Animals > Animal Diseases > Avian Influenza > British Columbia (2004)  

A Short Summary of the 2004 Outbreak of High Pathogenicity Avian Influenza (H7N3) in British Columbia, Canada

June 30, 2004

Outbreak Overview | Outbreak Summary Report

1. Background

Previous cases of avian influenza in Canada

Canada has not previously reported a case of highly pathogenic avian influenza to the OIE. However, low pathogencity avian influenza (LPAI) was recognized in turkeys during the 1960's, when it was common for them to be raised in outdoor ranges. An avian influenza virus, isolated from two extensive turkey breeding establishments with common ownership in Ontario in 1966, was later found to meet the modern criteria of a highly pathogenic influenza (HPAI) virus.

Since then, turkeys have been raised in closed poultry houses under more stringent biosecurity precautions with the result that cases of low pathogenicity avian influenza are now rare. Canada has uncovered three cases of low pathogenicity H5 and H7 in domestic poultry since 1975. The latest was an H7N1 strain, isolated from turkeys in Ontario in November 2000. The turkey flock of origin experienced a drop in egg production, along with a slight increase in mortality.

1.1 Description of the poultry industry in the Fraser Valley (see attachment 1)

The poultry industry in the Fraser Valley of British Columbia is typical of the national poultry marketing system. Chickens and turkeys are grown to produce meat and eggs under a supply-managed quota administered by the province. The poultry industry in the Fraser Valley is self-contained with 11 hatcheries and 10 abattoirs (8 under federal inspection, 2 under provincial inspection) and supplies a large local consumer market (the city of Vancouver).

2. Description of the outbreak (see attachments 2,3)

2.1 Infected Premises 1

Highly pathogenic avian influenza (H7N3) was confirmed on March 8, 2004, from samples collected on a farm north of Abbotsford, British Columbia. The farm was a chicken broiler breeder operation that supplied hatching eggs to a local hatchery. Two flocks of birds were on the farm when the disease first appeared, an older flock of 9200 birds (52 weeks of age) in one barn, and a younger flock of 9030 birds (24 weeks of age) in another.

The first signs of illness were a mild drop in egg production and feed consumption, and a slight increase in mortality (6 birds per day) in the older flock. The farm's veterinarian and the feed company representative investigated the case and samples were submitted to the British Columbia Ministry of Agriculture, Food and Fisheries (BCMAFF) diagnostic laboratory for routine post-mortem examination on February 9. Pathologic signs included meaty lungs and inflamed tracheas.

The clinical illness in the older flock appeared to resolve slightly over the next few days. However, on February 16 BCMAFF reported to CFIA that avian influenza virus had been detected on a polymerase chain reaction (PCR) test and within a day of this report mortality in the younger flock began to climb dramatically - from 25 dead birds one day, to 930 the next, and over 1500 on the third. Allantoic fluid specimens from both flocks were forwarded for further testing to the Canadian reference laboratory, the National Centre for Foreign Animal Disease (NCFAD) in Winnipeg, and on February 19 the H7 subtype of avian influenza was confirmed.

The farm had been placed under quarantine on February 16 and when the diagnosis was confirmed the entire population of approximately 16,000 birds were destroyed on February 19-20. A surveillance program, based on oropharyngeal and cloacal swabs and blood samples, was initiated on commercial farms within 5 km of the infected flock.

Ongoing work at NCFAD characterized the virus from the older flock as belonging to the H7N3 subtype (reported on February 23, 2004). It was further determined on March 1 as being a variety exhibiting low pathogenicity, with an intravenous pathology index (IVPI) of 0. On March 4 (and further confirmed on March 8) the isolate from the younger flock was found to be highly pathogenic, with an IVPI index of 2.96. Gene sequencing analysis demonstrated the presence of a seven amino acid insertion within the cleavage site of this isolate.

2.2 Infected premises 2

On March 11, a second flock within the surveillance zone, approximately 3 km west of the first infected premises, was confirmed as also being infected with H7N3 avian influenza. The only clinical sign observed was increasing mortality in one barn - beginning at 5-10 dead birds per day, peaking at 150 on March 9, and then dropping back to 10 dead birds on the day before depopulation. As part of CFIA's investigation into this increased mortality, on March 9 samples were collected and the farm was placed under quarantine. The birds were destroyed a day after the laboratory confirmation of avian influenza was received.

