H5 Test Failure in Winnipeg Raises Surveillance Concerns
Recombinomics Commentary
June 21, 2006
The Winnipeg lab will attempt to grow virus from samples from the index bird in order to characterize the virus. This process will take up to two weeks. Ultimately, it may not be possible to gain further information about the virus. This situation is not unusual as was evidenced in the 2005 survey of AI in wild birds.
The above comments from the Canada Food Service Program compare the testing of samples collected in August 1005 from healthy young wild birds to the testing of samples from a dead free range goose collected on Prince Edward Island on June 5, 2006.
The results from the August collections are at the Canadian Wildlife Health Center. The data indicated bird flu was frequently detected in the wild birds and almost 20% of the samples were positive for H5. Thus, the testing showed that samples properly collected and stored for several months could still yield positive PCR data as well as isolated virus.
The detection frequency was among the highest in countries that have increase surveillance of wild birds and screening for H5N1. Only one country, Russia, has been able to detect H5N1 in live birds. Collections of healthy birds shot by hunters have tested positive for the Qinghai strain of H5N1. In Europe many countries have detected H5N1 in wild birds, but the positive samples are from birds that have died. Initial positives were in mute swans near the Volga Delta or Danube Delta. The swans are easily noticed and deaths of swans were observed by local residents prior to testing. Hundreds of dead swans have tested positive fro H5N1 and virus has been frequently isolated. Several European countries have also found H5N1 in dead or dying birds on farms, but these tests were in countries that had already found H5N1 in wild birds.
Other countries, however, initially found H5N1 in birds on farms. Most of these countries had poor surveillance programs. Many found no bird flu in wild birds, indicating the collection testing process was flawed because low pathogenic avian influenza is common in wild birds and the low path virus should have been detected if samples were properly collected and tested.
In some countries, like India and Indonesia, a small number of samples are collected, which decreases the likelihood of positives. When the first human cases in Indonesia were identifies in July of 2005, animal samples were collected in neighboring communities. However, the number of samples collected was small, indicating the effort was not serious. The number of human cases has steadily increased and surveillance issues in Indonesia are still of concern.
In India, the failure to find H5N1 in wild birds is also suspect. Many instances of unexplained wild bird deaths have been investigated, but the number of samples tested are minimal. Until this year, India denied H5n1 in poultry or people, yet serum samples collected in 2002 from poultry workers were positive fro H5N1. Moreover, the Qinghai strain of H5N1 was identified in bar headed geese at Qinghai Lake in 2005. These birds winter I Indian so H5n1 in bar-headed geese in India is likely. This year H5N1 infections in bar headed geese in Qinghai province was found again, yet India still maintains that H5N1 is not in wild birds. H5N1 has now been acknowledged in domestic poultry, but media reports indicate the sequences in the domestic poultry match wild bird isolates in Qinghai and Xinjiang provinces in China.
Countries in Africa had also had difficulties in detecting H5N1 in wild birds. Initial detection has almost exclusively been on farms. This limitation is almost certainly linked to a poor surveillance system. Detection on farms is easier because the dead or dying birds are rich sources of H5N1and the deaths usually involve multiple birds.
Thus, the failure of Winnipeg to characterize H5 from a dead goose on a farm on Prince Edward Island is cause for concern. The failure reflects a poor decision to collect samples from a single dead bird followed by delays in shipment to Winnipeg. These delays almost certainly contributed to sample degradation and failure to detect the confirmed H5 in the index goose. The detection failure suggests that culturing will also be unsuccessful. In may instances PCR positive samples fail to generate isolated virus and the frequency of viral isolation from negative samples is low.
The Winnipeg failure raises surveillance issues. The proper collection, packaging, and/or shipment procedures appear to be lacking, generating false negatives which are cause for concern.
Media Sources
Phylogentic Tree