Optimization of antiretroviral therapy for HIV infected patients by simultaneous analysis of immune restoration and serious side effects

Abstract

The development of potent antiretroviral drugs has substantially improved life expectancy and quality of life for HIV infected patients. However, current HIV drugs do not induce complete viral eradication, requiring long periods of treatment. Continuous administration of antiretroviral drugs has led to serious drug toxicity and side effects resulting in forced therapy cease and consequent viral rebound. As a result, clinicians encounter an optimization problem: how to best control viral replication while maintaining low antiretroviral drug toxicity levels. In this thesis, the HIV drug dosage problem is solved based on a continuous time model predictive approach for nonlinear systems with previous engineering applications. HIV infection is modeled by two distinct mathematical models, while serious side effects are represented by a simple mechanism for drug toxicity that is based on a gradual liver dysfunction due to drug therapy.