Secondary Logo

Journal Logo


Current Clinical Status of Pulsatile Pediatric Circulatory Support

Reinhartz, Olaf*; Stiller, Brigitte; Eilers, Rolf; Farrar, David J.§

Author Information
  • Free


Mechanical circulatory support in children is being performed for reasons similar to those in adults: to bridge end-stage congestive heart failure patients to transplantation and to support patients in acute reversible cardiogenic shock or with failure to wean off bypass. However, in the United States, the current options for smaller children are limited, as there is no specific pulsatile pediatric size device approved for clinical use. Adult size paracorporeal devices, predominantly the Thoratec ventricular assist device (VAD) (Pleasanton, CA), despite the size discrepancy, are being implanted into intermediate size children for lack of better fitting pumps. Despite generally good results, this appears to be associated with increased risk of thromboembolic complications, especially when atrial cannulation is used. 1 For very small patients, centrifugal pumps or extracorporeal membrane oxygenation (ECMO) remain the only options in this country, both of which are poorly suitable for longer term support.

In Europe, a few centers have gained considerable experience with two pneumatic paracorporeal systems (Berlin Heart; Medos) that feature a variety of pump and cannulae sizes down to neonatal dimensions. However, the early clinical impression had been that morbidity and mortality in neonates and infants was increased over that of older children.

This article analyzes the existing clinical data with both adult and pediatric pulsatile VADs in children in the United States and Europe and outlines the specific risks associated with either therapy.


Thoratec Ventricular Assist Device

The Thoratec VAD is a pneumatically driven paracorporeal pump with mechanical valves, with a maximum stroke volume of 65 ml. In 57 centers worldwide, between November 1982 and September 2001, according to the voluntary company registry, it was implanted in a total of 101 children and adolescents under 18 years of age (64 boys, 37 girls). The registry data used for this analysis are from 90% to 100% complete; outcome data are 100% complete. Mean patient age was 13.7 years (range: 7–17 years), mean patient weight was 54.4 kg (range: 17–110 kg), mean patient body surface area (BSA) was 1.53 m2 (range: 0.7–2.2 m2), and median duration of support was 25 days (range: 0–434 days). Extensive retrospective data were provided by the implanting centers for a subgroup of the first 58 patients, which allowed for a more extensive analysis. 1

In the majority of cases, the device was used in the volume (“full to empty”) mode throughout most of the support. However, in eight children, the fixed rate mode with rates between 40 and 70 beats per minute was used, which allows for delivery of a smaller stroke volume.

For anticoagulation on the device, heparin was used during the initial postoperative period, with a goal pulmonary transit time (PTT) of 60 to 80 seconds. With mobilization of the patient, this was generally switched over to coumadin to achieve an international normalized ratio (INR) of 2.5 to 3.5. This management was no different from the standard protocol in adults using this device.

Berlin Heart Ventricular Assist Device

The Berlin Heart assist device (Berlin, Germany) (Figure 1) is also a pneumatically driven paracorporeal pump. Blood pumps are available in sizes of 10, 25, 30, 50, 60, and 80 ml. The 50 to 80 ml pumps are equipped with tilting disk valves, whereas the miniature pumps (10, 25, and 30 ml) use specially designed trileaflet polyurethane valves. There is a variety of cannula sizes and configurations, 2 including left ventricular apical cannulae for all ages. Between 1990 and September 2001, 45 patients between 2 days and 16 years of age (mean 7.6 years) have been supported with this device in one single center alone (Deutsches Herzzentrum Berlin), with another 40 patients in 15 other hospitals in 10 countries worldwide. Only data on the Deutsche Herzzentrum patients were sufficient to be used for statistical analysis and are 100% complete. Patient weight ranged from 2.2 to 81 kg (mean 25.3 kg), and duration of support ranged from 1 to 111 days (mean 19 days; median 10 days).

Figure 1
Figure 1:
Three year old patient (weighing 11 kg) using a 25 ml Berlin Heart ventricular assist device, supported for 27 days.

With exception of the first 17, all of these pumps were heparin coated (Carmeda, Stockholm, Sweden). Anticoagulation was achieved with IV heparin throughout support to maintain activated clotting time (ACT) levels between 140 and 160 seconds or a PTT between 60 and 80 seconds. Antithrombin III (heparin cofactor) was replaced when the serum level was below 70% of normal. In some patients, aspirin (1–3 mg/kg/day) was added.

Medos Ventricular Assist Device

The Medos assist device (Stolberg, Germany) is also paracorporeal and pneumatically driven; however, it uses trileaflet polyurethane valves. It is available with stroke volumes of 10, 25, 60, and 80 ml for left ventricular support, with 10% reduced volumes for right sided support, and in several different cannula sizes. 3 The data presented in this paper are provided by Medos and stem from the voluntary company registry. Outcome data are 92% complete; data on demographics and indications are between 77% and 100% complete. In several European centers, the device has been implanted in 64 children between 3 days and 16 years of age. Duration of support was 1 to 97 days (median 5.5 days).

