Evaluation of a new automated, rapid, colorimetric culture system using solid medium for laboratory diagnosis of tuberculosis and determination of

Transcription

1 INT J TUBERC LUNG DIS 8(6): IUATLD Evaluation of a new automated, rapid, colorimetric culture system using solid medium for laboratory diagnosis of tuberculosis and determination of anti-tuberculosis drug susceptibility O. Baylan, O. Kisa, A. Albay, L. Doganci Department of Microbiology and Clinical Microbiology, Gulhane Military Medical Academy, Ankara, Turkey SUMMARY SETTING: Department of Microbiology and Clinical Microbiology, Gulhane Military Medical Academy, Ankara, Turkey, a tertiary referral hospital in a region endemic for tuberculosis. OBJECTIVE: To evaluate the performance of the Culture System (CS), a new, rapid, automated colorimetric culture system. DESIGN: CS results were compared with routinely used Löwenstein Jensen (LJ) medium and Bactec 460 TB. RESULTS: In this study, 449 specimens, mostly sputum samples obtained from 348 patients, were evaluated. Mycobacteria were isolated from 31 (6.9%), 23 (5.1%), 18 (4.0%) and 21 (4.7%) of the specimens using Bactec 12B, LJ, Medium and SLC (selective), respectively. The mean time to detection of growth of 13 isolates by Bactec 12B, Medium, SLC and LJ medium was respectively 8.9, 15.1, 17.0 and 26.1 days. CONCLUSION: may be a practical and rapid culture system for daily use. However, the manufacturer should improve the system to minimise the effects of manipulation errors. Comparative studies with a larger number of isolates are needed to standardise drug concentrations used in anti-tuberculosis drug susceptibility testing. KEY WORDS: tuberculosis; Culture System; Löwenstein Jensen medium; Bactec 460 TB THE INCREASING INCIDENCE of tuberculosis (TB) has made it essential for laboratories to quickly detect and identify mycobacteria from human clinical material. 1 Culture methods still represent the gold standard for a definitive diagnosis of TB, providing isolates for identification and drug susceptibility testing (DST), but the delay in obtaining results remains a problem. 2,3 Despite the advantages of broth-based culture systems, traditional solid media still play a role in the recovery of mycobacteria from clinical samples, and are recommended by the Centers for Disease Control and Prevention (CDC) for use along with liquid media. 4,5 Culture System (CS) (Diomed Inc., Istanbul, Turkey) includes Medium, SLC (selective), PNB (p-nitrobenzoic acid), Dio- TK INH (isoniazid), RMP (rifampicin), Dio- TK SM (streptomycin) and EMB (ethambutol) media. Medium is a rapid solid culture medium, with multiple colour dye indicators, that enables early detection of mycobacterial growth. During the incubation period, the original red colours of media turn yellow with mycobacterial growth and green with many other bacterial or fungal species. The colour change depends on the metabolites and enzymes produced by different species, and occurs well before the colonies become visible. The CS uses neither radioactivity nor ultraviolet light, and it has the ability to differentiate mycobacterial growth from contamination. Scan detects colour changes on the Dio- TK media and automatically incubates the tubes at 37 C. Contamination can be detected without microscopic examination. SLC contains antibacterials and an antifungal agent (PolyPANT: 5 g/ml polymixin B, 50 g/ml piperacillin, 25 g/ml amphotericin B, 20 g/ml nalidixic acid, 2 g/ml trimethoprim) which inhibit the growth of many contaminant organisms and provide a better opportunity for mycobacterial isolation. 6 PNB contains p- nitrobenzoic acid (750 g/ml), which selectively inhibits the growth of mycobacteria belonging to the Correspondence to: Orhan Baylan, MD, Assistant Professor, Department of Microbiology and Clinical Microbiology, Gulhane Military Medical Academy and Medical School, Etlik, Ankara, Turkey. Tel: ( 90) Fax: ( 90) obaylan@gata.edu.tr Article submitted 6 August Final version accepted 24 November 2003.

