advent of microsphere-based multiplex assays, allow-
pool in the absence of an antibody titer may be protective, but ing the quantification of multiple soluble T-cell factors in small
only if the specific recall reaction after exposure is fast enough analyte volumes with high sensitivity and reproducibility. Several
to boost antibody production to clear the pathogen [34] . human studies assessing cytokine profiles after vaccination or
natural infection, using either soluble or cell-based assays, have
T-cell parameters as correlates of protection suggested the importance of the multifunctionality of T cells
As another component of the adaptive immune response, T-cell for protection or antigen load. Ex vivo levels of IFN-γ, IL-10
responses have been demonstrated to be induced by most protein and granzyme B in influenza virus-stimulated PBMCs were a
or live vaccines routinely administered today. Cytotoxic CD8 + better measures of risk for influenza illness than antibody tit-
T cells (cytotoxic T lymphocytes [CTLs]) and CD4 + Th cells ers in elderly vaccinees [40] . Furthermore, Seder and colleagues
exert their specific effector functions after recognizing pathogen- have related the production of three or more cytokines by single
derived protein fragments in the context of MHC class I or II CD8 + T cells to maximal functional differentiation and disease
molecules, respectively, on the outside of antigen-presenting outcome in various viral infections [41] .
cells. CTLs are probably unable to prevent infection but are
important in controlling an established infection. Alternatively, Quality control from process development perspective
CD4 + Th cells secrete cytokines and provide cognate help needed Product assays & process development
for the full maturation and maintenance of other antigen-spe- Since adequately monitoring the quality (potency and safety) of
cific immune cells, such as CTLs and B cells. The only clini- a candidate vaccine is quite difficult, the tendency is to follow
cal evidence that CTLs contribute to protection was reported the ‘process = product’ paradigm that demands the completion
by McMichael et al., who demonstrated an inverse correlation of process scale-up and validation before starting clinical trials [1] .
between virus-specific cytotoxicity and virus shedding in experi- However, this approach presents an unacceptable risk, because
mentally influenza virus-infected individuals who had no virus- huge investments might be wasted after a negative Phase I result.
specific anti­bodies [37] . Similarly, a number of studies have shown The exact epitopes that determine product quality are often not
that Bacillus Calmette–Guérin, the only licensed TB vaccine, defined and, for a subunit vaccine, usually adjuvants with only
relies on the generation of CD4 + T-cell responses and not antibody partially understood mechanisms of action have to be added. This
responses in human infants (reviewed in [38]). The major effector creates difficulties for the development of appropriate potency
function of these CD4 + T cells seems to be the delivery of mac- assays. It is also clear that, with an animal test that only requires
rophage-activation signals, predominantly IFN-γ. Nevertheless, 2 months, there will be no efficient exploration and confirmation
despite solid theories behind the importance of CD8 + and CD4 + of the design space for a production process. Therefore, it seems
T-cell-based immunity against a number of infectious diseases, best for the vaccine-development process to develop ‘marker’ assays
which are often extrapolated from experimental animal stud- for product quality that can be linked to the response in animals
ies, no human T-cell-related correlates of protection have been and, after clinical trials, the vaccine performance in humans. This
elucidated so far. This lack of clearly defined cellular correlates information on the product quality should be linked to a better
of protection illustrates a major challenge in vaccinology – that scientific understanding of the production process, as promoted
is, to fill the gaps in our knowledge on T-cell-based protective by the Process Analytical Technology (PAT) initiative of the US
immune mechanisms against a number of pathogens, such as FDA [42] . To this end, fingerprinting methods to characterize
HIV, influenza virus and Mycobacteriuim tuberculosis. Numerous the product prove very valuable for demonstrating consistency in
assays for the quantification and functional characterization of product production. This approach can be applied during many
T cells responding to defined antigens are cuurently based on stages of the production process: from the cultivation stage [43,44]
flow cytometry [39] . These include single cell-based fluorescence- to the inactivation step [45,46] . A marker assay that is used to moni-
activated cell sorting (FACS) ana­lysis of intracellular cytokines tor the quality of intermediate product of a whole-cell bacterial
and effector molecules, such as perforins and granzymes, MHC vaccine is a microarray for the fermentation process of Bordetella
multimer binding or the upregulation of activation markers and pertussis. Since the virulence factors of this bacterium are arranged
costimulatory molecules, such as CD40 ligand (CD40L) and in one gene cluster that is either on or off, the RNA microarray
molecules of the B7 family of immunoregulatory ligands. Signals presents a good marker assay for the presence of virulence factors

