Abstract

Objectives Most data for Central Nervous System Tuberculosis (CNS-TB) derive from high-incidence, resource-limited countries. We sought to determine the presentation, management and outcomes of CNS-TB in a low-incidence setting with accessible healthcare. Methods We undertook a retrospective, observational study of CNS-TB in adults at a single tertiary-referral London hospital (2001–2017). Cases were categorised as either TB meningitis (TBM) or TB mass lesions without meningitis (TBML), applying novel criteria for definite, probable, and possible TBML. Results We identified sixty-two cases of TBM (37% definite; 31% probable; 32% possible) alongside 14 TBML cases (36% definite; 29% probable; and 36% possible). Clinical presentation was highly variable. Among CSF parameters, hypoglycorrhachia proved most discriminatory for “definite” TBM. Neurosurgical intervention was required for mass-effect or hydrocephalus in 16%. Mortality was higher in TBM versus TBML (16% vs. 0%) but overall morbidity was significant; 33% of TBM and 29% of TBML survivors suffered persisting neurological disability at 12-months. In TBM, hydrocephalus, infarct, basal enhancement and low CSF white cell count were independently associated with worse neurological outcomes. Conclusion Although mortality was lower than previously reported in other settings, morbidity was significant, highlighting the need for improved CNS-TB diagnostics, therapeutics and interventions to mitigate neurological sequelae.