General description of procedure, equipment, technique

Traditional management of heart failure (HF) includes medical therapy, implantable cardiac defibrillators (ICDs), and biventricular pacemakers. Despite such advances, HF continues to carry high morbidity and mortality. Five-year mortality for all patients with HF is estimated to be 50%.

For those who have progressed to end-stage HF, the 1-year mortality is as high as 50%. Cardiac transplantation is an option for end-stage HF, but remains available only to a select few. Less than 2,500 transplants are performed in the United States annually, with organ availability being the limiting factor.

Mechanical circulatory support devices including ventricular assist devices (VADs) and total artificial hearts have emerged as an option for the management of end-stage HF. With improvements in technology, approval by the US Food and Drug Administration, and expanding indications, VAD implantation is growing. The Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) has reported the implantation of greater than 4,500 VADs between 2006 and 2011.

Basic components and function of a ventricular assist device

The purpose of VADs is to provide adequate cardiac output to perfuse target end-organs, while unloading the heart. There are four basic components to any VAD:

inflow cannula (implanted in the ventricle)
pumping chamber
outflow cannula (implanted into the pulmonary artery or aorta)
drive line (extracorporeal control and power lines that extend from the pumping chamber to outside the body)

VADs are most commonly implanted in the left ventricle, but can be implanted in the right ventricle or even both ventricles. The result of the VAD is a circuit that receives blood from the heart and pumps it into an arterial circulation. The VAD circuit is in parallel to the native cardiac circuit, but the ventricle itself generally functions as a passive conduit due to poor contractility, decompression of the ventricle, and at times, surgical closure of the aortic/pulmonic valve.

Types of ventricular assist devices

Generally speaking, VADs can be categorized by the type of pumping chamber used to provide cardiac output. There are two major types of pump—pulsatile displacement pumps and continuous flow pumps. Pulsatile displacement pumps use pneumatically or electromagnetically driven pumps that mimic the native ventricle’s cyclic delivery of output.

The cycle includes a loading phase of the pump which mimics diastole and a pulsatile delivery of a stroke volume which mimics systole. The cycle of the VAD is not synchronous with the native cardiac cycle, but still results in a pulsatile flow and a resultant pulse pressure.

The rate of the cycle can be adjusted manually, or based on the activity levels and volume status of the patient. Pulsatile displacement VADs are considered first generation technology and are currently infrequently used due to large size, the requirement for extensive dissection for intra-abdominal pumping chamber implant, and the clinical superiority of continuous flow VADs.

Continuous flows VADs utilize constantly moving rotors to provide continuous ventricular unloading and cardiac output. They lack cyclic systole and diastole, which results in nonpulsatile flow and absence of a pulse pressure. Cardiac output can be adjusted by changing the speed of the rotor.

There are two types of rotors that are used in continuous flow VADs. The first is the axial-flow VAD that utilizes an impeller to propel blood along a single axis. Axial-flow pumps are implanted intra-abdominally, but are much smaller and easier to implant than the pulsatile flow VADs. Clinical trials have shown superiority of axial-flow HeartMate II VAD (Thoratec, Pleasanton, CA) when compared with pulsatile flow VADs. Axial flow VADs are considered second-generation technology.

Another type of continuous rotor is used in centrifugal-flow VADs. They propel blood in a circumferential manner. Centrifugal-flow VADs are considered to be third-generation technology and are even smaller than the axial-flow VADs. This allows for implantation of the device in the pericardial space, thus avoiding the need for an intra-abdominal dissection or implant. Initial experiences with the Heartware Ventricular Assist System (HeartWare Inc, Framingham, Massachusetts) centrifugal-flow have been published, and further clinical trials of this device are ongoing.

Continuous flow VADs are the most commonly implanted type of VAD at present time. More than 98% of VADs implanted in 2010 were continuous flow VADs. The remainder of this review will focus on the management of continuous flow left ventricular assist devices (LVADs).

Many pulsatile and continuous flow VADs are approved by the FDA as a bridge to transplantation. Currently, only the Heartmate XVE and the Heartmate II (Thoratec, Pleasanton, CA) have been approved for use as destination therapy. However, there continue to be ongoing studies for the evaluation of other VADs both for bridge to transplantation and as destination therapy.

Indications and patient selection

Indications

VAD implantation is indicated in patients with end-stage, symptomatic heart failure (American College of Cardiology/American Heart Association [ACC/AHA] Stage D) despite optimal medical therapy. Indications for VAD implant can be divided into three major clinical scenarios:

Bridge to transplantation (BTT): used in patients who are or will likely be listed for cardiac transplantation, but are awaiting unfavorable wait times or organ availability.

Destination therapy (DT): used in patients who are not transplantation candidates or who are uninterested in transplantation. For patients who get a VAD implanted for DT, the VAD is intended as the final surgical therapy for their end-stage HF.

Bridge to recovery: used in patients with potentially reversible cardiac dysfunction.

The indication for VAD can be altered based on changes in a patient’s clinical status. For example, a patient who underwent VAD implant for DT due to seemingly irreversible comorbidities could have such a marked clinical improvement, that transplantation becomes an option. This patient could be listed for transplantation and ultimately be considered as a BTT.

Patient selection and timing of ventricular assist device implantation

Appropriate patient selection for VAD therapy is critical to ensure both appropriate use of the technology and successful patient outcomes. Patient selection includes assessment of several factors:

Severity of HF

Evaluation of comorbidities

Evaluation for transplantation candidacy

Evaluation of psychosocial factors

Severity of HF should be assessed on maximum tolerated medical therapy. There are several methods for assessing severity of HF, including hemodynamic assessment via right heart catheterization, maximal oxygen consumption during cardiopulmonary exercise testing, New York Heart Association (NYHA) functional classification, and use of multivariate clinical models such as the Heart Failure Survival Score and Seattle Heart Failure Model.

