| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1999;68:756-760
© 1999 The Society of Thoracic Surgeons
a University of Arizona Health Sciences Center, Tucson, Arizona, USA
Address reprint requests to Dr Copeland, Cardiovascular and Thoracic Surgery, The University of Arizona Health Sciences Center, 1501 N. Campbell Ave, Rm 4402, Tucson, AZ 85724
e-mail: jgcbriez{at}aol.com
Presented at the Fourth International Conference on Circulatory Support Devices for Severe Cardiac Failure, Houston, TX, Oct 3–5, 1997.
Abstract
Background. We hypothesized that bridge to transplantation with the CardioWest Total Artificial Heart would succeed in a large percentage of patients. Further, we hypothesized that this success rate would not be significantly decreased by infection or thromboembolism.
Methods. From 1993 to March 1999, 24 patients received implants with the intention of bridge to transplantation. Data were collected prospectively. Heparin, coumadin, aspirin, ticlopidine, dipyridamole, and pentoxifylline were used for anticoagulation.
Results. Four patients died while on device support. Nineteen of 23 patients (83%) were transplanted. All 19 survived long term. One patient remains on CardioWest Total Artificial Heart support 6 weeks after implant. There was one stroke on the day of transplantation. There was a second stroke on the day of implantation. Neither stroke caused significant residual deficits. Both were in close relationship to an operative procedure. There were no serious device-related infections.
Conclusions. The CardioWest Total Artificial Heart salvaged 20 of 24 critically ill patients. Neither infections nor neurologic problems were significant. We believe it is the device of choice for decompensating patients with biventricular failure who have adequate body and heart size.
Since January 1993, we have been the principal investigators in a national trial of the CardioWest Total Artificial Heart (TAH; CardioWest, Tucson, AZ) conducted under an Investigational Device Exemption from the Food and Drug Administration (FDA). The CardioWest TAH, formerly known as the Symbion Jarvik-70 TAH, changed ownership in 1991. Then, after minor modifications, it was approved by the FDA for further investigation in 1993. The only changes were elimination of the skin button in favor of a simple Dacron velour cover of the drive lines and addition of one drop of silicone oil to the housing side of the diaphragms to prevent sticking. We began our investigation in January 1993.
Our previous experience [1, 2], like that of others [3, 4], with the Symbion TAH was sufficiently positive to encourage further work. To update our 1996 report of the international experience [5], there have been 114 implants, including 9 patients still on device support with 72 of 105 (69%) transplanted, and 66 of 105 discharged from the hospital (63% of those implanted and 92% of those transplanted). Our recently reported national study [6] included 27 patients with implants, 25 (93%) of whom received a transplant and 24 of whom were discharged (89% of the total, 96% of those transplanted), and a group of historical control patients with a 39% discharge rate (7 of 18 patients). The actuarial survival rate for the implant group was significantly better (p < 0.0001) than that for the control group. The accumulated evidence suggests efficacy and safety for this device. Our own experience supports this.
Patients and methods
Inclusion criteria for our patients included age between 18 and 59 years, body surface area between 1.7 and 2.5 m2, acceptable or accepted for heart transplantation by standard criteria [7], evidence of hemodynamic decompensation including central venous pressure at least 18 mm Hg, systolic blood pressure 90 mm Hg or less, cardiac index less than 2.0 L/m2 per minute, or requirement for high dose inotropic therapy with two inotropic agents (ie, dobutamine dose of more than 10 µg/kg per minute and dopamine dose more than 10 µg/kg per minute) or need for intraaortic balloon pump plus one inotropic agent, or failure to wean from cardiopulmonary bypass or extracorporeal membrane oxygenation. Exclusion criteria included evidence of an active infection, renal or hepatic failure, a panel reactive antibody percentage of more than 10%, or the presence of any recognized contraindication to cardiac transplantation.
