INT J TUBERC LUNG DIS 12(11):1306–1312

© 2008 The Union

Comparison of MAS-PCR and GenoType MTBDR assay for the

detection of rifampicin-resistant Mycobacterium tuberculosis

D. Q. Tho,* D. T. M. Ha,*† P. M. Duy,* N. T. N. Lan,† D. V. Hoa,† N. V. V. Chau,‡ J. Farrar,*§ M. Caws*§

* Oxford University Clinical Research Unit, Hospital for Tropical Diseases, Ho Chi Minh City, † Pham Ngoc Thach Hospital
for Tuberculosis and Lung Diseases, Ho Chi Minh City, ‡ Hospital for Tropical Diseases, Ho Chi Minh City, Vietnam;
§ Centre for Clinical Vaccinology and Tropical Medicine, Churchill Hospital, Oxford, UK

SUMMARY

SETTING: Pham Ngoc Thach Hospital for Tuberculosis tests was respectively 83.7% (95%CI 75.1–90.2) and
and Lung Diseases, the tertiary referral hospital for tu- 93.3% (95%CI 86.6–97.3). Both tests were 100% spe-
berculosis (TB) in Southern Vietnam. cific. The negative predictive value was 74.6% (95%CI
O B J E C T I V E : To develop and evaluate a simple, rapid 65.3–83.1) for the MAS-PCR and 87.7% (95%CI 80.0–
and accurate multiplex allele specific polymerase chain 93.6) for the MTBDR test.
reaction (MAS-PCR) test to detect rifampicin (RMP) re- C O N C L U S I O N : The MTBDR test, although more sensi-
sistance point mutations at codons 516, 526 or 531 in tive, is currently prohibitively expensive in resource-poor,
the rpoB gene of Mycobacterium tuberculosis. high-burden settings. The MAS-PCR described here
D E S I G N : The novel MAS-PCR was compared with the presents a less laborious economic alternative. A suscep-
commercial M. tuberculosis Drug Resistance (MTBDR) tible result returned by either test cannot be used to ex-
test in 104 RMP-resistant and 50 RMP-susceptible rou- clude multidrug-resistant TB.
tine isolates, defined by conventional 1% phenotypic K E Y W O R D S : tuberculosis; rifampicin; resistance; PCR;
susceptibility testing. MTBDR
R E S U LT S : The sensitivity of the MAS-PCR and MTBDR

ONE THIRD of the world’s population is infected Detecting drug resistance, particularly MDR-TB,
with Mycobacterium tuberculosis, 5–10% of whom before the instigation of inappropriate chemotherapy
will develop active disease. Each year the disease kills reduces morbidity, mortality, economic costs and un-
approximately 1.7 million people worldwide, 1.4 mil- necessary treatment with ineffective drugs. Conven-
lion of whom (72%) are in 22 high-burden countries.1 tional phenotypic drug susceptibility testing (DST)
Standard tuberculosis (TB) treatment regimens for only returns results after several weeks of culture in
fully susceptible TB require multidrug regimens over a Mycobacterial Growth Indicator Tube or Löwenstein-
minimum of 6 months. The rise of multidrug-resistant Jensen (LJ) media.
tuberculosis (MDR-TB), defined as M. tuberculosis Two commercial molecular kits12 for MDR-TB are
resistant to at least isoniazid (INH) and rifampicin currently available: the INNO-LiPA RMP TB test (In-
(RMP), threatens the success of TB control programmes nogenetics, Zwijnaarde, Belgium)13 and the Geno-
in many regions of the world.2–4 It has been demon- Type MTBDR assay kits (Hain Lifesciences, Nehren,
strated that MDR-TB can successfully transmit within Germany).14 These kits perform well in research set-
a community;5,6 early detection of infectious cases is tings15–20 but, at over US$50 per test strip, are too ex-
therefore vital to interrupt chains of transmission.7,8 pensive to be used routinely in developing countries
MDR-TB is difficult to treat, with treatment dura- such as Vietnam. It is necessary to develop low-cost
tions of 18 months or longer, using more toxic and less techniques that are accurate, simple and rapid for the
potent second-line drugs.9,10 Furthermore, the cost of identification of drug-resistant TB.
treating a patient with MDR-TB is many times the The authors of the present study21 and others22,23
cost of treating of a patient with fully susceptible TB, have shown that RMP resistance is a surrogate marker
even with the help of the Green Light Committee, of MDR in some settings where RMP monoresistance
which has negotiated significant reductions in costs is rare. A molecular technique for the rapid detection
of second-line drugs for government programmes.9,11 of RMP resistance can therefore be used to screen for

