Discordance between Genotypic and Phenotypic Method For The Detection of Rifampicin and Isoniazid Resistant Mycobacterium Tuberculosis and Its Effect on Patient Treatment Outcomes

Abstract

Background: World Health Organization (WHO) recommended both genotypic and phenotypic methods for M. tuberculosis (MTB) drug susceptibility testing (DST). Discordance between genotypic and phenotypic methods for the detection of rifampicin (RIF) and isoniazid (INH) resistant MTB is poorly documented in Southwestern Oromia, Jimma, Ethiopia. This study aimed to determine discordance between genotypic and phenotypic method for the detection of RIF and INH resistant MTB and its effect on patient treatment outcomes in Southwestern Oromia, Jimma, Ethiopia. Methods: A cross-sectional study was conducted on 57 sputum Xpert MTB/RIF assay (Xpert) positive MTB isolates (45 RIF resistant from new and retreatment TB patients, and 12 RIF susceptible from retreatment TB patients) stored at Jimma University Mycobacteriology Research Center (JUMRC) laboratory. All isolates were sub-cultured on Lowenstein–Jensen (LJ) medium. DNA was extracted from LJ culture positive isolates for the detection of RIF and INH resistance by line probe assay (LPA). In parallel, phenotypic DST for RIF, INH, streptomycin (STR), and ethambutol (EMB) was performed by MGIT 960. Socio-demographic and clinical data were extracted from JUMRC laboratory registration book. Treatment outcomes data were extracted from patient medical records. Online GraphPad prism software was used to analyze the discordance between the 2 DST methods. Results: Overal1, 6 (10.5%) isolates had discordant results between genotypic and phenotypic DST method for RIF and or INH (3 for RIF, 1 for INH, and 2 for both RIF and INH). There was higher agreement between LPA and phenotypic DST for both RIF and INH (kappa=0.86), but lower between Xpert and phenotypic DST (kappa=0.76). Four Xpert RIF resistant isolates were RIF susceptible by phenotypic DST. Two RIF resistant isolates on LPA were RIF susceptible by phenotypic DST. For INH, 3 INH susceptible isolates on LPA were INH resistant by phenotypic DST. A total of 11 (19.3%) patients had unfavorable outcome: 3 (5.3%) were loss to follow up, 5 (8.7%) died, and 3 (5.3%) failed treatment. Out of these, 3 (27.3%) had discordant results for RIF and or INH, and all of them were died. Conclusion: Presence of discordance between genotypic and phenotypic DST for RIF and INH has clinical impact on patient treatment outcome. Thus, parallel use of rapid molecular assay with phenotypic DST would help to improve patient outcome.