- Email: [email protected]
Modifications in surgical implantation of the Penn State Electric Total Artificial Heart
Modifications in Surgical Implantation of the Penn State Electric Total Artificial Heart Eric M. Hoenicke, MD, Robert G. Strange, Jr, MD, William J. Weiss, PhD, Alan J. Snyder, PhD, Marjorie A. Rawhouser, PhD, Gerson Rosenberg, PhD, G. Allen Prophet, BS, Walter E. Pae, Jr, MD, Benjamin C. Sun, MD, and William S. Pierce, MD Department of Surgery and Artificial Organs, Penn State University College of Medicine, Hershey, Pennsylvania
Background. Two modifications of the surgical implantation protocol for the Penn State Total Artificial Heart (ETAH) were evaluated: Phrenic nerve ischemia was prevented by minimizing dissection and traction; and hemostasis was augmented and ETAH cuff anastomoses reinforced by using fibrin glue. Methods. Thirteen Holstein calves underwent orthotopic surgical implantation of the Penn State ETAH between February 1998 and August 2000. Mean hemodynamic and laboratory chemistry variables from the first postoperative week were compared between calves receiving the original (n ⴝ 7) and modified (n ⴝ 6) protocol. Results. Calves assigned to the modified protocol displayed an improvement in the PO2/FiO2 ratio compared to
original (419.4 ⴞ 17.5 vs 336.3 ⴞ 35.4, respectively; p ⴝ 0.05). All additional parameters were equivalent between groups. The percent survival of animals receiving the modified protocol at 2, 4, and 12 weeks was higher than that of animals that underwent the original protocol. Original-protocol calf deaths consisting of hemothorax (n ⴝ 3), and respiratory failure (n ⴝ 1) were prevented in the modified protocol. Conclusions. Our results suggest that manipulations in surgical protocol may promote increased survival in calves implanted with the Penn State ETAH.
H
inhibition of phrenic nerve blood supply. Fibrin glue was used to reinforce right atrial, left atrial, pulmonary artery, and aortic cuff anastomosis to promote more aggressive hemostasis. The hypothesis of this study was that phrenic nerve sparing and aggressive hemostasis augmented by fibrin glue could improve long-term survival by reducing pulmonary and bleeding complications.
eart transplantation remains the definitive treatment for end-stage heart disease despite decreasing organ supply. Alternative means of treating endstage heart disease include ventricular assist and total artificial heart devices, both of which have been under development at Penn State University College of Medicine over the past three decades [1–3]. The Penn State Electric Total Artificial Heart (ETAH) is currently undergoing preclinical testing, including in vitro mock circulatory loops and in vivo surgical implantation in calves. Many efforts to improve long-term survival of calves receiving ETAH implantation have been aimed at reducing pulmonary postoperative complications [4, 5]. These complications have been associated with prolonged cardiopulmonary bypass times and appear similar to the adult respiratory distress syndrome, presumably from overstimulation of systemic cytokine cascades [5]. Although cardiopulmonary bypass times have recently been decreased, little effect has been noted on animal survival. To facilitate improved postoperative survival of calves receiving implantation of the Penn State ETAH, two major modifications of the original surgical procedure were introduced. Previous efforts to surgically dissect the long right phrenic nerve were minimized to prevent Presented at the Fifth International Conference on Circulatory Support Devices for Severe Cardiac Failure, New York, NY, Sept 15–17, 2000. Address reprint requests to Dr Pierce, Department of Surgery and Artificial Organs, Penn State University College of Medicine, 500 University Dr, Hershey, PA 17033; e-mail: [email protected].
© 2001 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
(Ann Thorac Surg 2001;71:S150 –5) © 2001 by The Society of Thoracic Surgeons
Material and Methods Holstein calves weighing between 85 and 104 kg were used in this study. All of the animals involved in this study received humane care in Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) International–accredited (00036), USDAregistered (52-R-007) facilities in compliance with the “Principles of Laboratory Animal Care” prepared by the Institute of Laboratory Animal Resources and published by the National Institutes of Health (NIH publication 85-23, revised 1985). The Principal Investigator, Dr Pierce, and Dr Weiss, Dr Rosenberg, Dr Snyder, Mr Prophet, Dr Sun, Dr Pae, and Penn State University College of Medicine own shares in ABIOMED, Inc, the licensee of the Penn State Total Artificial Heart technology. Dr Rawhauser is an employee of ABIOMED, Inc.
0003-4975/01/$20.00 PII S0003-4975(00)02640-0
Ann Thorac Surg 2001;71:S150 –5
CIRCULATORY SUPPORT HOENICKE ET AL PENN STATE ELECTRIC TOTAL ARTIFICIAL HEART
S151
internal battery are contained in a titanium canister that also contains all of the components for energy transmission regulation and telemetry. The control system allows adjustments in cardiac output based on variations in venous return and pulmonary and systemic vascular resistance by a control algorithm developed at Penn State and described in detail previously [2].
Surgical ETAH Implantation
Fig 1. Penn State Electric Total Artificial Heart system (ETAH). From left to right (clockwise): intrathoracic compliance chamber, ETAH energy converter/blood pumps, compliance chamber infusion port, implanted electronics assembly energy transmission coils, electronic controller unit with internal batteries.
Penn State Completely Implantable Electric Total Artificial Heart System The Penn State completely implantable electric total artificial heart system (ETAH), shown in Figure 1, has been described in detail previously [6, 7]. In brief, the Penn State ETAH system is comprised of an energy converter with left and right blood pumps that are alternately compressed by pusher plates. The energy converter, positioned between the two pumps, houses a brushless DC motor and roller screw actuator (Fig 2). This assembly weighs 510 g. Approximately 4.5 revolutions of the motor are converted to a 1.9-cm linear stroke. The motor position is detected by three Hall effect sensors. Each blood pump contains flexible, seamless, segmented polyether polyurethane– urea blood sacs and 25-mm outlet and 27-mm inlet Delrin monostrut tilting disk valves (Arrow, Reading, PA). Passive filling of the pumps is facilitated because the pusher plates are free of attachments to the blood sacs. This prevents atrial suction and allows the control system to actively adjust left and right stroke volumes. The displacement of the pumps is 70 mL, and the dynamic stroke volume is 55 to 60 mL. Additional components of the ETAH system include an implanted electronic controller, compliance chamber with infusion port, energy transmission coils, external energy transmission electronics, and a portable power pack. All energy needed by the implanted system is supplied through the transcutaneous energy transmission system (TETS). The electronic controller and the
After rapid induction of anesthesia with methohexital (375 mg) and endotracheal intubation, calves were placed on the operating room table in the left lateral decubitus position. Mechanical ventilation was initiated and anesthesia was maintained with isoflurane (1% to 2%). A right thoracotomy was performed through the fourth intercostal space. In animals receiving implantation of the ETAH between February 1998 and June 1999 (group I), the right phrenic nerve was dissected free of adjacent anatomic structures beginning at the level of the proximal azygos vein and ending at the level of the diaphragmatic surface of the pericardium. This approximately 16-cm portion of the phrenic nerve was retracted freely out of the operating field. Animals receiving ETAH implantation between August 1999 and August 2000 (group II) did not undergo phrenic nerve dissection except for a 1-cm section where umbilical tape was passed underneath the superior and inferior vena cava for cardiopulmonary bypass (CPB) pump cannula securing. After stable CPB had been accomplished, the native calf heart was totally excised. Atrial connector cuffs and arterial connector grafts were sutured and leak tested. Leaks were repaired by additional hemostatic stitches in group I animals. All animals in group II underwent application of fibrin glue (Baxter AG Tisseel Fibrin Sealant, Baxter Healthcare, Glendale,
Fig 2. Computer aided design rendering of the Electric Total Artificial Heart blood pump/energy converter.
