INT J TUBERC LUNG DIS 16(11):14711476 2012 The Union http://dx.doi.org/10.5588/ijtld.11.

0602 E-published ahead of print 12 September 2012

The Xpert MTB/RIF assay evaluation in South Korea, a country with an intermediate tuberculosis burden
S. Y. Kim,* H. Kim,* S. Y. Kim,* E. K. Ra,* S-I. Joo,* S. Shin, M-W. Seong,* C-G. Yoo, E-C. Kim,* S. S. Park*
* Department of Laboratory Medicine, Seoul National University Hospital, Seoul, Department of Laboratory Medicine, National Medical Center, Seoul, Department of Laboratory Medicine, Seoul National University Boramae Hospital, Seoul, Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, Lung Institute of Medical Research Center, Seoul National University College of Medicine, Seoul, South Korea SUMMARY

A central hospital laboratory in South Korea. To evaluate the usefulness of the Xpert MTB/RIF assay in a country with an intermediate tuberculosis burden. D E S I G N : A total of 71 real-time polymerase chain reaction-positive sputum sediments were tested within 24 h by the Xpert MTB/RIF assay. Mycobacterium tuberculosis detection was compared with smear microscopy and culture. Rifampicin (RMP) resistance was compared with a culture-based method and rpoB gene sequencing. We also assessed the limit of detection for mutant proportions and time savings in diagnosis. R E S U LT S : The Xpert MTB/RIF assay detected M. tuberculosis in 71 (100%) specimens (32 smear-positive, 39 smear-negative). This assay showed 100% (62/62) concordance with drug resistance confirmed by culture
SETTING: OBJECTIVE:

and 98.4% (61/62) concordance with sequencing. A specimen containing approximately 50% of mutant p.His526Tyr was falsely interpreted as wild-type bacilli by this assay. The minimal detection ratio was 5:1 of mutant vs. wild-type cells. The median time saved was 18.5 days (range 930) for the diagnosis of M. tuberculosis and 81.5 days (65136) for RMP susceptibility in smear-negative, culture-positive patients. C O N C L U S I O N S : The Xpert MTB/RIF assay showed high sensitivity in detecting M. tuberculosis with information on RMP resistance, and had a more rapid time to diagnosis compared to conventional tests; however, the location and amount of mutation may affect test sensitivity. K E Y W O R D S : M. tuberculosis; rifampicin resistance; automated real-time PCR

TUBERCULOSIS (TB) is a major health problem worldwide. The global prevalence of Mycobacterium tuberculosis infection was approximately 14 million in 2009.1 Multidrug-resistant TB (MDR-TB) comprised 4.8% of the total number of TB cases.2 Despite the prevalence of TB and of drug-resistant forms of the disease, current diagnostic methods, including smear microscopy, culture and nucleic acid amplification tests, have not provided sensitive, accurate and rapid information for timely diagnosis and treatment of the disease. The World Health Organization recently recommended that the Xpert MTB/RIF assay (Cepheid, Sunnyvale, CA, USA) be utilised as the initial diagnostic test in individuals suspected of MDR-TB or co-infected with the human immunodeficiency virus (HIV) and TB due to the low sensitivity of smear microscopy. This assay is a cartridge-based test that consists of a hands-free sputum processing system and an automated real-time hemi-nested polymerase chain reaction (PCR) that amplifies the 81 base pair (bp)

rifampicin resistance-determining region (RRDR) of the rpoB gene to simultaneously detect M. tuberculosis and rifampicin (RMP) resistance.3,4 Previous research regarding the implementation of this assay has been performed primarily in countries with limited resources, high TB burden and high HIV prevalence.5,6 Nevertheless, the rapid and independent testing format of the Xpert MTB/RIF assay suggests that the assay can be an effective routine diagnostic test for central clinical laboratories in industrialised countries with a lower burden of TB and HIV. South Korea has an intermediate TB burden, with an incidence rate of 90 per 100 000 population and 38 741 case notifications in 2009.7 There is a relatively high prevalence of MDR-TB, with 1700 notified cases, including 2.7% of newly diagnosed TB and 14% of retreated cases.8 HIV prevalence is low. The number of newly detected TB-HIV cases was 0.095/100 000 persons in 2005.9 Accurate and rapid laboratory tests are required in Korea for the detection of both M. tuberculosis and drug resistance.

