Author + information
- Received November 8, 2001
- Revision received February 26, 2002
- Accepted March 1, 2002
- Published online June 5, 2002.
- Michael P Siegenthaler, MD*,* (, )
- J.ürgen Martin, MD*,
- Andreas van de Loo, MD†,
- Torsten Doenst, MD*,
- Wolfgang Bothe, MD* and
- Friedhelm Beyersdorf, MD*
- ↵*Reprint requests and correspondence:
Dr. Michael P. Siegenthaler, Department of Cardiovascular Surgery, University of Freiburg, Hugstetterstrasse 55, 79106 Freiburg, Germany.
Objectives We sought to evaluate the surgical results and effects of continuous support with the permanent Jarvik-2000 left ventricular assist device (LVAD). We report the early outcomes.
Background A shortage of transplant donors necessitates the testing of alternative treatments. The Jarvik-2000 is an axial flow pump with a percutaneous retro-auricular power connector, designed for permanent use.
Methods Patients with severe heart failure (HF), unsuitable for heart transplantation or conventional LVAD support, were offered implantation. The surgical approach included a left lateral thoracotomy. The device was implanted into the left ventricular apex on femoro-femoral bypass. It is set to allow pulsatile flow with an aortic valve opening. Anticoagulation is adjusted the same as for patients with a heart valve.
Results Between May 2001 and August 2001, we implanted the Jarvik-2000 in two patients with dilated cardiomyopathy and in one with cardiac amyloidosis, all with severe HF (cardiac index 1.8 ± 0.3 l/m2per min). One patient required preoperative inotropic support. All patients did well, with no repeat operations or infections. Patients received 4.3 ± 3.2 packed red blood cells and were intubated at 14 ± 3 h, and the intensive care unit stay was 7.0 ± 0.5 days. The cardiac index increased from 3.7 ± 1.5 l/min per m2at 8,000 rpm to 5.9 ± 2.9 l/min per m2at 12,000 rpm. All patients currently have mild hemolysis not requiring transfusion. The following postoperative events were recorded: a transient ischemic attack with complete recovery, a short re-intubation due to ventricular arrhythmia, loss of consciousness with a battery change while standing, knee-joint effusion after ergometry training, a minor wound problem and a short hospital re-admission due to dehydration. Patients were discharged home after 49 ± 7 days; one has returned to work. All quality-of-life scores have improved.
Conclusions The permanent Jarvik-2000 appears safe. It can be used for dilative or restrictive disease. The Jarvik-2000 might prove a valid option for the long-term treatment of patients with severe HF.
We tested the Jarvik-2000 permanent left ventricular assist device (LVAD) as a potential alternative to heart transplantation. The Jarvik-2000 is a novel axial flow LVAD (1,2)designed for permanent use (Fig. 1A). Conventional LVADs have been shown to improve survival in terminal heart failure (HF) patients awaiting transplantation (3). However, LVADs are associated with significant complications (4). Over 50% of patients treated with conventional left ventricular (LV) support will have a life-threatening event (5–9). Conventional assist devices are impractical for daily life. Taking all these problems into consideration, LVADs are not an equivalent alternative to heart transplantation.
The Jarvik-2000 is small and has several advantages in its design, so as to reduce complications associated with conventional mechanical circulatory support. It is designed for long-term use. The power cable is connected to a percutaneous retro-auricular skull-mounted pedestal (Fig. 1B). The portable batteries and control unit weigh ∼1.5 kg, allowing patients to lead an independent life-style. We report our early experience with this device.
The study protocol and patient information brochure have been reviewed and approved by the local Ethics Committee. The patient selection and preoperative work-up were in accordance with predetermined guidelines.
Patients with end-stage HF, who had no conventional treatment option, were considered for Jarvik-2000 implantation. Table 1lists the inclusion and exclusion criteria for implantation of the Jarvik-2000 LVAD. Heart failure criteria for transplantation or mechanical LV support had to be present. Patients who did not qualify for heart transplantation, because of advanced age or elevated pulmonary vascular resistance, or patients unsuitable for a conventional LVAD because of a small body frame, with a body surface area ≤1.5 m2, were eligible for Jarvik-2000 implantation. We did not offer the Jarvik-2000 to patients with fixed pulmonary vascular resistance >7 wood units (WU), because of the high risk of postoperative right ventricular (RV) failure. Only patients with a Columbia University risk score index ≤5 were considered for implantation (10). This score, used for patient risk stratification, is based on the following clinical variables: central venous pressure, urine output, respiratory status, hepatic synthetic function, previous cardiac surgery and leukocyte count. Most patients with a score >5 will die after LVAD implantation (10).
