CARC4: Incidence, Outcomes, and Risk Factors of Acute Kidney Injury in Patients on Percutaneous Mechanical Circulatory Support with Impella 5.0 and 5.5 for Cardiogenic Shock

Abstract

Purpose: Use of temporary mechanical circulatory devices (tMCS) such as the transaxial flow pump (ImpellaTM. Abiomed, Danvers, MA) for CS-HF is growing. Despite advances in tMCS investigations have found AKI is highly prevalent and strongly associated with mortality in CS patients supported with MCS. We identified pre and post implant risk factors of AKI in CS-HF patients supported with Impella 5.0 and 5.5. Methods: Single center retrospective analysis of 144 Impella 5.0/5.5 devices from 01/2012 to 07/2022 for CS-HF. Patients with non heart failure etiology (n=5) and pre-implant renal replacement therapy (RRT. n=44) were excluded. Demographics and pre-implant chacteristics of patients) as well as post-implant outcomes including mortality, heart replacement therapy (HRT) with ventricular assist device (VAD) or heart transplant (HTX), and native heart survival at discharge were compared between patients with and without acute kidney injury (AKI). AKI was defined as rise in Creatinine 1.5 x baseline or new need for RRT. Logistic regression was used to find independent predictors of post-implant AKI and spearmann correlation was used to find associations with 90-day kidney function defined as fold change in creatine from baseline to peak 90 day values). Results: The final cohort consisted of 96 patients with mean duration of Impella support of 10±12 days.. Of the remaining 96 devices 57 (59.4%) developed AKI with 9 requiring RRT (9.4%). Patients with AKI had higher mortality (42.1% vs 17.9%, p=0.04), lower HRT (43.9% vs 56.4%, p=0.04) and less native heart survival (discharged without HRT) (14.0% vs 25.6%, p=0.04) than those without AKI. Demographics and pre-implant characteristics are shown in in Table 1. Pre-implant creatinine, lactate dehydrogenase (LDH), total bilirubin, Model for End Stage Liver Disease (MELD), and mean right atrial pressure to pulmonary capillary wedge pressure (mRAP/PCWP) as well as post-implant hemolysis (rise in plasma free hemoglobin > 20 or LDH 2.5 x upper limit of normal) were higher in patients with AKI. On logistic regression with backwards selection independent predictors of AKI included pre-implant mRAP/PCWP (OR 33.5 95%CI 1.6-708.8, p=0.02) and post-implant hemolysis (OR 5.6 95%CI 2.1-31.5, p=0.045) (Table 2). On spearman correlation (Table 3) only pre-implant mRAP/PCWP and post-implant LDH:Upper limit of normal LDH were associated with worse long term kidney function defined as fold change of 90-day peak creatinine from baseline. Conclusions: AKI is associated with higher mortality after Impella 5.0 and 5.5 insertion. The etiology of AKI in patients with cardiogenic shock supported by Impella 5.0 and 5.5 is multifactorial but clearly related to venous congestion and hemolysis. Optimization of patient’s fluid status when the device is inserted and avoidance of hemolysis during device support play key roles in reducing AKI in this group of high risk patients.