Candidates for VAD implantation must have advanced symptoms (NYHA Class IIIB or IV) despite medical therapy. All methods of HF assessment have limitations, but multivariate clinical models may be superior to cardiopulmonary exercise testing to assess the prognosis in patients with advanced HF.

Beyond these models, assessment of target-organ dysfunction is important in HF prognosis. INTERMACS has further categorized profiles of patients with end stage HF (Table I). This ranges from scores of 1 (“crash and burn,” critical cardiogenic shock, most severe) to 7 (Advanced NYHA III, least severe). Patients who have profile 3 (“Stable, but inotrope dependent”) or worse HF are appropriate candidates for VADs.

Table I.

Profile	Description
1	Critical, cardiogenic shock
2	Progressive decline
3	Stable, but inotrope-dependent
4	Recurrent, advanced heart failure
5	Exertion intolerant
6	Exertion limited
7	Advanced NYHA functional classification

NYHA, New York Heart Association.

However, registry data shows that patients who are INTERMACS profile 1 have a worse prognosis, suggesting that the presence of target organ dysfunction is a predictor of poor outcomes. Thus, the challenge is to identify patients who have end-stage HF, but who do not have advanced or irreversible target organ dysfunction (renal dysfunction, liver dysfunction, etc.). (Table I)

The appropriate timing of VAD implantation is also challenging. If a patient’s target organ dysfunction can be improved with inotropic or temporary mechanical support (such as intra-aortic balloon pumps), their candidacy for a VAD implant may also improve. However, if delaying VAD implantation in a stable patient results in progression to a state of cardiogenic shock, the prognosis of the VAD implantation will worsen.

A thorough assessment of comorbid conditions must be done during the evaluation of a patient for VAD candidacy. Guidelines from the International Society for Heart and Lung Transplantation recommend consideration of several noncardiac factors including age, body size, renal function, pulmonary function, hepatic function, coagulation disorders, infectious disease, immunoinflammatory disease, neurologic disorders, malignancy, and nutritional status.

Additionally, a full cardiac assessment including valvular function, arrhythmia, and right ventricular function are recommended. Right ventricular dysfunction is a strong risk factor for poor outcomes after LVAD implantation (Table II). There are several risk scores for right ventricular failure that can used to help decide whether an LVAD or a biventricular VAD should be implanted.

Table II.

Method	Parameters
Echocardiography	Tricuspid annular plane excursion, RV dilation, tricuspid insufficiency
Cardiac MRI	3D volumetric assessment at end-systole and end-diastole for RV size and ejection fraction
LV Echo for RV Failure Score (Columbia University)	LV end-diastolic diameter (LVEDD); LV ejection fraction; ratio of left atrial dimension/LVEDD
RV Failure Risk Score(University of Michigan)	Use of vasopressors; bilirubin ≥2.0mg/dL; aspartate aminotransferase ≥80IU/L; creatinine ≥2.3mg/dL
Risk Score for Biventricular Mechanical Support(University of Pennsylvania)	Cardiac index ≤2.2 L/min/m²; RV stroke work index ≤0.25mm Hg L/m²; severe pre-VAD RV dysfunction; creatinine ≥1.9mg/dL; previous cardiac surgery; SBP ≤96mm Hg

LV, left ventricle; MRI, magnetic resonance imaging; RV, right ventricle; SBP, systolic blood pressure; VAD, ventricular assist device.

In patients at moderate risk for right heart failure after VAD implantation, implantation of an LVAD alone may be feasible. In those patients with right ventricular failure after LVAD implantation running the VAD at lower speeds, slow titration off inotropic therapy (typically with the guidance from a pulmonary artery catheter), and use of pulmonary vasodilator therapy may be useful. In patients at high risk for right ventricular failure, implantation of a biventricular VAD, listing for heart transplantation, or listing for heart-lung transplantation should be considered. Anatomical considerations should be reviewed for the feasibility of VAD implantation, particularly in patients with congenital heart disease or with prior cardiac surgeries (Table II).

Due to the considerable burdens of postdischarge VAD management, a full psychosocial assessment is an important component of VAD candidacy evaluation. This should include an assessment of demographics, relationship status, family/support systems, educational level, financial resources, coping mechanisms, history of substance abuse, history of nonadherence, and goals of care.

This should be done by a social worker with familiarity with the challenges and requirements of inpatient and outpatient VAD management. While there is little published data on these issues, they undoubtedly have an impact on a patient’s ability to thrive after VAD implantation.

Finally, it is important to consider the indications for VAD implantation. If a patient is being considered for a VAD as a BTT, then a comprehensive evaluation for transplant candidacy should be done as well.

Contraindications

Assessment for contraindication to VAD implantation is critical, considering the cost, burden, and risk of VAD implantation. The absolute contraindications for a VAD implant can be divided into cardiac and noncardiac conditions. Cardiac conditions include:

Reversible cause of heart failure

High risk anatomy for surgical implantation such as complex congenital heart disease or severe abnormality of the aorta

Severe mitral stenosis

Severe, fixed pulmonary hypertension

Pulmonary hypertension is a common finding in patients with end-stage HF and must be assessed carefully when evaluating a patient for LVAD support. Patients may have pulmonary arterial hypertension (PAH), pulmonary venous hypertension secondary to high left-sided filling pressures, or a combination of both.