All patients were excluded from left ventricular assist device support on the basis of one of the following criteria: mechanical valve previously implanted, right heart failure indicated by central venous pressure of 18 mm Hg or more, high pulmonary vascular resistance (> 4.5 Wood units), on extracorporeal membrane oxygenation or cardiopulmonary bypass, or miscellaneous reasons including no left ventricular assist device available (one case) and right ventricle injured in redo operation (one case). Biventricular assist device support was reserved for patients who did not meet the CardioWest TAH body surface area size criterion of at least 1.7 m2.
In addition to survival, adverse events were monitored prospectively. Definitions for these events used in our study have been published previously [6].
The following studies were used to evaluate coagulability: prothrombin time, partial thromboplastin time, platelet count, fibrinogen, platelet activation, bleeding time, and thromboelastogram [4]. Anticoagulation consisted of dipyridamole at doses from 100 to 400 mg four times per day by nasogastric tube starting immediately after implantation. Higher doses were used in situations when the patient was judged hypercoagulable by thromboelastography or had exceptionally high platelet count (ie, > 400,000/mL). Heparin in low dose to keep the partial thromboplastin time at 55 ± 5 seconds was generally started on postoperative day 2 or when chest tube drainage was serous and continued until coumadin therapy was therapeutic. Coumadin was started as soon as the patient could take oral medication and modulated to maintain INR of 2.5 to 3.5. Aspirin was started in low dosage (81 mg per day) simultaneously with coumadin and increased to 325 mg per day when the platelet count exceeded 100,000/mL. Generally, patients were given an additional 325 mg of aspirin for each additional increase of 100,000 platelets per milliliter (eg, 650 mg for a count of 200,000). Antiplatelet therapy was assessed by platelet activation studies; we sought a +/+ reaction to collagen and a +/- reaction to adenosine diphosphate, arachidonic acid, and epinephrine. We also attempted to keep the bleeding time, measured twice per week, above 10 minutes (the upper limit of normal) and below 20 minutes. Finally, we used the thromboelastographic determinations of maximal amplitude and thrombodynamic potential index attempting to keep the maximal amplitude between 50 and 60 mm and the thrombodynamic potential index in the normocoagulable range of 5 to 15. Pentoxifylline was used in a dose of 400 mg four times per day and was increased in relation to the increases in maximal amplitude and fibrinogen. Ticlopidine was used sparingly (250 mg per day), most often in addition to aspirin when the aspirin dose exceeded 985 mg per day.
Results
Twenty-four patients received an implant for a total of 1,246 implant-days (mean duration, 53 days; median, 32 days; range, 3 to 186 days) with 1 patient currently on support for 42 days. The patient profile is shown in Table 1. The mean age was 47 years. Nearly all patients (20 of 24) were men. The mean body surface area was 2.1 ± 0.2 m2. All patients were dependent on inotropic agents. Preoperative status included 2 patients who failed to wean from cardiopulmonary bypass, 2 patients on extracorporeal membrane oxygenation support, 5 patients with preimplant cardiac arrests, 11 patients on ventilators, and 10 on intraaortic balloon pumps. Risk factors included previous cardiac operations in 13 patients, smoking history in 10, diabetes mellitus in 6, and history of hypertension of 4.
|
The mean cardiac index on TAH support was 3 L/m2 per minute. The mean perfusion pressure (mean arterial pressure minus central venous pressure, in mm Hg) increased from 49 mm Hg before implant to 68 mm Hg at 4 hours after implant and remained 20 to 22 mm Hg above the preoperative level for the first 24 hours. There were only two exceptions (both included in the overall mean), one who died of preimplant sepsis, and the second who died of multiorgan failure. The average time from implant to discharge from the intensive care unit was 11 ± 6 days and from transplant to discharge from the intensive care unit was 5 ± 3 days, and for transplant to discharge home was 17 ± 16 days. After the implant, patients were extubated at 4 ± 4 days, and were walking over 100 feet at 9 ± 5 days. At 30 days after transplantation, 17 patients were in New York Heart Association functional class I, and 2 were in class II.
