Pandemrix and Narcolepsy

In 2010, following the H1N1 pandemic and the vaccination campaign to reduce its impact, researchers noted a significant increase in a rare neurological disorder, narcolepsy, in Sweden and Finland. Since then researchers have been studying a possible association between a specific H1N1 flu vaccine, Pandemrix by Glaxo-Smith-Kline (GSK) and a sudden onset of the sleep disorder narcolepsy. In those two countries the association seems strong, but the full story is still complicated with many unknowns.

Narcolepsy is a neurological disorder marked by excessive sleepiness, cataplexy (sudden loss of muscle tone, usually triggered by emotions) and disordered sleep. Almost all cases are associated with low levels of hypocretin in the hypothalamus – this is a hormone involved in sleep regulation.  Further there is a strong HLA (human leukocyte antigen) association – specifically DQB1*0602. HLA is a group of proteins involved in regulating immune activity. An HLA association strongly suggests that narcolepsy may be an auto-immune disease.

The current synthesis of this information is that narcolepsy occurs in genetically susceptible individuals after some environmental trigger, such as in infection, that causes the immune system to attack and destroy hypocretin cells in the brain.

It is currently not known what the precise mechanism is for the increase in narcolepsy in the two “signal countries,” Sweden and Finland. Possibilities include immune stimulation from the vaccine itself, increased immune activity from the adjuvant ASO3, immune activation from H1N1 infection, or a combination of these factors, perhaps with some other infection present at the time.

Epidemiologically, at this time it seems that the increase in narcolepsy in Finland and Sweden among children <20 (but not in adults) is a fairly strong signal and has been reasonably confirmed. This does not establish by itself a specific cause and effect, but other factors such as timing and the relative risk of vaccinated vs unvaccinated children make a causal link to the Pandemrix vaccine plausible. The increased risk from the Swedish data is about 1 case of narcolepsy per 27,800 vaccinations.

The Pandemrix vaccine was used throughout Europe and Scandinavia, and in a few South American countries, but not in the US (which used a vaccine without adjuvant). A related vaccine was used in Canada during the H1N1 outbreak. It is still uncertain whether other European countries also had an increase in narcolepsy. Data from France is conflicting. A 2011 study found no reports of narcolepsy following the vaccine through June 2010. A later report found a possible increase, but the pattern does not match the patterns seen in Sweden and Finland, with more adults affected in France.

There is still a debate as to whether there is a signal of increased narcolepsy in other countries that received the Pandemrix vaccine, but as yet this has not been confirmed. A recent review of the evidence concluded:

According to the strictest primary analysis, the kind of assessment designed to avoid most biases like media and diagnostic awareness biases, no significant risk was found in children and adolescents in other countries included in the study – Denmark, Italy, France, the Netherlands, Norway and the United Kingdom (non-signalling countries).

But there may be a signal when only data from prior to media attention are considered. This likely means that if there were an effect, it is much smaller than what was seen in Finland and Sweden. This leads to the question of why such a dramatic difference in different countries receiving the same vaccine. This cannot be explained by genetics alone, although this is a likely factor. This suggests that other environmental factors may be important, and even necessary, for narcolepsy risk to occur. There also may be confounding factors that have not yet been identified.


It now seems likely that some combination of factors, including but not limited to the Pandemrix vaccine, led to a spike in narcolepsy cases among children but not adults in Finland and Sweden.  The same effect has not been confirmed in other countries, but a smaller effect has not been ruled out.

Use of Pandemrix has been suspended because of this concern, which is appropriate. It must be emphasized, however, that this effect, if confirmed, has no relevance to the flu vaccine in general. It has led to increased surveillance for narcolepsy following vaccination, to be cautious, but at present there is no reason to be concerned about this risk with the seasonal flu vaccine, even those targeting H1N1.

Depending on the results of further investigation, this episode might lead to genetic typing for the DQB1*0602 variant and vaccine recommendations based upon this genetic type. It must be recognized, however, that infection can also trigger narcolepsy in susceptible individuals, and therefore all factors will be considered in making recommendations. At present there are still too many unknown variable to say what the implications of the Finland and Sweden experience are.

Posted in: Vaccines

Leave a Comment (11) ↓

11 thoughts on “Pandemrix and Narcolepsy

  1. geack says:

    How do we make this the mandatory worldwide template for all emerging health issue reporting?

  2. Janet says:

    This is fascinating and such a good exercise in how to evaluate a piece of medical information that one might encounter in a headline or “health” article.

    Thanks for the details; this is bound to pop up in anti-vax circles.

