Introduction
The use of long-term mechanical circulatory support (MCS), predominately in the form of ventricular assist devices (VADs), has undergone a major transformation in the past 15 years. In the early 2000s, device use was uncommon. Support was typically for short duration, in patients with cardiogenic shock, and pediatric-specific device use, especially in the United States, was exceptionally rare. In the initial report on VADs from the Pediatric Heart Transplant Study database, covering children listed for heart transplantation from 1993 to 2003, only 4% of patients were supported with a VAD and only one Berlin Heart EXCOR VAD was used. This is in contrast to the current state where nearly 50% of pediatric patients with dilated cardiomyopathy (DCM) in the International Society for Heart and Lung Transplantation Registry are bridged to transplant with a VAD. In most pediatric transplant centers, VADs would currently be considered a standard part of advanced heart failure management, predominately being used as a bridge to transplant.
Patient Selection
Long-term MCS is generally used when patients have failed medical therapy for severe heart failure. Heart failure in children is not uncommon, with more than 14,000 hospitalizations annually in the United States, with diverse etiologies including congenital heart disease (CHD), cardiomyopathy, and myocarditis. The majority of VADs in children nowadays are used in patients with DCM. Although CHD accounts for 60% to 70% of heart failure–related hospitalizations, only 20% of VADs are implanted in children in CHD. The underlying disease and size of the patient are important considerations for the type of VAD used.
Disease severity is important for consideration and timing of MCS. Many children with less advanced stages of heart failure can be managed on oral therapy as outpatients with few cardiac-related symptoms. Indeed, a substantial minority of children with DCM will have improvement and even normalization of their ventricular function over time. However, heart failure is a severe, progressive disease that carries a high risk of major morbidity and mortality. Many patients will develop symptoms that are refractory to medical therapy, either after a period of long-term medical therapy or during the first presentation of acute heart failure. It is in these patients that placement of a VAD is considered.
Decisions about the timing of VAD placement can be among the most challenging in caring for children with advanced heart failure, but there is growing evidence that earlier placement, before the patients are in shock with multiorgan system failure, leads to improved outcomes. In the most recent annual report from Pedimacs, a large North American registry of pediatric VAD patients, patients who were Pedimacs profile 1 (critical cardiogenic shock) at the time of VAD implant had significantly inferior survival compared with those that were less ill at the time of VAD implant ( Fig. 66.1 ). Additional major morbidities that can complicate heart failure include respiratory failure, renal insufficiency, liver dysfunction, and malnutrition/growth failure. These major morbidities are also associated with outcome after VAD implantation. Data from INTERMACS, a large registry of adult VAD patients, evaluating more than 10,000 adult patients, found that preoperative morbidity, including being on a ventilator and having an elevated bilirubin level, a lower albumin level, and renal insufficiency, are associated with a significantly greater risk of death after VAD implantation. Unfortunately, these morbidities are common in children, with 31% Pedimacs profile 1, nearly 50% intubated, and 35% dependent on total parental nutrition at time of VAD implantation. Fortunately, much of the end-organ dysfunction, growth failure, and debilitated condition of many patients can be reversed after VAD placement and prior to heart transplantation and VAD explant. Indeed, the outcome after transplantation among patients supported with VADs is equivalent to patients who come to transplant on medical therapy alone. This would almost certainly not be the case if patients were transplanted in cardiogenic shock or multiorgan system dysfunction, as is frequently the case for VAD implantation.
There is neither widespread agreement nor evidence-based guidelines on the optimal timing for placement of VADs in children. The International Society for Heart and Lung Transplantation published consensus guidelines for MCS in 2013 and recommended that long-term VADs be considered for patients whose ventricular function is unlikely to recover without long-term support, who are inotrope dependent, too ill to maintain normal hemodynamics and organ function without temporary mechanical support or inotropes, who have the capacity for meaningful recovery of end-organ function and quality of life, and who have a high risk of 1-year mortality without VAD support. These are also reasonable guidelines for children, although clearly more data are needed to help understand the optimal time of VAD placement.
