Because the number of patients who die while waiting for cardiac transplantation continues to increase, many ventricular assist devices (VADs) have been developed as life saving therapy and bridges to transplantation or myocardial recovery. 1 Although the Rematch Trial demonstrated a survival benefit and an improved quality of life for patients with advanced heart failure undergoing VAD support in comparison with an optimal medical management, 2 currently available VADs are associated with a considerable number of serious complications: thromboembolic events and bleeding complications, 3–8 infections, 9–11 and device malfunctions. 4,12–14 Therefore, the improvement of currently used or the development of new VADS is necessary to reduce the number of complications. This article reports on the first animal tests with a newly developed microdiagonal blood pump (MDP) for partial unloading of the left ventricle.
Materials and Methods
Six adult female sheep weighing 80 to 90 kg were used in this experiment. All animals received humane care in compliance with the Guide for the Care and Use of Laboratory Animals as published by the National Institutes of Health (NIH publication No. 85–23, revised 1985). The experiments were approved by the district government of Cologne, Germany (No. 50.203.2-AC 14, 06/02).
After introduction of anesthesia, a transthoracic echocardiography was performed. The operation started with insertion of the pulmonary artery and the central venous lines to measure the hemodynamic parameters and to take blood samples. Next, the MDP was implanted as described in the following sections. After implantation of the pump, another transthoracic echocardiographic study was performed, the hemodynamic parameters were measured, and blood samples were taken again. Anesthesia was stopped, and the animal was extubated after return of sufficient spontaneous breathing. During the next week, blood samples were taken daily in the morning. The technical parameters of the pump were recorded continuously. After 7 days, anesthesia was introduced again. After another transthoracic echocardiography, measurement of hemodynamic parameters, and hemanalysis, the animals were euthanized with pentobarbital. The pump was explanted, and the internal organs were excised for histologic examinations.
The MDP was developed at the Institute for Biomedical Engineering (Aachen, Germany), and was designed as a left VAD for partial unloading of the left ventricle.
The Rotary Pump
The rotary pump (Figure 1) consists of a brushless direct current–motor that drives the mixed flow impeller by means of a magnetic coupling system. Inside the pump, the blood stream serves to cool the integrated electric motor. A central feature of this implantable pump is the use of blood-immersed bearings, thus eliminating the need for seals, which are not appropriate for long-term applications. To enhance the fluid motion around the bearings and to prevent stagnation of blood flow in these critical regions, washout holes have been placed in the rotor hub.
Anesthesia for Implantation of the MDP and Hemodynamic Measurements
After intramuscular premedication with atropine (0.1–0.15 mg/kg, Atropinum sulfuricum solutum 1%, Wirtschaftsgenossenschaft deutscher Tierärzte eG, Garbsen, Germany) and xylazine (0.15–0.2 mg/kg, Rompun 2%, Bayer Vital GmbH, Leverkusen, Germany), a 20 Gauge (G) cannula (Abbocath-T, Abbott, Sligo, Ireland) was inserted into an auricular vein. General anesthesia was induced with pentobarbital (3–8 mg/kg, Narcoren, Merial GmbH, Hallbergmoos, Germany), and endotracheal intubation (inner diameter of the tube 7.5–8.5 mm, Hi-Lo Lanz, Mallinckrodt Medical Inc., Athlone, Ireland) was performed without muscle relaxation. Anesthesia was then maintained with isoflurane (0.5–1.0 minimally alveolar concentration, Forene, Abbott GmbH, Wiesbaden, Germany) and analgesia was supplemented by metamizole (15–20 mg/kg, Novalgin, Aventis Pharma Deutschland, Frankfurt a. M., Germany) at the beginning and the end of surgery.
Routine respiratory and hemodynamic monitoring were performed with pulse oximetry, capnography, electrocardiography, and continuous measurement of arterial blood pressure (S/5 Kompaktmonitor Intensiv, Datex Ohmeda GmbH, Duisburg, Germany and PM 8050 Atemgasmonitor, Dräger Medizintechnik GmbH, Lübeck, Germany). For arterial blood pressure monitoring, a 22 G cannula was inserted into an auricular artery. Tidal volume and inspired oxygen concentration (FiO2) were adjusted to maintain normal arterial partial pressure of carbon dioxide levels and arterial oxygen saturation greater than 98%.
