H5N1 Human to Human Transmission Media Myth
Recombinomics Commentary
December 21, 2007
Limited human-to-human bird flu transmission may have occurred in Pakistan, but no new infections have been reported for two weeks and there appears to be no threat of further spread, a top World Health Organization official said Friday.
A WHO team has finished its initial investigation in Pakistan after up to nine patients, including several family members, were suspected of being infected with the H5N1 bird flu virus in areas north of Islamabad. They were the country's first reported human cases.
The experts were expected back in Geneva to begin piecing together how the virus may have spread, but they found no evidence of anyone currently sickened by the virus, Dr. David Heymann, the WHO's top flu official in Geneva, told The Associated Press by telephone.
Most cases have been linked to contact with infected birds, but scientists believe limited human-to-human transmission has occurred a few times before among blood relatives who had close contact.
Five brothers were sickened last month in the small city of Abbotabad, about 50 kilometers (30 miles) north of Islamabad. One was a veterinarian who was involved in slaughtering sick poultry infected with bird flu. Two of his brothers fell ill and died, one of whom was buried before tests were conducted. The veterinarian and his other two brothers recovered.
Up to five other people in the same area also tested positive for the virus in preliminary tests.
The above comments perpetuate a number of media myths concerning human to human H5N1 transmission. The myths are clearly contradicted by WHO confirmed cases. These statements have no scientific basis, but the propagation by the popular press continues.
Although WHO has yet to put out a detailed situation update on the confirmed cases, a consensus from media reports strongly suggests that this is the longest and most sustained H5N1 infection reported to date. The longer chain appears to involve five brothers and a cousin, as well as one or two health care workers. Media reports indicate the index case developed symptoms on October 25. Two brothers were fatally infected after visiting their brother in the hospital. The two brothers died November 19 and 29. Media reports suggest two health care workers also tested positive or had bird flu symptoms and appear to be affiliated with the same Kyhber Teaching Hospital that treated several family members.
The health care workers, who are unlikely to be blood relatives, are not the first unrelated members of a cluster. A family of five in Haiphong, Vietnam was confirmed H5N1 positive. The husband and wife were both positive. Another cluster in Vietnam included a brother and sister, a confirmed health care worker, and an unconfirmed health care worker with symptoms. A cluster in Azerbaijan included a friend in addition to family members.
However, in addition to clusters involving contacts who were not blood relatives, there have been many clusters involving family members with disease onset dates with gaps of 5-10 days. Many of the earlier clusters in Vietnam and Thailand were documented in a peer reviewed publication authored by WHO consultants.
Similar peer reviewed publications by WHO consultants have describe three likely H2H clusters in Indonesia in 2005, as well as the Azerbaijan cluster in 2006. These clusters were in addition to the Karo cluster in Indonesia in 2006 and the cluster in Thailand in 2004.
Most clusters involve two family members, and in many instances samples are not collected from the index case prior to death. In other instances, such as the many clusters in Turkey, the samples that were H5N1 positive in local testing in Turkey, were not confirmed after shipment to a WHO reference center in Mill Hill, England. The pairing of the positives in Turkey strongly suggested the positives were true positives, and the reduction of 21 positives in Turkey to 12 positives in Mill Hill was due to sample degradation. Moreover, sequences from only four isolates from the 12 positives were released, providing additional evidence for sample degradation and false negatives.
The number of H5N1 human clusters has been large, and most clusters have gaps in disease onset dates.
These growing transmission chains provide additional evidence for a more efficiently transmitted H5N1 that is not linked to genetic factors.
The latest cluster is almost certainly linked to the Qinghai clade 2.2 sub-clade, which has PB2 E627K fixed, which allows the H5N1 to replicate more efficiently at lower temperatures. Human clusters in 2006 and 2007 have additional changes in the receptor binding domain (including S227N, N186K, N186S, Q196R, V223I, M230I) , which would also facilitate a more efficient transmission.
Changes in the receptor binding domain many play a role in the current extended cluster. Sequence data on the cases and clusters in Pakistan would be useful.
Media Links
Recombinomics Presentations
Recombinomics Publications
Recombinomics Paper at Nature Precedings