Background: Binding of antiretrovirals to plasma proteins is a factor that has been associated with lack of efficacy of some drug candidates. Most HIV-protease inhibitors (PIs) bind to a₁-acid glycoprotein (AAG). The effect of this binding is quantified by measuring the activity of the inhibitors against wild-type HIV-1 in the presence of AAG. Because the kinetics of the binding of PIs to AAG show a saturable interaction, the influence of AAG may be overestimated as it is determined at concentrations (wild type EC₅₀) where the inhibitor free active fraction is low. Therefore we measured the influence of AAG on the anti-HIV potency of PIs at higher, physiologically achieved concentrations by using virus strains with different levels of resistance to each of them.

Methods: The influence of AAG, human serum albumin (HSA), or human serum (HS) on the activity of PIs has been measured in a cell-based antiviral assay. Virus strains resistant to PIs have been obtained by the recombinant virus technology used in the Antivirogram assay.

Results: The anti-HIV activity of several PIs and the drug candidate TMC114 has been assessed in the presence of AAG (1 mg/mL), HSA (45 mg/mL) or HS (50%). With the exception of indinavir, which remains unaffected, all tested PIs show a decrease in potency against wild-type HIV-1 in the presence of AAG and HS, but not of HSA. The decrease ranges from 5- to 75-fold and is proportional to the AAG concentration. Virus strains with various levels of resistance against each of the PIs have been used in similar experiments with AAG. The results show that the decrease in potency observed in the presence of AAG for the tested PIs is inversely proportional to the EC₅₀ in the absence of the protein. At micromolar concentrations (1-5 µM), saquinavir, ritonavir, nelfinavir, amprenavir and TMC114 show only a less than 5-fold decrease in potency in the presence of 1 mg/mL AAG.

Conclusions: Binding of PIs to human plasma protein, and more precisely to AAG, is a factor often taken into account when determining the target trough levels for new drug candidates. We here show, however, that the influence of AAG decreases with increasing concentrations of drug, consistent with micromolar equilibrium dissociation constants for PIs binding to AAG. Hence at physiologically achieved, low micromolar plasma drug concentrations, binding to plasma proteins is a far less relevant factor to include in IQ calculations for the PIs tested than presumed before. Combination of this observation with the unique antiviral profile of TMC114 against multi-PI resistant HIV variants provides a further rationale for developing this compound also for use in PI-experienced patients.