Paradigm Shift Intervention Monitoring
twitter
Commentary
trH3N2 False
Negatives Raise IHR Compliance Concerns
Recombinomics Commentary 11:30
September 29, 2011
However, the removal of the two unsubtypables suggests the current assays the CDC use in routine analysis / reporting fails to identify trH3N2 cases because the week 30 update has no H3N2 cases (and A/Indiana/08/2011 was identified in samples collected during week 30) and the updated table in week 37 has removed the unsubtypable from week 33, when the sample for A/Pennsylvania/09/2011 was collected, and the number of H3N2 cases remained unchanged at five.
Thus, it appears that the 2011 trH3N2 cases are not reclassified as seasonal H3N2 cases, but instead are simply not reported, suggesting the samples fail to meet the CDC requirement for classification as influenza A positive, so instead of being reported as “unsubtypable”, these samples are being falsely reported as “negatives”, even though trH3N2 sequences have been identified in the samples from the Pennsylvania cases, and trH3N2 virus has been isolated from the Indiana case (and the sequence of the trH3N2 isolate matches the direct sequencing of the Indiana and Pennsylvania samples).
Thus, the CDC has reported the recent cases in its MMWR and has noted the cases in text in the CDC FluView, yet is reporting these samples as “negatives” in its updated subtyping table and graph, raising concerns that the vast majority of trH3N2 are being reported as negatives.
These concerns were raised last year when Pennsylvania was reporting unsubtypables in its weekly reports, but these unsubtypables failed to appear in the CDC weekly reports. Pennsylvania leads the nation in the number of trH3N2 isolates from infections in 2010 with two (A/Pennsylvania/14/2010 and A/Pennsylvania/40/2010) and also leads the nation in sequences from 2011 (A/Pennsylvania/09/2011, A/Pennsylvania/10/2011, and A/Pennsylvania/11/2010).
One of the 2010 cases, A/Pennsylvania/40/2010, was not reported until 2011 because it was initially reported as seasonal H3N2 and routine testing led to its identification as trH3N2. However, the reporting was delayed because the more advanced testing of antigen characterization and sequencing required viral isolation, which was delayed due to technical issues.
The H3 sequences from the four 2011 trH3N2 cases are closely related to three of the 2010 isolates (A/Pennsylvania/40/2010, A/Wisconsin/12/2010, A/Minnesota/11/2010), but the additional evolution in 2011 has moved these isolates into the unsubtypable category for H3 and apparent also creates influenza A detection issues leading to false negative representations in the CDC FluView.
These detection failures raise concerns that the number of trH3N2 is far greater than the 6 cases reported in 2011. One of those cases was the daughter of the index case in the Minnesota cluster. She was trH3N2 lab confirmed by antibody testing and other family members were “inconclusive” in such test, but had flu symptoms similar tow the two confirmed cases. As noted above, the H3 sequence from the index case, A/Minnesota/11/2010, was virtually identical to the other two 2010 sequences and was a precursor to the four 2011 cases signaling human to human transmission for these isolates.
Moreover, the 2011 sequences include the M gene from pandemic H1N1, which was deemed critical for human to human transmission, and the 2011 Indian case has no history of swine contact. His caretaker has a history of swine contact, but no symptoms or virus has been reported for the caretaker or associated swine, which is also true for the swine at the Washington County fair.
Thus, the failure to identify a source for any of the 2011 cases, coupled with the present of pandemic H1N1 M gene segment, strongly suggests that the number of cases in Indiana and Pennsylvania was far higher than the four reported cases.
This concern is increased by the CDC request for samples from Pennsylvania patients with swine contact, when the CDC assays fail to identify sources for the four known cases in spite of an “intense” investigation. In the week 37 FluView the were only 3 influenza A positive samples for the entire country in spite of increases in ILI reports or Google FluTrend values in multiple cities in recent weeks.
Thus, the CDC “adjustment” in the cut-off for influenza A positive is leading to a large number of false negatives and failure to detect or report trH3N2 cases, as mandated under IHR regulations.
An explanation for the elimination of unsubtypables from FluView reports, as well as an analysis of 2010 unsubtypables in Pennsylvania, is long overdue.
Recombinomics
Presentations
Recombinomics
Publications
Recombinomics
Paper
at Nature Precedings