Reduction in Bleeding after Heart-Lung Transplantation

Reduction in Bleeding after Heart-Lung Transplantation*

The Importance of Posterior Mediastinal Hemostasis IUchard]. Novick, M.D., MSc., F.C.C.P.; Alan H. Menkis, M.D., F.C .C.P.; F. Neil McKenzie, M.D., F.C.C.P.; Ken R. Reid, M.D .; liner W Pflugfelder, M.D.; William]. Kostuk, M.D.; and Dildar Ahmad, M.D., F.C.C.P.

To reduce perioperative hemorrhage following heart-lung transplantation, several technical modifications were introduced in June 1988 to secure better posterior mediastinal hemostasis. The intraoperative and postoperative use of blood and blood products, as well as the chest tube drainage in the first 24 hours postoperatively, were compared in the seven patients operated on since June 1988 with the nine patients operated on before that date. Significant (p<0.05) reductions were demonstrated in the intraoperative and postoperative transfusion of packed cells, in the postoperative administration of fresh frozen plasma, and in the

technique of heart-lung transplantation (HLT) T ishe well standardized.• Attention has since been

directed to the proper selection of donor and recipients for this procedure2-4 and, most recently, to improving the long-term functional results. 4-11 A review of published series of HLT, however, reveals a hospital mortality rate of up to 42 percent,9 although one-year survival rates of 85 percent are being achieved currently in experienced centers. 10 Not infrequently, bleeding has been a major cause of early postoperative morbidity and mortality, 24 •9 •11 • 15 and has been the initiating factor in a deteriorating clinical course that results in the need for multiple transfusions, deterioration in allograft function, multiorgan failure, and death. We have performed 16 HLTs since May 1983, with one-month and total hospital mortality rates of 6 percent and 12.5 percent, and one- and two-year actuarial survival rates of 87 ± 8 percent and 69 ± 13 percent, respectively. In June 1988, we initiated several technical modifications to secure better posterior mediastinal hemostasis intraoperatively. These changes have resulted in a marked decline in postoperative chest tube drainage and the need for perioperative transfusions and form the basis of this report.

*From the Divisions of CardiovascuJar..Thoracic Surgery, Cardiology, and Respiratory Medicine, University Hospital, London, Ontario, Canada. Manuscript received February 20; revision accepted June 6. Reprint requests: Dr. Novick, Division of Chest Surgery, PO Bor 5339, University Hospital, London, Ontario, Canada N6A 5A5

chest tube drainage within the first 24 hours postoperatively. The one-month and total hospital mortality rates were 6 percent and 12.5 percent, respectively. It is concluded that newer techniques to obtain optimal posterior mediastinal hemostasis have significantly reduced blood loss following heart-lung transplantation in our experience and have contributed to our excellent early postoperative results. (Chest 1990; 98:1383-87) HLT =heart-lung transplantation

PATIENTS AND METHODS

The 16 patients undergoing HLT all had advanced cardiorespiratory failure and were oxygen dependent. The recipients were 12 female and four male subjects, ranging in age from 9 to 50 years (mean, 31±3 years). The preoperative diagnoses of the HLT recipients are shown in Thble 1. Only one patient, operated on in November 1989, had undergone a previous thoracotomy; none had had a previous sternotomy. The technique of the donor heart-lung procurement was identical to that described by Jamieson et al,' except that an iso-oncotic 4°C solution of 5 percent albumin and Ringer's lactate solution was administered as a pulmonary 8ush, instead of modi6ed Collins solution. After the administration of cardioplegia and the pulmonary 8ush solution, the heart-lung block was excised, dividing the trachea at least five rings above the carina. The organs were then placed in a basin containing sterile saline solution at 4 to 8°C and were transported to the recipient operating room. The recipient operation was performed in a manner similar to that described in the above-mentioned report.' Accessible pleural adhesions were divided with the electrocautery prior to heparinization and cannulation for cardiopulmonary bypass. After separate excision of the heart and both lungs, and removal of the left atrial,

Table !-Preoperative DitJgnosea of Heart-Lung '&ansplont Recipienta Diagnosis Group A (May 1983-June 1988) Primary pulmonary hypertension Eisenmenger's syndrome Fibrosing alveolitis Pulmonary thromboemboli Immotile cilia syndrome Group B Oune 1988-November 1989) Primary pulmonary hypertension Fibrosing alveolitis, right ventricular infarct

CHEST I 98 I 6 I DECEMBER, 1990

No.

