Asian Cardiovasc Thorac Ann 2005;13:38-41
© 2005 Asia Publishing EXchange Ltd
Postoperative Mechanical Circulatory Support with Biomedicus Centrifugal Pump
Murat Mert, MD,
Atif Akcevin, MD1,
Cenk E Yildiz, MD,
Kaya Suzer, MD
Department of Cardiovascular Surgery, Institute of Cardiology, Istanbul University
1 Department of Cardiovascular Surgery, Istanbul American Hospital, Istanbul, Turkey
For reprint information contact: Murat Mert, MD Tel: 90 532 231 6666 Fax: 90 216 435 8600 Email: mmert{at}superonline.com, Ortaklar Cad. Kantasi apt 47/3 daire 4, Mecidiyekoy - 80290, Istanbul, Turkey.
 |
ABSTRACT
|
|---|
Despite advances in surgical techniques, myocardial protection, and management protocols, approximately 1% of patients undergoing open heart operations still need mechanical circulatory support for severe cardiac dysfunction. The Biomedicus centrifugal pump, available in most cardiovascular centers, is a highly effective and relatively inexpensive system compared to other more sophisticated devices for the same purpose. Of 10 patients aged 5 to 61 years who were supported for 22 to 168 hours with a Biomedicus centrifugal pump, 7 (70%) were weaned from support, there was one hospital death, and 6 patients were discharged from hospital. Two sudden deaths occurred in the first 8 months after discharge. Four patients (40%) were still alive after follow-up of 11–55 months, with no restriction in their daily activities. The centrifugal pump is a very cost-effective support system with survival rates comparable to those of more sophisticated devices in short-term ventricular assistance.
|
INTRODUCTION
|
|---|
Despite improvements in surgical techniques, methods of myocardial protection, and patient management protocols, post-cardiotomy cardiac failure occurs in 2%–6% of patients, and 1% of them will require mechanical circulatory support with a cardiac assist device in addition to an intraaortic balloon pump (IABP).1 Of the various ventricular assist devices presently available, the centrifugal pump has been used most often.2 This may be due to the fact that centrifugal pumps are available in most cardiovascular surgery centers and are relatively simple and inexpensive compared to other mechanical assist devices. The Biomedicus Bio-Pump (Medtronic, Inc., Eden Prairie, MN, USA) is the most widely used centrifugal pump for cardiac assistance. We present our clinical experience with this device as a short-term means of ventricular (or ventricular and pulmonary) support after cardiac surgery.
|
PATIENTS AND METHODS
|
|---|
Between 1998 and 2002, of 3,600 patients operated under cardiopulmonary bypass (CPB), 10 (0.28%) were supported with a Bio-Pump. The mean age of the supported patients was 42.04 ± 18.37 years (range, 5 to 61 years) and the male/female ratio was 8/2. The Bio-Pump was used as a left ventricular (LV) assist device in 8 patients, as a right ventricular (RV) assist device in 1 patient, and with the addition of an oxygenator for biventricular assist and extracorporeal membrane oxygenation in 1 patient. Support was commenced in the operating room in 8 cases because of inability to wean from CPB, and in 2 patients in the intensive care unit (ICU) because of acute hemodynamic deterioration. All but 3 of the patients (1 with valve disease and 2 with congenital defects) were supported with an IABP prior to centrifugal pump assistance. Patient details are summarized in Table 1
. The Bio-Pump was used after coronary artery bypass grafting (CABG) in 6 patients, after combined aortic valve replacement and CABG in one, after mitral valve replacement in one, and after surgical correction of congenital pathologies in 2.
