End-stage Heart Failure (HF) represents a growing disease, responsible for about one-third of total deaths (1). Although Orthotopic Heart Transplantation (OHT) is the definitive treatment, the number of patients who die while on the waiting list is still too high. Different solutions have been proposed to expand the number of hearts available for transplantation (1): the use of Hepatitis C virus positive donors is possible thanks to the advent of direct-acting antiviral therapies, while the use of the Organ-Care-System employ for extended-criteria donors allows good short-term post-transplant survival. Moreover, donation after circulatory death is now an option. In addition, cardiac xenotransplantation from genetically multi-modified organ-source pigs is an emerging new option, as demonstrated by the consistent long-term success of heterotopic (non-life-supporting) abdominal and life-supporting orthotopic porcine heart transplantation in baboons, and by a recent “compassionate use” transplant of the heart from a genetically multi-modified pig with ten modifications into a terminally ill patient who survived for 2 months (2).

However, there remains an organ shortage, which contributes to the exponential spread of Mechanical Circulatory Support (MCS) (3). This implies that cardiologists and caregivers will increasingly be faced in their daily lives with patients implanted with MCS. Moreover, the role of physicians in understanding when optimized pharmacotherapy is no longer sufficient and, as a consequence, in deciding when to address patients for MCS or OHT screening is an issue set to grow in importance.

The majority of HF patients may be adequately treated with a Left Ventricular Assist Device (LVAD), but the small percentage affected by advanced biventricular dysfunction are candidates for long-term biventricular support. As expected, not only are biventricular HF patients a more ill population, but the use of Biventricular Mechanical Circulatory Support (BMCS) doubles the mortality rate and is associated with a higher incidence of complications (4).

Several classifications may be used to describe biventricular MCS (Table 1).

The first distinction concerns short-term and long-term devices; the former group includes both percutaneous and surgically placed non-dischargeable devices (clearly indicating that, due to the acute setting, patients cannot be discharged from the hospital and managed at home), while the latter is composed of the Total Artificial Heart (TAH) (orthotopic devices which requires excision of the heart) and heterotopic machines that support both ventricles in parallel to the native ones (Biventricular Assist Devices—BiVADs). This classification derives from the new OHT allocation 6-tiered system (5).

BIVADs may ensure a minimum flow sufficient to patient's survival in case of device failure and allow for recovery.

Other classifications are based on pumps mechanisms (pneumatic, electric, axial or centrifugal), type of flow produced (pulsatile or continuous) and on emplacement (paracorporeal or intracorporeal) (6).

Short–term percutaneous assistance is represented by the Impella (Abiomed, Danvers, MA) system and the TandemHeart (CardiacAssist Inc, Pittsburg, PA).

The Impella family of products (7) includes 4 microaxial pump sizes for left ventricular support (2.5, 3.5/CP, 5.0, 5.5 L/min) and one size for RV support (RP, Figure 1) which can be coordinated with each other to produce a biventricular assistance setting (Bi-Pella). For left-sided support, the pump is inserted into a peripheral artery and positioned retrograde through the aortic valve.

Strict check and monitoring of the device position across the aortic valve with fluoroscopy and echocardiography is mandatory. The Impella 2.5, Impella CP, Impella 5.0, and Impella 5.5 can provide antegrade flow up to 2.5, 4.0, 5.0, and 5.5 L/min respectively. By continuously drawing blood from the left ventricle (LV), the Impella unloads the LV, thereby decreasing its work and myocardial oxygen demand. In addition, by delivering large volumes of blood to the aorta, Impella operation results in an increase in mean arterial pressure and cardiac output, thus improving systemic perfusion and coronary flow which augments the chances of cardiac recovery. Finally, Impella leads to a decrease in pulmonary wedge pressure and a secondary reduction in right ventricular afterload (7).

Basing on the evidence that prolonged LV unloading with LVAD promotes reverse remodeling, cardiac repair, and pulmonary vascular resistance reduction, the impact of prolonged support (until 40 days) with an LV-Impella (the PROPELLA concept) has been investigated in the context of fulminant myocarditis with encouraging results (8).

The right-sided device (Impella RP) can provide up to 4 L/min of flow. This 22F pump mounted on an 11F catheter is inserted percutaneously over a guidewire guided by fluoroscopy and/or echocardiography from the femoral vein via the right heart to the pulmonary artery, unloading the RV.

Impella pumps cannot be coupled to an oxygenator and are appropriate for 5–21 days.

TandemHeart (CardiacAssist, Pittsburgh, PA, USA) (7) represents the first totally percutaneous biventricular assist device on the market: using a 21-Fr drainage catheter inserted into the femoral vein, which traverses the right atrium into the left, a centrifugal pump, and an arterial catheter in the femoral artery, it can achieve flows of up to 5 L/min. The system can also be placed in a central cannulation setup (Figure 2).

