- Email: [email protected]
Multicenter clinical evaluation of the HeartMate 1000 IP left ventricular assist device
Multicenter Clinical Evaluation of the HeartMate 1000 IP Left Ventricular Assist Device 0. H. Frazier, MD, Eric A. Rose, MD, Quentin Macmanus, MD, Nelson A. Burton, MD, Edward A. Lefrak, MD, Victor L. Poirier, BSME, MBA, and Kurt A. Dasse, PhD Department of Cardiovascular Surgery, Texas Heart Institute, Houston, Texas; Division of Cardiothoracic Surgery, Columbia Presbyterian Hospital, New York, New York; Virginia Heart Center at Fairfax Hospital, Falls Church, Virginia; and Department of Research and Development, Thermo Cardiosystems Inc, Woburn, Massachusetts The Thermo Cardiosystems Inc (Woburn, MA) HeartMate 1000 IP left ventricular assist device (LVAD) has been evaluated as a bridge to transplantation in 34 patients for up to 324 days at seven clinical centers in the United States. Sixty-five percent of the patients underwent transplantation, 80% of whom were discharged from the hospital. Six additional control patients, transplant candidates who met the entrance criteria but who did not receive the device, were also included in the study. Although 3 (50%) of the control patients received transplants, all 6 died within 77 days of having met the LVAD inclusion criteria (100% mortality). Complications resulting from use of the device were comparable with those previously reported for all ventricular assist devices, except for thromboembolic events: bleeding, 39%; infection, 25%; and right heart failure, 21%. No devicerelated thromboembolic events occurred, although 1 patient experienced an event related to a mechanical aortic valve in the native heart. None of the complications had a significant negative association with outcome of the patient except for right heart failure. All survivors had a significant improvement in hepatic function before trans-
plantation. Total bilirubin values were reduced by 60% during LVAD support. No significant differences were observed when total bilirubin values were compared at 30 and 60 days after LVAD support and at 30 and 60 days after transplantation in a cohort of 15 patients (p > 0.05). The improvement in renal function was less predictable than that of hepatic function. Creatinine values decreased significantly before transplantation; however, the values measured at 30 and 60 days after transplantation were higher than those measured at the same intervals after LVAD support had been initiated, and this increase is presumably related to the immunosuppressive drugs. In conclusion, the HeartMate 1000 IP LVAD has been shown to be effective in supporting end-stage cardiomyopathy patients to transplantation. Thromboembolism, previously regarded as a serious complication with such devices, has not been a problem with this device. Additional patients are being enrolled into the study to further document the safety and effectiveness of this technology.
A
are undergoing clinical evaluation under Food and Drug Administration-approved investigational device exemptions [12, 131. Both devices have been successfully employed as a bridge to transplantation for durations exceeding 10 months, with the majority of patients receiving transplants and achieving long-term survival. The purpose of this report is to summarize the results obtained with one of these devices, the Thermo Cardiosystems Inc HeartMate 1000 IP pneumatically driven LVAD, which has been under evaluation since August 1985. To extend the use of LVADs to the clinical community, sufficient data must be gathered and presented to the Food and Drug Administration to prove, scientifically, that the device is safe and effective. To be considered safe, the device must be reliable and must not jeopardize the ability to perform transplantation in the patient; that is, the device should not be associated with unacceptable levels of adverse effects, such as infection, bleeding, hemolysis, end-organ dysfunction, or thromboembolic complications. To be considered effective, the device must improve the hemodynamic status of the patient, enhance the likelihood of survival, and, ideally, promote a higher
lthough cardiac transplantation has become a widely accepted therapy for patients with end-stage cardiac disease, the demand for donor organs far exceeds the supply [l].Consequently, a substantial percentage of transplant candidates die while waiting for a donor heart [Z, 31. Temporary mechanical circulatory support for patients waiting for donor organs was introduced by Cooley and co-workers in 1969 [4]. Subsequent research led to the development of ventricular assist devices for use as a bridge to transplantation [5-81. The results with these devices have been highly encouraging for the support of postcardiotomy [9, lo], postinfarction [ll],and end-stage cardiomyopathy patients [12, 131. Among the currently available left ventricular assist devices (LVADs), two are specifically designed for longterm mechanical circulatory support. Thermo Cardiosystems Inc (Woburn, MA) and Novacor (Oakland, CA) have developed implantable, pusher-plate-type LVADs that Accepted for publication Feb 14, 1992 Address reprint requests to Dr Frazier, Texas Heart Institute, PO Box 20345, Houston, TX 77225-0345.
0 1992 by The Society of
Thoracic Surgeons
(Ann Thorac Surg 1992;53:1080-90)
0003-4975/92/$5.OO
Ann Thorac Surg 1992;53:108C-90
FRAZIER ET AL CLINICAL EVALUATION OF THE HEARTMATE
1081
quality of life than would be otherwise possible without the device. The following discussion summarizes data obtained to demonstrate the safety and efficacy of this technology to date.
Material and Methods Description of the Device The HeartMate 1000 IP LVAD consists of a pusher-plate blood pump driven by a portable external console via a percutaneous driveline. The pump consists of a titanium housing (titanium, 90%; aluminum, 6%; vanadium, 4%) measuring 11.2 cm in diameter and 4.0 cm in thickness. Inside, a flexible Biomer polyurethane diaphragm is bonded to a rigid pusher plate. The diaphragm divides the pump into two halves: a blood chamber and an air chamber. Programmed pulses of air are delivered from the console to the air chamber behind the pusher-plate diaphragm. As the air accumulates, the diaphragm is displaced, propelling the blood through the outflow graft 'Orcine xenograft into the (25 mm) are placed in the inlet and outlet conduits to ensure unidirectional blood flow. Device implantation is accomplished through a median sternotomy, with the incision extending just above the umbilicus, and cardiopulmonary bypass is instituted through standard techniques. The left ventricular apex is cored, and the opening is reinforced with a sewing ring. The pump is then placed in the left abdominal cavity, below the diaphragm. An incision is made in the diaphragm, through which the inlet conduit is passed, then secured to the apical opening. The outflow graft (preclotted Dacron) is passed over the diaphragm and anastomosed end-to-side to the ascending aorta. Finally, the pneumatic drive line is tunneled through a stab incision in the left lateral abdominal wall, above the iliac crest. After the pump has been implanted, the patient is gradually weaned from bypass, and the median sternotomy is closed. A more detailed description of the device and the implant protocol has been published [12]. The portable console, which operates on batteries or standard alternating current power, drives the blood pump directly without the need for tanks of compressed air. Both asynchronous and synchronous control modes of operation are available. The asynchronous modes include fixed-rate and pump-on-full options. If set to operate in the pump-on-full mode, the device will automatically eject when the pump is approximately 90% full. Synchronization of the pump is possible using an external QRS detector. The combination of an implanted blood pump with a portable external console allows the patient to ambulate and exercise (Fig 1). Patients are encouraged to begin an exercise program consisting of walking and stationary bicycling as soon as possible after implantation to improve their general physiologic status before transplantation.
Fig 1. Patients are encouraged to begin ambulating and exercising as soon as possible after LVAD implantation to improve their physiological status before transplantation,
Blood-Contacting Surfaces and Recommended Anticoagula tion The HeartMate LVAD employs textured biomaterials to interface with the blood. Sintered-titanium microspheres are used on the pump housing and conduits, and integrally textured polyurethane is used on the flexing pusher-plate diaphragm. The rationale for using textured surfaces is to encourage the formation of a thin, well-adhered, pseudointimal lining on the inside of the pump. The resultant biologic lining serves as the primary interface with the blood throughout implantation. The need for anticoagulation with this device has been greatly reduced by use of porcine bioprosthetic valves in combination with the textured surfaces. In the majority of patients, an antiplatelet regimen consisting of 80 mg aspirin, once a day, and 75 mg dipyridamole, three times daily, has been used beyond the intraoperative period. Heparin, sodium warfarin (Coumadin), or both have been used only during implantation and in patients with mechanical valves in the native heart.
Patient Selection Use of the device was limited to approved transplant candidates who met the hemodynamic indications for use and who were not subject to exclusionary criteria. The hemodynamic indications included a pulmonary capillary wedge pressure of 20 mm Hg or greater coupled with either a cardiac index of 2.0 L min-' m-' or less or a systolic blood pressure of 80 mm Hg or less. The exclusion criteria included chronic, irreversible hepatic, renal, and respiratory failure; severe blood dyscrasia; and right heart failure. All data were obtained in compliance with proto-
-
1082
FRAZIER ET AL CLINICAL EVALUATION OF THE HEARTMATE
Ann Thorac Surg 1992;53:108C-90
Table 1. Patient Demographics and Results of Treatment with the HeartMate LVAD Patient No.
Age Sex
(Y)
Admission Diagnosis
Implant Duration (days)
Patients W h o Met the Study Selection Criteria 1 M 47 Isch CMP 2 M 48 Idio CMP 3 M 17 Viral CMP 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
M M M M M M M M M M M M M M M M M M M M M M F
22 39 43 42 55 41 49 55 44 55 37 48 47 47 22 46 57 50 39 60 45 46 35
Idio CMP Isch CMP Idio CMP Idio CMP Idio CMP Idio CMP Isch CMP Isch CMP Idio CMP Idio CMP Isch CMP Isch CMP Dil CMP Idio CMP Viral CMP Isch CMP MI Idio CMP Dil CMP Idio CMP Isch CMP Idio CMP Viral CMP
41 37 7
Yes Yes Yes
35 132 114 233 189 31 220 14 1 61 153 10 1 38 84 15 21 66 6 4 Ongoing Ongoing Ongoing
Yes Yes Yes Yes Yes Yes Yes No No Yes Yes No No Yes Yes Yes Yes Yes No No
...
... ...
5 19 25 1 84 1 1 1
No Yes Yes No Yes No No No
No No (multiorgan failure, sepsis) No (donor heart failure) No Yes No No No
Patients Who Did Not Meet the Study Selection Criteria 1 M 55 Idio CMP 2 M 53 Idio CMP 3 M 37 Idio CMP 4 M 59 Isch CMP 5 M 52 Isch CMP 6 M 51 Idio CMP 7 M 55 MI 8 M 38 MI a Cause of death appears in parentheses. CMP = cardiomyopathy; Dil = dilated;
Idio
=
idiopathic;
SUrviVOP (>60 days)
Transplanted
Isch = ischemic;
cols approved by the investigational review boards at each of the respective institutions. Between August 1985 and February 1991,34 patients at seven medical centers were treated with the HeartMate 1000 IP LVAD as a bridge to transplantation. All but 1 of the patients were male, and their ages ranged from 17 to 62 years. Of the 34 patients, 26 met the study selection criteria, and 8 did not for various reasons (Table 1).Whereas all 34
...
