Early antiretroviral therapy favors post-treatment SIV control associated with the expansion of enhanced memory CD8+ T-cells

Abstract

HIV remission can be achieved in some people, called post-treatment HIV controllers, after antiretroviral treatment discontinuation. Treatment initiation close to the time of infection was suggested to favor post-treatment control, but the circumstances and mechanisms leading to this outcome remain unclear. Here we evaluate the impact of early (week 4) vs. late (week 24 post-infection) treatment initiation in SIVmac₂₅₁-infected male cynomolgus macaques receiving 2 years of therapy before analytical treatment interruption. We show that early treatment strongly promotes post-treatment control, which is not related to a lower frequency of infected cells at treatment interruption. Rather, early treatment favors the development of long-term memory CD8⁺ T cells with enhanced proliferative and SIV suppressive capacity that are able to mediate a robust secondary-like response upon viral rebound. Our model allows us to formally demonstrate a link between treatment initiation during primary infection and the promotion of post-treatment control and provides results that may guide the development of new immunotherapies for HIV remission.

Fig. 1. Early therapy is associated with a higher rate of post-treatment controllers.

Plasma viral load kinetics A prior to ART initiation and during ART; and B post-ATI in W4-treated (n = 11) and W24-treated CyMs (n = 11). Plasma viral load kinetics during the initial six months following infection in animals not receiving ART (n = 17) shown as reference. Medians and IQR are shown. C Comparison of plasma viral load levels between W4- and W24-treated CyMs at the peak (acute infection), at the time of ART initiation and prior to treatment interruption. D Comparison of plasma viral load levels between W4- and W24-treated CyMs in the early days following ATI. The magnitude of the plasma viral load post-ATI in W4- and W24-treated CyMs is indicated by (E) the viral load peak and by (F) the cumulative pVL post-ATI (area under the curve - AUC, considering all pVL measurements until 6 months post-ATI). C–F Values for individual animals (W4 n = 11, W24 n = 11) and medians are shown. *p < 0.05, **p < 0.01; ***p < 0.001; ns non-significant; 2-sided Mann‒Whitney U test. G Kaplan‒Meier analyses of maintaining no viral rebound (time without pVL > 400 copies/ml) following ART interruption (left) and achievement of post-treatment control (time to durable pVL <400 copies/ml after viral rebound) (right) in W4- and W24-treated macaques (n = 11 for each). Frequency of CyMs spontaneously controlling plasma viremia (<400 copies/mL) in untreated SIV infection and after antiretroviral treatment interruption (middle and right panels). The frequencies of posttreatment controllers among W4- and W24-treated CyMs are shown. Mantel Cox log-rank test was used to compare the two groups. Source data are provided as a Source Data file.

Fig. 2. Post-treatment controllers maintain low number of infected cells and high CD4+ T cell frequencies.

Kinetics of SIV-DNA levels in A blood and in B PLNs in W4-treated and W24-treated CyMs. The results are expressed as copies of SIV-DNA/million CD4⁺ T cells. C Levels of SIV-DNA in mesenteric and inguinal lymph nodes at euthanasia. The results are expressed as copies SIV-DNA/million cells. Longitudinal evolution of CD4⁺ T cells in (D) blood and in (E) PLNs in W4- and W24-treated CyMs. The results are shown as absolute CD4⁺ T-cell counts in blood and as the proportion of CD4⁺ T cells among CD3⁺ lymphocytes in PLNs. F Proportion of CD4⁺ T cells among CD3⁺ lymphocytes in BM, MLN, spleen, colon mucosa and BAL at euthanasia. Longitudinal evolution of the CD4/CD8 ratio in (G) blood. H The ratio of CD4/CD8 in PBMCs, BM, PLNs, MLNs, spleen, colon mucosa and BAL at euthanasia. The dashed line indicates ratio = 1. I Proportion of central memory (CD45RA^-CD27⁺CCR7⁺) CD4⁺ T cells in PBMCs, BM, PLNs, MLNs, spleen, colon mucosa and BAL at euthanasia. A–I Individual values (n = 6 or 11 for each group depending on sampling during the pVISCONTI-1 and pVISCONTI-2 study phase). Medians are shown. *p < 0.05, **p < 0.01; ***p < 0.001; ns non-significant; Two-sided Mann‒Whitney U test. Source data are provided as a Source Data file.

