Paradigm Shift Intervention Monitoring
twitter
Commentary
CDC Obfuscation
of H3N2v Sub-clade In 2012 Human Cases
Recombinomics
Commentary 13:45
August 16, 2012
* Case in Utah occurred in April 2012 and is not part of the ongoing outbreaks.
The above notation on the CDC webpage entitled “Information on Influenza A (H3N2) Variant Viruses (“H3N2v”)” encapsulates the CDC position on H3N2v evolution, which maintains that H3N2v is infecting humans via jumps from swine, and the multiple genetic changes associated with H3N2v adaptation, other than the acquisition of the H1N1pdm09 M gene are of little or unknown consequence.
The above page represents the nexus of the CDC obfuscation on the novel sub-clade responsible for the explosion of H3N2v. It creates significant confusion for physicians and the general public on these increases. The page has a table with a column for 2011 cases (which begins with the first human case with an H1N1pdm09 M gene) as well as the 2012 cases, as listed by state. However, as noted above, the first case in 2012 (A/Utah/10/2012) has been excluded.
However, sequence analysis of the public data clearly demonstrates an lineage specific change in the NA gene, beginning with the last 2 cases in 2011 (A/West Virginia/06/2011 and A/West Virginia/07/2011) and extending to all public 2012 H3N2 human sequences, which includes the excluded Utah case.
The current table list 12 cases for 2011 and 154 cases for 2012, although recent reports from Ohio and Indiana raise the total to 195 for 2012 based on Indiana cases through Monday, and Ohio cases through Tuesday. Moreover, two additional states have reported cases (1 in Michigan and three in West Virginia). The CDC will be updated each Friday, and the totals this Friday will clearly show more than 200 confirmed cases in 2012.
The reasons for these dramatic increases are seen in key changes in the evolving H3N2v sequences, which appear in human cases prior to detection in swine cases, clearly supporting human to human transmission.
Although the table begins with 2011 H3N2v cases, the first case in the United States was identified in the summer of 2009 in a Kansas case (A/Kansas/13/2009), which was followed by a second 2009 summer case (A/Iowa/16/2009), which was followed by a Minnesota spring case in 2010, A/Minnesota/09/2010. However, interest in the US H3N2v cases increased significantly in the fall of 2010 when WHO issued a pager alert when two cases were identified in residents of Illinois and Pennsylvania. The Illinois case had attended an agricultural fair in Wisconsin, A/Wisconsin/12/2010, while the Pennsylvania case denied swine exposure, but lived in rural Pennsylvania, A/Pennsylvania/14/2010. The CDC noted that the two sets of sequences were similar to the earlier US H3N2v cases, but the degree of similarity could not be independently confirmed because only sequences from the Kansas case had been made public (at GISAID).
However, an analysis of the Kansas sequences indicated that the internal genes were related to swine H1N1 sequences from the 2007 Huron County Fair. Prior to the 2009 pandemic the CDC had identified 13 human cases with swine triple reassortant sequences (12 H1N1 and one H1N2), but only two of the confirmed cases were from the same outbreak, and the two cases, A/Ohio/01/2007 and A/Ohio/02/2007 were from an exhibitor and her father at the 2007 Huron County fair. Moreover, 26 attendees had ILI, which is unusual for Ohio in August (the 2012 edition of the Huron County Fair began August 13).
In addition to the close relationship to the sequences from the Ohio fair, the PB1 sequence from the Kansas H3N2v case had a modification (E618D) which was not present in swine sequences, but was present in all H1N1pdm09 sequences, signaling human adaptation.
Shortly after the WHO pager alert, the CDC released the sequences from the 13 human cases identified prior to the 2009 H1N1pdm09 pandemic, as well as the H3N2v sequences from cases identified after the pandemic. These sequences indicated that the PB1 in all of the 2010 H3N2v human cases had E618D, providing more evidence for human adaptation.
Moreover, most of the genes of most of the human cases matched each other as well as the internal genes from human and swine isolates from the 2007 Huron County Fair. However, E618D was not found in swine samples collected prior to the fall of 2010 (other than swine infected with H1N1pdm09).
In addition to the sequences released in 2010, developments in early 2011 provided additional evidence for H3N2v human adaptation. A second Pennsylvania case, A/Pennsylvania/40/2010, had been initially serotyped as a season H3N2 case. Subsequent analysis showed that the isolate was an H3N2v case with a constellation which matched the earlier sequences from A/Wisconsin/12/2010, and diseason onset dates were less than a week apart. This match in two cases that were not epidemiologically linked, raised concerns that this constellation signaled more human adaptation.
The release of sequences from the second Pennsylvania case was followed by a confirmation of the sixth H3n2v cases for 2010. However, the lab confirmation was relied on serology on the daughter of the second Minnesota case, A/Minnesota/11/2010, who was symptomatic. Other contacts were also symptomatic, but serological data was characterized as “inconclusive”. This cluster support human to human transmission, because the daughter had no swine contact.