Correspondence to: Sung Sup Park, Department of Laboratory Medicine, Seoul National University Hospital, Seoul, South Korea. Tel: (+82) 2 2072 3206. Fax: (+82) 2 747 0359. e-mail: sparkle@snu.ac.kr Article submitted 2 September 2011. Final version accepted 1 June 2012.

1472

The International Journal of Tuberculosis and Lung Disease

However, most real-time PCRs adopt batch-testing protocols requiring intensive labour and biosafety facilities, thus impeding rapid diagnosis and easy laboratory implementation. The aim of this study was to assess the Xpert MTB/RIF assay as a routine hospital laboratory diagnostic test in areas with an intermediate TB burden.

served for M. tuberculosis growth for a total of respectively 8 and 6 weeks. Growing microorganisms were confirmed as M. tuberculosis by PCR targeting the IS6110 and MPB64 genes (Seeplex MTB ACE Detection, Seegene, Seoul, South Korea). Phenotypic drug resistance determination for the Xpert MTB/RIF assay Among the 71 specimens tested by the Xpert MTB/ RIF assay, phenotypic drug susceptibility testing (DST) results were available for 62. The drug susceptibility of 35 isolates from culture-positive specimens was determined. Among 27 specimens, 10 patients had been diagnosed with RMP-resistant TB prior to the study by phenotypic determination using isolates grown before the study period. Five had been tested for phenotypic resistance within 30 days of submitting specimens to the study. Twelve were diagnosed with RMPsusceptible TB prior to the study using isolates obtained before the study period (range 4 months3 years prior to the study); during follow-up visits, all 12 patients were successfully treated using anti-tuberculosis drugs, including RMP. Phenotypic drug resistance was measured in Lwenstein-Jensen media using the absolute concentration method.13 The critical concentration of RMP resistance was 40.0 g/ml. Drugs used for resistance determination included RMP, isoniazid (INH), ethambutol (EMB), pyrazinamide (PZA), streptomycin, rifabutin, capreomycin, kanamycin, amikacin, ofloxacin, moxifloxacin, gatifloxacin, levofloxacin, prothionamycin, cycloserine and p-aminosalicylic acid. Sequence analysis of the rpoB gene for RMP resistance The 81-bp RRDR in the rpoB gene was directly sequenced in 62 specimens (42 cultured isolates and 20 culture-negative sediment specimens). All sediments and their DNA extracts were preserved at 70C after Xpert assay. Sequences were analysed after obtaining all culture results. The PCR primers were as follows:5,14 primer pair 5-CGACCACTTCGGCAA CCG and 5-TCGATCGGGCACATCCGG for singleround PCR for colony isolates; primer pairs 5AGATCCGGGTCGGCATGT and 5-CACATCCGG CCGTAGTGC, and 5-CGTGGAGGCGATCACACC GCAGAC and 5-AGCTCCAGCCCGGCACGCTCA CGT for nested PCR for sediments. Amplified products were sequenced using the ABI PRISM 3730xl Genetic Analyzer (Applied Biosystems, Foster City, CA, USA) and analyzed by Sequencher 4.6 software (Gene Codes Corporation, Ann Arbor, MI, USA). In vitro assessment of the Xpert MTB/RIF assay: minimal detection ratio for RMP-resistant mutant proportions The minimal detection ratio of the Xpert MTB/RIF assay for RMP-resistant mutant proportions was determined using two strains of M. tuberculosis cells: wild-type cells (M. tuberculosis H37Ra, American Tissue Culture Collection [ATCC] 25177) and clinical

