0
Clinical Investigations: TUBERCULOSIS |

Identification of Mycobacterium Species in Contaminated Cultures by Polymerase Chain Reaction* FREE TO VIEW

Danielle Malta Lima, MSc; Valdes Roberto Bollela, MD, PhD; Beatriz Junqueira Tavares Jácomo, BS; Roberto Martinez, MD, PhD; Benedito Antônio Lopes da Fonseca, MD, PhD, MPH
Author and Funding Information

*From the Department of Internal Medicine, School of Medicine of Ribeirão Preto, University of São Paulo, Brazil.

Correspondence to: Benedito Antônio Lopes da Fonseca, MD, PhD, MPH, Departamento de Clínica Médica, Faculdade de Medicina de Ribeirão Preto, Universidade de São Paulo, Avenida dos Bandeirantes, 3900, Ribeirão Preto, São Paulo, CEP 14049–900, Brazil; e-mail: baldfons@fmrp.usp.br



Chest. 2005;127(4):1283-1288. doi:10.1378/chest.127.4.1283
Text Size: A A A
Published online

Study objectives: The detection of Mycobacterium sp on a culture remains the “gold standard” technique for the diagnosis of mycobacterial infections. A small percentage of these cultures, however, may be contaminated by other nonfastidious microorganisms, making accurate diagnosis difficult. We evaluated the use of a polymerase chain reaction (PCR) protocol that was specific for the genus Mycobacterium, and specifically for Mycobacterium tuberculosis, Mycobacterium avium, and Mycobacterium intracellulare, for the identification of Mycobacterium sp growing on contaminated cultures.

Design: This prospective study was designed to identify Mycobacterium sp growing on mycobacterial cultures contaminated with other microorganisms.

Samples and patients: Twenty-six samples, taken from 23 patients with probable mycobacterial disease, that resulted in Mycobacterium growth but were contaminated during their processing were evaluated in this study. Clinical data and the clinical status of each patient were used to ascertain the final diagnosis.

Results: All samples studied here exhibited Mycobacterium growth on solid media but were contaminated by nonfastidious bacteria, compromising the biochemical identification of the Mycobacterium sp. PCR correctly identified the genus Mycobacterium in all samples. M tuberculosis was identified in 14 samples, and M avium in 10 samples. No amplification of M intracellulare was obtained, and in two samples there was amplification only for the genus Mycobacterium. In the cultures of those patients in whom a mycobacterial infection was evident, PCR identified M avium and M tuberculosis in samples from 6 and 12 patients, respectively. However, PCR identified M avium (two patients) and M tuberculosis (two patients) in the cultures of four patients for whom a mycobacterial disease could not be confirmed by our case definition. Finally, in two samples from one patient only the genus Mycobacterium was amplified by PCR.

Conclusion: PCR, with its advantages of greater speed and effectiveness than conventional detection methods, was successfully used to identify the Mycobacterium sp growing on contaminated cultures.

Tuberculosis (TB) remains a major cause of morbidity and mortality worldwide. The rapid and accurate detection of Mycobacterium tuberculosis is of paramount importance in the effective treatment and control of TB.1

Due to HIV infection, other species of Mycobacterium, such as Mycobacterium avium and Mycobacterium intracellulare, have become important in clinical practice. M avium and M intracellulare, which are frequently identified as the M avium complex (MAC), have overlapping phenotypic properties that make their speciation difficult to determine by conventional procedures. Although MAC strains can occasionally cause infections in the general population, the frequency of MAC infection among AIDS patients may be as high as 30 to 80%.2