...
No (liver failure) Yes No (respiratory failure; adverse OKT3 reaction) Yes Yes Yes Yes Yes Yes Yes No No Yes Yes No No Yes Yes Yes Yes Yes No No
...
MI = myocardial infarction.
patients were considered in the overall evaluation of survival and device safety, only the 26 who met study selection criteria were considered in the analyses of hemodynamic, hematologic, hepatic, and renal response to the pump. Of these 26 patients, 15 were long-term survivors (>60 days) of cardiac transplantation and were discharged from the hospital. Data collection in these patients allowed further comparison of hemodynamic, hematologic, hepatic, and renal values at 30 and 60 days
FRAZIER ET AL CLINICAL EVALUATION OF THE HEARTMATE
Ann Thorac Surg 1992;53:1080-90
1083
Table 2. Patient Demographics and Results of Treatment in Six Control Subjects ~
Patient
~~
Age
Admission Diagnosis
No.
Sex
(Y)
1 2
M M F F M M
21 38 31
Idio CMP Isch CMP
44
Isch CMP Isch CMP
3
4 5 6 a
52 57
Survivor"
Transplanted
(>60 days)
Yes
No (multiorgan failure) No No (allograft rejection) No
No Yes No Yes No
PPart CMP
Isch CMP
Yes No
Cause of death appears in parentheses.
CMP
=
cardiomyopathy;
Idio
=
idiopathic;
Isch = ischemic;
Wart = postpartum.
of device support and at the same times after transplantation.
Control Population Six historic controls were entered into the study (Table 2).
They were identified by searching the transplant database for candidates who would have met the patient selection criteria for LVAD support but who did not receive treatment with the device because it was not available.
Data Acquisition Hemodynamic data, including the cardiac index, pulmonary capillary wedge pressure, and blood pressure, were collected. Hematocrit levels, plasma free hemoglobin levels, and platelet counts were monitored to assess the hematologic response to circulatory assistance. Hepatic and renal function were assessed through total bilirubin levels, serum glutamic-oxaloacetic transaminase and serum glutamic-pyruvic transaminase values, and creatinine and blood urea nitrogen levels. These data were then compared before and during ventricular assistance, as well as after transplantation. The change in New York Heart Association functional status was also assessed before support and at 60 days after transplantation in the 15 patients who survived more than 60 days. Safety data pertaining to the incidence of bleeding, hemolysis, infection, right heart failure, peripheral end-organ failure, and mechanical failure were acquired for all of the patients.
Statistical Analyses The hemodynamic, hepatic, and renal function entrance criteria were analyzed using nonpaired t tests. Nonpaired t tests were also used to compare hepatic and renal function in survivors and nonsurvivors before and after LVAD treatment. Paired t tests were used to compare LVAD and transplantation data in a sample of 15 patients who were treated with the LVAD and then underwent transplantation. Finally, the survival and complication data were analyzed with the Fisher's exact probability test. A p value of less than 0.05 was considered significant in all analyses.
LVAD patients. Three of the 34 patients remained on LVAD support at the time of study. Twenty (65%)of the remaining 31 patients received transplants, 16 (80%) of whom were discharged from the hospital. Survival in the 16 patients ranges up to 3 years after the transplant procedure. Of the 6 control patients, 3 (50%)died before transplantation at 5, 12, and 34 days after having met the hemodynamic indications for LVAD support. The remaining 3 patients underwent transplantation but died at 2, 21, and 77 days after operation. Therefore, all 6 patients died within 77 days of having met the indications for LVAD support, regardless of whether they underwent transplantation. A Fisher's exact probability test revealed that the survival rate of the LVAD-treated group was significantly greater than that of the control group ( p < 0.05). HEMODYNAMIC COMPARISON OF THE STUDY GROUPS AT
The hemodynamic status of the 26 LVAD patients was documented at the time they met the indications for use and within 24 hours before implantation. These data were then compared with the hemodynamic values of the control patients when they met the study entrance criteria. No differences in cardiac index or systolic blood pressures were noted between the LVAD and control patients either at the time of meeting the study entrance criteria or at the time of LVAD implantation (p > 0.05). However, the control group had a significantly lower pulmonary capillary wedge pressure than the LVAD group at both points. When the data from the control patients were compared with those of the LVAD patients at the time of implantation, the average hemodynamic values (f the standard deviation) were, respectively, as follows: pulmonary capillary wedge pressure, 23 f 2 versus 28 8 mm Hg; cardiac index, 1.9 f 0.8 versus 2.1 f 0.6 L * min-l m-'; and systolic blood pressure, 90 f 19 versus 94 2 19 mm Hg. These data imply that the two groups were hemodynamically comparable, although the control group had a slightly lower wedge pressure. ENTRANCE.
*
-
Figure 2 illustrates the average pump index (pump flowbody surface area) and aortic pressures for the 26 LVAD patients who met the entrance criteria. The average pump index for all patients was 2.86 L min-' m-', which was approximately 30% greater than the average cardiac index at the HEMODYNAMIC PERFORMANCE OF THE LVAD.
Results Evaluation of Device Effectiveness SURVIVAL DATA. Table 1 lists the age, sex, admission diagnosis, implant duration, and outcome for the 34
-
-
1084
FRAZIER ET AL CLINICAL EVALUATION OF THE HEARTMATE
Ann Thorac Surg 1992;53108CL90
Avg. Pump Index (Llmlnlm2)
l
+
80
.i
*
*
8o
2
ol
60
100
160
200
Postoperative Days 3oo iWg. Aortic Prerrun (mmHg)
2501
' 0
-
+* ++
I
,
+
+++++ + +++ ++ , +++++++ + ++,++++++ ,
++
250
200
0
t
20
0
PHgb
8ool 800
50
100
150
200
"
250
0
Postoperative Days
50
100
150
200
250
Postoperative Days
Fig 2 . Hemodynamic measurements during LVAD support in 26 patients. The average pump index (pump flowlbody surface area) was 2.86 L * min-' * m-'. The average pressures were as follows: systolic, 119 mm Hg; diastolic, 71 mm Hg; and mean pressure, 95 mm Hg.
Fig 3 . Effects of LVAD support on hematologic indices in 26 patients. Average hematocrit (HCT) was 34%; average plasma free hemoglobin (PHgb) level, 8.7 mgldL; and average platelet count, 249,000lmL.
time of implantation ( p < 0.05). The systolic, diastolic, and mean aortic pressures were 119 mm Hg, 71 mm Hg, and 95 mm Hg, respectively. The device was found to be fully capable of generating adequate pressures and flows.
Further evidence that hepatic function improved during LVAD support is seen in Figure 6, where the values of total bilirubin before implantation were compared with the final values obtained either just before transplantation or at the time of death. In this analysis, data from 3 of the
EFFECTS OF LVAD SUPPORT ON HEMATOLOGIC INDICES. The average plasma free hemoglobin level was 8.7 mg/dL in the 26 LVAD patients, suggesting that a mild, but clinically acceptable level of hemolysis occurred in these patients (Fig 3). This finding was further confirmed by hematocrit values averaging 34% and platelet counts averaging 249,00O/mL. Adequate circulatory support was provided without serious damage to the blood.
FUNCAll but 2 of the 26 LVAD patients had elevated total bilirubin values (21.4 mg/dL) or serum glutamicoxaloacetic transaminase and/or serum glutamic-pyruvic transaminase values (250 U/L) before or during device support. Figures 4 and 5 illustrate the average hepatic values as a function of time. All three parameters tended to increase during the first month, then returned to normal after approximately 2 months of augmented perfusion. The values remained within normal limits for the remaining period of support. EFFECTS OF LVAD SUPPORT ON HEPATIC AND RENAL
TION.
Avg. Total Bilirubin (mgldl)
lo
I
8-
*
z x
4 - *
* * 20
** *** 60
100
150
200
250
Postoperative Days Fig 4 . Effect of LVAD support on hepatic function in 26 patients. The average total bilirubin value initially increased, then returned to normal by day 50.
FRAZIER ET AL CLINICAL EVALUATION OF THE HEARTMATE
Ann Thorac Surg 1992;53:1080-90
Avg. SQOT (UIL) 600,
1
500t* 400
*
y
-
300 200
w" - *
100
-
#
0
* ****x
'
...........................
600
500
-u
400
-
300
-
*
200
- **
100
-
c*
+*
X*
*
Y X"********f****XY
0
**** *****
'
1085
either failed to decrease or increased in all nonsurvivors. Hepatic function was improved in all patients with successful transplantations, and 3 patients remained on device support at the time of study. The average bilirubin value obtained just before pump removal divided by the baseline value yields a ratio that indicates whether the total bilirubin increased (>1.0), decreased (<1.0), or stayed the same (1.0) in response to the treatment. The bilirubin ratio for the LVAD survivors (0.40) was significantly reduced compared with that of the LVAD nonsurvivors (3.6; p = 0.03), as well as the control patients (1.3; p = 0.004). In contrast, no significant difference existed between the controls (1.3) and the LVAD nonsurvivors (3.6; p = 0.09). These data demonstrate that LVAD support leads to improvement in hepatic function in the majority of patients, providing that dysfunction is reversible. Hepatic function in the nonsurvivors failed to improve. These patients had significantly higher total bilirubin values at implantation, and the duration of support may have been too short for them to benefit from improved perfusion. The hepatic function of the controls also failed to improve in the absence of mechanical circulatory support. Figure 7 illustrates the change in serum creatinine and blood urea nitrogen values in response to ventricular assistance. In contrast to the hepatic response, the renal parameters did not transiently increase after LVAD implantation. Moreover, the renal parameters stabilized within a shorter period of time. COMPARISON OF HEPATIC A N D RENAL FUNCTION DURING
In a cohort of 15 LVAD patients surviving more than 60 days after transplantation, three hepatic and two renal parameters were measured at LVAD implantation, at days 30 and 60 during support, immediately before transplantation, and at days 30 and 60 after transplantation. A t test was then performed to compare the effects of ventricular assistance and transplantation on hepatic and renal function. The only significant difference in hepatic function during LVAD support versus that after transplantation was seen in the total bilirubin level. The LVAD patients had a significantly greater total bilirubin value at implantation compared with that on the day of transplantation. Figure 8 illustrates the change in total bilirubin values after implantation of the LVAD versus after transplantation. The baseline values (t = 0) obtained just before implantation of the LVAD were significantly greater than those obtained just before transplantation ( p < 0.05). Hepatic function was markedly improved after ventricular assistance, as was the overall physiologic status of the patients before transplantation. No significant differences were noted in the hepatic parameters at days 30 and 60 of LVAD support or at the same times after transplantation. Figure 9 illustrates the change in renal values as a function of LVAD treatment versus transplantation. The creatinine values were significantly greater before LVAD implantation than before transplantation, demonstrating that the LVAD improved the condition of the patient LVAD SUPPORT AND AFTER TRANSPLANTATION.
patients were excluded because treatment was ongoing. The remaining 23 patients were classified as survivors (n = 15) and nonsurvivors (n = 8). Whereas the total bilirubin values decreased in all survivors, the values Total Bilirubin (mgldi) NONSURVIVORS
25 20
10 5 n "
17
18
10
21
02
05 04
08
07
08
09
10
13
20
22
U
01
03
ti
11
15
23 w)
Patient Number
aPRE-IMPLANT
FINAL
Fig 6 . Total bilirubin values were significantly reduced after LVAD support in the survivors. Hepatic function continued to deteriorate in the nonsurvivors despite restoration of blood flow.