232 Expert Rev. Vaccines 8(2), (2009)

 Quality-control issues & approaches in vaccine development Review

that determine product quality in this phase of the production guideline to assist sponsors in this transition from nonclinical to
process. Another example is the fingerprinting method that was early clinical development [48] . The trigger for this new guideline
applied for diphtheria toxoid [45,46] . A combination of five physi- has been the development of serious adverse reactions during the
cochemical or immunochemical techniques can predict the qual- first-in-human clinical trial in March 2006 with a T-cell ago-
ity of the diphtheria toxoid vaccine. This approach needs to be nist monoclonal antibody TGN1412 developed to treat chronic
completed by the monitoring of product stability during further inflammatory diseases and leukemia. All six healthy volunteers
process steps and by demonstrating a correlation between newly who received the antibody rapidly developed multisystem failure
developed assays and the mandatory tests in animal models [44] . and became critically ill within a few hours after taking the
drug, probably due to a cytokine-release syndrome [49] . Several
Interaction between assay & process development questions were raised after this disaster. Why was the drug tested
The development of a new vaccine from concept to licensure goes on healthy volunteers with a normal functioning immune system
through several phases (Figure 1) . During the preclinical develop- rather than on patients with an impaired immune system, why
ment, a first product batch is made with cells from an initial R&D were all volunteers treated at the same time, and what preclinical
seed lot, with an often incomplete target process. After several information was available? The national competent authori-
iterations, a product for Phase I/II studies can be produced using ties in the EU member states, together with EMEA and the
cells from an approved master or working-seed lot, with a stand- European Commission, responded to this case with the afore-
ardized production process according to cGMP guidelines, with mentioned guideline, for potential high-risk medicinal products,
the product material used for toxicity studies as a reference stand- in an attempt to minimize the risk of such serious adverse reac-
ard. As long as the assumptions defined at the start of the project tions occurring. A draft version was published 1 year after the
seem correct and the feasibility criteria for both assay develop- trial. Most of the recommendations made in the guidance were
ment and process development are met, the project moves forward based on the expert report established in the UK in the wake
according to plan with smooth interactions between assays and of the failed trial. The guideline contains recommendations on
process steps. However, if unexpected disappointing results are quality aspects, risk assessment and communication, preparatory
found, the approach has to be evaluated, raising questions such as: preclinical work and experimental design of the first-in-human
does the cell line produce the correct product? Is the (intermedi- clinical trials with particular attention to dosing. It also includes
ate) product stable? Are the used assays reliable? When problems a statement that an investigational product with a high a priori
are persistent in this phase, the development can show similar- assessed risk rules out the use of healthy volunteers. Finally, it
ity with a failure investigation trying to find a root cause for an is acknowledged that, if the nature of the drug target is human
unwanted result that blocks the development. The first question specific, the ability of nonclinical studies to predict safety in
to be answered in this situation is: can the assays be trusted? Once humans may be limited. In these cases, the use of homologous
it is certain that the assays will provide reliable results, the quest proteins or relevant transgenic mice expressing the human tar-
to tackle the problems becomes worthwhile. get, or the use of in vitro human cell systems, could provide
To reduce risk during the early exploration phase for proc- relevant additional information.
ess development and assay development, it helps to use platform Another challenge in clinical research with vaccines is how to
technology and to clearly define assumptions and boundary show that a vaccine is effective in protection against a disease
conditions at the start of the development project (Figure 4) . If with a low incidence rate or if a similar vaccine is already avail­
state-of-the-art standard assay techniques are used, the variance able on the market. Vaccine efficacy is defined as the percentage
of the assay will be known and a blueprint for qualification and of reduction in the incidence rate of disease in vaccinated com-
validation will be available for the new vaccine product, based pared with unvaccinated individuals of the target group. If the
upon similar previous products. If assumptions on the stability of vaccine is being developed against a disease with a low incidence
raw materials and (intermediate) product are defined and result in rate, efficacy trials are difficult to perform, since large numbers
clear boundary conditions for sample treatment and the storage of participants would be required. Farrington and Miller pre-
of media, buffers, products and assay reagents, this can not only sented a sample-size calculation for a clinical trial with a menin-
provide information for later validation studies, but also assist in gococcal serotype B (MenB) vaccine [50] . To confirm efficacy
finding potential causes for problems. in a 2-year follow-up trial in infants, approximately 12% of the
A challenge for the future would be to develop a modular total birth cohort of the UK (n = 78,000 infants) would have to
approach for the assay and process development to support a participate to show efficacy greater than zero, to show efficacy is
fast-track approval for a new variant of a pathogen, similar to the greater than 50%, 68% of the UK birth cohort (n = 442,000,
yearly program related to the development and manufacturing of which is twice the Dutch birth cohort) is needed.
seasonal influenza vaccines using eggs [47] . To overcome this problem, ‘correlates of protection’ have
been identified to be used as indications for efficacy during the
Clinical development clinical development of a vaccine (see also the section entitled
A difficult decision during vaccine development is when to move ‘Animal models and other functional tests’). Levels of correlates
from preclinical studies, such as animal or in vitro studies, to the of protection induced by new vaccines should be compared with
first tests in humans. Recently, the EMEA has adopted a new levels induced by licensed vaccines that have been evaluated

www.expert-reviews.com 233
 Review Metz, van den Dobbelsteen, van Els et al.

