The epidemiology, pathogenesis and control of highly pathogenic avian influenza (HPAI) in ducks in Indonesia and Vietnam

Objective 1. To understand the epidemiology of highly pathogenic avian influenza (HPAI) H5N1 virus (e.g. seroprevalence, duration of virus shedding) in small holder duck production systems in Indonesia and Vietnam)
This objective will be achieved through field investigations. The field sites are located in Java in Indonesia and in the southern provinces of Vietnam to coincide with the largest duck populations and where there has been evidence of HPAI infection. A survey strategy for a longitudinal study that provides statistical confidence and considers the cost of sampling and the resources available was developed together with staff of the research organisations in Indonesia and Vietnam. An appropriate study design has been chosen and the sample sizes were estimated. Sampling frames were developed and study units from these sampling frames selected. In both countries 16 villages are involved in the study. A total of 80 duck-owning households in Vietnam and 96 duck-owning households in Indonesia were selected and farmers will be interviewed at two-monthly intervals over a period of 12 months. In addition blood samples and cloacal swabs will be collected from ducks and chickens and tested for haemagglutination inhibition antibody titres and for H5N1 viral RNA by real-time PCR.
Collaborating staff and sample collectors were trained in data collection, which commenced in Indonesia in March 2007 and in Vietnam in May 2007.
In addition in Vietnam a case-control study was conducted to identify risk factors associated with the HPAI outbreaks in the Mekong Delta from December 2006 to January 2007. A questionnaire was developed to investigate these outbreaks in five affected provinces and to collect information about potential risk factors for the emergence and spread of HPAI. In total 23 case farms were selected, along with 46 respective control farms. The questionnaire data was collected in February/ March 2007. The data obtained are currently being analysed.

Objectives 2&3. To understand the role of ducks as maintenance hosts and amplifiers of H5N1 virus during and after infection, and to understand the pathogenesis of HPAI H5N1 infection in ducks
The first activity under these objectives is the isolation of HPAI H5N1 viruses in Indonesia and Vietnam from samples collected in the longitudinal studies. Discussions have taken place, reagents purchased and personnel selected to undertake this activity in each of the research institutions in Indonesia and Vietnam.

Significant progress was made during this reporting period. The longitudinal studies conducted in Indonesia and Vietnam progressed well. In Indonesia, serological data from four bi-monthly samplings from March 2007 to September 2007 indicated that the seroprevalence of HPAI infection in unvaccinated ducks and in-contact chickens was relatively low, ranging from 0 to 6% in ducks and 0 to 10% in chickens between districts and samplings. This suggests that birds were highly susceptible to infection. Although the seroprevalence differed between districts, there was no clear temporal pattern apparent over this sampling period. No birds were vaccinated against H5N1 in Indonesia. Positive H5 PCR results were obtained from 28 ducks and 5 chickens during the same sampling period indicating that virus was circulating among these small-holder flocks. This was further reflected in a number of disease outbreaks that occurred in the study villages. H5 virus was isolated from 43 ducks and 52 chickens that died during these outbreaks.
In Vietnam, most of the birds in the longitudinal study were reported to be vaccinated. Although serological results from two samplings in May and July 2007 indicated that only 60% of vaccinated ducks and 39% of vaccinated chickens had protective titres, no mortality due to HPAI was reported in any of the study villages suggesting that the flock level of protection was adequate. When the antibody response to vaccination over time was analysed, it appeared that chickens responded later than ducks. Chickens had lower antibody titres than ducks at both under 3 weeks and over 15 weeks post vaccination. The sentinel unvaccinated birds in Viet Nam showed higher seroprevalences than the equivalent samples from birds in Indonesia, ranging from 2-26% in ducks and 3-21% in chickens between provinces and samplings. These results from Viet Nam suggest that despite not achieving protective levels of immunity in all birds, vaccination against HPAI can decrease the likelihood of disease outbreaks and reduce mortality in the presence of active exposure to the virus. Analysis of the case-control study conducted to identify farm-level risk factors associated with the HPAI outbreaks that occurred from December 2006 to January 2007 in the Mekong Delta region was continued during this period. In a preliminary analysis it appeared that vaccination status, other poultry species kept on the farm and the sharing of scavenging areas could be potential risk factors. A descriptive pilot study of nomadic duck farms was also conducted to describe this management system in detail and to evaluate the feasibility of conducting a longitudinal study on nomadic duck flocks.
Using H5N1 virus isolates from Indonesia and Viet Nam, isolated in 2006 and 2004 respectively, domestic Pekin ducks were infected by the mucosal route (oral, nasal, ocular). Viral shedding patterns were determined by virus isolation and titration of oral and cloacal swabs. Virus was shed from both the oral and cloacal routes; it was first detected 24 hours after challenge and continued to day 5 after challenge. Virus shedding was detected in all but two birds challenged with the Vietnamese strain, and in 10 of the 15 challenged with the Indonesian strain. The Vietnamese strain caused severe morbidity with fever and depression, whereas the Indonesian strain caused only transient fever. Both viruses had a predilection for a similar range of tissue types, but the quantity of tissue antigen and tissue virus titres appeared to be considerably higher with the Vietnamese strain.
John Allwright Fellowships were awarded to four candidates aligned to the project and they will commence their postgraduate studies in Semester 2, 2008 or in Semester 1, 2009. One project team member from Indonesia received training in molecular diagnostics at the Australian Animal Health Laboratory, Geelong. The project has been represented at a number of scientific meetings during this reporting period and two papers in conference proceedings have been published or accepted. In addition four journal articles are in final stages of preparation. Project meetings were held in Brisbane on 7 January 2008 and in Yogyakarta on 27 March 2008, to discuss progress and present results of the project.