1086
FRAZIER ET AL CLINICAL EVALUATION OF THE HEARTMATE
Ann Thorac Surg 1992;53108&90
64-
2t
4
01
3-
*************** * ******* ~********* '
SO
0
100
150
200
I
250
Postoperative Days
0'
0
m
90
20
40
50
60
Postoperative Days
Avg. BUN (mgldl)
2 8
Avg. Total Bilirubin (mgldl)
P08TTRANIPLANT
I 0
50
100
150
200
250
Postoperative Days
0 0
10
20
30
40
50
60
Postoperative Days
Fig 7. Effect of LVAD support on renal function in 26 patients. The average creatinine and blood urea nitrogen (BUN) values decreased after implantation and remained within normal limits for the remainder of implant duration.
Fig 8. Comparison of hepatic function after implantation and after transplantation. The average total bilirubin value was significantly greater at LVAD implantation (top) compared with that just before transplantation (bottom) in a cohort of 15 patients. There was no significant difference on days 30 and 60 after each procedure.
before transplantation ( p < 0.05). Unlike the hepatic indicators, the creatinine values were significantly elevated at days 30 and 60 after transplantation compared with those at the same intervals after LVAD implantation. The elevated creatinine levels were thought to have been a result of immunosuppressive therapy. In general, peripheral end-organ function was adequately supported after both LVAD treatment and transplantation. However, the physiologic status of the transplant candidates was markedly improved after LVAD support, enhancing the likelihood of success of transplantation.
multitude of environmental factors that can have a positive or negative impact on the performance of the device. Adverse effects were monitored for all patients. Of the 34 patients bridged to transplantation, 3 were treated for less than a day and were considered compassionate exemptions, and these patients experienced no adverse effects. Three additional patients remained on LVAD support at the time of study. The safety data reported in this article were collected from the remaining 28 LVAD patients who completed the study. Data are also presented for the six control patients entered into the study. Each category of adverse effects is described in the following sections.
Safety Evaluation A number of risks are associated with the use of ventricular assistance technology. The principal risks currently being monitored include bleeding, hemolysis, infection, right heart failure, peripheral end-organ dysfunction, thromboembolism, and mechanical failure. The frequency and nature of adverse effects resulting from the use of these devices depend on numerous factors, including the basic design of the device, the skill and experience of the clinical team, the physiologic status of the patient, and a
BLEEDING. Bleeding occurred in all of the patients. However, the bleeding was considered serious only if it was severe enough to require returning the patient to the operating room. None of the patients bridged to transplantation experienced device-related bleeding. Eleven patients (39%) experienced patient-related bleeding, such as that resulting from cardiac tamponade. Seven (64%)of these patients underwent transplantation, 5 (71%) of whom were long-term survivors. Three (50%)
FRAZIER ET AL CLINICAL EVALUATION OF THE HEARTMATE
Ann Thorac Surg 1992;531080-90
Avp. Cnatinina (mgldi)
6
fever, combined with the need for antimicrobial treatment. An infection was considered device-related if the specific organism that was cultured had not been detected preoperatively. Typical device-related infections involve the drive line or positive cultures isolated from the blood pump, coupled with the symptoms already described. Based on the specified criteria, 7 (25%) of the 28 patients bridged to transplantation experienced a device-related infection. Six (85%)of the patients underwent transplantation, 4 (67%) of whom were long-term survivors. No significant difference existed between the outcome of infected and noninfected patients according to a Fisher’s exact probability test ( p > 0.05), suggesting that infection may not be a critical determinant of outcome. Two of the 6 control patients (33%)experienced infections while awaiting transplantation. Although these patients did not receive an LVAD, the potential for infection was increased because of diagnostic instruments. The number of control patients was low, but the incidence of infection was high. It remains to be determined whether the presence of an LVAD further increases the risk of infection.
ON LUD
-
0
10
20
SO
40
50
60
Postoperative Days 6
Avg. Creatinino (mgldl) POSTTRANSPLANT
st
I 0
10
20
SO
40
50
1087
END-ORGAN DYSFUNCTION. As stated, all but two of the patients bridged to transplantation experienced renal or hepatic dysfunction or both before or during LVAD support. The dysfunction was not considered device-related in any of the patients.
60
Postoperative Days
Fig 9. Average creatinine values were significantly greater at LVAD implantation compared with before transplantation. Creatinine levels were significantly greater on days 30 and 60 after transplantation than during ventricular support, presumably caused by immunosuppression.
of the 6 patients who died had been treated simultaneously with a Biomedicus (Biomedicus Inc, Eden Prairie, MN) right ventricular assist device. A statistical comparison of the patients who did versus those who did not bleed revealed no significant difference in outcome between the two groups with respect to transplantation and survival rates (Fisher’s exact probability test; p < 0.05). HEMOLYSIS. The average plasma free hemoglobin values for all the patients was 8.7 mg/dL. The hemoglobin concentration was 11.0 g/dL. Only 1patient, a 50-year-old man who had a plasma free hemoglobin value of 40 mg/dL just before LVAD implantation, experienced hemolysis. His plasma free hemoglobin level exceeded 40 mg/dL on two consecutive occasions after implantation. He remained on the device for 66 days, underwent successful transplantation, and is currently alive and well several months after cardiac transplantation. Why hemolysis occurred before and after LVAD implantation is uncertain. INFECTION. Infection was defined through detection of a positive culture, elevated white blood cell count, and
RIGHT HEART FAILURE. Six (21%)of the 28 patients bridged to transplantation either required right ventricular assistance or exhibited symptoms of serious right ventricular dysfunction after implantation of the LVAD. Two of the patients experienced unanticipated right ventricular failure at the time of LVAD implantation but did not receive a right ventricular assist device. Although the LVAD is capable of generating flow rates of 6.0 L/min with 0.0 mm Hg inlet pressure, a low-flow condition persisted in both of the patients because of poor filling, and they died in the operating room. The inability to generate adequate flow under these circumstances was thought to be caused by an elevated pulmonary vascular resistance. The remaining 4 patients required right ventricular assistance, which was accomplished with the extracorporeal Biomedicus centrifugal pump. Two of the patients were electively weaned from the Biomedicus pump after 5 and 6 days, and 1of these patients received a transplant. The other 2 patients requiring right ventricular assistance experienced progressively increasing pulmonary vascular resistance secondary to bleeding. All patients requiring right heart assistance ultimately died. Among the adverse effects, right heart failure was the only complication that had a significantly negative correlation with outcome. When the patients with versus without right heart failure were compared, the transplantation rate and the survival rate were significantly lower for the patients who experienced right heart failure ( p < 0.05).
No device-related thromboemboli were observed in any of the patients. One patient, how-
THROMBOEMBOLISM.
1088
FRAZIER ET AL CLINICAL EVALUATION OF THE HEARTMATE
ever, did experience an embolic event related to a mechanical aortic valve located in the natural heart. The patient recovered, underwent successful transplantation after 173 days of LVAD support, and was discharged home. Bowel adhesions to the Dacroncovered drive line occurred in 2 patients in whom the lead was allowed to course intraabdominally. Both patients underwent transplantation, but these two incidences emphasize the need to minimize contact between the bowel and the textured materials, such as Dacron.
BOWEL ADHESIONS.
MECHANICAL FAILURE. In 1988, a threaded outflow connector loosened after LVAD implantation in 1 patient, who required reoperation to tighten the connector. The design of the device was modified to incorporate a locking mechanism, and no additional adverse effects have been observed since that incident. This was the only failure in 7 patient-years of experience.
OVERALL. In summary, the frequency of device-related adverse effects has been low. Although a variety of patient-related adverse effects were observed, none aside from right heart failure had a negative correlation with outcome.
Quality of Life The New York Heart Association classification system was used as a measure of quality of life in response to device support. All the LVAD-treated patients were in functional class IV at the time of implantation. Sixteen of the LVAD patients were alive at 60 days after transplantation, at which time 15 were in class I and 1 was in class 11. At the time of meeting study entrance criteria, 5 of the 6 control patients were in class IV, and the sixth patient was in class 111. Only 1 of the control patients survived to 60 days after transplantation and was determined to be in class IV at that time. These data demonstrate that the quality of life of the LVAD-treated patients, based on functional class, was significantly improved during ventricular assistance and after transplantation. Also, the functional class of the LVAD-treated patients was markedly improved when compared with that of the control patients.
