Narcolepsy and Pandemic Influenza Vaccination

What We Need to Know to be Ready for the Next Pandemic

Abstract

After the initial identification of the H1N1 pandemic influenza strain in Mexico in April 2009 and its subsequent global spread, several monovalent influenza vaccines were developed as part of the pandemic response. Three of these vaccines, Pandemrix, Arepanrix and Focetria were adjuvanted. One of these, the AS03-adjuvanted Pandemrix vaccine, was primarily used in Europe. Following widespread Pandemrix vaccine administration in Scandinavia, an increased risk of narcolepsy was noted in observational studies. Subsequently, this increased risk was also reported in other European countries as well. In contrast, studies from Canada of a similar AS03-adjuvanted vaccine, Arepanrix, did not demonstrate a similar increased risk of narcolepsy. No studies have identified an increased risk of narcolepsy following the MF59-adjuvanted Focetria vaccine. For many potential pandemic influenza strains, adjuvants might be required to solicit a protective immune response. Thus, it is critical that we understand the nature of the association between adjuvanted vaccine receipt and narcolepsy. Here, we present a potential hypothesis for narcolepsy seen during the 2009 H1N1 pandemic in AS03-adjuvanted influenza vaccine recipients.

First identified in April 2009 in Mexico, the H1N1 A/California pandemic influenza strain rapidly spread through North America and then globally.¹ As part of a World Health Organization-coordinated pandemic response, manufactures developed 3 adjuvanted vaccines: Pandemrix AS03-adjuvanted vaccine utilized primarily in Europe, Arepanrix AS03-adjuvanted vaccine utilized primarily in Canada and Focetria MF59-adjuvanted vaccine utilized globally.² The possibility of an increased risk of narcolepsy associated with Pandemrix was first identified in Finland and subsequently observed in Sweden.³ Since that time, many studies have been conducted to evaluate this association.^4–11 Thus, at this time, one is left with multiple studies in Europe showing an increased risk with Pandemrix, a single analysis in 1 small study in Quebec demonstrating a low level of increased risk following Arepanrix¹² and other studies showing no risk with either the closely related Arepanrix vaccine or the MF59-adjuvanted H1N1 vaccine.¹³ Because it is critical to better understand these data in preparation for the potential use of adjuvanted vaccines in a future pandemic, we have reviewed the existing literature and propose a hypothesis linking receipt of AS03-adjuvanted Pandemrix vaccine and natural influenza infection to narcolepsy.

INFLUENZA PANDEMICS

Influenza pandemics occur when a new influenza A strain evolves and circulates in humans not previously exposed to this strain. This frequently leads to widespread disease in all ages, often associated with high rates of morbidity and mortality. At the onset of a pandemic, it is not possible to predict the ultimate severity of the outbreak; therefore, the rational public health response is to rapidly produce and administer pandemic strain vaccines as widely in the population as possible. Many vaccines were developed at the onset of the 2009 H1N1 A/California pandemic. Ultimately, there were 8 pandemic vaccines licensed in Europe, with 3 containing adjuvants, as outlined above, to enhance immune responses and to allow antigen sparing.¹⁴ Limitations of influenza vaccine production during the pandemic resulted in a 5-month lag period between the identification of the pandemic strain and the availability of the pandemic vaccines. Thus, wild-type virus was no longer circulating extensively in some countries, including North American when vaccine became available. In contrast, in other countries, including Europe, vaccination and wild-type virus circulation occurred simultaneously.

ADJUVANTED VACCINES

Pandemrix, one of the 8 vaccines produced in Europe, contained the adjuvant, Antigen System 03 (AS03). This vaccine, produced by GlaxoSmithKline (Rixensart Belgium), was composed of 3.75 µg per dose of H1N1 A/California hemagglutinin combined with the AS03 adjuvant.¹⁵ The AS03 adjuvant contained squalene and alpha-tocopherol. The Pandemrix AS03-adjuvanted influenza vaccine was widely distributed in Europe with an estimated 30.8 million doses administered. In several European countries, including Sweden and Finland, this was the only pandemic vaccine available and uptake was very high. A similar AS03-adjuvanted pandemic vaccine produced at a different GSK manufacturing site with a slightly different manufacturing process called Arepanrix was used in Canada. The third adjuvanted vaccine, Focetria, included 7.5 µg per dose of H1N1 A/California hemagglutinin combined with the adjuvant MF-59, containing only squalene, was used in some countries in Europe, but far fewer doses were distributed there. Adjuvants are added to vaccines to enhance the immune response to the antigen by serving as a depot for the antigen, stimulating immune cells, or enhancing antigen uptake, often allowing dose sparing of the vaccine antigen. This is particularly appealing in a pandemic when vaccine antigen may be in short supply.¹⁶