In 2011 the initial human cases defined a novel sub-clade formed by reassortment between sequences seen in the two matching human 2010 cases (WI/12 and PA/40) and H1N2 sequences circulating widely in swine. This novel sub-clade had 5 genes from the dominant sequences in human cases in 2010 (PB2, PA, HA, NP, NS), and three genes from the swine H1N2 (PB1, NA, MP). The PB1 was closely related to the 2007 H1N1 sequences, which lacked E618D, but the MP gene was from H1N1pdm09. The N2 was a lineage that was widespread in swine H1N2, but traced back to seasonal H3N2 which was circulating in 2003.
The first case was from an Indiana child who had no swine contact, but his caretaker had exhibited swine. However, she was asymptomatic as were her swine. This July isolate, A/Indiana/08/2011, was followed by three cases from the Washington County Fair in August, A/Pennsylvania/09/2011, A/Pennsylvania/10/2011, A/Pennsylvania/11/2011), which had the same constellation. These cases had swine exposure or contact at the fair, but no symptomatic swine were reported. Subsequent cases were identified in Maine (A/Maine/06/2011 and A/Maine/07/2011), Indiana (A/Indiana/10/2011) and Iowa (A/Iowa/07/2011, A/Iowa/08/2011, A/Iowa/09/2011). The Iowa cluster had no swine exposure and were epidemiologically linked. The father and brother of the index case were symptomatic, and two classmates were H3N2v confirmed. Moreover, the three sets of sequences from the confirmed cases were virtually identical, signaling human to human transmission. This cluster created considerable concern because it involved 5 people and no swine.
However, concerns were significantly increased by the cluster at a day care center in Mineral County, West Virginia. This cluster also had no swine contact, but the two confirmed cases (A/West Virginia/06/2011 and A/West Virginia/07/2011 developed symptoms two weeks apart, indocating the second case was not infected by the index case. Initial results from an epidemiological study identified ILLI in 23 of the 70 contacts of the index case.
Moreover, the West Virginia sequences defined a novel sub-clade which replaced the N2 gene (which had a swine H1N2 lineage), with an N2 gene with a swine H3N2 pedigree. This swap was easily identified and clearly distinguished the two sub-clades.
This cluster led to a CDC conference call involving all 50 state labs, as well as a series of CDC updates calling for enhanced surveillance. However, this enhanced surveillance failed to identify any human cases in January or February, which represent the peak of the 2011/2012 flu season.
The first 2012 H3N2v case was identified at the end of March in Utah, A/Utah/10/2012, which matched the novel sub-clade identified at the West Virginia sub-clade (but excluded from the CDC count of 2012).
The CDC count for 2012 begins with a series of cases in July from Hawaii, Indiana, and Ohio. Sequences from the first four outbreaks were made public (at GISAID), and all were closely related to each other and matched the sub-clade first identified in West Virginia. The Hawaii sequence (A/Hawaii/03/2012) was from a farmer on Maui, which was collected Juyl 12. Four cases (collected July 16) are from the LaPorte County Fair in northwest Ohio (A/Indiana/06/2012, A/Indiana/07/2012, A/Indiana/08/2012, A/Indiana/09/2012). A fifth case (A/Indiana/12/2012) is from the Jackson County Fair in south central Indiana (collected July 27), while the Ohio case is from the Butler County Fair (A/Ohio/13/2012) in southwestern Ohio (also collected July 27). All of these july collections are closely related to each other, even though one case is from Hawaii.
The match of the seven July isolates with the March isolate from Utah and the West Virginia cluster from collections in November and December, 2011 signals further human adaptation and transmission as seen in the large size of the West Virginia cluster which was not linked to swine exposure, as well as the size of the clusters in Indiana and Ohio, which involve record numbers of confirmed cases (5 for LaPorte, 4 for Jackson, and 17 for Butler) as well as an explosion of Indiana and Ohio clusters which are larger. The Gallia cluster in southeastern Ohio involved more than 200 symptomatic attendees of the Gallia Junior County Fair, and now West Virginia has confirmed 3 cases at the Mason County fair, five miles east of the Gallia Fair, supporting human transmission.
Moreover, the human evolution and transmission is supported by the swine data, which creates serious discordance with the human data. A recent paper involving 387 HA and NA sequences (and 674 MP sequences) from swine isolates from 2009-2011 only found one match (based on HA, NA, and MP sequences) with the 2011 human sub-clade. More recent sequences generated under the USDA expanded surveillance identifies 26 isolates which matched the 2011 sub-clade (seen in the first 10 human cases in 2011) while only 2 matched the 2012 human cases (as well as the 2011 West Virginia cluster). Thus even though the earlier sub-clade is widespread in swine, including 2012 isolates from Indiana and Ohio, none of the 2012 human cases match this earlier sub-clade. Moreover, none of the swine matches were from samples collected prior to the July 2011 case in Indiana, and both swine matches for the 2012 human cases were collected in 2012, after the 2011 West Virginia cluster.
These data demonstrates clear differences between the two sub-clades identified in human cases in 2011 and 2012, and the strong discordance between human and swine isolates signals human adaptation and transmission for the more recent sub-clade, which has been found in all 2012 public human H3N2v sequences.
The CDC H3N2v landing page which removes the first 2012 case from the explosion of cases in July and August is significant and required for the CDC’s obfuscation on the two distinct sub-clades and the involvement of the 2011 West Virginia novel sub-clade in the human to human spread in H3N2v.
Recombinomics
Presentations
Recombinomics
Publications
Recombinomics
Paper
at Nature Precedings