Ventricular assist device (VAD) therapy has improved waitlist mortality for pediatric heart transplant candidates1; however, in the current era, pediatric patients with congenital heart disease remain at a greater than twofold increased risk of waitlist mortality compared with those with cardiomyopathy.1 To a degree, this disadvantage arises from the limited experience with VAD support in pediatric patients who have congenital heart disease, and particularly in patients who have single ventricle (SV) heart disease.2 Of 364 pediatric patients on VAD support enrolled in the Pediatric Interagency Registry for Mechanical Circulatory Support (Pedimacs) between 2012 and 2016, 61% had cardiomyopathy, 21% had congenital heart disease, and only 48 patients had SV heart disease.3 Survival is notably worse for pediatric patients with congenital heart disease on VAD support; among patients < 5 years old, 6 month survival was 45% for those with congenital heart disease versus 67% for those with cardiomyopathy (P = 0.008).3
The experience of VAD support in SV patients previously has been limited to case reports or small case series.4–17 Small case series are to be expected given that most pediatric VAD centers are small in volume; of 42 centers participating in Pedimacs, 35 have enrolled fewer than 16 VAD patients over 4 years, and only three centers have enrolled over 25 patients in the entire 4 year period.3 More recently, SVAD patients have been identified from registries and databases.3,18–22 Although registry data provide larger sample size and generalized survival information, they lack the granular detail provided by case series that is important in the management of a diverse and highly complex patient population.
The primary goal of this study was to describe the experience of VAD support of pediatric patients with SV heart disease at a pediatric center that has one of the higher VAD volumes, with an average of 19 pediatric VAD implants per year. We also sought to identify potential factors to target for improved patient selection and improved management of this challenging population.
All patients with SV heart disease who received VAD support at Lucile Packard Children’s Hospital at Stanford University were identified and medical records were reviewed retrospectively. SV heart disease was based on physiologic, not anatomic, diagnosis, with SV physiology defined as parallel (rather than serial) pulmonary and systemic blood flow, or if surgically palliated, passive pulmonary blood flow via a cavopulmonary anastomosis (Glenn) or Fontan palliation. Patients requiring extracorporeal membrane oxygenator support without VAD support were not included. Devices included temporary and durable systems, as well as pulsatile and continuous flow devices (Table 1). The HeartWare HVAD (Medtronic, Minnesota, USA) was the only durable intracorporeal device used during the study period in SV patients. The Berlin Heart EXCOR (Berlin Heart, Inc, Texas, USA) and Thoratec Pedimag (Thoratec, California, USA) are the only two devices currently approved by the Food and Drug Administration for use in pediatric patients; other devices included in this study are used off-label in children. This study was approved by the Institutional Review Board at Stanford University.
Anticoagulation regimens were determined by the pediatric VAD service and varied by specific devices. In general, at our institution, patients with the HeartWare HVAD are anticoagulated with heparin (target heparin activity level of 0.3–0.5 units/mL) or warfarin (target international normalized ratio of 2–3), and aspirin (target 10 mg/kg or 162 mg/dose maximum). Patients with para-corporeal continuous flow and pulsatile flow devices are anticoagulated with either heparin or enoxaparin (target heparin activity level of 0.35–0.5 units/mL or anti-Xa 0.6–1.0 units/mL) and a triple antiplatelet regimen with aspirin, dipyridamole, and clopidogrel. Details of the anticoagulation regimen routinely used at our institution for pulsatile flow devices have been previously published.23
Adverse events were defined according to Pedimacs definitions (see Appendix, Supplemental Digital Content 1, http://links.lww.com/ASAIO/A400), with neurologic dysfunction defined as an ischemic or hemorrhagic cerebrovascular accident with any new, temporary or permanent, focal or global neurologic dysfunction. All neurologic events were confirmed with computed tomography scans. Cardiac catheterization reports closest to the time of implant for patients with HeartWare HVADs, both pre- and postimplant, were reviewed. Filling pressures were recorded as direct atrial or wedge pressures. Cardiac indices were determined by thermodilution technique or were estimated by the Fick method if thermodilution was not performed. Cardiac catheterizations were not systematically performed for patients with other types of devices and thus were not collected. Preimplant systolic dysfunction and atrioventricular valve regurgitation (AVVR) were defined by qualitative reporting of moderate-to-severe ventricular dysfunction, or moderate-to-severe regurgitation, on echocardiogram closest to time of VAD implant.
