The TandemHeart^TM is a percutaneous left atrium to iliac artery bypass device. Powered by a centrifugal pump, it can provide up to 4 L/min in forward flow. Insertion technique required venous access and transseptal puncture, with a 21F inflow cannula. Using arterial access a 15F or 17F outflow cannula is placed in the ipsilateral common iliac artery. In the initial human pilot studies, the TandemHeart decreased LV end-diastolic pressure and oxygen demand while simultaneously increasing mean arterial pressure [36]. A small multicenter randomized trial followed comparing the device to the IABP [39]. However, a subsequent meta-analysis of two trials did not show any difference in mortality [32].

Given the demands of transseptal puncture and the large cannulas insertion times in excess of 30 min have been reported in literature [40]. In addition to bleeding and limb ischemia, specific complications of the TandemHeart include arrhythmias, tamponade, atrial perforation, as well as residual atrial septal defects. Specific contraindications include a ventricular septal defect as well as aortic insufficiency or severe peripheral vascular disease. With worsening LV function, hemodynamic support becomes an overriding concern, and it is in this situation that the TandemHeart has definitive indications. The Impella 5.0 may deliver as much support as the TandemHeart but cannot be inserted percutaneously.

Hemodynamic effects of the TandemHeart device include augmentation of both cardiac output as well as blood pressure, effectively raising the cardiac power output. In simulated models, the CPO from a TandemHeart is between the Impella 2.5 and the Impella 5.0 devices, irrespective of the hemodynamic status [41]. Oxygen consumption is directly lowered by the device. In cardiogenic shock the increase in afterload would counter the decrease in myocardial oxygen demand thereby leading to a net neutral effect on myocardial oxygen demand. In animal models of acute myocardial infarction, the device has been demonstrated to restore epicardial as well as microvascular blood flow [42].

At the time of this writing, the US Food and Drug Administration is reviewing a proposal to study the ability of TandemHeart to reduce infarct size when it is placed prior to primary PCI. The TandemHeart to Reduce Infarct Size (TRIS) trial will assign STEMI patients randomly to primary PCI with or without TandemHeart support prior to recanalization of the culprit vessel. Shock patients will be excluded from this trial, but if results are encouraging, additional study in the shock population would be pursued (personal communication, 2011, Howard A. Cohen MD).

The Impella consists of a small axial-flow pump that directly unloads the left ventricle by attaching to a pigtail catheter and moving blood from the left ventricle into the ascending aorta. Two devices are currently available with maximal flow rates of 2.5 L/min (Impella 2.5) or 5 L/min (Impella 5.0). The Impella requires a single femoral arteriotomy and the insertion of a 13F (Impella 2.5) catheter. After crossing the aortic valve, an exchange length 0.018″ stiff wire is used to place the device in the LV. Implantation times around 10 min have been documented [43]. A complication specific to the Impella device is hemolysis that can occur in 5–10 % of patients [44]. Aortic valve disease and LV thrombi are specific contraindications to insertion of the Impella device. The Impella has more supportive clinical data for efficacy in the failing ventricle than does the TandemHeart. The first US feasibility trial (PROTECT I) was a multicenter study of 20 sequential patients treated with Impella 2.5 device in a nonrandomized fashion [44]. All patients had poor left ventricular function (ejection fraction ≤35 %) and underwent PCI on an unprotected left main coronary artery or last remaining patent coronary conduit. Patients with acute myocardial infarction or cardiogenic shock were excluded from this study. Two patients died at days 12 and 14, and two patients developed transient hemolysis. The European AMC MACH I trial produced very similar results [37]. The AMC MACH 2 trial looked at patients who were experiencing an acute anterior wall myocardial infarction without cardiogenic shock. Compared to IABP, improvement in the LVEF was noted [43]. The ISAR SHOCK trial randomized patients with an acute anterior wall myocardial infarction with cardiogenic shock and demonstrated improved hemodynamics when compared to the IABP [45]. In the PROTECT II trial, the Impella 2.5 device was compared to IABP in patients with low LV ejection fractions undergoing non-emergent but high-risk PCI with an expectation that benefit would be seen, using a large composite endpoint made up of ten individual clinical outcomes. [46] The study was terminated prematurely because of futility. An a priori analysis showed no difference in mortality between the Impella and the IABP. Despite these disappointing findings, some investigators and many operators appear to believe the results were confounded, as evidenced by the continued (indeed, expanding) use of this device in such patients. In the Europella registry, the Impella device was used in elderly patients with an ejection fraction of less than 30 %. Mortality at 30 days was 5.5 % and hemolysis was seen in about 1 % of patients [47]. It should be noted that the majority of the clinical data are derived from the Impella 2.5 device and that the surgically placed Impella 5.0 device has yet to be subjected to rigorous clinical trial analysis. The IMPRESS in STEMI (European Impella vs. IABP Reduces Infarct Size in STEMI Patients Treated with Primary PCI) is an ongoing clinical trial that compares the Impella 2.5 with the IABP in pre-shock STEMI patients. Although these studies do not target overt heart failure or cardiogenic shock patients specifically, the results of these trials will help determine the best populations to target for use of these advanced devices.