Prospective Trial of a Pediatric Ventricular Assist Device

Fraser CD, Jaquiss RD, Rosenthal DN, Humpl T, Canter CE, Blackstone EH, Naftel DC, Ichord RN, Bomgaars L, Tweddell JS, Massicotte P, Turrentine MW, Cohen GA, Devaney EJ, Pearce FB, Carberry KE, Kroslowitz R, Almond CS.

New England Journal of Medicine 2012; 367:532-41

Reviewers: Jennifer E. Lam, DO and Lori A. Aronson, MD
Cincinnati Children’s Hospital Medical Center

Ventricular assist devices (VAD) have been used for years as a bridge to transplantation in adults with severe heart failure. However, due to size constraints and issues with thromboembolism, VADs have been of limited use in young children. Thus, the majority of pediatric patients awaiting transplantation have been placed on extracorporeal membrane oxygenation (ECMO) while awaiting a donor, a process that has been associated with a mortality of 43% and is generally effective for only 1-3 weeks when the average wait for an infant heart is over 100 days (1). The introduction of the pneumatically driven pediatric-specific VAD, the EXCOR Berlin Heart, available in a variety of sizes to accommodate a child as small as 3kg, has given physicians another option as a bridge to transplant with promising results.

Methods
This study is a prospective, multicenter (17 pediatric cardiac centers in the US and Canada), single-group cohort study comparing the use of a VAD versus ECMO for children as a bridge to heart transplantation. It included children 16 years of age and younger between 3 and 60 kg with two-ventricle circulation and severe heart failure on the waiting list for cardiac transplantation. Cohort 1 included a body-surface area of less than 0.7m2 and cohort 2 had a body-surface area of at least 0.7m2 but less than 1.5m2. A historical control group of children receiving ECMO was selected from the Extracorporeal Life Support Organization registry and a propensity-score analysis was used to match each patient who received a VAD to two patients who had received ECMO support. LVAD versus BiVAD placement was algorithm–based and a standardized anti-thrombotic regimen was employed.

The primary end point for the VAD group was the time to death or weaning from the device secondary to an unacceptable neurologic outcome. The primary end point for the ECMO group was time to death. Death was defined for all groups as occurring while the child required mechanical circulatory support (MCS) or within 30 days from weaning from MCS or before hospital discharge. A secondary outcome assessment for the VAD group included: death during receipt of circulatory support with the device, heart transplantation, failure to wean, and successful weaning. The outcome events tracked for the ECMO group included: death during device placement with ECMO, death within 30 days after weaning from ECMO, and removal of ECMO (without death); however, neurologic status was not reported.

Results
A total of 48 children were enrolled, 24 in each cohort. The median age and weight in cohort 1 were 1 year and 9kg. The median age and weight in cohort 2 were 9 years and 31kg. The majority of cardiac failure was due to either cardiomyopathy or myocarditis. The median duration of support for cohorts 1 and 2 for the VAD group was 28 days and 43 days, respectively. The median duration of support for the ECMO group was 5 days for both cohorts. The longest duration of support in either VAD group was 192 days versus 28 days for ECMO.

The median time to the primary end point for cohort 1 in the VAD group had not been reached at 174 days. The median time for the matched ECMO group was 13 days (p<0.001). In cohort 2, the primary end point was reached at a median time of 144 days for the VAD group versus 10 days for the matched ECMO group (p<0.001). In cohort 1 for the VAD group, at 174 days, 88% of the patients had undergone successful transplantation and 12% had died or had an unacceptable neurologic outcome upon weaning. In the matched ECMO group in cohort 1, 25% of the patients had died by 21 days and none were alive and still receiving ECMO support. In cohort 2 for the VAD group, at 192 days, 92% of the patients had undergone successful transplantation or had been weaned from the device, and 8% had died. In the matched ECMO group for cohort 2, 33% of the patients died, and none were alive and still receiving ECMO support. Efficacy outcomes were approaching statistical significance for cohort 1 VAD compared to its ECMO matched group and was significant for cohort 2 VAD versus its ECMO matched group.

