The impact of pharmacodynamic parameters on prediction of clinical outcome in HIV-1 infected patients and HCV patients with HIV co-infection

The objective of this project was to use pharmacometric approaches to investigate the impact of in vivo pharmacodynamic (PD) parameters on clinical outcome in patients infected with human immunodeficiency (HIV) or hepatitis C (HCV) virus. Pharmacometrics is an emerging science based on developing and applying mathematical/statistical methods to characterize and predict pharmacokinetic (PK) and PD behavior, and to quantify uncertainty inherent in information about that behavior.;The degree to which viral infection is inhibited at a given antiviral drug concentration depends on the IC50, the concentration inhibiting replication by 50%, and the slope of the relationship between drug effectiveness and concentration (gamma). In vitro experiments demonstrated that each antiretroviral (ARV) class has a characteristic gamma associated with inhibition of HIV replication. However, whether gamma is simply a shape factor that improves description of PK/PD data, or a fundamental parameter that can characterize in vivo efficacy, was not known. The first set of studies used integrated population PK-PD/HIV viral dynamic models to reveal that higher gamma values are associated with better clinical outcome (larger log10 viral load decline; higher proportion of patients with undetectable viral load). However, the impact of gamma became insignificant upon emergence of drug resistance. These studies also demonstrated that inclusion of gamma improved PD models, although accurately estimating inter-individual variability in gamma with a short-term monotherapy study design in HIV-infected patients was not possible. The second set of studies used integrated population PK-PD/HCV viral dynamic models to explore short-term clinical outcome associated with pegylated interferon (PEG-INF)-based treatment. These studies demonstrated no difference in gamma between patients who attained sustained virologic response (SVR) and those who did not (NR). In contrast, estimates of IC 50 and death rate for infected cells (delta) were significantly different between SVRs and NRs. Experiments also demonstrated that long-term HCV treatment outcome can be predicted using a population PK-PD/HCV viral dynamic model based on data obtained from a two-week PEG-INF-based treatment. Taken together, the results of this dissertation project indicate that pharmacometric approaches are useful in revealing the impact of in vivoPD parameters on clinical outcome in HIV and HCV infection.