Circulatory Assist Devices in Heart Failure

Acute Cardiogenic Shock

During the past decade, the incidence of acute cardiogenic shock has doubled in the United States and remains an important cause of cardiovascular morbidity and mortality. Most commonly, cardiogenic shock results from left ventricular (LV) failure after acute myocardial infarction (MI), or a mechanical complication following MI such as ventricular septal defect or mitral insufficiency ( see also Chapter 19 ). However, other conditions may present with similarly deranged hemodynamics, such as acute viral myocarditis ( see also Chapter 28 ), giant cell myocarditis, or acute aortic insufficiency (AI) ( see also Chapter 26 ). Postcardiotomy shock has been reported as a complication of cardiac surgery in 0.2% to 6% of cases and is associated with high short-term mortality risk without mechanically assisted circulation.

Despite advances in coronary reperfusion, including a focus on early intervention, post-MI cardiogenic shock is associated with high short-term mortality. The SHOCK II-IABP trial examined the impact of the intra-aortic balloon pump (IABP) in patients with cardiogenic shock following acute MI. The 30-day mortality rate was 40% in both the IABP and medical therapy arms of the trial despite revascularization and contemporary medical therapy.

The approach to acute cardiogenic shock requires rapid integration of clinical information targeted at determining the etiology, the severity of hemodynamic compromise, and the therapeutic options that address the physiologic needs of the individual patient ( Fig. 45.1 ). A directed history, physical examination, and electrocardiogram (ECG) are critical elements of the initial evaluation. If the cause or severity of the heart failure is not evident following the aforementioned, echocardiography and/or coronary angiography should be performed to evaluate ventricular and valvular function. Endomyocardial biopsy should also be considered in new-onset, nonischemic cardiomyopathy but should probably be limited to centers with expertise in the performance of the procedure and interpretation of the histology.

Initial interventions should include appropriate volume resuscitation, vasodilators in selected patients, and inotropic agents if the patient remains in shock. Placement of a pulmonary artery catheter ( see also Chapter 34 ) has been advocated to guide volume administration and vasoactive drug therapy. Mechanical circulatory support (MCS) should be considered in patients with persistent evidence of shock despite the aforementioned interventions. Device selection should be tailored to each patient’s unique hemodynamic abnormalities and the need for respiratory support.

Devices

Intra-Aortic Balloon Pump

Over the past 50 years, the IABP has been the most commonly used MCS device. The IABP is generally inserted retrograde in the aorta via the femoral artery and positioned with the distal tip just beyond the left subclavian artery ( Fig. 45.2 ). Balloon filling is triggered from the ECG or from the arterial pressure trace; the balloon inflates during diastole and deflates during systole. The favorable physiologic effects of diastolic augmentation include enhanced coronary blood flow and reduced left ventricular afterload.

The effectiveness of the IABP is highly dependent on proper timing of the balloon inflation and deflation ( Fig. 45.3 ). Optimal timing results in IABP inflation just after the dicrotic notch in the aortic pressure tracing and deflation before the pressure upstroke of ventricular systole. The hemodynamic and physiologic benefits of IABP support include elevation of systemic blood pressure relative to unassisted beats and reduction of LV afterload, LV wall stress, and myocardial oxygen demand. Inappropriate timing with early inflation or late deflation results in balloon expansion during ventricular systole increasing the afterload against which the ventricle is ejecting. Late balloon inflation or early deflation limits the hemodynamic benefits of the therapy. The hemodynamic effectiveness of the IABP may be limited by tachycardia, such as atrial fibrillation with rapid ventricular response. More than mild aortic valve insufficiency is likely to limit the hemodynamic benefits of IABP therapy by increasing LV loading and is a contraindication to therapy. Significant aortic or iliofemoral atherosclerotic disease is also a relative contraindication to IABP support and has led some to propose alternative insertion strategies, including subclavian artery or direct aortic access used in the context of cardiac surgery.

The IABP has been used as an adjunctive therapy for many cardiac conditions, including acute MI, postinfarction VSD, acute mitral insufficiency with compromised hemodynamics, and cardiogenic shock. However, improved outcomes with IABP therapy in clinical trials have been difficult to demonstrate. Perhaps the most validated use of the IABP is as adjunctive therapy for the treatment of acute MI treated with thrombolytic therapy. In this setting, the use of prophylactic IABP was associated with an 18% reduction in all-cause mortality. However, the SHOCK II-IABP trial failed to demonstrate improved survival in a more contemporary cohort of patients with acute MI and cardiogenic shock treated with IABP compared with those supported medically.