Descriptive data are presented as median (range) and number (percentage). Comparisons were made using Fisher exact test for categorical variables, Wilcoxon rank sum for continuous variables, and Wilcoxon signed rank test for matched-pair variables, where appropriate. Survival to transplant was examined using Kaplan–Meier method. Statistical analysis was performed on STATA/IC 13 (College Station, Texas), and p-values < 0.05 were considered significant.
Between May 2009 and July 2017, 104 patients at our institution underwent VAD implantation. Among these, 14 were patients with SV heart disease, supported for a total of 1,112 patient-days. Preimplant characteristics of the 14 patients are summarized in Table 1 (Supplemental Digital Content 2, http://links.lww.com/ASAIO/A401), with individual patient details shown in Table 2. All patients had acute systolic heart failure at time of VAD implant and were admitted in the intensive care unit; 12 (86%) patients had moderate-to-severe systolic dysfunction by echocardiogram and two patients with low normal systolic function by echocardiogram had severe AVVR. Five patients were on VAD support with pre-Glenn physiology—two were term infants (2 weeks old) who had pulmonary artery (PA) banding procedures performed 2 weeks before VAD implant, one was 1 year old who had underwent a Glenn takedown to a modified Blalock-Taussig shunt 3 months before VAD implant (and had been discharged home for 1 month before readmission for heart failure), one was a 7-month-old who had a Glenn takedown at the time of VAD implant (he had been discharged home after his Glenn palliation), and one was 10 years old unbalanced AV canal with PA banding who was not a Glenn candidate due to diastolic and systolic heart failure. There were five patients supported with Glenn physiology—four were more than 5 years status post Glenn palliation and one was a 7-month-old who underwent Glenn palliation from a modified BT shunt at the time of VAD implant. The four patients supported with Fontan physiology were 3–19 years status post Fontan palliation; none had protein-losing enteropathy or plastic bronchitis.
Device and Surgical Strategies
Over the years, device selection and surgical strategies evolved, depending on patient factors and underlying physiology. For smaller patients with Glenn physiology, our earliest experiences were with pulsatile devices; however, after unsuccessfully supporting two Glenn patients with the Berlin Heart EXCOR (patients 2 and 3), we shifted our subsequent approach to Glenn takedowns with continuous flow devices (patients 4 and 5). Both these patients achieved adequate hemodynamic support; one was successfully supported to transplant after 67 days but the other died on POD 42 from neurologic complications. In the neonates, we initially tried a strategy with the Berlin Heart EXCOR (patient 10), but as the patient was not successfully supported, a continuous flow device was used for patient 11, who was successfully supported for 137 days to transplant. The larger patients (patients 6–9 and 12–14) allowed for use of the HeartWare HVAD. Our first SV HVAD was implanted in the ventricle (patient 6), but we subsequently switched to atrial implantation, as this allowed for less surgical dissection, shorter time on pump, and improved decompression of the heart, especially in the setting of AVVR. Ventricular implants were performed in patients with anatomic challenges to atrial implant (patients 12 and 13).
