Literature review of the epidemiology of influenza B disease in 15 countries in the Asia‐Pacific region

Influenza control strategies focus on the use of trivalent influenza vaccines containing two influenza A virus subtypes and one of the two circulating influenza type B lineages (Yamagata or Victoria). Mismatches between the vaccine B lineage and the circulating lineage have been regularly documented in many countries, including those in the Asia‐Pacific region. We conducted a literature review with the aim of understanding the relative circulation of influenza B viruses in Asia‐Pacific countries. PubMed and Western Pacific Region Index Medicus were searched for relevant articles on influenza type B published since 1990 in English language for 15 Asia‐Pacific countries. Gray literature was also accessed. From 4834 articles identified, 121 full‐text articles were analyzed. Influenza was reported as an important cause of morbidity in the Asia‐Pacific region, affecting all age groups. In all 15 countries, influenza B was identified and associated with between 0% and 92% of laboratory‐confirmed influenza cases in any one season/year. Influenza type B appeared to cause more illness in children aged between 1 and 10 years than in other age groups. Epidemiological data for the two circulating influenza type B lineages remain limited in several countries in the Asia‐Pacific, although the co‐circulation of both lineages was seen in countries where strain surveillance data were available. Mismatches between circulating B lineages and vaccine strains were observed in all countries with available data. The data suggest that a shift from trivalent to quadrivalent seasonal influenza vaccines could provide additional benefits by providing broader protection.


| INTRODUCTION
Epidemic influenza causes global public health burden each season.
The World Health Organization (WHO) estimates that influenza severely affects between three and five million individuals each year and causes between 250 000 and 0.5 million deaths. 1 The influenza attack rate is highest in children, while complications including hospitalization and death occur most frequently in elderly individuals. 1 Other specific high-risk groups prioritized by WHO for vaccination include pregnant women, the highest priority group, followed by individuals with a compromised immune system and individuals with comorbidities such as pulmonary or cardiac disease. 2 Influenza type A and B viruses cause the vast majority of influenza disease in humans, and infection is preventable by vaccination. The relative proportion of influenza cases caused by type A and type B strains varies annually, reflecting antigenic drifts in the predominant strains and the host's level of immunity. In the last decade, influenza A viruses rep- in up to 60% of samples in Europe during the same period, with a seasonal average of 24% and 23% of samples, respectively. 7 Seasonal influenza vaccines are modified annually to include those antigenic variants that are likely to predominate in the following influenza season. Vaccine strain selection is performed by the WHO using data from the Global Influenza Surveillance and Response System, a network of over 140 institutions in 111 countries. 8 The B/Yamagata and B/Victoria influenza strains are antigenically distinct, and vaccines using one lineage induce only low levels of cross-protection to the other lineage. 9,10 Trivalent seasonal influenza vaccines only contain one influenza B lineage, and it is not always possible to predict which B lineage will predominate during the next influenza season. 11 Mismatch between the vaccine lineage and circulating influenza B lineage has occurred regularly, which can have a significant impact on influenza vaccine efficacy. 4,[12][13][14] Since 2012, the WHO has recommended the inclusion of strains from both B lineages in quadrivalent seasonal influenza vaccines. 15 Co-circulation of both influenza type B lineages has also been documented throughout South-East Asia and Oceania. 5 The use of influenza vaccine in many Asia-Pacific countries is limited, and the potential impact of quadrivalent influenza vaccines on illness and hospitalization rates in these countries is not known, but is also likely to be low. 16 To obtain an epidemiological view of influenza type B in the Asia-Pacific region, we conducted a review of the available literature.
We attempted to identify periods of influenza B lineage mismatch between vaccine and circulating strains in 15 countries within the Asia-Pacific region to better inform health authorities of the potential benefits of quadrivalent influenza vaccines for protection against seasonal influenza.