The limitations of the IABP coupled with a lack of positive outcome studies has resulted in the proliferation of other percutaneous approaches for the treatment of cardiogenic shock and support of complex cardiac procedures, such as high-risk percutaneous coronary interventions and ventricular tachycardia ablations. These devices can be rapidly inserted and are approved for short-term (hours) support.

TandemHeart

The TandemHeart (CardiacAssist, Pittsburgh, PA) is an extracorporeal centrifugal continuous flow pump that receives blood from a 21-F cannula inserted in the femoral vein and passed into the left atrium via a transseptal puncture (see Fig. 45.2 ). The TandemHeart returns the blood to the arterial circulation via a 17-F catheter inserted in the iliofemoral system. In this configuration, the device can provide up to 5 L/min of flow and is approved for short-term support. The hemodynamic effects of the TandemHeart were compared with IABP in two small, randomized clinical trials that demonstrated superior improvements in cardiac index and the lowering of intracardiac filling pressures with the TandemHeart pump. A nonrandomized, experiential series described the potential benefits of the TandemHeart in patients with cardiogenic shock. In this series, 117 patients with clinical evidence of shock (including almost 50% who were receiving or had just received cardiopulmonary resuscitation) were treated with the device. The median cardiac index increase from 0.5 to 3.0 L/min/m ² was associated with improvement in serum lactate and creatinine. The 30-day survival in this cohort was 60% and largely dependent on candidacy for another treatment such as implantable left ventricular assist device (LVAD). Limitations of the TandemHeart device include the transseptal puncture, which adds technical complexity and may require surgical closure if the patient is transitioned to surgical LVAD. In addition, the 17-F arterial cannula in the femoral artery can result in limb ischemia and often requires surgical closure. More recently, the TandemHeart pump has been used in conjunction with a novel dual-lumen catheter (Protek Duo) that allows withdrawal of blood from the right atrium and delivery of blood to the pulmonary artery, providing isolated right heart support. The TandemHeart systems for left and right heart support provide reasonable ventricular unloading and increased cardiac output without the need for major thoracic incisions that were previously necessary for temporary VAD applications. Importantly, these percutaneous systems use smaller cannulas relative to surgically placed devices, and this may result in limited flow and increased risk for hemolysis.

Impella

This miniaturized, microaxial flow pump is incorporated into a catheter-based technology and is available in several sizes capable of producing flows from 2.5 to 5.0 L/min (see Fig. 45.2 ). The smaller Impella (ABIOMED, Danvers, MA) pumps (9 F) can be inserted percutaneously via the femoral artery, whereas the larger device capable of greater blood flow requires surgical implantation techniques. Impella withdraws blood from the distal port in the LV and delivers it to the ascending aorta. This device has been demonstrated to improve cardiac output and reduce left ventricular filling pressures to a greater degree than IABP. Impella 5.0 was studied in a prospective registry that included 16 patients with postcardiotomy shock. Following implantation, the mean arterial pressure increased by 12 mm Hg and the mean cardiac index increased from 1.65 to 2.7 L/min/m ² . There were two primary safety events in this study, one stroke and one death, and the 30- and 180-day survival rates were 94% and 81%, respectively. The Impella EUROSHOCK Registry retrospectively examined 120 patients with cardiogenic shock following MI treated with Impella 2.5. Less than half of the patients were able to be weaned from support, with an associated 30-day mortality rate of 64%. Furthermore, 15% of the patients experienced a major cardiac or cerebrovascular adverse event. Finally, a randomized trial of Impella CP versus IABP was conducted in patients with cardiogenic shock following acute MI. No difference was observed between treatment group in either 30- or 60-day mortality rates.

The design of Impella has been reconfigured to allow percutaneous right-sided support. The Impella RP features a 22-F pump mounted on an 11-F catheter that withdraws blood from the right atrial/inferior vena caval junction and delivers the blood to the pulmonary artery. The RECOVER RIGHT trial prospectively examined the outcomes of 30 patients with right heart failure following LVAD, cardiotomy, or an MI who were treated with Impella RP. The hemodynamic benefits of Impella RP support included clinically meaningful improvements in central venous pressure and cardiac output. The 30-day survival rate was 73% in this cohort.

