Utilizing an Opportunistic Clinical Study and Population-Based Pharmacokinetic Models to Identify Rational Empiric Dosing Regimens for Piperacillin-Tazobactam in Critically Ill Patients

Abstract

Determining an effective dosing regimen for piperacillin-tazobactam in critically ill patients is challenging due to substantial pharmacokinetic variability caused by complex pathophysiological changes. To address this need, a prospective clinical study was conducted, which enrolled 112 critically ill patients and employed an opportunistic sampling strategy. Population modeling and simulation were performed to characterize the pharmacokinetics (PK) and probability of target attainment (PTA) of piperacillin-tazobactam under various dosing regimens. Both piperacillin and tazobactam final models were one-compartment models with zero-order input and first-order elimination. Significant covariates included lean body weight for piperacillin and creatinine clearance along with continuous renal replacement therapy (CRRT) for both drugs. Monte Carlo simulations demonstrated that continuous infusion can achieve higher PTA than intermittent and extended infusions. When considering the minimum inhibitory concentration (MIC) of 16 mg/L for Pseudomonas aeruginosa (a frequently encountered bacterial pathogen among critically ill patients) and a PK/PD target of 100% fT >MIC, continuous infusion of 6 g/day is recommended for critically ill patients with a CLcr <60 mL/min, 9 g/day for patients with CLcr in the range of 60 to 129 mL/min, and 12 g/day for patients with a CLcr ≥130 mL/min. In addition, extended infusion represents a good alternative, especially the 3 g q6h or 4 g q6h regimens which can achieve the designated European Committee on Antimicrobial Susceptibility Testing (EUCAST) non-species-related PK/PD breakpoint of 8 mg/L. Our study provided valuable insight into PTA outcomes, which, together with individual renal function of future patients and institution-specific piperacillin susceptibility patterns, may assist physicians when making dosing decisions.

Keywords: empiric antibiotic dosing regimen selection; opportunistic clinical study; pharmacometric modeling; piperacillin–tazobactam; population pharmacokinetics.

Figure 1

Individual observed (symbol) and model predicted (line) for piperacillin and tazobactam concentration versus time profiles in six representative ICU patients.

Figure 2

Prediction‐corrected visual predictive check (pc‐VPC) of (a) piperacillin and (b) tazobactam final models. Black dots, prediction‐corrected observed values; solid red line, median (50th percentile); dashed red lines, 5th and 95th percentiles for the prediction‐corrected observed values. The shaded regions are the 95% confidence intervals of the simulated 5th, 50th, and 95th percentiles.

Figure 3

PTA of piperacillin versus MIC following various dosing regimens with the target of 100% fT > MIC; intermittent infusion (upper panel), extended infusion (middle panel), and continuous infusion (lower panel).

Figure 4

Heatmap of PTA of piperacillin at different MIC values following different dosing regimens in all subjects as well as subjects with different renal functions, with the PK/PD targets of 100% fT > MIC (green, PTA ≥90%; yellow, PTA 80%‐89%; orange, PTA 50%‐79%; red, PTA <50%).

Figure 5

Simulated tazobactam‐free plasma concentration–time profiles following (a) tazobactam 0.5 g every 8 h (intermittent infusion), (b) tazobactam 0.5 g every 8 h (extended infusion), and (c) tazobactam 1.5 g daily dose (continuous infusion) against the MCC threshold of 4 mg/L.