Background: Studies comparing efficacy of PI- and NNRTI-based initial ART regimens in children have shown inconsistent results. We investigated virological response in children from 18 cohorts across 16 European countries and Thailand.
Methods: Time of virological suppression (''response'') was estimated as the midpoint between the first VL< 400c/mL and previous VL≥400c/mL. Multivariable stepwise logistic regression models (backwards elimination, exit probability p=0.05) were used to identify factors associated with virological suppression by 12 months from ART initiation (baseline).
Results: 2202 children initiating ART < 18 years with a boosted PI or NNRTI plus >2 NRTI had baseline VL and ≥1 VL measurement available within the first 15 months of ART. 91% were perinatally infected with median[IQR] age 3.8[0.8,8.5] years at HIV diagnosis and 6.4[1.6,10.7] years at ART initiation. Median[IQR] baseline CD4% was 16[8,24]. 91% achieved virological response by 12 months. In multivariable analysis, the effect of ART regimen on virological response varied by age at ART initiation (Table): in children < 3 years, compared to a PI-based regimen, odds of response were significantly reduced for those starting a NVP+2NRTI regimen, but not significantly different for those initiating on NVP+3NRTI; conversely, children ≥3 years had higher odds of virological response compared to those < 3 years, with no significant differences between regimens. Additionally, across all ages, starting an abacavir-containing regimen, VL < 100,000 copies/mL, and more recent calendar year at ART initiation were associated with increased odds of virological response. Children in Russia/Ukraine had reduced odds of response. There was no significant effect of gender, being born abroad, baseline CDC clinical stage, CD4%, HCV/HBV-status or mode of infection.

Table: Factors at ART initiation associated with virological response by 12 months
[Table: Factors at ART initiation associated with virological response by 12 months]

Conclusions: Most children achieve virological suppression by 12 months. Response was more likely in recent years and less likely in children starting NVP+2NRTI < 3 years, possibly due to under-dosing. Further work will explore predictors of virological failure.

Download the e-Poster