Sero-epidemiological studies of 2009 influenza A(H1N1) pandemic – A WHO Review and a subsequent publication from New ZealandArchived

Sero-epidemiological studies of 2009 influenza A(H1N1) pandemic – A WHO Review and a subsequent publication from New Zealand 

Seroepidemiological studies of pandemic influenza A(H1N1) 2009 virus Weekly Epidemiological Record (WER); 11 June 2010, No. 24, 2010, 85, 229–236.

This review published by the WHO aims at answering important questions among the Known Unknowns areas which were unclear at the beginning of the 2009 pandemic but which are now being answered with relevant analyses data, particularly regarding the infectivity and epidemiology of the pandemic.(1,2)  The review notes how that under pre-existing surveillance systems  the incidence of infection for many countries was calculated using data that captured only the number of people seeking medical care for illness, while those with asymptomatic or mild infections were uncounted. This led to an underestimation of the actual number of infections and an over-estimation of severity.  Influenza-specific antibodies are regarded as markers of past infection, and the detection of significant antibody titres against a specific strain of the virus may also reflect partial or complete protection against that virus, thus, measurements of antibody in serum samples obtained from populations before and after circulation of the 2009 pandemic influenza A(H1N1) virus can be used to estimate the extent of infection in the population, especially when they are not complicated by immunisation. The review’s authors examined the methods and results of 9 studies with seroprevalence findings relating to the 2009 pandemic influenza A(H1N1) virus that had been published in peer-reviewed journals.  In addition another study has been subsequently published by the authorities in New Zealand too late to be included in the review.(3)

The review’s studies come from 7 countries: China, the Chinese Province of Taiwan, Finland, Germany, Singapore, the United Kingdom and the United States of America. There were more than one study from most countries but there are detailed Tables giving the results in the review. The studies were mostly opportunistic using convenience samples of sera. The analyses were designed to determine the prevalence of cross-reactive antibodies to the 2009 pandemic virus prior to its appearance and to estimate the proportion of the population that was infected. The result of the baseline surveys of samples collected before the start of the 2009 pandemic provided evidence that the proportion of individuals with pre-existing antibodies that cross-reacted with 2009 pandemic influenza A(H1N1) viruses increased with age. Two studies performed in Singapore found no individuals - or only a low proportion of individuals - with prior cross-reactive antibodies whereas four other studies (from Finland, Germany, the United Kingdom and the United States) reported a relatively high proportion of individuals aged >60 years with pre-existing cross-reactive antibodies when compared with younger age groups.

In only one geographical area of study (from cohort studies in Singapore) were samples taken during the acute phase of illness and during convalescence.  In other areas (Germany, the UK and the US) studies compared the proportion of the population that was seropositive for the 2009 pandemic influenza A(H1N1) virus before and after virus circulation to estimate the proportion that became infected during the pandemic. However some of the studies were undertaken before the autumn/winter wave had run its course in the Northern Hemisphere.  Finally, the review notes that a number of the studies reported seropositivity rates after the 2009 pandemic influenza A(H1N1) virus had been circulating but did not strictly have baseline comparisons.

A finding common to all the studies and this was that young children and young adults had the highest rates of seropositivity to the 2009 pandemic influenza A(H1N1) virus after the entry of the virus in the community, or the greatest increase in seropositivity rates compared with baseline.

