Skip Navigation


Antigenic Drift is Not the Sole Factor Driving Evolution of the Influenza Virus

By Amesh A. Adalja, MD, October 30, 2009

During interpandemic years, evolution of seasonal influenza viruses has been thought to accrue via minute mutations in the genetic code of the hemagglutinin (HA) protein, a process referred to as antigenic drift. The major change that leads to pandemics, called antigenic shift, is traditionally thought to be the product of reassortment between 2 or more strains. Evolution of interpandemic (seasonal) influenza A viruses has long been attributed solely to antigenic drift; however, a new study by a team of researchers from the National Institutes of Health (NIH) demonstrates that the process is more complex than previously thought. Results of the NIH team's study of the 2003/2004 and 2004/2005 flu seasons indicate that reassortment between co-circulating strains and genome-wide “selective sweeps” may be involved in determining which strains predominate from year to year.1

The Unexpected Viral Evolution in the 2003/2004 and 2004/2005 Influenza Seasons

The 2003/2004 flu season, which was particularly severe for an interpandemic year, was dominated by a lineage of H3N2, called Fujian-like. That virus emerged suddenly and unexpectedly and was poorly matched by the season’s vaccine. Studies revealed that the virus was the product of reassortment between 2 different H3N2 subtypes (clades A and B) that had been co-circulating for several years. 

The following flu season, 2004/2005, was characterized by another surprise. The Fujian-like strain and other related co-circulating clade A and B strains were suddenly and nearly completely supplanted by a new clade B lineage, called California-like. In addition to antigenic changes, this lineage had acquired mutations to the genes encoding the ribonucleoprotein (RNP) complex. This new virus, termed clade B’, completely replaced the prior year’s strains. This near total replacement of multiple co-circulating strains by a single clonal strain is referred to as “genome-wide selective sweep.” The NIH researchers hypothesized that these genetic changes in influenza would have phenotypic correlates that would explain the dominance of the B’ strain.

The 2004/2005 Clade B’ Viruses Were More Fit

In the current article, the NIH researchers detail several experiments that established important differences in the clade B’ strain:

  • Viral growth of clade B’ viruses was 1-2 logs higher than both clade A and B viruses in cell culture and 10- to 100-fold higher in ferret nasal tissues.

  • Clinical studies in ferrets demonstrated more severe clinical disease—including enhanced nasal discharge—with clade B’ viruses.

  • Pathological examination of ferrets infected with the viruses demonstrated more extensive disease in those infected with the clade B’ strain

  • The RNP of clade B’ viruses was demonstrated to be more highly active than the other clades, a marker of replicative fitness.

Influenza Strain Variants May Arise Independent of Antigenic Drift of the HA Protein

This study demonstrates that newly discovered mechanisms play a role in the year-to-year evolution of influenza. Significant changes in the dominant strain may result from reassortment between related strains, from mutations of internal proteins that enhance viral fitness, or from antigenic drift. Therefore, it is important to monitor genetic changes accruing in more than just the antigenic portions of HA protein. As the 2009 H1N1 influenza virus evolves, it will be important to study the entire genetic characteristics of its descendants and any other influenza strains that may co-circulate with it and to observe their replicative fitness.  


Memoli MJ, Jagger BW, Dugan VG, et al. Recent human influenza A/H3N2 virus evolution driven by novel selection factors in addition to antigenic drift. J Infect Dis 2009; 200:1232-41.
. Accessed October 26, 2009.