The diagnosis of M tuberculosis-induced disease is most commonly made by direct sample examination and culture. Acid-fast bacilli (AFB) detection has low sensitivity and specificity, and, at best, can only provide a preliminary diagnosis. Culture is still the “gold standard” for the diagnosis of TB, but, in addition to the long time necessary for M tuberculosis growth, nonfastidious microorganisms will eventually contaminate a small percentage of cultures. Therefore, there is an urgent need for a rapid, safe, and reliable method to establish the diagnosis of TB.3The potential advantages of using molecular biology techniques for the diagnosis of a number of diseases have been widely discussed lately. In the field of infectious disease, for example, these approaches have been successfully applied for the detection of a variety of infectious agents.4 The most promising technique for approaching this diagnostic dilemma is polymerase chain reaction (PCR). PCR has been used to amplify different regions of the mycobacterial genome, making it a good candidate for assisting with species identification in a variety of specimens. Indeed, several research groups have described different PCR protocols for M tuberculosis genome amplification and/or performed clinical studies based on PCR detection of this agent. Based on these results, the US Food and Drug Administration has approved the use of PCR as an aide for the diagnosis of TB in clinical samples. Although these studies agree that PCR may improve the diagnosis of TB, they report variable specificity and sensitivity.,3 Because the sensitivity and specificity of PCR remain uncertain, PCR cannot be fully recommended for the diagnosis for M tuberculosis.,5

Consequently, we still have to rely on culture results to make the correct diagnosis of a mycobacterial infection. However, about 5% of the broth cultures for M tuberculosis are contaminated by nonfastidious bacterial species, making it difficult to confirm the diagnosis. Many of these contaminated cultures require decontamination and reincubation before appropriate identification, significantly increasing the time required to obtain culture results.1 Due to its ability to selectively differentiate mycobacteria from other bacteria, PCR may be useful in the accurate and timely identification of M tuberculosis. To date, however, there have been few studies,6 that have used PCR for the identification of mycobacteria in contaminated cultures.

Thus, we hereby describe the use of PCR for the identification of Mycobacterium sp in cultures that have been contaminated with nonfastidious microorganisms, and our results showed that we were able to achieve the correct species identification in the majority of the samples.

Study Design

The samples evaluated in our work were all of the samples that were contaminated by nonfastidious bacteria or fungi during a 7-month period of study. They were obtained from patients presenting with symptoms compatible with a mycobacterial infection and were inoculated in Löweinstein-Jensen (L-J) medium as soon they reached the microbiology laboratory. After the detection of M tuberculosis growth and contamination by other microorganisms (bacteria or fungi), the samples were sent to our laboratory for examination, using PCR for the identification of mycobacterial species.

DNA Extraction

A small portion of each contaminated mycobacterial culture was suspended in 500 μL of phosphate-buffered saline (PBS) solution and boiled at 100°C for 10 min. After boiling, the sample was cooled on ice and centrifuged at 13,500 revolutions per minute for 5 min.

PCR
M tuberculosis:

The PCR protocol that was used to detect the M tuberculosis has been described previously.78 PBS and extracted DNA from M tuberculosis were used as negative and positive controls, respectively. Amplicons were detected by agarose gel electrophoresis and staining with ethidium bromide, with subsequent visualization under ultraviolet light.

Genus Mycobacterium, M avium, and M intracellulare:

PCR was done for M tuberculosis using 50 pmol/L each primer, with each reaction specific for the genus Mycobacterium, and specifically for M avium and M intracellulare.2 PCR was conducted in separate tubes according to the following protocol: 94°C for 1 min; 55°C for 2 min; and 72°C for 6 min. These cycles were repeated 40 times with a final step at 72°C for 10 min, and then the samples were stored at 4°C. The specific amplicon for the genus Mycobacterium had 1,030 base-pairs, and those for M avium and M intracellulare had 1,279 base-pairs. These amplicons were separated by a 2% agarose gel electrophoresis and were detected as described above. PBS and extracted DNA from M avium and M intracellulare isolates were used as negative and positive controls, respectively.

Diagnostic Confirmation

In order to define whether a patient had a mycobacterial disease, the charts of all patients whose samples had been contaminated during mycobacterial cultivation were reviewed. The record review took into consideration the period spanning from the first medical appointment to 12 months after conducting the PCR protocol on the original sample. The diagnosis of TB was confirmed when either AFB or M tuberculosis growth was detected in another sample from the same patient, or when the patient showed a positive response to standard TB therapy. The diagnosis of MAC disease was confirmed when other samples from the same patient showed evidence of MAC growth or the patient had a positive response to clarithromycin and ethambutol, usually after the failure of TB therapy.

