Population pharmacokinetics of piperacillin at two dose levels: influence of nonlinear pharmacokinetics on the pharmacodynamic profile

Abstract

Piperacillin in combination with tazobactam is one of the most commonly used intravenous antibiotics. There is evidence for a possible saturable elimination of piperacillin. Therefore, the saturable elimination and its impact on the choice of optimal dosage regimens were quantified. In a randomized crossover study, 10 healthy volunteers received 1,500 mg and 3,000 mg of piperacillin as 5-min intravenous infusion. Population pharmacokinetics based on plasma and urine data were determined utilizing NONMEM and S-ADAPT. Probabilities of target attainment (PTAs) were compared for different models and dosage regimens, based on the target time of the non-protein-bound concentration above the MIC of at least 50% of the dosing interval. Total clearance of piperacillin was 18% (geometric mean ratio, 90% confidence interval, 11 to 24%) lower (P < 0.01), and renal clearance was 24% (9 to 37%) lower (P = 0.02) at the high compared to the low dose. The final model included first-order nonrenal elimination and parallel first-order and mixed-order renal elimination. Nonrenal clearance was 5.44 liter/h (coefficient of variation, 18%), first-order renal clearance was 4.42 liter/h (47%), and the maximum elimination rate of mixed-order renal elimination was 219 mg/h (84%), with a Michaelis-Menten constant of 36.1 mg/liter (112%). Compared to models with saturable elimination, a linear model predicted up to 10% lower population PTAs for high-dose short-term infusions (6 g every 8 h) and up to 4% higher population PTAs for low-dose continuous infusions (6 g/day). While renal elimination of piperacillin was saturable at therapeutic concentrations, the extent of saturation of nonrenal clearance was small. The influence of saturable elimination on PTAs for clinically relevant dosage regimens was relatively small.

Fig 1

Average (± SD) profiles of piperacillin in healthy volunteers after 5-min infusion of 1,500 mg or 3,000 mg piperacillin. (A) Plasma drug concentrations; (B) cumulative amounts excreted unchanged in urine. The dashed line in panel B represents the amounts excreted in urine after the 1,500-mg dose multiplied by 2.

Fig 2

Renal and nonrenal clearance from noncompartmental analysis after administration of 1,500 mg or 3,000 mg piperacillin to healthy volunteers. The boxes show the 25th, 50th, and 75th percentiles, and whiskers are the 10th and 90th percentiles; filled circles are the minima and maxima.

Fig 3

Linear renal clearance, saturable renal clearance, linear nonrenal clearance, and total body clearance at various plasma concentrations of piperacillin for model 3 (Table 2). The approximate range of clinically encountered piperacillin concentrations was 45 to 80 mg/liter (the 10th to 90th percentiles from the Monte Carlo simulation) for continuous infusion of 18 g/day, 0 to 60 mg/liter for a 4-h infusion of 3 g q8h, and 0 to 600 mg/liter for a 0.5-h infusion of 6 g q8h.

Fig 4

Visual predictive checks for plasma drug concentrations and amounts of drug excreted unchanged in urine for model 3 (see Table 2). The plots show the observed data and the 80% prediction interval (i.e., between the 10th to 90th percentiles) and the interquartile range (i.e., between the 25th and 75th percentiles) from 4,000 simulated subjects. Ideally, 50% of the observed data points should fall inside the interquartile range at each time point, and 80% of the observed data should fall inside the 80% prediction interval.

Fig 5

Probabilities of target attainment for the four population PK models (Table 2) and different dosage regimens of piperacillin (PK/PD target, fT_>MIC of ≥50%). ▲, first-order renal and first-order nonrenal elimination (model 1); □, mixed-order renal and first-order nonrenal elimination (model 2); ◊, parallel first- and mixed-order renal and first-order nonrenal elimination (model 3); and ○, parallel first- and mixed-order renal and mixed-order nonrenal elimination (model 4).