There was one death in a 44-year-old patient whose Batista procedure failed on the 124th day after implant because of device failure from rupture of the first layer of the four-layered diaphragm. A second death occurred 22 days after implantation in a 43-year-old man after a 5-vessel coronary artery bypass grafting procedure. Operative mediastinal cultures at implantation grew multiple gram-negative bacteria. He died of sepsis. A third death occurred from multiple organ failure 3 days after implant in a 54-year-old man (2.1 m2). At the time of implant, he had hepatic dysfunction (total bilirubin, 3.3 mg/dL) and renal dysfunction (serum creatinine 2.4 mg/dL) in addition to severe congestive heart failure. After implantation, he required veno-venous extracorporeal membrane oxygenation in an attempt to overcome what appeared to be adult respiratory distress syndrome. The fourth death was the result of a central line problem. A percutaneous central line was inserted in the patient’s right arm with precautions (x-ray confirmation) to cut the catheter short enough to avoid the device’s right-sided inflow valve. Unfortunately, the line migrated into that valve freezing the disc in a closed position and causing instantaneous pump shutdown and death on the 14th day after implant.
Adverse events were recorded prospectively by a nurse practitioner and checked by an investigator. Careful documentation and strict use of the definitions for adverse events disclosed 128 events, among which 18 were judged serious because they contributed to death (n = 11), led to chest reoperation (six events, four reoperations with two overlapping definitions), or had neurologic residual (n = 2). Thirty-two adverse events were labeled device related, eight were clustered related to the one episode of device diaphragm rupture. Individual adverse events are summarized as follows. There were five episodes of bleeding. Four occurred during the first week after implantation.
Six episodes of device malfunction were noted. Four were inconsequential (two transient drive line kinks, one low-pressure compressed air tank, one hypovolemic patient). One was the diaphragm tear described above, which was diagnosed on the 123rd day after implant. Direct examination as well as microscopic and scanning electron microscopic examination of the explanted diaphragm failed to suggest a manufacturing defect. Review of all data on this type of diaphragm dating back to the early 1980s and including all Jarvik and Symbion implants failed to find a similar event. The investigation of this complication concluded that this was a unique event and that it appeared to have occurred on a random basis, not related to any systematic error in manufacturing technique. Another episode was the catheter obstruction of the TAH right-sided inflow valve.
There were four instances of fit complication when the device was repositioned, two occurring intraoperatively at the time of initial implantation and two requiring reoperation, one on day 1 and one on day 3. In all cases the problem was compression of the inferior vena cava by the right atrial quick connect anastomosis. Repositioning was accomplished by right ventriculopexy to the middle of the left sixth rib. All of these patients are currently surviving after transplantation.
There were ten instances of hemodynamic insufficiency, eight of which were transient drops in device output or in blood pressure that were immediately corrected. The two serious episodes were premortem.
Nine instances of hepatic dysfunction were during the first 3 days after implant, related to preoperative hepatic dysfunction, and reversed while on device support. One instance was late premortem.
There were 42 infections, including 13 upper respiratory, seven urinary tract, six drive line, five decubitus ulcers, two mediastinal (both in the patient who died of sepsis), two stool, one bronchial, one great toe, one lip herpes, one vaginal yeast, and one infected cutdown site.
Of the 13 neurologic adverse events, there were seven transient ischemic attacks, two strokes, one anoxic encephalopathy, one retinal hemorrhage, one syncope, and one seizure. The first stroke, right parietal infarct by single-photon emission computed tomography scan and symptoms, occurred at transplant and resolved at 3 months. The second stroke occurred at the time of implantation. The patient developed a mild expressive aphasia that has persisted. Her computed tomographic scan was negative. The linearized stroke rate was 0.05/month or 0.6/year. Four of these adverse events, including the first stroke, occurred in our first 2 patients prior to the current anticoagulation protocol.
