Development of an urban molecular xenomonitoring system for lymphatic filariasis in the Recife Metropolitan Region, Brazil

INTRODUCTION: Molecular xenomonitoring (MX)—pathogen detection in the mosquito rather than human—is a promising tool for lymphatic filariasis (LF) surveillance. In the Recife Metropolitan Region (RMR), the last LF focus in Brazil, Culex quinquefasciatus mosquitoes have been implicated in transmitting Wuchereria bancrofti parasites. This paper presents findings on the ideal mosquito collection method, mosquito dispersion, W. bancrofti infection in mosquitoes and W. bancrofti antigen in humans to aid MX development. METHODS: Experiments occurred within two densely populated urban areas of Olinda, RMR, in July and August 2015. U.S. Centers for Disease Control and Prevention (CDC) light traps were compared to battery-powered aspirators as collection methods, and mosquito dispersion was measured by mosquito mark release recapture (MMRR). Female Cx. quinquefasciatus were tested by PCR for W. bancrofti infection, and study area residents were screened by rapid tests for W. bancrofti antigen. RESULTS: Aspirators caught 2.6 times more total Cx. quinquefasciatus, including 38 times more blood-fed and 5 times more gravid stages, than CDC light traps. They also collected 123 times more Aedes aegypti. Of the 9,644 marked mosquitoes released, only ten (0.01%) were recaptured, nine of which were < 50m (34.8m median, 85.4m maximum) from the release point. Of 9,169 unmarked mosquitoes captured in the MMR, 38.3% were unfed, 48.8% blood-fed, 5.5% semi-gravid, and 7.3% gravid. PCR on 182 pools (1,556 mosquitoes) found no evidence of W. bancrofti infection in Cx. quinquefasciatus. Rapid tests on 110 of 111 eligible residents were all negative for W. bancrofti antigen. CONCLUSIONS: Aspirators were more effective than CDC light traps at capturing Ae. aegypti and all but unfed stages of Cx. quinquefasciatus. Female Cx. quinquefasciatus traveled short (< 86m) distances in this urban area. Lack of evidence for W. bancrofti infection in mosquitoes and antigen in humans in these fine-scale studies does not indicate that LF transmission has ceased in the RMR. A MX surveillance system should consider vector-specific collection methods, mosquito dispersion, and spatial scale but also local context, environmental factors such as sanitation, and host factors such as infection prevalence and treatment history.