Accepted_test

Objective: It remains unclear why HIV persists in most untreated individuals, and why a small minority of individuals can control the virus, either spontaneously or after an early treatment. Striking differences have been discovered between patient cohorts in CD4 T cell avidity but not in CD8 T cell avidity. The present work has the aim to explain the diverse outcome of infection and identify the key virological and immunological parameters predicting the outcome. Design and Method: A mathematical model informed by these experiments and taking into account the details of HIV virology is developed. Results: The model predicts an arms race between viral dissemination and the proliferation of HIV-specific CD4 helper cells leading to one of two states: a low-viremia state (controller) or a high-viremia state (progressor). Helper CD4 cells with a higher avidity favor virus control. The parameter segregating spontaneous and post-treatment controllers is the infectivity difference between activated and resting CD4 T cells. Conclusion: The model is shown to have a better connection to experiment that a previous model based on T cell “exhaustion”. Using the model informed by patient data, the timing of antiretroviral  therapy can be optimized.