| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2006;82:917-925
© 2006 The Society of Thoracic Surgeons
Deutsches Herzzentrum Berlin, Berlin, Germany
Accepted for publication March 10, 2006.
* Address correspondence to Dr Hetzer, Deutsches Herzzentrum Berlin, Augustenburger Platz 1, Berlin 13353, Germany (Email: hetzer{at}dhzb.de).
Presented at the Forty-first Annual Meeting of The Society of Thoracic Surgeons, Tampa, FL, Jan 24–26, 2005.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
METHODS: Since 1990 and as of January 1, 2005, 62 Berlin Heart Excor systems have been implanted in patients below 18 years of age at our institution. The patients were divided into two groups according to the period of treatment: period 1, devices implanted between 1990 and 1998 (n = 34), and period 2, devices implanted between 1999 and 2004 (n = 28). We compared our experience during the earlier and later periods.
RESULTS: There were no significant differences in the preoperative patient data between the two periods except for time of support (17.9 ± 27.7 days versus 53.2 ± 83.9 days, p = 0.001). In period 1, more patients needed a biventricular VAD whereas in period 2, more patients were effectively treated with a left VAD (p = 0.05). In the later period, the chest could be primarily closed in a significantly higher percentage of infants (0% versus 89%, p = 0.012), and more infants could be extubated on the VAD (0% versus 55%, p = 0.16). Discharge from the hospital after either weaning from the system or heart transplantation was achieved for 35% in period 1 and for 68% in period 2 (p = 0.029). Whereas in period 1 there were no survivors in the group of children younger than 1 year old, during period 2, survival in this age group was similar to that of the two groups of older children (p = 0.024). There was a significant improvement in the discharge rate in period 2 in patients with cardiomyopathy (43% versus 76%, p = 0.045) and postcardiotomy heart failure (0% versus 57%, p = 0.01).
CONCLUSIONS: Earlier implantation of VADs, heparin coating of the blood pumps, and substantial modifications in cannula design, anticoagulation, and the coagulation monitoring regimen have led to a significant increase in the survival and discharge rate, especially among children under 1 year of age. The pediatric size Berlin Heart Excor VAD is a valuable option as a bridge to heart transplantation or recovery for children suffering from cardiogenic shock.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
This report analyzes our experience during the period since 1999 (period 2) in comparison with the earlier period between 1990 and 1998 (period 1).
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
The Berlin Heart System
The Berlin Heart ventricular assist device (Berlin Heart AG, Berlin, Germany) consists of a paracorporeal, pneumatic compressor-operated diaphragm pump with valves, silicone cannulas, and the IKUS stationary driving unit. In the larger pumps and with driving pressures lower than 250 mm Hg, the Excor mobile driving unit (Berlin Heart AG) can be utilized.
Anticoagulation Therapy
During the whole period of mechanical support, small infants receive continuous heparin infusion. We have no experience with newborns or young infants administered warfarin while on an assist device, in contrast to older children. The anticoagulation management was changed during the years of this investigation. In period 1, the activated clotting time (ACT) was measured every 2 hours, the target range being 140 s to 160 s. Different results led to frequent changes of the heparin dose, with subsequent coagulation disorders and bleeding requiring rethoracotomy. In the later period, the activated partial thromboplastin time was analyzed every 4 to 6 hours in the first few days and, after stabilization, changed to twice daily, with a target range of 50 s to 70 s. In both groups, antithrombin III was substituted if it fell below 70%. In both groups, older children received phenprocoumon with a target international normalized ratio (INR) level of 3 to 3.5. Further changes in period 2 are additional administration of dipyridamol and aspirin, starting after the first week of support and removal of the chest drains, analysis of the thrombelastogram, and platelet function tests [5, 6].