The patient was evaluated the same as for heart transplantation (11). This involves right and left heart catheterization, transesophageal echocardiography (TEE), carotid duplex scanning, abdominal ultrasonography, pulmonary function testing, exercise ergonometry and a full laboratory work-up. In addition, a computed tomographic (CT) scan of the head and chest was obtained. The CT scan allows site selection for the outflow graft anastomosis and placement of the skull-mounted pedestal in an area with thick bone (≥4 mm). A full evaluation by a neurologist was performed. Preoperative quality of life was assessed by use of a standardized questionnaire (12,13).
Patient and family information includes detailed discussions and a brochure in layman language explaining the potential risks, benefits and alternatives of the planned procedure. Sufficient time is allowed for questions and the decision-making process before full informed consent is obtained.
Intraoperative monitoring included TEE and a Swan-Ganz catheter. Aprotinin and nitric oxide (5 to 80 ppm) were used routinely. The retro-auricular pedestal site was prepared. A left postero-lateral thoracotomy incision was performed. The internal power cable of the device was passed through the posterior apex of the chest and tunneled to the pedestal, which was attached to the skull (Fig. 1B, inset). Then the outflow graft was sewn to the descending thoracic aorta in partial occlusion. Conventional femoro-femoral bypass was instituted. A sewing ring was attached to the left ventricular apex. The heart was fibrillated. The device was implanted into the LV apex and secured to the sewing ring. Complete de-airing was confirmed with TEE before unclamping the outflow graft. Cardiopulmonary bypass was discontinued, and meticulous hemostasis was secured before thoracotomy closure.
The pump was set to the minimal speed, keeping the LV in a partially loaded state, with aortic valve opening. Infection prophylaxis included 48 h of perioperative antibiotics, antibiotic-coated catheters, catheter removal within two days, early extubation followed by incentive spirometry and aggressive mobilization. In the absence of bleeding, intravenous heparin (500 IU/h) was started 12 h after the operation. On the second postoperative day, heparin was adjusted to full anticoagulation, with an activated partial thromboplastin time of 50 to 70 s. Coumadin (warfarin), as well as the pharmacologic HF treatment, including angiotensin-converting enzyme inhibitors and beta-blockers, was started as soon as the patient could tolerate oral intake. Diuretics could be gradually reduced. Patients visited a formal two-day teaching course for Coumadin management and received a device to check their international normalized ratio (INR) at home. Rehabilitation was started in the hospital and consisted of bicycle ergometry, walking and strengthening exercises with physical therapy. Systematic teaching of the functions and handling of the device components, as well as emergency algorithms, was provided to the patients and their relatives. After hospital discharge, a multidisciplinary team provided outpatient follow-up.
Between May 2001 and August 2001, we implanted the Jarvik-2000 in three patients. Table 2lists the patients’ pre-implantation characteristics. Two patients had dilated cardiomyopathy and one patient had amyloidosis. One patient required preoperative inotropic support with 14 μg/kg body weight per min of dobutamine. Follow-up included over 350 patient-days (91, 93 and 170 days). All patients are doing well. There were no device failures.
Table 3shows the intraoperative data. No problems related to the cardiovascular procedure were encountered. Operating times were 4 to 5 h; cardiopulmonary bypass times decreased slightly as we gained experience. The patient with amyloidosis had friable tissues, which led to several pleural tears with lung lacerations; treatment included fibrin glue and TachoComb (Nycomed, Munich, Germany). There were no further sequelae. Only one patient with a preoperative hemoglobin level of 9.8 gm/dl required intraoperative transfusion (2 U packed red blood cells and 4 U fresh-frozen plasma).
