Neonatal acute kidney injury: Reassessing diagnostic criteria for acute kidney injury in neonates with hypoxic-ischaemic encephalopathy undergoing therapeutic hypothermia.

Abstract

Background: Acute kidney injury (AKI) remains common in neonates with hypothermia-treated hypoxic-ischaemic encephalopathy (HIE). Although the modified neonatal KDIGO (nKDIGO) criteria form the diagnostic standard, their reliance on serum creatinine (SCr) increase and urine output (UO) is particularly challenging in early onset neonatal AKI. The Gupta definition, based on insufficient SCr-decline, has been proposed to identify neonates with clinically significant AKI uncaptured by nKDIGO.

Methods: This retrospective cohort included 55 near-term/term neonates with hypothermia-treated HIE in Stockholm (2007-2009). Applying nKDIGO and Gupta criteria, infants were categorised as nKDIGO-AKI, Gupta-AKI, or no AKI. Primary outcome was AKI-incidence, secondary outcomes included mortality, organ support, length of stay, and signs of possible impaired kidney function at 10-12 years.

Results: In our cohort, 31 (56%) suffered AKI: 17 (31%) fulfilled both definitions, 7 (13%) Gupta-only, and 7 (13%) nKDIGO-only. Need for inotropic support was more common among infants with nKDIGO-AKI compared to those without nKDIGO-AKI (p = 0.03). In a subanalysis, a significantly larger proportion of infants with Gupta-AKI needed inotropic support compared to those without Gupta-AKI (p = 0.03). Gupta-AKI was associated with a significantly higher cystatin C at 10-12 years (p = 0.03) compared to those without Gupta-AKI.

Conclusions: The Gupta definition identified additional infants with AKI beyond nKDIGO. For both definitions, a larger proportion of infants needed inotropic support in the AKI-group compared to those without AKI. Children with a history of Gupta-AKI had a significantly higher cystatin C in early adolescence compared to those without. Incorporating Gupta criteria may improve identification of infants with significant AKI.