03 Bouma.doc:chevalier 24/11/04
Rev. sci. tech. Off. int. Epiz., 2008, 27 (3), 633-642
Field trial for assessment of avian influenza
vaccination effectiveness in Indonesia
A. Bouma (1)*, A. Teguh Muljono (2), A. Jatikusumah (2), A.J. Nell (3),
S. Mudjiartiningsih (4), I. Dharmayanti (5), E. Sawitri Siregar (6),
I. Claassen (7), G. Koch (8) & J.A. Stegeman (1)
(1) Faculty of Veterinary Medicine, Department of Farm Animal Health, Marburglaan 2,
3584 CN Utrecht, the Netherlands
*Corresponding author: E-mail: a.bouma@uu.nl
(2) Center for Indonesian Veterinary Analytical Studies, Jl. Ismaya II No. 2 Rt 04/16, Indraprasta I,
Bogor, West Java. Indonesia 16000
(3) Wageningen International (Wageningen University and Research Centre), Lawickse Alee 11,
6701 AN Wageningen, the Netherlands
(4) Provincial Animal Health Laboratory, Jl Raya Tangkuban Prahu, KM 22.2, Cikole,
Lembang – Bandung 40391, Indonesia
(5) Bbalitvet, Jl R.E. Martadinata 30, Bogor 16114 – West Java, Indonesia
(6) Campaign Management Unit, Directorate of Animal Health, Ministry of Agriculture, Building C,
9th Floor, Jl Harsono RM Kav 3, Ragunan, Jakarta-Selatan, Indonesia
(7) Partnership Office, Indonesia-Netherlands Partnership on HPAI Prevention Control, CMU,
Jl Harsono RM Kav 3, Ragunan, Jakarta-Selatan, Indonesia
(8) Central Veterinary Institute of Wageningen University and Research Centre, Houtribweg 39,
Lelystad, the Netherlands
Submitted for publication: 22 February 2008
Accepted for publication: 22 July 2008
Summary
The aim of this field study was to determine the efficacy of vaccination against
highly pathogenic avian influenza (HPAI) virus strain H5N1 in Indonesia. A
limited, prototype clinical trial was performed using a standardised treatment
group, in which poultry flocks were vaccinated at least twice with a selected
H5N1 vaccine, and a control group comprising flocks treated with non-
standardised procedures chosen by the farmer. Each group consisted of six
flocks comprising either layers or native chickens. Haemagglutination inhibition
(HI) antibody levels were determined by regular serum sampling, and outbreak
surveillance relied on non-AI-vaccinated sentinel birds. After three vaccinations
high antibody titres were produced in the treatment group, and the percentage
of layers with an HI titre > 40 was approximately 90%. Although no conclusions
can be drawn regarding reduction of virus transmission, this study demonstrated
that 11 farms remained free from AI during the observation period, and that a
surveillance programme based on differentiating infected from vaccinated
animals (DIVA) can be implemented.
Keywords
DIVA – Highly pathogenic avian influenza – Indonesia – Non-AI-vaccinated sentinel –
Surveillance – Vaccination.
Introduction
Southeast Asia, Africa and Europe (2, 3). Outbreaks are
controlled by stamping out infected poultry, sometimes
In 1997 an outbreak in Hong Kong of highly pathogenic
followed by pre-emptive culling of contiguous flocks (11).
avian influenza (HPAI) strain H5N1 was reported to the
In Europe, Malaysia, Japan, South Korea and Thailand, this
World Organisation for Animal Health (OIE) (22). Since
strategy was successful, as the number of poultry outbreaks
then, the virus has spread to many other countries in
remained rather limited, and affected countries were
634
Rev. sci. tech. Off. int. Epiz., 27 (3)
declared free from the virus (3, 22). In some countries,
improved, as also indicated by the Indonesian government.
such as China, Vietnam and Indonesia, the virus spread
Using current surveillance data, the effectiveness of
rapidly and many regions have become endemically
vaccination in the field has been doubted, as suggested by
infected. The fear of a human pandemic (20), but also
low coverage and low haemagglutination inhibition (HI)
concern about food security and protection of livelihoods
antibody titres. However, the basic question is whether
are important drivers for various organisations, such as the
flocks were not vaccinated properly or whether the
Food and Agriculture Organization of the United Nations
vaccines were of insufficient quality. Moreover, it has been
(FAO), the World Health Organization (WHO), and the
questioned whether vaccination could reduce horizontal
United States Agency for International Development
transmission of AI between vaccinated flocks. Finally, a
(USAID), to participate in the control of HPAI in these
good surveillance system for AI in vaccinated poultry
affected countries. Control strategies are being developed,
flocks in Indonesia has been lacking.
and surveillance programmes such as participatory disease
surveillance and response (PDS/R) (7) have been
Therefore, a limited, prototype clinical trial was carried out
implemented. Despite all such efforts, the disease is still
to measure antibody titres induced by vaccination,
present, and human cases of H5N1 infection are still being
the success of immune coverage within flocks, and the
reported (21).
