HIV-Specific, Ex Vivo Expanded T Cell Therapy: Feasibility, Safety, and Efficacy in ART-Suppressed HIV-Infected Individuals

Abstract

Adoptive T cell therapy has had dramatic successes in the treatment of virus-related malignancies and infections following hematopoietic stem cell transplantation. We adapted this method to produce ex vivo expanded HIV-specific T cells (HXTCs), with the long-term goal of using HXTCs as part of strategies to clear persistent HIV infection. In this phase 1 proof-of-concept study (NCT02208167), we administered HXTCs to antiretroviral therapy (ART)-suppressed, HIV-infected participants. Participants received two infusions of 2 × 10⁷ cells/m² HXTCs at a 2-week interval. Leukapheresis was performed at baseline and 12 weeks post-infusion to measure the frequency of resting cell infection by the quantitative viral outgrowth assay (QVOA). Overall, participants tolerated HXTCs, with only grade 1 adverse events (AEs) related to HXTCs. Two of six participants exhibited a detectable increase in CD8 T cell-mediated antiviral activity following the two infusions in some, but not all, assays. As expected, however, in the absence of a latency reversing agent, no meaningful decline in the frequency of resting CD4 T cell infection was detected. HXTC therapy in ART-suppressed, HIV-infected individuals appears safe and well tolerated, without any clinical signs of immune activation, likely due to the low residual HIV antigen burden present during ART.

Keywords: HIV; cell therapy; latency.

Figure 1

Characterization of HXTC Products (A) HXTC products expanded to clinically relevant levels after two stimulations (n = 6). (B) All HXTC products produced IFNγ in response to HIV Gag, Neg, Pol (GNP) pepmix stimulation, as measured on ELISpot. (C) Product phenotying by flow to show T cells (CD45⁺, CD3⁺), CD4 T cells (CD3⁺CD4⁺), CD8⁺ T cells (CD3⁺CD8⁺), NK cells (CD3⁻CD56⁺CD16⁺), natural killer T (NKT) cells (CD3⁺CD56⁺CD16⁺), monocytes (CD45⁺CD14⁺), dendritic cells (CD3⁻CD83⁺, HLADR⁺), and K562 cells (CD45⁺CD562⁻, CD16⁻, CD56⁻, CD32⁺, CD83⁺). (D) HXTC products displayed minimal exhaustion on CD3⁺ T cells. (E) ⁵¹Chromium cytotoxicity assay shows specific lysis of HXTC products against autologous PHA blast target cells pulsed with no peptides or HIV peptides (gag, nef, pol peptide pools) at an E:T ratio of 20:1. Each data point represents the mean of three technical replicates. *p < 0.05 by Wilcoxon signed rank test.

Figure 2

Homing Markers of PBMCs and HXTCs PBMCs and HXTCs from three participants chosen based on cell availability were thawed, rested overnight, and stained for the described cell surface markers. Gating strategy is shown in (A). The % of live, CD8⁺ memory T cells for the indicated homing marker is shown for both PBMCs and HXTCs (B). Data points represent the mean of technical duplicates.

Figure 3

Antiviral Activity following HXTC Infusion (A–F) Viral inhibition assay against autologous activated CD4 T cells infected with JR-CSF and co-cultured with CD8 T cells at an effector:target ratio of 1:10 obtained from participants at baseline (averaged from 4 and 0 weeks before infusion) and 1, 2, 3, and 13 weeks post-infusion. HXTC-02 (A), HXTC-03 (B), HXTC-04 (C), HXTC-05 (D), HXTC-06 (E), HXTC-07 (F). Infusions occurred at weeks 0 and 2, denoted by arrows. Results are displayed as %p24 production, normalized to the p24 obtained in the absence of any CD8 effector cells. Error bars represent SEM of n = 4 replicates per time point per participant. Arrows denote HXTC infusion. *p < 0.05 by one-way ANOVA with Dunnett’s post-test correction for multiple comparisons.

Figure 4

HIV-Specific T Cell Responses following HXTC Infusion by IFNγ Release PBMCs isolated from leukapheresis products were stimulated with overlapping peptides spanning HIV Clade B Gag (A), Pol (B), or Nef (C) overnight in an ELISpot assay. Resulting IFNγ spot-forming units per 1 × 10⁵ PBMCs are displayed. Error bars represent SEM of n = 4 replicates.

Figure 5

Neither the Latent HIV Reservoir Nor Residual Viremia Decreased following HXTC Infusion (Left) QVOA was performed to determine the frequency of latently infected resting CD4 T cells using fresh resting CD4 T cells obtained from the baseline week 0 (pre-infusion) and week 13 (post-infusion). Results are presented as infectious units per million (IUPM) CD4 T cells. (Right) The number of copies of HIV-1 Gag RNA/mL of plasma is shown at the indicated time points. Gray shaded symbols indicate that results were below the limit of detection of the assay (values imputed were the threshold values).