Commentary
Paradigm Shift on Influenza Research Collaboration?
Recombinomics Commentary 17:34
May 14, 2008
Keiji Fukuda, coordinator of WHO's global influenza programme, said that more sharing of research was required, as happened informally between experts during the deadly SARS epidemic in 2003 which spread from southern China to Canada.
"In developing a WHO public health research agenda (on influenza) we are trying to push for a paradigm change," Fukuda told the final session of a four-day WHO meeting on bird flu.
"What we hope to improve is the kind of sharing and flow of information and take it to another level," he said.
Most research is driven by individual researchers but depends on the source of funding, so there is also a need to sensitise donors about priorities, according to Fukuda.
"It is really quite secretive until the information is published. It is like a poker game," he told Reuters.
The above comments on the need for a paradigm shift on sharing of influenza research data is welcome. Since some of the largest hoarders of information are WHO collaborating centers and consultants, WHO has considerable influence, but this situation has remained unchanged for years.
WHO operates multiple restricted database that are only accessible by a small number of WHO affiliates. The most well known database is of H5 sequences, but WHO also has a similar database for swine influenza sequences and likely has another for human influenza sequences. Although WHO has talked of making these sequences public, there has been little movement on this front, and the number of hidden sequences is likely to be at record levels. In addition to the sequences hidden at WHO databases are additional sequences withheld by individual sequencers. These withheld sequences create large holes in the sequence database, which limits research and predictions on the evolution of influenza, including H5N1.
It is not clear if recent events have precipitated the above comments. H5N1 outbreaks have recently been at record levels in India, Bangladesh, and South Korea. None of these countries have released data from bird isolates, and all have denied human cases, although there is little doubt that there have been human H5N1 infections in all three countries.
The most glaring has been South Korea, where a soldier has tested positive for high path H5. He developed pneumonia and was unconscious while culling H5N1 confirmed poultry. Korea did not deny confirmation of H5, but a lack of confirmation of the N1 led to an announcement that the soldier tested negative for H5N1. Korea has redefined the definition of a confirmed H5N1 case and has not reported the infection. Instead it issued press releases that have been picked up by wire services and the wire service reports have been cited as evidence that there have been no human H5N1 cases in Korea. This misinformation is being distributed to US military personnel stationed in South Korea.
However, recent media reports have suggested that the H5N1 in Korea may be related to H5N1 in Vietnam. Samples have been sent to the CDC in Atlanta, although it is unclear if the samples sent were from poultry, the soldier, or both. In any event, none of these agencies, including those analyzing samples from H5N1 infected whooper swans in Japan have released any sequence data or provided a clearer description of the relationship between the H5N1 in circulation in South Korea / Japan and earlier clade 2.3, 2.2, or 1.
The hoarding of sequences however extends well beyond the record outbreaks in India, Bangladesh, and Korea. WHO collaborating centers frequently get samples collected by local public health agencies which have detected H5 in suspect human or bird samples. Material is mailed to WHO centers for confirmation, which leads to sequencing of the samples and control of that information. The WHO centers the hold the data for a number of reasons, which includes publication. These sequences can also be used to show that H5N1 surveillance is poor, and the poor surveillance extends to the collaborating center. Similarly, the data can be used to show that the analysis done by the centers is either outmoded or incorrect, which are additional reasons for withholding of the data.
Recently, samples collected two years ago in Nigeria were released. These sequences which were collected in January, 2006 demonstrated that sequences collected at later dates in Europe and Egypt were similar, indicating H5N1 had migrated into these upstream locations in the fall of 2005 and then migrated to western Africa. However, fall 2005 infections were not reported by any country in central or western Europe or the Middle East, other than Turkey.
The WHO collaborators and consultants are also in denial about the role of recombination in influenza evolution in general and H5N1 evolution in particular. They maintain that genetic drift is due to random copy errors and therefore is not predictable. These views greatly limit vaccine targeting, which currently involve chase evolving H5N1 after it has emerged and become established. These outmoded views can only be maintained in the absence of sequence data, so the longer the sequences are withheld, the longer the outmoded approaches remain viable or defendable
The call for a paradigm shift is welcome, and WHO should, begin implementing the change by targeting its consultants and collaborative centers.
Media Links
Recombinomics Presentations
Recombinomics Publications
Recombinomics Paper at Nature Precedings