Evidence of nonlinearity in digoxin pharmacokinetics.

Six normal male volunteers received 0.5 mg label doses of digoxin as (a) a bolus intravenous injection over 2 min, (b) a constant rate intravenous infusion over 1 hr, (c) a constant rate intravenous infusion over 3 h, and (d) a solution in 5% dextrose given orally. Plasma concentrations of digoxin were measured by radioimmunoassay for a 4 day period and urinary excretion for a 6 day period after the single doses. The mean (coefficient of variation) total areas under the plasma concentration-time curves per 0.5 mg of digoxin were (a) 35.55 (14.8%), (b) 30.20 (27.7%), (c) 25.80 (35.5%), and (d) 15.47 (49.9%); the means differed significantly (0.01 greater than p greater than 0.005). The mean (coefficient of variation) total amounts excreted in the urine as a fraction of the dose were (a) 0.689 (6.31%), (b) 0.517 (20.4%), (c) 0.588 (16.8%), and (d) 0.374 (23.4%); the means differed significantly (p less than 0.001). Both the total clearance and the nonrenal clearance of digoxin differed significantly with the method of intravenous administration. The slower the rate of input of digoxin to the body, the greater were both the total clearance and the nonrenal clearance of the drug, which strongly suggests nonlinear pharmacokinetics.