Evaluation of a novel line-probe assay for genotyping-based diagnosis of Mycobacterium tuberculosis in Thailand.

SETTING: The Supranational Tuberculosis Reference Laboratory (NTRL), Bangkok, and Chiangrai Prachanukroh Hospital, Chiangrai, Thailand OBJECTIVE: To evaluate the diagnostic performance of newly developed line-probe assay (LiPA) kits in tuberculosis (TB) endemic settings. DESIGN: LiPA kits were used to evaluate 404 clinical isolates of Mycobacterium species and 163 sputum samples in Thailand. RESULTS: LiPA kits were able to identify M. tuberculosis, M. avium, M. intracellulare and M. kansasii with 100% sensitivity and specificity when compared with the commercially available AccuProbe assay. Testing of the LiPA kits for their ability to detect mutations in clinical isolates resistant to anti-tuberculosis drugs such as rifampicin, isoniazid, pyrazinamide and fluoroquinolones showed that the assay had very high sensitivity (65.9-100%) and specificity (98.2-100%) compared with drug susceptibility testing and DNA sequencing. LiPA had a sensitivity of 75.0-85.7% and a specificity of 96.4-100% in testing clinical sputum samples. CONCLUSION: The novel LiPA kits have high sensitivity and specificity, and may enhance the rapid detection of first- and second-line anti-tuberculosis drug resistance, improving the selection of suitable chemotherapy agents to treat multidrug-resistant and extensively drug-resistant TB.

Diagnostic value of blood gene expression signatures in active tuberculosis in Thais: a pilot study.

Tuberculosis (TB) is a major global health problem. Routine laboratory tests or newly developed molecular detection are limited to the quality of sputum sample. Here we selected genes specific to TB by a minimum redundancy-maximum relevancy package using publicly available microarray data and determine level of selected genes in blood collected from a Thai TB cohort of 40 active TB patients, 38 healthy controls and 18 previous TB patients using quantitative real-time PCR. FCGR1A, FCGR1B variant 1, FCGR1B variant 2, APOL1, GBP5, PSTPIP2, STAT1, KCNJ15, MAFB and KAZN had significantly higher expression level in active TB individuals as compared with healthy controls and previous TB cases (P<0.01). A mathematical method was applied to calculate TB predictive score, which contains the level of expression of seven genes and this score can identify active TB cases with 82.5% sensitivity and 100% specificity as compared with conventional culture confirmation. In addition, TB predictive scores in active TB patients were reduced to normal after completion of standard short-course therapy, which was mostly in concordant with the disease outcome. These finding suggested that blood gene expression measurement and TB Sick Score could have potential value in terms of diagnosis of TB and anti-TB treatment monitoring.