Tropical Medicine and International Health doi:10.1111/j.1365-3156.2008.02038.x volume 13 no 2 pp 286–290 february 2008 Desiccated vector mosquitoes used for the surveillance of Japanese encephalitis virus activity in endemic southern India S. C. Tewari, V. Thenmozhi, N. Arunachalam, P. Philip Samuel and B. K. Tyagi Centre for Research in Medical Entomology, Madurai, India Summary To monitor Japanese encephalitis virus (JEV) activity in endemic areas of Tamil Nadu, southern India, desiccated vector mosquitoes were screened for JEV antigen using ELISA, from 1996. A total of 133 233 specimens from eight index villages comprising 2816 pools (mainly Culex vishnui subgroup) were tested. Of these, 59 pools (2.1%) were positive for JEV antigen. Control measures were undertaken in positive villages accordingly. The average annual minimum infection rate was 0.8 at the beginning of the study and remained lower for nearly 8 years. A declining trend in JE cases was recorded. keywords desiccated mosquitoes, Culex vishnui subgroup, surveillance, Japanese encephalitis virus Introduction Japanese encephalitis (JE) is a mosquito-transmitted arboviral disease mainly affecting children <15 years of age. It is a major public health problem worldwide because of its high mortality and morbidity coupled with neurological sequelae during epidemics (Solomon 1997; Gajanana 1998). The epidemiology of JE is complex due to involvement of several vertebrate and invertebrate hosts. JE virus (JEV) is transmitted naturally between ardeid birds and pigs by mosquito species belonging to the Culex vishnui subgroup, comprising Culex tritaeniorhynchus, Culex vishnui and Culex pseudovishnui, in India (Reuben et al. 1994). Domestic pigs are considered as amplifying, humans as accidental and cattle as dead end hosts (Reuben et al. 1992; Solomon et al. 2000). A sound surveillance system is an integral part of arboviral disease control programmes. At present, surveillance of JE in endemic areas is mainly based on clinical case reporting and monitoring abundance of vector mosquitoes. JEV activity has been monitored using pigs (Thenmozhi et al.1991), donkeys (Mani et al. 1991) and goats (Rajendran et al. 2003) as sentinel animals. As the pig is an amplifying host for JEV, the introduction of non-immune animals to endemic areas as sentinels is unethical. Further, the pig may serve as a sensitive indicator of JEV activity in a given area, but may not be a good predictor of the risk to humans (Peiris et al. 1993). Donkeys are rarely available in the villages and may die of JEV infection. Vector infection and abundance were found to be good indicators of JE occurrence in surveillance studies con286 ducted in South India (Gajanana et al. 1997). The antigencapture enzyme-linked immunosorbent assay (ELISA) is a rapid and sensitive method for the detection of JEV infection in wild-caught vector mosquitoes. The assay was validated in large-scale field trials in southern India (Gajanana et al. 1997). Hence, it was considered useful for surveillance of JEV activity in endemic areas, but transportation of field-collected live specimens under cold-chain conditions to the central laboratory was a major limitation. Peiris et al. (1993) found JEV antigen in dead mosquitoes by ELISA. Tewari et al. (1999) applied this observation to develop an improved surveillance system in nine districts of Tamil Nadu from March 1996 to February 1999 (Tewari et al. 1999). The study was extended for 5 years and the consolidated results of the study from March 1996 until December 2004 are presented in this paper. Materials and methods The details of the study area and the methods employed are described elsewhere (Tewari et al. 1999). Briefly, the (CRME) formed a JE surveillance network with the Zonal Entomological Team (ZET) of the Department of Public Health & Preventive Medicine, Goverment of Tamil Nadu, of nine districts for JE. Of