Support Strategies
The vast majority of pediatric patients are placed on long-term MCS as a bridge to transplantation. Among 109 patients supported with continuous flow (CF) durable VADs reported to Pedimacs, only six had their VAD implanted as destination therapy (i.e., VAD implant with no plan to list for heart transplantation). This is in contradistinction to adult VAD patients, where nearly 50% of patients are currently implanted either as destination therapy or where a bridge to transplant is deemed unlikely. The use of VADs for destination therapy has been performed in some populations such as neuromuscular disease and will likely represent a growing support strategy in children over the next decade. Myocardial recovery and successful VAD explant without transplantation is uncommon with long-term MCS, especially outside of myocarditis.
Current Devices for Short-Term Mechanical Circulatory Support in Pediatrics
Extracorporeal membrane oxygenation (ECMO) has been the primary means of short-term MCS for many years due to its familiarity and ease of rapid deployment. Although ECMO is useful in certain circumstances such as the need for emergent support or simultaneous pulmonary support, throughout its 25 years of use we have yet to see significantly more than half the cardiac patients survive to hospital discharge reported in multiinstitutional studies. Furthermore, ECMO as a bridge to transplant is a well-established independent risk factor for posttransplant mortality. For isolated heart failure, other forms of short-term MCS have been developed and are preferred in most situations.
Historically, adult short-term left VADs (LVADs) (e.g., BVS 5000) were occasionally used in larger children, with cannula and device size/output being factors limiting their widespread use. Currently, temporary devices are extracorporeal centrifugal pumps used with cardiopulmonary bypass cannulas, such as the ROTAFLOW (Maquet Cardiovascular) and the CentriMag/PediMag (Abbott). These centrifugal pumps are now found in most pediatric ECMO programs and therefore are familiar and accessible to the vast majority of pediatric heart programs. Another pump, the TandemHeart (CardiacAssist), has percutaneously placed cannulas to the left atrium and femoral artery, whereas the Impella (Abiomed) is a percutaneous, rotary heart pump that sits across the aortic valve in the left ventricular outflow tract. Both devices, especially the TandemHeart, require adolescent-sized, if not adult-sized, children. Cannula movement with minimal patient manipulation requiring cannula repositioning makes the use of the TandemHeart in smaller patients quite challenging. Until recently, there was a paucity of pediatric data regarding the outcomes of temporary devices and support, but studies are beginning to accumulate.
The notion of temporary support revolves around a quick and simple cannulation strategy meant to briefly sustain cardiac output (CO) in a patient with a reversible cause of heart failure or in urgent need of MCS. For temporary support, the left atrium can be rapidly cannulated along with the aorta, using bypass cannulas and without the need for cardiopulmonary bypass. This strategy is useful for patients with severe graft rejection or fulminant myocarditis so that perfusion is normalized and end-organ function is supported until the inflamed state of the heart can resolve and function can, hopefully, return. The device can then be removed. However, it can also be used to get a patient out of INTERMACS 1 so they become a better long-term support candidate or in a patient who needs support to allow time to determine etiology of heart failure, neurologic status, genetic issues, and so forth as a bridge to decision.
The PediMag and ROTAFLOW have historically been synonymous with temporary support and were connected to patients with bypass cannulas in a temporary cannulation configuration. However, in the past few years, surgeons have begun connecting the same device pumps to EXCOR cannulas as a bridge to transplantation. This is significantly different because EXCOR cannula placement is a more permanent cannulation technique requiring a more involved surgery and cardiopulmonary bypass. By virtue of using EXCOR cannulas with a pump, these devices are no longer used only as a means of temporary support because it is the cannulas that determine the duration for which support can be provided, not the pump itself. This combination of EXCOR cannulas with CF pumps is being done as a bridge to transplant in smaller children to simplify management of their anticoagulation during their postoperative inflammatory state. Furthermore, centrifugal pumps may require a lower level of anticoagulation, but if they do become thrombosed, they are easier and cheaper to replace. Once the patient stabilizes, the pump can be exchanged for an EXCOR pump to increase patient mobility if desired. A similar strategy is becoming the standard of care for smaller patients (<2 years old) with single ventricle physiology (SVP) and pulmonary artery banding, aortopulmonary collaterals, or a shunted physiology. The ability of the centrifugal pump to accommodate frequent changes in preload and flow at a high output make it a better fit for these single ventricle patients whose pulmonary blood flow can be quite varied, especially soon after surgery. The EXCOR has a set rate and fixed output unless one manipulates the settings, which cannot be done on a minute-to-minute basis. After a single ventricle patient has been supported for some time and is in a steady state (i.e., extubated, on enteral feeds, on stable anticoagulation, and off vasoactive intravenous medications), their preload and CO stabilize and conversion to an EXCOR pump at the bedside is possible. Our current algorithm for determining type of MCS is shown in Fig. 66.2 .
