HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Jack J. Curtis Joseph T. Walls Richard A. Schmaltz Todd L. Demmy Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Curtis, J. J. Articles by Mann, F. A. Search for Related Content PubMed PubMed Citation Articles by Curtis, J. J. Articles by Mann, F. A.

Ann Thorac Surg 1999;68:666-671
© 1999 The Society of Thoracic Surgeons

---

Description of Devices and Surgical Ttechniques

Centrifugal pumps: description of devices and surgical techniques

^a Division of Cardiothoracic Surgery, University of Missouri School of Medicine, Columbia, Missouri, USA

Address reprint requests to Dr Curtis, Division of Cardiothoracic Surgery, MA312 HFC, University Hospital, One Hospital Dr, Columbia, MO 65212
e-mail: curtisj{at}health.missouri.edu

Presented at the Fourth International Conference on Circulatory Support Devices for Severe Cardiac Failure, Houston, TX, Oct 3–5, 1997.

Abstract

Background. Because of simplicity of application, universal access, and low cost, centrifugal pumps are commonly used for short-term mechanical cardiac assist. Indications and techniques for application of this technology continue to evolve.

Methods. The clinical experience with 151 patients undergoing centrifugal mechanical cardiac assist at the University of Missouri-Columbia has been reviewed. We have compared commonly available centrifugal pumping systems in vitro and in vivo for characteristics that might distinguish them.

Results. Centrifugal pumps have been found to be well suited for use in surgery on the thoracic aorta, for extracorporeal membrane oxygenation and for postcardiotomy cardiac mechanical assist. Complications associated with centrifugal mechanical assist are predictable and common but potentially can be reduced by improved surgical techniques and anticoagulation strategies. In vitro and in vivo experimentation with available centrifugal pumps reveals nuances characteristic of each of the devices.

Conclusions. All centrifugal pumps presently available are less destructive to blood cellular elements compared with roller pumps. With familiarity, all can function satisfactorily for short-term mechanical assist with no compelling evidence that favors any particular centrifugal pump system clinically available. Centrifugal pumps are ideally suited for left heart bypass during surgery on a thoracic aorta and for short-term application as may be required for postcardiotomy mechanical assist. Centrifugal pump technology should be part of the armamentarium of all cardiothoracic surgeons.

Of the various ventricular assist devices presently available, centrifugal pumps have been used most commonly [1]. This may be due to the fact that centrifugal pumps are available to all surgeons, are relatively simple to operate, and are inexpensive compared with other mechanical assist devices. At the University of Missouri-Columbia, since 1986, we have had experience with 201 centrifugal pumps in 151 patients for a variety of indications, including surgery on the thoracic aorta, support of cardiac allograft, extracorporeal membrane oxygenation (ECMO) for a variety of indications, and for postcardiotomy cardiogenic shock. In addition to the clinical experience, we have had the opportunity to evaluate a variety of centrifugal pumps in vitro and in vivo in our laboratory. In this report, we will describe and compare centrifugal pumps available to the surgeon in the U.S. and address the question as to which of the centrifugal pump systems is best for mechanical assist. We will review our surgical techniques for application of centrifugal pumps with emphasis on avoiding the considerable morbidity inherent in their use.

Description

Worldwide, numerous centrifugal pumps are available or are in development for clinical use. However, in the US, until recently, only three centrifugal pumps have been commercially available. All are disposable, cost less than $200.00 per unit, and are relatively simple to operate.

The Sarns centrifugal pump (3-M Health Care, Ann Arbor, MI), shown in Figure 1, uses a spinning impeller system to impart a rotary motion to incoming perfusate. The St. Jude Medical Lifestream centrifugal pump (St. Jude Medical, Inc, Cardiac Assist Division, Chelmsford, MA) is shown in Figure 2. The Lifestream centrifugal pump employs a curved vane design and angled egress blood flow path that purports to minimize turbulence, decrease hemolysis, and reduce periods of flow stasis. The BioMedicus BioPump centrifugal pump head manufactured and marketed by Medtronic BioMedicus, Inc (Eden Prairie, MN) is shown in Figure 3. The BioPump consists of valveless rotator cones that are made to impart a circular motion to incoming blood by viscous drag and constrained vortex principles generating pressure and flow. The Carmeda BioMedicus BioPump (also referred to as BioActive BioPump) has the same appearance as the BioPump, but has heparin covalently bonded to the blood exposed surfaces (Medtronic BioMedicus, Inc).

View larger version (107K): [in this window] [in a new window]   Fig 1. Sarns centrifugal pump head.