The serological findings in the New Zealand report is probably the first national representative seroprevalence cross-sectional study of the 2009 pandemic in the general population following a full pandemic wave. (3) That survey was undertaken to evaluate  immunity and incidence of infection in populations, identification of protective or risk factors (including the groups at higher risk) and the provision of evidence for decisions on effective vaccination and other public health interventions. The authors collected 1696 serum samples and individual risk factor data by questionnaire between the months of November 2009 and March 2010. The sample (individuals >1 year of age) was taken randomly from selected general practices across the country and hospitals from just one region (Auckland). To establish the baseline immunity necessary for the study, 521 pre-pandemic sera samples were collected from the previous 5 years. All the samples were tested for hemagglutination inhibition (HI) antibody to the 2009 pandemic influenza A(H1N1) virus.  The titre threshold for seroprotection or, in other words, possession of 2009 pandemic immunity was taken to be 1 in 40.  The overall community seroprevalence for the study population was 26.7% but varied with age and ethnicity; for instance, children aged 5-19 years had the highest seroprevalence (46.7%), significantly higher than their baseline immunity of 14%; they were closely followed by pre-school children aged 1-4 years, which seroprevalence was 29.5%, also high when compared to their age-specific baseline immunity of 6%; finally, the seroprevalence and the baseline immunity observed in people aged 60 or more were of similar value around 24%. Among ethnic-minority groups, the Pacific Peoples had the highest seroprevalence (49.5%) followed by the Maori (36.3%). It was noted that these two ethnic minorities also had much higher hospitalisation and intensive care unit admission rates than European and other groups.  The report notes that seroprevalence among neither primary care nor secondary care HCWs differed from community participants and was typically around 25-30%. Age was found to be the most important risk factor followed by ethnicity and a history of previous seasonal influenza vaccination – demonstrated by multivariate analysis techniques. The likelihood of immunity to 2009 pandemic influenza among the age group 5-19 and age group 1-4 was, respectively, 5.3 times and 3.5 times higher compared to that of the age group 40-59. Similarly the likelihood of immunity to pandemic influenza was 2.2 times higher among the Pacific People compared to that of the other groups. In relation to previous seasonal influenza immunization, the study showed that participants with a history of any past seasonal influenza vaccination were 1.8 times more likely to have seroprotective immunity levels compared with the participants who had never got vaccinated before; in addition, this protective effect was more evident in the age groups 1-4 and 5-19 years, but the authors stress that these levels would be insufficient to provide individual effective protection against 2009 pandemic influenza A(H1N1) among individuals belonging to these groups under 60 years, which could only be provided after effective vaccination with the 2009 pandemic vaccine.  Based on the questionnaire with the survey, a high percentage of the seropositive individuals (45.2%) had no symptoms, giving an indication of a relatively ‘silent’ spread of the disease, which the authors note would have major implications when implementing countermeasures to limit the spread of the disease in further 2009 pandemic waves.(3)  Based on these findings the authors estimate a hospitalization rate of 262 per 100,000 estimated symptomatic cases and a case-fatality rate of 8.2 per 100,000 estimated symptomatic cases. Presumably the case fatality rate would be about half this for all infections.  The authors conclude that the 2009 pandemic influenza A(H1N1) virus was highly infectious resulting in substantial proportions of both symptomatic and asymptomatic infections, and based on age and ethnicity standardisation to the national population, by March 2010, an estimated 29.5% of the population of New Zealand (1.3 million people) had immunity to the 2009 pandemic influenza A(H1N1) and an estimated 18% (800,000 people) were infected with the virus during the first wave - including an approximate 33% of all the children in New Zealand.(3)  Subsequently the New Zealand authorities have used the results to inform the 2010 immunisation campaign emphasising the need to protect vulnerable people.

ECDC Comment

For large outbreaks, epidemics and pandemics it is impossible and undesirable to confirm most cases by laboratory diagnosis.  Seroepidemiological studies like these allow for refinement of estimates of the numbers of people who are at risk of infection, obtain hospitalization and case fatality rate estimates and inform policy on vaccination needs. On the basis of the current literature of the seroepidemiological studies of the 2009 influenza A(H1N1) pandemic, some consistent observations emerge, although there are limitations in the data, and a general lack of standardisation of tests deployed which are also labour-intensive and so expensive.

One of the limitations of the studies is the heterogeneity in populations and in most of the studies children were under-represented. Exceptions are the New Zealand and UK studies.(3,4)  Most studies that looked at the seroprevalence prior to pandemic, found that the cross-protective immunity from previous infections or vaccinations increases with age, being highest in the >60 age group, a finding noted early in the pandemic (2,5). The studies with data from after the pandemic, consistently found that the seroconversion and infection rates were highest in school-aged children and teenagers (5-19 years olds) . In the New Zealand study there were no differences between regions but specific population groups based on the ethnic background found that the two recognised ethnic minority groups Pacific and Maori peoples had highest seroprevalence and higher hospitalisation or intensive care unit admission rates in comparison to general population.  In contrast the UK study, which was undertaken before its autumn/winter wave found a marked regional difference with seroprevalence being highest in the London and the West Midland regions. (4)

There are some differences in the findings for heath care workers.  The New Zealand study did not find any higher seroconversion rates in the health-care workers than in the general population . In the study conducted in Singapore, where hospital staff and long-term care facilities cohorts were used it was found  that these cohorts had lower infection rates than the general population.(6) However, the vaccination background of the different HCW cohorts is not known.(1) Furthermore, in Singapore, only 13% of the community cohort seroconverted in comparison to the 26.7% in New Zealand.(3,6)  In contrast a study in Taiwan found indications of higher prevalence of infection in HCWs with exposure to patients than in other HCWs or the general public.(7) 

The 2009 pandemic influenza A(H1N1) virus infection resulted in substantial proportions of both symptomatic and asymptomatic infections. The high proportion of asymptomatic infections has been seen in both the New Zealand and the Singapore studies that used a questionnaire to assess the symptoms of the blood donors during the testing period. Based on the surveys, about half the participants who seroconverted reported no respiratory illness or febrile episode. Taken together with the UK data it seems that approximately 30% of the children were infected in various countries and different continents.(1,3,4) The adult and elderly populations have had cross-reactive antibodies from previous vaccinations and infections that have protected them from illness. It seems therefore that  school-aged children played an especially crucial role in the transmission of influenza in the 2009 pandemic. (1,3,4)

The major limitations of comparative analyses of these studies are the differences in study populations and timing of sampling related to the pandemic. Lack of standardization of the haemagglutination and microneutralization assays is one issue together with different cut-off values for positivity. Both assays may also give cross-reactive results with other H1N1 subtype antibodies. All the published studies use slightly different settings of the assay as well as reference sera and virus strains. Also many of the baseline seropositivity estimates are based on blood bank or other convenience samples that may not represent the general population, especially children. Some of these issues are being addressed by use of standard panels of sera.

