| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2005;80:15-21
© 2005 The Society of Thoracic Surgeons
a Division of Cardiothoracic Surgery, Doernbecher Children’s Hospital, Oregon Health and Science University, Portland, Oregon
b Department of Pediatric Cardiology, Doernbecher Children’s Hospital, Oregon Health and Science University, Portland, Oregon
Accepted for publication February 1, 2005.
* Address reprints to Dr Ungerleider, Doernbecher Children’s Hospital, Oregon Health Sciences University, 3181 SW Sam Jackson Park Rd, DC 8 South, Portland, OR 97239 (Email: ungerlei{at}ohsu.edu).
Presented at the Fiftieth Annual Meeting of the Southern Thoracic Surgical Association, Bonita Springs, FL, Nov 13–15, 2003.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
METHODS: Between January 2001 and October 2003, 671 patients underwent surgery for congenital heart disease at our institution. We retrospectively reviewed the hospital and clinic charts of the patients who required extracorporeal life support postoperatively, and studied the factors associated with survival.
RESULTS: Thirty-six patients (5.36%) received extracorporeal life support after surgery, between 1 day and 8 years of age (age < 30 days, n = 34). We divided the patients into four groups. Group 1 consisted of 13 patients who were electively placed on ventricular support without an oxygenator (univentricular assist device) after repair of single-ventricle disease. Group 2 consisted of 16 patients who required extracorporeal membrane oxygenation after surgery for failed hemodynamics. Group 3 consisted of 2 patients who required left ventricle support (left ventricular assist device) after surgery for two-ventricle disease but who did not require biventricular (extracorporeal membrane oxygenation) support. Group 4 consisted of 5 patients who required conversion from ventricular assist device to extracorporeal membrane oxygenation. Overall, 28 patients were weaned successfully (78%), and 24 survived to discharge (67%). Hospital survival in groups 1, 2, 3, and 4 was 100%, 50%, 100%, and 20%, respectively. Univariate factors associated with survival were age, weight, ventricular assist device type, duration, single-ventricle disease, reexploration, number of complications, and specific complications such as sepsis, renal failure, and pulmonary failure.
CONCLUSIONS: Extracorporeal life support utilization was expanded to include different applications with different outcomes. The extracorporeal life support registry should be altered to reflect those changes.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
Extracorporeal life support is used in congenital heart surgery for several indications, including failure to wean from cardiopulmonary bypass (CPB), hemodynamic collapse before or after surgery, pulmonary hypertension, as a bridge for transplantation, and finally, in some centers including ours, routinely after surgical repair of hypoplastic left heart syndrome (HLHS) and its variants after successful weaning from CPB.
The indications, timing, and threshold for utilization of ECLS are different in many centers, as is the type of support chosen between different programs. Hospital survival is variable depending on the support required. In addition, multiple other factors that have been identified in previous studies affect the outcome of patients after ECLS.
The results of ECLS from different programs are combined together at the ECLS registry database, and the outcomes of different types and indications of ECLS are mixed together. We propose to separate the outcomes of various types and applications of ECLS to predict results more accurately and help toward establishing useful guidelines. We summarize our experience with ECLS in the last 2 years after repair of congenital heart disease. Our purpose is to identify factors affecting survival, including the various types and indications of support postoperatively.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
We divided the patients in our study into four groups on the basis of the type of ECLS required. Group 1 included 13 patients who received elective mechanical support after repair of single-ventricle disease (eg, HLHS). The shunt was left open and the lungs remained ventilated; thus, an oxygenator was not used. After completion of surgical repair, patients in group 1 were rewarmed, separated from CPB, and treated with a period of modified ultrafiltration. Atrial and neoaortic cannulas were then attached to a ventricular assist device (VAD) circuit, and the flow rate was slowly increased to 200 mL·kg–1·min–1. Protamine sulfate was administered, and hemostasis was achieved. Sternums were left open and covered with Eshmark dressing, and the patients were transported to the ICU. Heparin intravenous infusion was restarted in the ICU after several hours once hemostasis was complete. Activated clotting time was maintained between 160 and 180 seconds. The systemic-to-pulmonary shunt was left open in all cases [3]. The patients were ventilated to achieve an arterial oxygen partial pressure between 30 and 45 mm Hg and an arterial carbon dioxide partial pressure between 35 and 45 mm Hg. Once the systemic perfusion by physical examination and the lactate level became normal (<2 mmol/L), VAD flow rate was weaned to zero, and the cannulas were removed. Chests were closed in the ICU. If chest closure resulted in significant elevation of the central venous pressure, it was delayed for 1 to 2 days after decannulation.
