The Use of Hypothermic Circulatory Arrest During Heart Transplantation Does Not Worsen Post-Transplant Survival

Abstract

BACKGROUND

Hypothermic circulatory arrest (HCA) has been used as an adjunct to cardiopulmonary bypass for decades, both electively and emergently, in order to facilitate a bloodless operative field while maintaining cerebral protection. The aim of this study is to determine the impact of HCA during heart transplantation on post-transplant outcomes.

METHODS

All adult patients undergoing orthotopic heart transplant at our institution between 2000–2012 were retrospectively reviewed. Patients were stratified based on need for HCA during surgery; patients who required HCA (Group HCA, n = 25) and those who did not (NoHCA, n=903). The primary outcomes of interest were 30-day and 1-year mortality and postoperative complication rate.

RESULTS

Indications for HCA included control of significant hemorrhage (n=9), need for distal aortic procedures (n=9), or as an aid in difficult mediastinal dissection (n=7). Mean duration of HCA was 22 ± 18 minutes at a mean temperature of 24.5 ± 5.5 °C. Significantly more patients in Group HCA underwent transplant for congenital heart disease (16.0% HCA vs. 2.8% NoHCA, p = 0.006) and patients in HCA had undergone more prior sternotomies [1 (IQR 1–2) HCA vs. 1 (IQR 0–1) NoHCA, p < 0.001]. There was no statistical difference in 30-day (8.0% HCA vs. 4.2% NoHCA, p = 0.29) or 1-year (8.0% HCA vs. 12.3% NoHCA, p = 0.76) mortality. Group HCA experienced higher rates of reoperation for mediastinal bleeding and postoperative respiratory failure.

CONCLUSIONS

The need for HCA during heart transplant is rare but, when required, it is frequently a life-saving adjunct to cardiopulmonary bypass. However, patients who require HCA experience higher rates of postoperative complications. Risk factors for needing HCA during transplant include congenital heart disease and multiple prior sternotomies.

Hypothermic circulatory arrest (HCA) is a useful adjunct to cardiopulmonary bypass (CPB) for control of mediastinal structures not possible with standard CPB by allowing for a bloodless operative field while still maintaining cerebral protection by cooling the brain (1). Elective experience with this technique in adults undergoing thoracic aortic surgery, including repair of aortic arch aneurysms, is robust and has generally shown acceptable rates of perioperative morbidity and mortality as well as good late cognitive outcomes (2–6). In addition to elective uses, HCA is also particularly useful in the emergency control of massive hemorrhage from the major cardiac or head vessels (7–10), making it a useful tool in complex and reoperative cardiac surgery.

Patients undergoing heart transplantation in the current era have frequently had multiple prior sternotomies, a left ventricular assist device (LVAD) implanted, or have significantly altered cardiac anatomy from congenital heart defects, all of which contribute to the formation of dense mediastinal adhesions and increase the risk of vascular injury on sternal reentry or dissection. Additionally, outflow grafts of ventricular assist devices placed on the ascending aorta occasionally necessitate a more distal donor-recipient aortic anastomotic line. For these reasons, HCA is occasionally needed, either electively or emergently, during heart transplantation. However, studies of HCA in the transplant population are lacking, and a better understanding of how HCA specifically affects transplant outcomes may allow for expanded utilization, and potentially lower surgical operative risk. Therefore, the aim of this study is to evaluate the rate of HCA in heart transplantation and to determine its impact on the rate of post-transplant morbidity and mortality.

PATIENTS and METHODS

Patient Selection

All adult patients (age ≥ 18 years) undergoing orthotopic heart transplantation at our institution between 2000–2012 were retrospectively reviewed. Operative notes and cardiopulmonary perfusion records were reviewed for the need for HCA as an adjunct to CPB during transplant surgery. Patients were then stratified into two groups based on the need for HCA; patients who required HCA (Group HCA, n = 25) and those who did not require HCA (Group NoHCA, n = 903). The study was approved by the Columbia University Institutional Review Board and need for informed consent was waived.

