Early antiretroviral therapy for HIV-infected patients admitted to an intensive care unit (EARTH-ICU) : a randomized clinical trial

Abstract

Background: Highly active antiretroviral therapy (HAART) has reduced HIV-related morbidity and mortality at all stages of infection and reduced transmission of HIV. Currently, the immediate start of HAART is recommended for all HIV patients, regardless of the CD4 count. There are several concerns, however, about starting treatment in critically ill patients. Unpredictable absorption of medication by the gastrointestinal tract, drug toxicity, drug interactions, limited reserve to tolerate the dysfunction of other organs resulting from hypersensitivity to drugs or immune reconstitution syndrome, and the possibility that subtherapeutic levels of drug may lead to viral resistance are the main concerns. The objective of our study was to compare the early onset (up to 5 days) with late onset (after discharge from the ICU) of HAART in HIV-infected patients admitted to the ICU. Methods: This was a randomized, open-label clinical trial enrolling HIV-infected patients admitted to the ICU of a public hospital in southern Brazil. Patients randomized to the intervention group had to start treatment with HAART within 5 days of ICU admission. For patients in the control group, treatment should begin after discharge from the ICU. The patients were followed up to determine mortality in the ICU, in the hospital and at 6 months. The primary outcome was hospital mortality. The secondary outcome was mortality at 6 months. Results: The calculated sample size was 344 patients. Unfortunately, we decided to discontinue the study due to a progressively slower recruitment rate. A total of 115 patients were randomized. The majority of admissions were for AIDS-defining illnesses and low CD4. The main cause of admission was respiratory failure. Regarding the early and late study groups, there was no difference in hospital (66.7% and 63.8%, p = 0.75) or 6-month (68.4% and 79.2%, p = 0.20) mortality. After multivariate analysis, the only independent predictors of in-hospital mortality were shock and dialysis during the ICU stay. For the mortality outcome at 6 months, the independent variables were shock and dialysis during the ICU stay and tuberculosis at ICU admission. Conclusions: Although the early termination of the study precludes definitive conclusions being made, early HAART administration for HIV-infected patients admitted to the ICU compared to late administration did not show benefit in hospital mortality or 6-month mortality.