Correspondence to: Dau Quang Tho, Oxford University Clinical Research Unit, Hospital of Tropical Disease, 190 Ben
Ham Tu, District 5, Ho Chi Minh City, Vietnam. Tel: (+84) 8 923 9211. Fax: (+84) 8 835 3904. e-mail: dauquangtho@
yahoo.com
Article submitted 11 January 2008. Final version accepted 26 June 2008.
 MAS-PCR for RMP-resistant TB 1307

MDR-TB. As RMP monoresistance occurs more fre- lates were sequenced in the RRDR of the rpoB gene,
quently in human immunodeficiency virus (HIV) pos- as previously described.21 All 154 isolates were used
itive TB cases,24 in areas of high HIV prevalence RMP to evaluate the test described here by comparing it
resistance may not be applicable as a surrogate marker with the commercial MTBDR test against the pheno-
of MDR-TB. typic DST results.
Resistance to RMP is conferred by single-step mu-
tation in the 81 base pair (bp) RMP resistance deter- Drug susceptibility testing
mining region (RRDR) of the rpoB gene which encodes DST was performed at PNT, a World Health Organiza-
the beta subunit of the RNA polymerase. Single-base tion (WHO) international reference laboratory, for
substitutions, insertions or deletions in the rpoB gene INH (0.2 μg/ml) and RMP (40 μg/ml) on LJ media
can lead to RMP resistance.25 with the standard 1% proportion method according
Studies in the literature have used various molecular to the WHO protocol.
mutation detection techniques to identify MDR-TB,
such as real-time polymerase chain reaction (PCR),26,27 DNA extraction
single strain conformation polymorphism28 or sequenc- DNA extraction was performed using the cetyl tri-
ing to detect mutated codons.29,30 However, these methyl ammonium bromide (CTAB) method after cul-
techniques require expensive equipment and techni- ture on LJ. The purified DNA was dissolved in tris-
cal expertise. EDTA (TE) buffer, quantified, diluted to the final 15 ng/
To minimise post-amplification processing, we used μl working concentration and stored at −20°C.31
three multiplex allele specific (MAS) PCR reactions for
each isolate, with two outer primers as positive con- Sequencing
trol for each sample and one inner primer to detect One hundred and four isolates were sequenced in the
the mutation. In a previous study in Vietnam, we dem- RRDR and N-terminal of the rpoB gene with primers
onstrated that the mutations in RRDR of the rpoB TB146F and TB146R (Table 1), as previously de-
gene are located mostly at three hot spot codons 531 scribed.21 Half the volume of the CEQ Quick Start kit
(43%), 526 (31%) and 516 (15%), concurring with (Beckman Coulter, Fullerton, CA, USA) was used to
prevalence reported worldwide.23,25 sequence from both strands of a 350-bp fragment con-
The MAS-PCR was compared with the commer- taining an 81-bp sequence of RRDR (from position
cial reverse-hybridisation MTBDR test in a collection 760990 to 761339, H37Rv, Blast NC_000962.2).
of 104 RMP-resistant isolates with known RRDR se-
quence and 50 RMP-susceptible isolates. The MTBDR GenoType MTBDR
test also detects INH resistance-conferring mutations The GenoType MTBDR test was performed accord-
in the katG gene. Phenotypic DST results were con- ing to the manufacturer’s protocol. Briefly, 15 ng of
sidered the gold standard. sample DNA was used as a template to amplify rpoB
and katG target sequences with biotin-labelled prim-
METHODS ers. The amplification product was analysed by reverse
hybridisation to membrane bound probes. Colori-
Samples metric detection gave purple bands, read by eye. Neg-
One hundred and four consecutive isolates routinely ative controls (water) were included with each run.
identified as RMP-resistant (102 [98.1%] of which
were MDR) and 50 RMP-susceptible isolates were MAS-PCR
collected at Pham Ngoc Thach Hospital for TB and Three separate PCR reactions were performed with
Lung Diseases (PNT), Ho Chi Minh City, Vietnam, the same outer primers to amplify the RRDR of rpoB,
between January and March 2005. RMP-resistant iso- and one of three inner primers targeted to wild type

