| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Pediatric Cardiothoracic Surgery, University of Arkansas for Medical Sciences, Arkansas Children's Hospital, Little Rock, Arkansas
b Pediatric Cardiothoracic Anesthesiology, University of Arkansas for Medical Sciences, Arkansas Children's Hospital, Little Rock, Arkansas
c Pediatric Cardiology, University of Arkansas for Medical Sciences, Arkansas Children's Hospital, Little Rock, Arkansas
Accepted for publication March 17, 2009.
* Address correspondence to Dr Imamura, Pediatric Cardiothoracic Surgery, Arkansas Children's Hospital, 800 Marshall St, Slot 677, Little Rock, AR 72202 (Email: imamuramichiaki{at}uams.edu).
Presented at the Forty-fifth Annual Meeting of The Society of Thoracic Surgeons, San Francisco, CA, Jan 26–28, 2009.
| Dr Imamura discloses that he has a financial relationship with ZymoGenetics.
|
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
Methods: Data for 55 consecutive children undergoing BTT between 2001 and 2008 were abstracted from an institutional database. The analysis excluded 13 patients because EXCOR was not used for acute postcardiotomy BTT. Patients were divided into ECMO (n = 21) and EXCOR groups (n = 21). Specific end points included survival to transplant, overall survival, and bridge to recovery. Incidences of adverse events and the duration of support were determined.
Results: Groups were similar in weight, age, and etiologies of heart failure. Likewise, the incidences of stroke and multisystem organ failure were similar. Survival to transplant, recovery, or continued support was 57% in ECMO and 86% in EXCOR (p = 0.040). EXCOR patients had overall significantly better survival (p = 0.049). Two ECMO patients and 1 EXOR patient were bridged to recovery. The mean duration of support was 15 ± 12 days in the ECMO group and 42 ± 43 days in the EXCOR group (p < 0.001).
Conclusions: In children requiring BTT, EXCOR provided substantially longer support times than ECMO, without significant increase in the rates of stroke or multisystem organ failure. Survival to transplant and long-term survival was higher with EXCOR.
Cardiac transplantation is the definitive therapy for end-stage heart failure in children. A large number of these patients die before undergoing heart transplantation because of the scarcity of donor organs. Children with the most severe heart failure have been supported with extracorporeal membrane oxygenation (ECMO). This support modality is associated with significant morbidity and mortality, especially as the duration of support extends beyond 2 to 3 weeks. Severe complications such as cerebral infarction, brain hemorrhage, renal failure, and multiorgan system failure are often fatal or preclude transplantation. Recently, adult ventricular assist devices (VADs) have been used in larger children for the management of end-stage heart failure [1]. The large cannula sizes and stroke volumes of adult VADs have limited the usage of these devices in small children [2].
Several groups in Europe have developed VADs specifically designed for small children [3, 4]. One of these, the Berlin Heart EXCOR device (Berlin Heart AG, Berlin, Germany), has become available for use in North America in the last 5 years on a compassionate-use basis or as part of an investigational new device trial [5–7]. At the same time, the National Heart, Lung, and Blood Institute has sponsored an initiative to support United States domestic development of pediatric-specific devices [8].
Our center has relatively extensive experience with ECMO and its use as a bridge to cardiac transplantation [9] but limited experience with the use of VADs in children [10]. In April 2005, the Berlin EXCOR was added to our armamentarium of bridge technology for children awaiting cardiac transplantation [5]. Like ECMO, VAD implantation is potentially associated with significant morbidity and mortality but may permit longer bridge time and even allow pretransplantation rehabilitation. We undertook the present study to compare the clinical courses of two groups of patients who were supported by EXCOR and ECMO.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
|
A left ventricular assist device (LVAD) was used in 14 patients and a biventricular assist device (BiVAD) in 7. Initially we always tried to implant the LVAD alone. When appropriate filling of the LVAD pump was not obtained and adequate systemic perfusion was not present, despite maximal medical support of the right ventricle (including nitric oxide, inotropic support, and high central venous pressure), a right ventricular assist device (RVAD) was added.
