Diagnostic performance of whole-genome sequencing for identifying drug-resistant TB in Thailand.

BACKGROUND: Whole-genome sequencing (WGS) is a promising tool for the detection of drug-resistant TB (DR-TB). To date, there have been few comparisons of diagnostic performance of WGS and phenotypic drug susceptibility testing (DST) in DR-TB.METHODS: We compared drug resistance-conferring mutations identified by WGS analysis using TB-Profiler and Mykrobe with phenotypic DST profiles based on the Löwenstein-Jensen proportion method using drug-resistant Mycobacterium tuberculosis (n = 537) isolates from across Thailand. Based on available phenotypic DST results, diagnostic performance was analysed for resistance against isoniazid, rifampicin, ethambutol (EMB), streptomycin, ethionamide (ETH), kanamycin, capreomycin (CPM), para-aminosalicylic acid, ofloxacin and levofloxacin.RESULTS: High agreement between the two methods was observed for most drugs (>91%), except EMB (57%, 95% CI 53-61) and ETH (70%, 95% CI 66-74). Also, low specificity was observed for EMB (49%, 95% CI 44-54) and ETH (66%, 95% CI 61-71). Sensitivity was high for most drugs (range 83-98%), except CPM (77%, 95% CI 59-88).CONCLUSION: Low agreement between WGS and phenotypic tests for drug resistance was found for EMB and ETH. The current genomic database is insufficient for the identification of CPM resistance. Challenges remain for routine usage of WGS-based DST, especially for second-line anti-TB drugs.

Genetic polymorphisms of CCL2 associated with susceptibility to latent tuberculous infection in Thailand.

BACKGROUND: The extent to which individuals exhibit genetic susceptibility to tuberculosis (TB) is still unclear. Genetic variations in chemokine genes might influence the early clearance of Mycobacterium tuberculosis, affecting TB susceptibility. OBJECTIVES: To study single nucleotide polymorphisms (SNPs) of chemokine genes CCL2, CXCL9, CXCL10 and CXCL11, and their association with TB susceptibility. DESIGN: Of 248 participants enrolled, 49 had active TB, 43 had latent tuberculous infection (LTBI) and 156 were non-infected, including 24 healthy controls with no known TB exposure. These populations were divided into two groups based on TB exposure: susceptible (n = 92) and resistant (early clearance) (n = 132). RESULTS: Only CCL2 SNPs (-2518A/G) were significantly associated with increased TB susceptibility. Based on adjusted multivariate analysis, persons with the GG genotype at this SNP were twice as susceptible to TB as those with the AA genotype (P = 0.018, OR 2.880, 95%CI 1.201-6.903). Risk of LTBI was three times higher among those with GG (P = 0.003, OR 3.358, 95%CI 1.525-7.396 for AA+AG vs. GG and P = 0.012, OR 3.706, 95%CI 1.340-10.254 for AA vs. GG). Persons with the GG genotype produced significantly lower CCL2 levels in response to M. tuberculosis antigen stimulation (AA+AG vs. GG, P = 0.038). CONCLUSION: The CCL2 polymorphism (-2518A/G) was associated with susceptibility to LTBI in a North-East Thai populations.