these, eight districts namely Tiruchirapalli, Virudhunagar, Dindigul, Thirunelvelli, Cuddalore, Vellore, Coimbatore and Salem are endemic, whereas one district namely Thanjavur is non-endemic. The endemic districts had reported JE human cases at least once in 3 years; the non-endemic district had reported none in the past 10 years. One village with at least one case or ª 2008 Blackwell Publishing Ltd Tropical Medicine and International Health volume 13 no 2 pp 286–290 february 2008 S. C. Tewari et al. JEV surveillance in southern India death due to JE in the past 3 years in each district was selected as the index village. Every 3 years, the allocation of index villages was reviewed and if no case ⁄ death was reported consecutively for 3 years, the index villages used to be shifted to another nearby one where cases occurred recently. Specimens were received throughout the year as ZET visits each index village monthly for the longitudinal study to collect specimens. Field staff of the ZETs collected the mosquitoes at dusk and after identification, desiccation and pooling by species (7–50 ⁄ pool), put them in polythene packets and posted them to the laboratory in CRME, Madurai. In CRME, the vectors were re-identified. Mutilated specimens which could not be identified to species level were pooled and classified as Cx. vishnui subgroup. Subsequently each mosquito pool was screened by antigen-capture ELISA (Gajanana et al. 1995), and JEV identified by inhibition ELISA (Tsai et al.1987). Results of positive pools were communicated to the health authorities within a fortnight of receipt of the mosquito vector samples. The infection in vector was recorded as the minimum infection rate (MIR; MIR = [No. of positive pools ⁄ Total no. of specimens tested] · 1000). Statistical analysis was carried out using SPSS 11.5. Results From March 1996 to December 2004, a total of 2816 pools (133 233 mosquito specimens) were received from nine ZETs. A total of 429 pools (21 005 specimen) were Cx. tritaeniorhynchus, 61 pools (2787 specimen) were Cx. vishnui, three pools (120 specimen) were Cx. pseudovishnui, 2142 pools (101 707 mutilated specimens) were Cx. vishnui subgroup, 177 pools (7485 specimen) were Culex gelidus, three pools (104 specimen) were Culex fuscocephala and one pool (25 specimen) was Culex bitaeniorhynchus. Fifty-nine pools (Cx. tritaeniorhynchus 13, Cx. vishnui 1, Cx. vishnui subgroup 41 and Cx. gelidus 4) tested positive for JEV antigen by ELISA (Table 1). Among the 41 positive pools of Cx. vishnui subgroup, the probability would be maximum number of Cx. tritaeniorhynchus pools, although the number of pools was lower in the beginning of the study as implementation of the programme took time to gear up the peripheral health workers. Most of the positive pools were recorded during JE transmission season, and at least one positive pool was recorded each year (except 1998) with the highest number of positive pools (27) in 2004. The annual MIR was 0.8 at the beginning of the study (1996) and remained lower than this value for nearly 8 years (up to 2004). There was no correlation between MIR and JE cases during the study period (Table 1). In 2003 (v2 11.42, P < 0.001), ª 2008 Blackwell Publishing Ltd Table 1 Japanese encephalitis virus antigen detected in desiccated vector mosquitoes and human JE cases recorded in Tamil Nadu state, 1996–2004 Year 1996 1997 1998 1999 2000 2001 2002 2003 2004 Total No. pools tested (mosquitoes specimens) 25 145 79 202 399 410 316 464 776 (1250) (6880) (3784) (9656) (18391) (18515) (14363) (21711) (38683) 2816 (1,33,233) No. positive pools ⁄ minimum infection rate (MIR) No. of cases (death)* 1 ⁄ 0.8 2 ⁄ 0.29 0⁄0 5 ⁄ 0.52 2 ⁄ 0.11 11 ⁄ 0.59 2 ⁄ 0.14 9 ⁄ 0.41 27 ⁄ 0.70 111 89 25 14 4 0 0 163 19 59 ⁄ 0.44 425 (132) (53) (42) (14) (5) (0) (9) (9) *Courtesy of Annual Report Min. Hlth. Govt. of India. there was a sudden rise in MIR (1.3) in the