for efficacy against disease. For a number of vaccines, such as QC aspects after licensing
diphtheria and tetanus vaccines, serum antibody titers above a Once a product has been licensed, the quality of each individual
certain threshold have been generally accepted to correlate with batch should be tested by the QC laboratory in accordance with
protection against disease. An anti-toxin level below 0.01 IU/ml the marketing authorization. This irrevocably implies the QC
is correlated with susceptibility for diphtheria, 0.1 IU/ml is con- laboratory to run under a stringent quality system based on the
sidered a protective level of circulating anti-toxin and levels of GMP guidelines in force. Approved, written procedures should
1.0 IU/ml and above are associated with long-term protection. be available ensuring that, among other things, the necessary
The limit for susceptibility to diphtheria has been set based tests are carried out, describing how to deal with foreseen and
on the border between a positive and a negative Schick test. unforeseen deviations, ensuring out-of-specification situations are
The Schick test assesses erythema and induration 48 h after prepared for, validation of test methods is performed, as well as
intradermal injection of a minute amount of diphtheria, and the trending and training of personnel. Tests, performed by the
positivity has been defined as erythema and/or induration of QC laboratory, should always be validated specifically for testing
1 cm or more. the quality of the licensed product.
For pneumococcal vaccines, a WHO working group has The implementation of the quality system, both during pro-
recommended that IgG antibody concentrations, as measured duction as well as at the QC laboratory, will lead to the manu-
by ELISA in sera collected 1 month after a primary series of facturing of a consistent product with ‘built-in quality’ and the
three vaccinations, should be chosen as the main licensing performance of validated tests resulting in consistent results.
criterion, with a threshold anticapsular antibody concentra- Although implementation of a quality system is costly and time
tion of 0.35 µg/ml for all pneumococcal serosubtypes. IgG demanding, it is a necessary contributor to the release of high-
antibody concentrations should be used in the noninferiority quality vaccines and will, in the end, enable a manufacturer to
ana­lysis of a new vaccine when compared with a licensed vac- taste the benefits.
cine against invasive pneumococcal disease [51] . This reference Immediately after licensing, the QC will usually consist of a
antibody concentration has been determined by meta-ana­lysis complete set of tests, as laid down in the marketing authoriza-
of data from three double-blind controlled efficacy trials with tion. However, for various vaccines, the European Pharmacopoeia
invasive disease: two with a heptavalent vaccine and one with does allow the deletion of certain tests whenever it has been dem-
a nonavalent pneumococcal conjugate vaccine. The threshold onstrated that the production process consistently ensures com-
level of 0.35 µg/ml was the antibody concentration achieved by pliance with the specifications and is approved by the relevant
93% of subjects. Currently, the appropriateness of the serum regulatory authorities [53] . For the traditional vaccines, such as
bactericidal activity (SBA) threshold is debated. Several regula- diphtheria and tetanus vaccines, the specific toxicity test and
tory agencies requested that the WHO reconsiders their rec- flocculation test, for example, belong to this category, and are
ommendations, which were initiated in July 2008. For other based on a validation study, which is no longer performed in the
vaccines, such as the MenB vaccine, the use of correlates is not early stage of the production process by some manufacturers.
generally accepted. Moreover, it is recognized that the SBA This saves time, costs and test animals. Reduction is not always
may underestimate meningococcal vaccine efficacy, as there a consequence of economics and can also originate from a gen-
may be other mechanisms of protection involved in immunity eral concern regarding animal welfare. Most potency assays, for
to disease, such as opsonophagocitic activity (OPA), which example, either based on challenge or serology, consist of the
measures functional activity and herd immunity. Currently, the immunization of animals with graded doses of test and reference
relationship between OPA and efficacy has not been established vaccine. A simple and common applied approach is to reduce
in clinical trials. the number of dilutions for both vaccines to one [54,55] . Based
Both the induction and the maintenance of functional serum on available data obtained with the validated assay, critical test
antibody titers have a cellular basis, indicating the added value parameters, such as linearity and parallelism, are guaranteed and,
of the assessment of cellular immune responses. This has also as long as the vaccine and critical reagents used in the assay are not
been recognized by the EMEA, since the guideline on clinical modified, vaccine potency can be estimated by simply comparing
evaluation of new vaccines [52] , effective since February 2007, the relevant response of just one vaccine dilution.
states that an assessment of the cell-mediated immunity com- Part of the procedure leading to deletion of a test should be
ponent of the immune response to a novel antigen is consid- a risk assessment [6,56] , which can also be used to evaluate and
ered important and, for some types of antigens, it is essential. justify the frequency of testing or the deletion of tests additionally
Therefore, the understanding of immunity to the pathogen performed during vaccine development, but that are not strictly
is essential, so that only the relevant immune responses are required for lot-release testing.
studied. It is recommended that studies monitor the quantity
and quality of T-cell responses, such as antigen-specific T-cell Revalidation
frequencies, Th1, Th2, T regulator cells, memory T cells and Based on the ICH guidelines [57] , revalidation of a method is
relevant cytokines. Today, cellular immunity measurements are required whenever major changes are implemented. These com-
recommended. It is expected that, in the near future, they will prise changes in the production of the active vaccine component
become essential. or modification in the composition of the vaccine. In addition,

