Epidemiological and genomic determinants of tuberculosis outbreaks in First Nations communities in Canada

BACKGROUND: In Canada, tuberculosis disproportionately affects foreign-born and First Nations populations. Within First Nations’ peoples, a high proportion of cases occur in association with outbreaks. Tuberculosis transmission in the context of outbreaks is thought to result from the convergence of several factors including characteristics of the cases, contacts, the environment, and the pathogen. METHODS: We examined the epidemiological and genomic determinants of two well-characterized tuberculosis outbreaks attributed to two super-spreaders among First Nations in the province of Alberta. These outbreaks were associated with two distinct DNA fingerprints (restriction fragment-length polymorphisms or RFLPs 0.0142 and 0.0728). We compared outbreak isolates with endemic isolates not spatio-temporarily linked to outbreak cases. We extracted epidemiological variables pertaining to tuberculosis cases and contacts from individual public health records and the provincial tuberculosis registry. We conducted group analyses using parametric and non-parametric statistical tests. We carried out whole-genome sequencing and bioinformatic analysis using validated protocols. RESULTS: We observed differences between outbreak and endemic groups in the mean number of total and child-aged contacts and the number of contacts with new positive and converted tuberculin skin tests in all group comparisons (p < 0.05). Differences were also detected in the proportion of cases with cavitation on a chest radiograph and the mean number of close contacts in selected group comparisons (p < 0.02). A phylogenetic network analysis of whole-genome sequencing data indicated that most outbreak and endemic strains were closely related to the source case for the 0.0142 fingerprint. For the 0.0728 fingerprint, the source case haplotype was circulating among endemic cases prior to the outbreak. Genetic and temporal distances were not correlated for either RFLP 0.0142 (r(2) = − 0.05) or RFLP 0.0728 (r(2) = 0.09) when all isolates were analyzed. CONCLUSIONS: We found no evidence that endemic strains acquired mutations resulting in their emergence in outbreak form. We conclude that the propagation of these outbreaks was likely driven by the combination of characteristics of the source cases, contacts, and the environment. The role of whole-genome sequencing in understanding mycobacterial evolution and in assisting public health authorities in conducting contact investigations and managing outbreaks is important and expected to grow in the future. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1186/s12916-018-1112-9) contains supplementary material, which is available to authorized users.