Clinical Performance of the Amplified Mycobacterium tuberculosis Direct Test for the Detection of Mycobacterium tuberculosis in Non-respiratory Specimens

Kim, Sung Ryul; Shin, Jeong Hwan; Jeong, Joseph; Lee, Seon Ho; Chang, Chul Hun; Son, Han Chul

Abstract

BACKGROUND
The Amplified Mycobacterium tuberculosis Direct Test (AMTDT) has been developed for the direct detection of M. tuberculosis complex in respiratory specimens. Traditional methods for diagnosis of extrapulmonary tuberculosis such as the acid-fast bacilli (AFB) stain have their well-known limitations. We investigated the usefulness of the AMTDT for a wide range of non-respiratory specimens to establish early diagnosis of extrapulmonary tuberculosis.
METHODS
346 specimens (219 urine, 117 pleural fluid, 6 ascitic fluid, 2 lymph node, 1 gastric aspirate, and 1 pus specimens) from 340 patients referred from November 1997 to September 1998 were tested by the AMTDT. The AMTDT results were evaluated by comparing with clinical diagnosis and smear results.
RESULTS
The overall sensitivity, specificity, and positive and negative predictive values of the AMTDT were 82.9%, 93.8%, 64.2%, and 97.6%, respectively. There were no difference in sensitivity and specificity between pleural fluid and urine specimens. In 31 specimens from tuberculosis patients concurrently tested with AMTDT and stain, 15 were only AMTDT positive and 4 were only stain positive. Among the results considered to be false positive, 47.2% of cases were shown as being less than 150,000 relative light units (RLU). In 30 specimens from tuberculosis patients during or after treatment, all six of the patients with reactivation or aggravation were AMTDT positive, and one case was considered to be false positive.
CONCLUSIONS
Our study demonstrates the efficacy of the AMTDT in diagnosing extrapulmonary tuberculosis. Prudent interpretation of the AMTDT's results is recommended in case of that being less than 150,000 RLU.