Click on the 'Resources' tab to download the ePoster. 

Clinical Evaluation of the Xpert^® MTB/XDR Assay in China and South Africa

Q. Jiang¹, H. Zhu², S. Gong², W. Wang² , J. Gao², D.T. Armstrong³, X. Liu⁴, X. Yu³, X. Yuan⁴, G. Zhang⁵, R.Y. Chen³, Q. Gao¹, C.E. Barry 3^rd,3, N.A. Ismail⁶, L.E. Via³, S.V. Omar⁶

¹Fudan University, Shanghai, China; ²Sino-US Tuberculosis Collaborative Research Program, Zhengzhou, Henan, China; ³Tuberculosis Research Section, National Institute of Allergy and Infectious Diseases, U.S. National Institutes of Health, Bethesda, MD, USA; ⁴Henan Provincial Chest Hospital, Zhengzhou, Henan, China; ⁵Henan Provincial Institute of Tuberculosis and Prevention, Henan Center for Disease Control, Zhengzhou, Henan, China; ⁶Centre for Tuberculosis, National TB Reference Laboratory, National Institute for Communicable Disease, Republic of South Africa.

Multidrug resistant tuberculosis (MDR-TB) cases are increasing worldwide. New chemotherapeutics useful for both drug susceptible and resistant disease as a PAN-TB regimen are in testing. Until these can be established, accurate and rapid diagnostics both to conclusively identify infection with Mycobacterium tuberculosis (MTB) and detect resistance to drugs in current TB regimens are vital to successful treatment. We assessed the diagnostic accuracy of a cartridge-based molecular assay for the detection of MTB and resistance to aminoglycosides (SLI), ethionamide (ETH), fluoroquinolones (FLQ), and isoniazid (INH) relative to known reference results for MTB culture, phenotypic drug susceptibility testing (pDST) and DNA sequencing of selected targets aligning to the assay. We further assessed the ability of the cartridge to differentiate low and high level INH resistance. The positive and negative percent agreement (PPA and NPA) for the Xpert MTB/XDR assay was determined relative to the Xpert MTB/RIF and Xpert MTB/RIF Ultra assays. The XDR assay accuracy study was conducted on 531 eligible raw sputum samples or sputum sediments archived specifically for this purpose. The sensitivity of the Xpert MTB/XDR test relative to pDST for drug resistance ranged from 65% for ETH to 93.1% for FLQ. The specificity was ≥98.3% for all drug resistance targets. The sensitivity of the Xpert MTB/XDR test relative to sequencing for drug resistance was ≥96.3% for all drugs except FLQ which had a sensitivity of 93.3%. Specificity was 100% except for INH, which had a specificity of 98.7%. Concordance for the detection of low INH resistance detection was 97.6% compared to sequencing data (41/42) with the overall concordance at 99.6% for INH resistance. In terms of comparing the new assay to the assays currently deployed, PPA and NPA of the Xpert MTB/XDR test as compared to the Xpert MTB/RIF test for MTB detection were 98.9% and 93.8%, respectively. PPA and NPA of the Xpert MTB/XDR test as compared to the Xpert MTB/RIF Ultra test for MTB detection were 99.5% and 100.0%, respectively. The lower performance of the assay for ETH was expected as mutations conferring ETH resistance are reported to be in other genomic regions not targeted by the assay. Discordant results were examined and found most likely to result from very low or trace levels of bacteria in the samples as the limit of detection for both the Xpert MTB/RIF and MTB/RIF Ultra test is lower than the Xpert MTB/XDR test. The Xpert MTB/XDR assay accurately detected MTB mutations associated with resistance to INH, FLQ, and SLI, as well as distinguishing low and high level INH resistance. The MTB/XDR assay holds promise as a rapid point-of-care test to guide regimen composition for patients with drug-resistant tuberculosis.