This second broiler breeder premises had four barns, two with pullets and two with laying birds. Only one barn of 13 week old pullets showed clinical signs and virus was isolated only from that flock. Again, clinical signs were not consistent with a highly virulent virus - exhibiting less than 10% mortality, but IVPI and sequencing analysis demonstrated the same seven amino acid insertion, confirming that this isolate was also highly pathogenic.

2.3 Progression of the outbreak during March

On March 11, the Minister of Agriculture and Agri-food made a declaration, pursuant to subsection 27(1) of the Health of Animals Act, implementing a Control Area in the Fraser Valley of British Columbia that restricted the movement of any captive birds, including day-old chicks and hatching eggs, any products or byproducts of birds, and anything (i.e. equipment) that had been exposed to a bird. At the same time, the CFIA implemented 3 disease control zones. The 5 km zone around the original index case was designated as the High Risk Region (HRR). This was surrounded by a larger Surveillance Region (SR), and the Fraser Valley south of the Fraser River was designated as the Control Area (CA). Movement controls were established to prohibit removal of birds or infected material from the High Risk Region and cleaning and disinfection (C&D) stations were established at the perimeter.

Commercial and backyard flocks within the High Risk Region and commercial flocks within the Surveillance Region that were designated as high risk contacts were targeted for regular active surveillance. A sampling plan was implemented to enable detection if disease prevalence on the farm was 5% or higher. On all premises, a minimum of 60 birds were sampled; on those with 3 or more barns, 25 samples per barn were taken. The owners of other commercial flocks in the Surveillance Region were interviewed by telephone and dead birds were collected and swabbed weekly. In addition, all birds intended for slaughter were tested (pre-slaughter) according to the same protocol.

The disease spread 2 km south and west to a cluster of three farms within the High Risk Region that tested positive between March 13 and 19. In the face of the outbreak, depopulation decisions were based on a positive PCR result regardless of whether or not clinical signs were expressed or virus was isolated.

On March 24, after 5 commercial farms and 2 backyard flocks had tested positive for avian influenza, the outbreak was declared to be an emergency and the CFIA activated its National Emergency Operations Centre in Ottawa. A decision was made to pre-emptively slaughter all poultry (275,000 birds) within the High Risk Region and active surveillance activities were redirected to dead bird pickup within Surveillance Region.

Sporadic cases were diagnosed over the next few days among a few outlying farms outside the High Risk Region and within a second small cluster of farms 7 km to the west, along North Lehman Road. These did not exhibit the dramatic clinical evidence that is typical of highly pathogenic avian influenza, but were discovered during the course of routine surveillance testing of dead birds.

By the end of March, a total of 20 commercial farms had been found to be positive by PCR testing - 13 within the High Risk Region, 5 within the surrounding Surveillance Region and 2 outside the Surveillance Region but within the Control Area. In addition to these, there were 6 positive backyard flocks - 4 within the High Risk Region and 2 within the Surveillance Region.

2.4 Progression of the outbreak during April-May

On April 5, a decision was made to depopulate the entire Control Area of poultry - an estimated 19 million birds. Further movement restrictions were imposed to stop the movement of poultry, poultry products and contaminated equipment within the Control Area and legal steps were taken to require poultry owners to control access to their property to prevent unauthorized entry.

By mid-April a third cluster of positive farms had emerged 5 km away, on South Lehman Road and Columbia Street. These were located in a very high density poultry area south of Abbotsford. Barns were often located within several hundred meters of each other and, once introduced, the disease appeared to spread locally from one farm to the next. Rapid depopulation of poultry barns within 3 km of infected premises became a primary focus of the disease control efforts.

The progression of the outbreak slowed towards the end of April, approximately 2 months after confirmation of infection on the index farm. It appears that movement controls and pre-emptive slaughter of all birds within 3 km of an infected farm were effective in limiting spread of the outbreak. In a number of cases, surveillance conducted during pre-emptive depopulation revealed flocks that were positive to the PCR test but were not showing clinical signs.

