Intended for healthcare professionals

Letters Narcolepsy and pandemic A/H1N1 2009 influenza vaccine

Is the adjuvant solely to blame?

BMJ 2013; 346 doi: (Published 17 April 2013) Cite this as: BMJ 2013;346:f2375
  1. Theodore Tsai, head, scientific affairs 1,
  2. Giuseppe Del Giudice, head, translational medicine1,
  3. Antonio Crucitti, risk-benefit physician1,
  4. John Weil, head, epidemiology1,
  5. Vas Narasimhan, global head, development1
  1. 1Novartis Vaccines and Diagnostics, Cambridge, MA 02139, USA
  1. theodore.tsai{at}

We read with interest Miller and colleagues findings on the association between narcolepsy and the AS03 adjuvanted pandemic A/H1N1 2009 vaccine in England and, in particular, their call for investigations of other A/H1N1 2009 vaccines to ascertain the risk of narcolepsy.1 Risk estimates above 10 are unusual, and in pharmacoepidemiology lead to the suspicion of unknown confounding factors. The English data, confirming previous reports,2 3 considerably lessens the likelihood of confounding. However, the mechanisms by which this vaccine’s components may have contributed to an increased risk for narcolepsy are unclear and may be difficult to dissect, as the still uncertain pathogenesis of Guillain-Barré syndrome after flu vaccination illustrates.

Previous observations suggest that similar vaccines may have subtle differences that manifest themselves in different ways or in different seasonal formulations. Oculorespiratory syndrome has been associated with one split seasonal vaccine product but not with others.4 Perhaps appropriate in the context of narcolepsy cases occurring in 2009-10, the incidence of Guillain-Barré syndrome associated with seasonal flu vaccines has fluctuated, although never to as high as that seen in 1976, in association with the A/New Jersey/76 (Hsw1N1) pandemic vaccine.5 In addition, as Miller and colleagues noted, one observation suggested that the pandemic infection itself may have triggered an increase in narcolepsy cases.6 7 Thus, the pandemic virus, or its antigens, could have contributed to pandemic vaccine associated cases of narcolepsy in 2009-10.

Novartis Vaccines and Diagnostics distributed about 107 million doses of two pandemic vaccines containing the emulsion adjuvant MF59 (unpublished data, Novartis Vaccines and Diagnostics). These vaccines, containing egg or cell culture derived viral subunit antigens, were distributed in more than 17 countries, although the number of doses administered and their deployment by age are unknown. The association between narcolepsy and these vaccines was investigated through active case finding in the Netherlands in all 16 referral sleep centres, and no cases of associated narcolepsy were identified during the primary study period.8 An estimated 654 885 people aged 5-50 years (154 622 aged 5-19 years of age) with risk factors were vaccinated.9

Passively acquired data come from Argentina, where only an MF59 adjuvanted vaccine (10.8 million doses distributed) was used in the national pandemic response and high coverage was achieved in high risk people aged 5-64 years, but no cases were reported to the Administración Nacional de Medicamentos, Alimentos y Tecnología Médica.10 A search of the Novartis pharmacovigilance database disclosed one spontaneously reported case from Switzerland that met a level 2 Brighton Collaboration and European Narcolepsy Network case definition of narcolepsy, and another with insufficient confirmatory information classified as greater than level 3 (Novartis data on file).11 12 13 Other events of special interest that included certain autoimmune and neurological diseases were not spontaneously reported at unusual frequency in connection with the adjuvanted pandemic vaccines or MF59 adjuvanted subunit seasonal vaccine, which was first licensed in 1997.14 15

The absence of a narcolepsy signal in connection with pandemic vaccine adjuvanted with MF59 may reflect differences in adjuvant compositions (table) and mechanisms of action.16 17 Notably, MF59 is composed of a single emulsified oil, squalene, and does not contain α tocopherol. Although MF59 is a local activator only of the mouse transcriptome in muscle at the injection site, it does not upregulate genes controlling innate immunity distally, in draining lymph nodes.16 By contrast, AS03, which contains squalene coemulsified with the immunomodulator α tocopherol, activates innate immune responses in the draining lymph nodes of immunised mice, independently of antigen.17 Its more generalised immune potentiation is mediated directly by α tocopherol, which amplifies the adjuvant activity of AS03, making it more potent than a squalene-only adjuvant.17 But, as alluded to above, other factors may be involved. The association of narcolepsy, which is thought to be an autoimmune disorder, with the AS03 adjuvanted vaccine may not be strictly due to the adjuvant. Reporting rates for autoimmune disorders (not including narcolepsy) within EudraVigilance were no greater for adjuvanted pandemic vaccines (including MF59 and AS03 adjuvanted ones) than for non-adjuvanted vaccines.18

Composition of MF59 and AS03 adjuvants that are components of licensed adjuvanted vaccines (mg/dose)

View this table:

In summary, the available but limited observations have not signalled an increased risk of narcolepsy associated with MF59 adjuvanted pandemic vaccines. The cases in 2009-10 seem to have complex causes, and contributory roles of the pandemic virus or vaccine antigens cannot be excluded. However, it may be premature to attribute the increased risk associated with the AS03 adjuvanted vaccine solely to the adjuvant.


Cite this as: BMJ 2013;346:f2375


  • Competing interests: All authors are full time employees of Novartis Vaccines and Diagnostics, a manufacturer of adjuvanted pandemic influenza vaccines.


View Abstract