MATERIALS AND METHODS

Prescreening for clinical specimens This study included fresh sputum samples from 991 patients who visited or were admitted to a university hospital in South Korea. Specimens were consecutively submitted to a central hospital laboratory for the detection of M. tuberculosis from July to September 2009. On the day of submission, the fresh sputum specimens were liquefied and decontaminated using equal volumes of 4% sodium hydroxide, and centrifuged at 3500 rpm for 20 min to prepare sediment. Each sediment was divided for the Xpert MTB/RIF assay, smear miscroscopy, solid/liquid medium culture, and sequence analysis. Before the Xpert MTB/RIF assay, an aliquot from each sediment was prescreened for the presence of M. tuberculosis with commercial real-time PCR targeting insertion sequence (IS) 6110 (AdvanSure TB/ NTM, LG Life Science Diagnostics, Seoul, South Korea).10,11 A total of 71 real-time PCR-positive sediments from 71 patients were selected for the Xpert MTB/RIF assay: 53 were previously diagnosed and 18 were newly suspected TB patients. None of them had symptoms or signs of HIV infection. The HIV antigen/ antibody was absent in 44 patients (Abbott Laboratories Diagnostic Division, Abbott Park, IL, USA). The study design and ethical considerations were reviewed and approved by the institutional review board of the Seoul National University Hospital. The Xpert MTB/RIF assay Each Xpert MTB/RIF assay was performed within 24 h of specimen submission. We reduced the sediment sample volume from the recommended 500 l to 200 l. The assay procedure and interpretation were performed according to the manufacturers protocol. The assay contained five rpoB probes for real-time PCR. M. tuberculosis was considered to be detected when at least two probes showed cycle threshold (Ct) values within a two-cycle difference of each other. For RMP resistance, the first and the last amplified probes should show more than 3.5-cycle differences.3,12 Conventional microbiology tests for M. tuberculosis detection Smear microscopy was performed using auramine-O fluorochrome staining. Solid and liquid medium cultures were grown in respectively 3% Ogawa medium and BACTEC MGIT (Mycobacteria Growth Indicator Tube) 460 (BD Microbiology Systems, Cockeysville, MD, USA). Solid and liquid cultures were ob-

Evaluation of the Xpert MTB/RIF assay

1473

isolates carrying a p.Ser531Leu mutation. Sputum sediments were collected, preserved at 70C and pooled after confirming their M. tuberculosis negativity by smear microscopy, AdvanSure real-time PCR and culture using both solid and liquid media. Cells were grown on BACTEC MGIT 460, suspended in sterile phosphate-buffered saline (PBS), sonicated twice for 60 s (Model 3510, Branson Ultrasonics Corporation, Danbury, CT, USA), and adjusted to the concentration of McFarland 0.5 (approximately 108 cells/ml). These cell suspensions were serially diluted tenfold using PBS. Each 25 l of the two cell suspensions was spiked into a 200 l of pooled sediment to make different mutant proportions, with final concentrations of ~106 cells/ml sputum and mutant: wild-type ratios of 100:1, 10:1, 5:1, 1:1 and 1:10. The Xpert MTB/RIF assay was quintuplicated for each ratio. Statistical analyses were performed using SPSS for Windows version 12.0 (SPSS Inc, Chicago, IL, USA). Assessment of time saving in diagnoses The turnaround time (TAT), defined as the time from specimen submission to result reporting, was expressed as median and range. The report time was collected through electronic medical records for each patient. The M. tuberculosis TAT was compared between the Xpert MTB/RIF assay and culture. The DST TAT was compared between the Xpert MTB/RIF assay and phenotypic determination.

Table 1 Comparisons of conventional microbiology tests for M. tuberculosis and the Xpert MTB/RIF assay among 71 clinical specimens
Culture Smear M. tuberculosis AFB Positive Positive Negative Negative Failure* Total Positive Negative Positive Negative Positive Xpert MTB/RIF Tuberculosis susceptible/ diagnosis resistant new/previous 16/12 11/4 0/2 21/3 2/0 50/21 9/19 8/7 0/2 0/24 1/1 18/53

n 28 15 2 24 2 71

* One smear-positive specimen was contaminated with bacteria in culture, and one other showed growth of non-tuberculous mycobacteria. In both specimens, M. tuberculosis was detected by the Xpert MTB/RIF assay and the M. tuberculosis RRDR sequence was confirmed using sequence analyses. AFB = acid-fast bacilli; RRDR = rifampicin resistance-determining region.