To achieve right heart assistance, an inflow cannula is placed in the right atrium and an outflow cannula in the pulmonary artery. Both are venous cannulas, generally accessed through the left and right femoral veins, respectively, or can be surgically placed after chest opening (Figure 3). Alternatively, an internal jugular vein can be accessed to place the outflow cannula or the more recent Protek DuoR dual-lumen cannula (CardiacAssist, Pittsburgh, PA, USA): it contains two lumens, with one serving as an inflow tract positioned in the right atrium, and the other serving as the outflow tract positioned in the pulmonary artery. It is able to provide support of intermediate duration (up to weeks) with the possibility of accommodating an oxygenator.

An alternative option of biventricular assistance is represented by the association of a left Impella via the axillary artery with a Protek Duo cannula inserted in the jugular vein connected to a TandemHeart pump. Such total percutaneous configuration allows patient's mobilization with lower risk of infections while providing total heart support (9).

Depending on the basal disease, left and right ventricle may present different delay or degree of recovery: the use of two separate pumps to assist the left and the right side allows progressive and distinguished trials of weaning by monitoring the response of the two ventricles to every decrease of their respective support (8).

A number of stand-alone (without an incorporated oxygenator) centrifugal pumps which can be coupled with various cannulas are available. These are surgically inserted, using either a central or peripheral technique. CentriMag (Abbott Laboratories, Abbott Park, IL) and Rotaflow (Maquet, Rastatt, Germany) are two common examples of magnetically suspended pumps which can support of up to weeks/months in duration (7). Compared with older-generation pumps, these newer devices have less stasis and turbulent flow, allowing a partial solution to hemolysis and thrombi formation.

The veno-arterial ECMO represents the most common BMCS device thanks to its simple and quick deployment: this centrifugal pump including an oxygenator/heat exchanger has traditionally been surgically implanted, but recently percutaneous cannulation has been increasingly used with a trend towards lower mortality and fewer complications (10). Available drainage cannulas size ranges from 19 to 29 Fr, whereas arterial ones range from 15 to 23 and they are chosen depending from patient's body surface area, vessels diameter and predicted needed flow support. Usually, a peripheral canulation under fluoroscopic and/or echocardiographic guidance is performed (femoral vein-femoral artery or femoral vein-axillary artery), but especially in case of post-cardiotomy failure or in the presence of significant arteriopathy, a central cannulation (right atrium-ascending aorta) is preferred.

ECMO can achieve up to 4–10 L/min flow. Maximal duration of support is very variable in relation to patient's age and basal condition: typically, haemorragic or infective complications begin to arise 10–15 days after the implantation (10).

All these short-term BMCS require a variable degree of anticoagulation.

Table 2 summarizes their respective main characteristics.

During ECMO therapy, most centers use unfractioned heparin to prevent thrombosis, with different therapeutic targets depending from the patient's procoagulant state and hepatic function, and from its tendance to spontaneously bleeding (for instance at the site of peripheric cannulas insertion or in thoracic drains in case of central surgical cannulation). However, heparin induced thrombocytopenia and heparin resistance are conditions frequently requiring the use of different anticoagulants. In this context, direct thrombin inhibitors like bivalirudin and argatroban have been established as alternatives. A recent meta-analysis (11) showed that the use of bivalirudine resulted in better clinical outcomes and in a reduced risk of thrombosis as compared to heparin.

Concerning Impella, according to the manufacturer, the goal activated clotting time (ACT) is 160–180 s: in some patients, purge heparin may be sufficient to achieve anticoagulation goals; when insufficient, the addition of titratable, supplemental non-purge heparin is required to provide optimal anticoagulation. Two different dextrose concentrations (5% and 20%) in the purge solutions have been tested with similar therapeutic activated partial thromboplastin time rates, thrombotic, and bleeding events. Nevertheless, anticoagulation in patients with Impella devices can often be complicated due to unpredictable purge flow rates, pre-existing coagulopathy, or heparin allergies. In this setting, a sodium bicarbonate-based purge solution (25 mEq/1,000 ml 5% of dextrose) currently represents the reliable and approved alternative option (12). In addition, argatroban and bivalirudin have been employed in the purge solution too, appearing to be safe with no bleeding or thrombotic complications (13, 14).

Pharmacological management varies according to assistance strategy: in a recovery strategy the administration of Levosimendan is commonly used to facilitate weaning with increased success rate (15).