234 Expert Rev. Vaccines 8(2), (2009)

 Quality-control issues & approaches in vaccine development Review

changes in the analytical procedure itself, such as the critical rea- of protection do not initially exist for new vaccines. In the best
gents of a method, inevitably require revalidation. Furthermore, case, an animal model can be developed in which some kind of
a method transfer of an externally validated method into the QC functional immune response can be generated, but this may not
laboratory requires revalidation. be indicative for efficacy. With regard to vaccine safety and sta-
Notwithstanding the use of well-standardized and validated test bility, there are similar problems. Safety- and stability-indicating
methods, results can be obtained that do not fulfil the requirements tests are difficult to develop. A lack of correlates of protection
set for the particular product. Only when all validity criteria of the and assays indicating safety and stability make vaccine devel-
method are met, this can be considered as an out-of-specification opment a risky undertaking, because failures become apparent
(OOS) situation [58] . An OOS situation can have a tremendous after Phase II or Phase III studies, which are the most expen-
impact on the production process and should be reported imme- sive phases of the development. A better understanding of the
diately to all parties involved. The first step in an OOS situation underlying immune mechanisms, the use of fast and accurate
is to assess the accuracy and validity of the test data. Whenever in vitro assays next to in vivo studies, as well as robust process
possible, reagents and test samples should not be discarded during development using ‘omics’ approaches and process analytical
this phase, enabling retesting with the same materials in case of technology, can reduce the risk of failures in the late stages of
laboratory error or instrument malfunctions. If no clear abnormali- vaccine development.
ties are observed, a full-blown OOS investigation will be com-
menced, according to a predefined procedure. The investigation Five-year view
should focus on the root cause of the failure, the possible impact To bear the increasing costs for vaccine development, methods to
on already distributed batches and a review of production and reduce the number and size of clinical studies must be developed.
sampling procedures. Generally, additional tests are performed The ability to dissect the immune response and to unravel the
to investigate and identify potential problems. Retesting is only mechanisms of action of adjuvants will lead to the design of vac-
permitted when the standard operating procedure of the specific cines with a higher therapeutic index. This will make them safer
assay allows it. Due to their inherent variation, methods in which and more efficacious. New methods will become available to
test animals are used often do allow a retest. In all other situations, measure more immune parameters, including cellular responses,
no retest is permitted for OOS based on a valid assay. in a smaller volume of serum. This will allow smaller initial
clinical studies and will provide a better rationale to proceed to
Use of trending larger trials. It will be a major challenge for the coming decade
Within the QC setting, the consistent performance of assays is to identify numeric and qualitative B- and T-cell para­meters as
crucial and should be safeguarded. A common tool is the use of correlates of protection in the most important infectious diseases
Shewhart control charts [54] . All critical test parameters, including and to refine methods to be applicable to small blood samples
reference and standard preparation, calculated results and posi- or samples from more unconventional lymphoid origin, such
tive and negative controls, should be monitored using Shewhart as the mucosal system. Mucosal effector mechanisms may be
control charts [59] . These charts are based on obtained data and important co-correlates of protection to prevent colonization of
take into account the variation in the respective parameter. many pathogens at the site of their entry [2] .
When correlates of protection have been established, the next
Securing product & assay knowledge in a step can be the development of in vitro assays instead of ani-
routine environment mal experiments for testing new vaccine candidates or improve-
Often, assays are developed by the R&D department and subse- ments of existing vaccines. It will be easier to develop production
quently transferred to the QC department, where the test will be processes within tight specifications due to the use of design-of-
performed on a routine basis under good practices conditions. Test experiment approaches and genome, proteome and metabolome
transfer consists of the training of QC personnel with a standard ana­lysis. Experimental products obtained in this way can be char-
operating procedure and qualification or validation report as the acterized in vitro with a panel of fast in vitro assays, limiting the
main documentation. Each technician should be qualified for the use of animal studies.
test to be performed according to a prescribed training procedure,
authorized by the responsible manager. Once qualified, the tech- Acknowledgements
nician will run the assay on a routine basis for years. This might The authors would like to thank DR Mekkes for critical reading of this
result in a lack of background knowledge and a lack of capacity to manuscript.
solve problems or even to detect them. Therefore, measures should
be taken to ensure relevant know-how and skills are embedded in Financial & competing interests disclosure
the training of technicians and to keep them motivated. The authors have no relevant affiliations or financial involvement with any
organization or entity with a financial interest in or financial conflict with
Expert commentary the subject matter or materials discussed in the manuscript. This includes
The quality of vaccines is difficult to determine, because it is employment, consultancies, honoraria, stock ownership or options, expert
often not exactly known what the critical parameters are and testimony, grants or patents received or pending, or royalties.
to what extent these factors influence the efficacy. Correlates No writing assistance was utilized in the production of this manuscript.

www.expert-reviews.com 235
 Review Metz, van den Dobbelsteen, van Els et al.