ECMO
All patients but one were cannulated through a right cervical approach using the right carotid artery and jugular vein. One patient, who was scheduled for EXCOR implant but collapsed before its arrival, was cannulated through a median sternotomy using the ascending aorta and the right atrium. Most ECMO support used a standard roller pump, venous line, collapsible venous reservoir, silicone membrane oxygenator, heat exchanger, and arterial line. More recently, some patients were supported using a centrifugal pump. In the ECMO group, 11 patients had balloon atrial septostomy [11].
Anticoagulation
In EXCOR patients, postoperative anticoagulation was started with intravenous heparin 24 hours after arrival to the intensive care unit. The goal for the partial thromboplastin time was 1.5 to 2.0 times the normal value. Antithrombin III was replaced when the serum level was below 70% of normal. Subsequently, aspirin and dipyridamole were started, guided by the results of platelet mapping by thromboelastogram TEG 5000 (Haemoscope Corp, Niles, IL). Later, low molecular heparin or oral warfarin was started. A heparin-induced antibody was identified in 1 patient, and argatroban was used for anticoagulation during EXCOR support [12].
In ECMO patients, intravenous heparin was adjusted to keep the active clotting time level between 160 and 200 seconds.
Statistical Analysis
The end point of the study was the time to death. For descriptive parameters, variables were analyzed with the 
2 test or the Fisher exact test. For comparison of means, the t test was used. Survival estimates were made with the Kaplan-Meier method, and comparisons between survival distributions were made with the Breslow generalized Wilcoxon method. Mean, median, range, and standard deviations were also calculated. A value of p < 0.05 was considered statistically significant. All data were analyzed using SPSS 16.0 software (SPSS Inc, Chicago, IL).
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
Table 4
summarizes the comparison of postsupport profiles, including complications.
|
|
|
|
One late death in the EXCOR group was due to infection at 602 days after transplantation. There were two late deaths in the ECMO group, one from infection at 1112 days and one from rejection at 1437 days.
In the EXCOR group, 16 patients received a transplant and 2 recovered and were weaned from EXCOR. One patient remains on support. All transplant recipients were discharged. Of the two patients weaned from EXCOR, one died early from sepsis, and the other is alive and well at home. A flow chart of ultimate outcomes for EXCOR patients is shown in Figure 2.
|
|
Overall survival for the two groups is compared in Figure 4 and shows significantly better survival in the EXCOR group with Breslow generalized Wilcoxon test (p = 0.049). Hospital survival was not different between LVAD only (12 of 14) and BiVAD (6 of 7). Figure 5A illustrates the interaction of accumulating experience and survival for the EXCOR. Figure 5B displays similar information for CVA during EXCOR support, demonstrating no reduction in the rate of CVA in the increasing center experience. In the EXCOR group, 8 patients had CVAs. Two events were devastating, and support was withdrawn. The other 6 patients underwent transplantation and were discharged. Only 1 of these patients had apparent neurologic impairment at discharge.
|
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
Since coming into widespread usage as the pediatric bridge modality in the 1990s, ECMO has been the mainstay of therapy for children with end-stage cardiac failure who require mechanical support. A number of single-center reports [9, 14–20] have confirmed the utility of ECMO in this application, with a survival to hospital discharge of about 50%, although higher rates have been reported [15]. The 57% survival to discharge for ECMO-bridged patients in the present study is consistent with this experience, although slightly less than in our previous report [9], of which 28% of the patients were listed for transplant but recovered. The limitations of ECMO as a bridge modality, predominantly relating to an accelerating rate of transplant-disqualifying and fatal complications after approximately 14 days of support, have been well described and were also seen in this study. Lesser, but still important, drawbacks to ECMO are the requirement for sedation, mechanical ventilation, and the necessity of keeping the patients bedbound while on support.
Against this backdrop, Hetzer and colleagues [3], at the Deutches Herzzentrum, developed the Berlin Heart EXCOR VAD, which is a pneumatically driven pulsatile device capable of univentricular or biventricular support [3]. The device is available in a variety of sizes that permit its application in neonates as well as older children and teenagers. The device was first used in the United States in 2000 and has come into increasingly widespread use in a variety of centers since 2005. A number of single-center reports have appeared [5–7], with survival rates and support times that have generally exceeded typical results with ECMO. Our report confirms this experience with an intracenter comparison with ECMO, with the major finding that the EXCOR provides longer waiting times with consequent improvement in the proportion of patients who received a transplant and, most importantly, improvement in survival.