Fig. 3. The anti-SIV antibody levels after treatment interruption are associated with the magnitude of antigen stimulation.

Kinetics of plasma anti-gp140 SIV IgG (A) and IgA (E) in W4-treated and W24-treated CyMs. Comparison of anti-gp140 SIV IgG (B) and IgA (F) levels between W4- and W24-treated CyMs at baseline, on day 28 p.i.; at 6 months p.i. for the W24-treated group, and prior to ART interruption. C The magnitude of humoral response post-ATI is indicated by the cumulative plasma anti-gp140 SIV IgG (C) and IgA (G) measurements (area under the curve - AUC until 6 months post-ATI). Spearman correlation between pVL AUC post-ATI and anti-gp140 SIV IgG (D) and IgA (H) AUC post-ATI. A–H Individual values (n = 11 animals per group are shown. B, C, F, G Medians are shown, *p < 0.05, **p < 0.01; ***p < 0.001; ns non-significant; Two sided Mann‒Whitney U test. Source data are provided as a Source Data file.

Fig. 4. CD8+ T cells mobilized after interruption of early treatment have enhanced SIV suppressive capacity.

A Kinetics of CD8⁺ T-cell-mediated SIV-suppressive activity in the blood of W4-treated and W24-treated CyMs at primary SIV infection and post-ATI. B Comparison of CD8⁺ T-cell-mediated SIV-suppressive activity between W4- and W24-treated CyMs at baseline, on day 28 p.i.; at 6 months p.i. for the W24-treated group, and prior to ART interruption. C Comparison of the maximal CD8⁺ T-cell-mediated SIV-suppressive activity during the period following treatment interruption in W4- and W24-treated CyMs. D The magnitude of CD8⁺ T-cell-mediated SIV-suppressive activity is indicated by the cumulative measurements post-ATI (area under the curve – AUC of all measurements until 6 months post-ATI). E Kinetics of CD8⁺ T-cell-mediated SIV-suppressive activity in PLNs of W4- and W24-treated CyMs at the time of ART initiation, 14 days post-ATI and at euthanasia. The results are shown as log p27 decrease in the presence of CD8⁺ T cells. Spearman correlation between (F) pVL AUC post-ATI and blood CD8⁺ T-cell-mediated SIV-suppressive activity AUC post-ATI, G pVL and PLN CD8⁺ T-cell-mediated SIV-suppressive activity at euthanasia, (H) SIV-DNA levels in blood CD4 at euthanasia and blood CD8⁺ T-cell-mediated SIV-suppressive activity AUC post-ATI, I SIV-DNA levels at euthanasia and PLN CD8⁺ T-cell-mediated SIV-suppressive activity at euthanasia. (A–I) Individual values (n = 11 animals per group, except for (E), where n = 5–10 per group were analyzed depending on sample availability) are shown. (B–E) Medians are shown; *p < 0.05, **p < 0.01; ***p < 0.001; ns non-significant; Two sided Mann‒Whitney U test. Source data are provided as a Source Data file.

Fig. 5. CD8+ T cells expanding after interruption of early treatment express low levels of activation markers.

Expression levels of CD38 and HLA-DR activation markers (A) and intracellular levels of Ki67 proliferation markers (E) in blood CD8⁺ T cells during primary SIV infection and post-ATI in W4-treated (n = 11) and W24-treated CyMs (n = 11). Cumulative CD38, HLA-DR (B) and Ki67 (F) expression levels in blood CD8⁺ T cells during the first 4 weeks following SIV infection and post-ATI. Comparison of CD38 and HLA-DR (C) and Ki67 (G) expression levels in PLN CD8⁺ T cells between W4- and W24-treated CyMs at the time of cART initiation, 14 days post-ATI and at euthanasia. The results are shown as percent frequencies of CD8⁺ T cells or AUC; Spearman correlation between cumulative (AUC) CD38 and HLA-DR (D) and Ki67 (H) expression levels post-ATI and blood CD8⁺ T-cell-mediated SIV-suppressive activity AUC post-ATI. (A-H) Individual data are shown (n = 6 animals per group); (B–D, F, G) Medians are shown; *p < 0.05, **p < 0.01; ***p < 0.001; ns non-significant; Two sided Mann‒Whitney U test. Source data are provided as a Source Data file.