Pulmonary venous hypertension secondary to high left-sided filling pressures is not a contraindication to VAD implantation, and actually, VAD therapy will likely improve the pulmonary artery pressures by decreasing pulmonary venous pressures. However, elevated transpulmonary gradient (TPG) and pulmonary vascular resistance (PVR) on right heart catheterization suggest the presence of PAH.

Traditionally, PAH could be further classified as fixed or reversible based on the presence of reversibility with pulmonary vasodilator testing during right heart catheterization. There has been concern for poor outcomes in patients with severe, fixed PAH undergoing a heart transplant or isolated LVAD therapy.

However, multiple studies have reported improvement in TPG and PVR in patients with fixed PAH (mean PVR 3.5–4.8 Wood units) after prolonged LVAD support. Thus, paradoxically, there is a role for LVAD support in some patients with “fixed” PAH. However, in patients with extremely elevated PVR (>6 Wood units), or fixed PAH associated with right ventricular dysfunction, alternative therapies such as a biventricular assist device or heart–lung transplantation should be considered. Currently, the approach to PAH in VAD candidates remains center specific.

Noncardiac contraindications include severe, irreversible manifestations of systemic comorbidities and conditions that limit patient ability for self-care after VAD. They include:

Severe renal dysfunction (including chronic dialysis dependence)

Chronic liver failure

Severe bleeding diathesis, including inability to tolerate anticoagulation

Heparin-induced thrombocytopenia

Severe lung disease (with forced expiratory volume in 1 second [FEV1] <1)

Active or uncontrolled infection

Acute stroke, including intracranial hemorrhage

Neurologic disorder that limits the patient’s ability for self-care (e.g., advanced dementia, neuromuscular disease)

Coexisting terminal disease (e.g., metastatic cancer)

Additionally, risk scores have been developed and can be used to determine the risk for complications (including mortality) after VAD implantation. These scores can be used to identify patients who have a very high level of risk that is prohibitive to VAD implantation or who require further optimization prior to VAD implantation.

An example is the Lietz–Miller score which was developed using preoperative clinical characteristics in a DT population from the Randomized Evaluation of Mechanical Assistance for the Treatment of Congestive Heart Failure (REMATCH) trial (using pulsatile flow VADs). A very high score (>19) predicts a risk of in-hospital mortality at 90 days of 70%, while low scores (<9) are associated with less than 1% risk (Table III). Risk scores incorporating the current generation of continuous flow VADs are in development (Table III).

Table III.

Clinical characteristic	Weighted risk score
Platelet count ≤148×10³/μL	7
Serum albumin ≤3.3g/dL	5
International normalized ratio >1.1	4
Vasodilator therapy	4
Mean pulmonary artery pressures ≤25mm Hg	3
Aspartate aminotransferase >45U/mL	2
Hematocrit ≤34%	2
Blood urea nitrogen >51U/dL	2
No intravenous inotropes	2

Advanced age, obesity, poor nutritional status, limited financial resources, and poor social support may also be considered contraindications for VAD implantation. However, the thresholds for each are center-specific. Finally, it is important to recognize that the contraindications for VAD implantation may be different if the VAD is implanted for BTT versus DT. For example, end-stage renal disease on dialysis is typically a contraindication for DT VAD implantation. However, if a patient were listed for a combined heart–kidney transplant, then dialysis would not contraindicate VAD implantation as a BTT.

Details of how the procedure is performed

Presurgical management and education

Presurgical management prior to VAD implantation includes optimization of the preimplantation medical condition and education of patients and their providers. Because of the poor outcomes associated with critical cardiogenic shock, patients should be hemodynamic optimized.

Pulmonary artery catheters, inotropes, and percutaneous mechanical support (with intra-aortic balloon pump or percutaneous VAD) can be used to improve cardiac output and restore compensation of target organs. Optimization of volume status with diuretics and ultrafiltration is also important to minimize the risk of right ventricular dysfunction due to volume overload in the perioperative setting.

Medications to decrease pulmonary vascular resistance (such as sildenafil, inhaled nitric oxide, prostaglandins, and milrinone) may also be useful to decrease right ventricular afterload perioperatively. If the risk for right ventricular failure is still high despite medical therapy, implantation of a right ventricular assist device (RVAD) should be considered.

Potentially reversible comorbidities should be addressed prior to VAD implantation. Some comorbidities (including renal and hepatic dysfunction) can be managed by improving cardiac output. Active infections should be treated to cure prior to VAD implantation.

Prophylactic antibiotics are also given in the perioperative setting. Bleeding both early and late after VAD implantation is common. Abnormalities in coagulation (medication-induced and pathologic) should be reversed with vitamin K or fresh frozen plasma prior to implantation.

Antiplatelet agents should also be held in the perioperative setting. Sources of bleeding (particular gastrointestinal sources) should also be investigated and treated prior to VAD implantation. Finally, the patient’s nutritional status should be assessed and optimized with dietary supplementation. A prealbumin level greater than 15mg/dL has been suggested prior to VAD implantation.

Patients must be thoroughly educated on the functions and potential complications of VADs in order to be self-reliant after device implant. The preimplantation period is an optimal time for education as patients are free from postsurgical pain, analgesics, and sedatives.

Many programs provide individual training and education by a dedicated, experienced VAD nurse. Preimplantation material should include information on basic physiology, components of the VAD, and complications that can arise after VAD implantation. Education should be extended to patient caregivers, family members, and community personnel including fire departments, electric companies, and first responders (such as community emergency room personnel and emergency medical technicians).

Postsurgical management, rehabilitation, and education

The surgical considerations and challenges of VAD implantation have been described. The immediate postimplantation management includes monitoring and treatment of postsurgical complications, rehabilitation, and education.