RESULTS
M. tuberculosis detection Among the prescreened 71 specimens from 71 patients, the Xpert MTB/RIF assay detected M. tuberculosis in all specimens (Table 1). Smear microscopy detected acid-fast bacilli in 32 (45.1%) specimens, while solid/liquid culture detected M. tuberculosis in 43 (60.6%). There were 53 (74.6%) previously diagnosed and 18 (25.4%) newly suspected TB patients. Among the 15 smear-negative, culture-positive patients, more than 50% (8/15) were newly diagnosed with pulmonary TB using the Xpert MTB/RIF assay on the test day. All 26 culture-negative specimens were from TB patients who were taking medication at the time of testing (median 18 months, range 730, n = 5 for RMP-resistant patients; median 5 months, range 1 41, n = 21 for RMP-susceptible patients). RMP resistance determination The Xpert MTB/RIF assay identified RMP resistance in 21 (29.6%) patients and RMP susceptibility in 50 (70.4%). All patients with RMP-resistant M. tuberculosis had been diagnosed with TB and treated with anti-tuberculosis drugs. The Xpert MTB/RIF assay unambiguously showed RMP resistance. The mean Ctmax (cycle threshold dif-

ference between the earliest and latest signals among rpoB probes) in 50 RMP-susceptible specimens was 1.8 (range 1.22.9). Among 21 RMP-resistant specimens, 20 showed Ct 0 from mutation-located probes. One specimen showed Ct 38.4 from the mutated probe, with Ctmax11.6. The Xpert MTB/RIF assay and phenotypic DST showed 100% (62/62) concordance in 21 RMPresistant isolates and 41 RMP-susceptible isolates. Among the 21 RMP-resistant isolates, 95.2% (20/21) were MDR and six were extensively drug-resistant. INH- or RMP-monoresistant isolates were identified in 4.8% (3/62): one RMP-monoresistant and two INH-monoresistant isolates. The Xpert MTB/RIF assay and rpoB RRDR sequence analysis showed 98.4% (61/62) concordance for mutations (Table 2). The Xpert MTB/RIF assay

Table 2 The rpoB mutation spectrum of RMP-resistant M. tuberculosis in South Korea

Results from the Xpert MTB/RIF assay Mutations by sequence analysis p.Asp516Tyr (GAC>TAC) p.Asp516Val (GAC>GTC) p.Gln517del (delCAG) p.Met515_Asn518delinsIle (delGGACCAGAACinsT) p.Asn518Asp (AAC>GAC) p.His526Tyr (CAC>TAC) p.His526Asp (CAC>GAC) p.His526Leu (CAC>CTC)# p.Ser531Leu (TCG>TTG)# Wild sequence Occurrence* Delayed RMP n (%) probe interpretation 1 (1.6) 2 (3.1) 1 (1.6) 1 (1.6) 1 (1.6) 3 (4.7) 1 (1.6) 1 (1.6) 11 (17.2) 42 (65.6) B B B B B D D D E None Resistant Resistant Resistant Resistant Resistant Resistant Resistant Resistant Resistant Susceptible

* Among 62 specimens, two specimens harboured double mutations in their RRDR sequence: a specimen with p.Asp516Val (GAC>GTC) and p.Asn518Asp (AAC>GAC); and a specimen# with p.His526Leu (CAC> CTC) and p.Ser531Leu (TCG>TTG). No mutation was identified in the region of probe A (amino acids 507512). Other probe locations were as follows: Probe B = amino acids 511518; Probe C = amino acids 518523; Probe D = amino acids 522527; Probe E = amino acids 528533. Novel deletion mutation which spanned probe B and C. This mutation was located in the common site of probe B and C. RMP = rifampicin.

