Circulation of Influenza Type B Lineages in Greece During 2005

Abstract

The past decades influenza B lineages Victoria and Yamagata cocirculated. Our aim was to estimate the distribution of the two lineages circulating in Greece and any possible mismatching with vaccine influenza B strains. We studied 490 laboratory-confirmed influenza B nonsevere acute respiratory infection (non-SARI) cases diagnosed in the two National Influenza Reference Laboratories by reverse transcriptase polymerase chain reaction from July 1, 2005 to June 30, 2015 and 100 influenza B SARI cases diagnosed from July 1, 2011 to June 30, 2015. Median matching between the circulating influenza B lineages and the vaccine influenza B strains was 19.2% (range: 0–100%) for non-SARI cases during 2005–2015 and 67.6% (range: 41.2–94.1%) for SARI cases during 2011–2015. In two influenza seasons (2005–2006 and 2006–2007), complete lineage mismatch between influenza B non-SARI cases and influenza B vaccine strains was found. We estimated that 5, 12, or 16 laboratory-confirmed SARI cases could have been prevented by quadrivalent influenza inactivated vaccine (QIV) during the 2011–2012 season and 1, 2, or 3 SARI cases during the 2014–2015 season, with a vaccination coverage rate of 70% and a vaccine effectiveness of 20%, 50%, or 70%, respectively. Significant cocirculation of Victoria and Yamagata B strains and mismatching with vaccine influenza B strains were found during 2005–2015 in Greece. The wide use of a QIV instead of a TIV will confer additional immunity and therefore protection against influenza B, and it is expected to prevent several SARI cases annually. Our findings strongly support the recommendations for using QIV.

Introduction

Influenza is a significant cause of morbidity, health care seeking, hospitalization, and mortality globally. A recent study of the European Centre for Disease Prevention and Control (ECDC) showed that influenza is responsible for 30% of the total burden of 31 infectious diseases on the health of Europeans; this is attributed to the increased incidence and early mortality associated with influenza (6). Influenza is intrinsically characterized by the unpredictable predominance of one or cocirculation of several influenza A subtypes and/or influenza type B (hereafter referred to as influenza B) lineages in any influenza season; therefore, the influenza burden may differ substantially from year to year and potentially fluctuate from age group to age group and from region to region (2,5,7,11,14). Two antigenically distinct lineages of influenza B viruses have circulated globally since 1985 and have cocirculated since 2001, namely lineage Victoria and lineage Yamagata (3,5,17).

In 2012, the first quadrivalent influenza inactivated vaccine (QIV) was licensed in the United States and in 2017 in Greece. In addition to the two A subtypes (H1N1 and H3N2), QIVs include two influenza B strains, lineage Victoria and lineage Yamagata. The second influenza B strain was added to mitigate the risk linked to influenza B-lineage mismatch or cocirculation, which leads to suboptimal vaccine effectiveness and protection (3). A model-based study showed that a universal vaccination program with QIV could prevent a mean of ∼30,000 influenza cases, 3,500 hospitalizations, and 700 deaths associated with influenza in the United States each year (7). There are few published data about the potential impact of QIV, and no data from Greece (8,15). The aim of this study was to estimate the impact of influenza B in Greece during 2005–2015 based on nationwide virological data.

Materials and Methods

Objectives

The primary objective was to describe the seasonal distribution of influenza B lineages Victoria and Yamagata among cases diagnosed in the National Influenza Reference Laboratories in Greece and to estimate the mismatch between the circulating influenza B strains and the vaccine strains from the 2005–2006 influenza season through the 2014–2015 influenza seasons (10 seasons). The secondary objective was the estimation of the distribution of influenza B lineages Victoria and Yamagata and the percentage of lineage mismatch between the circulating influenza B strains and the vaccine influenza B strains in severe acute respiratory infection (SARI) cases in Greece from the 2011–2012 through the 2014–2015 influenza season (four seasons). An additional secondary objective was the estimation of the potentially prevented SARI cases by use of QIV.

