TRANSACTIONSOFTHEROYALSOCIETYOFTROPICALMEDICINEANDHYGIENE(1999)93,255-260 The unique riverine ecology of hepatitis E virus transmission in South-East Asia Tarot Winarno4, Maid: yP.Putri’,KantiLaras’, Andrew L Corwin’, Ngc lven _ T K Tien2 Khanthonn ~~_Bounlu3 ~ ~~~~ l~-.Bia P Larasati’, Nono Sukri’, Timothy’Endy’, H A Sulaiman6 and Kenneth C Hyams’ I US NavalMedical Research Unit No 2, Jakarta, Indonesia; 2Pasteur Institute, Ho ChiMinh City, Viet Nam; ‘Center-for National Laborato y and Epidemiology, Ministry of Health, Lao PDR; 4Provincial Health Service, Sintang District, West Kalimantan, Indonesia; ‘Armed Forces Research Institute of Medical Sciences, Bangkok, Thailand; 6Medical Faculty, University of Indonesia, Jakarta, Indonesia; 7Naval Medical Research Institute, Bethesda, Ma yland, USA Abstract The ecology of hepatitis E virus (HEV) transmission in South-East Asia was assessedfrom a review of published and unpublished NAMRU-2 reports of hepatitis outbreak investigations, cross-sectional prevalence studies, and hospital-based case-control studies Findings from Indonesia and Viet Nam show epidemic foci centred in jungle, riverine environments In contrast, few casesof acute, clinical hepatitis from cities in Indonesia, Viet Nam and Laos could be attributed to HEV When communities in Indonesia were grouped into areas of low (60%) prevalence of anti-HEV antibodies, usesof river water for drinking and cooking, personal washing, and human excreta disposal were all significantly associatedwith high prevalence of infection Conversely, boiling of river drinking water was negatively associatedwith higher prevalence (P < 0.01) The protective value ofboiling river water was also shown in sporadic HEV transmission in Indonesia and in epidemic and sporadic spread in Viet Nam Evidence from Indonesia indicated that the decreaseddilution of HEV in river water due to unusually dry weather contributed to risk of epidemic HEV transmission But river flooding conditions and contamination added to the risk of HEV infection in Viet Nam These findings attest to a unique combination of ecological and environmental conditions predisposing to epidemic HEV spread in South-East Asia Keywords: hepatitisE virus, hepatitisE virus antibodies,prevalence,ecology,epidemiology,South-EastAsia,Indonesia, Viet Nam, Laos Introduction Hepatitis E virus (HEV) transmission occurs predominantly in the developing world First identified during the 1956 outbreak in India, involving over 29 000 cases, epidemic and sporadic HEV infections have since been recognized in many developing regions except for South America, although antibody to the virus was recently detectedfromBrazil (VISWANATHAN,~~~~~;BALAYAN, ~~~O;BRADLEY, 1992; PANG etal., 1995; PARANA etal., 1997) In China, some 120 000 cases of HEV were associated with an epidemic in Xinjiang during 198688 (AYE et al, 1992) In South-East Asia, HEV in epidemic form has been reported from Myamnar, Viet NamandIndonesia (BALAYAN,~~~O;BRADLEY,~~~~; CORWIN et al., 1996a; MAST et al., 1996) HEV infections reuorted from develoued countries in contrast are generally acquired from tiavel-related exposures Only l-2% of US blood donors have been found to be HEV-seroreactive, and few infections have been documented in Europe In the USA and Singapore, acute, clinically recognized hepatitis E has been attributedtoforeigntravel (BALAYAN,~~~O;BRADLEY,~~~~; MAST et al., 19%‘) HEV is a small, single-stranded RNA virus similar in structure to the caliciviruses HEV spread is generally water-borne, particularly in epidemics Unlike hepatitis A virus (HAV), there is little evidence to suggestpersonto-person transmission Because HEV is principally spread by human and possibly animal faecal contamination of water resources, transmission is associated with poor water-related hygiene and sanitary conditions (VISWANATHAN,~~~~~;KANE etaZ., 1984; B~y~N,1990; BRADLEY, 1992; MAST&ALTER, 1993).Alsoin 1991; 'BRAD&Y, PREVENTION, This review, mostly of the literature as well as a few unpublished research findings from studies conducted or supported by the US Naval Medical ResearchUnit No (NAMRU-2) in Jakarta, Indonesia, is intended to show that in South-East Asia the epidemic form of HEV transmission may be a function of a unique, predominantly rural, riverine ecology The data also indicate that extremes of rainfall conditions, both flooding and drought, combined with a background of specific river usage practices, contribute to the risk of HEV transmission Methods and Analysis Examules of both enidemic and sooradic HEV transmission ‘were identifieh from our inbestigations of rural and urban settings in South-East Asia First, recognized foci of epidemic HEV transmission in jungle, riverine areas were investigated in Indonesian Borneo (West Kalimantan)and Eastern