Tuberculosis of the central nervous system: current concepts in diagnosis and treatment

Purpose of review The outcome of central nervous system (CNS) tuberculosis has shown little improvement over several decades, with diagnosis remaining unconfirmed in nearly half of the cases. This review highlights current insights and advancements in the diagnosis and treatment of CNS tuberculosis. Recent findings Miliary pulmonary tuberculosis is often linked to CNS tuberculosis and is associated with a worse prognosis. Complications, such as, optochiasmatic arachnoiditis, strokes, and transverse myelitis severely affect prognosis and quality of life. Nearly half of tuberculous meningitis patients exhibited impaired cognition. Diagnosing CNS tuberculosis is challenging because of the low accuracy of standard tests. Advanced techniques like metagenomic and nanopore sequencing enhance detection but are hindered by high costs and limited access. Treatment outcomes remain suboptimal but approaches such as higher drug doses, novel medications, and host-directed therapies are being explored. Drug-resistant tuberculous meningitis is increasingly recognized, posing significant challenges to both diagnosis and treatment. Artificial intelligence (AI) enhances care by enabling early diagnosis, disease monitoring, and personalized treatments, improving outcomes. Summary CNS tuberculosis diagnosis faces challenges due to limited sensitivity and delayed results of available tests. Treatments remain suboptimal, with multidrug-resistant cases posing high mortality risks. AI aids in early diagnosis and personalized care.