After induction of anesthesia and establishment of routine monitoring, a gastric tube, an esophageal temperature probe, and a urinary catheter were placed, and echocardiography (CFM 800, Vingmed Sound AB, Horten Norway) was performed transthoracically. The left jugular vein was surgically prepared for insertion of a central venous line (8.5 French, Arrow Deutschland GmbH, Erding Germany), as well as a 7 French Swan-Ganz catheter (Becton, Dickinson Critical Care Systems Pte Ltd., Singapore). The catheters were tunneled to the back of the sheep and fixed to ensure that they did not change their position during the study.
Arterial and mixed venous blood samples were obtained for determination of hemoglobin and blood gas analysis (ABL 520 and EML 100, both Radiometer Medical A/S, Copenhagen, Denmark). Routine hemodynamic variables (heart rate, mean arterial, central venous, pulmonary artery, and pulmonary capillary wedge pressure) were obtained after induction of anesthesia as well as after thoracotomy, just before and after the implantation of the MDP, and at the end of surgery. Measurements of cardiac output were performed by five valid bolus injections of 10 mL ice cooled sodium chloride solution (0.9%) into the right atrium. All hemodynamic values were recorded by a polygraph.
If deemed necessary, colloid (10% hydroxyethylstarch, Fresenius Kabi Deutschland GmbH, Bad Homburg v. d. H., Germany) or crystalloid fluids (Ringer solution, Delta Pharma GmbH, Pfullingen, Germany) were given to reach adequate intraoperative hemodynamic stability, defined as mean arterial pressure greater than 60 mm Hg and cardiac output greater than 4 L/minute. During the entire surgical procedure, the sheep were bedded on warming blankets (Astopad, System Medical GmbH, Bad Münder, Germany), and core temperatures were maintained greater than 36°C. Input and loss of fluids were recorded, and cumulative fluid balances were performed at the end of surgery. Perioperative antibiotic prophylaxis was performed with 1500 mg cefuroxime (Zinacef, GlaxoSmithKline GmbH & Co KG, München Germany) after induction of anesthesia and just before the end of surgery.
After the end of surgery, echocardiography was performed again. The gastric tube, the esophageal temperature probe, and the cannula in the auricular vein were removed. The central venous and the pulmonary artery catheter, and the auricular arterial line were left in place. Inhaled isoflurane concentrations were reduced. After return of spontaneous breathing, the sheep were extubated and brought into a special observation cage.
Postoperative analgesia was achieved with intravenously administered metamizole (15–20 mg/kg every 4 hours) and subcutaneously given buprenorphine (3–5 μg/kg every 12 hours, Temgesic, Essex Pharma GmbH, München, Germany). Additionally, the incisions were infiltrated with lidocaine 1% (Xylocain, Astra Zeneca GmbH, Wedel, Germany) postoperatively. The sheep were allowed to drink and eat immediately after they could stand alone.
Anesthesia for explantation of the MDP was the same as for the implantation procedure. Routine hemodynamic variables were obtained, and cardiac output measurements were performed after induction of anesthesia, after thoracotomy, and just before and after explantation of the MDP. After that, the sheep were euthanized with pentobarbital (50–80 mg/kg).
At first, the pulmonary artery and central venous lines were introduced via the external jugular vein on the left side. Then the thorax was opened using a left sided thoracotomy in the fourth intercostal space. Anticoagulation was done by heparin (400 U/kg, Heparin-Natrium, B Braun, Melsungen Germany) with an activated coagulation time of more than 400 s intra-operatively and 200–250 s postoperatively (measured with the Hemochrom 401 and FTCA 510, International Technidyne Corporation, Edison, New Jersey, USA). The descending aorta was side clamped, incised, and anastomosed to the outflow graft. For blood flow measurements, a flowprobe (Transsonic 10C, flowmeter: T110R, Transsonic Systems Inc., Ithaca, NY, USA) using ultrasonic signals was connected to the outflow graft. After that, the inflow conduit (24 French, 135°, ⅜″, Medos AG, Aachen, Germany) was inserted into the left atrium via the left atrial appendage. Both conduits were connected to the MDP after extensive de-airing. The pump was placed in the subcutaneous tissue over the tenth to twelfth ribs. The driving line was pulled through the subcutaneous tissue to the back of the animal and connected to the external console. The pump was started, and, if there was a significantly prolonged activated coagulation time or an enhanced bleeding tendency, heparin was antagonized partly by Protamine (Protamin ICN, ICN Pharmaceuticals GmbH, Frankfurt/Main, Germany). Pump flow was adjusted to 2–3 L/minute for partial unloading of the heart (6.000–6.500 rotations per minute). Surgery was terminated by closure of the thoracotomy.