This issue of defining temporary support versus temporary VADs (i.e., centrifugal pumps) was demonstrated in a recent review of the Pedimacs database by Lorts et al. Of the 63 devices implanted under the classification of “temporary” VAD as their first device, 40% were placed with a strategy of bridge to transplant, whereas 60% were more classic temporary strategies (i.e., bridge to candidacy/recovery). Patient median age was 3.7 years, with 41% having CHD and 40% having cardiomyopathy. The median duration of support was 15 days, and 61% of patients received greater than 10 days of support. The median duration of support for transplanted patients was 47 days (interquartile range, 10 to 227), and five patients remained on temporary VAD longer than 5 months. Overall, this multiinstitutional study had a positive result (bridge to transplant/recovery/durable device or alive) in 71% of patients implanted with a temporary device pump.
Several other recent studies have shown good outcomes with longer times on temporary device pumps and as bridge to transplant. A review of the Organ Procurement and Transplantation Network data found that prior to 2011, fewer than three of these devices were used per year as a bridge to transplant, with an explosion in use to 50 of these temporary devices being used as bridge to transplant in 2015. In this review, CentriMag/PediMag was by far the most common, used in 65% of patients, followed by the TandemHeart pump (not system), used in 18%. Importantly, in comparison to the ECMO cohort, a propensity score-matched short-term MCS cohort had longer survival to transplant, as well as longer overall survival. Data have also shown that, if temporary support is used for a chronic heart failure patient with acute cardiogenic failure, the vast majority of these patients will not recover and will require conversion to long-term VAD support.
Current Devices for Long-Term Mechanical Circulatory Support in Children
Berlin Heart EXCOR
The Berlin Heart EXCOR is a paracorporeal pulsatile, pneumatically compressed, volume displacement pump. It is available in 10-, 15-, 25-, 30-, 50-, and 60-mL blood chamber sizes ( Fig. 66.3 ). It is the most studied pediatric VAD, with more than 1800 implants, the only to be FDA approved, and the most commonly used worldwide. Use of customized polyurethane valves has allowed the manufacture of smaller blood pump sizes than are available with “adult” pumps, which have used commercially available mechanical valves. Its first reported successful use as a bridge to transplantation occurred in 1990. The longest known support time has been 877 days to successful transplant. The experience at the Berlin Heart Institute has been extensively reported. A total of 74 pediatric patients were supported there from 1990 to 2006, with a mean age of 7.6 years (range, 2 days to 17 years) and a mean support time of 36 days. Fifteen percent of these patients were weaned from the device, with 43% receiving a heart transplant; 41% of the children died during VAD support. The group notes significant improvement from the year 2000 onward, with a 74% survival rate to transplantation or discharge despite a younger patient population. The authors note that changes in their decision-making process from earlier to later in the experience included earlier implementation of support before the development of significant end-organ failure, improvements in cannula design, apical left ventricle rather than atrial cannulation, fewer biventricular VADs (BiVADs), and a focus on the anticoagulation protocol.
The device was first used in North America in 2000, and by 2004 there was widespread use. The initial EXCOR experience in the United States from 2000 to 2007 ( n = 73) was reported to have a positive outcome in 77% of patients and paved the way for a prospective, multicenter, clinical cohort investigational device exemption (IDE) study. The IDE study enrolled 48 patients from 2007 to 2010 and conclusively demonstrated EXCOR to be superior to ECMO in bridging children to transplantation. The outcomes of all 204 children who underwent EXCOR implantation during the duration of the IDE study was also published and was believed to represent the real-world use of EXCOR in North America. In this study, 75% of patients survived to transplantation or recovery, with lower weight, higher bilirubin values, and BiVAD support predicting early mortality (<2 months). Bilirubin extremes and decreased glomerular filtration rate at implant predicted late mortality (>2 months). The most common cause of death was stroke (33%), most often of thromboembolic origin. Lower weight at implantation and large pump size to body surface area (BSA) have been associated with neurologic injury in other studies. Importantly, only one death in the North American experience was attributed to device malfunction. Recent studies comparing posttransplant survival in those bridged to transplant with EXCOR versus no MCS showed comparable survival between the groups, with 94%, 90%, and 72% survival at 30 days, 1 year, and 5 years, respectively, in the EXCOR group. This was congruent with previous findings, which additionally identified similar 1-year survival between EXCOR patients with and without complications. The latter study also noted a significantly higher incidence of death after transplant in patients with CHD compared with those with cardiomyopathy (26.1% vs. 7.2%).
