Description: BACKGROUNDThe incidence of tuberculosis infection in the UK is rising. Disease control depends on case detection by epidemiological investigation, currently aided by MIRU-VNTR genotyping. Although transmission between cases can be reliably ruled out where genotypes differ, matching genotypes do not by themselves confirm transmission. Whole-genome sequencing (WGS) measures the genetic distance between Mycobacterium tuberculosis strains precisely, promising to accurately delineate outbreaks without prior epidemiological data. Moreover, as accumulated mutations are rarely reversed, there exists the potential to infer direction of transmission from the evolving pattern of mutations, allowing the identification of potential super-spreaders.METHODSWe used Illumina technology to sequence 390 M. tuberculosis genomes from 254 patients from the Midlands region of the UK. We measured pairwise nucleotide differences between genomes within hosts (79 individuals) and between hosts within 25 household outbreaks (63 individuals), using the findings to interpret genetic trees constructed from 11 MIRU-VNTR-based community clusters (168 patients) with variable epidemiological data.FINDINGSWe estimated a molecular clock of 0·5 SNPs/genome/year from longitudinal isolates from 30 individuals and 25 families that suggested divergence over three years is rarely >5 SNPs. Indeed, only 5/114 (4%) paired isolates from individuals and households differed by >5 SNPs. A threshold of ≤5 SNPs was applied to 11 MIRU-VNTR-based community clusters: 0/69 epidemiologically-related pairs, 2/13 possibly-related pairs and 13/75 pairs with no known epidemiological relationship were separated by >5 SNPs (p<0·0001). The topology of the genetic trees signalled two super-spreaders, both supported epidemiologically: one was successfully predicted from the first isolates sequenced from this cluster.INTERPRETATIONWGS can delineate outbreaks with unprecedented resolution and allows inference on direction of transmission between cases. It can signal potential “super-spreaders” and predict the existence of undiagnosed cases, potentially leading to earlier treatment of infectious patients and their contacts.

Data type: Other

Sample scope: Monoisolate

Release date: 2012-11-15