Extracorporeal Membrane Oxygenation

Extracorporeal membrane oxygenation (ECMO) is a temporary strategy to provide circulatory and/or respiratory support to critically ill patients. The ECMO circuit consists of a cannula inserted either percutaneously or centrally in the venous system for device inflow. ( Fig. 45.4 ). A centrifugal flow pump moves the blood through an oxygenator and returns it to the body via a cannula placed in the arterial system (venoarterial ECMO for cardiorespiratory failure) or to the venous system (venovenous ECMO for respiratory failure). Flow rates of 4 to 6 L/min are typical for most adult patients. ECMO can be initiated rapidly, and peripheral cannulation allows its use in many settings, including the cardiac catheterization laboratory, the intensive care unit, and the operating room. Overall, application of both venovenous (VV) ECMO and venoarterial (VA) ECMO has increased in the United States, related mainly to improvements in safety and durability of the oxygenators. In the setting of cardiogenic shock, establishing hemodynamic stability with ECMO allows time to assess cardiopulmonary recovery and improvement in end-organ function. ECMO is generally considered useful for short periods (days to weeks). An important complication of peripheral ECMO that limits longer-term benefit is a lack of direct LV unloading, with resultant ventricular distention and pulmonary venous hypertension. Furthermore, extended support is undesirable because the patient is typically confined to bed and the incidence of adverse events, including bleeding, hemolysis, thrombocytopenia, limb ischemia, vascular injury, and stroke, is related to the duration of support. Thus, after stabilization for a brief period, the clinical team must decide on the next step in the patient’s care. In some cases, ECMO can be weaned and the patient separated from the system. In other cases, it serves as a bridge to another procedure such as permanent MCS or transplantation. There are limited outcomes data examining the role of ECMO for the treatment of heart failure and cardiogenic shock. Survival following ECMO support appears to be strongly related to the underlying cause of the ventricular dysfunction, as well as the timing of application, with patients placed on ECMO following cardiac arrest faring poorly. ECMO-supported patients still have a 50% in-hospital mortality, with 6-month survival rates as low as 30%. ECMO has also been used to provide hemodynamic support during high-risk procedures such as percutaneous coronary interventions and ventricular tachycardia ablations.

Indications for Implantable Mechanical Circulatory Support Devices

Decision-making regarding implantation of durable MCS devices is dependent on the clinical status of the patient and the recognized indications for the therapy. Historically, there are two recognized indications for implantable LVADs: as a means to support critically ill patients until they can receive cardiac transplantation (bridge to transplant [BTT]) or as permanent therapy in non–transplant candidates (destination therapy [DT]). This narrowly focused paradigm is not aligned with contemporary use of these devices, and the following definitions are commonly used by clinicians:

Bridge to bridge is a strategy in which a short-term circulatory support device is used until a more definitive procedure can be performed. This is typically used for patients in cardiogenic shock who require rapid hemodynamic restoration to reverse the shock state and/or improve end-organ function. Device selection depends on the severity of hemodynamic compromise, the presence or absence of biventricular heart failure, and the anticipated duration of this approach. In many cases, percutaneous devices or ECMO are used.

Bridge to recovery may be used in disease processes anticipated to recover with a period of hemodynamic support, such as acute myocarditis, peripartum cardiomyopathy, cardiac transplant rejection with hemodynamic compromise, or postcardiotomy shock. Selection of the most appropriate device typically involves determination of the need to provide partial or full hemodynamic support and the projected duration of therapy.

Bridge to decision acknowledges that transplant candidacy is frequently confounded by potentially reversible comorbidities when the decision for durable MCS is made. The favorable hemodynamic impact of LVAD support commonly improves end-organ function, lowers pulmonary artery pressures, and allows the patient to become physically and nutritionally rehabilitated before consideration of transplantation. However, if the patient does not achieve these milestones, he or she may remain on mechanically assisted circulation for prolonged periods or indefinitely.

BTT is reserved for device implantation in patients listed for transplant at high priority who are failing optimal therapies. DT designates LVAD implantation in a patient with advanced heart failure who is currently ineligible for transplantation. The DT criteria are aligned with the inclusion criteria from clinical trials and include an ejection fraction less than 25%, NYHA class IIIb to IV symptoms, objective functional impairment with a maximal oxygen consumption of less than 14 mL/kg/min (or <50% predicted), and treatment with either optimal medical therapy for 45 of the past 60 days, intravenous inotropic support for 14 days, or an IABP for 7 days. During deliberations for LVAD financial coverage in the United States, the Centers for Medicare and Medicaid Services was unable to agree on the definition of NYHA class IIIb symptoms and subsequently supports only payment for patients with NYHA class IV functional limitations.

More recently, a clinical trial was completed that redefined LVAD implantation into either short- or long-term support. This approach is more aligned with contemporary clinical practice because it is less dependent upon future events (such as transplantation).