Based on these reports, it is possible to conclude some trends in the seroprevalence of the 2009 pandemic influenza A(H1N1) virus in certain countries and regions but not a global attack rate of the pandemic. To estimate a global attack rate, if such a concept is desirable, would need studies from all over the world covering the majority of the population and ethnic groups of the globe.  There are more than twenty seroepidemiological studies under way and those will provide valuable information on the true attack rates of the pandemic virus. A number of these studies are underway in Europe and were discussed, along with their limitations and the research and development needs  at the annual EISN meeting in Sofia, Bulgaria in June. The standardization and improvement of the laboratory methods to detect subtype specific antibodies without cross-reactivity and to allow comparison of the results between the laboratories worldwide, is crucial and will benefit the whole influenza laboratory network. ECDC and WHO are taking measures together with the laboratories in the Community Network Reference Laboratory to provide guidance and training as well as sharing of viruses and standards to achieve best possible standardization of the methods.

These seroprevalence studies support vaccination of specific population groups such as children and certain ethnic groups or people with underlying conditions that are more in the risk of disease and severe outcomes. The vaccination of military camps is also justified based on the studies as they show high transmission of the virus in garrisons (6). Whether or not HCWs are at higher risk is unclear.(1,7) However, their influenza  vaccination is important  to protect patients, reduce occupational risk and minimize the risk for absenteeism. The latter is also true for those in laboratory work.  The slow progress of these studies is unfortunate.  There is a need to improve the preparedness for seroprevalence studies in general, so that the studies could be launched, conducted and the results communicated faster in future outbreaks and pandemics.  Many laboratories that would have had the capacity for seroprevalence studies during the pandemic, were hampered by the regulatory processes of the clinical studies and by the intense work-load from the pandemic.  This should be overcome by preparation of study plans and approvals for upcoming pandemics simplification and automation of tests and pre-planning of alternative laboratory sites . The current labour-intensive methods should eventually be replaced by high-through-put technologies and improvements in quality assurance. The better understanding of antigenic epitopes of influenza virus will also contribute to development of more specific tests and use of modelling tools in the prediction of protection by specific antigenic properties of the circulating virus. Finely more timely sharing of the analyses and data should be a fundamental principle in future work in Europe and elsewhere.(2)

References:

(1) WHO Seroepidemiological studies of pandemic influenza A(H1N1) 2009 virus.  Weekly Epidemiological Record (WER); 11 June  2010, (24), 2010, 85, 229–236.

(2) Nicoll A, Ammon A, Amato Gauci A, Ciancio B, Zucs P, Devaux I, Plata F, Mazick A Mølbak K , Asikainen T, Kramarz P. Experience and lessons from surveillance and studies of the 2009 pandemic in Europe. Public Health 2010 124:14–23. Available here

(3) Bandaranayake D, Huang S, Bissielo A and Wood T. Seroprevalence of the 2009 influenza A(H1N1) pandemic in New Zealand. Ministry of Health, Client report FW10057, http://www.moh.govt.nz/moh.nsf/pagesmh/10124/$File/seroprevalence-flu-2009.pdf

(4) Miller E, Hoschler K, Hardelid P, Stanford E, Andrews N, Zambon M. Incidence of 2009 pandemic influenza A H1N1 infection in England: a cross-sectional serological study. Lancet. 2010 Mar 27;375(9720):1100-8. Epub 2010 Jan 21.

(5) US Centers for Disease Control and Prevention. CDC Serum Cross-Reactive Antibody Response to a Novel Influenza A (H1N1) Virus after vaccination with Seasonal Influenza Vaccine. MMWR May 22, 2009/58(19);521–524.

(6) Chen M, Lee V, Lim W et al. 2009 influenza A(H1N1) seroconversion rates and risk factors among distinct adult cohorts in Singapore. JAMA; 303: 1383-91.

(7) Chan YJ, Lee CL, Hwang SJ, Fung CP, Wang FD, Yen DH, Tsai CH, Chen YM, Lee SD. Seroprevalence of antibodies to pandemic (H1N1) 2009 influenza virus among hospital staff in a medical center in Taiwan. J Chin Med Assoc. 2010 Feb;73(2):62-6