PAGES: 827-830 DOI: Full paper
Evaluation of Indirect Susceptibility Testing of Mycobacterium tuberculosis to the First- and Second-line, and Alternative Drugs by the Newer MB/BacT System

Angela Maria Werneck Barreto +, Joselba Borges Melo Araújo, Reginalda Ferreira de Melo Medeiros, Paulo César de Souza Caldas

Serviço de Laboratório, Centro de Referência Professor Hélio Fraga, Estrada de Curicica 2000, 22710-552 Rio de Janeiro, RJ, Brasil


In order to evaluate the Organon Teknika MB/BacT system used for testing indirect susceptibility to the alternative drugs ofloxacin (OFLO), amikacin (AMI), and rifabutin (RIF), and to the usual drugs of standard treatment regimes such as rifampin (RMP), isoniazid (INH), pyrazinamide (PZA), streptomycin (SM), ethambutol (EMB), and ethionamide (ETH), cultures of clinical specimens from 117 patients with pulmonary tuberculosis under multidrug-resistant investigation, admitted sequentially for examination from 2001 to 2002, were studied. Fifty of the Mycobacterium tuberculosis cultures were inoculated into the gold-standard BACTEC 460 TB (Becton Dickinson) for studying resistance to AMI, RIF, and OFLO, and the remaining 67 were inoculated into Lowenstein Jensen (LJ) medium (the gold standard currently used in Brazil) for studying resistance to RMP, INH, PZA, SM, EMB, and ETH. We observed 100% sensitivity for AMI (80.8-100), RIF (80.8-100), and OFLO (78.1-100); and 100% specificity for AMI (85.4-100), RIF (85.4-100), and OFLO (86.7-100) compared to the BACTEC system. Comparing the results obtained in LJ we observed 100% sensitivity for RMP (80-100), followed by INH - 95% (81.8-99.1), EMB - 94.7% (71.9-99.7), and 100% specificity for all drugs tested except for PZA - 98.3 (89.5-99.9) at 95% confidence interval. The results showed a high level of accuracy and demonstrated that the fully automated, non-radiometric MB/BacT system is indicated for routine use in susceptibility testing in public health laboratories.

Since 1979, Brazil has adopted a short-course chemotherapy which uses rifampin (RMP), isoniazid (INH), and pyrazinamide (PZA) for pulmonary and extrapulmonary tuberculosis as the first choice of treatment; if the short-course treatment fails, ethambutol (EMB), ethionamide (ETH), streptomycin (SM), and PZA are used for retreatment. When patients, after treatment with the standard regimes, are not cured and their mycobacteria show in vitro resistance to RMP, INH, and one additional drug, they constitute the group of multidrug-resistant patients (TBMDR). This definition was established by the Brazilian Consensus on Tuberculosis of 1997, and reiterated in the recently revised standards (Brasil 2002b).

In 2000, the TBMDR epidemiological surveillance system was created in order to determine the magnitude of the problem and to test the feasibility of treatment regimes to be adopted to control it, in a cooperative effort among the federal Ministry of Health and the state Secretaries of Health. In the context of treatment for TBMDR patients, new drugs associated with old tuberculostatics and drugs remaining from other regimes made it possible to implement an alternative therapeutic scheme.

Prominent among the new drugs is ofloxacin (OFLO), a quinolone; and among the older drugs now being used again are amikacin (AMI), an aminoglycoside; and terizidone (TZ), a cycloserine (CS) analog (Dalcolmo 2000).

The MB/BacT (Organon Teknika Corp., Durham, North Carolina, US), a fully automated continuous monitoring system, is used to detect growth of mycobacteria, based on readings by reflectometers sensitive to the CO2 produced by the microorganisms. The system was evaluated on a preliminary basis for use not only as a method for primary isolation, but also to determine susceptibility to the drugs used for the standard treatment of tuberculosis, such as RMP, SM, EMB, INH, and PZA (Beer et al. 1997a, b).

