What Should Be Done for Improving the Outcomes of Salvage from Cardiogenic Shock on ECMO Support?

Abstract

Purpose Background: Survival rates for patients requiring extracorporeal membrane oxygenation (ECMO) support due to cardiogenic shock remain poor because of difficulties in decision making on optical timing of ECMO removal or conversion to ventricular assist devices (VAD). Methods From 2005 to 2018, 37 patients supported with ECMO due to circulatory deterioration were referred to our institution for implantation of VAD. The outcomes were analyzed using multi-variate analysis to assess the risk factors of VAD implantation, and we adopted a decision-tree to improve the outcomes. Results Six patients were withdrawn from VAD therapy due to stroke and infection. ECMO could be removed in 6 patients (6.1 days of ECMO support), but one of them was estimated to need VAD support due to the IABP-dependent circulatory condition. In total, 26 patients needed VAD support after 5.7 days of ECMO support. Four of them reached to heart transplantation, and VAD was explanted in 2 patients. Six patients remain on VAD. Statistically significant risk factors by multi-variate analysis on 2-year mortality were preoperative renal replacement (p=0.006) and T.Bil (p=0.051, Threshold 4.0 mg/dl). Two-year survival of patients without end-organ dysfunction as mentioned above was excellent (83.3%). However, that of patients with end-organ dysfunction was miserable (23.1%). Based on these findings, we applied to a following decision-tree. Step I: end-organ dysfunction; if yes, additional introduction of more effective circulatory support with left ventricular venting. Step II: aortic valve opening: if no, emergent/urgent VAD implantation. If yes, wait for cardiac functional recovery on ECMO. Step III: cardiac function; if no recovery, VAD implantation, and if recovered, ECMO removal. Since this decision-tree was applied, ECMO support was initiated in four patients with end-organ dysfunction before VAD implantation. After T.Bil decreased from 15.9mg/dl to 4.6 mg/dl and the value of Creatinine from 4.0 mg/dl to 1.5mg/dl, those patients underwent VAD implantation and survived. Conclusion Our experiences of salvage from ECMO suggest that preoperative renal replacement and T.Bil less than 4.0 mg/dl was a strong risk factor for mid-term mortality. For those patients, recovery from end-organ dysfunction before VAD implantation is essential.