Ann Thorac Surg 1997;63:117-123
© 1997 The Society of Thoracic Surgeons
Original Articles: Cardiovascular
Previous Open Heart Operation: A Contribution to Impaired Outcome After Cardiac Transplantation?
Kay Uthoff, MD,
Thorsten Wahlers, MD,
Joachim Cremer, MD,
Hans-Georg Borst, MD
Division of Thoracic and Cardiovascular Surgery, Surgical Department, Hannover Medical School, Hannover, Germany
Accepted for publication July 19, 1996.
 |
Abstract
|
|---|
Background. There is still debate about whether previous cardiac operations are a risk factor for patient outcome after cardiac transplantation. As waiting lists for cardiac transplantation increase, adverse outcome criteria should be identified.
Methods. To assess this problem, we retrospectively analyzed 53 patients with previous cardiac operations before heart transplantation and compared them with 53 control patients matched for sex and age. Patient groups were analyzed regarding their preoperative, intraoperative, and postoperative variables and survival.
Results. Ischemic times were comparable in both groups, but the duration of the operation was significantly longer in the study group (206.5 ± 62.5 minutes, versus 156.0 ± 36.7 minutes in controls; p < 0.05). In addition, postoperative blood loss was greater for the patients with previous cardiac operations (1,360 ± 260 mL, versus 730 ± 310 mL for controls; p < 0.01). Postoperatively, the rate of rejection episodes and the incidence of graft atherosclerosis were comparable within the first 2 years. However, survival was significantly reduced in the study group (60.1%) after 4 years (versus 83.1% for controls; p < 0.05).
Conclusions. Heart transplantation in patients with previous cardiac operations will lead to an impaired overall outcome. In addition, these patients have more postoperative complications.
|
Introduction
|
|---|
The favorable long-term results of cardiac transplantation have widened the indications for this operation [1–5]. The recipient population has increased by including older and younger age groups, as well as patients who have had previous cardiac operations. However, studies have suggested that patients with previous cardiac operations have a poorer outcome [3, 6]. In particular, previous blood-product transfusion may lead to increased allogenic antibody formation, which may produce a more severe rejection process. Scar formation from a previous operation requiring extended adhesiolysis also may contribute to increased perioperative bleeding. Prolonged operation times may lead to an increased inflammatory response caused by longer cardiopulmonary bypass times. This might result in an increased postoperative blood loss for operative or hemostatic reasons. An adequate supply of screened blood products improves organizational efforts. These immunologic and hemostatic risk factors may prolong the postoperative intensive care course and hospitalization.
Therefore, in this retrospective study, we evaluated a history of cardiac operation as a risk factor for a prolonged perioperative course and impaired survival after cardiac transplantation.
|
Material and Methods
|
|---|
Patients
The study was retrospective and investigated the perioperative and postoperative risk factors and survival of the study group as compared with the control group. All patients underwent orthotopic heart transplantation at the Medical School Hannover between 1985 and 1991. The study group consisted of 53 patients with a previous cardiac operation (pCO) before transplantation. Forty-eight patients were male and 5 were female. The mean age was 49.2 ± 7.5 years (Table 1
). Fifteen patients had dilated cardiomyopathy or cardiomyopathic courses after a valve replacement; 38 patients had ischemic heart disease. Previous cardiac operations included coronary artery bypass grafting (CABG) in 23 patients, CABG and aneurysmectomy in 14, and various other procedures in 17 patients: 1 aneurysmectomy, 1 CABG + implantable cardioverter-defibrillator, 3 endocardial resections, 1 endocardial resection + His bundle ablation, 1 cryoablation, 1 mitral valve replacement (MVR) + implantable cardioverter-defibrillator, 2 MVRs, 1 MVR + myotomy, 1 MVR + aortic valve replacement, 1 open mitral commissurotomy, 2 aortic valve replacements, 1 ventricular septal defect closure, and 1 complex reconstruction of congenital heart defects.
By random computer selection, a control group of patients without pCO before cardiac transplantation was obtained. The control group consisted of 53 patients matched for sex and age. The mean age was 48.5 ± 7.9 years. Twenty patients had dilated cardiomyopathy; 33 patients had end-stage ischemic heart disease (see Table 1
). There were 51 patients receiving anticoagulant therapy in the study group, and 48 patients in the control group. All pCO patients had received previous blood product transfusion, compared with 30.1% in the control group.
|
Investigated Indices
|
|---|
Preoperative data consisted of the standard serologic laboratory data, immunologic and virologic status, and hemodynamic indices. The perioperative variables were taken from the patients' records and included aortic cross-clamp and cardiopulmonary bypass times. The postoperative course was examined for complications and the course of rejection. Long-term outcome was assessed from the annual cardiac catheterization data to establish the incidence of chronic transplant atherosclerosis, and the actuarial survival was calculated.
