Performance of the low-cost phenotypic thin-layer agar MDR/XDR-TB Colour Test (first generation, 1G, Color Plate Test) for identifying drug-resistant Mycobacterium tuberculosis isolates in a resource-limited setting

Abstract Background The accessible, easy to use and timely, diagnosis of tuberculosis (TB) drug-susceptibility, is often challenging, particularly in resource-constrained settings. We therefore evaluated the phenotypic thin-layer agar based MDR/XDR-TB Colour Test, also known as the “First Generation (1G) Color Plate Test (TB-CX)” performance for detecting resistance of Mycobacterium tuberculosis ( Mtb ) isolates to selected anti-TB drugs versus other tests routinely used in our setting. Methods A cross-sectional study was conducted on Mtb clinical isolates stored at the Armauer Hansen Research Institute TB laboratory in Addis Ababa, Ethiopia. Drug-susceptibility testing was performed on 78 Mtb isolates for isoniazid, rifampicin, and moxifloxacin using the Colour Test and the Indirect Proportional Method (IPM) “in house” assay. Isoniazid and rifampicin were also evaluated by the Mycobacterial Growth Indicator Tube (MGIT) commercially available assay. Test accuracy was calculated as % agreement with 95% confidence intervals (95%CI). Results The median (range) times in days determining Mtb resistance or susceptibility for the Colour Test, IPM and MGIT assays were of 9 (5–18), 15 (13–18) and 19 (14–21) days, respectively. The Colour Test provided results significantly ( p  < 0.001) more rapidly than the IPM or MGIT assays. The colour test showed a sensitivity and specificity of 91%(95% CI: 87–96) and 87%(95% CI:75–95) for detecting isoniazid resistance,and 93%(95% CI:81–99) and 92%(95% CI:82–97) for detecting rifampicin resistance, respectively, when compared to MGIT DST. For detecting MDR-TB the sensitivity and specificity were 90%(95% CI:76–97) and 96%(95% CI:88–99),respectively. The colour test showed a sensitivity of 97%(95%CI = 87–100) and specificity of 89% (95%CI = 79–96) for detecting isoniazid resistance while for rifampicin resistance,it showed a sensitivity of 82%(95%CI = 64–93)and a specificity of 80%(95% CI = 68–90) rifampicin resistance. Colour Test accuracy compared to IPM to detect isoniazid, rifampicin resistance and MDR-TB was 92% (95%CI = 86–98), 81% (95%CI = 72–90), and 90% (95%CI = 83–96). IPM test accuracy compared to MGIT DST for detecting isoniazid and rifampicin resistance and MDR-TB was 91% (95%CI = 85–97), 83% (95%CI = 75–92), and 85% (95%CI = 77–93), respectively. Moxifloxacin drug-susceptibility testing could not be assessed because only two isolates showed evidence of resistance. Conclusion The accuracy of Mtb drug-susceptibility testing was similar comparing: Colour Test versus IPM, Colour Test versus MGIT; and comparing IPM versus MGIT. The Colour Test was easy to use and determined drug-susceptibility significantly more rapidly than the IPM and MGIT assays. Thus, implementing the Colour Test in clinical settings could make drug-susceptibility testing more accessible and rapid in high TB burden, and resource-constrained settings, including in Ethiopia.