Figure

Figure. . Areas sampled and location of subtyped avian influenza viruses (H5N1), Laos, 2009–2010. Provinces that had previous outbreaks of highly pathogenic avian influenza and were part of the survey are indicated...

Serum samples were collected in 9 of 17 provinces in Laos from healthy ducks and chickens in live-bird markets, village backyard flocks, and layer duck farms. Cloacal, tracheal, and environmental (fecal and water) swab specimens were also collected and placed immediately in transport medium (Figure). Swab specimens were screened in pools of 4 by using a real-time reverse transcription PCR for the matrix (M) gene segment (6). Positive pools were reextracted individually, retested, tested for hemagglutinin 5 (H5) by real-time reverse transcription PCR (7), and injected into 10–11-day-old embryonated chicken eggs.

Sequencing was conducted by using an Illumina (San Diego, CA, USA) platform (swabs and isolates) (8,9) and conventional Sanger sequencing (isolates). Illumina reads were first mapped to a database of publicly available human, avian, and swine influenza virus reference sequences from the Western Hemisphere Americas and Eurasia, and then remapped against references with the highest number of reads and best average coverage. Final average coverage varied between samples and segments with ranges of 123–24,163 (9 samples), 21–18,671 (12 samples) and 9–436 for sample A/duck/Lao/670/10. Isolate genotypes were verified by using Sanger sequencing. Mixed infections could not be excluded for direct sequencing in the absence of an isolate.

Phylogenetic analysis (ClustalW [http://www.clustal.org/], neighbor-joining analysis, 1,000 bootstrap tests, maximum composite likelihood, pairwise deletions) was conducted by using MEGA version 5.02 (10). Sequences are available in GenBank (CY098294–CY098334, CY098336–CY098340, CY098342–CY098368, and CY098370–CY098464). Serum samples were screened by using an ELISA (FlockChek MultiS Screen; IDEXX Laboratories, Westbrook, ME, USA), and antibody-positive serum samples were tested by using a hemagglutinin inhibition assay for subtypes H3, H4, H5 (clades 2.3.2 and 2.3.4), H6, and H9 (lineages G1 and Y280) as described (11).

Figure A1

Figure A1. . Phylogenetic relationships of hemagglutinin 5 (H5) gene segments of avian influenza viruses (H5N1), Laos, 2009–2010. The tree was rooted in A/goose/Guangdong/1/1996 (nt 91–919, H5 numbering). Surveillance sequences are indicated in...

During March 2010, a total of 3,695 swab specimens were collected (1,928 duck and 279 chicken cloacal samples, 446 duck tracheal samples, 675 fecal samples, and 367 water samples). M gene prevalence was 4.0% (ducks), 1.8% (chickens), and 0.3% (environment samples). Five isolates were obtained (Figure A1). All M gene–positive swab specimens were collected in 13 locations (8 backyards, 2 markets, and 3 farms) (Table 1; Figure) Sample protection was suboptimal, and only 21 samples could be subtyped by real-time RT-PCR and sequencing (Table 1).

Phylogenetic analysis identified 3 groups of viruses: clades 2.3.2.1, 2.3.4.1, and 2.3.4.2 (Figure A1). Two samples were closely related to A/chicken/Lao/LH1/2010–like virus (outbreak in Vientiane in April–May 2010) and to A/chicken/Laos/C100209–194-PTK/2009–like virus (outbreak in Phongsaly in February 2009) (clade 2.3.4.1) (Figure A1; Table 1). These viruses were closely related to A/Guizhou/1/2009 and A/chicken/Vietnam/NCVD-404/2010 (Figure A1).

Sixteen surveillance viruses of clade 2.3.4.2 were highly homogeneous (Figure A1; Table 1) and closely related to A/environment/Guizhou/4/2009 and A/chicken/Vietnam/NCVD-394/2010–like viruses, although they contained the polymerase basic 2 (PB2) gene of a clade 2.3.2.1 donor virus and were therefore interclade reassortants (Figure A1; Table 1). The genotype of the 4 isolates was verified by using Sanger sequencing. The PB2 gene was most closely related to A/grey_heron/Hong_Kong/1046/2008–like viruses. These reassortants were detected in 3 locations in northern (backyard), central (backyard), and southern (farm) Laos, suggesting multiple introductions of reassortants into Laos (Figure A1; Table 1).

Three clade 2.3.2.1 viruses were detected in 2 ducks and 1 environmental sample (same trader). These viruses were interclade reassortants. Two (1 isolate and 1 direct sequence) contained 6 or 7 segments that were A/whooper swan/Mongolia/6/2009 like (clade 2.3.2.1), and the nucleoprotein gene was A/tree_sparrow/Jiangsu/1/08 like (clade 2.3.4). The environmental sample (direct sequence) contained A/whooper swan/Mongolia/6/2009–like hemagglutinin and M genes and A/Guizhou/1/2009–like PB2, nucleoprotein, and neuraminidase genes (clade 2.3.4.1) (Table 1). The genotype of the isolate was verified by using Sanger sequencing. Hemagglutinin segments of the reassortants were identical (100% nt identity). This identity and the source of the 3 samples (1 trader) suggest that reassortment occurred recently, likely in Laos.

Antibody titers to H5 and H9 and a subtype H3N8 virus isolate have been reported in Laos (4,5). Subtypes H4 and H6 also circulate in this region (1,2,12). For this study, 2,148 serum samples (1,899 from ducks, 200 from chickens, and 49 from unspecified species) were collected and 267 antibody-positive (seroprevalence 14%) duck and 15 (7.5%) chicken serum samples were detected. Hemagglutination inhibition testing for specific antibodies against H3, H4, H5, H6, and H9 detected all antibodies but to H3 (Table 2). Antibodies against H9 were detected at highest titers and most frequently (1.1% of ducks for each H9 lineage), although some cross-reactivity between G1 and Y280 likely occurred (Table 2). These ducks were most widely distributed (18 [19%] of 97 locations in all 9 provinces). Antibodies against H5 clade 2.3.4 were found at the detection limit in few serum samples (0.4% of ducks in 3 northern and 2 southern provinces). Antibodies against H5 clade 2.3.2 were detected in 0.4% of ducks in 4 northern provinces and the capital of Vientiane.

Among ducks, 34 (1.7%) were either shedding or had antibodies against avian influenza virus (H5N1), and there was >1 virus-positive or antibody-positive duck in each of the 9 provinces sampled. One village had 36% of sampled ducks exposed to this virus; 18% shed 2.3.4 virus and 18% had antibody to 2.3.2 virus (Table 2). One duck in this village had antibodies against 2.3.2 and 2.3.4 clade viruses. Exposure to 2.3.2 and 2.3.4 viruses was evident in ducks from locations in 3 other provinces (1 district each in Champasak and Xiengkhoung, and several districts in Luang Prabang).

This study showed that 3 groups of avian influenza viruses (H5N1) were likely introduced into Laos in 2009–2010, one of which resulted in 2 outbreaks (2009, 2010). In all 9 provinces where surveillance was conducted, ducks had been exposed to this virus. Evidence of clades 2.3.2 and 2.3.4 virus activity was detected in 4 provinces. Several interclade reassortants were identified, demonstrating the high genetic mobility of these viruses in the region. Since 2004, Laos has had repeated outbreaks of highly pathogenic avian influenza viruses, which have also been detected in China and Vietnam. There is no evidence that a particular virus lineage has established itself in Laos. The frequency of introduction, diversity, and extent of these viruses in Laos suggests considerable movement of viruses into the country from surrounding territories (China and Vietnam, but not Cambodia) and within the country.