Paradigm Shift Intervention Monitoring
twitter
Commentary
CDC trH3N2 Testing Bias Raises Pandemic Concerns
Recombinomics Commentary 12:50
September 7, 2011
The above comments from ProMED cite an incorrect suggestion for the trH3N2 origin in the Washington County Fair cluster. The detection of three trH3N2 confirmed cases is driven by testing, not transmission from animals. There are many more people at fairs than swine (or animals), but fair attendees are far more likely to be tested for trH3N2, especially when presenting with flu-like symptoms in August, when seasonal H3N2 frequencies are low. Without the specific testing for trH3N2, these cases would test as seasonal H3N2 because the H3 and N2 in trH3N2 is human and recognized by serological reagents targeting human H3. This heavy bias is also reflected in the latest request for samples by the CDC, which is targeting patients with flu-like symptoms and a swine link.
Media reports and the CDC frequently cite the 21 cases of swine origin cases reported between 2005 and 2010, but this number is also driven by testing. 13 of the 21 cases are H1 and all were identified prior to the 2009 H1N1 pandemic.
Identification of these cases was facilitated by the swine H1, which would be unsubtypable by the human H1 reagents. Thus, these patients would be influenza A positive, but unsubtypable by reagents that target human H1 or H3. Further analysis would lead to the trH1N1 (12 cases) or trH1N2 ( 1 case) diagnosis. After the emergence of pandemic H1N1, which also has a swine H1, there have been no reported cases of trH1N1 or trH1N2 other than pandemic H1N1 and one cluster of 3 farm workers in Saskatchewan in the spring of 2009, which had human H1 and N1 from the Brisbane/59 sub-clade, which had H274Y).
However, after the pandemic H1N1 emerged, trH3N2 was detected in the United States, and almost all cases had acquired PB1 E618D, via recombination from pandemic H1N1. These isolates were closely related and clustered in human trH3N2 infections, signaling human transmission that produced cases detected at a low frequency. The most dramatic example was the cluster in Minnesota, involving a father. A/Minnesota/11/2010, and lab confirmation of trH3N2 in his daughter who did not have an exposure to swine and was assumed to have been infected by her father, or other family members who also had flu-like symptoms (serological testing on the other family members was inconclusive, but the daughter was serologically confirmed).
The transmission to humans was also enhanced by another acquisition from pandemic H1N1, the M gene segment, which was acquired via reassortment and critical for aerosal transmission. This acquisition was present in the isolate, A/Indiana/08/2011, from the Indiana case (1M). The isolate also had a PB1 from swine which was similar to the PB1 present in the large cluster from the Huron County fair in 2007, including A/Ohio/01/2007 and A/Ohio/02/2007 isolated from a presenter (10F) and her father (36M). In addition to these two confirmed cases, there were two dozen attendees who had flu-like symptoms, which was rare in August in Ohio.
Thus, the recent Indiana case has two acquired gene segments, M from pandemic H1N1 and PB1 from an isolate with PB1 closely related to the Ohio cases, which did not have E618D.
The constellation of genes in the sequences from the Pennsylvania cluster is unclear, since these sequences have not been released by the CDC, even though sequencing of the first Pennsylvania case was completed last month, and sequecing on the two more recent cases has identified the pandemic H1N1 origin of the M gene segment. The MMWR on Friday, and Pennsylvania Department of Health press release on Monday indicated all three cases had the M gene segment from H1N1, signaling more efficient transmission. However, the CDC also cited differences between the initial Pennsylvania case from Schuylkill Country (2F) and the Indiana case, which may reflect the difference in PB1. Thus, the Pennsylvania case(s) may have the PB1 with E618D, as well as the M gene segment from pandemic H1N1.
However, the H3 sequence may have also changed. The H3 sequence from Indiana is similar to the H3 sequence from earlier trH3N2 cases (most closely related to A/Minnesota/11/2010, as well as A/Wisconsin/12/2010 and A/Pennsylvania/40/2010) and typable because the H3 is from a human H3N2 in swine in the early 1990's. However, the CDC recently added an unsubtypbale to week 33 in its table of serotyping results, and the sample from the Pennsylvania case was collected in week 33. It was identified as influenza A by the rapid test, but identified as trH3N2 by a PCR test, suggesting the H3 was unsubtypable, in contrast to the earlier trH3N2 H3 sequences, including the Indiana isolate.
If the Pennsylvania H3 is unsubtypable, detection will rapidly increase because these cases will be influenza A positive but H3 negative (because the H3 has evolved away from the specificities on the human H3 serotyping reagents) and not pandemic H1, because it is the more evolved H3.
Thus, this evolution will lead to the detection of trH3N2 cases not linked to fairs or swine, because testing will produce a signature that heralds in a new pandemic (influenza A positive, but H3 and pandemic H1 negative).
Recombinomics
Presentations
Recombinomics
Publications
Recombinomics
Paper
at Nature Precedings