Influenza virological surveillance in Greece

Nasopharyngeal swab samples collected from outpatient or inpatient patients with influenza-like illness (ILI) are sent either to the National Influenza Reference Laboratory of Southern Greece in Athens or to the National Influenza Reference Laboratory of Northern Greece in Thessaloniki (covering a population of ∼7 and 3.8 million people, respectively) (10). Starting from the 2009 influenza A/HN1N1pdm09 pandemic, samples from SARI cases are also tested for influenza. Samples are sent to the laboratories along with a standardized form with the following information: age, sex, region of residence, underlying diseases, date of sample collection, vaccination status, health care facility where sample was collected, admission in intensive care unit, and patient outcome. The samples are tested for influenza by real-time reverse transcriptase polymerase chain reaction (RT-PCR) within 1–2 days. After the detection of influenza viruses, RNA is stored in deep freeze, while the clinical samples are also stored.

Study population

Influenza B cases diagnosed by real-time RT-PCR in the two National Influenza Reference Laboratories from the 2005–2006 influenza season to the 2014–2015 influenza season.

Sample size estimation

On average ∼145 influenza B strains per season are diagnosed in the National Influenza Reference Laboratory for Southern Greece and ∼35 influenza B strains per season in the National Influenza Reference Laboratory for Northern Greece. Therefore, 1,800 influenza B strains from 2005–2006 through 2014–2015 were expected to be in the national virological surveillance repository. A sample of 49 influenza B cases per influenza season will provide, for a proportion of 0.5, a 95% confidence interval (CI) ranging from 0.38 to 0.62 using the normal approximation interval for a binomial proportion, with finite population correction applied to the formula. All influenza B cases from both National Influenza Reference Laboratories were merged, and 49 cases per influenza season were selected by simple random selection (excel tool). For the secondary objective, 50% of a total of 200 stored samples from influenza B SARI cases from 2011–2012 through 2014–2015 were selected by simple random selection (excel tool). A total of 590 influenza B strains isolated during the 10 studied seasons, of which 100 in patients admitted in an intensive care unit (ICU), was randomly selected, stratified by NUTS1 geographical region (9). In case too few strains were available for some influenza seasons, due to low circulation of influenza B, proportionally more strains were selected from the other seasons.

Statistical analysis

Demographic characteristics of patients selected are summarized in contingency tables, and compared with the Mann–Whitney test for continuous variables and Fisher's exact test for categorical variables. For each season, the proportion of trivalent influenza vaccine (TIV)-matched B strains was calculated for ICU and non-ICU cases separately, including exact binomial 95% CIs. Geographically weighted proportion of matched strains by season were additionally calculated, with 95% CIs estimated by bootstrapping. The number of influenza B SARI cases potentially avoidable by using a QIV instead of a TIV was estimated as follows: the percentage of mismatched influenza B SARI strains per influenza season was multiplied by the total number of influenza B SARI cases determined through influenza surveillance. Then, this was multiplied by different plausible estimates of influenza vaccine effectiveness (IVE) against mismatched lineage (20%, 50%, and 70% defined as low, medium, and high IVE, respectively), assuming a vaccination coverage of 70%, to estimate the number of influenza B SARI cases potentially prevented by QIV vaccination under different IVE scenarios. All analyses were performed in the R software environment (version 3.6.0) (16).

Laboratory testing

After RNA extraction, one step real-time PCR was applied in the clinical samples of the 590 selected influenza B cases for the subtyping of influenza B viruses to lineages Victoria and Yamagata, according to the World Health Organization (WHO) protocol (19). Purified PCR products were sequenced, using the same primer pair for each amplified product.

Definitions

An influenza B case was defined as any person with ILI and RT-PCR-confirmed influenza B, diagnosed in one of the two National Influenza Reference Laboratories in Greece from July 1, 2005 to June 30, 2015. A case of influenza B SARI was defined as an influenza B case admitted in ICU and/or resulting in death of the patient. Until 2013–2014, the old WHO case definition if ILI was used as follows: sudden onset of fever and a temperature >38°C and cough or sore throat in the absence of another diagnosis. From 2014–2015 onward, the new ECDC definition was used as follows: sudden onset of symptoms, and at least one of the following four systemic symptoms: fever or feverishness, malaise, headache, myalgia; and at least one of the following three respiratory symptoms: cough, sore throat, and shortness of breath.

