Antigenic Variation of Influenza Viruses Infection
B. Antigenic Variation
The explosive nature of epidemic influenza and the specific
clinical features of this disease have given reliable
epidemiological records of this infection since the beginning of
the nineteenth century. Several epidemics were recorded during
the nineteenth century but the first pandemic was not accurately
recorded until 1889-92. A second pandemic, probably originating
in Europe, occurred in 1918-19, and is known as Spanish
Influenza. It is responsible for 20-25 million deaths,
principally in young adults. It was suggested that this strain
had unusual virulence. Alternatively, the large numbers of deaths
may be due to the debilitating conditions as a result of the
First World War. In fact, the number of deaths recorded may only
be a fraction of the true number.
Pandemics continued to occur regularly after the Spanish
influenza, in 1932-33, 1947-48, 1957 and 1968. The next pandemic
is thought to be overdue. These latter pandemics resembled the
pandemic of 1890, affecting millions of people with a mild URTI
and a small number of deaths. The H1N1 (swine) viruses probably
appeared in 1918 and continued to circulate until 1957, at which
time they were supplanted by the H2N2 (Asian) viruses. The H2N2
viruses were prevalent until 1968, when H3N2 (Hong Kong) strains
appeared. The H1N1 virus reappeared in 1977 and did not replace
the H3N2 subtype and both subtypes continued to cocirculate.
1. Antigenic Shift
The recorded patterns of influenza A infection contain 2
phenomena; the first being the almost identical annual epidemics
which occur in most countries, and the second are the extensive
pandemics which occur approximately every 10 - 12 years. Analysis
of virus isolates since 1933 showed that viruses isolated in the
years 1933-46, 1947-56, 1957-67, and from 1968 onwards
demonstrated wide variation. Virus specific sera raised from
ferrets did not cross react in HAI tests. It is apparent that
pandemics are due to the appearance of new influenza A subtypes
against which the population has no immunity. This phenomenon is
known as antigenic shift. As immunity to the new subtype builds
up, further epidemics are more limited. The appearance of a new
influenza virus subtype is paralleled by the disappearance of the
old subtype (an exception occurred in recent times, when 2 virus
subtypes have circulated concurrently) The HA antigen is always
involved in antigenic shift as it is responsible for eliciting
virus-neutralizing antibodies. The neuraminidase may be affected
as well. The origin of antigenic shift has been a subject of
intensive research but has yet to be resolved. There are 3
theories as to how antigenic shift arise and they are not mutally
exclusive: there is evidence for all 3 theories in past pandemics
Reassortment - the most widely held view is that
the new virus subtypes are reassortant viruses resulting
from double infection, so that that 8 RNA segments of
each strain reassortment with each other, producing a new
virus. As the appearance of the new subtype is paralleled
by the disappearance of the old subtype, it is unlikely
that the dual infection was by 2 human influenza A
subtypes. However, influenza A viruses infect other
species of animals, such as horses and birds. It was
postulated that dual infection with human and animal or
bird viruses could result in the production of a
reassortant virus. Indeed, such reassortant viruses have
been produced in the laboratory from human and animal
parents. Influenza A viruses can cross species barrier.
The pig is postulated as the most likely "mixing
vessel" as it can be infected by both human and
avian viruses. Such an event is most likely in occur in
SE Asia, in particular China, where humans and animals
live in close proximity. There is no pretty good evidence
that such an event occurred in 1947 when H1N1 was
replaced by H2N2, and in 1968 when H2N2 was replaced by
H3N2.. Influenza B viruses do not occur in animals and do
not exhibit antigenic shift. This has been put forward as
indirect evidence for recombination as the mechanism for
the emergence of new influenza A subtypes.
Recirculation of existing subtypes - It has also
been suggested that there are a limited number of
influenza A subtypes which are recycled in the human
population. Evidence for this theory comes from
seroepidemiological studies of antibody to influenza
viruses in sera taken at different times from subjects of
different ages. Antibody to a human influenza A subtype
was often found in the sera of elderly persons taken
years before the appearance the appearance of the same
subtype as a cause of pandemic infection. It was
suggested that all influenza A subtypes exist in nature,
and emerge when the antibody status of the population has
fallen to levels which allow pandemic infection: a cycle
of approximately 70 years. However, the evidence
supporting this theory is very fragile. In 1977 H1N1
reappeared which was very similar to the strain which
circulated before 1957. It was widely believed that it
may have escaped from a laboratory. As a sizable
proportion of the population had already been exposed to
the H1N1 virus before 1957, it did not cause a pandemic.