Key issues
• Vaccine development is a complex challenge and, because there is often a lack of knowledge on how vaccines exert a protective
response in humans, they are a heterogeneous product group as a result of enormous pathogen diversity; and extensive clinical testing
is required for market approval.
• A panel of high-quality assays in the developing stage of a vaccine can reduce the risk of failure during clinical trials.
• Physicochemical and immunochemical assays can assure the integrity of the antigenic structure that is often important for the potency
of vaccines, although a direct link between antigen structure and potency or efficacy is often not known.
• Marker assays that can be linked to potency are important for process development, reducing the developing time of a vaccine.
• Identification of an immune response as a correlate of protection against a certain disease would greatly facilitate the rational
development of effective vaccines.
• A challenge in clinical vaccine research is to show that a vaccine is effective against a disease with a low incidence or a similar vaccine is
already available on the market. In the first case, extensive serology is paramount. In the latter case, serological data from vaccinees
immunized with the existing vaccine may provide indication for protective antibody levels.
• After licensing, deletion of quality-control tests is allowed whenever it has been demonstrated that the production process consistently
ensures compliance with the specifications and is approved by the regulatory authorities.

the structure and stability of model protein 20 Gorringe AR, Reddin KM, Funnell SG,
References antigens. J. Biol. Chem. 280(14), Johansson L, Rytkonen A, Jonsson AB.
1 Buckland BC. The process development 13406–13414 (2005). Experimental disease models for the
challenge for a new vaccine. Nat. Med. 12 Peek LJ, Martin TT, Elk Nation C, assessment of meningococcal vaccines.
11(4), S16–S19 (2005). Pegram SA, Middaugh CR. Effects of Vaccine 23(17–18), 2214–2217 (2005).
2 Plotkin SA. Vaccines: correlates of stabilizers on the destabilization of proteins 21 Mrkic B, Pavlovic J, Rulicke T et al.
vaccine-induced immunity. Clin. Infect. upon adsorption to aluminum salt adjuvants. Measles virus spread and pathogenesis in
Dis. 47(3), 401–409 (2008). J. Pharm. Sci. 96(3), 547–557 (2007). genetically modified mice. J. Virol. 72(9),
13 Edwards KM. Development, acceptance, 7420–7427 (1998).
3 Peter G, des Vignes-Kendrick M, Eickhoff
TC et al. Lessons learned from a review of and use of immunologic correlates of 22 Johansson L, Rytkonen A, Bergman P et al.
the development of selected vaccines. protection in monitoring the effectiveness CD46 in meningococcal disease. Science
National Vaccine Advisory Committee. of combination vaccines. Clin. Infect. Dis. 301(5631), 373–375 (2003).
Pediatrics 104(4 Pt 1), 942–950 (1999). 33(Suppl. 4), S274–S277 (2001). 23 Nomura T, Tamaoki N, Takakura A,
4 Sheridan C. The business of making 14 Taffs RE. Potency tests of combination Suemizu H. Basic concept of development
vaccines. Nat. Biotechnol. 23(11), vaccines. Clin. Infect. Dis. 33(Suppl. 4), and practical application of animal models
1359–1366 (2005). S362–S366 (2001). for human diseases. Curr. Top. Microbiol.
15 Postema AS, Myers MG, Breiman RF. Immunol. 324, 1–24 (2008).
5 ICH guideline Q7. Good manufacturing
practice for active pharmaceutical Challenges in the development, licensure, 24 Lal G, Balmer P, Joseph H, Dawson M,
ingredients. 1–49 (2000). and use of combination vaccines. Clin. Borrow R. Development and evaluation of
Infect. Dis. 33(Suppl. 4), S261–S266 (2001). a tetraplex flow cytometric assay for
6 ICH guideline Q9. Quality risk quantitation of serum antibodies to
management. 1–19 (2005). 16 Gnanashanmugam D, Falkovitz-Halpern
MS, Dodge A et al. Shedding and reversion Neisseria meningitidis serogroups A, C, Y,
7 ICH guideline Q8. Pharmaceutical of oral polio vaccine type 3 in Mexican and W-135. Clin. Diagn. Lab. Immunol.
development. 1–9 (2006). vaccinees: comparison of mutant analysis 11(2), 272–279 (2004).
8 Dellepiane N, Griffiths E, Milstien JB. by PCR and enzyme cleavage to a real-time 25 Lal G, Balmer P, Stanford E, Martin S,
New challenges in assuring vaccine quality. PCR assay. J. Clin. Microbiol. 45(8), Warrington R, Borrow R. Development
Bull. World Health Organ. 78(2), 155–162 2419–2425 (2007). and validation of a nonaplex assay for the
(2000). Lu Z, Douthitt MP, Taffs RE, Ran Y, simultaneous quantitation of antibodies to
17
9 Lebron JA, Wolf JJ, Kaplanski CV, Norwood LP, Chumakov KM. nine Streptococcus pneumoniae serotypes.
Ledwith BJ. Ensuring the quality, potency Quantitative aspects of the mutant analysis J. Immunol. Methods 296(1–2), 135–147
and safety of vaccines during preclinical by PCR and restriction enzyme cleavage (2005).
development. Expert Rev. Vaccines 4(6), (MAPREC). PCR Methods Appl. 3(3), 26 Smith JF, Kowalski R, Esser MT,
855–866 (2005). 176–180 (1993). Brown MJ, Bryan JT. Evolution of
10 Capelle MA, Brugger P, Arvinte T. 18 Diphtheria vaccine (adsorbed). type-specific immunoassays to evaluate the
Spectroscopic characterization of In: Monograph 0443. European Pharmacopeia functional immune response to Gardasil: a
antibodies adsorbed to aluminium Commission (5th Edition). 660–661. Council vaccine for human papillomavirus types 16,
adjuvants: correlation with antibody of Europe, Strasbourg, France (2005). 18, 6 and 11. Hum. Vaccin. 4(2), 134–142
vaccine immunogenicity. Vaccine 23(14), (2008).
19 Schalk JA, de Vries CG, Jongen PM.
1686–1694 (2005). 27 van Gageldonk PG, van Schaijk FG,
Potency estimation of measles, mumps and
11 Jones LS, Peek LJ, Power J, Markham A, rubella trivalent vaccines with quantitative van der Klis FR, Berbers GA. Development
Yazzie B, Middaugh CR. Effects of PCR infectivity assay. Biologicals 33(2), and validation of a multiplex immunoassay
adsorption to aluminum salt adjuvants on 71–79 (2005). for the simultaneous determination of