4 2

1 1 1

6 1

1383

-- ---------- --- LEFT IAIIGUS NERVE

SUPERIOR VENA CAVA --------- -

RIGHT MAfl SA0NCHUS --------- - 1:.,._,.~-I
POSTERIOR· ·------PERICARDIUM

pulmonary artery, and main bronchial remnants, hemostasis was secured in the posterior mediastinum. Prior to June 1988, in nine patients (group A), this was accomplished by the use of vascular clips and electrocautery, with care taken not to traumatize the vagus nerves. After June 1988, in the last seven transplants (group B), several technical modifications were implemented. First, as much posterior pericardium as possible was left in the recipient. Second, the mediastinal parietal pleura was sutured to the lateral edge of the pericardium on each side using a running 4--0 polyplbpylene suture (Fig 1). Third, in addition to the use of electrocatHery and vascular clips in the superior aspect of the posterior mediastinum, further hemostasis was secured by taking large bites of the posterior mediastinal tissues using a running 3--0 polydioxanone suture while performing the posterior wall of the tracheal anastomosis. The anterior wall of the tracheal anastomosis was performed using interrupted 3--0 polydioxanone sutures. After completion of the tracheal anastomosis, the right atrial anastomosis was performed using running 3--0 polypropylene, incorporating the entire interatrial septum, including any left atrial remnants, in the suture line. Finally, the aortic anastomosis was completed using continuous 4--0 polypropylene. Following de-airing of the heart, the vent site in the left atrial appendage and the hole for the pulmonary Hush catheter in the pulmonary artery were closed. After the successful discon· tinuation of cardiopulmonary bypass, protamine sulfate was admin·

1384

FIGURE 1a, (upper). Appearance of the chest following removal of the recipient's heart and both lungs. -· LEFT PARIETAL As much posterior pericardium as possible is left in PLEURA situ. Both main bronchi will be subsequently excised and the trachea trimmed to one ring above the carina. FIGURE 1b, (lower) Using a running 4-0 polypropylene suture, the posterior mediastinal pleura is sutured to the posterior pericardium from the diaphragmatic surface to the cephalad aspect of the pericardial reflection on each side. Subsequently, large bites of the posterior mediastinal tissues will be incorporated in the posterior aspect of the tracheal anastomosis to obtain better hemostasis in that region. istered in a dose necessary to return the activated clotting time to its prebypass controL Blood and blood products were administered to maintain a hemoglobin greater than 90 giL. Angled tubes were inserted into both pleural spaces, and angled and straight mediastinal tubes were also placed. After surgery, the patients were monitored in the intensive care unit, and then transferred to a specialized multiorgan transplant unit within 24 hours of extubation. Packed cells were administered to maintain a hemoglobin greater than 90 giL, and , if the total chest tube drainage was in excess of 150 mllh, fresh frozen plasma and platelets were infused to normalize the coagulation status. Pleural and mediastinal tubes were removed when they were no longer draining greater than 10 ml per tube per hour over a 12-hour period. Further details of the postoperative care, immunosuppression, and serial bronchoscopic follow-up of our patients are described else· where.•• Statistical Analysis

Numeric data are expressed as mean± SEM _ Actuarial survival ligures and their SEs were calculated according to the method of Cutler and Ederer. 17 The statistical significance of differences between the two groups of patients was determined by the twotailed t testReduction in Bleeding after Heart-Lung Transplantation (Novick et al)

Table 2-Preoperative lbtient Data• Group A May 1983June 1988 Age, yr Weight, kg Hemoglobin, giL Prothrombin time, s Partial thromboplastin time, s

32::!:3 59::!:7 155::!:9 14::!: 1.2 36::!:2.4

Group B June 1988November 1989 NS NS NS NS NS

30::!:5 55::!:6 157::!:8 13::!:0.7 33::!: 1.6

*::!: =standard error of the mean. NS = p>O.OS between adjacent variables. RESULTS