Total Artificial Heart
The total artificial heart (TAH; SynCardia) is an implantable biventricular, pneumatic compression device that replaces both ventricles anatomically and mechanically ( ). It is available with a 70-mL chamber, which is FDA approved for use in patients with a BSA greater than 1.7 m 2 . The TAH was introduced in 1982 by Robert Jarvik as the Jarvik-7. It has since been known as the Symbion, CardioWest, and is currently the SynCardia TAH. In 1985 the Jarvik-7 bridged a patient 9 days to transplant, and an IDE for bridge to transplant was initiated. By 1993, CardioWest owned the Jarvik-7, and an IDE study was underway for its use as bridge to transplantation or destination therapy. In 2001, SynCardia took over the Jarvik-7, and in 2004, the FDA approved the SynCardia TAH for bridge to transplantation. Unfortunately, only 30 drivers, known as “Big Blue” (190 kg), existed worldwide, which limited use of the TAH until 2009, when the much smaller Companion Driver was introduced and quickly followed by the portable Freedom Driver (weighing approximately 6 kg) in 2010. Since the introduction of the Freedom Driver, which can be carried in a backpack, TAH use has tripled, and more than 40% of the more than 1750 implants have been performed in the past 6 years compared with the previous 25 years.
In pediatrics, the TAH is not meant to replace existing devices in the care of children but help care for those children who do not have a great mechanical support option. First, by removing both ventricles and atrioventricular valves, there is no right heart failure, atrioventricular valve regurgitation, cardiac arrhythmias, ventricular clots, intraventricular communications, or low blood flow. This, combined with a high CO, makes it ideal for patients with chronic rejection postorthotopic heart transplant (where immunosuppression can be stopped), a late failing Fontan circulation with end-organ dysfunction, chronic right heart failure with LVAD, significant biventricular restrictive disease, a large ventricular clot burden, primary arrhythmia-induced heart failure, cancer, or multiple defects that may require repair prior to VAD placement. There is very little known about long-term outcomes in pediatrics for the TAH, but a recent compilation of the world experience of patients aged 21 years or younger reported 43 devices implanted from 2005 to 2015. Positive outcomes were reported in 63% of the patients at 90 days, which is similar to the adult TAH experience.
Because of improving results and the increased use of the TAH in pediatrics and congenital heart patients, the SynCardia 50/50-cc TAH was developed. It is currently the subject of a clinical trial for BSAs ranging from 1.2 to 1.85 m 2 or those who can be demonstrated to fit by virtual implantation technique, which represents the first time that virtual fit has been accepted as a criterion by the FDA. Studies such as those by Moore et al. have demonstrated that virtual fit changed the eligibility of 33% of patients and allowed fit down to a BSA of 0.9 m 2 . This has greatly expanded its use in pediatric patients. The ongoing IDE trial of this device presently has eight patients in the primary and secondary arms, with a survival of greater than 80%. However, the worldwide experience of 60 patients has demonstrated that the 50/50-mL device has had its desired effect, although numbers remain small, of allowing TAH support to increase in congenital patients (4% to 9%), pediatrics (4% to 13%), and perhaps most significantly in women (12% to 70%).
Continuous Flow Ventricular Assist Devices
The introduction of CF devices to the pediatric field led to several important changes in device design, not least of which was smaller device sizes, allowing for intracorporeal support in pediatrics. The CF devices also came with a lower morbidity profile and less pump thrombosis. These advantages, coupled with the ability to go home, led to supporting patients on these devices regardless of whether or not they were listed for transplantation. With a CF device in place, patients can be discharged, resume most “normal” life activities, and be evaluated at a later date for potential heart transplant. However, only approximately 50% of CF VAD pediatric patients are discharged home.
