Evidence for Human to Human H5N1 Transmission in China
Recombinomics Commentary 16:41
January 26, 2009
of January 2009, the 1st in China since February 2008, occurred in 4
different provinces and appear to be unconnected. In all cases, there
is evidence of contact with diseased poultry. There is no evidence of
human-to-human transmission of infection from affected individuals to
their contacts or care givers.
The above remarks from a ProMED commentary on the cases in China are curious and unfortunate. ProMED is an infectious disease newsletter that is widely read by media writers and the lay public, who have minimal background in infectious diseases. As the commentator knows, the strongest data for human to human transmission are gaps in disease onset dates between two or more contacts who have symptoms and/or are lab confirmed.
The evidence for human to human transmission is the same evidence that has been repeated dozens of times since H5N1 exploded out of China in late 2003. The clusters of human H5N1 have obvious gaps in disease onset dates and have been reported dozens of times. Testing remains at the abysmal level. Samples are frequently not collected from the index case, and false negatives and misdiagnosis are common. In the latest cases, discussed above, the child (2F) from Hunan developed symptoms when her mother died of pneumonia. The child was H5N1 lab confirmed. In Beijing, multiple media reports indicated the nurse linked to the fatal H5N1 case (19F), developed symptoms, but recovered. One media report indicated the nurse was confirmed and none of the reports indicated the nurse tested negative for H5N1. Moreover, extensive testing failed to identify diseased poultry. The last confirmation of H5N1 in poultry in China was in December in Jiangsu, which was not directly linked to any of the subsequent confirmed cases in China.
However, the history of H5N1 clusters since late 2003 provides the most compelling argument that the recent clusters in China are virtually identical in time and space to prior clusters.
One of the first reported clusters was in Vietnam in early 2004. A recently married groom developed symptoms and died with bird flu symptoms in the hospital, but was never tested. His two sisters who cared for him developed symptoms on the same day, initially tested as “inconclusive”, were hospitalized on the same day, were retested and H5N1 confirmed on the same day, and died within an hour of each other. The gap in disease onset dates indicated both were infected by their brother, although the index case was not confirmed.
Later that year one of the most cited clusters was in Thailand and written up in the New England Journal of Medicine. The index case was living with her aunt, hundreds of miles from Bangkok, where her mother worked in an office and had no exposure to poultry. The index case developed symptoms after burying a pet bird, but was diagnosed as dengue fever. Consequently, the mother had no protection during hospital visits. After the index case died, the mother returned to Bangkok and developed symptoms, but also was not tested, even though her daughter had just died with symptoms. The aunt also developed symptoms and also tested negative, but was positive on a re-test. The mother was serendipitously identified just before cremation, and fixed tissue yielded H5N1 sequences. The index case was never tested for H5N1.
The index case for the first confirmed case in Cambodia collected dead chickens in early 2005. He developed symptoms and died, but was never tested. His sister subsequently developed symptoms and subsequently was hospitalized in Vietnam, where she was lab confirmed after she died.
In 2005 there were multiple clusters in Vietnam. The clusters in 2004 and 2005 were written up and published by the CDC. The vast majority of the clusters had gaps in disease onset dates.
The first confirmed case in Indonesia was also a cluster. The index case was initially diagnosed as having bacterial pneumonia. When she was finally tested for H5N1, her tittered had already risen to 64, as confirmed by the CDC and Hong Kong. A seconad sample, collected three days later had an even higher titer, 128, but since the titer wasn’t four fold higher than the initially collection, which should have been collected on admission, the case was not confirmed. Her sister subsequently developed symptoms and died, but was never tested. Her father was PCR confirmed and the isolated H5N1 is the current vaccine target for clade 2.1.
Clusters are common in Indonesia, and all initial confirmations were from clusters. The Tangerang cluster above was followed by a cluster involving an index case and her nephew. The aunt died and H5N1 was isolated, but her nephew’s case was mild. H5N1 was detected because he was among contacts tested, but H5N1 wasn’t isolated. That cluster was followed by a mild cluster on Sumatra, but virus was not isolated from any of the confirmed cases, and most were confirmed by antibody level. These three clsuters were also reported in the New England Journal of Medicine.
Subsequent reported cases from Indonesia had a higher case fatality rate. On Sumatra in 2006 six of seven members of the Karo cluster died. That cluster involved the index case infecting five family members including a brother who infected his son. That H2H2H cluster was easily identified, but no sample was collected from the index case.
An even larger cluster was a Garut the following year. Three independent clusters were identified with an H5N1 confirmed case in each cluster. However, none of the index cases fro the three clusters were tested and all died. Subsequent contacts developed symptoms and were hospitalized, but use of a Tamiflu blanket lowered viral loads and all symptomatic contacts tested negative.
More recently, Indonesia has ignored obvious misdiagnosis in cluster members, even after H5N1 has been confirmed. A series of such clusters involving fatal infections in index cases as lung inflammation, dengue fever, and typhus preceded the news blackout and reporting delays fro H5N1 cases.
False negatives have also been generated by degraded samples. In Turkey, almost all 21 cases confirmed locally were in clusters, but only 12 were confirmed in England. Similarly, only one of the cluster members in Pakistan tested positive by PCR, but four were subsequently confirmed by antibody increases.
Thus, the number of negatives due to failures to collect or preserve samples or limits in testing procedures, coupled with misdiagnosis has led to a gross undercount of official clusters.
However, recent clusters such as those in Hunan and Vietnam have clear evidence of human to human transmission, even though the samples from the index cases were not collected.
These H2H clusters are further supported by the explosion of confirmed cases in China (see updated map), which are causes of increased concern.
Media Links
No comments:
Post a Comment