Blood Pumps
The blood pumps were available in both periods with volumes of 10, 25, 30, 50, 60, and 80 mL (see Fig 1A). In period 1, in addition, modified pumps with 12- and 15-mL stroke volume were applied in some cases. The pump consists of a translucent, semirigid housing of polyurethane [7], and the pump chamber is divided into a blood chamber and an air chamber by a flexible diaphragm in three layers. The two diaphragm layers facing the air chamber serve as driving membranes; the seamless cast-on blood membrane is passively moved by the driving membranes. A deairing nipple, into which a cannula can be introduced to eliminate residual air after pump implantation, is integrated into the blood chamber. To direct the blood flow, trifleaflet polyurethane valves are mounted in the pediatric pumps (10, 25, 30 mL), whereas either tilting disc valves (Sorin Biomedical, Torino, Italy) or trileaflet polyurethane valves are available for the adult pump sizes (50, 60, 80 mL). The cross-section of the pump is shown in Figure 1B.
|
Cannulas
The Berlin Heart Excor cannulas are made of silicone rubber with an extremely smooth internal surface. Standard cannulation of the right heart is accomplished by drainage of the right atrium and ejection of the right pump blood volume into the pulmonary artery. On the left side, cannulation of either the left atrium or the left ventricular apex is possible (Fig 2). Pump outflow is directed into the ascending aorta. For special indications in adult patients, a system for left ventricular apex drainage and connection to the descending aorta is available.
|
|
All cannulas have a Dacron felt covered sewing ring. The suture technique includes 8 to 10 Teflon (Impra, subsidiary of L. R. Bard, Tempe, Arizona) felt enforced sutures and an additional pursestring suture. The infant and neonate arterial cannulas have a short titanium enforced tip for insertion into the blood vessel wall, the so-called "press-button" tip. These cannulas produce a considerably reduced afterload to the native heart in comparison with conventional heart-lung machine cannulas.
All Berlin Heart Excor cannulas are made to exit the body through the upper abdominal wall.
Driving Unit
The Berlin Heart Excor VAD for smaller children is powered by the IKUS driving unit. Compressed air moves the pump diaphragm into its endsystolic position, thereby ejecting the blood volume. In pump diastole, negative pressure is created to assist filling of the pump.
The driving unit consists of three separate compressor units: one for the left pump, one for the right pump, and one back-up compressor. In the case of malfunction of one compressor unit, the back-up unit will take over automatically and without delay. Should two compressor units simultaneously stop working, an acceptable pump output for both pumps is produced by the third compressor unit with a pump rate of 90 beats per minute.
The compressor and pressure/vacuum regulator actions are controlled by two redundant internal computers. User control is enabled through a personal computer (notebook). An internal battery provides DC power for up to 1 hour.
The maximum positive driving pressure is 350 mm Hg, and the maximum negative driving pressure is minus 100 mm Hg. These excessive pressures are necessary to overcome the high flow resistance of the pediatric cannulas. The pump rate can be adjusted to between 30 and 150 beats per minute, and the relative systolic duration can be varied from 20% to 70%.
The system may be operated in univentricular or biventricular (BVAD) mode. A special feature is the option of independent control for each side in terms of pump rate, systolic pump pressure, diastolic pump pressure, and length of systole.
Pump action in biventricular mode may be set to "synchronous" for simultaneous, "asynchronous" for alternating, and "separate" for independent pump action. The possibility of separate rates for the left and right pumps is especially important in the case of the right ventricle recovering on BVAD support, where the output of the right pump may have to be reduced to prevent pulmonary edema [7].
Patients
Since 1990 and as of January 1, 2005, Berlin Heart Excor systems have been implanted in 62 patients less than 18 years of age. The patients were divided into two groups according to the period of treatment: period 1, devices implanted between 1990 and 1998 (n = 34), and period 2, devices implanted between 1999 and 2004 (n = 28; Table 1). The primary outcomes are survival (defined as 30-day survival, heart transplantation, or myocardial recovery) and hospital discharge, namely, discharge home or to a rehabilitation center. The pump size (stroke volume) was chosen according to weight and anatomy. Table 2
shows the pump sizing in different age groups.
|
|
2 test was applied to test differences between treatment periods. A p value of less than 0.05 was considered statistically significant. A cumulative sum plot of all patients with cumulative mortality on the y-axis and patient numbers on the x-axis was constructed to show the improvement in mortality rate. |
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
There were more left ventricular assist devices (LVAD) implanted in the later treatment period. The differences reached statistical significance for infants less than 1 year old (p = 0.011; Table 1).
