.
H5N1 Answers and Questions - Human Transmission
Recombinomics Commentary
May 19, 2005
The WHO Manila report answers several questions and the answers raise several questions. The start of a bird flu pandemic is largely dependent on efficient human-to-human transmission. The latest announcement cites several reads why human-to-human transmission has moved into the probable rather than possible category.
However, human-to-human H5N1 transmission in Vietnam and Thailand has been probable since last season. A significant clue came from a cluster in northern Vietnam (Thai Binh) in January of 2004. A groom had developed bird flu symptoms. He was hospitalized and died before samples could be collected. However, his two sisters who cared for him developed symptoms about five days after the index case. They showed symptoms on the same day, were hospitalized on the same day, initially tested as inconclusive for H5N1, subsequently tested positive fro H5N1 and died the same day, within 1 hour of each other.
Distinguishing bird flu infection due to human-to-human transmission is heavily dependent on disease onset dates. Cases within a family will be likely be infected by a similar virus, regardless of transmission route. Even if the virus is different, H5N1 infection are frequently heterogeneous, so differences could arise via either transmission mode.
In this cluster there was no sample from the index case, and the two sisters did have different clinical presentations (one was respiratory and the other gastro-intestinal) and different viruses were isolated. However, the onset and progression dates made a common source more likely and the common source was probably their brother.
However, as single cluster is not definitive, but there were additional small familial cluster at the beginning of 2004 and all had a bimodal distribution of onset dates, indicating human-to-human transmission was probable. This probability moved to near certainty last summer.
At the end of July there was a familial cluster near Hau Giang in Vietnam that involved cousins who developed symptoms are about the same time. Again there wee no samples collected, but the older sister of one of the index cases developed symptoms after her brother died and she tested positive for H5N1. Thus, the bimodal distribution again made human-to-human transmission very likely more all or most of the clusters.
A few weeks later there was another familial cluster of three. This cluster was in Thailand and was written up in the New England Journal of Medicine. It was easier to classify as human-to-human transmission because the mother of the index case was an office worker hundreds of miles away from her daughter who was staying with the aunt of the index case. The mother developed symptoms after visiting her daughter, as did the aunt and both adults tested positive for H5N1 (again no samples had been collected from the index case). Human to human transmission was considered likely because of the distance between the index case and her mother at the time of onset, although the bimodal distribution and positive results in the secondary transmission matched the earlier clusters, raising the likelihood on human to human transmission to near certainty.
At the beginning of this year, similar clusters began to appear, especially in northern Vietnam. They also were bimodal, and the clusters became more frequent, grew in size and transmission chain length, and the number of cases in clusters grew to over one third of the total.
However, even with this overwhelming evidence for human-to-human transmission, WHO only said such transmission was possible and such transmission were dead ends. Unfortunately for the family members, many were literal dead ends, and there was not warning that human-to-human transmission within families was common. Most media reports in fact cited the cluster in Thailand as the only example and indicated that almost all cases were from exposure to birds.
Recently, human-to-human transmission of H5N1 was termed probable instead of possible, however such transmissions were virtually certain last season, and quite common at the beginning of this season.
Similarly, although the genetic changes were publicly described I detail for the first time in the Manila report, references to changes were made earlier and clearly the data supporting the sequence changes has also been know for some time, although none of the 2005 sequences are publicly available at GenBank, although at this time last year over 100 2004 sequences were publicly available.
Media link