New Trivalent Vaccine Targets H1N1 Tamiflu Resistance
Recombinomics Commentary 14:33
September 24, 2008

Last year, Baker explained, two of the three strains of influenza did not match well, so the vaccine only matched the strains going around by 44 percent, as opposed to typical goals of 70 percent to 90 percent.

As a result, all three of the flu strains whose components are included in this year's vaccine are new.

The above comments on the new seasonal flu trivalent vaccine are confusing, because the statements imply that one of the three targets was a match, yet it is being changed this season.  The confusion however, is due to the characterization of last years H1N1 target, Solomon Island/3 (clade 2A), as a match, when there was little Solomon Island in circulation last year.  It had been replaced by Brisbane/59 (clade 2B) and Hong Kong/2562 (clade 2C).  However, all three clade 2 sub-clades were grouped together because of cross reactivity with ferret antisera, which was linked to growth of the target viruses in eggs.  Brisbane/59 grown in eggs cross reacted with antisera against clade 2A, 2B, and 2C.  However, when the same virus was grown in mammalian MDCK cells, the titer against Brisbane (2B) was 320 and dropped to 40 for 2C and was below detection for clade 2A and clade 1.  Thus, when grown in mammalian cells the clade 2 sub-clades were antigenically distinct, as expected from phylogenetic analysis, yet all were classified as “Solomon Island-like” and considered a match.  However, since they were not a match, and Solomon Island was no longer circulating, this seasons target was changed to Brisbane/59 (clade 2B).

The H1N1 mismatch last year may have contributed to the dominance of Tamiflu resistance, which was primarily circulating on a Brisbane/59 sub-clade.  This sub-clade became dominant in northern Europe, leading to Tamiflu resistance above 60% in Norway and high levels in other European countries including France and western Russia.  Recently released phylogenetic trees of 2008 isolates from the northern and southern hemisphere indicate that the sub-clade is still dominant and the level has risen to 100% in South Africa and Australia.

The circulation of this sub-clade, coupled with rapid change near the receptor binding domain in isolates in South Africa and Seychelles, raises concerns that the northern hemisphere vaccine target of Brisbane/59 will have limited activity because of these changes which appeared in the summer of 2008.

The target for the 2009 southern hemisphere have been announced, and the H1N1 target remains Brisbane/59, raising concerns that the reduced effectiveness of the vaccine for this seasons northern hemisphere will be extended to next seasons southern hemisphere, keeping H1N1 evolution one step ahead of the H1N1 vaccine.  These concerns are compounded by current statements that clade 2B and clade 2C are antigenically indistinguishable, which is not the case when targets are grown on mammalian cells.  Thus, the reliance on antigenic cross reactivity, in spite of clear phylogenetic differences as well as differences in isoaltes grown on mammalian cells, increases concerns that vaccine targets will continue to trail influenza evolution.

Media Links

Recombinomics Presentations

Recombinomics Publications

Recombinomics Paper at Nature Precedings