The history of mechanical circulatory support (MCS) parallels that of cardiac surgery and began with the introduction of the Gibbon bubble oxygenator.¹ That revolutionary device opened up the field of cardiac surgery and was one of the major medical advances of the last century. Soon after the use of cardiopulmonary bypass (CPB) became commonplace, failure to wean from CPB became recognized as a problem whose solution required temporary cardiac support to enable postcardiotomy cardiac recovery. Spencer and coworkers utilized postoperative femoral CPB in 3 patients, one of whom survived and ushered in the modern era of temporary MCS.² This was soon followed by the use of the first extracorporeal mechanical assist device by DeBakey et al. in 1964 and the subsequent development of the intra-aortic balloon counterpulsation device, variations of which are still in use today. Kantrovitz, Cooley, Oyer, Devries, and others were responsible for other notable firsts including the first successful bridge to transplant (BTT) and first successful total artificial heart (TAH) implant.^3–7 Notable clinical landmarks in the history of MCS are shown in Table 51-1. In this chapter, we review the major indications for MCS as well as the various options available.

Generally speaking, there are three primary indications for the use of MCS: “bridge to recovery” (BTR), “bridge to transplant” (BTT), and “destination therapy” (DT). Data on MCS devices implanted in the United States are kept by the Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS).⁸ According to the second INTERMACS annual report that provides information on FDA-approved device implantations from 2006 to 2009, the most common indication for MCS was BTT (45.4 percent), followed by DT (9.2 percent), and BTR (2.3 percent) (Fig. 51-1). Bridge to candidacy for transplantation, or “bridge to decision” is a fourth category that has emerged and includes patients whose candidacy for transplantation has not been decided for medical or social reasons. This accounted for 41.9 percent of all device implantations during this time period. Within each of these categories, specific device therapy can be broken down into short-term (days to weeks) and long-term (weeks or longer) therapy.

BTR is used in patients who have severe, acutely decompensated heart failure from potentially reversible causes who may eventually recover myocardial function but whose native heart is unable to provide sufficient end organ perfusion without assistance. This therapy serves to unload the damaged heart to allow remodeling and recovery while providing improved tissue perfusion to vital organs. Indications for BTR therapy include postcardiotomy support following cardiac surgery, acute viral cardiomyopathic syndromes, hemodynamic instability, and myocardial infarction. For BTR, short-term devices are typically used as their cost is lower and implantation is easier. In certain patients in extremis and in whom neurologic status and candidacy for cardiac transplantation or longer term mechanical assist therapy is unknown, a short-term device can potentially serve as a bridge to a long-term device in a “bridge to a bridge” scenario in which the short-term device is replaced by a long-term device for DT or BTT prior to cardiac transplantation. BTR has also been used in patients with chronic advanced CHF with the hope that a period of unloading will enable ventricular remodeling and recovery, allowing device removal.

BTT is currently the most common indication for MCS today. The intent of device therapy for this group of patients is not only to provide temporary support for a failing heart but also to improve end organ function, making the potential recipient healthier prior to transplantation. Studies have demonstrated improved survival to transplantation and improved posttransplant survival compared with inotrope supported patients.⁹

DT is the phrase coined to describe the use of left ventricular assist devices (LVADs) as a final therapy to support patients until their death. These patients are those without expected recovery of myocardial function, but who are not candidates for cardiac transplantation due to older age, obesity, smoking, cancer, pulmonary hypertension, social concerns, or other advanced organ dysfunction. Longer term support devices intended for extended treatment periods are utilized for this purpose, and this is currently the second most common indication for device implantation today.

MCS devices can be classified by their mechanism of action (Table 51-2). These include the first-generation pulsatile devices, the second-generation continuous axial flow devices, the third-generation centrifugal pumps, and TAHs.¹⁰ Some FDA-approved devices are listed in Table 51-3.

Early Devices

Early devices for MCS were easy to insert, relatively inexpensive, and the most easily available of the assist devices. They have been designed for short-term support lasting hours to days, and are most often employed for postcardiotomy failure.

Nonpulsatile, Centrifugal/Rotodynamic Radial Flow Pumps

Blood flow for these devices is generated by rotating blades or impellers.¹¹ These devices do not have valves or multiple moving or occluding parts (Fig. 51-2), theoretically reducing the potential for hemolysis. These devices require priming with volumes differing from pump to pump (52–87 mL). Devices of this class are capable of providing high flow rates with low rises in pressure. Several commercial devices are available, including the Carmeda Bio-Pump and the BIO-PUMP (Medtronic Inc, Minneapolis, MN), the Sarns pump (Terumo Cardiovascular Systems Ann Arbor, MI), and the Nikkiso pump (Nikkiso Pumps America Inc., Plumsteadville, PA). These devices are not designed for long-term use. Indications include left heart bypass for thoracic aortic surgery, postcardiotomy ventricular failure, extracorporeal membrane oxygenation (ECMO), and bridge to transplantation or to another VAD. Cannulation sites commonly used for left heart bypass include the left superior pulmonary vein or left atrium (inflow) and the ascending aorta or femoral artery (outflow). For right heart bypass, inflow is obtained via right atrial cannulation with outflow directed into the pulmonary artery. Cannula fixation is important with these and other devices, since most hemorrhagic complications are due to cannula dislodgement. Anticoagulation is required after device insertion and must be monitored closely.