Survival to Transplant
Over a median duration of 68 (3–219) days of VAD support, eight (57%) patients survived to heart transplant. The Kaplan–Meier survival to transplant curves is shown in Figures 1 and 2. Preimplant comparisons between those who died and were transplanted are shown in Table 2 (Supplemental Digital Content 3, http://links.lww.com/ASAIO/A402). The six patients who died were smaller (10.1 kg vs. 28.3 kg, P = 0.04). While nonsignificant, the patients who died were more likely to have a Glenn palliation (80% of Glenn patients died vs. 0% of Fontan patients, Figure 2), were implanted in an earlier era (60% died in 2009–2013 vs. 33% in 2014–2017), and were sicker at time of implant (60% of those who died were Intermacs 1 vs. 40% of those who survived). Postimplant outcomes are shown in Table 3. Patients who died were supported for a relatively brief interval of time, with a median of 16 (3–89) days support, compared with a median of 107 (35–219) days of support for those who survived to transplant (P = 0.01).
Cardiac catheterization data for the seven patients with HeartWare HVADs are shown in Table 2 (Supplemental Digital Content 3, http://links.lww.com/ASAIO/A402)and F igures 3 and 4. Post-VAD catheterizations were performed at a median 14 (2–52) days postimplant, varying depending upon patient status. Postimplant, cardiac index increased significantly in all patients, from a median 2.6 (1.0–3.8 L/min/m2) to 3.8 (3.1–6.7) L/min/m2 (Figure 3, P = 0.03). Filling pressures decreased postimplant for five of the seven patients (Figure 4), although overall, filling pressure did not change significantly (median 11 (7–21) mm Hg preimplant to 12 (4–13) mm Hg postimplant, P = 0.11). Neither cardiac index nor filling pressures, pre- and postimplant, were associated with survival to transplant. Patients with Glenn palliation had the highest filling pressures (12 mm Hg vs. 7 mm Hg, P = 0.07) along with the highest cardiac indices (5.5 L/min/m2vs. 3.5 L/min/m2, P = 0.03) on VAD support, reflecting attempts to decrease filling pressure by increasing cardiac output with increases in VAD speed settings. Two patients with Glenn palliation developed moderate-to-severe aortic regurgitation postimplant, whereas the other patients did not.
Medical Care Deescalation and Hospital Discharge
The eight patients who survived to transplant were all extubated, at a median 4 (1–23) days, and five were transferred out of the intensive care unit, at a median 20 (14–27) days. In comparison, only two of the six patients who died were extubated, at a median 32 (31–33) days (P = 0.04), and none left the intensive care unit. There were five patients with HeartWare HVADs who were potential hospital discharge candidates. The first HeartWare HVAD patient (patient 6) was stable for hospital discharge, participating fully in rehabilitation and attending the hospital school, but as the first SV patient with a HeartWare HVAD at our institution, we decided to keep her inpatient until her transplant on postoperative day 76. Two subsequent patients were discharged from the hospital; one was a patient with a central shunt and PA band (patient 9) discharged on postoperative days 61 and transplanted on postoperative day 219, and another was a Glenn patient (patient 13) discharged on postoperative day 72 and transplanted on postoperative day 174. One patient (patient 14) underwent heart transplant on the day of planned discharge, on postoperative day 69. One additional patient (patient 7) was expected to be discharged, but remained in the hospital because of on-going concerns regarding the family’s ability to care for the patient, waiting on the floor unit and participating fully in rehabilitation and hospital school until transplant on postoperative day 165. After transplant, patients were discharged at a median 21 (range 12–196) days posttransplant.
Eleven patients (79%) experienced at least one adverse event. The most common adverse event was infection, occurring in 50% of patients (n = 7), followed by neurologic adverse event in 36% (n = 5) and bleeding in 29% (n = 4). There were no differences in adverse event incidence between those who died and those who survived to transplant. The five neurologic events occurred at a median 29 (range 5–63 days) post VAD implant. There were four subdural hematomas; one occurred in the setting of supra-therapeutic anticoagulation and one required neurosurgical intervention. All four patients with subdural hematomas survived to transplant without neurologic deficits. One patient had an ischemic stroke with multiple foci; support was withdrawn because of poor neurologic prognosis and the patient died. Four patients were being treated actively for infections at the time of death, including one patient with confirmed Candida albicans infection. Causes of death are described in Table 2.
