Recommended β-lactam regimens are inadequate in septic patients treated with continuous renal replacement therapy

Sepsis is responsible for important alterations in the pharmacokinetics of antibiotics. Continuous renal replacement therapy (CRRT), which is commonly used in septic patients, may further contribute to pharmacokinetic changes. Current recommendations for antibiotic doses during CRRT combine data obtained from heterogeneous patient populations in which different CRRT devices and techniques have been used. We studied whether these recommendations met optimal pharmacokinetic criteria for broad-spectrum antibiotic levels in septic shock patients undergoing CRRT.

This open, prospective study enrolled consecutive patients treated with CRRT and receiving either meropenem (MEM), piperacillin-tazobactam (TZP), cefepime (FEP) or ceftazidime (CAZ). Serum concentrations of these antibiotics were determined by high-performance liquid chromatography from samples taken before (t = 0) and 1, 2, 5, and 6 or 12 hours (depending on the β-lactam regimen) after the administration of each antibiotic. Series of measurements were separated into those taken during the early phase (< 48 hours from the first dose) of therapy and those taken later (> 48 hours).

A total of 69 series of serum samples were obtained in 53 patients (MEM, n = 17; TZP, n = 16; FEP, n = 8; CAZ, n = 12). Serum concentrations remained above four times the minimal inhibitory concentration for Pseudomonas spp. for the recommended time in 81% of patients treated with MEM, in 71% with TZP, in 53% with CAZ and in 0% with FEP. Accumulation after 48 hours of treatment was significant only for MEM.

In septic patients receiving CRRT, recommended doses of β-lactams for Pseudomonas aeruginosa are adequate for MEM but not for TZP, FEP and CAZ; for these latter drugs, higher doses and/or extended infusions should be used to optimise serum concentrations.