There are no firm criteria for device weaning and removal, but the achievement of hemodynamic stability and a documented cardiac recovery dictate the removal strategy. Typically a gradual reduction in support is initiated over hours to days with device removal once minimal support is required. When a device of this class fails, it is often due to breakdown of the pump head seal, leading to the entry of fluid into the magnetic chamber.¹² Clinical experience has shown that these pumps are especially well suited for left heart bypass and for short-term postcardiotomy left ventricular failure. In one study, 62 patients were supported using the centrifugal pump for postcardiotomy failure; 22 patients required left ventricular support, 9 required right ventricular support, and 31 required biventricular support. Support was extended for up to 19 days, with 42 patients weaned successfully, 27 patients discharged home, and 18 patients surviving more than 1 year.¹³ Other studies have reported similar results for short-term support.^14,15

Paracorporeal Pulsatile Devices

Abiomed BVS 5000.

This device was approved by the FDA in 1992 for use in all types of recoverable heart failure. It is an external pulsatile VAD capable of supporting the right and/or left ventricles for days to weeks. This device is a pneumatically driven, asynchronous, pulsatile, automated, self-regulating, polycarbonate-housed dual-chamber pump system (Fig. 51-3). The atrial chamber of this two-chamber pump fills passively by gravity drainage; the volume is usually about 100 mL, and the internal chamber is made of flexible polyurethane. The atrial chamber drains into the ventricular chamber through a trileaflet valve made of proprietary material. During systole, compressed air is delivered into the ventricular chamber; it compresses the polyurethane bladder and ejects its contents toward the patient. The pumping unit and console are located outside the body; the inflow and outflow cannulas are tunneled subcutaneously into the mediastinum. Each pump is capable of delivering a maximal output of 6 L/min at a constant stroke volume of 80 mL.

The implantation technique for the BVS 5000 varies from center to center. For right ventricular support, the inflow cannula is placed into the mid-right atrium and secured firmly with pledgeted purse-string sutures. Another alternative is to insert this cannula directly into the right ventricle through its diaphragmatic surface, particularly if there are adhesions from previous surgery. The outflow cannula is anastomosed to the main pulmonary artery. For left ventricular support, the inflow cannula is placed into the left atrium via the right superior pulmonary vein, left atrial dome or directly into the left ventricle via the apex. Left ventricular apical cannulation permits complete ventricular decompression, optimizing its chance for recovery. Left ventricular apical cannulation is also advantageous if a mechanical mitral prosthesis is in place, since it permits opening and closing of the leaflets and may reduce thrombotic complications. The outflow cannula is anastomosed to the ascending aorta.

Heparin anticoagulation is required during BVS 5000 support and must be closely monitored; an activated clotting time (ACT) of 180 to 200 s is the goal. The pump is easy to manage and operated based on the principles of preload and afterload management. Limitations of this device include limited patient mobility, restricted flow capacity, and heparinization requirements.

A prospective multicenter trial evaluated the safety and efficacy of the BVS 5000 in 31 patients with hemodynamic instability despite maximal pharmacologic and IABP support. There was an improvement in mean arterial pressure from 50.1 ± 15.3 mm Hg prior to support to 77.1 ± 8.0 mm Hg after support was initiated (p <0.01). The cardiac index increased from 1.6 ± 0.6 L/min/m² before support to 2.3 ± 0.3 L/min/m² after support was initiated (p <0.01). A total of 76 percent of the cohort (42 patients) experienced bleeding; 17 (55 percent) were successfully weaned from support, and 9 (29 percent) were discharged.¹⁶

In a retrospective review, the BVS 5000 was implanted in 47 patients exhibiting postcardiotomy acute heart failure. There were 38 patients in the BTR group and 9 in the BTT group. A total of 25 patients (66 percent) in the BTR group were weaned, and 16 patients (42 percent) went on to discharge. In the BTT group, 1 patient recovered myocardial function and 1 died while awaiting transplantation, although 7 patients (77 percent) underwent successful cardiac transplantation with a posttransplant survival rate of 66 percent. Patients were supported in the isolated left ventricular (28 percent), biventricular (45 percent), and right ventricular modes (28 percent).¹⁷ In a more recent study involving 202 patients, survival at 3 days, 30 days, and 5 years was 76, 38, and 24 percent, respectively. Patients surviving 30 days experienced a 63-percent 5-year survival. A total of 48 patients (21 percent) were bridged to transplantation, and 71 patients (35.5 percent) were weaned with intent for survival.¹⁸ Similar results are reported from other centers.¹⁹ The wean and discharge rates is generally around 60 and 40 percent, respectively, and this approach has been shown to be cost-effective as well.²⁰

Abiomed AB5000.

The next-generation device to follow the BVS 5000 was the Abiomed AB5000 (Fig. 51-4). This device was an updated, improved version of the BVS 5000. This pneumatically driven device utilizes the same cannulas as the BVS 5000, allowing an easy transition to this device. The AB5000 provides flows up to 6 L/min and is capable of right- or left-sided support and has a console that enables patients to become mobile within the hospital setting.

The experience with transition from BVS5000 to AB5000 in the same patient has also been good. A report of 50 patients who initially underwent BVS5000 implantations at primary hospitals who were then transferred to tertiary care centers and were converted to the AB5000 device reported a 42 percent survival rate to either recovery, DT implantation, or transplant.²¹ Of the surviving 21 patients, 18 were alive 30 days after device removal with 61 percent able to be weaned, 5.6 percent receiving a DT device, and 33 percent proceeding to transplantation.

The initial published experience with primary device implantation in patients with AMI and hemodynamic instability or postcardiotomy shock was encouraging with several patients able to be weaned from their support prior to successful hospital discharge.²² A more recent report from the national AB5000 device registry among 100 patients who underwent AB5000 implantation for cardiogenic shock following AMI demonstrated a 40 percent 30-day survival with 63 percent successfully weaned from support, 17 percent transitioned to another support device, and 20 percent proceeded to transplantation. Of those who achieved myocardial recovery and device explantation, 2-year survival was 78 percent.²³

First-Generation, Pulsatile Devices

The first-generation devices have a pulsatile mode of action with flow originating from a pumping chamber separated by two valves, one for inflow and another for outflow. These are pneumatically or electrically driven and have a cyclical mode of filling and emptying.

