|
|
||||||||
Ann Thorac Surg 2001;72:86-90
© 2001 The Society of Thoracic Surgeons
Accepted for publication April 5, 2001.
Address reprint requests to Dr Acker, Division of Cardiothoracic Surgery, 6th Floor Silverstein, Hospital of the University Pennsylvania, 34th and Spruce St, Philadelphia, PA 19104-4283
e-mail: macker{at}mail.med.upenn.edu
| Abstract |
|---|
|
|
|---|
Methods. A respective review of ECMO and VAD data registries was instituted.
Results. From May 1996 to July 2000, 23 patients were placed on ECMO support for profound cardiogenic circulatory failure. Eleven patients (47%) were withdrawn from support due to severe neurologic injury or multisystem organ failure. Three patients (13%) were weaned off ECMO with good outcome. Nine patients (39%) were transferred to a VAD. Two patients expired while on VAD support, and 7 of the VAD-supported patients (78%) survived to transplantation. Overall survival was 43%.
Conclusions. Emergent ECMO support is a salvage approach for cardiac resuscitation once conventional measures have failed. In neurologically intact patients, the early transfer to a VAD quickly stabilizes hemodynamics, avoids complications, and is essential for long-term circulatory support before definitive treatment with cardiac transplantation.
| Introduction |
|---|
|
|
|---|
ECMO circulatory support successfully maintains hemodynamic function, however the deleterious effects on the complement, renal, and pulmonary systems limit its long-term utilization. In addition, complications such as hemorrhage, pulmonary edema, and thromboembolic events limit its use. In adults, the application of medium- to long-term ECMO support has frequently been applied to maintenance of oxygenation in patients with severe respiratory failure rather than circulatory dysfunction. The development of newer univentricular and biventricular assist devices has provided a means to support hemodynamic function for several months before return of cardiac function or "bridge" to cardiac transplantation.
Patients presenting with severe cardiogenic shock have few options once conventional cardiopulmonary support has failed. Restoration of adequate circulatory function with conventional cardiopulmonary resuscitation (CPR) is limited in patients with intractable nonpulsatile ventricular tachycardia, or ventricular fibrillation (VT/VF). To restore circulatory stability in these patients, we have instituted the early aggressive application of ECMO. Once circulatory hemodynamics are restored and stabilized, long-term circulatory function is maintained with an early bridge to univentricular or biventricular assist device. Definitive treatment is achieved with cardiac transplantation. This report details the successful use of this "double bridge" to cardiac transplantation that utilizes both ECMO and VAD support for resuscitation of the "near dead" in intractable cardiogenic shock.
| Material and methods |
|---|
|
|
|---|
Treatment algorithm
Concurrent with failure of standard Advanced Cardiac Life Support (ACLS-CPR) resuscitation techniques, arterial-venous cannulation was initiated and patients were placed onto full ECMO support, most with CPR in progress (Fig 1). To minimize the delay in perfusion, percutaneous femoral cannulation was preferentially attempted. Exceptions included patients who were postcardiotomy where access to the great vessels could be obtained without delay.
|
Two ECMO circuits were used in our algorithm. If patients were cannulated outside of the operating room, the Bard Cardiopulmonary Perfusion System (CPS) mobile ECMO circuit was utilized. This consisted of arterio-venous lines, hollow-fiber membrane oxygenator (Bard, Haverhill, MA), and centrifugal pump (BP80; Biomedicus, Eden Prairie, MN). Patients who were placed onto ECMO in the operating room utilized our standard ECMO circuit (Baxter, Deerfield, IL), consisting of arterio-venous lines, solid silicone membrane oxygenator (Avecor, Minneapolis, MN), and centrifugal pump (BP-80; Biomedicus). Neither system utilized cardiotomy or venous reservoirs.
After ECMO cannulation early assessment of neurologic status and aggressive metabolic support was initiated (Fig 1). Neurologic status was ascertained with serial formal neurologic examination and computed tomography of the brain to assess for hypoxic injury or massive cerebral hemorrhage. Attempts to achieve negative fluid balance with hemofiltration and attention to pulmonary compliance were aggressively instituted in all patients. There were no attempts to decompress the left ventricle with venting or intraatrial shunt. Patients deemed neurologically intact, who demonstrated viable hepatic, renal, and pulmonary functions, were transferred to univentricular or biventricular support devices. Intraoperative transfer for ECMO to VAD was achieved in most cases with aorta-right atrial recannulation. In several cases, the venous ECMO cannula was retracted into the interior vena cava (IVC) when a right ventricular assist device (RVAD) was planned to facilitate right atrial cannula placement. In these cases, a superior vena cava cannula was added during cardiopulmonary bypass to facilitate venous drainage during placement of the RVAD. After placement of ventricular assist devices, patients underwent aggressive cardiac rehabilitation. Definitive long-term survival treatment was achieved after cardiac transplantation.