|
Preoperative Data Assessment
|
|---|
All patients underwent right and left heart catheterization. Right heart catheterization determined the pulmonary artery pressure, pulmonary capillary wedge pressure, and cardiac output before and after application of nitroglycerin 0.5 mg intravenously. Pulmonary vascular resistance and systemic vascular resistance were calculated according to standard formulas. Left heart catheterization was used to analyze left ventricular ejection fraction, wall motion, and coronary angiography. All patients had a full blood count including hemoglobin, hematocrit, and differential white blood count. Liver function was assessed by measuring the transaminases (alanine aminotransferase, aspartate aminotransferase, 
-glutamyl transferase) and cholinesterase. Coagulation status was assessed by the Quick test, by partial thromboplastin time (PTT), and by measuring the factor II and V levels.
Renal function was assessed by measuring plasma creatinine, urea, and creatinine clearance. All patients were screened for cytomegalovirus, herpesvirus, Epstein-Barr virus, human immunodeficiency virus, and hepatitis A and B by immunoglobulin G and immunoglobulin M antibody determinations. Patients' sera were screened for cytotoxic antibodies, and the serum reactivity was tested against a nonspecific panel of different donor sera. Cross-match results were achieved by testing donor-specific lymphocytes against recipients' serum after transplantation. Preoperative cross-matches were defined as positive when a serum reactivity panel of more than 10% was present.
|
Operative Technique
|
|---|
The operations were performed according to standard techniques [7]. Cardiac preservation was achieved by either St. Thomas Hospital Solution (1,500 mL, 4°C) or University of Wisconsin Solution (1,000 mL, 4°C). The operation time was defined as the interval between skin incision and skin closure. The ischemia time was the interval between aortic cross-clamping in the donor and release of the cross-clamp in the recipient.
|
Perioperative Complications
|
|---|
In all patients, postoperative blood loss was measured, and the necessity for reoperation for bleeding or other causes was noted. Rhythm disorders were assessed by 12 hourly electrocardiographic-recordings. Respiratory insufficiency was defined as respiratory failure necessitating reintubation after extubation, which was performed according to common extubation criteria [8]. Infectious complications included those of respiratory, urologic, cardiovascular, and gastrointestinal causes. Bacteriologic cultures were performed according to standard procedures. All infections requiring antibiotic therapy after positive cultures were noted. Multiorgan failure was defined as failure of at least two organ systems, with refractoriness to therapeutic intervention. Cerebrovascular complications were defined as changes in the neurologic status. The diagnosis was based on examination by a neurologist, supported by an abnormal computed tomographic or magnetic resonance imaging scan.
|
Postoperative Data
|
|---|
The stay in the intensive care unit and the total hospitalization time were recorded. The incidence of infection or of rejection requiring treatment was established. Rejection was diagnosed using endomyocardial biopsies taken at weekly intervals for the first 8 weeks and at 2- to 3-week intervals thereafter. Histologic grading of biopsy specimens was performed according to the definitions of the International Society for Heart and Lung Transplantation [9].
|
Immunosuppressive Induction Therapy
|
|---|
All patients were treated by triple-drug immunosuppression: cyclosporin A, prednisolone, and azathioprine. Perioperatively, all patients received methylprednisolone, 125 mg intravenously, at 12, 24, and 36 hours. Induction therapy was with anti–T-lymphocyte globulin, 100 mg for 4 days. In the postoperative period, azathioprine dosages were adjusted to achieve a white blood cell count of 4,000 cells/µL. Prednisolone administration was started orally after extubation with 0.5 mg/kg and was tapered slowly down to 0.1 mg/kg within 6 months. The cyclosporine dose was determined by radioimmunoassay to achieve a specific serum level of 150 to 300 ng/L. Forty-three of 53 patients (84.3%) in the study group received induction immunosuppressive therapy with anti–T-lymphocyte globulin, compared with 90.2% in the control group. The remaining 10 patients in the study group received early intravenous application of cyclosporin A, compared with 5 patients (9.8%) in the control group. The immunosuppressive regimen was modified when infectious complications occurred or organ impairment was noted.
|
Statistical Analysis
|
|---|
Statistical analysis was performed using the SPSS computer program version 5.1. Data were analyzed by applying Student's t test, analysis of variance, 
2 test, and Kaplan-Meier actuarial survival curves. The log-rank test was used to determine statistically significant differences between the curves. Values were expressed as mean ± 1 standard deviation, and p values less than 0.05 were considered significant.