2 Culture System 773 Mycobacterium tuberculosis complex (MTC), and enables rapid differentiation of these from nontuberculous mycobacteria (NTM). 6 8 Different antituberculosis drugs, including INH, RMP, SM, and EMB media, enable rapid DST to be performed. 6 Scan, the automated incubator reader for CS, is the first system to show the graphics of mycobacterial growth and contamination and to differentiate contamination from actual mycobacterial growth by continuous colorimetric analysis. The software analysis provides descriptive growth curves for rapid and easy evaluation. The system guides the user in diagnosis and provides individual and group statistical analysis of the samples. The mobile incubator, similar in size and shape to a refrigerator, has four shelves with access for computerised monitoring. Each shelf holds up to 117 culture tubes, with a total capacity of 468 tubes. There is a scanner at the bottom of each shelf, and the tube racks directly above have transparent bases; this is the hardware of the system. The main working principle of the system is the imaging of the materials placed over the scanner. This image is divided into squares reflecting the coordinates of each tube. Properties such as the colour of each square and scanning time are linked to the patient data and stored in the system. Periodic scans are made at predefined intervals. Data obtained from the scans are added to each patient s file and stored. Further data are produced by the colour changes, which are time dependent. The system s software provides statistical analyses and produces patient reports based on data evaluation. The system is practical and user friendly. 6 Dio-Safeprocess (Diomed Inc) is a decontamination and concentration kit consisting of a sample collection tube containing sterile glass beads and N- acetyl-l-cysteine (NALC), a tube containing phosphate buffer saline (PBS), a tube containing 3% sodium hydroxide (NaOH) solution, and a tube containing antiseptic solution. The kit allows clinical samples to be processed safely and easily for microscopy, culture (conventional and radiometric) and molecular methods for isolation and identification of mycobacteria. Samples can be collected directly into the collection tube or transferred from other tubes in which samples were previously collected. NaOH solution is added to the sample. Processing with NaOH-NALC decontaminates the samples by killing microorganisms susceptible to NaOH. Shaking with glass beads enables quick homogenisation. The suspension, which has a very alkaline ph due to the addition of NaOH, should be neutralised using the PBS. The mixture is then centrifuged to sediment the bacilli and the supernatant is discarded. The sediment in which the bacilli are concentrated can be used for microscopy, culture and molecular methods. The antiseptic solution is poured into the contaminated sample tubes to kill any potential pathogens, and the kit can then be disposed of safely. The Dio-Safeprocess kit eliminates the risk of serial contamination of samples. 6 The aim of the present study was to evaluate the efficacy (recovery rate, time to detection and DST of mycobacteria) of CS, a new colorimetric, automated rapid culture system, and to compare it with routinely used conventional media, Löwenstein Jensen (LJ) (Diomed Inc) and the Bactec 460 TB (B460) culture system (Becton Dickinson Diagnostic Instruments, Sparks, MD, USA) in terms of laboratory diagnosis of TB and DST. MATERIALS AND METHODS Patients and specimens In this study, 449 specimens, mostly sputum samples, obtained from 348 patients from different clinics were evaluated for the diagnosis of TB between March 2002 and April 2003 in the Mycobacteriology Laboratory of the Department of Microbiology and Clinical Microbiology, Gulhane Military Medical Academy. Culture procedures A Class IIA/B3 biological safety cabinet (Forma- Scientific, Inc., Marietta, OH, USA) was used for the study in the mycobacteriology laboratory. A maximum of 10 ml of sputum and other viscous samples were collected into Dio-Safeprocess plastic 50 ml conical centrifuge tubes with glass beads. The same amount of NaOH-NALC solution was added to the tubes. After homogenisation of the specimens by vortexing, the tubes were incubated at room temperature for min. After the addition of PBS (ph 6.8) to reach 50 ml solution in the tubes, each specimen was concentrated by refrigerated centrifugation at 4000 g for min. The supernatant was then discarded into the emptied PBS tube, leaving the glass beads in the tube. The sediment remaining among the glass beads was vortexed to obtain a suspension for microscopy and culture. After the processing of the sample, approximately half of the antiseptic solution was poured