Prevention of Ischemic Injury in Cardiac Surgery

Monitoring for ischemia

Routine intraoperative monitoring during cardiac surgery includes temperature, pulse oximetry, capnography, surface electrocardiography, and noninvasive blood pressure monitoring. More invasive methods typically employed during cardiac surgery include an arterial line for continuous blood pressure monitoring and repeated blood sampling, a central venous catheter to measure central venous pressure, and use of a Swan-Ganz pulmonary artery catheter (PAC). Transesophageal echocardiography (TEE) is also an invaluable adjunct in the operating room, especially with valvular surgery.

The PAC provides potentially valuable information regarding pulmonary arterial pressure, pulmonary capillary wedge pressure (a surrogate of left-sided filling pressure), cardiac output, mixed venous saturation, and both systemic and pulmonary vascular resistance, all of which can be used to guide management directed at improving perfusion or optimizing hemodynamic performance after cardiotomy. Nevertheless, use of the PAC has been a topic of great debate ever since its introduction into clinical practice, and the controversy continues to rage today. Almost 20 years ago, several published studies failed to detect a benefit from the use of PACs in the setting of acute myocardial ischemia/infarction. Similarly, in 1989, Tuman and colleagues published their study showing no differences in outcome when PACs were used during CABG. In 1996, the SUPPORT trial (Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments) reported that PAC use was associated with longer hospital and intensive care unit (ICU) stays, significantly increased costs, and increased mortality, including a 1.5-fold increase in the relative risk of death in postoperative patients. In the ensuing discussions, some called for a moratorium on continued use of the PAC. However, attachment to the device by most practitioners proved strong, and, with evidence that the PAC was a useful tool for identifying perioperative ischemia, a more measured approach was adopted. Nevertheless, even the staunchest of PAC supporters recognized that nonselective, routine use of the PAC was not justified. Today, the decision to place a PAC preoperatively should be made between the surgical and anesthetic teams, and important factors such as the type of operation and baseline cardiac function will help guide this decision. TEE provides important data on valvular and cardiac function and is a common adjunct used in the cardiac operating room. Advocates of TEE point to the fact that it is generally safe and can provide nearly all the same data as the PAC, plus additional information on wall motion, ejection fraction, stroke volume, volume status, and valvular function not ascertainable by any other method. Small, nonrandomized studies of TEE in the setting of CABG have shown that TEE may be more important than PAC in guiding interventions involving fluid administration, vasoactive medications, and other anti-ischemia therapies, and that among high-risk patients undergoing CABG, data provided by TEE affected anesthetic or surgical management 50% of the time.

The reported safety and benefits of TEE notwithstanding, debate similar to that surrounding the PAC has emerged. Critics of TEE point to studies suggesting that unsuspected TEE findings of major significance occur in less than 2% of cases, that intraoperative interpretation by cardiac anesthesiologists is widely variable and often does not concur with later interpretation by cardiologists, and that, like PAC, TEE has not been shown to improve outcomes. Critics also point to the fact that performing TEE may divert the anesthesiologist from performing other critical tasks; in one study, anesthesiologists’ response time to an alarm light was 10 times slower when performing TEE than during monitor observation.

In sum, a range of modalities to detect ischemia are available to the modern cardiac anesthesiologist, and the wealth of information provided by them is vast. In valvular cardiac surgery, the importance of TEE is well established. However, in the setting of myocardial revascularization, serious questions remain unanswered as to whether the information these devices provide, and the way that clinical care is guided by that information, is helpful as determined by measuring important outcome variables such as morbidity, mortality, and cost. Thus, judicious rather than routine use is warranted.

Anesthetic agents

The goals of cardiac anesthesia are to maintain hemodynamic stability and myocardial oxygen balance, minimize the incidence and severity of ischemic episodes, and facilitate prompt and uncomplicated separation from CPB and assisted ventilation. Throughout the relatively short history of cardiac surgery, numerous anesthetic regimens have been developed, each with its own proponents and each with purported advantages and shortcomings. To date, however, no evidence exists that any one technique can be claimed to be superior for patients with cardiovascular disease.

Anesthesia during the early years of cardiac surgery consisted primarily of high-dose opioids, first morphine and later synthetic opioids. In the 1970s, fentanyl gained widespread acceptance for both induction and maintenance of anesthesia because of its improved hemodynamic stability. Because opiate-alone anesthesia was associated with an unacceptably high risk of patient awareness (which can lead to hypertension, tachycardia, and an attendant increase in myocardial oxygen consumption), benzodiazepines were added. More recently, propofol and volatile anesthetics have become a mainstay in both the induction and maintenance of cardiac anesthesia. Although some centers have gone to an all-intravenous anesthetic technique, the reported preconditioning protection provided by inhaled anesthetics has led some to advocate their continued and routine use. In practice, the importance of anesthesia in minimizing ischemic risk is perhaps most evident during induction, when the potentially wide variation in blood pressure may put overwhelming strain on an already stressed heart. Standard therapy in the modern era, therefore, may include preinduction use of beta-blockade in addition to anxiolytics. Typical induction agents include a combination of paralytic, analgesic, and anesthetic agents. Anesthesia is maintained with a combination of analgesics and anesthetic agents.