Transesophageal and transthoracic echocardiographic measurements and electrocardiogram triggered loops were performed before and after MDP implantation as well as before explantation of the MDP. Measurements using two-dimensional and M-Mode echocardiography were performed for dimensions of the left ventricle, left atrium, left ventricular outflow tract, and aorta. Pulsed doppler and continuous doppler modes were used for velocity measurements in typical locations, especially in the left ventricular outflow tract for calculation of the cardiac output (CO) of the left ventricle; two-dimensional mode and M-mode were used to evaluate ejection fraction (by the Simpson method) and fractional shortening.
Echocardiographic measurements and imaging were performed with GE Vingmed CFM 800 (GE Ultraschall Deutschland GmbH, Solingen, Germany). Calculations were performed with the GE Vingmed ECHOPAC System.
Evaluation of Blood Samples
Blood samples were taken daily in the morning. The platelet count and hemoglobin values were measured by the Celltek M-Device (Bayer AG, Leverkusen, Germany). As parameters for the function of internal organs, the creatinine and bilirubin values were controlled (Vitros 250, Ortho-Clinical Diagnostics, Johnson–Johnson Company, Neckargemünd, Germany). Extent of hemolysis was assessed by the free hemoglobin (measured with the Uvikon 930 Spectrophotometer, Bio-Tek Kontron Instruments, Winoosci, Vermont, USA).
Autopsies were performed immediately after termination of the trials. The most important internal organs (lung, heart, brain, bowels, kidneys, spleen, liver) were evaluated macroscopically. Tissue samples were taken from each organ and from areas suggestive of infarctions, bleeding, or other lesions. These samples were routinely fixed in formaldehyde and embedded in paraffin. Sections 3–5 μm thick were stained by hematoxylin and eosin for histologic evaluation.
After implantation of the MDP, mean arterial and mean pulmonary arterial pressures as well as the pulmonary capillary wedge pressure remained stable during the entire test period. The data are presented in Table 1.
Because of partial unloading of the heart after MDP implantation, left ventricular end diastolic and end systolic diameters as well as the ejection fraction decreased and did not change during the test period. Cardiac output of the left ventricle decreased postoperatively, indicating a partial unloading of the heart. Neither a mitral nor an aortic valve incompetence was observed. Left ventricular mass remained unchanged during the test period. The most important parameters are depicted in Table 1.
Evaluation of Blood Samples
Hemoglobin and the number of platelets decreased perioperatively for surgical reasons and increased continuously in the postoperative course. Free hemoglobin was not enhanced in the postoperative course. Function of peripheral organs (kidney, liver) was slightly impaired, with a maximum value on the third postoperative day, but recovered completely within the test period. All values are shown in Table 2.
All hearts revealed chronic pericarditis caused by the surgical procedure. The left lungs showed some degree of atelectasis of subpleural areas, also for operative reasons. There were no pathologic findings in the brain, spleen, liver, kidney, and bowels.
Because of the ongoing shortage of donor hearts, the concept of partial unloading of the heart as a bridge to myocardial recovery is gaining more and more interest. Explantation of LVADs in patients with heart failure appears to be more successful after partial rather than complete cardiac unloading because prolonged total unloading of the heart may cause myocyte atrophy. 15 This may explain why patients with partially decompressed ventricles demonstrated significantly better exercise capacities than patients with fully decompressed ventricles. 15 Maybaum et al.15 conclude that some ventricular unloading may be beneficial for myocardial function, blood flow, and metabolism. Hemodynamic and echocardiographic data with the MDP demonstrated an effective partial unloading of the heart (decline of CO of the left ventricle, left ventricular enddiastolic diameter (LVEDD) (4.1 → 3.6 cm, 12.2 %), left ventricular endsystolic diameter (LVESD), ejection fraction, Table 1) with simultaneously stable hemodynamics (no significant changes of mean arterial pressure (MAP), mean pulmonary arterial pressure (MPAP), pulmonary capillary wedge pressure (PCWP), central venous pressure (CVP), Table 1). In addition, Frazier et al.16 reported improvements of hemodynamics after partial unloading of the heart: The average LVEDD decreased by 16% during left VAD support accompanied by a stable arterial blood pressure. No reduction of the left ventricular mass was observed, which may be explained by the short test period of 7 days.