NARCOLEPSY

Narcolepsy is a rare sleep disorder characterized by excessive daytime sleepiness that persists for at least three months. In children, narcolepsy is commonly associated with cataplexy—the abrupt loss of muscle tone especially in association with emotional situations. The diagnosis of narcolepsy-cataplexy is confirmed by polysomnography and multiple sleep latency testing (MSLT). The diagnosis of type 1 narcolepsy is confirmed with a cerebral spinal fluid determination of a hypocretin concentration of <110 pg/mL, whereas in type 2 narcolepsy, there is an absence of cataplexy and hypocretin levels are normal.¹⁷ However, the cerebrospinal fluid hypocretin test is not widely available. The incidence of narcolepsy is reported to be 1/100,000, with an equal distribution between type 1 and type 2. It also can vary by the ethnicity of the population with Asians having a lower incidence. The diagnosis is often delayed, often for years, making precise incidence calculations and identification of secular trends problematic. The peak age at onset is 15 years, with the syndrome highly correlated with presence of the HLA DQB1*0602 haplotype. Over 90% of the subjects with narcolepsy-cataplexy possess this haplotype.

NARCOLEPSY AND VACCINATION

After the announcement of the association of the Pandemrix vaccine with narcolepsy from the Swedish Medical Products Agency, the Finnish National Institute for Health and Welfare launched a retrospective cohort study. That study determined that the relative risk for narcolepsy after Pandemrix in children was 12.7 with a 95% confidence interval of 6.1–30.⁸

Recently, Sarkanen et al summarized the various narcolepsy studies that were conducted in Europe.¹³ The overall assessment from the combined studies included in that systematic review indicated that an elevated risk was only associated with Pandemrix and that the attributable risk for narcolepsy was 1 in 18,400 doses. Interestingly, the risk of narcolepsy appeared reduced when wild-type infection was temporally further removed from vaccine administration, raising questions surrounding the potential of an interaction between wild-type influenza infection and vaccination.

The Centers for Disease Control and Prevention-funded Systematic Observational Method for Narcolepsy and Influenza Immunization Assessment (SOMNIA) study was conducted in 13 different study sites in 9 countries.¹⁸ The rationale of that study was to evaluate the risk of narcolepsy following both AS03- and MF59-adjuvanted pandemic influenza vaccines and to assess for narcolepsy in other locales outside of Europe. SOMNIA used electronic health databases to conduct a retrospective cohort study to assess narcolepsy incidence rates before and during natural 2009 pH1N1 virus circulation, and after pH1N1 vaccination campaigns in Canada, Denmark, Spain, Sweden, Taiwan, the Netherlands and the United Kingdom. The study also used a case-control study design to evaluate the risk of narcolepsy following AS03- and MF59-adjuvanted pH1N1 vaccines in Argentina, Canada, Spain, Switzerland, Taiwan and the Netherlands with the primary case definition based on a confirmatory MSLT with cases identified through sleep centers at the study sites. A total of 360 cases were identified, 150 in children and 210 in adults. In the case-control study, there was no increased risk of narcolepsy observed for Pandemrix, Arepanrix or Focetria but the ability to assess risk following Pandemrix was very limited because of the low level of Pandemrix usage at the selected study sites. Comparing incidence rates of narcolepsy before and after administration of any of the adjuvanted vaccines, there was no overall change in incidence of narcolepsy after vaccination, except in Sweden. Sweden was included in the SOMNIA study as a signaling country and thus a known positive comparator. Of note, in Canada, a single analysis in 1 small study in Quebec demonstrated low level of increased risk following Arepanrix,¹² but this was not demonstrated in other studies, despite widespread use of the AS03-adjuvanted Arepanrix vaccine. Interestingly, in Taiwan there was an increased rate of narcolepsy following the circulation of the wild-type H1N1 virus in both children and adults, but no increased risk following vaccination.¹⁹

WHAT IS THE POTENTIAL PATHOGENESIS?

