Background. Perinatal treatment with lopinavir boosted by ritonavir (LPV/r) is associated with steroidogenic abnormalities. Long-term effects in infants have not been studied.

Methods. Adrenal-hormone profiles were compared at weeks 6 and 26 between human immunodeficiency virus (HIV)-1-exposed but uninfected infants randomly assigned at 7 days of life to prophylaxis with LPV/r or lamivudine (3TC) to prevent transmission during breastfeeding. LPV/r in vitro effect on steroidogenesis was assessed in H295R cells.

Results. At week 6, 159 frozen plasma samples from Burkina Faso and South Africa were assessed (LPV/r group: n = 92; 3TC group: n = 67) and at week 26, 95 samples from Burkina Faso (LPV/r group: n = 47; 3TC group: n = 48). At week 6, LPV/r-treated infants had a higher median dehydroepiandrosterone (DHEA) level than infants from the 3TC arm: 3.91 versus 1.48 ng/mL (P < .001). Higher DHEA levels (>5 ng/mL) at week 6 were associated with higher 17-OH-pregnenolone (7.78 vs 3.71 ng/mL, P = .0004) and lower testosterone (0.05 vs 1.34 ng/mL, P = .009) levels in LPV/r-exposed children. There was a significant correlation between the DHEA and LPV/r AUC levels (rho = 0.40, P = .019) and C-trough (rho = 0.40, P = .017). At week 26, DHEA levels remained higher in the LPV/r arm: 0.45 versus 0.13 ng/mL (P = .002). Lopinavir, but not ritonavir, inhibited CYP17A1 and CYP21A2 activity in H295R cells.

Conclusions. Lopinavir was associated with dose-dependent adrenal dysfunction in infants. The impact of long-term exposure and potential clinical consequences require evaluation.