Whether the left atrium or the left ventricular apex should be cannulated for insertion of the inflow cannula remains an ongoing topic of discussion. If the “bridge to recovery” concept is intended, the left atrial cannulation appears to be more reasonable because left ventricular cannulation results in damage of the apex and regional hypokinesis around the apex, thereby limiting ultimate recovery. 17,18 Conversely, some authors report that left atrial cannulation may not allow complete left ventricular decompression caused by the pliable walls of the left atrium, 19 whereas other authors are of the opinion that left atrial as well as left ventricular cannulation may provide adequate circulatory support. 18,20 Because complete left ventricular unloading is not intended in the “bridge to recovery concept,” for the previously mentioned reasons, the left atrial cannulation site may be the cannulation site of choice, if myocardial recovery is expected.
Despite all efforts to achieve an effective anticoagulation in LVAD patients, the risk of thromboembolic events is approximately 30%. 3,4 This frequency of thromboembolic events is caused mainly by the anticoagulation management, the device design, changes in blood flow, altered rheologic conditions, and number/activation status of blood cells. 5,7 Furthermore, thrombus formation may take place within the native heart. 21 No cerebral infarctions were observed, which may be explained by two reasons: partial unloading of the heart and anastomosing the outflow conduit to the descending aorta. In this setting, there is an antegrade blood flow in the ascending aorta and the aortic arch (provided by the native heart) so that thrombi developed within the pump are embolized to peripheral organs distal to the aortic arch.
Beside thromboembolic events or bleeding complications, the degree of blood damage and recovery of peripheral organ function in patients suffering from a cardiogenic shock before LVAD implantation are important issues. The MDP demonstrated no hemolysis; the free hemoglobin level did not increase after MDP implantation (Table 2). The decrease of the hemoglobin level and the platelet count (Table 2) must be explained by operative reasons (blood loss caused by minor bleedings, suction, and the like). Similar results were reported by other groups. 22–25 The low degree of blood damage by the MDP also is demonstrated by normalization of the platelet count and an increase of the hemoglobin level up to day 7 after MDP implantation. In contrast to Ochiai et al.23 and Nojiir et al.,24 the authors of this study observed a slight impairment of the renal and hepatic function (measured by the creatinine and bilirubin values, Table 2) on day 3 after MDP implantation. A similar slight increase of the creatinine value to the third postoperative day was described by Deleuze et al., 25 but urine output was always normal for all animals. This finding also may be explained by operative reasons (for example, intervals of low systemic blood pressure during side clamping of the descending aorta for the anastomosis with the outflow graft). However, these findings were not clinically relevant (e.g., no impairment of the diuresis) and recovered within the test period of 7 days. Similar results of recovery of end organ function were reported by Dasse et al.26 Enhanced values of bilirubin and creatinine before LVAD implantation in heart failure patients normalized after LVAD implantation and improved the physiologic status of the patients before cardiac transplantation.
The MDP appears to be a promising LVAD for partial unloading of the left ventricle. Long-term experiments are planned to fulfill the requirements for clinical approval.
1. Reul HM, Akdis M: Blood pumps for circulatory support. Perfusion
15: 295–311, 2000.
2. Rose EA for the Rematch Study Group: Long-term use of a left ventricular assist device for end-stage heart failure. N Engl J Med
345: 1435–1443, 2001.
3. Christiansen S, Jahn UR, Meyer J, et al:
Anticoagulative management of patients requiring left ventricular assist device implantation and suffering from heparin-induced thrombocytopenia type II. Ann Thorac Surg
69: 774–777, 2000.
4. Christiansen S, Van Aken H, Breithardt G, et al:
Successful cardiac transplantation after four cases of DeBakey LVAD failure. J Heart Lung Transplant
21: 706–709, 2002.
