Seroprevalence of Trypanosoma cruzi infection in French Guiana

EPIDEMIOLOGY

Seroprevalence of Trypanosoma cruzi infection in French Guiana

Christine Aznar^I,1; Guy La Ruche^I; Stéphane Laventure^II; Bernard Carme^I; Pascale Liegeard^III; Mireille Hontebeyrie^III

^ILaboratoire Hospitalo Universitaire de Parasitologie et Mycologie, UFR de Médecine Antilles-Guyane and Centre Hospitalier Andrée Rosemon, Equipe EA 3593, BP 6006, Rue des Flamboyants, 97300 Cayenne, French Guiana

^IILaboratoire d'Entomologie Médicale, Institut Pasteur de la Guyane, Cayenne, French Guiana

^IIIDépartement de Biologie Structurale et Chimie, Institut Pasteur, Paris, France

ABSTRACT

A survey was carried out on 1487 individuals to assess the seroprevalence of Trypanosoma cruzi infection in French Guiana. The overall prevalence of T. cruzi specific IgG was 0.5%. In multivariate analysis, residence in areas where housing is favorable for the presence of triatomine bugs was the only factor associated with the presence of T. cruzi antibodies. These results have implications for public health since blood donors are not routinely screened for T. cruzi infection in French Guiana.

Key words:Trypanosoma cruzi - Chagas disease - seroprevalence - French Guiana

Chagas disease (American trypanosomiasis) is caused by the protozoan Trypanosoma cruzi, mainly transmitted to humans by blood-sucking triatomine bugs (Hemiptera, Triatominae) but also transmitted by blood transfusion from infected donors, and occasionally by transplacental mother-to-child transmission. It is estimated that 16-18 million people are infected in Latin America (WHO 1991). However, relatively few cases have been reported from the Guyanas (Guyana, Suriname, and French Guiana) (Luquetti & Schofield 2003).

In French Guiana, only 17 clinical cases of Chagas disease have been reported since 1939 (Floch & Tasque 1940, Beaudet et al. 1985, Esterre & Dedet 1987), includ-ing four cases of acute chagasic myocarditis diagnosed between 1994 and 1996 (Carme et al. 2001a). A serological study conducted in 1983-1985, using an indirect immuno-fluorescence test, showed that only one of 301 individuals (0.3%) had antibodies to T. cruzi (Esterre & Dedet 1987). We present here a cross-sectional study designed to assess the seroprevalence of T. cruzi infection in various areas of French Guiana.

MATERIALS AND METHODS

French Guiana (Figure) is located in the Amazonian forest between Brazil and Suriname. It is technically a Department of France and has the highest living standard in South America. In the 1990 census there were 114,678 inhabitants, and in 1999 were around 157,274 inhabitants. There is massive immigration from neighboring countries which is difficult to quantify, but it is estimated that illegal immigrants represent up to a third of the total population.

Map of French Guiana (areas of study)

For this retrospective study, sera were selected randomly from several large banks of sera from "Centre National de Référence de Virologie et du Paludisme" (Collection Biological samples used for infectious disease surveillance) under the Ministry of Health. These sera are available without need for specific ethical clearance. Sera samples were stored at –80°C, without glycerol, for 3 to 4 years and were thawed once. These samples had originally been collected either for epidemiological studies on human T-cell leukemia/lymphoma virus type I, during routine testing of pregnant women, or during previous studies on viral hepatitis or arboviroses. The samples included most of the ethnic groups and regions of French Guiana. For each sample, basic demographic data included sex, age, place of current residence, and ethnic group (Table I). However, Cayenne and its surroundings was the only region where sera could be selected to include a sufficient number of each ethnic group; thus, the seropreva-lence presented in this paper is calculated based on the respective percentages of ethnic groups observed in this city.

Sera were tested with a simple T. cruzi enzyme-linked immunoassay (ELISA) using antigen prepared from trypomastigotes of the Brener CL strain of T. cruzi (Aznar et al. 1997). Briefly, all samples were diluted 1:200 in reference to a previous study (Aznar et al. 1997). Samples were then assayed in duplicate and one negative and one positive control were included on each microplate (Nunc, Roskilde, Denmark). Fifty µl of a 1:1000 dilution of goat peroxidase-conjugated anti-human IgG (Sanofi Diagnostics Pasteur, Paris, France) were added to each well. After incubation, peroxidase was revealed with 50 µl of a 1/1 mixture of 2.2'-azino-di-3-ethyl-benzothiazoline sulfonate and hydrogen peroxide (ABTS Peroxidase Substrate System, Kirkegaard Perry Laboratories Inc., Gaithersburg, MD). Optical density (OD) was measured at 492 nm for peroxidase activity in an automated ELISA reader (Titertek Multiskan MCC/340 - Lab Systems, France). Sera were considered as positive when the mean of duplicate optical density values reached the mean value obtained from 68 negative controls + 3 standard deviations. Controls were chosen from among blood donors from mainland France (Aznar et al. 1997). The cut-off value for anti-T. cruzi IgG was determined as OD = 0.108.