236 Expert Rev. Vaccines 8(2), (2009)

 Quality-control issues & approaches in vaccine development Review

serum antibodies to Bordetella pertussis, 39 Appay V, van Lier RA, Sallusto F, Roederer Medicine: Meningococcal Vaccines.
diphtheria and tetanus. J. Immunol. M. Phenotype and function of human T Pollard AJ, Maiden MCJ (Eds). Humana
Methods 335(1–2), 79–89 (2008). lymphocyte subsets: consensus and issues. Press, NJ, USA 371–393 (2001).
28 Finne J, Leinonen M, Makela PH. Cytometry A 73(11), 975–983 (2008). 51 WHO. Recommendations for the
Antigenic similarities between brain 40 McElhaney JE, Xie D, Hager WD et al. T production and control of pneumococcal
components and bacteria causing cell responses are better correlates of conjugate vaccines. WHO Technical
meningitis. Implications for vaccine vaccine protection in the elderly. Report Series 927, Annex 2 (2005).
development and pathogenesis. Lancet J. Immunol. 176(10), 6333–6339 (2006). 52 EMEA. Committee for medicinal products
2(8346), 355–357 (1983). 41 Seder RA, Darrah PA, Roederer M. T-cell for human use (CHMP). Guideline on
29 Alaedini A, Latov N. Antibodies against quality in memory and protection: clinical evaluation of new vaccines. EMEA/
OspA epitopes of Borrelia burgdorferi implications for vaccine design. Nat. Rev. CHMP/SWP/164653/2005 (2006).
cross-react with neural tissue. Immunol. 8(4), 247–258 (2008). 53 European Pharmacopeia Commission. Edition
J. Neuroimmunol. 159(1–2), 192–195 42 US FDA. Guidance for Industry. PAT – A 6.3. Council of Europe, Strasbourg (2009).
(2005). framework for innovative pharmaceutical 54 WHO. Recommendations for diphtheria,
30 Arigita C, Kersten GF, Hazendonk T, development, manufacturing, and quality tetanus, pertussis and combined vaccines.
Hennink WE, Crommelin DJ, Jiskoot W. assurance. US Food and Drug WHO technical report series, No. 927,
Restored functional immunogenicity of Administration, 1–19 (2004). Annex 5 (2003).
purified meningococcal PorA by 43 Streefland M, van de Waterbeemd B, 55 Akkermans AM, Hendriksen CF,
incorporation into liposomes. Vaccine Happe H et al. PAT for vaccines: the first Marsman FR, van de Donk HJ.
21(9–10), 950–960 (2003). stage of PAT implementation for Evaluation and validation of a single-
31 Kim HW, Canchola JG, Brandt CD et al. development of a well-defined whole-cell dilution potency assay based upon
Respiratory syncytial virus disease in vaccine against whooping cough disease. serology of vaccines containing diphtheria
infants despite prior administration of Vaccine 25(16), 2994–3000 (2007). toxoid: analysis for consistency in
antigenic inactivated vaccine. Am. J. 44 Streefland M, van de Waterbeemd B, Kint production and testing. Dev. Biol. Stand.
Epidemiol. 89(4), 422–434 (1969). J et al. Evaluation of a critical process 86, 323–324 (1996).
32 Trotter CL, Andrews NJ, Kaczmarski EB, parameter: oxygen limitation during 56 Quality risk management. In: Eudralex
Miller E, Ramsay ME. Effectiveness of cultivation has a fully reversible effect on (Volume 4); EU guidelines to Good
meningococcal serogroup C conjugate gene expression of Bordetella pertussis. Manufacturing Practice Medicinal Products