Of the 16 HLT recipients, one died within the first 30 days postoperatively, due to respiratory failure at 36 hours. Another patient developed of a false aneurysm at the aortic anastomosis as well as cytomegalovirus pneumonia, and died five weeks postoperatively. The 30-day operative and total hospital mortality rates were, therefore, 6 percent and 12.5 percent, respectively. Eleven of the HLT recipients are still living, and our current one- and two-year actuarial survival rates are 87 ± 8 percent and 69 ± 13 percent, respectively. There were no significant differences in preoperative patient variables, including hemoglobin, prothrombin time, and partial thromboplastin time, between the group A and group B patients (Table 2). In addition, there were no significant differences in preoperative renal and hepatic function between the two groups. The warm ischemic times of the heartlung graft were similar in groups A and B, but because of the more recent practice of distant donor procurement, the cold and total ischemic times were significantly longer in group B (Table 3). The time on cardiopulmonary bypass was slightly more prolonged in group A patients, reflecting the additional effort required to obtain posterior mediastinal hemostasis prior to the discontinuation of cardiopulmonary bypass. The intraoperative and postoperative transfusion of blood and blood products was often considerable and highly variable in group A patients but was markedly reduced in the seven most recently operated-on patients (Table 4). The combined mediastinal and chest

Table 4-Units of Packed Cella, Freah Frozen Pl&ma (FFP), and Platelet& (PLTS) 'Iranafiued Intraoperatively and Postoperaticely• Intraoperative Packed Cells

Cold ischemic time, min Total ischemic time, min Pump time , min

40::!:8 101::!: 11 248::!: 15

Group B June 1988November 1989 X X X

93::!:20 153::!: 15 201::!: 11

*::!: =standard error of the mean. X= p~O.OS between adjacent variables.

PLTS

Packed Cells

FFP PLTS

21.3::!: 10 20::!:9 17::!:7 16.6::!:5 9::!:3 14::!:8 Group A (May 1983X NS X NS X X June 1988) 1.7::!:0.6 5.6::!: 1 3.4::!: 1 3::!:0.5 1.5::!: I 0::!:0 GroupB Oune 1988November 1989) *::!: =standard error of the mean. NS = p>O.OS between adja(:ent variables. X=p~O.OS between adjacent variables.

tube drain~ge at 24 hours postoperatively was significantly less in group B, although the difference in reexploration rates for hemorrhage between the two groups did not reach statistical significance (Table 5). In our last three HLT recipients, two did not require any intraoperative transfusions, and the third received only a single unit of packed cells during the procedure. DISCUSSION

Heart-lung transplantation is nearing the end of its first decade of clinical application. Because the number of satisfactory heart-lung grafts is limited, only a few centers in North America have acquired a significant experience and hence technical facility with the procedure. 3 •9 Intraoperative hemorrhage and postoperative hemorrhage from pleural adhesions and the posterior mediastinum have been problematic in the early experience with HLT, 2-u.n- 15 and with infection have been a major cause of morbidity and m'ortality postoperatively. . Several reports emphasize the importance of proper recipient selection in improving the results of HLT. 2 •3 •9 •18 In the early Pittsburgh experience, 9 no patient with a previous sternotomy survived, and those who had undergone previous thoracotomy had Table 5-Mediaatinal and Pleural Tube Drainage at 24 Houn and the &operation Hate for Postoperative Hemorrhage•

Table 3-Intraoperative Data• Group A May 1983June 1988

FFP

Postoperative

Group A (May 1983June 1988) GroupB Oune 1988November 1989)

Chest Tube Drainage at 24 Hours, L

Reoperation for Bleeding

4.3::!: 1.4

319

X

NS

1.4 ::!:0.3

017

*::!: =standard error of the mean. NS = p>O.OS between adjacent variables. X= p:S0.05 between adjacent variables. CHEST I 98 I 6 I DECEMBER, 1990

