Longitudinal impact of temporary mechanical circulatory support on durable ventricular assist device outcomes: An IMACS registry propensity matched analysis.

ABSTRACT

BACKGROUND: Patients with advanced heart failure and cardiogenic shock (CS) often require temporary circulatory support (TCS) as a bridge to durable ventricular assist devices (dVADs). We aim to characterize longitudinal outcomes of patients with and without CS. METHODS: Between 2013 and 2017, 13,813 adult patients classified as Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) Profiles 1 to 3 with continuous flow left ventricular assist devices or biventricular assist devices were registered into the International Society for Heart and Lung Transplantation Registry for Mechanically Assisted Circulatory Support. Patients were sub-grouped according to support type (extracorporeal membrane oxygenation [ECMO], intra-aortic balloon pump [IABP], and other TCS). Other TCS included all other surgical and percutaneous TCS devices. Estimated survival was compared based on need for pre-operative TCS and by profile. RESULTS: Pre-operative TCS was used in 5,632 (41%) cases. Of these, ECMO was used in 1,138 (20%) cases, IABP in 3,901 (69%) cases, and other TCS in 595 (11%) cases. Patients requiring ECMO had greater needs for biventricular support after dVAD (22% ECMO, 5% IABP, and 7% other TCS; p < 0.001) with longer post-implant intensive care stays (ECMO 24 days, IABP 14 days, and other TCS 12 days; p < 0.001). INTERMACS Profile 1 to 3 patients with pre-implant ECMO had the lowest longitudinal survival (82% at 1 month and 44% at 48 months) compared with IABP (93% at 1 month and 51% at 48 months), other TCS (92% at 1 month and 52% at 48 months), and non-TCS (95% at 1 months and 55 % at 48 months) (p < 0.0001). Propensity score matching analysis of the pre-implant ECMO INTERMACS Profile 1 group when compared with alternative pre-implant TCS strategies had an associated higher hazard impacting early phase survival vs other TCS (hazard ratio, 1.80; p < 0.01) and IABP (hazard ratio, 1.65; p < 0.01). CONCLUSIONS: In advanced heart failure with patients with CS, the use of ECMO before dVAD was associated with lower longitudinal survival and increased utilization of biventricular support compared with alternative TCS strategies. Research focused on longitudinal profiling in CS and pre-implant TCS is warranted to further understand these differences.