Comment In contrast to extracorporeal LVADs that have been largely employed to support patients in acute cardiogenic shock or postinfarction, the HeartMate device has been used primarily for the extended support of patients with chronic cardiomyopathy who require a bridge to transplantation. Efficacy has been demonstrated by our study, in which the device provided adequate hemodynamic support until transplantation, increased survival, and improved the New York Heart Association functional class status of the patients. In comparison with the unsupported controls, the LVAD-treated patients had a
Ann Thorac Surg 1992;53:108C-90
better outcome. The majority of the LVAD-treated patients received transplants (65%), and 80% were discharged from the hospital after transplantation. The ongoing clinical study will determine if the HeartMate LVAD performs in a safe and reliable fashion. Adverse effects, including bleeding, hemolysis, infection, peripheral end-organ dysfunction, and thromboembolic events, have not had a negative correlation with the rate of transplantation or survival. In contrast, acute perioperative right heart failure was associated with a significantly greater mortality. In comparing complication rates with the HeartMate device with those reported in the Combined Registry [14], two important observations were noted. First, the average duration of support for the HeartMate patients was almost twice as long as that for the patients included in the Combined Registry report: 53 days versus 27.7 days. Second, the HeartMate patient population was composed primarily of end-stage cardiomyopathy patients, whereas those included in the Registry report were a combination of postcardiotomy, postinfarction, and cardiomyopathy patients. In general, the patients included in the Combined Registry report were experiencing acute left ventricular failure, whereas the HeartMate population was composed primarily of patients experiencing chronic left ventricular failure. The nature and frequency of complications with the HeartMate were comparable with those reported for other devices, although the incidence of thromboembolic episodes was lower with the HeartMate. Previous reports estimate patient-related bleeding to occur in approximately 40% of LVAD-supported patients [13, 151, compared with a 39% rate of patient-related bleeding after implantation of the HeartMate. Infectious complications were more difficult to compare because no uniformity for reporting these data exists. Reported rates of infection range from approximately 20% to 75%, although not all of the infections were device-related [13, 15, 161. In patients treated with the HeartMate, the incidence of devicerelated infection was 25%, yet 85% of the patients underwent transplantation, and their survival rate was 67%.It is noteworthy that a recent review of the literature from 12 transplant centers, encompassing 384 transplant patients treated with cyclosporine, showed that 57.6% of the patients experienced infections, with 5.2%dying of infection-related causes [17]. Two (33%)of the 6 control patients in the current study also had infections. These data appear to indicate that although the rate of infectious complications in both the LVAD-treated patients and the transplant patients is high, the number of infectionrelated deaths associated with both forms of treatment was relatively low. Whereas mediastinal infection was the primary cause of death in 40% of patients receiving a total artificial heart as a bridge to transplantation [18, 191, the risk of infection-related death with ventricular assistance appears significantly lower. Thromboembolic complications have been a topic of great concern in the past decade, particularly in regard to the Jarvik-7total artificial heart [20,21]. Numerous reports of thromboembolic complications have also appeared in the ventricular assist device literature, with the rate rang-
Ann Thorac Surg 1992;53:1080-90
ing from 33% to 47% [15, 16, 22, 231. No device-related thromboembolic events were observed in the HeartMate patients despite the use of minimal anticoagulative therapy. The only incident that occurred was related to a mechanical aortic valve in the native heart. The presence of a mechanical aortic valve may be considered a contraindication for use of an LVAD. The absence of thromboembolic complications with the HeartMate represents a major step forward for circulatory support systems toward minimizing the risk of device-related strokes. Although it has been stated that left ventricular assistance may be potentially detrimental to right heart function [24, 251, additional reports have shown that impaired right ventricular function improves during left ventricular assistance [26]. Right ventricular function was indeed found to improve with left ventricular support in the majority of patients in this study, as evidenced by increased right ventricular ejection fractions. However, approximately 20% of the patients experienced severe right ventricular dysfunction at implantation, which increased mortality and morbidity in comparison with patients who had no right ventricular involvement. The inability to fill the pump in the presence of right ventricular failure has been attributed to concomitant elevated pulmonary vascular resistance. Use of prostaglandins [27] combined with other pharmacologic measures may aid in reducing pulmonary vascular resistance, but additional study is required. Currently, right ventricular dysfunction appears to be a significant contraindication for left ventricular support. The onset of renal failure shortly before or during ventricular assistance has been suggested to be highly predictive of a poor prognosis [28]. However, both the Thermo Cardiosystems Inc and Novacor devices have successfully supported patients who were anuric before implantation and on hemodialysis for prolonged periods (months) followed by successful transplantation and survival [13]. Renal failure has been regarded as a contraindication for patients in acute left ventricular failure, but it is not clear that it should be considered a contraindication for chronic cardiomyopathy patients. Hepatic function was shown to improve significantly with hemodynamic support, but up to 2 months of LVAD support was required for hepatic function to return to normal. The patients had better hepatic and renal function just before transplantation than before LVAD implantation. This improvement was a result of using the device. Beyond implantation, no difference in hepatic function was observed when the two treatments were compared. In contrast, creatinine values were shown to be significantly higher in the transplant patients after 30 to 60 days of the operation than in the LVAD patients at comparable times after initiation of ventricular assistance. The rise in creatinine level was attributed to the use of immunosuppressive drugs after transplantation. In conclusion, the HeartMate 1000 IP LVAD has been shown to provide an effective means of supporting patients for periods in excess of 300 days. The study will continue to be expanded to include additional patients and controls to further assess safety and effectiveness of this technology as a bridge to transplantation.
FRAZIER ET AL CLINICAL EVALUATION OF THE HEARTMATE
1089
References 1. Evans RW, Manninen DL, Garrison LP Jr., Maier AM. Donor availability as the primary determinant of the future of heart transplantation. JAMA 1986;255:1892-8. 2. Copeland JG, Emery RW, Levinson MM, Copeland J, McAleer MJ, Riley JE. The role of mechanical support and transplantation in treatment of patients with end stage cardiomyopathy. Circulation 1985;72(Suppl2):7-12. 3. Schuler S, Parnt R, Warnecke H, Matheis G, Hetzer R. Extended donor criteria for heart transplantation. J Heart Transplant 1988;7326-30. 4. Cooley DA, Liotta D, Hallman GL, Bloodwell RD, Leachman RD, Milam JD. Orthotopic cardiac prosthesis for two-staged cardiac replacement. Am J Cardiol 1969;24:723-30. 5. Bernhard WF, Clay W, Shoen FJ, et al. Clinical and laboratory investigations related to temporary and permanent ventricular bypass. Heart Transplant 1983;3:16-25. 6. Hill JD, Farrar DJ, Hershon JJ, et al. Use of a prosthetic ventricle as a bridge to cardiac transplantation for postinfarction cardiogenic shock. N Engl J Med 1986;314626-8. 7. Kanter KR, McBride LR, Pennington DG, et al. Bridging to cardiac transplantation with pulsatile ventricular assist devices. Ann Thorac Surg 1988;46:134-40. 8. Champsaur G, Ninet J, Vigneron M, Cochet P, Neidecker J, Boissonnat P. Use of the Abiomed BVS System 5000 as a bridge to cardiac transplantation. J Thorac Cardiovasc Surg 1990;100:122-8. 9. Pierce WS, Parr GVS, Myers JL, et al. Ventricular-assist pumping in patients with cardiogenic shock after cardiac operations. N Engl J Med 1981;305:1606-10. 10. Pennington DG, Samuels LD, Williams G, et al. Experience with the Pierce-Donachy ventricular assist device in postcardiotomy patients with cardiogenic shock. World J Surg 1985;9:3746. 11. Pae WE Jr, Pierce WS. Temporary left ventricular assistance in acute myocardial infarction and cardiogenic shock: rationale and criteria for utilization. Chest 1981;79:692-5. 12. McGee MG, Parnis SM, Nakatani T, et al. Extended clinical support with an implantable left ventricular assist device. ASAIO Trans 1989;35:614-6. 13. Portner PM, Oyer PE, Pennington DG, et al. Implantable electrical left ventricular assist system: bridge to transplantation and the future. Ann Thorac Surg 1989;47142-50. 14. Miller CA, Pae WE Jr, Pierce WS, et al. Combined Registry for the Clinical Use of Mechanical Ventricular Assist Pumps and the Total Artificial Heart in Conjunction with Heart Transplantation: fourth official report-1989. J Heart Transplant 1990;9:453-8. 15. Pennington DG, McBride LR, Kanter KR, et al. Bridging to heart transplantation with circulatory support devices. J Heart Transplant 1989;8:116-23. 16. Didisheim P, Olsen DB, Farrar DJ, et al. Infections and thromboembolism with implantable cardiovascular devices. ASAIO Trans 1989;35:54-70. 17. Linder J. Infection as a complication of heart transplantation. J Heart Transplant 1988;7390-4. 18. Griffith BP, Kormos RL, Hardesty RL, Armitage JM, Dummer JS. The artificial heart: infection-related morbidity and its effect on transplantation. Ann Thorac Surg 1988;45:409-14. 19. Rice LB, Karchmer AW. Artificial heart implantation: what limitations are imposed by infectious complications? JAMA 1988;259:894-5. 20. Levinson MM, Smith RG, Cork RC, et al. Thromboembolic
1090
21. 22. 23. 24.
Ann Thorac Surg 1992;53:108&90
FRAZIER ET AL CLINICAL EVALUATION OF THE HEARTMATE
complications of the JaMk-7 total artificial heart case report. Artif Organs 1986;10:23W. DeVries WC, Anderson JL, Joyce LD, et al. Clinical use of the total artificial heart. N Engl J Med 1984;310:27M. Termuhlen DF, Swartz MT, Pennington DG, et al. Thromboembolic complications with the Pierce-Donachy ventricular assist device. ASAIO Trans 1989;35:616-8. Icenogle TB, Smith RG, Cleavinger M, et al. Thromboembolic complications of the Symbion AVAD system. Artif Organs 1989;13:532-8. Farrar DJ, Compton PG, Hershon JJ, Foner JD, Hill JD. Right ventricular pressure-dimension relationship during left ventricular assistance in dogs. ASAIO Trans 1984;30:121-33.
25. Elbeery JR, Owen CH, Savitt MA, et al. Effects of the left ventricular assist device on right ventricular function. J Thorac Cardiovasc Surg 1990;99:80!9-16. 26. Kormos RL, Gasior T, Antaki J, et al. Evaluation of right ventricular function during clinical left ventricular assistance. ASAIO Trans 1989;35:547-50. 27. Starnes VA, Oyer PE, Portner PM, et al. Isolated left ventricular assist as bridge to cardiac transplantation. J Thorac Cardiovasc Surg 1988;9662-71. 28. Kanter KR, Swartz MT, Pennington DG, et al. Renal failure in patients with ventricular assist devices. ASAIO Trans 1987; 33:426-8.
