Mycobacterial culture and PCR remain the gold standard for the confirmation of TB meningitis. Molecular (PCR-based) TB assays continue to have suboptimal sensitivity though (although they are increasingly used for CSF analysis, including the Xpert MTB/RIF assay), which means that a negative test result does not rule out TB meningitis. If the history (including contact history), immunological TB tests (TST or IGRA) and/or CSF findings (ie WBC, protein, glucose) are suggestive of TB, the safest strategy is to commence anti-TB treatment despite absence of microbiological confirmation.
One brief addition to my previous reply - the WHO guidance on using PCR-based assays in the context of suspected TB meningitis can be found here: http://apps.who.int/iris/bitstream/10665/112472/1/9789241506335_eng.pdf .