The displacement of phenylbutazone-14C and warfarin-14C from human albumin by various drugs and fatty acids

Valproic acid binding to human serum albumin and determination of free fraction in the presence of anticonvulsants and free fatty acids

The interaction between valproic acid (VPA) and human serum albumin (HSA) was investigated using the equilibrium dialysis technique under various conditions. Solutions of VPA in HSA (2 x 10(-4) M) were dialyzed against isotonic phosphate buffer at 37 degrees C. Protein and buffer compartments were assayed for VPA by GLC. The free fraction (alpha) of VPA increased from 0.13 at 27 microgram/ml to 0.49 at 103 microgram/ml. Scatchard plots were linear, indicating the existence of one type of binding site. The mean (+/- % SD) number of binding sites per macromolecule was 2.06 +/- 3.7% and the mean (+/- % SD) association constant was 2.69 x 10(4) +/- 15.0% liters/mole. The effects of three anticonvulsants (phenytoin, phenobarbital, and carbamazepine) and four major free fatty acids (FFA) (stearic, palmitic, oleic, and linoleic) on alpha were studied. The free fraction, 0.18, was not affected by phenobarbital (20 and 40 microgram/ml), carbamazepine (10 and 20 microgram/ml) or phenytoin (20 and 40 microgram/ml). Each of the four FFA caused a significant increase in alpha: 19--48% increase at 100 microgram/ml of FFA and 88--118% at 200 microgram/ml.

The fatty acid content and drug binding characteristics of commercial albumin preparations

The fatty acid contents of thirteen commercial preparations of human albumin were found to be in the range 0.03 to 9 mol of fatty acid/mol albumin. Marked differences were found between the preparations in the binding of the fluorescent probes, 5-dimethylamino-naphthalene-1-sulfonamide (DNSA) and dansylsarcosine. The displacement of these probes by ibuprofen and phenylbutazone also showed marked differences between preparations. The differences between the commercial albumin samples correlated well with their fatty acid contents and were abolished by treatment with charcoal. They were similar to the changes observed when oleic acid was added to fatty acid free albumin. The source and fatty acid content of commercial albumin preparations should be considered when comparing studies of the binding of drugs to human albumin.

The effect of bezafibrate on the fibrinolytic enzyme system and the drug interaction with racemic phenprocoumon

The influence of a new hypolipidaemic agent, bezafibrate, on anticoagulant requirements and fibrinolysis was studied in 15 patients with hyperlipidaemia on long-term treatment with racemic phenprocoumon. Our results suggest a dose-dependent augmentation of the anticoagulant response to the coumarin drug. Treatment with bezafibrate at 450 and 600 mg daily required a reduction of the phenprocoumon dose by 18.5 and 33.5%, respectively. Correspondingly, the serum level of phenprocoumon decreased by 11.6 and 35.3%. No evidence for an altered drug elimination of racemic phenprocoumon could be found during treatment with bezafibrate. The results support the hypothesis that bezafibrate and analogous hypolipidaemic drugs enhance the response to oral anticoagulant drugs by increasing the affinity of the receptor site for coumarins or the rate of degradation of the vitamin-K-dependent clotting factors. The investigation of the fibrinolytic enzyme system demonstrated an increase of the fibrinolytic activity by enhancing the activity of the plasminogen activator. The lysis time for euglobulin clot was reduced significantly, plasma fibrinogen only moderately. The antiplasmin activity could not be altered substantially by a decrease of alpha1-antitrypsin and a slight increase of alpha2-macroglobulin. In contrast with the inhibition of platelet function the effect of bezafibrate on the fibrinolytic enzyme system showed no dose dependence.

Renal function, protein binding and pharmacological response to diazoxide

1 The effect of rapid (10s) injections of diazoxide was studied in ten hypertensive patients with varying degrees of impairment of renal function. 2 There was a significant correlation between the patient's plasma urea concentration and reduction in mean arterial blood pressure. Diazoxide was also shown to be less highly protein bound in patients with renal failure. 3 It is suggested that the explanation for the increased hypotensive effect of diazoxide observed in patients with reduced renal function is related to higher unbound drug concentrations.

Lack of relationship between serum free fatty acids and impaired plasma protein binding of diphenylhydantoin in chronic renal failure

The plasma protein binding of diphenyldantoin (DPH) in 13 patients with varying degrees of renal failure was considerably less than in normal healthy subjects confirming early studies. The fraction of unbound DPH was correlated with serum creatinine (r = 0.81, p less than 0.001), but there was no significant correlation with the serum concentration of free fatty acids, triglycerides or cholesterol.