S152
CIRCULATORY SUPPORT HOENICKE ET AL PENN STATE ELECTRIC TOTAL ARTIFICIAL HEART
CA) to all anastomoses. Each of the four atrial or arterial conduits contained end bolts that were attached to the pump assembly and tightened. Meticulous deairing of the right and left pumps was accomplished, and pumping was initiated slowly as CPB was weaned. The chest, flank, and carotid incisions were closed after the placement of three standard thoracostomy tubes (left pleura, right pleura, and right pump) and placement of one Jackson-Pratt drain deep into the controller unit.
Cardiopulmonary Bypass Circuit The CPB system was primed with lactated Ringer’s solution and mannitol. A standard membrane oxygenator was employed. Venous cannulas supplied the CPB circuit from the superior and inferior vena cava (singlestage 34F, 36F, or 40F Medtronic DLP cannula (Medtronic, Minneapolis, MN), Grand Rapid, MI). Oxygenated arterial blood was returned from the circuit into the right common carotid artery (5.8- or 6.4-mm stainless steel cannula). Calves were cooled systemically and maintained at temperatures in the range of 28°C to 30°C. The lungs were ventilated every 20 minutes during CPB. Methylprednisolone sodium succinate (1 g) was slowly administered intravenously 2 minutes before initiation of CPB. Animals underwent full heparinization with confirmation of activated clotting time (ACT) more than 600 seconds before CPB.
Postoperative Animal Care Postoperative care for each animal receiving ETAH implantation was similar to that for human patients. All calves were transported to the recovery area after completion of the operation. Calves were extubated after they were able to stand and after arterial Po2 was greater than or equal to 100 mm Hg on T-piece respirations. Extubation generally occurred 1 to 4 hours after the surgical operation. Vital signs, arterial blood gas hemodynamic data, and ETAH systems were monitored. Routine hematologic and biochemical analyses were assessed periodically. Prophylactic antibiotics consisting of firstgeneration cephalosporin (kefzol) and aminoglycosides (gentamicin or tobramicin) were initiated preoperatively and continued for approximately 1 week postoperatively. Coumadin was administered to maintain prothrombin times approximately 1.5 to 2.0 times baseline values. All calves were sacrificed 12 weeks after ETAH implantation.
Ann Thorac Surg 2001;71:S150 –5
postoperative days 0 to 7 was calculated using the following formula: positive water balance ⫽ (infused water ⫹ transfused blood ⫹ oral water) ⫺ (urine ⫹ drain outputs). Drain output represented the total output of two right chest thoracostomy tubes plus one left thoracostomy tube plus one flank Jackson-Pratt drain. Transfused blood was the cumulative total of whole blood transfused during the initial postoperative week. Urine output was compared as mean daily urine output and mean total urine output, representing cumulative urine production during the first postoperative week. Total drain outputs for the three thoracostomy tubes and Jackson-Pratt drain was averaged during postoperative days 0 through 7.
Survival and Causes of Death Studies on 13 calves were carried out between February 1998 and August 2000. All calves underwent routine orthotopic implantation of the Penn State 70-mL ETAH with the sealed system without percutaneous leads. Seven calves received implantation of the ETAH in group I (February 1998 to June 1999), and six calves underwent ETAH implantation with the two previously mentioned surgical modifications in group II (August 1999 to August 2000). Causes of death were compared between groups I and II. All calves underwent autopsy to reveal the cause of death. If the calves survived the 12-week protocol, death was characterized as a scheduled termination. Other causes of death included massive hemothorax, gastrointestinal bleed, early device failure (on-table), late device failure, idiopathic pulmonary failure, and superior vena cava syndrome with subsequent cerebrovascular accident. Percentages were calculated to represent percent survival during three time periods: 2 weeks, 4 weeks, and 12 weeks. Intraoperative and immediate postoperative deaths were excluded from percent survival calculations. Two group I animals died of intraoperative device failures and one group II animal died immediately postoperatively.
Statistical Analysis Data are presented as mean ⫾ standard error of the mean. A t test was used to compare all group I and group II parameters. A p value of 0.05 or less was considered indicative of a statistically significant difference.
Study Endpoints Survival days, CPB time, and anesthesia time were compared between group I and group II calves. Mean hematologic and biochemical data were also compared between the two groups. These data represented mean values measured during the first postoperative week after ETAH implantation. Mean ETAH pump flow, central venous pressure, Pco2, Po2/Fio2 ratio, respiratory rate, hematocrit, plasma hemaglobin, leukocyte count, total protein (percent baseline), albumin (percent baseline), serum glutamic oxaloacetate transaminase, and creatinine were compared. Cumulative water balance on
Results Anesthesia and CPB Times Total CPB and anesthesia times were not statistically different between groups I and II (Table 1).
System Parameters and Hemodynamics of ETAH Mean ETAH pump flow was not statistically significantly different between groups I and II (Table 1). Likewise, central venous pressure was not significantly different between groups.
Ann Thorac Surg 2001;71:S150 –5
CIRCULATORY SUPPORT HOENICKE ET AL PENN STATE ELECTRIC TOTAL ARTIFICIAL HEART
Table 1. Mean Hemodynamic and Hematologic Parameters in Calves Surgically Implanted With the Penn State Electric Total Artificial Heart Endpoint Survival times (days) CPB time (min) Anesthesia time (min) Flow (L/min) CVP (mm Hg) (⫹) Water balance (L) Respiratory rate (respirations/min) Pco2 (mm Hg) Po2/Fio2 ratio Transfused blood (L) Leukocyte count (⫻ 103/mm3) Hematocrit (%) Plasma hemaglobin (mg%) Total protein (% baseline) Albumin (% baseline) SGOT (IU) Creatinine (mg/dL)
Group I (n ⫽ 7)
Group II (n ⫽ 6)
p Value
25.6 ⫾ 16.3 65.0 ⫾ 14.0 142.2 ⫾ 6.6 135.5 ⫾ 5.4 370.5 ⫾ 15.1 375.8 ⫾ 29.5 6.3 ⫾ 0.7 7.2 ⫾ 0.2 7.9 ⫾ 1.8 9.2 ⫾ 0.9 5.7 ⫾ 3.9 11.7 ⫾ 3.3 32.2 ⫾ 2.4 30.2 ⫾ 1.6
0.10 0.45 0.88 0.20 0.50 0.27 0.51
48.2 ⫾ 2.9 46.4 ⫾ 2.3 336.3 ⫾ 35.4 419.4 ⫾ 17.5 8.0 ⫾ 2.0 7.5 ⫾ 2.5 17.3 ⫾ 5.5 13.3 ⫾ 3.4 27.4 ⫾ 1.7 23.9 ⫾ 1.8 5.1 ⫾ 1.5 4.1 ⫾ 0.5 64.5 ⫾ 3.1 63.8 ⫾ 2.5 67.2 ⫾ 6.5 63.3 ⫾ 3.1 228.4 ⫾ 38.5 220.6 ⫾ 55.5 0.9 ⫾ 0.1 0.9 ⫾ 0.1
0.63 0.05 0.89 0.55 0.20 0.55 0.87 0.56 0.91 0.83
All values represent mean ⫾ standard error of the mean during the first postoperative week. CPB ⫽ cardiopulmonary bypass; CVP ⫽ central venous pressure; Pco2 ⫽ arterial partial pressure of carbon dioxide; Po2/Fio2 ratio ⫽ arterial pressure of oxygen-to-fraction of inspired oxygen; SGOT ⫽ serum glutamic oxaloacetic transaminase.
