Early and reliable diagnosis of tuberculous meningitis (TBM) still poses a great challenge. One of the underlying difficulties is due to the fact that tubercle bacilli are mainly not present in the cerebrospinal fluid (CSF) but in the phagocytotic macrophages. The present study was designed to demonstrate early secretory antigenic target 6 (ESAT-6), a mycobacterium-specific antigens, in the macrophages in infected CSF samples and compare the efficiency of this antigen in the laboratory diagnosis of TBM.

Tuberculosis is a disease that carries a global health burden, and tuberculous meningitis (TBM) is one of the most serious clinical manifestations of extrapulmonary tuberculosis. The failure to promptly recognize and treat patients with TBM enables the disease to exact a devastating toll around the globe despite the availability of effective chemotherapy. Therefore, accurate and timely identification of infected individuals and subsequent prompt initiation of effective anti-tuberculosis chemotherapy are essential in reducing the mortality of TBM.1 However, the disease is often difficult to diagnose with certainty, especially at early stages,2 as the presenting clinical and cerebrospinal fluid (CSF) features of TBM are non-specific and have to be differentiated from a plethora of other infectious meningitides such as viral, bacterial or cryptococcal meningitis, which frequently result in diagnostic confusion. Consequently, TBM is often diagnosed when irreversible neurologic damages have already taken place and immediate anti-tuberculosis therapy is therefore recommended regardless of the results of individual tests if TBM is seriously contemplated.

The confirmation of TBM depends on the demonstration of Mycobacterium tuberculosis in the CSF by bacteriological methods. Smear acid-fast bacillus microscopy is rapid and inexpensive, but has a very low sensitivity (10-20%),3 and culture also lacks sensitivity and is time-consuming. A number of strategies have been attempted to improve the laboratory diagnosis of TBM. Nucleic acid amplification tests (NAAT) show potential roles in confirming the diagnosis of TBM, but they, too, suffer from the problem of overall low sensitivity (0.46-0.66), partly due to a low bacillary load in the CSF.4-5 Immunodiagnostic methods have also been evaluated, but the heterogeneity of immune responses in TBM patients at the different stages of the infection poses a major obstacle to the detection of mycobacterial antibodies in CSF samples.6-7 In general, patients at the chronic stages of TBM have a myriad of antibody responses to all major antigens of Mycobacterium tuberculosis, while patients at the early stages have scarcely any detectable antibody response. In addition, interpretation of mycobacterial antibodies in the CSF must take into account the contribution of antibodies from the plasma, therefore, jeopardizing the use of these antibody-based immunodiagnostic methods in the location where they are most needed.7 One approach to provide direct evidence of existing infection is the detection of the presence of specific antigens in the circulating CSF. The first generation of these tests employed non-specific antigens as exemplified by bacille Calmette-Guerin (BCG) or purified protein derivative (PPD).8-10 Genomic analysis and antigen mining of Mycobacterium tuberculosis have yielded novel, more specific antigens, such as early secretory antigenic target 6 (ESAT-6), 38-kD antigen, and Ag85 complex.11 Nevertheless, the diagnostic efficiency of this antigen-based approach is still unsatisfactory.5-6 Mycobacterium tuberculosis belongs to the group of intracellular bacteria, which replicate within resting macrophages. During the early stages of the central nervous system (CNS) infection, the tubercle bacilli in the CSF are immediately phagocytosed by the macrophages,12-13 leading to the scarcity of mycobacterial markers in the circulating CSF. Thus, no such tests have yet become available for early diagnosis of active TBM with the requisite sensitivity and specificity.

The investigators hypothesized that, in contrast to the scarcity of mycobacterial markers in the circulating CSF, macrophages in infected CSF may carry a high load of mycobacterial antigens, which could lend themselves to the development a cell-based diagnostic approach for TBM. Here, the investigators sought to develop a smear immunocytochemical method to improve the accuracy of early diagnosis of TBM by examining early secretory antigenic target 6 (ESAT-6), a mycobacterium-specific antigens, in the macrophages in CSF specimens from patients with TBM and patients with infectious meningitis other than TBM. The investigators also compared the efficiency of this cell-based approach for detecting mycobacterial antigens with microbiological and NAAT method in the laboratory diagnosis of TBM. Further, on the basis of the clinical presentations, CSF and etiological findings in this study, the investigators proposed novel diagnostic criteria for TBM.

Inclusion Criteria:

• Patients older than 14 years of age
• Suspected meningitis within the previous four weeks

Exclusion Criteria:

• Entertained a final diagnosis other than infectious meningitis
• Were under anti-tuberculosis therapy on admission
• Not obtained consent from either the patients or their legal surrogates

All the patients described in this study were admitted to the Department of Neurology, Xijing Hospital of the Fourth Military Medical University, Xi'an, Shaanxi, China.