For anticoagulation, heparin was used throughout support to achieve ACT levels between 180 and 220 seconds. In patients supported for longer periods of time, aspirin was added at 5 mg/kg/day.

Patient Groups

Because of the diversity of patient sizes and indications for support, we divided the children and adolescents into four size groups. Group S contained neonates and infants with a body weight of less than 10 kg. Children in group M were toddlers and young schoolchildren with weights ranging from 10 to 20 kg. Patients in these two groups were supported only by designated pediatric size pumps, with the exception of the smallest patient on a Thoratec device, who weighed 17 kg. Group L comprised older children and adolescents weighing between 20 and 50 kg, and group XL included children and adolescents of almost adult body size (weighing more than 50 kg).

In regards to indications for support, we have divided the study population into four major groups: cardiomyopathies, acute myocarditis, chronic heart failure caused by congenital disease, and failure to wean off bypass after repair of congenital defects.

Statistical Analysis

Outcome differences among groups were evaluated using the Fisher exact test for 2 × 2 contingency tables and χ2 analysis for more complex circumstances, and p < 0.05 was considered significant. In a subgroup of 58 patients with Thoratec VADs, the Cox proportional hazard model was used to assess predictors. 1


Thoratec Ventricular Assist Device

Overall Outcome.

Five of the 101 patients supported with Thoratec devices were still supported by the device at the time of this analysis. Of the remaining 96 patients, 66 survived to either transplantation (56) or recovery of the native heart (10), respectively, which resulted in an overall survival rate of 68.8%. Survival to discharge was 65.6%, with one patient dying in hospital posttransplantation and two died postrecovery. Posttransplantation survival, therefore, was 98.2% of the patients.

Size and Survival.

Figure 2 shows the numbers of survivors and nonsurvivors in each of the size groups. The survival rate in group L was 68.3% and not significantly different from group XL (67.9%). A previously published analysis for the first 58 patients did not show age, weight, height, or BSA to be a significant risk factor for survival. 1

Figure 2
Figure 2:
Survival of patients using Thoratec ventricular assist devices, separated by body weight. Group S, less than 10 kg; M, 10–20 kg; L, 20–50 kg; XL, greater than 50 kg.

Indication for Support and Survival.

In Figure 3, the outcomes for different groups of indications is shown. The survival rate for patients supported for congenital defects, both for chronic heart failure and postcardiotomy, was significantly worse than for patients with cardiomyopathies or myocarditis.

Figure 3
Figure 3:
Patients using Thoratec ventricular assist devices, grouped by indication for support and survival. CMP, cardiomyopathies; myoc., myocarditis; cong. CHF, chronic heart failure caused by congenital heart disease; cong. postc., congenital heart disease postcardiotomy.

Neurologic Events.

Data on neurologic complications are only available for the first 58 patients and have been previously published. 1 Eighteen events occurred in 15 of these patients (27%): 3 transient ischemic attacks, 10 thromboembolic cardiovascular accidents (CVAs) (1 fatal), and 5 intracerebral hemorrhages (all fatal). The risk for embolic stroke was significantly lower in patients cannulated using the left ventricle (5 of 38 patients; 13%) as opposed to the left atrium (5 of 17 patients; 29%).

Berlin Heart Ventricular Assist Device

Overall Outcome.

Of the 45 patients supported at the German Heart Center in Berlin, 17 survived to transplantation and 5 to recovery of the native heart, resulting in an overall survival rate of 48.9%. 17 of these 22 survivors were discharged (posttransplantation/weaning survival 77.3%).

Size and Survival.

The relationship between survival rate and patient size is illustrated in Figure 4. Survival rates for the four groups S, M, L, and XL were 21.4%, 71.4%, 55.5% and 66.7%, respectively. The survival rate for group S was significantly lower than in all other groups (group S versus groups M, L, and XL). However, after excluding postcardiotomy patients, this difference lost statistical significance.

Figure 4
Figure 4:
Survival of patients using Berlin Heart ventricular assist devices, separated by body weight. Group S, less than 10 kg; M, 10–20 kg; L, 20–50 kg; XL, greater than 50 kg.

Indication for Support and Survival.

Figure 5 shows the outcomes for different groups of indications for support. The survival rate for patients supported for congenital defects, chronic heart failure, and postcardiotomy combined was 21.4%, which was significantly lower than for patients with cardiomyopathies and myocarditis (66.7% and 71.4%, respectively).