We present the results of VAD support in SV heart disease at a single pediatric center. Several themes emerge from this early experience and the lessons we have learned are summarized in Table 4.
First, and most basic, this experience affirms that VAD support can be effectively provided to patients with SV heart disease in a variety of anatomic configurations and at different stages of surgical palliation. This is consistent with, and supportive of, other single-center and multicenter studies published in the past few years.4–22 Also consistent with other reports is the observation that results from VAD support in SV heart disease lag far behind those obtained for patients with dilated cardiomyopathy,1,3,20–22 but may be similar to the results that can be expected with support of restrictive cardiomyopathy or hypertrophic cardiomyopathy, situations where support has been particularly challenging.24–28 The data are insufficient for a formal comparison of outcomes in these conditions; it may only be stated that for dilated cardiomyopathy, successful support of a suitable candidate should be expected, whereas for the other conditions, it is not so certain. This observation may have relevance in transplant listing prioritization schemes, which currently do not distinguish between different anatomic substrates for VAD support as bridge to transplantation.
The second conclusion that can be drawn from this report is that hemodynamic stability and full deescalation of medical care is feasible in VAD support of some patients with SV heart disease. In this regard, we demonstrate these results for patients implanted with a durable intracorporeal continuous flow devices. This observation is important particularly in the early reporting of VAD outcomes, as it is sometimes difficult to distinguish successful bridge to transplantation from successful VAD support (because a transplant donor offer can sometimes arrive in a timely manner). The hemodynamic results and the level of stability achieved for the patients supported with the HeartWare HVAD give a measure of assurance that longer-term support would also be feasible and that hospital discharge is a realistic objective for such patients. Hospital discharge is an important goal for these patients because VAD support ultimately can, and should, be used not only as a rescue therapy in bridging to transplant, but as a therapy for making patients better transplant candidates. To do so requires longer support durations, ideally out of the hospital, to allow for nutritional, physical, and psychologic rehabilitation. According to Pedimacs, over half of eligible pediatric VAD patients are discharged from the hospital,3 and some centers report discharge rates up to 85%.29 There are pediatric institutions that routinely delay waitlist activation, up to 3 months postimplant, to specifically allow more time for such rehabilitation30; Although this has not been standard practice at our institution, it is an objective worth considering for SVAD patients.
In evaluating the hemodynamic results, it is noteworthy that the relief of congestion is more variable than the improvement in cardiac index. Filling pressures did not change significantly from pre- to postimplant, whereas the cardiac index changed significantly from 2.6 L/min/m2 to 3.8 L/min/m2. It is likely that this is a result of our management practice, in which diuretics are used extensively to control congestion before VAD implantation, while VAD output is increased postimplant for this same objective. It is also interesting to note that the cardiac index achieved with support is quite high, up to 6.7 L/min/m2 in patients with Glenn physiology. The experience reported here is insufficient to make a definitive statement in this regard, but it should be considered possible that SV patients, particularly those with Glenn physiology, may require high flow rates for adequate support, as compared with flows used for dilated cardiomyopathy patients on VAD support, and that device selection may need to provide unusually high rates for successful hemodynamic stabilization.
The mortality that was observed in this report appears to represent two overlapping issues. For some patients, hemodynamic support was not adequate. In the very early experience using para-corporeal pulsatile devices, this likely resulted from a mismatch between the patients’ circulatory requirements and the device capability. This is supported by an early report of successful VAD support in SV heart disease using oversized cannulae and devices15 and is further supported by modeling studies that demonstrate the superior flow characteristics of continuous flow compared with pulsatile flow systems, for a given cannula configuration.7,31,32 In addition to the patients with inadequate hemodynamic support, some portion of mortality was related to early multisystem organ dysfunction that suggests an important role of patient selection. Both of these issues appear amenable to a learning curve, particularly the matter of patient selection. Adequate hemodynamic support is largely a matter of matching the patient to the appropriate device capability, and of optimizing the inflow cannula placement. Oversizing of devices (as compared with the expectation for a comparable dilated cardiomyopathy patient) is now a routine component of our preimplantation planning, including the use of preimplantation CT scanning to determine probable device fit. We do not use a formal process for device fit-testing, as has been reported by others,33 but acknowledge that it might be a useful adjunct to preimplant planning. We believe baseline cardiac catheterization has also been helpful in understanding the hemodynamic requirements. Additionally, this affords the opportunity to minimize collateral vessels that might complicate implantation or VAD support. In our institution, we have largely shifted to atrial implantation of the inflow cannula,34 which has allowed for good flow rates. This is consistent with reports of atrial implantation from other centers,16 but the ideal implantation site is not yet known and may be approached in different manners at other centers.
