1. Home>
2. Specialties>
3. HIV/AIDS>
4. Feature

Diagnosis and Detection of Drug-Resistant Strains of M. tuberculosis

Although tuberculosis (TB) continues to affect an estimated 10% to 50% of the world population, with three million tuberculosis-related deaths annually, the disease had been widely believed to be controlled in the United States. Improved nutrition, higher standards of living for the poor, and the post-World War II discoveries of antituberculosis antibiotics dramatically decreased the incidence of active disease. The rate of decline in incidence in the U.S. had been approximately 6% per year, reaching an all-time low of only 22,000 reported cases in 1984.^1,2 This decline in diagnosed disease was the foundation for the forecast that complete eradication of tuberculosis could be achieved in the United States by the end of the 20th century.

However, beginning in the 1960s, responsibility for managing remaining tuberculosis cases shifted from specialized hospital units to less-well-equipped outpatient settings. Medical school teaching of tuberculosis-related issues declined, and research funding for the disease was dramatically reduced.² In the 1980s, large-scale immigration from regions endemic for tuberculosis introduced new cases into the U.S., and the increase in homelessness and AIDS in American cities has created a dangerous and self-perpetuating pocket of disease in two populations that are particularly at risk. Finally, the ease with which the organism can spread within institutionalized populations has made outbreaks of tuberculosis common in prisons, nursing homes, and hospitals.

As a result of these many factors, a deadly resurgence of this ancient scourge is taking place, with an estimated 10 to 15 million Americans currently infected by tuberculosis.¹ Two recent studies from San Francisco and New York revealed startling findings about the nature of the latest epidemic.^3,4 Using DNA fingerprinting to evaluate the relative contribution of recent infection to the overall incidence of TB, these studies found that at least 33% of current cases can be attributed to new infection rather than reactivation of latent disease. It was also found that approximately 10% of patients are highly infectious, and these individuals may be responsible for the infection of a significant number of the new cases of TB.

This latest epidemic is particularly dangerous because of an associated rise in the number of drug-resistant strains. This is abetted by faltering patient compliance and lack of clinical follow-up. From a previous prevalence of only 3%, strains resistant to at least one drug now account for 14% of infections in the U.S., with at least one-quarter of this 14% resistant to both isoniazid and rifampin. In New York City, 33% of reported cases are resistant to at least one drug, and 19% to two drugs.² Furthermore, outbreaks in New York and Florida have shown additional resistance to ethambutol, streptomycin, ethionamide, kanamycin, and rifabutin.

Strains resistant to at least isoniazid and rifampin are considered multidrug resistant and pose a significant menace to treatment because alternative drug regimens achieve variable response rates and can induce undesirable side effects. Although multidrug-resistant strains are not intrinsically more virulent, patients with multidrug-resistant tuberculosis have a disproportionately high fatality rate of 40% to 89%.² AIDS patients and other severely immunocompromised individuals are the most susceptible to death related to multidrug resistance. The disease progresses rapidly in these patients and is often fatal before final laboratory evaluations are available.

The high risk of fatality in such cases, potential community exposure, and documented transmission to health care workers have made rapid diagnosis of tuberculosis and identification of multidrug-resistant tuberculosis strains a matter of national urgency.

Laboratory Diagnosis of Tuberculosis

Expedient diagnosis is one of the most vital factors in controlling the spread of tuberculosis, and one of the most challenging. The number of today's physicians who fail to do sputum smears and cultures for coughing patients without obvious tuberculosis risk factors suggests that health care training in the 1970s and 1980s did not adequately emphasize suspicion of tuberculosis infection. As tuberculosis has resurged and multidrug-resistant strains have evolved, clinical laboratories around the U.S. have attempted to discard slow, outdated methods of identifying Mycobacterium tuberculosis and have begun to require testing of all tuberculosis isolates for antibiotic resistance.

With traditional methods, identification of M. tuberculosis and determination of drug susceptibilities require two to four months. More sophisticated technologies combine the use of liquid media, automated radiometric detection methods for identification of positive cultures and nucleic acid probe testing for specific identification of organisms. With these methods, results are available within three to six weeks.

In some labs positive acid-fast staining results on specimens received during the week can be reported within 24 hours of receipt, and definitive identification of mycobacteria achieved from positive cultures within 7 to 14 days; drug susceptibility profiles on positive cultures can be generated in 5 to 12 days of a positive culture.

Specimen Collection

Recovery of M. tuberculosis from clinical specimens is optimized if the following guidelines are observed.