Hematologic and Biochemical Endpoints Mean values for postoperative days 0 to 7 were not statistically significantly different between groups with regard to leukocyte count, hematocrit, plasma hemaglobin, total protein, albumin, serumglutamic oxaloacetic transaminase (SGOT), creatinine, and arterial Pco2 (Table 1). The Po2 to Fio2 ratio was statistically significantly improved in group II compared to group I ( p ⫽ 0.05) (Table 1).
Water Balance, Urine Output, and Drain Output Table 2 shows that there was no statistically significant differences between positive water balance, total urine output, mean daily urine output, and volume of transfused blood in the first postoperative week. There were no significant differences observed in total drain output for left pleural tube, right pleural tube, right pump tube, and right flank JP drain (Table 2). Preoperative weights were also not significantly different between groups (Table 2).
S153
Table 2. Mean Water Balance, Surgical Drainage Output, and Volume of Transfused Blood in Calves Surgically Implanted with the Penn State Electric Total Artificial Heart Endpoint Preoperative weight (kg) (⫹) Water balance (L) Total urine total (L) Daily urine output (L) Total left pleural tube (L) Total right pleural tube (L) Total right pump tube (L) Total right flank JP drain (L) Transfused blood (L)
Group I (n ⫽ 7)
Group II (n ⫽ 6)
p Value
90.9 ⫾ 1.0 5.7 ⫾ 3.9 22.5 ⫾ 7.5 5.9 ⫾ 1.3 2.8 ⫾ 1.4 1.2 ⫾ 0.6 2.4 ⫾ 1.0 1.0 ⫾ 0.7 8.0 ⫾ 2.0
94.8 ⫾ 2.6 11.7 ⫾ 3.3 33.2 ⫾ 7.4 5.6 ⫾ 0.5 2.7 ⫾ 0.7 0.9 ⫾ 0.3 2.5 ⫾ 0.7 0.4 ⫾ 0.1 7.5 ⫾ 2.5
0.19 0.27 0.34 0.83 0.92 0.71 0.90 0.40 0.89
All values represent mean ⫾ standard error of the mean during the first postoperative week. JP drain ⫽ Jackson-Pratt drain.
additional calf died because of pulmonary failure of unknown cause. One calf survived in group I until scheduled termination at 12 weeks. Group II deaths resulted from mediastinal hematoma resulting in an acute superior vena cava syndrome and subsequent massive cerebrovascular accident (6 hours postoperatively), gastrointestinal bleed (19 days postoperatively), and device failure (46 days postoperatively) (Fig 3). The deaths of the remaining three calves were 12-week scheduled terminations (Fig 3). Excluding intraoperative and immediate postoperative deaths, the 2-week survival was 40% in group I and 100% in group II (Fig 4). One-month survival was 40% in group I and 80% in group II. The percentage of animals surviving to the 12-week scheduled termination was 20% in group I and 60% in group II.
Comment Circulatory support devices designed to augment or to completely replace hearts in patients with end-stage heart disease have been under development at Penn State over the past three decades [1–3]. Developmental and engineering milestones were initially achieved with the Thoratec (Pierce-Donachy) ventricular assist device as treatment for severe cardiac stunning during cardiopulmonary bypass and, later, as a bridge to transplanta-
Causes of Death and Survival Group II animals survived longer than those in group I (65.0 ⫾ 14.0 days vs 25.6 ⫾ 16.3 days, respectively); however this was not statistically significant ( p ⫽ 0.10) (Table 1). Figure 3 compares the causes of death in both groups I and II. Two calves in group I died on-table from device failures. Three group I calves died of massive hemothorax on postoperative days 1, 6, and 34. An
Fig 3. Causes of deaths in calves surgically implanted with the Penn State Electric Total Artificial Heart.
S154
CIRCULATORY SUPPORT HOENICKE ET AL PENN STATE ELECTRIC TOTAL ARTIFICIAL HEART
Fig 4. Comparison of calf survival between group I and group II 2, 4, and 12 weeks postoperatively.
