Episodes of HIV Viremia and the Risk of Non-AIDS Diseases in Patients on Suppressive Antiretroviral Therapy

Background: Several studies reported an association between immunodeficiency and non-AIDS-defining diseases. We investigated whether nonstructured treatment interruptions and episodes of viremia during suppressive combination antiretroviral therapy were independently associated with non-AIDS diseases. Methods: Six thousand four hundred forty patients with viral suppression (<50 copies/mL) within 48 weeks of starting combination antiretroviral therapy were selected from the Dutch ATHENA cohort. In proportional hazards models, associations between treatment interruptions, viral suppression, low-level (50-400 copies/mL), and high-level viremia (>400), and serious non-AIDS diseases (cardiovascular disease, chronic renal failure, liver fibrosis/cirrhosis) were investigated by including time-updated cumulative exposure to either viremia and interruptions or HIV RNA >400 copies per milliliter. Results: During 24,603 person-years, of which 88.5% occurred during viral suppression, 102 patients developed cardiovascular disease, 54 chronic renal failure, and 70 liver fibrosis/cirrhosis. Overall incidence of non-AIDS diseases ranged from 1.41 (95% confidence interval: 0.73 to 2.46) per 100 person-years for CD4 counts <200 to 0.71 (0.49 to 1.00) for CD4 >= 500 cells per cubic millimeter. Compared with viral suppression, high-level viremia was associated only with cardiovascular disease (relative hazard: 1.37, 1.04 to 1.81 per year longer), whereas interruptions and low-level viremia were not associated with non-AIDS diseases. Relative hazards for cumulative exposure to RNA >400 versus <= 400 copies per milliliter were 1.32 (1.01 to 1.73) for cardiovascular disease, 1.13 (0.66 to 1.92) for renal failure, and 0.86 (0.51 to 1.44) for fibrosis/cirrhosis. Conclusions: Lower CD4 counts are associated with increased risk of non-AIDS diseases, whereas high-level viremia seems to be independently associated with cardiovascular disease. However, the power to detect associations with viremia or interruptions may have been limited as most events occurred during viral suppression.