Epidemiologic and Virologic Investigation of Hand, Foot, and Mouth Disease, Southern Vietnam, 2005

Human enterovirus 71, but not coxsackievirus A16, is strongly associated with acute neurologic disease.

H and, foot, and mouth disease (HFMD) is a common febrile illness of early childhood, characterized by 3-4 days of fever and the development of a vesicular enanthem on the buccal mucosa, gums, and palate and a papulovesicular exanthem on the hands, feet, and buttocks (1). HFMD is caused by acute enterovirus infections, particularly by viruses belonging to the human enterovirus A (HEVA) species (1).
Although all HEVA viruses can cause HFMD, infection with HEV71 is also associated with a high prevalence of acute neurologic disease (4). Despite their close genetic relationship to HEV71, the HEVA CVA viruses rarely cause acute neurologic disease. HEV71 infection is associated with a wide spectrum of acute central nervous system syndromes, including aseptic meningitis, poliomyelitislike paralysis, brainstem encephalitis, and acute neurogenic pulmonary edema (4). Children <5 years of age are particularly susceptible to HEV71-associated acute neurologic disease, which may occasionally cause permanent neurologic disability or death (4).

Epidemiologic and Virologic
Investigation of Hand, Foot, and Mouth Disease, Southern Vietnam, 2005 Before 1999, most cases of encephalitis in southern Vietnam occurred in children >5 years of age, of which ≈60% were identifi ed as Japanese encephalitis (diagnostic records of the Pasteur Institute, Ho Chi Minh City, Vietnam). Since 2002, however, viral encephalitis has increasingly been observed in younger children, particularly in those <4 years. Furthermore, since 2002 <27% of encephalitis cases have been confi rmed as Japanese encephalitis, which indicates that the epidemiology of viral encephalitis in southern Vietnam may be changing. This situation led us to consider other possible causes for viral encephalitis.
In 2003, we isolated HEV71 (at the Pasteur Institute, Ho Chi Minh City, Vietnam) from 12 patients with encephalitis, who sought treatment at the hospital during an HFMD outbreak in southern Vietnam. To our knowledge, this was the fi rst identifi cation of HEV71 in Vietnam. Although laboratory surveillance has been shown to provide adequate warning of impending outbreaks of HEV71-associated acute neurologic disease (18), laboratory surveillance for HEV71 has not yet been established in Vietnam.

Study Participants and Specimen Collection
Children <15 years of age were admitted to a large pediatric hospital in Ho Chi Minh City, Vietnam. This hospital serves ≈70% of the city's pediatric population; 764 children with HFMD were enrolled in the study. HFMD was defi ned as a febrile illness (>37.5°C), accompanied by a papulovesicular rash in a characteristic distribution (oral mucosa, extremities of limbs, buttocks). A total of 1,928 specimens were collected from the children on the day of admission. Each child had at least 1 specimen collected from vesicle fl uid, throat swab, or stool. Children who also exhibited acute neurologic disease had a cerebrospinal fl uid specimen collected. All specimens were extracted with chloroform (1:10 in phosphate-buffered saline) before virus isolation in cell culture.

Virus Isolation
Virus isolation was undertaken in cell culture by using both human rhabdomyosarcoma (RD) (ATCC CCL136) and African green monkey kidney (Vero) (ATCC CCL81) cell lines. Each specimen underwent at least 2 cell culture passages in RD and Vero cells before being reported as negative. Samples demonstrating viral cytopathic effect (CPE) were screened for enterovirus RNA by reverse transcription-PCR (RT-PCR), as outlined in the following section.

RNA Extraction from Cell Culture Supernatants
Total cellular RNA was extracted from cell culture supernatants that demonstrated CPE; Tri-reagent (Ambion, Austin, TX, USA) was used. The RNA obtained from 250 μL of infected cell culture supernatant was suspended in 30 μL RNase-free water and stored at -80°C before use.

