Burden of influenza B virus infections in Scotland in 2012 / 13 and epidemiological investigations between 2000 and 2012

H Harvala (heli.simmonds@hotmail.com)1, D Smith2, K Salvatierra2, R Gunson3, B von Wissmann4, A Reynolds4, C Frew3, A MacLean3, A Hunt5, D Yirrell6, P Simmonds2, J McMenamin4, K Templeton1 1. Specialist Virology Centre, Department of Laboratory Medicine, Royal Infirmary Edinburgh, Edinburgh, United Kingdom 2. Centre for Immunology, Infection and Evolution, University of Edinburgh, Edinburgh, United Kingdom 3. West of Scotland Specialist Virology Centre, Glasgow, United Kingdom 4. Health Protection Scotland, Glasgow, United Kingdom 5. Department of Medical Microbiology, Aberdeen Royal Infirmary, Aberdeen, United Kingdom 6. Department of Medical Microbiology, Ninewells Hospital and Medical School, Dundee, United Kingdom


Introduction
Both influenza A and B viruses show antigenic variation with sequence variability in the haemagglutinin (H) and neuraminidase (N) genes, altering neutralising properties and protection from reinfection [1].However, they differ in the dynamics of turnover of antigenic variants and in their propensity for reassortment [2].Variants of influenza A viruses are classified into subtypes based on the H and N genes; two subtypes (H1N1 and H3N2) are currently found in humans in Europe.Although influenza B viruses are not formally classified into subtypes, two antigenically distinct lineages defined by the reference strains B/Victoria/2/87 (Victoria lineage) and B/Yamagata/16/88 (Yamagata lineage) have been circulating globally since 1983 [3].Influenza B viruses of the Victoria and Yamagata lineages have been shown to circulate simultaneously or individually in the past and switching between lineages occurs almost every other year [2].In contrast, antigenic shift of influenza A, characterised by changes in the circulating H and N types occurs less frequently.Because several influenza variants are present in the same season, influenza vaccines currently in use in the United Kingdom (UK) and elsewhere in Europe are trivalent, containing two influenza A subtypes (H1N1 and H3N2) and one influenza B lineage.However, it provides only limited immunity against the influenza B strains of the other lineage [4].
In order to improve our understanding of the burden and epidemiology of influenza B, we have reviewed the influenza B viruses circulating in Scotland over the period 2000 to 2013 and estimated the clinical burden of influenza B virus infections in more detail for the season 2012/13.

Routine laboratory testing
In Scotland, diagnostic testing by reverse transcription-polymerase chain reaction (RT-PCR) for influenza B virus forms part of multiplex real-time screening for a range of respiratory viruses.PCR-based influenza B virus screening has been carried out in National Health Service (NHS) laboratories in Glasgow since early 2000 [5], in Edinburgh since 2005 [6], in Dundee since 2007 and in Aberdeen since 2008.Thus Figure 1 from 2000 to 2005 is based on Glasgow results only.In addition to individuals presenting to hospital with presumed respiratory virus infection, a sample of those presenting in the community is tested for respiratory viruses including influenza B virus via the Scottish sentinel surveillance scheme [7].

Prospective data collection
All reports of influenza B virus detection from NHS laboratories in Scotland are collated centrally by the national public health body Health Protection Scotland (HPS) via the Electronic Communication of Surveillance in Scotland (ECOSS) non-mandatory reporting system.The population of Scotland was estimated to be 5,295,400 in 2011 [8].
General practitioners (GP) participating in the Scottish sentinel surveillance scheme are requested to systematically submit nose and throat swabs from the first five patients per week consulting with influenza-like illness or acute respiratory tract infection (based on clinical judgement) [7].Supplementary data on demography, clinical signs and risk group as well as vaccination status are collected for each patient.Since season 2000/01 between 18 and 89 GPs per season have participated in the scheme, submitting between 600 and 5,000 samples per season.Furthermore, the Scottish Severe Acute Respiratory Illness (SARI) surveillance system collects individuallevel data on confirmed influenza cases admitted to an intensive care unit (ICU) during the influenza season from all 30 hospitals with intensive care provision across the 14 NHS boards.

