Comparison of polymerase chain reaction assay, bacteriologic culture, and serologic testing in assessment of prevalence of urinary shedding of leptospires in dogs

Objective-To compare results of polymerase chain reaction (PCR) testing of urine samples, serologic testing, and bacteriologic culture of urine to determine prevalence of urinary shedding of leptospires in dogs. Design-Serial case study. Animals-500 dogs evaluated serially without regard to health status. Procedure-Urine samples were examined via PCR assay and bacteriologic culture for leptospires. Blood samples were analyzed for antibodies against serovars canicola, bratislava, pomona, icterohemorrhagiae, grippotyphosa, and hardjo. Results-Titers greater than or equal to 1:100 against at least 1 serovar were detected in 104 (20.8%) dogs, and titers greater than or equal to 1:400 were detected in 41 (8.2%) dogs. High titers were detected most commonly to serovar grippotyphosa, followed by icterohemorrhagiae, canicola, pomona, bratislava, and hardjo. High titers to > 1 serovar were detected in 14 dogs. A positive PCR assay result was obtained in 41 (8.2%) dogs, only 9 of which had a titer greater than or equal to 1:100. Leptospires were not cultured from the urine of any dog. Only 4 dogs had clinical leptospirosis. Overall disease prevalence was 0.8% for the 6-month evaluation period. Compared with PCR assay, serologic testing for predicting shedding had a sensitivity of 22%, specificity of 79%, positive predictive value of 9%, and negative predictive value of 92%. Conclusions and Clinical Relevance Irrespective of health status, 8.2% of dogs were shedding pathogenic leptospires. Serologic testing was a poor predictor of urinary shedding. Clinically normal dogs that shed leptospires may pose a zoonotic risk to their owners.