During ECMO support, several patients demonstrated return of intrinsic cardiac function and were weaned off of mechanical support (Fig 1). Criteria for weaning off ECMO included maintenance of cardiac index greater than 2.0 L/m2 and mixed venous saturation greater than 65% with the ECMO circuit flow turned down in the intensive care unit (ICU).
Data analysis
Data from our institution was collected in both a retrospective and concurrent manner. The tracking of each patients clinical course during ECMO, VAD, and posttransplant treatment was recorded. Vascular, neurologic, hemorrhagic, renal, and perfusion system complications were documented. Survival was defined as discharge from the hospital.
| Results |
|---|
|
|
|---|
The majority of patients (n = 15) presented with a VT/VF arrest (65%). Three of these patients were postcardiotomy, and 2 patients were postcardiac catheterization. The remaining 35% of patients were placed onto ECMO for cardiogenic shock after myocardial infarction. Two patients were transferred to our institution on full ECMO support. Eighteen of 23 (78%) patients had ongoing CPR during ECMO cannulation.
ECMO was established in the operating room (36%), surgical intensive care unit (22%), cardiac catheterization laboratory (17%), emergency department (12%), cardiac care unit (7%), and the nursing floor (4%). There were no differences in survival based on the location where ECMO support was established. Before the initiation of ECMO support, 9 patients had ongoing IABP counterpulsion to improve left ventricular function for cardiogenic shock.
Outcome
Twenty-three patients were placed on ECMO, and 11 patients (47%) were withdrawn from support due to severe neurologic injury (n = 4) or multisystem organ failure (n = 7). Three patients (13%) were weaned off ECMO with good outcome. The remaining 9 patients (39%) were transferred to a VAD. Two patients (8.6%) expired while on VAD support due to multisystem organ failure, and 7 of the VAD-supported patients (78%) survived to transplantation and were discharged home. Overall survival was 10 patients (43%) (Table 1).
|
|
Figure 2 represents duration of treatment and overall survival as a function of time on ECMO, VAD support, and time to transplant. By day 3, almost all patients had either been transferred to VAD or had been withdrawn from ECMO support. The largest change in mortality occurred on day 3. After this time, there was minimal change in survival. The majority of viable ECMO patients who did not recover intrinsic cardiac function were transferred to VAD in less than 24 hours. All of these patients were bridged to VAD by 3 days.
|
Of the 9 patients who were bridged to VAD support, 4 patients in this series had left ventricular assist device (LVAD) support only (Heartmate-Thermo Cardiosystems Inc, Woburn, MA), and 5 patients had biventricular assist device (BiVAD) implantation (4; Thoratec, Pleasantown, CA, and 1; Abiomed, Danvers, MA). The Heartmate was preferentially placed for left ventricular failure in patients large enough to accept the device. The Thoratec system was preferentially used for biventricular failure associated with multiorgan dysfunction. Decision for LVAD versus BiVAD support was based on clinical hemodynamic and echocardiographic evaluation of cardiac performance. Biventricular support was used in patients with sustained VT/VF arrest or severe multiorgan dysfunction with right ventricular failure. Right ventricular performance was evaluated by transesophageal echocardiography (TEE) with ECMO flows turned down in the ICU. The final decision to place a BiVAD was made intraoperatively using a combination of direct visualization of the heart with the chest open combined with intraoperative TEE and central venous pressure monitoring.
Few complications were encountered while patients were on ECMO or mechanical support. One patient developed oxygenator failure secondary to clot, which required interchange. One patient on VAD developed an intracerebral thromboembolic event, and 1 patient developed RVAD thrombus. While on VAD support, 3 of 9 patients developed renal failure. Two patients presented with transient acute tubular necrosis that required temporary hemofiltration. One patient developed progressive renal failure requiring hemodialysis and ultimately expired.
Early in this series, 8 patients developed lower extremity ischemia during ECMO perfusion. This complication was not encountered after the institution of ipsilateral limb perfusion. However, the large diameter of the arterial cannula frequently damaged the femoral artery, and the majority of patients required surgical reconstruction after decannulation to maintain patency.
| Comment |
|---|
|
|
|---|
After restoration of cardiac function, patients underwent an aggressive "early triage" for viability. Serial neurologic evaluations within 12 hours of institution of ECMO allowed for establishment of neurologic viability. In most cases, the decision to terminate support or proceed to long-term mechanical assist was made within 24 hours. This triage system allowed for inexpensive initial support for all patients. After a cardiac arrest, the immediate transfer to VAD would allocate expensive unnecessary resources for patients who would not have survived or been viable transplantation candidates. In addition, direct transfer to VAD would have prolonged critical ischemic time until perfusion could be reestablished and may have affected viability in these critically ill patients.
Three patients in this series were weaned directly off ECMO with restoration of good native cardiac function. Aggressive early ECMO support prevented multisystem organ failure and poor perfusion secondary to intractable ventricular arrhythmia. It allowed for ventricular recovery once intrinsic rhythm had been restored. Once native rhythm and function had been optimized, patients were decannulated and made good recovery.
