Urine NGAL adds to serum creatinine in predicting cefepime clearance in critically ill children at high risk of acute kidney injury.

INTRODUCTION

Cefepime clearance (CL) depends on renal function, but serum creatinine (SCr) lags behind real-time changes in kidney function. Urinary neutrophil gelatinase-associated lipocalin (uNGAL) is a biomarker of real-time tubular injury, though its role in predicting cefepime pharmacokinetics (PK) is unclear.

METHODS

We conducted a prospective study in paediatric intensive care unit (PICU) patients at high risk for acute kidney injury who received cefepime within 48 h of admission and had uNGAL measured. A validated population PK model incorporating SCr-estimated GFR (eGFR) was adapted to test the added predictiveness of uNGAL. Inter-occasion variability (IOV) assessed whether CL and central volume (V1) differed after 48 h of PICU admission. Monte Carlo simulations (n = 10 000) evaluated the probability of pharmacodynamic (PD) target attainment (PTA) for various cefepime regimens and PK/PD targets across weight groups (5-40 vs. 40-100 kg), kidney function strata (eGFR 30-60, 60-90, 90-150 mL/min/1.73m²), and uNGAL thresholds (≥500 vs. <500 ng/mL).

RESULTS

Fifty patients (median age 11.5 y) were enrolled. PICU admission uNGAL ≥500 ng/mL was associated with ∼35% lower CL, independent of eGFR, with IOV at 4 h post-PICU admission significant for CL and V1. Patients with low uNGAL had higher CL at a given eGFR with consequently lower PTA for a given dosing regimen, necessitating extended or continuous infusions to reach PK/PD targets.

CONCLUSIONS

Integrating uNGAL improves the prediction of cefepime CL beyond SCr alone. Patients with low uNGAL despite elevated SCr risked subtherapeutic exposure. Urine biomarker-guided dosing may help optimise PTA and warrants further study.