tion. More recently, the Arrow–Penn State Lionheart II LVD (Arrow, Reading, PA) has been introduced with success as a permanent left ventricular assist device. The production of a totally implantable artificial heart has proved to be a challenging endeavor, which has resulted in the design and fabrication of the Penn State Electric Total Artificial Heart (ETAH) at our institution. The Penn State ETAH is currently undergoing both in vitro and in vivo preclinical evaluation. The long-term survival of animals undergoing surgical implantation of total artificial hearts has been limited by various postoperative complications. Earlier reports of the survival of calves receiving ETAH implantation focused on deaths related to pulmonary complications [5]. A significant number of postoperative deaths were attributable to pulmonary complications. It was demonstrated that longer cardiopulmonary bypass times were associated with decreased 2-week survival. Kuroda and colleagues [5] argued that these early postoperative deaths were produced by a syndrome similar to adult respiratory distress syndrome (ARDS). This clinical entity was believed to be a product of harmful complement activation associated with long (3 hours or more) cardiopulmonary bypass times. Because Kuroda and colleagues [5] demonstrated an association between shorter cardiopulmonary bypass (CPB) times and improved survival, [5] subsequent experiments focused on minimizing CPB times. Compared to long-term survivors in the study of Kuroda and colleagues [5], CPB times for all calves in the present study were significantly decreased (156.0 ⫾ 6.4 minutes vs 138.8 ⫾ 4.2 minutes, p ⫽ 0.032). With the shorter CPB times in the current study, the number of postoperative ARDS-like deaths was decreased compared to earlier reports [5]. Despite these shorter CPB times, overall postoperative survival between 1998 and 1999 was low. To improve the results of these calf studies, our group modified the surgical implantation protocol. To decrease further the likelihood of postoperative pulmonary complications, all efforts to skeletonize and retract the phrenic nerve were minimized. Our group reasoned that
Ann Thorac Surg 2001;71:S150 –5
dissection of the right phrenic nerve would add to CPB time and potentially result in ischemic phrenic nerve damage. The importance of minimizing traction and dissection of the phrenic nerve has been previously described by Lick and colleagues [8] in a simplified technique of heart–lung transplantation. Vascular or traction injury to the phrenic nerve may result in significant perturbations in pulmonary function [9]. The second surgical modification was the introduction of fibrin glue to reinforce right and left atrial, pulmonary artery, and aortic anastomoses. Our group reasoned that CPB time could be minimized by reducing the time repairing anastomotic bleeding. Moreover, the risks of postoperative hemothorax would be minimized. The comparison of hemodynamic and laboratory values of group I calves (before modifications) to group II calves (after modifications) did not yield many differences. The avoidance of phrenic nerve dissection and fibrin glue application did not reduce the time spent on CPB, nor did these modifications augment postoperative flow or central venous pressure. Water balance and laboratory values indicative of infection, bleeding, hemolysis, liver function, and renal function were also not different between groups. Our results did indicate a significant difference between oxygenation in the two groups. Oxygenation was measured by comparing the ratio of arterial Po2 to fraction of inspired oxygen (P:F ratios). Animals that were spared phrenic nerve dissection displayed Po2/Fio2 ratios that were virtually normal, and animals that underwent phrenic nerve dissection exhibited lower ratios. Although this difference could be attributable to spared postoperative right phrenic nerve function, a difference in ventilation between the two groups would also have been anticipated with impaired ventilation in group I compared to group II. The results show that ventilation was not improved in group II compared to group I because the Pco2 values were statistically equivalent. If postoperative tidal volumes were pathologically reduced because of phrenic nerve injury, an increase in respiratory rate would also be anticipated; however, respiratory rate was also equivalent between groups. Hematological differences attributable to improved postoperative hemostasis from fibrin glue application were also not observed in the results. Improved hemostasis may have been confirmed by decreased volume of postoperative blood transfusions, decreased thoracostomy tube drainage, and higher hematocrit values. Comparison of these parameters between group I and group II animals yielded no statistically significant differences. Occult hemothoraces resulting from nonfunctioning thoracostomy tubes or hemothoraces occurring after removal of thoracostomy tubes may manifest clinically as hypoventilation syndromes. However, as previously mentioned, no significant differences in arterial Pco2 and respiratory rates were observed between groups. Caution should be applied postulating a causative link between the two surgical modifications and improved Po2/Fio2 ratios. The small numbers involved in the current study do not allow for speculation that right phrenic
Ann Thorac Surg 2001;71:S150 –5
nerve sparing or fibrin glue–augmented hemostasis resulted in improved oxygenation. Because multiple variations in surgical protocol were introduced in the present study, the causal relationship between each modification and improved oxygen ratios is not possible. However, as there was a trend of increased survival days in group II compared to group I (65 ⫾ 14 days vs 26 ⫾ 16 days, respectively; p ⫽ 0.10), the Po2/Fio2 ratio may serve as a positive prognostic indicator for improved long-term survival. Despite the small number of laboratory and hemodynamic differences observed between group I and group II animals, comparison between causes of death in groups I and II revealed many interesting trends. In group I, one calf died on postoperative day 1 of pulmonary complications. Figure 1 shows this death to be related to idiopathic pulmonary failure. This animal showed signs of possible phrenic nerve damage, as its mean Pco2 was 57.4 with a mean respiratory rate of 37 and mean P:F ratio of 208, and eventually succumbed to respiratory failure and death. Although this pattern of death approximates the ARDSlike syndrome described by Kuroda and colleagues [5], this animal was not exposed to a prolonged CPB time (CPB time, 129 minutes). No animals in group II showed this pattern of death. The major cause of death in group I calves was massive hemothorax. Three of 7 calves died of massive hemothorax in group I. Two of these deaths were early postoperative deaths (postoperative days 1 and 6) and the other occurred much later (postoperative day 34). No animals in group II died of massive hemothorax. We may speculate that the use of fibrin glue in Group II decreased the frequency of irreversible massive hemothorax and subsequent death; however, the hematologic and thoracostomy drainage data did not reveal any substantiating differences. Overall, calves in group II showed overall improved survival at 2, 4, and 12 weeks after surgical implantation of the Penn State ETAH. Although the improved results do seem to correlate with the two surgical modifications, it is difficult to explain these improved results given the lack of differences observed in hemodynamic, laboratory, and intrathoracic drain data. It is unclear whether improved early postoperative oxygenation was related to the surgical modifications. It is also unknown whether the improved Po2/Fio2 ratios underlie improved survival. These studies are ongoing and, as more ETAHs are implanted, differences in these endpoints may be elaborated. A larger number of studies will need to be performed to address these issues.
CIRCULATORY SUPPORT HOENICKE ET AL PENN STATE ELECTRIC TOTAL ARTIFICIAL HEART
S155
Conclusions Two modifications in the surgical protocol of implantation of the Penn State Electric Total Artificial Heart were initiated to promote improved long-term survival in calves. Both minimizing phrenic nerve dissection and applying fibrin glue correlated with improved survival of calves implanted with the ETAH. Previous deaths related to hypoventilation and massive hemothorax were prevented in calves that underwent the modified surgical protocol; however, analysis of hemodynamic and hematologic variables did not fully explain these survival differences. Early postoperative oxygenation was improved in the calves receiving the modified surgical protocol. It is not possible to ascertain whether improved oxygenation accounts for the differences in overall calf survival.
Supported by USPHS contract 1 HV-38130 and USPHS grant 1 RO1 HL 60276-01A1.
References 1. Pierce WS, Sapirstein JS, Pae WE Jr. Total artificial heart: from bridge to tranplantation to permanent use. Ann Thorac Surg 1996;61:342– 6. 2. Weiss WJ, Rosenberg G, Snyder AJ, et al. Steady state hemodynamic and energetic characterization of the Penn State/3M Health Care Total Artificial Heart. ASAIO Journal 1999;45:189–93. 3. Pierce WS, Rosenberg G, Snyder AJ, Pae WE, Waldhausen JA. An electric artificial heart for clinical use. Ann Surg 1990;212: 339– 43. 4. Honda T, Kito Y, Gibson W, Nemoto T, Cockrell JV, Akutsu T. Lung problem and liver damage in total artificial heart and their prevention. ASAIO Trans 1974;20:27–32. 5. Kuroda H, Rosenberg G, Snyder AJ, et al. Postoperative pulmonary complications in calves after implantation of an electric total artificial heart. ASAIO J 1998;44:M613– 8. 6. Weiss WJ, Rosenberg G, Snyder AJ, et al. Recent improvements in a completely implanted total artificial heart. ASAIO J 1996;42:M342– 6. 7. Rosenberg G, Snyder AJ, Weiss WJ, Sapirstein JS, Pierce WS. In vivo testing of a clinical-size totally implantable artificial heart. In: Unger F, ed. Assisted circulation 4. Berlin: SpringerVerlag, 1995, 236 – 48. 8. Lick SD, Copeland JG, Rosado LJ, Arabia FA, Sethi GK. Simplified technique of heart-lung transplantation. Ann Thorac Surg 1995;59:1592–3. 9. Abd AG, Braun NMT, Baskin MI, O’Sullivan MM, Alkaitis DA. Diaphragmatic dysfunction after open heart surgery: treatment with a rocking bed. Ann Intern Med 1989;111: 881– 6.