Figure 5
Figure 5:
Patients using Berlin Heart ventricular assist devices, grouped by indication for support and survival to transplantation (transplant) or recovery of the native heart (weaned). CMP, cardiomyopathies; myoc., myocarditis; cong. CHF, chronic heart failure caused by congenital heart disease; cong. postc., congenital heart disease postcardiotomy.

Neurologic Events.

Five cerebral events in patients on support (three thromboembolic CVAs and two fatal intracerebral hemorrhages) account for a neurologic complication rate of 11.1% overall. Two of the three strokes occurred in patients on pumps that had not yet been coated with Carmeda.

Myocarditis Subgroup.

Table 1 focuses on the subgroup of patients who were being supported for myocarditis. Four of seven patients were weaned from the device after recovery of their native hearts and were discharged home; they are alive after up to 7 years.

Table 1
Table 1:
Characteristics and Outcome of Myocarditis Patients Supported with Berlin Heart Devices

Medos Ventricular Assist Device

Overall Outcome.

Of the 56 patients with known outcome who were being supported with Medos devices, 11 survived to transplantation and 10 could be weaned off the device because of recovery of the native heart, resulting in an overall survival rate of 37.5%.

Size and Survival.

Patient weight information was only available in a minority of these patients; however, pump size was known in 95%. Survival by pump size is plotted in Figure 6 and was not significantly different among the groups (9/10 ml, 37.5%; 22/25 ml, 34.8%; 54/60 ml, 37.5%).

Figure 6
Figure 6:
Survival of patients using Medos ventricular assist devices, grouped by pump size (ml).

Indication for Support and Survival.

This analysis had to be limited to patients with sufficient data. Ten of 16 patients (62.5%) with cardiomyopathies survived through support, whereas only 6 of 17 patients (35.3%) postcardiotomy survived, showing a trend towards a better survival rate in the former group (p = 0.08).


Experience with pulsatile circulatory support in children is increasing and clinical results are improving, even though progress in this field is still behind that of adult support. This may be caused by having a smaller number of patients in need, and therefore less incentive for investigators and industry to market devices. Despite earlier attempts to scale down adult blood pumps, 4 there is still no pediatric device available in the United States. However, the manufacturer of the Berlin Heart is currently applying for approval by the Food and Drug Administration. For now, we will be able to learn from the extensive European experience included in this report. The data presented here represent an almost complete update on pulsatile support in the pediatric population. Only a limited number of patients are not included: some Thoratec Heartmate and Novacor devices have been implanted in adolescents, one series of which has recently been published. 5 Up to May 2001, 44 patients aged 17 or younger have been reported to the Heartmate voluntary company database (Thoratec Corporation, unpublished data, 2001). However, these fully implantable devices are generally only suitable for patients with BSAs greater than 1.5 m2 (equivalent to group XL in this article). There are only anecdotal data on use of the Abiomed biventricular support 2000 VAD in children. 6,7 Finally, in Japan, a small, rather unsuccessful series of implantations of the pediatric version of the Toyobo pump was reported several years ago. 8

We will not comment on the differences in survival rates or other outcome data among the three devices. These patient groups vary significantly in patient size and indication for support, and they may have further differences in other variables that we did not assess. They have been treated in very different centers, and the data acquisition was different. Both the Thoratec and Medos data were generated in a variety of centers with very different experience levels, whereas the Berlin Heart data stem from a single institution (data on 40 patients supported at other centers were not available for analysis). We therefore believe that the data presented here are not sufficient to make comparisons among these devices. However, the fact that certain observations appear similar in all groups allows for some conclusions to be made with sufficient certainty.

The outcome on pulsatile support is similar for patients of all body sizes supported with the Thoratec device. It is also similar for groups M to XL supported with the Berlin Heart VAD, although lower for neonates and infants (group S). This difference may be caused by a variety of factors. It may be confounded by the fact that group S consists of relatively more postcardiotomy patients than the other groups. Indication for support, particularly in the smaller children and in the earlier experience, may have often been determined too late in the disease process, when end-organ failure was already irreversible. A number of early patients may have been treated with ECMO as the more established device, rather than with a pulsatile VAD, only because of their size. In addition, technical improvements, such as apical cannulation and specially optimized small cannulae for infants, only became available in recent years. In the patients who were supported earlier, cannulae placement may have often been suboptimal, and the chest frequently had to be left unapproximated because of the bulkiness of the hardware.

Indication for support could be determined as a major factor in outcome: For all devices, support for cardiomyopathies and myocarditis was associated with much higher survival rates than those seen for congenital defects and postcardiotomy situations. Patients with cardiomyopathies, regardless of size, are likely to be supported for longer durations. They are therefore also more likely to benefit from the advantages of pulsatile devices, such as fewer severe infectious complications, less bleeding complications, better mobilization, and the chance for extubation. A reliable neurologic examination before listing to transplant is important for parents and children, especially in children who have undergone previous resuscitation.