Data was collected from a review of the electronic medical record and survival follow-up for patients who were followed at other transplant centers was obtained from the United States Social Security Death Index. Variables collected included recipient age, sex, body mass index, etiology of heart failure, medical co-morbidities, number of prior sternotomies, CPB time, aortic cross-clamp time, graft ischemic time, intraoperative transfusion rate, minimum CPB temperature, HCA variables including indication, duration, urgency of cannulation, and cannulation strategy, and post-transplant outcomes including 30-day and 1-year mortality and postoperative complication rate. Respiratory failure was defined as intubation > 72hrs or tracheostomy placement and post-operative stroke was defined as the presence of a new, focal, reproducible neurologic deficit lasting > 24 hours with or without presence of a cerebral lesion on neuroimaging as determined by a neurologist.

Operative Procedure

All recipients in the current study underwent bicaval orthotopic heart transplant with cold storage of the graft during transport to our center. A standard sternal saw was used for sternotomy in patients with no prior sternotomy. In patients with a prior sternotomy, an oscillating saw was used for sternal division. In the absence of the need for HCA, arterial cannulation was placed high on the ascending aorta to facilitate proper orientation of the graft-recipient aortic anastomosis. Venous cannulation was accomplished in the distal superior vena cava and inferior vena cava and snares were used to completely isolate the cardiopulmonary circulation. Target skin temperature during CPB without HCA was approximately 32°C, with slow rewarming after release of the cross-clamp.

In cases that required HCA, cannulation site was left to the discretion of the operating surgeon. When used electively due to an anticipated difficult sternal reentry or emergently due to significant hemorrhage encountered during sternal reentry, the femoral artery and vein were the most common site for arterial and venous cannulation. Target cooling temperature was also left to the discretion of the operating surgeon and depended on the anticipated duration of HCA. In all cases of HCA, the head was packed in ice and barbiturates and glucocorticoids were administered for cerebral protection during the period of circulatory arrest. Continuous electroencephalography was not routinely used.

Statistical Analysis

All analysis was conducted using SPSS version 22 (IBM Corp., Armonk, NY). Continuous variables are reported as mean ± standard deviation and compared using the independent samples t-test or as median and interquartile range and compared using the Mann-Whitney U test for variables with non-normal distributions. Categorical variables are presented as frequency and percentage of total group and compared using Pearson’s chi-squared test or Fisher’s exact test where applicable. Kaplan-Meier analysis was used to compare post-transplant survival at 5 years. Survival curves are compared using the log-rank test. All p-values ≤ 0.05 are considered statistically significant.

RESULTS

Baseline Demographics

Patient demographics, etiology of heart failure, and co-morbidities are presented in Table 1. Mean age for the entire cohort was 53.0 ± 12.3 years and was not significantly different between groups. All patients in Group HCA were male which was significantly higher than the 77.4% of males in Group NoHCA (p = 0.007). The rate of patients undergoing heart transplant for congenital heart disease (3 single ventricle patients, 1 transposed great vessels status post mustard) in Group HCA was significantly higher than patients in NoHCA (16.0% HCA vs. 2.8% NoHCA, p = 0.006). Additionally, there was a trend towards a lower incidence of HCA use in patients with dilated cardiomyopathy (1.6%) compared with other etiologies. There were no differences in medical co-morbidities including prior stroke, coronary artery disease, hypertension, diabetes, or peripheral vascular disease. Overall, 325 (35%) of patients had a left ventricular assist device at the time of transplant with no significant difference between groups. Of patients in HCA, 7 had an implantable LVAD, 2 had a temporary LVAD, and 3 had temporary BiVADs in place at the time of transplant. Patients in HCA had undergone prior sternotomy more frequently than NoHCA [1 (IQR 1–2) HCA vs. 1 (IQR 0–1) NoHCA, p < 0.001].

Table 1.