Gradual adaptation of animal viruses to human
transmission - A third mechanism for antigenic shift
is the gradual adaptation of avian viruses to human
transmission. There is now evidence that this might have
happened in the 1918 pandemic: that the pandemic virus
was directly descended from an avian ancestor.
Reassortment of human H2 with avian H3 virus: there is
strong evidence that this occurred with the emergence of the
H3N2 pandemic virus in 1968.
2. Antigenic Drift
In addition to the large pandemics due to antigenic shifts for
influenza A viruses seen every 10 - 12 years, smaller epidemics
occur regularly in the intervening years. The viruses isolated
from such epidemics showed strain differences when compared in
the HAI tests ie. although the viruses belong to the same
subtype, they do not cross react completely. These lesser
antigenic changes are known as antigenic drift. Antigenic drift
is thought to arise through natural mutation, and selection of
new strains takes place by antibody pressure in an immune or
partially immune population. Epidemics due to new virus strains
arising due to antigenic drift is not as great as for those
showing antigenic shift, since partial immunity is present in
persons with cross- reacting antibody induced by previous
infection.
C. Pathogenesis
Virus infection is spread via respiratory droplets. The virus
particles binds to cells of the respiratory epithelium which are
rich in viral receptors. Neuraminidase present on the virus
particles aid the infectious process by releasing virus particles
which have been bound by the mucous present on the surface of
epithelial cells. Because of the generalized symptoms present,
viraemic spread form the respiratory tract has been suspected,
although there is no conclusive evidence.
D. Clinical Features
Following a typical incubation period of 48 hours, the typical
symptoms of influenza appears. The onset is abrupt with a marked
fever, headache, photophobia, shivering, a dry cough, malaise,
myalgia, and a dry tickling throat. The fever is continuous and
lasts around 3 days. Influenza B infection is similar to
influenza A, but infection with influenza C is usually
subclinical or very mild in nature.
Complications
1. Tracheobronchitis and bronchiolitis - A small
proportion of patients develop more sever respiratory
symptoms where rales and rhonchi are heard but the chest
is radiologically clear. These symptoms are more commonly
seen in the elderly and patients with COAD.
2. Pneumonia - primary viral pneumonia or a
secondary bacterial pneumonia may develop. Primary viral
pneumonia is relatively uncommon, but cases have been
demonstrated in many influenza epidemics. It may occur in
previously young and healthy persons, but are commonly
associated with patients with preexisting cadiovascular
disease such as Rheumatic fever. Secondary bacterial
pneumonia is more common than primary viral pneumonia. It
was speculated that the high incidence of deaths in young
people during the Spanish influenza pandemics of
1917-1918 may have been due to secondary bacterial
pneumonia in a population generally debilitated by the
effects the WWI.
Secondary bacterial pneumonia - usually
occurs late in the course of disease, after a period of
improvement has been observed for the acute disease. The
symptoms and signs are that of a typical bacterial
pneumonia. S. aureus is most commonly involved
although S. pneumoniae and H. influenzae
may be found. There appears to be a good reason why S.
aureus is so commonly found in cases of secondary
bacterial pneumonia. Infection of cells by influenza A
requires cleavage of the virus haemagglutinin by
proteases, and some strains of S. aureus produces
such enzymes. Thus S. aureus and influenza may
promote infection by the other. Influenza A by damage to
the healthy respiratory epithelium.
Myositis and myoglobinuria - In addition to
myalgia, which is characteristic of acute influenza
infection, clinical myositis and myoglobinuria may occur.
Reye's syndrome - Reye's syndrome is characterized
by encephalopathy and fatty liver degeneration. The
disease has a 50% mortality amongst hospitalized cases
and had been associated with several viruses; such as
influenza A and B, Coxsackie B5, echovirus, HSV, VZV, CMV
and adenovirus.
Other complications - influenza infection have
been implicated in acute viral encephalitis and
Guillain-Barre syndrome. Influenza A was also associated
with the cot death syndrome.