Data were analyzed using Epi-Info and Egret software. The presence of T. cruzi-specific IgG (without specific IgM, data not shown) was considered to denote a chronic infection (Medrano-Mercado et al. 1996). From the results in the different areas of French Guiana, we extrapolated the global seroprevalence using the 1990 national census. Demographic factors related to T. cruzi infection were determined by univariate analysis using Chi-square or Fisher's exact test, with a significance level of 5%. Multivariate analyses using stepwise backward logistic regression were subsequently performed and strength of associations was estimated by the adjusted odd ratios (OR) with 95% confidence interval (CI).

RESULTS

A total of 1487 sera was tested for IgG antibodies to T. cruzi (40.3% from men and 59.7% from women, with a median age of 26 years, range 1 to 93). Of these sera, 324 (21.8%) were from patients living in the littoral plain or in Saint-Laurent, originally taken for diagnosis of dengue fever. The remaining 78.2% were taken from pregnant women or during epidemiological studies of populations living in the forest areas of the Oyapock or Maroni River regions.

The seroprevalence rates for T. cruzi antibodies ranged from zero to 7.3% according to place of residence, age class, and ethnic group (Tables I, ^II). Seroprevalence was below 3% except in three areas where the prevalence of antibodies was greater than 6%: in Antécume-Pata in the southern part of the Maroni river (2 of 33 sera, 6.1%), in Cacao in the forest area (3/41, 7.3%), in Camopi/Trois-Sauts in the southern part of the Oyapock river (6/82, 7.3%). Globally, the seroprevalence of T. cruzi antibodies extrapolated for the entire population of French Guiana was 0.5% (95% CI, 0.1-0.8%) (Table I).

^{Table II} shows univariate comparisons of the demographic factors and the context of collection of sera between T. cruzi seronegative and seropositive patients. The only factors statistically associated with the presence of anti-T. cruzi IgG were residence in Camopi/Trois-Sauts, in Cacao or in Antécume-Pata. The prevalence of T. cruzi antibodies did not differ between males and females. Among 321 subjects with fever as the reason for sera collection, the prevalence of T. cruzi antibodies was not associated with Dengue fever (data not shown).

After multivariate analysis, the place of residence remained the only factor significantly associated with T. cruzi antibodies. The context of collection of sera (sera collected for fever, year of collection) was not associated with T. cruzi antibodies. The presence of anti-T. cruzi IgG was independently associated with living in Camopi/Trois-Sauts (adjusted OR = 21.8, 95% CI = 5.7-83.0, p < 0.001), in Antécume-Pata (adjusted OR = 21.4, 95% CI = 3.8-121.3, p < 0.001) and in Cacao (adjusted OR = 18.0, 95% CI = 3.2-101.1, p = 0.001).

DISCUSSION

Our study suggests that 0.5% of the general population in French Guiana – approximately 785 people – may have T. cruzi specific IgG antibodies. Such a seropreva-lence is 1.75 times higher than that previously observed in French Guiana (Esterre & Dedet 1987), and is similar to seroprevalence levels recorded for children in Brazil and Uruguay prior to the Southern Cone Initiative against Chagas disease (WHO 1997, 1998, Schofield & Dias 1999).

By contrast, very few cases of clinical Chagas disease have been observed in French Guiana during the last 50 years (Carme et al 2001b). Several hypotheses could explain this paradox. Firstly, our survey could have overestimated the actual seroprevalence of T. cruzi infection. However, the method that we used, ELISA, offers a high reliability and is concordant with other serological methods such as direct agglutination test or indirect immuno-fluorescence antibody test (Añez et al. 1999) and PCR (Brenière et al. 2002). Furthermore, T. cruzi trypomastigote extract, with almost no cross-reactivity with other protozoan infections (Aznar et al. 1997), has been shown to be an excellent reagent for detecting the infection in areas of low prevalence, and appears to be better than epimastigote extracts.

