The leishmaniases are a complex of parasitic diseases which cause
approximately 30,000 deaths annually. In Brazil, leishmaniasis is a significant burden (Bezerra et al., 2018). Dogs are the main reservoir host for Leishmania (Le.) infantum parasites and are the source of infection for the human population. Le. infantum is transmitted to people via the bite of an infected Lutzomyia (Lu.) longipalpis sandfly. To disrupt this transmission cycle, it is important to identify Leishmania infected dogs and remove them from the population. However, current diagnostic methods are complex, can be inaccurate and therefore limit effective reservoir control.

The aim of this study was to determine whether Leishmania infantum
infected dogs could be discriminated from uninfected dogs using three different methodologies; a VOC Analyser, behavioural bioassay and gas chromatography mass spectrometry (GC/MS). Previous studies have shown that animals (e.g. dogs, hamsters) infected with Leishmania parasites produce odours that are different to those produced by uninfected animals (De Oliveira et al., 2008; Magalhães-Júnior et al., 2014b). In this study the VOC analyser demonstrated the ability to reliably discriminate between Leishmania infected and uninfected dogs, with high sensitivity (97-100%) and specificity (95-100%). The accuracy demonstrated of this diagnostic is an improvement on the current in-field DPP test used for cVL diagnosis, with a sensitivity of 75-89% and specificity of 56-70% having been recently reported (Figueiredo et al., 2018). This result suggests the potential for this approach for the rapid, non-invasive POC diagnosis of dogs infected with leishmaniasis. Studies on the behavioural response of male and female to the odour of infected and uninfected dogs showed that female Lu. longipalpis were significantly more attracted to infected odour whereas males were not. This result may suggest that Leishmania could be manipulating canine host odour in order to aid its own transmission. Increased attractiveness of infected dog odour suggests the potential of whole dog odour baited traps as a novel vector control methodology. Further GC/MS analysis confirmed the differences observed between the odour profiles of infected and uninfected dogs through both the VOC Analyser and behavioural bioassays. This analysis also allowed for the identification of the chemical structure and composition
of compounds present in infected dogs which have the potential to not only enhance the VOC Analyser as a cVL specific diagnostic, but also could be investigated as artificial compounds for odour baited sandfly traps.