Table 1 Primers for sequencing and MAS-PCR

Temperature Concentration Target in Product length

Primer Sequence (5′–3′) °C nM rpoB gene bp
RPOBF GGG AGC GGA TGA CCA CCC A 68 125 Flank RRDR 350
RPOBR30 GCG GTA CGG CGT TTC GAT GAA C 68* 125*
TB146-F CTT CTC CGG GTC GAT GTC GTT G 64 300 N-terminal region 365
TB146-R32 CGC GCT TGT CGA CGT CAA ACT C 64 300
RIF2R GAC AGC GGG TTG TTC TGG T 68 300 Codon 516 139†
RIF3F CCG CTG TCG GGG TAG ACC CA 68 400 Codon 526 219‡
RIF4R CCG CCG GGC CCC ATC GCC G 68 300 Codon 531 184†
* The temperature and concentration for sequencing with RPOBR + RPOBF were respectively 65°C and 150 nM.
† Amplicon generated with primer RPOBF.
‡ Amplicon generated with primer RPOBR.

MAS-PCR = multiplex allele specific polymerase chain reaction; bp = base pair; RRDR = resistance determining region.
1308 The International Journal of Tuberculosis and Lung Disease

Figure 1 Amplicon profiles in MAS-PCR of isolates carrying mutations at each site. * a, codon
516; b, codon 526; c, codon 531. Samples: 1, mutation at codon 516; 2, mutation at codon 526;
3, mutation at codon 531; 4, codon 516 and 526 mutated; 5, wild type control sample; 6, nega-
tive control. M = 100 bp ladder; MAS-PCR = multiplex allele specific polymerase chain reaction;
bp = base pair.

at hot spot codons 516, 526 and 531 (Table 1). The actions were amplified under the same thermocycling
results were read with two bands for wild type isolates programme: 95°C for 1 min, followed by 35 cycles of
and only one band for mutant isolates (Figure 1). 95°C for 10 s, 68°C for 20 s, 72°C for 20 s and a final
The 350 bp fragment of the rpoB gene was ampli- step of 72°C for 5 min. A negative control (water) and
fied using outer primers RPOBF (5′-GGGAGCGGA a positive control for each of the three target codons
TGACCACCCA-3′) and RPOBR (5′-GCGGTACGG was included with each batch processed. PCR products
CGTTTCGATGAAC-3′). The inner primers used to were analysed by electrophoresis on 1% agarose gel
detect mutations at codon 516, 526 and 531 were at 150V for 30 min.
RIF2R (5′-GACAGCGGGTTGTTCTGGT-3′), RIF3F All sensitivities and specificities were calculated in
(5′-CCGCTGTCGGGGTAGACCCA-3′) and RIF4R comparison with conventional 1% phenotypic DST
(5′-CCGCCGGGCCCCATCGCCG-3′), respectively as gold standard. Discrepant results were evaluated by
(Table 1). The primers were designed using Primer analysis of sequencing data.
Express version 2.0 software (Applied Biosystems Inc,
Foster City, CA, USA).
RESULTS
Each PCR reaction contained 0.75 U Taq DNA
polymerase (Bioline, Taunton, MA, USA), 0.2 mM The results of MAS-PCR, sequencing and the com-
of each dNTP (Roche, Basel, Switzerland) and 125– mercial MTBDR test are summarised and shown in
400 nM of primers (Table 1) and 2 mM MgCl2. Fifteen Table 2. Sequencing showed that among 104 RMP-
nanograms of genomic DNA was added. All of the re- resistant isolates, 92 isolates (88.5%) had mutations