Background. Two modifications of the surgical implantation protocol for the Penn State Total Artificial Heart (ETAH) were evaluated: Phrenic nerve ischemia was prevented by minimizing dissection and traction; and hemostasis was augmented and ETAH cuff anastomoses reinforced by using fibrin glue. Methods. Thirteen Holstein calves underwent orthotopic surgical implantation of the Penn State ETAH between February 1998 and August 2000. Mean hemodynamic and laboratory chemistry variables from the first postoperative week were compared between calves receiving the original (n ⴝ 7) and modified (n ⴝ 6) protocol. Results. Calves assigned to the modified protocol displayed an improvement in the PO2/FiO2 ratio compared to
original (419.4 ⴞ 17.5 vs 336.3 ⴞ 35.4, respectively; p ⴝ 0.05). All additional parameters were equivalent between groups. The percent survival of animals receiving the modified protocol at 2, 4, and 12 weeks was higher than that of animals that underwent the original protocol. Original-protocol calf deaths consisting of hemothorax (n ⴝ 3), and respiratory failure (n ⴝ 1) were prevented in the modified protocol. Conclusions. Our results suggest that manipulations in surgical protocol may promote increased survival in calves implanted with the Penn State ETAH.
H
inhibition of phrenic nerve blood supply. Fibrin glue was used to reinforce right atrial, left atrial, pulmonary artery, and aortic cuff anastomosis to promote more aggressive hemostasis. The hypothesis of this study was that phrenic nerve sparing and aggressive hemostasis augmented by fibrin glue could improve long-term survival by reducing pulmonary and bleeding complications.
eart transplantation remains the definitive treatment for end-stage heart disease despite decreasing organ supply. Alternative means of treating endstage heart disease include ventricular assist and total artificial heart devices, both of which have been under development at Penn State University College of Medicine over the past three decades [1–3]. The Penn State Electric Total Artificial Heart (ETAH) is currently undergoing preclinical testing, including in vitro mock circulatory loops and in vivo surgical implantation in calves. Many efforts to improve long-term survival of calves receiving ETAH implantation have been aimed at reducing pulmonary postoperative complications [4, 5]. These complications have been associated with prolonged cardiopulmonary bypass times and appear similar to the adult respiratory distress syndrome, presumably from overstimulation of systemic cytokine cascades [5]. Although cardiopulmonary bypass times have recently been decreased, little effect has been noted on animal survival. To facilitate improved postoperative survival of calves receiving implantation of the Penn State ETAH, two major modifications of the original surgical procedure were introduced. Previous efforts to surgically dissect the long right phrenic nerve were minimized to prevent Presented at the Fifth International Conference on Circulatory Support Devices for Severe Cardiac Failure, New York, NY, Sept 15–17, 2000. Address reprint requests to Dr Pierce, Department of Surgery and Artificial Organs, Penn State University College of Medicine, 500 University Dr, Hershey, PA 17033; e-mail: [email protected].
© 2001 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
(Ann Thorac Surg 2001;71:S150 –5) © 2001 by The Society of Thoracic Surgeons
Material and Methods Holstein calves weighing between 85 and 104 kg were used in this study. All of the animals involved in this study received humane care in Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) International–accredited (00036), USDAregistered (52-R-007) facilities in compliance with the “Principles of Laboratory Animal Care” prepared by the Institute of Laboratory Animal Resources and published by the National Institutes of Health (NIH publication 85-23, revised 1985). The Principal Investigator, Dr Pierce, and Dr Weiss, Dr Rosenberg, Dr Snyder, Mr Prophet, Dr Sun, Dr Pae, and Penn State University College of Medicine own shares in ABIOMED, Inc, the licensee of the Penn State Total Artificial Heart technology. Dr Rawhauser is an employee of ABIOMED, Inc.
0003-4975/01/$20.00 PII S0003-4975(00)02640-0
Ann Thorac Surg 2001;71:S150 –5
CIRCULATORY SUPPORT HOENICKE ET AL PENN STATE ELECTRIC TOTAL ARTIFICIAL HEART
S151
internal battery are contained in a titanium canister that also contains all of the components for energy transmission regulation and telemetry. The control system allows adjustments in cardiac output based on variations in venous return and pulmonary and systemic vascular resistance by a control algorithm developed at Penn State and described in detail previously [2].
Surgical ETAH Implantation
Fig 1. Penn State Electric Total Artificial Heart system (ETAH). From left to right (clockwise): intrathoracic compliance chamber, ETAH energy converter/blood pumps, compliance chamber infusion port, implanted electronics assembly energy transmission coils, electronic controller unit with internal batteries.
Penn State Completely Implantable Electric Total Artificial Heart System The Penn State completely implantable electric total artificial heart system (ETAH), shown in Figure 1, has been described in detail previously [6, 7]. In brief, the Penn State ETAH system is comprised of an energy converter with left and right blood pumps that are alternately compressed by pusher plates. The energy converter, positioned between the two pumps, houses a brushless DC motor and roller screw actuator (Fig 2). This assembly weighs 510 g. Approximately 4.5 revolutions of the motor are converted to a 1.9-cm linear stroke. The motor position is detected by three Hall effect sensors. Each blood pump contains flexible, seamless, segmented polyether polyurethane– urea blood sacs and 25-mm outlet and 27-mm inlet Delrin monostrut tilting disk valves (Arrow, Reading, PA). Passive filling of the pumps is facilitated because the pusher plates are free of attachments to the blood sacs. This prevents atrial suction and allows the control system to actively adjust left and right stroke volumes. The displacement of the pumps is 70 mL, and the dynamic stroke volume is 55 to 60 mL. Additional components of the ETAH system include an implanted electronic controller, compliance chamber with infusion port, energy transmission coils, external energy transmission electronics, and a portable power pack. All energy needed by the implanted system is supplied through the transcutaneous energy transmission system (TETS). The electronic controller and the
After rapid induction of anesthesia with methohexital (375 mg) and endotracheal intubation, calves were placed on the operating room table in the left lateral decubitus position. Mechanical ventilation was initiated and anesthesia was maintained with isoflurane (1% to 2%). A right thoracotomy was performed through the fourth intercostal space. In animals receiving implantation of the ETAH between February 1998 and June 1999 (group I), the right phrenic nerve was dissected free of adjacent anatomic structures beginning at the level of the proximal azygos vein and ending at the level of the diaphragmatic surface of the pericardium. This approximately 16-cm portion of the phrenic nerve was retracted freely out of the operating field. Animals receiving ETAH implantation between August 1999 and August 2000 (group II) did not undergo phrenic nerve dissection except for a 1-cm section where umbilical tape was passed underneath the superior and inferior vena cava for cardiopulmonary bypass (CPB) pump cannula securing. After stable CPB had been accomplished, the native calf heart was totally excised. Atrial connector cuffs and arterial connector grafts were sutured and leak tested. Leaks were repaired by additional hemostatic stitches in group I animals. All animals in group II underwent application of fibrin glue (Baxter AG Tisseel Fibrin Sealant, Baxter Healthcare, Glendale,
Fig 2. Computer aided design rendering of the Electric Total Artificial Heart blood pump/energy converter.