into the sample collection tube and the other half into the PBS tube; both tubes and their contents were discarded after 15 min. The same amount of suspension (100 l) was inoculated into LJ solid medium, and SLC solid media. For the Bactec 12B liquid medium the amount inoculated was 500 l. All the media were incubated at 37 C. The incubation periods for Bactec 12B, LJ (in 5 10% C0 2 ), Medium (in Scan) and SLC (in Scan) were 6 weeks, 8 weeks, 30 days and 30 days, respectively. Smears were prepared from the residual suspension for Ehrlich-Ziehl-Neelsen (EZN) staining for microscopic examination. All the patient data were entered into the Scan software. Culture of the samples by B460 and preparation of inocula for DST were performed according to the

3 774 The International Journal of Tuberculosis and Lung Disease manufacturer s instructions. 9,10 If MTC growth was detected in the Bactec 12B vials, DST of these microorganisms against four first-line anti-tuberculosis drugs (INH, RMP, EMB, SM) was performed. The antibiotic concentrations tested were 2.5 g/ml for EMB, 0.1 g/ml for INH, 2.0 g/ml for SM, and 2.0 g/ ml for RMP. 10 LJ medium was examined once a week for 8 weeks. Smears of typical colonies on LJ medium were stained using EZN and examined microscopically. Scan detected mycobacterial growth and contamination automatically by distinguishing the colour changes in the tubes. EZN staining was also used for confirmation of growth in media. If growth was detected in any of the tubes ( Medium, SLC), only one of the tubes was used for identification and DST. Bacterial suspension was prepared by adding 100 l of suspension fluid to a growth culture tube and mixing gently using a wire loop. The suspension was transferred back into the suspension tube, which was vortexed for homogenisation. The inoculum for DST was prepared by adding the bacterial suspension to a new tube containing suspension fluid (roughly comparable with No 1 McFarland standard). After thorough mixing, 100 l of this suspension was inoculated into the drug-containing media tubes, a control medium without drug and PNB, which were part of the Anti TB kit. The drug concentrations were as follows: INH (0.2 g/ml), RMP (1.0 g/ml), SM (2.0 g/ml), and EMB (40.0 g/ml). Patient data were recorded by the Dio- TK Scan. After 1 week s incubation in PNB medium, the isolate was differentiated according to its colour; a change from red to yellow showed the presence of NTM, and no colour change indicated that it was a member of the MTC. In media containing anti-tuberculosis drugs, colour change from red to yellow indicated drug resistance and no colour change indicated drug susceptibility. 6 RESULTS In this study, 449 samples (205 sputum, 95 bronchoalveolar lavage fluid, 61 urine, 43 pleural fluid, 12 cerebrospinal fluid, 10 peritoneal fluid, nine gastric lavage, six abscess material, four synovial fluid, four urethral flux) obtained from 348 patients (254 males and 94 females) were evaluated. The results of microscopic examination of stained smears were negative for 438 specimens (97.6%) and positive for 11 specimens (2.4%). Positive and negative results of Bactec 12B were compared with the results of EZN staining and positive, negative and contamination results of LJ, Medium and SLC media (Table 1). Mycobacteria were isolated from respectively 31 (6.9%), 23 (5.1%), 18 (4.0%) and 21 (4.7%) of the Table 1 Comparison of Bactec 12B results with AFB, LJ, Medium and SLC results Samples (n 449) DIO-TK DIO-TK AFB LJ Medium SLC Cont Cont Cont Bactec (n 31) Bactec (n 418) AFB acid-fast bacilli; LJ Löwenstein-Jensen; SLC selective; n number of samples; Cont contamination. specimens in Bactec 12B, LJ, Medium and SLC. Two samples were acid-fast bacilli (AFB) smear-positive but negative on all culture methods. Thirty samples were MTC and one was NTM on Bactec 12B. Of 18 isolates grown on Medium, 16 were MTC and two were NTM, while 19 of 21 isolates grown on SLC were MTC and two were NTM. Respectively five and four isolates were positive by SLC and Medium, but were negative by B460. Two of four isolates detected on Medium and three of five isolates detected on SLC were identified as MTC using the PNB medium. Isolates that could not be detected by Bactec 12B, but were detected by Medium and SLC were confirmed as AFB by EZN staining. By contrast, 17 specimens were positive on B460, but negative on Medium, and 15 isolates were detected by B460 but not by SLC. The contamination ratios for Bactec 12B, Medium, SLC and LJ media were 1.3% (n 6), 3.8% (n 17), 1.6% (n 7) and 2.4% (n 11), respectively. The mean times to detection of isolates by Bactec 12B (n 31), Medium (n 18), SLC (n 21) and LJ (n 23) were 12.5 (range 2 27), 16.0 (5 24), 18.6 (7 27) and 