Thoratec VAD

The Thoratec VAD (Thoratec, Pleasanton, CA) was approved by the FDA in 1996 for use for BTT and subsequently in 1998 for BTR; this is the only approved dual-use device on the market. It was first used clinically in 1982 for postcardiotomy failure. Initially labeled the Thoratec VAD, with the subsequent development of the Thoratec implantable VAD (iVAD), this device became known as the Thoratec paracorporeal VAD (pVAD) and will be further referred to as such.

The Thoratec pVAD device is flexible and can be employed for short- or long-term use in both large and small patients because of its paracorporeal design. It is a pneumatically driven paracorporeal pump that uses Delirin tilting monostrut mechanical disk valves for maintaining unidirectional blood flow (Fig. 51-5). There is a flexible blood sac housed in a rigid outer casing; this sac is compressed by air to achieve ejection. Separate pumps are needed for each side of the heart. Three modes of operation are available: fill-to-empty, fixed-rate, and electrocardiogram (ECG)-synchronous modes. The most commonly used mode is the fill-to-empty mode, which pumps at a rate determined by VAD filling. The inflow and outflow cannulas being the only internal components, this pump can be used for small patients, including children. The cannulas are made of polyurethane except for the distal part, which is made of polyester Dacron. Usually CPB is needed for implantation of the cannulas. For left ventricular support, inflow can be achieved via the left ventricular apex or the left atrium; outflow is usually to the ascending aorta. For right heart support, inflow is via the right atrium and outflow is directed into the main pulmonary artery. Proper positioning of the cannulas is essential for reducing undue kinking of the cannulas and optimal device positioning on the abdominal wall. Anticoagulation with heparin followed by warfarin is required. The main limitations of this device include rather bulky paracorporeal pumping chambers and restricted patient mobility due to a large drive console.

More than 4000 patients have undergone Thoratec pVAD device implantation for left, right, and biventricular support with the longest duration of support of 1204 days. This device has been particularly useful for short-term BTT in patients suffering from biventricular failure.²⁴ In an analysis of 828 BTT patients, this system was used for biventricular support in 472 cases, left ventricular support in 326 cases, and right ventricular support in 30 cases for up to 515 days. During the support periods, cardiac indices increased significantly from 1.4 ± 0.8 to 3.0 ± 0.5 L/min/m² (with biventricular assistance and left ventricular cannulation). Of the 828 patients, 60 percent underwent transplantation, and the posttransplant survival rate was 86 percent. In the 195 patients who needed postcardiotomy support, the device was used for up to 80 days for cardiac recovery. Thirty-eight percent of the patients were weaned from the device, with 59 percent of these patients discharged. Of 49 postcardiotomy patients considered for transplantation, 32 underwent transplantation and 23 were discharged. Other series reflect similar results.^25,26

Newer series also validate the earlier experiences with the pVAD. Among 84 patients who underwent pVAD implantation, 56 percent were able to wean from support or proceed to transplantation with actuarial survivals of 79, 73, and 63 percent at 1, 3, and 5 years after device implantation.²⁷ Another recent series with 71 patients with a total of 5193 patient-days of support had 65 percent of patients either bridged to recovery or to transplantation with actuarial survivals of 76, 68, and 62 percent at 1, 2, and 3 months, respectively.²⁸

An experience with a new portable drive unit designed to allow home discharge demonstrated excellent clinical outcomes with an average discharge time of 62 days (range 16–242) and 89 percent of patients surviving to myocardial recovery and device explantation or transplantation.²⁹ The majority of postimplantation time (83 percent) was outpatient with significant improvement in CHF symptoms to NYHA I or II following device implantation.

Use of the pVAD for biventricular support has also been examined. Among a series of 73 patients who had ThoratecpVAD implanted as both right ventricular assist device (RVAD) and LVAD, 84 percent went on to transplantation, 10 percent were weaned and had successful device explantation, and 6 percent remained with BIVAD support.³⁰ Overall survival was 69 percent and 5-year survival for those receiving a heart transplant was 58 percent. For patients with high likelihood of right ventricular failure after LVAD implantation, the decision to perform both RVAD and LVAD implantation at the same time leads to improved outcomes compared with delayed BIVAD implementation.³¹ Among 99 patients who underwent Thoratec BIVAD placement, there was a trend toward improved bridging to transplantation (65 vs 45 percent, p = 0.10) and a significantly improved survival to discharge (51 vs 29 percent, p <0.05) among patients with immediate BIVAD placement compared with those who underwent delayed RVAD implantation after initial LVAD implantation. In a separate study, risk factors for eventual need of RVAD following LVAD placement included cardiac index <2.2 L/min/m², RV stroke work index <0.25 mm Hg · L/m², severe preoperative RV dysfunction, preoperative creatinine >1.9 mg/dL, reoperative cardiac surgery, and systolic blood pressure <96 mm Hg.³²

In 2004, the FDA licensed the Thoratec iVAD for BTT, with identical features to that of the pVAD but with an intracorporeal placement (Fig. 51-6). Results from two groups have been published.^33,34 Slaughter et al. reported 39 patients who underwent placement of the iVAD who were supported for 3938 patient-days. This group was compared with a historical group of 100 patients who underwent implantation of the Thoratec pVAD device. Survival outcomes were similar for both iVAD and pVAD for BTT (70 vs 69 percent survival to transplantation) and slightly better for BTR (67 vs 48 percent survival to explant and myocardial recovery) indications, respectively. Berman et al. reported 24 patients who were supported for 2308 patient-days.³⁴ Survival to transplantation was 71 percent in this cohort. Taken together, these results suggest that the Thoratec iVAD has results similar to the pVAD with the advantages of its intracorporeal position. To date, more than 500 patients have undergone implantation of the iVAD device with the longest duration of support of 979 days.

