Fast atom bombardment mass spectrometric determination of the molecular weight range of uremic compounds that displace phenytoin from protein binding: absence of midmolecular uremic toxins

Comparisons of Four Protein-Binding Models Characterizing the Pharmacokinetics of Unbound Phenytoin in Adult Patients Using Non-Linear Mixed-Effects Modeling

Background and objective

Phenytoin is extensively protein bound with a narrow therapeutic range. The unbound phenytoin is pharmacologically active, but total concentrations are routinely measured in clinical practice. The relationship between free and total phenytoin has been described by various binding models with inconsistent findings. Systematic comparison of these binding models in a single experimental setting is warranted to determine the optimal binding behaviors.

Methods

Non-linear mixed-effects modeling was conducted on retrospectively collected data (n = 37 adults receiving oral or intravenous phenytoin) using a stochastic approximation expectation–maximization algorithm in MonolixSuite-2019R2. The optimal base structural model was initially developed and utilized to compare four binding models: Winter–Tozer, linear binding, non-linear single-binding site, and non-linear multiple-binding site. Each binding model was subjected to error and covariate modeling. The final model was evaluated using relative standard errors (RSEs), goodness-of-fit plots, visual predictive check, and bootstrapping.

Results

A one-compartment, first-order absorption, Michaelis–Menten elimination, and linear protein-binding model best described the population pharmacokinetics of free phenytoin at typical clinical concentrations. The non-linear single-binding-site model also adequately described phenytoin binding but generated larger RSEs. The non-linear multiple-binding-site model performed the worst, with no identified covariates. The optimal linear binding model suggested a relatively high binding capacity using a single albumin site. Covariate modeling indicated a positive relationship between albumin concentration and the binding proportionality constant.

Conclusions

The linear binding model best described the population pharmacokinetics of unbound phenytoin in adult subjects and may be used to improve the prediction of free phenytoin concentrations.

Electronic supplementary material

The online version of this article (10.1007/s40268-020-00323-2) contains supplementary material, which is available to authorized users.

Usefulness of monitoring free (unbound) concentrations of therapeutic drugs in patient management

Drugs are bound to various serum proteins in different degrees and only unbound or free drug is pharmacologically active. Although free drug concentration can be estimated from total concentration, for strongly bound drugs, prediction of free level is not always possible. Conditions like uremia, liver disease and hypoalbuminemia can lead to significant increases in free drug resulting in drug toxicity even if the concentration of total drug is within therapeutic range. Drug-drug interactions may also lead to a disproportionate increase in free drug concentrations. Elderly patients may have increased free drug concentrations due to hypoalbuminemia. Elevated free phenytoin concentrations have also been reported in patients with AIDS and pregnancy. Currently free drug concentrations of anticonvulsants such as phenytoin, carbamazepine and valproic acid are widely measured in clinical laboratories. Newer drugs such as mycophenolic acid mofetil and certain protease inhibitors are also considered as candidates for monitoring free drug concentration.

Clinical utility of free drug monitoring

Most drugs are bound to serum proteins to a various degree. Only unbound or free drug is pharmacologically active. Usually total drug is measured for therapeutic monitoring because there is equilibrium between bound and free drugs, and concentration of free drug can be predicted from total drug concentration. However, under certain conditions this equilibrium is disturbed and the measured free drug concentration can be significantly higher than expected from total drug concentrations, especially for strongly protein-bound drugs. In such case a patient may experience drug toxicity even if the total drug concentration is within the therapeutic range. Conditions like uremia, liver disease and hypoalbuminemia can lead to significant increases in free drug concentration. Therefore, monitoring free phenytoin and free valproic acid is recommended in these patients. Drug-drug interactions can also lead to a disproportionate increase in free drug concentration. One strongly protein-bound drug can significantly displace another strongly protein-bound drug if both drugs share the same binding site. Several over-the-counter pain medications such as salicylate, naproxen, and ibuprofen can cause significant displacement of both phenytoin and valproic acid from albumin binding site. Interestingly, such interactions are absent in uremic patients. Elderly patients may have increased free phenytoin or valproic acid due to hypoalbuminemia. Elevated free phenytoin concentrations have also been reported in patients with AIDS. Although digoxin is 25% bound to protein, monitoring free digoxin is useful in patients with elevated endogenous digoxin-like immunoreactive substances or in patients overdosed with digoxin and being treated with digibind. Monitoring free digoxin can also eliminate interference of Chinese medicines Chan Su and Danshen in serum digoxin measurement by certain immunoassays. However, free drug monitoring is not a routine procedure in clinical laboratories due to technical difficulties and lack of established reference ranges for free drugs.

Uremic sera contain inhibitors that block digitoxin-valproic acid interaction

BACKGROUND

Digitoxin and valproic acid show strong binding to serum albumin. Thus, when present simultaneously in serum, digitoxin and valproic acid compete for binding sites. We studied digitoxin-valproic acid interaction in normal and uremic sera.

METHODS

Fluorescence polarization immunoassays were used for measuring total digitoxin and total valproic acid concentrations. We used a modified protocol of improved sensitivity to measure free digitoxin concentrations. We supplemented 2 normal and 2 uremic pools with digitoxin and then aliquots of these pools were further supplemented with various concentrations of valproic acid. After incubation at 37 degrees C for 2 hours in a water bath, specimens were allowed to re-equilibrate at room temperature for 20 minutes. Free digitoxin concentrations were measured. We also investigated digoxin-valproic acid interaction using 1 normal and 1 uremic serum pool.

