Evaluation of the broth microdilution plate methodology for susceptibility testing of Mycobacterium tuberculosis in Peru

Abstract

Background: Tuberculosis (TB) is a communicable, preventable and curable disease caused by the bacterium Mycobacterium tuberculosis (MTB). Peru is amongst the 30 countries with the highest burden of multidrug-resistant tuberculosis (MDR-TB) worldwide. In the fight against drug-resistant tuberculosis, the UKMYC6 microdilution plate was developed and validated by the CRyPTIC project. The objective of the study was to evaluate the use of the broth microdilution (BMD) plate methodology for susceptibility testing of drug-resistant MTB strains in Peru. Methods: MTB strains isolated between 2015 and 2018 in Peru were used. 496 nationally-representative strains determined as drug-resistant by the routine 7H10 Agar Proportion Method (APM) were included in the present study. The Minimum Inhibitory Concentration (MIC) of 13 antituberculosis drugs were determined for each strain using the UKMYC6 microdilution plates. Diagnostic agreement between APM and BMD plate methodology was determined for rifampicin, isoniazid, ethambutol, ethionamide, kanamycin and levofloxacin. Phenotypes were set using binary (or ternary) classification based on Epidemiological cut-off values (ECOFF/ECV) proposed by the CRyPTIC project. Whole Genome Sequencing (WGS) was performed on strains with discrepant results between both methods. Results: MIC distributions were determined for 13 first- and second-line anti-TB drugs, including new (bedaquiline, delamanid) and repurposed (clofazimine, linezolid) agents. MIC results were available for 80% (397/496) of the strains at 14 days and the remainder at 21 days. The comparative analysis determined a good agreement (0.64 ≤ k ≤ 0.79) for the drugs rifampicin, ethambutol, ethionamide and kanamycin, and the best agreement (k > 0.8) for isoniazid and levofloxacin. Overall, 12% of MIC values were above the UKMYC6 plate dilution ranges, most notably for the drugs rifampicin and rifabutin. No strain presented MICs higher than the ECOFF/ECV values for the new or repurposed drugs. Discrepant analysis using genotypic susceptibility testing by WGS supported half of the results obtained by APM (52%, 93/179) and half of those obtained by BMD plate methodology (48%, 86/179). Conclusions: The BMD methodology using the UKMYC6 plate allows the complete susceptibility characterization, through the determination of MICs, of drug-resistant MTB strains in Peru. This methodology shows good diagnostic performances for rifampicin, isoniazid, ethambutol, ethionamide, kanamycin and levofloxacin. It also allows for the characterization of MICs for other drugs used in previous years against tuberculosis, as well as for new and repurposed drugs recently introduced worldwide.