Drug interactions and clinical pharmacokinetics

Some drugs influence the gastro-intestinal absorption, distribution , metabolism or renal excretion of other drugs, i.e., processes involved in pharmacokinetic interactions. The clinical consequences of pharmacokinetic drug-drug interactions will be either an increase or a decrease in known therapeutic or toxic effects of the interacting drug. In order to evaluate the importance of drug interaction affecting gastro-intestinal absorption, it is necessary to distinguish between interactions which alter the rate of absorption of another drug and those which alter the amount of drug absorbed. Many drugs displace other drugs from their protein binding sites in vitro. This may cause an increase in the pharmacological effect of the displaced drug. However, much discrepancy exists between in vitro findings. In some cases, the enhanced effect only seems to be a temporary phenomenon. The degree of protein binding and the size of apparent volume of distribution (Vd) must also be taken into consideration. Perhaps the importance of interaction involving protein binding has been overemphasized. Barbiturates, glutethimide, rifampicin and phenytoin increase the rate of drug metabolism in man. The most important interactions reported are between oral anti-coagulants and barbiturates. After withdrawal of these hepatic microsomal enzyme inducing drugs, it takes 2 to 3 weeks before the rate of drug metabolism reaches the pretreatment level. In this period, risk of haemorrhage exists. Induction seems to be dose-dependent, but not all persons are inducible. Many drugs compete for the drug metabolising enzyme system in the liver and consequently some drugs inhibit the biotransformation of other drugs. The time course of these interactions depends on the pharmacokinetic properties of the drug involved, and these interactions also seem to be dose-dependent. The most important of such interactions, clinically involved the oral sulphonylurea hypoglycaemic drugs and the antiepilepic drug phenytoin. Drugs are eliminated by urinary excretion through three mechanisms: glomerular filtration, tubular reabsorption, and active tubular secretion. The most important interactions seem to be those involving competition for tubular secretion.

The effects of allopurinol and clofibrate on the elimination of coumarin anticoagulants in man

The effects of chronic treatment with allopurinol and clofibrate on the elimination from plasma of warfarin and dicoumarol were examined in man. In addition, the binding of warfarin to human serum proteins was measured with and without clofibrate and its metabolite chlorophenoxyisobutyric acid (CPIB), both present in therapeutic concentrations. Treatment with allopurinol and clofibrate did not alter the rate of elimination of warfarin from plasma. In addition, clofibrate and CPIB caused no significant displacement of warfarin from serum proteins. This evidence supports the conclusion that the clinically significant potentiation of warfarin activity by clofibrate in man is due to an interaction at the receptor site. In contrast, treatment with allopurinol resulted in significant prolongation of the plasma half-life of dicoumarol in one of three subjects. These data are consistent with inhibition of the metabolism of dicoumarol by allopurinol in some individuals.

Failure of indomethacin and warfarin to interact in normal human volunteers

In young, normal, male Caucasian volunteers, two different double-blind, placebo-controlled, randomized studies were performed to determine whether indomethacin interacts in a clinically significant fashion with warfarin. In the first study, stabilized hypoprothrombinemia was achieved by daily oral warfarin administration for at least 11 days; then indomethacin was administered concurrently with warfarin for five more days. During this five-day period failure of indomethacin to alter the degree of hypoprothhrombinemia induced by warfarin indicated that indomethacin does not interact with warfarin in vivo in a clinically significant fashion. A similar conclusion was drawn from results of the second experiment, in which plasma warfarin concentrations and prothrombin times were measured after a single oral dose of warfarin administered both before and after ten consecutive days of oral indomethacin administration. Thus, these studies suggest that warfarin may be given to patients receiving indomethacin and indomethacin may be administered to patients on warfarin without intensification or diminution of the normally expected hypoprothrombinemie effects of the dose of warfarin in that individual.

Spectroscopic techniques in the study of protein binding. A fluorescence technique for the evaluation of the albumin binding and displacement of warfarin and warfarin-alcohol

1. The binding of racemic mixtures of warfarin and warfarin-alcohol to human serum albumin (HSA) is accompanied by an increase in the fluorescence quantum yield of these compounds. This property has been used to measure the characteristics of the binding of warfarin and warfarin-alcohol to HSA at 22 degrees C and 37 degrees C. Within the limits of the technique, no significant differences between the number of binding sites and strength of binding at the tight site at either temperature were observed. 2. The fluorescence of warfarin and warfarin-alcohol was used to label their binding site on HSA and to study the effects of other drugs on their binding. The results indicate that these two molecules are bound to the same site on HSA. 3. The validity of using changes in the fluorescence of warfarin as a measure of its displacement from HSA was investigated. Good correlations were observed between drug-induced decreases in the fluorescence of bound warfarin and displacement as measured by equilibrium dialysis. The displacement of warfarfin, as detected by fluorescence, correlates well with the increase in free warfarin resulting from addition of therapeutic drug concentrations to undiluted human serum. 4. The most potent displacing agents, by all the methods used, were iophenoxic acid, phenylbutazone and oxyphenylbutazone. The first of these is no longer used clinically, but the latter two are and have been reported to cause hypoprothrominaemia by displacing warfarin from HSA. The present study indicates that changes in the fluorescence of warfarin bound to HSA can be used to measure displacement of bound warfarin and to screen drugs that may cause clinically significant interactions by this mechanism.

Benzodiazepines: specific competitors for the binding of L-tryptophan to human serum albumin

By means of the gel filtration technique, the effect of nine benzodiazepine derivatives on the binding of L-tryptophan to human serum albumin was investigated. Using equimolar tryptophan and benzodiazepine concentrations, all benzodiazepines with binding constants higher than 10(4) (M(-1), displace L-tryptophan from its binding site to a high degree. The mechanism of the displacement was characterized as a competition for a common binding site. Some of the benzodiazepines displace L-tryptophan to a greater extent than salicylic acid. The benzostereospecific binding to human serum albumin. This study shows that there is only one binding site on the human serum albumin molecule, which binds tryptophan and the benzodiazepines in a highly stereospecific manner. Therefore it is concluded that the benzodiazepines and L-tryptophan must have similarities in their molecular structure, so that both can bind to the common binding site in such specific manner. These considerations are discussed in regard to the known influence of benzodiazepine derivatives on the L-tryptophan metabolism in brain. A direct involvement of the reported displacement in the pharmacological actions of the drugs seems not to be relevant because of their small therapeutical plasma levels.