Background
We modelled patterns of virological response to Dolutagravir (DTG)-based ART in a cohort of children and adolescents living with HIV in West Africa.
Methods
All CALHIV aged 0 to 19 years initiating DTG-based ART between 2019-2023 and enrolled in the West African pediatric International epidemiological Database for Evaluating AIDS (IeDEA pWADA) were included and followed up to 24 months since DTG initiation (baseline). Baseline viral load (VL) was the closest measurement within 6 months prior and 2 weeks post DTG initiation. We used group-based trajectory modelling to assign each CALHIV to a group based on their VL accross visits; a logistic trajectory was used to identify groups with similar 24-month longitudinal patterns of VL based on dectable VL at a 1000-copy threshold. Multinomial regression analysis adjusted for age, sex and program was used to identify factors associated with each VL trajectory.
Results
Overall, 2,319 CALHIV (46% female) were included; 1,649 (71%) were aged 10-19 years. Median follow-up time on DTG was 15 months [IQR 5 – 29]; 0.6% died and 6.4% were lost to follow-up (≥7 months since last visit). At baseline, 1,149 (50%) were in viral suppression (VS, VL<200 copies), 100 (4%) had low level viremia (LLV: 200 <VL≤1000 copies) and 450 (19%) were in viral failure (VF: VL>1000 copies); 622 (27%) had no available baseline VL. Four trajectory groups were identified over 24 months (Figure 1): those with sustained VS (79%), decreasing probability for VF over time (5%), increasing probability for VF over time (9%) and high probability for VF (7%). Among those in VF at baseline, 54% reached VS but 46% remain in the high probability group, with sustained VF throughout follow-up. Age and sex were not associated with VL trajectories. However, those in care in Nigeria were least likely to be in either the low probability for VF by 24 months (p<0.001), moderate probability for VF (p<0.001) or high probability for VF (p=0.004) compared to Côte d’Ivoire.
Conclusions
We identified four different patterns among CALHIV initiating DTG. Those in VF at DTG initiation remained at moderate or high probability for VF, unable to achieve VS by 24 months on DTG. While our current model was not able to capture individual factors associated with HIV VL trends, adherence and resistance studies are needed to explain VF at DTG initiation and guide interventions to improve clinical outcomes among CALHIV.