29.6 (14 49) days, respectively. The mean times to detection of 13 isolates (grown by all four culture methods) without considering the AFB smear results, by Bactec 12B, Medium, SLC and LJ were 8.9, 15.1, 17.0 and 26.1 days, respectively. The average time required for the indication of mycobacterial growth Table 2 The mean time to detection (days) of mycobacteria in different media Bactec 12B Medium SLC AFB variable* (n 31) (n 14) (n 16) (n 23) AFB variable (n 13) AFB positive (n 7) * The mean growing time of the mycobacteria isolated in various numbers was calculated separately for each culture. All of the isolates were culture positive, but negative or positive for AFB. All of the isolates were positive both on the four culture methods and for AFB. SLC selective; LJ Löwenstein-Jensen; AFB acid-fast bacilli. LJ

4 Culture System 775 Table 3 DST results of isolates (n 16) isolated in CS and B460 DIO-TK CS INH RMP EMB SM n 16* S R S R S R S R Bactec 460 INH S 13 R 3 RMP S 15 1 R EMB S 15 1 R SM S 16 R * n number of isolates isolated in CS and B460. DST drug susceptibility testing; INH isoniazid; RMP rifampicin; EMB ethambutol; SM streptomycin; S susceptible; R resistant. in AFB-positive samples from which mycobacteria were isolated on all four types of media (n 7) was 8.6 days for Bactec 12B, 13.6 days for Medium, 14.9 days for SLC and 25.7 days for LJ (Table 2). DST of members of the MTC was performed using B460 (n 30) and the Anti TB kit (n 19). All of the strains (n 30) were susceptible to SM, EMB and RMP and three were resistant to INH by B460. Sixteen of 19 MTC isolates detected by Dio- TK Medium or SLC were susceptible to INH, SM, EMB and RMP, and one was resistant to INH, one was resistant to RMP and another was resistant to EMB with the Anti TB kit. Compared to the DST results of 16 MTC isolates detected by B460 and CS, three MTC isolates were resistant to INH with B460 and were susceptible with CS. On the other hand, two isolates resistant to EMB and RMP separately by CS were susceptible to these drugs with B460. Eleven isolates susceptible to four drugs in CS were also susceptible in B460. There was therefore a discrepancy between the DST results of B460 and CS for five isolates (Table 3). DISCUSSION A variety of different media for culture of mycobacteria have been described, but only a few are in use today. The Bactec 460 culture system has gained a reputation as a standard for comparison with other, newer systems. 5 The present study was undertaken to compare the newly developed automated CS against B460. To date, to our knowledge, only one report has been published on the evaluation of this new culture system, although some abstracts are available (Table 4) In our study, the highest isolation rates for mycobacteria were detected by B460, followed by LJ, Dio- TK SLC and Medium, respectively. However, five isolates were determined only by CS. The number of positive mycobacteria samples was very low in our study; the rate of AFB true-positive samples was only 2% (n 9) in 449 samples. Some of the mycobacteria could only be isolated by using some of the culture systems from the same inoculum, and thus different isolation rates were obtained for different culture systems. This may be due to low numbers of mycobacteria in patient samples which resulted in insufficient inocula in some of the media. In addition, the higher isolation rate in B460 may be due to the procedure of repeat decontamination of contaminated cultures, which is not done for the other culture media. Application of the same re-digestion procedure for LJ and media may also increase the isolation rate in these media. The CDC has recommended radiometric methods for DST because they can provide results for evaluating first-line anti-tuberculosis drugs more quickly than conventional testing on solid media. 14 In the last decade, the increasing number of multidrug-resistant strains of M. tuberculosis has stimulated efforts to develop rapid and accurate non-radiometric methods for DST. 15 In a study of 78 M. tuberculosis isolates, Bicmen et al. found resistance ratios for INH, RMP, EMB and SM of respectively 24.6%, 22.6%, 22.9% and 21.0% in CS and 18.3%, 15.6%, 11.0% and 16.6% in LJ medium. 16 Using, contamination ratios of 5%, 3.8%, 5% and 3.8% were detected for INH, Table 4 Studies of CS reported by other authors Patients n Samples n AFB positivity Bactec LJ Mean time to detection (days) Medium SLC Kocagoz et al. 11 (n 50) * 25.7* 9.6* 10.7* Ercis et al. 12 (n 15) Bicmen et al Present study (n 13) * Samples AFB-positive. Number of isolates detected not given. Samples with AFB variable. n number of isolates detected in all cultures; CS culture system; AFB acid-fast bacilli; LJ Löwenstein-Jensen; SLC selective.