Baseline Characteristics

	NoHCA	HCA	p-value
Demographics
Total, n	903	25	---
Age, years (mean ± SD)	53.8 ± 12.3	50.4 ±15.2	0.29
Male, n (%)	699 (77.4)	25 (100)	0.007
BMI, kg/m2 (mean ± SD)	26.2 ± 11.8	25.3 ± 5.2	0.72
Etiology of Heart Failure, n (%)
Dilated CM	426 (47.2)	7 (28.0)	0.06
Ischemic CM	350 (38.8)	13 (52.0)	0.18
Valvular CM	26 (2.9)	1 (4.0)	0.53
Restrictive CM	33 (3.7)	0 (0)	1.00
Congenital heart disease	25 (2.8)	4 (16.0)	0.006
Medical/Surgical History
Stroke, n (%)	108 (12.0)	3 (12.0)	1.00
Coronary artery disease, n (%)	388 (43.0)	12 (48.0)	0.62
Hypertension, n (%)	370 (41.0)	8 (32.0)	0.37
Diabetes, n (%)	262 (29.0)	2 (8.0)	0.02
Peripheral vascular disease, n (%)	52 (5.8)	1 (4.0)	1.00
Prior LVAD implant, n (%)	313 (34.7)	12 (48.0)	0.17
Prior sternotomies, n (median, IQR)	1 (0–1)	1 (1–2)	< 0.001

Abbreviations: BMI=body mass index, CM=cardiomyopathy, LVAD=left ventricular assist device

Operative Details

General CPB data and intraoperative transfusion data are presented in Table 2. CPB time (243 ± 73 mins HCA vs. 164 ± 48 mins NoHCA) and aortic cross-clamp time (115 ± 34 mins HCA vs. 96 ± 28 mins NoHCA) were both significantly longer in Group HCA (p < 0.001 and p = 0.001, respectively). However, graft cold ischemic time was not significantly longer in Group HCA. Additionally, patients in HCA required significantly more intraoperative transfusions than NoHCA patients. Not surprisingly, minimum bypass temperature was significantly lower in Group HCA.

Table 2.

Operative Variables

	NoHCA	HCA	p-value
CPB time, mins (mean ± SD)	164 ± 48	243 ± 73	< 0.001
Cross-clamp time, mins (mean ± SD)	96 ± 28	115 ± 34	0.001
Graft ischemic time, mins (mean ± SD)	193 ± 55	219 ± 69	0.11
Intraoperative PRBCs, units (mean ± SD)	1.8 ± 2.2	4.0 ± 3.9	0.01
Minimum temperature, °C (mean ± SD)	31.6 ± 2.3	24.5 ± 5.5	< 0.001

Abbreviations: CPB=cardiopulmonary bypass, PRBC=packed red blood cells

Circulatory arrest details for patients in Group HCA are presented in Table 3. The rate of HCA use during heart transplant was 2.7%. Mean duration of HCA time was 22 ± 18 minutes (range 1–58 minutes) and mean temperature for the HCA period was 24.5 ± 5.5 °C (range 18–32 °C). Overall, 9 patients (36%) required HCA in order to control significant hemorrhage, either upon sternal reentry or during mediastinal dissection (right ventricle × 2, innominate vein × 2, superior vena cava, ascending aorta × 3, prior bypass graft), 9 patients (36%) for distal aortic procedures, and 7 patients (28%) as an aid in an anticipated difficult sternal reentry or cardiac dissection. Of patients undergoing HCA for aortic procedures, there were 3 iatrogenic aortic dissections (1 hemiarch replacement, 2 ascending aortic replacements), 3 aortic atheromas or calcifications that required resection distally and extended donor aortic distal anastomosis, 1 ascending aortic aneurysm that required dacron graft replacement, 1 area of denuded aortic wall that required cormatrix patching, and 1 patient that required a particularly distal aortic anastomosis due to positioning of an LVAD outflow graft. Urgent or emergent cannulation (72%) was far more common than elective/non-urgent cannulation. Finally, central arterial and venous cannulation was the most common strategy (68%) followed by peripheral arterial and venous cannulation (24%). The femoral artery was used in all cases of peripheral arterial cannulation.

Table 3.

