A new method for arriving at individualized doses of phenytoin (PHT) is described. This method, designated the plusmn; method, operates on two measurements of unbound plasma PHT. One plasma sample is collected 48 h after a single, oral PHT dose (sim;4.5 mg/kg). From the single dose, single sample measurements of plasma unbound PHT individual estimates of intrinsic unbound PHT clearance (CLuint) can be made. With the use of CLuintand weight-adjusted mean population values of Vmaxpredicted plasma concentrations of unbound PHT for 11 subjects exhibited a mean prediction error (mpe) of 0.07 mg/L with 95percnt; confidence limits of minus;0.11 to 0.26 mg/L and an error coefficient of variation root mean squared error (rmse)/mean unbound concentration at steady state (Cuss) of 0.4. The second plasma unbound PHT is collected at steady state. From CLuintand one Cussmeasurement, Vmaxand Kmcan be directly calculated. Doses calculated by the plusmn; method to produce plasma unbound PHT concentrations of sim;1.0 mg/L were identical to those calculated by the Bayesian method to yield plasma total PHT concentrations of sim;10 mg/L. When those doses were administered to six subjects, the plasma unbound PHT concentrations predicted by the plusmn; method exhibited a mpe of ndash; 0.13 mg/L, with 95percnt; confidence limits ranging from minus;0.33 to 0.06 mg/L and an error coefficient of variation of 0.2. The plasma PHT concentrations predicted by the Bayesian method exhibited an mpe of ndash; 1.30 mg/L, with 95percnt; confidence limits of minus;4.02 to 1.42 mg/L and an error coefficient of variation of 0.3. Values of Vmaxand doses of PHT calculated to produce a Cussof 1.0 mg/L calculated by the plusmn; method were nearly identical to Vmaxvalues and doses of PHT calculated by Bayesian double feedback (based on two different Cssvalues) to produce a Cssof 10.0 mg/L.
展开▼