Group 2 included 16 patients who required extracorporeal membrane oxygenation (ECMO) for hemodynamic or pulmonary support after surgery for single-ventricle or double-ventricle defects. Group 3 was confined to 2 patients in whom ECLS was used to support the left ventricle (left VAD) after repair of double-ventricle anatomy. Left VAD was provided for group 3 patients using a roller pump with flows maintained at 100 mL·kg–1·min–1, or at whatever level return to the left atrium allowed. After hemostasis was adequate, heparin was administered in the ICU to keep activated clotting time between 180 and 200 seconds, sternums were left open, and the incision was covered with Eshmark. Flows were weaned to zero, and the patients had their cannulas removed and chests closed in the ICU when the left ventricle recovered. Finally, group 4 included 5 patients who needed conversion from VAD to ECMO by adding an oxygenator to the ECLS circuit postoperatively (VAD-ECMO) for either single-ventricle or double-ventricle anatomy.
In the ECMO groups (2 and 4), the cannulation sites were the right or common atrium for venous outflow and the aorta or neoaorta for arterial inflow. Atrial and aortic cannulas were then attached to an ECLS circuit, which consisted of a roller pump and membrane oxygenator. The flow rate ranged between 100 and 200 mL·kg–1·min–1 depending on whether the patient had a patent shunt with systemic and pulmonary circuit in series (100 mL·kg–1·min–1) or parallel (200 mL·kg–1·min–1). Sternums were left open and covered with Eshmark. Heparin was administered to maintain activated clotting time between 180 and 220 seconds. Weaning was done by decreasing flow rates, institution of inotropic support, and adjusting ventilation with observation of the hemodynamics, saturations, acid-base status, and perfusion and echocardiographic findings. The patients had their cannulas removed in the ICU, and the chest was closed at a later date if needed in the ICU as well.
We performed data analysis using SPSS for Windows, version 11.0 (SPSS, Inc, Chicago, IL). Frequency distributions were analyzed for all data. Normally distributed continuous data were expressed as mean ± standard deviation. Categorical data were expressed as counts and proportions. Unrelated two-group comparisons were performed with unpaired, two-tailed Student’s t tests for means of normally distributed variables with Wilcoxon rank sum tests for skewed data. The 
2 or Fisher exact test was used to analyze differences among the categorical data.
Bivariate comparisons were then made between the following variables and 30-day mortality, our main outcome measure using the 2 test: age less than 10 days, weight less than 3.0 kg, single-ventricle versus double-ventricle disease, ECLS duration less than 180 hours, ECLS type, need for reexploration, and number and type of complications. Because of the small number of patients, the power to detect a significant difference is limited, and a multivariate analysis was not performed.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
The most common preoperative diagnosis was HLHS variant in 16 patients (44%). Other common diagnoses included variants of transposition of the great arteries in 5 patients, 3 patients with variants of double-outlet right ventricle, 2 patients with tricuspid atresia, 2 patients with tricuspid atresia and transposition of the great arteries, 2 patients with total anomalous pulmonary venous connection, 2 with unbalanced atrioventricular canal, and 4 other causes including Ebstein anomaly, mitral valve deformity, tetralogy of Fallot, and a patient after the Glenn procedure for a single ventricle.
Indications for ECLS support included failure to wean from bypass in 17 patients (47%), cardiac arrest in the ICU after surgery in 4 patients (11%), and finally routine use of ECLS after successful weaning from CPB in 15 (42%) patients. Standard ECLS indications (failure to wean from CPB and cardiac arrest) were present in 21 patients (3.12% of total patients after cardiotomy).