Myocarditis also appears to be an excellent indication for a pulsatile device, in that it promotes better ventricular unloading and therefore probably increases the chance for myocardial recovery. The Berlin group has an outstanding series of patients with this condition that were weaned from the device after recovery of their native hearts 9 (Table 1).

On the other hand, ECMO or centrifugal pumps remain the methods of choice for postcardiotomy patients that cannot be weaned from extracorporeal circulation but have a potential for early myocardial recovery. Survival for these patients on the devices reported here was between 13% and 47%, which is not superior to survival on nonpulsatile devices. 10–12 Obviously ECMO is the only option if intracardiac shunts persist or respiratory failure complicates the situation.

There is increasing evidence for an advantage to left ventricular apical cannulation in pulsatile support in children and adolescents. The risk for neurologic events in children and adolescents on the Thoratec is elevated with left atrial cannulation (29%), although it is fairly low in ventricularly cannulated patients (13%), and this is independent of body size. The risk of stroke on the Berlin heart is also low (11.1%), despite recent introduction of ventricular cannulation in the Berlin heart (the latest four patients were ventricularly cannulated). This difference may be caused by a number of possible factors: multiple unknown patient variables that this study does not control for, differences in anticoagulation, or differences in pump properties. The introduction of Carmeda coating in the Berlin heart might have further decreased the incidence of embolic neurologic events. Two of the three strokes occurred in bigger patients with 50 to 60 ml pumps, which also makes small patient size unlikely to be a risk factor. Other advantages of ventricular cannulation are higher VAD flows and more efficient unloading of the heart, which appears to be associated with increased potential for myocardial recovery. For these reasons, ventricular cannulation can be recommended for the majority of children supported with any pulsatile device. Apical cannulae are available for the Thoratec VAD and, in a variety of sizes, for the Berlin Heart. Medos is planning to release apical cannulae for its device in the near future.

In conclusion, pulsatile support of the failing heart in children can be performed with overall similar morbidity and mortality as in adults. In neonates and infants, however, the survival rate is still somewhat lower than for older children. Implantation for cardiomyopathies or myocarditis is associated with much better survival rates, compared with implantation for congenital defects and postcardiotomy. Introduction of a specifically designed pediatric pulsatile pump in the United States could improve outcome, especially for smaller children with cardiomyopathies or myocarditis.


1. Reinhartz O, Keith FM, El-Banayosy A, et al: Multicenter experience with the Thoratec ventricular assist device in children and adolescents. J Heart Lung Transplant 20: 439–448, 2001.
2. Hetzer R, Loebe M, Potapov EV, et al: Circulatory support with pneumatic paracorporeal ventricular assist device in infants and children. Ann Thorac Surg 66: 1498–1506, 1998.
3. Herwig V, Severin M, Waldenberger FR, Konertz W: Medos/HIA-assist system: First experiences with mechanical circulatory assist in infants and children. Int J Artif Organs 20: 692–694, 1997.
4. Daily BB, Pettitt TW, Sutera SP, Pierce WS: Pierce-Donachy pediatric VAD: progress in development. Ann Thorac Surg 61: 437–443, 1996.
5. Helman DN, Addonizio LJ, Morales DL, et al: Implantable left ventricular assist devices can successfully bridge adolescent patients to transplant. J Heart Lung Transplant 19: 121–126, 2000.
6. Sadeghi AM, Marelli D, Talamo M, Fazio D, Laks H: Short-term bridge to transplant using the BVS 5000 in a 22-kg child. Ann Thorac Surg 70: 2151–2153, 2000.
7. Ashton RC Jr., Oz MC, Michler RE, et al: Left ventricular assist device options in pediatric patients. ASAIO J 41: M277–280, 1995.
8. Takano H, Nakatani T: Ventricular assist systems: experience in Japan with Toyobo pump and Zeon pump. Ann Thorac Surg 61: 317–322, 1996.
9. Stiller B, Daehnert I, Weng Y, Hennig E, Hetzer R, Lange PE: Children may survive severe myocarditis with prolonged use of biventricular assist devices. Heart 82: 237–240, 1999.
10. Karl TR, Horton SB: Options for mechanical support in pediatric patients, in Goldstein DJ, Oz MC (eds), Cardiac Assist Devices. Armonk, NY: Futura, 2000, pp. 37–62.
11. Meliones JN, Custer JR, Snedecor S, Moler FW, O’Rourke PP, Delius RE: Extracorporeal life support for cardiac assist in pediatric patients. Review of ELSO Registry data. Circulation 84: III168–172, 1991
12. Duncan BW, Hraska V, Jonas RA, et al: Mechanical circulatory support in children with cardiac disease. J Thorac Cardiovasc Surg 117: 529–542, 1999.
Copyright © 2002 by the American Society for Artificial Internal Organs