Cefepime plasma concentrations and clinical toxicity: a retrospective cohort study.

OBJECTIVES

Cefepime remains an important antibiotic for severe bacterial infections, yet some meta-analyses have shown elevated mortality among patients randomized to it. Therapeutic drug monitoring (TDM) of β-lactam antibiotics is increasing, but optimal plasma concentrations remain unknown. We examined clinical outcomes of patients undergoing cefepime TDM in an initial effort to define the drug's toxicity threshold.

METHODS

In this single-centre retrospective cohort study, we enrolled all adult hospitalized patients receiving cefepime and undergoing TDM from January 2013 through July 2016. The primary outcome was the incidence of clinical toxicity; a secondary outcome was clinical failure. Plasma samples were analysed via high-performance liquid chromatography with ultraviolet detection.

RESULTS

A total of 161 cefepime concentrations were drawn from 93 patients. Roughly half (82/161, 51%) and one-third (49/161, 30%) were trough and steady-state levels from patients receiving intermittent and continuous infusions, respectively; median concentrations were 17.6 mg/L (IQR 9.7-35.2) and 29.2 mg/L (IQR 18.9-45.9). Ten patients (11%) experienced a neurologic event considered at least possibly related to cefepime; neurotoxicity was associated with poorer renal function (median creatinine clearance 54 (IQR 39-97) vs. 75 mL/min/1.73 (IQR 44-104)) and longer cefepime durations (mean 8.3 (SD±6.7) vs. 13.3 days (± 14.2), p = 0.071). Patients with trough levels >20 mg/L had a fivefold higher risk for neurologic events (OR 5.05, 95% CI 1.3-19.8).

CONCLUSIONS

Neurotoxicity potentially related to cefepime occurred at plasma concentrations >35 mg/L. For those receiving intermittent infusions, trough concentrations >20 mg/L should be avoided until further information is available from prospective studies.