Seroprevalence of Chikungunya Virus (CHIKV) Infection ... - CiteSeerX

Am. J. Trop. Med. Hyg., 78(2), 2008, pp. 333–337 Copyright © 2008 by The American Society of Tropical Medicine and Hygiene

Seroprevalence of Chikungunya Virus (CHIKV) Infection on Lamu Island, Kenya, October 2004 Kibet Sergon, Charles Njuguna, Rosalia Kalani, Victor Ofula, Clayton Onyango, Limbaso S. Konongoi, Sheryl Bedno, Heather Burke, Athman M. Dumilla, Joseph Konde, M. Kariuki Njenga, Rosemary Sang, and Robert F. Breiman* Field Epidemiology and Laboratory Training Program, Nairobi, Kenya; Disease Outbreak Management Unit, Ministry of Health, Nairobi, Kenya; US Army Medical Research Unit Kenya, Nairobi, Kenya; International Emerging Infections Program, Nairobi, Kenya; Ministry of Health, Lamu, Kenya; Central Bureau of Statistics, Mombasa, Kenya; Kenya Medical Research Institute, Nairobi, Kenya

Abstract. An outbreak of Chikungunya virus (CHIKV) disease associated with high fever and severe protracted arthralgias was detected in Lamu, Kenya, peaking in July 2004. At least 1,300 cases were documented. We conducted a seroprevalence study to define the magnitude of transmission on Lamu Island. We conducted a systematic crosssectional survey. We administered questionnaires and tested 288 sera from Lamu residents for IgM and IgG antibodies to CHIKV. Chikungunya virus infection (seropositivity) was defined as a person with IgG and/or IgM antibodies to CHIKV. IgM antibodies to CHIKV were detected in 18% (53/288) and IgG antibodies in 72% (206/288); IgM and/or IgG antibodies were present in 75% (215/288). The seroprevalence findings suggested that the outbreak was widespread, affecting 75% of the Lamu population; extrapolating the findings to the entire population, 13,500 (95% CI, 12,458– 14328) were affected. Vector control strategies are needed to control the spread of this mosquito-borne infection. INTRODUCTION

Lamu Island beginning May 2004 and peaked in July 2004. Illness caused by Chikungunya had not been recognized previously on Lamu Island. After that outbreak, other associated outbreaks occurred in Mombasa, Kenya, between November and December 2004, Comoros Islands from January to May 2005 (unpublished data), Reunion Island,19 other islands in the Indian Ocean,20 and in India21 in 2006. The Ministry of Health, Lamu district, documented at least 1,300 patients meeting a clinical case definition of Chikungunya illness with no deaths. Of 130 specimens collected and sent to KEMRI for testing, IgM antibodies against CHIKV were detected in 60, and CHIKV was isolated by viral culture or detected by polymerase chain reaction in an additional 22 specimens. The extent and scope of CHIKV infection on Lamu Island was unknown. A seroprevalence study was carried out to define the magnitude of the outbreak and make recommendations for prevention and control interventions. This paper describes the findings of the serosurvey.

Chikungunya virus (CHIKV) is a positive single-stranded RNA virus belonging to the family Togaviridae, genus Alphavirus, and the Semliki Forest virus antigenic complex.1–3 The Semliki forest complex constitutes CHIKV, O’nyong nyong virus, Ross River, and Barmah Forest viruses from the South Pacific, which are associated with similar clinical manifestations. O’nyong’nyong, Mayaro, and Ross River viruses are closely related to CHIKV antigenically.4 Chikungunya virus was first isolated from the blood of a febrile patient in Tanzania in 1953.5 “Chikungunya” in the local dialect in Tanzania means stooping or bending, describing the position assumed by patients with the illness.2,5 Chikungunya virus illness is associated with fever, severe arthralgias, rash, headache, and malaise. Other symptoms include muscle aches and retro-orbital pains. Arthralgia can be debilitating and prolonged.4,6,7 Chikungunya disease is rarely fatal but is associated with significant morbidity. Chikungunya illness has an approximate incubation period of 1–2 weeks. The vectors principally responsible for transmission of the virus are Aedes mosquitoes.6,8 In Africa, CHIKV apparently is maintained in a sylvatic transmission cycle involving primates and forest-dwelling Aedes mosquitoes.9 Sylvatic vectors that have been implicated in transmission include Ae. africanus in East Africa,10 Ae. furcifer, Ae. taylori, Ae. delzieli, and Ae. luteocephalus in West Africa,6,11 and Ae. taylori and Ae. codellieri in South Africa.12 In contrast, transmission of CHIKV in Asia has been documented to occur mainly in urban areas where Ae. aegypti and Ae. albopictus are the identified vectors.13–15 Chikungunya is a re-emerging disease of public health importance in both African and Southeast Asian countries, causing major outbreaks. Outbreaks in Democratic Republic of Congo,1 Malaysia,16 Indonesia,17 and Senegal18 were reported earlier. An outbreak of Chikungunya occurred on