Ethical issues

The study was approved by the scientific and ethics committees of the Hellenic Pasteur Institute and the University of Thessaloniki. Informed consent was not required, since the data were anonymous and collected as part of the routine virological surveillance.

Influenza vaccination policy in Greece

Influenza vaccination is recommended for high-risk groups, including people with chronic cardiovascular, renal, respiratory, and/or neurologic diseases, pregnant women regardless of trimester, children on aspirin treatment, persons >60 years, and health care personnel.

Results

Table 1 shows the characteristics of the 490 influenza B non-SARI cases compared with the 100 influenza B SARI cases. SARI cases were significantly older (median age of 60 years compared with 18.5 years, respectively; p-value <0.001), more frequently had an underlying disease (35% vs. 3.7%; p-value <0.001), a history of influenza vaccination (4% vs. 2.9%; p-value = 0.039), or a fatal outcome (19% vs. 0.4%; p-value <0.001) compared with non-SARI cases. No other differences were found.

Table 1.

Characteristics of Influenza B Cases (n = 590)

Characteristic	Non-SARI	SARI	p
Characteristic	n = 490 (%)	n = 100 (%)	p
Male sex	256 (52.5)	53 (53.0)	0.91
Age group, years
0–4	70 (14.3)	4 (4.0)	<0.001
5–14	153 (31.4)	6 (6.0)
15–64	218 (44.7)	51 (51.0)
≥65	47 (9.6)	39 (39.0)
Median age (IQR), years	18.5 (7–46)	60 (39.5–70.25)	<0.001
Region of residence (NUTS1)
Attica (EL3)	278 (56.7)	56 (56)	0.13
Aegean Islands, Crete (EL4)	10 (2.0)	6 (6.0)
Northern Greece (EL5)	171 (34.9)	30 (30)
Central Greece (EL6)	31 (6.3)	8 (8.0)
Underlying disease^a	18 (3.7)	35 (35)	<0.001
Influenza vaccination	14 (2.9)	4 (4)	0.039
Fatal outcome	2 (0.4)	19 (19)	<0.001

Chronic cardiovascular disease, chronic respiratory disease, chronic metabolic disease, immunosuppression, chronic neurologic disease, chronic renal disease, obesity, pregnancy.

IQR, interquartile range; NUTS1, nomenclature of territorial units for statistics; SARI, severe acute respiratory infection.

Table 2 shows the distribution of lineages Victoria and Yamagata strains in the non-SARI cases diagnosed in Greece during 2005–2015. Lineage Victoria strains were the only isolates in the 2005–2006 influenza season; lineage Yamagata strains predominated in six influenza seasons (2006–2007, 2007–2008, 2010–2011, 2012–2013, 2013–2014, and 2014–2015); and significant cocirculation was documented in two seasons (2008–2009 and 2011–2012). As expected, circulation of influenza B was negligible during 2009–2010, since the latter season was dominated by the 2009 pandemic A1/H1N1pdm09 strain. In terms of lineage matching with vaccine influenza B strains, the median lineage matching was 19.2% during the 10-year study period, ranging from 0% to 100%. A 100% lineage matching was found in two seasons only (2012–2013 and 2013–2014), while 100% lineage mismatching was documented in 2005–2006 and 2006–2007.

Table 2.

Distribution of Influenza B Lineages Victoria and Yamagata and Matching with Vaccine Influenza B Strains by Influenza Season, Greece 2005–2015