Heartmate LVAD

The HeartMate LVAD (Thoratec Corp., Pleasanton, CA) was designed in 1975 as an implantable, pulsatile, pneumatically actuated (HeartMate IP), intracorporeal left ventricular assist system.³⁵ The pneumatically driven unit is made of sintered titanium and houses a flexible, textured polyurethane diaphragm. It is driven by a pusher-plate mechanism that, in turn, is driven by compressed air and controlled by a console (Fig. 51-7). The design was modified to a vented electrical system (HeartMate VE) in 1991 with a much smaller console, which confers greater patient mobility.³⁶ The inflow and outflow conduits utilize porcine xenograft valves (Medtronic-Hancock, Minneapolis, MN). The maximum stroke volume generated is about 85 mL and maximum output can reach up to 11 L/min. The blood-contacting surfaces, unlike those of other devices, are made up of textured polyurethane, which facilitates deposition of a fibrin-collagen matrix, forming a pseudointimal layer. This feature eliminated the anticoagulation requirement, calling for aspirin only. To accommodate the device, the patient’s body surface area (BSA) must to be at least 1.5 m². The pump is inserted either pre- or intraperitoneally in the left upper quadrant of the abdomen. There are two modes of operation: fixed mode (20–140 bpm for IP and 50–120 bpm for VE) and automatic mode. In the latter mode, the pump senses when its chamber is full and activates the pusher-plate mechanism (fill-to-empty). The battery life usually ranges from 4 to 6 h, with a portable hand pump backup available for device malfunction or power failure. Following the successful REMATCH trial, the enhanced version of the HeartMate SNAP-VE, the HeartMate XVE LVAS, was approved by the FDA for DT in 2002 (Fig. 51-8). The notable modification was the inflow valve conduit, which was designed to be six times more durable than in previous versions. To date, more than 1300 patients have undergone HeartMate IP LVAS implantation, although close to 2500 patients have received the HeartMate VE LVAS, with cumulative patient-years of 330 and 964 years, respectively. Over 4500 patients have received the newest HeartMate XVE device with the longest duration of support of 1864 days.

Experience from Columbia University of 243 patients undergoing implantation of HeartMate devices included 52 pneumatic, 17 dual-lead vented electrical, and 174 single-lead vented electrical devices. Overall actuarial survival rates at 1, 3, 5, and 10 years posttransplant were 90.5, 85.1, 69.6, and 39.6 percent, respectively. The overall incidence of infection was 17.7 percent (n = 43). Device malfunctions occurred in 32 patients.^37,38 A pivotal prospective multicenter clinical trial was conducted at 24 centers in the United States. There were 280 transplant candidates (232 men, 48 women; median age 55 years) unresponsive to inotropic drugs, intraaortic balloon counterpulsation, or both who received the device. These patients were compared with a control group of 48 patients who were not supported with a device. HeartMate VE support lasted an average of 112 days, with 54 patients supported for more than 180 days. Device-related adverse events included bleeding in 31 (11 percent), infection in 113 (40 percent), neurologic dysfunction in 14 (5 percent), and thromboembolic events in 17 (6 percent) of patients. Successful BTT was observed in 71 percent of HeartMate patients, compared with 33 percent of the control group (p <0.001). A total of 198 patients survived, with 188 patients undergoing cardiac transplantation. One-year survival after transplantation was significantly better in patients who underwent LVAD implantation compared with the control population [84 percent (158 of 188) vs 63 percent (10 of 16); log rank analysis p = 0.0197)].³⁹

A more recent report by Long et al. compared 42 patients who underwent HeartMate XVE implantation to those who received HeartMate VE devices from the REMATCH trial.⁴⁰ This study showed that there was improved survival with the newer HeartMate XVE compared with the VE (30 day: 90 vs 81 percent, 1 year: 61 vs 52 percent), as well as a decrease in adverse events. Whether the improvement in outcome was due to the advancements in technology or an improvement in patient selection and postoperative care is not known. Despite these outcomes, the HeartMate XVE has had issues with durability with device exchange or failure of 18 and 73 percent at 1 and 2 years, respectively.³⁸

Novacor LVAS

The Novacor LVAS (World Heart Corp., Ottawa, Canada) was developed in collaboration with Stanford University and was originally planned as a totally implantable system. It was the first device to bridge a patient successfully to transplant in 1984.⁴¹ Over the years it evolved into a console-based VAD. The pump drive unit incorporated a dual pusher-plate sac-type pump with a smooth blood-contacting surface. The stroke volume was about 70 mL. It had two porcine-valved polyester inflow and outflow conduits. The system had a high-efficiency linear motor using a pulsed solenoid energy converter with a two-armature assembly. Thus it required no gears, cams, or intermediate hydraulic conversion, reducing the potential for mechanical failure. There were three modes of operation: single-stroke, fixed-rate, and automatic or synchronous (fill-to-empty) modes. The pump filled passively, aided by the large circular dual-pusher-plate surface area (Fig. 51-9). The power packs had 6 h of work life and could be worn on a belt, vest, shoulder bag, or backpack. Anticoagulation with warfarin was necessary and was closely monitored. The primary cause of mechanical failure was wearing out of the energy converter, although this usually was detected at least 3 months prior to its ultimate failure.⁴²