vaccine 4 years after introduction. Lancet Biotechnol. Bioeng. 102(1), 161–167 for Human and Veterinairy Use. Annex 20
364(9431), 365–367 (2004). (2008). (2008).
33 Crotty S, Ahmed R. Immunological 45 Metz B, Jiskoot W, Hennink WE, 57 ICH Topic Q2. Validation of analytical
memory in humans. Semin. Immunol. Crommelin DJ, Kersten GF. procedures: text and methodology (2005).
16(3), 197–203 (2004). Physicochemical and immunochemical
58 US FDA. Guidance for Industry.
34 Kelly DF, Pollard AJ, Moxon ER. techniques predict the quality of diphtheria
Investigating out-of-specification (OOS) test
Immunological memory: the role of B cells toxoid vaccines. Vaccine 22(2), 156–167
results for pharmaceutical production. 1–14.
in long-term protection against invasive (2003).
US Food and Drug Administration (2006).
bacterial pathogens. JAMA 294(23), 46 Metz B, Jiskoot W, Mekkes D et al.
59 Shewhart WA. Statistical method from the
3019–3023 (2005). Quality control of routine, experimental
viewpoint of quality control. The Graduate
35 Clutterbuck EA, Oh S, Hamaluba M, and real-time aged diphtheria toxoids by in
School, the Department of agriculture,
Westcar S, Beverley PC, Pollard AJ. vitro analytical techniques. Vaccine
WA, USA (1939).
Serotype-specific and age-dependent 25(39–40), 6863–6871 (2007).
60 Meiring HD, Soethout EC, de Jong AP,
generation of pneumococcal 47 Fedson DS. Vaccine development for an
van Els CA. Targeted identification of
polysaccharide-specific memory B-cell and imminent pandemic: why we should worry,
infection-related HLA class I-presented
antibody responses to immunization with a what we must do. Hum. Vaccin. 2(1),
epitopes by stable isotope tagging of
pneumococcal conjugate vaccine. Clin. 38–42 (2006).
epitopes (SITE). Curr. Protoc. Immunol.
Vaccine Immunol. 15(2), 182–193 (2008). 48 EMEA. Committee for medicinal Chapter 16, Unit 16 13 (2007).
36 Nanan R, Heinrich D, Frosch M, Kreth products for human use (CHMP).
61 Wu XD, Kircher RA, McVerry PH,
HW. Acute and long-term effects of booster Guideline on strategies to idenitfy and
Malinzak DA. Characterization of
immunisation on frequencies of antigen- mitigate risks for first-in-human clinical
post-translationally modified recombinant
specific memory B-lymphocytes. Vaccine trials with investigational medicinal
protein using liquid chromatography/mass
20(3–4), 498–504 (2001). products. EMEA/CHMP/SWP/28367/07
spectrometry. Dev. Biol. (Basel) 103, 61–67
37 McMichael AJ, Gotch FM, Noble GR, (2007).
(2000).
Beare PA. Cytotoxic T-cell immunity to 49 Suntharalingam G, Perry MR, Ward S
62 Williams JN, Skipp PJ, Humphries HE,
influenza. N. Engl. J. Med. 309(1), 13–17 et al. Cytokine storm in a Phase 1 trial of
Christodoulides M, O’Connor CD,
(1983). the anti-CD28 monoclonal antibody
Heckels JE. Proteomic analysis of outer
38 Kaufmann SH, McMichael AJ. Annulling TGN1412. N. Engl. J. Med. 355(10),
membranes and vesicles from wild-type
a dangerous liaison: vaccination strategies 1018–1028 (2006).
serogroup B Neisseria meningitidis and a
against AIDS and tuberculosis. Nat. Med. 50 Farrington C, Miller E. Meningococcal lipopolysaccharide-deficient mutant. Infect.
11(4 Suppl.), S33–S44 (2005). vaccine trials. In: Methods in Molecular Immun. 75(3), 1364–1372 (2007).

