As my last research job involved vaccine development and safety, I get a lot of emails and Facebook posts relating to the safety of particular vaccines. I often get sent links from Natural News, which is renowned for its biased and inaccurate reporting of anything scientific. In particular, there was a lot of chatter about the H1N1 vaccine and its link to narcolepsy in Europe (http://www.naturalnews.com/039732_flu_vaccines_narcolepsy_children.html). The truth of the situation is, of course, a bit murky and the nuances of how science has pulled apart this issue get lost. A recent retraction of a key paper in the story is complicating the situation.

Soon after the 2009 H1N1 outbreak, medical authorities in Sweden and Finland documented an increase in the number of cases of narcolepsy. Narcolepsy is a disorder in which the sleep-regulation hormone hypocretin is too low, resulting in sleepiness, cataplexy (sudden loss of muscle tone) and disordered sleep. The reason for the increase in Sweden and Finland, (subsequently also seen in the UK) was associated with the use of Pandemrix, a particular formulation of the H1N1 vaccine made by Glaxo-Smith-Kline. Importantly, this particular vaccine was not used in Canada or the US.

In this case, the risk of developing narcolepsy for children who received pandemrix was around 0.36  (Science, 345(6196):498) to 6.67 per 100,000. In contrast, the risk of hospitalization for all types of flu run between 66 and 99 per 100,000 cases . In other words, the increased risk of narcolepsy from is negligible compared to the risk of developing complications from the flu, a relatively common disease during Canadian winters.

Still, a side-effect from vaccine is a serious issue. The association between Pandemrix administration and the rise of narcolepsy in Europe suggests that Pandemrix was a likely cause, but much work needs to be done in order to determine why, including an examination of the molecular mechanisms behind the side effect. Vaccines induce immunity against an infectious organism by presenting your body with a tiny bit of that bug such as a bit of a cell wall or a bit of a protein. This is called an antigen, and presentation of just the antigen allows your immune system to mount a response against it without you having to suffer the illness. In the case of Pandemrix, it’s increasingly looking like narcolepsy was induced by an auto-immune reaction: somehow, the small bit of the flu particle included in the Pandemrix vaccine looked like the hypocretin molecule. Thus the immune system of susceptible kids who received Pandemrix would suddenly attack the hypocretin producing cells, thinking that they were the flu.

However, one of the key papers demonstrating this phenomenon was retracted this week because the key experiments could not be replicated (Science, 345(6196): 498). The original paper (Herran-Arita, 2013. “CD4+ T Cell Autoimmunity to Hypocretin/Orexin and Cross-Reactivity to a 2009 H1N1 Influenza A Epitope in Narcolepsy.” Sci. Trans. Med. 5(216): 216ra176) demonstrated how particular immune cells attack hypocretin. It was a key paper supporting the hypothesis that narcolepsy is an auto-immune disorder, and that Pandemrix triggered this reaction in susceptible kids.

What does this mean for parents trying to decide if vaccines are safe? It is still accepted that Pandemrix contributed to the rise of narcolepsy, and that the antigen in Pandemrix was the cause. However, it will take years to understand all the molecular details. In the meantime, vaccine designers are increasingly aware of the potential for side effects if the chosen antigen is too similar to a host protein, though now that the Mignot and Mellins paper has been retracted there are no documented cases of an antigen triggering an autoimmune reaction. The story of Pandemrix does show how vaccines can have side effects. However, the rate of those side effects is still far lower than the rate of complications with vaccine preventable diseases.

There are risks to any medical intervention; the job of the parent is to knowingly weigh those risks, which is very difficult. We like to think that we behave in rational ways, but we don’t. We are extremely poor at understanding risk, and because of this people are much more comfortable committing a sin of omission than commission. When applied to vaccination, it is much easier for parents to accept making a mistake by not vaccinating their children, than by vaccinating them and risking a side-effect. That’s why it’s so important to quantify risks. In this case, it’s clear that the risks of developing narcolepsy is roughly 11-300 times lower than the risk of complications from the flu, and furthermore, those risks are only for one type of vaccine. H1N1 vaccines in Canada and the US did not have the same side effect. Meanwhile, the discovery of a side effect can hit the media full force, generating fears among the public. It’s extremely common for the media to report the results of “an important new study,” as if it is definitive on any subject. The retraction of the Mignot and Mellins paper demonstrates how dangerous this is; science is a slow and careful practice. It’s important to remember that demonstrating cause can take years of difficult and complicated study and require multiple verifiable experiments. Nearly 100 years of use have shown that vaccines are safe and effective. When reports of side effects surface, weigh the risks, and remember that one study is never enough.

 Why Retractions?

Websites like Retraction Watch (retractionwatch.com) have shown that the rates of retraction have been increasing in recent years. Part of the reason for this increase is the high pressure for investigators to publish or perish, particularly for young investigators for whom a high profile publication can set up their career. But this retraction is particularly interesting: the original authors themselves could not replicate their experiment. The authors of the paper (corresponding authors Emmanuel Mignot and Elizabeth Mellins, both at Stanford) deserve commendation for their honesty. The immunological assays that they were working with are notoriously tricky; when, after publishing the first paper they returned to the experiment to build on the results, they couldn’t get it to work again (http://retractionwatch.com/2014/07/30/authors-retract-paper-confirming-that-narcolepsy-is-an-autoimmune-disease/). Where other scientists may have thrown up their hands, they notified Science Translational Medicine and retracted their work.