Java (Bondowoso), and in the An Giang Province of the Mekong River Delta region of Viet Nam which borders Cambodia (CORWIN et al., 1995,1996a, 1997; unpublished data; HAU et al., 1999) Second, the importance of HEV in acute, sporadic hepatitis was examined in a multi-hospital-based study of HEV that included the cities of Takarta Indonesia (unpublished data), Hanoi and Ho C& Ming, Viet Nam (CORWIN et al., 1996b) and Vientiane, Lao PDR (BO~NLU et al., 1998) Data from studies conducted in Indonesia, Viet Nam, and Laos were included in the analysis con- trast to HAV, HEV is characterized by (i) a longer incubation period ofup to weeks, (ii) prolonged clinical course of illness, and (iii) poor protective value of serum immune globulin Particularly notable and unique relative to HAV is the high case-fatality rate (CFR) in pregnant women (1O-24%) associated with H&V ir&ections KI-IUROO ~~~O;DECOCK etal 1987: BALAYAN 1990;‘ REYES 8( BA&JDY, CENTERS FOR DISEASE CONTROLAND 1993;MAsT&AL~~~,1993) 1942, Correspondingauthor: CDR Andrew Lee Corwin, US NAMRU-2, Box Unit 8132, APO AP96520-8132; phone +62 21 421 4457 to 4463, fax +62 21424 4507, e-mail corwin@smtp.namru:!.go.id Epidemiological approach Three approaches were used to investigate HEV transmission in reported studies: (i) outbreak investigations; (ii) cross-sectional prevalence studies; and (iii) hospital-based, case-control studies Data pertaining to demographics, the environment, medical history, and risk-related behaviours had been obtained bv use of a standardized questionnaire administered by trained interviewers Except for hosnital-based studies and HEV outbreak investigation described from Vie; Nam, the household served as the principal sampling unit Outbreak investigation In Indonesia, an investigation of a 1991 HEV outbreak was carried out among com- ANDREW L 256 munities on the Inggar River of Borneo, providing a retrospective epidemiological evaluation (CORWINet al., 1995) Another outbreak of hepatitis, that affected rural villages along the Balut River in eastern Java, was investigated in 1998, and subsequently attributed to HEV (unpublished data) From Viet Nam, an HEV outbreak investigation was conducted in 1994 among communities along a 60-km stretch of the Hau River, a tributary of the Mekong River (CORWIN et al., 1996a) Cross-sectional seroprevalence studies Communities in Indonesian Borneo (contiguous to the 1991 outbreak area described above) located along a 150~km stretch of the Kapuas River (into which the Inggar River flows) were surveyed in 1995 Targeted for study purposes was the affected population from a 1987 HEV outbreak where HEV was first recognized in Indonesia For comparison a population further downstream of the Kapuas River that was more urban was studied (CORWIN et al., 1997) In the Mekong Delta river region of Viet Nam, the serourevalence of anti-HEV antibodies was studied in communities surveyed from the area adjacent to the foci of epidemic HEV transmission in 1994 (HAU et aZ., 1999) Hospital-based, case-control studies of acute hepatitis A standardized case-control design and data collection instrument were used to evaluate the HEV infection in acute, clinically recognized hepatitis, from urban centres (unpublished data; CORWIN et aZ., 1996b; unpublished data; BOUNLU et aZ., 1998) in South-East Asia countries: Indonesia, Viet Nam, and Laos The duration of each study was 12 consecutive months: the first study began in 1993 and the last in 1997 Cases (175-200) were selected on the basis of clinical criteria for suspected acute hepatitis, and controls were selected to match the demographic characteristics of cases I Serological tests Laboratory testing was coordinated by the US NAMRU-2, Jakarta, Indonesia Serawere tested by commercial enzyme immunosorbent assayfor IgG antibody to HEV (Abbott Laboratories, Abbott Park, IL) and IgGantibody to HAV (Abbott Laboratories) Serafrom acute hepatitis studiesalsoweretestedforIgMantibodytoHAVandIgM antibody to hepatitis B core antigen (HBcAg) (Abbott Laboratories) Sera from patients with signs and symptoms of acute hepatitis that were positive for IgG antibody to HEV were further tested for IgM antibody to HEV (Genelabs Diagnostics PTE, Singapore) n Fig Geographical Specimens found positive for IgG antibody to HEV were also tested by polymerase chain reaction (PCR) at the Armed Forces Research Institute of Medical Sciences (AFRIMS), Bangkok, Thailand (TAM et al., 199 1) The oligonucleotide primers used in HEV PCR were HEV I’4653 (5’-TTT-TCA-GGT-GGC-TGC-C3’) and HEV P4874 (5’-GGG-CCC-CAA-TTC-TTCT-3’) for RT-PCR; and HEV I’4704 (5’-ATA-CCGTCA-GAG-CCC-AGG-3’) and HEV I?4845 (5’-GGCGAA-GCG-CAC-GAC-ATC-3’) for nested PCR Reliability evaluation using positive and negative controls conducted at AFRIMS indicated high