IgG avidity index and complete blood count as biomarkers of clinical disease in naturally infected dogs with Leishmania infantum.

Canine visceral leishmaniosis (CVL), a parasitic disease caused by Leishmania infantum, may evolve to a chronic condition and lead to death. Evaluation of infected dogs is important to establish the clinical and laboratory parameters involved in the evolution of the disease. The objectives of the present study were to discriminate a canine population (n = 52) into sub-clinical and clinically affected dogs based on signs and scores, to evaluate the hematological, biochemical, histopathological and parasitological parameters of the two dog groups, and to analyze the results by multivariate regression analysis with the aim of establishing biomarkers of CVL clinical disease. The most common signs observed in the clinically affected dogs (n = 29) were hyperkeratosis, weight loss, onychogryphosis, pale mucosa and lymphadenomegaly. In the multivariate analysis, animals presenting high IgG avidity index and low red blood, lymphocyte and eosinophil counts, and low serum urea concentration had an increased probability of being classified as clinically affected (p <  0.05). All five parameters were considered to be strong biomarkers for monitoring the clinical disease, while IgG avidity percentage was strongly correlated with the number of clinical signs and could function as an indicator of the duration of infection. This is the first report on the application of IgG avidity and of multivariate regression analysis in establishing associations between the clinical signs of CVL and host biomarkers. Since avidity index (AI) percentages were strongly correlated with the number of clinical signs, it could be useful in clinical practice for auxiliary diagnosis of CVL and monitoring disease progression. A limitation of this study is the lack of information on co-infections by Anaplasma platys, Babesia canis vogeli, Ehrlichia canis and Hepatozoon canis. Therefore future studies should evaluate the influence of such co-infections on the associations studied using multivariate methods with larger samples.