Epidemiologic relationship between Toscana virus infection and Leishmania infantum due to common exposure to Phlebotomus perniciosus sandfly vector.

Sand flies are recognised vectors of parasites in the genus Leishmania and a number of arthropod-borne viruses, in particular viruses within the genus Phlebovirus, family Bunyaviridae. In southern France, Toscana phlebovirus (TOSV) is recognized as a prominent cause of summer meningitis. Since Leishmania and TOSV have a common vector (Phlebotomus perniciosus), an epidemiologic link has been assumed for a long time. However, there is no scientific evidence of such a link between human leishmaniosis and phleboviral infections. To identify a possible link, we investigated the presence and distribution of antibodies against these two microorganisms (i) in individuals and (ii) at a spatial level in the city of Marseille (south-eastern France). Five hundred sera were selected randomly in the biobank of the Department of Parasitology of the Public Hospitals of Marseille. All sera were previously tested for IgG against Leishmania by Western Blotting, and TOSV IgG were detected by indirect immunofluorescence. The seropositivity rates were 21.4% for TOSV and 28% for Leishmania. Statistical analysis demonstrated that seropositivity for one pathogen was significantly associated with seropositivity to the other pathogen. This result provided the first robust evidence for the existence of an epidemiological relationship between Leishmania infantum and TOSV. Addresses of tested patients were geolocalized and integrated into Geographical Information System software, in order to test spatial relationship between the two pathogens. Spatial analysis did not allow to identify (i) specific patterns for the spatial distribution of positive serological results for TOSV or Leishmania, and (ii) a spatial relationship between Leishmania and TOSV positive serological results. This may reflect the fact that the sample studied was not powerful enough to demonstrate either a spatial clustering or co-location, i.e. that the actual risk exposure area is smaller than the mean of distance between patients in our study (245 m).