Patients

The ages of the 23 patients ranged from 6 months to 67 years. Table 1 displays the clinical characteristics, PCR results, and the definitive diagnosis for each patient. Thirteen of the 23 patients were infected with HIV, 11 of them with CD4+ counts < 200 cells/μL. All of the patients in PCR-confirmed TB cases (12 patients) who received specific treatment achieved remission of their disease. All patients were adults, with the exception of a 6-month-old child who was born to an HIV-infected mother with TB but was not infected with HIV. Two samples collected from a single patient were PCR-positive for mycobacteria but were negative for all three Mycobacterium sp tested. This patient was an HIV-positive man who had been previously diagnosed with and treated for pulmonary TB in 1999, with reported clinical and microbiological remission. The radiograph and the CT scan of this patient, in the period of this study, showed a residual cavitation in the upper left lobe of the lung with the presence of a fungus ball. This patient was not treated for TB and is currently doing well. Of the six patients with a confirmed diagnosis of MAC disease, five had AIDS, all with < 50 CD4+ cells/μL. The sixth patient was a 67-year-old HIV-negative patient presenting with a chronic pulmonary disease, and liver and spleen enlargement for 11 months at the time of diagnosis. He had been treated for TB without response during the investigation of his disease. The contaminated L-J slope, on which the PCR was performed, had been collected 7 months after the beginning of his disease investigation. Despite the positive PCR result, the patient was not treated for MAC disease until there was a mycobacterial growth in a subsequent sputum sample. The patient was then treated with clarithromycin and ethambutol for 10 months, and achieved clinical and microbiological remission. A mycobacterial infection was not confirmed in the other four patients included in this study.

Contaminated Cultures

According to our data, these contaminated samples represent 6% of all cultures (26 of 433 cultures) yielding mycobacterial growth during the study period. In this study, 26 samples were assayed by PCR to identify the most clinically important mycobacteria species. These samples were collected from 23 patients who had been admitted to our hospital during the study period and presented with clinical findings of a mycobacterial disease. Three patients had two samples each. In each one of the 26 samples, a Mycobacterium sp was successfully cultivated in L-J medium, but each sample was also contaminated by nonfastidious microorganisms. Of the 26 studied samples, 16 consisted of sputum, 3 of gastric lavage fluid, 2 of BAL fluid, 2 of lymph node biopsy specimens, 1 of skin biopsy specimen, 1 of feces, and 1 of abscess secretion (ie, pus).

PCR

All 26 samples showed PCR amplification for the genus Mycobacterium. Fourteen samples were identified as M tuberculosis and 10 samples as M avium, and there was no amplification of M intracellulare. There were two samples with amplification only of the genus Mycobacterium, probably representing other mycobacterial species not evaluated in this study.

Clinical Diagnosis

A medical record review revealed mycobacterial infection in 19 patients, 12 with TB, 6 with MAC disease, and 1 with an unidentified Mycobacterium sp. The clinical diagnosis could not be confirmed in four patients. All of the 12 patients with a definitive diagnosis of TB had a positive PCR for M tuberculosis. The six patients with a definitive diagnosis of MAC infection had a positive PCR for M avium. Of the four patients without an established mycobacterial disease, two had a positive PCR for M tuberculosis, and two had a positive PCR for M avium. Finally, in two samples from a single patient (Table 1), PCR was only positive for the genus Mycobacterium and did not identify the mycobacterial species.

The analysis of the patients’ charts showed that the final diagnosis of those four patients in whom mycobacterial disease was not confirmed were a bacterial pneumonia that responded to clarithromycin, Staphylococcus aureus sepsis, bronchiolitis obliterans-organizing pneumonia (BOOP), and cutaneous histoplasmosis.

