Outbreak Investigation Report on H5N2 Avian Influenza in Ontario, 2015
3. Overview of Outbreak

3.1 Initial Detection

On April 3rd, 2015, samples from IP1 were sent by a private veterinary practitioner to the provincial Animal Health Laboratory (ALH) in Guelph, Ontario. The veterinarian had been called to the farm following a pattern (first noticed on April 1st, 2015) of increasing mortality in one barn, which contained the oldest birds. On April 5th, 2015, the AHL reported the detection of an H5 NAI in the submitted samples. Samples were then sent to the CFIA National Centre for Foreign Animal Diseases (NCFAD) in Winnipeg, Manitoba, which confirmed the virus was H5N2 on April 8th, 2015.

3.2 Findings on Infected Premises 1

The first infected premises (IP1) is a meat turkey operation located near Woodstock, Ontario. The operation is composed of four barns with birds at multiple stages of production, from 6.5 to 16.5 weeks of age. Each barn was initially stocked with 10,000 to 12,000 birds, all of which were around six weeks of age and obtained from a nearby brooder farm. All birds from this operation went directly to slaughter and all birds were obtained from the same brooder farm.

As noted previously, the premises contained four barns. Increasing mortality was noted in the first barn (B1), with approximately 4,500 birds remaining out of 9,900 birds initially placed. Birds in barns 2, 3 and 4 were 14.5, 8.5 and 6.5 weeks of age, respectively, with no significant mortalities.

Following a detailed epidemiological investigation, one high-risk premises was identified. This premises was the brooder barn that provided turkeys for IP1. The manager of this farm was a relative of the managers on IP1 and also assisted on the property as needed. As a result, this farm was placed under quarantine and monitoring for 21 days (three viral incubation periods) following the last contact between the premises. No evidence of infection was detected during this period.

The likely source of infection for this premises was a breach in biosecurity. CFIA staff on site reported significant numbers of small rodents within the barns as well as at least seven wild birds observed in one of the barns following destruction. Additionally, there were reports of wild waterfowl on the property in the weeks preceding infection.

3.3 Findings on Infected Premises 2

IP2 was a broiler breeder farm approximately 38 km from IP1 and was therefore outside the 10-km control zone.  Following a pattern of increasing mortality in one barn, samples were initially submitted to AHL by a private veterinary practitioner on April 17, 2015 and the subsequent H5 detection was reported to CFIA on April 18, 2015. H5N2 was confirmed by CFIA National Centre for Foreign Animal Disease (NCFAD) on April 19, 2015. 

The farm consisted of two connected barns with 12,000 to 13,000 birds per barn.  A ratio of 90% hens to 10% roosters was maintained, and all birds were approximately 29 weeks of age.

Following a detailed epidemiological investigation, no linkages between IP1 and IP2 were identified.

It is believed that this farm was also infected via a breach in biosecurity, b the evidence for this is not as strong as for IP1. In addition to farm management, five employees were active on this farm. All were interviewed and reported good biosecurity practices (boot wash / clothing change).  Some wild birds were observed on a compost pile located between the barns, but there were no reports of rodents or wild birds in the barns, so they cannot be confirmed as the source of infection.

The private veterinary practitioner who initially took samples for submission had performed initial post-mortem diagnostics in his clinic. This clinic was placed under quarantine until appropriate cleaning and disinfection could be verified. Also placed under quarantine were two broiler flocks visited by the veterinarian subsequent to the post mortem. The quarantine on these birds was lifted following negative diagnostic sampling.

3.4 Findings on Infected Premises 3

On April 23rd, 2015 CFIA was contacted by a turkey breeder operation with high relative mortality in one barn. Samples were collected and taken to AHL by CFIA staff and a resulting H5 positive was reported on the same day. H5N2 confirmatory results were reported by CFIA NCFAD on April 25, 2015. IP3 is located in close proximity to IP2, with an estimated 600 meters separating the closest barns.

This farm was composed of approximately 8,000 breeding turkeys located in four barns. Barns 1 and 2 were located near the road, and barns 3 and 4 were set approximately 400 meters further back.  Due to the nature of production, birds of different ages were maintained in the barns.

Given the nature and value of the birds located on IP3, and the inability of a genetics facility to make use of an all-in, all-out production cycle, an extremely stringent biosecurity protocol was described. A relatively large number (approximately 20) of staff were present on this site, all of whom were solely employed on this site.

The most likely source of infection for this premises was windborne transmission from IP2. As noted previously, the distance between the two properties was approximately 600 m. Contemporaneous records from nearby weather stations (in London and Woodstock, Ontario) indicated a strong wind in a south-easterly direction, supporting of a direct wind connection between IP2 and IP3, an observation shared by CFIA staff on IP2 on April 18, 2015 (five days prior to onset of clinical signs on IP3). This connection is further supported by the pattern of infection observed, wherein birds in the barn closest to IP2 were the first to show clinical signs. Further analysis from the CFIA NCFAD laboratory confirmed a higher degree of homology between the virus detected at IP2 and IP3 than between either site and IP1, though all strains were highly homologous with only minor variation between the locations thought to be due to random mutation. A breach in biosecurity is a less likely source of infection on this farm due to the stringent biosecurity protocol for this premises and its temporal and geographical proximity to IP2.

3.5 Laboratory Findings

Sequencing of the H5N2 virus obtained from samples of poultry on infected farms and analysis of the results indicated that it was a reassortant virus. The genome of all Influenza A viruses contains eight RNA gene segments, including hemagglutinin (H) and neuraminidase (N) genes, of which the H is the main gene segment involved in determining the pathogenicity. Based on sequence analysis, avian influenza viruses can be broadly categorized into Eurasian (Europe and Asia) and North American lineages. The new H5N2 virus detected in BC contains five of eight gene segments from highly pathogenic Eurasian H5N8 virus, including the H5 gene, and three of eight segments from typical North American viruses, including the N2 gene (originating in wild birds). It was the first time a Eurasian HPAI H5 lineage virus had been a cause of outbreaks of avian influenza in domestic poultry in North America. In addition, it appears that this particular reassortant virus with Eurasian and North American avian influenza virus gene segments had not been observed anywhere before, although all eight gene segments have been identified in either Asia/Europe or North America.

3.6 Outbreak Conclusion and Post-Outbreak Surveillance

Following destruction of the birds on each IP, carcasses, products and by-products were composted on site to deactivate the virus via bio-heat treatment (BHT). This was followed by cleaning and disinfection of the barns, then a 21-day fallow period. Cleaning and disinfection was approved for IP1 and IP3 on June 29th, 2015 and for IP2 on July 8th, 2015. The post-outbreak surveillance period started on July 8th, 2015 and concluded on October 8th, 2015.