Abstract
BACKGROUND: Adequate exposure to antiretroviral drugs is necessary to achieve and sustain viral suppression. However, the target antiretroviral concentrations associated with long-term viral suppression have not been adequately defined in children. We assessed the relationship between plasma lopinavir or nevirapine concentrations and the risk of subsequent viremia in children initially suppressed on antiretroviral therapy. METHODS: After an induction phase of antiretroviral treatment, 195 children with viral suppression (viral load ≤400 copies/mL) were randomized to continue a lopinavir/ritonavir-based regimen or to switch to a nevirapine-based regimen (together with lamivudine and stavudine). Viral load and lopinavir or nevirapine concentrations were measured at clinic visits 4, 8, 12, 16, 20, 24, 36, 52, 64 and 76 weeks post randomization. Cox multiple failure event models were used to estimate the effects of drug concentrations on the hazard of viremia (viral load >50 copies/mL) RESULTS:: At randomization, the median (interquartile range) age, CD4 T-Lymphocyte percentage, weight-for-age and weight-for-height z scores were 19 (16-24) months, 29% (23-37), -0.6 (-1.3 to 0.2) and -3.2 (-4.1 to -2.1), respectively. The proportion of children with viral load 51-400 copies/mL at randomization was 43%. The hazard of subsequent viremia during follow-up was increased for lopinavir concentrations <1 versus ≥1 mg/L [adjusted hazard ratio 0.62 (95% confidence interval, 0.40-0.94)] and for children with viral loads 51-400 copies/mL at randomization. Nevirapine concentrations were not significantly associated with subsequent viremia. CONCLUSIONS: Plasma lopinavir concentrations predicted viral outcomes in children receiving lopinavir-based antiretroviral therapy. Our findings support a minimum target concentration of ≥1 mg/L of lopinavir to ensure sustained viral suppression.