5 776 The International Journal of Tuberculosis and Lung Disease RMP, EMB and SM, while with LJ they were 16.6%, 11.5%, 11.5% and 12.8%, respectively. All the isolates susceptible to four drugs in media were also susceptible in LJ medium. Mycobacteria did not grow from five of the specimens in LJ medium; the drug susceptibilities of these five were therefore studied in Medium. Two of the isolates were resistant to all four drugs, one was susceptible to all the drugs and the other two were resistant to both RMP and EMB. To confirm the DST results between B460 and CS, a third independent method, such as the Middlebrook agar dilution test, should ideally be performed using a larger number of isolates. Some limitations of CS were detected in our study. Media are ph sensitive, and if the NaOH treated samples are not properly neutralised, detection of growth by the colour change from red to yellow takes longer. The risk of contamination is much greater during the inoculation procedure due to the rubber stoppers used with the tubes compared to screw caps or vials. The colour change from red to yellow may sometimes be due to Gram-positive bacteria; any growth indication by colour change should be confirmed by AFB staining to prevent falsepositive results. Opening the tubes for staining purposes creates an additional risk of contamination. The DST results for some isolates were different in the comparison of B460 and CS. While B460 detected mycobacteria most rapidly, CS was approximately 2 weeks faster than LJ medium. A comparison of CS and nonradioactive rapid mycobacterial culture systems would be very useful. As conventional media such as LJ are those most commonly used worldwide, CS may be a good alternative in these laboratories, saving about 2 weeks in primary isolation and 10 days in DST. While visual evaluation of CS makes it a good candidate for low-income countries, where the TB problem is greatest, its sophisticated automated system is promising for modern mycobacteriology laboratories in industrialised countries. In summary, our data suggest that despite its shortcomings, the many advantages of the CS over other mycobacterial culture systems make it a practical and rapid system for daily use, and a suitable alternative to other currently available solid media, such as LJ, for recovery rate and time to detection of mycobacteria and DST. When used in combination with broth-based media, the performance of CS is comparable to that of LJ for detection of mycobacteria. Our repeated decontamination approach for contaminated cultures, which may have led to a higher isolation rate in B460, may be beneficial if applied to other culture systems and may increase the mycobacterial isolation rate. However, the manufacturer should improve the system to minimise the effects of misapplication in creating false results and to standardise the drug concentrations of DST. DST of a larger number of isolates and comparison with the results obtained from another gold standard medium such as Middlebrook is crucial for standardisation of drug concentrations. Acknowledgements This study was granted by DiaMed Inc which was supported by the Technology Development Foundation of Turkey and the Scientific and Technical Research Council of Turkey. The study was presented at the 14th European Congress of Clinical Microbiology and Infectious Diseases, Prague, Czech Republic, 4 May References 1 Hanna B A, Ebrahimzadeh A, Elliott L B, et al. Multicenter evaluation of the Bactec MGIT 960 system for recovery of mycobacteria. J Clin Microbiol 1999; 37: Mirovic V, Lepsanovic Z. Evaluation of the MB/BacT system for recovery of mycobacteria from clinical specimens in comparison to Lowenstein-Jensen medium. Clin Microbiol Infect 2002; 8: Brunello F, Favari F, Fontana R. Comparison of the MB/BacT and Bactec 460 TB systems for recovery of mycobacteria from various clinical specimens. J Clin Microbiol 1999; 37: Kent B D, Kubica G P. Public Health Mycobacteriology: a guide for the level III laboratory. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control, Centers for Disease Control and Prevention (CDC). Essential components of a tuberculosis prevention and control program. MMWR 1995; 44: RR Diomed Inc. Culture System: Rapid automated mycobacterial culture system. Product Insert. Istanbul, Turkey:. Diomed Inc., 2002 ( Accessed 13 April Rastogi N, Goh K S, David H L. Selective inhibition of the Mycobacterium tuberculosis complex by p-nitro-alpha-acetylaminobeta-hydroxypropio phenone (NAP) and p-nitrobenzoic acid (PNB) used in 7H11 agar medium. Res Microbiol 1989; 140: Collins T, Levett P N. Radiometric studies on the use of selective inhibitors in the identification of Mycobacterium spp. J Med Microbiol 