The last commercial premises infected with H7N3 was discovered on May 13. A number of this owner's flocks were found to be positive when tested at depopulation, which was staged over several days. The last positive backyard flock was found on May 18. Of the 42 commercial farms that were depopulated on the basis of a positive PCR test, 11 were found during the course of surveillance for pre-emptive slaughter or depopulation. No clinical signs were seen in these flocks, and further work was undertaken to determine if active infection with H7N3 was present. By June 23, the laboratory had isolated H7 virus from 28 of the 42 commercial premises and 2 of the 11 backyard flocks deemed to be infected on the basis of a positive PCR test. Highly pathogenic avian influenza virus isolates were confirmed on 27 commercial farms (Attachment 2).

One flock of ducks, not reported in the Table of Infected Premises, tested positive for the H11N9 strain of avian influenza. These birds had been destined for slaughter for meat and were uncovered as a result of regular pre-slaughter surveillance. This finding was considered incidental to the outbreak; therefore, the ducks were slaughtered and their meat was processed for human consumption. Similarly, there was no regulatory action taken when a flock of geese were found to be serologically positive for the H6 strain of the virus.

By May 21, 2004, the last of the farms within 3 km of all infected premises had been depopulated, and on the week of May 23, the last commercial flock was slaughtered (70,000 ducks that were not infected, but that were designated as a dangerous contact to the 42^nd infected farm). By then, approximately 17 million commercial poultry (90% of the estimated population) had been slaughtered within the Control Area. At the end of May, a few farms with replacement pullets coming into production were allowed to move birds within the Control Area under conditions specified in a permit. These conditions stated that:

• birds must originate from influenza-negative flocks (as confirmed through testing blood and swab samples),
• birds could not be moved or placed within 3 km of an infected premises, and
• the flock must participate in an ongoing dead-bird surveillance program.

2.5 Conclusion of the outbreak in June

On June 3, twenty-one days had elapsed since the last infected farm was discovered. This gave additional confidence that the outbreak had been contained, so that on June 4 all depopulation actions were officially suspended and on June 10 a new High Risk Region (HRR) was named - the municipal boundary of the City of Abbotsford. This allowed for orderly repopulation of farms that were located within the Control Area, but outside the new HRR. Hatching eggs could be moved into, but not out of the HRR. In addition, chicks from approved hatcheries could be placed on farms that were located within the Control Area, but outside the HRR. Other controls were maintained prohibiting the movement of hatching eggs, birds for slaughter and pet birds out of the Control Area. Dead bird surveillance was continued on 12 layer and breeder farms within the new High Risk Region with the understanding that any flocks representing a new outbreak of disease would be depopulated immediately.

By June 11, all sites where dead birds and/or manure were composted had achieved the time and temperature requirements to be deemed virus-free. Compost piles containing dead birds were maintained until complete carcass breakdown had occurred. Cleaning and disinfection procedures were concluded on 41 of the infected premises by June 18. The remaining farm, outside the HRR, was to remain under quarantine until the cleaning and disinfection was completed (projected to be the end of June).

3. Laboratory findings

NCFAD's role in the B.C. avian influenza outbreak initially involved:

• the characterization of the virus isolated by the Animal Health Centre, BCMAFF that originated from an older flock located on the index premise, and
• isolation of virus from tissue specimens sampled from a younger flock located on the index premise.

Characterization of the original allantoic fluid specimen obtained from BCMAFF employed:

• real-time RT-PCR assays specific for H5 and H7 hemagglutinin sub-types,
• H-typing by hemagglutination-inhibition assay,
• N-typing by neuraminidase-inhibition assay,
• conventional RT-PCR and cycle-sequencing to determine the amino acid sequence of the hemagglutinin cleavage site, and
• intravenous pathogenicity indexing.

The isolate derived from the older flock on the index premise was an H7N3 virus that possessed the typical waterfowl cleavage sequence PENPKTR*GLF. This isolate was predicted to be of low pathogenicity based on the cleavage site, and was subsequently shown to have an IVPI of 0.0. Tissue specimens derived from the younger flock were processed and inoculated into 9 day old embryonating chicken eggs. The virus isolated was also H7N3, however, the cleavage site differed by possessing a 21 nucleotide/7 amino acid insert. It had an IVPI of 2.96 and produced plaques on QT-35 cell monolayers in the absence of exogenously added trypsin.