Notice From the Southern Thoracic Surgical Association The Thirty-ninth Annual Meeting of the Southern Thoracic Surgical Association will be held at the Saddlebrook Golf and Tennis Resort, Wesley Chapel (near Tampa), Florida, November 5-7, 1992. The Postgraduate Course will be held the morning of Thursday, November 5, 1992, and will provide in-depth coverage of thoracic surgical topics selected primarily as a means to enhance and broaden the knowledge of practicing thoracic and cardiac surgeons. Applications for membership should be completed by
August 1, 1992, and forwarded to John P. Clarke, MD, Membership Committee Chairman, Southern Thoracic Surgical Association, 401 North Michigan Avenue, Chicago, IL 60611-4267. Hendrick B. Burner, M D Secretary-Treasurer Southern Thoracic Surgical Association 401 North Michigan Avenue Chicago, IL 60611-4267
plantation. Total bilirubin values were reduced by 60% during LVAD support. No significant differences were observed when total bilirubin values were compared at 30 and 60 days after LVAD support and at 30 and 60 days after transplantation in a cohort of 15 patients (p > 0.05). The improvement in renal function was less predictable than that of hepatic function. Creatinine values decreased significantly before transplantation; however, the values measured at 30 and 60 days after transplantation were higher than those measured at the same intervals after LVAD support had been initiated, and this increase is presumably related to the immunosuppressive drugs. In conclusion, the HeartMate 1000 IP LVAD has been shown to be effective in supporting end-stage cardiomyopathy patients to transplantation. Thromboembolism, previously regarded as a serious complication with such devices, has not been a problem with this device. Additional patients are being enrolled into the study to further document the safety and effectiveness of this technology.
A
are undergoing clinical evaluation under Food and Drug Administration-approved investigational device exemptions [12, 131. Both devices have been successfully employed as a bridge to transplantation for durations exceeding 10 months, with the majority of patients receiving transplants and achieving long-term survival. The purpose of this report is to summarize the results obtained with one of these devices, the Thermo Cardiosystems Inc HeartMate 1000 IP pneumatically driven LVAD, which has been under evaluation since August 1985. To extend the use of LVADs to the clinical community, sufficient data must be gathered and presented to the Food and Drug Administration to prove, scientifically, that the device is safe and effective. To be considered safe, the device must be reliable and must not jeopardize the ability to perform transplantation in the patient; that is, the device should not be associated with unacceptable levels of adverse effects, such as infection, bleeding, hemolysis, end-organ dysfunction, or thromboembolic complications. To be considered effective, the device must improve the hemodynamic status of the patient, enhance the likelihood of survival, and, ideally, promote a higher
lthough cardiac transplantation has become a widely accepted therapy for patients with end-stage cardiac disease, the demand for donor organs far exceeds the supply [l].Consequently, a substantial percentage of transplant candidates die while waiting for a donor heart [Z, 31. Temporary mechanical circulatory support for patients waiting for donor organs was introduced by Cooley and co-workers in 1969 [4]. Subsequent research led to the development of ventricular assist devices for use as a bridge to transplantation [5-81. The results with these devices have been highly encouraging for the support of postcardiotomy [9, lo], postinfarction [ll],and end-stage cardiomyopathy patients [12, 131. Among the currently available left ventricular assist devices (LVADs), two are specifically designed for longterm mechanical circulatory support. Thermo Cardiosystems Inc (Woburn, MA) and Novacor (Oakland, CA) have developed implantable, pusher-plate-type LVADs that Accepted for publication Feb 14, 1992 Address reprint requests to Dr Frazier, Texas Heart Institute, PO Box 20345, Houston, TX 77225-0345.
0 1992 by The Society of
Thoracic Surgeons
(Ann Thorac Surg 1992;53:1080-90)
0003-4975/92/$5.OO
Ann Thorac Surg 1992;53:108C-90
FRAZIER ET AL CLINICAL EVALUATION OF THE HEARTMATE
1081
quality of life than would be otherwise possible without the device. The following discussion summarizes data obtained to demonstrate the safety and efficacy of this technology to date.
Material and Methods Description of the Device The HeartMate 1000 IP LVAD consists of a pusher-plate blood pump driven by a portable external console via a percutaneous driveline. The pump consists of a titanium housing (titanium, 90%; aluminum, 6%; vanadium, 4%) measuring 11.2 cm in diameter and 4.0 cm in thickness. Inside, a flexible Biomer polyurethane diaphragm is bonded to a rigid pusher plate. The diaphragm divides the pump into two halves: a blood chamber and an air chamber. Programmed pulses of air are delivered from the console to the air chamber behind the pusher-plate diaphragm. As the air accumulates, the diaphragm is displaced, propelling the blood through the outflow graft 'Orcine xenograft into the (25 mm) are placed in the inlet and outlet conduits to ensure unidirectional blood flow. Device implantation is accomplished through a median sternotomy, with the incision extending just above the umbilicus, and cardiopulmonary bypass is instituted through standard techniques. The left ventricular apex is cored, and the opening is reinforced with a sewing ring. The pump is then placed in the left abdominal cavity, below the diaphragm. An incision is made in the diaphragm, through which the inlet conduit is passed, then secured to the apical opening. The outflow graft (preclotted Dacron) is passed over the diaphragm and anastomosed end-to-side to the ascending aorta. Finally, the pneumatic drive line is tunneled through a stab incision in the left lateral abdominal wall, above the iliac crest. After the pump has been implanted, the patient is gradually weaned from bypass, and the median sternotomy is closed. A more detailed description of the device and the implant protocol has been published [12]. The portable console, which operates on batteries or standard alternating current power, drives the blood pump directly without the need for tanks of compressed air. Both asynchronous and synchronous control modes of operation are available. The asynchronous modes include fixed-rate and pump-on-full options. If set to operate in the pump-on-full mode, the device will automatically eject when the pump is approximately 90% full. Synchronization of the pump is possible using an external QRS detector. The combination of an implanted blood pump with a portable external console allows the patient to ambulate and exercise (Fig 1). Patients are encouraged to begin an exercise program consisting of walking and stationary bicycling as soon as possible after implantation to improve their general physiologic status before transplantation.
Fig 1. Patients are encouraged to begin ambulating and exercising as soon as possible after LVAD implantation to improve their physiological status before transplantation,
Blood-Contacting Surfaces and Recommended Anticoagula tion The HeartMate LVAD employs textured biomaterials to interface with the blood. Sintered-titanium microspheres are used on the pump housing and conduits, and integrally textured polyurethane is used on the flexing pusher-plate diaphragm. The rationale for using textured surfaces is to encourage the formation of a thin, well-adhered, pseudointimal lining on the inside of the pump. The resultant biologic lining serves as the primary interface with the blood throughout implantation. The need for anticoagulation with this device has been greatly reduced by use of porcine bioprosthetic valves in combination with the textured surfaces. In the majority of patients, an antiplatelet regimen consisting of 80 mg aspirin, once a day, and 75 mg dipyridamole, three times daily, has been used beyond the intraoperative period. Heparin, sodium warfarin (Coumadin), or both have been used only during implantation and in patients with mechanical valves in the native heart.
Patient Selection Use of the device was limited to approved transplant candidates who met the hemodynamic indications for use and who were not subject to exclusionary criteria. The hemodynamic indications included a pulmonary capillary wedge pressure of 20 mm Hg or greater coupled with either a cardiac index of 2.0 L min-' m-' or less or a systolic blood pressure of 80 mm Hg or less. The exclusion criteria included chronic, irreversible hepatic, renal, and respiratory failure; severe blood dyscrasia; and right heart failure. All data were obtained in compliance with proto-
-
1082
FRAZIER ET AL CLINICAL EVALUATION OF THE HEARTMATE
Ann Thorac Surg 1992;53:108C-90
Table 1. Patient Demographics and Results of Treatment with the HeartMate LVAD Patient No.
Age Sex
(Y)
Admission Diagnosis
Implant Duration (days)
Patients W h o Met the Study Selection Criteria 1 M 47 Isch CMP 2 M 48 Idio CMP 3 M 17 Viral CMP 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
M M M M M M M M M M M M M M M M M M M M M M F
22 39 43 42 55 41 49 55 44 55 37 48 47 47 22 46 57 50 39 60 45 46 35
Idio CMP Isch CMP Idio CMP Idio CMP Idio CMP Idio CMP Isch CMP Isch CMP Idio CMP Idio CMP Isch CMP Isch CMP Dil CMP Idio CMP Viral CMP Isch CMP MI Idio CMP Dil CMP Idio CMP Isch CMP Idio CMP Viral CMP
41 37 7
Yes Yes Yes
35 132 114 233 189 31 220 14 1 61 153 10 1 38 84 15 21 66 6 4 Ongoing Ongoing Ongoing
Yes Yes Yes Yes Yes Yes Yes No No Yes Yes No No Yes Yes Yes Yes Yes No No
...
... ...
5 19 25 1 84 1 1 1
No Yes Yes No Yes No No No
No No (multiorgan failure, sepsis) No (donor heart failure) No Yes No No No
Patients Who Did Not Meet the Study Selection Criteria 1 M 55 Idio CMP 2 M 53 Idio CMP 3 M 37 Idio CMP 4 M 59 Isch CMP 5 M 52 Isch CMP 6 M 51 Idio CMP 7 M 55 MI 8 M 38 MI a Cause of death appears in parentheses. CMP = cardiomyopathy; Dil = dilated;
Idio
=
idiopathic;
SUrviVOP (>60 days)
Transplanted
Isch = ischemic;
cols approved by the investigational review boards at each of the respective institutions. Between August 1985 and February 1991,34 patients at seven medical centers were treated with the HeartMate 1000 IP LVAD as a bridge to transplantation. All but 1 of the patients were male, and their ages ranged from 17 to 62 years. Of the 34 patients, 26 met the study selection criteria, and 8 did not for various reasons (Table 1).Whereas all 34
...
...
No (liver failure) Yes No (respiratory failure; adverse OKT3 reaction) Yes Yes Yes Yes Yes Yes Yes No No Yes Yes No No Yes Yes Yes Yes Yes No No
...
