ABSTRACT
Leptospirosis is a vaccine-preventable bacterial zoonotic disease caused by pathogenic Leptospira species. The efficacy of Leptospira canine vaccines is assessed by challenging vaccinated and control dogs with virulent serovars of Leptospira, followed by detection of Leptospira in blood and urine. We assessed the consistency between results obtained for urine and blood samples from clinical studies with species-specific real-time quantitative PCR (qPCR) targeting the lipL32 gene and those obtained with the reference culture method. The specificity of the qPCR assay was confirmed by negative results for nonpathogenic Leptospira and for several canine viruses, bacteria, and parasites. The results from the two methods were compared using McNemar's test, kappa coefficient (κ), and percentage of agreement analyses. The results for numbers of positive and negative dogs were similar, with no false-negative results with the qPCR assay. For both blood and urine, there was strong agreement between the culture method and qPCR results (κ = 0.68 [95% confidence interval (CI), 0.62 to 0.74] and κ = 0.65 [95% CI, 0.59 to 0.71], respectively). However, there was a statistically significant difference between blood samples (P < 0.001) and urine samples (P = 0.028). The negative percentage agreements were 97% and 84% and the positive percentage agreements were 68% and 83% for blood and urine samples, respectively. Although the cell culture method is the recommended gold standard, our results show that qPCR assay is a valid alternative method for the rapid and specific detection of pathogenic Leptospira spp. in urine and blood samples during vaccine efficacy studies, without loss of sensitivity.