Mortality was particular pronounced in our Glenn patients. The reason for poor outcomes with Glenn physiology is likely multifactorial. For one, VAD support in the Glenn physiology only directly decompresses the systemic venous system of the lower body. The pressure differential between the upper and lower venous system, unique in the setting of a Glenn, may be further exaggerated while on VAD support, leading to further hypoxia with development of veno-venous collaterals. Chronic severe hypoxia is problematic in VAD support given the impact on end-organ function, polycythemia, and anticoagulation. To overcome the issues associated with VAD support of the Glenn physiology, we have shifted to conversion of the Glenn anastomosis to shunt physiology at the time of VAD implant. For various reasons, including concern for elevated pulmonary pressures and to minimize surgical dissection, we considered but did not perform Fontan completions at the time of VAD implant for several of the older Glenn patients. However, other centers have reported success with such a strategy35 and moving forward, we would strongly consider Fontan completion at the time of VAD implant for Glenn patients.
The matter of patient selection remains challenging, especially the optimization of timing. In a situation where current results from VAD support are not optimal, and devices are being used as bridge to transplant, there is always the hope that a good donor might be obtained with a short additional period of delay. This hope often works in opposition to the principles of good patient selection for VAD implantation, and this tension cannot be easily or simply resolved. Frequent reassessment of medical status and VAD candidacy is necessary to address this issue, combining input from the heart failure service, the cardiovascular intensive care service, and the VAD surgeon.
The experience that we report does not permit a formal analysis of factors associated with good outcomes. In a qualitative manner, we would note that results appear to be improving in the more recent era that Fontan patients are among the more successful populations for VAD support and that use of continuous flow devices can result in improved hemodynamics. On the other hand, VAD support of the Glenn circulation remains very problematic. This is likely a matter of patient selection and timing, given the hemodynamics that can be achieved in the Fontan circulation but may also reflect the more limited device availability in smaller patients. Although our cardiac catheterization data demonstrate adequate hemodynamic support with the HeartWare HVAD, it remains unclear if similar support can be achieved with pulsatile devices, particularly in smaller patients. Because there continues to be reports of successful outcomes with pulsatile devices,22 we continue to consider the Berlin Heart EXCOR for our smaller patients.
As a single-center series, patients we present here may not be representative of those seen and treated elsewhere. For example, we have moved toward atrial cannulation for SV patients at our institution, a practice that may differ from other institutional approaches. Also as a single-center series, our study had limited power for statistical comparisons, and some of our conclusions are drawn from qualitative data review. Larger, multicenter reports on VAD support in SV patients will certainly be helpful, but such studies may lack the granularity of data from single-center reports. Details pertaining to patient selection and timing of device implant, critical to successful VAD outcomes, can often be better assessed from detailed reports from single-center experiences. Another important limitation to our study is that 86% of the patients we present, and in particular all the Fontan patients included in our series, had significant systolic dysfunction. However, various phenotypes of Fontan failure have been described,34 and we are unable to comment on outcome of VAD support in patients with other presentations of failing single physiology, such as in patients with protein-losing enteropathy and preserved systolic function. It is likely that this particular population of SV patients—those with preserved systolic function—will remain one of the more challenging populations to successfully bridge with VAD support to transplant.