MI = myocardial infarction.
patients were considered in the overall evaluation of survival and device safety, only the 26 who met study selection criteria were considered in the analyses of hemodynamic, hematologic, hepatic, and renal response to the pump. Of these 26 patients, 15 were long-term survivors (>60 days) of cardiac transplantation and were discharged from the hospital. Data collection in these patients allowed further comparison of hemodynamic, hematologic, hepatic, and renal values at 30 and 60 days
FRAZIER ET AL CLINICAL EVALUATION OF THE HEARTMATE
Ann Thorac Surg 1992;53:1080-90
1083
Table 2. Patient Demographics and Results of Treatment in Six Control Subjects ~
Patient
~~
Age
Admission Diagnosis
No.
Sex
(Y)
1 2
M M F F M M
21 38 31
Idio CMP Isch CMP
44
Isch CMP Isch CMP
3
4 5 6 a
52 57
Survivor"
Transplanted
(>60 days)
Yes
No (multiorgan failure) No No (allograft rejection) No
No Yes No Yes No
PPart CMP
Isch CMP
Yes No
Cause of death appears in parentheses.
CMP
=
cardiomyopathy;
Idio
=
idiopathic;
Isch = ischemic;
Wart = postpartum.
of device support and at the same times after transplantation.
Control Population Six historic controls were entered into the study (Table 2).
They were identified by searching the transplant database for candidates who would have met the patient selection criteria for LVAD support but who did not receive treatment with the device because it was not available.
Data Acquisition Hemodynamic data, including the cardiac index, pulmonary capillary wedge pressure, and blood pressure, were collected. Hematocrit levels, plasma free hemoglobin levels, and platelet counts were monitored to assess the hematologic response to circulatory assistance. Hepatic and renal function were assessed through total bilirubin levels, serum glutamic-oxaloacetic transaminase and serum glutamic-pyruvic transaminase values, and creatinine and blood urea nitrogen levels. These data were then compared before and during ventricular assistance, as well as after transplantation. The change in New York Heart Association functional status was also assessed before support and at 60 days after transplantation in the 15 patients who survived more than 60 days. Safety data pertaining to the incidence of bleeding, hemolysis, infection, right heart failure, peripheral end-organ failure, and mechanical failure were acquired for all of the patients.
Statistical Analyses The hemodynamic, hepatic, and renal function entrance criteria were analyzed using nonpaired t tests. Nonpaired t tests were also used to compare hepatic and renal function in survivors and nonsurvivors before and after LVAD treatment. Paired t tests were used to compare LVAD and transplantation data in a sample of 15 patients who were treated with the LVAD and then underwent transplantation. Finally, the survival and complication data were analyzed with the Fisher's exact probability test. A p value of less than 0.05 was considered significant in all analyses.
LVAD patients. Three of the 34 patients remained on LVAD support at the time of study. Twenty (65%)of the remaining 31 patients received transplants, 16 (80%) of whom were discharged from the hospital. Survival in the 16 patients ranges up to 3 years after the transplant procedure. Of the 6 control patients, 3 (50%)died before transplantation at 5, 12, and 34 days after having met the hemodynamic indications for LVAD support. The remaining 3 patients underwent transplantation but died at 2, 21, and 77 days after operation. Therefore, all 6 patients died within 77 days of having met the indications for LVAD support, regardless of whether they underwent transplantation. A Fisher's exact probability test revealed that the survival rate of the LVAD-treated group was significantly greater than that of the control group ( p < 0.05). HEMODYNAMIC COMPARISON OF THE STUDY GROUPS AT
The hemodynamic status of the 26 LVAD patients was documented at the time they met the indications for use and within 24 hours before implantation. These data were then compared with the hemodynamic values of the control patients when they met the study entrance criteria. No differences in cardiac index or systolic blood pressures were noted between the LVAD and control patients either at the time of meeting the study entrance criteria or at the time of LVAD implantation (p > 0.05). However, the control group had a significantly lower pulmonary capillary wedge pressure than the LVAD group at both points. When the data from the control patients were compared with those of the LVAD patients at the time of implantation, the average hemodynamic values (f the standard deviation) were, respectively, as follows: pulmonary capillary wedge pressure, 23 f 2 versus 28 8 mm Hg; cardiac index, 1.9 f 0.8 versus 2.1 f 0.6 L * min-l m-'; and systolic blood pressure, 90 f 19 versus 94 2 19 mm Hg. These data imply that the two groups were hemodynamically comparable, although the control group had a slightly lower wedge pressure. ENTRANCE.
*
-
Figure 2 illustrates the average pump index (pump flowbody surface area) and aortic pressures for the 26 LVAD patients who met the entrance criteria. The average pump index for all patients was 2.86 L min-' m-', which was approximately 30% greater than the average cardiac index at the HEMODYNAMIC PERFORMANCE OF THE LVAD.
Results Evaluation of Device Effectiveness SURVIVAL DATA. Table 1 lists the age, sex, admission diagnosis, implant duration, and outcome for the 34
-
-
1084
FRAZIER ET AL CLINICAL EVALUATION OF THE HEARTMATE
Ann Thorac Surg 1992;53108CL90
Avg. Pump Index (Llmlnlm2)
l
+
80
.i
*
*
8o
2
ol
60
100
160
200
Postoperative Days 3oo iWg. Aortic Prerrun (mmHg)
2501
' 0
-
+* ++
I
,
+
+++++ + +++ ++ , +++++++ + ++,++++++ ,
++
250
200
0
t
20
0
PHgb
8ool 800
50
100
150
200
"
250
0
Postoperative Days
50
100
150
200
250
Postoperative Days
Fig 2 . Hemodynamic measurements during LVAD support in 26 patients. The average pump index (pump flowlbody surface area) was 2.86 L * min-' * m-'. The average pressures were as follows: systolic, 119 mm Hg; diastolic, 71 mm Hg; and mean pressure, 95 mm Hg.
Fig 3 . Effects of LVAD support on hematologic indices in 26 patients. Average hematocrit (HCT) was 34%; average plasma free hemoglobin (PHgb) level, 8.7 mgldL; and average platelet count, 249,000lmL.
time of implantation ( p < 0.05). The systolic, diastolic, and mean aortic pressures were 119 mm Hg, 71 mm Hg, and 95 mm Hg, respectively. The device was found to be fully capable of generating adequate pressures and flows.
Further evidence that hepatic function improved during LVAD support is seen in Figure 6, where the values of total bilirubin before implantation were compared with the final values obtained either just before transplantation or at the time of death. In this analysis, data from 3 of the
EFFECTS OF LVAD SUPPORT ON HEMATOLOGIC INDICES. The average plasma free hemoglobin level was 8.7 mg/dL in the 26 LVAD patients, suggesting that a mild, but clinically acceptable level of hemolysis occurred in these patients (Fig 3). This finding was further confirmed by hematocrit values averaging 34% and platelet counts averaging 249,00O/mL. Adequate circulatory support was provided without serious damage to the blood.
FUNCAll but 2 of the 26 LVAD patients had elevated total bilirubin values (21.4 mg/dL) or serum glutamicoxaloacetic transaminase and/or serum glutamic-pyruvic transaminase values (250 U/L) before or during device support. Figures 4 and 5 illustrate the average hepatic values as a function of time. All three parameters tended to increase during the first month, then returned to normal after approximately 2 months of augmented perfusion. The values remained within normal limits for the remaining period of support. EFFECTS OF LVAD SUPPORT ON HEPATIC AND RENAL
TION.
Avg. Total Bilirubin (mgldl)
lo
I
8-
*
z x
4 - *
* * 20
** *** 60
100
150
200
250
Postoperative Days Fig 4 . Effect of LVAD support on hepatic function in 26 patients. The average total bilirubin value initially increased, then returned to normal by day 50.
FRAZIER ET AL CLINICAL EVALUATION OF THE HEARTMATE
Ann Thorac Surg 1992;53:1080-90
Avg. SQOT (UIL) 600,
1
500t* 400
*
y
-
300 200
w" - *
100
-
#
0
* ****x
'
...........................
600
500
-u
400
-
300
-
*
200
- **
100
-
c*
+*
X*
*
Y X"********f****XY
0
**** *****
'
1085
either failed to decrease or increased in all nonsurvivors. Hepatic function was improved in all patients with successful transplantations, and 3 patients remained on device support at the time of study. The average bilirubin value obtained just before pump removal divided by the baseline value yields a ratio that indicates whether the total bilirubin increased (>1.0), decreased (<1.0), or stayed the same (1.0) in response to the treatment. The bilirubin ratio for the LVAD survivors (0.40) was significantly reduced compared with that of the LVAD nonsurvivors (3.6; p = 0.03), as well as the control patients (1.3; p = 0.004). In contrast, no significant difference existed between the controls (1.3) and the LVAD nonsurvivors (3.6; p = 0.09). These data demonstrate that LVAD support leads to improvement in hepatic function in the majority of patients, providing that dysfunction is reversible. Hepatic function in the nonsurvivors failed to improve. These patients had significantly higher total bilirubin values at implantation, and the duration of support may have been too short for them to benefit from improved perfusion. The hepatic function of the controls also failed to improve in the absence of mechanical circulatory support. Figure 7 illustrates the change in serum creatinine and blood urea nitrogen values in response to ventricular assistance. In contrast to the hepatic response, the renal parameters did not transiently increase after LVAD implantation. Moreover, the renal parameters stabilized within a shorter period of time. COMPARISON OF HEPATIC A N D RENAL FUNCTION DURING
In a cohort of 15 LVAD patients surviving more than 60 days after transplantation, three hepatic and two renal parameters were measured at LVAD implantation, at days 30 and 60 during support, immediately before transplantation, and at days 30 and 60 after transplantation. A t test was then performed to compare the effects of ventricular assistance and transplantation on hepatic and renal function. The only significant difference in hepatic function during LVAD support versus that after transplantation was seen in the total bilirubin level. The LVAD patients had a significantly greater total bilirubin value at implantation compared with that on the day of transplantation. Figure 8 illustrates the change in total bilirubin values after implantation of the LVAD versus after transplantation. The baseline values (t = 0) obtained just before implantation of the LVAD were significantly greater than those obtained just before transplantation ( p < 0.05). Hepatic function was markedly improved after ventricular assistance, as was the overall physiologic status of the patients before transplantation. No significant differences were noted in the hepatic parameters at days 30 and 60 of LVAD support or at the same times after transplantation. Figure 9 illustrates the change in renal values as a function of LVAD treatment versus transplantation. The creatinine values were significantly greater before LVAD implantation than before transplantation, demonstrating that the LVAD improved the condition of the patient LVAD SUPPORT AND AFTER TRANSPLANTATION.
patients were excluded because treatment was ongoing. The remaining 23 patients were classified as survivors (n = 15) and nonsurvivors (n = 8). Whereas the total bilirubin values decreased in all survivors, the values Total Bilirubin (mgldi) NONSURVIVORS
25 20
10 5 n "
17
18
10
21
02
05 04
08
07
08
09
10
13
20
22
U
01
03
ti
11
15
23 w)
Patient Number
aPRE-IMPLANT
FINAL
Fig 6 . Total bilirubin values were significantly reduced after LVAD support in the survivors. Hepatic function continued to deteriorate in the nonsurvivors despite restoration of blood flow.
