Vaccine 21 (2003) 1776–1779
The annual production cycle for influenza vaccine
Catherine Gerdil∗
Aventis Pasteur S.A., Campus Mérieux 1541, Av. Marcel Mérieux, F-69280 Marcy l’Etoile, France
Abstract
Influenza disease impacts every year up to 10% of the world’s population, i.e. up to 500 millions of people. The international surveillance
network organised by WHO allows the detection of important antigenic variation of the virus in humans but also animals (especially birds
and pigs). Epidemiological data are summarised twice a year in two meetings (one in February, the other in September) allowing WHO to
propose new recommendations for vaccines formulations for each northern and southern hemisphere influenza season annual vaccination.
According to such recommendation, vaccine manufacturers proceed each year with two different vaccine production campaigns in order to
match the vaccine composition with global epidemiological surveillance data and so, develop each time an updated vaccine formulation.
To date within 6 months production periods, almost 250 million of doses are brought annually on market in more than 100 countries.
In such a tight schedule, the annual production requires careful coordination of a highly complex process involving both public health
laboratories and vaccine companies, in order to provide on time safe and effective vaccines. Major steps of this vaccine production process
including the milestones and the bottlenecks need to be detailed in order to well understand difficulties and hurdles towards increase of
global capacity and introduction of new vaccines.
© 2003 Elsevier Science Ltd. All rights reserved.
Keywords: Influenza vaccine; Annual production cycle; Acute respiratory illness
Key messages
The WHO international surveillance system identifies
new strains of influenza virus that need to be included
in influenza vaccines.
The twice yearly, 6-month vaccine production cycle
includes:
• Preparing high-growth reassortants for influenza A
(H3N2) and A (H1N1).
• Growing vaccine virus in embryonated eggs.
• Inactivating, purifying, splitting, formulating and filling the vaccine.
• Testing its immunogenicity and safety in clinical studies (Europe only).
Currently, approximately 250 million doses of inactivated influenza vaccine are produced each year.
1. Introduction
The acute respiratory illness known as influenza appears
to have afflicted humans since ancient times. Hippocrates
∗
Tel.: +33-4-3737-3365; fax: +33-4-3737-9967.
E-mail address: [email protected] (C. Gerdil).
described an influenza-like epidemic in 412 BC and numerous similar episodes have been described since then. They
were once widely believed to be the result of adverse astrological alignments or other occult influences. In 1933, a
virus was identified as the causative agent by Smith et al.
[1]. The first inactivated influenza vaccines were introduced
in the 1940s. Since then, vaccine companies have developed
expertise in a safe and effective vaccine manufacturing process, allowing production of an increasing number of doses
of the influenza vaccine each year. This article briefly reviews this production process.
2. WHO epidemiological surveillance
The hemagglutinin (HA) and neuraminidase (NA) surface glycoproteins of influenza viruses change frequently
as a result of genetic mutation. For this reason, the World
Health Organization (WHO) coordinates an international
surveillance system to monitor the epidemiology of influenza viruses [2]. This system, established in 1947, allows
for the detailed analysis of circulating influenza viruses isolated from both humans and animals, especially birds and
pigs, and is able to detect newly evolved antigenic variants
of the influenza A (H3N2 and H1N1) and B strains to which
human populations are likely to be susceptible. For surveillance in human populations, sentinel physicians obtain
0264-410X/03/$ – see front matter © 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0264-410X(03)00071-9
C. Gerdil / Vaccine 21 (2003) 1776–1779
1777
Fig. 1. Influenza vaccine production timetable.
nasopharyngeal swabs from patients with influenza-like illness. The samples are sent to the National Influenza Centers
(110 centers in over 80 countries) for primary isolation and
identification. When a new virus strain is detected, it is sent
for detailed antigenic and molecular analysis to one or more
of the four Collaborating Centers for Influenza Reference
and Research located in London, Atlanta, Melbourne and
Tokyo.
Twice yearly, the surveillance data are carefully reviewed
by WHO Collaborating Center investigators. Epidemiological surveillance has shown that in temperate climates influenza viruses circulate primarily during the winter months.