www.expert-reviews.com 237
 Review Metz, van den Dobbelsteen, van Els et al.

63 Turula VE, Kim J, Michon F, Pankratz J, outer membrane protein containing vesicle 84 Metz B, Kersten GF, Baart GJ et al.
Zhang Y, Yoo C. An integrity assay for a vaccine. Vaccine 14(10), 1001–1008 Identification of formaldehyde-induced
meningococcal type B conjugate vaccine. (1996). modifications in proteins: reactions with
Anal. Biochem. 327(2), 261–270 (2004). 74 Lorent E, Bierau H, Engelborghs Y, insulin. Bioconjug. Chem. 17(3), 815–822
64 Mawas F, Ho MM, Huskisson R, Verheyden G, Bosman F. Structural (2006).
Saydam M, Corbel MJ. Physico-chemical characterisation of the hepatitis C 85 Davidsen J, Rosenkrands I, Christensen D
characterisation and immunogenicity of a envelope glycoprotein E1 ectodomain et al. Characterization of cationic liposomes
multi-valent candidate vaccine against derived from a mammalian and a yeast based on dimethyldioctadecylammonium
non-typeable Haemophilus influenzae and expression system. Vaccine 26(3), 399–410 and synthetic cord factor from
Moraxella catarrhalis. Vaccine 25(25), (2008). M. tuberculosis (trehalose
4801–4808 (2007). 75 McHugh CA, Tammariello RF, 6,6´-dibehenate)-a novel adjuvant inducing
65 Frech C, Hilbert AK, Hartmann G, Millard CB, Carra JH. Improved stability both strong CMI and antibody responses.
Mayer K, Sauer T, Bolgiano B. of a protein vaccine through elimination of Biochim. Biophys. Acta 1718(1–2), 22–31
Physicochemical analysis of purified a partially unfolded state. Protein Sci. (2005).
diphtheria toxoids: is toxoided then 13(10), 2736–2743 (2004). 86 Meyer BK, Ni A, Hu B, Shi L.
purified the same as purified then 76 Zheng Y, Lai X, Bruun SW et al. Antimicrobial preservative use in parenteral
toxoided? Dev. Biol. (Basel) 103, 205–215 Determination of moisture content of products: past and present. J. Pharm. Sci.
(2000). lyophilized allergen vaccines by NIR 96(12), 3155–3167 (2007).
66 Rich RL, Myszka DG. Spying on HIV spectroscopy. J. Pharm. Biomed. Anal.
with SPR. Trends Microbiol. 11(3), 46(3), 592–596 (2008).
124–133 (2003). 77 Lai X, Zheng Y, Sondergaard I et al.
Affiliations
67 Turbyfill KR, Kaminski RW, Oaks EV. Determination of aluminium content in • Bernard Metz
Immunogenicity and efficacy of highly aluminium hydroxide formulation by Research and Development Unit,
purified invasin complex vaccine from FT-NIR transmittance spectroscopy. Netherlands Vaccine Institute, Bilthoven,
Shigella flexneri 2a. Vaccine 26(10), Vaccine 25(52), 8732–8740 (2007). The Netherlands
1353–1364 (2008). bernard.metz@nvi-vaccin.nl
78 Lamb DH, Summa L, Lei QP. Capillary
68 Jiang G, Joshi SB, Peek LJ et al. Anthrax electrophoretic analysis of meningococcal • Germie van den Dobbelsteen
vaccine powder formulations for nasal polysaccharide–diphtheria toxoid conjugate Research and Development Unit,
mucosal delivery. J. Pharm. Sci. 95(1), vaccines. Dev. Biol. (Basel) 103, 251–258 Netherlands Vaccine Institute, Bilthoven,
80–96 (2006). (2000). The Netherlands
69 Amorij JP, Meulenaar J, Hinrichs WL et al. 79 Mellado MC, Franco C, Coelho A, • Cécile van Els
Rational design of an influenza subunit Alves PM, Simplicio AL. Sodium dodecyl Research and Development Unit,
vaccine powder with sugar glass sulfate-capillary gel electrophoresis analysis Netherlands Vaccine Institute, Bilthoven,
technology: preventing conformational of rotavirus-like particles. J. Chromatogr. A The Netherlands
changes of haemagglutinin during freezing 1192(1), 166–172 (2008). • Johan van der Gun
and freeze-drying. Vaccine 25(35), 80 Stewart-Tull DE, Bleakley CR, Galloway TS. Quality Control Unit, Netherlands Vaccine
6447–6457 (2007). Characteristics of Vibrio cholerae proteinases: Institute, Bilthoven, The Netherlands
70 Jedrzejas MJ, Lamani E, Becker RS. potential, candidate vaccine antigens. Vaccine • Lonneke Levels
Characterization of selected strains of 22(23–24), 3026–3034 (2004). Quality Control Unit, Netherlands Vaccine
pneumococcal surface protein A. J. Biol. 81 Delgado M, Yero D, Niebla O et al. Institute, Bilthoven, The Netherlands
Chem. 276(35), 33121–33128 (2001). Lipoprotein NMB0928 from Neisseria • Leo van der Pol
71 Salnikova MS, Joshi SB, Rytting JH, Warny meningitidis serogroup B as a novel vaccine Research and Development Unit,
M, Middaugh CR. Physical characterization candidate. Vaccine 25(50), 8420–8431 Netherlands Vaccine Institute, Bilthoven,
of Clostridium difficile toxins and toxoids: (2007). The Netherlands
effect of the formaldehyde crosslinking on 82 Khemiri A, Galland A, Vaudry D et al. • Nynke Rots
thermal stability. J. Pharm. Sci. 97(9), Outer-membrane proteomic maps and Registration and Medical Unit,
3735–3752 (2008). surface-exposed proteins of Legionella Netherlands Vaccine Institute, Bilthoven,
72 Belnap DM, McDermott BM Jr, pneumophila using cellular fractionation The Netherlands
Filman DJ et al. Three-dimensional and fluorescent labelling. Anal. Bioanal.
• Gideon Kersten
structure of poliovirus receptor bound to Chem. 390(7), 1861–1871 (2008).
Research and Development Unit,
poliovirus. Proc. Natl Acad. Sci. USA 97(1), 83 Smith TG, Lim JM, Weinberg MV, Netherlands Vaccine Institute, Bilthoven,
73–78 (2000). Wells L, Hoover TR. Direct analysis of the The Netherlands
73 Claassen I, Meylis J, van der Ley P et al. extracellular proteome from two strains of
Production, characterization and control of Helicobacter pylori. Proteomics 7(13),
a Neisseria meningitidis hexavalent class 1 2240–2245 (2007).

238 Expert Rev. Vaccines 8(2), (2009)

View publication stats