Azlocillin

Improved detection of Mycobacterium spp. using the Bactec (R) MGIT(TM) 960 system

Abstract: Until 1987, the notification rate for mycobacterial infection was on the decline; however, it now appears to be increasing once more. The reason for this may be multifactoral and include improved reporting of diagnosed cases, increased infection of an ageing population, homelessness, immunosuppression (e.g. due to human immunodeficiency virus infection), and immigration of people from countries where tuberculosis is endemic. This rising incidence and the increasing importance of resistant organisms mean that rapid identification by the clinical microbiology laboratory is required, and this is where an automated detection system can be an advantage. Over a two-year period, 2743 clinical specimen were examined for Mycobacterium spp. using the Bactec MGIT(TM) 960, and 286 were positive. Time to detection ranged from three to 14 days (mean: 9.3 days), and M. tuberculosis was recovered from 214 (75.5%). Contamination rate was higher (8.6%) than with manual methods, however. On balance, the Bactec(R) MGIT(TM) 960 system proved a valuable tool in the routine microbiology laboratory.

Key words: Automation. Culture media. Mycobacteriaceae.

Introduction

Of all the known infectious diseases, tuberculosis (TB) is the leading cause of morbidity and mortality in humans, and may affect individuals of any age.1 Until 1987, the notification rate for mycobacterial infections was in decline, but it now appears to be on the increase again (Table 1). Cases of TB in London, UK, notified to the Office of Population Censuses and Surveys rose from 1445 in 1987 (21 per 100 000) to 2130 in 1994 (31 per 100 000).2

The reason for this may be multifactoral and include improved clinical and laboratory reporting of diagnosed cases, increased infection in an ageing population, homelessness, immunosuppression (e.g. due to human immunodeficiency virus [HIV] infection) and immigration of people from countries where TB is endemic.3,4 In the USA, the rise began in the mid-- 1980s, where, in addition to the above, there was a deterioration in control programmes.5 In 1993, the World Health Organisation (WHO) recognised that TB had been neglected and was out of control in many countries, and classed it a global emergency.

A third of the world's population is thought to be infected with TB, although not all will develop the disease, and a further 300 million individuals will become infected - 90 million of whom will develop the disease and 30 million will die - in the next ten years. As with many infectious diseases, TB is a particular problem in developing countries, where it is responsible for over 25% of all preventable adult deaths.1,7In recent years, both in the UK and other developed countries there has been an increase in ethnically diverse populations. Many of these show a higher rate of TB carriage, and generally have added to the overall increase in TB cases seen.

Clinical and radiological findings permit only presumptive diagnosis of TB; however, there are several methods available for laboratory diagnosis of infection, and these include standard TB culture using Lowenstein-Jensen (LJ) medium (egg yoke with pyruvate) and a smear stained either using Ziehl-Neelsen (ZN) or auramine, and more recent technology such as the Bactec(R) system or the polymerase chain reaction (PCR).

Conventional microscopy (ZN or auramine stain) for acid/alcohol-fast bacteria (AAFB) is rapid but has a low sensitivity, and is unable to give information on susceptibility of the organism.8 Application of molecular biology techniques in mycobacteriology promised radical changes, but the routine diagnostic use of such procedures has a number of major drawbacks, including high variation in sensitivity, detection of non-viable bacteria, inability to obtain susceptibility data, and the high cost of the tests.9,10

Whichever method is used, rapid diagnosis of TB is important so that the necessary control and prevention steps can be taken to limit the spread of the disease, the administration of inadequate therapy be avoided, and the costs of hospitalisation reduced.

The traditional method for isolating Mycobacterium tuberculosis from human specimens is by culture on solid media, and this can take up to eight weeks and may lack sensitivity. The Becton-Dickinson Bactec(R) MGIT(TM) 960 non-radiometric fluid-culture system has demonstrated new advances, showing considerable reduction in the time required to detect mycobacteria, and an increase in the sensitivity of isolation. 11-13

Here, we assess the MGIT(TM) TI 960 system and its ability to improve the recovery rate and detection times for isolation of mycobacteria from clinical specimens.

Materials and methods Specimens

A total of 2743 clinical specimens were received for routine mycobacterial investigation. The majority were sputa (n = 1068), with the remainder including bronchial washings (n = 301), early morning urines (n = 420), sterile fluids (n = 391; cerebrospinal fluid [CSF], pleural, ascitic, synovial and pericardial fluids), abscess/pus (n = 144), and miscellaneous samples (n = 419; biopsy material, lymph nodes, wound swabs and gastric aspirates). Specimens were stored at 40C prior to processing.