Circulatory Arrest Characteristics

Indication, n (%)
Hemorrhage	9 (36)
Distal aortic procedure	9 (36)
Mediastinal dissection aid	7 (28)
Urgency of cannulation, n (%)
Elective/Non-Urgent	7 (28)
Urgent/Emergent	18 (72)
Cannulation Strategy, n (%)
Central A+V	17 (68)
Peripheral A+V	6 (24)
Central A/Peripheral V	1 (4)
Peripheral A/Central V	1 (4)
Duration, mean ± SD (min)	22 ± 18
Range, min	1 – 58

Abbreviations: A=arterial, V=venous

Outcomes

Post-transplant outcomes are presented in Table 4. There was no difference in 30-day mortality (8.0% HCA vs. 4.2% NoHCA, p = 0.29) or 1-year post-transplant mortality (8.0% HCA vs. 12.3% NoHCA, p = 0.76) between groups. Patients in Group HCA experienced higher rates of reoperation for mediastinal bleeding and postoperative respiratory failure than patients in NoHCA. Rates of postoperative stroke, new need for continuous renal replacement therapy, need for permanent pacemaker implantation, and deep sternal wound infection were similar between groups. Additionally, there was no difference in median postoperative hospital length of stay. At 5 years post-transplant, there was no significant difference in survival between groups (Figure 1, log rank p = 0.29). Finally, there were no differences in perioperative mortality or post-operative complications between patients cannulated for HCA in a non-urgent fashion and those cannulated urgently.

Table 4.

Post-transplant Outcomes

	NoHCA	HCA	p-value
Mortality, n (%)
30-day	38 (4.2)	2 (8.0)	0.29
1-year	111 (12.3)	2 (8.0)	0.76
Postoperative Complications, n (%)
Reoperation for bleeding	103 (11.4)	7 (28.0)	0.02
Respiratory failure	65 (7.2)	6 (24.0)	0.009
New need for CRRT	106 (11.7)	4 (16.0)	0.53
Stroke	16 (1.8)	1 (4.0)	0.37
Deep sternal wound infection	57 (6.3)	1 (4.0)	1.00
New PPM implant	14 (1.5)	0 (0)	1.00
Postoperative LOS, days (median, IQR)	17 (13–26)	18.5 (14–40.8)	0.35

Abbreviations: CRRT=continuous renal replacement therapy, LOS=length of stay, PPM=permanent pacemaker

Figure 1.

Kaplan-Meier analysis of 5-year survival stratified by treatment groups.

NOTES

Article Type = Original Article

IC by Kilic to come.

COMMENT

Hypothermic circulatory arrest as an adjunct to CPB during adult cardiac surgery has been used for decades, particularly in the field of aortic surgery (11). Arresting the circulation allows for a bloodless operative field and, in combination with hypothermia, protects the brain tissue by significantly decreasing metabolic demand. Outside of the elective realm, the use of HCA to facilitate repair of major vascular injuries leading to exsanguinating hemorrhage, especially in reoperative cardiac surgery, is a life-saving technique (12). Despite the increasing complexity of heart transplant surgery due to a rising frequency of LVAD and transplant candidates who have undergone prior cardiac surgery, no studies evaluating the effect of HCA on post-transplant outcomes currently exist. Therefore, our aim in this study was to evaluate our experience with HCA in the transplant population and determine its impact on transplant recipient morbidity and mortality.

Our analysis has shown that HCA is used in approximately 3% of heart transplants and is employed for 3 major indications; 1) to control and repair significant hemorrhage either upon sternal entry or during mediastinal dissection, 2) to perform distal ascending aortic/arch procedures including repair of iatrogenic acute aortic dissections, and 3) as an aid in particularly difficult mediastinal dissection, often after multiple prior sternotomies. There was no significant difference in perioperative or late mortality between patients who required HCA and those who did not. However, patients who required HCA during transplant experienced significantly higher rates of intraoperative transfusion, reoperation for postoperative bleeding, and respiratory failure. Analysis of preoperative risk factors shows that prior sternotomy and congenital heart disease are significantly associated with the need for HCA during heart transplantation.

The perception among the cardiac surgical community is that if HCA is needed during surgery, particularly in an unplanned setting, patients are at increased risk of complications, and our analysis does confirm this to be true in the transplant population. While we found no difference in perioperative or late mortality, we found that significantly more complications were experienced by patients requiring HCA such as transfusion requirement, postoperative bleeding requiring operative intervention, and higher rates of respiratory failure. This suggests that, while HCA is frequently lifesaving during its use, it does lead to a significantly higher utilization of healthcare resources during the index hospitalization.