Right ventricular failure is a common complication after LVAD implantation. The function of the right ventricle is critical as it provides the preload required for the LVAD circuit. Pharmacologic support with inotropes and pulmonary vasodilators are commonly used to improve right ventricular function.

These agents must be weaned delicately, usually with the assistance of a pulmonary artery catheter. Insufficient right ventricular output can manifest with target-organ damage (such as renal dysfunction). However, it can also create dysfunction of the LVAD itself due to insufficient filling of the left ventricle.

This can result in the LVAD sucking onto a small, underfilled, and decompressed left ventricle causing “suction events” and ventricular arrhythmias. The rotor speed of continuous flow VADs may need to be reduced in this scenario to allow for less mechanical left ventricular decompression. While LVAD recipients are often volume overloaded, a daily assessment of volume status (with or without the assistance of a pulmonary artery catheter) must be done to avoid volume depletion (which may also cause insufficient left ventricle preload).

After VAD implantation, patients are started on anticoagulation (typically warfarin) and antiplatelet therapy (aspirin) to prevent thrombotic complications of VADs. However, postoperative bleeding is another common postoperative complication. The evolution of the smaller continuous flow VADs has reduced the risk of bleeding complications from VAD implantation, but clinical trials report bleeding intrathoracic or intraperitoneal cavities requiring reoperation in up to 30% of patients.

Platelet aggregation becomes impaired due to an acquired von Willebrand syndrome due to shear. Complete blood counts and drain outputs must be monitored closely after VAD implant. Postoperative bleeding is managed with fluid resuscitation, blood products, and surgery when necessary.

Infection is a relatively uncommon, but serious complication in the immediate postimplantation period. The incision site, VAD drivelines, invasive lines, and endotracheal tubes may serve as a nidus for infection. Antibiotics with gram-positive coverage (including Staphylococcus aureus) are typically given prophylactically in the perioperative period.

In addition, strict hand washing protocols and meticulous monitoring for opportunities to remove invasive lines may help prevent infectious complications of VADs. Treatment of VAD infections usually involves prolonged courses of antibiotics. Exchange of hardware is less frequently done due to the inherent risks of reoperation.

Postoperative management also includes optimization of nutritional status and initiation of physical rehabilitation. Cachexia is a common metabolic state in end-stage heart failure and is associated with poor outcomes. Treatment of nutritional deficiencies may reduce the risk of postoperative infections and may improve wound healing after VAD implantation. Laboratory tests including albumin, prealbumin, cholesterol and C-reactive protein may be helpful in this assessment. Enteral nutrition is generally recommended, but an individual treatment plan should be formulated after VAD implantation with the assistance of a nutritionist.

The implantation of a VAD can result in improvement in functional status (based on NYHA class) and performance during cardiopulmonary exercise. However, end-stage HF patients are often chronically debilitated and deconditioned. Thus, physical rehabilitation is recommended to optimize functional recovery.

Outpatient and inpatient rehabilitation centers may be needed to facilitate a level of focused rehabilitation greater than can be achieved in a hospital setting. Many VAD programs form partnerships and provide education for regional rehabilitation centers to ensure appropriate postdischarge care for their patients. Without proper rehabilitation, patients may continue to be chronically debilitated, but have the added responsibility of management of their VAD.

Finally, patients and their caregivers must continue to receive dedicated education after VAD implantation. This education should be focused on transition from the hospital to home. Patients must learn how to manage their VAD including changing batteries, cleaning dressings, and troubleshooting alarms.

They must receive education on the symptoms of potential complications and must have plans in store for managing emergencies. Finally, they must learn how to do the everyday activities of life such as bathing, dressing, and cooking without compromising the VAD. Not all nuances of living with a VAD can be anticipated or taught prior to hospital discharge, so patients should have contact information for VAD coordinators and their VAD manufacturers’ help lines for when new questions arise.

Outcomes (applies only to therapeutic procedures)

Multiple well-designed clinical trials have been performed to evaluate the performance of VADs. The REMATCH trial was a pivotal, randomized trial comparing the first-generation pulsatile-flow HeartMate XVE with optimal medical therapy in 129 patients who were not transplant candidates (DT population). The results showed a 48% reduction in all-cause mortality and greater improvement in quality of life with VAD support. However, survival rates at 1 and 2 years were poor at 52% and 23%, respectively.

Further studies have shown the superiority of subsequent generation of VADs. The HeartMate II investigators compared an axial continuous-flow VAD (HeartMate II) to the HeartMate XVE in 200 DT subjects. Their primary end-point was a composite of survival free from disabling stroke and reoperation to repair/replace the device at 2 years. The superiority of continuous-flow VADs was shown as 46% of subjects in the continuous-flow group met the primary end-point compared with 11% in the pulsatile-flow VAD group. Additionally, survival at 2 years was 58% in the continuous-flow group compared with 24% in the pulsatile-flow device group.

The HeartMate II has also been studied in patients awaiting cardiac transplantation. At 6 months, survival with VAD implantation, transplantation, or myocardial recovery was 75%. Of those who did not receive transplantation, the 1-year survival rate with continuous VAD support was 68%. Quality of life and functional status was also improved after VAD implantation.

The potential benefit of the latest generation centrifugal continuous-flow VADs have also been reported. The experience with the HeartWare HVAS in 50 patients (both BTT and DT) has been promising, with 1- and 2-year survival rates of 84% and 79%, respectively. However, randomized trials are currently in progress to evaluate further the benefit of centrifugal continuous flow VADs.

