Abstracts

Rationale: Phenytoin has a narrow therapeutic window and interpretation of phenytoin serum levels are complicated by the drug s high degree of protein binding. Only 10-30% of the total serum phenytoin measured in the body circulates as the free unbound pharmacologically active drug. In normal healthy adults, monitoring of total serum phenytoin levels suffice. However, due to altered protein binding pharmacokinetics, total serum phenytoin levels do not accurately represent therapeutic phenytoin levels in the elderly, malnourished, and critically ill. Various equations to estimate a corrected total serum phenytoin level in these patients have been proposed. The objective of our study was to determine the percent of phenytoin protein-binding in our veteran population and to determine whether various equations used to estimate a corrected total serum phenytoin levels provide an accurate estimate in our population.Methods: This was a retrospective chart review. Patients with documented free and total phenytoin levels taken at VA West Los Angeles Medical Center in 2007 were identified. Percent phenytoin protein binding was determined by the following equation: % Bound Phenytoin = 100 - (total phenytoin/free phenytoin). The following equations were used to estimate a normalized total serum phenytoin level were evaluated: 1) C(normal) = [C(observed)/(0.1 x albumin + 0.1)]; 2) C(normal) = [C(observed)/(0.2 x albumin + 0.1)]; 3) C(normal) = [C(observed)/(0.25 x albumin + 0.1)]; where C(normal) = a normalized serum phenytoin concentration that would be expected in non-hypoalbuminemic patients and where C(observed) = the observed serum phenytoin concentration in hypoalbuminemic patientsResults: 210 sets of free and total phenytoin levels were evaluated. 42% of patients were greater than 65 years of age. 59% of patients were classified as hypoalbuminemic with albumin levels less than 3.2 g/dL. Over 50% of patients were hospitalized in the intensive care unit (ICU) at the time of phenytoin level collection. The mean percent phenytoin binding in our patient population overall was 83.6%. Further subgroup analysis showed decreased phenytoin binding in the renally impaired (79%), elderly (81%), hypoalbuminemic (82%), and critically ill (79%). Normal protein binding was observed in a small group of patients who were ambulatory and who had normal albumin levels. Equations 1, 2 and 3 did not estimate corrected total phenytoin levels well. All of the equations failed to provide estimates of normalized total phenytoin level that were within +/- 2.0 mcg/mL of what was expected based on observed free phenytoin levels. Equation 3 most often provided estimates of normalized total phenytoin levels that were within +/- 2.0 mcg/mL of what was expected, but this occurred less than 50% of the timeConclusions: Phenytoin protein binding is variable depending on the population studied. Common equations used to estimate a normalized total phenytoin total levels are not accurate. More accurate equations for estimation of normalized total phenytoin levels or obtaining free phenytoin levels in hospitalized patients may be warranted