Among the adults who required support after CPB, one was moved from the catheterization laboratory directly to the operating room under resuscitation and IABP support after failed percutaneous transluminal angioplasty due to left anterior descending coronary artery dissection. Three others with very recent myocardial infarction underwent emergency CABG after thrombolytic therapy, because of deterioration in hemodynamic status and lack of response to medical therapy. In these patients, weaning from CPB was impossible despite full inotropic and IABP support, possibly due to the loss of all myocardial energy stores in the preoperative ischemic period. Centrifugal pump support was instituted to allow the myocardium to recover following complete revascularization. A patient who had undergone CABG and resection of a large LV aneurysm could not be weaned from CPB due to hypokinesis of all segments of the heart. It was considered that this global hypokinesia was due to the sudden reduction of the LV cavity, and the patient might be weaned if the resting segments of his ventricles could recover during a period of support. After atrioventriculoplasty for Ebstein’s anomaly, transesophageal echocardiography showed severe tricuspid regurgitation, so the tricuspid valve was replaced with a 33 mm mechanical prosthesis. As the patient could not be weaned from CPB and RV contractions were very poor at this stage, RV assistance was started. After surgery for ventricular septal defect closure and pulmonary stenosis, a patient with corrected transposition of the great arteries was taken back to the operating room with severe pulmonary regurgitation. The pulmonary valve was replaced but the patient could not be weaned from CPB. Transesophageal echocardiography did not show any additional cardiac pathology that could explain the global hypokinesia, so LV support was initiated. In a young woman who had undergone mitral valve replacement, LV assistance was decided on due to deterioration of her hemodynamic status in the ICU. Her low cardiac output was attributed to inadequate myocardial protection during surgery, as retrograde cardioplegia had been used, and it was thought that her heart might recover after a period of support. Another patient who was weaned from CPB with IABP support after CABG and LV aneurysm resection, deteriorated in the ICU possibly due to acute dilatation of the resting ventricular cavity following volume reduction surgery. Support was instituted to relieve acute volume overload on the resting LV cavity. In both of these patients, the decision for support was made after transesophageal echocardiography.
Left ventricular support was achieved by cannulating the left atrium via the right superior pulmonary vein for inflow to the pump, and the ascending aorta for pump outflow. In one patient, RV support was implemented by cannulating the right atrium and the main pulmonary artery. In the single patient who had biventricular support and extracorporeal membrane oxygenation, the right atrium and ascending aorta were cannulated for inflow and outflow, respectively. Inflow was via a 32F right-angled metal-tipped cannula in the left atrium and a 32F two-stage cannula in the right atrium; 26F standard aortic cannulas were used for outflow. The sternum was left open in all cases in order to intervene quickly in the event of pump failure, a cannula problem, or acute bleeding, and to eliminate the possibility of myocardial compression by the sternum. The subcutaneous tissues and skin were closed in the usual manner.
All patients were ventilated using a volume-controlled ventilator with a tidal volume of 12 mL·kg–dose inotropic drugs were used to maintain contractility. An IABP was kept in all but 3 of the patients, to reduce afterload, improve coronary perfusion, and add a pulsatile component to the circulation. The RV and LV filling pressures were continuously monitored with a Swan-Ganz catheter inserted via the right internal jugular vein and with a left atrial catheter inserted via the right superior pulmonary vein. Filling pressures of 14–18 mm Hg were considered optimal (mean pressure in each atrium was monitored). On initiation of support, pump flow was regulated to maintain the mean arterial pressure between 70 and 80 mm Hg. The patients were anticoagulated with continuous heparin infusion to maintain a partial thromboplastin time in the range of 50–70 seconds. Continuous diuretic or insulin infusions were started in the event of decreasing urine output or hyperglycemia. In the last 6 patients, venovenous hemofiltration via the femoral vein was instituted to reduce the third-space fluid.
Formal weaning trials were not usually attempted during the first 24 hours of support. During weaning trials, pump flow was gradually reduced and cardiac function was monitored by bedside echocardiography and filling pressures. If evidence of poor ejection was noted (poor contractility on echocardiography and ventricular filling pressures > 20 mm Hg, cardiac index < 2.0 L·min–;m–2, mean arterial pressure < 60 mm Hg), pump flow was promptly restarted. The dosage of inotropic drugs was increased when pump flow was reduced to 2 L·min–1; pump flow was kept at 1.5 L·min–1 for an additional hour before the decision to terminate support. If hemodynamically stable during this period, the patient was transferred to the operating room for device removal. Pump flow was stopped and the heart was inspected for a brief period before removal of the cannulas. The inflow cannula was removed before the outflow cannula. During weaning, the anticoagulation regimen was not increased if the partial thromboplastin time was within the prescribed range.
