Announcement

**NS1** · May 22, 2013, 07:32 PM

Re: N Engl J Med. Live-Animal Markets and Influenza A (H7N9) Virus Infection

Two Months
is
Too Long

Summary

A finding of 228S on the emergent human H7N9 Hemagglutinin, an experimental potentiator of human transmission, apparently de-motivated the authors to publish immediately. Intentionally withholding this data, from a hospitalisation on 2013-03-23 and an admitted tracheal sample taken on 2013-03-28, until now is clearly faulty thinking if, in fact, the reported genetic results are validated. The paper appears to have been written some time ago because the analyses, including the phylogenetic trees, are missing current sequences for comparative purpose.

Aged Phylogenetic Trees

Figure 1. Phylogenetic Tree of Full-Length Matrix Protein (M) Gene of Novel H7N9 Virus. (clickable from original paper) caption (emphasis ours):

The two isolated viral strains of H7N9 — Env/Nanjing and Nanjing/1 — had a remarkable degree of separation from the three other available human strains (Shanghai 1, Shanghai 2, and Anhui 1). Phylogenetic analysis showed that the M genes in the novel virus were derived from avian H9N2 viruses, among others.

The authors did create the illusion of currency by updating the first paragraph with case count and fatality count statistics prior to publication, but the phylogenetic tree caption in Figure 1 and the trees for all segments discussed in the supplement indicate a very early date of production. The comparative set of 2 early Shanghai and the single Anhui human sequences represented the full, public human set for only 5 days, from 2013-03-31 until 2013-04-04. On April 5, 2013, ChinaHangzhou1_C1_38M_2013_03_24_f was deposited at GISAID, a sequence sampled only four days before the Nanjing1 sequence that is now being profiled today for the first time (New England Journal of Medicine).

Furthermore, we would estimate that the authors of this current paper had earlier access than the public to all sequences produced by the CDC units of China. At any rate, the comparatives found on their trees (if restricted to public data) are suggestive of an analytic run prior to April 5, 2013, more than 6 weeks prior to the publication.

Viral researchers and epidemiologists are well aware of what can happen in 30 days during a pandemic emergence. Perhaps we should take caution prior to our next standing ovation for the ostensible transparency and cooperation of these organisations that withhold critical data at inflection points of human history. Who, in the public health leadership, would possibly advise a group of bench scientists to withhold such data? And who in that group of bench scientists would ever allow such information to be withheld?

The narrative indicates HA 228S, but the actual chart of polymorphisms per segment found in the supplement shows that the two sequences under discussion do NOT have the 228S, but remain at 228G wildtype? The Env/Nanjing and Nanjing/1 columns next to HA G228S are notated as 'G', the same as all other H7N9 sequences.

Narrative

From Page 1, Paragraph 4, Sentence 1 of the original paper (emphasis ours):

Two viral strains, one from the patient (Nanjing/1) and one from poultry-cage specimens obtained in the neighboring stall (Env/Nanjing), were successfully isolated (GenBank numbers, KC896771-KC896778 for the case specimen and KC896763-KC896770 for the environmental specimen).

From Page 1, Paragraph 4, Sentence 3 of the original paper (emphasis ours):

Both isolates had potentially functional amino acid sites related to mammal- or human-adapting substitution T189A, Q226L, and G228S (H3 numbering) in the receptor-binding site of hemagglutinin.

Data in Supplement

From Page 12 of the supplement document (labeled 'Page 11'), we have excerpted the lines of interest (emphasis ours, dots employed for web spacing):

Table S1. Analysis of the Key Amino Acid Positions in Different Viral Proteins Related with Interspecies Transmission or Drug Resistance.

. . . . . . . . . . . . . . . . . . . . . . . . . . . Env/Nanjing Nanjing/1 Shanghai/1 Shanghai/2 Anhui/1
HA Receptor binding site (RBS) G228S . . . . . G . . . . . . . G . . . . . . G . . . . . . G . . . . . G . . .

Is this disjunction between the narrative and the actual data yet another publication error in a peer-reviewed journal concerning this potential H7N9 pandemic emergence? Does the peer-review process, at some point, involve a reading of the paper?

