WORKING PAPER

RESEARCH FOR THE DEVELOPMENT

OF AN INFLUENZA PANDEMIC VACCINE
Working Group 3 of the Consultation for the development of a global action plan to
increase supply of influenza pandemic vaccines, WHO, Geneva, 2-3 May 2006

Introduction
Vaccination is considered a key component of the strategy to respond to an influenza
pandemic, with the aim to reduce morbidity and mortality, and to curb the spread of
the disease
Currently available inactivated vaccines are effective in the majority of populations
when they contain antigens that are well matched with those in the circulating viral
strains. A modification of vaccine composition is necessary almost every year to
follow antigenic changes of circulating viruses and annual immunization is
recommended for groups decided to be at increased risk of severe disease. While
considerable effort is being made to develop and evaluate candidate H5N1 vaccines
based on existing technologies, there is a need for the development of vaccines that
can overcome some of the limitations of the existing candidates. In 2004, the WHO
Initiative for Vaccine Research established a research programme with the following
objects: the development of new influenza vaccines that induce broad-spectrum and
long-lasting immune responses and to coordinate research projects that will contribute
to the acceleration of the development of influenza pandemic vaccines.

Generation of new vaccines

The main objective is to develop safe and effective influenza vaccines that induce
broad-spectrum (cross protection within subtypes and/or heterosubtypic immunity),
long-lasting protection against influenza and its sequelae preferably with a single
dose. .
Vaccines that induce broad immunity would probably not have to be updated
annually, since they would protect against viruses with antigenically distinct HA
molecules. This would allow the preparation of vaccine lots in advance and would
make the logistics for vaccine production easier. The development of some vaccines
may also result in improved efficacy in certain high-risk groups, such as the elderly.
There are several strategies that can be used for the development of such broadly
protective vaccines.

Live attenuated vaccines

A review of the data on live attenuated vaccines (LAIV) demonstrates a history of
safety, protection against homologous virus and minor variants, evidence of herd
immunity through vaccination of children and the early onset of protection possibly
due to an interferon response. There is some evidence that LAIV might be more
efficacious than inactivated vaccines. In animal studies it was shown that LAIV
induced broad immunity against various variants of the H5 subtype and further
evaluation of candidate pandemic vaccines for H5 and H9 viruses in humans is now in
progress. However the safety and efficacy of LAIVs in patients with asthma, the
immunocompromised, the very young or elderly awaits confirmation. It is unclear,

1
whether pre-pandemic use of live vaccines containing novel HA subtypes could
represent a risk of generating undesirable reassortants with pandemic potential. Also,
the use of reassortant DNA technology to prepare LAIV presents regulatory issues in
some countries.

Vaccines that target conserved influenza virus proteins

Some viral proteins like the matrix proteins M1 and M2 and the nucleoprotein (NP)
are relatively conserved and are considered potential targets for the induction of broad
protective immunity. Studies in animals have shown that experimental vaccines based
on these proteins can induce cross-protective immunity against different subtypes of
human influenza A viruses.
M2 protein
The M2 protein is expressed on the membrane of infected cells and protection
induced by this protein is probably mediated through antibody dependent cell
mediated cytotoxicity (ADCC). A number of approaches have been used to enhance
immune responses to M2, including fusion with various proteins and delivery using
viral vector or in virus-like particles. In animals, M2-based vaccines induced broad
protection against different subtypes of influenza A virus. However, M2 is not present
in influenza B viruses and will thus not contribute to protective immunity against this
group of viruses. In addition, while the sequence of the M2 protein is highly
conserved between influenza A viruses, it demonstrates species specificity. Further
studies are required to determine whether an M2-based vaccine could be sufficient to
protect against severe influenza and death or whether they should be considered as a
supplement to existing inactivated vaccines.
M1 and NP proteins
M1 and NP are major targets for virus-specific cytotoxic T lymphocytes (CTL) and
considered candidates for the induction of cross-protective immunity. Indeed it has
been shown in animal models and humans that virus-specific CTL contribute to
protective immunity against influenza virus infection. NP-based vaccines provided
protection against different subtypes of influenza A viruses in mice and NP-
expressing plasmid vaccines induced virus specific CTL and protective immunity
against challenge infection. For the induction of CTL responses it is essential that the
protein of interest is processed in antigen presenting cells via the endogenous route of
antigen processing. The required delivery of protein into the cytoplasm can be
achieved with live viral vector, allowing de novo synthesis of protein in the
cytoplasm, or the use of specialized antigen delivery forms.

Mucosal application of inactivated vaccines

Alternative routes of vaccination include the mucosal delivery, which has the
potential advantages of being non invasive and inducing both local and systemic
immune responses. Mucosal vaccination with live or inactivated vaccines is better for
the induction of local IgA responses in the upper respiratory tract, which contributes
to protective immunity and might be useful in strategies using vaccine to limit
transmission of influenza. In addition, it has been suggested that IgA antibody
responses are more broadly reactive than IgG antibody responses. The potential value
for aerosol vaccination that would deliver vaccine to both the upper and lower
respiratory tracts is being assessed in some studies.

2
New adjuvants
The development of new inactivated pandemic vaccines and especially those based on
purified antigens has been hampered for a long time by low immunogenicity of
antigens and by the lack of safe and efficacious adjuvants. Many studies are in
progress to identify appropriate immune-stimulating substance and their advantages
and drawbacks becomes more apparent. There is need to have more information on
mechanisms of action of particular adjuvants because it could open the way for
combining some of them to reach synergistic effects on immune responses and
eliminate adverse reactions.

