Dynamics of HIV treatment and social contagion

Abstract

Modern-day management of infectious diseases is critically linked to the use of mathematical models to understand and predict dynamics at many levels, from the mechanisms of pathogenesis to the patterns of population-wide transmission and evolution. This thesis describes the development and application of mathematical techniques for HIV infection and dynamics on social networks. Treatment of HIV infection has improved dramatically in the past few decades but is still limited by the development of drug resistance and the inability of current therapies to completely eradicate the virus from an individual. We begin with a synthesis of the important evolutionary principles governing the HIV epidemic, emphasizing the role of modeling. We then describe a modeling framework to study the emergence of drug-resistant HIV within a patient. Our model integrates laboratory data and patient behavior, with the goal of predicting outcomes of clinical trials. Current results demonstrate how pharmacologic properties of antiretroviral drugs affect selection for drug resistance, and can explain drug-class-specific resistance risks. Thirdly, we describe models for a new class of drugs that aim to eliminate cells with latent viral infection. We provide estimates for the required efficacy of these drugs and describe the potential challenges of future clinical trials. Finally, models and mechanisms for understanding viral dynamics are increasingly finding applications outside traditional virology. They can be used to study the dynamics of behaviors, to help predict and intervene in their spread. We describe techniques for applying infectious disease models to social contagion, drawing on techniques for network epidemiology. We use this framework to interpret data on the interpersonal spread of health-related behaviors.