5. Copeland JG: Thromboembolism and bleeding: Clinical strategies. Ann Thorac Surg
61: 376–377, 1996.
6. Gross DR: Concerning thromboembolism associated with left ventricular assist devices. Cardiovasc Res
42: 45–47, 1999.
7. Jahn UR, Christiansen S, Van Aken H, et al:
Platelet activation status in patients with left ventricular assist device. Anesth Analg
90: S52, 2000.
8. Schmid C, Scheld HH, Hammel D: Control of perigraft bleeding during ventricular assist device implantation. Ann Thorac Surg
69: 958–959, 2000.
9. Deng MC, Weyand M, Hammel D, et al:
Selection and outcome of ventricular assist device patients: The Muenster experience. J Heart Lung Transplant
17: 817–825, 1998.
10. Schmid C, Deng M, Hammel D, et al:
Emergency versus elective/urgent left ventricular assist device implantation. J Heart Lung Transplant
17: 1024–1028, 1998.
11. Tjan TD, Asfour B, Hammel D, et al:
Wound complications after left ventricular assist device implantation. Ann Thorac Surg
70: 538–541, 2000.
12. Hammel D, Tjan DT, Scheld HH, et al:
Successful treatment of a Novacor LVAD malfunction without repeat sternotomy. Thorac
46: 154–156, 1998.
13. Jahania S, Onsager DR, Weigel TL, et al:
Diagnosis and successful treatment of a unique form of malfunction of the heartmate left ventricular assist device. J Thorac Cardiovasc Surg
114: 143–144, 1997.
14. Scheld HH, Soeparwata R, Schmid C, et al:
Rupture of inflow conduits in the TCI-Heartmate system. J Thorac Cardiovasc Surg
114: 287–289, 1997.
15. Maybaum S, Epstein S, Beniaminovitz A, et al:
Partial loading of the left ventricle during mechanical assist device support is associated with improved myocardial function, blood flow and metabolism and increased exercise capacity. J Heart Lung Transplant
21: 446–454, 2002.
16. Frazier OH, Myers TJ, Gregoric ID, et al:
Initial clinical experience with the Jarvik 2000 implantable axial-flow left ventricular assist system. Circulation
105: 2855–2860 (Discussion 2808–2809), 2002.
17. Duncan JM, Baldwin RT, Igo SR, et al:
Myocardium-sparing cannulation technique for left ventricular assist device support. Ann Thorac Surg
52: 565–566, 1991.
18. Sethia B, Martin W, Wheatley DJ: The effects of left atrial and left ventricular cannulation on left ventricular function. Int J Artif Organs
8: 331–334, 1985.
19. Holman W, Bourge RC, Murrah CP, et al:
Left atrial or left ventricular cannulation beyond 30 days for a Thoratec ventricular assist device. ASAIO J
41: M517–M522, 1995.
20. Tevaearai HAT, Mueller XM, Jegger D, et al:
Atrialo, ventricular, or both cannulation sites to optimize left ventricular assistance. ASAIO J
47: 261–265, 2001.
21. Scheld HH: Mechanical support—Benefits and risks. Thorac Cardiovasc Surg
45: 1–5, 1997.
22. Fukamachi K, Ochiai Y, Doi K, et al:
Chronic evaluation of the Cleveland Clinic CorAide left ventricular assist system in calves. Artif Organs
26: 529–533, 2002.
23. Ochiai Y, Golding LAR, Massiello AL, et al: In vivo
hemodynamic performance of the Cleveland Clinic CorAide blood pump in calves. Ann Thorac Surg
72: 747–752, 2001.
24. Nojiir C, Kijima T, Maekawa J, et al:
Terumo implantable left ventricular assist system: results of long-term animal study. ASAIO J
46: 117–122, 2000.
25. Deleuze PH, Liu Y, Tixier D, et al:
Centrifugal pump or pneumatic ventricle for short-term mechanical support?Int J Artif Organs
12: 327–332, 1989.
26. Dasse KA, Frazier OH, Lesniak JM, et al:
Clinical responses to ventricular assistance versus transplantation in a series of bridge to transplant patients. ASAIO J
38: M622–M626, 1992.