Outcomes of extracorporeal membrane oxygenation with or without left ventricular unloading devices in patients with non-myocardial infarction cardiogenic shock

Abstract Background Venoarterial extracorporeal membrane oxygenation (VA-ECMO) is being increasingly used to treat cardiogenic shock, however its effect on increasing left ventricular (LV) afterload may slow myocardial recovery and negatively affect survival. Percutaneous mechanical support devices have been utilized for LV unloading by reducing afterload in an attempt to improve outcomes. While the use of LV unloading devices remains debatable, its use has not been specifically studied in patients with non-acute myocardial infarction cardiogenic shock (non-AMICS). Purpose To study the outcomes of VA-ECMO with or without LV unloading devices in patients with non-AMICS patients. Methods National inpatient sample database from years 2015 to 2018 was queried to select patients admitted with non-AMICS. Patients were included in the study if they underwent VA-ECMO during admission and later categorized into 3 groups i.e. VA-ECMO, VA-ECMO plus Impella and VA-ECMO plus intra-aortic balloon pump (IABP). Baseline demographics and in-hospital outcomes were compared between the 3 pre-specified groups. Statistical significance was assigned at p<0.05. Results 178,605 patients met criteria for non-AMICS. Of these, 2190 (1.23%) patients received VA-ECMO alone, 965 (0.54%) received VA-ECMO plus IABP and 414 (0.23%) received VA-ECMO plus Impella. On univariate analysis, patients who received VA-ECMO alone had higher rates of inpatient mortality as compared to those who received VA-ECMO plus IABP or VA-ECMO plus Impella (39.04%, 33.72% and 25.81% respectively, p=0.001). On multivariate analysis, the patients who received VA-ECMO plus IABP or VA-ECMO plus Impella had lower odds of mortality when compared to VA-ECMO alone (OR: 0.61 (0.39–0.96), p=0.03), OR: 0.51 (0.23–1.08), p=0.08). The length of stay and cost were significantly higher for patients with VA-ECMO with unloading devices (IABP or Impella) compared with VA-ECMO alone (24.77±2.44 and 27.74±3.55 days vs 23.70±1.25 days respectively. p=0.001, $846,404±71169 and 860,999±121942 vs $740,274±43644 respectively, p=0.001). Conclusions Non AMICS patients who received VA-ECMO along with LV unloading devices (esp IABP) had lower in-hospital mortality as compared to those who received VA-ECMO alone despite having longer length of stay and higher cost. Use of LV unloading devices like IABP or Impella may improve outcomes in patients requiring VA-ECMO support for non-myocardial infarction cardiogenic shock. Further studies are needed to identify specific patient subsets that may benefit from this approach. Funding Acknowledgement Type of funding sources: None.