Colistin-induced acute kidney injury in neonates with sepsis and its association with plasma colistin concentration: a prospective cohort study.

BACKGROUND: Colistin is increasingly used as a reserve antibiotic for multidrug-resistant (MDR) Gram-negative infections. However, colistin-induced acute kidney injury (AKI) is under-reported in neonates with sepsis. OBJECTIVES: To evaluate the association and causality between colistin use and AKI in neonatal sepsis, and to explore the correlation between steady-state plasma colistin concentration and AKI development. METHODS: In this prospective cohort study, neonates receiving colistin for proven or probable sepsis were enrolled. Exclusion criteria included major congenital anomalies and death within 72 h of life. Causality of colistin-induced AKI was assessed using the World Health Organization-Uppsala Monitoring Centre (WHO-UMC) criteria. Steady-state colistin levels were measured after the sixth dose. AKI was defined and staged according to KDIGO guidelines. RESULTS: A total of 40 neonates were enrolled (mean gestational age: 29.3 ± 3.1 weeks). The mean duration of colistin therapy was 10.1 ± 0.8 days. AKI occurred in 28/40 (70%), of whom 13 had moderate-to-severe AKI. According to WHO-UMC criteria, 10/28 (35%) AKI cases were classified as “possible” colistin-induced nephrotoxicity. The median (IQR) steady-state colistin concentration was 1.08 (0.53,2.53) mg/L overall. Plasma colistin levels were higher in neonates with “possible” nephrotoxicity (median: 2.09 mg/L) compared with those classified as “unlikely/unclassifiable” (0.95 mg/L) and the non-AKI group (1.04 mg/L) (p = 0.3). Thirteen neonates (33%) died during hospitalization. CONCLUSION: Possible colistin induced AKI was seen in nearly one third cases. Higher steady-state plasma colistin levels was seen in the possible colistin-induced AKI group, highlighting the need for therapeutic drug monitoring and cautious use in this vulnerable population.