Season	Vaccine influenza B strain	Non-SARI, total	Victoria	Yamagata	Matching (%)	Matching, weighted
2005–2006	Yamagata	55	55	0	0/55, 0.0%, 95% CI 0.0–6.5%	0.0%, 95% CI 0.0–0.0%
2006–2007	Victoria	25	0	25	0/25, 0.0%, 95% CI 0.0–13.7%	0.0%, 95% CI 0.0–0.0%
2007–2008	Victoria	38	2	36	2/38, 5.3%, 95% CI 0.6–17.7%	5.7%, 95% CI 0.0–14.3%
2008–2009	Yamagata	82	67	15	15/82, 18.3%, 95% CI 7.0–20.8%	19.2%, 95% CI 10.8–28.3%
2009–2010	Victoria	2	0	2	0/2, 0.0%, 95% CI 0.0–84.2%	—
2010–2011	Victoria	71	9	62	9/71, 12.7%, 95% CI 6.0–22.7%	12.9%, 95% CI 5.7–22.0%
2011–2012	Victoria	87	47	40	47/87, 54.0%, 95% CI 43.0–64.8%	53.4%, 95% CI 43.0–63.7%
2012–2013	Yamagata	11	0	11	11/11, 100.0%, 95% CI 71.5–100.0%	100.0%, 95% CI 100.0–100.0%
2013–2014	Yamagata	31	0	31	31/31, 100.0%, 95% CI 88.8–100.0%	100.0%, 95% CI 100.0–100.0%
2014–2015	Yamagata	88	6	82	82/88, 93.2%, 95% CI 85.7–97.5%	93.2%, 95% CI 87.5–97.8%

CI, confidence interval.

Table 3 shows the distribution of lineage Victoria and Yamagata strains in SARI cases diagnosed with influenza B in Greece during 2011–2015. Lineage Victoria and Yamagata strains cocirculated in 2011–2012, while Yamagata strains predominated during the 2014–2015 season. The distribution of lineage Victoria and Yamagata strains and the lineage matching estimates of SARI cases during 2011–2015 were comparable with those of non-SARI cases for the respective season. In particular, a 41.2% lineage matching with vaccine influenza B strains was found in the 2011–2012 influenza season and 94.1% in the 2014–2015 influenza season, comparable with the lineage matching estimates in non-SARI cases for the respective influenza seasons. Estimation was not feasible for 2012–2013 and 2013–2014 due to the very small numbers of influenza B SARI cases.

Table 3.

Distribution of Influenza B Lineages Victoria and Yamagata in Severe Acute Respiratory Infection Cases and Matching with Vaccine Influenza B Strains by Influenza Season, Greece 2011–2015

Influenza season	ICU, total	Victoria	Yamagata	Matching	Matching, weighted
2011–2012	46	19	27	19/46, 41.3%, 95% CI 27.0–56.8%	41.2%, 95% CI 28.0–56.3%
2012–2013	1	0	1	1/1, 100.0%, 95% CI 2.5–100.0%	—
2013–2014	2	0	2	2/2, 100.0%, 95% CI 15.8–100.0%	—
2014–2015	51	3	48	48/51, 94.1%, 95% CI 83.8–98.8%	94.1%, 95% CI 86.5–100%

ICU, intensive care unit.

Subsequently, we estimated the number of influenza B SARI cases potentially avoidable by using a QIV instead of TIV during the 2011–2012 and the 2014–2015 influenza seasons, where important circulation of influenza B strains was detected. The proportion of lineage mismatch between the SARI strains and the vaccine influenza B strains was 58.7% (95% CIs: 43.2–73%) during 2011–2012 and 5.9% (95% CIs: 1.2–16.2%) during 2014–2015. It is estimated that during the 2011–2012 influenza season, with a 70% vaccination coverage rate among high-risk group patients, 5 SARI cases (95% CIs: 3–6 cases) could have been saved using a QIV with an IVE of 20%, 12 cases (95% CIs: 9–15 cases) using a QIV with an IVE of 50%, and 16 cases (95% CIs: 12–20 cases) using a QIV with an IVE of 70%. The respective estimates for saved SARI cases were 1 case (95% CIs: 0–2 cases) with an IVE of 20%, 2 cases (95% CIs: 0–6 cases) with an IVE of 50%, and 3 cases (95% CIs: 1–8 cases) with an IVE of 70% during the 2014–2015 season. Estimation of the avoidable SARI cases using a QIV instead of TIV was not feasible for the 2012–2013 and 2013–2014 influenza season, because of the very small number of SARI influenza B cases (three cases in each season).