An understanding of the pathogenesis of narcolepsy has been sought, particularly as it relates to pandemic influenza vaccines. What triggers the damage to the hypocretin neurons? It is often stated that narcolepsy is an autoimmune disease and that the damage to the neurons is immune-mediated. This is largely based on the association of the disease with the HLA-DQB1*06:02 haplotype. However, there are limited data available on the role of cellular immunity in the precipitation of narcolepsy. Similarly, the search for narcolepsy autoantibodies has not been productive, with no autoantibodies consistently found.²⁰ In addition, studies in pregnant women with narcolepsy have not demonstrated maternal-fetal transfer of pathogenic autoantibodies and their infants have not been born with narcolepsy symptoms.²¹

This discussion is further complicated by the observation that narcolepsy has been associated with both administration of adjuvanted vaccine and with wild-type influenza infection. A case for the role of molecular mimicry in the pathogenesis of narcolepsy has recently been proposed.²² If these results are confirmed, it could explain the association of narcolepsy with both wild-type infection and vaccination. However, questions remain. Is this damage vaccine enhanced, wild-type virus related, or is it a combination of both vaccine and influenza infection? These questions remain largely unanswered and are critical for formulating public health policy in future pandemics.

As noted above, there is a strong association of narcolepsy with HLA-DQB1*06:02, but these same HLA alleles are also associated with vigorous responses to influenza vaccines. Conversely, HLA-DQB1*06:03 appeared to be associated with a protection against narcolepsy and poor responsiveness to influenza vaccines.²³ Hence, some have suggested the role of wild-type virus infection in the related pathology.

It is known that influenza A viruses, especially H1N1 strains, can infect the olfactory receptor neurons and slowly move to the olfactory bulb with occasional anterograde axonal transport into the brain.²⁴ Hypocretin producing neurons have projections in the olfactory bulb with fibers mainly seen in the anterior olfactory nuclei, in the glomerular layer and in the internal granular layer.²⁵ In immunodeficient mice, experimental intranasal administration of wild-type H1N1 was shown to infect the olfactory bulb and the hypocretin-producing neurons in the lateral hypothalamus.²⁶ In addition, it has also been demonstrated that the transfer of hemagglutinin-specific CD8 T cells (but not CD4 T cells) to transgenic mice with hypocretin producing neurons expressing influenza hemagglutinin, led to the destruction of these specific neurons causing manifestations similar to human narcolepsy.²⁷

Based upon these studies, a “double hit” hypothesis has been proposed to explain the disparate risk data following receipt of AS03 vaccines. It is proposed that in some patients infected by wild-type H1N1/09 influenza, viruses may migrate through the olfactory pathway to the hypothalamus and infect hypocretin producing neurons. By itself this may cause some neuronal damage, likely amplified by natural CD8 responses to viral antigens. However, the administration of a strongly adjuvanted influenza vaccine concomitantly or soon after infection could markedly amplify the CD8 response and its pathogenic effects (Figure 1). This hypothesis would provide a possible explanation for the differences in the observed association between narcolepsy and AS03-adjuvanted H1N1 vaccine in Scandinavia, where wild-type H1N1 infection circulated during vaccination, and its absence in Canada, where circulation of wild-type virus occurred at a time distant from the vaccination program. However, pending further study, this remains only a hypothesis.

FIGURE 1.: The two hit hypothesis—a possible mechanism for an enhanced risk for narcolepsy following 2009 H1N1 pandemic influenza vaccine.

