Report on the association between the 2009 pandemic vaccine (Pandemrix®) and narcolepsy – Republic of IrelandArchived
Investigation of an increase in the incidence of narcolepsy in children and adolescents in 2009 and 2010 - Final Report of National Narcolepsy Study Steering Committee, Department of Health, Ireland
Investigation of an increase in the incidence of narcolepsy in children and adolescents in 2009 and 2010 - Final Report of National Narcolepsy Study Steering Committee, Department of Health, Ireland.
In late August 2010 Swedish and Finnish authorities let it be known internationally that they were investigating cases of narcolepsy reported by health care professionals as a possible adverse event following the use of the Pandemrix® vaccine used during the 2009 influenza A(H1N1) pandemic which had ended earlier in August. See ECDC Pandemic Safety Time Lines. The Finnish Agency stopped recommending use of the specific vaccine (it was then used little with the pandemic over and seasonal vaccines available again) and the European Medicines Agency (EMA) undertook an initial review requesting further research and has been monitoring since then.
As documented in the Time Lines there have been many developments since including publication of the Finnish findings in medical journals.(1,2) ECDC and a number of other countries started studies including Ireland where in March 2011, the Irish Medicines Board (IMB) received reports of two confirmed cases of narcolepsy following vaccination with the 2009 pandemic vaccines.
Since then the Irish Health Service Executive Health Protection Surveillance Centre (HPSC) has been leading intensive work for the Department of Health with the IMB and national clinical experts to undertake a cohort study researching all narcolepsy report back the findings. On April 19th the Irish Ministry of Health has made public the initial findings.(3) This Public Health Development summarises and comments on those results of such further Irish research on narcolepsy.
The researchers undertook a retrospective population-based cohort study to calculate the incidence of narcolepsy in vaccinated and non-vaccinated individuals. Narcolepsy cases, with a symptom onset from April 2009 were identified through active case finding, by contacting all sleep clinics, neurologists, paediatricians, general practitioners, psychiatrists, psychologists and public health nurses in Ireland. Vaccination status, date of vaccination and vaccine brand were cross-checked in the vaccination databases.
Two experts, who were blinded to the vaccination status of the cases and involved with the diagnosis or management of the cases included in the study, reviewed their clinical history in order to confirm the diagnosis and classify them using the internationally agreed Brighton case definition for narcolepsy.(4) The primary follow-up time was 01/04/2009 to 31/12/2010 but different study time were employed in sensitivity analyses. Further important details are in the publication.(3)
The results showed that 32 cases met the case definition for narcolepsy with a symptom onset from April 2009. Twenty eight of these 32 cases occurred in children/adolescents aged 5-19 years old. Based on the first contact with health care because of narcolepsy symptoms, the incidence of narcolepsy during the primary follow-up time (01/04/2009 to 31/12/2010) was 5.8 [95% CI: 3.5-9.0] per 100,000 person years in the vaccinated and 0.5 [95% CI: 0.2-1.0] per 100,000 person years in the unvaccinated individuals. Thus was a highly statistically significant 13-fold higher risk of narcolepsy in vaccinated compared to unvaccinated individuals. The 95% confidence interval was 4.8 to 34.7). The absolute increased risk associated with the vaccine was five narcolepsy cases per 100,000 vaccinated children and adolescents. The authors identified only 4 cases in adults with new onset of narcolepsy symptoms from April 2009. The median delay found between the vaccination and the first reported symptom of narcolepsy was 2.2 months (min: 2 weeks; max: 12.8 months).
The authors particularly considered whether there was a potential source of bias in case selection from the extensive media attention within Finland and Sweden following August leading to increased care-seeking in Ireland, though they could not find any evidence of this in the Irish data. Indeed going for a shorter period of study (April 2009 to August 2010) slightly increased the relative risk. They also estimated that they would have needed to have missed an additional 55 cases in the unvaccinated cohort of young people to bring the relative risk down to unity. It seemed highly unlikely that so many cases would have been missed given the intensive case finding.