The Novacor LVAS was successfully used for the past two decades as a bridge to cardiac transplantation in patients with end-stage congestive heart failure. Stanford University reported a series of 53 Novacor patients (48 male, 5 female) with a mean age of 44 years (range of 16–62 years) and a mean support time of 56 days (range of 1 to 374). Complications consisted of bleeding (43 percent), infection (30 percent), and embolic cerebrovascular events (24.5 percent). Sixty-six percent of the supported patients were successfully bridged to cardiac transplantation.⁴³ A recent study suggested that, in terms of durability, the Novacor device exceeded the HeartMate VE at 2 months (93.5 percent). Novacor’s durability at 3 years was 85.9 percent, with 78 percent of supported patients surviving to transplantation.⁴⁴ Another study reported improvement in cardiac output, wedge pressure, pulmonary vascular resistance, and mean pulmonary pressure after Novacor implantation. Most of the complications observed were related to thromboembolism and occurred in the first 3 months after implantation.⁴²

The Novacor was discontinued in 2008 with later-generation devices from the same manufacturer (Levacor LVAD) taking its place.

Lionheart LVAS

Developed at Penn State University, the LionHeart LVAS (Arrow International, Reading, PA) was specifically designed for DT in patients with end-stage chronic heart failure.⁴⁵ The system was fully implantable without any external conduits for power. This was the first system to not require a percutaneous driveline and gave hope to the eventual development of a totally implantable device. It utilized a brushless motor that actuated a pusher plate using a roller-and-screw mechanism and Delirin disk monostrut valves. The blood-contact surfaces consisted of polyurethane sacs, which filled passively. The stroke volume was approximately 65 mL and the maximum achievable flow rates were 8 L/min. The LionHeart required a gas-filled compliance chamber implanted in the left pleural space to accommodate pump volume displacement (Fig. 51-10). In addition, there was a subcutaneous infusion port that was replaced every 2 to 4 weeks. The battery and system controller were all implanted in the right lower quadrant of the abdomen. The LionHeart’s pumping characteristics, based on end-diastolic volumes, were automatically adjusted by a software algorithm, which attempted to provide maximal filling of the pump with each stroke by altering pump speed. The unique feature of this LVAS system was the transcutaneous energy transmission system (TETS), in which electrical energy was supplied to the pump’s drive unit by radiofrequency induction. Warfarin anticoagulation was required with this device.

The first clinical implantation of the LionHeart was performed in 1999; the first implantation in the United States was performed in 2001.⁴⁶ The Clinical Utility Baseline Study (CUBS) was designed to determine the safety and performance of the LionHeart LVAS to serve as a permanent mode of circulatory support for patients with end-stage heart failure who were ineligible for heart transplantation. As a part of this trial, 26 male patients underwent implantation at seven European Centers; actuarial survival rates were 86 percent at 1 month, 45 percent at 6 months, 41 percent at 1 year, and 34 percent at 2 years.

The LionHeart LVAS was discontinued in 2005 by the manufacturer in favor of development of the newer CorAide device.

Second-Generation, Rotary Axial Pumps

The second-generation mechanical circulatory assist devices sought to improve upon the design of the first-generation devices and to address some of the potential complications with the earlier technology. These are smaller, valveless, implantable devices that provide continuous axial flow (Fig. 51-11). An internal rotor suspended by contact bearings within the device is electrically driven by an external power source.

The first-generation pulsatile LVADs are hampered by percutaneous drivelines and bulky consoles and support systems. In addition, these assist devices are generally not suitable for patients with body surface areas of less than 1.5 m² or in pediatric patients with the exception of the ThoratecpVAD. Other characteristics of these devices include bleeding complications and difficulty in explantation. These limitations prompted the development of axial flow pumps. Technically, nonpulsatile flow pumps are more attractive than pulsatile systems because they have one moving part, permitting compactness in design, ease of insertion, and lower energy consumption. These devices are inserted into the ventricular apex (inflow) and anastomosed to the thoracic ascending or descending aorta (outflow). They do not require valves and compliance chambers, affording additional simplicity. Despite these advantages, several potential drawbacks exist. Device failure could potentially result in significant backflow of 1.5 to 2.0 L/min from the aorta into the left ventricle. Also, blood-washed bearings were not studied prior to the HeartMate II experience to adequately determine thrombogenic potential and wear characteristics. Early research showed little bearing wear and a reasonable thrombogenicity profile.⁴⁷ Furthermore, earlier concerns about the nonphysiologic nature of nonpulsatile blood flow persists, although results from rotary axial flow pumps have been encouraging. Table 51-4 summarizes the clinical data available for the second-generation devices.

Debakey VAD

The DeBakey VAD (MicroMed Technology Inc., Houston, TX) resulted from a collaboration between the Baylor College of Medicine and NASA; it is undergoing clinical trials in United States and European centers.^48,49 A small (30-mm diameter, 76-mm length, 95-g weight) titanium axial flow pump, the DeBakey VAD is based on an axial impeller–inducer supported by an inflow straightener and outflow diffuser housing ceramic blood-lubricated bearings. The pump produces a flow of 5 to 6 L/min and generates a pressure of 100 Torr at 10,000 rpm. An ultrasonic flow probe continuously measures the pump output. The inflow cannula is placed in the left ventricular apex and the outflow cannula can be anastomosed to the ascending or descending aorta. The pump is placed in a small preperitoneal pocket. A percutaneous driveline connects the power supply and control to the pump from an external console. Pump flow is nonpulsatile and requires warfarin anticoagulation. Laboratory indices reveal no significant hemolysis or changes in plasma free hemoglobin.