View larger version (115K): [in this window] [in a new window]   Fig 2. St. Jude Medical Lifestream centrifugal pump head (formerly Aires).

View larger version (96K): [in this window] [in a new window]   Fig 3. BioMedicus BioPump centrifugal pump head. The Carmeda BioPump has the same appearance as the BioPump but has been heparin bonded.

In vitro comparison of centrifugal pumps

We have compared the four centrifugal pump systems described above with a standard roller pump in an in vitro trial designed to assess blood cell element destruction [2]. Identical circuits (n = 7) primed with fresh, unpooled citrated bovine blood were used to test each of the five systems for 24 hours at a flow of 4.5 L/min. There were no mechanical mishaps observed during 24 hours of pumping with any of the five systems. Platelet counts decreased similarly among the perfusion systems. Of the monitored blood parameters, there were no significant differences observed during the first 4 hours of pumping. However, by 5 hours, plasma-free hemoglobin and lactate dehydrogenase rose significantly with the roller pump and the Sarns centrifugal pump (Fig 4). With additional pumping, evidence of continued red blood cell destruction with the roller pump was observed. There was no apparent advantage of heparin bonding in this 24-hour in vitro test. Others have found similar comparative results with in vitro testing of these pumps [3].

View larger version (25K): [in this window] [in a new window]   Fig 4. Change in lactate dehydrogenase (LDH) expressed as percent of baseline, for each of test systems over the 24-h experimental protocol. Plasma-free hemoglobin changes mirrored lactate dehydrogenase changes. (Used with permission from Curtis and associates [2]. From the International Journal of Angiology 3:128–134 (1994). Reproduced with permission of the copyright owner; International College of Angiology, Inc, Neconset, NY, USA. All rights reserved.

We have compared the centrifugal pumping systems described above in a 96-hour left heart assist calf model [4] with emphasis on mechanical function, hematologic effects, and thromboembolism [5]. Left atrial to thoracic aorta bypass was accomplished with Sarns, St. Jude, BioPump, and Carmeda BioPump (n = 5, each group). Heparin was used in the initial prime only. Left heart bypass was maintained for 96 hours at 3.5 L/min with continuous hemodynamic monitoring and frequent blood sampling. Compared with sham-operated controls, platelet counts dropped significantly but similarly with each of the pumps. There were no significant differences noted among the centrifugal pump groups for plasma-free hemoglobin nor lactate dehydrogenase. All pumps functioned for 96 hours. Inspection of the pump heads at the end of 96 hours revealed pump thrombosis in four of five pump heads in each the Sarns, St. Jude, Medtronic BioPump, and Carmeda BioPump. Gross and histopathological findings at postmortem examination revealed frequent thromboembolism as has been observed with clinical centrifugal mechanical cardiac assist [6].

Surgical techniques

The surgical techniques for instituting centrifugal mechanical assist are simple but details are important. Techniques are dependent upon the indication for application and predicated on knowledge of anticipated morbidity. Cannula selection and cannulation site options have been described in detail [7] and have evolved during our experience. Wire-reinforced cannulas are preferred. To establish left ventricular assist in the setting of inability to wean from cardiopulmonary bypass, the existing aortic cannula is used for patient flow ingress, with the left atrium being cannulated at the junction of the right superior pulmonary vein for patient blood egress. We do not attempt to cross the mitral valve. For elective left ventricular assist in other settings, ie, delayed postcardiotomy shock, we would attempt to cannulate the aortic arch for the theoretical advantage of reducing cerebral embolization. For right ventricular assist in the setting of inability to wean from cardiopulmonary bypass, we no longer use the two-stage cannula routinely placed through the right atrial appendage, which was used for the cardiac surgical procedure. Rather, we recannulate the right atrium near the inferior vena cava directing the single-stage 36 French cannula cephalad, allowing easy horizontal exit of the cannula through the abdominal wall. Return of blood from the centrifugal pump to the patient is by way of cannulation of the proximal pulmonary artery.

In our experience, the most common complication associated with centrifugal mechanical cardiac assist is bleeding, and the most common site of bleeding at reexploration is at one of the cannulation sites. Meticulous cannulation technique is the single most important factor for avoiding postoperative bleeding. At all cannulation sites, concentric purse-string sutures are used tying the innermost and retaining the other for tying at the time of device explantation. We now avoid exiting cannulas through intercostal spaces, because fixation here may result in movement with respiration. Rather, the aortic cannula exits superior to the median sternotomy with the remaining three cannulas for biventricular support exiting through the abdominal wall.