EPIDEMIOLOGIC STUDIES SUPPORT A POTENTIAL ROLE OF WILD-TYPE VIRUS IN THE CAUSATION OF NARCOLEPSY

Studies by Han et al in Beijing have demonstrated a seasonal distribution of narcolepsy cases with increased rates of narcolepsy being reported following other influenza epidemics and with a larger number of cases reported after the 2009 H1N1 pandemic.²⁸ More recently, the Taiwan Centers for Disease Control investigated the relationship between narcolepsy and wild-type H1N1 exposure using patients recruited from sleep centers following the 2009 pandemic.¹⁹ In their assessment, they identified 64 cases that were HLA DQB1*0602 positive, met the MSLT criteria for narcolepsy and were clinically characterized as having cataplexy. They noted a significant increase in narcolepsy occurrence following circulation of the wild-type H1N1 virus (but not following vaccination). However, in the United States no association has been found. For example, population-based studies conducted in the Vaccine Safety Datalink in the United States did not demonstrate such an association.²⁹ However, it has been proposed that a potential reason for this lack of association with wild-type virus in the United States is that in the United States there is a long lag time between onset of symptoms and the diagnostic MSLT test, often many months or years. Whereas in China and Taiwan this interval is very short, usually just a few weeks to months. Longer lag times make it much more difficult to detect a temporal association between wild-type influenza circulation and narcolepsy and may explain differences in the observed results among the different countries.³⁰

During the 2009 pandemic with the delay in the availability vaccines, there was considerable circulation of the pandemic virus circulation in some countries before the vaccine was introduced. Thus, it was extremely difficult to sort out whether many of the immunized subjects had previously been infected with the wild-type H1N1 pandemic strain, and particularly in those individuals who ultimately developed narcolepsy. To address this issue, the Finland National Institute of Health and Welfare assessed whether patients who fell ill with narcolepsy after vaccination with AS03-adjuvanted Pandemrix had specific antibody responses to nonstructural protein 1 from the H1N1pdm09 virus, which was not a component of Pandemrix vaccine.³¹ Serum specimens from 45 narcoleptic patients were collected during 2011, 2 years after circulation of the first wave of the pandemic. Paired serum specimens from 28 adults suffering from a clinical, laboratory-confirmed H1N1pdm09 virus infection were collected acutely and convalescent samples were obtained 14–21 days later. The study used quantitative Western blot analysis and only 2 of the 45 (4.4%) Pandemrix-vaccinated narcoleptic patients showed specific antibody responses against the nonstructural protein 1 from the H1N1pdm09 virus 2 years after the exposure. The paired serum samples from patients who suffered from a laboratory confirmed H1N1pdm09 infection, showed high levels of antibody against the nonstructural protein 1 in the convalescent samples obtained soon after the infection. In contrast, another study using published and unpublished H1N1pdm seroepidemiologic data reported that 47% of children 5- to 19-years old from 19 different countries were infected with wild-type virus during the pandemic.³² Also in Norway, using a mathematical model of A/H1N1p transmission, developed by calibrating the projected number of symptomatic A/H1N1p cases to the number of laboratory-confirmed A/H1N1p cases reported to the surveillance system, over half of school children were projected to be infected before vaccination based upon serum samples collected before vaccination.³³ Hence, given our current knowledge, it is difficult to definitively exclude an interaction between wild-type influenza virus infection and vaccination in the pathogenesis of narcolepsy and to prove or disprove the double hit hypothesis.

SUMMARY AND POSSIBLE NEXT STEPS

Thus, while the association between the receipt of Pandemrix and the development of narcolepsy is real, other studies with similar adjuvanted vaccines have not demonstrated an increased risk. The reasons for these differences are poorly understood. These differences may be related to the timing of the pandemic curve in relation to the receipt of vaccine or to other unidentified factors. The data from both Taiwan and China point to an association between wild-type H1N1 virus and narcolepsy, which needs to be further explored. If the wild-type influenza virus infection is confirmed in other countries as a risk factor for narcolepsy, studies undertaken to determine the mechanism for this association could further our understanding of the role of the influenza immune response to wild-type infection and vaccines in causation of narcolepsy in general.^19,28 Such studies could include studying the CD4 and CD8 responses and dissecting the genetic factors involved.

The “double-hit hypothesis,” meaning natural infection closely followed by vaccination, should be further explored in both animal models and human studies. Findings of such studies are essential to inform the optimal public health response during the next pandemic. If the double hit hypothesis is confirmed, adjuvanted vaccines might be withheld in populations especially susceptible to narcolepsy, such as adolescents and young adults, during the period following wild-type virus circulation, particularly when this virus is causing a relatively mild disease.

Finally, one of the lessons learned from the 2009 pandemic was that there is a need for an internationally coordinated vaccine safety infrastructure so that the safety of pandemic vaccines, which are rapidly deployed in large numbers, can be rapidly evaluated. We know another pandemic is coming, but we do not know when. Hence, the time to expand our understanding of the risk of narcolepsy following adjuvanted pandemic vaccines is now.