Genotyping of Mycobacterium leprae for better understanding of leprosy transmission in Fortaleza, Northeastern Brazil

Leprosy is endemic in large part of Brazil with 28,761 new patients in 2015, the second largest number worldwide and reaches 9/10.000 in highly endemic regions and 2.7/10.000 in the city of Fortaleza, Ceará, Northeast Brazil. For better understanding of risk factors for leprosy transmission, we conducted an epidemiologic study supplemented by 17 locus VNTR and SNP 1–4 typing of Mycobacterium leprae in skin biopsy samples from new multibacillary (MB) patients diagnosed at a reference center in 2009 and 2010. Among the 1,519 new patients detected during the study period, 998 (65.7%) were MB and we performed DNA extraction and genotyping on 160 skin biopsy samples, resulting in 159 (16%) good multilocus VNTR types. Thirty-eight of these patients also provided VNTR types from M. leprae in nasal swabs. The SNP-Type was obtained for 157 patients and 87% were of type 4. Upon consideration all VNTR markers, 156 different genotypes and three pairs with identical genotypes were observed; no epidemiologic relation could be observed between individuals in these pairs. Considerable variability in differentiating index (DI) was observed between the different markers and the four with highest DI [(AT)15, (TA)18, (AT)17 and (GAA)21] frequently demonstrated differences in copy number when comparing genotypes from both type of samples. Excluding these markers from analysis resulted in 83 genotypes, 20 of which included 96 of the patients (60.3%). These clusters were composed of two (n = 8), three (n = 6), four (n = 1), five (n = 2), six (n = 1), 19 (n = 1) and 23 (n = 23) individuals and suggests that recent transmission is contributing to the maintenance of leprosy in Fortaleza. When comparing epidemiological and clinical variables among patients within clustered or with unique M. leprae genotypes, a positive bacterial index in skin biopsies and knowledge of working with someone with the disease were significantly associated with clustering. A tendency to belong to a cluster was observed with later notification of disease (mean value of 3.4 months) and having disability grade 2. A tendency for lack of clustering was observed for patients who reported to have lived with another leprosy case but this might be due to lack of inclusion of household contacts in the study. Although clusters were spread over the city, kernel analysis revealed that some of the patients belonging to the two major clusters were spatially related to some neighborhoods that report poverty and high disease incidence in children. Finally, inclusion of genotypes from nasal swabs might be warranted. A major limitation of the study is that sample size of 160 patients from a two year period represents only 15% of the new patients and this could have weakened statistical outcomes. This is the first molecular epidemiology study of leprosy in Brazil and although the high clustering level suggests that recent transmission is the major cause of disease in Fortaleza; the existence of two large clusters needs further investigation.