Commentary
Vaccine Mismatch Drives Spread of H1N1 Tamiflu Resistance
Recombinomics Commentary 19:51
August 25, 2008
Although A/Solomon Islands/3/2006-like viruses and A/Brisbane/59/2007-like A/H1N1 viruses were represented only by minor clades during the 2006–2007 season (H and F, respectively), Solomon Islands-like viruses achieved global A/H1N1 dominance by the start of the 2007–2008 season, and the reassortant clade of Brisbane-like viruses rose to dominance later during the 2007–2008 season [39].
The above comments from today’s PLOS paper on H1N1 and H3N2 circulation in the United States during the 2006/2007 season is based on the MMWR characterization of H1N1 in the 2007/2008 season, which is misleading and adds to the confusion regarding the emergence of oseltamivir resistance in H1N1.
The emergence is readily followed through phylogenetic analysis which distinguishes four different sub-clades, New Caledonia/20 (clade 1), Solomon Island/3 (clade 2A), Brisbane/59 (clade 2B), and Hong Kong/2652 (clade 2C). In the United States in 2006/2007, clade 1 was dominant, and there were small numbers of clade 2A and clade 2B in circulation. However, clade 1 and clade 2A were replaced by clade 2B and clade 2C in 2007/2008, but these isolates were initially called Solomon Island–like because of cross reactivity of ferret anti-sera with isolates grown in chicken eggs. The differences between the titers from Solomon Island/3 and Brisbane/59 was only a factor of two, so all isolate were initially classified as Solomon Island/3–like and a match for the new vaccine, which used Solomon Island/3 as the target.
However, when Brisbane/59 grown in mammalian MDCK cells was used as a target, the dramatic differences in antigenicity were clear. The titer for Brisbane/59 (clade 2B) was 320, but the level dropped to 40 for the anti-sera against Hong Kong/2652 (clade 2C), and was below the limits of detection (40) for the anti-sera against Solomon Island/3 (clade 2A) or New Caledonia/20 (clade 1).
Thus, the differences between the sub-clades were significantly different by both phylogenetic analysis and serology, but the poor serological data of egg grown tragets led to an initial grouping of all three clade 2 sub-clades into the Solomon Island/3–like group. However, though there was no reported Solomon Island/3–like H1N1 in circulation in the 2007/2008 season. The Solomon Island/3 vaccine target was a mismatch, but the confusion generated by the earlier MMWR still persists in peer reviewed publications as well as this month’s European surveillance report.
In the upcoming flu season in the northern hemisphere, the Solomon Island/3 target is being replaced by Brisbane/59, but recent sequence data from South Africa raises concerns that the Brisbane/59 vaccine will have limited utility in the fall, because the dominant H1N1 strain in South Africa is oseltamivir resistance and already has acquired a cluster of non-synonymous changes flanking the receptor bind domain at position 190 (H3 numbering). Initially, South Africa reported that the first 23 isolates that were sequenced all had H274Y. More recently, the 100% resistance was extended to the first 107 isolates. Similarly, the first 10 H1N1 isolates from Australia this season have H274Y.
Although the isolates in South Africa can be divide into two groups, all isolates link back to the clade 2B linear in circulation in the northern hemisphere. Thus, the emergence and spread of H274Y can be easily followed via phylogenetic analysis, which applies to the three gene segments which had been recently sequenced, HA, NA, and MP, although there is more genetic diversity in the HA and NA genes and there fore easier to see independent introductions.
H274Y was first reported in clade 2C in China in the 2005/2006 season, which was followed by acquisition of H274Y by clade 1 in the 2006/2007 season in United States. In October, 2007, the first report of H294Y in the clade 2B was reported in two isolates in Hawaii. All of the above isolates had identical sequences downstream from H274Y. The polymorphisms was subsequently acquired in a related clade 2B sequence which then became dominant and was a precursor for the dominant sequence in South Africa.
Thus, the H1N1 mismatch last year may have contributed to the dominance of the clade 2B sequence, which has subsequently led to increases in oseltamivir resistance to 100% of H1N1 isolates in South Africa and Australia.
Media Links
Recombinomics Presentations
Recombinomics Publications
Recombinomics Paper at Nature Precedings