Selection of appropriate serological tests to measure the incidence of natural Leishmania infantum infection during DNA/MVA prime/boost canine vaccine trials

In response to the increasing need for field trials of experimental DNA vaccines against zoonotic visceral leishmaniasis in dogs, our aim was to validate the use of ELISA protocols which will be suitable for detection of natural infection in vaccinated dogs. We have previously demonstrated that DNA/modified vaccinia virus Ankara (MVA) vaccine expressing tryparedoxin peroxidase (TRYP) induced high titres of TRYP antigen-specific IgG in immunized dogs. Here we report our findings that seroconversion to an unrelated diagnostic antigen rK39 did not occur in vaccinated dogs, and that responses to crude Leishmania infantum promastigote antigen (CLA) were weak and short-lived. This is in contrast to strong responses to both antigens shown in naturally infected dogs. To select an appropriate serological test for measurement of infection incidence, we also tested longitudinal samples from an immunologically well-characterized cohort of naturally infected dogs. The sensitivity of CLA ELISA was superior to that of rK39 in early stage infection (from 2 months before, to 2 months after the first detection of infection by PCR or parasitological culture), and more sensitive than rK39 in cross-sectional sampling (81.0% vs 61.9%). We conclude that CLA ELISA will provide sensitive estimates of L. infantum infection incidence in DNA/MVA vaccinated dogs, though optimal testing would include rK39, or a similar recombinant antigen, to improve overall specificity.