Enhanced HIV-1 Control After Antibody Therapy is Associated with Autologous Antibodies and Reservoir Clearance in the RIO trial

Technical Summary: Enhanced HIV-1 Control After Antibody Therapy is Associated with Autologous Antibodies and Reservoir Clearance in the RIO Trial

Problem Statement
Despite the efficacy of antiretroviral therapy (ART) in suppressing HIV-1 replication, a latent reservoir of integrated proviruses persists primarily in resting memory CD4+ T cells, preventing a cure. Upon ART interruption, these reservoirs typically lead to rapid viral rebound. While broadly neutralizing antibodies (bNAbs) have shown promise in delaying rebound and inducing post-treatment control in small studies, the mechanisms driving this control and the impact of bNAb therapy on reservoir dynamics in humans remain incompletely understood. Specifically, it is unclear how baseline reservoir characteristics, autologous antibody responses, and the selection of resistant variants interact to determine the duration of ART-free control.

Methodology
This study reports pre-specified secondary and exploratory analyses from the RIO trial, a double-blind, randomized, placebo-controlled phase II study. The trial enrolled 68 participants who initiated ART during primary or early HIV infection (defined by a nadir CD4+ T-cell count >500 cells/µL). Participants were randomized 1:1 to receive two long-acting bNAbs, 3BNC117-LS and 10-1074-LS (Arm A), or saline (Arm B) two days before an analytical treatment interruption (ATI).

The study employed a multi-faceted analytical approach:

• Reservoir Quantification: The size and composition of the HIV-1 reservoir were measured at baseline and follow-up using digital droplet PCR (ddPCR) and quadruplex 4-probe real-time quantitative PCR (Q4PCR) to distinguish between intact and defective proviruses.
• Genomic Sequencing: Intact proviral sequences were recovered via whole-genome amplification and sequencing (Q4PCR), while rebound viruses were sequenced using single-genome amplification (SGA). Phylogenetic trees were constructed to compare reservoir and rebound viral populations.
• Neutralization Assays: Pseudotyped viruses derived from baseline reservoir sequences and rebound viruses were tested for sensitivity to the infused bNAbs (3BNC117 and 10-1074) and to autologous IgG antibodies purified from participants' baseline serum.
• Longitudinal Monitoring: Viral loads and bNAb concentrations were monitored throughout the ATI period. Time-to-event analyses (Cox proportional hazards models) were used to correlate baseline reservoir sensitivity with the time to ART restart.

Key Results

• Clinical Outcomes: Consistent with the primary trial results, Arm A participants experienced significantly delayed viral rebound compared to the placebo group (Arm B). At 96 weeks, 24% of Arm A participants remained off ART compared to 6% in Arm B.
• Reservoir Characteristics: Baseline reservoir sizes were small and comparable between arms, consistent with early ART initiation. Intact proviruses showed high clonality (median 57%), yet there was no significant correlation between baseline reservoir size and the time to ART restart.
• Viral Rebound and Resistance: In Arm A, rebound viruses showed significant selection for resistance to 10-1074 (geometric mean IC80 increased from 0.09 to 14.7 µg/ml), whereas resistance to 3BNC117 was less pronounced. In contrast, Arm B participants showed no significant shift in bNAb sensitivity between reservoir and rebound viruses.
• Role of Autologous Antibodies: A significant correlation was observed in Arm A between higher baseline autologous neutralizing activity against reservoir viruses and a longer time to ART restart. Similarly, greater baseline sensitivity to 10-1074 correlated with delayed rebound. Participants with pre-existing autologous neutralizing activity who received bNAbs maintained control for a median of 108 weeks, compared to 27.5 weeks for those without such activity.
• Reservoir Dynamics: In Arm A, the intact proviral reservoir showed a significant decrease over time (estimated half-life of 36 weeks) compared to baseline, whereas no measurable change was observed in Arm B. This decay was specific to intact proviruses; defective proviruses remained stable.
• Fluctuating Viremia: A distinct phenotype of "fluctuating viremia" (sustained low-level viremia without meeting ART restart criteria) was observed in 38% of Arm A participants. These individuals harbored viruses resistant to 10-1074 and autologous antibodies but retained sensitivity to 3BNC117.

Significance and Claims
The authors conclude that the RIO trial data establishes that a significant fraction of people living with HIV-1 who receive bNAb therapy can partially control viremia for prolonged periods. The study identifies pre-existing autologous neutralizing antibodies and baseline reservoir sensitivity to 10-1074 as key correlates of prolonged control.

The paper posits that prolonged control is likely driven by a combination of factors:

1. Combined Antibody Pressure: Pre-existing autologous antibodies may act as an adjunct to therapeutic bNAbs, creating a higher barrier for viral escape.
2. Reservoir Clearance: The observed accelerated decay of the intact reservoir in the bNAb arm suggests that prolonged exposure to LS-bNAbs may enhance the clearance of infected cells, potentially through mechanisms such as antibody-dependent cellular cytotoxicity (ADCC) or enhanced antigen presentation, although the study notes that reservoir decay alone does not fully explain the delay in rebound.
3. Selection Dynamics: The emergence of resistance to 10-1074 but not 3BNC117 suggests that rebound often occurs under functional monotherapy conditions when 3BNC117 concentrations become subtherapeutic.

The authors emphasize that while the associations are correlative, the findings support the hypothesis that enhancing host immunity (via vaccination or other means) alongside bNAb therapy could increase the proportion of individuals achieving post-treatment control. The study also highlights "fluctuating viremia" as a reproducible virologic phenotype following bNAb therapy, distinct from complete suppression or rapid rebound.

Limitations
The authors acknowledge limitations including the inability to sequence the reservoir in all participants due to small reservoir sizes, the restriction of sampling to peripheral blood (which may not reflect tissue reservoirs), and the correlative nature of the findings regarding autologous antibodies. Additionally, the study was not powered for sex-based subgroup analyses.