The two groups had an identical proportion of patients with congenital heart disease; however, the ECMO group also included 3 patients with single-ventricle physiology, but the EXCOR group did not. During the period of the study, we considered single-ventricle anatomy to be a contraindication to EXCOR use.
Comparison with ECMO, which always provides biventricular support even if the right ventricular function is preserved, the EXCOR device requires a decision about whether to provide univentricular support or biventricular support. As has been well demonstrated in the adult experience, this decision is not always straightforward in patients with RV dysfunction of a moderate degree. In certain circumstances, the provision of LV support does not provide enough relief to allow the dysfunctional RV to improve. In other circumstances, the reduction in left atrial pressure and RV afterload by placement of an LVAD, along with adjuncts such as inotropic medications and inhaled nitric oxide, have allowed for univentricular support with gradual improvement in RV dysfunction. This latter scenario has been the rule rather than the exception in the experience described in the present study.
Our default approach for children with myocarditis or dilated cardiomyopathy not supported on ECMO at the time of device placement has been to implant an LVAD only, and we did not have to initiate another operation to implant an RVAD in any patient treated in this fashion. In contrast, for children on ECMO at the time of implantation, we have tended to proceed with biventricular support on the assumption that the RV has become deconditioned during ECMO. In children on ECMO supported with LVAD alone, our impression has been that the degree of early RV dysfunction has been much more significant than in children not on ECMO before LVAD implantation. It must be acknowledged that ECMO-supported patients may simply have had worse RV function to begin with, rather than having RV deconditioning as the explanation.
The exception to this approach of LVAD-only for children with dilated cardiomyopathy has been in the setting of extremely elevated pulmonary vascular resistance. In this type of patient, we have pursued the strategy described by Gandhi and colleagues of BiVAD implantation with aggressive treatment of pulmonary resistance. The magnitude and rapidity of decline in pulmonary vascular resistance has been remarkable, allowing patients whose pulmonary resistance was originally prohibitive to go on to successful transplantation. It is unclear how long to persist with this approach before reassessment of the pulmonary resistance and it is also unclear whether all patients will respond. However, our preliminary experience is certainly encouraging in this group of patients for whom no other reasonable treatment is possible.
A disappointing aspect of our experience with the EXCOR device has been a relatively high incidence of neurologic injury, which at 40% is less than that described by the group at Stanford [6] and more than described by the group at Washington University [7]. The mechanism and timing of stroke in our patients has not followed a consistent pattern; thromboembolism and intracranial hemorrhage have both occurred. The episodes of thromboembolism have been particularly frustrating because they have occurred in the setting of full anticoagulation and antiplatelet therapy, guided by conventional anticoagulation monitoring as well as thromboelastography and platelet function mapping. Furthermore, none of the events were preceded by the appearance of visible debris in the transparent blood pump.
Although potentially more devastating, hemorrhagic stroke has at least been more readily explicable. The consequences of both types of stroke have been significant: 2 patients died of hemorrhagic stroke, and all of the patients with embolic stroke have required extensive rehabilitation. As a testimony to the plasticity of the juvenile central nervous system, only 1 patient with embolic stroke has any significant residual dysfunction.
Despite the present analysis that contrasts ECMO and VAD as bridge technologies, it is important to emphasize that the two modalities are actually complementary rather than competitive, as evidenced by the fact that one-third of the VAD patients were converted from initial ECMO support. It is clear that a portion of patients requiring mechanical support may present in such dire hemodynamic condition that emergency initiation of support is necessary. In the foreseeable future, this will mean using ECMO for small children. Older, adult-sized children may receive emergency percutaneous VAD support (TandemHeart; Cardiac Assist Inc, Pittsburgh, PA).