Sequencing
Nucleotide sequences of the influenza B haemagglutinin gene were obtained for each influenza season from a representative subset of archived nucleic acid samples derived from anonymised respiratory samples from patients with confirmed influenza B (collected since 2007).The haemagglutinin gene was amplified using hemi-nested primers (sense: TACTACTCATGGTAGTAACATCC; outer antisense: TTTCGTTRTGSAGTTCATCCATSGC; inner antisense: AATCATK CCTTCCCAKCCKCCT).Nucleotide sequences were aligned with reference sequences using SSE v1.1 [9] and phylogenetic analysis was performed using Mega 5.1 [10].In addition, influenza B virus-positive samples tested in Glasgow in 2012/13 were typed using lineage-specific real-time PCR [11], and sequences from 39 previously characterised influenza B virus strains from England were obtained from the database of the Global Initiative on Sharing All Influenza Data (GISAID; http://platform.gisaid.org/epi3/frontend#48fe7c).The authors gratefully acknowledge the originating and submitting laboratories who contributed these sequences to GISAID, in particular Public Health England (formerly Health Protection Agency) and the National Institute for Medical Research in the UK (n=37 sequences) and the National Institute for Medical Research and the Centers for Disease Control and Prevention in the United States (US) (n=2 sequences).

Epidemiology of influenza B virus during 2000 to 2013
During the period January 2000 to April 2013, several distinct waves of influenza A and B viruses infections were observed in Scotland (Figure 1

Burden of influenza B infections in the 2012/13 season
More detailed information on influenza virus incidence and clinical outcomes was available for the season 2012/13.A total of 22,015 respiratory samples were tested in Scotland for influenza A and B viruses between November 2012 and April 2013.The overall detection frequency of influenza viruses was 14.4%, varying from 2.3% in November to 20.8% in February.Influenza B virus infections accounted for 44.4% of all influenza detections (1,405/3,163), exceeding the number of influenza A infections in February (Figure 2).

Community versus hospital
Compared with 1,405 laboratory detections of influenza B virus in the 2012/13 season, 1,279 episodes of influenza B virus infection were reported to HPS during the same time period, showing the robustness of automated reporting (data not shown).The discrepancy probably reflects individuals who tested positive for influenza B more than once.Of the 1,279 reported infections, 394 related to samples collected by GPs and represented influenza B infections in the community, whereas the remaining 885 influenza B-positive samples were obtained during hospital visits (Figure 2).
The community-based influenza B infections will not be discussed further; the remaining analysis focuses on hospital-based influenza B infections only.

Age distribution
In the 2012/13 season, hospital-based influenza A and B infections were widely distributed between age groups (Table 1

Severity of influenza B
In 2012-13, a total of 37 influenza B virus-infected individuals admitted to ICU were reported through the Scottish SARI surveillance system; three of them died.In comparison, 21 H1N1 and 60 H3N2 influenza A virusinfected individuals were admitted to ICU, of whom nine and 22 died, respectively.Most patients admitted to ICU with influenza B were older than 37 years (n=20), whereas a small number (n=9) of ICU admissions were observed in those younger than 20 years (Figure 4).The three mortalities occurred in the age groups 15-19, 20-24 and >65 years.
Only 16 of 27 influenza B-infected individuals were recorded to have been treated with antiviral drugs in the ICU (data not available for the remaining 10 individuals).Of the 37 influenza B patients treated in ICU, 24 belonged to groups recommended for vaccination (65 years and older or a clinical risk group), but only six had actually received the seasonal influenza vaccination.

Discussion
This epidemiological study of influenza B virus infections in Scotland is timely in view of the recent changes in the influenza vaccination programme in the UK and the possible introduction of a tetravalent influenza vaccine containing influenza B viruses of both the Victoria and Yamagata lineage.Our study reveals that influenza B virus was the predominant overall cause of influenza in four of the 13 influenza seasons from 2000/01 through 2012/13, and cocirculation of the Yamagata and Victoria lineages was observed in two of the six seasons investigated.Similar patterns of influenza B virus circulation have been described in the US, Europe and Hong Kong [12][13].

Laboratory testing
With influenza diagnostic screening now improved through faster turnaround times and the capacity for screening larger numbers of samples in all four centres (a sevenfold increase during the past 10 years), virological data currently obtained through routine testing of samples from patients presenting to hospitals can be used effectively and directly to support influenza surveillance.April 2013.These screening results can therefore be an indicator for influenza A and B virus incidence and circulation in that age group irrespective of whether infections lead to hospital admission.However, this approach is not routinely used elsewhere in Scotland: Only 20% of samples submitted for virological screening in Glasgow were obtained from children under the age of five years.This is likely to reflect different screening policies adopted in hospitals (i.e.screening only those admitted vs everybody screened).

