Binding of ketoprofen enantiomers in various human albumin preparations.

Published data conflict with respect to the enantioselective protein binding parameters of R(-) and S(+) ketoprofen. We studied whether differences in experimental conditions used and/or presence of interfering compounds could provide a possible explanation for these discrepancies. Equilibrium dialysis, supported by ultrafiltration (67 mM Sörensen phosphate buffer pH 7.4, 580 microM HSA, 37 degrees C) allowed the characteristics of the binding sites to be determined according to Scatchard's analysis. (R) and (S)-ketoprofen concentrations were measured by HPLC. The free (R)-ketoprofen/free (S)-ketoprofen (F(R)/F(S)) concentration ratio was calculated. The effect of octanoic acid (OA) found in currently marketed intravenous HSA solutions, and hippuric acid (HA), on F(R)/F(S) concentration ratio was considered. Two classes of binding sites were characterized for both enantiomers. The free (S)-ketoprofen concentrations remained equal to those of the (R)-antipode at low concentrations of racemate (2-35 microg ml(-1)) indicating non-stereoselective albumin binding over the therapeutic range. From 35 microg ml(-1), the free (S)-ketoprofen concentrations were slighty greater than those of its antipode. Both OA and HA induced an increase of the free fraction of the enantiomers by a two-fold to a 15-fold order of magnitude. OA, but not HA, showed a more pronounced effect for the (S)-form leading to a marked decrease in F(R)/F(S) concentration ratio (0.61). Differences in HSA preparations used and/or the presence of interfering compounds may explain the variability in the reported protein binding characteristics of ketoprofen enantiomers.