Alternative sources of HA antigen

Traditionally, influenza vaccines are produced in embryonated chicken eggs, which is
not conducive to up scaling of production volume. In addition, the production
capacity of the vaccine manufacturers is limited and insufficient to allow a global
vaccination campaign and an increase in global production capacity and investment in
alternative production systems is very desirable. Recently, cell culture systems have
been approved for vaccine production, may which allow flexible vaccine production
and upscaling. In addition, recombinant HAs expressed by baculoviruses are
currently being evaluated and these preparations induced antibody responses and
protective immunity comparable to that induced by conventional vaccines. Recent
studies in mice have shown that an adenovirus-based vaccine harbouring expression
cassettes for H5N1 antigen provided protection against challenge with homologous
virus. Caution is required in interpreting these results as history has shown that mouse
data does not always predict the performance of a vaccine in humans.

Regulatory issues
There are a range of regulatory issues in gaining acceptance of new influenza
vaccines, many of these defined in existing WHO Technical Reports and other
documents. Greater international consistency in licensing requirements is desirable.
WHO can facilitate coordination among regulatory authorities to develop harmonized
procedures for vaccine registration.

Immunological Issues

Improvement of priming of immune responses

A current concern is that it may be difficult to prepare effective vaccines containing
HA molecules of a subtype to which no immunity exists in the general population.
Further studies should help to identify methods to effectively induce and measure
priming. There is a great demand for adjuvant systems that could improve the primary
immune responses induced by the current or future influenza vaccines and that are
safe.

Immunological mechanisms of protection, laboratory correlates of protection

Traditionally, hemagglutination-inhibiting (HI) antibodies have been used as a
correlate of protection and titers of > 40 are considered protective. For novel subtypes
like H5, HI assays have proven difficult and virus neutralization assays (VN) are
being performed although the levels that correlate with protection have not been
established. Since cell mediated immunity is also a component of protection and can
be induced by certain types of vaccines (such as LAIV which generally produce
modest HI or VN responses), the demonstration of virus-specific T-cell mediated

3
effector function is desirable. Further vaccination-challenge experiments and host-
virus interaction with regard to antibody and cell- mediated immunity, may help to
define additional correlates of protection. An important and urgent subject for further
exploration is whether correlates for conventional vaccines will be applicable for the
evaluation of new subtypes vaccines such as H5N1.

Improvement and standardization of immunological assays for evaluation of new

influenza vaccines
As indicated above, VN assay is an alternative for the HI assay and is more sensitive
than the HI assay for the detection of antibodies against the H5N1 subtype and an
international collaborative study is in progress to evaluate VN assay methods and to
develop a standardized protocol which may allow assessing the utility of VN antibody
titers as a correlate of protection against infection/disease. Several techniques are
available to measure T cell responses. However, most methods are complex, time
consuming and require expensive equipment and reagents and are also hard to
standardize. New assays are under development.

Questions/issues for discussion

• What targets should be set for the performance of new vaccines with broader
and longer lasting immunity? e.g., duration of protection or extent of
protection against drift mutants;
• What studies should be recommended for the development of influenza
vaccines using the following strategies:
- conserved viral proteins (M2, M1, NP) as vaccines
- mucosal delivery of inactivated vaccines
- adjuvants and alternative sources of viral antigen;
• What studies should be done in the area of live attenuated vaccines;
• Are there any others technologies, that should be explored;
• What studies are recommended to improve our knowledge on mechanisms of
immunological protection and establish laboratory correlates of protection
against influenza viruses and what strategies should be envisaged to measure
immunogenicity in context of licensure of vaccines;
• What studies should be recommended to improve immunological priming as
the way for the development of effective one dose vaccine;
• Are there any other immunological issues to be considered in relation to the
development of new vaccine;
• What research studies should be recommended as short-, mid-, and long term
priorities of Global Action Plan.

4
 Summary on strategies to develop new influenza vaccines

vaccine timelines Potential advantages challenges potential for scaling up

technologies until production (500 million doses
license per year)
live attenuated licensed in not injectable, induce local and safety for asthmatic,
two systemic immune responses; immunocompromised, infants
countries improved protection against drift and elderly need to be confirmed;
mutants efficacy is elderly unknown;
possibility of pre-pandemic
distribution unclear;

conserved 3-5 years provide broad protection against M2 demonstrates species Scale-up for M2-based vaccines
proteins (M1, M2, different subtypes in animals specificity, safety and sufficient relatively straight forward.
NP) immunity in man should be
confirmed
mucosal 3-5 years not injectable , induce local and need safe adjuvants
administration of systemic immune responses
inactivated which are broadly protective
vaccines
proteins 2-3 years may be easier to produce poorly immunogenic in naïve high
(baculovirus subjects
system)
virus (adeno) - 5-10 years good inductor of T cell immune difficulty to boost immune
vectored proteins responses responses, vector immunity
(HA)

5
 Summary on new strategies to develop pandemic influenza vaccines (… continued)

DNA vaccine 5-10 years immunogenic in immune proof of safety and efficacy of
humans, induce T cell immune DNA vaccines in human not yet
responses, easy to produce established
epitope-based 5-10 years demonstrate potential for study in other animal models are
vaccine generation broad immune requested to confirm safety and
responses in mice immunogenicity, haplotype
restriction

6
7