1. Zafar F, Castleberry C, Khan MS, et al. Pediatric heart transplant
waiting list mortality in the era of ventricular assist devices. J Heart Lung Transplant 2015.34: 82–88.
2. Carlo WF, Villa CR, Lal AK, Morales DL. Ventricular assist device use in single ventricle congenital heart disease
. Pediatr Transplant 2017;21(7)
3. Blume ED, VanderPluym C, Lorts A, et al. Second annual Pediatric Interagency Registry for Mechanical Circulatory Support (Pedimacs) report: Pre-implant characteristics and outcomes. J Heart Lung Transplant 2018.37: 38–45.
4. VanderPluym CJ, Rebeyka IM, Ross DB, Buchholz H. The use of ventricular assist devices in pediatric patients with univentricular hearts. J Thorac Cardiovasc Surg 2011.141: 588–590.
5. Mackling T, Shah T, Dimas V, et al. Management of single-ventricle patients with Berlin Heart EXCOR Ventricular Assist Device: single-center experience. Artif Organs 2012.36: 555–559.
6. Brancaccio G, Gandolfo F, Carotti A, Amodeo A. Ventricular assist device in univentricular heart physiology. Interact Cardiovasc Thorac Surg 2013.16: 568–569.
7. Lal AK, Chen S, Maeda K, et al. Successful bridge to transplant with a continuous flow ventricular assist device in a single ventricle
patient with an aortopulmonary shunt. ASAIO J 2014.60: 119–121.
8. Valeske K, Yerebakan C, Mueller M, Akintuerk H. Urgent implantation of the Berlin Heart Excor biventricular assist device as a total artificial heart in a patient with single ventricle
circulation. J Thorac Cardiovasc Surg 2014.147: 1712–1714.
9. Weinstein S, Bello R, Pizarro C, et al. The use of the Berlin Heart EXCOR in patients with functional single ventricle
. J Thorac Cardiovasc Surg 2014.147: 697–704.
10. Halaweish I, Ohye RG, Si MS. Berlin heart ventricular assist device as a long-term bridge to transplantation in a Fontan patient with failing single ventricle
. Pediatr Transplant 2015.19: E193–E195.
11. Niebler RA, Shah TK, Mitchell ME, et al. Ventricular assist device in single-ventricle heart disease and a superior cavopulmonary anastomosis. Artif Organs 2016.40: 180–184.
12. Arnaoutakis GJ, Blitzer D, Fuller S, et al. Mechanical circulatory support as bridge to transplantation for the failing single ventricle
. Ann Thorac Surg 2017.103: 193–197.
13. Woods RK, Ghanayem NS, Mitchell ME, Kindel S, Niebler RA. Mechanical circulatory support of the Fontan patient. Semin Thorac Cardiovasc Surg 2017.20: 20–27.
14. Hoganson DM, Boston US, Gazit AZ, Canter CE, Eghtesady P. Successful bridge through transplantation with berlin heart ventricular assist device in a child with failing Fontan. Ann Thorac Surg 2015.99: 707–709.
15. Irving CA, Cassidy JV, Kirk RC, Griselli M, Hasan A, Crossland DS. Successful bridge to transplant with the Berlin heart after cavopulmonary shunt. J Heart Lung Transplant 2009.28: 399–401.
16. Gazit AZ, Petrucci O, Manning P, et al. A novel surgical approach to mechanical circulatory support in univentricular infants. Ann Thorac Surg 2017.104: 1630–1636.
17. Boston U, Sun J, Kumar TKS, Knott-Craig C. An innovative ventricular assist device strategy as a bridge-to-recovery in an infant with Glenn physiology [published online ahead of print September 28, 2018]. ASAIO J 2018.
18. Poh CL, Chiletti R, Zannino D, et al. Ventricular assist device support in patients with single ventricles: the Melbourne experience. Interact Cardiovasc Thorac Surg 2017.25: 310–316.