1474

The International Journal of Tuberculosis and Lung Disease

Figure RMP-resistant mutations in Korean isolates not detected by the Xpert MTB/RIF assay. A. The Xpert probes are aligned with the rpoB RRDR sequence in M. tuberculosis. Locations of each nucleotide change are indicated by arrows and lowercase letters. A patient harboured double mutations with p.Asp516Val (a) in the middle of probe B and p.Asn518Asp (b) located at both ends of probe B and C: RMP resistance in this patient was detected by probe B delay (Ct = 0). However, probe C failed to make a signal delay > 3.5 Ct (Ct = 3.2 from the fastest probe A). Another patient contained a p.His526Tyr mutation (c). B. The Xpert MTB/RIF assay failed to detect approximately 50% of this p.His526Tyr mutation. The upper chromatogram is from a patient with RMP-susceptible M. tuberculosis, with a wild-type sequence of p.His526. The lower chromatogram is from a patient with mixed mutant and wild-type sequences of p.His526Tyr and p.His526. RMP = rifampicin; RRDR = rifampicin resistance-determining region; Ct = cycle threshold. This image can be viewed online in colour at http://www.ingentaconnect.com/content/iuatld/ijtld/2012/00000016/00000011/art00010

failed to detect a mixed mutant peak, which was the p.His526Tyr mutant, comprising approximately 50% of peak height in sequencing electropherogram from a culture-negative sediment of a patient on treatment (Figure). One RMP-resistant specimen showed concordant results between the Xpert MTB/RIF assay and the sequence analysis, but it showed a probe error in assay: this specimen harboured double mutations (p.Asp516Val in probe B and p.Asn518Asp in the ends of both probe B and C, Figure A). Resistance was detected in the Xpert MTB/RIF assay by only probe B delay (Ct = 0). We identified nine types of mutation in Korean patients, including the novel mutation p.Met515_ Asn518delinsIle (delGGACCAGAACinsT). The most common mutation was p.Ser531Leu. In vitro assessment of the Xpert MTB/RIF assay: minimal detection ratio for RMP-resistant mutant proportions The RMP-resistant subpopulation in patients on treatment could not be detected by the Xpert MTB/RIF assay, which adopts amplification signal delays by mutation. We used p.Ser531Leu in our assessment as the most frequently encountered mutation in Korea (Table 2). We assessed the minimal detection limit of the minor mutant population as intact cell specimen, and the Xpert MTB/RIF assay detected 5:1 mutant:

wild-type or more larger mutant proportion in specimens (95% confidence interval 3:1279:1). Assessment for time saving in diagnoses Among 39 smear-negative specimens, 15 were culturepositive 19 days after submission (range 948; Table 1). In the ~50% (8/15) of smear-negative, culturepositive specimens that were from first-visit patients, the Xpert MTB/RIF assay diagnosed TB on the test day. This assay saved diagnostic TAT a median of 18.5 days (range 930, n = 8) compared to culture. This assay also saved 81.5 days (range 65136, n = 6) for RMP susceptibility information compared with phenotypic DST. Among 28 smear-positive, culturepositive specimens, the Xpert MTB/RIF assay diagnosed all those with TB on the test day, and saved 69 days (range 57113, n = 10) for RMP susceptibility information compared with the phenotypic test.

DISCUSSION
This study showed that the Xpert MTB/RIF assay detected M. tuberculosis with adequate sensitivity for the laboratory diagnosis of TB in an area with an intermediate TB burden. The M. tuberculosis detection sensitivity of this assay was at least equivalent to the sensitivity of a commercial real-time PCR using multicopy IS6110 targets (90.7% sensitivity and