The virus strains circulating in the Northern hemisphere are
reviewed in February in order to determine which variants
are likely to be the cause of human disease the following
winter. Likewise, viruses circulating in the Southern hemisphere are reviewed each September. Based on these reviews, the WHO experts decide which variants should be
included in the next season’s influenza vaccine anticipating
an accurate cross match between vaccine composition and
circulating strains. From the time these decisions are made,
vaccine companies have approximately 6 months to produce
their vaccines and deliver them to health care providers.
factors govern the vaccine production timetable. First, manufacturers must undertake two production cycles each year;
one each for the Northern and Southern hemispheres. Second, timing the delivery of vaccine is crucial because annual
vaccination programs are conducted during the few months
preceding the expected circulation period of the influenza
virus. Third, because all doses of influenza vaccine are produced in embryonated eggs, the number of eggs needed to
produce the number of doses expected by the market must
be anticipated well in advance of beginning the production
cycle. Every step in the process must be carefully planned
with respect to both timing and production capacity (Fig. 1).
In addition, for European vaccine manufacturers, whenever
WHO recommends that a new strain be included in the vaccine, a revalidation of the manufacturing process for the new
strain must be undertaken for regulatory authorities. Submission of this variation file must include not only a validation
of the manufacturing process but also an evaluation of the
immunogenicity and safety of the new vaccine formulation.
4. Preparing seed strains, producing the vaccine and
getting it licensed
3. Vaccine manufacturing timetable
4.1. Characterizing the strains used for vaccine
production
Once the virus strains to be included in next season’s
vaccine have been determined, candidate high-growth seed
strains must be prepared by the WHO Collaborating Centers,
tested by the manufacturers and put into production. Several
Currently, two subtypes of influenza A virus (H3N2 and
H1N1) and one strain of influenza B virus are responsible
for outbreaks of human disease and are, therefore, included
in influenza vaccines. During the period since influenza A
1778
C. Gerdil / Vaccine 21 (2003) 1776–1779
(H3N2) viruses re-emerged in 1968, the A (H3N2) strains
chosen for the vaccine have been changed twice as often
as have the strains for the other two influenza viruses. During the 21-years period from 1980/1981 to 2001/2002, 14
strains changes were made in the vaccine formulation for influenza A (H3N2), eight for influenza A (H1N1) and 10 for
influenza B.
4.2. Preparing reassortant viruses as seed strains for
vaccine production
Since the early 1970s, the influenza A seed strains used
for vaccine production have been prepared by genetic reassortment using the field strains chosen by the WHO experts and an A/PR8/34 or PR8-like master strain that grows
to high titer in embryonated eggs [3]. Each of these strains
is closely similar antigenically to the prototype wild virus
strain and is isolated in embryonated eggs, a step that is believed to reduce the possibility of contamination by extraneous infectious agents also carried in the human respiratory
tract. These high-growth reassortants are tested to confirm
the absence of genes coding for PR8 or PR8-like surface glycoproteins. Seed strains for producing influenza B vaccines
are field isolates because no master B strain has yet been
found that improves the growth performance of influenza B
viruses in egg-based production systems.
4.3. Standardizing the seed strains for vaccine
production
Once the candidate seed strains for vaccine production
have been prepared, their surface antigens undergo identity testings and sequence analyses by WHO Collaborating
Centers in order to confirm their similarity to the reference
strains. Successful candidates are then sent to each vaccine
producer in order to evaluate their suitability for vaccine
production. This includes satisfactory yields when grown in
embryonated eggs and antigenic stability throughout serial
passage in eggs as well as the inactivation and purification
processes. The results obtained by the individual producers
and the official WHO laboratories are reconciled in several
meetings held in Western Europe (the European Medicines
Evaluation Agency (EMEA)), the US (the Food and Drug
Administration (FDA)) and Australia (XXX (TGA)). These
meetings are usually held 1 month after the initial WHO
decision on vaccine formulation.
4.4. Producing each vaccine component
Once the production strain for each vaccine component
has been selected, bulk vaccine production can begin. All
seed lots are qualified according to official requirements
and internal specifications [4,5]. Vaccine virus is grown
in the allantoic cavity of embryonated eggs, harvested, inactivated by formalin or ␤-propriolactone and purified by
ultra-centrifugation. The virus particles are then either split
using ether and a detergent or solubilized using a detergent
alone. European regulatory authorities consider the inactivation and splitting of virus particles critical steps in the
manufacturing process and require that they be validated for
each new strain introduced into the vaccine. Each producer
has to demonstrate that these processes have no impact on
the HA and NA antigens and each has to confirm by identity
testing the similarity of each antigen at each of these critical
steps. Stability studies are initiated on each new formulation
for every vaccine presentation (e.g. 10-dose vial, ampoule
or single-dose syringe). All production data are summarised
in the pharmaceutical variation file submitted for regulatory
authorisation each year. This usually occurs approximately
5 months after production begins (i.e. in June for the Northern hemisphere).