Influenza B, a paediatric infection?
Influenza is generally recognised as an important disease with high mortality rate among the elderly, although children have been shown to have an important role in the dissemination of influenza [14][15].In contrast to influenza A(H3N2) virus, influenza B virus predominantly infects children and young adults [16][17].In the current study, influenza B contributed to 45% of influenza-related hospital visits overall during the 2012/13 season, as much as 60% for children aged six to 15 years and 75% for young adults aged 21 to 36 years.The detection rate for influenza A dropped remarkably from those younger than five years to those aged six to 10 years, whereas only a slight drop was seen for influenza B infections (Figure 3) This is consistent with previous studies [12,18] and may be due to children accumulating natural immunity to influenza B more slowly than to influenza A [19].

Severity of influenza B infections
Data on severe influenza B virus infections and mortality are limited.A large study from the US covering the period from 1976 to 1999 reported that 25% of all influenza-related mortality could be attributed to influenza B virus.This is substantially higher than seen in this study where influenza B infections accounted for ca 9% of all influenza virus-related deaths (3/34) during the 2012/13 season.Among those admitted to ICU with influenza B infections, we saw two deaths in young adults (15 to 24 years; one without underlying health problems).Whether they and others who were admitted to ICU with influenza B virus infection were genetically predisposed for severe infection remains to be investigated further [20].

Prevention and treatment
Influenza immunisation has been recommended in the UK since the late 1960s and has been targeted to those predisposed towards greater morbidity following influenza.Although influenza vaccine uptake in Scotland during the 2012/13 season was 77% for those older than 65 years and 56% for those 65 years and younger in a risk group [21], only a quarter of those treated in ICU who should have been vaccinated actually had received the seasonal influenza vaccine.Most individuals (16/18) in ICU who had missed their vaccination were younger than 65 years but should have been vaccinated due to an underlying chronic medical condition.Whether the mechanism to target these individuals for influenza vaccinations could be improved remains to be studied.
The neuraminidase inhibitors zanamivir and oseltamivir have been shown to be effective in the treatment of influenza [22][23], and are recommended for those hospitalised with influenza.Only 60% of individuals admitted to ICU with influenza B were treated with antiviral drugs; none of those who died were reported to have received antiviral treatment.However, the effectiveness of neuraminidase inhibitors against influenza B virus has been questioned and needs to be investigated further because inhibitory concentrations in vivo differ by up to 10-fold [24].
As the disease burden of influenza B is higher in children and young adolescents than in older age groups and because of recent cocirculation of influenza B viruses of both lineages, tetravalent vaccines may need to be combined with the newly introduced vaccination programme for children to improve the control of influenza B. The national programme aims to offer annual influenza vaccination to all school children across Scotland by autumn 2015, but there is no defined time frame for the introduction of the tetravalent vaccine in this group.Our study demonstrates that influenza B virus infections are associated with substantial morbidity and that influenza surveillance and interventions including vaccination and treatment are still suboptimal.
Years when more than 30% of all influenza infections were caused by influenza B virus are shown in orange.
).The highest frequency of influenza A virus detection (over 80% typed as H3N2) was seen in individuals over 65 years of age (23.5%; 451/1,916), whereas influenza B detections were significantly underrepresented in this age group in Scotland (7.0%, 134/1,916; p<0.0001 by chi-square test of association).When comparing the age-specific incidence of Detection frequency of influenza A and B virus infections and total number of samples submitted for testing by general practitioners and hospitals, Scotland, November 2012-April 2013 (n=22,015) Virology testing in Scotland is performed in four NHS laboratories, located in Dundee, Aberdeen, Edinburgh and Glasgow.

Table 1
Age distribution of influenza A and B virus infections in individuals tested in Edinburgh and Glasgow, November 2012-April 2013 (n=16,146) [8]idences are calculated based on and census data from 2011[8].Line graphs refer to the right y-axis and show the percentage of samples tested per age group vs the percentage of each age group in the population of southern and central Scotland.

Table 2
Number of influenza B isolates grouping with the Victoria lineage (Brisbane/06) or Yamagata lineage (Florida/04 or Wisconsin/10), by season and location, Scotland, November 2007-April 2013 (n=186) Additional typing results based on lineage-specific real-time PCR.The vaccine strain used in each season is indicated.The authors acknowledge the laboratories that submitted 39 UK sequences used in this Table to GISAID: Public Health England (formerly Health Protection Agency) and the National Institute for Medical Research in the UK (n=37 sequences) and the National Institute for Medical Research and the Centers for Disease Control and Prevention in the United States (n=2 sequences). a