The outcome in the two different treatment periods is shown in Table 3. In the later treatment group, the chest could be primarily closed after VAD implantation in a significantly higher percentage of infants less than 1 year old (0% versus 89%, p = 0.012), and more infants could be extubated on the VAD (0% versus 55%, p = 0.156). Among the surviving patients, 5 of 15 patients were extubated on the device in period 1, and 13 of 24 in period 2.
|
|
|
|
In the earlier period, reexploration for bleeding was necessary in 8 patients compared with 7 patients in the later period. The pump exchange rate per month for thrombus formation was 1.0, 0.3, and 0.4 in the early period, and 0.36, 1.1, and 0.27 in the later period in the age groups of less than 1 year, 1 to 8 years, and more than 8 years, respectively.
|
Comment |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
In comparison with patients with other devices such as centrifugal pumps and extracorporeal membrane oxygenation (ECMO), which we also use in patients with potential for fast myocardial recovery and therefore expected short-term support [11–15], patients with the Berlin Heart VAD can be mobilized, extubated, and fed orally (Fig 7), which is advantageous when waiting periods of more than 3 or 4 weeks must be anticipated [16–19]. Further, patients on VAD required less transfusion of blood products, with a correspondingly decreased risk of infection and of developing HLA antibodies [6, 20]. The advantages of pulsatile VAD application in children have been recognized, and an increasing number of centers are using this type of support system in small children and adolescents [21–23]. Application of the Medos VAD, which is also available for small children, has been reported to be successful with shorter support times [4], but use of the Thoratec and the MicroMed DeBakey VAD is limited to children weighing more than 15 to 20 kg [19, 24–28].
|
Four patients were supported by ECMO before VAD implantation. The decision to convert ECMO to the VAD was taken when it became clear that longer-term support was needed and that the patient had not suffered damage by ECMO. One of the indications was continuous bleeding on ECMO or the cardiopulmonary bypass circuit.
As with any kind of support system, making the decision in favor of earlier implantation has yielded better results. This effect was most pronounced in the patient group less than 1 year of age. In this group during the earlier period, the majority of patients were placed on the system in a state of advanced circulatory failure, characterized by irreversible organ shock sequelae and unresponsiveness of the peripheral circulation to alpha-stimulants, so-called vasoplegia. Consequently, in period 1, no patient less than 1 year of age survived longer than 30 days after VAD implantation, whereas after a policy of earlier placement on the system was established, more than three quarters of the infants could finally leave the hospital alive.
The introduction into the pediatric field of the criteria that have become well established in the adult population, namely, VAD implantation before shock organ failure sets in or, at the latest, at the very first signs of such organ failure [16, 29–33], has fulfilled its promise of better survival. The criteria to be observed when making the decision for VAD implantation in our experience are the following, whereby downhill course plays an important role: rapid deterioration of the circulation; critical peripheral perfusion; metabolic acidosis; cardiac index less than 2.0 L · min-1 · m-2; mixed venous saturation less than 40%; signs of beginning renal and hepatic failure; patient on respirator with mounting FiO2; and massively impaired cardiac function as shown by echocardiography.
We strongly believe that, with more confidence in the system, these criteria will be applied less restrictively in the future. However, increased survival and the fact that more patients were treated with an LVAD alone, were extubated on the assist device, and did not require delayed sternum closure indicate the positive effects of employing these criteria. (It is our policy to close the chest immediately whenever possible, to control infection.)
Earlier decision-making for VAD support is most important in the post–cardiopulmonary bypass group. During period 1, in most cases this decision was made after protracted courses in the intensive care unit, often after repeated cardiopulmonary resuscitation. There were no long-term survivors. Now, VADs are already implanted during the initial operation when it becomes apparent that the cardiac function cannot be stabilized postoperatively.
In patients with chronic end-stage heart failure, measurement of levels of natriuretic peptides together with markers of inflammation may provide better prediction of the optimal time-point for VAD implantation, as has been shown in the adult population [34].
Some of the technical improvements have had great impact on the patient courses. For instance, the introduction of left apical cannulation in 1998, leading to superior ventricular drainage as compared with atrial cannulation, has made myocardial recovery more probable. Also, apical drainage has made it more likely that the circulation can be sustained with an LVAD alone in comparison with atrial drainage, in which case BVAD support more often became inevitable.