Bactec(R) MGIT(TM) 960

The MGIT(TM) 960 is an automated instrument used to determine the presence of mycobacteria in clinical specimens, and alerts the operator to the presence and location of the positive tubes by means of indicator lights positioned on the front of the instrument. The tube stations are displayed on a video display screen and an LED display indicates the appropriate positive tube station.

A fluorescent compound is embedded in silicone on the bottom of 16 x 100 mm tubes and is sensitive to the presence of oxygen dissolved in the broth. Initially, the large amount of dissolved oxygen quenches emission from the compound and little fluorescence can be detected. Actively respiring microorganisms consume the oxygen and permit the fluorescence to be detected.

Specimen processing

Specimens were processed twice a week in a modified Class 2 microbiological safety cabinet. All fluids were concentrated by centrifugation prior to digestion. Lymph node and tissue specimens were cut into pieces and homogenised in a Griffith's tube before processing. Specimens collected from sterile sites were centrifuged and processed without decontamination.

All other specimens were decontaminated with N-acetyl-L-lysine (NALC)/2% NaOH,14 according to standard procedures. NaOH is effective both as a mucolytic and decontaminating agent. As a mucolytic agent, it is most effective at a final specimen concentration of 2%; however, this concentration is toxic both to contaminants and some mycobacteria. NALC is also a mucolytic agent.

In the BBL(TM) (Becton-Dickinson, USA) Mycoprep reagent bottle, NALC is combined with 2% NaOH. When this is diluted with an equal volume of specimen, it provides effective digestion and decontamination, and the final concentration of 1% NaOH is less toxic to mycobacteria. Sodium citrate is included in the reagent to bind heavy metal ions which may be present in the specimen and which can inactivate NALC.

All specimens were washed with sterile phosphate buffer (BBL(TM) Mycoprep phosphate buffer) and centrifuged at 3000 rpm for 15 min. The supernatant was discarded and the pellet resuspended in fresh phosphate buffer to a final volume of 1.5 mL. The mixture was used to prepare both smears and cultures.

Media and culturing methods

Samples (0.5 mL) were inoculated into the BBL(TM) Mycobacteria Growth Indicator Tubes (MGIT(TM), and a further 0.-140.25 mL inoculated onto LJ pyruvate slopes. All slopes were incubated at 37 deg C for eight weeks and inspected daily.

BBL(TM) MGIT(TM) tubes contain 110 (mu)L fluorescent indicator and 7 mL broth. The indicator contains Tris 4,7-diphenyl-1,10-phenanthroline ruthenium chloridepentahydrate in the silicone rubber base. The tubes are flushed with 10% CO^sub 2^ and covered with polypropylene caps. The broth consists of a modified Middlebrook 7H9 broth base containing casein (5.9 g), peptone (1.25 mL) and glycerol (3.1 mL).

To reduce contamination, the broth was supplemented with the BBL(TM) MGIT(TM) OADC enrichment/BBL(TM) MGIT(TM) PANTA antibiotic mixture prior to inoculation. The enrichment supplement contains 15 mL Middlebrook OADC (BBL'- OADC) enrichment mixture (bovine albumin [50 g], dextrose [20 g], catalase [0.03 g] and oleic acid [0.6 g]) The BBL(TM) MGIT(TM) PANTA vial contains a lyophilised mixture of antimicrobial agents (polymyxin B [6000 u], amphotericin B [600 (mu)g], nalidixic acid [2400 (mu)g], trimethoprim [600(mu)g] and azlocillin [600(mu)g]).

Initially, a processed sediment sample (0.5 mL) was added to 0.8 mL enrichment/antibacterial mixture (OADC + PANTA) in each MGIT(TM) tube. The tubes were incubated at 36(+/-1) deg C in the MGIT(TM) 960 instrument for 42 days. They were monitored automatically every 60 min for an increase in fluorescence.

Positive samples

Samples identified as positive by the system were removed from the instrument and a smear prepared, stained with auramine-phenol and examined for the presence of AAFB. The time to detection of a mycobacterium-positive sample was based on the date of the earliest positive indication that correlated with the AAFB smear.14

Positive cultures that failed to reveal AAFB in a smear were Gram stained; if this showed contaminating organisms they were eliminated from the study. If no contaminating organisms were seen, the tubes were reincubated. According to the protocol, if the same specimens were positive twice more and continued to be clear of AAFB or contaminating organisms they were eliminated from the study.