Pan Enterovirus RT-PCR Assay, 5′ Untranslated Region (UTR)
Briefl y, cDNA was prepared in a 10-μL reaction mixture containing 6 μL RNA template, 0.5 mmol/L dNTP, 200 U Moloney murine leukemia virus reverse transcriptase (M-MuLV RT) (Promega, Madison WI, USA), and M-MuLV RT buffer (Promega). cDNA synthesis was performed for 1 h at 42°C. In the PCR step, the 5′UTR was amplifi ed by using 2 μL of cDNA in a 20-μL reaction volume, as described by Romero and Rotbart (19). The PCR products were examined by gel electrophoresis. Oligonucleotide primers for this assay (forward primer MD90, reverse primer MD91) fl ank a conserved nucleotide sequence in the 5′UTR of the enterovirus genome and amplify an expected product size of 154 bp.

HEV71-specifi c RT-PCR Assay
The HEV71-specifi c RT-PCR was performed as described (21,22) to provide rapid identifi cation of HEV71 in cell culture supernatants that were positive in the screening RT-PCR assay. First, strand cDNA was prepared as outlined above. In the PCR step, the VP1 gene was amplifi ed by using 2 μL of cDNA in a 20-μL reaction volume, as described (22). The PCR products were examined by gel electrophoresis. Oligonucleotide primers for this assay (forward primer MAS01S, reverse primer MAS02A) fl ank a region within the VP1 gene unique to HEV71 and amplify an expected product size of 376 bp.

Partial VP1 RT-PCR Assay
To identify HEV viruses that were not detected by the VP4 RT-PCR screening assay, a molecular serotyping method based on RT-PCR amplifi cation and sequencing of a portion of the VP1 gene was performed as described (23). An ≈340-bp fragment was amplifi ed by RT-PCR by using the forward primer 292 (5′-MIGCIGYIGARACNGG-3′) and reverse primer 222 (5′-CICCIGGIGGIAYRWACAT-3′), under conditions exactly as described by Oberste et al. (23). PCR products were examined by gel electrophoresis and purifi ed by using the GENECLEAN III kit (Qbiogene).

Nucleotide Sequencing of HEV71 VP4 and VP1 Gene Amplicons
Enterovirus VP4 gene amplicons were sequenced on both strands by using the PCR primers. HEV71 VP1 gene amplicons were sequenced on both strands by using the PCR primers and internal VP1 primers 161 and 162, described by Brown et al. (24). Sequencing was performed by using the Big Dye Cycle Sequencing kit version 3.0 and an ABI377 automated DNA sequencer (Applied Biosystems, Foster City, CA, USA). The SeqMan software module in the Lasergene suite of programs (DNASTAR, Madison, WI, USA) was used to format the nucleotide sequences. Partial VP1 and VP4 sequences for 173 HEV71 strains and 214 CVA16 strains have been submitted to the European Molecular Biology Laboratory database (partial VP1 gene accession nos. EU072122-EU072195; VP4 gene accession nos. EU051005-EU051317).

HEV71 VP1 Gene Nucleotide Sequence Data from GenBank
In addition to 23 VP1 gene sequences from HEV71 strains isolated in Vietnam, 26 VP1 gene nucleotide se-quences of HEV71 strains available in the GenBank database were included in this analysis, allowing the generation of a dendrogram containing 49 strains isolated between 1970 and 2005 ( Table 1). The strains used to reproduce the

Phylogenetic Analysis
VP1 and VP4 gene sequences were subjected to nucleotide-nucleotide BLAST analysis (blastn) by using the online server at the National Center for Biotechnology Information (www.ncbi.nlm.nih.gov/blast). Alignment of the 23 HEV71 complete VP1 gene sequences was undertaken by using the ClustalW program (25). A dendrogram was constructed by using the neighbor-joining method with PHYLIP version 3.5 (26) and drawn by using TreeView (27). Bootstrap analysis with 1,000 pseudoreplicates was performed by using the program Seqboot (28). Coxsackievirus A16 (CVA16), strain G10 (29), was used as an outgroup in the analysis.

Statistical Methods
Differences between proportions were tested by using the χ 2 test with Yates correction or Fisher exact test. Epi Info version 6 (Centers for Disease Control and Prevention, Atlanta, GA, USA) was used for the analysis.