|
RESULTS
|
|---|
The mean duration of the support was 92.4 ± 47.71 hours (range, 22 hours to 7 days). The 7 patients (70%) who could be weaned from the device had a mean support duration of 89 ± 57 hours. The mean duration was 99.5 ± 47.7 hours in 3 patients who could not be weaned; a decision to terminate support was made after 64, 70, and 168 hours (Table 1
). Support was terminated because of severe hemodilution in one of these patients (no. 3), because of lack of improvement in LV contraction on periodic echocardiography in another (no. 4), and because of severe sepsis leading to multiorgan failure in the 3rd patient (no. 10).
One of the 7 weaned patients expired 7 days later. This patient had been admitted from the catheterization laboratory under resuscitation, he was weaned from the ventricular assist device after 96 hours, however he had a severe neurologic sequela possibly due to a hypotensive period preoperatively. It was not possible to wean him from the ventilator and he died from sepsis and multiorgan failure. The other 6 patients (60%) were discharged from the hospital in good condition. There were 2 late, sudden deaths (20%). The patient with corrected transposition of the great arteries died after 3 months, and the patient with Ebstein’s anomaly died after 8 months. Four of the weaned patients were still alive and in New York Heart Association functional class I after a mean follow-up period of 42.5 ± 22.0 months. All have returned to work and have no limitations in their daily activities.
|
DISCUSSION
|
|---|
Temporary cardiac support will continue to be required for management of cardiac failure after cardiac surgery. The ideal strategy for short-term cardiac assistance should combine low cost, adaptability to diverse applications and patient requirements, and rapid and easy deployment.3 In our experience, most of the assisted patients (7/10) had been treated for coronary artery disease, however, we also used the Bio-Pump in 2 patients with congenital heart disease and one with valvular heart disease. Although the 2 late deaths occurred in this group, we were able to wean all 3 of these patients, thus mechanical support after cardiac operations should not be reserved for coronary artery disease patients. The current challenges are to determine which patients will benefit from support and when to begin it, as no hospital can afford to provide assist devices in all cases of failure after cardiac surgery. Patients selected for cardiac support should primarily have a technically satisfactory operation with good preoperative ventricular function and no pre-existing organ dysfunction. It is very important to differentiate irreversible myocardial damage from hibernating or stunned myocardium, for a successful outcome of support. In all but one case, myocardial hibernation or stunning was thought to be the cause of cardiac failure, and 7/9 of these patients were weaned. In patient no. 4 with a very large perioperative myocardial infarction, assistance was started because of his young age, but no improvement could be achieved. We do not select patients over 70 years of age even if they meet the assist criteria. In accordance with other reports, we think that these patients have very limited physiologic reserve to overcome the assist period.4 The timing of device insertion is equally important. If weaning from CPB is unsuccessful and there is no surgically correctable problem, volume optimization, maximal inotropic support, and an IABP should be attempted. If such measures are ineffective, a decision to initiate ventricular assistance should be made without delay. We believe that prolonged CPB and several attempts to wean from CPB are responsible for multiple complications and poor results of mechanical ventricular assistance for cardiac failure.5