Quality control

M. tuberculosis (ATCC 27294) reference strain was used as a positive control and an uninoculated tube as a negative control. As each new batch of MGIT(TM)tubes and reagents were opned, a positive and a negative control were run in parallel.

Original smears prepared from all specimens were stained with auramine-phenol and examined using fluorescence microscopy. Positive slides were confirmed by overstaining with ZN.

Results

Over a two-year period, 2743 specimens were examined for Mycobacterium spp. Of these, 286 were found to be culture-positive. A breakdown by specimen type is shown in Table 2. None of the wound swabs were positive. The various Mycobacterium spp. recovered can be seen in Table 3. A comparison of culture results against the microscopy can be seen in Table 4.

The MGIT(TM) 960 system recovered all 286 mycobacterial species (recovery rate: 10.42%), detected 61 (100%) of the isolates from smear-positive specimens and 225 from smear-negative specimens. Of the 61 smear-positive cases, 46 grew M. tuberculosis; and this bacterium was recovered from 168 of the smear-negative cases. Non-tuberculosis mycobacteria (NTM) were isolated from 72 samples (smear-positive: 15; smear-negative: 57).

Speed of detection by the MGIT(TM) 960 system is summarised in Figure 1. Almost 80% of mycobacterial isolates grew within two weeks of inoculation; however, a known M. avium intracellularae grew within fours hours. Mean detection time for smearpositive samples was 6.1 days (range: 2-10), whilst for smear-negative samples this was slightly longer - 10.5 days (range: 7-42).

Using the MGIT(TM) 960 system, contamination was seen in 237 samples (8.6%). In three cases where the MGIT(TM)- 960 gave a positive signal, two showed no growth and the third harboured a slow-growing Nocardia sp. Contamination with fungal and other highly resistant organisms was seen commonly in sputum specimens, and was determined by microscopic examination of Gram-stained smears.

Discussion

Rapid diagnosis of TB is need so that appropriate antibiotic therapy can be initiated and further spread prevented. More importantly, multiresistant strains must be identified quickly and appropriate preventative measures taken. In the past, ZN and auramine stains have been used to obtain a rapid provisional result; however, such techniques lack sensitivity. Culture on standard agar (LJ slopes) is adequate, although some strains grow slowly, and liquid culture using Middlebrook formulation is compromised by difficulty in detecting growth and a high rate of contamination. In the past, the Bactec(R) 460 system proved versatile but had some drawbacks, not least of which was the use of radioactive isotopes.

Continuous monitoring is one of the advantages of the automated MGIT(TM)960 system whereby the process vials are read every 60 minutes, compared with LJ slopes that are read either daily or weekly depending on the standard operating procedure employed. The identification system used in the MGIT(TM)- permits detection of metabolic changes that occur before visible bacterial growth is evident on an LJ slope.15

In our study, the MGIT(TM)- 960 detected all the smear-positive M. tuberculosis isolates (n = 61) and 225 out of 227 (99.1 %) smear-negative specimens that grew mycobacteria. Two smear-negative samples that grew M. tuberculosis on LJ slopes after seven weeks' incubation, but were not detected by the automated system, were from Pakistani women in their thirties. Neither had received treatment prior to specimen collection, and the apparent lack of growth in the MGIT(TM)- 960 remains unexplained perhaps the organisms were very scanty or of a slow-growing strain.

The second important criterion of automated system performance is incubation time to growth detection. Here, mean time to detection of M. tuberculosis from smear-positive specimens was 6.1 days; for smear-negative specimens this was slightly longer at 10.5 days. Mean time to detection of smear-positive and smear-negative NTM was 7.3 and 13.3 days, respectively. Overall, mean time to detection was 9.3 days (Table 5).

Of particular interest during this study was the very early detection of one isolate, which occurred within four hours of inoculation. Subsequent identification showed it to be M. avium intracellularae, and was from a 60-year-old female in-patient of European origin.

Disadvantages of using the automated system included a higher contamination rate, seen predominantly in sputum samples (12.7%). Reasons for this could include delays in sample processing or inadequate prior decontamination. Unlike LJ slopes, which contain inhibitors that prevent the growth or organisms other than mycobacteria, the detection system employed in the MGTT 960 measures oxygen depletion produced by the growth of any organism present in the tube.

