College of American Pathologists
Printable Version



Testing for tuberculosis

Back to Cover Story

January 2000

Amplification assays are used primarily to identify viral pathogens, but they are also applied to a few slow-growing or difficult-to-grow bacteria. Chief among these is Mycobacterium tuberculosis.

Two tests are available to address this area, both of which are done directly on respiratory specimens, according to Gail Woods, MD, director of the Division of Microbiology, Department of Pathology, University of Texas Medical Branch, Galveston. One is a Roche PCR assay, approved only for smear-positive respiratory specimens, and the other is the GenProbe transcription-mediated amplification kit, which is approved for smear-positive and smear-negative samples.

In smear-positive cases, both assays approach 100 percent sensitivity. For smear-negative samples, sensitivity is lower and varies from site to site. Dr. Woods has reported 83 percent sensitivity with the GenProbe enhanced assay using growth in culture as the gold standard. Yield with the Roche assay is lower.

"False-positive results do occur," Dr. Woods said. "But in a majority of cases, it isn’t a fault of the assay but of the people doing the test." The test should be employed by skilled, experienced technologists who take precautions against contamination.

You really want to use the amplification test on "a person who has signs and symptoms of TB," Dr. Woods said. Even for standard mycobacterial testing of a sputum specimen from an inpatient, the University of Texas requires that the patient be in airborne precautions. Clinicians need to understand that the test should only be used for outpatients when they have a moderate to strong suspicion of tuberculosis, Dr. Woods emphasized.

Dr. Woods’ laboratory and two other sites are now participating in a project to evaluate the utility and cost-effectiveness of nucleic acid amplification tests. The project is sponsored by the Centers for Disease Control and Prevention. In the current phase, the GenProbe assay for Mtb is being studied. Patients are randomized to get the amplification test or not to get it, regardless of their smear result. "I think the benefit is there," Dr. Woods said, "but we will see the evidence."

William Check, PhD