Table 2 Comparison between MAS-PCR test and commercial MTBDR test in

104 RMP-resistant isolates
MTBDR test MAS-PCR RMP test
Isolates Resistant Susceptible Resistant Susceptible
Mutation n n n n n
S531L 46 46 0 45 1
H526D/F/L /R /S/Y 28 28 0 28 0
D516V 9 9 0 8 1
D516V, H526Q 1 1 0 1 0
D516A, H526N 1 1 0 1 0
L511P, H526N 1 1 0 1 0
Q513K, H526D 1 1 0 1 0
D516G, 522::SGL 1 1 0 1 0
D516A, L533P 1 1 0 1 0
Q513K /L 3 3 0 0 3
D516G, L533P 2 2 0 0 2
L511P, D516G 1 1 0 0 1
L511P, S512G 1 1 0 0 1
L533P 2 1 1 0 2
514::F 1 0 1 0 1
V146F 1 0 1 0 1
WT 4 0 4 0 4
Total, n (%) 104 97 (93.27) 7 (6.7) 87 (83.65) 17 (16.3)
MAS-PCR = multiplex allele specific polymerase chain reaction; MTBDR = GenoType MTBDR assay; RMP = rifampicin.
 MAS-PCR for RMP-resistant TB 1309

Table 3 Sensitivity of MAS-PCR and MTBDR with reference to 1% conventional phenotypic

drug susceptibility testing as gold standard
MAS-PCR MTBDR
Resistant Susceptible Resistant Susceptible
Conventional DST n (%) n (%) n (%) n (%)
RMP
Resistant (n = 104) 87 (83.65) 17 97 (93.3) 7 (6.7)
Susceptible (n = 50) 0 50 (100) 0 50 (100)
INH
Resistant (n = 132) NA NA 109 (82.6) 23 (17.4)
Susceptible (n = 22) NA NA 0 22 (100)
MAS-PCR = multiplex allele specific polymerase chain reaction; MTBDR = GenoType MTBDR assay; DST = drug sus-
ceptibility testing; RMP = rifampicin; INH = isoniazid; NA = not applicable.

located at the three target codons: 531 (43%), 526 100% and 48.9% (95%CI 38.9–59.2). Of 154 isolates
(31%), 516 (15%). Eight (7.7%) had mutations at in this study, 102 (66%) were MDR; the PPV and NPV
other codons in the RRDR of the rpoB gene. Four for MDR-TB in this sample set were respectively 100%
(3.8%) had no mutation in the regions that were and 70.3% (95%CI 60.0–78.8).
sequenced. MAS-PCR detected 87/104 (83.7%) RMP-resistant
Using the MTBDR test, 97 isolates (93.3%) were isolates. Of the 17 discrepant results, four of the iso-
detected as resistant to RMP. Of the seven resistant lates carried no mutations in any sequenced region,
isolates returning a susceptible result, four had no mu- eight had mutations in the rpoB gene but outside the
tations in RRDR (Table 2), two had mutations outside three target codons (Q513K/L ×3, L533P ×2, 514::F
the location of the primers (L533P and V146F) and ×1, V146F ×1 and L511P/S512G ×1) (Table 2), two
the remaining isolate had three nucleotides inserted at showed evidence of a mixed population of resistant
the end of a probe (514::F); thus, hybridisation still and susceptible isolates on sequencing with double
occurred, returning a wild type result. The sensitivity peaks at a single allele (Figure 2), which would be de-
and the specificity of MTBDR for RMP resistance fined as wild type by MAS-PCR due to the generation
were respectively 93.3% (95% confidence interval [CI] of the wild-type band. The remaining three isolates
86.6–97.3) and 100%. The positive predictive value had a D516G mutation, which primer RIF2R hybri-
(PPV) was 100%, with a negative predictive value dises weakly, generating a weak wild-type band and
(NPV) of 87.7% (95%CI 80.0–93.6). returning a false-susceptible result.
MTBDR also detects mutations at katG315 that The specificity of the MAS-PCR test was 100%,
are related to INH resistance. Of 154 isolates tested, with no susceptible isolates returning a false-positive
132 isolates were resistant (102 RMP-resistant and resistant result (Table 3). Sensitivity was 83.7%
30 RMP-susceptible) and 22 isolates were susceptible (95%CI 75.1–90.2). The PPV and NPV for RMP re-
to INH by DST. MTBDR detected INH resistance in sistance in this sample set were respectively 100%
109/132 (82.6%) INH-resistant isolates, all of which and 74.6% (95%CI 65.3–83.1).
carried mutations at codon 315, of which 98 (72.2%)
were katGS315T mutants. The sensitivity and speci-
DISCUSSION
ficity of MTBDR for INH resistance were respectively
82.6% (95%CI 75.0–88.6) and 100% (Table 3). The Neither molecular test was 100% sensitive for the de-
PPV and NPV for INH resistance were respectively tection of RMP-resistant TB; however, both tests are

