Accuracy of the M-kit absolute concentration method for detection of drug-resistant mycobacterium tuberculosis in Uganda

Abstract

With an increasing incidence of multidrug-resistant tuberculosis (MDR-TB), accurate drug susceptibility testing (DST) of Mycobacterium tuberculosis to first-line and second-line plus new individualized anti-TB drugs has become crucial for proper patient management. Diagnosis of DR-TB is mainly performed using Lowenstein Jensen (LJ) and BACTEC MGIT 960 phenotypic DST methods at the Uganda National TB Reference Laboratory (NTRL). These methods have high skills demand and some require 4-6 weeks to report results with high contamination rates. Although WHO recently recommended molecular DST assays like GeneXpert and Hein MDRplus and MDRsl Line Probe Assay, these methods also present performance challenges with low coverage for drug-resistance-conferring mutations and few anti-TB agents tested. This study evaluated the diagnostic accuracy of the Multiplexing Kit (M-Kit) absolute concentration method in the detection of Drug-Resistant Mycobacterium Tuberculosis in Uganda, against the conventional LJ proportion phenotypic DST methods. A total of 212 freshly grown MTB Isolates from Tuberculosis patients referred to NTRL, were tested for drug resistance for Isoniazid, Rifampicin, Ethambutol, Moxifloxacin, Levofloxacin, Amikacin and P-Nitro Benzoic acid, on M-Kit and LJ proportion DST assay. Data obtained was entered in Microsoft Excel-2010 and statistical data analyzed using SPSS version 27 software; whereas inferential statistics, sensitivity, specificity, positive predictive value (PPV), negative predictive values (NPV) and Kappa value for M-Kit with respect to the gold standard method(s) were calculated using the MedCalc software. The sensitivity and specificity findings of the M-Kit assay for different studied drugs was; Isoniazid 98.3%, 97.9%, Rifampicin 94.8%, 96.9%, Ethambutol 68.6%, 96.7%, Moxifloxacin 100%, 99.6%, Levofloxacin 100%, 99.0% respectively, and Amikacin with specificity of 100%. The accuracy for various drugs studied ranged between 87.9 % to 99.9%, with an excellent concordance between the two methods showed by kappa values ranging from 0.806 to 0.962. The M-kit absolute concentration method was found to be highly accurate in susceptibility testing of M. tuberculosis against various first and second line drugs; Isoniazid, Rifampicin, Moxifloxacin, Levofloxacin, and Amikacin, with suboptimal performance for Ethambutol, when compared with the LJ proportion method in Uganda.