Table 3lists the postoperative data. All patients were extubated on the first postoperative day (14 ± 3 h). Postoperative blood requirements were low with the Jarvik-2000 LVAD, compared with traditional LVAD implantation. Swan-Ganz catheters were removed early (2.3 ± 0.5 days). Figure 2shows the cardiac index on the first postoperative day, as a function of the pump setting. The intensive care unit stay was seven days, due to a prolonged need for vasomotor support with alpha-agonists and vasopressin (Pitressin; 4.2 ± 1.5 days) to maintain a mean arterial pressure of 65 mm Hg. All patients presently have mild hemolysis not requiring transfusion, stable hemoglobin levels, increased lactate dehydrogenase levels, slightly increased reticulocytes and low haptoglobin levels, as illustrated in Figure 3. The end-diastolic diameters of the two patients with dilated cardiomyopathy decreased with postoperative unloading. There were two major postoperative events. One patient had a transient ischemic attack manifested by right arm weakness, which completely resolved within 30 min. The CT scan and examination, performed by a neurologist, were normal. No intracardiac thrombus was found on the echocardiogram. The other event included one 7-h re-intubation in the early postoperative period, due to ventricular arrhythmia. The patient, with known Lown IVa arrhythmia, had an episode of sustained ventricular tachycardia, during which he remained awake. He was sedated and intubated for electrical defibrillation. Several minor events were recorded, including an episode of loss of consciousness with a battery change while standing, a significant knee effusion after vigorous ergometry training, a large skin abrasion from adhesive tape in the patient with amyloidosis and a one-day hospital re-admission due to dehydration during the summer heat. All patients were fully ambulatory within 10 days and were able to climb stairs within two to three weeks. Two patients required postoperative psychological therapy. All patients were discharged home after a postoperative hospital stay of 49 ± 7 days, and they live independently. One patient has returned to work. All patients’ quality-of-life scores improved two months after implantation.
Current treatment options
Heart transplantation, against which alternative methods should be measured, remains the standard therapy for patients with end-stage HF. Because of a shortage of donors, multiple alternative methods of treatment have been investigated (14–16), including mechanical assist systems that were introduced in the 1960s. In theory, mechanical assist systems could offer advantages over heart transplantation, including the avoidance of complications related to immunosuppression and its availability to more patients in need. The initial enthusiasm for assist systems was followed by a long period of discouragement, due to high mortality and frequent complications. The main problems associated with LVADs include selection of appropriate patients, bleeding, RV failure, infection and thromboembolic complications (4). Over 50% of patients receiving a conventional LVAD will have one or more potentially life-threatening complications (5–9). Despite these problems, conventional LVADs have improved survival in terminal HF patients awaiting transplantation (3). The Jarvik-2000 LVAD offers new concepts in its design, potentially reducing device-associated complications. If the results continue to be as encouraging as those in our early experience, this device may become a true option for the long-term treatment of terminal heart disease.
Choosing the right patients for LVAD support poses a difficult problem. Oz et al. (10), at Columbia University, have established a score for risk stratification of patients, which we have used for bedside decision-making. Implantation of the LVAD in patients with multi-system organ failure will result in a prohibitively high mortality rate, regardless of the device chosen (10). In the early phase of device testing, we preferred to offer this therapy to patients with a risk score index ≤5 in order to avoid offering this expensive therapy to patients who are unlikely to benefit and to recognize the potential morbidity inflicted by the device. It is important to emphasize that the patients in our series were in better preoperative condition than many patients requiring mechanical support. Yet, all of the patients were in dire need of LV support and had contraindications to heart transplantation or conventional ventricular support, as illustrated in Table 2.
Bleeding and RV failure
Conventional LVAD implantation is associated with a high rate of bleeding (4,8), often requiring a repeat operation, thus increasing the risk of postoperative RV failure (17). In our patients, we noted no bleeding complications, no RV failure and a low transfusion requirement after Jarvik-2000 implantation. Several factors might have been responsible for this: the surgeons’ awareness of the high morbidity associated with bleeding, the routine use of aprotinin, a smaller surgical procedure, compared with conventional LVAD implantation, using a lateral thoracotomy, short cardiopulmonary bypass times and the absence of severe preoperative coagulopathy. Aprotinin reduces the transfusion requirements and mortality after LVAD implantation and also potentially reduces the incidence of RV failure (17). In addition, unloading of the right ventricle with routine use of nitric oxide (5 to 80 ppm) decreases the pulmonary artery pressure and improves LVAD flow, although no beneficial effect on mortality has been shown (18,19).
Infections occur in 30% to 50% of patients in a series of conventional LVAD implantations (5,7,20). Drive-line and peri-implant infections, as well as LVAD endocarditis, cannot generally be cleared without removing the device at the time of transplantation. The majority of infections associated with LVAD do not appear to have a negative effect on the outcomes of heart transplantation (5). We have encountered no infectious complications after Jarvik-2000 implantation. Factors that may potentially reduce the risk of infection include a strict infection control policy and a small amount of surgically implanted foreign material. The different power supply by means of an immobile skin exit site in the left retro-auricular area probably has been the most important factor (Fig. 1B). Clinical experience with early cochlear implants, using similar percutaneous skull-mounted pedestals for power supply, demonstrated a low rate of complications, including infection (21). Other advantages of this mode of power supply include only minimal interference with daily activities and the waterproof power plug, which allows the patients to shower. This skull-mounted power connector will soon be released for clinical testing in the United States.