During the course of the outbreak the laboratory employed a number of diagnostic assays. The majority of samples submitted to the laboratory for analysis were oropharyngeal and cloacal swab specimens stored in avian transport medium. In the early part of the investigation these were processed for inoculation of 9 day old embryonating chicken eggs and testing by a real-time RT-PCR assay targeting the matrix gene. Submissions that tested positive by the matrix real-time RT-PCR assay were tested using the H7 type-specific RT-PCR assay. Virus isolates were H-typed and N-typed, the hemagglutin cleavage site sequenced, and the pathogenicity confirmed by IVPI. As the investigation progressed, the matrix real-time RT-PCR was used as a screening assay. Positive reactors were subsequently processed for virus isolation, H-typing, and nucleic acid sequencing of the H gene segment.

By June 23 H7 viruses had been isolated from 28 infected commercial premises and 2 backyard flocks. All of the highly pathogenic isolates possessed a 21 nucleic acid/7 amino acid insert at the cleavage site. To date, six PENPK variants have been identified that have arisen by point mutations within the insert. An additional isolate, a PENPR variant, contains an point mutation at the cleavage site outside of the insertion.

Nevertheless, variants identified to date are highly pathogenic based on the intravenous inoculation of 4 to 6 week old specific pathogen free chickens. The majority of isolates possess IVPIs in the 2.9 to 3.0 range. One isolate had a slightly lower IVPI of 2.17.

4. Epidemiological Investigation

4.1 Information on tracing forward and backward at the outset of the outbreak

Because the index case was a broiler breeder farm, a potentially significant part of the follow up investigation included tracing of hatching eggs, which were all being sent to a single hatchery located within the Control Area. Chicks hatched from eggs laid from January 12 through to February 16 were tested at hatch or, for chicks in broiler barns, monitored and tested in the barns. This procedure was also followed for the second infected premises, which was also a chicken broiler breeder farm. In the light of the uniformly negative results of testing of eggs hatched from these two infected broiler breeder flocks, the CFIA concluded that the movement of chicks and hatching eggs did not pose a risk. It is generally agreed that if influenza virus is present in the egg the embryo would be killed by the virus and if the virus is on the surface of the egg it would be killed during the incubation process.

4.2 Subsequent traceout investigations

Traceout investigations have been conducted on all infected farms for the movement of products, birds, equipment and people within 21 days preceding the onset of clinical signs, or a positive test. Farm owners/operators were interviewed using a detailed questionnaire and, from the information obtained, a time line of events constructed for each farm. Attachment 4 shows the time line diagrams for all 12 farms infected in the critical period between January 12 (3 weeks in advance of clinical signs on the index premises) and March 31 (3 weeks after the imposition of movement controls on March 11).

The poultry industry in B.C. is highly integrated and localized in the Fraser Valley. During its investigation, the CFIA found only one example of potential risk products leaving the Control Area - eggs sent to hatcheries in another part in BC from a Leghorn breeder farm in the High Risk Region. The flock was depopulated on the basis of a positive PCR test, but virus could not be isolated and the eggs hatched outside the control area were deemed to be low risk

4.3 Transmission

Wild birds, especially waterfowl, act as reservoirs of avian influenza virus and may be the initial source of infection to domestic birds through direct contact or contamination of feed and water. Low pathogenicity viruses can acquire virulence over time due to antigenic drift (small mutations) or antigenic shift (larger genetic changes). Because the disease is highly contagious, it spreads rapidly through a confined commercial poultry flock.

High concentrations of virus are present in the respiratory and digestive tracts of infected birds. Fecal material from infected birds may contain up to 16 x 10 ⁶ virions/gm of feces and one gram contains enough virus to infect one million birds.

The routes of transmission of the virus between infected premises appear to be similar to those reported in outbreaks of highly pathogenic avian influenza overseas. Movement of people, equipment or birds is considered as the primary mode of transmission to new locations within the control area. Two positive farms have common ownership. Common usage of hatcheries, feed companies, and poultry industry service companies may also have contributed to the spread of disease. The investigation of each positive premises was based upon a veterinary visit, responses to a questionnaire (attachment 5) and follow up of movements of risk products and other potential avenues of disease transmission (attachments 6,7).