reduction in the number of outbreaks of H5N1 in
vaccinated flocks. The results should contribute
In endemically affected areas it is not feasible to apply pre-
to improving the control strategy and surveillance systems
emptive culling or a stamping-out strategy on a large scale
in Indonesia, and may benefit the health status of both
because many flocks are affected. Poultry farmers in these
poultry and humans. In addition, the difficulties with
areas lose income when their flock is infected, either owing
implementing such a trial, and the use of a surveillance
to reduced production or to increased mortality of their
system and vaccination campaign in sectors three and four
poultry. Consequently, the disease in these countries is
of the poultry production industry are discussed in this
mainly controlled by vaccination, using various types
paper. (The FAO has grouped all systems of poultry
of vaccines, with the aim of preventing production losses if
production into four operational levels – sector one:
HPAI occurs within a flock. However, farmers are still
industrial, integrated farms with a high level of biosecurity;
confronted with outbreaks, even in birds on vaccinated
sector two: non-integrated farms with moderate to high
farms. Explanations that have been given are the use
biosecurity levels; sector three: small and medium-scale
of poor quality vaccines or improper vaccination
poultry farms with limited biosecurity; and sector four:
technique (1).
backyard poultry with no biosecurity measures in
place [7]).
The question is, therefore, whether it is possible to
eradicate the virus by means of a proper vaccination
schedule using adequate vaccines, or whether the
application of vaccines will induce clinical protection for
Materials and methods
poultry but will facilitate silent spread of the infection and
human exposure to the virus. Experimental studies have
Study area
shown that various vaccines (either heterologous or
homologous) are able to induce good clinical protection
The trial was carried out in three sub-districts within the
against a challenge infection with H5N1 strains of HPAI (5,
district of Sukabumi in the province of West Java on Java,
8, 14, 15, 18). It is, however, questionable whether
Indonesia. Selection criteria for these sub-districts
experimental data can be extrapolated to the field, because
included: recent outbreaks of H5N1 HPAI in the area, the
vaccination is often less effective when applied under field
presence of commercial layer flocks of sector three,
conditions. Field data from Vietnam have shown that the
adequate veterinary infrastructure, and access to laboratory
number of human cases of AI decreased following the
facilities.
wide-scale application of vaccines (21). However, whether
this reduction was attributable to the vaccination
A survey was carried out by the Centre for Indonesian
campaign, or to increased biosecurity measures or public
Veterinary Analytical Studies (CIVAS), an Indonesian non-
awareness, has not been established, because no control
governmental organisation, to locate poultry flocks, and to
group was included. It remains, therefore, unclear whether
provide technical data on these farms. The minimum
vaccination can be effective with respect to reduction of the
number of flocks to be included in this clinical trial to
number of outbreaks.
provide sufficient power to demonstrate a difference in
effectiveness between a treatment and a control group was
In Indonesia, a large number of human cases have been
calculated to be approximately 200, with 100 flocks per
reported and the virus is still circulating despite the wide
group. As expected, the number of flocks in the selected
application of vaccines. The control and surveillance
sub-districts was not sufficient to meet this requirement.
programme for AI in this country should therefore be
Therefore, the trial was initiated with a lower number of
Rev. sci. tech. Off. int. Epiz., 27 (3)
635
flocks. It was considered a demonstration project to gain
Table I
insight into the adequacy of laboratory procedures, to
Vaccination and sampling scheme for flocks in the treatment
improve vaccination coverage and surveillance
group
programmes, to build experience in the design and
Farm type
Vaccination at age
Sampling at age
implementation of vaccination and surveillance
programmes, to extend the ability to conduct field
Layer
4, 10, 17 weeks
8, 14, 21 weeks
research, and to contribute to capacity building in
Native chicken
10, 30 days
28, 70 days
Indonesia.
broilers were vaccinated twice, at 10 and 30 days old
Criteria for inclusion of farms
(Table I).
The farms to be included in the trial were those producing
commercial layers and native-breed broilers, as the
Backyard (sector four) poultry within a radius of 1 km
production cycles of these two types would last long
around each flock in the treatment group were vaccinated
enough for the chickens to develop a vaccine-induced
every four months by Dinas Peternakan and CIVAS staff
immune response. The production period of the layer
using the same vaccine batches as for flocks of the
flocks was approximately 1 year and that of the native
treatment group. This area is referred to as ‘the ring’. The
broiler chickens was 3 months. Farms of sectors one and
aim of vaccination in this area was to reduce the chance of
two were not included because government veterinarians
infection occurring in chickens near to a flock in the
generally do not have access to these farms.
treatment group, which could increase the risk of
introduction of virus into the study flock. This is not
Two groups were formed: a treatment group and a control
unlikely, because the biosecurity level of these flocks is
group. These two groups differed in the vaccination
not very high.
