Paradigm Shift Intervention Monitoring
twitter
Live feed of underlying pandemic map data here
Commentary
Emergence and
Evolution of Pandemic Tamiflu
Resistance
Recombinomics
Commentary 12:33
December 22, 2009
In seasonal H1N1 the role of H274Y was largely that of a hitchhiker. It moved from one H1N1 genetic background to another via recombination, and eventually paired up with the receptor binding domain change, A193T, which led to increases in H274Y in the summer of 2008 in the southern hemisphere and dominance in the fall in the northern hemisphere. In 2008/2009 virtually all Brisbane H1N1 had A193T and H274Y, but there were multiple variation with changes flanking A193T at positions 187, 189, and 196. These acquisitions, as well as early acquisitions, were drive by recombination and most newly acquired polymorphism were acquired from the co-circulating clade 2C (Hong Kong/2652/2006).
The evolution of H274Y began earlier and was characterized by its presence in patients who had not received Tamiflu, demonstrating that the H1N1 with H274Y was evolutionarily fit. The first examples were in the summer of 2006 in China and involved clade 2C. In the 2006/2007, H274Y jumped to clade 1 (New Caledonia/20/1999), but was relative rare in the US and UK, although most positives were in patients not taking Tamiflu. In 2007/2008 the frequency began to rise in clade 2B, which had displaced clade 1 and clade 2A (Solomon Islands/3/2006). The frequency varied, but the highest frequencies were in countries which had one particular clade 2B sub-clade, which was present in 70% of the H1N1 in Norway. The discovery of H274Y in 70% of H1N1 samples in Norway led to widespread screening. H274Y was found in multiple clade 2B sub-clades, but one sub-clade dominated and it acquired additional changes from clade 2C via recombination, including A193T.
This sub-clade was established in South Africa and Australia in the summer of 2008, which set the stage for emergence in the fall of 2008, when H274Y levels in H1N1 approached 100%. This high frequency in seasonal H1N1 led to concerns that H274Y would then jump to pandemic 2009, which emerged in the spring of 2009. Some expected the acquisition to be via reassortment, where pandemic H1N1 would acquire the Brisbane/59 N1. However, there have been no reports of acquisition of a human N1 or any other human flu gene. The pandemic H1N1 has maintained its original constellation of gene segments, with one human (PB1), two avian (PA and PB2) and 5 swine. However, H274Y has jumped from seasonal to pandemic H1N1 via recombination and is currently jumping from one pandemic genetic background to another.
Although H274Y is following the same evolutionary path in pandemic H1N1, the close monitoring of patients, especially those receiving Tamiflu, has led to early detection in patients receiving prophylactic Tamiflu because of linkage to an H1N1 confirmed contact, or therapeutic Tamiflu because of an immune-compromised state. The initial cases were in contracts being treated prophylactically, which led to speculation by Roche that the resistance was "spontaneous" and similar to previously described resistance in children in Japan. However, those earlier examples were linked to sub-optimal dosing and largely limited to H3N2. The resistant H3N2 had changes at multiple positions and did not transmit. Consequently, the switch to optimal doing eliminated the resistance, and there have been no recent examples of H3N2 Tamiflu resistance in Japan, or anywhere else In the world. In contrast, Tamiflu resistance in H1N1 was limited to a single nucleotide which produced H274Y. That same nucleotide change was in H274Y in H5N1 as well as all seasonal H1N1 clades (clade 1, 2B, and 2C) as well as pandemic H1N1. Moreover, the pandemic H1N1 patients receiving prophylactic Tamiflu developed symptoms on day 5 which was close to the normal incubation period of 2-4 days, indicating the H274Y was from a co-circulating H1N1 with H274Y and not due to a spontaneous or de novo mutation. The rapid appearance of H274Y was also noted in patients receiving Tamiflu treatments. In the first case in Singapore, the initial sample was wild type, but two days later H274Y was dominant.
However, the H274Y example that was most analogous to seasonal H1N1 was in patients who had not received Tamiflu. The first example of an evolutionarily fit H274Y was found in a San Francisco traveler to Hong Kong. The June isolate, A/Hong Kong/2369/2009, was a pandemic H1N1 that had also acquired a receptor binding domain change, D225E. Although the initial attempts to find the subclade in northern California failed, the recently released sequences from Japan have an August isolate, A/SHIGA/43/2009, has H274Y and D225E as well as additional HA and NA markers linking the two isolates to the same sub-clade. Moreover, a September isolate from Tennessee, A/ Tennessee/17/2009, also has H274Y with the same series of markers, indicating the sub-clade is widely transmitted over an extended time frame.
In addition to the above sub-clade, additional clustering of H274Y has been noted. One sub-clade includes the isolates from immuno-suppressed patients in Seattle, Washington, as well as the isolate from Illinois, A/Illinois/10/2009, which has also acquired another receptor binding domain change, D225G. This combination of H274Y with D225G has also been described for a fatal case in France, but the sequences have not been made public.
However, the multiple examples of clustering of H274Y in a series of pandemic H1N1 sub-clades demonstrates that H274Y in pandemic H1N1 is following the same path as H274Y in seasonal flu. It has appeared on multiple pandemic H1N1 sub-clades and has also paired up wuth two receptor binding domain changes, D225E and D225G.
Moreover, recently released data from Vietnam has described evolutionarily fit H1N1 in Vietnam in July, leading to infection of 7 healthy students on a train. Similarly, outbreaks of Tamiflu resistant H1N1 have been reported in hospitals in Wales and North Carolina. In addition, the reported number of H274Y cases has exploded in the United States and the Netherlands in recent weeks, and these new isolates will undoubtedly grow the sub-clades with clusters of H274Y.
Thus, the evolution of H274Y in pandemic H1N1 is following the same path as H1N1 in seasonal flu, which involves H274Y jumping to multiple sub-clades via recombination, leading to an emerging sub-clade with H274Y and receptor binding domain changes.
Media Links
Recombinomics
Presentations
Recombinomics
Publications
Recombinomics
Paper
at Nature Precedings