Influenza is a contagious respiratory disease caused by the influenza virus, a member of the Orthomyxoviridae family of RNA viruses which have a segmented genome. There are three types of influenza viruses (A, B and C), originally distinguished on the basis of their internal nucleoprotein and matrix proteins. Influenza virus A and B cause annual, recurrent outbreaks in tropical and sub-tropical countries leading to high morbidity and mortality [1], especially in children and the elderly. It is estimated that one million deaths each year are due virus infection [2]. According to the World Health Organization (WHO), about20%of all deaths among children aged <5 years are attributable to influenza-like illnesses (ILI) [3], and the majority of those deaths occur in Africa and Southeast Asia [3, 4].

Seasonal influenza viruses are a population of virus variants, mainly existing as a result of antigenic drift, the accumulation of point mutations in the hemagglutinin (HA) and neuraminidase (NA) surface glycoprotein due to the lack of virus RNA polymerase complex pro of reading activity [5]. The HA is known to be the main antigenic determinant of the host’s neutralizing immune response [6,7,8] and is under selective pressure for antigenic shift, the process by which new virus variants emerge from genetic reassortment among influenza viruses infecting human and animal populations. Antigenic shift results in virus evasion of the preexisting immunity of the host [9]. In addition, genetic mechanisms like nucleic acid substitution, gene insertion and deletion have been reported to contribute with influenza virus variability [9, 10].

Previous studies have shown that influenza is a major contributor of respiratory illness in tropical and subtropical regions worldwide, with more complex patterns of influenza seasonality in Africa and southern Asia [11, 12, and 13]. Tropical countries closer to the equator have year-round influenza activity without a distinct seasonal peak [14, 15, 16, 17]. However, South Africa, Zambia [18], and Senegal [19] seem to have well-defined patterns of seasonality similar to other countries with temperate climates [17]. Furthermore, Madagascar described a yearly circulation of influenza with two distinct peaks, one that corresponds to the northern hemisphere winter (January–March) and a second that corresponds to the southern hemisphere winter (June–October) [18]. Unfortunately, the limited data available from sub-Saharan Africa makes it difficult to comprehensively map and define the seasonality of influenza in the region. The seasonality of influenza coupled with virus genetic and antigenic variation present challenges to the selection of potential virus candidates to be used in the formulation of influenza vaccines. Timely and geographically comprehensive virologic surveillance involving genetic and antigenic analysis of circulating strains is critical to maintaining the effectiveness of the influenza vaccine [20].

Although some countries, such as Madagascar and Senegal, have led national influenza surveillance for many years [21, 22], several countries in the African continent began conducting influenza surveillance recently. In DRCongo, the influenza surveillance system based on laboratory analysis was introduced in 2008 and reported that influenza is present year-round associated with some peaks related to the rainy season [23]. Despite the increased interest on improving influenza surveillance in Africa, the available genetic and epidemiological data are still insufficient to understand disease dynamics in the region. Due to the limited knowledge on influenza virus dynamics in Africa and to contribute in the improvement of the DRC surveillance system, we sought to identify and characterize for the first time the influenza viruses circulating in the Democratic Republic of Congo during the period from August to December of 2014.

From August to December 2014, 806 samples were collected in DRCongo from which 61 were positive for influenza A (Table 1, Figure 1). During this reporting period, no influenza B specimens were detected (Table 1). Only 32 of the 61 influenza A positive samples from different surveillance sites in DRC had a Ct < 30 and these samples were selected and shipped to CDC in Atlanta for antigenic and genetic characterization. Results from this study showed that among the 32 samples tested: 26 were influenza A(H3N2), 2 were influenza A(H1N1)pdm09, 2 had insufficient volume for testing and 2 samples were negative for influenza by RT-PCR, as shown in Table 2.

The main objective of this study was to identify and characterize influenza viruses circulating in the Democratic Republic of Congo in 2014. To the best of our knowledge, this kind of work has never been done in the DRCongo, while other African countries like Madagascar, Morocco and Egypt, have worked on the improvement of their surveillance system capacity to understand the molecular epidemiology and disease burden caused by circulating influenza viruses [35].