• Proper specimen containers -- Sputum specimens should be collected in a sterile, nonwaxed, leakproof container and be free of preservatives such as formalin.
  
• First morning specimen -- The specimen that yields the greatest sensitivity is the first sputum the patient produces upon awakening from the night's sleep, as organisms have accumulated in the sputum during rest.
  
• Induced sputum and biopsy -- If expectorated sputum fails to support a diagnosis of tuberculosis in the presence of high clinical suspicion, induced sputum and biopsy may be required. Again, formalin and other fixatives should not be used in the handling of these specimens.
  
• Immediate transport -- The specimen should be transported immediately to the laboratory, optimally within 60 minutes of collection. Specimens that cannot be transported immediately should be refrigerated at 48°C.
  
• Specimens other than sputum -- Blood cultures must be collected in special lysis-centrifugation Isolator tubes or in liquid media designed specifically for mycobacterial blood culture. Urine, cerebrospinal fluid (CSF), and tissue can also be cultured in the laboratory for mycobacteria when a diagnosis of renal, meningeal, or tissue tuberculosis is suspected. Waxed containers and swabs should never be used, since the bacteria will stick to waxy surfaces. Urine specimens should be first morning urine obtained on five to six consecutive mornings. Five to ten ml of CSF must be collected to maximize the likelihood of detection. Tissue specimens should not be collected in formalin or any other fixative.
  

Identification of M. tuberculosis

Because the incidence of tuberculosis infection is three times higher in laboratory personnel who work with mycobacteria than in the general population, all specimens should be regarded as potentially infectious and processed according to strict guidelines.^5,6

Direct microscopic examination (smear) is the initial step in the evaluation of a specimen for tuberculosis. For screening, the acid-fast fluorochrome dye auramine O is used rather than the conventional Ziehl-Neelsen stain, because its bright fluorescence under UV-microscopy allows for greater sensitivity and ease of detection at lower power (16X for auramine O compared with 100X for Ziehl-Neelsen). However, the Ziehl-Neelsen stain remains the stain of choice for examination of organisms recovered from culture; it is superior for visualizing organisms' morphology and therefore has greater specificity for M. tuberculosis.

Overall, the sensitivity of the direct acid-fast smear is lower than that of culture, as 5000 to 10,000 bacilli per ml of sputum are required for visualization by direct microscopy, compared with 10 to 100 bacilli per ml for generation of a positive culture.^5,6 A recent study using clinical criteria for tuberculosis diagnosis as the standard revealed the sensitivity of fluorochrome staining of three consecutively induced sputums to be only 53.7%, compared with 87.5% from culture results of the same three specimens.⁷ In the hands of a trained technologist, the specificity of both smear and culture is greater than 99%, although any acid-fast dye will also stain nontuberculous strains of mycobacteria.

Definitive differentiation among mycobacterial species can be achieved by culture followed by biochemical or nucleic acid probe testing. Two types of culturing methods can be used in parallel. The traditional method of growth on solid media (Lowenstein-Jensen slant) can require as long as 3 to 8 weeks to yield positive cultures. In many labs this is now used primarily as a backup. The preferred culture method is one using a liquid growth medium, such as the BACTEC automated radiometric system (Becton-Dickinson). The BACTEC has two advantages: growth in a liquid rather than a solid medium, for faster mycobacterial replication; and detection by a highly sensitive radiometric technique. Organisms are grown in liquid medium containing ¹⁴C-labelled palmitic acid as a substrate; they are periodically assessed for ¹⁴C-labelled carbon dioxide release, which is produced by the organisms from the palmitic acid substrate during rapid growth.

An optimum sequence of events is as follows:

• Positive cultures are first verified by Ziehl-Neelsen staining for typical acid-fast morphology.
  
• The mycobacterial species is then identified by nucleic acid probe testing. This is based on hybridization to species-specific DNA probes directed against rRNA genes (Gen-Probe/Accu-Probe). Species that can be identified by hybridization include: M. tuberculosis, M. avium intracellulare, M. kansasii, and M. gordonae. This process, in which chemiluminescent DNA probes for classification of organisms isolated from culture, yields 95% sensitivity, and >99% specificity; results can be obtained within 60 minutes.
  

Thus, with newly available methods, detection and positive identification of M. tuberculosis can be achieved in as little as 7 to 10 days.

Susceptibility Testing

In some protocols, all isolates identified as M. tuberculosis are tested for resistance to the five to ten most commonly used antituberculosis drugs (Table 1). One form of testing is the agar dilution method, in which organisms are grown on solid media containing antibiotic that inhibits the growth of 99% of nonresistant organisms. Optimally, about ten antibiotics are tested. However, this method is extremely laborious and results can require two to three months.