ECDC Comment (19 April 2012)
These finding of a powerful epidemiological association between this specific pandemic vaccine and narcolepsy in children and adolescents, but not in adults are similar to those reported by Finland and Sweden.(1-3) However as stressed by the Finnish researchers this does not mean that the vaccine alone was causative. It may be that there were other factors involved such as an interaction of vaccination with other factors such as pandemic influenza, another as yet unidentified circulating infection, genetic or environmental factors.[1,2] Though this is the first analysis showing an association of Pandemrix® and narcolepsy outside the Nordic countries the results are not unexpected as Ireland was one of the countries that successfully immunised many healthy children and adolescents during the pandemic, and particularly used this vaccine.[4,5]
Vaccines are regulated and authorised at a European level by the European Medicines Agency (EMA) and the European Commission as well as at a national level. All EU Member States are required to provide to EMA any reports of any adverse events following immunisation (AEFIs) through an EMA system called Eudravigilance. That should includes narcolepsy or symptoms suggestive of narcolepsy. However such spontaneous reporting is known to have its limitations; therefore it can be especially problematic with newly recognised, vague or emerging syndromes such as narcolepsy. Following the reports of narcolepsy in Finland and Sweden, the European Medicines Agency (EMA) initiated a review of Pandemrix® and narcolepsy at the request of the European Commission under Article 20 of Regulation (EC) No 726/2004 and in July 2011 EMA recommended the use of Pandemrix® in people under 20 years of age only if the seasonal trivalent influenza vaccines were not available. In fact except where there were shortages in seasonal vaccines the use of any pandemic vaccine in the EU/EEA countries fell to near zero after the new seasonal vaccines (containing the pandemic antigen) became available in the early autumn of 2010. In Ireland itself there was some use of Pandemrix in the early autumn of 2010, but on a precautionary basis it was not given to children and adolescents.
Since the first signals in Sweden and Finland, ECDC and its VAESCO study have been extending studies around vaccine safety to test whether a true association between infections and vaccines including Pandemrix and narcolepsy is being seen in Europe. The studies follow an established methodology that ECDC and VAESCO have agreed upon and used earlier. More details on the studies are available on the ECDC web-site.[6,7] A study on Guillain-Barre syndrome has already been completed and published. The VAESCO narcolepsy study involves eight countries and its main objective is using a case-control methodology to provide results confirming or rejecting the hypothesis raised in Sweden and Finland. Cases for analysis have come in at different paces by country with the most data coming from Sweden and Finland. These confirm the association in Sweden and Finland, on the same cases but using a different methodology. However the data received are complex and incomplete and a more in depth analysis is therefore needed before publication. In addition there are also additional independent national studies underway at least in Norway and the United Kingdom.
Narcolepsy is proving a challenge to study. All cases have to be carefully validated against agreed international case definitions preferably by independent experts unaware of the vaccination status. Onset of the condition is usually gradual and so ascribing a date of onset for comparison with vaccination date. Possible diagnostic bias and care seeking behaviour is a major methodological issue. Media attention and professional awareness could have had an effect on shortening lag times from onset of disease to investigation and on increased diagnosis rates. It is also possible to imagine that where there is awareness families whose children have been immunised looking more carefully for symptoms. Because of the lag time from first symptoms to diagnosis it is unclear when the ‘time at risk’ begins, following immunisation. Hence the ‘risk window’ remains unknown presenting a difficulty to establish which cases could be linked to vaccination, even if diagnosed shortly after administration of the vaccine. The national studies and the VAESCO studies are endeavouring to overcome these challenges. Two consequences of this is that studies cannot be rushed and that even after a completion date cases continue to accumulate meaning there will then need to be follow-up analyses.
Comments or questions on this public heath development are welcome and should be addressed to: firstname.lastname@example.org
Understanding the association between narcolepsy and one of the 2009 adjuvanted influenza A (H1N1) vaccines
The roles of influenza virus antigens and the AS03 adjuvant in the 2009 pandemic vaccine associated with narcolepsy needs further investigation