programme for AI, as described in the next paragraph. In
the treatment group there were more flocks than there
were farms: either there were more sheds on the farm or
Control group
the production period of native chickens was shorter than
It was considered unethical to include a group of
the duration of the trial, so when a flock was slaughtered it
intentionally unvaccinated flocks in this trial, as the area
was replaced by a new flock. The owners might have been
was endemically infected. Any introduction of virus would
more interested in participating a second time than owners
probably result in high mortality, and a high virus load,
of the control native farms.
resulting in loss of income for the owner, and increased
risk of exposure for humans. The control group consisted
of flocks that were vaccinated by the farmers according to
Vaccine and vaccination scheme
their own vaccination scheme. Vaccines were purchased by
Treatment group
the farmers and applied by farm staff. Vaccination
schedules on control farms differed. Layer flocks were
The vaccine used for the treatment group was a locally
generally vaccinated twice, at 4 and at 16 to 18 weeks of
produced commercially available vaccine based on the
age; native chicken flocks were vaccinated once at 10 days
HPAI H5N1 seed strain A/chicken/Legok/03. The vaccine,
of age. Information about the vaccines used was either not
Medivac AI vaccine, was produced by PT Medion
provided by the farmers or was incomplete. The vaccines
(Bandung, Indonesia) according to OIE guidelines (22).
could include any vaccine available, either legally or
Medivac AI contains field isolated avian influenza virus of
illegally, including Medivac AI. No ring vaccination around
H5 subtype. The inactivated viruses are emulsified in
farms in this group was applied by Dinas Peternakan or
mineral oil adjuvant to enhance and prolong the efficacy of
CIVAS.
the vaccine. Each dose contains at least 50 times the 50%
protective dose (PD50) (5).
Surveillance
The vaccine was injected intramuscularly into the leg
according to the recommendation of the manufacturer. The
Two variables were determined:
vaccination was carried out by farm staff or by para-
– antibody responses after vaccination
veterinarians of the Dinas Peternakan (Government
Livestock Services) supervised by veterinarians from
– the number of H5N1 outbreaks in each flock.
CIVAS.
In a vaccinated population, the surveillance strategy
The aim was to vaccinate layers in the treatment group
should be based on virological and/or serological methods
three times, at 4, 10 and 17 weeks of age. Native chicken
and clinical surveillance (22). In flocks where homologous
636
Rev. sci. tech. Off. int. Epiz., 27 (3)
AI vaccines are used, the OIE (22) suggests a surveillance
vaccination; native chickens were sampled at the end of the
system based on the DIVA principle. A DIVA strategy is
production period (at the age of 2 to 3 months). Serum
defined as one that differentiates vaccinated from infected
samples from ten randomly selected sentinels were
animals (12, 19), and it involves the use of accompanying
collected at the same sampling points to demonstrate that
serological tests, or sentinel birds that are not vaccinated
they were still seronegative for H5, and to monitor
against AI (hereafter referred to as ‘sentinels’) (22). Such a
subclinical infections with low pathogenic avian influenza
surveillance system is, however, currently lacking in
(LPAI) H5 strains. At the end of the production period,
Indonesia.
sentinels were swabbed and serum samples were taken to
demonstrate the absence of infection with AI virus.
As a homologous vaccine was used, at least in the
treatment group, the surveillance had to be based on
Two months after each vaccination of backyard chickens in
sentinels, according to the DIVA principle as described by
the ring around the flocks in the treatment group, serum
the OIE (22). This excludes the possible contribution of
samples were taken from 20 randomly selected birds per
infection to antibody titres and allows detection
ring to monitor antibody titres.
of incursions and subsequent outbreaks of H5N1
virus strains.
Laboratory tests
Sentinels were placed in each flock in both the treatment
and control groups. The sentinels originated from the same
The swabs were transported in medium to the Bbalitvet
farm as the one on which they were placed. The sentinel
reference laboratory in Bogor, Indonesia. The samples were
birds in the native chicken flocks were labelled with a
stored at –70ºC until they were tested for the presence
metal ring around one leg, and were easily identifiable in
of H5 AI antigen. A polymerase chain reaction (PCR) kit
the flock; the birds in the layer flocks were housed in cages,
(InvitrogenTM) (9) was used to detect H5 RNA in each
as were the other birds in the flock. The sentinels received
sample according to the manufacturer’s instructions.
equivalent treatment with respect to vaccinations against
Positive samples were cultured in embryonated specific
all diseases except AI. Within each shed, 10 to 50 sentinels
pathogen free (SPF) eggs according to the standard
were maintained, depending on the number of birds in
procedures of Bbalitvet, which follow the method
each shed. The sentinels were housed in the same way as
described in chapter 2.1.14 of the OIE Manual
the other birds: either they could mingle freely within the
of Diagnostic Tests and Vaccines for Terrestrial Animals
group (native chickens) or were put in individual cages
(Terrestrial Manual) (22).