The design of the cannulas has been shown to be crucial for optimal arterial return. In the earlier period, cannulation of the great vessels, in particular the aorta, was done with an indwelling cannula, as for standard extracorporeal circulation. It was observed that this sometimes led to obstruction of the aorta, increasing the afterload for the natural heart, precluding any recovery and possibly contributing to pulmonary congestion. With the "press-button" cannulas introduced at the end of 1996, this problem was alleviated.
Anticoagulation and its monitoring remain a major problem despite significant progress in this field [6, 9, 35–38]. Heparin coating alone has not resolved all the problems. The anticoagulation regime we use has previously been described in detail by Stiller and colleagues [5, 6]. The strategy now employed in our institution is as follows: first, close anticoagulation monitoring with activated partial thromboplastin time instead of activated clotting time in the early postoperative period, with a target of between 60 s and 80 s is preferable. Second, thrombelastography helps to identify the coagulation status and the impact of heparin. Antithrombin III should also be closely monitored and substituted if the level falls below 70%. Later on, after treatment with aspirin and dipyridamol is initiated, platelet aggregation tests should be performed on a weekly basis with target activation of 30% [35]. Adolescents who were discharged home on a VAD received phenprocoumon with a target international normalized ratio of 3 to 3.5. Because the pump housing is translucent, thrombi can be detected at an early stage. In both periods, pumps were exchanged for significant thrombus formation that became visible. This rate has decreased to a certain extent during period 2.
The effects of the changes described here are seen in the cumulative sum analysis (Fig 6). This clearly shows the impact of changes in system design, implantation technique, patient selection, and postoperative care on survival.
In conclusion, earlier implantation of VADs, heparin coating of the blood pumps, and substantial modifications in cannula design, anticoagulation, and the coagulation monitoring regimen have led to a significant increase in the survival and discharge rate, especially among children less than 1 year of age. Now the pediatric size Berlin Heart Excor VAD offers a valuable option as a bridge to transplantation or recovery for children suffering from cardiogenic shock.
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
DR HETZER: Well, as far as I understand, this is still according to compassionate-use requests. And there are some efforts being made to introduce it as a humanitarian device.
DR CHRISTO I. TCHERVENKOV (Montreal, Quebec, Canada): I congratulate you for your pioneering efforts and for setting the gold standard with your large experience. I am very pleased to report that the 2-year-old patient we put on the Berlin Heart biventricular support 3 years ago is doing extremely well 2 and a half years after his heart transplant.
We recently implanted our third Berlin Heart at Montreal Children's Hospital in a 4-year-old boy who had a Norwood operation for hypoplastic left heart syndrome. He had undergone a bidirectional Glenn at around 6 months. At that time he had good ventricular function and almost no tricuspid regurgitation. Over the next 3 years he developed a moderate to severe tricuspid regurgitation. I repaired his tricuspid valve, reducing the tricuspid regurgitation from severe to mild. I decided not to complete the Fontan operation, leaving him with a bidirectional Glenn physiology. A month later, he had developed very severe right ventricular dysfunction and dilatation with mild tricuspid valve regurgitation. He was listed for transplantation, and after 3 weeks he was supposed to be intubated with a mixed venous saturation of 20%. After some deliberation and trying to get some feedback about experience with this kind of support in single-ventricle patients, 9 days ago we placed the Berlin Heart, with cannulation of the diaphragmatic surface of the right ventricle for outflow and neoaortic cannulation for the arterial inflow.
I would like to ask you a few questions with respect to the patient I just described. First is the issue of cannulation for outflow. Since you have a right atrium that drains both the pulmonary venous return and the inferior vena cava, should we have cannulated the right atrium in this patient as opposed to the right ventricle?
And the second point, I would like to know from your series what is your experience with the Berlin Heart in patients with single-ventricle physiology? We had a lot of trepidation in our patient as to how the bidirectional Glenn physiology would perform after the Berlin Heart implantation. We were relieved that our patient was extubated 36 hours postoperatively, and the oxygen saturation increased from 60% to 80% after the Berlin Heart was implanted. Of course, we are waiting for a transplant and the story is not finished yet.
DR HETZER: I must confess that I cannot give you any answer to that because we have not supported a Fontan patient with the system up to now. I think any kind of patient with a right-to-left or left-to-right flow of the shunt may not be really very suitable for such an assist device. There must be a completely separated circulation, also without any kind of regurgitation, as you mentioned, which, of course, can be created at the time of assist implantation. But we did not have any case like that.