Figure 2 Double peak in the sequencing trace from sample DQ786 showing a heterogeneous
population of susceptible strain (wild type, D516) and resistant mutant (D516V).
1310 The International Journal of Tuberculosis and Lung Disease

100% specific and therefore useful as rapid detection mutations outside the three hot spot codons 516, 526
tests for positive identification of RMP resistance and and 531. Three had a D516G mutation, which is not
by proxy for MDR-TB. A negative result from either detected by this test. The two remaining isolates prob-
test cannot be used to rule out MDR-TB. The com- ably contained a heterogeneous population of suscep-
mercial MTBDR test is more sensitive (93.3% vs. tible and resistant isolates as they had a double peak
83.7%) and allows the direct detection of INH- at a resistance allele on the sequencing trace. These iso-
resistant TB, directly confirming the diagnosis of MDR- lates were defined as wild type by the MAS-PCR test
TB in the majority of cases (n/N = 80/102, 78.4%). due to generation of the wild-type band. The MTBDR
The MTBDR test used in this study was only able to test is able to detect the resistant population in these
detect INH resistance mutations at katG315. The new isolates, as hybridisation with the resistant band will
version of this test, MTBDRplus,33 has additional occur. It is likely that for clinical purposes these iso-
probes for mutations at inhA-15 (the second most lates should be treated as resistant because a resistant
common INH resistance mutation) and for additional population is emerging. However, it is possible that
regions of the rpoB RRDR. Sensitivity for MDR-TB the patient may be successfully treated on a standard
detection should therefore be higher with the MTB- regimen, particularly if the organism remains INH-
DRplus test. The test format is user-friendly and inter- susceptible, and therefore the clinical implications of
pretation of the test is simple. However, at more than such results remain unclear.
US$50 per test strip, routine use of this kit in develop- Molecular tests will not have 100% accuracy until
ing settings such as Vietnam is not currently feasible. all the relevant mutations have been described. A sus-
The MTBDRplus test strips are one of the technolo- ceptible result returned by a molecular test therefore
gies under evaluation in the Foundation for Innova- cannot yet be used to exclude RMP resistance. Ideally
tive New Diagnostics (FIND) portfolio, which aims DST should be performed on every patient at the time
to evaluate novel diagnostic technologies for poverty- of diagnosis; however, in high-volume developing set-
related diseases and promote uptake of those that are tings where this is not feasible, priority should be
effective through public-private partnerships.* If suc- given to cases where the index of suspicion is high,
cessful, the FIND demonstration projects for MTBDR such as HIV-positive patients, MDR contact cases and
should lead to wider availability at significantly lower retreatment/relapse cases, where a susceptible result
costs (approximately US$5/test). returned by molecular screening methods should be
The MAS-PCR test evaluated here offers several ad- confirmed by DST or other phenotypic methods. How-
vantages over the MTBDR test strips in a high-volume ever, with high specificity (100%), molecular tech-
setting. First, the cost for a batch of 10 samples is niques allow the early detection of MDR cases, poten-
around US$16 using reagents from local suppliers tially reducing morbidity, mortality and interrupting
without bulk discounts (US$1.6 per sample), exclud- chains of transmission when combined with effective
ing salary and capital equipment costs. The equipment MDR treatment programmes.
required is basic PCR equipment—a thermocycler, gel
electrophoresis equipment and ultraviolet transillu-
minator—which is now available in many reference References
TB laboratories. There is minimal post-amplification 1 World Health Organization. Global TB control: surveillance,
processing involved, considerably less hands-on time planning, financing. WHO report 2006. WHO/HTM/TB/2006.
362. Geneva, Switzerland: WHO, 2006.
and the MAS-PCR does not require adherence to a 2 Emergence of Mycobacterium tuberculosis with extensive resis-
strict time protocol for post-amplification process- tance to second-line drugs—worldwide, 2000–2004. MMWR
ing, which increases flexibility to incorporate the test Morb Mortal Wkly Rep 2006; 55: 301–305.
into a routine diagnostic laboratory schedule. After 3 Zignol M, Hosseini M S, Wright A, et al. Global incidence of
DNA extraction, the MAS-PCR can be completed in multidrug-resistant tuberculosis. J Infect Dis 2006; 194: 479–
485.
2 h, whereas the MTBDR test requires 6 h of post- 4 World Health Organization. Anti-tuberculosis drug resistance
amplification processing (including PCR, electropho- in the world: the WHO/International Union Against Tubercu-
resis and hybridisation). losis and Lung Disease global project on drug resistance sur-
The performance of the MAS-PCR test in routine veillance, 1994–1997. WHO/TB/97.229. Geneva, Swizerland:
use as an MDR detection tool should be evaluated WHO, 1997.
5 van Doorn H R, de Haas P E, Kremer K, Vandenbroucke-
both on cultures and directly on clinical samples. How-
Grauls C M, Borgdorff M W, van Soolingen D. Public health
ever, molecular tests are only appropriate diagnostic impact of isoniazid-resistant Mycobacterium tuberculosis strains
tools in laboratories that have sufficient resources and with a mutation at amino-acid position 315 of katG: a decade
expertise to implement quality control and assurance, of experience in The Netherlands. Clin Microbiol Infect 2006;
particularly regarding contamination. 12: 769–775.
6 Sun Y J, Lee A S, Wong S Y, et al. Genotype and phenotype re-
There were 13 discrepant cases between sequenc-
lationships and transmission analysis of drug-resistant tuber-
ing and rpoB MAS-PCR. Eight of these isolates have culosis in Singapore. Int J Tuberc Lung Dis 2007; 11: 436–
442.
* www.finddiagnostics.org 7 Cohen T, Murray M. Modeling epidemics of multidrug-resistant
 MAS-PCR for RMP-resistant TB 1311

