Population-based meta-analysis of furosemide pharmacokinetics.

Furosemide is a loop diuretic frequently used to treat fluid overload conditions such as hepatic cirrhosis and congestive heart failure (CHF). A population-based meta-analysis approach in NONMEM® was used to develop a PK model characterizing the time-course of furosemide in plasma and excretion into the urine for healthy subjects and fluid overload patients. Furosemide PK data from healthy subjects receiving 80 mg of oral furosemide were supplemented with additional individual and aggregate plasma concentration and urinary excretion versus time data from the literature after intravenous (i.v.) or oral furosemide administration (10-500 mg) to healthy subjects or fluid overload patients. A three-compartment model with zero-order input following i.v. administration (or first-order absorption using a Weibull function after oral administration) and first-order elimination best described furosemide PK. A covariate analysis identified creatinine clearance (CL(CR)) as a statistically significant predictor of renal clearance (CL(R)), with a population mean CL(R) of 4.67, 3.11, 1.95 and 1.17 l/h for a subject with normal renal function (CL(CR) = 120 ml/min) or mild (CL(CR) = 80 ml/min), moderate (CL(CR) = 50 ml/min) or severe (CLCR  = 30 ml/min) renal impairment. Oral bioavailability was 59.1% and non-renal clearance was 2.02 l/h. A PC-VPC and other model diagnostics demonstrated that the population PK model can reasonably predict the rate of urinary furosemide excretion over time using dosing history and commonly available demographic data, allowing for convenient assessment of PK-PD relationships for furosemide when given alone or in combination with other agents used to treat fluid overload conditions.