I was speculating about this. And certainly we would at this time, at the time of implantation, do a modification or close any kind of regurgitant flow either in the vessels or across a shunt.
DR TCHERVENKOV: So you have no experience with the Berlin Heart in patients with single-ventricle physiology?
DR HETZER: Right. Exactly.
DR ALI DODGE-KHATAMI (Zurich, Switzerland): I congratulate you on your pioneering work with the Berlin Heart, as you've got the biggest overall experience with it. We just started with our experience in Zurich. I have three questions. First, concerning myocarditis or in other patients where you're doing this as a bridge to recovery and not to transplantation, what would be your weaning criteria to come off the Berlin Heart? Do you do it similar to an ECMO wean with stress echocardiograms?
And the second question would be, we've noticed in some small children in whom we would want to be having this Berlin Heart for quite a while, that these connectors are very long, and that the actual pumping chambers are in the pelvic area. Are we allowed to shorten these connectors so that this is a bit more comfortable for the child?
And the last question is, in your experience, you mentioned stroke. Have your complications been more with bleeding or with thromboembolism?
DR HETZER: The first question I think is very important. We have weaned altogether 12 of those patients, and out of those, there were 4 with acute viral myocarditis. And in the very first case, we were not sure whether we could just take the system out, you know, explant the system. So in this case, the first case, we reduced the flow and then we switched to ECMO for another 5 days; and this, of course, could be gradually reduced.
Now, according to echocardiography studies, when you can see that you have full contractility of the ventricle and also good wall thickness and thickness changes, we have now recently gone ahead and explanted the system. And in the viral myocarditis cases this was very fast. The support time was between 10 and 21 days only.
The second question—according to the cannula—of course, these cannulas can be shortened as to suitability and we do it all the time. However, in our opinion, it's very important to have this felt-covered part of the cannula crossing the skin. Because, of course, we agree then, when you go for longer term support, infection of the system is a major hazard.
And as for the third question, maybe my conclusions were not quite correctly formulated. We have seen thrombi in the pumps. The pump has the advantage that you can look into the pump and you can see even small deposits, in which case we would either clean the pump or exchange the pump. But there were only two significant strokes. But I must say that in the earlier period of our experience, of course, some of those children, especially after cardiopulmonary bypass, were in such a bad condition that they had brain damage for a variety of reasons, which also could have been attributed to stroke. Bleeding certainly has been much reduced with the present anticoagulation system.
DR DODGE-KHATAMI: On a long-term basis, would you have a tendency to overanticoagulate or underanticoagulate?
DR HETZER: I think in the past we have anticoagulated too early. Of course, in children in a more advanced condition of shock, there are spontaneous coagulation disturbances. But I think in the beginning, we have overanticoagulated those patients.
DR PEER M. PORTNER (Stanford, CA): Someone raised the issue of acquiring the pediatric Berlin Heart VAD for use in the United States. As you commented, we recently had a very good experience with this system at Stanford, successfully bridging a 7 kg infant to transplantation. It may be of interest to know that the process to acquire this device is not terribly burdensome. A compassionate-use request to the Institutional Review Board and then to the Food and Drug Administration, followed by Berlin Heart company contact and shipment were accomplished fairly quickly.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. Januszewska, E. Malec, J. Birnbaum, M. Loeff, R. Sodian, C. Schmitz, H. Netz, and B. Reichart Ventricular assist device as a bridge to heart transplantation in children Interactive CardioVascular and Thoracic Surgery, November 1, 2009; 9(5): 807 - 810. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Imamura, A. M. Dossey, P. Prodhan, M. Schmitz, E. Frazier, U. Dyamenahalli, A. Bhutta, W. R. Morrow, and R. D.B. Jaquiss Bridge to Cardiac Transplant in Children: Berlin Heart versus Extracorporeal Membrane Oxygenation. Ann. Thorac. Surg., June 1, 2009; 87(6): 1894 - 1901. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Nicolini and T. Gherli Alternatives to transplantation in the surgical therapy for heart failure Eur. J. Cardiothorac. Surg., February 1, 2009; 35(2): 214 - 228. [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 |
