Research Article: Patterns of influenza B circulation in Latin America and the Caribbean, 2010–2017

Date Published: August 8, 2019

Publisher: Public Library of Science

Author(s): Rakhee Palekar, Angel Rodriguez, Cinthia Avila, Gisela Barrera, Miriam Barrera, Hebleen Brenes, Alfedo Bruno, Nathalie El Omeiri, Rodrigo Fasce, Walquiria Ferreira de Almeida, Danilo Franco, Maribel Huaringa, Jenny Lara, Roxana Loayza, Irma Lopez-Martinez, Terezinha Maria de Paiva, Jose Medina, Jenny Ojeda, Alba Maria Ropero, Viviana Sotomayor, Cynthia Vazquez, Marta Von Horoch, Jeffrey Shaman.


There are limited published data about the circulation of influenza B/Victoria and B/Yamagata in Latin America and the Caribbean (LAC) and most countries have a vaccine policy that includes the use of the trivalent influenza vaccine. We analyzed influenza surveillance data to inform decision-making in LAC about prevention strategies, such as the use of the quadrivalent influenza vaccine.

There are a total of 28 reference laboratories and National Influenza Centers in LAC that conduct influenza virologic surveillance according to global standards, and on a weekly basis upload their surveillance data to the open-access World Health Organization (WHO) platform FluNet. These data include the number of specimens tested for influenza and the number of specimens positive for influenza by type, subtype and lineage, all by the epidemiologic week of specimen collection. We invited these laboratories to provide additional epidemiologic data about the hospitalized influenza B cases. We conducted descriptive analyses of patterns of influenza circulation and characteristics of hospitalized cases. We compared the predominant B lineage each season to the lineage in the vaccine applied, to determine vaccine mismatch. A Chi-square and Wilcoxan statistic were used to assess the statistical significance of differences in proportions and medians at the P<0.05 level. During 2010–2017, the annual number of influenza B cases in LAC was ~4500 to 7000 cases. Since 2011, among the LAC-laboratories reporting influenza B lineage using molecular methods, both B/Victoria and B/Yamagata were detected annually. Among the hospitalized influenza B cases, there were statistically significant differences observed between B/Victoria and B/Yamagata cases when comparing age and the proportion with underlying co-morbid conditions and with history of oseltamivir treatment (P<0.001). The proportion deceased among B/Victoria and B/Yamagata hospitalized cases did not differ significantly. When comparing the predominant influenza B lineage detected, as part of surveillance activities during 63 seasons among 19 countries, to the lineage of the influenza B virus included in the trivalent influenza vaccine used during that season, there was a vaccine mismatch noted during 32% of the seasons analyzed. Influenza B is important in LAC with both B/Victoria and B/Yamagata circulating annually in all sub regions. During approximately one-third of the seasons, an influenza B vaccine mismatch was identified. Further analyses are needed to better characterize the medical and economic burden of each influenza B lineage, to examine the potential cross-protection of one vaccine lineage against the other circulating virus lineage, and to determine the potential impact and cost-effectiveness of using the quadrivalent vaccine rather than the trivalent influenza vaccine.

Partial Text

Seasonal influenza viruses circulate annually and cause disease in humans. In temperate countries, there is typically one influenza season annually, while in tropical countries, the seasonality of influenza can vary with typically two periods of peak virus activity annually. There are four groups or types of seasonal influenza viruses—influenza A, B, C and D. Type C influenza viruses cause mild human infection and are associated with sporadic cases and as such, active surveillance is not conducted for influenza C viruses. Influenza D viruses primarily affect cattle and are not known to infect or cause illness in people [1]. Among influenza A and B viruses, influenza A viruses tend to predominate and during each season, a variable percent of influenza A viruses are of the H1 subtype and of the H3 subtype. While influenza A viruses predominate during influenza seasons compared to influenza B viruses, influenza B viruses are detected globally each season in both tropical and temperate countries [2,3].

The Global Influenza Surveillance and Response System (GISRS) is a network that as of 2017, included 141 National Influenza Centers (NICs), six Collaborating Centers (CCs), and four Essential Regulatory Laboratories that conduct influenza virologic surveillance according to terms of reference established by the World Health Organization (WHO)[12]. As of 2017 in LAC, there were a total of 24 NICs and an additional four laboratories [13] that conducted influenza surveillance according to these global standards, covering a total of 34 LAC countries and territories. These laboratories receive samples from sentinel sites conducting influenza surveillance, primarily in hospitals, but also in some ambulatory primary-care settings. They also receive samples collected upon clinicians’ request for diagnostic testing. Laboratories use a combination of indirect immunofluorescence and real-time polymerase chain reaction (rRT-PCR) to test for the presence of influenza viruses. Real-time RT-PCR testing is conducted using molecular detection kits provided by the WHO CC for Influenza Surveillance at the U.S. Centers for Disease Control and Prevention (CDC) [13,14]. Influenza A viruses are further subtyped, and influenza B viruses are tested for lineage, by rRT-PCR. The WHO CC at the U.S. CDC began to provide influenza B lineage detection molecular kits in 2013 to countries in LAC free-of-charge. The percent of the population covered by the surveillance in each country has not been systematically documented, but in most of the countries, samples are collected from all parts of the country for testing at the NIC.

During 2010–2017, 28 LAC laboratories detected n = 182,813 influenza A and 38,456 influenza B cases (Fig 1). While influenza A cases predominated during the period of analysis, compared to influenza B, there were a total of 4500–7000 influenza B cases annually (Table 1). Influenza B cases were detected in all geographic subregions of LAC (Fig 2). Among all influenza cases detected, the percentage of influenza samples in a given year that were reported to be influenza B varied from 1% to 54% (Table 1). There was year-to-year variability in all sub-regions (S1 Table). The timing of the circulation of influenza B varied by year and subregion (Fig 2). In the tropical Andean and Central American subregions, influenza B was detected year-round while in the temperate Southern Cone, influenza B was mostly detected at the end of the influenza season (Fig 2).

This analysis of the patterns of influenza B circulation in Latin American and the Caribbean provides important information to guide influenza prevention and control strategies. First, influenza B viruses represented 11–26% of all influenza cases in LAC annually. In some sub-regions, and during certain years, they represented as much as 43% of all influenza cases. These findings are similar to those of studies from other regions, reporting influenza B cases to represent 20–50% of influenza cases detected during a season ([3,4]. These results suggest that influenza B is an important contributor to influenza disease. Countries estimating influenza burden should contemplate estimates by type of influenza virus. Second, while influenza B has often been described as circulating at the end of the influenza season in temperate countries, we found, especially in tropical countries, that influenza B circulates year-round. This pattern of year-round circulation in tropical countries has been described in other regions globally [3,7] and as such, typing of influenza viruses, as part of virologic surveillance, should be conducted throughout the year [16].