1385

development of postoperative hemorrhage from pleural and mediastinal adhesions. In the Stanford experience, as well, the highest risk of operative mortality occurred in patients with a previous thoracic operation, pleural adhesions, or impaired hepatic function. 2 These factors were strongly correlated with increased perioperative hemorrhage and poor postoperative allograft function due to multiple transfusions. Like the Stanford group, 3 we no longer consider a previous thoracotomy to be a contraindication for HLT, but we review these patients carefully on a case-by-case basis. Our most recent HLT recipient had undergone a previous left thoracotomy and had minimal blood loss perioperatively. Although increasing operative experience in HLT may play a role in improved results in our group B patients, this series represents the experience of several surgeons, each of whom noted a marked reduction in perioperative hemorrhage when the above-described techniques for the reduction of mediastinal bleeding were adopted. Similarly, although patients with Eisenmenger's syndrome may be considered to be at higher risk for hemorrhage due to more prominent posterior mediastinal collaterals, the two patients with this condition in our series had a mean perioperative blood loss and transfusion requirement that was not significantly different from others in group A. A description of the standard technique of HLT mentions the importance of securing the posterior mediastinal vessels with clips during resection of the recipient's heart and lungs. 1 In addition, the importance of including the entire interatrial septum, including all left atrial remnants, in the right atrial suture line has been emphasized.' The Papworth group reinforced the tracheal anastomosis with remnants of the recipient's pulmonary artery in their early experience; currently, as is our practice, they incorporate large bites of the posterior mediastinal tissue in the posterior aspect of the tracheal anastomosis to minimize the possibility of an air leak and to secure better hemostasis in the superior aspect of the posterior mediastinum.4 The technique of HLT has been modified by a number of surgeons to diminish the risk of uncontrollable perioperative hemorrhage. In a retrospective case report, Losman 19 advocated leaving the recipient pulmonary artery bifurcation and posterior wall of the left atrium in place and oversewing the orifices of the pulmonary veins. In addition, he referred to the practice of other surgeons in leaving the posterior wall of the tracheal carina and main bronchi in situ. A more innovative approach to minimizing technical complications following HLT has been reported by Vouhe and DartevelleOO in 1989. The authors emphasized the importance of a limited dissection of the posterior mediastinum, exposing only the distal tra1318

chea and both bronchi. They performed extrapericardial pneumonectomies and used surgical stapling as much as possible in dividing the left atrial wall and in creating appropriate-sized passages for the donor lungs. They made a routine practice of leaving the posterior wall of the left atrium, the pulmonary artery bifurcation, and the transpericardial aspect of the pulmonary pedicles in situ. Prior to implantation of the donor heart-lung graft, fibrin glue was placed over the posterior mediastinum to achieve optimal hemostasis. Using this approach in 21 patients, the authors recorded a mean blood loss during the first 24 postoperative hours of 800 ± 350 mi. No patient in their experience or in our most recently operated-on group required reoperation for excessive postoperative hemorrhage. In summary, bleeding has been a major complication of HLT because of the presence of large posterior mediastinal collaterals, coagulopathy from hepatic congestion due to right-sided heart failure, adhesions from previous thoracic operations, and the long car~ diopulmonary bypass times that are required. The techniques that we have adopted should decrease the risk of hemorrhagic complications and the risks associated with multiple transfusions after HLT, and thus improve the early outcome after this procedure. ACKNOWLEDGMENT: The authors thank Anne-Marie Powell for her assistance in chart review and Heather Motloch for her expertise in typing and preparing this manuscript.