The INTERMACS group has collected real-world experience with VADs, and has reported outcomes that mimic trial-based data. The Fourth Annual Report from INTERMACS has reported on more than 3,400 LVAD implants, and survival to transplantation, recovery, or continued support is 80% at 1 year and 69% at 2 years. Additionally, survival has significantly improved with use of a continuous flow VADs compared with pulsatile-flow VADs. INTERMACS continues to collect and publish data on an annual basis.

Complications and their management

Understanding the major concepts of long-term management of VAD patients is important to optimize outcomes and to avoid complications of VAD therapy. The principles of long-term care include the assessment and management of VAD function, hemodynamics, anticoagulation, arrhythmias, infections, and psychosocial factors.

Ventricular assist device performance

Assessment of continuous flow VAD performance begins with evaluation of the four basic distinct VAD parameters:

Rotor speed—an adjustable parameter (in rotations per minute) that regulates the forward flow of the VAD

Power—directly measured (in watts) by the VAD

Flow—unmeasured; rather, calculated parameter of output based on algorithm incorporated rotor speed and power

Pulsatility—reflects the dynamic nature of the pressure gradient between the left ventricle and aorta during the cardiac cycle

The speed of the rotor regulates the forward flow of the VAD and is the sole adjustable parameter of the VAD. Inadequate rotor speed can result in low, absent, or even retrograde VAD flow. Excessive rotor speed can lead to high negative pressure in the left ventricle, resulting in a small, overly decompressed left ventricle with a leftward septal shift. In extreme cases, the left ventricle wall can be sucked into the inlet cannula, resulting in ventricular arrhythmias and low VAD flow (commonly referred to as “suction events”).

Additionally, the septal shift may distort the geometry of the right ventricle as well and adversely affect performance by both chamber dilation and annular dilation (causing increase in tricuspid insufficiency). Real-time invasive hemodynamic data and echocardiographic data can be used to select an optimum rotor speed that maximizes cardiac output while maintaining an appropriate left ventricle size.

The frequency of the aortic valve on echocardiography is also used to determine the optimal rotor speed. Opening of the aortic valve with each cardiac cycle suggests inadequate rotor speed. Intermittent opening of the aortic valve is generally recommended, but not achievable in some patients due to surgical closure of the aortic valve, aortic stenosis, or severely reduced native left ventricle contractility.

Flow is also dependent upon the pressure gradient between the aorta and the left ventricle and the structural integrity of the VAD. Because the VAD circuit must pump against the gradient between the left ventricle and aorta, greater differences between aortic and left ventricle pressures (due to high Doppler pressures or low left ventricle preload) results in reduced flow. A diminished difference between aortic and left ventricle pressures (due to lower Doppler pressures or higher left ventricle preload) will result in an increase in VAD flow.

In the presence of normal VAD function, increases in power correlate with increases in actual flow (cardiac output). However, estimates of flow are not validated at extremely high or low pump powers, so some interrogators will merely report flow as (– – –) or (+++) at extreme power levels. Furthermore, there are situations (such as VAD rotor thrombus) in which a resultant increase in power does not correlate with increase in actual flow (if anything, there will be a decrease in actual flow). VADs cannot detect this and will report a misleading high flow state. Thus, the calculated flow should be taken into clinical context, rather than be interpreted independently.

The degree of pulsatility is the last major parameter of continuous flow VADs. Pulsatility is displayed differently by various VAD manufacturers. It can be displayed visually or quantified as an indexed average between the peak and nadir of flow (pulsatility index).

Pulsatility is directly related to the degree of left ventricle contractility and inversely related to the degree of assistance by the VAD. When VAD flow is high, left ventricle preload and left ventricle contractility (by Frank–Starling mechanism) decrease. This results in a low pulsatility.

Conversely, low relative VAD flow or high left ventricle contractility will result in high pulsatility. Lack of pulsatility suggests excessive VAD speeds or low left ventricle preload (due to right ventricle dysfunction, dehydration, etc.). High pulsatility suggests that VAD speeds can be increased or that left ventricle contractility has improved.

Abnormalities in the components of the VAD itself may present as abnormalities in VAD parameters (Table IV). The inflow cannula, outflow cannula, and the pump itself can be compromised by thrombus, kinking, and pannus ingrowth. Obstruction involving either the inflow or outflow cannula presents with a decrease in flow and power.

Table IV.

Abnormality	Presentation	Comments
Outflow/inflow cannula obstruction	Lower power and flow	Echo and CT imaging helpful for diagnosis
VAD rotor malfunction	High power and flow	The VAD reports flow as high, while true cardiac output is reduced. CT and echo may be helpful
Suction event	Sudden, transient drops in power and flow	Treatment options include increasing LV preload or decreasing VAD rotor speed
Low LV filling/preload	Low flow, power, and pulsatility	May be due to low volume status, RV dysfunction, or excessive VAD rotor speed

CT, computed tomography; LV, left ventricle; RV, right ventricle; VAD, ventricular assist device.

Echocardiography or a computed tomography scan can assist in further differentiation. Thrombus of the rotor itself typically presents with high power and erroneously reported high flow. This can be difficult to diagnose, but is critically important and can necessitate exchange of the VAD pump. Finally, sudden transient drops in power and flow are suspicious for suction of the left ventricle (due to an overly decompressed left ventricle). This can be further investigated using echocardiography and can be treated by decreasing VAD rotor speed (Table IV).

Hypertension

Hypertension is a common comorbidity in patients with heart failure and is common in patients after VAD implantation. Because the loading conditions have an impact on the performance of a VAD, blood pressure must be monitored and hypertension should be treated.

However, traditional methods of blood pressure assessment cannot be used since continuous flow significantly limits pulse pressure (and results in lack of a palpable pulse). A mean arterial pressure, measured by an invasive arterial line, is used in the immediate postoperative period.