There were three clusters of infected farms all within a 20 kilometre radius:

• the High Risk Region, northwest of the City of Abbotsford
• a cluster directly west of the original high risk area, and
• a cluster south of the City of Abbotsford

In a few cases, outlying farms were positive, based on PCR tests, but did not show clinical signs and the disease did not spread locally. This was evident in the farms east and west of the Abbotsford region.

These three clusters have a diameter of 5-6 km. In this outbreak, it appears that once a farm within a densely populated poultry area became infected, nearby farms (often within several hundred meters of each other) also become infected. The majority of the Fraser Valley poultry producers (95%) use wood shavings and the remainder use sawdust for bedding material. Although wood shavings contain mainly coarse particles, they often contain extremely fine and light sawdust particles, which can readily become airborne if the bedding is disturbed. The large exhaust fans on modern poultry barns can evacuate high volumes of air and airborne dust could readily be drawn into the intakes of adjacent barns. Further studies are underway on the role of local, airborne spread of the virus on feathers or dust particles.

The working hypothesis of transmission in this outbreak is that LPAI was initially introduced into the older flock on the index farm. The virus then mutated into HPAI and infected the younger flock on the same farm. Transmission from farm to farm over large distances was likely due to the movement of infected birds (early in the outbreak) or the result of mechanical spread by movement of people and contaminated equipment.. However once infected, a flock of 8-10 thousand birds essentially becomes a "virus factory". If located in a densely populated area, it may produce enough virus to infect nearby flocks through airborne transmission of virus on dust particles or feather debris.

5. Control Actions

CFIA controls were based on three major disease control principles, namely :

• rapid detection of infected flocks (surveillance)
• rapid destruction of infected flocks, high risk contact flocks and proximal flocks (within 3 km)
• effective biocontainment of infective material (carcasses, manure and feed).

5.1 Surveillance

Initially, a surveillance plan was developed for all commercial farms within 5 km of the index premises. Once the High Risk and Surveillance Regions were designated, the surveillance protocol was expanded to include all commercial farms within these regions. When the plan to depopulate the entire Control Area was announced, active surveillance targeted commercial farms within 3 km of an infected premises.

Active surveillance initially employed serological methods surveillance, but this was found to be operationally difficult and was considered to be relatively ineffective in detecting recently acquired infections.

Swabbing the oropharynx and cloaca was adopted as the method of choice for testing live and dead birds. PCR-based tests can be performed quickly and are very sensitive. Swabbing dead birds from commercial flocks on a weekly (or more frequent) basis was both operationally and epidemiologically efficient.

In addition to these active surveillance sampling strategies, a heightened level of passive surveillance was implemented throughout the area during the entire outbreak. Public information campaigns were used to educate poultry owners on the signs of avian influenza and to encourage prompt reporting of sick birds or increased mortality.

The poultry industry throughout Canada was also on the alert for AI, as evidenced by the marked increase in submissions to the National Centre for Foreign Animal Disease (NCFAD) from flocks outside the Control Area. In the period March 16 to April 22, NCFAD received 13 AI suspect submissions from Manitoba and Alberta, as compared to 15 submissions for the 5 years 1997-2002.

5.2 Destruction of infected and high risk flocks

Rapid destruction of high risk flocks was essential in controlling the spread of disease. High risk flocks included those with common ownership or shared equipment and those within a 3 km radius of the infected flock. The goal was to destroy these flocks within 24-48 hours of finding a positive flock.

Humane euthanasia of commercial flocks was accomplished in situ through the rapid application of carbon dioxide gas after sealing the barn. To facilitate disposal, layer hens were removed from cages before they were killed. A small portable carbon dioxide gas chamber was used to euthanise small backyard flocks. This method of destruction was used for all but one infected flock.

5.3 Disposal

Incineration, landfill and composting were the methods used for disposal of carcasses of infected birds. Some carcasses, most often layer hens, were removed from the barns and placed in biosecure containers to be transported to incineration or landfill sites. Wherever possible, birds, eggs and litter/manure were composted/bio-heat treated in the barn. Once the compost had been held at 30 degrees Celsius for a minimum 3 consecutive days, additional composting could take place outside the barn. Wood shaving or sawdust bedding was moistened to minimize aerosols prior to movement of the compost material out of the barn.