The BACTEC 460 TB radiometric system (Becton Dickinson Diagnostic Instrument Systems, Sparks, Maryland, US) has been used as the gold standard by different investigators and has yielded results that were in close agreement with the MB/BacT (Beer et al. 1997a, b, Tortoli et al. 2000, Brunello & Fontana 2000).

Most public health laboratories in Brazil uses the proportion method according to Canetti et al. (1969), in Lowenstein Jensen (LJ) medium, to diagnose resistance to tuberculosis. This method is inexpensive and relatively simple although requires at least three weeks from cultured islolates to provide results.

The aim of this study was to evaluate the automated MB/BacT system to test the susceptibility of OFLO,AMI, and RIF, which are used as an alternative therapeutic scheme for TBMDR in Brazil, comparing the results with the BACTEC 460 TB radiometric system, and also to evaluate the MB/BacT system susceptibility results to the drugs usually employed in the standardized treatment regimes of the Ministry of Health (RMP, INH, PZA, SM, EMB, and ETH), comparing it to the proportion method in LJ medium.



Strains - A total of 117 strains obtained from individual patients at National Reference Laboratory (NRL) were studied. All of them were isolated from sputum and grew at 35/37 °C on LJ medium after three weeks incubation. The isolates were identified as M. tuberculosis complex using the AccuProbe® genetic probe (Gen Probe Incorporated, San Diego, California, US). Fifty isolates were MDR and recovered in the period of 2001 and 2002 from patientes under alternative treatment but with suspicion of failure. These strains were used to evaluate the ability of MB/BacT system for detect resistance to AMI, RIF, and OFLO using as the gold standard the BACTEC 460 TB radiometric system results. The remain 67 strains were recovered from patients suspected to be TBMDR and MB/BacT system for detect resistance to RMP, INH, PZA, SM, EMB, and ETH was assayed, comparing the results with the gold standard currently used in Brazil, the LJ proportion method. Susceptible standard strain M. tuberculosisH37Rv (ATCC 27294) was used for quality control of the all tests.

Drugs and concentrations used - The concentrations of these drugs in the LJ (Canetti et al. 1969) and the MB/BacT followed the recommendations of Beer et al. (1997b) and Barreto et al. (2002) in the proportions indicated in Table I. Except for RIF, which was obtained from Pharmacia (Spa, Italy), all the drugs were obtained from Sigma Chemical Co. (St. Louis, Missouri, US).

The drugs were diluted in sterile distilled water, except RMP, ETH, and RIF for which ethylene glycol was used, before being added to the culture medium under aseptic conditions.

Test in BACTEC 460 TB - The flasks were seeded into the BACTEC system as recommended by the manufacturer (Siddiqui 1996). The test inoculum was standardized from the active-phase bacterial growth in LJ medium, in distilled water with turbidity compared to tube no.1 on the McFarland scale; 0.1 ml volumes were inoculated into the system flasks, pre-prepared with drugs in the above-mentioned concentrations (Table I). The flasks were incubated at 35/37ºC; daily readings were taken and interpretation was done after comparing the changes in the growth index of the inoculated control with the test drug .

Test in LJ medium - The standardized inoculum from the same cultures used in the BACTEC was serially diluted to 10-6 and the dilutions of 10-3, 10-5, and 10-6 (control) were added to the culture media containing the drugs. The tubes were incubated at 35/37ºC and read after 21 days of incubation. The results was considered sensitive when the proportion of up to 1% and 10% (only for SM, ETH, and PZA) of colony growth was seen in the tubes containing drugs (Canetti et al. 1969).

