Ventricular assist devices (VADs) can be invaluable tools for the management of end-stage heart failure in children. While development of such devices in adults has grown significantly since the 1960s and has led to several generations of VAD enhancements, the development of such devices in children has not grown at the same rate. This slower evolution has been due to a number of factors, such as the inherent differences in anatomic and physiologic parameters in children, especially in patients with congenital heart disease involving a variety of defects, as well as developmental changes in the coagulation system. The last 10 years, however, have witnessed an increased attention for VAD development and use in the pediatric population. This increase is due to the fact that while heart transplantation remains the ultimate therapy for children with advanced heart failure, the number of pediatric heart transplants has not increased over the past 10 years to meet the increase in demand. In fact, of all patients on the waiting list for solid-organ transplantation in the United Sates, children listed for heart transplantation face the highest waiting list mortality regardless of age.^1,2 Therapeutic options, therefore, become limited at the point of end-stage heart failure, particularly with pediatric heart transplant wait times often exceeding 4 to 6 months, especially for patients 5 years of age and younger. Due to these factors, it is anticipated that more children with both congenital heart disease and cardiomyopathy-associated heart failure will require long-term VAD support in the coming decade as a bridge to transplantation or recovery.

Historically, extracorporeal membrane oxygenation (ECMO) was the only readily accessible form of mechanical assist for both long and short-term (i.e., postcardiotomy) support. Long-term use of ECMO was limited by complications and often lack of rehabilitation of patients in preparation for transplantation. There were limited options for other long-term assist devices until 2000, when the Berlin Heart EXCOR pulsatile VAD was first utilized in a US pediatric patient under compassionate use approval by the FDA. Over the ensuing 10 years, the Investigational Device Exemption (IDE) study was conducted by Berlin Heart, Inc., eventually leading to FDA approval in late 2011.^3,4 Over this time, there have been advancements in the design of such devices, which permit their placement in neonates and older children in univentricular or biventricular support configuration. There is also a wealth of research currently underway for the development and evaluation of new VADs to further enhance outcomes for pediatric patients requiring these devices as part of the PumpKIN (Pumps for Kids, Infants and Neonates) trial sponsored by the NIH NHLBI.

While pediatric indications for mechanical circulatory support are often very different from those of the adult population, the same goal of bridging the failing heart to recovery or transplant may be accomplished. Successful myocardial recovery may occur with temporary mechanical assist for myocarditis, failure to wean from bypass (i.e., postcardiotomy) and heart transplant graft failure, through the use of ECMO, Tandem Heart or Rotaflow, to name a few examples, with ECMO remaining the most commonly used short-term device at present. More long-term mechanical support for heart failure due to cardiomyopathy, some cases of myocarditis or various congenital heart lesions often requires a VAD such as Berlin Heart EXCOR for infants and children and Thoratec HeartMate II or HeartWare for older children and adolescents. Most commonly, long-term pediatric VADs are placed with the intent of successful bridging to transplantation, although recovery has been reported.^5–8 The results of several studies indicate a definite benefit in this patient population, with a recent study revealing significantly higher survival rates with the Berlin Heart EXCOR device than with ECMO for pediatric patients with heart failure. However, VADs can be associated with significant morbidity and mortality, most commonly due to infection, hemorrhage and/or thromboembolism resulting in stroke.²

Mechanical support for failing pediatric hearts must be customized to the individual’s need, size, anatomy and physiology. Certainly, in children from neonatal age to adolescence, one size does not fit all. In the United States, the initial reports of mechanical pediatric cardiac support entailed utilization of devices initially intended for the adult patient population. Therefore, many of these devices were too large to accommodate the specific requirements of the small child and ECMO was the mainstay of circulatory assist in this particular group. Fortunately, with the development of the EXCOR Berlin Heart VAD and FDA approval for study in the United States, it has become the most commonly used VAD for pediatric patients less than or equal to body surface area (BSA) 1.5 m². (Of note, Berlin Heart EXCOR has pumps available for adult size patients, but is not FDA approved for use above 1.5 m² at this time.) For older patients requiring VAD support, Heartmate II and Heartware have become viable options and allow patients to resume normal daily activities, such as participation in school.

