Methods: A study of adult patients with suspected meningitis whose cerebrospinal fluid specimens were tested by PCR-TB (Roche AMPLICORE) and culture for M.tuberculosis were retrospectively reviewed. The sensitivity and specificity of combined Thwaites’ diagnostic index and PCR based assay in diagnosis of TBM were evaluated as compared to culture (definite TB) and clinical diagnosis using predefined criteria (probable TB). Drug-resistance profiles and treatment outcome of TB cases were also recorded.
Results: From 2008 to 2010, 160 adult hospitalized patients with suspected meningitis were reviewed. Of which, 29 were classified as definite TBM, 15 were classified as probable TBM and 116 were classified as non-TB. Nineteen of 44 (43.2%) patients were HIV seropositive. Thwaites’ diagnostic index had a high sensitivity (92%) but very low specificity (6%) in diagnosis of TBM among patients suspected of meningitis. In contrast, PCR-based assay gave a low sensitivity (18%) but high specificity (100%). A sequential testing of Thwaites’ diagnostic index and PCR-based assay had net sensitivity of 17.3% and net specificity of 100%. Forty-two patients (95.5%) were treated with anti-tuberculosis drugs, 33 of which (75%) received corticosteroid treatment. Of 29 culture-proven cases, 8 (27.6%) isolates were resistant to at least one first-line agents: 5 were mono-resistant strain and 3 were poly-resistant or multidrug-resistant TB. Overall survival to hospital discharge of patients with TBM was 66%.
Conclusion: Given a high sensitivity of Thwaites’ diagnostic index, it can be used to rule out tuberculous meningitis while PCR-based assay should be used to confirm diagnosis. Owing to emerging problem of drug-resistant tuberculous meningitis, molecular genetic assay for identification of resistance to standard anti-TB drugs is greatly needed.
V. Srinontprasert, None
S. Foongladda, None