How long does it take the flu to evade a vaccine?

With all the recent Presidential attention to the threat from the H5N1 Avian Flu, and the many billions now earmarked for stockpiling vaccines and drugs, it seems like a good idea to ask on what time scale the virus might be able to evade these countermeasures.  Vaccination is of particular interest, as it is regarded as by far the best tool to combat viral infection at the population level.

Last February, I examined how little is known about the evolution of pandemic strains of influenza ("Avian Flu Uncertainties").  At the time, there was very little data concerning either the origin of pandemic strains or how often they should arise.  Forensic work on the 1918 pandemic flu suggested that virus was entirely avian in origin, which has now been confirmed by sequence analysis.  The WHO Global Influenza Program Surveillance Network published a paper in the journal Emerging Infectious Disease in October, "Evolution of H5N1 Avian Influenza Viruses in Asia", which states that:

Genomic analyses of H5N1 isolates from birds and humans showed 2 distinct clades with a nonoverlapping geographic distribution. All the viral genes were of avian influenza origin, which indicates absence of reassortment with human influenza viruses.

The WHO team also notes that:

Genetic and antigenic analyses have shown that, compared to previous H5N1 isolates, 2004-2005 isolates share several amino acid changes that modulate antigenicity and perhaps other biological function.  Furthermore, our molecular analysis of the HA from isolates collected in 2005 suggests that several amino acids located near the receptor-binding site are undergoing change, some of which may affect antigenicity or transmisibility.

That is, the currently circulating strain of H5N1 is in the process of becoming something else.  Of course, everything subject to variation and selection (e.g. evolution) is in the process of becoming something else, but the future of H5N1 is of particular interest since understanding it may help us plan for a pandemic.  (Incidentally, is just now sporting the most excellent headline, "WHO: Human flu pandemic inevitable".)

Despite the recent "Isolation of drug-resistant H5N1 virus," (Nature, 20 October 2005), the WHO Global Influenza team determined that recent isolates are "sensitive to 2 neuraminidase inhibitors that are recommended for prophylactic or therapeutic intervention against human infections."  So obviously the specific details of which viral isolate one is working with determine its sensitivity to drugs.  In the case of the drug-resistant strain, it was isolated in February 2005 from a Vietnamese girl who may have contracted the virus while she cared for her brother.  Human-to-human transmission is supported by the girl's lack of contact with poultry and the fact that the neuraminidase gene from virus isolated from the girl was virtually identical to that isolated from her brother.  Thus there are evidently drug resistant strains running around in the wild.

As far as I can tell, it just isn't yet clear how fast drug-resistance traits can spread.  There is now at least a little data about how long it takes a particular viral strain to find work-arounds for vaccines.  A recent initial effort to sequence many flu strains in parallel indicates that mutations that provide ways around vaccines can rapidly dominate a population of flu viruses. 

In, "Large-scale sequencing of human influenza reveals the dynamic nature of viral genome evolution," (Nature, 25 October, 2005) Ghedin et al., find:

Perhaps the most dramatic finding in our data is the discovery of an epidemiologically significant reassortment that explains the appearance, during the 2003-2004 season, of the 'Fujian/411/2002'-like strain, for which the existing vaccine had limited effectiveness. ...Phylogenetic analysis of 156 H3N2 genomes from our project revealed the clear presence of multiple, distinct clades circulating in the population. Through a reassortment event, a minor clade provided the haemagglutinin gene that later became part of the dominant strain after the 2002-2003 season.

That is, exchange of gene segments between subpopulations within a particular strain can provide the means for the strain to escape a previously effective vaccine.  Here is the important bit:

This finding illustrates not only that the influenza virus population contains multiple lineages at any given time, but also that alternate, minor lineages can contribute genetic variation to the dominant lineage, resulting in epidemiologically significant, antigenically novel strains. It is worth emphasizing that our sequence-based sampling approach--in contrast to traditional serologically based sampling--will reveal co-circulating strains even before they become antigenically novel.

In other words, the authors assert that amongst viruses that we give the same name there is considerable variation that may be hard to distinguish using traditional techniques.  Sequencing the genomes of many isolates can provide a map of how a population of viruses is changing in response to vaccines.  Ghedin, et al., note that their work demonstrates significant change of the dominant flu strain even within the 2003-2004 flu season.  That variation appeared to originate and then dominate the population of viruses within 12 months.

Which brings me to the core of this post, namely that we now have real-world data demonstrating that flu viruses can escape vaccines in less than a year, far shorter than the time it takes to produce significant quantities of effective vaccines.  It is important to note that this is a different problem than producing a new vaccine for new annual flu strains every year.  If a pandemic strain emerges, our problem is not planning ahead just far enough to deploy a vaccine for next year's strain, but rather to combat a strain already killing people worldwide.  It took most of a year for the CDC and Sanofi to come up with a hypothetical H5N1 vaccine, and the new National Strategy for Pandemic Influenza contains the expectation of years to accumulate enough vaccine to be useful for large populations.  Never mind that the existing whole virus vaccine may not be effective against a pandemic strain.

There is another significant point of concern embedded in the Ghedin paper:

The fact that the minor Fujian-like clade has donated its HA to the previously dominant strain rather than itself becoming the dominant circulating virus indicates that there may be important amino acid co-substitutions in the other proteins essential for viral fitness.

Which needs to be combined with another important bit:

...Even within a geographically constrained set of isolates, we have found surprising genetic diversity, indicating that the reservoir of influenza A strains in the human population -- and the concomitant potential for segment exchange between strains -- may be greater than was previously suspected.

We cannot think of the H5N1, or any other strain, as either a clonal population experiencing selection or as a bunch of individuals producing descendants that may accumulate mutations leading to a pandemic strain.  Rather, flu viruses exists as elements of a population that appear to be constantly innovating and trading parts.  This is a critical distinction, particularly in light of rapid human and avian intercontinental travel.  Not only have we now learned that there is greater variation in any given set of geographically linked isolates, but because of human travel we can expect all kinds of novel parts to show up in populations that were otherwise isolated and appeared to be of no immediate threat.

The Ghedin paper doesn't necessarily teach us directly about the evolution of pandemic flu strains, but it does suggest our current plan for pandemic vaccination is not well suited for the problem at hand.

After studying synthetic vaccines for Bio-ERA, amongst other clients, I think DNA vaccines are the best bet for rapid response on a time scale shorter than flu strains seem to evolve.  I've a draft paper on synthetic vaccines in for consideration at Biosecurity and Bioterrorism, and will shortly embark on another paper specifically about distributed manufacture of DNA vaccines.  PowderMed is waiting for publication of their (already accepted) first paper on their plasmid vaccine for the annual flu and will be starting trials of an H5N1 DNA vaccine early next year.  Unfortunately, it seems the folks in DC aren't taking this technology seriously, and instead blowing billions on developing cell culture production of whole virus vaccines.  Even the folks who manufacture vaccines in cell culture acknowledge this will only cut a month or two off the response time.