There were two splenic emboli noted in our first patient by abdominal computed tomographic scan. There was also one late retinal embolus diagnosed by ophthalmologic examination. Neither embolism influenced survival.
There were 6 patients with renal dysfunction, all within the first 10 days after implantation, 2 of whom died. The other four events were reversed while patients were on TAH support.
The general reoperations were all minor operations. There were eight bronchoscopies, six chest tube placements, three upper gastrointesitnal endoscopies, and one each of transesophageal echo, thoracentesis, and placement on extracorporeal membrane oxygenation support.
There were four chest reoperations, three for bleeding or repositioning (all alive) and one for mediastinal sepsis (died).
There were nine reintubations for renal dysfunction, three in patients who died. The other six were early postimplant and temporary.
Thirty-two of the total 128 adverse events were judged to be device related. The adverse events were eight neurologic events including two strokes, five device malfunctions including the fatal diaphragm rupture, five hemodynamic insufficiency, four fit complications, four infections (all involving drive lines), three peripheral emboli (two splenic and one retinal), and one each of reoperation (transesophageal echo), respiratory dysfunction, and hepatic dysfunction. The patient who died from a diaphragm rupture accounted for 11 of these device-related events. Our first patient accounted for eight additional events.
Comment
The most striking observation in this study was that critically ill patients were routinely salvaged. Times to extubation and intensive care unit discharge, 4 and 11 days, respectively, were not excessive. Cardiac outputs on TAH support were routinely in the 6 to 7 L/minute range. End organ dysfunction was routinely reversed because of excellent cardiac outputs and blood pressures. We never used inotropic agents (other than renal-dose dopamine) or had to treat right ventricular failure or arrhythmias. Patients became candidates for transplantation, in fact, better candidates than our routine patients in view of their 100% survival rate although the transplantation was admittedly difficult because of adhesions from a mean implant time of 53 days.
Only four postimplant bleeding episodes and three reoperations for bleeding or repositioning accompanied implantation of the TAH. Thus, technical difficulties of implantation have been mostly overcome despite the critical and often technically difficult condition of the patients. We credit this to an improved implantation technique [8], intraoperative thromboelastography, and the use of high-dose aminocaproic acid at implant and aprotinin at transplant. Our implant technique is fairly simple compared with implantation of a left ventricular assist device, which requires a large abdominal pocket. We use Teflon felt buttresses placed with a whip stitch around the perimeter of the atrial cuffs to help anchor the atrial quick connectors and prevent atrial suture line bleeding. Simple end-to-end graft to great vessel anastomoses constitute the two other suture lines. Transesophageal echocardiography is an integral part of all procedures and is especially important in monitoring for trapped air and venous obstruction. Gore-Tex (W.L. Gore & Assoc, Flagstaff, AZ) membranes are used to close over the device joining the left and right sides of the pericardium and covering the device thus preventing adhesions to the lungs and chest wall. We prime the oxygenator with fresh frozen plasma at the time of transplant. We also routinely administer platelets and use in-room thromboelastography as a rapid guide to any other replacement therapy.
Twenty (83%) of 24 patients are surviving, including 1 who is on device support and all 19 (100%) of those transplanted. The longest survivor received an implant in January 1993. At 30 days after transplant, 17 of the patients had returned to normal function and two were functional class II.
Neurologic and peripheral embolic events resulted neither in patient death nor in significant permanent deficits. Our first patient, anticoagulated with subcutaneous heparin for 186 days, had two transient ischemic attacks and two peripheral emboli. Our next patient, still not treated with our present diagnostic and therapeutic protocol, had one stroke and one transient ischemic attack. The other patient with a stroke, like our second patient, had the event at the time of operation. Thus, in both cases, a relationship strictly to the device was not evident. However, we have adopted the convention of attributing both of those events to the device even though it might not be true.
