Résumé
BACKGROUND: Frequency of adrenal insufficiency in patients with tuberculosis varies from 0 to 58%; however, all published series excluded severely ill patients. Our objective was to investigate adrenal insufficiency with the low-dose cosyntropin test in patients with severe active tuberculosis. METHOD: From two large university affiliated hospitals, 18 patients with tuberculosis and criteria of sepsis or severe sepsis according to SCCM/ACCP criteria, defined by the present authors as severe active tuberculosis, participated in the study. A low-dose ACTH test with 10 mg of ACTH was performed. After ACTH test, all patients received a stress dose of hydrocortisone (240 mg/day) during their entire hospitalization along with four antituberculous drugs. Abnormal response was considered when elevation of serum cortisol was <7 microg/dl with respect to basal level, 60 min after ACTH administration. RESULTS: Adrenal insufficiency was found in seven patients (39%); no clinical or laboratory data were associated with the presence of abnormal adrenal response. Except in one patient with HIV infection, all the signs and symptoms improved after antituberculous and hydrocortisone treatment. The increment in serum cortisol value post-ACTH test was lower in patients with hypoalbuminemia. CONCLUSIONS: Adrenal insufficiency is frequent in severe active tuberculosis. The efficacy and security of supplemental steroid treatment in severe active tuberculosis should be established by a randomized clinical trial.