Extracorporeal life support was initiated in the operating room in 32 patients (88.9%) and in the pediatric ICU in 4 patients (11.1%). In addition, 1 patient required ECMO for the second time after weaning from ECLS that was started in the operating room the first time. Table 1 shows the survival of patients in relation to their preoperative diagnosis.
|
|
|
Complications recorded in our series included arrhythmias requiring intervention in 16% of our patients, renal failure requiring the addition of a hemofilter in 22%, clinical sepsis in 25%, pulmonary hemorrhage and failure in 19%, bleeding requiring reexploration in 25%, intracranial hemorrhage in 8%, gastrointestinal bleeding in 5%, mechanical problems requiring changing parts of the ECLS circuit in 14%, mediastinitis in 3%, and finally 1 patient who died as a result of heparin-induced thrombocytopenia complications. Table 2 summarizes these complications.
|
|
|
Reexploration for bleeding alone was not a significant factor affecting survival; however, reexploration for surgical reintervention (residual defect, shunt revision, and so forth) affected survival significantly. Table 3 shows the effect of multiple variables on survival.
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
The results for cardiac ECLS in neonates and children are inferior to those of pulmonary ECLS in the same age group [2, 19]. In the most recent ECLS registry database (July 2003), the survival rate for neonates and children after cardiac ECLS was 37% and 42%, respectively, compared with survival of 77% and 55% after respiratory ECLS in neonates and children, respectively.
Several factors that influence survival have been identified in previous reports [4, 8, 11–14]. Those include patient factors such as age, weight, and diagnosis (eg, ALCAPA [anomalous origin of left coronary artery from pulmonary artery], HLHS). Some studies found single-ventricle disease to be associated with worse prognosis whereas others found that biventricular disease patients had lower survival rate [11, 12].
Patients requiring ECLS without the use of an oxygenator (VAD) as in groups 1 and 3 had fewer complications compared with patients requiring ECMO and the use of an oxygenator. The presence of an oxygenator increases the foreign body surface area of the system, resulting in increased activation of the inflammatory cascade, consumption of coagulation factors, hemolysis, and generation of thromboemboli, all contributing to end-organ damage [3, 20, 21].
In addition, when the circuit does not include an oxygenator, it is possible to postpone heparinization to achieve hemostasis. Anticoagulation can be provided several hours after surgery, once hemostasis seems to be secure [22]. This significantly decreases bleeding requiring reexploration, as noted in our series. Only 1 patient in groups 1 and 3 required reexplorations for bleeding compared with 8 patients in groups 2 and 4. The decreased bleeding rate and therefore the transfusion requirements, consumption of coagulation factors, and tissue edema probably contribute to improved outcome in the VAD groups.
Finally, as noted in laboratory and clinical series, maintaining the shunt open in the univentricular shunt-dependent group (versus the formerly recommended protocols of occluding the shunt) maintains the perfusion of the lungs and decreases lung injury after ECLS and therefore pulmonary complications after ECLS [22, 23].
Most of the studies showed that prolonged duration of ECLS and emergence of complications, especially renal failure and sepsis, were all factors that have a significant negative impact on survival. Some studies have reported that the institution of ECLS in the operating room for failure to wean from CPB was associated with better survival [11], whereas others reported that patients who had a brief period of hemodynamic stability before the need for ECLS had better prognosis [4, 12, 16]. The difference in the outcomes of all those studies reflects the different practices of ECLS among different centers and also the disadvantages of single-center retrospective studies, including lack of randomization and small sample sizes.
Our experience with applying ECLS for different requirements provides some explanation of the variable results and conclusions reported in previous studies. It seems clear that when ECLS is provided to support ventricular function without the need for an oxygenator, prognosis in terms of complications and hospital survival are significantly better than when patients require biventricular support and the addition of an oxygenator.
This is not intended to simply be a review of VAD versus ECMO. In 1998, Thuys and colleagues [24] reported only moderate success with VAD after cardiac surgery for infants, and their results were not substantially different from results from other applications of ECLS. The number of patients who required VAD in our series (group 3, 2 patients) is too small to make conclusions about the efficacy of VAD versus ECMO.
Patients in our group 1 (VAD applied electively for HLHS and its variants) have been reported elsewhere and constitute a unique group. If we exclude them from our analysis, then our results would be very similar to other reported studies with 15 of 23 (65%) patients being weaned and 11 of 23 (48%) patients surviving to discharge. Nevertheless, this group represents a group of patients who may confound database information as the use of ECLS to provide elective hemodynamic support after cardiac surgery expands in the future.