TB is one of the major threats to public health worldwide, affecting more that one third of the world’s population. In the last few decades, TB has been one of the main causes of death in the world, being responsible for almost 3 million deaths annually.9Experts in the field suggest that TB control depends on rapid diagnosis and effective treatment. The emergence of drug-resistant strains in recent years has increased concern about the control of this disease. Moreover, given the increasing number of patients infected with HIV or undergoing immunosuppressive treatment, M tuberculosis is not the only Mycobacterium sp capable of causing human disease. Indeed, several nontuberculous mycobacteria, primarily M avium, M intracellulare, and Mycobacterium kansasii can cause disease in immunosuppressed patients. Clinical presentation and history can help in making a presumptive diagnosis. Definitive diagnosis, however, requires a species identification of the organism. This is particularly important because medical treatment regimens and isolation needs differ for the various mycobacterial species.10

Consequently, there is a pressing need for a rapid and accessible diagnostic method that could be applied either directly on clinical samples or at an early stage of a Mycobacterium culture in order to identify the many species of mycobacteria isolated in the microbiology laboratory. In addition, the accurate identification of mycobacteria in cultures may be further complicated by contamination with nonfastidious bacteria.6 Contamination of mycobacterial cultures may lead to errors in the diagnosis and delay in the administration of appropriate treatment. We hypothesized that PCR could be used in this situation, leading to a rapid identification of the mycobacteria and to more effective treatment.

As cited by Allen and Swaffield,11 as early as 1958 Marshall et al described culture contamination when they examined an L-J slope that showed a color change from pale-green to blue-green when subcultured onto blood agar. The change of color was found to be due to acid production by contaminating Gram-positive cocci. Although no attempt was made to identify these organisms in detail, they were found to have characteristics typical of streptococci.

Allen and Swaffield11 approached culture contamination by subculturing contaminated cultures on selective oleic-acid albumin agar culture slopes. They reported the successful recovery of M tuberculosis in 93.5% of their samples.

Cormican et al6 evaluated the potential of PCR for the identification of mycobacterial species directly from mycobacteria grown on a noncontaminated medium. Their culture results indicated that species identification by this method corresponded to that achieved by conventional methods in the majority of the cases.

Trying to solve the contamination problem that occasionally occurs in our microbiology laboratory, we used a PCR protocol to identify the mycobacteria growing on contaminated cultures. Using this technology, we were able to detect the growing Mycobacterium in 100% of the contaminated cultures. Furthermore, we were able to define the growing Mycobacterium sp in approximately 96% of the patients. The results of this study show that PCR is more sensitive than the method of Allen and Swaffield11 and has the additional advantage of being able to reach the diagnosis in a shorter period of time.

In our study, the presence of a Mycobacterium was detected in two contaminated sputum samples from one patient, but PCR did not identify this mycobacterium as any of the three tested species. They were from a young man with AIDS and a history of treated TB, which had achieved clinical and microbiological cure. At the time of the study, the patient was being treated for pulmonary aspergillosis in a residual pulmonary cavitation. This patient is currently well and did not receive any new treatment for TB. The amplified Mycobacterium probably belonged to a different species than those tested for here.

Moreover, we observed that there was an agreement between PCR identification of the mycobacteria growing on contaminated cultures and the final diagnosis. However, of the 23 patients enrolled in the study, a mycobacterial infection was not confirmed in four cases. This unexpected finding may be explained by the presence of “Mycobacterium colonization” in these patients, or, more likely, it is the evidence for cross-contamination in the microbiology laboratory.

In two patients with probable M avium disease, this diagnosis could not be confirmed. One patient was HIV-infected and had received a diagnosis of cutaneous histoplasmosis by histopathology of the skin. Unfortunately, this patient was lost to follow-up. The other patient was a 2-year-old child with achondroplasia and BOOP in long-term use of corticosteroids. There was no clinical evidence of M avium disease in this case, but harmless colonization of the respiratory tract by mycobacteria other than TB has already been demonstrated.12

