Utilizing whole genome sequencing to delineate relapse and reinfection tuberculosis on the Canadian prairies

Abstract

RATIONALE Recurrent tuberculosis (TB) accounts for 5% of the Canadian TB burden. Recurrence can occur from relapse or reinfection. Identifying reinfection has implications for developing control policies. OBJECTIVE The objectives of this study were to quantify reinfection in recurrent TB using whole genome phylogenetic and single nucleotide polymorphism (SNP) and compare with epidemiological and clinical parameters from the Canadian prairies. METHODS DNA sequences of Mycobacterium tuberculosis isolates from recurrent TB cases along with epidemiological and clinical parameters were collected from Alberta and Saskatchewan. Inclusion criteria were two or more culture positive notifications age ≥17 years. SNP and phylogenetic tree scale differences (TIP) were used to determine the relapse and reinfection categories. RESULTS Of 7,627 notifications of TB disease, 533 were recurrent. 93 pairs (180 cases) were culture positive from which 26 (50 cases) were available for sequencing. 19 cases with SNP and TIP values ≤25 and ≤.001 were classified relapse. Seven SNP and TIP values >160 and >.001 were classified reinfection. Five of seven reinfections were Indigenous cases from high TB incidence areas. The non-sequenced and sequenced pairs differed only in age. CONCLUSIONS Recurrent culture positive TB is uncommon on the Canadian prairies and is more likely to be relapse. Reinfection is more likely in Indigenous persons living in high TB incidence communities. A limitation was the risk of selection bias since only 28% of eligible cases were sequenced. Since we did not know the non-sequenced reinfection risk, we could only conclude that the non-sequenced and sequenced categories were similar.