M. tuberculosis of heterogeneous fitness. Nat Med 2004; 10: Mycobacterium tuberculosis isolates from a hospital in Vietnam.
1117–1121. J Clin Microbiol 2006; 44: 2333–2337.
8 Blower S M, Chou T. Modeling the emergence of the ‘hot 22 Johansen I S, Lundgren B, Sosnovskaja A, Thomsen V O. Direct
zones’: tuberculosis and the amplification dynamics of drug re- detection of multidrug-resistant Mycobacterium tuberculosis in
sistance. Nat Med 2004; 10: 1111–1116. clinical specimens in low- and high-incidence countries by line
9 Resch S C, Salomon J A, Murray M, Weinstein M C. Cost- probe assay. J Clin Microbiol 2003; 41: 4454– 4456.
effectiveness of treating multidrug-resistant tuberculosis. PLoS 23 Van Rie A, Warren R, Mshanga I, et al. Analysis for a limited
Med 2006; 3: e241. number of gene codons can predict drug resistance of Myco-
10 Caminero J A. Treatment of multidrug-resistant tuberculosis: bacterium tuberculosis in a high-incidence community. J Clin
evidence and controversies. Int J Tuberc Lung Dis 2006; 10: Microbiol 2001; 39: 636–641.
829–837. 24 Sandman L, Schluger N W, Davidow A L, Bonk S. Risk factors
11 Gupta R, Irwin A, Raviglione M C, Kim J Y. Scaling up treat- for rifampicin-monoresistant tuberculosis: a case-control study.
ment for HIV/AIDS: lessons learned from multidrug-resistant Am J Respir Crit Care Med 1999; 159: 468– 472.
tuberculosis. Lancet 2004; 363: 320–324. 25 Gillespie S H. Evolution of drug resistance in Mycobacterium
12 Baltussen R, Floyd K, Dye C. Cost effectiveness analysis of tuberculosis: clinical and molecular perspective. Antimicrob
strategies for tuberculosis control in developing countries. BMJ Agents Chemother 2002; 46: 267–274.
2005; 331: 1364. 26 Sajduda A, Brzostek A, Poplawska M, et al. Molecular charac-
13 Morgan M, Kalantri S, Flores L, Pai M. A commercial line terization of rifampicin- and isoniazid-resistant Mycobacterium
probe assay for the rapid detection of rifampicin resistance in tuberculosis strains isolated in Poland. J Clin Microbiol 2004;
Mycobacterium tuberculosis: a systematic review and meta- 42: 2425–2431.
analysis. BMC Infect Dis 2005; 5: 62. 27 Wada T, Maeda S, Tamaru A, Imai S, Hase A, Kobayashi K.
14 Hillemann D, Weizenegger M, Kubica T, Richter E, Niemann S. Dual-probe assay for rapid detection of drug-resistant Myco-
Use of the genotype MTBDR assay for rapid detection of rifam- bacterium tuberculosis by real-time PCR. J Clin Microbiol 2004;
picin and isoniazid resistance in Mycobacterium tuberculosis 42: 5277–5285.
complex isolates. J Clin Microbiol 2005; 43: 3699–3703. 28 Bobadilla-del-Valle M, Ponce-de-Leon A, Arenas-Huertero C,
15 Makinen J, Marttila H J, Marjamaki M, Viljanen M K, Soini H. et al. rpoB gene mutations in rifampicin-resistant Mycobacte-
Comparison of two commercially available DNA line probe rium tuberculosis identified by polymerase chain reaction single-