test sensitivity and specificity (unpublished data) Additionally, IgM antibody to HEV was detected from sera using Western blotting as previously described (HE et aZ., 1993) HEV studies in South-East Indonesia Asia A 199 HEV outbreak in Sintang, West Kalimantan, affected over 2500 neonle An attack rate of 90 cases/1000 population and case-fatality rate (CFR) of 14% among pregnant women were estimated (CORWIN et al., 1995) Notable was the relatively high seroprevalence of IgG antibody to HEV years following this outbreak Overall, anti-HEV prevalence was 59% among 445 study subjects There was no significant difference (P > 0.05) in HEVprevalence between cases(72%) with a history of acute jaundice and subjects without jaundice (61%) Similarly, the prevalence of anti-HEV years after the 1987 epidemic was 50% in an area adjacent to the 1991 outbreak (CORWIN et al 1997) (Fig 1) Finally, ‘first time’ epidemic HEV transmission h Indonesia outside of Borneo was recognized March/April 1998 A preliminary attack rate of 13% was estimated, ranging by community from 10% to 19% (unpublished data) However 415 (43%) of the 962 studv subiects surveyed were positive for’IgG antibody to*HEV, of which 49% had serological evidence of a recent infection based on RT-PCR testing In the area of the 1987 West Kalimantan outbreak, continuinn (snoradic) HEV transmission was found bv comparini I&V infection in the population living (aged 27 years: prevalence 53%) during the epidemic with young children born after the event (aged positive family member There was no anoarent increased risk of HEV infection associated w&*&creased family size (CORwIN et al., 1995) Finally, findings from a survey conducted in 1994 from West Kalimantan region of the 1987 outbreak provided a more complete picture of transmission resulting from both epidemic and sporadic HEV transmission: 94% of 178 households in the outbreak area had > family member positive for IgG antibody to HAV compared with 52% for IgG antibody to HEV (P < 0.0001) (CORWIN et aZ., 1997) In urban Jakarta, prevalences of background HEV infections (positive for IgG antibody to HEV) among cases and control subjects (non-jaundiced, hospital inpatients) were very low, and differed little: 5%.and 2% resnectivelv (urmublished data) No data are available for-IgM an&odi to HEV (Tadle) Indo-china (Viet Nam and Laos) Another study of epidemic HEV transmission in South-East Asia involved an outbreak during 1994 in An Giang Province, Viet Nam (Fig 1) IgG antibody to HEV was recognized in 76% of cases compared with 36% of matched community and geographical controls (P < 0.001) As in Indonesia, there was no evidence of familial clustering of seropositive individuals in Viet Nam (CORWIN et al., 1996a) Data presented in Table show low prevalence of antiHEV antibodies in acute, clinically recognized hepatitis from urban settings In Hanoi and Vientiane, 2% of 375 and 4% of 52 acute episodes of suspected hepatitis, respectively, were attributed to recent HEV infections (CORWIN~~~Z., 1996b;BOUNLUetaL, 1998).Noserum sample from hospitalized case subjects in Ho Chi Minh or Jakarta was positive for IgM antibody to HEV (unpublished data) Background IgG antibody to HEV, reflecting previous infections, varied little between acute jaundiced case and control populations at each study location in Indo-china Notable was the high proportion of controls positive for IgG antibody to HEV similar to cases, from Hanoi and Vientiane: 14% ofco&rolsvs 12% of&es and 17% ofcontrolsvs 16% of cases, respectively (CORWIN et al., 1996b; BOIJNLU et al., 1998) Role of river ecology Indonesia Water use related to hygiene and sanitation was clearly associated with the risk of HEV infection In the 1991 West Kalimantan outbreak area, HEV prevalence increased with the usage of river water for drinking and Table Study of acute clinical East Asia Locality Vientiane Jakarta IgG anti-HEV 16% hepatitis IgG anti-HEV” E in South- 16% (n $16) Hanoi (n = ;21) 14% Ho Chi Minh City (n 187) (n = 6) IgM anti-HEV ‘In the absence of IgM anti-HAV and IgG anti-HBc bNot available cooking (P < O.OOl), bathing (P < O.OOOl), and excrement disposal (P< 0.001; Fig 2) Conversely, the prevalence of anti-HEV antibodies decreased as the practice of boiling drinking water increased (P = 0.02; Fig 2) (CORWIN et al, 1995) Adding to the risk of epidemic and sporadic HEV transmission in the 1987 outbreak affected and comparison study areas of West Kalimantan were the practices of bathing, human waste disposal, drinking, and laundering Usage of river water for bathing and human waste disposal was significantly associated with positivity for IgG anti-HEV antibodies in both outbreak and comparison areas In the comparison area, drinking and washing with river water were found to be strongly associated (P