Open-Chest Management after Heart Transplantation

Abstract

Postcardiotomy open-chest management has been widely used in cardiac surgery. Although this strategy can be applied to heart transplantation, the use of immunosuppressants in transplant recipients raises particular concerns about sternal wound infection and impaired healing.

We performed a retrospective review of 403 patients who had undergone 410 heart transplantations at our institution from 1985 through 2004. Among them, 9 patients (2.2%) had open-chest management postoperatively. There were 8 men and 1 woman, with a mean age of 58 ± 7 years. The graft ischemic time ranged from 130 to 374 minutes (mean, 218 ± 99 min), and the cardiopulmonary bypass time ranged from 98 to 360 minutes (mean, 210 ± 69 min). In all cases, the reason for open-chest management was hemodynamic lability that precluded chest closure after transplantation. One patient also experienced postoperative bleeding. All patients underwent delayed sternal closure between postoperative days 1 and 11 (median, 4 days). Delayed sternal closure did not cause any significant hemodynamic changes. One patient died of stroke on postoperative day 22. No patient had sternal wound infection or impaired wound healing during the follow-up period.

We conclude that, when required, open-chest management is an effective and safe measure for hemodynamically unstable heart transplant patients.

Open-chest management (OCM) with subsequent delayed sternal closure (DSC) is an accepted technique in treating postcardiotomy heart failure. This concept was first described in 1975 by Riahi and associates.¹ There have been several reports of its efficacy in both adult and pediatric patients.^2–5 Open-chest management provides an edematous heart with additional space after cardiopulmonary bypass and can minimize the harmful effects of tamponade. In case of postcardiotomy failure, this technique may be useful until DSC is possible. This management strategy has been used in 1.2% to 4.2% of adult cardiac patients and in 9% to 62% of pediatric cardiac patients.^2–5

Postoperative cardiac failure is among the most serious early complications of heart transplantation. In addition to the well-known factors that can cause cardiac dysfunction in association with other cardiac surgical procedures, prolonged graft ischemic time, donor–recipient mismatch in size, and elevated pulmonary vascular resistance can cause cardiac failure after heart transplantation. As a result, OCM can be a helpful tool in acute postoperative management after heart transplantation. Concern regarding postoperative infection may have limited the use of this technique thus far. The objective of the present study was to review the incidence, patient characteristics, and outcomes of OCM and DSC in heart transplant recipients in our heart transplantation center.

Patients and Methods

From 1985 through 2004, 410 heart transplants were performed on 403 recipients at our institution. We retrospectively studied the records of these recipients in search of information on the incidence of OCM and on the characteristics and outcomes of patients who had been managed by this technique. Nine patients were identified as having been treated with OCM. Recipients who had their chests reopened for management of sternal infection (but who were not initially managed with OCM) were excluded from this analysis.

Transplant Surgery. Donor hearts were harvested by the standard technique and were preserved with University of Wisconsin solution, Collins solution, Stanford solution, or cardioplegic solution. Heart transplantations were performed by our standard technique, which involved biatrial anastomoses until 1995, and bicaval anastomoses thereafter.⁶

Open-Chest Management and Delayed Sternal Closure. The decision to keep the sternum open was made by the attending surgeon who performed the transplantation, primarily on the basis of the hemodynamic condition of each patient. In all instances, the wound was closed with synthetic material, and endotracheal intubation was maintained while the chest was left open. Inotropic agents were tapered, the patient's hemodynamic condition was kept stable, and the hemodynamic values were evaluated daily in deciding whether to perform DSC. All patients received immunosuppression in accordance with our protocol: azathioprine, starting with a single preoperative dose and continuing until the patient is extubated and able to take oral medications, at which time a change is made to mycophenolate mofetil; corticosteroids, starting at the time of cross-clamp removal; a calcineurin inhibitor (tacrolimus or cyclosporine), once renal function has stabilized; and induction therapy using rabbit antithymocyte globulin (ATG). The ATG is given 1.5 mg/kg at the time of cross-clamp removal and then 1 mg/kg on postoperative days 1, 3, and 5.

Data Analysis. All data are reported as mean ± SD. Postoperative days were defined as the number of days after transplantation. Statistical analysis was done using a paired Student's t-test.

Results

A total of 410 heart transplants were performed at our institution. Three hundred nine recipients were men and 94 were women. The mean age at transplantation was 50 ± 14 years. A left ventricular assist device was implanted before transplantation in 55 patients, and the mean graft ischemic time was 191 ± 87 minutes. Nine patients (8 men and 1 woman) underwent OCM (2.2%). The ages of these patients at transplantation ranged from 42 to 68 years (mean, 58 ± 7 years). The origins of cardiomyopathy were ischemic in 7 patients and idiopathic in the other 2. Prior sternotomy had been performed in 7 patients. Five patients had undergone pre-transplant circulatory support: with an intra-aortic balloon pump in 2 and a left ventricular assist device in 3. The graft ischemic time ranged from 130 to 374 minutes (mean, 218 ± 99 min), and the cardiopulmonary bypass time ranged from 98 to 360 minutes (mean, 210 ± 69 min). The indication for OCM was hemodynamic instability in all patients. In 7 of the 9 patients, an attempt at chest closure resulted in hemodynamic collapse, which was reversed by reopening the chest. In 2 patients, poor graft function or a need for high levels of inotropic agents indicated a need for OCM, and chest closure was not attempted. In 1 patient, substantial bleeding was associated with hemodynamic instability.

