Sensitivity Analysis of the Effect of Bioavailability or Dosage Form Content on Mean Steady State Phenytoin Concentration

Summary: Stochastic simulations were used to examine the sensitivity of mean phenytoin steady state concentrations (Css) to changes in the effective dosing rate produced by differences in average product content or bioavailability. Changes of +4%, + 6%, + 8% and + 10% in the effective dosing rate (based on a starting dose yielding a Css of 15 mg/L) were examined. Monte Carlo simulations were performed for each change in dosing rate assuming a one-compartment open model with parallel Michaelis-Menten and first-order elimination. Parameter sets were comprised of a combination of values for maximal rate of saturable elimination (410 or 510 mg/day), the concentration at which the rate of saturable elimination is half maximum (Km, 4.4 or 5.7 mg/L), and linear clearance (Ç, 0.15 or 1.5 L/day). These parameters were assumed to be log-normally distributed with coefficients of variation of 30%, 50%, and 15%. The percentages of “individuals” who would be predicted to have Css of >10 mg/L following a reduction in the effective dosing rate increased with decreasing Km and Ç values. For a Km of 5.7 mg/L and Ç of 1.5 L/day, 5% of the “individuals” had Css values of > 10 mg/L with an 8% decrease in the dosing rate. If the dosing rate was reduced by 10%, then 14–16% of the “individuals” were predicted to have concentrations of > 10 mg/L. All other combinations of Km and Ç values yielded higher percentages of “individuals” with Css of > 10 mg/L. When the effective dosing rate increased, the percentages of the “individuals” who would be predicted to have a Css of < 20 mg/L increased with decreasing Km and Ç values. For a Km of 5.7 mg/L and Ç of 1.5 L/day, 10–12%, 32–38%, and 60–64% of the replicates would be predicted to have a Css of<20 mg/L if the dosing rate increased by 6%, 8%, and 10%. The other parameter sets showed more sensitivity to changes in the dosing rate. These simulations illustrate the importance of product uniformity that must be maintained from lot to lot of the same product and from one manufacturer to another. The sensitivity of Css to the change in effective dosing rate is influenced by a parallel, first-order elimination pathway, the magnitude of which is unknown. A definitive study is needed to substantiate these simulation-based results.