The sequences are not yet available at GenBank for verification.

We acknowledge the authors, originating and submitting laboratories of the sequences from GenBank & from GISAID’s EpiFlu™ Database on which this research is based. An additional list is detailed in the linked PDF entitled "GISAID_Citations_H5N1_2011" at Is H7N9 Spreading from Human to Human in China? Post#164

GISAID Citations

GeneWurx Cross Serotype Homology Analysis, Open-Access, Full-Text version
Human Emergent H7N9 from Zhejiang Province
Environment Emergent H7N9 from Zhejiang
Fatal H5N1 Homology to Emergent H7N9 from Shanghai in March
Fatal H5N1 Karo Cluster Homology to Emergent H7N9 from First Fujian Case
Receptor Binding Site Novelty from First Taiwan H7N9 Case
China - H7N9 Human Isolates on Deposit at GISAID, Comprehensive Genetics Discussion

**NS1** · May 23, 2013, 12:56 PM

Re: N Engl J Med. Live-Animal Markets and Influenza A (H7N9) Virus Infection

The two referenced Nanjing sequences have been made available at GenBank now and do NOT show HA G228S.

**NS1** · May 24, 2013, 12:59 AM

Re: N Engl J Med. Live-Animal Markets and Influenza A (H7N9) Virus Infection

New England Journal of Medicine
Retraction of HA G228S Claim
without
Erratum

GeneWurx comments were originally written on an earlier thread and in earlier posts on this thread concerning a joint production of various CDC units of China that was web published on 2013-05-22 in the New England Journal of Medicine Correspondence entitled Live-Animal Markets and Influenza A (H7N9) Virus Infection:

While we are grateful and relieved to see the Chinese CDC (joint team) remove the claim of HA G228S from the peer-reviewed New England Journal of Medicine discussion, we are also dismayed that no erratum is notated nor any versioning indicated so that those who were mis-led by the error may be suitably relieved of their first impression concerns and may properly cascade their retractions. The small print at NEJM does, however, indicate that an update took place on May 23, but does not document the subject matter or attribution of the revision.

This letter was published on May 22, 2013, and updated on May 23, 2013, at NEJM.org

Narrative

From Page 1, Paragraph 4, Sentence 3 of the original paper (emphasis ours):

Original Publication

Both isolates had potentially functional amino acid sites related to mammal- or human-adapting substitution T189A, Q226L, and G228S (H3 numbering) in the receptor-binding site of hemagglutinin.

Current Web Image

Both isolates had potentially functional amino acid sites related to mammal- or human-adapting substitution T189A and Q226L (H3 numbering) in the receptor-binding site of hemagglutinin.

BP2 versus PB2

If interest remains toward accuracy, the authors may wish to involve their scribes in reading the next sentence after today's correction. Making another update there will add to readability of the segment name that carries the very important mammalian determinant at Polymerase Basic 2 aa627. BP2 should, perhaps be restated as the more common abbreviation, "PB2"? Secondarily, it's never too late for the editorial team at NEJM to have a reading of the full text.

From Page 1, Paragraph 4, Sentence 4 of the updated paper (emphasis ours):

The two isolates showed E rather than K at position 627 of BP2.

PB2 627E has been and remains the standard amino for non-mammal-hosted emergent H7N9, so finding that value in a poultry cage should be expected and is not remarkable. On the other hand, their finding of a human sequence without PB2 627K or PB2 701N was NOVEL and is most certainly a matter of international interest though the specimen was taken from an upper respiratory location. Broncho-alveolar lavage, the gold standard for emergent H7N9, may have produced a more representative Influenza organism with PB2 mammalian adaptations.

No explanation by any of the joint authors from the various Chinese CDC organisations has been added to the discussion concerning the extensive delay in publication of these sequences.

Announcement

N Engl J Med. Live-Animal Markets and Influenza A (H7N9) Virus Infection

N Engl J Med. Live-Animal Markets and Influenza A (H7N9) Virus Infection

Comment

Comment

Comment