MATERIALS AND METHODS Study design. We conducted a cross-sectional seroprevalence study during October 5–9, 2004 (9 weeks after the peak of the outbreak). Serum samples were collected from selected inhabitants whose residences were distributed throughout the entire Lamu Island. We sampled from all 30 preexisting census enumeration areas (EAs) on the island. Sample size in each EA was determined using probability proportional to size sampling (PPS), in which the probability that a particular sampling unit was to be selected in the sample was proportional to the population size of the sampling unit. In each EA, the first household was randomly selected; every subsequent eighth household was chosen for the survey. In each selected household, one resident, irrespective of whether ill or not, was chosen randomly as the survey participant for interview and blood collection, by blindly picking a numbered slip of paper from those placed within a container. Infants (< 1 year old) were excluded from the survey. Standardized questionnaires were administered by field workers to collect data on demographics, symptoms, and treatment of members of each household. The questionnaires

* Address correspondence to Robert F. Breiman, International Emerging Infections Program, CDC-KEMRI, Nairobi, Kenya. Email: [email protected]

333

334

SERGON AND OTHERS

FIGURE 1. Lamu Island where CHIKV outbreak occurred in 2004 on the map of Kenya (Source of map: Central Bureau of Statistics, Kenya).

were translated into Kiswahili, the local dialect. With informed consent, questionnaires were administered and blood specimens were collected. The Kenya Ministry of Health determined that this study was part of its public health response to the outbreak and that review by an Ethical Review Committee was not required. Study site. Lamu Island covers an area of 102.4 km2 along the Kenyan coast in the Indian Ocean (Figure 1). The projected population for 2004 was 18,000 based on the 1999 census. The Lamu population is predominantly urban and is

FIGURE 2.

mostly composed of Swahili-speaking people who are Muslims of Arabic origin. Definition of CHIKV infection. A person with CHIKV infection (seropositive) was defined as any person in whom IgM or IgG antibodies to CHIKV were detected in serum. Laboratory diagnosis. Sera were separated from whole blood specimens and tested at Kenya Medical Research Institute (KEMRI) for IgM and IgG antibodies to CHIKV using antibody capture enzyme-linked immunosorbent assay (MAC ELISA). All sera were heat inactivated at 56°C for 30 minutes before testing for presence of CHIKV IgM using either indirect ELISA22 or IgG using direct ELISA.17 For both tests, we used a positive control serum specimen obtained from a previous Chikungunya virus outbreak in East Africa; a negative control serum specimen was from a person from a non-endemic region who tested negative for arbovirus infection. For CHIKV-specific IgM detection, 96-well polystyrene ELISA plates were coated with 1:1,000 dilution of anti-human IgM (Kirkegaard and Perry Laboratories, Gaithersburg, MD) overnight at 4°C. After five washes with phosphate-buffered saline (PBS) with 0.05% Tween-20 (PBS-T), non-specific binding was blocked by adding 5% non-fat dry milk in PBS with 0.5% Tween-20 and incubated for 30 minutes at room temperature. Test serum was added at 1:400 dilution and incubated for 60 minutes at 37°C. Each diluted test serum sample was added in quadruplicate, with two wells serving as positive control (with CHIKV antigen) and two wells serving as negative control. After adding a 1:40 dilution of CHIKV antigen (S-27; Centers for Disease Control and Prevention, Fort Collins, CO), plates were incubated overnight at 4°C before adding a horseradish peroxidase– conjugated alphavirus-specific monoclonal antibody 2A2C-3 (Centers for Disease Control and Prevention) at 1:6,000 dilution and further incubated for 60 minutes at 37°C. Presence of CHIKV antibodies was detected by adding ABTS [2,2’ amino-bis(3-ethylbenthiazoline-6-sulfonic acid)] substrate (Kirkegaard and Perry Laboratories), and the absorbance was read at 405 nm. Positive samples required a mean optical density (OD) value ⱖ 0.2 above that of the negative control for each sample. The range of OD values for positive serum specimens in this study was 0.3–1.5 above that of the negative control. For virus-specific IgG detection, plates were coated with 1:2,000 dilution of CHIKV antigen (United States’ Naval