In another two patients, M tuberculosis was identified in one sputum sample and in one gastric lavage fluid specimen, but the diagnosis of TB was not confirmed. Severe community-acquired pneumonia was diagnosed in one patient and was treated with clarithromycin, but the patient was lost to follow-up after being discharged from the hospital. The other patient received a diagnosis of S aureus septicemia and recovered totally after specific treatment with oxacillin. These cases were considered to be false-positive TB cases, most probably due to cross-contamination in the laboratory during specimen manipulation. Considering this hypothesis, we were astonished by the possible high rates of cross-contamination in our microbiology laboratory, and then we decided to examine these samples using molecular methods in order to investigate whether they had resulted from a common source of contamination. Unfortunately, by the time that we went back to the microbiology laboratory to ask for these samples, they had been discarded, making it impossible to prove our assumption. These results were unwanted but not totally unexpected since the same findings have been published by other authors. Burman and Reves13 reviewed reports of false-positive cultures for M tuberculosis and found that in 13 of the 14 studies (93%) that were evaluated, > 100 patients showed some degree of false-positive results. The mean false-positive rate was 3.1%, ranging from 2.2 to 10.5%. A total of 236 false-positive TB cases were identified, and 158 (67%) were treated, resulting in adverse effects from therapy, and unnecessary hospitalizations, blood tests, and contact investigations in a number of cases. Thus, the contamination of mycobacterial cultures has a tremendous impact on the safety and cost of misdiagnosed patients. Based on the findings from our study, we suggest that PCR may improve the identification of mycobacteria species even in the context of contaminated cultures. PCR has the additional advantages of being faster, easier to perform, and more effective than conventional methods for confirming the diagnosis of TB and other disease-causing mycobacteria.

Abbreviations: AFB = acid-fast bacilli; BOOP = bronchiolitis obliterans-organizing pneumonia; L-J = Löweinstein-Jensen; MAC = Mycobacterium avium complex; PBS = phosphate-buffered saline; PCR = polymerase chain reaction; TB = tuberculosis

Table Graphic Jump Location
Table 1. Clinical Characteristics, PCR Results, and the Final Diagnosis of the Patients Participating in the Study*
* 

The definitive diagnoses based on chart analysis and the Mycobacterium sp identification by PCR are shown. Note that most of the samples where a Mycobacterium sp was identified by PCR correlated well with the final diagnosis. The Mycobacterium sp detected on the samples from patient 14 probably is species not tested in this study. HAART = highly active antiretroviral therapy; CSF = cerebrospinal fluid; + = positive; − = negative; ID = identification; F = female; M = male.

The authors are indebted to Dr. Larry Scahill for his careful reading of the manuscript.

Zheng, X, Pang, M, Engler, HD, et al (2001) Rapid detection ofMycobacterium tuberculosisin contaminated BACTEC 12B broth cultures by testing with amplifiedMycobacterium tuberculosisdirect test.J Clin Microbiol39,3718-3720. [CrossRef] [PubMed]
 
Chen, ZH, Butler, WR, Baumstark, BR, et al Identification and differentiation ofMycobacterium aviumandM. intracelullareby PCR.J Clin Microbiol1996;34,1267-1269. [PubMed]
 
Beige, J, Lokies, J, Schaberg, T, et al Clinical evaluation of aMycobacterium tuberculosisPCR assay.J Clin Microbiol1995;33,90-95. [PubMed]
 
Santis, G, Legrand, V, Hong, SS, et al Molecular biology for the critical care physician: Part II. Where are we now?Crit Care Med1999;27,825-831. [CrossRef] [PubMed]
 
Tony, JC Impact of basic research on tomorrow’s medicine: applications and limitations of polymerase chain reaction amplification.Chest1995;108,1393-1404. [CrossRef] [PubMed]
 
Cormican, MG, Glennon, M, Ni Riain, U, et al Evalution of a PCR based method for identification of mycobacterial isolates.Diagn Microbiol Infect Dis1995;22,20-23
 
Eisenach, KD, Cave, MD, Bates, JT Polymerase chain reaction amplification of a repetitive DNA sequence specific forMycobacterium tuberculosis.J Infect Dis1990;160,977-981
 
Lima, DM, Colares, JK, da Fonseca, BA Combined use of the polymerase chain reaction and detection of adenosine deaminase activity on pleural fluid improves the rate of diagnosis of pleural tuberculosis.Chest2003;124,909-914. [CrossRef] [PubMed]
 
Bloom, BR, Murray, CJL Tuberculosis: commentary on a reemergent killer.Science1992;257,1055-1064. [CrossRef] [PubMed]
 
Kaul, KL Molecular detection ofMycobacterium tuberculosis: impact on patient care.Clin Chem2001;47,1553-1558. [PubMed]
 