Other strategies to avoid postoperative bleeding include administration of single donor-apheresed platelets and fresh-frozen plasma at the time of institution of mechanical assist because all patients in our experience with postcardiotomy centrifugal support have had an abnormal coagulation profile. When the sternum is not closed, warmed bone wax is applied to sternal edges because this may be a significant site of bleeding in the patient with coagulopathy, removing the wax at the time of decannulation. We also practice an aggressive mediastinal reexploration policy that is performed in our intensive care unit. We find this less cumbersome than attempting to take the patient with multiple devices back to the operating room. Although bleeding continues to be our most common complication, these strategies have been effective for decreasing postoperative bleeding with centrifugal mechanical assist [7].

Once uni- or biventricular assist has been instituted, the devices are monitored by our routine intensive care unit nursing staff with oversight by our cardiopulmonary perfusionists. Centrifugal pumps are inspected every 12 hours and are not changed unless some indication of malfunction presents.

Our anticoagulation regimen has become more aggressive with experience and with the realization that thromboembolism as discovered by autopsy is more prevalent than clinically appreciated [6]. We begin heparin when mediastinal bleeding is less than 150 mL/h, maintaining a partial thromboplastin time in the range of 40–60 seconds.

After bleeding, the next most common complication observed during centrifugal mechanical assist is renal failure. In our experience with 86 patients supported with centrifugal pump for inability to wean from cardiopulmonary bypass, 32 (37%) have experienced acute renal failure. Because the need for hemodialysis or ultrafiltration is common in these patients, we routinely interpose a plastic Luer-Lock (Braun, Lucerne, Switzerland) three-way connector between the cardiac cannula and the centrifugal pump bypass tubing. This allows one to exit intravenous tubing from the outflow limb of the ventricular assist device to an ultrafiltration or hemodialysis unit [8], obviating the need for establishing additional vascular access. If a right ventricular circuit is available, this limb is used. However, left ventricular limbs have been used with great caution to avoid air embolization.

Centrifugal mechanical assist for surgery on the thoracic aorta

Centrifugal pumps are ideally suited for facilitation of surgery on the arch or thoracic aorta, as demonstrated in Figure 5 by an actual case. These cases are managed by cannulating the left atrial appendage using a 36 Fr wire-reinforced cannula, gaining exposure through a small pericardiotomy. An alternative route for venous cannulation is the pulmonary vein at its junction with the left atrium. Using short sections of tubing, the centrifugal pump is used to return blood either to the femoral artery, as shown in Figure 6, or to the thoracic aorta distal to the lesion and cross-clamp. The right radial and right femoral arteries are used to monitor proximal aortic perfusion pressure provided by the heart and distal thoracic aortic perfusion pressure provided by the centrifugal pump during aortic cross-clamping. Afterload reduction is easily accomplished by adjusting the revolutions per minute of the centrifugal pump while distal aortic perfusion pressure is maintained by increasing preload with volume administration as necessary. By unloading the heart, clamp injury to the proximal aorta should be lessened while maintaining adequate perfusion to abdominal viscera. A perfusion cannula can be constructed from the pump egress limb of the centrifugal pump to perfuse the left carotid artery if necessary. In the case of traumatic thoracic aorta disruption associated with multisystem organ trauma, systemic heparinization is not employed [9, 10].

View larger version (144K): [in this window] [in a new window]   Fig 5. This thoracic arch pseudoaneurysm in this 77-year-old man demonstrates an ideal application of centrifugal pumps. The aorta was clamped distal to the left subclavian artery and graft interposition accomplished after establishing partial left heart bypass.

View larger version (29K): [in this window] [in a new window]   Fig 6. Left heart partial bypass is accomplished by proximal cannulation of the left atrium and cannulation of the femoral artery or distal thoracic aorta. The centrifugal pump allows controlled afterload reduction and ensures distal aortic perfusion while permitting good visualization of the operative site. (Used with permission from Walls and associates [9]. Sarns centrifugal pump for repair of thoracic aortic injury. J Trauma 1989:29:1283–5.)

A comparison of clinical outcomes with the use of a variety of centrifugal pumps for postcardiotomy mechanical assist is difficult. This is because the indication for instituting ventricular mechanical assist varies in different reported series and may include inability to wean from cardiopulmonary bypass, for support of poor hemodynamics that develops in the operating room or in the intensive care unit after successful wean from cardiopulmonary bypass, and when the centrifugal assist is used as a bridge to cardiac transplantation or to a long-term mechanical assist device. The outcome differs with each of these indications. Noon and colleagues reported their experience with 129 patients who received BioMedicus centrifugal support after failure to be weaned from cardiopulmonary bypass or progressive postcardiotomy cardiac failure shortly after arrival in the intensive care unit. Fifty-six percent of their patients were weaned from mechanical support and 21% were discharged from the hospital. Coagulopathy and renal insufficiency were the most common complications, occurring in 57% and 43% of their patients, respectively [11].