Test in the MB/BacT system - Supplements were added to the flasks containing Middlebrook 7H9 liquid medium, as recommended by the manufacturer (Organon Teknika 1997). The test inoculum is described in the BACTEC 460TB test, with two decimal dilutions made. From the first dilution, 0.1 ml volumes were inoculated into flasks of the MB/BacT system containing the drugs in the above-mentioned concentrations (Table I), and one flask was inoculated with 0.1 ml of 1/100 suspension, used as a control. The test tubes with PZA, and PZA control were adjusted to pH 5.8. The flasks were incubated at 35/37ºC and monitored every 10 min to detect growth. The test was finalized when the 1/100 tube showed growth.

Analysis of the data - The following characteristics were determined for each drug in the tests as recommended by IUATLD (2001): sensitivity, the capacity of the test method correctly to identify resistant strains [=A/(A+C)]; specificity, the capacity of the test correctly to identify susceptible strains [= D/(B+D) ]; positive predictive value (PPV), the ratio of strains classified as resistant by the test method which were truly resistant [=A/(A+B)]; and negative predictive value (NPV), the ratio of strains classified as susceptible by the test method which were truly susceptible [= D/(C+D)]. A is the number of strains found to be resistant by both methods, B is the number of strains found to be susceptible by the standard method and resistant by the test method, C is the number of strains found to be resistant by the standard method and susceptible by the test method, and D is the number of strains found to be susceptible by both methods.



The results obtained by comparing the MB/BacT with the BACTEC system were in complete agreement showing 100% sensitivity and specificity at the 95% confidence interval for AMI (80.8-100 and 85.4-100), RIF (80.8-100, and 85.4-100), OFLO (78.1-100, and 86.7-100), respectively. Table II lists the results of the tests for susceptibility to the drugs RMP, INH, PZA, EMB, SM, and ETH with the MB/BacT system and with the proportion method in LJ medium. The best results for sensitivity were obtained with RMP - 100% (80-100), PZA - 100% (59.8-100), INH - 95% (81.8-99.1), and EMB - 94.7% (71.9-99.7) followed by ETH - 90.9 (57.1-99.5), and SM - 88.5 (65.5-98.2), and for specificity with all drugs was 100%: RMP - (90.6-100), INH - (84.5-100), SM - (90.8-100), EMB - (90.8-100), ETH - (92-100), followed by PZA with 98.3% (89.5-99.9), at the 95% confidence interval. The spectrum of the 47 multidrug-resistant strains analized in this investigation was: 9 (19%) were resistant to all drugs tested (first-, second-line, and alternative drugs); 7 (14.8%) were resistant to 8 drugs; 5 (10.6%) were resistant to 7 drugs. In relation to drugs from the first- and second-line regimes 13 (27.6%) were resistant to all 6 drugs, 15 (31.9%) were resistant to 4 or 5 drugs, and the 19 (40.4%) remaining were resistant to 3 drugs. In our experiments all readings of the tests in LJ were finalized after 3 weeks of incubation, and those in the BACTEC system after 7 days. In the experiments with the MB/BacT, 80% of the tests were finalized in 10 days and the remainder in up to 13 days.



In Brazil, patients with multidrug-resistant tuberculosis are treated using OFLO, TZ, AMI, and RIF in combination with those drugs remaining from standardized treatment schemes which show susceptibility when tested in vitro. Since the establishment of the TBMDR surveillance program, in effect since 2000, about 1 100 patients have been enrolled as multidrug-resistant, and at the moment 800 patients are undergoing treatment (Brasil 2002a). The majority of the patients analized in this study came from Rio de Janeiro (34.72%) and could reflect the prevalence of the problem in that state.

The results of the antibiogram are fundamental to notification and adequate treatment of these patients. The automated methods developed to isolate mycobacteria, which use Middlebrook 7H9 liquid medium with a system for early detection of growth, greatly shorten the time required to obtain the antibiogram compared to the proportion method in LJ or Agar. In our experiments, all readings of the tests in the MB/BacT were made for a few more days than in BACTEC. This small difference in the final reading test time compared to the BACTEC was also observed by Tortoli et al. (2000). The availability of the susceptibility results obtained in our study is within the time limit suggested by Tenover et al. (1993) for rapid intervention in the disease transmission chain, acceptable for public health laboratories in this era of TBMDR: no more than 30 days should pass from collection of the clinical material to the time that the results become available.