|
Results
|
|---|
Preoperative Cardiac Catheterization Data, Serologic Status, and Immunologic Status
No significant differences were found between the study and the control groups for preoperative hemodynamic values, serologic status, or pretransplantation immunology, including a negative immunoglobulin M profile for acute infection (Table 2). In the study group, 8 patients were receiving coumarin therapy (3 patients in the control group) and 43 patients received aspirin (45 patients in the control group). Two patients in the study group had no anticoagulant therapy (6 patients in the control group). Renal function was compromised in both groups. Although not of statistical significance, the serum creatinine was markedly higher in the pCO-group, indicating an increased renal risk.
|
Graft Ischemia Time, Aortic Cross-Clamp Time, Bypass Time, and Operation Time
|
|---|
Mean graft ischemic time in the study group was 151.7 ± 45.8 minutes, versus 154.8 ± 38.8 minutes in the control group (p = not significant). The mean aortic cross-clamp time was 42.1 ± 8.4 minutes for the study group and 36.8 ± 8.8 minutes in the control group (p = 0.003). In the study group, the mean bypass time was 91.2 ± 32.1 minutes, compared with 77.5 ± 30.9 minutes in the control group (p < 0.05). The mean operation time in the study group was 206.5 ± 62.5 minutes, compared with 156.0 ± 36.7 minutes in the controls (p < 0.05) (Table 3).
|
Perioperative Immunology
|
|---|
One patient in the study group (2%) showed a positive lymphocyte cross-match, versus no control-group patients. One hyperacute rejection causing death occurred in the study group, and 1 occurred in the control group. Preoperative panel reactive cytotoxic antibodies greater than 25% were found in 4 pCO patients (7.5%) and in 6 control patients (11.3%). None of the patients with preformed cytotoxic antibodies had hyperacute rejection after cardiac transplantation.
|
Perioperative Complications
|
|---|
Substantial postoperative bleeding leading to reexploration occurred in 8 study patients (17.0%), compared with none of the control patients (p < 0.01). The mean postoperative blood loss within the first 24 hours was 1,360 ± 260 mL for the study group, versus 730 ± 310 mL in the control group (p < 0.01). The mean perioperative blood product transfusion was 1,026 ± 210 mL in the study group versus 497 ± 150 mL in the control group (p < 0.05) (see Table 2). Nine patients in the study group (17.0%) required reoperation within the first 24 hours. Eight of these were due to postoperative bleeding. In the control group, only 1 patient required reexploration for right heart failure.
Twenty patients in the study group (37.7%) had perioperative infections, versus 9 control-group patients (17.0%; p < 0.05). One of the study-group patients died of severe pulmonary infection. No deaths occurred in the control group. Multiorgan failure developed in 5 patients in the study group (9.4%; p < 0.05), of whom 4 died (3 patients after retransplantation; p < 0.05). Cerebrovascular complications occurred in 3 patients in the study group (5.7%). All were due to ischemia, and the patients survived. Two patients in the control group (3.8%) had neurologic complications (one major and one minor stroke), leading to death in 1 patient. Five patients (9.4%) in the study group experienced respiratory failure after extubation, which required reintubation, versus 2 patients (3.8%) in the control group. Seven patients in the study group (13.2%) required pacemaker implantation because of rhythm disturbances, and there were 10 such patients (18.9%) in the control group; p = not significant. Five patients in the study group (9.5%) died perioperatively, compared with 1 patient (1.89%) in the control group (Table 4).
|
Hospital Stay
|
|---|
Hospital stay was longer in the study group than in the control group (intensive care unit stay: 4.8 ± 7.1 versus 3.5 ± 3.5 days; p = not significant). The mean hospital stay was 24 ± 19 days in the study group, versus 20.2 ± 5.6 days for the control group (p = not significant) (see Table 3).
|
Acute Rejection
|
|---|
The rejection rate within the first 2 years postoperatively was comparable between pCO patients and controls (p = not significant) (Table 5).
|
Cardiac Catheterization Data: Chronic Rejection
|
|---|
Forty pCO-patients (75.5%) underwent cardiac catheterization after 1 year, as did 50 patients (94.3%) in the control group. The mean left ventricular ejection fraction and cardiac index did not differ significantly between the groups. These findings were consistent for the following 2 years (see Table 5).