Fig. 6. Expansion of central memory CD8 + T cells is observed in blood and lymph nodes after interruption of early treatment.

Evolution of central memory (CM) (CD45RA^-CD27⁺CCR7⁺), effector memory (EM) (CD45RA⁻CD27⁻CCR7⁻) and effector (CD45RA⁺CD27^-CCR7^-) CD8⁺ T-cell subsets in blood at primary SIV infection (top) and early post-ATI (bottom) in W4- (A) and W24-treated (B) CyMs. Frequency of CM, EM and effector CD8⁺ T cells in PLNs of W4- C and W24-treated D CyMs prior to and 14 days post-ATI. E Comparison of intracellular levels of Ki67 in blood (left) and PLN (right) CM CD8⁺ T cells at baseline, at the time of ART initiation, prior to and post-ATI in W4- and W24-treated CyMs. F Spearman correlation between blood CD8⁺ T-cell-mediated SIV-suppressive activity AUC post-ATI and the frequency of CM CD8⁺ T cells 28 days post-ATI in blood (left) and 14 days post-ATI in PLNs (right). A–F Individual data are shown (n = 6 animals per group); A–E Medians are shown; *p < 0.05, **p < 0.01; ***p < 0.001; ns non-significant; A, B Friedman test with Dunn’s correction; C,D Two sided Wilcoxon matched-pairs rank test. E Two sided Mann‒Whitney U test. Source data are provided as a Source Data file.

Fig. 7. SIV-specific CD8⁺ T cells with memory-like characteristics are mobilized after interruption of early treatment.

A INFγ, TNFα, IL-2, and CD107a production by SIV-specific CD8⁺ T cells and the total SIV-specific response in the blood of W4-treated and W24-treated CyMs at the time of ART initiation and post-ATI. The results are shown as proportions among CD8⁺ T cells. Negative responses were given an arbitrary value of 0.01. B Frequencies of SIV-specific CD8⁺ T cells with 1-to-4 functions based on the expression of IFNγ, TNFα, IL-2, and/or CD107a in W4- and W24-treated CyMs at the time of ART initiation and post-ATI. Negative polyfunctional responses are indicated as 0.001. C CD127 expression levels in SIV-specific CD8⁺ T cells at the time of ART initiation and post-ATI. The results are shown as the median fluorescence intensity in SIV-specific CD8⁺ T cells. D Proportion of SIV-specific memory CD8⁺ T cells expressing PD-1, CD39, CCR7 or TCF1 in the spleen at the end of the study. E Division index (left) and proliferation index (right) of CD8⁺ T cells at the time of ART initiation and post-ATI. F Left panel: UMAP plot of n = 2979 SIV-specific CD8⁺ T cells from the blood of W4- and W24-treated CyMs at the time of ART initiation and post-ATI. Phenotypically distinct clusters defined by Phenograph are represented with different colors. Right panel: Heatmap showing the relative expression of differentiation markers, cytokine secretion and markers of mTORC1 (pS6) and mTORC2 (pAKT) pathway activation for each phenotypically distinct cluster shown in (F). G Dynamics of Clusters 1, 2 and 3 at the time of ART initiation and post-ATI in W4-treated CyMs. H) Violin plots comparing the median fluorescence intensity (MFI) of the differentiation markers, cytokine secretion and markers of mTORC1 and mTORC2 pathway activation among Clusters 1, 2 and 3 in W4-treated CyMs; one-way ANOVA, *p < 0.05, **p < 0.01; ****p < 0.0001. A–E, G, H Individual data are shown. A–C, E, G, H n = 6 animals per group. D n = 12 animals per group. A–E Medians are shown; *p < 0.05, Two-sided Mann‒Whitney U test. G Two-sided Wilcoxon matched-pairs rank test. Source data are provided as a Source Data file.