Abstract Body

Background: There are no comprehensive analyses of HIV proviral DNA in Primary HIV Infection (PHI) that include inflammation, HIV-specific immunity and clinical progression. We present analyses from the SPARTAC trial characterizing immunological and virological associations with proviral DNA in PHI, and exploring the impact of proviral DNA on clinical progression and virological rebound on stopping ART. Methodology: Participants with PHI in SPARTAC were analysed for HIV proviral DNA (‘integrated’ and ‘total’ in CD4 T cells), HIV CD4 and CD8 T cell immune responses (by ELISpot), immune activation (HLA DR, CD38), immune exhaustion (PD-1, TIM-3, LAG-3) and generalized markers of inflammation (D-dimer, IL-6). Outcome measures included time to trial primary endpoint and time to plasma viral load rebound on stopping ART. Statistical associations were determined using linear regression and Cox proportional hazard models. Results: Baseline samples at trial enrolment were available from up to 154 participants with subtype B infection. ‘Integrated’ and ‘Total’ HIV DNA levels were closely correlated (r=0.85; P<0.001) and associated with baseline CD4 count (P<0.001) and plasma viral load (P<0.001). CD8 and CD4 T cell activation (CD38, HLA DR) were significantly associated with HIV DNA (P<0.01), the association being stronger for CD8s. There was a strong association with HIV DNA and D-dimer (P<0.01), but not IL-6. Both CD8 PD-1 and Lag-3 (not TIM-3) were significantly associated with HIV DNA, but there was no evidence for an association with proviral DNA and CD4 T cell exhaustion. Participants making a CD4 T cell ELISpot response to HIV Gag had significantly lower HIV DNA levels (p<0.05; Mann Whitney), with weaker evidence for an association with CD8 T cell immunity (P=0.03 and 0.05 for responses to 0 vs 3 Gag overlapping peptides for ‘total’ and ‘integrated’ DNA, respectively). In univariable analyses of untreated individuals at PHI, both ‘total’ and ‘integrated’ DNA were significant predictors of clinical progression. In multivariable analyses, ‘total’ DNA remained strongly associated with trial outcome (HR 3.6; P=0.002), whereas integrated DNA and plasma viral load did not. In patients receiving 48 weeks of suppressive ART, total HIV DNA at time of stopping therapy was predictive of trial end-point (HR 3.6; p=0.01) and time to plasma viral load rebound (HR 2.5; P=0.007). Conclusions: This is the first comprehensive analysis of the HIV reservoir in the context of HIV-specific immunity, inflammation and clinical progression in a large randomized clinical trial in PHI. The data confirm the close relationship between immunity, inflammation and the reservoir in PHI, and show that reservoir size is a key predictor not only of progression but also of viraemic control on stopping ART.