Our findings showed that the influenza activity in DRCongo increased during the month of November of 2014, reaching its peak on December the same year. Of note is that influenza activity was concomitant with the September - April rainy season in DRC, as previously described for the 2009-2011 seasons [36]. Similar results were found in Senegal (Dakar), Côte d’Ivoire (Abidjan), and Niger (Niamey) probably because of lower temperatures, as described elsewhere for Niger in 2010 [36]. This same pattern has also been observed in the north eastern part of tropical Brazil (Fortaleza) [37].

A total of 32 virus samples collected at the DRC from August to December of 2014 were shipped to the CDC in Atlanta, from which 20 were genetically characterized by either full genome sequencing or limited pyrosequencing. Limited pyrosequencing was done in A(H3N2) virus isolates which had insufficient HA titer for HI testing or when viruses were not recovered in cell culture.

Sixteen of the A(H3N2) viruses isolated at the DRC belonged to genetic group3C.2a and two isolates to group 3C.3. Four of these isolates were antigenically similar to the A/Switzerland/9715293/2013 virus, the recommended A(H3N2) vaccine component during the 2015-2016 influenza season. In 2013, the majority of A(H3N2) viruses tested was antigenically similar to the A/Texas/50/2012 vaccine virus. After the WHO northern hemisphere vaccine selection meeting in February 2014, antigenic drift was detected in viruses from genetic group 3C, giving rise to subgroups 3C.2a and 3C. 3a. Ferret antisera raised against A/Texas/50/2012 vaccine virus showed a reduction in HA titer, with the concomitant worldwide increase of 3C.2a and 3C.3a groups from 2014 to date. Despite the genetic diversity, the majority of 3C.2a and 3C.3a viruses were antigenically similar to the new A/Switzerland/9715293/2013 vaccine virus.

Other A(H3N2) HA sequences from African countries, including Algiers, Dakar, Madagascar, Ethiopia, Ghana and Tanzania, available from the GISAID Epiflu database, were classified as 3C.3a, and one isolate from Egypt was found to cluster with other 3C.3b viruses (figure 2). No DRC A(H3N2) isolates from genetic groups 3C.3a and 3C.3b were collected in 2014. However, similarly to the DRC, 3C.3 viruses were also found in Zambia and Kenya after GISAID EpiFlu database searching. Based on the close proximity of East African countries with 3C.3a viruses in circulation, we expect the further expansion of the 3C.3a group to the DRC.

Two A(H1N1)pdm09 isolates from the DRC were classified under genetic group 6B and were antigenically indistinguishable to the A/California/07/2009 virus, which remains the WHO A(H1N1)pdm09 influenza vaccine component since 2009. The A(H1N1)pdm09 HA sequences from Ethiopia and Egypt, also available from the GISAID Epiflu database, clustered with other 6B virus sequences as well. Interestingly, Ghana was the only African country found to have viruses from genetic group 6C in circulation since the summer of 2014 and the reason why further expansion was not observed at the time remains to be elucidated.

Ferret antisera raised against the A/California/07/2009 vaccine virus showed reduced HA titers with a small number of isolates during the 2014-2015 influenza season. These A(H1N1)pdm09 low-reactors often had amino acid changes in HA positions 153-157 and may have acquired these changes through cell culture, which is consistent with results obtained since May 2009. African isolates were not found to have amino acid changes in HA positions 153-157, confirming that no antigenic change has taken place, as established by the antigenic analysis data.

Influenza A(H3N2) and A(H1N1)pdm09 viruses have continued to circulate worldwide since their emergence and are expanding at high levels in most countries. Molecular surveillance year-round, especially in sub and tropical areas of Africa, is critical for improving vaccine strategies for the prevention and control of influenza illness. It is also important in contributing to the understanding of genetic variations and in the early detection of antigenic drift.