Hemodynamic Changes. Detailed perioperative data were available for 6 of the 9 patients. Comparisons of hemodynamic data on these 6 patients were performed at 3 points: immediately after transplant surgery, just before DSC, and immediately after DSC. No significant differences among patients were noted at any point, once the decision for OCM had been made. No data were available to compare hemodynamic values before and after attempted chest closure. All patients were weaned from inotropic agents while acceptable hemodynamic values were maintained, before DSC. No important changes in hemodynamics were observed before and after DSC.

Outcomes. Table I shows the outcomes of the patients. The sternotomy wounds were closed in all patients between postoperative days 1 and 11 (median, 4 days). There was 1 death, which resulted in a mortality rate of 11%. This patient had a prolonged cardiopulmonary bypass time during the transplant surgery, to enable management of surgical bleeding with massive fluid administration. This resulted in abdominal compartment syndrome, for which the patient underwent decompression laparotomy in addition to OCM. Although the wounds were successfully closed on postoperative day 11, he died of multiple strokes and sepsis on postoperative day 22.

TABLE I. Outcomes of All Heart Transplant Recipients Who Underwent Delayed Sternal Closure after Open-Chest Management

Follow-up data were obtained for the remaining 8 patients. These patients were extubated between postoperative days 2 and 19 (mean, 9.7 ± 6.3 days), and were weaned from inotropic agents between postoperative days 5 and 18. The mean duration of intensive care unit stay was 15 ± 7.7 days, and the mean hospital stay was 23 ± 10.0 days. None of the patients, including the one who died, developed surgical-site infection during the short- and long-term follow-up periods. The follow-up period for survivors ranged from 4 to 64 months (mean, 39.6 ± 22.3 months). Wound-healing was satisfactory in all 8 survivors. Significant rejection episodes (myocardial biopsy grade >2, International Society for Heart and Lung Transplantation) occurred in 1 patient, who was successfully treated by increasing immunosuppression.

Discussion

In this retrospective study, we found that recipients in 9 of 410 heart transplants required OCM, resulting in an incidence of 2.2%. All patients successfully underwent DSC without hemodynamic compromise. Moreover, none developed sternal wound infection or impaired wound healing.

Postoperative low cardiac output is one of the biggest concerns during the early period after orthotopic heart transplantation. Patients who do not respond to intravenous inotropic drugs often require mechanical assistance with intra-aortic balloon counterpulsation, a left ventricular assist device, or nitric oxide. Closure of the sternal wound may further reduce already-impaired cardiac output, so OCM with DSC is sometimes used. Since Riahi's group first described the technique in 1975,¹ there has been increased use of OCM with DSC. This stratagem, used in about 1.2% to 4.2% of adult postcardiotomy patients and up to 62% of pediatric patients,^2–5 has become one of the most valuable tools for the management of postoperative low cardiac output.

In heart transplant patients, post-transplant heart failure may be attributable to prolonged graft ischemic time, size mismatch between donor and recipient, or elevated pulmonary vascular resistance. Acute right ventricular failure is a frequent cause of early morbidity and death.⁷ In these instances, OCM may be a valuable tool for the management of early hemodynamic instability after transplantation. We noted that none of our patients before 1998 required this management; perhaps post-transplant graft dysfunction is now more frequent, due to sicker recipients, longer graft ischemic time, and older donor age. In our series, 5 of 9 patients received pre-transplant circulatory support, and the graft ischemic times were long (a mean of 218 ± 99 min). Solomon and colleagues, from New Zealand, reported a change in the age and illness distribution of their donors and recipients that is compatible with our own observation.⁸ A review of the International Society for Heart and Lung Transplantation registry⁹ confirms this trend.

In most heart transplant centers, infectious complications are among the most common causes of death after transplantation. Despite the development of more effective immunosuppressants that are associated with fewer infectious complications, infection remains a major problem.¹⁰ Carrier and co-authors¹¹ reported the incidence and results of sternal wound infection after heart transplantation—a potential shortcoming of OCM. Sternal wound infection developed in 20 out of 226 patients at their institution, an incidence of 8.8%. At our institution, there were 9 cases of post-transplant sternal wound infection out of 410 heart-transplant procedures, an incidence of 2.2%. Sternal wound infection after DSC has been extensively reviewed in non-transplant cardiac surgeries. In series that involve adults, the incidence has been reported to be from 2.4% to 5%.^2,12 In the pediatric population, several groups^4,13,14 have reported OCM as a risk factor of sternal wound infection, with incidences ranging from less than 1% to 10.3%. In our series, no patient developed either superficial or deep sternal wound infection.

In conclusion, our study suggests that OCM followed by DSC is effective and safe in selected patients who have hemodynamic instability after orthotopic heart transplantation. Our observed lack of any increase in the rate of sternal infections supports the use of this management option in selected patients.