Allen, BW, Swaffield, JE Mycobacterium tuberculosis: recovery from contaminated culture media and identification of bacteria responsible for contamination.Med Lab Sci1982;39,11-13. [PubMed]
 
Katila, ML, Katila, P, Erkinjuntti-Pekkanen, R Accelerated detection and identification of mycobacteria with MGIT 960 and COBAS AMPLICOR systems.J Clin Microbiol2000;38,960-964. [PubMed]
 
Burman, WJ, Reves, RR Review of false-positive cultures ofMycobacterium tuberculosisand recommendations for avoiding unnecessary treatment.Clin Infect Dis2000;31,1390-1395. [CrossRef] [PubMed]
 

Figures

Tables

Table Graphic Jump Location
Table 1. Clinical Characteristics, PCR Results, and the Final Diagnosis of the Patients Participating in the Study*
* 

The definitive diagnoses based on chart analysis and the Mycobacterium sp identification by PCR are shown. Note that most of the samples where a Mycobacterium sp was identified by PCR correlated well with the final diagnosis. The Mycobacterium sp detected on the samples from patient 14 probably is species not tested in this study. HAART = highly active antiretroviral therapy; CSF = cerebrospinal fluid; + = positive; − = negative; ID = identification; F = female; M = male.

References

Zheng, X, Pang, M, Engler, HD, et al (2001) Rapid detection ofMycobacterium tuberculosisin contaminated BACTEC 12B broth cultures by testing with amplifiedMycobacterium tuberculosisdirect test.J Clin Microbiol39,3718-3720. [CrossRef] [PubMed]
 
Chen, ZH, Butler, WR, Baumstark, BR, et al Identification and differentiation ofMycobacterium aviumandM. intracelullareby PCR.J Clin Microbiol1996;34,1267-1269. [PubMed]
 
Beige, J, Lokies, J, Schaberg, T, et al Clinical evaluation of aMycobacterium tuberculosisPCR assay.J Clin Microbiol1995;33,90-95. [PubMed]
 
Santis, G, Legrand, V, Hong, SS, et al Molecular biology for the critical care physician: Part II. Where are we now?Crit Care Med1999;27,825-831. [CrossRef] [PubMed]
 
Tony, JC Impact of basic research on tomorrow’s medicine: applications and limitations of polymerase chain reaction amplification.Chest1995;108,1393-1404. [CrossRef] [PubMed]
 
Cormican, MG, Glennon, M, Ni Riain, U, et al Evalution of a PCR based method for identification of mycobacterial isolates.Diagn Microbiol Infect Dis1995;22,20-23
 
Eisenach, KD, Cave, MD, Bates, JT Polymerase chain reaction amplification of a repetitive DNA sequence specific forMycobacterium tuberculosis.J Infect Dis1990;160,977-981
 
Lima, DM, Colares, JK, da Fonseca, BA Combined use of the polymerase chain reaction and detection of adenosine deaminase activity on pleural fluid improves the rate of diagnosis of pleural tuberculosis.Chest2003;124,909-914. [CrossRef] [PubMed]
 
Bloom, BR, Murray, CJL Tuberculosis: commentary on a reemergent killer.Science1992;257,1055-1064. [CrossRef] [PubMed]
 
Kaul, KL Molecular detection ofMycobacterium tuberculosis: impact on patient care.Clin Chem2001;47,1553-1558. [PubMed]
 
Allen, BW, Swaffield, JE Mycobacterium tuberculosis: recovery from contaminated culture media and identification of bacteria responsible for contamination.Med Lab Sci1982;39,11-13. [PubMed]
 
Katila, ML, Katila, P, Erkinjuntti-Pekkanen, R Accelerated detection and identification of mycobacteria with MGIT 960 and COBAS AMPLICOR systems.J Clin Microbiol2000;38,960-964. [PubMed]
 
Burman, WJ, Reves, RR Review of false-positive cultures ofMycobacterium tuberculosisand recommendations for avoiding unnecessary treatment.Clin Infect Dis2000;31,1390-1395. [CrossRef] [PubMed]
 
NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

CHEST Journal Articles
CHEST Collections
PubMed Articles
Guidelines
  • CHEST Journal
    Print ISSN: 0012-3692
    Online ISSN: 1931-3543