Developmental pharmacokinetics of piperacillin and tazobactam using plasma and dried blood spots from infants

Abstract

Piperacillin-tazobactam is often given to infants with severe infection in spite of limited pharmacokinetics (PK) data. We evaluated piperacillin-tazobactam PK in premature and term infants of ages <61 days with suspected systemic infection. Infants received intravenous piperacillin-tazobactam (80 to 100 mg/kg of body weight every 8 h [q 8 h]) based on gestational and postnatal age. Sparse plasma samples were obtained after first and multiple doses. Drug concentrations were measured by tandem mass spectrometry. PK data were analyzed using population nonlinear mixed-effect modeling. Target attainment rates for the time unbound piperacillin concentrations remained above the MIC for 50% and 75% of the dosing interval at steady state were evaluated. Bias in population PK parameter estimates was assessed for dried blood spot (DBS) samples, and a comparability analysis was performed for DBS and plasma drug concentrations using linear regression. We obtained 128 plasma samples from 32 infants, median gestational age of 30 weeks (range, 23 to 40 weeks) and postnatal age of 8 days (range, 1 to 60). Piperacillin and tazobactam PK models included body weight (WT) and postmenstrual age (PMA) as covariates for clearance and WT for volume of distribution and were used to optimize dosing in infants. DBS drug concentrations were 50 to 60% lower than those in plasma, but when combined with plasma concentrations and a matrix effect, the data generated PK model parameters similar to those for plasma alone. With PMA-based dosing (100 mg/kg q 8 h, 80 mg/kg q 6 h, and 80 mg/kg q 4 h for PMA of ≤30, 30 to 35, and 35 to 49 weeks, respectively), 90% of simulated infants achieved the surrogate therapeutic target of time above the MIC (≤32 mg/liter) for 75% of the dosing interval.

FIG 1

Final piperacillin population PK model diagnostic plots: observed versus individual predictions (A), visual predictive check (B), and conditional weighted residuals (CWRES) versus population predictions (C), or time (D). For panel A, the line of identity is included as a reference. For panel B, solid black circles represent observed concentrations, the shaded gray area represents the 90% prediction interval, and solid and dashed black lines represent observed and predicted median concentrations, respectively. For panels C and D, a solid line at y = 0 is included as a reference.

FIG 2

Final tazobactam population PK model diagnostic plots: observed versus individual predictions (A), visual predictive check (B), and conditional weighted residuals versus population predictions (C), or time (D). For panel A, the line of identity is included as a reference. For panel B, solid black circles represent observed concentrations, the shaded gray area represents the 90% prediction interval, and solid and dashed black lines represent observed and predicted median concentrations, respectively. For panels C and D, a solid line at y = 0 is included as a reference.

FIG 3

Individual Bayesian piperacillin clearance estimates from the final model versus postmenstrual age.

FIG 4

Target attainment rates by MIC for concentrations at 75% (A) or 50% (B) of the dosing interval in 1,000 subjects randomly selected from the database. PMA, postmenstrual age dosing regimen. The solid red line represents 90%.

FIG 5

Relationship between plasma and dried blood spot piperacillin (A and B) or tazobactam (C and D) concentrations. The dashed line represents the line of unity, and the solid line represents weighted linear regression.