Evaluation of the Xpert MTB/RIF assay

1475

97.9% specificity).11 Xpert MTB/RIF assay was sensitive, even with lower sediment volume than that recommended by manufacturers. The Xpert MTB/ RIF assay also detected M. tuberculosis in two cases with unreliable culture results: in a specimen with non-tuberculous mycobacteria overgrowth and in a specimen with bacterial overgrowth. The determination of RMP resistance using the Xpert MTB/RIF assay was satisfactory among the clinical laboratory specimens. The current cut-off, >3.5-cycle difference between the first and the last amplified probes (Ctmax) for RMP resistance, was acceptable for interpretation. RMP resistance by this assay was concordant in all specimens with phenotypic DST results. However, prediction of MDR-TB based on RMP information from the Xpert MTB/RIF assay was accurate in 95.2% of the isolates. This study identified 4.8% INH- or RMP-monoresistant isolates. Resistance had been reported more frequently for INH than for RMP in Korea, among both new and previously treated TB patients.2 Additional probes for INH resistance may therefore improve the surveillance of MDR-TB in South Korea. The Xpert MTB/RIF assay showed 98.4% agreement with the rpoB RRDR sequence analysis. The most frequent mutations in Korean M. tuberculosis isolates occurred at the same codons observed worldwide, which are at rpoB codons 531, 526 and 516.15,16 The Xpert MTB/RIF assay consistently detected all of these mutations and the newly found mutation p.Met515_Asn518delinsIle. Care should be taken during laboratory diagnosis to avoid the rare mismatches between the Xpert MTB/ RIF assay and sequence analyses that were observed in this study. Both false-positive and false-negative RMP resistance by the Xpert MTB/RIF have been reported compared with phenotypic DST, sequence analysis or line-probe assays.1719 We observed that the Xpert MTB/RIF assay detected RMP resistance in a mutant ratio-dependent manner. Our in vitro assessment showed that the detection limit was five times greater in mutant M. tuberculosis than in wild-type M. tuberculosis in bacilli populations. This detection limit using intact p.Ser531Leu cells was similar to the detection limit shown in a previous study (95% confidence interval [CI] 56.988.4%), using extracted DNA with the same mutation.12 The Xpert MTB/RIF assay failed to detect approximately 50% of the p.His526Tyr mutants located at the sixth nucleotide of the probe D margin (Ct = 1.3 between probe D and the fastest probe A). This specimen showed no bacilli on smear microscopy and no growth on culture. This patient had been diagnosed with drug-susceptible TB. At the time of the Xpert MTB/RIF testing, the patient had been treated with INH, RMP, EMB and PZA for 2 months, followed by INH, RMP and EMB for an additional 2 months. Xpert MTB/RIF failed to detect RMP resis-

tance only in this case. We cannot estimate the exact clinical significance of low mutant DNA in a culturenegative specimen, particularly in a patient taking medication. The manufacturer warned that the test results might be affected by prior or current antituberculosis drug medication, and did not recommend this assay for patient follow-up. Our results demonstrate that the Xpert MTB/RIF assay may not detect minor proportions of mutant bacilli in patients who are on treatment or in the course of acquiring resistance. The Xpert MTB/RIF assay was also relatively insensitive to mutations located at the margin of the probe. A previous study reported that the assay could not detect the presence of mutant rpoB 633ccg unless 100% of the M. tuberculosis DNA was mutant.12 Our study identified a case with probe C failure (Ct = 3.2) for probe-end mutation p.Asn518Asp (AAC>GAC). Rare or novel mutations at probe margin may therefore cause false-negative results in this assay. The rapidity and independent testing format of the Xpert MTB/RIF assay was useful for the clinical hospital laboratory in terms of saving time and resources. For smear-negative patients, this assay provided an 18.5-day faster TB diagnosis in first-visit, culturepositive patients. It also saved 81.5 days for information on RMP resistance, without additional laboratory labour. For smear-positive patients, this assay saved 69 days for information on RMP susceptibility. We suggest that the Xpert MTB/RIF assay can be a good routine test for clinical laboratories handling specimens from patients who require rapid diagnosis of TB and detection of resistance.

CONCLUSIONS
In this study, the Xpert MTB/RIF assay showed high sensitivity in detecting M. tuberculosis and providing information on RMP resistance, and shortened the routine test time. The Xpert MTB/RIF assay could be a rapid and accurate diagnostic test in countries with an intermediate TB burden. This study also showed that the Xpert MTB/RIF assay required a high-mutant population for resistance detection. Mutations located on the probe-margin codons may not be detected by this assay. Acknowledgements
Partial financial support for this research was provided by Hoil Biomed Inc, Seoul, South Korea (grants no. SNUH0620071320 and SNUH0620113330), local distributor of Cepheid Inc.