Scatter distribution of IgG and IgM antibodies from study participants.

335

CHIKUNGUNYA SEROPREVALENCE—LAMU ISLAND 2004

Medical Research Unit, Bethesda, MD) overnight at 4°C. Heat-inactivated serum was added at 1:100 dilution and incubated for 60 minutes at 37°C. Each diluted test serum sample was added in quadruplicate, with two wells serving as a positive control (with CHIKV antigen) and two wells serving as a negative control. After adding a 1:3,000 dilution of horseradish peroxidase–conjugated anti-human IgG (Kirkegaard Perry Laboratories), presence of CHIV-specific IgGs was detected by adding ABTS substrate, and the absorbance was read at 405 nm. Positive samples required a mean OD value ⱖ 0.2 above that of the negative control for each sample. Data analysis. Data collected from the questionnaires were entered and analyzed using Epi Info 2002 software (Centers for Disease Control and Prevention). Age, sex, and locality distribution of the seropositive and seronegative participants were compared. Locality was defined as urban (living within Lamu town) or rural (living away from Lamu town). ␹2 was used for statistical testing.

we estimate that 13,500 people were infected on the island during the outbreak (95% CI, 12,458–14,328 persons). Seropositive and seronegative participants were similar regarding sex distribution (Table 1). The highest seropositivity ratio was in the 1- to 4-year-old group (86%) (Table 2), although participants ⱖ 15 years old were more likely to be seropositive than those < 15 years old (P ⳱ 0.04; Table 2). The proportion of participants who were IgG seropositive was higher than the proportion of participants who were IgM seropositive for all age groups. Among participants ⱖ 15 years old, 192 of 257 had CHIKV IgG antibodies detected, which was significantly higher than the proportion of participants < 15 years of age with IgG antibodies (P ⳱ 0.002). However, there was no statistical difference in the presence of IgM antibodies among persons ⱖ 15 years of age and the younger participants (19% versus 16%; P ⳱ 1.00). Although there were seropositive persons throughout 30 localities widely dispersed on Lamu Island, there was a higher (although not statistically significant) prevalence of seropositivity in urban areas than within rural areas on the island (77% versus 52%, P ⳱ 0.075; Table 3). Among the seropositive, the duration of joint pains ranged from 1 to 90 days, with a mean of 7 days (median, 3 days), principally affecting wrists, elbows, fingers, knees, and ankle joints. Myalgias were particularly noted in leg muscles (86%; Table 4). Most of the 118 febrile seropositive cases (86%) reported being hospitalized or confined to bed at home for a mean of 7 days (median, 7 days), with a range of 1–90 days. Among symptomatic seropositives, 84% (97/115) reported missing work or school for a mean of 7 days (range, 1–90 days; median, 7 days).

RESULTS

DISCUSSION

Among 445 households that were selected, 428 (96%) heads of household agreed to participate. Among the household members (one from each household was randomly selected for blood collection), 302 (71%) consented. Fourteen serum specimens had inadequate volume or were hemolyzed and were not tested; thus, 288 serum specimens were analyzed for IgG and IgM antibodies to CHIKV. Among 288 serum specimens tested, CHIKV IgG or IgM antibodies were detected in 215 (75%); 206 (72%) had IgG antibodies and 53 (18%) had IgM antibodies. Nine serum specimens with antibodies detected had only IgM antibodies (no IgG detected). For patients with detectable IgG or IgM antibodies, levels of IgG antibodies were higher, as indicated by higher OD values compared with IgM antibodies (Figure 2). Given a population of 18,000 and an attack rate of 75%,