E. Laboratory Diagnosis
During epidemics, a presumptive diagnosis can be made on the
basis of the clinical symptoms. However, influenza A and B can
co-circulate, and mixed infections of influenza and other viruses
have been reported. Isolated cases of suspected influenza should
be investigated for these may represent the first cases of an
impending epidemic.
Virus Isolation - Throat swabs, NPA and nasal
washings may be used for virus isolation. It is reported
that nasal washings are the best specimens for virus
isolation. The specimen may be inoculated in embryonated
eggs or tissue culture. 10-12 day embryonated eggs are
used for virus isolation. The specimen is inoculated into
the amniotic cavity. The virus replicates in the cells of
the amniotic membrane and large quantities are released
back into the amniotic fluid. After 2-3 days incubation,
virus in the amniotic fluid can be detected by adding
aliquots of harvested amniotic fluid to chick, guinea
pig, or human erythrocytes. Pathological specimens can be
inoculated on to tissue cultures of kidney, chicks or a
variety of other species. Rhesus monkey cells are the
most sensitive. Although no CPE is produced, newly
produced virus can be recognized by haemadsorption using
the cells in the tissue culture, and haemagglutination
using the culture medium which contains free virus
particles. Influenza B virus and occasionally influenza A
will produce a CPE in MDCK cells. Influenza viruses
isolated from embryonated eggs or tissue culture can be
identified by serological or molecular methods. Influenza
viruses can be recognized as A, B, or C types by the use
of complement fixation tests against the soluble antigen.
(A soluble antigen is found for all influenza A, B or C
type virus but antibody against one type does not cross
react with the soluble antigen of the other. The further
classification of influenza isolates into subtypes and
strains is a highly specialized responsibility of the WHO
reference laboratories. The HA type is identified by HAI
tests, the NA type is also identified.
Rapid Diagnosis by Immunofluorescence - cells from
pathological specimens may be examined for the presence
of influenza A and B antigens by indirect
immunofluorescence. Although many workers are convinced
of the value of this technique, others have been
disappointed with the specificity of the antisera and the
level of background fluorescence that makes the test
difficult to interpret. EIA tests for the detection of
influenza A viral antigens are available that are easier
to interpret than immunofluorescence. PCR assays for the
detection of influenza RNA have also been developed but
there usefulness in a clinical setting is highly
questionable.
Serology - Virus cannot be isolated from all cases
of suspected infection. More commonly, the diagnosis is
made retrospectively by the demonstration of a rise in
serum antibody to the infecting virus. CFT is the most
common method used using the type specific soluble
antigen. However, the CF test is thought to have a low
specificity. A more specific test is the HAI test.
Infection by influenza viruses results in a rise in serum
antibody titre, but the requirement for a 4-fold or
greater rise in titre of HI of CF antibody reflects the
inaccuracy of these tests for detecting smaller increases
in antibody. A more precise method for measuring antibody
is by SRH. SRH is more sensitive than CF or HAI tests and
has a greater degree of precision. A 50% increase in zone
area represents a rise in antibody and is evidence of
recent infection. Sera do not have to be pretreated to
remove non-specific inhibitors which plaque the HAI test.
SRH may well replace CF and HAI tests in diagnostic
laboratory in future.
F. Treatment
Influenza epidemics are responsible for massive disruption to
industry, and for a significant number of deaths, particularly in
the elderly and the very young. At present, treatment of
influenza is entirely symptomatic. Salicylates should be avoided
in children because of the link with Reye's syndrome. 2
compounds, amantidine and ribavirin, with antiviral activity
against influenza have been identified and may be of value.
Amantidine - this compound inhibit the growth of
influenza viruses in cell culture and in experimental
animals. Amantidine is only effective against influenza
A, and some naturally occurring strains of influenza A
are resistant to it. The mechanism of action of
amantadine is not known. It is thought to act at the
level of virus uncoating. The compound has been shown to
have both therapeutic and prophylactic effects.
Amantidine significantly reduced the duration of fever
(51 hours as opposed to 74 hours) and illness. The
compound also conferred 70% protection against influenza
A when given prophylactically. Amantidine can
occasionally induce mild neurological symptoms such as
insomnia, loss of concentration and mental
disorientation. However, these symptoms quickly developed
in susceptible individuals and cease when treatment is
stopped. The therapeutic and prophylactic activity of
amantidine is now generally accepted and numerous
analogues of this compound have been prepared.