Virus Isolation from HFMD Patients
An enterovirus was isolated from 411 (53.8%) of the 764 HFMD patients enrolled in the study. The number of CVA16, HEV71, and other enterovirus serotypes isolated from HFMD patients is presented in Table 2. CVA16 was identifi ed in 214 (52.1%) and HEV71 in 173 (42.1%) of the enterovirus-positive HFMD patients. Twenty-four (5.8%) enteroviruses of another serotype were also isolated from HFMD patients ( Table 2).
Procedures for the isolation and identifi cation of enterovirus strains obtained in the study are presented in a fl owchart (Figure 1). Of the 411 enteroviruses isolated in this study, 170 were identifi ed by using HEV71-specifi c primers. Another 3 were identifi ed as HEV71 when the VP4 and partial VP1 RT-PCR products were sequenced. We used the RT-PCR assay and sequencing of the VP4 gene as a screening tool because a single set of primers allowed us to obtain a preliminary identifi cation of HEV71 or CVA16. In our laboratory, 256 enterovirus isolates were sequenced in both VP1 and VP4, and 100% concordance was found between the VP1 and VP4 results for HEV71 (130 isolates) and CVA16 (61 isolates); only 28 (43%) of 65 other enteroviruses had concordant results in both the VP1 and VP4 sequences (unpub. data). Thus, 24 non-HEV71, non-CVA16 isolates were identifi ed as other enteroviruses.

Clinical Features of HFMD
The clinical features observed in HFMD patients enrolled in the study are presented in Figure 2,     The distribution of CVA16-and HEV71-associated HFMD cases by month during 2005 is presented in Figure 3, panel A. HFMD was identifi ed in southern Vietnam throughout the year; HEV71 and CVA16 were also isolated throughout the year. Two peaks of HFMD activity were observed during 2005. The fi rst peak occurred from March through May. CVA16 was the predominant virus during this time, accounting for 81.1% (116 cases) of HFMD compared to 18.9% (27 cases) for HEV71 (Figure 2, panel A). The second peak of HFMD activity occurred from September through December. HEV71 was the predominant virus during this time, accounting for 65.3% (128 cases) of HFMD compared to 34.7% (68 cases) for CVA16 ( Figure  3, panel A). Figure 4 depicts the geographic distribution of HFMD cases due to HEV71 (Figure 4, panel A) and CVA16 (Figure 4, panel B) who were brought for treatment to a major children's hospital in Ho Chi Minh City. Children admitted to this hospital are predominantly drawn from the urban area but were also referred from provinces surrounding Ho Chi Minh City.

Molecular Epidemiology of HEV71
The HEV71 isolates were further analyzed to determine the monthly distribution of viral subgenogroups in southern Vietnam during 2005 (Figure 3, panel B). This analysis was achieved by RT-PCR amplifi cation of complete VP4 and partial VP1 gene sequences, nucleotide sequencing, and BLAST analysis (20). Using these methods, we identifi ed 3 HEV71 subgenogroups, C1, C4, and a previously undescribed subgenogroup, C5. Two virus isolates (1.2%) belonging to subgenogroup C1 were identifi ed, 1 each in May and June. A total of 9 (5.2%) subgenogroup C4 strains were identifi ed; 7 were isolated from March through May and 1 each in October and November. Strains belonging to the new subgenogroup C5 (162 [93.6%]/173) were the predominant genetic lineage identifi ed in southern Vietnam during 2005. Subgenogroup C5 viruses were identifi ed in each month and were the primary cause of the large increase in HFMD from September through December.
Because we had identifi ed a putative new subgenogroup of HEV71 (C5) by analysis of complete VP4 and partial VP1 gene sequences (Figure 3, panel B), we conducted further nucleotide sequence analysis of the complete VP1 gene of 23 HEV71 isolates whose VP4 sequences were representative of all clusters observed in dendrograms generated from the screening data (9,24). Complete VP1 gene sequence analysis is considered the most rigorous method for determining the molecular phylogeny of HEV71 strains (6,24), and our analysis needed to be confi rmed with a subset of all the isolates ( Figure 5). We used previously published VP1 gene cDNA sequences to reconstruct the subgenogroup lineage structure of HEV71, fi rst identifi ed by Brown et al. (24) (Table 2).
Two of the Vietnamese HEV71 isolates clustered within subgenogroup C1; 5, within subgenogroup C4; and 16, within the new subgenogroup C5 ( Figure 5). The subgenogroup clustering of the HEV71 Vietnamese isolates is strongly supported by bootstrap analysis, which indicates that 3 independent genetic HEV71 lineages (C1, C4, and C5) circulated in southern Vietnam during 2005. This, together with the year-round isolation of CVA16 and HEV71 from HFMD patients (Figure 3, panels A, B), suggests that both viruses circulate endemically in southern Vietnam.