19. VanderPluym CJ, Cedars A, Eghtesady P, et al. Outcomes following implantation of mechanical circulatory support in adults with congenital heart disease
: An analysis of the Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS). J Heart Lung Transplant 2018.37: 89–99.
20. Lorts A, Eghtesady P, Mehegan M, et al. Outcomes of children supported with devices labeled as “temporary” or short term: A report from the Pediatric Interagency Registry for Mechanical Circulatory Support. J Heart Lung Transplant 2018.37: 54–60.
21. Dipchand AI, Kirk R, Naftel DC, et al. Ventricular assist device support as a bridge to transplantation in pediatric patients. J Am Coll Cardiol 2018.72: 402–415.
22. Dykes JC, Bleiweis MS, Maeda K, et al. Berlin heart outcomes in single ventricle
patients: where are we now? J Heart Lung Transplant 2018.37: S102.
23. Rosenthal DN, Lancaster CA, McElhinney DB, et al. Impact of a modified anti-thrombotic guideline on stroke in children supported with a pediatric ventricular assist device. J Heart Lung Transplant 2017.36: 1250–1257.
24. Dykes JC, Reinhartz O, Almond CS, et al. Alternative strategy for biventricular assist device in an infant with hypertrophic cardiomyopathy. Ann Thorac Surg 2017.104: e185–e186.
25. Patel SR, Saeed O, Naftel D, et al. Outcomes of restrictive and hypertrophic cardiomyopathies after LVAD: an INTERMACS analysis. J Card Fail 2017.23: 859–867.
26. Su JA, Menteer J. Outcomes of Berlin heart EXCOR pediatric ventricular assist device support in patients with restrictive and hypertrophic cardiomyopathy. Pediatr Transplant 2017.21: e13048.
27. Topilsky Y, Pereira NL, Shah DK, et al. Left ventricular assist device therapy in patients with restrictive and hypertrophic cardiomyopathy. Circ Heart Fail 2011.4: 266–275.
28. Wynne E, Bergin JD, Ailawadi G, Kern JA, Kennedy JL. Use of a left ventricular assist device in hypertrophic cardiomyopathy. J Card Surg 2011.26: 663–665.
29. Fraser CD Jr, Chacon-Portillo MA, Zea-Vera R, et al. Ventricular assist device support: single pediatric institution experience over two decades [published online ahead of print 2018]. Ann Thorac Surg 2018.107: 829–836.
30. Adachi I. Continuous-flow ventricular assist device support in children: A paradigm change. J Thorac Cardiovasc Surg 2017.154: 1358–1361.
31. Fujii Y, Ferro G, Kagawa H, et al. Is Continuous flow superior to pulsatile flow in single ventricle mechanical support
? Results from a large animal pilot study. ASAIO J 2015.61: 443–447.
32. Schmidt T, Rosenthal D, Reinhartz O, et al.; Modeling of Congenital Hearts Alliance (MOCHA)+ Investigators: Superior performance of continuous over pulsatile flow ventricular assist devices in the single ventricle
circulation: a computational study. J Biomech 2017.52: 48–54.
33. Moore RA, D’Souza GA, Villa C, Taylor MD, Morales DLS, Lorts A. Optimizing surgical placement of the HeartWare ventricular assist device in children and adolescents by virtual implantation. Progress in Pediatric Cardiology 2017.47: 11–13.
34. Ma M, Yarlagadda VV, Rosenthal DN, Maeda K. A novel inflow cannulation strategy for pediatric mechanical circulatory support in small left ventricles. J Thorac Cardiovasc Surg 2017.154: e47–e48.
35. Adachi I, Williams E, Jeewa A, Elias B, McKenzie ED. Mechanically assisted Fontan completion: a new approach for the failing Glenn circulation due to isolated ventricular dysfunction. J Heart Lung Transplant 2016.35: 1380–1381.
36. Book WM, Gerardin J, Saraf A, Marie Valente A, Rodriguez F 3rd. Clinical phenotypes of Fontan failure: implications for management. Congenit Heart Dis 2016.11: 296–308.