Rimantadine is not as effective as amantadine but is less
toxic. Prophylaxis with 200mg of amantadine per day for 5
to 6 weeks or for the duration of the influenza A
outbreak is not recommended for all persons. However,
elderly persons with chronic underlying disease,
institutionalized persons, staff and patients in
hospital, close contacts of an index case, and patients
who cannot receive influenza A vaccine due to sensitivity
to egg protein may benefit from prophylaxis. Amantadine
can also be used for therapy of uncomplicated influenza A
infections. The recommended dose is 200mg for 5 days.
Rimantadine may be used in place of amantadine for
prophylaxis and the treatment of uncomplicated influenza
A infections.
Rimantidine - this compound is similar
to amantidine but has fewer side effects. It is approved
by the FDA for the treatment and prophylaxis of influenza
A infection in persons one year or older. It should be
used for uncomplicated influenza A infections only since
it is thought to be less effective than amantidine.
Amantadine and rimantadine resistant viruses are readily
generated in the laboratory. Resistance has been linked
to changes in the M2 protein. To date, the emergence of
resistant influenza A has been documented primarily in
young children undergoing therapy with rimantadine. The
resistant viruses had been transmitted and caused
influenza. The universal susceptibility of all types of
naturally occuring influenza A isolated from man and
animals suggests that resistance will be found only in
individuals treated with the drug. The reason for the
natural selection of the susceptible phenotype of
influenza A in nature is not known.
Zanamivir - the rational approach to
drug design has led to the design of several potent
inhibitors of influenza neuraminidase. Zanamivir was the
first neuraminidase inhibitor available for clinical use
and is effective against both influenza A and B. Because
of its poor bioavailability, zanamivir must be
administered by inhalation. Zanamivir had been shown to
be effective and devoid of significant side effects in
clinical trials. It is now approved by the FDA for use as
treatment for influenza A and B in persons 12 years or
older but not for prophylaxis.
Oseltamivir - oseltamivir is another
neuraminidase inhibitor but unlike zanamivir, it can be
given orally. Like zanamivir, it had been shown to be
effective and devoid of significant side effects in
clinical trials. It is approved by the FDA for use as
treatment for influenza A and B in persons 18 years or
older. It is also approved for prophylaxis in persons 13
years or older. Its lack of side effects would make
particularly attractive in a family setting although its
higher cost compared to amantidine and rimantidine should
be taken into account.
G. Prevention
Vaccines against influenza have been around for 50 years.
Despite this, the efficacy of influenza vaccines is still
questioned, and the ability of vaccines to limit epidemic
infection has not been proven.
1. Immunity to Influenza - the results of challenge
studies indicated that immunity is induced by the host responses
to the virus haemagglutinin (HA) and to neuraminidase (NA).
Antibody against HA is the most important component in the
protection against influenza viruses. In addition to conferring
relative protection against infection, serum HI is reported both
to reduce the severity of infection and decrease virus spreading
in infected persons. Serum anti-neuraminidase Ab has also been
shown to contribute protection against influenza infection. A
generally held view is that serum HI antibody is more important
in determining immunity than anti-neuraminidase antibody. It is
clear that an influenza vaccine must contain both HA and NA
antigens in a form which will stimulate the production of
neutralizing antibody, local IgA antibody and possibly cellular
immunity.
2. Types of vaccine
Whole virus vaccines - whole inactivated virus
vaccines were the first influenza vaccines to be
produced. The currently circulating strain of influenza
is inoculated into embryonated eggs, harvested 2-3 days
later and inactivated. this vaccine confers protection in
60-90% of vaccinees and the protection lasts for 1-5
years, depending on the vaccine strain and the age of the
vaccinee. However, the subsequent infecting virus may
show slow antigenic drift and the vaccine induced
antibody will be less effective in conferring protection
against the new strains.
Split virus vaccines - Because of the high
incidence of reactions seen in vaccinees given whole,
inactivated virus vaccine, attempts have been made to
produce a vaccine which is less reactogenic but
conserving immunogenicity. Split vaccines were prepared
inactivated particles disrupted with detergents. These
vaccines have been shown to induce fewer side effects in
the vaccinees and are just are immunogenic as whole virus
vaccine. Whole virus vaccine should not be used in
children.