Conclusions

The results presented here indicate that the Bactec MGIT(TM) 960 system is a rapid and sensitive means to recover mycobacteria, proving more sensitive than routine culture methods. It offers several advantages, such as complete automation, continuous monitoring, better sensitivity, faster detection, and easier data management. Moreover, the system is non-radiometric. Although contamination rate was slightly higher than that seen with conventional methods, the faster detection time and sensitivity of the system outweighed the cost of reprocessing samples.

The authors would like to thank Dr Winning (Consultant Physician, West Middlesex University Hospital) for supplying specimens, and Dr Stephen Mortlock (Department of Molecular Biology. Quest Diagnostics) for helpful suggestions and review of the manuscript prior to publication.

References

1Clarke S. An introduction to tuberculosis. Biomed Sci 1996; 40: 670.

2CDSC. Tuberculosis surveillance control. Commun Dis Rep CDR Wkly 1996; 6: 2.

3CDSC. New UK guidance on the prevention and control of transmission of HIV-related and drug-resistant tuberculosis Common Dis Rep CDR Wkly 1999; 9: 6.

4CDSC. National survey of tuberculosis in England and Wales. Common Dis Rep CDR Wkly 1998; 8: 24.

5 CDSC. Tuberculosis in England and Wales 1982-1993; notifications exceeded predictions. Commun Dis Rep CDR Wkly 1995; 5: 3.

6 CDSC. Tuberculosis remains `the caption of all these men of death'. Commun Dis Rep CDR Wkly 1997; 7: 8.

7 Drobniewski FA, Yates MD. Multiple drug resistant tuberculosis. J Clin Pathol 1997; 50: 89-92.

8 Rohner P, Ninet B, Metral C, Emler S, Auckenthaler R. Evaluation of the MB/BacT system and comparison to the Bactec 460 system and solid media for isolation of mycobacteria from clinical specimens. J Clin Microbiol 1997; 35: 3127-31.

9 Somoskovi A, Magyar P. Comparison of the Mycobacteria Growth Indicator Tube with MB Redox, Lowenstein-Jensen, and Middlebrook 7h I media for recovery of mycobacteria in clinical specimens. J Clin Microbiol 1999; 37: 1366-9.

10 Kearns AM, Freeman R, Steward M, Magee JG. A rapid polymerase chain reaction technique for detecting M. tuberculosis in a variety of clinical specimens. J Clin Pathol 1998; 51: 922.

11 Hanna BA, Ebrahimzadeh A, Elliott LB et al. Multicenter evaluation of the Bactec MGIT(TM) 960 system for recovery of mycobacteria. J Clin Microbiol 1999; 37: 748-52.

12 Tortoli E, Cichero P, Piersimonia C, Simonetti TM, Gesu G, Nista D. Use of Bactec MGIT(TM)"' 960 for recovery of mycobacteria from clinical specimens: multicenter study. J Clin Microbiol 1999; 37: 3578-82.

13 Pfyffer GE, Welscher HM, Kissling P et al. Comparison of the Mycobateria Growth Indicator Tube (MGIT(TM)"') with radiometric and solid culture for recovery of acid-fast bacilli. J Clin Microbiol 1997; 35: 364-8

14 Pfyffer GE, Cieslak C, Welscher H, Kissling P, Rusch-Gerdes S. Rapid detection of mycobacteria in clinical specimens by using the automated Bactec 9000 MB system and comparison with radiometric and solid-culture systems. J Clin Microbiol 1997; 9: 2229-34.

15 Badak FZ, Kiska DL. Setterquist S, Hartley C, O'Connell MA, Hopfer RL. Comparison of Mycobacteria Growth Indicator Tube with Bactec 460 for the detection and recovery of mycobacteria from clinical specimens. J Clin Microbiol 1996; 34: 2236-9.

K. V. JAYAKUMAR, T. FORSTER and M. S. KYI

Department of Microbiology, Quest Diagnostics, Heston, Middlesex TWS 9QA, UK

(Accepted 27 January 2001)

Correspondence to: Mr K. V. Jayakumar, Microbiology Department, West Park Hospital, Horton Lane, Epsom, Surrey KT 19 8PB, UK.

Copyright Royal Society of Medicine Press Ltd. 2001
Provided by ProQuest Information and Learning Company. All rights Reserved