The findings of this serosurvey suggested that the magnitude of the CHIKV outbreak was substantially greater than what was predicted based on the number of cases detected in health facilities. The outbreak may have affected 75% of the population on Lamu Island. There is no documentation of previous outbreaks or cases of alphavirus infection on Lamu Island, although such occurrences would likely not have been studied. We cannot be certain that all seropositives were infected during the course of this outbreak; however, the disproportionately high IgG, but not IgM, seropositivity of persons ⱖ 15 years old compared with younger participants suggested that previous outbreaks of CHIKV or other alphaviruses (because antibodies to other alphaviruses may cross-react with CHIKV antibodies) or sporadic infections might have oc-

TABLE 1 Demographic characteristic of participants of Chikungunya virus disease serosurvey, Lamu Island, Kenya, October 2004 (N ⳱ 288)

Characteristic

Age Mean Median Range Sex Men Women

Seropositive (IgG or IgM positive) (n ⳱ 215)

Seronegative (IgG and IgM negative) (n ⳱ 73)

34 32 1–80

30 25 1–80

73 (34%) 142 (66%)

22 (30%) 51 (70%)

TABLE 2 CHIKV disease serostatus among age groups, Lamu Island, Kenya, October 2004 Age group (years)

Number sampled from study population (%) (n ⳱ 282)*

IgG positive, n (%)

IgM positive, n (%)

Seropositive, n (%)

Seronegative (%)

1–4 (n ⳱ 7) 5–14 (n ⳱ 18) 15–24 (n ⳱ 61) 25–49 (n ⳱ 153) > 50 (n ⳱ 43) < 15 (n ⳱ 25) ⱖ 15 (n ⳱ 257)

7 (3) 18 (6) 61 (22) 153 (54) 43 (15) 25 (9) 257 (91)

6 (86) 6 (33) 39 (64) 120 (78) 33 (77) 12 (48) 192 (75)

1 (14) 3 (17) 11 (18) 36 (24) 1 (2) 4 (16) 48 (19)

6 (86) 8 (44) 39 (64) 126 (82) 33 (77) 14 (56) 198 (77)

1 (14) 10 (56) 22 (36) 27 (18) 10 (23) 11 (44) 59 (23)

* Information about age was not collected for six participants.

336

SERGON AND OTHERS

TABLE 3 CHIKV disease serostatus by locality on Lamu Island, Kenya October 2004 Locality

Number tested

Seropositive

Seronegative

Urban Rural Totals

201 26 227

153 (76%) 15 (58%) 168

48 (24%) 11 (42%) 59

Information about locality was missing in 61 responses (79% response rate). Prevalence OR ⳱ 2.34 (0.93–5.85), p ⳱ 0.075.

curred on Lamu Island. However, our survey was conducted after the peak of the outbreak; thus, it is possible that IgM antibodies had already waned in many participants exposed during the outbreak peak. This hypothesis is supported by data showing higher levels of IgG than IgM in most of the participants with detectable antibodies (Figure 2). In addition, it is possible that persons ⱖ 15 years old were more likely to have behaviors that increased the risk of exposure to CHIKV-infected Aedes mosquitoes. It is possible that older residents may have been previously infected with CHIKV or other alphaviruses and boosted their IgG antibodies on exposure to CHIKV during this outbreak. At the moment, little is known about the lifeline of circulatory CHIKV specific antibodies, beyond what is generally known about the kinetics of IgG and IgM antibodies with other immune correlates. Although persons with IgG or IgM antibodies to CHIKV (seropositives) were distributed throughout Lamu Island, infection seemed to occur more commonly in urban areas. As with dengue, CHIKV vector is most frequently transmitted by the Ae. aegypti mosquito,13–15 which tend to thrive in urban areas where more favorable conditions for breeding and transmission exist, including the presence of water storage containers, discarded water holding containers, and other debris in which stagnant water can accumulate. Symptoms associated with CHIKV infection such as fever, joint pains, and myalgias are non-specific and could be mistakenly identified with a variety of other diseases including dengue, malaria, Rift Valley fever, and influenza. However, pronounced persistent severe joint pains that affect wrists, elbows, fingers, and knees in some patients should raise the suspicion of alphavirus infection, especially chikungunya disease or O’nyong nyong fever, which also occurred in epidemic form in East Africa in the late 1990s.4,7,23 Among those who were symptomatic, a large proportion

