Prediction of neurologic sequelae in childhood tuberculous meningitis: a review of 20 cases and proposal of a novel scoring system.

BACKGROUND: Despite effective antituberculous medications, the mortality and morbidity remain high in children with tuberculous meningitis (TBM). The traditional clinical staging for TBM developed by Lincoln et al in 1960 has been widely used to predict long term neurologic sequelae (NS). In the current era of critical care medicine and corticosteroid therapy, a new scoring system is needed to predict NS more accurately in children with TBM. METHODS: We reviewed all available cases of TBM in San Diego, CA, during 1991-2001 retrospectively, and we developed a novel scoring system to predict NS in children with TBM. We assessed a tuberculous meningitis acute neurologic (TBAN) score at day 0 and on day 3 of hospitalization, to compare children who subsequently developed severe NS with those who did not. RESULTS: Among 20 children with TBM, 7 children developed severe NS and 1 child died during hospitalization. The TBAN score was higher on day 0 in those with severe NS (5.5 versus 2.0, P = 0.09), and the difference became statistically significant by day 3 of hospitalization (5.5 versus 0.0, P = 0.02). Sensitivity and specificity of the TBAN score (> or =4) on day 0 (75 and 92%) and day 3 (88 and 100%) to predict severe NS were superior to the traditional clinical staging system on day 0 (63 and 58%). CONCLUSIONS: The TBAN score is an objective marker for predicting severe NS in children with TBM.

Tuberculous Meningitis in Children.

Initial empiric treatment for central nervous system (CNS) tuberculosis should include four antituberculous drugs until results of cultures and sensitivities are available. Treatment should include isoniazid, rifampin, pyrazinamide, and either ethambutol or streptomycin. Total treatment should extend for 12 months. Daily therapy should be used for the first 2 months, followed by either twice a week treatment or continued with daily therapy for the duration with directly observed therapy (DOT). Pyrazinamide should be included in all treatment regimes for the first 2 months of therapy. Corticosteroids should be used in the management of children with tuberculous meningitis. Corticosteroids have been shown to decrease mortality, long-term neurologic complications, and permanent sequelae. Prednisone is often used at a dosage of 1 to 2 mg/kg per day. Steroids should be used for 4 to 6 weeks, and then tapered over the next 2 to 3 weeks. Cerebrospinal fluid (CSF) cultures and other infected sites must be aggressively pursued in order to obtain an organism for identification and sensitivities testing. Cranial CT scans with contrast should be included in the early diagnostic work-up of a child with suspected CNS tuberculosis infection. Hydrocephalus is often an early finding and may be helpful in establishing the diagnosis of CNS tuberculosis. Treatment of CNS tuberculosis should be for 12 months. All children with CNS tuberculosis should be promptly reported to the local public health department. Public health will facilitate the case-contact study and assist with follow-up and DOT after discharge. Directly observed therapy should be given for the entire treatment course. This is best accomplished with the collaboration of local public health services. Children with tuberculous meningitis should be evaluated in follow-up monthly. Monitoring should include determining adherence to drug treatment, an interval history for signs and symptoms of disease progression, careful physical examinations and evaluation for adverse effects of drugs. Liver function tests should be obtained at baseline, 2-, 4-, 6-, and 8 weeks, and then monthly for the first several months of treatment. Children with tuberculous meningitis should be tested for HIV infection, including pre- and post-test counseling.