1989; 30: Siddiqi S H. Procedure for primary isolation of mycobacteria from clinical specimens. In: Bactec TB system product and procedure manual. Sparks, MD: Becton Dickinson. Revision E, May 1996: pp II Siddiqi S H. Drug susceptibility testing. In: BACTEC TB system product and procedure manual. Sparks, MD: Becton Dickinson. Revision E, May 1996: pp IV Kocagoz T, Alp A, Albay A. A new rapid non-radioactive medium for culturing mycobacteria, that also enables visual differentiation of mycobacterial growth from contamination [Poster]. Los Angeles, CA: American Society for Microbiology General Meeting, May 21 25, Ercis S, Alp A, Hasçelik G, Kocagöz T. Comparison of rapid mycobacterium culture system with BACTEC 460 TB system and Löwenstein Jensen medium in diagnosis of tuberculosis and detection of susceptibility of antituberculosis drugs [abstract]. Abant, Bolu, Turkey: 4th National Mycobacterium Symposium, October 31 November 2, 2002: Bicmen C, Coskun M, Senol G, Florat N, Kocagöz T. Comparison of and Löwenstein-Jensen media for primary culture of mycobacteria: a preliminary study for a new medium [Poster]. Glasgow, UK: 13th European Congress of Clinical Microbiology and Infectious Diseases, May 10 13, Tokars J I, Rudnick J R, Kroc K, et al. US hospital mycobacte-

6 Culture System 777 riology laboratories: status and comparison with state public health department laboratories. J Clin Microbiol 1996; 34: Ardito F, Posteraro B, Sanguinetti M, Zanetti S, Fadda G. Evaluation of BACTEC Mycobacteria Growth Indicator Tube (MGIT 960) automated system for drug susceptibility testing of Mycobacterium tuberculosis. J Clin Microbiol 2001; 39: Bicmen C, Senol G, Coskun M, Florat N, Kocagöz T. Comparison of and Löwenstein-Jensen media in antimycobacterial susceptibility testing for Mycobacterium tuberculosis by using the proportion method [Poster]. Glasgow, UK: 13th European Congress of Clinical Microbiology and Infectious Diseases, May 10 13, RÉSUMÉ CONTEXTE : Le département de Microbiologie et de Microbiologie Clinique, Académie Médicale Militaire de Gulhane, Ankara, Turquie, un hôpital tertiaire de référence dans une région où la tuberculose est endémique. OBJECTIF : Evaluer les performances du système de culture CS, un nouveau système de culture colorimétrique, automatisé et rapide. SCHÉMA : On a comparé les résultats du système de culture au système de culture sur milieu de Löwenstein Jensen (LJ) utilisé en routine et au Bactec 460 TB. RÉSULTATS : Dans cette étude, 449 échantillons ont pu être étudiés, principalement des échantillons d expectoration obtenus chez 348 patients. Les mycobactéries ont été isolées respectivement dans 31 cas (6,9%) par le Bactec 12B, dans 23 (5,1%) par le LJ, dans 18 (4,0%) par Medium, et dans 21 (4,7%) par le SLC (sélectif). La durée moyenne de détection de la croissance pour 13 isolats a été respectivement de 8,9, 15,1, 17,0 et 26,1 jours pour le Bactec 12B, le Medium, le SLC et le milieu de LJ. CONCLUSION : Le CS pourrait être une technique de culture pratique et rapide pour l utilisation quotidienne. Toutefois, le fabricant devrait améliorer cette technique de culture pour minimiser les effets des erreurs de manipulation. Des études comparatives portant sur un plus grand nombre d isolats sont nécessaires pour standardiser les concentrations des médicaments utilisés dans les tests de sensibilité à l égard des médicaments antituberculeux. RESUMEN CONTEXTO : El Departamento de Microbiología y Microbiología Clínica de la Academia Médica Militar de Gulhane, Ankara, Turquía, un hospital terciario de referencia en una región donde la tuberculosis es endémica. OBJETIVO : Evaluar el rendimiento del Sistema de Cultivo (CS), un nuevo sistema de cultivo colorimétrico, automatizado y rápido. DISEÑO : Los resultados del CS fueron comparados con los sistemas de cultivo en medio de Löwenstein- Jensen (LJ) y Bactec 460 TB, utilizados en la práctica de rutina. RESULTADOS : En este estudio se evaluaron 449 muestras, en su mayoría de esputo obtenidas de 348 pacientes. Se aislaron micobacterias de 31 (6,9%), 23 (5,1%), 18 (4,0%) y 21 (4,7%) muestras, con Bactec 12B, LJ, Medium y SLC (selectivo), respectivamente. El tiempo promedio de detección del crecimiento para 13 aislados fue de 8,9, 15,1, 17,0 y 26,1 días para el Bactec 12B, Medium, SLC y medio LJ, respectivamente. CONCLUSIÓN : El CS puede ser un sistema de cultivo práctico y rápido para uso cotidiano. Sin embargo, el fabricante debe mejorar el sistema para minimizar los efectos de los errores de manipulación. Se necesita efectuar estudios comparativos con un mayor número de aislados para estandarizar las concentraciones de las drogas utilizadas en los tests de sensibilidad a los medicamentos antituberculosos.