The evaluation of rapid screening of M/XDR-TB patients within a dedicated M/XDR-TB hospital in Gauteng, South Africa.

[摘要] BackgroundIn 2006, South Africa documented its first outbreak of extensively drug-resistant (XDR-TB) in Tugela Ferry, Kwa-Zulu Natal. Delayed diagnosis of XDR-TB increases mortality, emphasizing the need or rapid diagnostics. The MTBDRsl assay rapidly identifies resistance to fluoroquinolones (FLQ) and aminoglycosides/capreomycin (AG/CP). Hospitalization provides the ideal opportunity for nosocomial infections of TB, including M/XDR-TB, with its associated morbidity and mortality. Occupational exposures amongst healthcare workers are also of concern. Rapid knowledge of second-line resistance patterns of these in-patients would allow for quick stratification and administration of individualized anti-chemotherapy treatment by clinicians, thereby reducing morbidity and mortality within this population.Methods `A prospective cohort study was performed from admission sputum specimens collected from patients admitted to Sizwe Tropical Disease Hospital (Sizwe Hospital), the only M/XDR-TB referral hospital in Gauteng, South Africa. MTBDRplus (version 1) and MTBDRsl line probe assays (LPAs) were performed on all sputa, irrespective of smear positivity and subsequently on Mycobacteria Growth Indicator Tube (MGIT) isolates. The results from the LPAs were compared to the gold standard MGIT 960 culture and direct susceptibility testing (DST) results.ResultsFrom April 2011 to January 2012, 150 participants were recruited. Seventy-five (50.0%) were female, 91 (60.7%) were HIV-positive and 9 (6.00%) had an unknown HIV status. Of the MGIT cultures performed 71 (47.3%) were positive for Mycobacterium tuberculosis (MTB), 31 (20.7%) negative, 40 (26.7%) were contaminated and 8 (5.63%) were inadvertently discarded by the routinen laboratory. The proportion of smear-negative specimens in HIV- positive patients was not statistically significantly different to the smear-negatives in HIV-negative patients (57.1% vs. 46.0%; p=0.139). On phenotypic drug susceptibility testing, 47 (66.2%) were resistant to only isoniazid and rifampicin (MDR-TB), 10 (14.0%) were MDR-TB with added resistance to ofloxacin (pre-XDR-TB FLQ), 4 (5.63%) were MDR-TB with added resistance to kanamycin (pre-XDR-TB AG/CP), 4 (5.63%) were MDR-TB with added resistance to both ofloxacin and kanamycin (XDR-TB), 3 (4.23%) were mono-isoniazid resistant, 2 (2.82%) mono-rifampicin resistant, and 1 (1.41%) was TB-drug sensitive.No TB drug-sensitive sputum specimens were collected and used as a control group for the MTBDRplus (version 1) assay, thus only sensitivity indices and positive predictive values (PPVs) could be interpreted.All specimens collected were used as an inherent control group for the performance of the MTBDRsl assay, thus sensitivity and specificity indices together with the PPVs and NPVs, were calculated.From all sputum samples collected, including both smear-positives and smear-negative specimens, the sensitivity of the MTBDRplus (version 1) for RIF and INH was 98.1% for both drugs. Overall sensitivity for the MTBDRsl assay to detect ofloxacin (OFX) and kanamycin (KAN) resistance was 90.9% and 42.9%, respectively. The specificity was 100.0% and 95.7%, respectively.From smear-positive specimens, the sensitivity for OFX and KAN was 100.0% and 50.0%, respectively. The specificity was 100.0% and 96.8%, respectively. From smear-negative specimens the sensitivity was 50.0% and 33.3%, respectively. The specificity was 100.0% and 92.9%, respectively.From MGIT culture isolates, sensitivity for RIF and INH was 98.2% and 94.5% respectively. The sensitivity for OFX and KAN were 100.0% and 42.9%, respectively, whereas the specificity was 95.6% and 100.0%, respectively.ConclusionThe performances of both LPAs correlate with other local and international publications. On direct, smear-positive sputa both LPAs perform well. From these, the MTBDRsl assay is a good rule-out test for detection of resistance to FLQ and AG/CP. For FLQ, it’s a good rule-in test: this indicates that the MTBDRsl assay can be used as a rapid screen for the onset of XDR-TB. As predicted, MTBDRsl does not perform well on smear-negative specimens, whereas the MTBDRplus (version 1) does. On cultures, both LPAs perform well.

[发布机构] University of the Witwatersrand