S152
CIRCULATORY SUPPORT HOENICKE ET AL PENN STATE ELECTRIC TOTAL ARTIFICIAL HEART
CA) to all anastomoses. Each of the four atrial or arterial conduits contained end bolts that were attached to the pump assembly and tightened. Meticulous deairing of the right and left pumps was accomplished, and pumping was initiated slowly as CPB was weaned. The chest, flank, and carotid incisions were closed after the placement of three standard thoracostomy tubes (left pleura, right pleura, and right pump) and placement of one Jackson-Pratt drain deep into the controller unit.
Cardiopulmonary Bypass Circuit The CPB system was primed with lactated Ringer’s solution and mannitol. A standard membrane oxygenator was employed. Venous cannulas supplied the CPB circuit from the superior and inferior vena cava (singlestage 34F, 36F, or 40F Medtronic DLP cannula (Medtronic, Minneapolis, MN), Grand Rapid, MI). Oxygenated arterial blood was returned from the circuit into the right common carotid artery (5.8- or 6.4-mm stainless steel cannula). Calves were cooled systemically and maintained at temperatures in the range of 28°C to 30°C. The lungs were ventilated every 20 minutes during CPB. Methylprednisolone sodium succinate (1 g) was slowly administered intravenously 2 minutes before initiation of CPB. Animals underwent full heparinization with confirmation of activated clotting time (ACT) more than 600 seconds before CPB.
Postoperative Animal Care Postoperative care for each animal receiving ETAH implantation was similar to that for human patients. All calves were transported to the recovery area after completion of the operation. Calves were extubated after they were able to stand and after arterial Po2 was greater than or equal to 100 mm Hg on T-piece respirations. Extubation generally occurred 1 to 4 hours after the surgical operation. Vital signs, arterial blood gas hemodynamic data, and ETAH systems were monitored. Routine hematologic and biochemical analyses were assessed periodically. Prophylactic antibiotics consisting of firstgeneration cephalosporin (kefzol) and aminoglycosides (gentamicin or tobramicin) were initiated preoperatively and continued for approximately 1 week postoperatively. Coumadin was administered to maintain prothrombin times approximately 1.5 to 2.0 times baseline values. All calves were sacrificed 12 weeks after ETAH implantation.
Ann Thorac Surg 2001;71:S150 –5
postoperative days 0 to 7 was calculated using the following formula: positive water balance ⫽ (infused water ⫹ transfused blood ⫹ oral water) ⫺ (urine ⫹ drain outputs). Drain output represented the total output of two right chest thoracostomy tubes plus one left thoracostomy tube plus one flank Jackson-Pratt drain. Transfused blood was the cumulative total of whole blood transfused during the initial postoperative week. Urine output was compared as mean daily urine output and mean total urine output, representing cumulative urine production during the first postoperative week. Total drain outputs for the three thoracostomy tubes and Jackson-Pratt drain was averaged during postoperative days 0 through 7.
Survival and Causes of Death Studies on 13 calves were carried out between February 1998 and August 2000. All calves underwent routine orthotopic implantation of the Penn State 70-mL ETAH with the sealed system without percutaneous leads. Seven calves received implantation of the ETAH in group I (February 1998 to June 1999), and six calves underwent ETAH implantation with the two previously mentioned surgical modifications in group II (August 1999 to August 2000). Causes of death were compared between groups I and II. All calves underwent autopsy to reveal the cause of death. If the calves survived the 12-week protocol, death was characterized as a scheduled termination. Other causes of death included massive hemothorax, gastrointestinal bleed, early device failure (on-table), late device failure, idiopathic pulmonary failure, and superior vena cava syndrome with subsequent cerebrovascular accident. Percentages were calculated to represent percent survival during three time periods: 2 weeks, 4 weeks, and 12 weeks. Intraoperative and immediate postoperative deaths were excluded from percent survival calculations. Two group I animals died of intraoperative device failures and one group II animal died immediately postoperatively.
Statistical Analysis Data are presented as mean ⫾ standard error of the mean. A t test was used to compare all group I and group II parameters. A p value of 0.05 or less was considered indicative of a statistically significant difference.
Study Endpoints Survival days, CPB time, and anesthesia time were compared between group I and group II calves. Mean hematologic and biochemical data were also compared between the two groups. These data represented mean values measured during the first postoperative week after ETAH implantation. Mean ETAH pump flow, central venous pressure, Pco2, Po2/Fio2 ratio, respiratory rate, hematocrit, plasma hemaglobin, leukocyte count, total protein (percent baseline), albumin (percent baseline), serum glutamic oxaloacetate transaminase, and creatinine were compared. Cumulative water balance on
Results Anesthesia and CPB Times Total CPB and anesthesia times were not statistically different between groups I and II (Table 1).
System Parameters and Hemodynamics of ETAH Mean ETAH pump flow was not statistically significantly different between groups I and II (Table 1). Likewise, central venous pressure was not significantly different between groups.
Ann Thorac Surg 2001;71:S150 –5
CIRCULATORY SUPPORT HOENICKE ET AL PENN STATE ELECTRIC TOTAL ARTIFICIAL HEART
Table 1. Mean Hemodynamic and Hematologic Parameters in Calves Surgically Implanted With the Penn State Electric Total Artificial Heart Endpoint Survival times (days) CPB time (min) Anesthesia time (min) Flow (L/min) CVP (mm Hg) (⫹) Water balance (L) Respiratory rate (respirations/min) Pco2 (mm Hg) Po2/Fio2 ratio Transfused blood (L) Leukocyte count (⫻ 103/mm3) Hematocrit (%) Plasma hemaglobin (mg%) Total protein (% baseline) Albumin (% baseline) SGOT (IU) Creatinine (mg/dL)
Group I (n ⫽ 7)
Group II (n ⫽ 6)
p Value
25.6 ⫾ 16.3 65.0 ⫾ 14.0 142.2 ⫾ 6.6 135.5 ⫾ 5.4 370.5 ⫾ 15.1 375.8 ⫾ 29.5 6.3 ⫾ 0.7 7.2 ⫾ 0.2 7.9 ⫾ 1.8 9.2 ⫾ 0.9 5.7 ⫾ 3.9 11.7 ⫾ 3.3 32.2 ⫾ 2.4 30.2 ⫾ 1.6
0.10 0.45 0.88 0.20 0.50 0.27 0.51
48.2 ⫾ 2.9 46.4 ⫾ 2.3 336.3 ⫾ 35.4 419.4 ⫾ 17.5 8.0 ⫾ 2.0 7.5 ⫾ 2.5 17.3 ⫾ 5.5 13.3 ⫾ 3.4 27.4 ⫾ 1.7 23.9 ⫾ 1.8 5.1 ⫾ 1.5 4.1 ⫾ 0.5 64.5 ⫾ 3.1 63.8 ⫾ 2.5 67.2 ⫾ 6.5 63.3 ⫾ 3.1 228.4 ⫾ 38.5 220.6 ⫾ 55.5 0.9 ⫾ 0.1 0.9 ⫾ 0.1
0.63 0.05 0.89 0.55 0.20 0.55 0.87 0.56 0.91 0.83
All values represent mean ⫾ standard error of the mean during the first postoperative week. CPB ⫽ cardiopulmonary bypass; CVP ⫽ central venous pressure; Pco2 ⫽ arterial partial pressure of carbon dioxide; Po2/Fio2 ratio ⫽ arterial pressure of oxygen-to-fraction of inspired oxygen; SGOT ⫽ serum glutamic oxaloacetic transaminase.
