Emergence and Spread of Worldwide H1N1 Tamiflu Resistance
Recombinomics Commentary 18:31
September 30, 2008

The emergence and expansion of oseltamivir (Tamiflu) resistance in seasonal H1N1 influenza has baffled the experts, because the vast majority of cases are in patients who have not been recently treated with Tamiflu.  Moreover, recent reports of H1N1 in the flu season of the southern hemisphere indicate that the polymorphism, H274Y, is becoming fixed and many countries are reporting frequencies at or near 100%. Phylogenetic analysis of recent isolates indicates that most of the current activity is linked to an expansion of a subset of clade 2B (Bribane/59), which clearly does not have a fitness penalty linked to the acquisition of H274Y.

Earlier data had suggested that osletamivir resistance was linked to a fitness penalty, which would limit the spread of resistant viruses.  However, most of these studies focused on changes in N2 in H3N2, and the current resistance has been limited to H274Y in N1 in H1N1.

In 2005 there were reports of H274Y developing in H5N1 patients undergoing oseltamivir treatment, supporting the earlier results with H3N2 and fitness penalties.  However, in late 2005 H5N1 with H274Y was detected in isolates from wild birds, indicating fitness penalties were minimal, since the H5N1 was spreading in hosts not receiving oseltamivir.

These observations were then seen in H1N1 seasonal flu in the 2005/2006 season in China.  The clade 2C (Hong Kong/2562) isolates were found in patients not receiving oseltamivir, indicating H274Y did not produce a fitness penalty in seasonal flu.  However, clade 2C isolates with H274Y were not widely reported.

In the following season, clade 1 (New Caledonia/20) isolates were reported in cases in the United States.  Once again the number of reported cases was limited, and once again the patients had not been receiving oseltamivir.  Although the clade 1 and clade 2C genetic backgrounds differed, sequence identity adjacent to H274Y provided a mechanism for acquisition by homologous recombination.

In 2007, H274Y was reported in clade 2B (Brisbane/59) in isolates from Hawaii.  One again the acquisition was on another genetic background, but regions of identity facilitated the movement of H274Y in clade 1 to clade 2B, and once again the spread of this version of clade 2B was limited.  However, H274Y was subsequently reported on a clade 2B sub-clade, which was dominant in the United States and many other countries, including Norway.  The frequency of H274Y in Norway exceeded 65%, leading to widespread testing for H274Y.  This version was widespread in Europe, and many countries, including France, had levels approaching 50%.

Recent reports indicate that the level of H274Y is now approaching 100% in many countries, and recent phylogenetic analysis indicates these isolates are evolved versions of the dominant sub-clade in Europe and the United States.

This dominance and spread may have been facilitated by a mismatched H1N1 vaccine that was introduced in the 2007/2008 season.  The target for that season in both the northern and southern hemisphere was clade 2A (Solomon Island/3).  Although the clade 2B isolates were readily distinguished from clade 2A in phylogenetic analysis, initial serological produced cross-reactivitive between all three clade 2 sub-clades (2A, 2B, 2C) and all were classified as "Solomon Island –like", which could be distinguished from clade 1 (New Caledonia).  However, much of this cross reactivity was dependent on isolates grown in chicken eggs.  When clade 2B (Brisbane/59) was grown in mammalian cells, the level of cross reactivity was below detection levels for clade 1 and clade 2A, and at the lower limit of detection of a titer of 40 for clade 2C.

Thus, in the 2007/2008 season, the Solomon Island vaccine was a mismatch, because in 2007/2008 clade 2A was not in circulation because it had been replaced by clade 2B and 2C.  This season the H1N1 target has been changed to clade 2B (Brisbane), but many of the recent isolates had begun to evolve away from the Brisbane sequence, raising concerns of limit utility for the vaccine.

This concern has been heightened by the increase in frequency of H274Y.  In South Africa, 283 H1N1 isolates have been collected and the first 129 tested have had H274Y.  Similarly, 25 of 26 isolates in Australia have H274Y, as do 10 of 11 in the Philippines.  Other countries with lower numbers of samples tested that have similar frequencies, are Cameroon (2/2), El Salvador (7/7), United States (2/3), Uruguay (6/7), United Kingdom (1/1), New Caledonia (3/3), and New Zealand (1/1).

In the recently release NA phylogram which included 11 isolates (see list below) collected since the 2nd quarter of 2008, all 11 mapped to the dominant sub-clade that was in circulation in Europe and the United States.  These recent isolates included some from the northern hemisphere (France and Russia) raising concerns that variations of this sub-clade will emerge in the upcoming flu season in the northern hemisphere.

Moreover, the fixing of H274Y in N1 in H1N1 raises concerns that a high level in the N1 genetic reservoir will increase the likelihood of H274Y appearing in H5N1, if H5N1 is readily transmissible in humans.

Recent isolates with H274Y

Cameroon/073/2008
Cameroon/080/2008
Chubut/102197/2008
Ghana/100/2008
Johannesburg/35/2008
Johannesburg/47/2008
Kursk/1/2008
Lyon/1059/2008
Seychelles/2010/2008
St Petersburg/75/2008
St Petersburg/85/2008

Media Links

Recombinomics Presentations

Recombinomics Publications

Recombinomics Paper at Nature Precedings