Heartmate II
Thoratec HeartMate II has a rotary, axial flow pump that can provide CF at 2.5 to 10 L/min. Due to its size, the HeartMate II is approved for bridge to transplant or as destination therapy in pediatric patients with a BSA of 1.4 m 2 or greater and can be used for discharge home. This device has been tried and improved in adult patients, with more than 25,000 implants worldwide. Of those, 6451 devices have been in place greater than 3 years, 2513 devices more than 5 years, and 46 for more than 10 years. This device has a low incidence of thromboembolic events and excellent outcomes in pediatrics, with greater than 95% of patients either transplanted, recovered, or remaining on the device at 6 months after implant in a review of INTERMACS data. The most frequent complication was bleeding (21%), and stroke and sepsis occurred in 7% and 11%, respectively. Even in this early series (2008–2011) with only 28 patients followed over 6 months, the results of this study are very encouraging for this device. As of 2017, Abbott has acquired Thoratec and with it, the HeartMate 3, which was FDA approved for “short-term hemodynamic support (e.g., bridge to transplant or bridge to myocardial recovery)” in late 2017, although it has been available for short- and long-term use in Europe since 2015. The HeartMate 3 uses complete magnetic levitation technology to provide frictionless rotor movement, reducing hemolysis and thrombosis. As with former CF devices, we anticipate the HeartMate 3 will begin to see use in select pediatric patients ( ).
HVAD
HeartWare HVAD (Medtronic) is an intracorporeal centrifugal LVAD that has allowed for discharge to home in children down to a BSA of 1.0 m 2 . This device is intracorporeal and is smaller than the HeartMate II, improving device comfort and making it useful in patients with poor healing, such as Duchene muscular dystrophy. An early multicenter study from 2015 reported excellent outcomes in 12 patients from around the world who were discharged home with a mean support time of 290 days, while commenting on at-home management. Patients ranged from 8 to 15 years of age, and the smallest was an 18-kg child with a BSA of 0.8 m 2 , although implantation down to a BSA of 0.6 m 2 has been reported. Eight of the children returned to school. Thirty readmissions occurred in 10 patients (0.02 events/person per month) most commonly for driveline infection (22%). No pump thrombosis or neurologic thrombotic episodes occurred. At the conclusion of the study, five of the patients had been bridged to transplantation, five were awaiting transplant, one device was explanted after recovery, and the final family elected to remain on the device as destination therapy.
In an analysis of the global experience with the HeartWare HVAD, Conway et al. reported on 205 pediatric patients in 2017. Males were more frequently implanted (61%), and cardiomyopathy was the most common diagnosis (82%). More than 50% of patients were discharged home, and this number improved with experience. At 1 year, 65% had been transplanted, 21% remained on the device, 11% had died, and 3% had recovered, for an overall positive outcome of 89%. The authors found that the HVAD system in the pediatric population was associated with low mortality in the majority of patients supported for 1 year. This is comparable to adult data, although temporary right VAD support and pump exchange remain risk factors for poor outcomes in the pediatric population. Although very encouraging, the dataset did not allow an analysis of the morbidities, which will need to be studied.
Prior to 2012, only one pediatric VAD program had an intracorporeal CF device. Now, 5 years later, every program uses intracorporeal CF devices routinely. With recent improvements, CF VADs are now placed more often that EXCORs in pediatrics, according to most recent Pedimacs report. This shift toward CF VADs in pediatrics follows the previous trend in adults.
Future Devices
The National Heart, Lung, and Blood Institute started the Pediatric Circulatory Support Program for development of novel MCS devices for infants and children in 2004. Careful selection led to more than $20 million of support granted to five contractors: PediaFlow pediatric VAD (University of Pittsburgh), PediPump (Cleveland Clinic), pediatric cardiopulmonary assist system (Ension, Inc.), Pediatric Jarvik 2000 (Jarvik Heart, Inc.), and pediatric VAD (Pennsylvania State University). This was followed with funding for the Pumps for Kids, Infants, and Neonates (PumpKIN) preclinical program in 2010 to further develop four MCS devices. Four contracts totaling $23.6 million were allotted for continuation of three of the Pediatric Circulatory Support Program devices (pediatric cardiopulmonary assist system, Jarvik 2000, and PediaFlow), as well as the pediatric pump-lung (PediPL) (Levitronix, LLC). Of the initial devices funded, only the Jarvik 2015 is still undergoing testing. It is an intracorporeal, axial CF VAD no larger than an AA battery, designed to support pediatric patients between 8 and 20 kg. After a series of changes to the original device, most recently to decrease hemolysis, the device has passed animals studies, and a randomized trial that began in 2017 is enrolling 88 patients to be compared 1 : 1 versus the EXCOR.
