Clinical features, diagnosis, and treatment of tuberculosis in infants.

OBJECTIVE: Despite the recent resurgence of tuberculosis among children in the United States, no series of infants < 1 year of age with tuberculosis has been reported in the last 20 years. This study was undertaken to describe the epidemiology, clinical, and radiographic manifestations, and response to therapy in infants < 1 year of age with tuberculous disease. METHODS: The medical records were reviewed for all infants age 12 months or less with a diagnosis of tuberculosis and cared for at the Children's Tuberculosis Clinic at Ben Taub General Hospital in Houston, Texas between January 1, 1985 and June 30, 1992. RESULTS: Of the 47 infants identified, 51% were female. The median age at diagnosis was 8 months (range 3.5 to 12 months). Fifty-one percent of the infants were African-American and over one-third were Hispanic. All patients were born in the United States. Diagnosis resulted from the examination of an ill infant in 79% of cases, a case contact investigation of an adult suspected of having tuberculosis in 19%, and routine tuberculin skin testing in 2%. An adult with infectious tuberculosis who had contact with the infant was identified in 68% of cases. Intrathoracic disease alone was present in 70% of cases. Fourteen (30%) infants had extrapulmonary tuberculosis (11 central nervous system disease, 2 disseminated disease, and 1 cervical adenitis). Gastric aspirate cultures yielded Mycobacterium tuberculosis from 75% of the infants with isolated intrathoracic disease. Forty-five infants successfully completed therapy and only one death was directly related to tuberculosis. Forty-eight percent of the infants with pulmonary tuberculosis were treated with a 6-month regimen consisting of isoniazid and rifampin supplemented during the first 2 months by pyrazinamide. Eighteen infants received some twice weekly directly observed therapy mainly due to documented or suspected nonadherence. Treatment was well-tolerated; one patient (2%) developed hepatotoxicity due to isoniazid. No infant had a relapse or recurrence of disease in 6 months to 7 years follow-up for a median of 3 years (range, 6 months to 7 years). CONCLUSION: Most infants with tuberculosis are symptomatic at the time of diagnosis, and bacteriologic confirmation was obtained in 70% of cases. A contact investigation of the household should be initiated immediately when an infant is suspected of having tuberculosis because valuable information needed to establish the diagnosis and guide therapy in the infant may be obtained. Intensive 6-month and twice weekly directly observed therapy appear to be well-tolerated and effective for the treatment of pulmonary tuberculosis in infants.