Once the arterial line is removed, a Doppler method is used to assess blood pressure. This is done by inflating a manual blood pressure arm cuff and then deflating it until the first Korotkoff sound by Doppler can be recorded. This single value is referred to as the “Doppler pressure” and serves as the assessment of blood pressure.

The goal Doppler pressure for VAD patients is between 70 to 90 mm Hg. Hypertension should be managed with vasodilators, as needed. If a VAD is implanted as a “bridge to recovery”, use of ACE-inhibitors, beta-blockers, and aldosterone antagonists is recommended. Adjusting the VAD rotor speed is not helpful in management of hypertension after VAD implantation.

Thrombosis and anticoagulation

Thrombotic complications of VADs can be catastrophic and may make patients ineligible for cardiac transplantation. Due to the risk of VAD thrombosis causing device dysfunction and thromboembolic events (particularly strokes), all VAD patients must be treated with antithrombotic therapy.

However, this must be balanced against the risk of bleeding due to the use of anticoagulation therapy and acquired von Willebrand syndrome. A combination of aspirin and warfarin (with a targeted international normalized ratio, INR goal of 2.0–3.0) was used in the HeartMate II trial, which compared the pulsatile flow Heartmate XVE with the axial, continuous flow Heartmate II.

However, the rates of bleeding in this trial were almost 10 times the rate of thromboembolic rates. Thus, many centers target an INR goal of 1.5 to 2.5 instead for patients with Heartmate II in the absence of atrial fibrillation, history of stroke, or previous venous thromboembolism. However, each VAD has its own recommended INR goal. Antithrombotic therapy is typically started as soon after VAD implantation as is feasible and safe.

Bleeding is the most common complication after VAD implantation and is the primary cause in about 10% of cases of VAD mortality. Early bleeding is often related to surgical issues, but patients with continuous flow VADs have high rates of gastrointestinal bleeding over time as well.

Bleeding is treated based on the source of the bleed. Due to the low day-to-day risk of thromboembolic events after VAD implantation, anticoagulation can be reversed and has even been safely withheld for up to months at a time in the case of life threatening bleeds.

Bleeding often requires treatment with blood transfusions and operations as well. Unfortunately, the need for repeated blood transfusions may have consequences in patients with VAD for “bridge to transplantation”, as they may face cross-match incompatibility while awaiting a donor organ.

If a source of bleeding can be identified and controlled, patients may resume their previous anticoagulation. When a gastrointestinal bleeding source persists, but cannot be localized, options for treatment include octreotide, clotting factors (such as recombinant factor VIIa, von Willebrand factor), and decreasing VAD pump speed. Adjustment of VAD speed may help decrease shear of von Willebrand factor, a common finding after VAD implantation that results in an acquired von Willebrand deficiency. However, none of these strategies have been well studied. Ultimately patients whose bleeding cannot be controlled may require cardiac transplantation.

Arrhythmias

Ventricular arrhythmias (VAs) are frequent in patients with end-stage HF, necessitating treatment with antiarrhythmic medications or ICD. Despite the ventricular unloading provided by VADs, many patients continue to have VAs. Some data suggests that the frequency of VA increases after VAD implantation (affecting up to 50% of patients). Mechanisms for VAs after VAD implantation include underlying cardiomyopathy, scar-based re-entry from surgical incision into the left ventricle, prolongation of QT intervals due to remodeling, and suction events.

The clinical significance of VA is variable in patients after implantation of a continuous flow VAD. Since cardiac output depends little on phases of systole in diastole, VA is often well tolerated in patients with continuous flow VADs. They may be asymptomatic despite being in ventricular fibrillation. Some patients will have symptoms, including palpitations, lightheadedness, syncope, or ICD shocks. There is limited evidence on management of VA in patients with VADs. One study suggests that lack of beta-blocker use is a risk factor for development of post-VAD VA, but there are no prospective trials that prove a benefit. Antiarrhythmic medications can be used, but efficacy is unclear. An assessment of hemodynamic status, VAD function, and electrolytes should be done and may reveal reversible promoters of VA.

ICD therapy is common in VAD patients and is often initiated prior to VAD implantation. A retrospective study of 478 VAD patients suggests an improvement in survival in VAD patients who have concurrent ICD therapy (compared with those without ICDs), but prospective data is lacking.

However, optimal programming of ICDs after VAD implantation is unknown. ICD shocks (both appropriate and inappropriate) are common after VAD implantation (up to 40% of patients) and are associated with worse prognosis. It is possible that conservative programming of ICDs may reduce shocks and avoid symptoms of VA that are otherwise hemodynamically well tolerated.

Finally, there can be device-device interactions between VADs and ICDs. The risk of interference causing programming errors is rare, but both sensing and pacing thresholds may become impaired after VAD implantation. ICD testing after VAD implantation is recommended and lead replacement should be considered if clinically indicated.

Infections

VADs are also susceptible to infectious complications. Potential sources for infection of VAD include the driveline, the pump pocket, the pump itself, and the conduits. VAD-related infections can disseminate systematically as well, causing bloodstream infections and endocarditis. Thrombotic (or thromboembolic) complications of infections can also occur. Gram-positive organisms including S. aureus, S. epidermis, and Enterococcus are the most frequent organisms in driveline site infections. Fungal infections have also been reported and are associated with a high risk of mortality.

If infection is suspected in a VAD patient, a careful assessment of the location and extent of the infection must be done to guide therapy. Blood cultures should be checked if infection is suspected, and additional imaging via echocardiography, computed tomography scan, or ultrasound may be helpful.