Discussion
To our knowledge, this study provides the fi rst comprehensive epidemiologic and virologic survey of HFMD, CVA16, and HEV71 infection in Vietnam. Similar to the situation in other countries, HEV71 infection was associated with a subset of HFMD cases in which acute neurologic disease developed. Our epidemiologic and phylogenetic data suggest that both CVA16 and HEV71 circulate endemically in southern Vietnam.
Nearly one third of the HEV71-associated HFMD cases identifi ed in our study were complicated by acute neurologic disease. The case-fatality rates of 1.7% in all identifi ed HEV71 infections and 5.9% in HEV71 acute neurologic disease cases are higher than those observed in other studies (7,30,31). However, the case-fatality rates calculated in our study may overestimate the true values because only HFMD patients who were brought for treatment at a major children's hospital were included in the study. The best estimates of case-fatality rates for HEV71 infection have come from a large seroepidemiologic study of the 1998 HFMD epidemic in Taiwan (32); the authors estimated a case-fatality rate of 96.96 per 100,000 population in infants <1 year of age, declining to 6.64 per 100,000 population in children >5 years of age. To rigorously determine the incidence and case-fatality rate of HEV71 infection in southern Vietnam, a similar population-based seroepidemiologic study should be undertaken.
Although cases of HFMD were identifi ed throughout the year, 2 periods of increased prevalence were identifi ed-from March through May and from September through December. In southern Vietnam, these months are interim periods between the dry and wet seasons. CVA16 was the predominant virus isolated in the fi rst period, and HEV71 infection was the predominant virus isolated in the second period. Ongoing epidemiologic surveillance will be necessary to determine whether this pattern of HFMD and enterovirus activity recurs in a regular annual cycle.
Phylogenetic analysis based on nucleotide sequence alignment of the complete VP1 gene of 23 representative strains of HEV71 from southern Vietnam showed that they belonged to 3 subgenogroups, C1, C4, and to the previously undescribed subgenogroup C5. Since 1997, 2 genetically distinct major lineages (B, C) of HEV71 have circulated in different parts of the Asia-Pacifi c region (6,9). Viruses belonging to genogroup B have predominated in Southeast Asia, whereas viruses belonging to genogroup C have predominated in northern Asia (6,9,11,33). Before 1997, HEV71 strains belonging to subgenogroup C1 were identifi ed in several small outbreaks around the world (15,24). Since 1997, subgenogroup C1 viruses have circulated endemically in the Asia-Pacifi c region and have been found to cocirculate as a minor subgenogroup together with a predominant HEV71 subgenogroup during several outbreaks (6,11,34). In this study, subgenogroup C1 viruses comprised only 1.1% of HEV71 strains isolated, indicating low-level circulation. Viruses belonging to subgenogroup C2 have circulated widely in the Asia-Pacifi c region between 1998 and 2000 (9,11,16) and were responsible for the large outbreak in Taiwan in 1998 (6,8,9,33). Two new genetic lineages of genogroup C, subgenogroups C3 and C4, have emerged recently in northern Asia. Viruses belonging to subgenogroup C3 fi rst appeared in the People's Republic of China in 1998 (6) and reemerged in South Korea in 2000 (6,9). Viruses belonging to subgenogroup C4 were fi rst identifi ed in the People's Republic of China in 1998 and again in 2000 (35)  Our data indicate that the molecular epidemiology of HEV71 in southern Vietnam conforms to the northern Asian epidemiologic pattern of endemic circulation of genogroup C virus strains, with evidence of the ongoing evolution of new subgenogroups, similar to that observed for genogroup B HEV71 strains in Southeast Asia (6,9,33). Furthermore, the year-round isolation and circulation of multiple independent genetic lineages of HEV71 (36) suggest that this virus circulates endemically within the human population of southern Vietnam.
In conclusion, this study has established that HEV71 circulates endemically in southern Vietnam and thus represents a substantial threat to the health of children in this region. Improvements in public sanitation and personal hygiene alone are unlikely to prevent HEV71 transmission within the community. A vaccine is necessary to prevent HEV71-induced neurologic disease in susceptible children. However, until such a vaccine is available, virus activity in the community must be monitored through the establishment and maintenance of sentinel surveillance.