Antibody signatures of HIV treatment effectiveness: toward low-cost rapid tests for treatment monitoring

About This Grant

Polymerase chain reaction (PCR) testing for HIV viral load is a mainstay of HIV treatment monitoring, but has limitations including high costs, long turn-around times, and limited information that only reflects viral load “in the moment.” This study explores the hypothesis that low-cost rapid antibody tests can complement HIV PCR, analogously to how hemoglobin A1c testing complements “in-the-moment” glucose testing. First, we will consolidate existing quantitative antibody data across ≥13 antibody-based assays and ≥17 longitudinal HIV treatment cohorts spanning 10 countries and all major HIV subtypes. We will develop regression-based and mechanistic viral dynamics models of antibody trajectories and their determinants, hypothesizing that mechanistic models will out-perform regression. We will also explore latent trajectory models that account for unobserved heterogeneity. For example, it is known that some clients are more adherent to treatment in the days leading up to clinic visits, motivated in part by a desire for positive interactions with healthcare providers. Undetected between-visit viral rebounds can lead to HIV transmission and adverse health effects, suggesting a role for tests detecting viral rebound over a longer retrospective time window. Next, we will select several of the most promising, low-cost, widely-available antibody assays to test on ≥4 long-term treatment cohorts spanning a range of HIV acquisition modalities and viral subtypes, and which include individuals on treatment for >10 years. These newly-generated data will be used to augment the dataset, prospectively validate the trajectory models, and formally analyze performance characteristics (receiver operating characteristic curves) of the assays, alone or in combination, predicting viral rebound over different retrospective time windows. We will determine which assays best detect current and past viral rebounds. Finally, we will conduct individual- and population-level modeling of HIV treatment monitoring strategies that incorporate antibody assays. At individual levels, we will assess health impact and cost-effectiveness when antibody assays augment, replace, or partially replace PCR. We will also model hypothetical performance characteristics in order to establish target product profiles for future assays. At population levels, we will model how antibody assays could augment HIV epidemic goals such as “95-95-95” (diagnosing ≥95% of people living with HIV, providing treatment to ≥95% of those diagnosed, and achieving undetectable viral load in ≥95% of those on treatment). A potential “fourth 95” could involve maintaining long-term viral suppression. Monitoring this “fourth 95” with antibody assays could make population-level HIV studies more affordable, feasible, and useful as 95’s approach 100’s. The impact of this research is both scientific and translational. Scientifically, we will develop novel datasets and models to enhance understanding of antibody trajectories during HIV treatment. Translationally, we will pave the way for potentially game-changing diagnostics to reduce HIV care costs, improve turn-around time and convenience, and provide richer information for people living with HIV.