Abstract Body

Background

ART stops replication and suppresses HIV-1 RNA to undetectable levels, but a reservoir of infected cells persists. We, and others, have demonstrated that long-lived reservoirs are typically composed of cells infected near ART initiation. Here, we examine whether this pattern holds in a more clinically diverse cohort of men and women with a wide range of years untreated and CD4+ T cell nadirs.

Methods

Longitudinal blood samples were collected pre- and post-ART from 19 participants with HIV-1 in the MACS/WIHS Combined Cohort (MWCC; N=7 women, N=12 men). Median values of the duration of untreated infection and nadir CD4+ T cell counts were 7 years (range 2-16) and 258 cells/uL (range 12-606), respectively. HIV-1 RNA was isolated from pre-ART plasma, converted to cDNA, and 5 subgenomic regions were sequenced with Primer-ID Miseq deep sequencing. 3’ half genomes were amplified at limiting dilution from proviral DNA in PBMCs collected after 3-9 years of ART suppression and sequenced (PacBio with barcodes). Genetic similarity between proviruses and pre-ART HIV-1 RNA was used to estimate when reservoir cells were infected. Fig 1a shows a representative analysis of longitudinal partial env sequences from 1 participant.

Results

We examined a median of 42 unique reservoir sequences (range 32-59) per participant. A median of 64% of reservoir cells were infected in the year before ART (range 35 – 94%; Fig 1b). The percent of reservoir cells infected in the year before ART was positively correlated with viral load at ART initiation (p = 0.03, r = 0.49) and negatively correlated with CD4:CD8 ratio at both 6 (p = 0.03, r =0.52) and 36 months post-ART (p = 0.01, r = 0.59), but was not correlated with duration of untreated infection, nadir CD4+ T cell count or sex.

Conclusions

Across this clinically diverse cohort of people with HIV-1 on ART, we observed that cells infected near ART initiation typically make up most of the reservoir. The ubiquity of near-ART variants in the reservoir suggests that generalizable mechanisms drive the temporal dynamics of reservoir formation. Our results also suggest that interventions that prevent reservoir formation at ART initiation could broadly benefit people with diverse clinical histories.

Table or Figure