28 - Renal Recovery after Renal Replacement Therapy in Pediatric Extracorporeal Membrane Oxygenation

Poster Number: 28
Publication Number: 28.102

Amanda Ruth, Texas Children's Hospital, Houston, TX, United States; Sameer Thadani, Baylor College of Medicine, Houston, TX, United States; Dana Fuhrman, UPMC, Pittsburgh, PA, United States; Joseph A. Carcillo, University of Pittsburgh, Pittsburgh, PA, United States; Poyyapakkam Srivaths, Baylor College of Medicine, Houston, TX, United States; Ayse Akcan Arikan, Baylor College of Medicine, Houston, TX, United States

Background: There is little known regarding renal recovery in pediatric extracorporeal membrane oxygenation (ECMO) patients with acute kidney injury (AKI) supported on continuous renal replacement therapy (CRRT). Previously, using PELOD-2, higher P/F ratio and urine output pre-CRRT as surrogates, we described lower severity of organ dysfunction at CRRT start as independent predictors of renal recovery in pediatric CRRT.

Objective: We aim to describe renal recovery in ECMO patients on CRRT and to explore potential clinical risk factors associated with lack of renal recovery.

Design/Methods: Multicenter retrospective cohort analysis (all patients < 18 years) who required concurrent CRRT while on ECMO between February 2014 through February 2020. Primary outcome was Major Adverse Kidney Events (death, dialysis dependency, or persistent renal dysfunction) at 30 days (MAKE30).

Results: Seventy-five patients (median age 72 months (IQR 17-172); 53% male) were included. The median CRRT duration was 11 days (IQR 3-25); the median PELOD-2 score at ICU admission and CRRT start were 8 (IQR 5-10) and 9 (IQR 8-12), respectively. Median fluid overload at CRRT start was 9.4% (IQR 2.4% – 16.4%). Median number of organ failures were 3 (IQR 2-4) and 4 (IQR 3-5) at ICU admission and CRRT start respectively (p < 0.001). One third of the cohort (33%) had full renal recovery while 55/75 (73%) met MAKE 30, of which 27 (49%) patients died, 20 (36%) were dialysis dependent, and 8 (15%) with persistent renal dysfunction at the time of discharge. Only higher weight and higher mean arterial pressure at ICU admission were associated with higher odds of MAKE 30 (OR 1.02, 95% 1.00- 1.05, p=0.034; OR 1.04, 95% CI 1.00-1.08, p=0.044, respectively). None of the other covariates tested including PELOD-2 at ICU admission or CRRT start, age, diagnostic category, fluid overload, type of ECMO support, vasoactive infusion use, P/F ratio, or urine output production before CRRT were associated with MAKE 30.Conclusion(s): Pediatric ECMO population receiving CRRT have low rates of renal recovery at 30 days but seem to have distinctive risk factors associated with lack of renal recovery. Knowledge gained from other pediatric CRRT populations might not be transferable to this risk group and predictors for failure of renal recovery should be further studied to identify modifiable risk factors.