Alternatively, it may be that Chagas disease is significantly under-diagnosed in French Guiana, possibly as a result of frequent asymptomatic forms of T. cruzi infection or recent infections with clinical symptoms too mild to be recognized (Medrano-Mercado et al. 1996). Our results could therefore signify the emergence of a public health problem in French Guiana and the number of clinical cases could increase in the future. Indeed, our results indicate highest seroprevalence among people living in rural rain-forest villages (Upper-Maroni, Upper-Oyapock, Cacao) with traditional housing (wood and/or palms). Clinical cases in these areas may well go undiagnosed because of the basic level of health facilities there and because the Hmong and the Native American groups often have different heath-seeking behaviors than metropolitan or Creole French ethnic groups. Active detection campaigns should be implemented to establish the true prevalence of Chagas disease in areas of French Guiana with poor medical facilities. Potential vectors such as Rhodnius pictipes and R. robustus are recorded from such areas, and we have also found these species, and Panstrongylus geniculatus, infected with T. cruzi in urban areas of Cayenne and surrounding villages (Aznar C., unpublished observation). These observations lead us to further investigate the presence of infected vectors and domestic or wild reservoirs in rural and urban areas of French Guiana, and also indicate a possible risk of blood transfusional transmission in French Guiana since French law does not require any serological screening of blood donors for this infection.

ACKNOWLEDGMENTS

To Antoine Talarmin and Jean-Louis Sarthou for their support; to Alejandro Luquetti for helpful advice; to M Nacher and CJ Schofield for critical revision of the manuscript.

Received 18 May 2004

Accepted 14 October 2004

• Añez N, Carrasco H, Parada H, Crisante G, Rojas A, Gonzalez N, Ramirez JL, Guevara P, Rivero C, Borges R, Scorza JV 1999. Acute Chagas disease in Western Venezuela: a clinical, seroparasitologic, and epidemiologic study. Am J Trop Med Hyg 60: 215-222.
• Aznar C, Liegeard P, Mariette C, Lafon S, Levin MJ, Hon-tebeyrie M 1997. A simple Trypanosoma cruzi enzyme-linked immunoassay for control of human infection in nonendemic areas. FEMS Immunol Med Microbiol 18: 31-37.
• Beaudet B, Esterre P, Jacquet-Viallet P, Ganty J, Dedet JP 1985. Un cas de cardiopathie chagasique chronique autochtone en Guyane française. Bull Soc Path Exot 78: 56-63.
• Brenière SF, Bosseno MF, Noireau F, Yacsik N, Liegeard P, Aznar C, Hontebeyrie M 2002. Integrate study of a Bolivian population infected by Trypanosoma cruzi, the agent of Chagas disease. Mem Inst Oswaldo Cruz 97: 289-295.
• Carme B, Aune I, Nguyen G, Aznar C, Beaudet B 2001a. Four cases of acute chagasic myocarditis in French Guiana. Am J Trop Med Hyg 64: 162-163.
• Carme B, Aznar C, Pradinaud R 2001b. Absence of a proven resurgence of Chagas disease or cutaneous leishmaniasis in French Guiana over the last two decades. Ann Trop Med Parasitol 95: 623-625.
• Esterre P, Dedet JP 1987. Situação atual da doença de Chagas na Guiana Francesa. Rev Soc Bras Med Trop 20: 139-142.
• Floch H, Tasqué P 1940. Un cas de maladie de Chagas en Guyane française. Pub Inst Pasteur Guyane Fr Territ Inini 2: 1-4.
• Luquetti A, Schofield CJ 2003. Chagas disease in the Guyanas. Proceedings ECLAT-AMCHA International Workshop on Chagas Disease Surveillance in the Amazon Region, Universidad de Los Andes, Bogota, p. 28-29.
• Medrano-Mercado N, Luz MR, Torrico F, Tapia G, Van Leuven F, Araujo-Jorge TC 1996. Acute-phase proteins and serologic profiles of chagasic children from an endemic area in Bolivia. Am J Trop Med Hyg 54: 154-161.
• Schofield CJ, Dias JCP 1999 The Southern Cone Initiative against Chagas disease. Adv Parasitol 42: 1-27
• WHO 1991. Control of Chagas disease. WHO Technical Report Series 811
• WHO 1997. Chagas disease. Interruption of transmission. WHO Wkly Epidemiol Rec 72: 1-5.
• WHO 1998. Chagas disease. Interruption of transmission. WHO Wkly Epidemiol Rec 73: 1-4.

Publication Dates

• Publication in this collection
  04 Mar 2005
• Date of issue
  Dec 2004

History

• Accepted
  14 Oct 2004
• Received
  18 May 2004