Whole-genome sequencing differentiates relapse from re-infection in TB.

Distinguishing TB relapse from re-infection is important from a clinical perspective to document transmission patterns. We investigated isolates from patients classified as relapse to understand if these were true relapses or re-infections. We also investigated shifts in drug susceptibility patterns to distinguish acquired drug resistance from re-infection with resistant strains. Isolates from pulmonary TB patients from 2009 to 2017 were analysed using whole-genome sequencing (WGS). Of 11 patients reported as relapses, WGS results indicated that 4 were true relapses (single nucleotide polymorphism difference ≤5), 3 were re-infections with new strains, 3 were both relapse and re-infection and 1 was a suspected relapse who was later categorised as treatment failure based on sequencing. Of the 9 patients who went from a fully susceptible to a resistant profile, WGS showed that none had acquired drug resistance; 6 were re-infected with new resistant strains, 1 was probably infected by at least two different genotype strains and 2 were phenotypically misclassified. WGS was shown to distinguish between relapse and re-infection in an unbiased way. The use of WGS minimises the risk of false classification of treatment failure instead of re-infection. Furthermore, our study showed that strains without major genetic differences can cause re-infection.