Although a long-term study of clinically-effective ART showed no evidence of HIV evolution in 13 of 14 adults (Kearney, 2014), a recent study (Lorenzo-Redondo, 2016) of 3 persons concluded that HIV evolution occurs on ART at a rate of 6×10^-4 to 1×10^-3 mutations/site/month due to ongoing viral replication in the lymph nodes with subsequent trafficking of newly infected T-cells to blood. We therefore looked for evidence of HIV evolution in children on long-term suppressive ART. …
We studied children initiated on continuous ART in the CHER study (Cotton, Lancet 2013). Samples obtained near ART initiation and after 7 years of ART underwent single-genome sequencing of the p6-PR-RT region of the HIV genome. Sequences from each time point were compared for evidence of evolution by multiple, sensitive methods: 1) calculation of average pairwise distance (APD) for sequence diversification, 2) panmixia testing for sequence divergence, and 3) construction of maximum-likelihood (ML) trees to measure root-to-tip distances for emerging new variants.
12 children were studied: 10 who started ART ≤ 1 year and had viremia suppressed on ART for 7 years and 2 ‘replication-controls’ who had viremia for 1-3 years prior to or during ART. All children had very low viral diversity (median of 0.3%) at the time of ART initiation (Table), providing a low background on which to detect evolution. The 2 replication-controls showed obvious evidence of HIV evolution: increased viral diversity, a significant virus population shift by panmixia (Table), and longer root-to-tip distances in ML trees. In 8 of 10 children, there was no evidence of viral divergence on ART. A significant virus population shift by panmixia occurred in 1 child (PID 8) who had decreasing branch lengths on the ML tree suggesting decay of infected cells on ART. Another child (PID 9) with an HIV RNA blip to 56 copies/ml had a marginally significant shift in the virus population by panmixia from possible ongoing viral replication. The absence of viral evolution in 9 of 10 children is inconsistent with the 5-8% divergence expected from ongoing replication over 7 years predicted by Lorenzo-Redondo et al.
These data from early ART-treated children strongly refute the concepts that ongoing HIV replication is common on current ART regimens and that it replenishes the HIV reservoir.