P1.227: Urinary Biomarkers of Renal Injury and Association with Fluid Overload in Extremely Premature Neonates in First 72 Hours of Life

Background: Acute kidney injury (AKI) is common among extremely premature neonates and is clinically characterized by disordered fluid, electrolyte, and acid-base balance. Serum creatinine (sCr) is a key parameter among contemporary AKI definitions. Yet, sCr is confounded by maternal Cr in fetal blood in the transitional period (first 72-hours of life), and fluid overload (FO). As a result, sCr may underestimate and poorly delineate renal dysfunction in neonates. To address this knowledge gap, we sought to characterize alternative urinary biomarkers of renal injury (uBRI) in the transitional period and determine if they correlate with the percentage of FO among extremely premature neonates.

Methods: We performed a secondary analysis of cryopreserved urine specimens of neonates < 28 weeks GA collected serially at < 12h, 24h, 48h, and 72h ±6h. Targeted multiplex (ProcartaPlexTM Human Kidney Toxicity Panels 1 & 2) testing was performed in duplicate and adjusted to urine Cr as a pooled analysis and by discrete sample collection times. Analyte values were compared by the percent FO (≤-5% [FO-], -5 to +5%, and ≥5% [FO+]), calculated by the change in daily weight from birth. Demographics, anthropometrics, treatments, and inpatient clinical outcomes were summarized with frequencies and percentages, or with median and interquartile ranges by FO groups. Spearman’s Rank-Order Correlation analysis was performed to evaluate associations (ρ) between FO and each putative uBRI for all neonates and by FO group.

Results: Of 24 unique patients studied, 10 (42%) were FO-, 7 (29%) were FO+, and 2 (8%) fulfilled the nKDIGO definition of stage 1 AKI in the first 72h based on change in sCr (none met criteria by absolute sCr level or urine output). As compared to those that were FO-, those that were FO+ had lower birth weight Z-scores, were more frequently small for GA, more frequently received vasoactive infusions, and less frequently administered indomethacin and gentamicin (see Table 1). Regarding putative uBRI analytes, cystatin C and β-2-microglobulin were strongly positively correlated with FO+ (ρ: 0.82 and 0.79, p=0.02 and 0.04, respectively). α-1-microglobulin was moderately negatively correlated (ρ: -0.46, p=0.02) with FO. A number of putative uBRI (see Figure 1) were positively correlated with those FO+ (MCP-1, GSTA1, TIMP1) and negatively correlated with those FO- (renin, calbindin), but not significantly associated.

Conclusion: In this single center observational study, we identified several uBRI correlated with FO in extremely premature neonates during the transitional period. The discriminatory performance of these analytes (both independently and compared to current indices of renal dysfunction) warrant further validation and investigation in this population with an analytic focus on long-term outcomes and sequelae of renal dysfunction.