TABLE 4 Sites of joint pains and myalgias among CHIKV seropositive participants Joint pains Joint Wrist Elbow Fingers Knees Ankle

Frequency 85% (99/117) 82% (96/117) 75% (88/117) 75% (88/117) 58% (68/117) Myalgias

Location Legs Arms Back Neck

Frequency 86% (68/79) 6% (5/79) 5% (4/79) 3% (2/79)

(84%) reported absence from work or school because of the illness for prolonged periods. An epidemic of O’nyong-nyong virus infection in East Africa in 1959–1962 caused at least one quarter of employed adults to miss at least 5 days of work.24 Although not commonly associated with mortality, epidemics of Chikungunya disease present public health threats because of substantial morbidity, associated suffering, and loss of economic productivity. Participants were asked about exposures, which likely occurred months earlier, so there may have been problems with recall, resulting in the loss of important information about illnesses. The refusal rate (29.4%) for blood collection could have resulted in an overestimation of the magnitude of infection, if persons who had been ill or lived in close proximity to people who were ill (and were more aware and concerned about the outbreak) were more likely to agree to participate. In summary, Chikungunya virus infection during the outbreak on Lamu Island was widespread with a high attack rate, affecting an estimated 75% of the Lamu population. In addition to substantial illness-associated discomfort and suffering, the outbreak most likely had a significant economic impact. Received September 2, 2006. Accepted for publication July 2, 2007. Acknowledgments: We thank the people of Lamu Island for their participation in the study and the Lamu District Health Office for its tremendous support. We appreciate the many helpful suggestions that we received during the course of this study and manuscript review from Drs. Roy Campbell, Ned Hayes, Barry Miller, and Ann Powers and other staff within the Division of Vector-Borne Infectious Diseases, CDC, Ft. Collins, CO. Authors’ addresses: Kibet Sergon, Field Epidemiology and Laboratory Training Program, Nairobi, Kenya. Charles Njugana and Rosalia Kalani, Disease Outbreak Management Unit, Ministry of Health, Nairobi, Kenya. Victor Ofula, Clayton Onyango, Limbaso S. Konongoi, and Rosemary Sang, Kenya Medical Research Institute, Nairobi, Kenya. Sheryl Bedno, US Army Medical Research Unit, Nairobi, Kenya. Athman M. Dumilla, Ministry of Health, Lamu, Kenya. Joseph Konde, Central Bureau of Statistics, Mombasa, Kenya. Heather Burke, M. Kariuki Njenga, and Robert F. Breiman, International Emerging Infections Program, Nairobi, Kenya, Telephone: 254-20271-3008, x166, Fax: 254-20-271-4745, E-mail: [email protected].

REFERENCES 1. Pastorino B, Muyembe-Tamfum JJ, Bessaud M, Tock F, Tolou H, Durand JP, 2004. Epidemic resurgence of Chikungunya virus in Democratic Republic of the Congo: identification of a new central African strain. J Med Virol 74: 277–282. 2. Lumsden WHR, 1955. An epidemic of virus disease in Southern Province, Tanganyika Territory, in 1952-53, II. General description and epidemiology. Trans R Soc Trop Med Hyg 49: 33. 3. Powers AM, Brault AC, Tesh RB, Weaver SC, 2000. Reemergence of Chikungunya and O’nyong-nyong viruses: evidence for distinct geographical lineages and distant evolutionary relationships. J Gen Virol 81: 471–479. 4. Tesh RB, 1982. Arthritides caused by mosquito-borne viruses. Annu Rev Med 33: 31–40. 5. Robinson MC, 1955. An epidemic of virus disease in Southern Province, Tanganyika Territory, in 1952-53, I. Clinical features. Trans R Soc Trop Med Hyg 49: 28. 6. Jupp P, McIntosh B, 1988. Chikungunya virus disease. Monath T, ed. The Arboviruses, Epidemiology and Ecology. First edition. Boca Raton, FL: CRC Press, 137–157. 7. Jeandel P, Josse R, Bagambisa G, Durand JP, 2004. [Exotic viral arthritis: role of alphaviruses]. Med Trop (Mars) 64: 81–88. 8. Jupp PG, McIntosh BM, 1990. Aedes furcifer and other mosquitoes as vectors of Chikungunya virus at Mica, Northeastern Transvaal, South Africa. J Am Mosq Control Assoc 6: 415–420. 9. Vanlandingham DL, Hong C, Klingler K, Tsetsarkin K, McElroy

CHIKUNGUNYA SEROPREVALENCE—LAMU ISLAND 2004

10. 11. 12. 13.