Furthermore, children may present and require support in institutions that do not have VAD programs or transplant capability. In the present study, 9 of the 42 patients were transported by air to our facility by our mobile ECMO team [21]. Of these, 6 were placed on ECMO by surgeons at the referring facility, and 3 were placed on support by our mobile ECMO team. In fact, referral for VAD support and transplantation has become the leading indication for mobile ECMO in our program, which has transported more than 100 patients since its inception.
As experience with the Berlin Heart grows, a recurring and evolving question is which patients should be supported with which technology. For the few patients who have hemodynamic collapse, the answer is obviously ECMO; but for the remainder, the choice must be guided by other factors, including an assessment of the likely wait times for transplant. In contrast to the extremely short wait time of 7.5 days reported in the experience reported by Goldman and colleagues [16], waiting times in our center are much more frequently measured in weeks or even months. During the period examined in the present study, 2001 to 2008, we performed 115 transplants, with median waiting times for United Network for Organ Sharing status 1A of 22 days (range, 1 to 188; mean 36 ± 40 days). In this experience, no patient survived more than 34 days on ECMO, while 9 of the EXCOR patients were on support for more than 34 days. Thus, in the current era in the United States, it seems reasonable that most patients requiring bridge support will be better served by VAD, although we certainly agree that if the waiting times are predictably less than 3 weeks in another location, ECMO would be a reasonable alternative.
An additional consideration in selecting the type of support is the underlying cardiac diagnosis, with specific focus on whether the process may be reversible, and, if so, in what time frame. For example, some patients with acute myocarditis may recover very quickly with support times of days or, at most, a few weeks. However, it is virtually impossible to predict whether the recovery may take months and often just as difficult to distinguish established cardiomyopathy from the subacute phase of myocarditis. In the event that a recovery is thought to be possible within a period of weeks or months, the VAD is certainly the modality of choice. In the present study, 2 patients were weaned from VAD support. One died a few days later of sepsis, but the other is well with normal ventricular function at more than a year after decannulation. Similarly, 2 patients were weaned from ECMO support, of which 1 required transplantation during the same hospitalization, and the other was discharged.
Currently at our institution, for patients on ECMO who have biventricular physiology and are awaiting a transplant, we proceed to EXCOR implantation if the patients cannot be weaned from ECMO after approximately 7 to 10 days of support. For patients awaiting transplant, but not on ECMO, a requirement for escalating dose levels of inotropic support or worsening end-organ dysfunction (renal, hepatic, or pulmonary) is considered an indication for EXCOR implantation. The necessity of intensive care management or stable mechanical ventilation of regular fraction of inspired oxygen, peak end-expiratory pressure, or peak inspiratory pressure, is not of itself considered an indication for mechanical support.
The most important limitations in the present study are its retrospective and nonrandomized nature, although it is hard to imagine how a randomized comparison of ECMO and VAD could be conducted. An additional limitation is the relatively small study size limiting the statistical robustness of any inferences that may be drawn. The patient groups are also not perfectly contemporaneous, because the EXCOR device was only used during the latter half of the study. This means that the follow-up interval for the two groups is different, although it is likely that differences in outcome will be apparent in early follow-up.
Finally, the study represents a comparison between a mature technology, ECMO, with which our center has experience in more than 850 patients, and a novel technology with which we have significantly less experience. The influence of experience is graphically shown by the fact that all of the deaths in the EXCOR group occurred in the first 5 patients in the series. Viewed in this way, it is likely that the present study significantly underestimated the advantage of EXCOR vs ECMO.
In conclusion, patients supported with EXCOR achieved a longer period of support and improved survival compared with similar patients bridged to transplant with ECMO. Both groups had a disappointingly high rate of stroke, the incidence of which did not seem to decline during the study period. Because the waiting times for children listed for transplantation in the United States are unlikely to decline in the foreseeable future, it is likely that the EXCOR, although not risk-free, will assume an increasing role in helping to keep patients alive until a suitable donor organ is available and thus reduce waiting list mortality.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
DR IMAMURA: The first question, the two groups, some parts are pretty similar; however, these patients are not equal, because some patients had a single-ventricle physiology in ECMO group. We do not use EXCOR [Berlin Heart AG, Berlin, Germany] for the single-ventricle physiology patients. And as I showed, the one group of patients on EXCOR had the ECMO before. And at the time of ECMO, we select patients, if these patients fit for the EXCOR or not, so these groups are not totally comparable patient groups. And additionally, I think ECMO and EXCOR are not competitive modalities, but are complementary modalities to obtain a longer support period.