RESULTS

We observed significant increases in free digitoxin concentrations in normal sera in the presence of valproic acid. In contrast, we observed a slight decline in free digitoxin concentration in the presence of valproic acid in uremic sera. We speculated that uremic sera contained inhibitors that block digitoxin-valproic acid interaction and identified indoxyl sulfate as an inhibitor. However, another uremic compound, hippuric acid showed no inhibitory effect. Interestingly, we observed no significant interaction between digoxin and valproic acid in either normal or uremic serum pool. This is probably because of poor protein binding of digoxin.

CONCLUSION

We conclude that valproic acid significantly displaces digitoxin from protein binding sites in normal serum. However, uremic sera contain inhibitors that block digitoxin-valproic acid interaction.

Sensitive methods for determination of free digitoxin concentration using digitoxin immunoassays: demonstration of elevated free digitoxin concentration caused by digitoxin-phenytoin interaction by applying these new techniques

Digitoxin is very strongly bound to serum albumin. Although free digitoxin is pharmacologically active, it is not monitored because of the lack of a sufficiently sensitive technique. The concentration of free digitoxin in the protein-free ultrafiltrate is usually below the detection limit of digitoxin immunoassays. A modified technique is described by which free digitoxin can be routinely monitored using commercially available immunoassays. The fluorescence polarization immunoassay for determining total digitoxin concentration requires that 100 microL of serum be treated with 300 microL of methanol to precipitate proteins. It is demonstrated that free digitoxin can easily be measured by adding 100 microL of methanol to 300 microL of ultrafiltrate, thus improving the sensitivity of the assay three-fold. The free digitoxin concentration can easily be calculated by dividing the observed value by 3. An attempt to use only ultrafiltrate (no methanol added) caused significant bias in the result, probably as a result of a matrix problem. The chemiluminescent assay for digitoxin does not require any specimen pretreatment and requires only 10 microL of serum. The program was modified and used 50 microL of ultrafiltrate to improve the sensitivity of the free digitoxin assay. If the chemiluminescent assay is used to measure free digitoxin, the true free digitoxin concentration can be calculated by dividing the observed value by 4.3. The free digitoxin concentrations were comparable in eight patients receiving digitoxin as measured by both methods. To show an application of this technique, two serum pools were prepared from patients receiving digitoxin and supplemented with various concentrations of phenytoin. A significant increase in free digitoxin concentration was observed because of the displacement of digitoxin from protein binding sites by phenytoin.

Elevated free fosphenytoin concentrations in uremic sera: uremic toxins hippuric acid and indoxyl sulfate do not account for the impaired protein binding of fosphenytoin

Fosphenytoin is a new phosphate ester prodrug of phenytoin. Impaired protein binding of phenytoin in uremia has been extensively documented, which prompted us to investigate the protein binding of fosphenytoin in uremic sera. Also studied was the role of uremic toxins hippuric acid and indoxyl sulfate as potential inhibitor of the protein binding of fosphenytoin because these compounds impair protein binding of phenytoin in uremia. Five serum pools were prepared from normal volunteers and five pools from patients with uremia. None of them received phenytoin. The normal serum pools were diluted with saline to mimic the albumin concentration of uremic pool. Both the diluted normal pool and the uremic pool were supplemented with fosphenytoin; after incubation at room temperature for 30 minutes, total and free fosphenytoin concentrations as phenytoin equivalents were measured using fluorescence polarization immunoassay (Abbott Laboratories; Abbott Park, IL, U.S.A.). The authors observed significantly elevated free fosphenytoin concentration in uremic sera compared with that of normal sera in all cases. Because both normal and uremic sera had the same concentrations of albumin, the elevated free fosphenytoin concentration in uremic sera was not caused by hypoalbuminemia. Both indoxyl sulfate and hippuric acid cause significant displacement of phenytoin from protein binding. In contrast, none caused any displacement of fosphenytoin from protein binding.

Interaction of valproic acid with nonsteroidal antiinflammatory drugs mefenamic acid and fenoprofen in normal and uremic sera: lack of interaction in uremic sera due to the presence of endogenous factors

Valproic acid is an anticonvulsant that is strongly bound to serum albumin. The nonsteroidal antiinflammatory drugs mefenamic acid and fenoprofen are also strongly bound to albumin. We observed significant displacement of valproic acid from protein binding by mefenamic acid and fenoprofen at both therapeutic and slightly above therapeutic concentrations. The concentration of free valproic acid was higher in the uremic serum, as expected, but we observed no further displacement of valproic acid in the presence of mefenamic acid and fenoprofen. Known uremic compounds, hippuric acid and indoxyl sulfate, did not inhibit the interactions between valproic acid and mefenamic acid or fenoprofen. Treatment of uremic serum with activated charcoal at pH 3.0 removes endogenous interfering factors and corrects the binding defect of uremic serum for valproic acid. We observed significant displacement of valproic acid from protein binding by both mefenamic acid and fenoprofen in uremic serum after charcoal treatment. We conclude that endogenous factors are present in uremic sera that block interaction of valproic acid with mefenamic acid and fenoprofen.