Thromboembolic events are often devastating complications after successful LVAD implantation. To potentially reduce thromboembolic events, the Jarvik-2000 is implanted directly into the LV with no inflow cannula, and the outflow graft is anastomosed into the descending thoracic aorta. The device is set at a minimal pump speed to allow aortic valve opening, thus preventing potential thrombus formation in areas of stasis in the LV outflow tract and aortic root. All of these factors can potentially reduce the incidence of thromboembolism. One of our patients had a transient ischemic attack, likely due to a small embolus. At the time, his INR was subtherapeutic at 1.8. Based on this single event, we did not change the anticoagulation regimen, adjusting Coumadin to an INR of 2.0 to 2.5. To date, we intend to avoid giving platelet inhibitors together with Coumadin because of the increased risk of bleeding complications in other patient cohorts (22).
Chronic nonpulsatile flow
The effects of chronic nonpulsatile flow are unknown; therefore, we avoided exposing our patients to it. The Jarvik-2000 is set to allow aortic valve opening with a palpable pulse. The concept of this device is new, as it is truly assisting and not replacing LV function, keeping the ventricle in a slightly loaded state. It should therefore not be used in patients with virtually absent contractile function. Conventional LVADs are more suitable to support such patients. To be more accurate, they should probably be called “LV replacement systems,” as they completely replace native heart function, with the aortic valve permanently closed. Interestingly, the Jarvik-2000 alone appears to maintain, at least temporarily, sufficient cardiac output, as our patient with sustained ventricular tachycardia remained fully awake.
Hemolysis is one of the main concerns of axial flow pumps, given the devices’ tremendous rotational speeds (8,000 to 12,000 rpm). As illustrated in Figure 3, we found evidence of mild blood trauma, likely due to mechanical forces affecting the cellular blood components. Hemolysis does not appear to pose a clinical problem thus far, as no patient required a transfusion or had complications due to increased red-cell turnover. We will perform ultrasonography of the gallbladder every six months because of the increased risk of pigment gallstone formation and related complications.
Many patients require psychological therapy after LVAD implantation (23). Denial is a common coping mechanism, but patients with an LVAD cannot ignore or deny their condition. The patient has to change the batteries several times a day and is constantly reminded of the device. Professional psychological help, focusing on coping strategies and treating depression, was perceived as beneficial by two of our patients. All patients had an improved quality-of-life score (Table 3).
We can only speculate on the long-term prognoses of these patients. It will depend on the device’s durability, the incidence of late complications and the course of the underlying myocardial disease. The wear and tear of the device is very low according to the manufacturer (Jarvik Heart, Inc., New York, New York). We are attempting to keep late complications at a minimum, with close outpatient multidisciplinary follow-up. Long-term unloading could potentially allow partial myocardial recovery or stabilize the disease. Changes in myocardial gene expression have been described with long-term unloading (24). Only a few patients with a conventional LVAD implanted because of chronic HF had sufficient myocardial recovery to allow device explantation. Many of them later developed recurrent HF (25). Native myocardial function is crucial, as the Jarvik-2000 only assists the LV and depends on the heart’s contractility. Therefore, continuous HF treatment with administration of angiotensin-converting enzyme inhibitors and beta-blockers appears to be essential in the long-term management of these patients.
Our early experience with the permanent Jarvik-2000 LVAD has been encouraging. It can be used in patients with dilative or restrictive disease. There were no device failures. All patients were discharged home. We noted a low incidence of complications. The Jarvik-2000 might prove to be a valid tool in the physician’s armamentarium for the long-term treatment of severe HF.
- computed tomography
- heart failure
- international normalized ratio
- left ventricle/ventricular
- left ventricular assist device
- right ventricular
- transesophageal echocardiography
- wood units
- Received November 8, 2001.
- Revision received February 26, 2002.
- Accepted March 1, 2002.
- American College of Cardiology Foundation
- El-Banayosy A.,
- Deng M.,
- Loisance D.Y.,
- et al.
- Costanzo M.R.,
- Augustine S.,
- Bourge R.,
- et al.
- McCarthy J.F.,
- McCarthy P.M.,
- Starling R.C.,
- et al.
- Oz M.C.,
- Argenziano M.,
- Catanese K.A.,
- et al.
- Parkin J.L.