Other companies are looking to develop new types of MCS for children as well. VADovations is a private company that is currently testing intracorporeal CF devices in animals, focusing specifically on hemocompatibility, including preserving von Willebrand factor function to reduce nonsurgical bleeding and reducing platelet activation to limit thromboembolic events. These promising animal studies have the potential for use as a right VAD in adults and acute right-sided support for the Fontan circulation. Being only 8 mm in diameter and 50 mm long (size of AA battery), the device can easily sit inline with a blood conduit. Although these and other systems are years from widespread use in children, it reflects the increased interest and funding going toward developing long-term solutions for pediatric heart failure and CHD.
Special Populations
Bivad Versus LVAD Support
As seen in the adult VAD experience, the number of BiVADs always decreases as decision-making matures, which correlates with an increase in survival and decrease in morbidity. This is documented by multiple institutional series and by examination of the Pedimacs or INTERMACS registries. Stiller and Hetzer have written regarding their use of the Berlin Heart EXCOR VAD at the Berlin Heart Institute. Despite essentially similar patient populations over time, they have decreased the number of BiVADs, which has contributed to improved outcomes. Improved survival of LVADs over BiVADs was also demonstrated in the pediatric field by the early North American experience with the EXCOR (88% vs. 64% 6-month survival). A follow-up study of this cohort specifically attempted to identify those patient groups who would benefit from BiVAD support over LVAD support but unfortunately were unable to identify any patient cohort. However, this should not be taken to mean that there are no patients who benefit from BiVAD support because the lack of identifying a cohort was surely secondary to patient numbers. Most experts agree there are patients who benefit from BiVAD support over LVAD support, likely those with primary unremitting arrhythmia burden, certain severe biventricular restrictive disease, or certain congenital patients with biventricular failure.
Not only can improved decision making decrease the need for right VAD support, but perioperative techniques have also led to decreased need for BiVAD support. Right ventricular stress can be reduced by limiting circulating volume, reducing myocardial edema and cytokine overload with ultrafiltration, limiting bleeding and thus postoperative transfusions, and supporting the right heart aggressively in the immediate postimplant period (e.g., inhaled nitric oxide, milrinone, and/or epinephrine). Managing septal shift and minimizing tricuspid insufficiency via transesophageal echocardiography when separating from cardiopulmonary bypass and closing the sternum is important. The increasing use of intracorporeal CF devices in the pediatric population underlines how the maturation of the field has progressed from 40% of patients receiving BiVAD support to less than 20% presently. Ultimately, a small portion of patients (likely <10%) will and should be better supported with BiVADs compared with an LVAD, but this will likely be determined by etiology of heart failure as opposed to preoperative hemodynamics. More studies will need to focus on which pediatric patients are best served with BiVAD support.
Congenital Heart Disease
Long-term mechanical support of CHD presents unique clinical challenges and is associated with significantly worse outcomes than for cardiomyopathy patients. The reasons for this is likely multifactorial. Aside from anatomic and physiologic variations, many of these patients are not placed on support until after one or more failed operations and frequently in the immediate postcardiotomy period. Many of these patients also have intracardiac communications, mixed circulations, or SVP, which can make the use of VADs quite difficult. In patients with two-ventricle physiology, all intracardiac shunts should be closed at the time of LVAD implantation to avoid desaturation. In patients with systemic-to-pulmonary shunts, it is now believed best to leave the shunt open and run the VAD at higher flow rates. This has not been studied, yet it is the general feeling among most pediatric VAD programs.
