Extensively drug-resistant (XDR) tuberculosis (TB) cannot be easily or quickly diagnosed. We developed a rapid, automated assay for the detection of XDR-TB plus resistance to the drug isoniazid (INH) for point-of-care use. Using a simple filter-based cartridge with an integrated sample processing function, the assay identified a wide selection of wild-type and mutant sequences associated with XDR-TB directly from sputum. Four new large-Stokes-shift fluorophores were developed. When these four Stokes-shift fluorophores were combined with six conventional fluorophores, 10-color probe detection in a single PCR tube was enabled. A new three-phase, double-nested PCR approach allowed robust melting temperature analysis with enhanced limits of detection (LODs). Finally, newly designed sloppy molecular beacons identified many different mutations using a small number of probes. The assay correctly distinguished wild-type sequences from 32 commonly occurring mutant sequences tested in gyrA, gyrB, katG, and rrs genes and the promoters of inhA and eis genes responsible for resistance to INH, the fluoroquinolone (FQ) drugs, amikacin (AMK), and kanamycin (KAN). The LOD was 300 CFU of Mycobacterium tuberculosis in 1 ml sputum. The rate of detection of heteroresistance by the assay was equivalent to that by Sanger sequencing. In a blind study of 24 clinical sputum samples, resistance mutations were detected in all targets with 100% sensitivity, with the specificity being 93.7 to 100%. Compared to the results of phenotypic susceptibility testing, the sensitivity of the assay was 75% for FQs and 100% each for INH, AMK, and KAN and the specificity was 100% for INH and FQ and 94% for AMK and KAN. Our approach could enable testing for XDR-TB in point-of-care settings, potentially identifying highly drug-resistant TB more quickly and simply than currently available methods.KEYWORDS 10-color assay, three-phase PCR, XDR-TB, point-of-care test

eResistance to amikacin (AMK) and kanamycin (KAN) in clinical Mycobacterium tuberculosis strains is largely determined by specific mutations in the rrs gene and eis gene promoter. We developed a rapid, multiplexed sloppy molecular beacon (SMB) assay to identify these mutations and then evaluated assay performance on 603 clinical M. tuberculosis DNA samples collected in South Korea. Assay performance was compared to gold-standard phenotypic drug susceptibility tests, including LowensteinJensen (LJ) absolute concentration, mycobacterial growth indicator tubes (MGIT), and TREK Sensititre MycoTB MIC plate (MycoTB) methods. Target amplicons were also tested for mutations by Sanger sequencing. The SMB assay correctly detected 115/116 mutant and mixed sequences and 487/487 wild-type sequences (sensitivity and specificity of 99.1 and 100%, respectively). Using the LJ method as the reference, sensitivity and specificity for AMK resistance were 92.2% and 100%, respectively, and sensitivity and specificity for KAN resistance were 87.7% and 95.6%, respectively. Mutations in the rrs gene were unequivocally associated with high-level cross-resistance to AMK and KAN in all three conventional drug susceptibility testing methods. However, eis promoter mutations were associated with KAN resistance using the MGIT or MycoTB methods but not the LJ method. No testing method associated eis promoter mutations with AMK resistance. Among the discordant samples with AMK and/or KAN resistance but wild-type sequence at the target genes, we discovered four new mutations in the whiB7 5= untranslated region (UTR) in 6/22 samples. All six samples were resistant only to KAN, suggesting the possible role of these whiB7 5= UTR mutations in KAN resistance.T uberculosis (TB) was declared a global public emergency nearly 20 years ago (1). Although the rate of new cases of TB has been decreasing worldwide, the millennium developmental goal target of a 50% disease reduction by 2015 is unlikely to be achieved (1). An increase in the incidence of multidrug-resistant (MDR) and extensively drug resistant (XDR) TB is a serious threat to these reduction goals (1). Patients with drug-resistant TB are best identified as rapidly as possible so that appropriate infection control and treatments can be quickly initiated (2).Conventional phenotypic methods can take weeks to months to fully define the drug resistance pattern of Mycobacterium tuberculosis isolates due to the very slow growth of this bacterium (3-5). Molecular tests offer the promise of more rapid drug resistance detection. A number of tests are available to detect resistance to many of the first-line anti-TB drugs (6-8). However, there are fewer rapid tests available to test for resistance to the injectable second-line drugs amikacin (AMK), kanamycin (KAN), and capreomycin (CAP). DNA sequencing studies suggest that most cases of resistance to injectable drugs can be identified by detecting mutations in positions 1401, 1402, and 1484 in the M. tuberculosis 16S rRNA (rrs) gene and between nucleotides Ϫ...