Hematologic and Biochemical Endpoints Mean values for postoperative days 0 to 7 were not statistically significantly different between groups with regard to leukocyte count, hematocrit, plasma hemaglobin, total protein, albumin, serumglutamic oxaloacetic transaminase (SGOT), creatinine, and arterial Pco2 (Table 1). The Po2 to Fio2 ratio was statistically significantly improved in group II compared to group I ( p ⫽ 0.05) (Table 1).
Water Balance, Urine Output, and Drain Output Table 2 shows that there was no statistically significant differences between positive water balance, total urine output, mean daily urine output, and volume of transfused blood in the first postoperative week. There were no significant differences observed in total drain output for left pleural tube, right pleural tube, right pump tube, and right flank JP drain (Table 2). Preoperative weights were also not significantly different between groups (Table 2).
S153
Table 2. Mean Water Balance, Surgical Drainage Output, and Volume of Transfused Blood in Calves Surgically Implanted with the Penn State Electric Total Artificial Heart Endpoint Preoperative weight (kg) (⫹) Water balance (L) Total urine total (L) Daily urine output (L) Total left pleural tube (L) Total right pleural tube (L) Total right pump tube (L) Total right flank JP drain (L) Transfused blood (L)
Group I (n ⫽ 7)
Group II (n ⫽ 6)
p Value
90.9 ⫾ 1.0 5.7 ⫾ 3.9 22.5 ⫾ 7.5 5.9 ⫾ 1.3 2.8 ⫾ 1.4 1.2 ⫾ 0.6 2.4 ⫾ 1.0 1.0 ⫾ 0.7 8.0 ⫾ 2.0
94.8 ⫾ 2.6 11.7 ⫾ 3.3 33.2 ⫾ 7.4 5.6 ⫾ 0.5 2.7 ⫾ 0.7 0.9 ⫾ 0.3 2.5 ⫾ 0.7 0.4 ⫾ 0.1 7.5 ⫾ 2.5
0.19 0.27 0.34 0.83 0.92 0.71 0.90 0.40 0.89
All values represent mean ⫾ standard error of the mean during the first postoperative week. JP drain ⫽ Jackson-Pratt drain.
additional calf died because of pulmonary failure of unknown cause. One calf survived in group I until scheduled termination at 12 weeks. Group II deaths resulted from mediastinal hematoma resulting in an acute superior vena cava syndrome and subsequent massive cerebrovascular accident (6 hours postoperatively), gastrointestinal bleed (19 days postoperatively), and device failure (46 days postoperatively) (Fig 3). The deaths of the remaining three calves were 12-week scheduled terminations (Fig 3). Excluding intraoperative and immediate postoperative deaths, the 2-week survival was 40% in group I and 100% in group II (Fig 4). One-month survival was 40% in group I and 80% in group II. The percentage of animals surviving to the 12-week scheduled termination was 20% in group I and 60% in group II.
Comment Circulatory support devices designed to augment or to completely replace hearts in patients with end-stage heart disease have been under development at Penn State over the past three decades [1–3]. Developmental and engineering milestones were initially achieved with the Thoratec (Pierce-Donachy) ventricular assist device as treatment for severe cardiac stunning during cardiopulmonary bypass and, later, as a bridge to transplanta-
Causes of Death and Survival Group II animals survived longer than those in group I (65.0 ⫾ 14.0 days vs 25.6 ⫾ 16.3 days, respectively); however this was not statistically significant ( p ⫽ 0.10) (Table 1). Figure 3 compares the causes of death in both groups I and II. Two calves in group I died on-table from device failures. Three group I calves died of massive hemothorax on postoperative days 1, 6, and 34. An
Fig 3. Causes of deaths in calves surgically implanted with the Penn State Electric Total Artificial Heart.
S154
CIRCULATORY SUPPORT HOENICKE ET AL PENN STATE ELECTRIC TOTAL ARTIFICIAL HEART
Fig 4. Comparison of calf survival between group I and group II 2, 4, and 12 weeks postoperatively.
tion. More recently, the Arrow–Penn State Lionheart II LVD (Arrow, Reading, PA) has been introduced with success as a permanent left ventricular assist device. The production of a totally implantable artificial heart has proved to be a challenging endeavor, which has resulted in the design and fabrication of the Penn State Electric Total Artificial Heart (ETAH) at our institution. The Penn State ETAH is currently undergoing both in vitro and in vivo preclinical evaluation. The long-term survival of animals undergoing surgical implantation of total artificial hearts has been limited by various postoperative complications. Earlier reports of the survival of calves receiving ETAH implantation focused on deaths related to pulmonary complications [5]. A significant number of postoperative deaths were attributable to pulmonary complications. It was demonstrated that longer cardiopulmonary bypass times were associated with decreased 2-week survival. Kuroda and colleagues [5] argued that these early postoperative deaths were produced by a syndrome similar to adult respiratory distress syndrome (ARDS). This clinical entity was believed to be a product of harmful complement activation associated with long (3 hours or more) cardiopulmonary bypass times. Because Kuroda and colleagues [5] demonstrated an association between shorter cardiopulmonary bypass (CPB) times and improved survival, [5] subsequent experiments focused on minimizing CPB times. Compared to long-term survivors in the study of Kuroda and colleagues [5], CPB times for all calves in the present study were significantly decreased (156.0 ⫾ 6.4 minutes vs 138.8 ⫾ 4.2 minutes, p ⫽ 0.032). With the shorter CPB times in the current study, the number of postoperative ARDS-like deaths was decreased compared to earlier reports [5]. Despite these shorter CPB times, overall postoperative survival between 1998 and 1999 was low. To improve the results of these calf studies, our group modified the surgical implantation protocol. To decrease further the likelihood of postoperative pulmonary complications, all efforts to skeletonize and retract the phrenic nerve were minimized. Our group reasoned that
Ann Thorac Surg 2001;71:S150 –5
dissection of the right phrenic nerve would add to CPB time and potentially result in ischemic phrenic nerve damage. The importance of minimizing traction and dissection of the phrenic nerve has been previously described by Lick and colleagues [8] in a simplified technique of heart–lung transplantation. Vascular or traction injury to the phrenic nerve may result in significant perturbations in pulmonary function [9]. The second surgical modification was the introduction of fibrin glue to reinforce right and left atrial, pulmonary artery, and aortic anastomoses. Our group reasoned that CPB time could be minimized by reducing the time repairing anastomotic bleeding. Moreover, the risks of postoperative hemothorax would be minimized. The comparison of hemodynamic and laboratory values of group I calves (before modifications) to group II calves (after modifications) did not yield many differences. The avoidance of phrenic nerve dissection and fibrin glue application did not reduce the time spent on CPB, nor did these modifications augment postoperative flow or central venous pressure. Water balance and laboratory values indicative of infection, bleeding, hemolysis, liver