non-endemic zone Thanjavur, indicating an increase in the circulation of JEV (Table 2). However, no human case was reported in this area during the study period, and there was no significant difference (v2 = 0.05 P = 0.94 using Fisher’s exact test) in overall infection rate (MIR) between nonendemic and endemic areas between 2000 and 2004. Among the endemic zones, two zones, Tiruchirapalli and Cuddalore, were relatively more viraemic than the other endemic zones, as 90.24% (37 ⁄ 41) of positives were recorded from these two areas (Figure 1). Discussion This study indicated that the existing infrastructure with the health department could be effectively used for surveillance of JEV activity in endemic areas. ELISA is particularly useful for mass screening of mosquitoes quickly and communicating results to field staff. JEV antigen was detected in 59 pools of vector mosquitoes. Of these, 55 belonged to Cx. vishnui subgroup in which Cx. tritaeniorhynchus is a dominant species. Field-collected adults of Cx. vishnui subgroup are difficult to identify; nearly 88% could be identified up to species status (Reuben 1969). In this study, we received for analysis, species which had been transported via surface mail, which increases the mutilation of specimens. Hence, prior to ELISA test while re-identification most of the specimens were pooled as Cx. vishnui subgroup (mixed specimens of Cx. tritaeniorhynchus, Cx. vishnui and Cx. pseudovishnui). In these mixed pools of Vishnui subgroup, Cx. tritaeniorhynchus would be dominant, as studies in southern India already demonstrated that Cx. tritaeniorhynchus is a dominant species in rural environment of paddy 287 Tropical Medicine and International Health volume 13 no 2 pp 286–290 february 2008 S. C. Tewari et al. JEV surveillance in southern India Table 2 Comparison of JEV activity in non-endemic and endemic study areas (2000–2004) Places ⁄ year Non-endemic zone (1) Mosquitoes specimens ⁄ no. pools tested Positive pools MIR Endemic zones (8) Mosquitoes specimens ⁄ no. pools tested Positive pools MIR 2000 2001 2002 2003 2004 Total 3813 ⁄ 83 4125 ⁄ 90 4879 ⁄ 100 5400 ⁄ 113 6005 ⁄ 123 24222 ⁄ 509 0 0 0 0 0 0 7 1.3 3 0.5 10 0.41 14578 ⁄ 316 14545 ⁄ 319 6484 ⁄ 216 16251 ⁄ 351 32678 ⁄ 653 84536 ⁄ 1855 2 0.14 11 0.76 2 0.31 2 0.12 24 0.73 41 0.49 MIR, minimum infection rate; JEV, Japanese encephalitis virus. ecosystem where it was found ranging between 55 and 78% (Reuben 1971; Mani et al. 1991; Gajanana et al. 1997; Sunish & Reuben 2001). The JEV-antigen detection in desiccated vector pools in this study confirms that Cx. tritaeniorhynchus is the main vector of JEV in the study areas (Gajanana et al. 1997). However, in south India, all the three species had been identified as vectors of JE (Reuben et al. 1988, 1994). There was no significant difference (v2 = 0.05 p 0.94, Fisher’s exact test) in overall infection rate (MIR) between non-endemic and endemic areas, indicating occurrence of silent JEV transmission in the non-endemic Thanjavur zone. This phenomenon was also recorded earlier in a serological investigation of children, 1991–1993 (Vijayarani & Gajanana 2000). Hence, it confirms that the rural population of the non-endemic zone is getting sublethal doses of JEV. There has been a decline in deaths ⁄ cases due to JE in Tamil Nadu since 1999 onwards. However, JEV activity continued, although no case was reported in 2001 and 2002. This may be because of underreporting as laboratory diagnostic facilities are inadequate in the existing health system. In hospital-based studies in south India, 50% of suspected encephalitis cases were confirmed as JE (Gajanana et al. 1996; Kabilan et al. 2004). Another possibility is that the natural immunity level of the endemic population has increased due exposure to infective bites of JE vectors, which was estimated to be 0.53 from Cuddalore district (Gajanana et al. 1997). The JEV antigen was mainly