  3. windriven says:

    From the WHO*:

    “During 2009-2010 [Finnish authorities] found that the risk of narcolepsy among people aged 4-19 years old who had received pandemic influenza vaccine was nine times higher than that among those who had not been vaccinated. This corresponds to a risk of about 1 case of narcolepsy per 12,000 vaccinated in this age group. No increased risk has been seen in younger or older age groups.”

    In the same report Swedish authorities cited a 3 per 100,000 rate among vaccinated individuals.

    I was not easily able to find the vaccination rates for these countries, much less the rates for these particular demographics. Seems like a very small signal in a very noisy data set.

    The observation that the (HLA) DQB1*0602 genotype is nearly universal among those developing narcolepsy is interesting in a gross sense but without more information about the distribution of that genotype in the larger population it seems more a curiosity than a clue.

    I hope some of the commenters will have some additional light to shed.


  4. daedalus2u says:

    Flu vaccine is made from virus grown on individual chicken eggs. In times past, thimerosal was used to suppress bacterial growth during the incubation of the eggs and the growth of virus. It is my understanding that there were problems with how fast the virus grew and in yields of suitable virus for production of vaccine, which was one reason that adjuvant was used.

    This virus was expected to be extremely problematic, which was why the new virus was rushed into vaccine production.

    There are autoimmune neurological disorders that can be triggered by infections, for example PANDAS. Presumably it is not the infection per se, but rather the immune system response that triggers the autoimmunity.

    It may be that trace levels of specific bacterial antigens in different lots of vaccine were sufficient to trigger narcolepsy in susceptible individuals due to the conflation of multiple circumstances, including the immunological naivety of those affected. However there was also an association with infection by the virus which would imply a viral antigen, or virus infection acting as an adjuvant to cause sensitization to antigens from resident bacteria.

    It will be interesting to see if in the years to come, there is a reduction in the rate of new narcolepsy in this exposed population. In other words, did this exposure simply accelerate the development of narcolepsy into the period immediately following the exposure, or are these new cases that would not have happened without the exposure. But that will take pretty good followup because the rates are so low.

  5. daedalus – I wondered that to. Are the extra cases people who would likely have developed narcolepsy anyway. If so we should see a drop in cases from the prior baseline over the next 20 years. This question will take a long time, if ever, to sort out.

    My sense is, though – probably not. The pool of people with the HLA predisposition is much greater than the number of people who develop narcolepsy. So I don’t think this episode depleted the pool of susceptibles appreciably, but it would be interesting to see specific numbers on this.

  6. PernilleN says:

    Is PANDAS really a recognized disease? I thought it was rather hypothetical so far?

  7. PANDAS remains controversial. I personally don’t buy it, but nor would I rule it out just yet. PANDAS proponents have not put all their ducks in a row yet, IMO.

  8. ladentduchat says:

    Hello. A short comment from Sweden.

    The vaccination rate was about 60% in Sweden.* The possible problem with Pandemrix has been debated very much and still is. I am sorry to say that most of the public debate now focus on whether the government should be held responsible for the alleged raise in narcolepsy and thus “pay up”. I am not quite sure for what though, since we have (more or less) free health care, extra family expenses perhaps.

    I do not know if “genetics” could play a role at all in this case. But if so, it would be interesting to see some statistics about it. Sweden has been, and still is, mostly populated with people who is related to each other. Moreover, we did spread into Finland not that long ago. A rough guess would be that about 10% Finns is of Swedish heritage. **

    They have. More importantly, one of Sweden’s most wide spread daily “serious” news papers has been campaigning against the national vaccination policy (and at the same time scaring people due to blunt reporting). I have not yet seen any numbers if the campaign has had any effect on the different vaccination rates though.

    * The source has ISBN: 978-91-86585-91-4 (its in Swedish though =)
    ** I am way out of my comfort zone “knowledge-wise” here

  9. rbryant says:

    It will be interesting to see if in the years to come, there is a reduction in the rate of new narcolepsy in this exposed population. In other words, did this exposure simply accelerate the development of narcolepsy into the period immediately following the exposure, or are these new cases that would not have happened without the exposure.

    It may be that, as you suggest, the pool of genetically-susceptible people can be depleted by exposure, and that, following an extensive, population-wide exposure, the incidence of narcolepsy will decrease for some subsequent years.;jsessionid=36FA2AD0EACFD9C59D72E4D5AD23D808.d03t03

    BTW, Mignot et al. have shown that narcolepsy tracks (months later) with seasonal respiratory diseases such as influenza in China, where the incidence of narcolepsy tripled during the H1N1 pandemic:

  10. Badly Shaved Monkey says:

    What happened/is happening to those narcolepsy cases following onset? Is it a permanent defect or a resolving problem?

Comments are closed.