We observed hypercoagulability in this group of patients but believe that with our combined anticoagulant and antiplatelet therapy a balance can be attained. The linearized thromboembolic rate for this series, counting all transient ischemic attacks, strokes, and peripheral emboli is 3.5 events per patient-year. If we eliminate our first 2 patients, who account for three transient ischemic attacks, one stroke, and two peripheral emboli, we are left with a linearized rate for the remaining 22 patients treated with our current protocol of 2.14 events per patient-year. Both of these rates are in the range of linearized thromboembolic rates for currently used valve prostheses [9]. Our data on bleeding are also of interest. Only five episodes met the criteria for bleeding, four were early after implant and one was late in the patient who died of device malfunction and multiorgan dysfunction, who bled from a chest tube. These data add to the evidence that a balance in the hemostatic mechanism has been attained. Patients who were taking therapeutic warfarin, 650 to 875 mg/day of aspirin, 400 to 800 mg/day of dipyridamole, and 1,200 mg/day of pentoxifylline did not bleed.
A second adverse event of interest is infection. We did not have any device-related mediastinal infections or sepsis. The infections we observed were typical for critically ill patients after cardiac operations and were mostly upper respiratory. We did find six drive line infections, but in no instance did they ascend into the chest. Examination of all devices after explant failed to reveal a single instance of infection of the device, vegetations, or thrombi.
We report a 17% mortality rate in this series. Causes of death were TAH diaphragm rupture, infection from a prior operation, multiorgan failure, and catheter entrapment of a valve disc. We believe the rupture to be an extremely rare random event. The catheter entrapment is a preventable event. We suggest that great caution be used if a central line is necessary in a TAH patient and that, in general, central lines be avoided. The two other deaths were in our view not preventable and resulted from severe patient disease.
In our view, the CardioWest TAH has become the device of choice for bridge to transplant in patients with advanced heart failure that is unresponsive to medical therapy and presents as rapid decompensation involving biventricular failure. It must be used only in patients in whom it will fit. Generally, these are people with a body surface area of at least 1.7 m2, with a large heart by chest radiography, or with an anterior-posterior dimension of 10 cm at T-10 by computed tomographic scan.
We have been able to salvage and then successfully transplant most of our patients, with acceptable mortality and morbidity rates. These are patients who, in our judgement, would not have survived with left ventricular support alone. On the basis of this experience, we recommend more widespread use of the CardioWest TAH in the bridge to transplant setting.
Footnotes
Dr Copeland is a medical adviser to CardioWest. He does not sit as an official member of their board or receive any financial consideration from CardioWest.
References
This article has been cited by other articles:
|
|
 |
|
  J. C. Roussel, T. Senage, O. Baron, C. Perigaud, O. Habash, J. C. Rigal, M. Treilhaud, J. N. Trochu, P. Despins, and D. Duveau CardioWest (Jarvik) Total Artificial Heart: A Single-Center Experience With 42 Patients Ann. Thorac. Surg., January 1, 2009; 87(1): 124 - 130. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. G. Copeland, R. G. Smith, R. K. Bose, P. H. Tsau, P. E. Nolan, and M. J. Slepian Risk Factor Analysis for Bridge to Transplantation With the CardioWest Total Artificial Heart Ann. Thorac. Surg., May 1, 2008; 85(5): 1639 - 1644. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. G. Copeland, R. G. Smith, F. A. Arabia, P. E. Nolan, G. K. Sethi, P. H. Tsau, D. McClellan, M. J. Slepian, and the CardioWest Total Artificial Heart Investigator Cardiac Replacement with a Total Artificial Heart as a Bridge to Transplantation N. Engl. J. Med., August 26, 2004; 351(9): 859 - 867. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. H. Delgado, V. Rao, H. J. Ross, S. Verma, and N. G. Smedira Mechanical Circulatory Assistance: State of Art Circulation, October 15, 2002; 106(16): 2046 - 2050. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