Pediatric patients in acute need of ventricular assistance often differ from those who present with more chronic heart failure that has progressed despite maximal medical therapy. Indeed, the decision as to which device to use should be based on the indication and prognosis (Fig. 88-1). A patient who cannot be weaned from cardiopulmonary bypass might only need ventricular support for a few days while the myocardium recovers, while another patient with end-stage heart failure might require a VAD designed for longer use that is both portable and durable in anticipation of a lengthy wait on a heart transplant list.

New devices are constantly being developed, and adult VADs will continue to be adapted to the pediatric patient population. As the technology evolves and becomes more readily available, VADs will likely become a mainstay of the management of pediatric heart failure in both the acute and chronic settings.

Pediatric heart failure severe enough to necessitate a VAD often differs from adult heart failure (Table 88-1). For children, the most common etiologies of myocardial failure include postoperative failure to wean from cardiopulmonary bypass, congenital heart disease, cardiomyopathies, and acute myocarditis.^8,9 Children do not usually suffer from the typical adult problems of hypertension and ischemic heart disease. Congenital heart disease, for example, may entail elements of right and left ventricular (LV) and pulmonary dysfunction and cardiomyopathy with LV failure may result in elevated pulmonary vascular resistance (PVR). While biventricular support is available for such patients, the outcomes for patients with BiVAD (right and left) support are worse than those with univentricular support.² Furthermore, with the advent of pulmonary vasodilators, such as milrinone and sildenafil, patients with elevated PVR due to left heart failure may have near normalization of their pulmonary pressures following unloading of the diseased LV through the use of an isolated LVAD.

By assuming much of the responsibility for cardiac output, VADs may allow for the failing myocardium to recover or at least for the patient to have increased mobility and rehabilitation and improved nutrition if transplantation is required. Decompression of the cardiac chambers through mechanical assistance can lower wall stress, reduce myocardial oxygen consumption, and allow for recovery of the myocardium. Other advantages of VAD support include potentially lower requirements for inotropic and vasoactive agents that have, as some of their side effects, arrhythmias, peripheral vasoconstriction, and hyper- or hypotension. For those patients with reasonable pulmonary function, VADs offer isolated ventricular support without the need for oxygenation. In these patients (as opposed to patients on ECMO, for example), the benefits of not requiring an oxygenator in the circuit include potentially lower anticoagulation requirements, less platelet destruction, fewer bleeding complications, and the possibility of mobilization.

Congestive heart failure in children is primarily attributable to congenital heart defects, cardiomyopathies, and myocardial dysfunction following cardiac surgical repair. The yearly incidence of heart failure from structural defects is 0.1 to 0.2 percent of live births, while the yearly incidence of cardiomyopathy in infants and children has been estimated to be 0.6 in 100,000. The number of children being hospitalized with heart failure is increasing. In an analysis of 15 million pediatric hospitalizations in the US, it was reported that the number of children requiring hospitalization for heart failure increased by 25 percent between 2003 and 2006. During that time period, there was a 32 percent increase in patients hospitalized for heart failure whose underlying diagnosis was congenital heart disease, and a 14 percent increase in patients with cardiomyopathy. Furthermore, about 10 to 20 percent of those who have undergone complex repairs of congenital cardiac defects will develop heart failure by young adulthood.²

Aside from failure to wean from cardiopulmonary bypass, typical symptoms of heart failure in infants include tachypnea, tachycardia, failure to thrive, and poor feeding. Patients may develop signs of congestion, including an enlarged cardiac silhouette on chest radiographs, pulmonary edema, and hepatic enlargement. In toddlers and older children, jugular venous distention and edema might be findings observed in association with growth failure and fatigue. As children become adolescents, the severity of heart failure can be classified as it is in adults, typically with the New York Heart Association (NYHA) functional classification.

