ABSTRACT
BACKGROUND: Early initiation of antiretroviral therapy (ART) in human immunodeficiency virus (HIV)-infected patients who have tuberculosis reduces mortality among patients with low CD4 counts, but it increases the risk of paradoxical tuberculosis-associated immune reconstitution inflammatory syndrome (IRIS). METHODS: We conducted this randomized, double-blind, placebo-controlled trial to assess whether prophylactic prednisone can safely reduce the incidence of paradoxical tuberculosis-associated IRIS in patients at high risk for the syndrome. We enrolled HIV-infected patients who were initiating ART (and had not previously received ART), had started tuberculosis treatment within 30 days before initiating ART, and had a CD4 count of 100 cells or fewer per microliter. Patients received either prednisone (at a dose of 40 mg per day for 14 days, then 20 mg per day for 14 days) or placebo. The primary end point was the development of tuberculosis-associated IRIS within 12 weeks after initiating ART, as adjudicated by an independent committee. RESULTS: Among the 240 patients who were enrolled, the median age was 36 (interquartile range, 30 to 42), 60% were men, and 73% had microbiologically confirmed tuberculosis; the median CD4 count was 49 cells per microliter (interquartile range, 24 to 86), and the median HIV type 1 RNA viral load was 5.5 log10 copies per milliliter (interquartile range, 5.2 to 5.9). A total of 120 patients were assigned to each group, and 18 patients were lost to follow-up or withdrew. Tuberculosis-associated IRIS was diagnosed in 39 patients (32.5%) in the prednisone group and in 56 (46.7%) in the placebo group (relative risk, 0.70; 95% confidence interval [CI], 0.51 to 0.96; P=0.03). Open-label glucocorticoids were prescribed to treat tuberculosis-associated IRIS in 16 patients (13.3%) in the prednisone group and in 34 (28.3%) in the placebo group (relative risk, 0.47; 95% CI, 0.27 to 0.81). There were five deaths in the prednisone group and four in the placebo group (P=1.00). Severe infections (acquired immunodeficiency syndrome-defining illnesses or invasive bacterial infections) occurred in 11 patients in the prednisone group and in 18 patients in the placebo group (P=0.23). One case of Kaposi's sarcoma occurred in the placebo group. CONCLUSIONS: Prednisone treatment during the first 4 weeks after the initiation of ART for HIV infection resulted in a lower incidence of tuberculosis-associated IRIS than placebo, without evidence of an increased risk of severe infections or cancers. (Funded by the European and Developing Countries Clinical Trials Partnership and others; PredART ClinicalTrials.gov number, NCT01924286 .).