Where practical, tested free flocks were processed at designated slaughter facilities within the Control Area and their meat consigned direct to retail outlets within Canada. Poultry products processed within the Control Area could not be certified for export. If slaughter was not practicable, these flocks were euthanized on site and carcases taken to rendering, landfill or communal composting sites.

5.4 Cleaning and disinfection of facilities and equipment

Cleaning and disinfection (C&D) is physically and financially the responsibility of the poultry producer. Infected premises must be cleaned and disinfected before any restocking is allowed, and all producers have been provided with guidance and recommendations regarding the appropriate protocols to use. Equipment on infected premises must also be cleaned and disinfected and may not leave the premises until C&D has been approved by the CFIA. Farmers may chose to perform cleaning and disinfection procedures themselves or to hire the services of contracting agencies.

Before cleaning is commenced, a CFIA inspector visits the farm to make suggestions and to highlight any areas of particular concern. After the CFIA inspection team is satisfied with the cleanliness of the facility and equipment, disinfection is permitted. During disinfection CFIA inspectors ensure that Health Canada-approved virucidal disinfectants are applied according to label directions. Only then can the premises be officially declared as cleaned and disinfected.

5.5 Plans for Repopulation

After positive premises were satisfactorily cleaned and disinfected and CFIA was satisfied that there was no risk of reinfection, repopulation of poultry barns was scheduled to commence in the Fraser Valley Control Area. This process began on June 10, with the establishment of the new High Risk Region (City of Abbotsford) and the lifting of some movement restrictions (see attachment 10). Farms within the Control Area, but outside the new HRR were to be repopulated first (June 10). Farms inside the HRR could be restocked next (beginning July 9), 21 days after the last infected premises had satisfactorily completed cleaning and disinfection, on June 18.

All infected premises remained under quarantine and a surveillance plan was developed such that if poultry were restocked less than 60 days after the farm had been cleaned and disinfected, then birds would be tested at the end of the first, second, third and fourth weeks after the birds were placed. However, if the barns had been empty for more than 60 days after the farm’s C&D was approved, then testing would not be required. Similarly, surveillance testing would not be required on uninfected farms that were being repopulated.

5.6 Actions taken at the US border (see attachments 8, 9)

In addition to the prohibition on the movements of poultry and poultry products across the Canada-US border additional logistic issues had to be resolved. Since the main supply depot of carbon dioxide (used for euthanizing birds) was located just south of the border, in Washington State, a protocol was developed in conjunction with USDA to ensure that carbon dioxide delivery vehicles were thoroughly cleaned and disinfected between visits to infected premises and re-entry into the US. A transfer system was developed whereby a large tanker would haul carbon dioxide from the depot to two smaller tankers. These smaller tankers would then deliver the gas to the farms.

There was a CFIA-designated truck wash station in the control area, approximately 5 kilometres north of the U.S. border. Any vehicles servicing the live poultry industry (e.g. carrying feed, litter, cages, or equipment) had to be cleaned and had to obtain a CFIA certificate prior to crossing the border.

6. Ongoing investigations

In conjunction with other agencies, the CFIA continues to study topics relevant to the disease outbreak, including airborne transmission, the role of hatcheries, wild birds and service providers, disposal methods and database management/integration with GIS.

Prepared by:
Dr Wayne Lees (wlees@inspection.gc.ca) and Lawana Chown (chownlb@inspection.gc.ca)
Dr Carolyn Inch (cinch@inspection.gc.ca)
with help from: Drs Ken Moll, Peter Brassel and Tong Wu.
Edited by Dr Sarah Kahn (kahns@inspection.gc.ca)
Information on laboratory analysis contributed by Dr John Pasick (pasickj@inspection.gc.ca)

List of Attachments

1.  BC Poultry Industry Profile

2.  Table showing Infected Premises

3.  Map of Infected Premises

4.  Time lines - Infected Premises

5.  CFIA Producer Questionnaire

6.  High Risk Contact Determination

7.  Hatchery as a potential source of AI transmission

8.  Vehicle Cleaning and Disinfection Protocol

9.  CFIA Certificate of Cleaning and Disinfection

10. Map of New High Risk Region (June 10)