Patient Selection for Mechanical Circulatory Support

In general, patients considered for MCS have severely depressed ventricular function, have marked limitation in functional capacity, are treated with evidence-based medical and electrical therapies, and have a high residual mortality risk within the ensuing 1 to 2 years. Patient selection is critically important to achieving optimal postoperative outcomes. Selection criteria should identify patients with sufficient severity of illness to derive benefit from MCS while simultaneously avoiding those with a severity of illness or comorbidities that would compromise survival following implantation. Baseline characteristics of patients enrolled in LVAD trials demonstrated end-organ dysfunction with hyponatremia and elevated serum blood urea nitrogen and creatinine levels. In addition, the mean ejection fraction was less than 0.20 with elevated right- and left-sided cardiac filling pressures and mean cardiac index of 2.0 L/min/m ² despite treatment with continuous infusion intravenous inotropes in 80% to 90% of patients and IABP support in 20% to 40%.

DT was originally conceived as a treatment for patients with end-stage heart failure ineligible for cardiac transplantation. As a result, many of those being referred for DT LVAD are older than 65 years. Older age has been identified as an important predictor of adverse outcomes in the VAD population. The HeartMate II risk score demonstrated an increased postimplant mortality risk of 32% per decade. Data from the Interagency Registry for Mechanically Assisted Circulation (INTERMACS) also described older age as a risk factor for early mortality following LVAD placement and highlighted the important interaction between age and other risk factors for mortality, such as severity of illness. However, carefully selected patients older than 70 years appear to derive similar benefits with VAD as a younger cohort, raising the important concept of chronologic versus physiologic age in patient selection. Chronologic age is likely an imperfect surrogate for the true predictors of adverse outcomes in this population, which are more likely measures of frailty and debilitation.

Beyond age, other contraindications to implantable VAD therapy appear to influence short- and long-term outcomes and must be considered in the overall risk assessment of the candidate. INTERMACS developed a new nomenclature for classification of advanced heart failure that has been used to understand the impact of severity of illness on outcomes ( Table 45.1 ). Patients with INTERMACS profile 1 and 2 have a high early mortality hazard relative to MCS patients with lesser degrees of hemodynamic compromise, leading many centers to be highly selective in the use of durable implantable LVADs in these patient cohorts.

Right ventricular (RV) failure, defined as the need for prolonged inotropic therapy to support the right heart or a RV assist device, remains an Achilles heel of LVAD therapy and is associated with multisystem organ failure, prolonged hospitalization, and increased morbidity and mortality following LVAD implantation. Unfortunately, prediction of post-LVAD RV failure is challenging despite identification of individual parameters and multivariable models that provide insights into the likelihood of RV failure in larger patient populations. Predictors of RV failure following LVAD fall into three general categories: (1) echocardiographic measurements; (2) hemodynamic parameters; and (3) clinical features before LVAD insertion. Increased RV size and severe RV systolic function are associated with post-LVAD RV failure. Quantitative measures of RV performance, such as a tricuspid annular plane systolic excursion (TAPSE) of less than 7.5 mm, reduced RV peak longitudinal strain, and the severity of tricuspid insufficiency have been shown to be useful markers in the prediction of RV failure after LVAD. Hemodynamic variables such as a central venous pressure to pulmonary capillary wedge pressure ratio of greater than 0.63 or an RV stroke work index of less than 250 to 300 mm Hg × mL/m ² are linked to worse outcomes following LVAD placement. Finally, general clinical features such as preoperative mechanical ventilation and abnormal renal and hepatic function have been identified as risk factors for RV failure. A recent validation study of several published RV failure risk scores demonstrated only modest accuracy, highlighting the real clinical dilemma facing clinicians in the preimplant prediction of this important comorbidity.

Renal failure requiring dialysis is considered a strong relative contraindication to durable MCS. Significant renal dysfunction was an exclusion criterion in the clinical trials, so the benefit and potential incremental complications of implanting an LVAD in dialysis patients are unknown. However, 1-year survival in LVAD patients requiring renal replacement therapy is approximately 50% and significantly reduced compared with nondialysis patients in the INTERMACS registry. Furthermore, support with newer-generation LVADs that provide continuous flow results in a minimal (and often imperceptible) pulse pressure, making measurement of blood pressure difficult during hemodialysis.

Active systemic infection is a strong relative contraindication to LVAD implantation. Patients with fever or unexplained leukocytosis should undergo thorough evaluation, including blood and urine cultures, chest x-ray, and other diagnostic testing directed at potential sites of infection. Hospitalized patients and those with chronic indwelling catheters should have intravenous cannulae removed. Patients with pacing systems and unexplained bacteremia may require chest wall or transesophageal echocardiography to rule out pacemaker-associated endocarditis.