Significant progress was made during this reporting period. The longitudinal studies conducted in Indonesia and Vietnam were completed in early 2009.
In Indonesia, serological data show that the bird-level seroprevalence of HPAI infection in the sampled ducks and in-contact chickens was relatively low, ranging from 1 to 4% in ducks and 0 to 2% in chickens. This suggests that the birds are highly susceptible to infection. However, about 20% of the duck flocks and 2% of the in-contact chicken flocks had at least one seropositive bird, indicating that these flocks had been exposed to HPAI virus. A total of 96 households were visited seven times over the period of one year in Indonesia. In about 16% of flock-visits ducks were H5 antibody positive while chickens at the same sampling were negative, indicating that ducks had been exposed to H5 and might have been carriers of the virus. H5-positive PCR results were obtained from about 3% of the flocks monitored in Indonesia. During HPAI outbreak periods nearly 12% of flocks had healthy birds from which H5 virus was isolated. This indicated that virus shedding was increased during outbreak periods. Outbreaks were common with nearly half of the monitored flocks experiencing at least one outbreak over the 12-month study period. Mortality during the outbreaks was high among chickens.
In Vietnam, most of the birds in the longitudinal study of 80 household-flocks were reported to be vaccinated. Serological results indicated that only in half of the vaccinated duck and chicken flocks, 50% or more of birds per flock had protective levels of antibody. However, no mortality due to HPAI was reported in any of the study villages suggesting that the flock level of protection was adequate. When the response to vaccination over time was analysed, it appeared that only 50% of vaccinated birds had protective titres 4 weeks post vaccination. The sentinel unvaccinated birds in Viet Nam showed higher H5 bird-level seroprevalences than the equivalent birds in Indonesia, ranging from 21-29% in ducks and 13-20% in chickens. About 49% of the duck flocks and 26% of the chicken flocks were probably exposed to HPAI indicated by protective H5 titres found in at least one unvaccinated bird per flock. Flock-level virus prevalence in Viet Nam was low with only 0.2% of flocks having birds that shed H5 virus over the study period. These results from Viet Nam suggest that despite not achieving protective levels of immunity in all birds, vaccination against HPAI can decrease the likelihood of disease outbreaks and reduce mortality in the presence of active exposure to the virus.
Longitudinal studies on the nomadic duck management system were conducted over a period of six months in both countries. In these studies the HPAI field virus antibody prevalence and incidence in moving duck flocks will be determined, the patterns of movement will be described and risk factors associated with HPAI infection will be identified. Data from these studies are currently being compiled into databases.
Using H5N1 virus isolates from Indonesia and Viet Nam, isolated in 2006 and 2004 respectively, domestic Pekin ducks were infected by the mucosal route (oral, nasal, ocular). Both isolates had a predilection for muscle tissues and brain tissue, with little evidence of infection of epithelial surfaces. However, virus was also detected in feather shafts, which may have implications for virus transmission and could become an additional sampling source. Chickens infected with both of these virus isolates succumbed one day post infection. There was a predilection for blood vessel endothelium and lymphoid tissues. No strain differences in pathogenicity or tissue tropism were observed, but larger amounts of virus were shed from chickens infected with the Indonesian isolate compared to the Vietnamese isolate.
An informal project review was conducted during this reporting period and the project was considered to be producing excellent results in a strong collaborative environment between investigators and researchers from Indonesia, Vietnam and Australia. A financial extension until 30 September 2010 was granted. In this extension period the moving duck study will be completed and additional field studies on bio-security issues relating to moving duck flock management will be conducted. This will include surveys of people who provide transport for moving duck flocks, of rice paddy owners who provide scavenging areas and of hatcheries, which provide ducklings and purchase fertilized eggs from the moving duck flocks.
The project has been represented at a number of scientific meetings during this reporting period, three conference papers have been presented and three additional conference papers have been accepted. In addition two journal articles are published or in press and two additional papers are in the final drafting stage.