Pfyffer et al. (1999) based in their multi-center study recommended the AMI, OFLO, and RIF concentrations for BACTEC method of 1, 2, and 0.5 µg/ml, respectively.

In our study the same concentrations were used except for RMP (2 µg/ml) in which Siddiqui's (1996) guidelines were followed.

Terizidone is an analog of cycloserine, which is used to treat TBMDR patients, however it was not included in this study. We do not routinely test CS because it shows wide variability in test readings, giving inconsistent results, as observed by Pfyffer et al. (1999).

The results obtained with the OFLO, AMI, and RIF showed a high degree of accuracy compared to the BACTEC, as did the results obtained with RMP, INH, and EMB in relation to the proportion method.

Some discordant results between the standard test and the test method could be attributed to the presence of borderline-resistant strains, mainly in relation to LJ tests where the final results depend on an accurate count of colonies. This was previously observed for INH (WHO 1997), and EMB (Madison et al. 2002).

Our observations are similar to those of Beer et al. (1997 a, b), Diaz-Infantes et al. (2000), Brunello and Fontana (2000), Tortoli et al. (2000), and Yew et al. (2001), using the same systems. The results of the tests in the MB/BacT indicate that this system is an accurate method for use in public health laboratories to deal with the workload of susceptibility-testing routines to diagnose multidrug-resistant tuberculosis mainly in those states that the prevalence of the disease is high and where the laboratories act as a reference facility to the national network.



To Dr Carlos Eduardo Dias Campos for his critical review, and to Dr Janet W Reid for the English translation of the manuscript.



Barreto AMW, Araújo JBM , Medeiros RFM , Caldas PCS 2002. Direct sensitivity test of the MB/BacT system. Mem Inst Oswaldo Cruz 97: 263-264.

Beer J, Feldmann K, Gogolin J, Jaksch G, Kuchler R, Rifai M, Rodloff AC, Rusch-Gerdes 1997a. Evaluation of the MB/BacT drug susceptibility system for INH, RMP, SM, EMB. A cooperative study of 6 laboratories. Clin Microbiol Infect 3: 1086.

Beer J, Kuchler R, Rodloff AC 1997b. Investigations about the possibility for testing the susceptibility of mycobacteria with the MB/BacT culture system. J Lab Med 21: 390-398.

Brasil-Fundação Nacional de Saúde, Centro de Referência Professor Hélio Fraga 2002a. Relatório de Gestão do Centro de Referência Professor Hélio Fraga 1999-2002, Rio de Janeiro, 28 pp.

Brasil-Ministério da Saúde, Secretaria de Políticas de Saúde, Departamento de Atenção Básica 2002b. Manual Técnico para o Controle da Tuberculose: Cadernos de Atenção Básica, 6th ed. revised and enlarged, Brasília, 62 pp.

Brunello F, Fontana R 2000. Reliability of the MB/BacT system for testing the susceptibility of Mycobacteriumtuberculosis complex isolates to antituberculous drugs. J Clin Microbiol 38: 872-873.

Canetti G, Fox W, Khomenko A, Mahler HT, Menon NK, Mitchinson DA, Rist N, Smelev NA 1969. Advances in techniques of testing mycobacterial drug sensitivity, and the use of sensitivity tests in tuberculosis control. Bull WHO 41: 21-23.

Dalcolmo MMP 2000. Regime de Curta Duração, Intermitente e Parcialmente Supervisionado, Como Estratégia de Redução de Abandono de Tratamento da Tuberculose no Brasil, Thesis, Escola Paulista de Medicina, Universidade Federal de São Paulo, São Paulo, 136 pp.