References
1 World Health Organization. Global tuberculosis control. WHO report 2010. WHO/HTM/TB/2010.7. Geneva, Switzerland: WHO, 2010. 2 Wright A, Zignol M, Van Deun A, et al. Epidemiology of antituberculosis drug resistance 200207: an updated analysis of

1476

The International Journal of Tuberculosis and Lung Disease

10

11

the Global Project on Anti-Tuberculosis Drug Resistance Surveillance. Lancet 2009; 373: 18611873. Helb D, Jones M, Story E, et al. Rapid detection of Mycobacterium tuberculosis and rifampin resistance by use of on-demand, near-patient technology. J Clin Microbiol 2010; 48: 229237. Telenti A, Imboden P, Marchesi F, et al. Detection of rifampicinresistance mutations in Mycobacterium tuberculosis. Lancet 1993; 341: 647650. Boehme C C, Nabeta P, Hillemann D, et al. Rapid molecular detection of tuberculosis and rifampin resistance. N Engl J Med 2010; 363: 10051015. Boehme C C, Nicol M P, Nabeta P, et al. Feasibility, diagnostic accuracy, and effectiveness of decentralised use of the Xpert MTB/RIF test for diagnosis of tuberculosis and multidrug resistance: a multicentre implementation study. Lancet 2011; 377: 14951505. World Health Organization. Global tuberculosis control: epidemiology, strategy, financing. WHO report 2009. WHO/HTM/ TB/2009.411. Geneva, Switzerland: WHO, 2009. World Health Organization. Tuberculosis country profiles: Republic of Korea. Geneva, Switzerland: WHO, 2012. http://www. who.int/tb/country/data/profiles/en/index.html Accessed July 2012. Lee C H, Hwang J Y, Oh D K, et al. The burden and characteristics of tuberculosis/human immunodeficiency virus (TB/HIV) in South Korea: a study from a population database and a survey. BMC Infect Dis 2010; 10: 66. Cho S Y, Kim M J, Suh J T, Lee H J. Comparison of diagnostic performance of three real-time PCR kits for detecting Mycobacterium species. Yonsei Med J 2011; 52: 301306. Kim Y J, Park M Y, Kim S Y, et al. Evaluation of the perfor-

12

13

14

15

16

17

18

19

mances of AdvanSure TB/NTM real time PCR kit for detection of mycobacteria in respiratory specimens. Korean J Lab Med 2008; 28: 3438. Blakemore R, Story E, Helb D, et al. Evaluation of the analytical performance of the Xpert MTB/RIF assay. J Clin Microbiol 2010; 48: 24952501. Oh S H, Kim Y J, Park S K, et al. Comparison of antimycobacterial drug susceptibility test results by institutes and methods. Korean J Clin Microbiol 2008; 11: 4348. Halse T A, Edwards J, Cunningham P L, et al. Combined realtime PCR and rpoB gene pyrosequencing for rapid identification of Mycobacterium tuberculosis and determination of rifampin resistance directly in clinical specimens. J Clin Microbiol 2010; 48: 11821188. Ahmad S, Mokaddas E. Recent advances in the diagnosis and treatment of multidrug-resistant tuberculosis. Respir Med 2009; 103: 17771790. Adekambi T, Drancourt M, Raoult D. The rpoB gene as a tool for clinical microbiologists. Trends Microbiol 2009; 17: 37 45. Van Rie A, Mellet K, John M-A, et al. False-positive rifampicin resistance on Xpert MTB/RIF: case report and clinical implications. Int J Tuberc Lung Dis 2012; 16: 206208. Marlowe E M, Novak-Weekley S M, Cumpio J, et al. Evaluation of the Cepheid Xpert MTB/RIF assay for direct detection of Mycobacterium tuberculosis complex in respiratory specimens. J Clin Microbiol 2011; 49: 16211623. Theron G, Peter J, van Zyl-Smit R, et al. Evaluation of the Xpert MTB/RIF assay for the diagnosis of pulmonary tuberculosis in a high HIV prevalence setting. Am J Respir Crit Care Med 2011; 184: 132140.