Discussion

The impact of influenza B was estimated rather recently (1,4,12). Matias et al. found that influenza B accounted for 29% of the total influenza-attributable mortality in the United States during 1997–2007, particularly for 51–95% of deaths in 4 of 12 studied influenza seasons (12). Beyond morbidity and mortality, influenza B accounted for 17% of school absence due to acute viral respiratory infection among school-aged children in the United States during three influenza seasons (2012–2015) (13). A recent study based on data from five European countries (France, Germany, Italy, Spain, and the United Kingdom) during 2002–2013 estimated that an additional 1.03 million influenza cases, 453,000 consultations, 672,000 workdays lost, 24,000 hospitalizations, 10,000 deaths, and 242 million Euros on medical expenses and workdays lost could have been avoided in these five countries using a QIV compared with TIV over the 10-season period (18).

This is a nationwide study of influenza B lineages circulating in Greece during 2005–2015 and their impact. We found significant cocirculation of Victoria and Yamagata lineages or predominance of one of them in most influenza seasons. Overall, Yamagata lineage was the prevalent circulating influenza B lineage in Greece during 2005–2015. In a similar vein, Mosnier et al. found that lineages Yamagata and Victoria accounted for 62.8% and 37.2% of influenza B cases seeking primary health care during 2003–2013, respectively (14). However, virological surveillance data from 2000 to 2013 from 26 countries globally (200 influenza seasons) showed predominance of Victoria (64% of influenza B isolates) (5). Of note, lineages Victoria and Yamagata cocirculated in 32% of studied influenza seasons, accounting for ∼20% of isolates each (5).

In our study, the median lineage matching of circulating influenza B strains with vaccine influenza B strains during 2005–2015 was 19.2%, while complete mismatch was found in two influenza seasons. These findings indicate that TIV in most seasons conferred inadequate or no immunity, and thus protection against the circulating influenza B strains. Matching of SARI cases approximated those of non-SARI cases for the respective influenza seasons. In the study by Mosnier et al., mismatch with vaccine influenza B strains was found in three of five influenza seasons with influenza B activity of ∼10% (14). Mismatch between circulating influenza B lineages and vaccine strains was documented in 24% of influenza seasons in the study from 26 countries globally during 2000–2013, and this was even higher (31%) in northern hemisphere countries (5). We also found that the use of a QIV instead of a TIV could potentially prevent up to 16 SARI cases during the 2011–2012 season and up to 3 cases during the 2014–2015 season, depending on vaccine effectiveness.

Moreover, antigenic drift of influenza viruses should be considered concerning IVE. For example, during 2013–2015 influenza seasons, although there has been a “lineage match” to the circulating strains, there was considerable influenza. This was probably due to the antigenic drift of Yamagata lineage, as the vaccine strain changed from B/Massachusetts to B/Phuket in 2015.

Strengths of our study are as follows: the 10-year study period and the high virological, epidemiological, and geographical representativeness, since the vast majority of PCR testing for influenza in Greece is performed in the two National Influenza Reference Laboratories. Potential limitations include limitations of the surveillance system per se (e.g., in case there is a differential composition of the influenza B cases collected from cases not demanding medical attention and therefore not being eligible for swab collection) and the variation over the study period of surveillance practices (changes in ILI definition or potential testing practices).

In conclusion, our study provides for the first time an insight on influenza B and its impact in Greece. Considerable lineage mismatching was found between influenza B strains isolated from non-SARI and SARI cases and vaccine influenza B strains during the studied influenza seasons. The wide use of a QIV instead of a TIV is expected to save several SARI cases each year. Our findings strongly support the recommendation for using a QIV.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

Funding Information

Funding for this study was provided by GlaxoSmithKline Biologicals SA.

References

An der Heiden

, and Buchholz

. Estimation of influenza-attributable medically attended acute respiratory illness by influenzatype/subtype and age, Germany, 2001/02–2014/15. Influenza Other Respir Viruses, 2017; 11:110–121.

Anton

, Marcos

, Torner

, et al. Virological surveillance of influenza and other respiratory viruses during six consecutive seasons from 2006 to 2012 in Catalonia, Spain. Clin Microbiol Infect, 2016; 22:564.e2–564.e9.