Adverse events for the VAD patients in cohorts 1 and 2 included: major bleeding (42% and 50%), infection (63% and 50%), stroke (29% and 29%), and hypertension (50% and 33%), respectively. The primary safety endpoint, serious adverse events per patient-day of support with VAD, was prospectively defined as <0.25 events/day with a confidence interval of 95%. The actual rate of serious events in cohorts 1 and 2 were 0.07 events per patient-day and 0.08 events per patient-day respectively. Additionally, 46 pump changes occurred between both cohorts, usually secondary to thrombus formation.

Comments
This well-conducted study is the first multicenter study of its kind in the U.S./Canada comparing not only the survival outcomes, but also the safety and efficacy of the Berlin Heart EXCOR VAD versus ECMO as a bridge to transplantation in children with severe heart failure. The data supports previous institutional reports of substantially longer mechanical support times with the VAD and increased survival to transplant (1-4.)

The major limitations of this study are due to the use of historical data for the ECMO group and lack of randomization. This ECMO registry did not have information on adverse events or set end points for discontinuing ECMO. In addition, the authors admit to the possibility that the ECMO group may have been more ill. Of note, no data were available on the neurological outcomes of the ECMO group, as was evaluated in the VAD cohorts as a primary end point, and thus, the survival rate for the VAD group may be even higher than reported. Furthermore, since there was no age-matching or weight-matching for cohort 2, these patients may have differed substantially more from the larger VAD patients.

Although the authors have demonstrated a significant increase in survival with the use of the Berlin Heart EXCOR VAD in this population, the incidence of serious adverse events should be recognized. Previous studies have not shown a significant increase in the rates of stroke in VAD versus ECMO patients (1), however, Fraser et al study did not compare such data. They did note, however, that the sequelae of the strokes for the majority of the patients did not preclude eligibility for transplantation and were generally mild. It is yet to be determined if thrombosis and stroke are related to a low-flow pulsatile system or a suboptimal anticoagulation strategy in children. Further studies will hopefully clarify these points.

Conclusion
Prior to the introduction of the Berlin Heart EXCOR in the U.S. for pediatric patients in 2000, children in severe heart failure had few options other than ECMO as a bridge to transplantation with high waitlist mortality and a very limited period of support. This study by Fraser et al adds data to the growing evidence that survival rates are significantly higher with the use of a VAD as a bridge to transplant and provides a much longer support period than ECMO in pediatric heart failure. This longer period of support will hopefully afford physicians more time to optimize patient status to improve transplant candidacy and overall morbidity and mortality. More information needs to be gathered, however, regarding the rate of serious adverse events with this device.

References

Michiaki I, Dossey AM, Prodhan P, Schmitz M et al. Bridge to Cardiac Transplant in Children: Berlin Heart versus Extracorporeal Membrane Oxygenation. Ann Thorac Surg 2009;87:1894-1901.
Morales DLS, Almond CSD, Jaquiss RDB et al. Bridging children of all sizes to cardiac transplantation: The initial multicenter North American experience with the Berlin Heart EXCOR ventricular assist device. The Journal of Heart and Lung Transplantation 2011;30:1-8.
Sanjiv K, Huddleston CB, Balzer DT, Epstein DJ, Boschert TA, Canter CE. Cardiac Transplantation and Surgery for Heart Failure: Biventricular Assist Devices as a Bridge to Heart Transplantation in Small Children. Circulation 2008;118:S89-S93.
Jeewa A, Manlhiot C, McCrindle BW, Van Arsdell G, Humpl T, Dipchand A. Outcomes with ventricular assist device versus extracorporeal membrane oxygenation as a bridge to pediatric heart transplantation. International Center for Artificial Organs and Transplantation and Wiley Periodicals 2010 Dec;34(12):1087-91.