M ultidrug-resistant (MDR) and extensively drug-resistant (XDR) Mycobacterium tuberculosis is increasing worldwide (9, 27, 37). Rapid methods to detect drug resistance are needed to quickly identify drug-resistant strains and to implement appropriate therapy (4,16,36). M. tuberculosis does not naturally contain plasmids, and almost all cases of clinical drug resistance are caused by single-nucleotide polymorphisms (SNPs) or small insertions/deletions in relevant genes (28). In the case of rifampin resistance, 95 to 98% of rifampin-resistant clinical strains have mutations in the 80-bp rifampin resistance determining region (RRDR) of the M. tuberculosis RNA polymerase beta (rpoB) gene (9,11,12,15,20,28). PCR and probe-based molecular genotyping assays can be used to detect these resistance-inducing mutations. Such genotypic assays are potentially more rapid than labor-intensive culture-based drug susceptibility tests. Genotypic drug susceptibility testing has shown good overall concordance with the phenotypic antibiotic susceptibility tests of MDR and XDR clinical strains (6), and in a recent study, a genotypic test actually correlated better with clinical outcome than standard phenotypic susceptibility testing (36). When combined with automated sample processing systems, such as the Xpert MTB/RIF test (14), genotypic susceptibility tests can significantly reduce testing turnaround time, increasing patient notification rates and decreasing time to treatment (4, 32).The Xpert MTB/RIF assay is one example of a genotypic test that is being increasingly used to screen for rifampin resistance (33). However, the single-use cartridge design of the Xpert assay limits its use for laboratory-based high-throughput testing. Several widely used reverse blot hybridization assays, such as the INNO-LIPA Rif.TB assay (Innogenetics, Belgium) and the MTBDRplus (Hain, Germany) assay (1,5,15,19,21,23,29,34), are available for laboratory-based rifampin resistance screening; however, these assays are complicated by their open hybridization format. Open hybridization systems require a relatively cumbersome work process, including rigorous physical separation of different work areas (2, 23) due to the risk of handling open PCR amplicons in a molecular diagnostic laboratory. Open systems also require a relatively large number of probes to test for relevant resistanceassociated mutations. This requirement complicates assay chemistry and hybridization parameters. In contrast to reverse blot hy-

Several molecular assays to detect resistance to Rifampin, the Fluoroquinolones, and Aminoglycosides in Mycobacterium tuberculosis (M. tuberculosis) have been recently described. A systematic approach for comparing these assays in the laboratory is needed in order to determine the relative advantage of each assay and to decide which ones should be advanced to evaluation. We performed an analytic comparison of a Sloppy Molecular Beacon (SMB) melting temperature (Tm) assay and a Dual labeled probe (DLP) Tm assay. Both assays targeted the M. tuberculosis rpoB, gyrA, rrs genes and the eis promoter region. The sensitivity and specificity to detect mutations, analytic limit of detection (LOD) and the detection of heteroresistance were tested using a panel of 56 clinical DNA samples from drug resistant M. tuberculosis strains. Both SMB and DLP assays detected 29/29 (100%) samples with rpoB RRDR mutations and 3/3 (100%) samples with eis promoter mutations correctly. The SMB assay detected all 17/17 gyrA mutants and 22/22 rrs mutants, while the DLP assay detected 16/17 (94%) gyrA mutants and 12/22 (55%) rrs mutants. Both assays showed comparable LODs for detecting rpoB and eis mutations; however, the SMB assay LODs were at least two logs better for detecting wild type and mutants in gyrA and rrs targets. The SMB assay was also moderately better at detecting heteroresistance. In summary, both assays appeared to be promising methods to detect drug resistance associated mutations in M. tuberculosis; however, the relative advantage of each assay varied under each test condition.