An evaluation for cerebrovascular disease should be performed in at-risk patients using noninvasive imaging. The presence of a prior stroke does not preclude implantation of an LVAD, but consideration must be given to the potential for meaningful rehabilitation and the patient’s ability to interact with the device. For example, an individual with hemiparesis of a dominant arm may have difficulty making the electrical connections required to operate the VAD.

Other end-organ dysfunction may also limit favorable outcomes with VAD therapy and should be considered during the evaluation. Individuals with clinically significant chronic obstructive pulmonary disease whose FEV ₁ is less than 1 L are likely to have residual dyspnea despite hemodynamic improvement and may have difficulty weaning from the ventilator postoperatively. A VE/MVV ratio of more than 80% on a preoperative cardiopulmonary exercise test suggests a pulmonary component to dyspnea. Patients with long-standing right heart failure or other conditions associated with liver injury should undergo an evaluation for hepatic insufficiency. Serum transaminases, albumin, and imaging studies to examine the texture and contour of the liver may provide insights about the necessity for liver biopsy. The presence of an elevated model for end-stage liver disease (MELD) score has been linked to higher post-LVAD mortality. Careful evaluation of the coagulation system is warranted in individuals with a history of a bleeding diathesis or in those with unexplained thrombotic or thromboembolic events. Patients with a history of gastrointestinal bleeding or intolerance to systemic anticoagulation with warfarin should be carefully evaluated because of their high risk of rebleeding following LVAD implantation. Patients with a low platelet count and exposure to heparin should be screened for heparin-induced thrombocytopenia with a PF4 antibody and a serotonin release assay. To the extent possible, patients should have a normal coagulation profile before MCS surgery, because an elevated international normalized ratio (INR) at the time of LVAD implantation was identified as a risk factor for mortality. Correction of coagulopathy will reduce the likelihood of bleeding complications and associated perioperative morbidity. Malnutrition is considered an important risk factor for adverse outcomes, including infection, prolonged debilitation, and mortality. However, the ability to favorably impact nutrition in a critically ill heart failure patient is unclear. Instead, nutrition management should be a primary focus of the entire VAD team following device implantation. Supplemental enteral feedings may be required perioperatively, with additional support in the outpatient setting until nutritional deficits are corrected.

Disease processes with an anticipated survival of less than 3 years were an exclusion criterion in the DT clinical trials, so there are no data supporting the role for mechanically assisted circulation in the management of these patients.

Psychosocial factors also play a pivotal role in VAD outcomes. As part of the evaluation, patients should be seen by multiple health care providers, including those who focus primarily on prior history of compliance, substance use, health literacy, and the availability and abilities of family and friends who will participate in the ongoing outpatient management of the patient and the device. There is a high caregiver burden with MCS, including the need for device training, care of the percutaneous driveline, and companionship. These issues and expectations need to be clearly articulated by the team and agreed on by the patient and his or her caregivers prior to device implantation.

In an attempt to integrate large numbers of predictive clinical variables, the HeartMate II Risk Model was derived from a large clinical trials database and demonstrated that age, elevated INR, increased serum creatinine, and lower serum albumin were predictive of postimplant mortality. Follow-up analyses in institutional datasets suggest only modest predictive accuracy (C-statistic 0.6) for short- and long-term outcomes.

Recent successes in mechanically assisted circulation have resulted in acceptance of this approach as a useful therapy for the treatment of selected patients with advanced heart failure. The INTERMACS registry has captured almost all implants using FDA-approved MCS devices in the United States since 2006 and has carefully documented the growth of this field following the introduction of the new-generation continuous flow devices. The number of centers implanting long-term devices is increasing and has expanded from traditional transplant centers to programs that do not perform transplantation. The impact of center volume on outcomes was recently reported from INTERMACS and showed that both very low volume and high volume programs had higher perioperative and long-term mortality rates.

Mechanical Circulatory Support Devices

The mechanical blood pumps can be characterized in several ways: temporary versus permanent, intracorporeal versus extracorporeal, and pulsatile flow versus continuous flow. At present, the vast majority of clinically available pumps are continuous flow devices. Pulsatile flow pumps such as the ABIOMED 5000, Thoratec PVAD and IVAD, Novacor LVAD, and HeartMate XVE are of historical interest. However, their importance in supporting patients and forming the foundation of the principles of mechanically assisted circulation cannot be underestimated. For example, the HeartMate XVE ( Fig. 45.5 ) and Novacor LVAD were the original electric, implantable LVADs that were tested in clinical trials and shown to be superior to optimal medical heart failure treatment in patients either awaiting transplantation or as DT.