WHO Hoarding of H5N1 Sequences Raises Pandemic Concerns
Recombinomics Commentary
December 23, 2007
Rainford declined to predict today how soon confirmatory testing in the eight cases, which were identified in preliminary tests by Pakistan's national lab, will be completed. He said some samples from the patients would be sent to London for analysis.
"I hope we'll get some early results but question whether we'll have lab results that are what we need," he said. "We've got, hopefully, samples arriving in London sometime over the weekend that will provide a better environment to assess what we've got."
He said he expected both the NAMRU lab and the one in London will play a role in the testing: "NAMRU staff on site will be able to give us some information, but I guess there's a role for an additional lab to give us a full picture."
The above comments remain curious. It is not clear if the samples are being sent to London for receptor binding studies, sequencing data, or both. It is likely that the H5N1 sequences from Pakistan have receptor binding domain changes, which require affinity binding studies to help determine the significance. However, the implementation of a Tamiflu blanket may lower RNA levels in clinical samples which may be causing problems for the NAMRU-3 mobile lab, which may not be as fined tuned as the region labs in Cairo or Mill Mill / Weybridge, which may be leading to detection problems or a lack of sequencing information.
Sequence data coming out of London as led to extensive delays in release of the information. Many sequences from samples collected in 2005 and early 2006 still have not been released by Weybridge. Similarly, human H5N1 sequences generated by Mill Hill were withheld from the public for almost a year.
The paper published in Nature at this time last year described receptor binding domain changes in H5N1 in southeast Asia. The sequence data presented indicated the patients were infected with H5N1 mixtures with different combinations of receptor binding domain sequences. These mixtures were consistent with earlier data also published in Nature, showing that patients also had combinations of Tamiflu resistance markers also (H274Y and N294S). These data were important because N294S was found in the Gharbiya cluster in samples collected prior to Tamiflu treatment, indicating Tamiflu resistance markers were circulating in wild bird sequences in the region.
The paper on receptor binding domain changes in southeast Asia mentioned that the same changes had been seen in Iraq and Azerbaijan. However, these claims could not be independently confirmed because the sequences were being hoarded in the WHO private database. The sequences were subsequently released by Mill Hill and had an alarming number of receptor binding changes found in H5N1 from patient clusters, providing a genetic mechanism for the human to human transmission that was actively denied in WHO press releases.
The lack of transparency increased pandemic concerns, and the delays in the release of details, coupled with continued hoarding of sequences further increases the concerns.
Earlier this year there was an H5N1 outbreak in Kuwait that led to the culling of over 1 million birds. Those sequences have not been released, although comments from DEFRA indicated the sequences were closely related to the Uva Lake sequences released from three outbreaks over the summer. Similarly, the sequences from the Czech Republic and France reporting additional outbreaks over the summer have now been released. More recently, Weighbridge collected samples from the outbreak in England. Although DEFRA has indicated the sequences are closely related to the other Uva Lake sequences, the actually sequence data has been withheld.
Similarly, sequences from Pakistan have been withheld. OIE reports detail poultry outbreaks in 2006 and 2007, but no H5N1 sequences from Pakistan have been released. 2006 sequences from Afghanistan and India have been released, but both countries had outbreaks this year, as did neighboring Bangladesh and Myanmar, but none of those sequences have been release. Recently Myanmar reported the first human H5N1 case, but that sequence has not been released.
The past hoarding of sequences by WHO consultants using the WHO password protected sequence database has been discussed previously, and the increased hoarding this year has raised significant transparency issues.
Ironically, WHO has complained about the lack of transparency in Indonesia, while WHO consultants withheld record numbers of sequences.
This lack of transparency does not inspire confidence, press releases and speeches not withstanding.
The time for release of these sequences as well as those generated for Pakistan has long since passed.
Media Links
Recombinomics Presentations
Recombinomics Publications
Recombinomics Paper at Nature Precedings