REFERENCES 1 Jamieson SW, Stinson EB, Oyer PE, Baldwin JC, Shumway NE. Operative technique for heart-lung transplantation. J Thorac Cardiovasc Surg 1984; 87:930-35 2 Harjula A, Baldwin JC, Starnes VA, Stinson EB, Oyer PE , Jamieson SW, et al. Proper donor selection for heart-lung transplantation: the Stanford experience. J Thorac Cardiovasc Surg 1987; 94:874-80 3 Starnes VA, Baldwin JC, Harjula A. Combined heart and lung transplantation: the Stanford experience. J Appl Cardiol1987; 2:71-89 4 Hutter JA, Despins P, Higenbottam T, Stewart S, Wallwork J. Heart-lung transplantation: better use of resources. Am J Med 1988; 85:4-11 5 Dawkins KD , Jamieson SW, Hunt SA, Baldwin JC, Burke CM , Morris A, et al. Long-term results, hemodynamics, and complications after combined heart and lung transplantation. Circulation 1985; 71:919-26 6 Theodore J, Morris AJ, Burke CM, Glanville AR, VanKessel A, Baldwin JC, et al. Cardiopulmonary function at maximum tolerable constant work rate exercise following human heartlung transplantation. Chest 1987; 92:433-39 7 Glanville AR, Baldwin JC, Burke CM , Theodore J, Robin ED. Obliterative bronchiolitis after heart-lung transplantation: apparent arrest by augmented immunosuppression. Ann Intern Med 1987; 107:30(H)4 8 Starnes VA, Theodore J, Oyer PE, Billingham ME, Sibley RK, Berry G, et al. Evaluation of heart-lung transplant recipients with prospective, serial transbronchial biopsies and pulmonary function studies. J Thorac Cardiovasc Surg 1989; 98:683-90 9 Griffith BP, Hardesty RL. Trento A, Paradis IL, Duquesnoy RJ, Reduction in Bleeding after Heart-Lung Transplantation (Novick et aJ)

10

11 12

13 14

Zeevi A, et al. Heart-lung transplantation: lessons learned and future hopes. Ann Thorac Surg 1987; 43:6-16 Starnes VA, Theodore J, Oyer PE, Stinson EB, Moreno-Cabral CE, Sibley R, et al. Pulmonary infiltrates after heart-lung transplantation: evaluation by serial transbronchial biopsies. J Thorac Cardiovasc Surg 1989; 98:945-50 Jamieson SW, Baldwin J, Stinson EB, Reitz BA, Oyer PE, Hunt S, et al. Clinical heart-lung transplantation. Transplantation 1984; 37:81-4 McGregor CGA, Jamieson SW, Baldwin JC, Burke CM, Dawkins KD, Stinson EB, et al. Combined heart-lung transplantation for end-stage Eisenmenger's syndrome. JThorac Cardiovasc Surg 1986; 91:443-50 Copeland JG. Heart-lung transplantation: current status. Ann Thorac Surg 1987; 43:2-3 Diethrich EB, Bahadir I, Gordon M, Maki P, Warner MG, Clark R, et al. Postoperative complications necessitating right lower lobectomy in a heart-lung transplant recipient with

previous sternotomy. J Thorac Cardiovasc Surg 1987; 94:389-92 15 'Thzelaar HD, Yousem SA. The pathology of combined heartlung transplantation: an autopsy study. Hum Pathol 1988; 19:1403-16 16 Novick RJ, Menkis AH, McKenzie FN, Reid KR, Ahmad D. Should heart-lung transplant donors and recipients be matched according to cytomegalovirus serology? J Heart Transplant (in press) 17 Cutler SJ, Ederer F. Maximum utilization of the life table method in analyzing survival. J Chronic Dis 1958; 8:699-711 18 Zenati M, Dowling RD, Armitage JM, Kormos RL, Dummer JS, Hardesty RL, et al. Organ procurement for pulmonary transplantation. Ann Thorac Surg 1989; 48:882-86 19 Losman JG. Heart-lung transplantation. J Thorac Cardiovasc Surg 1985; 89:807-08 20 Vouhe PR, Dartevelle PG. Heart-lung transplantation: technical modifications that may improve the early outcome. J Thorac Cardiovasc Surg 1989; 97:906-10

Radiology Courses The Department of Radiology, University of California San Diego School of Medicine, will present the following postgraduate courses: Intraluminal Ultrasound, February 6, Le Meridien Hotel, Coronary (San Diego); Basic Introduction to Doppler Principles, February 6, Le Meridien Hotel, Coronado {San Diego); Duplex Imaging Thtorial Conftmmce and Exhibit, February 7-9. For information, contact Ryals and Associates, PO Box 1925, Roswell, Georgia 30077-1925 {404:641-9773).

CHEST I 98 I 6 I DECEMBER, 1990

1387