function, and renal function were also not different between groups. Our results did indicate a significant difference between oxygenation in the two groups. Oxygenation was measured by comparing the ratio of arterial Po2 to fraction of inspired oxygen (P:F ratios). Animals that were spared phrenic nerve dissection displayed Po2/Fio2 ratios that were virtually normal, and animals that underwent phrenic nerve dissection exhibited lower ratios. Although this difference could be attributable to spared postoperative right phrenic nerve function, a difference in ventilation between the two groups would also have been anticipated with impaired ventilation in group I compared to group II. The results show that ventilation was not improved in group II compared to group I because the Pco2 values were statistically equivalent. If postoperative tidal volumes were pathologically reduced because of phrenic nerve injury, an increase in respiratory rate would also be anticipated; however, respiratory rate was also equivalent between groups. Hematological differences attributable to improved postoperative hemostasis from fibrin glue application were also not observed in the results. Improved hemostasis may have been confirmed by decreased volume of postoperative blood transfusions, decreased thoracostomy tube drainage, and higher hematocrit values. Comparison of these parameters between group I and group II animals yielded no statistically significant differences. Occult hemothoraces resulting from nonfunctioning thoracostomy tubes or hemothoraces occurring after removal of thoracostomy tubes may manifest clinically as hypoventilation syndromes. However, as previously mentioned, no significant differences in arterial Pco2 and respiratory rates were observed between groups. Caution should be applied postulating a causative link between the two surgical modifications and improved Po2/Fio2 ratios. The small numbers involved in the current study do not allow for speculation that right phrenic
Ann Thorac Surg 2001;71:S150 –5
nerve sparing or fibrin glue–augmented hemostasis resulted in improved oxygenation. Because multiple variations in surgical protocol were introduced in the present study, the causal relationship between each modification and improved oxygen ratios is not possible. However, as there was a trend of increased survival days in group II compared to group I (65 ⫾ 14 days vs 26 ⫾ 16 days, respectively; p ⫽ 0.10), the Po2/Fio2 ratio may serve as a positive prognostic indicator for improved long-term survival. Despite the small number of laboratory and hemodynamic differences observed between group I and group II animals, comparison between causes of death in groups I and II revealed many interesting trends. In group I, one calf died on postoperative day 1 of pulmonary complications. Figure 1 shows this death to be related to idiopathic pulmonary failure. This animal showed signs of possible phrenic nerve damage, as its mean Pco2 was 57.4 with a mean respiratory rate of 37 and mean P:F ratio of 208, and eventually succumbed to respiratory failure and death. Although this pattern of death approximates the ARDSlike syndrome described by Kuroda and colleagues [5], this animal was not exposed to a prolonged CPB time (CPB time, 129 minutes). No animals in group II showed this pattern of death. The major cause of death in group I calves was massive hemothorax. Three of 7 calves died of massive hemothorax in group I. Two of these deaths were early postoperative deaths (postoperative days 1 and 6) and the other occurred much later (postoperative day 34). No animals in group II died of massive hemothorax. We may speculate that the use of fibrin glue in Group II decreased the frequency of irreversible massive hemothorax and subsequent death; however, the hematologic and thoracostomy drainage data did not reveal any substantiating differences. Overall, calves in group II showed overall improved survival at 2, 4, and 12 weeks after surgical implantation of the Penn State ETAH. Although the improved results do seem to correlate with the two surgical modifications, it is difficult to explain these improved results given the lack of differences observed in hemodynamic, laboratory, and intrathoracic drain data. It is unclear whether improved early postoperative oxygenation was related to the surgical modifications. It is also unknown whether the improved Po2/Fio2 ratios underlie improved survival. These studies are ongoing and, as more ETAHs are implanted, differences in these endpoints may be elaborated. A larger number of studies will need to be performed to address these issues.
CIRCULATORY SUPPORT HOENICKE ET AL PENN STATE ELECTRIC TOTAL ARTIFICIAL HEART
S155
Conclusions Two modifications in the surgical protocol of implantation of the Penn State Electric Total Artificial Heart were initiated to promote improved long-term survival in calves. Both minimizing phrenic nerve dissection and applying fibrin glue correlated with improved survival of calves implanted with the ETAH. Previous deaths related to hypoventilation and massive hemothorax were prevented in calves that underwent the modified surgical protocol; however, analysis of hemodynamic and hematologic variables did not fully explain these survival differences. Early postoperative oxygenation was improved in the calves receiving the modified surgical protocol. It is not possible to ascertain whether improved oxygenation accounts for the differences in overall calf survival.
Supported by USPHS contract 1 HV-38130 and USPHS grant 1 RO1 HL 60276-01A1.
References 1. Pierce WS, Sapirstein JS, Pae WE Jr. Total artificial heart: from bridge to tranplantation to permanent use. Ann Thorac Surg 1996;61:342– 6. 2. Weiss WJ, Rosenberg G, Snyder AJ, et al. Steady state hemodynamic and energetic characterization of the Penn State/3M Health Care Total Artificial Heart. ASAIO Journal 1999;45:189–93. 3. Pierce WS, Rosenberg G, Snyder AJ, Pae WE, Waldhausen JA. An electric artificial heart for clinical use. Ann Surg 1990;212: 339– 43. 4. Honda T, Kito Y, Gibson W, Nemoto T, Cockrell JV, Akutsu T. Lung problem and liver damage in total artificial heart and their prevention. ASAIO Trans 1974;20:27–32. 5. Kuroda H, Rosenberg G, Snyder AJ, et al. Postoperative pulmonary complications in calves after implantation of an electric total artificial heart. ASAIO J 1998;44:M613– 8. 6. Weiss WJ, Rosenberg G, Snyder AJ, et al. Recent improvements in a completely implanted total artificial heart. ASAIO J 1996;42:M342– 6. 7. Rosenberg G, Snyder AJ, Weiss WJ, Sapirstein JS, Pierce WS. In vivo testing of a clinical-size totally implantable artificial heart. In: Unger F, ed. Assisted circulation 4. Berlin: SpringerVerlag, 1995, 236 – 48. 8. Lick SD, Copeland JG, Rosado LJ, Arabia FA, Sethi GK. Simplified technique of heart-lung transplantation. Ann Thorac Surg 1995;59:1592–3. 9. Abd AG, Braun NMT, Baskin MI, O’Sullivan MM, Alkaitis DA. Diaphragmatic dysfunction after open heart surgery: treatment with a rocking bed. Ann Intern Med 1989;111: 881– 6.