For infections involving the driveline or driveline site, topical and systemic antibiotics are used. Antibiotics should include gram-positive coverage. If a pump pocket infection is suspected, then empiric broad-spectrum antibiotics (with both gram-positive and gram-negative coverage) are recommended.

The pump pocket may need surgical debridement, wash out, or drainage. Infection of the pump (sometimes referred to as pump endocarditis) itself is a serious condition that should initially be treated with broad-spectrum antibiotics (including gram-positive, gram-negative, and fungal coverage). Surgical replacement of the VAD rotor or consideration of heart transplantation may be necessary.

Several measures can be taken to minimize the risk of infection. Aseptic dressing changes must be performed meticulously during hospitalizations, and proper technique must be taught to patients and caregivers prior to discharge. The driveline’s exit site in the skin is particularly prone to infection and must be cleaned daily with antimicrobial soap. Patients and caregivers must monitor the site for early signs of infection and must notify their providers if there is concern for infection. Infectious risk may also decline in the future with development in technology, including smaller VADs and use of chlorhexidine or silver sulfadiazine coating on drive line sites.

Medical therapy for left ventricle recovery

While the majority of VAD implantation and VAD trials are performed for either BTT or DT, there is a select patient population who undergo VAD implantation with the hopes of left ventricle recovery (“bridge to recovery”). In addition, both VAD therapy and transplantation are associated with complications that could be avoided in cases of left ventricle recovery and VAD explantation (even if the original indication for VAD therapy was not “bridge to recovery”).

Approximately 1 to 2% of VADs are implanted with a strategy of “bridge to recovery” based on the most recent INTERMACS report. However, some argue that this is an underutilized strategy. One of the biggest limitations to “bridge to recovery” strategy is the lack of methods to identify patients who are likely to have sustained recovery of ventricular function after a period of VAD support. In case series, recovery of function has been seen in patients who have nonischemic etiologies of their cardiomyopathies.

In particular, patients with peripartum cardiomyopathy and myocarditis may have a higher likelihood of recovery. However, alternative clinical predictors of ventricular recovery with VAD support are currently lacking. While left ventricle recovery is difficult to predict prior to VAD implantation, institution of medical therapy and serial clinical and echocardiographic assessment can help identify patients who may have sustained reversal of HF after VAD implantation.

No randomized trial of medical therapy for VAD patients has been done, but a two-stage pharmacologic approach has been described in VAD patients with nonischemic cardiomyopathy.

Stage one: enhance reverse remodeling and improve neurohormonal impairmentLisinopril—titrate to 40mg dailyCarvedilol—titrate to 50mg twice a day
Spironolactone—titrate to 25mg daily

Losartan—titrate to 100mg daily

Stage two: instituted after maximal, stable left ventricle size regression.Carvedilol is replaced by bisoprololClenbuterol (a beta-2-agonist) started and titrate to max dose of 700μg, three times a day.

Clenbuterol was used due to its benefits myocardial metabolism in experimental models. However, it has not undergone rigorous clinical testing in HF.

After medical therapy has been initiated, echocardiography and hemodynamic assessment can identify patients in whom VAD explantation can be considered. Parameters that suggest feasibility of VAD explantation include:

Left ventricle end diastolic diameter >60 mm

Left ventricle end systolic diameter <50 mm

Left ventricle ejection fraction >45%

Left ventricle end diastolic pressure <12 mm Hg

Cardiac index >2.8 L per minute per body surface area

Maximal oxygen consumption > 16 mL/kg/min.

These studies should be done off VAD support. In patients with potentially reversible cardiomyopathies, use of medical therapy is recommended, and consideration of VAD explantation is feasible.

Psychosocial support

While appropriate use of VADs has been shown to improve HF outcomes and quality of life, living with a VAD can be a source of significant burden for patients, family members, and other caregivers. Patients face the possibility of long recovery times, frequent clinic visits, and rehospitalizations. They must handle their dressing changes, VAD alarms, and VAD-related complications.

There also may be a financial burden from dressing materials, medications, and hospital care. Even when symptoms of heart failure improve, VAD patients may have new or persistent symptoms of pain, depression, and organic mental syndromes. The burdens of VAD therapy also extend to caregivers, who may also be responsible for VAD care, transportation, and complication troubleshooting. Recognizing and addressing the psychosocial needs of VAD patients can potentially optimize quality of life.

Psychosocial assessment and interventions is a multidisciplinary process, and often begins prior to VAD implantation. Early on, it is important to identify a broad network of invested family members, friends, and neighbors to support patients and caregivers.

VAD teams should include social workers and VAD coordinators who can provide psychosocial support and resources. They can even provide insight on candidacy for VAD therapy.

Finally, palliative care teams are helpful in establishing goals of care prior to VAD implantation, and can help manage symptoms and end-of-life issues after VAD implantation. Preimplantation use of palliative care consultation may clarify post-VAD care and ease the burden of complications. After VAD implantation, treatment of comorbidities, such as depression are important, and may improve motivation in patients who have difficulties with nutrition or rehabilitation.

End-of-life

Both BTT and DT VAD patients can reach a terminal condition due to progression of native disease, multiorgan failure, or complications of VAD therapy. Data from the most recent INTERMACS report shows a wide spectrum of causes of death.

Death is not exclusively due to VAD dysfunction or cardiac failure, and thus patients may continue to have adequate cardiac output and hemodynamic parameters despite having a terminal illness or even brain death. In these cases, it is not uncommon for a patient or their caregiver to request withdrawal of VAD support.