Small studies prior to the North American Berlin Heart EXCOR experience had shown CHD to be a risk factor for mortality in heart failure patients on VAD support when compared with non-CHD patients. These data are similar to the European experience, which found only a 47% successful bridge for all CHD patients. In the combined EXCOR study, 59 of 204 (29%) of the patients had CHD. These patients were more likely to be mechanically ventilated, have had prior cardiac surgery, and have severe renal and hepatic dysfunction, which emphasizes the importance of end-organ function in patient selection. CHD patients were considerably less likely to successfully bridge to transplant or wean from VAD (80% vs. 48%) versus non-CHD patients. If ECMO prior to VAD was associated with congenital heart surgery, “salvage VAD” survival was 17% compared with 82% in patients requiring ECMO without prior cardiac surgery in the same admission. It is important to recognize that age also played an important role in survival, with 92% of neonates and infants dying, although the vast majority of these patients were salvage VADs compared with 60% of the children older than 1 year. This led the authors to conclude that EXCOR support may not provide any survival benefit in neonates and infants with preimplant congenital heart surgery and ECMO but that children older than 1 year with CHD can be successfully and consistently supported with the EXCOR. Again, the issue is patient selection and timing of implant. It has been documented repeatedly that a failed palliation to ECMO then to a VAD, especially in an infant, will very rarely be successful. Nonetheless, this sequence comprises most of the congenital data in infants. However, one can expect reasonable survival to transplant in a patient with CHD who is INTERMACS 2 or higher who is supported prior to end-organ dysfunction. In addition, despite being more ill (increased creatinine, hepatic dysfunction, more mechanical ventilation, etc.), congenital patients bridged to transplant with VAD support had similar posttransplant outcomes to congenital patients not bridged with VAD support. Therefore VAD support seemed to mitigate some of the usual risk factors for poor transplant outcomes.
Single Ventricle Hearts
Before entering into discussion about device outcomes in patients with single ventricle hearts (SVHs), it must be noted that we use the term SVP to describe a patient with (1) complete intracardiac mixing of the systemic and pulmonary venous returns; and (2) distribution of the CO between two parallel competing circuits (e.g., hypoplastic left heart syndrome, unrepaired aortopulmonary window, shunted patients). When possible, patients having undergone stage II (Glenn) or stage III (Fontan) repair are referred to separately. SVH refers to patients in the single ventricle pathway at any stage due to historical groupings in prior studies and simplicity for discussion; also, by the time these patients present in heart failure, they usually have significant aortopulmonary collaterals (systemic to pulmonary shunting).
The largest studies of VADs in SVH patients come from 26 patients in the North American EXCOR study and another 7 patients in a series by De Rita et al. All patients received EXCOR devices, with both studies having a survival to transplant of 42%. This was significantly lower than the 73% survival to transplant seen in the biventricular population. Weinstein et al. found that survival varied by stage. Eight out of nine patients receiving EXCOR after stage I palliation died, the only survivor being a noninfant who underwent Damus-Kaye-Stansel procedure and modified Blalock-Taussig shunt at 19 months of age. As of yet, no one has reported the successful support of a neonate after a Norwood operation to transplant and home using an EXCOR VAD. In contrast, 7 of 12 patients VAD supported after stage II and 3 of 5 patients VAD supported after Fontan operation survived to transplant. Therefore survival to transplant compared favorably to ECMO in this population, and importantly, survival after transplant was not different from patients transplanted without MCS. Patients after Glenn palliation tend to do well because they are generally patients whose circulation is failing over time as opposed to being placed on a VAD as a salvage procedure, immediately after surgery. Many times, Glenn palliated patients have been discharged home but develop end-stage heart failure prior to Fontan. There are always those patients in whom we hope their ventricular function will improve after Glenn and volume unloading, but unfortunately, some never do. In rare instances, the EXCOR can be set up in a BiVAD fashion in SVH. A recent report describes a novel four-stage transition from urgent peripheral ECMO to centrifugal VAD to EXCOR in a failing Fontan connection. This process allowed the lungs to heal following the initial central cannulation with centrifugal pump by maintaining peripheral ECMO cannulation prior to switching to full central support. Subsequently, the fenestration was closed and the oxygenator removed before converting to full EXCOR support. Most experts believe that the Fontan circulation is not well supported by peripheral ECMO, and most recommend that after initial stabilization with ECPR, central cannulation with common atrial decompression is an important step toward successful durable VAD support.
Failures of ECMO in Glenn physiology have been attributed to challenges with cannulation, high central venous pressures leading to neurologic injury, and its common use after cardiac arrest. Although Glenn physiology typically represents a “volume-unloaded” circulation compared with SVP, patients in heart failure often present with significant collateral circulation leading to a higher than expected CO demand than what would be anticipated for their BSA. Conversely, both the EXCOR and centrifugal VADs have been used successfully in Glenn physiology. This is in part because these patients tend to not be salvage VADs, rather a systemic ventricle that has failed over time.