(layers). Serum samples were taken from the sentinels at
the time of first vaccination of the flock, in order to
The serum samples were tested in the provincial animal
demonstrate the absence of antibodies against H5 at the
health laboratory of West Java in Cikole. Blood samples
start of the trial. It was assumed that birds were not
were centrifuged and serum was stored at 4ºC until tested.
infected with H5N1 if clinical signs were absent.
Tests were carried out shortly after arrival of the samples in
the laboratory. The sera were tested by an HI test using
twofold dilutions according to the procedures described in
the Terrestrial Manual (22), using 4 haemagglutination
Sampling
units (HAU) of H5N1 strain A/ch/Legok/03 as the antigen.
The farms included in the trial were visited regularly by
Erythrocytes from SPF chickens were provided by PT
CIVAS staff members for sampling and clinical inspection.
Medion, Bandung. A positive and a negative serum sample
Dead sentinels were reported immediately by the owners to
were included in each test, and tests were carried out in
CIVAS, and swab samples were collected from the trachea
duplicate. Titres were expressed as 2x of the serum dilution
and cloaca of every dead sentinel bird. If sentinels were
that caused complete inhibition of agglutination (22).
found dead by CIVAS, or if these birds or other birds in the
flock showed signs of AI, the same procedure was
followed. Swabs were placed in a tube containing transport
medium and were stored at 4ºC during transport to the
Results
laboratory, which took place within 24 h after collection.
Farms
Serum samples were collected regularly to monitor
vaccine-induced antibodies and introduction of H5N1
A total of twelve farms were included in the trial. The
virus. Serum samples from 20 randomly selected
treatment group consisted of two layer and four native
vaccinated birds were collected on the days on which
chicken farms, and the control group contained four layer
flocks of the treatment group were vaccinated; samples
and two native chicken farms. In the treatment group, for
were collected from flocks of the control group at similar
reasons already discussed, more flocks than farms were
times. Layers were also sampled one month after the last
included and used for the calculation of the average titre
Rev. sci. tech. Off. int. Epiz., 27 (3)
637
(Table II). The treatment was not allocated randomly, as
35
some farmers were willing to cooperate only when they
30
could apply their own vaccination programme. These
25
farms became part of the control group.
20
15
Table II
10
Technical data from the flocks in the treatment and control
groups
Percentage of birds sampled
5
0
Group
Farm type
Number of farms
Total number of flocks*
0
1
2
3
4
5
6
7
8
9
10
11
12
Haemagglutination inhibition titre
Treatment
Layer
2
5
Samples taken after the second vaccination
Native chicken
4
9
Samples taken after the third vaccination
Control
Layer
4
4
Fig. 1
Native chicken
2
2
Distribution of HI antibody titres in samples from layer
flocks in the treatment group that were collected after two
* In the treatment group there were more flocks than there were farms: one layer farm
had more sheds that were included. With respect to the native chicken farms, the
and three vaccinations
production period of poultry was shorter than the duration of the trial, so when a flock was
slaughtered it was replaced by a new flock
Control group
The average titre of birds in layer flocks after the second
vaccination was 25.8 (SD 23.7). This calculation was based
Serology
on samples collected from chickens of all four layer farms.
In one of these flocks, very few sampled birds developed a
The average titre was based on samples with a titre >21. So-
titre. In this flock a vaccine of unknown origin had been
called non-responders were not included in this
used. The average titre in this particular flock was 20.4
calculation. The vaccination coverage (percentage of birds
(SD 20.9) with a coverage of 0%. The average titre of the
with a predefined titre) was calculated for a cut-off of 24
three other flocks was 27.6 (SD 22.2) with a coverage (at a
and for a cut-off of 25 (10).
titre of 24) of approximately 93% after the second
vaccination.
Treatment group
The average coverage in the native chicken flocks was
In the layer farms, the average geometric titre of serum
50%. The average titre in the native chicken flocks in the
after two vaccinations was 25.9 (standard deviation [SD]
control group was 22.9 (SD 22.4).
21.7), and after the third vaccination it was 27.2 (SD 22.0).
The mean was based on samples from a total of five flocks
Ring zone
from two layer farms, collected in 2007. None of the
samples had a titre of 21, indicating that in these flocks no
The average titre in the 1 km ring around the flocks in the
non-responders were present after the second and third
treatment group after two vaccinations was 22.6 (SD 22.7).
vaccinations. The distribution of titres was more or less
normal (Fig. 1). Two birds out of a total of 120 birds
sampled from the treatment group had a titre below 24.
30
Assuming that a titre of 24 represented a positive response,
25
this finding means that the vaccine coverage, here defined
as the percentage of birds with a titre equal to or higher
20
than 24, was >95%. The vaccine coverage using a cut-off
value of 25 was approximately 90%.