1086
FRAZIER ET AL CLINICAL EVALUATION OF THE HEARTMATE
Ann Thorac Surg 1992;53108&90
64-
2t
4
01
3-
*************** * ******* ~********* '
SO
0
100
150
200
I
250
Postoperative Days
0'
0
m
90
20
40
50
60
Postoperative Days
Avg. BUN (mgldl)
2 8
Avg. Total Bilirubin (mgldl)
P08TTRANIPLANT
I 0
50
100
150
200
250
Postoperative Days
0 0
10
20
30
40
50
60
Postoperative Days
Fig 7. Effect of LVAD support on renal function in 26 patients. The average creatinine and blood urea nitrogen (BUN) values decreased after implantation and remained within normal limits for the remainder of implant duration.
Fig 8. Comparison of hepatic function after implantation and after transplantation. The average total bilirubin value was significantly greater at LVAD implantation (top) compared with that just before transplantation (bottom) in a cohort of 15 patients. There was no significant difference on days 30 and 60 after each procedure.
before transplantation ( p < 0.05). Unlike the hepatic indicators, the creatinine values were significantly elevated at days 30 and 60 after transplantation compared with those at the same intervals after LVAD implantation. The elevated creatinine levels were thought to have been a result of immunosuppressive therapy. In general, peripheral end-organ function was adequately supported after both LVAD treatment and transplantation. However, the physiologic status of the transplant candidates was markedly improved after LVAD support, enhancing the likelihood of success of transplantation.
multitude of environmental factors that can have a positive or negative impact on the performance of the device. Adverse effects were monitored for all patients. Of the 34 patients bridged to transplantation, 3 were treated for less than a day and were considered compassionate exemptions, and these patients experienced no adverse effects. Three additional patients remained on LVAD support at the time of study. The safety data reported in this article were collected from the remaining 28 LVAD patients who completed the study. Data are also presented for the six control patients entered into the study. Each category of adverse effects is described in the following sections.
Safety Evaluation A number of risks are associated with the use of ventricular assistance technology. The principal risks currently being monitored include bleeding, hemolysis, infection, right heart failure, peripheral end-organ dysfunction, thromboembolism, and mechanical failure. The frequency and nature of adverse effects resulting from the use of these devices depend on numerous factors, including the basic design of the device, the skill and experience of the clinical team, the physiologic status of the patient, and a
BLEEDING. Bleeding occurred in all of the patients. However, the bleeding was considered serious only if it was severe enough to require returning the patient to the operating room. None of the patients bridged to transplantation experienced device-related bleeding. Eleven patients (39%) experienced patient-related bleeding, such as that resulting from cardiac tamponade. Seven (64%)of these patients underwent transplantation, 5 (71%) of whom were long-term survivors. Three (50%)
FRAZIER ET AL CLINICAL EVALUATION OF THE HEARTMATE
Ann Thorac Surg 1992;531080-90
Avp. Cnatinina (mgldi)
6
fever, combined with the need for antimicrobial treatment. An infection was considered device-related if the specific organism that was cultured had not been detected preoperatively. Typical device-related infections involve the drive line or positive cultures isolated from the blood pump, coupled with the symptoms already described. Based on the specified criteria, 7 (25%) of the 28 patients bridged to transplantation experienced a device-related infection. Six (85%)of the patients underwent transplantation, 4 (67%) of whom were long-term survivors. No significant difference existed between the outcome of infected and noninfected patients according to a Fisher’s exact probability test ( p > 0.05), suggesting that infection may not be a critical determinant of outcome. Two of the 6 control patients (33%)experienced infections while awaiting transplantation. Although these patients did not receive an LVAD, the potential for infection was increased because of diagnostic instruments. The number of control patients was low, but the incidence of infection was high. It remains to be determined whether the presence of an LVAD further increases the risk of infection.
ON LUD
-
0
10
20
SO
40
50
60
Postoperative Days 6
Avg. Creatinino (mgldl) POSTTRANSPLANT
st
I 0
10
20
SO
40
50
1087
END-ORGAN DYSFUNCTION. As stated, all but two of the patients bridged to transplantation experienced renal or hepatic dysfunction or both before or during LVAD support. The dysfunction was not considered device-related in any of the patients.
60
Postoperative Days
Fig 9. Average creatinine values were significantly greater at LVAD implantation compared with before transplantation. Creatinine levels were significantly greater on days 30 and 60 after transplantation than during ventricular support, presumably caused by immunosuppression.
of the 6 patients who died had been treated simultaneously with a Biomedicus (Biomedicus Inc, Eden Prairie, MN) right ventricular assist device. A statistical comparison of the patients who did versus those who did not bleed revealed no significant difference in outcome between the two groups with respect to transplantation and survival rates (Fisher’s exact probability test; p < 0.05). HEMOLYSIS. The average plasma free hemoglobin values for all the patients was 8.7 mg/dL. The hemoglobin concentration was 11.0 g/dL. Only 1patient, a 50-year-old man who had a plasma free hemoglobin value of 40 mg/dL just before LVAD implantation, experienced hemolysis. His plasma free hemoglobin level exceeded 40 mg/dL on two consecutive occasions after implantation. He remained on the device for 66 days, underwent successful transplantation, and is currently alive and well several months after cardiac transplantation. Why hemolysis occurred before and after LVAD implantation is uncertain. INFECTION. Infection was defined through detection of a positive culture, elevated white blood cell count, and
RIGHT HEART FAILURE. Six (21%)of the 28 patients bridged to transplantation either required right ventricular assistance or exhibited symptoms of serious right ventricular dysfunction after implantation of the LVAD. Two of the patients experienced unanticipated right ventricular failure at the time of LVAD implantation but did not receive a right ventricular assist device. Although the LVAD is capable of generating flow rates of 6.0 L/min with 0.0 mm Hg inlet pressure, a low-flow condition persisted in both of the patients because of poor filling, and they died in the operating room. The inability to generate adequate flow under these circumstances was thought to be caused by an elevated pulmonary vascular resistance. The remaining 4 patients required right ventricular assistance, which was accomplished with the extracorporeal Biomedicus centrifugal pump. Two of the patients were electively weaned from the Biomedicus pump after 5 and 6 days, and 1of these patients received a transplant. The other 2 patients requiring right ventricular assistance experienced progressively increasing pulmonary vascular resistance secondary to bleeding. All patients requiring right heart assistance ultimately died. Among the adverse effects, right heart failure was the only complication that had a significantly negative correlation with outcome. When the patients with versus without right heart failure were compared, the transplantation rate and the survival rate were significantly lower for the patients who experienced right heart failure ( p < 0.05).
No device-related thromboemboli were observed in any of the patients. One patient, how-
THROMBOEMBOLISM.
1088
FRAZIER ET AL CLINICAL EVALUATION OF THE HEARTMATE
ever, did experience an embolic event related to a mechanical aortic valve located in the natural heart. The patient recovered, underwent successful transplantation after 173 days of LVAD support, and was discharged home. Bowel adhesions to the Dacroncovered drive line occurred in 2 patients in whom the lead was allowed to course intraabdominally. Both patients underwent transplantation, but these two incidences emphasize the need to minimize contact between the bowel and the textured materials, such as Dacron.
BOWEL ADHESIONS.
MECHANICAL FAILURE. In 1988, a threaded outflow connector loosened after LVAD implantation in 1 patient, who required reoperation to tighten the connector. The design of the device was modified to incorporate a locking mechanism, and no additional adverse effects have been observed since that incident. This was the only failure in 7 patient-years of experience.
OVERALL. In summary, the frequency of device-related adverse effects has been low. Although a variety of patient-related adverse effects were observed, none aside from right heart failure had a negative correlation with outcome.
Quality of Life The New York Heart Association classification system was used as a measure of quality of life in response to device support. All the LVAD-treated patients were in functional class IV at the time of implantation. Sixteen of the LVAD patients were alive at 60 days after transplantation, at which time 15 were in class I and 1 was in class 11. At the time of meeting study entrance criteria, 5 of the 6 control patients were in class IV, and the sixth patient was in class 111. Only 1 of the control patients survived to 60 days after transplantation and was determined to be in class IV at that time. These data demonstrate that the quality of life of the LVAD-treated patients, based on functional class, was significantly improved during ventricular assistance and after transplantation. Also, the functional class of the LVAD-treated patients was markedly improved when compared with that of the control patients.
