The role of molecular diagnosis of drug resistant tuberculosis

Abstract

Emerging multidrug-resistant tuberculosis (MDR-TB) is one of the most urgent global public health issues. Recent advances in molecular techniques should enable the development of different rapid detection tests for drug-resistant TB. Large-scale comparative studies on the diagnostic accuracy and turn-around-time (TAT) of these novel assays may promote their smooth implementation as routine tests for TB in diagnostic laboratories.

In a pilot evaluation of 30 clinical isolates and 202 sputum specimens, diagnostic performance of a novel in-house assay for MTB identification (IS6110 qPCR) was compared to a commercial COBAS TaqMan MTB test (Roche Diagnostics). The sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of IS6110 qPCR were 100%, 94.6%, 85.2% and 100%, respectively, compared to 94.7%, 100%, 100% and 98.6% for COBAS TaqMan MTB. Large-scale validation using 2,350 sputum specimens revealed the optimal cut-off crossing point (Cp) value of IS6110 qPCR was 29.61 with 97.3% sensitivity and 98.3% specificity determined by receiver operating characteristics (ROC) curve analysis. The median TAT for IS6110 qPCR and COBAS TaqMan MTB test to the reporting of results was 0.9 and 1.2 days, respectively.

Among the IS6110 qPCR-positive specimens in the large-scale validation, 287 samples were tested in-house by katG MAS-PCR and rpoB PCR sequencing assays and 159 samples were tested by GenoType® MTBDRplus assay (Hain LifeScience). The sensitivity and specificity of katG MAS-PCR for isoniazid (INH) resistance detection were 71.4% and 99.5%, respectively. The sensitivity and specificity of rpoB PCR sequencing for rifampicin (RIF) resistance detection were 100% and 99.6%, respectively. Commercial GenoType® MTBDRplus assay reached 100% sensitivity for both INH and RIF resistance detection at a specificity of 99.3% and 100%, respectively. The median TAT for the in-house assays and GenoType® MTBDRplus assay to the reporting of the results was 4.7 and 1.4 days, respectively.

The findings from this study provide different implementation strategies for diagnostic test combinations. The most cost-effective drug-resistant TB diagnosis cascade was IS6110 qPCR followed by GenoType® MTBDRplus assay. The TAT for results is 2.3 days at a cost of US$49.7. Despite an additional cost of US$24.6, COBAS TaqMan MTB test should replace IS6110 qPCR in populations with high prevalence of IS6110-negative strains. The in-house katG MAS-PCR and rpoB PCR sequencing assays should be used in developing countries instead of the expensive GenoType® MTBDRplus assay. Subsequently, accurate diagnosis of drug-resistant tuberculosis can be achieved in 4.5 days with a reasonable reagent cost of US$9.3. In conclusion, excellent diagnostic accuracy and shorter TAT of the molecular diagnostic cascade for drug-resistant TB, in particular IS6110 qPCR, can serve to guide physicians in the prompt choice of chemotherapy. This leads to timely delivery of anti-TB treatments to patients and holds the promise of easing the MDR-TB burden.