15
10
In the native chicken flocks (4 farms; 9 flocks in total),
Percentage of birds sampled
24% of the birds had a titre of 21 at the end of production,
5
after two vaccinations. We estimated the average titre of the
0
remaining 76% of the birds, which did respond (Fig. 2).
0
1
2
3
4
5
6
7
8
9
10
The average titre in these birds was 24.4 (SD 22.3). In this
Haemagglutination inhibition titre
group, 51% of samples had a titre above 24, thus a much
Fig. 2
lower coverage was achieved, and this was exacerbated
Distribution of HI antibody titres in samples from native chicken
because the number of non-responders was much higher
flocks in the treatment group after two vaccinations, collected
than among the layers in the treatment group.
at the end of the production period
638
Rev. sci. tech. Off. int. Epiz., 27 (3)
The vaccination coverage (percentage of birds with a titre
assumed that vaccinated zoo birds with a titre of 1:40 or
of 24 or above) was 38%. The coverage estimated using this
greater would be protected. After three vaccinations the
cut-off titre in the 1 km ring around the flocks in the
coverage obtained using this cut-off value was
control group was 12%.
approximately 90%. It would be helpful to have more data
on the relationship between titre and protection against
virus transmission, as has been described for Newcastle
Virus detection
disease (17).
An outbreak of H5 virus infection occurred on one farm in
An important aim of vaccination is to reduce the
the treatment group approximately 4 weeks after the
transmission of virus within and between flocks. The
second vaccination, as demonstrated by a positive PCR test
reproduction ratio R, the average number of secondary
and virus isolation in embryonated eggs. The farm had
cases caused by one infectious animal (6), is a frequently
95 sheds containing poultry, of which five were included in
used measure of transmission of pathogens. If R is larger
the trial. The outbreak seemed to have started in one of the
than one, the infection will spread. This measure can also
sheds containing growers that had not been included in
give an indication of the critical fraction of immune birds
the trial. Poultry in two sheds included in the trial were
that is required to prevent major outbreaks in a flock (6).
affected 1 week after the second vaccination (aged 12 to
If the percentage of vaccinated birds in a flock is 1−1/R, the
13 weeks). The average titre just before virus introduction
virus can be eliminated by vaccination. For H5N1 virus
was approximately 25.2 (SD 21.3); the percentage of samples
within a poultry flock the R has been estimated to be about
with a titre >24 was 80%. The other sheds in the trial were
2.3 to 2.6 (16). From the reproduction ratio, the critical
culled at 57 to 62 weeks of age, and no additional
fraction of immune birds is estimated to be approximately
information about these sheds was available. The virus that
60% if a perfect vaccine is used. Assuming that HI titres
caused the outbreak was isolated from the sentinel birds
>40 are sufficient to reduce the reproduction ratio to 0, the
and was confirmed to be HPAI H5 by PCR. Preliminary
coverage of 90% achieved in the layer flocks in
RNA sequence analysis performed at Bbalivet showed that
the treatment group of this study seems to be sufficient. It
the strain was similar to A/Ch/WJ/PWT-WIJ/06, which was
should be realised, however, that the vaccine is probably
isolated on Java in 2006 (13). No H5 virus was detected on
not perfect, and R values may differ between flocks. A level
the other farms that participated in the trial.
of coverage higher than 60% is therefore advisable.
The titres of the samples from native chickens were lower,
Discussion
and more birds on these farms did not have a detectable
titre. It has been suggested that native chickens may be less
responsive to vaccination. The lower responses were
The aim of this study was to study the effectiveness of
explained by concurrent diseases or immunosuppressive
vaccination against H5N1 AI under field conditions by
infections at the time of or after vaccination, but
measuring antibody titres in the HI test and detecting
the underlying mechanism of the lower response needs to
outbreaks of H5 virus infection. The vaccine coverage of a
be explored. It is necessary to study the transmission of
flock was estimated from the titre values. Sentinel birds
the virus in native chickens using vaccination challenge
that were not vaccinated against AI were used to record
experiments under controlled conditions.
major H5N1 outbreaks in the flocks included in the trial,
and were also necessary for correct interpretation of the
An outbreak of AI occurred in one of the layer flocks in the
antibody titres. Without this DIVA surveillance system it is
treatment group. It is believed that the outbreak started in
impossible to be sure whether the measured titres result
one of the sheds on the farm that was not included in the
from vaccination, infection with H5 subtype viruses, or a
treatment group. As a result of the increase in the number
combination of the two.
of birds in the affected shed that were shedding virus, the
dose of virus to which the birds in the flock that were
The average titre after two vaccinations of layers in the
included in the treatment group was exposed may have
treatment group was 25.9 and the coverage, here defined as
been very high. The level of virus challenge was probably
the percentage of samples with a titre of 24 or above, was
much higher than that to which birds in sheds located in a
more than 95%. This trial demonstrates that, with locally
remote area would have been exposed. In addition, this
produced homologous vaccines, high antibody titres and a
farm may have been infected by an antigenic mutant strain
high coverage could be reached, at least in layer flocks.