Worldwide experience with the DeBakey VAD includes 150 patients who underwent placement of the device as a BTT. In this series, 82 patients (55 percent) were bridged to transplantation/recovery or are currently supported and 68 patients (45 percent) died with the device.⁵⁰ Reoperation for bleeding was the most common adverse event in 48 patients (32 percent), followed by hemolysis (12 percent) and device infection (3.3 percent). Device failure occurred in four cases. Seventeen pumps (11 percent) were noted to have detectable thrombus; of these, eleven cases (64 percent) were successfully resolved with transplantation, pump exchange, or thrombolysis. Thromboembolic events occurred in 10.7 percent of patients. There is currently an ongoing clinical trial for DT (DELTA: Destination Evaluation Long-Term Assist) using the DeBakey VAD and Thoratec HeartMate XVE.

Flowmaker (Jarvik 2000)

The Jarvik 2000 (Jarvik Heart Inc., New York, NY) is a titanium axial flow pump (25 mm diameter, 51 mm length, 90 g weight) that is implanted in the left ventricular apex. The outflow graft is anastomosed to either the ascending or the descending aorta depending on the surgical approach (Fig. 51-12).⁵¹ The pump can be implanted by either a sternotomy or a thoracotomy, and with or without cardiopulmonary bypass. The lack of a pump pocket and the use of a unique postauricular connector that brings power through the infection-resistant scalp tissue behind the ear has reduced infection to less than 5% with infection rate of only .03 events/patient year. The Jarvik 2000 is available for BTT under a CAP approval and is used in a randomized destination therapy trial approved for 350 patients at up to 50 U.S. medical centers. It has CE mark and is available throughout Europe.

Three different options are available for energy transmission: a percutaneous lead, a skull-mounted titanium pedestal, and Transcutaneous Energy Transfer and Telemetry System (TETTS).⁵¹ The Flowmaker is capable of producing flows of 7 L/min at 8000 to 12,000 rpm and requires warfarin anticoagulation. Pump flow rates can be adjusted manually in times of increased activity or can be set at a fixed rate.

The first patient to receive a Flowmaker was a 61-year-old with dilated cardiomyopathy and NYHA class IV heart failure who was implanted in 2000 for DT.⁵² This patient was alive and was highly functional with an improvement to NYHA class II heart failure 6 years after implantation.⁵³

Frazier et al. described the experience of 26 patients who underwent implantation of a Jarvik 2000 device for BTT (22 patients) or DT (4 patients).⁵⁴ In this series 13 patients (55 percent) underwent successful transplantation after a mean support of 67 days.

Since these early reports, over 100 patients have been implanted with this device internationally.⁵⁵ Siegenthaler reported on a multicenter (European and American) experience in 102 patients who were implanted as BTT (83 patients) or DT (19 patients). There were no mechanical device failures either during in vivo use or in ex vivo bench testing over a total of 110 cumulative years (59 years in vivo, 51 years ex vivo). Mean device support times for BTT and DT were 159 and 551 days, respectively. Six of the 19 DT patients were alive for greater than 2 years and one patient nearly 5 years after implantation. The Flowmaker may prove to be a reliable, long-term solution for DT and a viable solution for BTT in appropriate patients.

Heartmate II

The HeartMate II (Thoratec Corp., Pleasanton, CA) was born out of research conducted at the University of Pittsburgh in collaboration with Nimbus and has been one of the most studied devices to date. It received FDA approval for DT in 2010. It has a 40-mm diameter, 70-mm length, 176-g weight.⁵⁶ The pump mechanism is similar to that of the DeBakey VAD and Jarvik 2000. It is capable of achieving flow rates of 10 L/min and can generate 120 Torr of aortic pressure, operating at 6000 to 13,000 rpm (Fig. 51-13). The pump receives its inflow via the left ventricular apex and delivers outflow to the ascending aorta; it is placed in a small preperitoneal pocket. The current system uses a percutaneous driveline to supply power and control from an external console. Patients are maintained on warfarin anticoagulation and aspirin therapy. The first human implant was performed in 2000, and since then, the device has been evaluated by many groups.⁵⁷ The ease of implantation, small size, and nonpulsatile flow has made this a popular device, and results from both single institution experiences and multicenter clinical trials have established the HeartMate II as an effective device for MCS, which has revolutionized modern LVAD therapy.

Miller et al. reported the results of their prospective, multicenter clinical trial with the HeartMate II among 133 patients who received device implantation as BTT.⁵⁸ One hundred patients went on to transplantation (56 patients), continued support (43 patients), or device explantation due to myocardial recovery (1 patient). Among these patients, median device support was 126 days with an 89-percent 1-month, 75-percent 6-month, and 68-percent 1-year survival. Following successful transplantation 94 percent of patients were alive at 1 month, and 80 percent at 1 year. Heart failure was improved on device therapy as measured by both quantitative and qualitative measures. Six-minute-walk tests increased from a mean of 42 m at baseline to a mean of 292 m after 3 months of device therapy. Both the Minnesota Living with Heart Failure and Kansas City Cardiomyopathy questionnaires were statistically significantly improved at this time point as well. The most frequent complication was bleeding with 53 percent requiring transfusions greater than two units in the postoperative period and 31 percent who required re-exploration for control of hemorrhage. Rates of both for ischemic (6 percent) and hemorrhagic (2 percent) stroke were low. Rates of infection were also low with device-related infection of only 14 percent, and nondevice related infection of 28 percent. Leading causes of death were stroke (6 percent), sepsis (4 percent), and multisystem organ failure (3 percent). Rates of right heart failure were 17 percent and pump thrombosis 2 percent. Two patients who required device replacement were alive at 216 and 367 days after replacement.