Because of the growth in demand for susceptibility testing for tuberculosis and the need for fast results, some labs determine resistance to five of the first-line drugs. Susceptibility testing is initiated as soon as sufficient growth is obtained by culture. Drug-susceptibility profiles can be determined within four to seven days of the time a culture becomes positive. In one procedure, organisms growing in culture are inoculated into new liquid broth containing ¹⁴C-labelled palmitic acid and a single antibiotic. Resistance of organisms to the antibiotic is inferrred if there is sustained production of ¹⁴C-labelled CO₂, as detected by the BACTEC radiometric detection system. A clinical isolate is considered resistant to an antibiotic if growth exceeds 1% of that found with the same isolate grown without the antibiotic.

The Future of Tuberculosis Testing

Although the BACTEC radiometric culture system, combined with DNA probe analysis, is now the most rapid and advanced method of testing, two to three weeks can still seem a long time. The organism is easily transmitted and difficult to eradicate, and increasing numbers of drug-resistant strains are being reported. Hence, there is now great interest in the development of laboratory assays that could be used directly on patient specimens and could yield definitive answers within hours.

Clinical laboratories may soon be able to perform direct DNA testing on submitted specimens using various amplification techniques currently awaiting FDA approval. One assay that has undergone extensive clinical evaluation is based on identification of M. tuberculosis by hybridization to rRNA genes. However, a second-generation test in development includes an amplification step prior to hybridization, in which mycobacterial rRNA is first converted to DNA by reverse transcriptase and then transcribed back to multiple copies of rRNA for detection.

Another assay awaiting FDA approval involves polymerase-chain-reaction (PCR) amplification of genes encoding species-specific rRNA. When used for direct testing of specimens, new methods involving amplification techniques show sensitivities of 96% to 100% and specificities of 99% to 100%. After positive identification of M. tuberculosis is made in this manner, drug-susceptibility testing can be performed on the remainder of the submitted specimen. Other strategies for determining drug resistance are currently under investigation. For example, in one innovative approach, Bloom and colleagues developed a highly sensitive method based on luminometry. This method allows for detection of as few as 100 viable resistant organisms within hours of the initiation of testing.⁸

If current clinical trials continue to generate reliable results, new methods for detection of mycobacteria and assessment of drug resistance could revolutionize the way M. tuberculosis infection is diagnosed in the laboratory. A slow and labor-intensive test could be converted into a rapid, highly sensitive assay that generates results within 24 hours of collection.

— MGH Clinical Microbiology Lab

This article was adapted with permission from "Turnaround Times," a publication of the Massachusetts General Hospital (MGH) Clinical Laboratories. Its authors are Jeannie T. Lee, MD, PhD; Wolfgang Klietmann, MD; and Mary Jane Ferraro, PhD, all of the MGH Department of Pathology.

Published in AIDS Clinical Care April 1, 1995

Citation(s):

1. National action plan to combat multidrug-resistant tuberculosis. MMWR Morb Mortal Wkly Rep 1992 Jun 19 41 RR-11 5-48.

2. Reichman LB. Multidrug-resistant tuberculosis: meeting the challenge. Hosp Pract 1994 May 15 29 85-96.

3. Small PM, et al. The epidemiology of tuberculosis in San Francisco. N Engl J Med 1994 Jun 16 330 1703-1709.

• Original article (Subscription may be required)
• Medline abstract (Free)

4. Alland D, et al. Transmission of tuberculosis in New York City: an analysis by DNA fingerprinting and conventional epidemiologic methods. N Engl J Med 1994 Jun 16 330 1710-1716.

• Original article (Subscription may be required)
• Medline abstract (Free)

5. Stratton CW. Tuberculosis, infection control, and the microbiology laboratory. Infect Control Hosp Epidemiol 1993 Aug 14 481-487.

• Medline abstract (Free)

6. Tenover FC, et al. The resurgence of tuberculosis: is your laboratory ready? J Clin Microbiol 1993 Apr 31 767-770.

• Medline abstract (Free)

7. Jonas V, et al. Detection and identification of Mycobacterium tuberculosis directly from sputum sediments by amplification of rRNA. J Clin Microbiol 1993 Sep 31 2410-2416.

• Medline abstract (Free)

8. Jacobs WR, et al. Rapid assessment of drug susceptibilities of Mycobacterium tuberculosis by means of luciferase reporter phages. Science 1993 May 7 260 819-822.

• Medline abstract (Free)