like the A/Ch/WJ/PWT-WIJ/06 strain. Vaccination
Philippa et al. (10) used a different cut-off value for vaccine
challenge studies have shown that most vaccines currently
effectiveness. They used a titre of >40, a value that is
licensed in Indonesia do not provide optimal protection
considered to be protective and to reduce virus replication
against this strain (13). Based on preliminary sequencing,
in humans, in a situation in which challenge experiments
the virus isolated from this flock was similar to the
were not ethical. For similar reasons Philippa et al. (10)
A/Ch/WJ/PWT-WIJ/06 strain. Given that an outbreak
Rev. sci. tech. Off. int. Epiz., 27 (3)
639
occurred only in a flock of the treatment group, one could
The small sample size means that it is impossible to draw
infer at first sight that vaccination in this group was less
conclusions about the effectiveness of vaccination in
effective than in the control group. This cannot be
relation to virus transmission between flocks, to perform
concluded, however. First, it is obvious that the number of
statistical analysis or to make statistical inferences from the
outbreaks is far too low to be of any statistical significance.
results of this trial. In addition, the owners of backyard
Second, farms were not randomly allocated to the two
chickens did not always respond to the vaccination
groups, either treatment or control, because participation
campaign, and there was a lack of support from the
was entirely voluntary. Therefore, we cannot exclude the
government of the villages. Moreover, flock owners were
possibility that, among other factors, farm management
not convinced that repeated vaccinations were necessary,
differed substantially between the two groups.
and were afraid that vaccination may lead to increased
mortality. The whole vaccination procedure was time
When monitoring HI antibodies induced by vaccination it
consuming because field officers had to catch many of the
is essential to ensure that the antibodies are the result of
chickens to be vaccinated.
vaccination only and are not caused by infection with an
H5 subtype virus, or a combination of vaccination and
Nevertheless, we learned important lessons from this trial
infection. The possible contribution of infection can only
about the antibody responses that can be induced by
be excluded by using the DIVA principle (12, 19).
vaccination in the field, about the quality of the laboratory
The DIVA strategy used in this trial was based on the use
test procedures, and finally about various aspects of the
of sentinel birds; we could not use antibodies against
organisation and implementation of a surveillance
the neuraminidase protein because a homologous
programme in Indonesia. The surveillance system, based
neuraminidase vaccine was used. It seems reasonable to
on the DIVA principle, also showed that farms can remain
assume that the risk of outbreaks of HPAI attributable
free from H5N1 virus during their production cycle. These
to the presence of sentinels is negligible, as only a few birds
and other lessons can be of use for other organisations
in each shed were sentinels. Moreover, direct contact
involved in the planning and implementation of
between the sentinels and the vaccinated birds did not
surveillance and control programmes for AI.
occur, as they were all housed in cages. Data from
transmission trials has shown that no indirect virus
transmission occurs (4); therefore it seems unlikely that the
sentinels on the outbreak farm initially contracted the
Conclusions
infection and were responsible for the subsequent
extensive spread of the virus. Many farmers are of the
A clinical trial was carried out to monitor vaccine-induced
opinion that sentinels are at high risk of infection with AI.
HI titres, vaccination coverage and the number of
The level of knowledge of farmers and veterinarians with
outbreaks of HPAI H5 in poultry flocks. The trial was only
respect to the risks inherent in the use of sentinels and the
small-scale, and this prevents the drawing of conclusions
value of adequate surveillance systems in the control of AI
about the effectiveness of vaccination in AI control in
should therefore be improved. This may contribute to
Indonesia. Nevertheless, valuable experience was obtained
the adoption of DIVA vaccination campaigns and
in the operation of veterinary services, sample handling,
improvement of the current surveillance strategies.
quality assurance of laboratory testing, and the
implementation of a DIVA surveillance strategy using
The number of farms included in this study was much
sentinel birds that were not vaccinated against AI. The
lower than calculated for the power analysis. One of the
results and experiences of this trial may help to develop
reasons was that farmers were reluctant to participate, and
and improve future surveillance and control strategies in
much time and effort was expended on visiting farms and
Indonesia and in other countries in the region.
informing the farmers of the trial. In addition to this, the
treatment was not allocated randomly, as some farmers
were willing to cooperate only when they could apply their
Acknowledgements
own vaccination programme. The programme set up to
inform farmers and try to convince them to join the trial
This research was funded by the Dutch Ministry of
experienced difficulties, especially in convincing farmers to
Agriculture, Nature and Food Quality and was carried out
implement the DIVA strategy using sentinel chickens.
for the Indonesian-Dutch Bilateral Programme on the
Farmers were worried about the risk of these sentinels
Control of HPAI in Indonesia
being a source of infection to other chickens in the flock.