detected from the collections made between August and January, which is the JE transmission season in southern India (Gajanana et al. 1997). Cuddalore and Tiruchirapalli, situated close to each other (Figure 1), were found to be the epicentre of JE in Tamil Nadu. There was a significant rise in MIR (1.3) in the hitherto non-endemic area, viz. Thanjavur district, which remains free of human JE cases (Vijayarani & Gajanana 2000). The 288 reasons would be (i) the characteristic of zoonotic diseases which do not spill over to humans, although JEV circulation in nature continued in vector and hosts; ii) a classical longitudinal study from 1991 to 1993 in Thanjavur area showed that the ratio of pig (amplifying host) to cattle (dumping host) was 1:400, whereas in the neighbouring JE endemic area it was 1:4 (Vijayarani & Gajanana 2000). Species of Cx. vishnui subgroup are zoophilic and JEV does not multiply in cattle. Hence JEV did not spill over to humans (Ilkal et al. 1988). The ELISA technology has many advantages over the conventional sentinel animal method to monitor JEV activity. It does not require a cold-chain for the preservation and transportation of vector mosquitoes from peripheral rural health centres to the central laboratory. Screening vector mosquito pools for JEV infection by using JE-antigen capture ELISA was the most rapid system compared to other highly sensitive systems such as insect bioassay (Toxo-IFA system) or IIF, which are cumbersome, time-consuming (Gajanana et al. 1995) and thus unsuitable for routine surveillance programmes where results must be disseminated to operational organizations within a week. This improved surveillance networking system was effective in identifying high-risk areas to initiate appropriate control measures and could be implemented elsewhere in JE prone areas in the country to avoid impending outbreaks. Acknowledgements We thank the Director General of the Indian Council of Medical Research for providing the required facilities and for his consistent encouragement. We also gratefully acknowledge the cooperation by the Director, Department of Public Health & Preventive Medicine (DPH&PM), Tamil Nadu. We are deeply grateful to Dr. N.C Appavoo (former Director DPH&PM, Tamil Nadu) and ª 2008 Blackwell Publishing Ltd Tropical Medicine and International Health volume 13 no 2 pp 286–290 february 2008 S. C. Tewari et al. JEV surveillance in southern India N INDIA H ADES A PR R ANDH Vellore A KARNATAK 24 Salem 2 Cuddalore im Co re to ba li 13 pal a r 11 i h uc Thanjavur r i T ALA KER 1 Dindigul 2 Virudhunagar Number of positive pools 6 Tirunelveli Figure 1 Japanese encephalitis virus activity in the different entomological zones of Tamil Nadu, India. Dr. A. Gajanana (Former Officer in-Charge, CRME) who initially formulated and implemented this surveillance network programme, and also critically reviewed the ª 2008 Blackwell Publishing Ltd manuscript. We thank the Centre for Disease Control, Fort Collins, USA for providing JEV monoclonal antibodies. We appreciate the excellent technical assistance rendered by 289 Tropical Medicine and International Health volume 13 no 2 pp 286–290 february 2008 S. C. Tewari et al. JEV surveillance in southern India staff members of CRME, Madurai Culex ecology and Serology sections and Mr. A. Venkatesh for desktop publishing and illustration. References Gajanana A (1998) Epidemiology and surveillance of Japanese encephalitis in Tamil Nadu. ICMR Bulletin 28, 33–37. Gajanana A, Rajendran R, Thenmozhi V, Samuel PP, Tsai TF & Reuben R (1995) Comparative evaluation of bioassay and ELISA for detection of Japanese encephalitis virus in field collected mosquitoes. Southeast Asian Journal of Tropical Medicine and Public Health 26, 91–97. Gajanana A, Philip Samuel P, Thenmozhi V & Rajendran R (1996) An appraisal of some recent diagnostic assays for Japanese encephalitis. Southeast Asian Journal of Tropical