| Search strategy and selection criteria
Articles were identified from PubMed and the Western Pacific Region Index Medicus (WPRIM) using search strings comprised of terms that identified influenza, the selected countries of interest in the Asia-Pacific region, and the epidemiology/burden of disease (Appendix S1). The search covered publication dates between January 1990 until 11 April 2016 and was limited to articles published in English language. Titles and abstracts were screened for relevance: That is, reported outcomes related to seasonal influenza in humans and in a country of interest. Articles were excluded if they reported pharmacokinetic or pharmacodynamic studies, case reports, case series, clinical trials, or meta-analyses. Articles were also excluded if they reported features of influenza pathophysiology, treatment, or diagnosis, or if they reported data in fewer than 30 patients. Publications without abstracts were only reviewed if the title fitted the review objectives.
F I G U R E 1 Countries included in the review according to climate. Northern Hemisphere: temperate = dark blue, subtropical = light blue, subtropical to tropical = yellow, tropical = orange. Southern Hemisphere: tropical = red, temperate = green Full-text articles were reviewed to assess their relevance and methodological quality. Articles were excluded if the method sections were insufficiently described; if the content did not provide relevant information to the review objectives; if the article reported "pneumonia and influenza" as a combined outcome (unless pneumonia was described as a complication of influenza), outcomes from mathematical models; and if no quantitative data could be retrieved. Gray literature including WHO websites, local ministries of health, and WHO vaccine recommendations was also assessed for relevance. Extracted data included information on epidemiology and circulating strains. We did not collect clinical criteria or clinical case definitions of influenza, influenza-like illness (ILI), febrile illness, acute respiratory infection (ARI) disease, or severe acute respiratory illness (SARI) used in individual studies. Nor did we specify the methods for selecting cases for specimen collection or the influenza-testing method used for laboratory diagnosis. In our review, "laboratory-confirmed influenza" or a "positive sample" refers to a case of influenza confirmed by the method stated in the reporting paper.
An influenza B mismatch was defined as the circulating influenza B virus lineage strains differing from the B lineage representative strain included in the WHO-recommended influenza vaccine composition for that season. When <20% of the circulating influenza B strains differed from the WHO-recommended vaccine strain, we arbitrarily considered the degree of mismatch to be "low." A difference between 20% and 40% was considered as partial mismatch. We considered a significant mismatch as >40% and complete mismatch when ≥95% of the circulating influenza B lineage strains did not belong to the trivalent vaccine lineage.
The initial literature search in 2013 was conducted by Pallas Health Research and Consultancy B.V., the Netherlands. Quality control activities included review of the first 30% of titles and abstracts and of the first 10% of full-text articles in duplicate by two independent researchers from Pallas. Any disagreements were adjudicated by a third researcher. The search was updated in 2016 by BM, and the articles were selected by BM and JC.
Ethics approval was not required for this study. The majority of relevant publications concerned surveillance or other observational epidemiological studies for which no standard quality checklists are currently available.

| RESULTS
There were 121 English language articles included in the review ( Figure 2), of which 120 articles provided information on influenza B strains as a proportion of all laboratory-confirmed influenza from data collected between 1990 and 2015 (Table 1) hospitalized for respiratory tract infections (RTI), patients in intensive care, and respiratory samples from in/outpatients with a broad range of underlying respiratory syndromes. In many studies, the age range of subjects/samples was not specified. Many studies included patients with diagnoses of low specificity for influenza such as "febrile illness," "ILI," and "acute lower respiratory tract infection" (ALRTI). There were 102 prospective studies (two studies included both prospective and retrospective components). Sample sizes in individual study groups ranged from 26 to more than 300 000.

| Influenza surveillance and vaccine coverage in Asia-Pacific countries
With the exception of Myanmar, all of the countries studied use a sentinel site approach for influenza surveillance ( In countries with a policy to provide free vaccine for at-risk groups, vaccine uptake among these groups has been substantial: Approximately 73% of 65+-year-olds in Australia, 67.5% in New Zealand, 80% of adults in Taiwan, and 82.5% in 65+-year-olds in South Korea were reported to have received influenza vaccine (Table 1). In China, influenza vaccine coverage was reported as 26% in children and 7.4% in 60+-year-olds, but regional differences may exist due to subsidization of influenza vaccines in some regions. 17  Twenty-three articles reported on influenza B in China from 1995 until 2014 ( Table 2). Most of the articles reported clinic/hospitalbased ILI surveillance in one city/province and provided aggregated data over multiple seasons.

| Cambodia
A study in Shenzhen in the southern province of Guangdong from 1995 to 2009 reported an annual influenza detection rate among ILI cases ranging from 0.2% in 1998 to 25% in 2009. The lowest proportion of influenza B among all influenza-positive samples was in 2008. 35

| Indonesia
We identified five articles with data relevant to Indonesia ( Table 2).
Three of these papers were based on ILI surveillance of multiple sea-

| Myanmar
Three

| New Zealand
Seven articles describing influenza in New Zealand between 1990 and 2015 were identified. National influenza surveillance data showed that during 1997-2008, on average 718 cases of laboratory-confirmed influenza were detected every year. 53 Influenza surveillance in 2005 recorded the highest influenza B activity since 1990 with co-circulation of influenza strains from B/Victoria and B/Yamagata lineages, which resulted in significant B strain mismatch with the recommended vaccine composition. 54 In 2015, 44.2% of patients with ILI and 23.6% SARI patients tested positive for influenza. 55 The proportion of influenza B among all influenza-positive samples was 13%, but half of them mismatched the B strain contained in the influenza vaccines for that season (Table 3).

| Papua New Guinea
In the one study identified as providing information on influenza in Papua New Guinea, 29.3% of samples from patients with ILI received by the Papua New Guinea National Influenza Centre in 2010 had laboratory-confirmed influenza, of which 43.2% were influenza type B. 56

| Singapore
In Singapore, the National Influenza Centre carries out influenza virus surveillance using samples from public hospitals and sentinel clinics. 62 Ten articles reporting data in Singapore from 1990 until 2012 were identified ( Table 2)

