Individualizing piperacillin/tazobactam dosing in adult patients with cystic fibrosis: can tobramycin measurements help?

Abstract

Empirical models to predict β-lactam pharmacokinetics (PK) using information from routine aminoglycoside therapeutic drug monitoring (TDM) have been proposed for critically ill patients; however, no such models exist for patients with cystic fibrosis (CF). To investigate whether PK parameters of tobramycin could be used to predict those of piperacillin. A non-interventional, open-label PK study was conducted in hospitalized adults treated with piperacillin/tazobactam and tobramycin for acute pulmonary exacerbations of CF. Six serum samples per patient were collected and analysed. One- and two-compartment population PK models with linear, Michaelis-Menten or mixed elimination were evaluated for both drugs within the PmetricsTM package for R. Models were developed and compared iteratively using the log-likelihood and Akaike information criterion (AIC) objective functions. Nine primarily female (n = 8) and Caucasian (100%) adult CF patients were enrolled. The median (IQR) age, height, weight and serum creatinine of included patients was 31 (27-32) years, 51.4 (49.9-55.8) kg, 162.6 (160.0-165.1) cm and 0.6 (0.5-0.6) mg/dL, respectively. The final model with the lowest objective function values consisted of one compartment with first-order elimination for tobramycin and two compartments with mixed-order elimination for piperacillin with the elimination rate constant of piperacillin modelled as a linear function of the elimination rate constant of tobramycin. A relationship was identified between the elimination rate constants of tobramycin and piperacillin. Validation of this relationship in larger studies of adult patients with CF is needed before application to the precision dosing of piperacillin/tazobactam in this patient population.