Among influenza, the vaccine is not a trendsetter

I’m still shaking off the remnants of what I suspect was an influenza infection that started almost two weeks ago, despite getting my annual flu shot like I always do. Stat news reported this week that this season’s vaccine is less effective than usual, information that I found equally comforting and annoying. The H3N2 strain is predominant this season but the H3N2 vaccine component does not provide as much protection from its target strain, for reasons that are purely speculative.

The reason we need a new flu shot every year is due to the antigenic drift of the virus- that is, the flu acquires changes that make it look different over time. The flu vaccine gives our immune system a picture of a suspect, but every season the suspect gets plastic surgery and a haircut!

Influenza has a short generation time (hours) and a high mutation rate, so within a single infected host we will find lots of viruses that all have very slight changes. Selective pressures within this infected host will prune away some of these variants. For example, if a mutation makes a virus suddenly unable to cope with the hazards of living in the human airway, that virus will either outright be destroyed or be unable to make more copies of itself. And we’ll never hear from that variant again.

But consider the immune system, an important pressure on the virus. Unvaccinated people can still fight and clear an influenza infection. If a mutation makes a virus more susceptible to the immune system, that virus is less likely to make it out of its host and propagate through the rest of the population. But a mutation that helps a virus escape the immune system (even better if it can escape the immune system of multiple people), it is more likely to become prevalent in the overall population of influenza circulating through a human community.

Vaccination gives a person a better, faster immune response to influenza infection. In a vaccinated person, you can imagine that the selective pressures exerted by the immune system are more strongly applied to the virus and at an earlier time during the course of the infection.

The overall diversity of influenza viruses within an infected person is driven by its fast mutation and growth rates, and limited by the selective pressures within a host. This process, and how it translates into the diversity of influenza in a group of people, is still largely a mystery.

Some researchers hypothesize that vaccination might accelerate the evolution of the virus into strains that are less recognizable by the immune system. Say that our virus starts out with short blue hair and a round face. Because it takes each unvaccinated immune system a while to even get a description of this suspect, at the end of the season the virus might end up with short light blue hair and a minor nose job. Compare this with the vaccinated immune systems that know immediately to look for short blue hair/round face, do we end the season with multiple different viruses with red hair and full face lifts?

A new PLoS Pathogens paper tested this hypothesis using the H3N2 strain of influenza. These researchers looked at flu virus in throat swabs that were part of a clinical trial on flu vaccine effectiveness in humans and developed a way to identify and quantify variant flu viruses within each infected person.

Once the variants were identified, they used the data set to ask if any specific variant virus existed that was specifically infecting and propagating within one of the vaccination groups. In fact they found no relationship between the variants found within the vaccination groups of the trial.

Next, the researchers looked at virus diversity at the individual level. Again, they found no difference in the number of variants found within vaccinated people compared to unvaccinated people. The researchers narrowed their focus from raw variant number to variants within antigenic regions — parts of the virus that are responsible for being recognized by the immune system. If vaccination was specifically driving viral evolution with more pressure than the unvaccinated immune system, we might expect to see more changes in antigenic regions. However, variants with mutations in antigenic regions were found similarly across both vaccinated and unvaccinated groups.

The results of this study provide evidence that exonerates the flu vaccine from significantly contributing to the steady evolution of influenza as it makes its rounds each winter. Not only is the flu vaccine not driving the creation of more mutant viruses within a single person, it also does not seem to specifically give rise to super viruses in the vein of multidrug-resistant bacteria. At the end of the season we’ll still end up with the plain old light blue-haired flu, regardless of how many people got their flu shots.

However, the conclusions of this paper also contradict some past studies. Research in this area is challenging because of the use of human subjects and the many confounding factors of human diversity, infection route, and prior flu history (both infections and vaccinations). I’d love to see this workflow applied to more trials from different parts of the world and different flu strains to see how well the conclusions hold up. It would be a big win for vaccine advocates to be able to point to multiple, robust human studies showing that the flu vaccine does not push influenza to be more dangerous within or between seasons.

I’m pretty excited about virology in general, but I also wanted to highlight this paper because its an excellent example of why negative data is important. It is critical that we understand how vaccines affect the evolution of their target virus especially in a case like flu, where ideally we would like to see just about everyone receive the vaccine. The literature on the subject may be divided, but imagine how much more uncertain we would be if the only studies that were published were the ones suggesting that such a link exists! It is often just as important to know that there’s no relationship between two things as it is to establish one.