Technical report In Brief: Narcolepsy in association with pandemic influenza vaccination

Conclusions/considerations

Background and subsequent incidence rates of narcolepsy by age and time

• The pooled background incidence rate of diagnosed narcolepsy was low and stable at around 1 per 100 000 PY between 2000 and 2010 (0.85/100 000 PY prior to the vaccination campaigns).
• Lower background rates of diagnoses were observed among children: <5 years and 5–19 years (0.12/100 000 PY and 0.56/100 000 PY, respectively
• No increased incidence rate of narcolepsy was observed in temporal association with the 2009 pandemic itself.
• The overall incidence rates of narcolepsy differed substantially between the signalling and non-signalling countries after the start of the vaccination campaigns: 1.67/100 000 PY vs. 0.95/100 000 PY, respectively.
•  In the signalling countries the following incidence rate patterns were identified:
  • In Finland, an increase in the incidence rate of narcolepsy diagnoses after September 2009 was observed in children and adolescents 5–19 years of age with a relative risk of 6.4 (95%CI 4.2-9.7).
  • − In Sweden, a similar increase was also observed after September 2009 in the 5–19  year age group with a relative risk of 7.5 (95%CI 5.2–10.7). The pattern was different in the other non-signalling five study countries during the period for which data could be
    
    analysed. While in Denmark a small increase in the incidence rate of narcolepsy was also observed, the upward trend started earlier, prior to the start of the vaccination campaign (focusing on risk groups only) and in a different age group. In the Netherlands, the UK and Italy no increase in incidence was seen, however vaccination coverage was low in all the non-signalling countries. Associations of narcolepsy with risk factors including influenza, other infections, influenza A(H1N1)pdm09 vaccination and other vaccinations.
• The primary analysis focusing on the primary study period before professional/regulatory and media attention in the signalling countries Finland and Sweden showed an association between pandemic influenza immunisation and narcolepsy in children and adolescents (5-19 years) but not in adults. This confirmed
  
  observations made with different methodologies in those countries in national studies.
• A similar association in the primary analysis was not found in the non-signalling countries.
• Sensitivity analyses highlight the importance of time-related factors for the strength of association. It should be noted that sensitivity analyses of different study periods and different index dates may contain a mixture of potential vaccine and/or regulatory and media attention effects.
• All epidemiologic studies investigating the association between the influenza A(H1N1)pdm09 vaccine and narcolepsy have the challenge to address a multitude of time-related biases, and results obtained beyond the primary study period should be interpreted cautiously.
• To increase statistical power for further association analyses on these observations, especially for cases diagnosed prior to regulatory/media attention the following possibilities should be considered:
  • pooling all data available across Europe (including data from studies when completed from the Swedish Medical Products Agency, Finnish National Institute for Health and Welfare, Irish Health Service Executive and Health Protection Surveillance Centre, and the UK Health Protection Agency)
  • finalising VAESCO study case inclusion especially in Finland, France, and Sweden  where many cases could not be included at the time of completion of the contract and which is particularly important for this investigation
  • including further European countries with significant vaccine coverage such as Ireland
  • extending the investigations beyond Europe in countries where narcolepsy cases could be diagnosed and where media attention was less pronounced but where Pandemrix and other pandemic vaccines (including the adjuvanted Arepanrix) were offered for children.
• The observations are sufficiently strong and consistent for children in the signalling countries using different methodologies to warrant further investigations as to a possible mechanism. While it is the case that the vaccine concerned is one with a novel adjuvant there is a confounding factor in Europe in that this was the only vaccine offered for children in any volume in Europe. Hence it cannot be concluded that the adjuvant is
  
  the cause of the observation. Equally there is also the possibility that it is the combination of vaccination TECHNICAL REPORT Narcolepsy in association with pandemic influenza vaccination and influenza transmission or another unrecognised infection or environmental factor in individuals with developing nervous systems. There will be a role also for animal models and mechanistic studies. As of August 2012 more than 600 narcolepsy cases (including more than 100 in adults) have been reported spontaneously to the EMA EudraVigilance database following the influenza A(H1N1)pdm09 vaccination. Cases are still being reported in 2012 but in lower numbers compared to 2010–2011. Surprisingly, more than 500 of the cases originate from Member States being part of the current VAESCO project. Continued investigations are needed in the VAESCO countries as well as in other countries where influenza A(H1N1)pdm09 vaccines were used.