Paradigm Shift Intervention Monitoring
twitter
Live feed of underlying pandemic map data here
Commentary
Duke D225G/N Death
Data Destroys Random Mutation Paradigm
Recombinomics
Commentary 11:37
February 15, 2010
The above comments from the WHO preliminary report on D225G (D222G in H1 numbering) represent the swan song for random mutations as a key driver of influenza antigenic drift. Recently released sequences by the CDC at GISAID from the death cluster at Duke did not have D225G or D225N in initial samples from four H1N1 positive patients, but viruses lab isolated from samples from these four patients were positive for D225G or D225N in 3 of 5 sequences, including D225G in isolates from two sequential samples from the same patient. These sequences all had the same rare genetic HA marker Y233H.
The detection of D225G and D225N in this fatal cluster was not unexpected. The four patients were on the same ward and infected with the same virus. Three of the four died, suggesting a viral genetic change was responsible. D225G and D225N have been frequently identified in direct sequencing of clinical samples, indicating such detections were not due to growth of the virus in the lab. Moreover, the isolates were clustered in time and space, supporting transmission of the lethal strain. However, the Duke cluster represented clear lethal transmission of the same H1N1 among contacts.
The detection of D225G and D225N in two of the four cases was difficult to explain by random mutation. The four patients were infected with H1N1 that had the rare marker Y233H in all HA sequences and H274Y in all NA sequences from the cluster, leaving little doubt that the virus was transmitting to all four patients. This sub-clade was distinct from the H1N1 in Ukraine and Russia, which also had a high frequency of the same two changes in fatal cases, including 11 cases in Ukraine that had both D225G and D225N. The presence of both of these changes in the same patient or same cluster is difficult to explain by random mutation, because these clusters also include sequences with wild type receptor binding domain sequences, which would therefore require multiple new introductions. Thus, for Duke one patient would require a copy error leading to D225N, while another would require a D225G error, which would have to happen twice in samples from the same patient if the detection was due to an in vitro effect.
Although this ad hoc tortured explanation is extremely unlikely, it is cited in the original WHO report, as quoted above, and is repeated in a subsequent WER distribution. This logic is clearly an act of desperation to shore up the WHO working hypothesis that D225G is generated by repeated copy errors that are spontaneous and sporadic, which is not supported by the actual data generated in multiple countries by multiple labs sequencing original clinical samples which show that D225G and D225N is clustered in space and time and found in phylogenetically related sequences, which would require multiple independent introductions of both markers, which is not credible.
However, the WHO pronouncements can have a significant effect on the world’s health. This week WHO is selecting the H1N1 target for the 2010/2011 vaccine for the northern hemisphere, and the tortured logic used in the WHO paper on D225G may be used to recommend the current vaccine target which lacks D225G in the killed vaccine.
Mill Hill has characterized a Ukraine isolate with D225G as a “low reactor” which indicates that target has a titer that is at least 4 fold lower than the vaccine target used to create the reference anti-sera. There is also evidence for an increase of D225G/N in fatal cases, raising concerns that D225G will be more dominant in the next wave, and a vaccine target lacking D225G will have reduced efficacy against H1N1 circulating next season in the northern hemisphere.
An independent analysis of WHO’s ability to interpret sequence data from pandemic H1N1,as well as a reliance on random mutations to explain influenza evolution is long overdue.
Media Links
Recombinomics
Presentations
Recombinomics
Publications
Recombinomics
Paper
at Nature Precedings