Protein binding in antiretroviral therapies.

There is marked variability in the extent to which the three classes of antiretroviral (ARV) drugs bind to plasma proteins (<5 to >99%). Protease inhibitors (PIs), with the exception of indinavir, are more than 90% protein bound, mainly to alpha1-acid glycoprotein (AAG). Efavirenz, a nonnucleoside reverse transcriptase inhibitor (NNRTI), is more than 99% bound, mainly to albumin. Nucleoside reverse transcriptase inhibitors (NRTIs) are not highly protein bound. The pharmacological activity of ARV drugs is dependent on unbound drug entering cells that harbor the human immunodeficiency virus (HIV). There has been concern that changes in protein binding could impact on antiviral activity and management. However, for PIs and NNRTIs, and for many drugs given orally, altered plasma binding would not be expected to influence the average exposure to unbound (active) drug after chronic oral dosing. Nevertheless, there will be a change in the relationship between total and unbound concentrations that will be important if, as part of therapeutic drug monitoring, the total rather than the unbound drug is measured. Measuring drug concentrations that are needed to inhibit different HIV strains (wild type and drug resistant) in vitro could also cause confusion because most methods employ bovine serum in the assay medium, and unbound concentrations are not directly measured. Estimating unbound drug concentrations in human plasma and in incubation media can be highly method dependent and thus may affect the calculated IC50 (the concentration of drug that results in 50% inhibition of viral replication). Because inhibitory quotients (IQs = C(trough)/IC50) are becoming part of pharmacokinetic/pharmacodynamic (PK/PD) analyses of clinical trial data, the strengths and weaknesses of the methods used for the determination of unbound drug concentration in plasma and in vitro systems--ultracentrifugation, ultrafiltration, and equilibrium dialysis--need to be understood. Consensus on standard procedures must be reached. In June 2002, a panel of experts assembled by the Forum for Collaborative HIV Research met in Washington, DC, to review the basic principles of protein binding of ARV drugs, and to discuss the impact that changes in plasma protein binding may have on the PKs and activity of ARV drugs as well as on therapeutic drug monitoring. The purpose of the meeting was to discuss the following topics: (1) basic principles of protein binding and how changes in binding can impact on drug PKs and drug exposure in vivo, (2) variability in plasma protein binding among patients taking ARV drugs, (3) the impact of HIV infection and concomitant diseases on the extent of plasma protein binding, (4) the likelihood of clinically relevant drug interactions at the level of plasma protein binding, (5) the evidence that measuring unbound concentrations of ARV drugs in the plasma of patients gives more meaningful information than total drug concentration and, therefore, should be considered in routine therapeutic drug monitoring of ARV agents, (6) optimal method(s) for measuring the unbound concentration of drugs in vitro (for IC50 determination) and in vivo, and (7) future studies that need to be considered to fully understand the importance of plasma protein binding in therapeutic drug monitoring. This report summarizes the topics discussed at this meeting. It guides the reader through the discussions that allowed the panel to formulate a series of statements regarding the significance of plasma protein binding of ARV drugs when studied in vitro and in vivo. The roundtable participants also identified research priorities that are important for understanding the sources of inter- and intraindividual variability in protein binding in patients. These include obtaining data on unbound as well as on total concentrations in PK studies; looking at variants of AAG and whether they differ in binding affinity; and emphasizing the importance of developing a standard procedure for drug susceptibility assays used to determine IC50 values.