Evaluation of variable number tandem repeat genotyping of Leptospira as a tool for epidemiological investigations of leptospirosis

Introduction: Leptospirosis is a zoonosis maintained by chronic renal infection of animals. The basic taxon is the serovar, however serotyping is limited to few reference laboratories. Our objective was to evaluate two genotyping methods based on Variable Number Tandem Repeat (VNTR) sequences, as tools for quality control in reference laboratories, and for identifying serovars in epidemiologic investigations.

Methods: We used two VNTR protocols (VNTR1 and VNTR2) to genotype 24 L. interrogans strains representing 20 serovars from a WHO reference collection, and 25 isolates from surveillance in Brazil and Columbia (11 human and 10 rat serovar Copenhageni isolates; 2 dog and 1 human serogroup Canicola isolates; 1 human serogroup Djasiman isolate). Sensitivity analysis was performed at the serovar level based on UPGMA trees.

Results: Overall discriminatory power of VNTR1 and VNTR2 was 62% and 92%. We found discrepancies between our and published results for 2 strains. Sensitivity analysis found that for the 95% similarity threshold, VNTR1 had a higher sensitivity than VNTR2 (98% vs. 90%) but a lower specificity (97% vs. 100%). No associations were found between genotypes and temporal or spatial clustering of cases and disease severity.

Discussion: This was the first study to evaluate two VNTR methods on the same set of reference and field strains. We recommend that either method should be implemented as a quality control measure for reference strains used in serologic testing. Although VNTR may be used to identify serovar for L. interrogans isolates, it lacks sufficient discriminatory power for molecular epidemiology applications.