In a follow-up to this trial, Pagani et al. reported the HeartMate II Investigators’ extended experience among 281 patients who underwent implantation of the HeartMate II as a BTT.⁵⁹ Again, there were improvements in heart failure following 6 months of therapy; 6-min-walk test was completed in 83 percent compared with 0 percent at baseline and NYHA heart failure improved from 0 percent class I or II at baseline to 83 percent. Overall survival at 18 months was 72 percent, with 56 percent of patients undergoing transplantation, 21 percent continuing on device support, and 3 percent who experienced myocardial recovery. Again, the most common adverse event was bleeding in 53 percent, and mediastinal re-exploration in 26 percent. Infection was the next most common with nondevice infection in 30 percent, device infection in 14 percent, and sepsis in 17 percent. Stroke rates were low at 8 percent (5 percent ischemic, 3 percent hemorrhagic). Right heart failure was 17 percent and pump thrombosis was only 1 percent.

Another trial reported by the HeartMate II investigators was for use of this device as DT among patients with advanced heart failure.⁶⁰ This trial compared the nonpulsatile HeartMate II and the older-generation, pulsatile HeartMate XVE. Patients with advanced heart failure refractory to maximal medical support and who were ineligible for cardiac transplantation were included. Patients with significant pulmonary, hepatic, or renal dysfunction were excluded. The primary endpoint was survival at 2 years without stroke or reoperation for device exchange. Secondary endpoints included actuarial survival, adverse events, functional status, and overall quality of life. The HeartMate II demonstrated a clear improvement in outcome compared with the HeartMate XVE with regard to the primary endpoint (46 vs 11 percent at 2 years). In addition, events that prevented patients from reaching the primary endpoint such as reoperation to repair or replace the pump (10 vs 36 percent) or any reason overall (54 percent vs 89 percent) were statistically different among the HeartMate II and HeartMate XVE groups, respectively. Actuarial survival rates at 1 and 2 years were significantly better with the HeartMate II than the HeartMate XVE as well; 68 percent vs 58 percent and 55 vs 24 percent, respectively, among the groups.

Both groups experienced improvements in NYHA class heart failure, increased performance on the 6-min-walk test, and improved quality of life following device implantation. At baseline, all patients had severe heart failure, including 75 percent who had NYHA class IV heart failure in the HeartMate II group and 78 percent in the HeartMate XVE group. For the HeartMate II group, after 12 months of device support 76 percent of patients were improved to NYHA class I or II, and by 24 months 80 percent of patients were improved to NYHA class I or II heart failure in 42 percent and 38 percent, respectively. In the HeartMate XVE group after 12 months, 61 percent were improved to NYHA class I or II, and by 24 months the remaining surviving patient had NYHA class I heart failure. Six-minute-walk times were improved from a mean of 182 to 372 m for the HeartMate II group and from 172 to 306 m for the HeartMate XVE group. Results from both the Minnesota Living with Heart Failure and Kansas City Cardiomyopathy questionnaires were also significantly improved from baseline, the former being statistically better with the HeartMate II and the latter approaching significance between the groups.

Rates of adverse events with the HeartMate II were also significantly decreased compared with the HeartMate XVE, including both device and nondevice infection, right heart failure, arrhythmias, and respiratory and renal failure. The leading causes of death were hemorrhagic stroke (9 vs 10 percent) and right heart failure (5 vs 8 percent) in both groups, HeartMate II and XVE, respectively. Bleeding requiring transfusions occurred in 81 and 76 percent and that requiring re-exploration in 30 and 15 percent among the HeartMate II and HeartMate XVE, respectively. Pump thrombosis was more common in the HeartMate II group at 4 percent, compared with no patients in the HeartMate XVE group.

Frazier et al. reported the Texas Heart Institute experience with the HeartMate II in 43 patients, 23 for BTT and 17 for DT.⁶¹ In this initial experience, the average support time was 258 days and 35 survived to hospital discharge including 3 patients who underwent transplantation and 4 patients who had myocardial recovery and pump explantation. Ongoing support was reported in 27 patients with the longest greater than 700 days. Cardiac index was significantly improved and pulmonary capillary wedge pressures were reduced in all patients, and inotropes were removed by 1 week following implantation. All patients improved to NYHA class I status following implantation. There were 9 deaths; 5 died from multifactorial causes (multisystem organ dysfunction, right heart failure, and hemorrhage) and two patients suffered strokes.

John et al. reported an experience in 47 patients who received the HeartMate II for BTT (32 patients), DT (7 patients), or exchange for a prior device (8 patients) demonstrating excellent results: 96-percent 30-day and 87-percent 6-month survival among patients receiving a HeartMate II and mean support time of 193 days with a 3-percent incidence of device malfunction.⁶² Among those patients who received a HeartMate II for BTT, 60 percent were transplanted and all survived to discharge. Renal and hepatic function were both significantly improved after 3 months of device therapy. The most common complications were bleeding in 32 percent, half of whom required re-exploration, followed by infection in 13 percent. Right ventricular failure and pump thrombosis were uncommon, each occurring in 6 percent of patients. Stroke was equally uncommon, occurring in only 6 percent of patients.

Lahpor et al. reported the European experience among three centers and 571 patients implanted as BTT (73 percent), DT (21 percent), and BTR (6 percent).⁶³ Mean support was 236 days with support greater than 6 months, 1 year, and 2 years of 61, 29, and 3 percent, respectively. Survival at 6 months was 74 percent and at 1 year was 72 percent. In this group, 23 percent proceeded to transplantation, 42 percent have continued support, and 4 percent underwent device explantation following myocardial recovery. The most frequent cause of death was multifactorial in nature (sepsis, right heart failure, followed by stroke). Bleeding occurred in 51 percent and was not a source of mortality. There were only three device thromboses, and right heart failure occurred in 20 percent.