If untreated, heart failure ultimately compromises end-organ perfusion, thus leading to the downward spiral of multi-organ system failure. Altered mentation, decreased urine output, rising creatinine, increasing liver function indices, and elevated lactic acid levels all point to inadequate perfusion. If cardiac output or perfusion pressures do not meet the metabolic demands of the body, persistent decline and circulatory collapse will ensue.

Initial therapy for pediatric heart failure will usually include diuretics (including aldosterone), and β-blockers as well as angiotensin-converting enzyme inhibitors and/or digitalis. As heart failure progresses, patients will require admission to the intensive care unit, along with inotropic and vasodilator support to maintain adequate cardiac output. Milrinone is most commonly used in these circumstances with the addition of low dose dopamine and/or low dose epinephrine if further inotropic support is necessary, realizing that all of these medications have the potential to result in development or worsening of arrhythmias. With worsening failure, patients may require intubation and ventilatory support, hemodynamic monitoring with a central venous or Swan-Ganz catheterization, close monitoring of renal function and placement of an indwelling urinary catheter. Adequate enteral nutrition must also be provided, usually by means of a feeding tube whenever possible.

The indications for VAD placement are much the same in larger children as in adults, even though the disease process is often different (Table 88-2). For smaller children, such as infants and toddlers, indications for VAD support are much more vague and often rely on clinical judgment. Signs such as recurrent ventricular arrhythmias, inability to feed and recurrent pulmonary edema are particularly ominous in this patient population. Some centers utilize involvement of more than one system as a potential trigger for VAD placement (Table 88-2). The inability to tolerate enteral feeds, for example, suggests worsening heart failure with decreased splanchnic perfusion. In pediatric patients, catastrophic collapse can occur with great rapidity and without the usual clinical warning signs often seen in adults. Near-death and resuscitation events, even when patients are stabilized, should not be given another chance to occur. In a medically optimized patient, elevated cardiac filling pressures along with persistent acidosis and hypotension indicate poor perfusion, often secondary to poor ventricular function. These are the types of “gray-zone signs” that should lead one to contemplate institution of VAD support. In essence, if clinical deterioration secondary to heart failure continues despite maximal inotropic support, strong consideration should be given to insertion of a VAD.

Pediatric VADs are by definition “bridging” devices to either recovery or transplantation.^2,10 The concept of “destination therapy,” which is gaining acceptance for the adult population, is not accepted for pediatric patients at this time. Ethical considerations (including futility and hopelessness) must also be addressed before placing a VAD, particularly in cases of cardiac arrest.

The indications for VAD placement can be divided into acute and chronic heart failure. This distinction may affect the choice of which device to place once the decision for VAD implantation has been made. For example, postcardiotomy patients can initially be maintained on VADs designed for short-term use. Options like ECMO may be equivalent if not better in this setting. Patients with chronic heart failure might benefit more from a continuous flow device (such as HeartMate II or HeartWare) or a pulsatile VAD (such as Berlin Heart EXCOR) that allow for earlier mobility and less lifestyle restrictions. There is a rising population of patients who will fail to wean successfully from “acute” devices and will need conversion to more “chronic” support.¹¹ This will increasingly become the case, as more pediatric VADs become available.

As pediatric VADs become more commonplace in clinical practice, new indications may arise for their utilization. Rather than merely representing a salvage procedure, VADs may become a standard adjunct for some complex congenital heart operations. One group of investigators has suggested routine ventricular assist after complex congenital reconstructions like the Norwood procedure, citing greater ease of postoperative management as an advantage.¹² With greater use and long-term follow-up, the indications for VAD placement will change and might be further broadened.