At the end of the first year, the incidence of moderate to severe tricuspid valve regurgitation (TR grade 3) in the control patients was 20.0% (n = 10/50), versus 40.5% (n = 17/40) in the study group (p < 0.05). After 4 years, 81.8% (n = 9/11) of pCO patients had TR, versus 44% (n = 15/34) in the control group. Mitral valve regurgitation (grade 3 to 4) occurred once in each group (2.9% in the study group versus 4.6% in the control group; p = not significant). The overall incidence of graft atherosclerosis as determined by coronary angiography after 1 year was 20.1% (n = 9/45) in the pCO group versus 20.0% (n = 10/50) in the control group (p = not significant). After 4 years, 9 pCO patients (80%) had graft atherosclerosis versus 13 (47%) in the control group (p < 0.05) (see Table 5).
|
Retransplantation
|
|---|
Among all patients, 6 had a repeat procedure after cardiac transplantation. All 6 patients belonged to the study group (p < 0.05 for pCO versus controls). Two patients underwent cardiac retransplantation because of severe chronic rejection and subsequent organ failure, and both patients died soon after this operation (1 died 10 days after retransplantation; the other died 50 days after retransplantation). Four patients had acute retransplantation after primary cardiac transplantation (all within the first postoperative week). Three of these patients died within 5 days after retransplantation; 1 patient is a long-term survivor. The mortality rate for this patient subgroup was 83.3% within the first postoperative year (Table 6).
|
Survival
|
|---|
The postoperative survival of pCO patients was lower than that in the control group. The actuarial survival curve revealed a perioperative survival rate of 98.1% in the control group (96.2% after 1 year, 84.6% after 2 years, and 83.1% after 4 years). In contrast, the pCO-patients had a perioperative survival of 91.5% (75.4% after 1 year, 71.6% after 2 years, and 60.1% after 4 years; p < 0.05) (Fig 1). For causes of death, see Table 7.
View larger version (18K):
[in this window]
[in a new window]
|
Fig 1. . Survival times determined by Kaplan-Meier actuarial curves. (HTx-Control = primary heart transplantation; HTx-pCO = previous cardiac operation before heart transplantation.)
|
|
|
Comment
|
|---|
Cardiac transplantation remains the ultimate therapy for end-stage cardiac disease. The annual registry of the International Society for Heart and Lung Transplantation shows an increase in the recipient pool, versus decreasing donor organ availability [6, 10]. Long-term results after cardiac transplantation have greatly improved. However, an increased number of transplantation candidates die before operation because the number of transplants has plateaued.
Attempts have been made to provide effective alternatives to cardiac transplantation, eg, cardiomyoplasty, high-risk cardiac operations, and implantation of automatic cardioverter-defibrillators [11–15]. Donor criteria also have been liberalized [16], but despite this, the donor organ supply remains unsatisfactorily low. Hence, only a selected group of candidates benefit from cardiac transplantation. Therefore, the selection of transplantation candidates warrants risk stratification to optimize postoperative patient outcomes.
We hypothesized that survival was impaired in patients undergoing heart transplantation with a history of cardiac operations. In contrast to Lammersmeier and associates [3], who did not show any significant impact on postoperative survival or rejection course after cardiac transplantation following pCO, we were able to demonstrate an enhanced risk profile for those patients undergoing repeat operations compared with a control group, matched for sex and age, having primary surgery. We also tried to match both groups for diagnosis, but the random computer selection could not achieve a complete match.
Our study showed that significantly increased operating times and blood loss were contributing factors to an increased incidence of operative reinterventions. Despite the perioperative use of aprotinin since 1989, blood loss in the study group was significantly higher. We believe that postoperative bleeding and blood product requirements were independent of preoperative anticoagulant therapy in those patients undergoing repeat operations, as compared with patients with preoperative coumarin or antiplatelet therapy in the control group. In a small series of unselected patients, Karck and Haverich [15] showed that preoperative coumarin therapy was associated with a significant increase in blood product requirement, but had no influence on bleeding-associated postoperative complications. Previous studies have demonstrated the relation between postoperative bleeding and an increased risk of infection. Our study confirmed that increased perioperative blood loss was associated with an increased infection rate. According to Lammermeier and associates [3], however, the increased infection rate in the pCO patient group may be due to a greater age of that patient group and a diminished tolerance for infectious complications.
In addition, although pCO has been identified as a risk factor for hyperacute rejection, we observed only 1 such case in the study group and 1 in the control group, confirmed by immunohistochemistry in a patient with a postoperatively positive lymphocyte cross-match. In a previous study, we described these 2 cases of hyperacute rejection (0.4%) in a series of 524 heart transplant recipients [17]. Two potential risk factors for hyperacute rejection were described: pCO and the persistence of identical viral genomes in the recipient and donor heart. In contrast, Lavee and colleagues [18] showed a tenfold higher incidence of hyperacute rejection (3.9%) in a series of 463 heart transplantations; although the retrospective lymphocyte cross-matches revealed positive results in 42 of 401 (10.4%) of the patients, there was no correlation with the incidence of hyperacute rejection. If a panel-reactive antibody test was positive, regardless of the panel-reactive antibody values, and the lymphocyte cross-match test was negative, the risk of death caused by rejection was not reduced. However, a panel-reactive antibody greater than 25% showed a significant correlation with the risk of death caused by any form of rejection.