Loforte et al. reported the Italian experience with the HeartMate II among their 18 patients who underwent implantation for BTT.⁶⁴ In this group, the median support time was 217 days with the longest supported patient having continued support 665 days after implantation. Fifty percent of patients proceeded to transplantation with 67 percent alive after transplantation at last following. There were no device failures, and the most common complications were bleeding in 33 percent, right ventricular failure in 28 percent, infection in 11 percent, and thrombosis in 6 percent. There were no reported device failures.

This device has also been reported as use for BTT in the pediatric population, with two adolescents (12 and 14 years of age) being supported for 85 and 128 days before successful implantation.⁶⁵

Third-Generation, Centrifugal Continuous Flow Pumps

The potential hemolytic consequences of having rotor bearings in contact with blood in the second-generation devices led to the development of the most recent, or third-generation devices. These have been designed without any components in direct contact with the bloodstream, using hydrodynamic or magnetic levitation to suspend an internal rotor that provides centrifugal, continuous flow. Table 51-5 summarizes the clinical data available for the second-generation devices.

Ventrassist LVAS

The VentrAssist (VentraCor Limited, Chatswood NSW, Australia) is a small implantable centrifugal pump designed for DT and BTT. It uses a hydrodynamic rotor suspension, which avoids areas of slow and turbulent blood flow. The absence of bearings reduces rotor wear, enhancing the device’s durability. It uses large rare-earth magnets with small gaps and a harmonic drive system. Weighing 298 g, it is 6 cm in diameter and 6.5 cm in length. The externally worn battery and controller provide power and control; the battery holds sufficient charge for up to 8 h. The small size of the device will permit its use in smaller adults as well as in children. The short inflow cannula connects to the left ventricular apex and the outflow cannula is anastomosed to the ascending aorta. It has a low potential for thrombosis due to a reduction of stasis and therefore requires minimal anticoagulation. This device has been studied in several multinational clinical trials.

Several smaller clinical trials evaluated the feasibility of using the VentrAssist for both BTT and DT (Pilot Trial, 9 patients; Pivotal Trial, 44 patients; US Feasability Study for BTT, 7 patients) from 2003 to 2005.⁶⁶ Results from these smaller trials were encouraging with no device failures over a cumulative device time of 33 years and a median support time of 180 days. Among the 33 patients who underwent device implantation for BTT, 82 percent were alive at the 5-month trial endpoint with 40 percent having undergone heart transplantation and another 42 percent alive and awaiting transplantation.⁶⁶ Results for the 16 patients who received a device as DT were equally good; 1- and 2-year survivals were 80 and 60 percent.⁶⁷ Recently the international experience with the VentrAssist device implanted in over 400 patients has been published.⁶⁸ Among 412 patients who underwent device implantation, 81 percent reached the endpoint of transplantation (36 percent), continued support (44 percent) or myocardial recovery (1 percent) and the longest period of support was in a patient who was alive with greater than 4 years of continuous support.

The future of this device, however, is uncertain. In 2009, following the deaths of three patients due to damaged electrical leads, the company temporarily stopped all new implantations. Following this event, the company manufacturing the device declared bankruptcy, and was liquidated as clinical trials were underway.

Duraheart LVAS

The DuraHeart LVAS (Terumo Heart, Inc., Ann Arbor, MI) is an active magnetically levitated centrifugal pump. The impeller is levitated by three electromagnets, and rotation is achieved by magnetic coupling between the impeller and the motor. The pump weighs 540 g, displaces 196 mL, and can generate flows up to 8 L/min (Fig. 51-14) at 120 mmHg head pressure with no residual left ventricular function.⁶⁹ The heparin-bonded inflow and outflow cannulas and pump blood chamber comprise the blood contacting surfaces. Warfarin anticoagulation is recommended. The device is commercially approved in Europe and Japan. Clinical trial enrollment was closed in December 2011 for a Pivotal BTT FDA investigation.

The initial clinical experience with the DuraHeart was a European, multicenter trial of 33 patients who underwent device placement for BTT.⁷⁰ This group had no device mechanical failures over a cumulative support duration of 28.7 patient-years.⁷¹ Forty-two percent of patients went on to receive a heart transplant at a mean of 185 days. For those that remained on support awaiting transplant, a 2-year survival was 57%. An updated experience reporting on 82 European patients who underwent implantation of the DuraHeart showed no pump replacements after 6 months over a cumulative support time of 78 years and a median duration of support of 261 days.⁷¹ Survival at 6 months and 1 year were 85 and 79%, respectively. The longest duration of support, at the time of the publication, was more than 4.1 years.

Heartware HVAD

The Heartware HVAD (Heartware International, Framingham, MA) is a continuous flow, centrifugal pump that is small enough (4 × 2 cm, 50 mL displacement) to be implanted in the pericardial space yet is capable of generating flows up to 10 L/min (Fig. 51-15). The only moving part is the impeller that is suspended directly in the bloodstream within the pump itself using magnetic and hydrodynamic thrust bearings, thus eliminating any physical contact between the impeller and the pump. The inflow cannula is integrated into the device itself that is placed directly into the apex of the heart, and the distal end is anastomosed to the ascending aorta. Implantation can be achieved through median sternotomy or left thoracotomy with or without the use of CPB. Warfarin anticoagulation is required. This device has been studied for use as both BTT and DT in the United States and internationally.