Modeling how memory CD8 T cells can elicit post-treatment control of HIV infection

While most people living with HIV suffer progressive disease following cessation of antiretroviral therapy, a small fraction elicits lasting post-treatment control. Understanding the mechanisms underlying this control is key to devising effective HIV remission strategies. Although recent studies implicate memory CD8 T cells, how these cells establish lasting viremic control remains unknown. Here, we combine mathematical modeling and analysis of data from SIV-infected non-human primates to elucidate the underlying mechanisms. We recognized that sustained antigenic stimulation leads to heritable epigenetic remodeling of the CD8 T cell pool, impairing memory cell survivability. Antiretroviral therapy rapidly suppresses viremia, thereby arresting antigenic stimulation and preserving memory potential. The greater this preservation is, the better would be the memory recall response following viral rebound post-treatment. Our mathematical model based on this hypothesis predicts that post-treatment control is an alternative steady state to progressive infection, realized by strong memory-driven recall responses. Our model fits longitudinal virological data spanning the pre-, during-, and post-antiretroviral treatment phases of infection, and recapitulates the outcomes of progressive disease and long-term remission realized, the latter predominantly with early treatment initiation. It shows, consistently with data, that memory CD8 T cells could drive post-treatment control independently of the size of the latent reservoir, explaining how such control may be realized more widely than estimated with prevalent hypotheses. Our model further explains the existence of a window of treatment initiation times that maximizes the chances of post-treatment control. Finally, model predictions inform interventions targeting memory CD8 T cells for HIV remission.