640
Rev. sci. tech. Off. int. Epiz., 27 (3)
Étude de terrain pour évaluer l’efficacité de la vaccination
contre l’influenza aviaire en Indonésie
A. Bouma, A. Teguh Muljono, A. Jatikusumah, A.J. Nell,
S. Mudjiartiningsih, I. Dharmayanti, E. Sawitri Siregar, I. Claassen,
G. Koch & J.A. Stegeman
Résumé
Une étude de terrain a été réalisée afin de déterminer l’efficacité de la
vaccination contre la souche H5N1 du virus de l’influenza aviaire hautement
pathogène (IAHP) en Indonésie. L’essai clinique, de portée limitée et basé sur un
prototype, a porté sur deux groupes de poulets, le premier comprenant des
cheptels traités suivant une procédure standardisée et ayant reçu au moins deux
doses vaccinales d’un vaccin H5N1 sélectionné, et le second (groupe de
contrôle) comprenant des cheptels traités suivant des procédures non
standardisées décidées par l’éleveur. Chaque groupe consistait en six cheptels
de poules pondeuses ou de poulets autochtones. Les titres d’anticorps inhibant
l’hémagglutination (IH) ont été déterminés en se basant sur un échantillonnage
régulier ; la surveillance du foyer reposait sur l’utilisation d’oiseaux sentinelles
non vaccinés contre l’influenza aviaire. Après trois vaccinations, le groupe traité
présentait des titres élevés d’anticorps, avec près de 90 % des poules
pondeuses exhibant un titre d’anticorps IH supérieur à 40. Bien qu’aucune
conclusion ne puisse être tirée quant à une éventuelle diminution de la
transmission virale, il a été constaté que 11 élevages sont restés indemnes
durant la période d’observation ; en outre, l’étude a montré la faisabilité d’un
programme de surveillance fondé sur la différenciation entre animaux infectés et
animaux vaccinés (DIVA).
Mots-clés
Différenciation entre animaux infectés et animaux vaccinés (DIVA) – Indonésie –
Influenza aviaire hautement pathogène – Oiseau sentinelle non vacciné contre l’influenza
aviaire – Surveillance – Vaccination.
Ensayos sobre el terreno para evaluar la eficacia de la vacunación
contra la influenza aviar en Indonesia
A. Bouma, A. Teguh Muljono, A. Jatikusumah, A.J. Nell,
S. Mudjiartiningsih, I. Dharmayanti, E. Sawitri Siregar, I. Claassen,
G. Koch & J.A. Stegeman
Resumen
Los autores describen un estudio sobre el terreno realizado en Indonesia para
determinar la eficacia de la vacunación contra la cepa H5N1 del virus de la
influenza aviar altamente patógena. Se llevó a cabo un ensayo clínico limitado,
con carácter experimental, utilizando un grupo de tratamiento normalizado, en el
que las bandadas fueron vacunadas como mínimo dos veces con una vacuna
Rev. sci. tech. Off. int. Epiz., 27 (3)
641
anti-H5N1, y un grupo de control con bandadas sometidas a un tratamiento no
normalizado, que se dejaba a discreción del criador. Cada grupo constaba de
seis bandadas, compuestas por aves ponedoras o pollos nativos. Tras un
muestreo sérico ordinario se determinaron los niveles de anticuerpos por
inhibición de hemaglutinación (IH) y se estableció una vigilancia de posibles
brotes con aves centinela no vacunadas contra la influenza aviar. Después de
tres vacunaciones se obtuvieron títulos elevados de anticuerpos en el grupo de
tratamiento: alrededor de un 90% de las ponedoras presentaba un título de IH
superior a 40. Aunque no cabe extraer conclusiones respecto a la reducción de
la transmisión del virus, el estudio demostró que 11 explotaciones permanecían
libres de influenza aviar durante el periodo de observación, y que es posible
instituir un programa de vigilancia basado en la discriminación entre animales
infectados y vacunados (DIVA: differentiating infected from vaccinated animals).
Palabras clave
Centinela no vacunado contra la influenza aviar – DIVA – Indonesia – Influenza aviar
altamente patógena – Vacunación – Vigilancia.
References
1. Adjid R.M.A., Nuradji H., Indriani R., Dharmayanti N.L.P.I.,
5. Claassen I.J.T.M., Jonas Jahja E., Sahesti A., Farida A., Balk F.
& Darminto (2007). – Evaluation on vaccination program in
& Bouma A. (2007). – PD50 experiments demonstrate that
backyard and in intensive native poultry under villages
avian influenza vaccines used in Indonesia differ greatly in
condition: a case study in Serang district. International
protective capacity. OIE/FAO/IZSVe Scientific Conference –
AI Vaccination Seminar, June, Jakarta. Indonesian Central
Vaccination: a tool for the control of avian influenza,
Research Institute for Veterinary Science, Agency
20-22 March, Verona, Italy.
for Agricultural Research and Development, Ministry
of Agriculture, Jakarta, Indonesia.
