(709c) Broadly Neutralizing Antibodies Enhance Clearance of HIV-Infected Cells In Vivo: Pharmacodynamic Analysis of Viral Load Kinetics with Validation in Humanized Mice

The primary role of antibodies typically involves neutralizing free virus particles in circulation by binding to specific regions of their cognate antigen. In the case of HIV, the causative agent of AIDS, the antigen is the envelope (Env) glycoprotein found on the surface of the virus particles. Broadly neutralizing antibodies (bNAbs), such as 3BNC117, are capable of neutralizing a large diversity of mutational variants of Env, leading to a rapid but transient reduction in viral load when passively administered. In vitro experiments have previously demonstrated the ability of antibodies to not only block new infection and reduce circulating virus concentrations, but also to engage particular classes of host immune effector cells by recognizing Env found on the surface of infected cells. This interaction is mediated by receptors on the host cell surface that bind to constant (Fc) regions on the antibodies. The host cells then mediate the death of these infected cells through mechanisms such as antibody-dependent cellular cytotoxicity (ADCC) or antibody-dependent cellular phagocytosis (ADCP). However, this interaction has not previously been demonstrated in vivo.

This work represents the first evidence of enhanced infected-cell clearance in vivo. Based on results from the first-in-man dose escalation phase 1 clinical trial of the bNAb 3BNC117, we found that the antibody kinetics upon passive administration are well-described by a simple two-compartment model where antibody can transport between blood plasma and lymph tissue (LT) compartments. We then coupled this transport model, with parameters inferred from healthy individuals, to a popular model of viral infection dynamics, including birth of virus from infected cells (which reside primarily in LT), and transport of virus between plasma and LT. This coupling took the form of an additional kinetic term corresponding to antibody binding free virus particles, leading to their subsequent elimination from the system. This simple model was unable to recapitulate the observed kinetics of viral load decay; in particular, the time scale and extent of decay could not both be fit simultaneously. However, by adjusting our model to include a mechanism whereby antibody causes death of infected cells, we substantially improved the fit to the viral load data because it causes a higher-order decay in the viral load over a longer time scale. Thus, the clinical patient data can not be explained when antibody exclusively neutralizes free virus particles. Motivated by these findings, our experimental collaborators demonstrated in a humanized mouse model that the clearance rate of infected cells is indeed increased, via a mechanism that involves engagement of Fc receptors (FcRs). These results have implications for passive immunotherapy (treatment or prevention) targeted against HIV-infected cells.