4.5. Quantifying antigen content
One critical bottleneck in the production cycle is quantifying the new antigens. This comes late in the production
process because it requires international reagents specific
to the selected seed with international calibration. These
reagents usually become available 3 months after WHO selects the candidate vaccine strains. HA antigen is quantified
by the single radial immuno-diffusion assay using standard antigens and specific sheep antiserum [6]. Once this
has been accomplished, the manufacturers have a precise
way of measuring the concentration of each HA antigen
produced and are able to begin the formulation and filling
steps.
Clinical studies are required by European regulatory
authorities. In Western Europe, the European Medicines
Evaluation Agency (EMEA) requires that the safety and
immunogenicity of each new influenza vaccine formulation
be evaluated in a clinical study [7]. The study is usually
completed by mid-July and satisfactory results are needed
in order for a market authorization to be obtained. The
study is performed in two groups of healthy volunteers:
one, 18–60 years in age and, the other, 60 years in age and
older. It must be undertaken within very tight time lines:
the study protocol must receive ethical committee approval
beforehand, recruitment of subjects must be accomplished
quickly and serological studies and their statistical analyses must be performed with dispatch. The immunogenicity
of one dose of vaccine is assessed for each of its three
antigenic components by hemagglutination-inhibition (HI)
and/or single radial hemolysis (SRH) tests. The study results must meet EMEA criteria and the data must be quickly
incorporated into the Clinical Expert report that is submitted to the reference Member State. These requirements are
uniquely European; for example no clinical study or process
validation for producing each new vaccine formulation is
required by regulatory authorities in the United States and
no clinical studies are required for Southern hemisphere
vaccines.
C. Gerdil / Vaccine 21 (2003) 1776–1779
1779
5. Conclusions
References
The production of influenza vaccines requires careful coordination of a highly complex process involving a wide
range of technical expertise in both public health laboratories and vaccine companies in order to provide on time safe
and effective vaccines. Its remarkable success is self-evident;
within two 6-month production periods, almost 250 million
doses are brought to market in more than 100 countries each
year [8]. The process of producing influenza vaccine can be
expected to become even more complex in the years ahead.
The use of reverse genetics to prepare vaccine seed strains
and the introduction of cell culture production systems will
introduce new safety and regulatory issues. Newer inactivated vaccines that contain adjuvants or new preservatives,
the introduction of live-attenuated vaccines and new inoculation routes will add to the complexity of both vaccine
production and standardization. The accelerating demand
for influenza vaccines throughout the world will challenge
the technical capacity, financial strength and strategic vision of vaccine companies. This challenge, however, will be
matched by the challenge to public health officials of effectively using these vaccines in the control of influenza.
[1] Smith W, Andrewes CH, Laidlaw PP. A virus obtained from influenza
patients. Lancet 1933;2:66–8.
[2] Hampson AW, Cox NJ. Global surveillance for pandemic influenza:
are we prepared? In: Brown LE, Hampson AW, Webster RG, editors.
Options for control of influenza, Part III. Amsterdam: Elsevier; 1996,
p. 50–9.
[3] Kilbourne ED, Shulman JL, Schild GS, Schoer G, Swanson J, Busher
D. Correlated studies of a recombinant influenza virus vaccine, Part
I. Derivations and characterization of virus and vaccine. J Infect Dis
1971;124:449–62.
[4] European Agency for Evaluation of Medicinal Products: note for
guidance and harmonization of requirements for influenza vaccines.
CPMP/BWP/214/96; 12 March 1997.
[5] Vaccinum influenza inactivation ex virosum fragmentis preparation.
European Pharmacopeia: 0158.
[6] Schild GC, Wood JM, Newman RW. A single radial-immunodiffusion
technique for the assay of influenza hemagglutinin antigen. Bull World
Health Org 1975;52:223–30.
[7] Wood JM, Lewandowski RA. The influenza vaccine licensing process.
Vaccine 2003; in press.
[8] Van Essen GA, Ambrosch F, Forleo E, Palache AM, Fedson DS.
Influenza vaccination in 2000: vaccination recommendations and
vaccine use in 50 developed and developing countries. Vaccine 2003;
in press.