Eleven articles reported the distribution of influenza A and B viruses
in South Korea between 1990 and 2013 (

| Taiwan
Thirteen studies conducted between 1995 and 2012 were identified for Taiwan (Table 2). Five articles reported data in children, and no studies specifically reported on influenza B in adults. All studies con-  (Table 3).

| Distribution of influenza B by age group
There were 22 articles that provided information on influenza B in different age groups ( Table 4). The grouping of ages differed between studies which hinders easy comparisons.

| Circulating influenza B strains
The results from co-circulation of several types/lineages, the degree of antigenic drift, combined with the hosts' immune status, which together ensure the continuing ability of the virus to cause illness. 9 Significant or complete mismatches between the circulating and trivalent vaccine type B strain were observed on numerous occasions in countries with available data. Our study was not designed to quantify the possible public health implications in seasons where vaccine mismatch existed. Influenza vaccine efficacy is reduced when there is a mismatch between vaccine and circulating strains, 14,86 suggesting that a mismatch season is likely to be associated with a higher clinical disease burden.
Influenza B causes similar morbidity as influenza A. 87 In mismatch seasons, hospitalization due to influenza type B can exceed that due to influenza A in all age groups. 71 The available data suggest that consistent with observations in other regions, 7,9 in the Asia-Pacific region, influenza B occurs more frequently in children aged between 1 and 10 years than in other age groups and causes more severe disease in this age group than influenza A. 88 Evidence suggests differences in the age distribution of patients infected with the B/Yamagata or B/Victoria lineages, with the latter appearing to be more frequently identified in younger age groups. [89][90][91][92][93] Although The articles included in this review varied with respect to their design, population characteristics (eg, age range, mild vs severe cases), the illness definition selected for study (specific vs non-specific diagnoses), the laboratory methods used to detect influenza, and the methods of case surveillance (population-based, laboratory-based, hospital or emergency department-based, sentinel general practice). Therefore, the results of individual studies cannot be easily compared and are unlikely to be broadly representative. In some articles, data were not stratified per year, and only an average proportion of influenza type B over the study period could be obtained. In many studies, proportions of influenza B among any laboratory-confirmed influenza cases were only available within a specified population. Most information was retrieved for China, Australia, South Korea, and Taiwan, but four articles or fewer were identified for Laos, Myanmar, the Philippines, and Papua New Guinea.
This review has identified important knowledge gaps within the region. In several countries, the epidemiology of influenza and influenza type B is not well described, little is known about the age groups most affected by influenza type B and the relative contribution of the two type B lineages to the disease burden. Epidemiological data for the two circulating influenza B lineages in the Asia-Pacific region are extremely limited. Co-circulation of Yamagata and Victoria lineages occurred in most countries where strain surveillance data were available, with considerable fluctuation from year to year. A variable, but substantial influenza B burden, as well as variable mismatch between circulating lineages and vaccine lineages was observed in all countries with available data regardless of geographical location, suggesting that a shift from trivalent to quadrivalent seasonal influenza vaccines that include both influenza B lineages would be beneficial in many seasons.
Establishing or enhancing existing influenza surveillance networks in individual countries across the Asia-Pacific region is needed to contribute to an improved understanding of the burden of influenza.
Education of the medical profession and public, vaccine implementation strategies including the development of specific national recommendations in countries where they are lacking, and improved access to influenza vaccines, are needed to improve influenza vaccine uptake and reduce the influenza disease burden in Asia-Pacific countries.
Few countries in the Asia-Pacific region have policies, recommendations, or funding methods in place supporting influenza immunization, and those that do limit publicly funded re-imbursement to at-risk groups. Our review provides evidence of a substantial influenza burden in the Asia-Pacific. The data suggest that countries in the Asia-Pacific stand to benefit from development of immunization policy targeting influenza prevention. Additionally, quadrivalent influenza vaccines that reduce the likelihood of vaccine mismatch among influenza type B strains are likely to provide improved protection against influenza type B infection.

CONTRIBUTORSHIP
LJ, QSH, and WJK participated in the design of the review. LJ also participated in assembling and analysis of data and provided Asia-Pacific regional knowledge for the interpretation of the published material collected. QSH, IB, P-IL, and WJK participated in assembling and interpreting the data. MS, P-IL, and WJK participated in the analysis and interpretation of data. PB and BAM participated in collecting, assembling, analysis, and interpretation of data. BAM also participated in the design of the review. JC participated in the design of the review, collecting the data, the supervision of the analysis, and the interpretation of data. All authors were involved in the development of this manuscript, had full access to the data, and gave final approval before submission.

SOURCES OF SUPPORT
GlaxoSmithKline Biologicals SA was the funding source and was in-