During the period of ECMO support, aggressive metabolic support and fluid management were instituted. In the majority of cases, continuous hemofiltration was used to maintain negative fluid balance, and to avoid pulmonary hypertension. Ratcliffe and associates [19] have documented that ECMO support results in ongoing left ventricle dilatation, pulmonary hypertension, edema, and hemorrhage. Other investigators utilizing ECMO support for continued circulatory support have performed atrial septostomy to reduce these complications [18]. In our experience, aggressive fluid management combined with minimal ECMO time alleviated these complications. In our series, we avoided venting of the left ventricle in order to maintain a closed chest until the patient was deemed suitable for either definitive VAD treatment or decannulation.
Since 1993 at the University of Pennsylvania, 113 VAD devices have been placed. These include 68 LVAD, 39 BiVAD, and 6 RVADs. Seventy-three percent of the LVADs, 46% of BiVADs, and 0% of RVADs have survived until transplantation. Five of the LVAD, 5 of the BiVAD, and 2 RVAD patients have been explanted with good recovery. The success to transplantation in this series of patients initially stabilized with ECMO is comparable with our overall survival once patients have been bridged to VAD (7 of 9, 78%).
Pagani and associates [18] have reported similar experiences using combination ECMO and VAD as an early resuscitative means to stabilize circulatory hemodynamics until transplantation could be achieved. In their series, 14 of 32 patients were transferred directly to ECMO for cardiac arrest (50%) or severe hemodynamic instability. Seven of these patients were transferred to VAD and 6 survived to transplant. One patient was bridged directly to transplant. In their series, the 1-year actuarial overall survival was 43% and the 1-year actuarial survival from the time of LVAD implant was 71%. Our series demonstrates similar outcomes for overall survival and VAD survival. However, several patients in our series were weaned directly off ECMO after restoration of hemodynamic stability. The duration for ECMO and VAD support in their series was 6 ± 5 and 124 ± 97 days, respectively. Cannulation was performed percutaneously in 50% of this series, and carotid artery and right internal jugular cannulation was used in the remainder. In addition, 5 of the 14 patients placed onto ECMO underwent atrial septostomy for left ventricular decompression.
The majority of patients in our series were placed onto ECMO while CPR was in progress for a ventricular tachycardia or fibrillation arrest (78%). We have found that percutaneous cannulation was technically feasible and allowed for continued CPR until adequate perfusion could be initiated. In patients who survived VAD support, our mean ECMO support time was 3.4 ± 0.5 days. The early transfer to VAD was performed to limit ECMO support and minimize complications. Because of aggressive ventilator and fluid management, we did not find that left ventricular decompression was necessary to maintain adequate perfusion or to limit pulmonary congestion in our series.
In this series, we experienced few complications that were associated with ECMO or prolonged mechanical assist. We contribute this success to the early transfer to ventricular devices. Our initial experience with distal limb ischemia has been circumvented by providing parallel perfusion to the distal extremity.
This treatment algorithm is an aggressive method for the stabilization and salvage of critically ill patients who have failed standard resuscitation techniques secondary to poor myocardial function or intractable ventricular arrhythmia. Utilizing early "on-site" cannulation and the early switch to VAD, we have reduced expected complications and increased survival in this critically ill group of patients. The early triage system avoids overutilization of resources, and effectively separates patients who would not be viable transplantation candidates. We advocate more aggressive application of this treatment protocol for patients who would otherwise not have survived.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
M. Tajik and M. G. Cardarelli Extracorporeal membrane oxygenation after cardiac arrest in children: what do we know? Eur. J. Cardiothorac. Surg., March 1, 2008; 33(3): 409 - 417. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. Hoefer, E. Ruttmann, G. Poelzl, J. Kilo, C. Hoermann, R. Margreiter, G. Laufer, and H. Antretter Outcome evaluation of the bridge to bridge concept in patients with cardiogenic shock. Ann. Thorac. Surg., July 1, 2006; 82(1): 28 - 33. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. G. Leshnower, T. G. Gleason, M. L. O'Hara, A. Pochettino, Y. J. Woo, R. J. Morris, T. J. Gardner, and M. A. Acker Safety and Efficacy of Left Ventricular Assist Device Support in Postmyocardial Infarction Cardiogenic Shock Ann. Thorac. Surg., April 1, 2006; 81(4): 1365 - 1371. [Abstract] [Full Text] [PDF] |
||||
![]() |
U. Mehlhorn, M. Brieske, U. M. Fischer, M. Ferrari, P. Brass, J. H. Fischer, and H.-R. Zerkowski LIFEBRIDGE: A Portable, Modular, Rapidly Available "Plug-and-Play" Mechanical Circulatory Support System Ann. Thorac. Surg., November 1, 2005; 80(5): 1887 - 1892. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. P. Anstadt and J. E. Lowe Cardiopulmonary Resuscitation Card. Surg. Adult, January 1, 2003; 2(2003): 471 - 494. [Full Text] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |