Circulation and characterization of seasonal influenza viruses in Cambodia, 2012‐2015

Abstract Background Influenza virus circulation is monitored through the Cambodian influenza‐like illness (ILI) sentinel surveillance system and isolates are characterized by the National Influenza Centre (NIC). Seasonal influenza circulation has previously been characterized by year‐round activity and a peak during the rainy season (June‐November). Objectives We documented the circulation of seasonal influenza in Cambodia for 2012‐2015 and investigated genetic, antigenic, and antiviral resistance characteristics of influenza isolates. Patients/Methods Respiratory samples were collected from patients presenting with influenza‐like illness (ILI) at 11 hospitals throughout Cambodia. First‐line screening was conducted by the National Institute of Public Health and the Armed Forces Research Institute of Medical Sciences. Confirmation of testing and genetic, antigenic and antiviral resistance characterization was conducted by Institute Pasteur in Cambodia, the NIC. Additional virus characterization was conducted by the WHO Collaborating Centre for Reference and Research on Influenza (Melbourne, Australia). Results Between 2012 and 2015, 1,238 influenza‐positive samples were submitted to the NIC. Influenza A(H3N2) (55.3%) was the dominant subtype, followed by influenza B (30.9%; predominantly B/Yamagata‐lineage) and A(H1N1)pdm09 (13.9%). Circulation of influenza viruses began earlier in 2014 and 2015 than previously described, coincident with the emergence of A(H3N2) clades 3C.2a and 3C.3a, respectively. There was high diversity in the antigenicity of A(H3N2) viruses, and to a smaller extent influenza B viruses, during this period, with some mismatches with the northern and southern hemisphere vaccine formulations. All isolates tested were susceptible to the influenza antiviral drugs oseltamivir and zanamivir. Conclusions Seasonal and year‐round co‐circulation of multiple influenza types/subtypes were detected in Cambodia during 2012‐2015.


Influenza viruses belong to the
Seasonal influenza epidemics occur every year in temperate regions during the winter months 3 : November to March/April in the northern hemisphere and May to September in the southern hemisphere. 4,5 Influenza seasonality is more variable in tropical/subtropical regions where circulation can be observed year-round, although activity is often more intense during rainy seasons. 6 In addition, influenza activity is punctuated by occasional pandemics arising from the introduction of novel influenza A viruses into human circulation.
These pandemics can significantly increase morbidity and mortality worldwide, with major economic impacts. 7 We have previously described the circulation and seasonality of influenza viruses in Cambodia during six consecutive years (2006)(2007)(2008)(2009)(2010)(2011) following the establishment of the Cambodian National Influenza Centre (NIC) in 2006. [8][9][10] These previous data demonstrated a peak in influenza circulation during the rainy season from June to November, which is consistent with influenza circulation in the southern hemisphere. However, year-round circulation was also described, characteristic of influenza seasonality in tropical/ subtropical regions, including some other Southeast Asian countries. 11,12 This current study furthers our understanding of influenza in Cambodia and describes the seasonal circulation, genetic and antigenic diversity, and antiviral drug susceptibility analyses of influenza viruses in Cambodia during four consecutive years (2012)(2013)(2014)(2015).

| Ethical statement
The Cambodian ILI surveillance system is a public health activity managed by the Ministry of Health in Cambodia and has a standing authorization from the National Ethics Committee for Human Research. Samples and patient information were anonymized for the purpose of this surveillance.

| Geographic background
Cambodia is a tropical climate country in Southeast Asia with more than 15.5 million people, situated in the southwestern part of the Indochina peninsula and sharing international borders with Thailand and Laos on the West and North, and Vietnam on the East and Southeast. 13 The country is affected by the Asian monsoon and is mostly hot and humid with a mean temperature of 27ºC and mean relative humidity of 77.5%. Similar to other subtropical/tropical areas, Cambodia has two distinct seasons: the dry season, which generally runs from November to April; and the rainy season, which starts in May-June and ends in October-November.  Figure 1.

| ILI surveillance system in Cambodia
The case definition for ILI was defined as previously described by WHO: sudden onset of fever (≥38ºC axillary temperature) and cough or sore throat in the absence of other diagnosis. 9,14

| Specimen collection
Between January 2012 and December 2015, specimens and surveillance data were collected from a subset of outpatients presenting with ILI at sentinel surveillance sites. During 2012, ILI samples were collected as described previously. 9  quality control testing) were forwarded to the NIC for confirmation and viral characterization.

| Laboratory methods
At the NIC, viral RNA was extracted using the QIAamp Viral RNA Mini Kit (Qiagen, CA, USA) and amplified using real-time RT-PCR to detect influenza A and B viruses using standard protocols. Influenza A viruses were subsequently subtyped using subtype-specific real-time reverse transcriptase polymerase chain reaction (RT-qPCR) assays targeting H1pdm, H1, H3, H5, H7, N1pdm, N1, and N2 genes. 8,9 All influenza primers were sourced from the International Reagent Resource (https ://www.inter natio nalre agent resou rce.org/Home.aspx).
Influenza viruses were isolated at the IPC laboratory by inoculation of the specimens that tested positive by real-time RT-PCR onto Madin-Darby canine kidney (MDCK) cells in an enhanced biosafety level 2 laboratory. 9 The influenza isolates were characterized by hemagglutination inhibition assay (HAI) using reference antigens and anti-sera provided by the WHO Collaborating Center (WHOCC) for Reference and Research on Influenza in Melbourne, Australia.
A representative number of influenza isolates were sent each year to the WHOCC in Melbourne for confirmation and further analysis, including antiviral testing and partial or full genome sequencing of representative viruses.

| Genome sequencing and phylogenetic analysis
Viral RNA extracted from MDCK supernatant was used to sequence the HA gene of all influenza isolates at the NIC laboratory using Sanger sequencing. At the WHOCC (Melbourne), a single-reaction, F I G U R E 2 Number of specimens positive for influenza by subtype in Cambodia 2012-2015 by week. Cambodian data accessed from the Global Initiative on Sharing All Influenza Data (GISAID-https ://www.gisaid.org/) multiplex RT-PCR method that amplifies the HA, NA, and M genomic segments of seasonal influenza A and B viruses for next-generation sequencing was used, as previously described. 15 Nucleotide sequences from the coding regions of the HA genes of A(H3N2), A(H1N1)pdm09, and influenza B viruses were aligned using the Mafft multiple aligner V1.3.7 in the Geneious V10.0.9 software package (www.genei ous.com). Sequences originating from Cambodia, surrounding countries, and representative reference sequences were downloaded from the EpiFlu™ Database (www.gisaid.org). Maximum likelihood trees were estimated using PhyML 3.0 16 with 1000 bootstrap replicates using the ATGC server (http://www.atgc-montp ellier.fr/phyml/ execu tion). The most appropriate nucleotide substitution method determined for each data set was the GTR + G model.
The complete matrix gene was sequenced from representative influenza A viruses using previously described methods, 15 to ascertain the presence of mutations (eg, Ser31Asn) associated with resistance to the adamantine class of inhibitors.

| Nucleotide sequence accession numbers
All Cambodian influenza A(H3N2), A(H1N1)pdm09, and influenza B viral sequences included in the analysis were submitted to the EpiFlu™ Database, and all of these sequences are available via the GISAID website (https ://www.gisaid.org/). Table S1 provides detailed information about all of the Cambodian isolates and sequences analyzed in this study.

| Antiviral susceptibility testing
All influenza isolates sent each year to the WHOCC in Melbourne were analyzed for neuraminidase (NA) inhibitor susceptibility testing using an enzyme inhibition assay utilizing the fluorescent substrate MUNANA as described previously. 17 The concentration of drug required to inhibit 50% of the NA activity (IC50) was calculated using the non-linear curve fitting function in the GraphPad Prism 4 package (GraphPad Software). The average IC 50 (nM) (± standard deviation) of two independent determinations was calculated for each virus. Outliers of more than two standard deviations from the overall mean were retested twice. 18 Antiviral susceptibility was classified according to the guidelines from the WHO working group on surveillance of influenza antiviral susceptibility. 18

| Statistical analysis
The comparisons between percentages and two means were tested by chi-squared (χ 2 ) and Student's t test, respectively. A p value < 0.05 TA B L E 2 The seasonal influenza strains circulating in Cambodia (2012Cambodia ( -2015 compared to the strains included in the WHOrecommended vaccine formulations for trivalent influenza vaccines; viruses in bold indicate where the dominant Cambodian strain matched the vaccine strain Year Virus

Trivalent inactivated influenza vaccine strains Cambodian circulating strains b (proportion of Cambodian isolates) Northern hemisphere a Southern hemisphere
was considered statistically significant. Proportions, means, and all statistical analyses were performed using STATA 9.0 (Statacorp).

| Influenza activity in Cambodia
During 2012-2015, 3,222 specimens were submitted to the Cambodian NIC and analyzed as part of the ILI surveillance system (Table 1) Table S2.

| Neuraminidase inhibitor susceptibility analysis
A total of 148 A(H3N2), 73 A(H1N1)pdm09, and 83 influenza B viruses were tested for susceptibility to the neuraminidase inhibitors oseltamivir and zanamivir. The analysis demonstrated that all of the tested isolates were sensitive to both drugs (Table S4). Full NA gene sequences were also generated for 68 A(H3N2), 25 A(H1N1)pdm09, and 48 influenza B viruses and confirmed that none contained mutations associated with NA inhibitor resistance (Table S3).

| Sequence analysis of the matrix gene for mutations associated with amantadine resistance
Sequencing of the matrix gene was completed for representative A(H1N1)pdm09 (n = 27) and A(H3N2) (n = 66) viruses from 2012 to 2015. Sequence analysis showed that all of the Cambodian isolates contained an amino acid change from serine to asparagine at position 31 (Ser31Asn) in the M2 protein, which is associated with resistance to the adamantine class of inhibitors (Tables S4 and S5).

| Phylogenetic analysis of A(H1N1) PDm09 Isolates
Phylogenetic analysis of the HA gene sequences was carried out for 70 representative Cambodian A(H1N1)pdm09 isolates from 2012 to 2015 ( Figure 3; GISAID accession numbers are listed in Table S1).
The    Vaccine strains began in 2006, 8,9 and the frequent mismatch between strains in- and continues to contribute to our knowledge of the regional and global circulation of seasonal influenza strains.