A pilot external quality assurance programme for line-probe assay detection of anti-tuberculosis drug resistance.

BACKGROUND: Multidrug-resistant tuberculosis (MDR-TB; resistance to isoniazid and rifampicin) is difficult to detect and control. Line-probe assays (LiPA) are widely used for the rapid detection of MDR-TB. OBJECTIVE: To ensure the quality of the test, a pilot external quality assurance (EQA) programme was initiated to assess the feasibility of running such a programme and the possibility of improving the proficiency of TB laboratories in performing the test. DESIGN: Prepared filter-paper-based Mycobacterium tuberculosis DNA samples were shipped to participant laboratories for LiPA EQA. The tests were performed blind, and the results were returned to the organising laboratory for comparison and analysis. RESULTS: A total of four rounds of EQA samples were dispatched to five laboratories in four countries. Overall inter- and intra-laboratory reproducibility was respectively 97% and 96%. The strengths and weaknesses of the participant laboratories in performing the test were discussed. CONCLUSION: A LiPA EQA programme can ensure quality and improve the performance of TB laboratories. This is a critical step during the initial stages at the time of setting up this method of testing.

Usefulness of resistant gene markers for predicting treatment outcome on second-line anti-tuberculosis drugs.

AIM: Mutations in rrs [nucleotide (nt) 1401], gyrA gene (codons 90, 91 or 94), tlyA, ethA and thyA genes of Mycobacterium tuberculosis (MTB) were evaluated for their usefulness in predicting treatment outcome of kanamycin (KM), capreomycin (CPM), ofloxacin (OFX), ethionamide (ETH) and para-aminosalicylic acid (PAS). METHODS AND RESULTS: DNA sequence analyses of these genes were performed against 188 MTB isolates obtained from patients put on second-line anti-TB drugs (SLDs) with well-documented clinical history and treatment outcome. Mutations in rrs and gyrA have 100% positive predictive value (PPV) in predicting treatment failure for KM and OFX, while 88·9 and 80% were obtained, respectively, when tlyA and rrs mutations were considered in CPM. For ETH and PAS, the PPV of using ethA and thyA mutations to predict treatment failure was 82·5 and 89·3%, respectively. CONCLUSIONS: Our study demonstrated high specificities of gene mutations in predicting poor treatment outcome; however, further technical advancement is required to make the molecular detection of resistances to other SLDs feasible in clinical laboratories. SIGNIFICANCE AND IMPACT OF THE STUDY: This is the first study to correlate different polymorphisms of major SLD resistance gene markers with predicted treatment outcome, using an international set of well-documented clinical MTB strains.