In the 1970s, the use and indications of intrathoracic balloon pump to support failing adult hearts was expanded as it was applied to more "salvageable" patients. of Intrathoracic balloon pump became an often-used and successful part of the cardiac surgery toolbox.
We believe that application of circulatory support after neonatal cardiac surgery has numerous benefits for patients that may not be measurable simply as survival. Brain recovery after CPB exposure in hypoxic patients (such as is typical after a Norwood procedure for HLHS) is flow-dependent, and the cardiac output support from routinely applied postoperative VAD may be beneficial. Our question is whether it confounds the ECLS database to lump these patients, who have a clearly different outcome than patients exposed to more conventional applications of ECLS, into the same data set.
This study also emphasizes that the prognosis for infants with single-ventricle disease who can be managed with an open shunt (group 1) and normal ventilation is substantially different from infants with single-ventricle disease who despite an open shunt require an oxygenator to support pulmonary function.
Finally, we present these data to illustrate that application of ECLS is changing. There are patients who can be managed with ECLS electively or semielectively (failing but not failed hemodynamics) who can have short (<180 hours) duration of ECLS and consequently fewer complications. When ECLS becomes prolonged, more complications ensue, and the prognosis for survival is markedly worse. These patients comprise a different population with respect to outcome and possibly should be recorded in some separate way.
The ECLS registry plays an excellent role in combining results from different, experienced centers and sharing results in terms of weaning and survival rates after surgery for different diseases, indications, and procedures. In addition, it provides information about the expected ECLS complications. It does not differentiate among different types of ECLS, timing, and clinical scenarios leading to ECLS. Combining the data from different institutions, taking into consideration the various applications of ECLS described in this review, would reflect more accurately the clinical course and outcome of cardiac support and may help in establishing guidelines and protocols that can be applied in diverse medical centers with reproducible results and therefore may help with better utilization of expensive resources.
The role of ECLS after pediatric cardiac surgery is expanding to include new and different applications. Its role is not limited to the support of children with refractory cardiac failure after repair of congenital heart disease. There are elective and semielective indications for the institution of ECLS postoperatively that influence not only patients’ survival but also recovery and neurologic outcome. Hospital survival after ECLS is variable depending on the patient’s condition and the type of support required. Patient factors as age, weight, and disease may affect outcome. The duration of support required exposes the patients to increased risk of complications and affects their outcome. Survival after VAD is better than survival after ECMO, and this is especially true when ECMO follows conversion from VAD.
It is extremely helpful to accumulate data from different centers and to combine the experience of different programs to learn more about the applications of ECLS and predict the outcome of patients requiring this support. However, national database entry could be altered to reflect different applications of ECLS. Doing so will help provide guidelines for the use of ECLS and better utilization of resources.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
I think you concluded that the result depended upon the type of support. I would challenge that a bit and say that the type of support depended upon what was required in the patient. For example, my suspicion is that the patients on this system without the oxygenator and the patients requiring only left ventricular assist devices may have had left ventricular failure without severe biventricular failure, and, of course, those are always the ones most likely to survive in any such series.
The other difference, of course, between this series and one you might have heard here 5 years ago is that you now have a large number of hypoplastic left heart patients. It wasn’t clear to me exactly the mortality rate in that group and whether that made a difference in terms of overall improvement in results.
I thought this was a great paper.
DR ALSOUFI: Thank you, Doctor Pennington. I agree with you that those who required extracorporeal membrane oxygenation (ECMO) probably are patients who have biventricular failure or combined cardiac and respiratory failure, and those patients are sicker to begin with. However, we don’t think that this is the only reason why patients requiring ventricular assist device (VAD) alone without the oxygenator have a better prognosis than patients with ECMO. We believe that ECMO patients have more complications compared with the VAD patients, and, as I mentioned, the number of complications affected the survival.
We believe that omitting the oxygenator will help us, first, to wait a few hours without starting the anticoagulation until hemostasis is achieved, thus decreasing the complication of bleeding. In addition to that, we use smaller circuits with less foreign surface and there is no oxygenator. This will help decrease the inflammatory cascade, the activation of coagulation factors, and thromboemboli. We believe that will decrease the end-organ damage that can happen in patients with extracorporeal life support (ECLS).