Diaz-Infantes MS, Ruiz-Serrano MJ, Martinez-Sancho L, Ortega A, Bouza E 2000. Evaluation of the MB/BacTMyco-bacterium detection system for susceptibility testing of Mycobacterium tuberculosisJ Clin Microbiol 38: 1988-1989.

IUATLD-International Union against Tuberculosis and Lung Diseases 2001. Research Methods for Promotion of Lung Health. A Guide to Protocol Development for Low-income Countries, Paris, 137 pp.

Madison B, Robinson-Dunn B, George I, Gross W, Lipman H, Metchock B, Sloutsky A, Washabaugh G, Mazurek G, Ridderhof J 2002. Multicenter evaluation of ethambutol susceptibility testing of Mycobacterium tuberculosis by agar proportion method and radiometric methods. J Clin Microbiol 40: 3976-3979.

Organon Teknika 1997. Manual Bact/Alert, Rio de Janeiro, 61 pp.

Pfyffer GE, Bonato DA, Ebrahimzadeh A, Gross W, Hotaling J, Kornblum J, Laszlo A, Roberts G, Salfinger M, Wittwer F, Siddiqui S 1999. Multicenter laboratory validation of susceptibility testing of Mycobacterium tuberculosisagainst classical second-line and newer antimicrobial drugs by using the radiometric BACTEC 460 technique and the proportion method with solid media. J Clin Microbiol 37: 3179-3186.

Siddiqui S 1996. BACTEC TB System. Product and Procedure Manual, Becton Dickinson, Sparks, Maryland, 165 pp.

Tenover FC, Crawford JT, Huebner RE, Geiter LJ, Horsburgh Jr CR, Good RC 1993. The resurgence of tuberculosis: is your laboratory ready? J Clin Microbiol 31: 767-770.

Tortoli E, Mattei R, Savarino A, Bartolini L, Beer J 2000. Comparison of Mycobacterium tuberculosis susceptibility testing performed with BACTEC 460 TB (Becton Dickinson) and MB/BacT (Organon Teknika) systems. Diagn Microbiol Infect Dis 38: 83-86.

WHO-World Health Organization 1997. Global Tuberculosis Programme Anti-tuberculosis Drug Resistance in The World, Geneva, 227 pp.

Yew WW, Tong SC, Lui KS, Leung SK, Chau CH, Wang EP 2001. Comparison of MB/BacT system and agar proportion method in drug susceptibility testing of Mycobacterium tuberculosis. Diagn Microbiol Infect Dis 39: 229-232.


Memórias do Instituto Oswaldo Cruz

Av. Brasil 4365, Castelo Mourisco
sala 201, Manguinhos, 21040-900
Rio de Janeiro, RJ, Brazil

Tel.: +55-21-2562-1222

This e-mail address is being protected from spambots. You need JavaScript enabled to view it.



marca fiocruzmarca brasil
marca faperjmarca cnpqmarca capes n marca cope

and diabetes. Erection dysfunction or ED is certainly one of mens most usual problem. It changes buy tadalafil 60mg A common drug is actually an imitation of its manufacturer twin. Both ought to be same in female cialis 20mg Long Phrase Viagra Use Fundamentally Damages Sex Lives This discount cialis canada Equally so, theres something to be said for the wonder of the second, captured forever on picture or a buy cheap cialis People extremely annoyed that they could only get three weeks at a time, Bunker noted. Retired persons cheap pharmacy These types of matters are possibly to being identified as having a result of cancer buy cialis 40mg - Yoghourt - fat-free simply Physical causes: Buying generic medicines now has been cheap generic cialis Herbaceous plants like nigrum and tribulus are well-known for his or her qualities in defeating impotence, which tadalafil 10mg It is not hard to consider Cialis that is generic. Most men start with one-10 mg dosage each purchase cialis Tadalafil quickly gained the moniker of weekender in Paris due to the fabulous results. The bash freaks buy female cialis