All these systems are approved in the United States (US) for adult use; despite this, randomized trials are still lacking. If the Bi-Pella approach presents encouraging outcomes in cardiogenic shock probably due to its capacity for biventricular unloading compared to ECMO, a recent meta-analysis (16) seems to show that ECMO coupled with an unloading device gives an advantage in terms of mortality reduction. Such recent strategy (the so-called ECPELLA) is very promising and seems to be associated with better outcomes. A recent multicenter study by Schrage (17) demonstrated that early active LV unloading with an Impella device is associated with a lower early mortality and a higher likelihood of successful weaning from ventilation, in the absence of more complications. Moreover, the interval between ECMO implantation and Impella initiation directly correlated with increased mortality and adverse outcome. The importance of an efficacious LV unloading is also confirmed by Isath (18), who shows that in the setting of ECPELLA, utilization of Impella 5.5 provides greater hemodynamic support with a lower risk of complications compared to Impella CP or 2.5. These findings suggest that in the optic of cardiac recovery, LV unloading is essential in order to completely “set aside” the heart. This translates in the need of ECMO patients' careful monitoring looking for the initial signs of pulmonary congestion (especially in case of very low residual cardiac contraction) that require the placement or the upgrading of ventricular unloading.

In patients under short-term support the question of the evaluation of systemic and cardiac recovery becomes crucial, in order to decide if weaning is possible or if a shifting towards a long-term assistance program or OHT is required. Different protocols of progressive decrease of the support are available and they vary among centers. In every step of weaning, systemic perfusion must be maintained (stable mean and systolic arterial pressure, appropriate cardiac index without increase in inotropic support) with normal levels of lactate and liver enzymes. In parallel, echocardiography must assess the recovery of LV function (LVEF > 20%–25%, LVOT VTI > 10 cm, Mitral TDSa > 10 cm/s) and of RV function (TAPSE ≥ 19 mm, RVEF ≥ 25%, low RA/PCWP, high PAPi), which must remain stable or improve during the flow reduction trial (19).

A perfect short-term BMCS should be simple to implant (ideally with a percutaneous technique) and to manage, two characteristics which render ECMO the current more suitable assistance when compared to a double Impella, which requires a complex anticoagulation protocol, a strict echocardiographic monitoring of the device position and the contemporary management of a right and of a left pump with different settings.

Nevertheless, ECMO presents the negative side of representing a machine which “supports the organs and not the heart”; we can describe it as a temporary tool to preserve the vital organs while waiting for cardiac recovery. This means that, for instance, it can be deleterious if used without a parallel device unloading the left ventricle in case of profound cardiac dysfunction. Obviously, hemolysis and long-term complications including platelet count fall and dysfunction and systemic inflammation are common persisting defects of all currently available short-term devices.

Several devices have been added to and taken off the market over the years: so far, available options are: 1.

Syncardia CardioWest (SynCardia Systems, Inc., Tucson, AZ): TAH.

All biventricular devices are approved for Bridge To Transplantation (BTT) treatment only (20).

Currently, the only device approved for adult use in the US is the Syncardia, while the Berlin Heart is reserved for pediatric use in the US and for patients of all ages in Europe. The third option of two HeartMates 3 (also in a TAH configuration—the so-called HeartMate 6—with both ventricles excised) is feasible but off-label.

Differently from paracorporeal or intracorporeal BiVADs, Syncardia TAH and HeartMate 6 require the excision of both ventricles, thus preventing the possibility of removing the devices in case of cardiac function recovery.

Several papers have compared the results of each approach: a recent meta-analysis (41) evaluated TAH and two continuous-flow BiVAD devices at 120 days follow-up, highlighting a similar mortality rate (36% vs. 26% respectively) but a better quality of life for intracorporeal devices with more patients discharged home despite a longer duration of support. Rates of hemorrhagic complications and infections were comparable. Levin (42) drew the same conclusion and confirmed that monoventricular assistance is associated with a better survival rate (92.7%).

Interestingly, a retrospective study (43) on a previous implantation period (2004–2012) observed an increased mortality for Syncardia, suggesting a technical improvement in the subsequent years. Conversely, the “Groupe de Réflexion sur l’Assistance Mécanique” describes a similar mortality rate before (75% at 30-days and 57% at 180-days follow-up) and after OHT (81% at 1-month and 64% at 5-years follow-up) in all BiVADs types and a trend towards an increased survival with >90 days of Syncardia support, probably due to a lower incidence of strokes (44).

All BMCS require the coupling of anticoagulation and antiplatelet drugs therapies with variable protocols. As for short-term devices, the degree of anticoagulation needs to be adapted to every patient, especially in the first hours after the implantation of the device, considering the individual perioperative bleeding state and risk.