At present, only a limited number of pediatric VADs are available in the United States. This should change in the near future, as promising results from European studies are being reported along with ongoing US studies as well. The range of mechanical assist devices for the pediatric heart includes ECMO circuits as well as pulsatile and nonpulsatile VADs. Intra-aortic balloon pumps (IABPs) are used much less frequently in the pediatric population. Although ECMO for cardiac support has been directly compared to nonpulsatile and pulsatile VADs in some series,¹³ there are few data directly comparing one VAD to another.

Intra-Aortic Balloon Counterpulsation

Although use of the IABP has greatly facilitated the management of adult patients with heart failure, adaptation to the pediatric population has not been as fruitful. The IABP is a biocompatible balloon that is introduced into the descending thoracic aorta and intermittently inflated/deflated (triggered by the electrocardiogram or blood pressure tracing) to augment diastolic blood pressure. By deflating prior to the initiation of systole, the device decreases afterload and thus ventricular work, by creating a “vacuum.” During diastole, inflation of the balloon improves coronary as well as splanchnic perfusion. Pediatric anatomy and physiology have posed many obstacles for the use of this technology. Difficulty with insertion remains a challenge even with smaller designs, while higher pediatric heart rates make timing a significant challenge, particularly in a 1:1 setting. Moreover, some feel that the pediatric aorta is considerably more elastic and compliant than that of the adult, which limits the counterpulsation effect; this assertion has nevertheless been challenged.¹⁴ Furthermore, most pediatric patients have unobstructed coronaries, rendering the increased coronary perfusion afforded by the IABP less relevant than it is in adults. Patients with IABP suffer from problems with thrombosis and emboli, as well as the destruction and consumption of cellular blood components, therefore requiring transfusion. For these reasons, the IABP has a limited role in the pediatric population.¹⁵ Although they may be reasonable for short-term support in larger children, IABPs have a very limited role in their long-term mechanical support or bridging to transplantation.

Extracorporeal Membrane Oxygenation

ECMO consists of a centrifugal or roller pump combined with an oxygenator, which is used for venoarterial or venovenous bypass (Fig. 88-2). Introduced in 1957, this modality has proven to be most useful for neonates with respiratory failure. However, it has also been utilized with success in pediatric cardiac failure. At this time, it is the most widely used method of pediatric ventricular support in the United States, mostly secondary to its availability and ease of placement. Unlike some other VADs, ECMO can be used for children of all sizes, from premature neonates to near-adults. It can be placed centrally (through the chest) or peripherally with either carotid artery and jugular vein or femoral artery and vein cannulation. Physiologically, venoaterial ECMO does result in a decrease in LV preload as well as an increase in LV afterload as ECMO flow is increased. Furthermore, without venting the left heart, it can result in increased ventricular wall stress because of the inability to decompress the left-sided cardiac chambers. Left atrial and ventricular distention occurs to a degree proportional to the number of aortopulmonary collaterals; it may be managed by venting the left atrium either directly or trans-septally via the right atrium.^16,17 ECMO by design improves arterial oxygenation and, if cannulated in a venoarterial fashion, acts to decompress and support both ventricles. For those patients with mainly pulmonary dysfunction (meconium aspiration, for example), ECMO can be instituted in a venovenous configuration, thereby improving venous blood oxygenation and myocardial oxygen delivery. Pulmonary hypertension resulting from low venous blood saturations may also be improved, allowing for some off-loading of the right heart. Since the ECMO circuit involves both a pump and an oxygenator, it is reasonable that the pump alone may suffice for those patients without pulmonary insufficiency. By eliminating the oxygenator, it is possible to decrease the amount of foreign surface contact in the circuit, thus decreasing the inflammatory response. Moreover, if the venous cannula is placed in the left atrium or ventricle, the same pump may be used as a left VAD. For this reason, the most commonly used true VAD in the United States has been the centrifugal pump. ECMO is discussed in further detail in Chapter 87. For the purpose of this chapter, we categorize VADs into nonpulsatile and pulsatile types.