assays for detection of multidrug-resistant Mycobacterium tu- stranded conformational polymorphism. Emerg Infect Dis 2001;
berculosis. J Clin Microbiol 2006; 44: 350–352. 7: 1010–1013.
16 Soudani A, Hadjfredj S, Zribi M, et al. Characterization of Tu- 29 Cavusoglu C, Hilmioglu S, Guneri S, Bilgic A. Characterization
nisian Mycobacterium tuberculosis rifampicin-resistant clinical of rpoB mutations in rifampicin-resistant clinical isolates of
isolates. J Clin Microbiol 2007; 45: 3095–3097. Mycobacterium tuberculosis from Turkey by DNA sequencing
17 Traore H, van Deun A, Shamputa I C, Rigouts L, Portaels F. and line probe assay. J Clin Microbiol 2002; 40: 4435– 4438.
Direct detection of Mycobacterium tuberculosis complex DNA 30 Kapur V, Li L L, Iordanescu S, et al. Characterization by auto-
and rifampicin resistance in clinical specimens from tuberculosis mated DNA sequencing of mutations in the gene (rpoB) encod-
patients by line probe assay. J Clin Microbiol 2006; 44: 4384– ing the RNA polymerase beta subunit in rifampicin-resistant
4388. Mycobacterium tuberculosis strains from New York City and
18 Somoskovi A, Dormandy J, Mitsani D, Rivenburg J, Salfinger Texas. J Clin Microbiol 1994; 32: 1095–1098.
M. Use of smear-positive samples to assess the PCR-based gen- 31 van Soolingen D, Hermans P W, de Haas P E, Soll D R, van
otype MTBDR assay for rapid, direct detection of the Myco- Embden J D. Occurrence and stability of insertion sequences in
bacterium tuberculosis complex as well as its resistance to iso- Mycobacterium tuberculosis complex strains: evaluation of an
niazid and rifampicin. J Clin Microbiol 2006; 44: 4459– 4463. insertion sequence-dependent DNA polymorphism as a tool in
19 Brossier F, Veziris N, Truffot-Pernot C, Jarlier V, Sougakoff W. the epidemiology of tuberculosis. J Clin Microbiol 1991; 29:
Performance of the genotype MTBDR line probe assay for de- 2578–2586.
tection of resistance to rifampicin and isoniazid in strains of 32 Heep M, Brandstatter B, Rieger U, et al. Frequency of rpoB
Mycobacterium tuberculosis with low- and high-level resistance. mutations inside and outside the cluster I region in rifampicin-
J Clin Microbiol 2006; 44: 3659–3664. resistant clinical Mycobacterium tuberculosis isolates. J Clin
20 Cavusoglu C, Turhan A, Akinci P, Soyler I. Evaluation of the Microbiol 2001; 39: 107–110.
genotype MTBDR assay for rapid detection of rifampicin and 33 Hillemann D, Rusch-Gerdes S, Richter E. Evaluation of the
isoniazid resistance in Mycobacterium tuberculosis isolates. genotype MTBDRplus assay for rifampicin and isoniazid suscep-
J Clin Microbiol 2006; 44: 2338–2342. tibility testing of Mycobacterium tuberculosis strains and clini-
21 Caws M, Duy P M, Tho D Q, Lan N T, Hoa D V, Farrar J. cal specimens. J Clin Microbiol 2007; 45: 2635–2640.
Mutations prevalent among rifampicin- and isoniazid-resistant