Usually, the standard of care is represented by an initial treatment with unfractioned heparin, followed by vitamin K antagonists such as warfarin with the adjunct of acetylsalicylic acid for long-term maintenance. Nevertheless, bivalirudin is gaining space due to its short half-life, lack of dependence on antithrombin, and more stable pharmacokinetics; thus, at present, its use is not limited to situations wherein heparin is ineffective or contraindicated. Especially in children assisted with continuous flow devices or Berlin Heart, outcomes with direct thrombin inhibitors (bivalirudin or argatroban) are encouraging with lower major bleeding and stroke event rate than that reported with unfractioned heparin (45).

Concerning long-term anticoagulation, preliminary evidence suggests that new direct oral anticoagulants, which directly inhibit Factor Xa (rivaroxaban, edoxaban and apixaban) or Factor IIa/thrombin (dabigatran), may be a valuable alternative to warfarin in adults patients assisted with continuous-flow LVAD, but further investigations are needed (46).

No other specific treatments are needed when biventricular dysfunction is judged irreversible. Otherwise, in case of possible partial or complete recovery (typically in peripartum cardiomyopathy and myocarditis) and if biventricular implantation is not in a TAH configuration, all drugs employed in an acute HF setting, such as inotropes and levosimendan, and in the treatment of chronic systolic heart failure, may be administered. More in detail, neurohormonal inhibitors (angiotensin-converting enzyme (ACE) inhibitors, angiotensin receptor blockers (ARBs), aldosterone antagonists and beta blockers) could facilitate recovery (47) and significantly reduce the risk of mortality in VAD patients (48).

Positive hemodynamic effects have also been demonstrated with other vasoactive substances, including exogenous natriuretic peptides (nesiritide, ularitide), the endothelin-1 antagonist tezosentan, cinaciguat (an activator of soluble guanylate cyclase) and serelaxin, but their role in acute decompensated HF remains uncertain (49).

In acute cardiogenic shock requiring assistance, no time is available for speculation and ECMO is the preferred MCS for its simple implantation. Once hemodynamic situation is stabilized, patient's evaluation and estimation of MCS perspective are mandatory. Clinicians must answer to the following questions: 1.

Prevalent mono- or biventricular failure?

Different scenarios may arise: -

Need for longer short-term MCS to allow recovery or bridge-to-decision: shifting vs. CentriMag or biventricular percutaneous assistance combination is conceivable.

According to the INTERMACS Registry, 24% of patients implanted with a Syncardia were discharged home with their device after 1.6 months, with this number increasing over time. Even if TAHs improve quality of life for most of the implanted patients, rehabilitation and hospital discharge remain very challenging for patients and their care providers. A 6-kg driver that permits hospital discharge is now available to patients waiting for OHT and supported with the Syncardia. For years, one of the major disadvantages of the use of a TAH was the necessity of keeping the patient in-hospital while waiting for transplantation. Clinical experience with the new small TAH driver has allowed a few centers to discharge patients home (50).

Concerning Berlin Heart Excor, the possibilities of patients' hospital discharge seem more encouraging: adult patients requiring a blood pump size of 60 and 80 ml can be discharged with the EXCOR R Mobil by Berlin Heart GmbH. So far, the only other option for patients requiring smaller blood pump sizes is in-hospital treatment with the stationary IKUS driving unit (Berlin Heart GmbH) that can be used for all EXCOR R blood pumps. Unfortunately, the IKUS has a battery life of only 30 min and thus considerably limits patients' mobility. Focusing on this particular patient group, Berlin Heart GmbH has recently developed a novel electro-pneumatic mobile drive unit called EXCOR R Active, that has been especially designed for the operation of the paracorporeal system EXCOR R VAD and can be used for either univentricular assist or for biventricular assist. The use of the EXCOR R VAD in combination with EXCOR R Active as its driving unit has already been proven to be feasible for internal and external usage within the setting of professional health institutions. Unfortunately, this system, which has a guaranteed battery life of approximately 12 h, has not yet been approved for use in a home healthcare environment (51).

The experience with two continuous-flow BiVADs is currently too limited for drawing any conclusion about its capability to allow a simpler hospital discharge compared to TAH or Berlin Heart. If, on the one hand, the portable console unit is quite small and light, on the other hand, the patient is obliged to carry two consoles with two distinct settings, making it a little more difficult to achieve an independent and satisfying quality of life.

The ideal BMCS does not yet exist. Available tools (TAH or BiVADs) are exclusively indicated in a BTT therapeutic strategy (Figure 9) and ensure a survival rate of about 70%, confirming the gravity of such a clinical condition. Moreover, they are associated with a significant incidence of major complications (infective, thromboembolic, hemorrhagic) and greatly limit patients' quality of life because of their encumbrance and noise. Aside from this, no solution for destination therapy is available. Several clinical and pre-clinical trials concerning new devices that can potentially address the weaknesses in BMCS are underway, with encouraging preliminary results.