Discordance between genexpert, line probe assay and drug susceptibility test in assessing drug-resistant tuberculosis

Abstract

Background: Rapid molecular diagnostics have potentially revolutionized early detection of drug-resistant tuberculosis (DR-TB) in Indonesia. However, there is discordance between conventional culture using drug susceptibility test (DST) and rapid diagnostic tools using GeneXpert and line probe assay (LPA). This discordance result can cause confusion to clinician in determining diagnosis of DR-TB. Objective: This study aimed to identify discordance between GeneXpert, LPA, and DST. Methods: A retrospective study was conducted at Dr. Soetomo General Academic Hospital, Surabaya, Indonesia. Data were collected based on medical record between third to fourth quarter of 2018. Rifampicin resistant-tuberculosis (RR-TB) proven by GeneXpert, were further analyzed with second-line LPA and DST. Discordance result between it was analyzed using McNemar. Results: Among 81 patients diagnosed with DR-TB, 59 patients RR-TB were eligible in this study. There were 13 (22.0%) DST rifampicin result showed sensitive but resistant to GeneXpert. Among 53 samples from LPA, there were 3 (5.7%) result showed LPA fluoroquinolone resistant but sensitive to DST ofloxacin; 10 (18.9%) result has LPA fluoroquinolone sensitive but resistant to DST ofloxacin; 2 (3.8%) has LPA second-line injectable drug (LPA SLID) resistant but sensitive to DST kanamycin; 5(9.4%) has LPA SLID sensitive but resistant to DST kanamycin. The McNemar analysis showed discordance between GeneXpert and DST rifampicin was 13 (22.0%; p=0.046); LPA fluoroquinolone and DST ofloxacin was 13 (24.6%; p=0.042); LPA SLID and DST kanamycin was 7 (13.2%; p=0.183). No variable that can be used to analyze discordance result between GeneXpert and LPA. Conclusion: There is significant discordance between GeneXpert and DST; LPA fluoroquinolone and DST ofloxacin, while neither LPA SLID nor DST kanamycin shows no significant discordance.