Adjunctive corticosteroid therapy for tuberculosis: a critical reappraisal of the literature.

An extensive, although largely forgotten, literature addresses the utility of adjunctive corticosteroid therapy in the management of tuberculosis. Corticosteroid therapy probably improves neurological outcomes of, and decreases mortality due to, tuberculous meningitis of moderate severity. Although therapy for tuberculous pericarditis is simplified (with less need for operative intervention) by adjunctive corticosteroid administration and there are fewer deaths, the incidence of subsequent constriction is not changed. The signs and symptoms of typical reactivation tuberculous pneumonia, tuberculous pleurisy, and probably primary tuberculous disease (with lymphadenopathy) seem to decrease rapidly with corticosteroid therapy, although no differences in final outcomes have been observed. Corticosteroid regimens used in most studies varied greatly in duration and dosage and generally caused significant side effects. Corticosteroids do not appear to diminish the efficacy of adequate antimycobacterial therapy. Adjunctive corticosteroid therapy appears to offer significant short-term but (other than for tuberculous meningitis and effusive pericarditis) minimal long-term benefit for patients with tuberculosis.