Over the past few years, VAD support for single ventricle patients (shunted, banded, or with significant collateral flow [e.g., failing Glenn]) has progressed and there is now an appreciation that the preload of such patients is quite variable, especially during the perioperative period. As such, support with the EXCOR, whose response to preload is not dynamic and requires manual manipulation of its parameters unlike centrifugal pumps, is not ideal. Therefore a practice gaining widespread application is the use of EXCOR cannulas in combination with a centrifugal pump as aforementioned (e.g., Pedi/CentriMag, ROTAFLOW). A centrifugal VAD’s output varies with the systemic resistance encountered, and they are able to self-adjust to preload demands secondary to changes in collateral flow or Q p :Q s , which is quite variable in the immediate postimplant, inflammatory milieu. Another advantage is if oxygenation becomes an issue; it is simple to temporarily add an oxygenator to the centrifugal system. Although patients can be extubated and be somewhat mobile with this support strategy, the EXCOR system does allow for more mobilization, rehabilitation, and the ability to be discharged from the intensive care unit, which is very rarely done with a centrifugal pump. Therefore, after reaching a stable state (i.e., extubated, no inotropes, stable anticoagulation, and tolerating enteral feeds) and understanding the CO required for that particular patient, exchanging the centrifugal pump to the EXCOR pump is a reasonable practice so that the patient can be maximally rehabilitated. EXCOR cannulas with a centrifugal pump is becoming the standard of care for VAD support of those smaller patients with SVHs.
The Fontan circulation may fail for a variety of reasons but rarely due to isolated systemic systolic ventricular failure, which is why the MCS experience in the Fontan has been inconsistent. When considering mechanical support of the Fontan circulation, one must understand if the issues are right sided, left sided, or more likely a combination of both. Thus without taking an inventory of all the causes of why a particular patient’s Fontan circulation is failing, one cannot understand the dominant cause of failure and thus whether mechanical support or what type of MCS could be helpful. In addition, as with all other types of heart failure, there are different stages of Fontan circulation failure for which different therapeutic surgical and medical therapies may apply.
Fig. 66.4 demonstrates a multimodality approach that takes into account the stage and type of Fontan failure occurring. Early failure: if an atrial Fontan is developing symptoms and/or intraatrial reentrant tachycardia or if a lateral tunnel or extracardiac conduit Fontan has an anatomic obstruction, Fontan revision can successfully palliate these patients for quite some time. Late failure: if a Fontan patient has progressed to arrhythmias (e.g., atrial fibrillation) recalcitrant to treatment or begins to show early signs of renal and/or liver dysfunction, one should start considering transplantation because waiting until there is significant end-organ dysfunction will significantly decrease posttransplant outcomes. In the case where systemic ventricular failure is felt to be the main cause, a VAD can improve overall circulation. However, one must know the end-diastolic systemic ventricular pressure prior to VAD implantation because if it is less than 12 mm Hg, the likelihood that VAD support will significantly help is low. However, a pressure less than 12 mm Hg does not guarantee success because post implant an infectious pulmonary process, transfusion-related lung injury, or any significant pulmonary process can significantly affect VAD preload and right-sided congestion. Even the well-supported Fontan patient with a VAD is in a delicate balance. A low end-diastolic systemic ventricular pressure indicates that the right-sided issues are likely the dominant cause of the failing circulation. If venous congestion is to blame, a subpulmonary VAD has been described and has on rare occasions successfully supported patients to transplant. However, isolated right-sided support is not a good long-term VAD strategy for a failing Fontan circulation because forcing blood through an abnormal pulmonary vasculature into what is often a restrictive systemic ventricle will not be sustainable. If a Fontan patient awaiting heart transplant develops progressively worsening end-organ dysfunction secondary to systemic ventricular systolic dysfunction, a VAD may improve the patient’s overall condition and allow for discharge home prior to transplant. The first report of a patient successfully discharged home after a failing Fontan supported by implantable VAD was published in 2011. Implantation of a HeartMate II allowed a 15-year-old boy to return home during 72 days of support prior to transplant and represented a huge step in the care of this patient cohort with applications for destination therapy as well ( Table 66.1 ). Overall, 66% of the reported Fontan patients undergoing device implantation have had a positive outcome.