Doctor Ungerleider’s group has proved in the past that leaving the shunt open will lead to excellent survival, and allows the patient’s lungs to act as their own "oxygenator." Following a Norwood operation, the patients who had the shunt open with omission of the oxygenator had excellent survival that approaches 90% currently. So we believe that the oxygenator itself and the ECMO circuits do have their own share of complications and they probably will affect survival, but I do agree with you, in most collected series, those patients are also sicker to begin with.
DR JAMES A. QUINTESSENZA (St. Petersburg, FL): Excellent presentation, and your group obviously is to be congratulated for your expertise with small children, especially single-ventricle babies and their ventricular support. The question I have is regarding renal failure in these patients.
I think your data show that there is an extremely high mortality once renal failure occurs, and that has been our experience as well. Can you give us a little insight into the way you manage renal failure or your use, for example, of ultrafiltration on the VAD or on the ECMO circuits and what kind of tricks you might have to prevent renal failure? Very nice talk.
DR ALSOUFI: Well, we do use the VAD in the beginning to optimize the cardiac output following surgery and we wean the VAD once we have the lactic acid level below 2 mmol/L. As end-organ perfusion improves, we normally see adequate renal function, especially in patients who have not had periods of hypoperfusion preceding the application of ECLS. If the patient has adequate perfusion on ECLS and there is persistent renal dysfunction, we may start a hemofilter to remove fluids or normalize electrolytes while nutrition is being increased. In patients with transient renal dysfunction, the VAD support of cardiac output seems to improve renal perfusion, and renal dysfunction is ordinarily not a problem.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  V. Bautista-Hernandez, R. R. Thiagarajan, F. Fynn-Thompson, S. K. Rajagopal, D. E. Nento, V. Yarlagadda, S. A. Teele, C. K. Allan, S. M. Emani, P. C. Laussen, et al. Preoperative extracorporeal membrane oxygenation as a bridge to cardiac surgery in children with congenital heart disease. Ann. Thorac. Surg., October 1, 2009; 88(4): 1306 - 1311. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. C. Oliver Anticoagulation and Coagulation Management for ECMO Seminars in Cardiothoracic and Vascular Anesthesia, September 1, 2009; 13(3): 154 - 175. [Abstract] [PDF]
|
|
|
|
 |
|
 B. Alsoufi, O. O. Al-Radi, C. Gruenwald, L. Lean, W. G. Williams, B. W. McCrindle, C. A. Caldarone, and G. S. Van Arsdell Extra-corporeal life support following cardiac surgery in children: analysis of risk factors and survival in a single institution Eur. J. Cardiothorac. Surg., June 1, 2009; 35(6): 1004 - 1011. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D Picarelli, C. Kreutzer, M Barboza, S Antunez, G Pose, G Touya, L Liguera, D Abdala, and G Echegaray Post-cardiotomy circulatory support with the Terumo Baby - Rx tm oxygenator in a newborn Perfusion, November 1, 2007; 22(6): 377 - 379. [Abstract] [PDF]
|
|
|
|
 |
|
 B. Alsoufi, O. O. Al-Radi, R. I. Nazer, C. Gruenwald, C. Foreman, W. G. Williams, J. G. Coles, C. A. Caldarone, D. G. Bohn, and G. S. Van Arsdell Survival outcomes after rescue extracorporeal cardiopulmonary resuscitation in pediatric patients with refractory cardiac arrest. J. Thorac. Cardiovasc. Surg., October 1, 2007; 134(4): 952 - 959.e2. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. K. Allan, R. R. Thiagarajan, P. J. del Nido, S. J. Roth, M. C. Almodovar, and P. C. Laussen Indication for initiation of mechanical circulatory support impacts survival of infants with shunted single-ventricle circulation supported with extracorporeal membrane oxygenation J. Thorac. Cardiovasc. Surg., March 1, 2007; 133(3): 660 - 667. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. L. Hannan, J. W. Ojito, M. A. Ybarra, M. C. O'Brien, A. F. Rossi, and R. P. Burke Rapid Cardiopulmonary Support in Children With Heart Disease: A Nine-Year Experience Ann. Thorac. Surg., November 1, 2006; 82(5): 1637 - 1641. [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 |