Although we have grouped all HCA patients into one group for the purposes of analysis, it must be kept in mind that the group is somewhat heterogeneous, namely based on the urgency with which HCA is implemented and etiology of underlying pathology. In cases where HCA is implemented at the start of the case because of an anticipated difficult sternal reentry, cannulation is done peripherally and in a controlled manner after which cooling can begin. Conversely, when HCA is implemented in an urgent or emergent fashion due to massive hemorrhage or aortic dissection, cannulation is undertaken in a more expedient manner and venous drainage may be suboptimal for a significant period of time if cardiotomy suction is being used as an adjunct to a true intravascular venous cannula. Thus, while we have grouped all HCA patients together for our analysis, there are truly different “sub-populations” within the overall group based on urgency of HCA implementation.

When cannulation for CPB/HCA is done electively or emergently during sternal reentry, peripheral cannulation is required as no other options exist at the time of cannula placement. However, 68% of our patients were cannulated centrally, which includes some patients who underwent HCA for bleeding control. Specifically, in 5 cases of hemorrhage that were cannulated centrally, 4 involved direct cannulation of a left ventricular assist device outflow graft with standard venous cannulation and 1 was due to a posterior superior vena cava injury that was digitally controlled until the target temperature was reached. If a sufficient amount of mediastinal dissection has occurred prior to the onset of hemorrhage and/or an LVAD outflow graft is accessible, central aortic or LVAD graft cannulation rather than the femoral vessel cannulation is likely more optimal given the risk of vascular complications or retrograde embolic events.

Target cooling temperature warrants some discussion. In the initial published experience with the use of HCA, cooling to deep hypothermic levels (~18°C) was typically used as it was shown that the majority of cerebral function ceased at this point and, therefore, offered the best cerebral protection (11). Subsequently, more analysis has been undertaken in the aortic surgery population, and several studies have shown that moderate hypothermia (22 to 26 °C) does not worsen postoperative cerebral event rates (13, 14). Although the optimal cooling temperature for HCA during heart transplant is beyond the scope of this paper, we do feel that cooling to higher temperatures, such as 25 °C, is appropriate for short periods of anticipated HCA (eg. control LVAD outflow graft), and cooling to deep hypothermic levels (18 to 20 °C) can be reserved for cases with particularly difficult mediastinal dissection, prior congenital cardiac operations, and major aortic injuries. Additionally, given that the majority of our experience was with urgent/emergent HCA, the option to use antegrade cerebral perfusion techniques would be cumbersome at best.

The study suffers from several limitations. This was a retrospective, single-center study and reflects the bias of our surgeons, anesthesiologists, and perfusionists. Our study is underpowered given the small sample size of patients who required HCA during the transplant procedure. Furthermore, there was no formal protocol for implementing HCA during the study period and cannulation strategy and target cooling temperature were left to the discretion of the operating surgeon. We acknowledge that some patients in our series underwent brief periods of circulatory arrest that may have had little effect on the overall outcome of the patient, however, we included these patients in order to provide a complete perspective on circulatory arrest use in the transplant population. Additionally, our definition of respiratory failure includes patients who remain intubated > 72 hours after the index operation, thus, if they required exploratory reoperation and remained intubated due to timing of reoperation rather than truly intrinsic respiratory failure, they were labeled as having respiratory failure. Finally, our criteria for inclusion into the study was actually undergoing HCA, so if a patient was cannulated for potential HCA use prior to incision but then did not require HCA, they are not included in our analysis.

In conclusion, we have found that the use of HCA, implemented either electively or emergently, during heart transplant surgery does not lead to worsened perioperative or late survival in comparison to patients who did not require HCA. However, the use of HCA is associated with higher rates of postoperative complications including higher rates and bleeding and respiratory failure, primarily due to more complex underlying anatomy. As cardiac transplantation surgery becomes increasingly complex given the rising use of LVAD therapy and the growing population of adults with congenital heart disease, the need for HCA during heart transplantation may rise in the coming decades, but will likely remain fairly rare. Nonetheless, the use of HCA during heart transplant surgery is frequently life-saving and will remain a necessary adjunct to cardiopulmonary bypass for the cardiac transplant surgeon.