Comment In contrast to extracorporeal LVADs that have been largely employed to support patients in acute cardiogenic shock or postinfarction, the HeartMate device has been used primarily for the extended support of patients with chronic cardiomyopathy who require a bridge to transplantation. Efficacy has been demonstrated by our study, in which the device provided adequate hemodynamic support until transplantation, increased survival, and improved the New York Heart Association functional class status of the patients. In comparison with the unsupported controls, the LVAD-treated patients had a
Ann Thorac Surg 1992;53:108C-90
better outcome. The majority of the LVAD-treated patients received transplants (65%), and 80% were discharged from the hospital after transplantation. The ongoing clinical study will determine if the HeartMate LVAD performs in a safe and reliable fashion. Adverse effects, including bleeding, hemolysis, infection, peripheral end-organ dysfunction, and thromboembolic events, have not had a negative correlation with the rate of transplantation or survival. In contrast, acute perioperative right heart failure was associated with a significantly greater mortality. In comparing complication rates with the HeartMate device with those reported in the Combined Registry [14], two important observations were noted. First, the average duration of support for the HeartMate patients was almost twice as long as that for the patients included in the Combined Registry report: 53 days versus 27.7 days. Second, the HeartMate patient population was composed primarily of end-stage cardiomyopathy patients, whereas those included in the Registry report were a combination of postcardiotomy, postinfarction, and cardiomyopathy patients. In general, the patients included in the Combined Registry report were experiencing acute left ventricular failure, whereas the HeartMate population was composed primarily of patients experiencing chronic left ventricular failure. The nature and frequency of complications with the HeartMate were comparable with those reported for other devices, although the incidence of thromboembolic episodes was lower with the HeartMate. Previous reports estimate patient-related bleeding to occur in approximately 40% of LVAD-supported patients [13, 151, compared with a 39% rate of patient-related bleeding after implantation of the HeartMate. Infectious complications were more difficult to compare because no uniformity for reporting these data exists. Reported rates of infection range from approximately 20% to 75%, although not all of the infections were device-related [13, 15, 161. In patients treated with the HeartMate, the incidence of devicerelated infection was 25%, yet 85% of the patients underwent transplantation, and their survival rate was 67%.It is noteworthy that a recent review of the literature from 12 transplant centers, encompassing 384 transplant patients treated with cyclosporine, showed that 57.6% of the patients experienced infections, with 5.2%dying of infection-related causes [17]. Two (33%)of the 6 control patients in the current study also had infections. These data appear to indicate that although the rate of infectious complications in both the LVAD-treated patients and the transplant patients is high, the number of infectionrelated deaths associated with both forms of treatment was relatively low. Whereas mediastinal infection was the primary cause of death in 40% of patients receiving a total artificial heart as a bridge to transplantation [18, 191, the risk of infection-related death with ventricular assistance appears significantly lower. Thromboembolic complications have been a topic of great concern in the past decade, particularly in regard to the Jarvik-7total artificial heart [20,21]. Numerous reports of thromboembolic complications have also appeared in the ventricular assist device literature, with the rate rang-
Ann Thorac Surg 1992;53:1080-90
ing from 33% to 47% [15, 16, 22, 231. No device-related thromboembolic events were observed in the HeartMate patients despite the use of minimal anticoagulative therapy. The only incident that occurred was related to a mechanical aortic valve in the native heart. The presence of a mechanical aortic valve may be considered a contraindication for use of an LVAD. The absence of thromboembolic complications with the HeartMate represents a major step forward for circulatory support systems toward minimizing the risk of device-related strokes. Although it has been stated that left ventricular assistance may be potentially detrimental to right heart function [24, 251, additional reports have shown that impaired right ventricular function improves during left ventricular assistance [26]. Right ventricular function was indeed found to improve with left ventricular support in the majority of patients in this study, as evidenced by increased right ventricular ejection fractions. However, approximately 20% of the patients experienced severe right ventricular dysfunction at implantation, which increased mortality and morbidity in comparison with patients who had no right ventricular involvement. The inability to fill the pump in the presence of right ventricular failure has been attributed to concomitant elevated pulmonary vascular resistance. Use of prostaglandins [27] combined with other pharmacologic measures may aid in reducing pulmonary vascular resistance, but additional study is required. Currently, right ventricular dysfunction appears to be a significant contraindication for left ventricular support. The onset of renal failure shortly before or during ventricular assistance has been suggested to be highly predictive of a poor prognosis [28]. However, both the Thermo Cardiosystems Inc and Novacor devices have successfully supported patients who were anuric before implantation and on hemodialysis for prolonged periods (months) followed by successful transplantation and survival [13]. Renal failure has been regarded as a contraindication for patients in acute left ventricular failure, but it is not clear that it should be considered a contraindication for chronic cardiomyopathy patients. Hepatic function was shown to improve significantly with hemodynamic support, but up to 2 months of LVAD support was required for hepatic function to return to normal. The patients had better hepatic and renal function just before transplantation than before LVAD implantation. This improvement was a result of using the device. Beyond implantation, no difference in hepatic function was observed when the two treatments were compared. In contrast, creatinine values were shown to be significantly higher in the transplant patients after 30 to 60 days of the operation than in the LVAD patients at comparable times after initiation of ventricular assistance. The rise in creatinine level was attributed to the use of immunosuppressive drugs after transplantation. In conclusion, the HeartMate 1000 IP LVAD has been shown to provide an effective means of supporting patients for periods in excess of 300 days. The study will continue to be expanded to include additional patients and controls to further assess safety and effectiveness of this technology as a bridge to transplantation.
FRAZIER ET AL CLINICAL EVALUATION OF THE HEARTMATE
1089
References 1. Evans RW, Manninen DL, Garrison LP Jr., Maier AM. Donor availability as the primary determinant of the future of heart transplantation. JAMA 1986;255:1892-8. 2. Copeland JG, Emery RW, Levinson MM, Copeland J, McAleer MJ, Riley JE. The role of mechanical support and transplantation in treatment of patients with end stage cardiomyopathy. Circulation 1985;72(Suppl2):7-12. 3. Schuler S, Parnt R, Warnecke H, Matheis G, Hetzer R. Extended donor criteria for heart transplantation. J Heart Transplant 1988;7326-30. 4. Cooley DA, Liotta D, Hallman GL, Bloodwell RD, Leachman RD, Milam JD. Orthotopic cardiac prosthesis for two-staged cardiac replacement. Am J Cardiol 1969;24:723-30. 5. Bernhard WF, Clay W, Shoen FJ, et al. Clinical and laboratory investigations related to temporary and permanent ventricular bypass. Heart Transplant 1983;3:16-25. 6. Hill JD, Farrar DJ, Hershon JJ, et al. Use of a prosthetic ventricle as a bridge to cardiac transplantation for postinfarction cardiogenic shock. N Engl J Med 1986;314626-8. 7. Kanter KR, McBride LR, Pennington DG, et al. Bridging to cardiac transplantation with pulsatile ventricular assist devices. Ann Thorac Surg 1988;46:134-40. 8. Champsaur G, Ninet J, Vigneron M, Cochet P, Neidecker J, Boissonnat P. Use of the Abiomed BVS System 5000 as a bridge to cardiac transplantation. J Thorac Cardiovasc Surg 1990;100:122-8. 9. Pierce WS, Parr GVS, Myers JL, et al. Ventricular-assist pumping in patients with cardiogenic shock after cardiac operations. N Engl J Med 1981;305:1606-10. 10. Pennington DG, Samuels LD, Williams G, et al. Experience with the Pierce-Donachy ventricular assist device in postcardiotomy patients with cardiogenic shock. World J Surg 1985;9:3746. 11. Pae WE Jr, Pierce WS. Temporary left ventricular assistance in acute myocardial infarction and cardiogenic shock: rationale and criteria for utilization. Chest 1981;79:692-5. 12. McGee MG, Parnis SM, Nakatani T, et al. Extended clinical support with an implantable left ventricular assist device. ASAIO Trans 1989;35:614-6. 13. Portner PM, Oyer PE, Pennington DG, et al. Implantable electrical left ventricular assist system: bridge to transplantation and the future. Ann Thorac Surg 1989;47142-50. 14. Miller CA, Pae WE Jr, Pierce WS, et al. Combined Registry for the Clinical Use of Mechanical Ventricular Assist Pumps and the Total Artificial Heart in Conjunction with Heart Transplantation: fourth official report-1989. J Heart Transplant 1990;9:453-8. 15. Pennington DG, McBride LR, Kanter KR, et al. Bridging to heart transplantation with circulatory support devices. J Heart Transplant 1989;8:116-23. 16. Didisheim P, Olsen DB, Farrar DJ, et al. Infections and thromboembolism with implantable cardiovascular devices. ASAIO Trans 1989;35:54-70. 17. Linder J. Infection as a complication of heart transplantation. J Heart Transplant 1988;7390-4. 18. Griffith BP, Kormos RL, Hardesty RL, Armitage JM, Dummer JS. The artificial heart: infection-related morbidity and its effect on transplantation. Ann Thorac Surg 1988;45:409-14. 19. Rice LB, Karchmer AW. Artificial heart implantation: what limitations are imposed by infectious complications? JAMA 1988;259:894-5. 20. Levinson MM, Smith RG, Cork RC, et al. Thromboembolic
1090
21. 22. 23. 24.
Ann Thorac Surg 1992;53:108&90
FRAZIER ET AL CLINICAL EVALUATION OF THE HEARTMATE
complications of the JaMk-7 total artificial heart case report. Artif Organs 1986;10:23W. DeVries WC, Anderson JL, Joyce LD, et al. Clinical use of the total artificial heart. N Engl J Med 1984;310:27M. Termuhlen DF, Swartz MT, Pennington DG, et al. Thromboembolic complications with the Pierce-Donachy ventricular assist device. ASAIO Trans 1989;35:616-8. Icenogle TB, Smith RG, Cleavinger M, et al. Thromboembolic complications of the Symbion AVAD system. Artif Organs 1989;13:532-8. Farrar DJ, Compton PG, Hershon JJ, Foner JD, Hill JD. Right ventricular pressure-dimension relationship during left ventricular assistance in dogs. ASAIO Trans 1984;30:121-33.
25. Elbeery JR, Owen CH, Savitt MA, et al. Effects of the left ventricular assist device on right ventricular function. J Thorac Cardiovasc Surg 1990;99:80!9-16. 26. Kormos RL, Gasior T, Antaki J, et al. Evaluation of right ventricular function during clinical left ventricular assistance. ASAIO Trans 1989;35:547-50. 27. Starnes VA, Oyer PE, Portner PM, et al. Isolated left ventricular assist as bridge to cardiac transplantation. J Thorac Cardiovasc Surg 1988;9662-71. 28. Kanter KR, Swartz MT, Pennington DG, et al. Renal failure in patients with ventricular assist devices. ASAIO Trans 1987; 33:426-8.
Notice From the Southern Thoracic Surgical Association The Thirty-ninth Annual Meeting of the Southern Thoracic Surgical Association will be held at the Saddlebrook Golf and Tennis Resort, Wesley Chapel (near Tampa), Florida, November 5-7, 1992. The Postgraduate Course will be held the morning of Thursday, November 5, 1992, and will provide in-depth coverage of thoracic surgical topics selected primarily as a means to enhance and broaden the knowledge of practicing thoracic and cardiac surgeons. Applications for membership should be completed by
August 1, 1992, and forwarded to John P. Clarke, MD, Membership Committee Chairman, Southern Thoracic Surgical Association, 401 North Michigan Avenue, Chicago, IL 60611-4267. Hendrick B. Burner, M D Secretary-Treasurer Southern Thoracic Surgical Association 401 North Michigan Avenue Chicago, IL 60611-4267