57 - Proenkephalin as a biomarker for early AKI in critically ill neonates: a pragmatic embedded study design

Poster Number: 57
Publication Number: 57.1700

Background: Neonatal acute kidney injury (AKI) is a potentially preventable mortality, morbidity, chronic lung disease, and chronic kidney disease risk factor with 30% incidence in the first 7 days of life (DOL). The current neonatal AKI definition uses serum creatinine (SCr) and urine output, which have several limitations. While AKI is potentially reversible, SCr lags by 48-72h. Serum Proenkephalin A 119–159 (PENK) at admission predicts AKI and 90-day mortality in adults. Timely identification of AKI using PENK could allow successful intervention to decrease morbidity in high-risk infants.

Objective: Our primary aim was to compare PENK in critically ill infants with and without early AKI (within 7 DOL). We hypothesize PENK will be significantly higher in those with AKI compared to those without AKI (defined by SCr).

Design/Methods: This is an observational pragmatic study that is BCM IRB approved. We obtained biospecimens collected for clinical purposes. We screened all hospitalized neonates for inclusion by high-risk criteria for AKI (hypoxic ischemic encephalopathy, low birth weight, prematurity, suspected sepsis, nephrotoxin exposure, mechanically ventilated, vasoactives, and congenital heart defects (CHD)) and exclusion criteria (>14 DOL, not expected to survive >24h, mortality < 48 HOL, known congenital anomalies of the kidney and urinary tract, or renal replacement therapy before labs were available). We recorded confounders such as maternal/infant demographics, pertinent clinical data/interventions, comorbidities, and nephrotoxin exposure. Where there was sufficient specimen, we analyzed PENK (Cusabio ELISA). We performed univariate and multivariate analysis for the relationship between PENK and AKI.

Results: 61 infants were included in this interim analysis. 11 (18%) did not have 2 SCr to determine AKI status. Of the remaining 50, 9 (18%) had AKI (56% Stage 3, 44% Stage 1) and 41 (82%) did not have AKI. Those with AKI were significantly more likely to have CHD, pulmonary hypertension, ECMO, and TPN (Table 1). PENK did not significantly differ based on early AKI (AKI: 82.72 [56.29, 148.09], Non-AKI: 78.73 [49.35, 118.08] pmol/L, p=0.99). Receiver operating characteristic curve (AUC 0.916) found infants with lower 5min APGAR (p=0.011) and higher gestational age (p=0.045) were more likely to develop AKI.

Conclusion(s): An opportunistic approach to study PENK in critically ill neonates using residual biospecimens can be successful. In this interim analysis, PENK did not significantly differ in infants with early AKI by KDIGO SCr criteria.