Prior treatment with non anti-TB antibiotics, and the duration of symptoms have no effect on diagnostics of tuberculous meningitis

Empirical treatment with non-anti-tuberculosis antibiotics decreased microbiological detection in cervical tuberculous lymphadenitis

Diagnosis of cervical tuberculous lymphadenitis (CTL), the most commonly occurring form of extrapulmonary tuberculosis, remains as a challenge in clinic. Detection of the presence of Mycobacterium tuberculosis (Mtb) in fine needle aspiration cytology (FNAC) samples is one golden criterion to confirm the CTL diagnosis. Due to the non-specific clinical presentation, CTL might be confused with other lymph node enlargement diseases; therefore empirical treatment with non-anti-TB antibiotics is often initially administered. However, it is still unclear whether this diagnostic antibiotic treatment affects the positivity of Mtb detection in FNAC. The demographics and clinical characteristics of 732 lymph node enlargement patients who had underwent FNAC were retrospectively analyzed and 605 (82.65%) of them were diagnosed as CTL. A total of 279 CTL cases (279/605, 46.11%) with completion of three Mtb tests (AFB, NAAT, and Mtb culture) in FNAC samples were selected for analyzing the effect of empirical antibiotic treatment on the positivity of Mtb tests. Compared to CTL patients without antibiotic treatment prior to FNAC, patients received empirical non anti-TB treatment had significantly lower positivity for acid fast bacilli staining (adjusted OR 0.11, 95% CI 0.06-0.21), nucleic acid amplification test (NAAT) (adjusted OR 0.38, 95% CI 0.21-0.71), and Mtb culture (adjusted OR 0.11, 95% CI 0.06-0.19). In conclusion, this study demonstrated that empirical non anti-TB antibiotic treatment reduced the opportunity to confirm CTL by microbiological analysis. Patients with cervical lymph node enlargement should undergo FNAC for Mtb tests prior to initiation of empirical non anti-TB treatment.

Diagnostic scoring system for tuberculous meningitis among adult patients with non-suppurative and non-bacterial meningitis

Tuberculous meningitis (TBM) is the most severe form of extra-pulmonary tuberculosis. The definite diagnosis of this disease is difficult and can result in delayed treatment. Conventional culture yields low sensitivity while high-sensitivity diagnostic techniques are costly and unpractical. Adenosine deaminase (ADA) is used to diagnose several settings of extra-pulmonary tuberculosis but it is limited in TBM especially among HIV-infected patients. We retrospectively reviewed the data of patients with non-suppurative meningitis and compared the patient data with TBM and other causes including carcinomatous, lymphomatous, lymphocytic and fungal meningitis. We found that HIV infection, diabetes mellitus, duration of symptoms <14 days, radiologic findings of hydrocephalus, and CSF ADA level >10 IU were associated with TBM. The scoring system based on these parameters and their coefficients in the final model achieved an area under the receiver operating characteristic curve of 0.95,625. The indices were HIV infection = 5, diabetes mellitus = 3, duration of symptoms <14 days = 5, hydrocephalus = 4, and ADA in CSF >10 IU = 5. Based on the assumed costs of the patients with false negative and false positive, an appropriate cut off value of 10 was selected and the sensitivity was 92% and specificity was 89%.