ABSTRACT
The resurgence of tuberculosis prompted the development of a number of new options for the rapid laboratory diagnosis of Mycobacterium tuberculosis (MTB). One of the most promising and exciting methodologies has been the introduction of assays employing amplification technology to detect MTB directly in clinical specimens. Although amplification assays hold significant promise to improve the laboratory diagnosis of tuberculosis, the decision to perform or not perform these assays is complicated. The performance of in-house polymerase chain reaction (PCR) assays and two commercially-prepared assays, GenProbe's AMTD test and Roche's AMPLICOR PCR assay are reviewed. Regardless of the amplification format, all assays have decreased sensitivity with specimens that are acidfast bacilli (AFB) stain-negative. Data from these studies and others indicate possible potential pitfalls of amplification assays, those being sampling errors, the presence of substances in clinical specimens that inhibit the amplification assay, and clinical utility.
