Early diagnosis and appropriate treatment of tuberculous meningitis (TBM) are crucial steps to reduce morbidity and mortality. The WHO recommended to use Xpert MTB/RIF assay to diagnose pulmonary TB, pediatrics TB, extra pulmonary TB and rifampicin resistance. However, the data of accuracy in diagnosis of TBM is still lacking. This study aimed to find out the diagnostic performance of Xpert MTB/RIF assay for the diagnosis of tuberculous meningitis, especially in patients who presented with subacute lymphocytic meningitis.

Tuberculous meningitis (TBM) was the most serious manifestation of Mycobacterium tuberculosis infection with high rates of morbidity and mortality if treatment is delayed, especially in HIV-infected patients. The fatality rates increased to 60 percent compared to about one third in non-HIV infection. Prevalence and incidence of TBM worldwide is unclear. In 2015, the incidences of TBM were approximately 6-10 percent. In Northern Thailand, from the national survey programs, the prevalence rates of TBM were about 9.7 per 100,000 people, and TBM was the most common cause of death in extra-pulmonary TB patients with mortality rates about 0.54 per 100,000 people. However, these may be underreported due to diagnostic problems.

Prompt and accurate diagnoses with early treatments, were the key elements to reduce deaths and disabilities in TBM. However, there were some major obstacles in the diagnostic procedures concerning the limitations, such as low sensitivity in conventional acid-fast bacilli (AFB) microscopy staining techniques or long turnaround times in gold-standard culture. Therefore, the delays in diagnoses lead to late managements and increased mortalities in TBM, especially in multidrug-resistance (MDR) cases. In the past, a lot of studies tried to solve these problems urgently, but the low frequency of confirmed cases and the absence of a consensus clinical case definition were the huge barriers causing a lack of progression in new clinical trials. In 2010, a uniform clinical case definition for use in clinical research about TBM was established. They used criteria from clinical characteristics, cerebrospinal fluid (CSF) examinations, neuroimaging studies, and evidence of tuberculosis elsewhere. Additionally, microbiological or pathological proof was used to classify the patients who had signs and symptoms of meningitis and were defined as definite, probable, possible, and not tuberculous meningitis. Since then, many clinical studies about TBM diagnosis were developed and made more convenient.

The Xpert MTB/RIF assay was the closed-cartridge-based system heminested PCR for the diagnosis of tuberculosis (TB), and detecting the rifampicin resistant gene in sputum and the CSF. This can be used as a marker for MDR-TB, focused on endemic areas and HIV coinfection. This test has better diagnostic threshold levels (80-100 cfu/ml) than the 5,000 cfu/ml in AFB microscopy staining techniques, shorter duration for the results than conventional TB cultures, and more simplification than nucleic-acid amplification tests. Many studies about the diagnostic accuracy revealed that sensitivity and specificity of this new test varied from 50-86 % and 94-98 % depended on method, population, and laboratory techniques. There were some discussions about the application practicality in the diagnosis of TBM, such as in a large-population study in Vietnam that questioned the lower sensitivity and cost-effectiveness compared with conventional staining and culture. Subsequently, this newer rapid test was approved by WHO for TBM diagnosis.

In Thailand, one of the countries with a high burden of tuberculosis, despite many advantages, use of this test was limited in CSF specimens due to the lack of data, including accuracy in CSF specimens for TBM diagnosis among the Thai population. This study aims to prospectively find out the diagnostic performance and accuracy of Xpert MTB/RIF in CSF samples from patients who presented with subacute lymphocytic meningitis in Northern Thailand, which can be the database for extending the study to cover patients in other parts of Thailand.

Inclusion Criteria:

1. Presence of meningitis symptoms and signs including one or more of the following: headache, irritability, vomiting, fever, neck stiff ness, convulsions, focal neurological deficits, altered consciousness, or lethargy at least 5 days prior to admission, plus

2. CSF with abnormalities of at least two of these three characters;

   • Lymphocytic pleocytosis (≥ 50%)
   • More than 40 mg/ml of protein
   • Lower than 0.5 of CSF: blood sugar ratio

Exclusion Criteria:

1. Contraindicated in lumbar puncture
2. Positive for cryptococcal antigen in the CSF.