Medicine and Public Health 27, 673–679. Gajanana A, Rajendran R, Philip Samuel P et al. (1997) Japanese encephalitis in South Arcot district, Tamil Nadu: A three-year longitudinal study of vector abundance and infection frequency. Journal of Medical Entomology 34, 651–659. Ilkal MA, Dhanda V, Rao BU et al. (1988) Absence of viraemia in cattle after experimental infection with Japanese encephalitis virus. Transactions of the Royal Society of Tropical Medicine and Hygiene 82, 628–631. Kabilan L, Vrati S, Ramesh S et al. (2004) Japanese encephalitis virus (JEV) is an important cause of encephalitis among children in Cuddalore district, Tamil Nadu, India. Journal of Clinical Virology 31, 153–159. Mani TR, Mohan Rao CVR, Rajendran R et al. (1991) Surveillance for Japanese encephalitis in villages near Madurai, Tamil Nadu, India. Transactions of the Royal Society of Tropical Medicine and Hygiene 85, 287–291. Peiris JSM, Amerasinghe FP, Arunagiri CK et al. (1993) Japanese encephalitis in Sri Lanka: comparison of vector and virus ecology in different agro-climatic areas. Transactions of the Royal Society of Tropical Medicine and Hygiene 87, 541–548. Rajendran R, Thenmozhi V, Tewari SC et al. (2003) Longitudinal studies in south Indian villages on Japanese encephalitis virus infection in mosquitoes and seroconversion in goats. Tropical Medicine and International Health 8, 174–181. Reuben R (1969) A redescription of Culex vishnui Colles and Culex tritaeniorhynchus Giles, from southern India. Bulletin of Entomological Research 58, 643–652. Reuben R (1971) Studies on the mosquitoes of North Arcot District, Madras state, India. Part 5. Breeding places of the Culex vishnui group of species. Journal of Medical Entomology 8, 363–366. Reuben R, Kaul H & Soman RS (1988) Mosquitoes of arboviral importance in India. Mosquito Borne Diseases Bulletin 5, 48–54. Reuben R, Thenmozhi V, Samuel PP, Gajanana A & Mani TR (1992) Mosquito blood feeding patterns as a factor in the epidemiology of Japanese encephalitis in southern India. American Journal of Tropical Medicine & Hygiene 46, 654–663. Reuben R, Tewari SC, Hiriyan J & Akiyama J (1994) Illustrated keys to species of Culex (Culex) associated with Japanese encephalitis in Southeast Asia (Diptera: Culicidae). Mosquito Systematics 26, 75–96. Solomon T (1997) Viral encephalitis in southeast Asia. Neurological Infections and Epidemiology 2, 191–199. Solomon T, Dung NM, Keen R, Gainsborough M, Wvaughn D & Khanh VT (2000) Japanese encephalitis. Journal of Neurosurgery and Psychiatry 68, 405–415. Sunish IP & Reuben R (2001) Factors influencing the abundance of Japanese encephalitis vector in rice fields in India- 1 Abiotic. Medical Veterinary Entomology 15, 381–392. Tewari SC, Thenmozhi V, Rajendran R, Appavoo NC & Gajanana A (1999) Detection of Japanese encephalitis virus antigen in desiccated mosquitoes: an improved surveillance system. Transactions of the Royal Society of Tropical Medicine and Hygiene 93, 525–526. Thenmozhi V, Samuel PP, Gajanana A & Reuben R (1991) Temporal relationship of swine infection of Japanese encephalitis virus to human cases in a Japanese encephalitis endemic area of Tamil Nadu. Virus Information Exchange Newsletter 8, 64. Tsai TF, Bolin RA, Montoya M et al. (1987) Detection of St Louis encephalitis virus antigen in mosquitoes by capture enzyme immunoassay. Journal of Clinical Microbiology 25, 370–376. Vijayarani H & Gajanana A (2000) Low rate of Japanese encephalitis infection in rural children in Thanjavur district (Tamil Nadu), an area with extensive paddy cultivation. Indian Journal of Medical Research 111, 212–214. Corresponding Author S.C. Tewari, Centre for Research in Medical Entomology (Indian Council of Medical Research), 4 Sarojini Street, Chinna Chokkikulam, Madurai 625 002, Tamil Nadu, India. Tel.: 91 452 2531430; Fax: 91 452 2530660; E-mail: crmeicmr@icmr.org.in 290 ª 2008 Blackwell Publishing Ltd