Tj. Brown et al., Evaluation of three commercial detection systems for Mycobacterium tuberculosis where clinical diagnosis is difficult, J CLIN PATH, 52(3), 1999, pp. 193-197
Citations number
18
Categorie Soggetti
Research/Laboratory Medicine & Medical Tecnology","Medical Research Diagnosis & Treatment
Aims-To assess the performance of three commercially available Mycobacteriu
m tuberculosis detection systems employing nucleic acid amplification, when
applied directly to respiratory and nonrespiratory specimens from patients
where the diagnosis of tuberculosis is difficult using clinical and tradit
ional bacteriological methods.
Methods-42 respiratory Emd 21 nonrespiratory specimens were concentrated, e
xamined with auramine staining, and cultured on Lowenstein-Jensen slopes. T
hese specimens were also assayed using the Amplicor Mycobacterium tuberculo
sis test (AM) (Roche Diagnostic Systems), the Amplified Mycobacterium tuber
culosis direct test (AMD) (Gen-Probe), and the LCx Mycobacterium tuberculos
is assay (LMA) (Abbott Laboratories).
Results-All three amplification systems used in this study gave specificiti
es of 100% when used on respiratory specimens. When used ion non-respirator
y specimens, AM and LMA gave specificities of 100% and AMD 75%. With respir
atory specimens the AIM, AMD, and LMA systems gave sensitivities of 75%, 65
.2%, and 79.2%, respectively. With nonrespiratory specimens the sensitiviti
es were 50%, 66.7%, and 60%, while with smear negative, culture positive sp
ecimens they were 33.3%, 66.7%, and 55.6%. Positive predictive values of 10
0% were seen with all specimens except nonrespiratory specimens assayed usi
ng AMD where the value was 66.7%.
Conclusions-The manufacturers of these systems recommend that they should o
nly be used for the direct analysis of respiratory specimens, and the US Fo
od and Drug Administration has approved them for use only with smear positi
ve specimens. This study confirms that sensitivities are lower for non-resp
iratory and smear negative specimens, but positive predictive values sere h
igh. Provided they are interpreted with caution, positive results with thes
e tests in respiratory and non-respiratory specimens are useful in tubercul
ous patients who are otherwise difficult to diagnose. Each amplification ha
s advantages and disadvantages com; pared with the others.