Joyce and colleagues reported their experience with the Sarns centrifugal pump with 43 patients for a variety of indications and found that 65% could be weaned or transplanted and 42% were dismissed from the hospital. When they analyzed subsets of patients, including 21 patients who could not be weaned from cardiopulmonary bypass, 62% could be weaned or transplanted and 33% were dismissed from the hospital [12].

At the University of Missouri in Columbia, we have had experience with 86 patients who could not be weaned from cardiopulmonary bypass despite multiple inotropes and intraaortic balloon pumping in all. Thirty-nine patients (45%) were weaned and 17 patients (19.8%) were dismissed from the hospital. Complications were ubiquitous, with bleeding being observed in 37 patients (43%) and renal failure in 32 patients (37%).

The International Registry for Mechanical Ventricular Assist Pumps and Artificial Hearts previously maintained at the Pennsylvania State University, as reported by Joyce and associates, listed 12,079 patients who received a variety of assist devices for postcardiotomy support, finding 584 (45.7%) were weaned from the devices or transplanted and 323 (25.3%) were discharged from the hospital [12].

Difficulty comparing various reported surgical series is punctuated by an inspection of our institution’s experience by individual surgeons’ outcome (Table 1). The outcome varied from a wean rate of 35% to 62% and a hospital survival rate of 0% to 43%. If one were selecting a centrifugal pump for postcardiotomy assist, one might choose the device used by the surgeon with the best outcome. However, all surgeons used the Sarns centrifugal pump exclusively and all devices were used for the purpose of inability to wean from cardiopulmonary bypass.

A centrifugal pump capable of left heart bypass was developed by Bernstein and associates [13] and used clinically in pioneering work by Golding and associates and Pennington and associates [14, 15]. The technology employed, while successful, had an unacceptably high hemolysis rate and that centrifugal pump is no longer used. Subsequently developed centrifugal pumps have generally touted their low hemolysis rate as a principal marketing strategy. Suffice it to say, all centrifugal pumping systems presently available for use in the US are kinder to blood cellular elements compared with roller pumps. It would seem prudent, therefore, to use centrifugal pump technology in any clinical application that will require 4 hours or more of blood pumping.

Centrifugal pumps are ideally suited for surgery on the thoracic aorta even though assist usually lasts less than 1 hour. This is because the length of bypass tubing, and therefore blood/foreign surface interface-associated pathophysiology, is minimized by placing the centrifugal pump near the operating table. For this and other reasons we have described in this report, we believe centrifugal left heart bypass is the technique of choice for surgery on the thoracic aorta.

Presently available centrifugal pump technology was not intended for long-term use. Despite this, centrifugal pumps have been used most commonly in the setting of postcardiotomy mechanical assist. Most of these patients have infarcted or stunned myocardium, and recovery is not predictable. In the setting of inability to wean from cardiopulmonary bypass, we have used centrifugal assist for up to 18 days. However, we have never had a survivor who required mechanical assist greater than 96 hours. We have shown in our in vivo studies, described in this report, that all centrifugal pumping systems presently available may function for 96 hours of mechanical assist. Since patients receiving uni- or biventricular assist cannot be ambulated or rehabilitated, centrifugal devices are suboptimal for longer-term use. After 96 hours, a decision should be made regarding replacement of the centrifugal pump with pneumatic or electrical mechanical assist devices.

We conclude that the centrifugal pump is the mechanical assist device of choice for ECMO and for surgery on the thoracic aorta. All centrifugal pumps presently available are less destructive to blood cellular elements compared with roller pumps, and all function satisfactorily for short-term mechanical assist as is required in the setting of inability to wean from cardiopulmonary bypass. There is no compelling evidence, either investigational or clinical, that favors one of the commercially available centrifugal pumps over the other. Centrifugal pump technology should be part of the armamentarium of all cardiothoracic surgeons.

Acknowledgments

The clerical assistance of Paula Threet is greatly appreciated. Special gratitude is expressed to the manufacturers of centrifugal pumps (Medtronic, Bio-Medicus, Inc, Eden Prairie, MN; Sarns 3-M Health Care, Ann Arbor, MI; and St. Jude Medical, Inc, Chelmsford, MA) for their support of our investigational and clinical efforts at the University of Missouri. Appreciation is expressed to Don Wood, Terry Bahler, Mark Mikesch, and Tom Weeks for their assistance in cardiopulmonary perfusion.