Baxter

. Evaluating the case for trivalent or quadrivalent influenza vaccines. Hum Vaccin Immunother, 2016; 12:2712–2717.

Beauté

, Zucs

, Korsun

, et al. Age-specific differences in influenza virus type and subtype distribution in the 2012/2013 season in 12 European countries. Epidemiol Infect, 2015; 143:2950–2958.

Caini

, Huang

, Ciblak

, et al. Epidemiological and virological characteristics of influenza B: results of the Global Influenza B Study. Influenza Other Respir Viruses, 2015; 9 (Suppl 1):3–12.

Cassini

, Colzani

, Pini

, et al. Impact of infectious diseases on population health using incidence-based disability-adjusted life years (DALYs): results from the Burden of Communicable Diseases in Europe study, European Union and European Economic Area countries, 2009 to 2013. Euro Surveill, 2018; 23: DOI: 10.2807/1560-7917.ES.2018.23.16.17-00454.

Clements

, Meier

, McGarry

, et al. Cost-effectiveness analysis of universal influenza vaccination with quadrivalent inactivated vaccine in the United States. Hum Vaccin Immunother, 2014; 10:1171–1180.

Domachowske

, Pankow-Culot

, Bautista

, et al. A randomized trial of candidate inactivated quadrivalent influenza vaccine versus trivalent influenza vaccines in children aged 3–17 years. J Infect Dis, 2013; 207:1878–1887.

European Commission. Eurostat. NUTS-Nomenclature for territorial units for statistics. https://ec.europa.eu/eurostat/web/nuts/background (accessed July 14, 2019 ).

10.

Lytras

, Kossyvakis

, Melidou

, et al. Influenza vaccine effectiveness against laboratory confirmed influenza in Greece during the 2013–2014 season: a test-negative study. Vaccine, 2015; 33:367–373.

11.

Maltezou

, Katerelos

, Mavrouli

, et al. Seroepidemiological study of pandemic influenza H1N1 following the 2009–2010 wave in Greece. Vaccine, 2011; 29:6664–6669.

12.

Matias

, Taylor

, Haguinet

, et al. Estimates of mortality attributable to influenza and RSV in the United States during 1997–2009 by influenza type or subtype, age, cause of death, and risk status. Influenza Other Respir Viruses, 2014; 8:507–515.

13.

McLean

, Peterson

, King

, et al. School absenteeism among school-aged children with medically attended acute viral respiratory illness during three influenza seasons, 2012–2013 through 2014–2015. Influenza Other Respir Viruses, 2017; 11:220–229.

14.

Mosnier

, Caini

, Daviaud

, et al. Ten influenza seasons in France: distribution and timing of influenza A and B circulation, 2003–2013. BMC Infect Dis, 2015; 15:357.

15.

Pepin

, Dupuy

, Borja-Tabora

CFC

, et al. Efficacy, immunogenicity, and safety of a quadrivalent inactivated influenza vaccine vaccine in children aged 6–35 months: a multi-season randomised placebo-controlled trial in the Northern and Southern Hemispheres. Vaccine, 2019; 37:1876–1884.

16.

R Core Team. R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. 2019. www.R-project.org/ (accessed July 14, 2019 ).

17.

Shaw

, Xu

, Li

, et al. Reappearance and global spread of variants of influenza B/Victoria/2/87 lineage viruses in the 2000–2001 seasons. Virology, 2002; 303:1–8.

18.

Uhart

, Bricout

, Clay

, et al. Public health and economic impact of seasonal influenza vaccination with quadrivalent influenza vaccines compared to trivalent influenza vaccines in Europe. Hum Vaccin Immunother, 2016; 12:2259–2268.

19.

World Health Organization. WHO information for the molecular detection of influenza viruses. www.who.int/influenza/gisrs_laboratory/WHO_information_for_the_molecular_detection_of_influenza_viruses_20171023_Final.pdf (accessed July 14, 2019 ).

Circulation of Influenza Type B Lineages in Greece During 2005–2015 and Estimation of Their Impact