14.

15. 16.

17.

KL, Powers AM, 2005. Differential infectivities of O’nyongnyong and chikungunya virus isolates in Anopheles gambiae and Aedes aegypti mosquitoes. Am J Trop Med Hyg 72: 616– 621. McCrae AW, Henderson BE, Kirya BG, Sempala SD, 1971. Chikungunya virus in the Entebbe area of Uganda: isolations and epidemiology. Trans R Soc Trop Med Hyg 65: 152–168. Diallo M, Thonnon J, Traore-Lamizana M, Fontenille D, 1999. Vectors of chikungunya virus in Senegal: current data and transmission cycles. Am J Trop Med Hyg 60: 281–286. Jupp PG, Kemp A, 1996. What is the potential for future outbreaks of chikungunya, dengue and yellow fever in southern Africa? S Afr Med J 86: 35–37. Mourya DT, Thakare JR, Gokhale MD, Powers AM, Hundekar SL, Jayakumar PC, Bondre VP, Shouche YS, Padbidri VS, 2001. Isolation of chikungunya virus from Aedes aegypti mosquitoes collected in the town of Yawat, Pune District, Maharashtra State, India. Acta Virol 45: 305–309. Myers RM, Carey DE, Reuben R, Jesudass ES, De Ranitz C, Jadhav M, 1965. The 1964 epidemic of dengue-like fever in South India: isolation of chikungunya virus from human sera and from mosquitoes. Indian J Med Res 53: 694–701. Sarkar JK, 1966. Virological studies of haemorrhagic fever in Calcutta. Bull WHO 35: 59. Lam SK, Chua KB, Hooi PS, Rahimah MA, Kumari S, Tharmaratnam M, 2001. Chikungunya infection–an emerging disease in Malaysia. Southeast Asian J Trop Med Public Health 32: 447–451. Laras K, Sukri NC, Larasati RP, Bangs MJ, Kosim R, Wandra T,

18. 19. 20. 21. 22.

23.

24.

337

Master J, Kosasih H, Hartati S, Beckett C, Sedyaningsih ER, Beecham HJ, Corwin AL, 2005. Tracking the re-emergence of epidemic Chikungunya virus in Indonesia. Trans R Soc Trop Med Hyg 99: 128–141. Thonnon J, Spiegel A, Diallo M, Diallo A, Fontenille D, 1999. Chikungunya virus outbreak in Senegal in 1996 and 1997. Bull Soc Pathol Exot 92: 79–82. Josseran L, Paquet C, Zehqnoun A, Caillere N, Le Tertre A, Solet JL, Ledrans M, 2006. Chikungunya disease outbreak, Reunion Island. Emerg Infect Dis 12: 1994–1995. Higgs S, 2006. The 2005-2006 Chikungunya epidemic in the Indian Ocean. Vector Borne Zoonotic Dis 6: 115–116. Kalantri SP, Joshi R, Riley LW, 2006. Chikungunya epidemic: an Indian perspective. Natl Med J India 19: 315–322. Thein S, La Linn M, Aaskov J, Aung MM, Aye M, Zaw A, Myint A, 1992. Development of a simple indirect enzyme-linked immunosorbent assay for the detection of immunoglobulin M antibody in serum from patients following an outbreak of Chikungunya virus in Yangon, Myanmar. Trans R Soc Trop Med Hyg 86: 438–442. Sanders EJ, Rwaguma EB, Kawamata J, Kiwanuka N, Lutwama JJ, Ssengooba FP, Lamunu M, Najjemba R, Were WA, Bagambisa G, Campbell GL, 1999. Nov. O’nyong-nyong fever in south-central Uganda, 199-1997: description of the epidemic and results of a household-based seroprevalence survey. J Infect Dis 180: 1436–1443. Johnson BK, 1988. O’nyong’nyong virus disease. Monath TP, ed. The Arboviruses: Epidemiology and Ecology. Boca Raton, FL: CRC Press, 217–223.