Evaluation of the Xpert MTB/RIF assay

RSUM
CONTEXTE :

Le laboratoire dun hpital central en Core

du Sud. Evaluer lutilisation du test Xpert MTB/ RIF dans un pays o le fardeau de tuberculose (TB) est du niveau intermdiaire. S C H M A : On a test dans les 24 h par le test Xpert MTB/RIF au total 71 sdiments de crachats positifs la raction de polymrase en chaine en temps rel. La dtection de Mycobacterium tuberculosis a t compare celle de lexamen microscopique des frottis et de la culture. On a compar la rsistance la rifampicine (RMP) dtermine par une mthode base sur la culture avec celle obtenue par le squenage du gne rpoB. Nous avons valu galement les limites de dtection pour les proportions de mutants et le temps gagn pour le diagnostic. R S U LTAT S : Le test Xpert MTB/RIF a dtect M. tuberculosis dans 71 chantillons (100%) des chantillons
OBJECTIF :

(32 frottis positifs, 39 frottis ngatifs). Ce test a concord 100% (62/62) avec la rsistance aux mdicaments base sur la culture et raison de 98,4% (61/62) avec le squenage. Un chantillon contenant environ 50% du mutant p.His536Tyr a t interprt erronment par ce test comme bacille de type sauvage. Le ratio minimal de dtection a t de 5:1 pour les cellules mutantes vs. les cellules de type sauvage. Chez les patients frottis ngatif mais culture positive, le temps mdian gagn a t de 18,5 jours (extrmes 930) pour le diagnostic de M. tuberculosis et de 81,5 jours (65136) pour la sensibilit la RMP. C O N C L U S I O N S : Le test Xpert MTB/RIF a dtect avec sensibilit M. tuberculosis en mme temps quune information sur la rsistance la RMP, et il a raccourci la priode avant diagnostic par comparaison avec les tests conventionnels. Toutefois, la localisation et la quantit de mutations pourraient modifier sa sensibilit.
RESUMEN

El laboratorio de un hospital central en Corea del Sur. O B J E T I V O : Evaluar la utilidad de la prueba Xpert MTB/RIF en un pas con una carga intermedia de morbilidad por tuberculosis (TB). M T O D O : Se practic en las muestras clnicas una prueba simultnea de reaccin en cadena de la polimerasa y en un lapso de 24 h se realiz la prueba Xpert MTB/RIF a los 71 sedimentos de esputo que haban obtenido un resultado positivo para micobacterias. La deteccin de Mycobacterium tuberculosis se compar con el resultado de la baciloscopia y el cultivo. La resistencia a rifampicina (RMP) se compar con un mtodo basado en el cultivo y la secuenciacin del gen rpoB. Se evalu adems el umbral de deteccin en funcin de la proporcin de mutantes y el tiempo ganado en el diagnstico. R E S U LTA D O S : La prueba Xpert detect el M. tuberculosis en 71 muestras (100%; 32 con baciloscopia positiva y 39 con baciloscopia negativa). Esta prueba exMARCO DE REFERENCIA:

hibi una concordancia de 100% (62/62) con el anlisis de resistencia basado en el cultivo y una coincidencia de 98,4% (61/62) con la secuenciacin. Una muestra que contena alrededor de 50% de mutantes p.His526Tyr se interpret errneamente como una cepa natural del bacilo con esta prueba. El umbral inferior de deteccin se observ con la proporcin de cinco mutantes por una clula de tipo natural. La mediana del tiempo ahorrado hasta el diagnstico de M. tuberculosis fue 18,5 das (entre 9 y 30) y 81,5 das (de 65 a 136) con respecto al diagnstico de sensibilidad a la RMP en los pacientes con baciloscopia negativa y cultivo positivo. C O N C L U S I N : Con la prueba Xpert se obtuvo una buena sensibilidad en la deteccin de M. tuberculosis con resistencia a RMP y se acort el lapso hasta el diagnstico, en comparacin con las pruebas clsicas; sin embargo, la localizacin de la mutacin y la proporcin de mutantes podran modificar la sensibilidad de la prueba.