Preoperative blood transfusions have been associated with improved allograft survival [15, 19]. We cannot confirm this finding. Our study group showed no significant difference in acute rejection compared with controls. However, chronic rejection, demonstrated by coronary angiography, was more frequent in the study group (80% for pCO patients versus 47% in controls after 4 years; p < 0.05). This demonstrates that the possibility of tolerance induction by previous blood transfusions is speculative; in contrast, a presensitizing mechanism may be associated with a more pronounced development of chronic rejection.
Perioperative morbidity in cardiac transplant recipients as reflected by bleeding complications requiring operative reintervention was significantly higher in pCO patients (17%) than in controls (1.6%) and repeat cardiac procedures (ie, repeat CABG, 4.59%) [20]. The rate of respiratory complications was 1.77% in repeat cardiac procedures (repeat CABG [20]), which is comparable with the results in our two groups.
In a retrospective analysis, Hausen and associates [21] demonstrated a significant correlation between rejection rate and the prevalence of tricuspid valve incompetence postoperatively after cardiac transplantation. After 4 years, the prevalence of TR grade 3 was 50%, and it increased thereafter. Only 4% of the general heart transplantation population remained free of TR at the fifth year. By the end of the third year, 45.7% of the patients who experienced five rejections in the first year (fewer than six rejections per year, 20%) had TR grade 3. During the first year, patients with TR grade 3 experienced an average of 4.8 ± 2.6 rejection episodes, versus an average of 2.5 ± 2.3 in those patients without TR. In the underlying study, the higher incidence of tricuspid valve incompetence in the study group suggested a higher rejection frequency in that group.
Long-term survival after primary heart transplantation was 81.1% at 4 years, compared with 60.1% in pCO patients, revealing a markedly impaired outcome in the latter. The impaired outcome in our study group is supported by data published by the International Society for Heart and Lung Transplantation in their annual report [10]: 4-year survival rate in the general cardiac transplantation population is reported as 70%. However, the difference in survival occurs early (0 to 9 months postoperatively), when the actuarial survival slopes diverge. The subsequent course shows parallel slopes. Taking a closer look at the early causes of death, it is generally noticed that a relatively high proportion of pCO patients die of early graft failure. The analysis of these patient courses revealed that 2 had right heart failure and 1 showed biventricular failure. Both cases of right heart failure had elevated preoperative pulmonary vascular resistance (>300 dynes • s • cm-5). The biventricular failure occurred in a patient with prolonged graft ischemia, where extensive preparation time because of multiple previous sternotomies was documented. All early donor organ failures were seen in patients who received donor organs with excellent or good performance before explanation. Optimized perioperative management (ie, nitric oxide inhalation) to avoid increased pulmonary resistance and subsequent right heart failure is suggested. In addition, longer preparation times should be expected when previous transsternal operations have been performed before heart transplantation; borderline transportation times thus should be avoided. The relatively high incidence of acute rejection leading to death was scattered throughout the average follow-up of 3 years in the study group. All of these patients had substantial renal impairment (serum creatinine level >200 mmol/L), and immunosuppression (cyclosporin A) was tapered to serum levels of 140 to 180 mg/dL (radioimmunoassay). New immunosuppressive agents with lower nephrotoxicity could provide a potent alternative, and inadequate immunosuppression therefore could be avoided.
The outcome of acute cardiac retransplantation was poor (20% survival in the first year). These results are similar to those in the published literature [22, 23].
In summary, our study has demonstrated that pCO patients undergoing cardiac transplantation have an increased operating time, increased blood loss, and greater risk for blood-borne infections because of increased transfusion requirements. Survival of such patients is significantly impaired. However, long-term survival seems not to be affected in pCO patients. Therefore, pCO should be considered an adverse risk factor for the early course after cardiac transplantation. Acute cardiac retransplantation in pCO patients has less than desirable results and should be avoided.
|
Footnotes
|
|---|
Address reprint requests to Dr Uthoff, Division of Thoracic and Cardiovascular Surgery, Surgical Department, Medical School Hannover, 30623 Hannover, Germany.
|
References
|
|---|
Top
Abstract
Introduction