References

1. Pae W.E., Jr, Miller C.A., Mathews Y., Pierce W.S. Ventricular assist devices for postcardiotomy cardiogenic shock. J Thorac Cardiovasc Surg 1992;104:541-553.[Abstract]
2. Curtis J.J., Wagner-Mann C.C., Turpin T.A., et al. In-vitro evaluation of five commercially available perfusion systems. Int J Angiol 1994;3:128-133.
3. Iatridis E, Chan T. An evaluation of vortex, centrifugal and roller pump systems. Proceedings of the International Workshop on Rotary Blood Pumps, Vienna, Austria, September 1991.
4. Mann F.A., Wagner-Mann C.C., Curtis J.J., Demmy T.L., Turk J.R. A calf model for left ventricular centrifugal mechanical assist. Artif Organs 1996;20:670-677.[Medline]
5. Curtis JJ, Wagner-Mann CC, Mann FA, et al. 96 Hour comparative study of centrifugal pumps used for left ventricular assist. Proceedings of the 5th Congress of the International Society for Rotary Blood Pumps, Marseille, France 1997:11.
6. Curtis J.J., Walls J.T., Boley T.M., Schmaltz R.A., Demmy T.L. Autopsy findings in patients with postcardiotomy centrifugal ventricular assist. ASAIO J 1992;38:M688-M690.[Medline]
7. Curtis J.J., Walls J.T., Schmaltz R.A., Demmy T.L., Wagner-Mann C.C., McKenney C., Nawarawong W. Improving clinical outcome with centrifugal mechanical assist for postcardiotomy ventricular failure. Artif Organs 1995;19:761-765.[Medline]
8. Curtis J.J., Deese L.R., Walls J.T., Boley T.M. Use of a rate-limited ultrafiltration circuit with centrifugal ventricular assist. Artif Organs 1994;18:465-466.[Medline]
9. Walls J.T., Curtis J.J., Boley T.M. Sarns centrifugal pump for repair of thoracic aortic injury. J Trauma 1989;29:1283-1285.[Medline]
10. Walls J.T., Boley T.M., Curtis J.J., Schmaltz R.A. Experience with four surgical techniques to repair traumatic aortic pseudoaneurysm. J Thorac Cardiovas Surg 1993;106:283-287.[Abstract]
11. Noon G.P., Ball J.W., Short H.D. BioMedicus ventricular centrifugal support for postcardiotomy cardiac failure. Ann Thorac Surg 1996;61:291-295.[Abstract/Free Full Text]
12. Joyce L.D., Kaiser J.C., Frasier L., et al. Experience with generally accepted centrifugal pumps. Ann Thorac Surg 1996;61:287-290.[Abstract/Free Full Text]
13. Bernstein E.F., Dorman F.D., Blackshear P.L., Scott D.R. An efficient, compact blood pump for assisted circulation. Surgery 1970;68:105-115.[Medline]
14. Golding L.R., Groves L.K., Peter M., et al. Initial clinical experience with a new temporary left ventricular assist device. Ann Thorac Surg 1980;29:66-69.[Abstract/Free Full Text]
15. Pennington D.G., Merhavy J.P., Swartz M.T., Willman V.L. Clinical experience with a centrifugal pump ventricular assist device. Trans ASAIO 1982;28:93-99.

This article has been cited by other articles:

				S. Aggarwal, F. Cheema, M. C. Oz, and Y. Naka Long-Term Mechanical Circulatory Support Card. Surg. Adult, January 1, 2008; 3(2008): 1609 - 1628. [Full Text]

				M. Mert, A. Akcevin, C. E Yildiz, and K. Suzer Postoperative Mechanical Circulatory Support with Biomedicus Centrifugal Pump Asian Cardiovasc Thorac Ann, March 1, 2005; 13(1): 38 - 41. [Abstract] [Full Text] [PDF]

				M Bennett, S Horton, C Thuys, S Augustin, M Rosenberg, and C Brizard Pump-induced haemolysis: a comparison of short-term ventricular assist devices Perfusion, March 1, 2004; 19(2): 107 - 111. [Abstract] [PDF]

				D. Paparella, T. M. Yau, and E. Young Cardiopulmonary bypass induced inflammation: pathophysiology and treatment. An update Eur J Cardiothorac Surg, February 1, 2002; 21(2): 232 - 244. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Jack J. Curtis Joseph T. Walls Richard A. Schmaltz Todd L. Demmy Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Curtis, J. J. Articles by Mann, F. A. Search for Related Content PubMed PubMed Citation Articles by Curtis, J. J. Articles by Mann, F. A.