6. Diekmann O. & Heesterbeek J.A.P. (2000). – Mathematical
epidemiology of infectious diseases. Model building, analysis
2. Alexander D.J. (2007). – An overview of the epidemiology
and interpretation. John Wiley & Sons, Chichester, United
of avian influenza. Vaccine, 25 (30), 5637-5644.
Kingdom.
3. Alexander D.J. (2007). – Summary of avian influenza activity
in Europe, Asia, Africa, and Australia, 2002-2006. Avian Dis.,
7. Food and Agriculture Organization of the United Nations
51 (1S), 161-166.
(FAO) (2007). – Programmes/Projects. Available at:
http://www.fao.org/ag/againfo/subjects/ en/health/diseases-
4. Bouma A., Tiensin T., Claassen I., Nielen M., Van Boven M.
cards/27septrecomm.pdf (accessed in June 2008).
& Stegeman A.S. (2007). – Estimation of the critical
proportion of chickens in a flock to be immunized to prevent
8. Kumar M., Chu H.J., Rodenberg J., Krauss S. & Webster R.G.
major outbreaks of HPAI H5N1. OIE/FAO/IZSVe Scientific
(2007). – Association of serologic and protective responses of
Conference – Vaccination: a tool for the control of avian
avian influenza vaccines in chickens. Avian Dis., 51 (1S),
influenza, 20-22 March, Verona, Italy.
481-483.
642
Rev. sci. tech. Off. int. Epiz., 27 (3)
9. Lee M.S., Chang P.C., Shien J.H., Cheng M.C. & Shieh H.K.
16. Tiensin T., Nielen M., Vernooij H., Songserm T., Kalpravidh
(2001). – Identification and subtyping of avian influenza
W., Chotiprasatintara S., Chaisingh A., Wongkasemjit S.,
viruses by reverse transcription-PCR. J. virol. Meth.,
Chanachai K., Thanapongtham W., Srisuvan T. & Stegeman
97, 13-22.
A. (2007). – Transmission of the highly pathogenic avian
influenza virus H5N1 within flocks during the 2004
10. Philippa J.D., Munster V.J., Bolhuis H., Bestebroer T.M.,
epidemic in Thailand. J. infect. Dis., 196 (11), 1679-1684.
Schaftenaar W., Beyer W.E., Fouchier R.A., Kuiken T. &
Osterhaus A.D. (2005). – Highly pathogenic avian influenza
17. Van Boven M., Bouma A., Fabri T.H.F., Katsma E., Hartog L.
(H7N7): vaccination of zoo birds and transmission to non-
& Koch G. (2008). – Flock immunity to Newcastle disease
poultry species. Vaccine, 23 (50), 5743-5750.
virus in poultry by vaccination. Avian Dis., 37 (1), 1-5.
11. Stegeman A., Bouma A., Elbers A.R.W., De Jong M.C.M.,
18. Van der Goot J.A., Koch G., de Jong M.C.M. & Van Boven M.
Nodelijk G., De Klerk F., Koch G. & Van Boven M. (2004). –
(2005). – Quantification of the effect of vaccination on
The avian influenza A virus (H7N7) epidemic in
transmission of avian influenza (H7N7) in chickens. Proc. natl
the Netherlands in 2003: course of the epidemic
Acad. Sci. USA, 102 (50), 18141-18146.
and effectiveness of control measures. J. infect. Dis.,
190, 2088-2095.
19. Van Oirschot J.T. (2001). – Present and future of veterinary
viral vaccinology: a review. Vet. Q., 23 (3), 100-108.
12. Suarez D.L. (2005). – Overview of avian influenza DIVA test
strategies. Biologicals, 33 (4), 221-226.
20. Webby R.J. & Webster R.G. (2003). – Are we ready for
pandemic influenza? Science, 302, 1519-1522.
13. Swayne D.E. (2007). – Overview of avian influenza vaccines
and vaccination. International Avian Influenza Vaccination
21. World Health Organization (WHO) (2006). – Outbreak
Seminar, June, Jakarta, Indonesia.
news. Weekly Epidemiol. Record, 81 (5), 41-48. Available at:
www.who.int/wer/2006/en/ (accessed in November 2006).
14. Swayne D.E. (2008). – Avian influenza vaccines and therapies
for poultry. Comp. Immunol. Microbiol. infect. Dis., Epub.
22. World Organisation for Animal Health (OIE) (2007). –
Manual of Diagnostic Tests and Vaccines for Terrestrial
15. Swayne D.E. & Kapczynski D.R. (2008). – Vaccines,
Animals. Available at: http://www.oie.int/eng/normes/
vaccination and immunology for avian influenza viruses
mmanual/A_summry.htm?e1d11 (accessed in November
in poultry. In Avian influenza (D.E. Swayne, ed.). Blackwell
2007).
Publishing, Ames, Iowa, 401-445.