RÉSUMÉ

C O N T E X T E : Hôpital Pham Ngoc Thach pour la tuber- le test commercial MTBDR (M. tuberculosis Drug Re-
culose (TB) et les maladies respiratoires, hôpital de ré- sistance) dans des isolats de routine résistants à la RMP
férence tertiaire pour la TB au Sud du Vietnam. (n = 104) ou sensibles à la RMP (n = 50) en fonction des
O B J E C T I F : Développer et évaluer un test de réaction en tests conventionnels phénotypiques de sensibilité à 1%.
chaîne par polymérase (PCR) spécifique à des allèles R É S U LTAT S : La sensibilité du test MAS-PCR a été de
multiples (MAS-PCR), simple, rapide et précis pour dé- 83,7% (IC95% 75,1–90,2) et celle du test MTBDR de
tecter les mutations ponctuelles liées à la résistance à la 93,3% (IC95% 86,6–97,3). La spécificité a été de 100%
rifampicine (RMP) aux codons 516, 526 ou 531 dans le pour les deux tests. La valeur prédictive négative a été de
gène rpoB de Mycobacterium tuberculosis. 74,6% (IC95% 65,3–83,1) pour le test MAS-PCR et de
S C H É M A : On a comparé le nouveau MAS-PCR avec 87,7% (IC95% 80,0–93,6) pour le test MTBDR.
1312 The International Journal of Tuberculosis and Lung Disease

C O N C L U S I O N : Le test MTBDR, bien que plus sensible, alternative moins difficile et économique. Un résultat de
est actuellement beaucoup trop coûteux dans les con- sensibilité provenant de l’un ou l’autre de ces tests ne
textes à faibles ressources et à haut fardeau de TB et peut pas permettre d’exclure une TB multirésistante.
pour cette raison le MAS-PCR décrit ici représente une

RESUMEN

MARCO DE REFERENCIA : El hospital de tuberculosis R E S U LTA D O S : La sensibilidad de la MAS-PCR fue

(TB) y enfermedades respiratorias Pham Ngoc Thach, 83,7% (IC95% 75,1–90,2) y la sensibilidad de la prueba
tercer hospital de referencia de TB en Vietnam del Sur. MTBDR fue 93,3% (IC95% 86,6–97,3). Ambas prue-
O B J E T I V O : Elaborar y evaluar una prueba de reacción bas tuvieron una especificidad de 100%. El valor de pre-
en cadena de la polimerasa (PCR) con amplificación dicción negativa fue 74,6% (IC95% 65,3–83,1) para la
múltiple (multiplex) de alelos específicos (MAS-PCR), MAS-PCR y 87,7% (IC95% 80,0–93,6) para la prueba
que sea sencilla rápida y precisa, con el fin de detectar MTBDR.
mutaciones puntuales en los codones 516, 526 o 531 del C O N C L U S I O N E S : Si bien la prueba de genotipo
gen rpoB de Mycobacterium tuberculosis, que se aso- MTBDR es más sensible, su costo es excesivo en la actu-
cian con la resistencia a rifampicina (RMP). alidad en medios con escasos recursos y alta carga de
M É T O D O S : La nueva MAS-PCR se comparó con la morbilidad ; la MAS-PCR descrita en el presente estudio
prueba comercial de genotipo M. tuberculosis drug re- constituye una opción económica y menos dispendiosa.
sistance (MTBDR), en 104 aislados clínicos recogidos Un resultado de susceptibilidad obtenido con cualquiera
en forma sistemática, resistentes a RMP y en 50 aislados de estos métodos no se puede utilizar con el fin de excluir
sensibles, definidos mediante el método convencional la TB multidrogorresistente.
fenotípico de susceptibilidad del 1%.