A. Bonington et al., USE OF ROCHE AMPLICOR MYCOBACTERIUM-TUBERCULOSIS PCR IN EARLY DIAGNOSIS OF TUBERCULOUS MENINGITIS, Journal of clinical microbiology, 36(5), 1998, pp. 1251-1254
Several nucleic acid-based amplification tests are available for the d
etection of Mycobacterium tuberculosis, but few data are available on
their use in the diagnosis of tuberculous meningitis (TBM). We perform
ed a prospective study to assess the Roche AMPLICOR Mycobacterium tube
rculosis PCR test (TB AMPLICOR) for use in the diagnosis of TBM and co
mpared it with direct Ziehl-Neelsen staining of smears, radiometric cu
lture for M. tuberculosis, and clinical and cerebrospinal fluid (CSF)
findings. Eighty-three CSF specimens collected from 69 patients with s
uspected meningitis in South Africa were tested by TB AMPLICOR. On the
basis of clinical and laboratory findings, 40 of these patients were
treated for TBM and 29 patients were not treated for TBM. Ten CSF samp
les from 10 patients were positive by TB AMPLICOR. Seven of these 10 p
atients were classified as having definite TBM, 2 were classified as h
aving probable TBM, and 1 was classified as having possible TBM. The s
ensitivity of TB AMPLICOR for detecting cases of definite and probable
TBM in patients from whom CSF specimens had been collected less than
10 days into antituberculosis treatment was 60.0%. Specimens from all
29 patients not treated for TBM were negative by the TB AMPLICOR, givi
ng a 100% specificity, TB AMPLICOR is therefore more sensitive than th
e combination of Ziehl-Neelsen staining of smears and radiometric cult
ure for M. tuberculosis and is a rapid and highly specific diagnostic
test for TBM.