DR CHRISTOPHER CALDARONE (Toronto, Canada): I have to say I share Jim's concern a little bit there because the differences in the population may make important difference in terms of the survival curves that you have shown in the abstract. Also, 7 of the ECMO patients crossed over to the EXCOR; presumably, those were the healthier patients or more stable patients, et cetera. So that is some selection process which may have made the ECMO results look worse, which may not be a good representation of what was going on there. So at first glance the abstract appears to suggest that ECMO is an inferior mode of support, but it sounds like you are saying that that is not your intended message.
DR IMAMURA: Right.
DR JOHN HAWKINS (Salt Lake City, UT): I think the central question is whether the Berlin Heart is the cause of the improved survival or the effect. And I just wonder also if the two populations are the same, because 9 of the 19 patients that were on ECMO died of multisystem organ failure and would therefore not be deemed probably salvageable to go onto the Berlin heart, which is not readily available on a very quick basis like ECMO is. And so the real question is whether these 2 patient groups are the same and that you didn't just select out the patients with less optimal outcomes by going through ECMO first. So if you separated them based upon initial therapy, would that ECMO really look as bad; instead of basing it upon ultimate therapy, which was how you divided the patients.
DR CALDARONE: Could you comment on a tough clinical dilemma, and that is, a patient with failing cardiomyopathy requiring some mild or moderate inotropic support in the ICU [intensive care unit]. There is always a question whether to let them continue to deteriorate, knowing that you could use ECMO to resuscitate them and then electively put in a Berlin heart, vs the strategy of operating while they are reasonably well and implanting a Berlin heart. How do you make that decision? That's a tough one that we have faced.
DR IMAMURA: Yes, during this period we are still trying to find out the indications for EXCOR. Since we still have a high incidence of neurologic complications, we hesitate to put the EXCOR to the patient in the unit and without intubation. However, in patients with another organ failure, like kidney failure or liver failure, we use the EXCOR.
DR CALDARONE: So the requirement for intubation you use is ...
DR IMAMURA: I do not use. We do not put EXCOR just because the patient is intubated.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. M. Chen, M. E. Richmond, K. Charette, H. Takayama, M. Williams, L. Gilmore, A. Garcia, and L. J. Addonizio A decade of pediatric mechanical circulatory support before and after cardiac transplantation J. Thorac. Cardiovasc. Surg., February 1, 2012; 143(2): 344 - 351. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Vasavada, Feng Qiu, and A. Undar Current status of pediatric/neonatal extracorporeal life support: clinical outcomes, circuit evolution, and translational research Perfusion, July 1, 2011; 26(4): 294 - 301. [Abstract] [PDF]
|
|
|
|
 |
|
 Y. Fan, Y.-G. Weng, Y.-B. Xiao, M. Huebler, N. Franz, E. Potapov, and R. Hetzer Outcomes of ventricular assist device support in young patients with small body surface area Eur J Cardiothorac Surg, May 1, 2011; 39(5): 699 - 704. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Stiller, C. Benk, and C. Schlensak Mechanical cardiovascular support in infants and children Heart, April 1, 2011; 97(7): 596 - 602. [Full Text] [PDF]
|
|
|
|
 |
|
 R. Henaine, M. Moutaouekkil, S. Di-Filippo, and O. Bastien Right ventricular recovery after cervical extra corporeal membrane oxygenation in a four-month-old male leading to left Berlin Heart Excor assistance Interact CardioVasc Thorac Surg, December 1, 2010; 11(6): 840 - 842. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Arain, R. Bryant III, R. Ameduri, D. Gruenstein, E. Braunlin, L. Joyce, and J. St. Louis Successful Coil Embolization of a Large Ascending Aortic Pseudoaneurysm Following Explantation of the EXCOR Pediatric Ventricular Assist Device in a Patient With Acute Fulminant Myocarditis World Journal for Pediatric and Congenital Heart Surgery, July 1, 2010; 1(2): 259 - 261. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