Ethical issues may arise in these situations (particularly if a caregiver requests withdrawal of VAD support), so the optimal time to discuss end-of-life issues and advanced directive is prior to VAD implantation. Palliative care consultation can be helpful in addressing end-of-life issues. Ultimately, withdrawal of VAD can be wholly appropriate. If the decision has been made to withdraw VAD support, an organized approach that includes silencing of alarms, deactivation of ICDs, removal of invasive lines, deactivation of VAD, and removal of invasive airways should be done to avoid an unnecessarily disordered and unpleasant end-of-life event.

What’s the evidence?

High impact reviews about ventricular assist device management

Slaugher, MS, Pagani, FD, Rogers, JG. “Clinical management of continuous-flow left ventricular assist devices in advanced heart failure”. J Heart Lung Transplant. vol. 29. 2010. pp. S1-S39. (This article provides a thorough review on management of continuous flow ventricular assist devices in the preoperative and immediate post-operative periods.)

Wilson, SR, Givertz, MM, Stewart, GC, Mudge, GH. “Ventricular assist devices: the challenges of outpatient management”. J Am Coll Cardiol. vol. 54. 2009. pp. 1647-59. (This article describes chronic management of continuous flow ventricular assist devices.)

Recent statistic and outcomes of the INTERMACS registry

Kirklin, JK, Naftel, DC, Kormos, RL. “The fourth INTERMACS annual report: 4,000 implants and counting”. J Heart Lung Transplant. vol. 31. 2012. pp. 117-26. (The INTERMACS registry is a nationwide registry of ventricular assist device implants. The most current outcomes data from this registry is published on an annual basis.)

Patient selection

Deng, MC, Loebe, M, El-Banayosy. “Mechanical circulatory support for advanced heart failure: effect of patient selection on outcome”. Circulation. vol. 103. 2001. pp. 231-37. (This study identifies risk factors for mortality after ventricular assist device implant. This data can be used to risk stratify and select appropriate patients for ventricular assist device implant.)

Wilson, SR, Mudge, GH, Stewart, GC, Givertz, MM. “Evaluation for a ventricular assist device: selecting the appropriate candidate”. Circulation. vol. 119. 2009. pp. 2225-32. (An overview of the process for evaluating candidates for ventricular assist device implant.)

Risk scores for preoperative assessment of risk and right ventricle function

Lietz, K, Long, WL, Kfoury, AG. “Outcomes of left ventricular assist device implantation as destination therapy in the post-REMATCH era: implications for patient selection”. Circulation. vol. 116. 2007. pp. 497-505. (The Lietz-Miller score assesses risk and outcomes on patients undergoing ventricular assist device implant. The data are derived from the REMATCH trial of patients with pulsatile flow ventricular assist devices.)

Matthews, JC, Koelling, TM, Pagini, FD, Aronson, KD. “The right ventricular failure risk score”. J Am Coll Cardiol. vol. 51. 2008. pp. 2163-72. (A risk score for predicting right ventricular failure after ventricular assist device implant using common clinical parameters.)

Fitzpatrick, JR, Frederick, JR, Hsu, VM. “Risk score derived from preoperative data analysis predicts the need for biventricular mechanical circulatory support”. J Heart Lung Transplant. vol. 27. 2008. pp. 1286-92. (A risk score for predicting the need for biventricular mechanical support when considering implant of a ventricular assist device.)

Kato, TS, Farr, M, Schulze, PC. “Usefulness of two-dimensional echocardiographic parameters of the left side of the heart to predict right ventricular failure after left ventricular assist device implantation”. Am J Cardiol. vol. 109. 2012. pp. 246-51. (A description of echocardiographic findings that can be used to predict right ventricular failure after ventricular assist device implant.)

Major trials of ventricular assist devices

Rose, EA, Gelijns, AC, Moskowitz, AJ. “Long-term use of a left ventricular assist device for end-stage heart failure”. N Engl J Med. vol. 345. 2001. pp. 1435-43. (The REMATCH trial was the first trial to randomize patients to optimal medical therapy versus ventricular assist device. The trial demonstrated improvement in both survival and quality of life with ventricular assist devices.)

Miller, LW, Pagani, FD, Russell, SD. “Use of a continuous-flow device in patients awaiting heart transplantation”. New Engl J Med. vol. 357. 2007. pp. 885-96. (The HeartMateII-Bridge to Transplant study demonstrated the outcomes of using a continuous flow ventricular assist device as a bridge to transplantation.)

Slaughter, MS, Rogers, JG, Milano, CA. “Advanced heart failure treated with continuous-flow left ventricular assist device”. N Engl J Med. vol. 361. 2009. pp. 2245-51. (This is a randomized trial comparing continuous flow ventricular assist devices versus pulsatile flow ventricular assist devices in a destination therapy population. The results showed significant improvement in survival time free from stroke or device failure with use of a continuous flow ventricular assist device.)

Strueber, M, O’Driscoll, G, Jansz, P. “Multicenter evaluation of an intrapericardial left ventricular assist system”. J Am Coll Cardiol. vol. 57. 2011. pp. 1375-82. (This is the first multicenter report of the Heartware HVAD centrifugal flow ventricular assist device.)

Left ventricular assist device for reversal of heart failure

Birks, EJ, Tansley, PD, Hardy, J. “Left ventricular assist device and drug therapy for the reversal of heart failure”. N Engl J Med. vol. 355. 2006. pp. 1873-84. (A description of patients and medical therapy in a population of patients who had ventricular assist device therapy used as a bridge to recovery of left ventricular function.)

No sponsor or advertiser has participated in, approved or paid for the content provided by Decision Support in Medicine LLC. The Licensed Content is the property of and copyrighted by DSM.

Managing the Patient with a Ventricular Assist Device