We compared 2 novel loop-mediated isothermal amplification (LAMP) assays that target either the 16S ribosomal RNA ( rrs) gene or the gene encoding a 32-kDa leptospiral lipoprotein ( lipL32) in order to assess the effect of the target on the accuracy of the LAMP assays. The most sensitive assay was the rrs assay with a limit of detection (LOD) of 1.2 × 101 genome equivalents per reaction. The novel lipL32 assay showed an LOD of 1.2 × 102 genome equivalents per reaction. Both assays showed adequate specificity when tested against a collection of bacteria commonly found in voided canine urine. However, when field samples were assayed, the rrs assays gave many false-positive results and a poor positive predictive value of 8.33%. In conclusion, even if the LAMP assay is used in low prevalence areas, the lipL32 assay would be preferable. Conversely, the higher analytical sensitivity of the rrs assay could be effectively used as a screening test in endemic areas with high disease prevalence, followed by confirmation of the positive results using the lipL32 assay.
Keywords: 16S rRNA; Leptospira; dogs; lipL32; loop-mediated isothermal amplification; urine.