Subunit virus vaccines - subunit vaccines have
been prepared which contained only the HA and NA
antigens. These are used in aqueous suspension or may be
absorbed to carriers such as alhydrogel. Volunteers given
subunit vaccines experienced fewer reactions than those
given whole virus vaccines and those given aqueous
vaccine experienced fewer reactions than those given the
absorbed subunit vaccine. therefore, the best vaccines
available at present are the aqueous subunit vaccines,
although some authorities have questioned the
effectiveness of subunit vaccines.
Live attenuated vaccines - there is experimental
evidence that immunization with live, attenuated
influenza virus vaccines induce a solid immunity than do
inactivated vaccines. Normal methods for attenuation,
such as repeated passages and temperature adaptation
require a long period to complete, and probably too long
for the vaccine to become available for immunization
against the current influenza strain. To circumvent this
problem , already attenuated strains have been mixed with
wild-type virus to produce recombinants which contain the
RNA fragments which code for wild-type HA and NA, and all
the other genetic material form the attenuated strain.
These recombinants can be produced relatively quickly.
When given intranasally, produced few side effects.
Although research to develop live attenuated vaccines has been
pursued for 20 years, basic problems remains particularly in the
area of purification. The vaccine must also be shown to be
attenuated and safe. It is estimated that if the safeguards are
to be satisfied, 2 years would be needed for the development of
an attenuated vaccine. This makes their development impractical ,
since by the time the vaccine virus can be made available, the
epidemic strain against which the vaccine has been prepared would
have disappeared. For an attenuated vaccine to be a practical
proposition, the development time must be down to 6-9 months.
3. Recommendations - At present, no live attenuated
vaccine is available for general use. The vaccines that are
currently available are produced from virus grown in embryonated
eggs. The aqueous subunit vaccine is the most acceptable
formulation. These vaccines produce few reactions and confer
protection in 60-90% of vaccinees. Vaccination is recommended for
the elderly and individuals at risk for severe infection. In
addition, key personnel n industry and social and medical
services. It is highly debatable whether the vaccine should be
given to the general population. Influenza immunization is
strongly recommended for adults and children with any of the
following:
1. Chronic respiratory disease
2. Chronic heart disease
3. Chronic renal failure
4. Diabetes mellitus and other endocrine disorders
5. Immunosuppresion due to disease or treatment
Immunization is also recommended for residents of nursing
homes and old peoples' homes and other long stay facilities where
rapid spread is likely to follow. Two types of vaccines are
available in the UK; "split virus vaccines", and
"surface antigen" vaccine which contains highly
purified HA and NA antigens prepared from disrupted virus
particles. Both vaccines are suitable for use in children.
4. WHO Influenza Surveillance Program
The WHO has a network of around 110 influenza centres
worldwide that regularly submit new influenza isolates to the 4
WHO collaborating centres (US, UK, Japan, Australia) for
analysis. The aim is to detect new and potentially
dangerous strains of influenza at the earliest moment so that
measures can be enacted in the event of a pandemic. The strains
used in current influenza vaccines are supplied to the vaccine by
the WHO. It is normally a trivalent vaccine: one H3N2, one H1N1
and one influenza A sutype. The sutypes selected are those that
are normally the most antigencally diverse strains considered to
have to greatest epidemic potential.
H. The H5N1 avian influenza outbreak in Hong Kong 1997
In the latter half of 1997, an outbreak occurred in Hong Kong
whereby 18 persons were infected by an avian influenza A,
serotype H5N1. Of these 6 died, and 3 others were severely ill.
The source of the outbreak was infected chickens and the outbreak
stopped after all the chickens were slaughtered in the territory.
Large-scale serological studies carried out showed that workers
in the poultry industry were particularly at risk of infection
although none complained of any symptoms. There was evidence of
limited human to human transmission. It was postulated that the
strain of avian influenza involved was unusually virulent; it had
multiple basic amino acids near the cleavage site of the
haemagglutinin protein, which as a result may render the
haemagglutinin susceptible to a wider range of proteases. Since
that outbreak, no more cases have occurred. In 1999, there were
reports of human infections by avian influenza A H9N2 in Hong
Kong and in Mainland China. However, all these cases were very
mild and it is thought that the virus was unlikely to pose a
large public health risk.