Once support is established and hemodynamic stabilization has been achieved, we prefer to lower and then stop inotropic drugs, keeping only low-dose dopamine infusion to increase renal blood flow. The main reasons for stopping inotropics on initiation of support is to relieve the extra load and reduce oxygen consumption due to beta-adrenergic stimulation, and to allow the heart to rest for the first 24 hours. Then, inotropics are restarted and increased gradually during the weaning period. Except in children, we prefer to keep an IABP in place during the assist period to add pulsatility to the circulation and minimize the negative effects of continuous flow.6 Venovenous hemofiltration via the femoral vein at the start of ventricular assist markedly improved the management of fluid balance. Routine installation of hemofiltration is recommended to reduce pulmonary and other complications. We prefer anticoagulation by continuous heparin perfusion, but there are reports of no anticoagulation used with the heparin-coated Medtronic Carmeda pump.5
The results in this limited study are in accordance with those of larger series with weaning rates of 45%–60% and patient discharge rates of 20%–42%.2,7 In addition, most of the patients successfully discharged have a satisfactory long-term survival.8 Two of our patients died suddenly during follow-up and we think they might have suffered fatal arrhythmias as a complication of their congenital pathology. The most impressive finding in our experience with Bio-Pump support post-CPB was that survival and discharge rates were very close to those of more expensive and sophisticated devices, proving once again the cost effectiveness of a centrifugal pump in short-term cardiac support.9,10 Another advantage of this type of support is its very rapid installation. In the 2 patients with acute hemodynamic deterioration in the ICU, support was initiated immediately at the bedside; both were alive after 41 and 60 months. Patients needing longer support or bridging to cardiac transplantation always retain the option of being switched to a more sophisticated device. With careful patient selection, good timing in initiating support, and careful handling of the assist period, one can expect 40% to 60% survival of patients who would otherwise have died. Thus, one of the oldest assist devices, the Bio-Pump, still retains its importance among the available cardiac support systems.
Presented at the 52nd International Congress of the European Society for Cardiovascular Surgery, Istanbul, Turkey, November 7–10, 2003.
|
REFERENCES
|
|---|
- Noon GP, Lafuente JA, Irwin S. Acute and temporary ventricular support with BioMedicus centrifugal pump. Ann Thorac Surg
1999;68:650–4.[Abstract/Free Full Text]
- Curtis JJ, Walls JT, Wagner-Mann CC, Schmaltz RA, Demmy TL, McKenney CA, et al. Centrifugal pumps: description of devices and surgical techniques. Ann Thorac Surg
1999;68:666–71.[Abstract/Free Full Text]
- Magovern GJ Jr, Simpson KA. Extracorporeal membrane oxygenation for adult cardiac support: the Allegheny experience. Ann Thorac Surg
1999;68:655–61.[Abstract/Free Full Text]
- Pennington DG, Farrar DJ, Loisance D, Pae WE Jr, Emery RW. Circulatory support 1991. The Second International Conference on Circulatory Support Devices for Severe Heart Failure. Patient selection. Ann Thorac Surg
1993;55:206–12.[Medline]
- Noon GP, Ball JW Jr, Short HD. Bio-Medicus centrifugal ventricular support for postcarditomy cardiac failure: a review of 129 cases. Ann Thorac Surg
1996;61:291–5.[Abstract/Free Full Text]
- Dernevik L, Arvidsson S, William-Olsson G. Cerebral perfusion in dogs during pulsatile and non pulsatile extracorporeal circulation. J Cardiovasc Surg (Torino)
1985;26:32–7.[Medline]
- Joyce LD, Kiser JC, Eales F, King RM, Overton JW Jr, Toninato CJ. Experience with generally accepted centrifugal pumps: personal and collective experience. Ann Thorac Surg
1996;61:287–90.[Abstract/Free Full Text]
- Kanter KR, Ruzevich SA, Pennington DG, McBride LR, Swartz MT, Willman VL. Follow-up of survivors of mechanical circulatory support. J Thorac Cardiovasc Surg
1988;96:72–80.[Abstract]
- Farrar DJ, Holman WR, McBride LR, Kormos RL, Icenogle TB, Hendry PJ, et al. Long-term follow-up of Thoratec ventricular assist device bridge-to-recovery patients successfully removed from support after recovery of ventricular function. J Heart Lung Transplant
2002;21:516–21.[Medline]
- Samuels LE, Holmes EC, Thomas MP, Entwistle JC 3rd, Morris RJ, Narula J, et al. Management of acute cardiac failure with mechanical assist: experience with the ABIOMED BVS 5000. Ann Thorac Surg
2001;71:S67–72.[Abstract/Free Full Text]
TOP
ABSTRACT
INTRODUCTION