Systemic temperature

With the passing of Wilfred Gordon Bigelow in 2005, cardiac surgery lost the man referred to as “the father of heart surgery in Canada.” After training at Johns Hopkins, Bigelow returned to the University of Toronto and the Banting Research Institute in 1947, where over the next 2 decades, he and his colleagues performed much of the initial research that identified hypothermia as a practical means by which the body could be protected during the brief periods of circulatory arrest required for relatively simple cardiac repairs. Prior to his investigations, heart surgery was performed in an environment approaching normothermia. The early CPB circuits did not include a heat exchanger, and most of the temperature change that did occur was passive. Indeed, if any attempt was made to regulate the patient’s body temperature, it was to maintain normothermia. Hypothermia was seen as detrimental to the sick and wounded patients because of its adverse effects on coagulation and because metabolic rates were actually known to increase as patients got colder, largely caused by shivering. Bigelow’s systematic studies of surface-induced hypothermia, primarily using a canine model, were the first to show that with shivering minimized by adequate anesthesia, metabolism actually decreased in a linear relationship to core body temperature, with each 10°C temperature drop corresponding to a roughly 50% reduction in oxygen consumption.

Although metabolism could be essentially halted at low-enough temperatures, providing excellent protection for the brain, temperatures below 28°C to 32°C led to cardiac and pulmonary failure, limiting the early use of hypothermic protection alone to cases requiring only very brief periods of cardiac arrest. With the advent of CPB, however, temperatures could be reduced to 10°C, thereby lengthening the window of protection. Following these early observations on systemic hypothermia, numerous others expanded on the use of hypothermia and its application in cardiac surgery. Shumway and colleagues soon demonstrated the additional benefit gained by topical cooling of the heart using ice-cold saline lavage. This topical effect probably had more impact on the more anteriorly situated, thinner-walled right ventricle more prone to rewarming under the operative lights and ambient room temperature than the thicker left ventricle better protected by intravascular protection strategies. Other advances, including introduction of the heat exchanger in the CPB circuit, made hypothermia a mainstay of myocardial protection during the early years of cardiac surgery.

Despite significant advances in our understanding of the molecular processes involved in tissue metabolism, hypothermia continues to be a central component of myocardial protection strategies, particularly in the transplant realm. The use of hypothermia, however, remains relatively individualized between surgeons and institutions. Some centers routinely actively cool during CPB; others simply let the patient’s temperature drift downward. Almost all centers use topical cooling of the heart, although since the early 1990s, a few investigators have advocated “warm heart surgery.”

Although significant debate persists, consensus is building that rather than the creation of severe hypothermia or active maintenance of normothermia, the appropriate temperature for the systemic circulation may be mild to moderate hypothermia (approximately 28°C to 34°C). A number of studies suggest that such a tepid strategy provides the best level of myocardial protection without the consequences of deep hypothermia.

Myocardial acid–base management

The influence of intraoperative systemic pH status on CABG outcomes has been studied primarily in relation to its influence on neuroprotection. Most cardiac surgeons, and certainly cardiac anesthesiologists, are familiar with the concept of α-stat versus pH-stat management strategies, a detailed discussion of which is beyond the scope of this chapter. Less well understood is the relationship between ischemia, myocardial acid–base management, and non-neurologic outcomes.

Normal myocardial pH (7.2) is lower than systemic pH, and various studies have demonstrated that mild acidosis (6.8 to 7.0) may actually protect the myocardium during ischemia by decreasing cardiomyocyte energy demands. However, myocardial acidosis during CABG may be much more severe, with typical measurements of pH 6.5 or lower, and may trigger cell death via apoptosis. Decreased myocardial pH is a consequence of inadequate coronary blood flow, which results in decreased oxygen delivery, decreased washout of hydrogen ions, and an attendant rise in myocardial tissue partial pressure of carbon dioxide, and as such, it may be used as a surrogate marker for myocardial ischemia. During cardiac surgery, low coronary flow may occur as a result of preexisting severe coronary artery stenosis, ineffective cardioplegia during CPB, or inadequate revascularization. Myocardial acidosis may be the sentinel event in a potentially devastating cycle in which inadequate oxygen delivery leads to depressed myocardial function. This in turn necessitates the use of inotropes, which themselves can lead to increased myocardial oxygen consumption. If not reversed, irreversible cardiac injury can occur. As an indicator of underlying ischemia, measurement of myocardial acidosis represents a potentially valuable monitoring tool for guiding the care of the cardiac surgery patient. Regular, repeated arterial blood gas measurements are standard practice in cardiac surgery, but they represent only static images of a constantly changing landscape and may not accurately reflect the condition of the myocardium. Current myocardial protection strategies based on these intermittent measurements may be insufficient. Several attempts have therefore been made to develop real-time, continuous myocardial pH-measuring devices. Continuous blood gas monitors, for example, have been credited with decreasing the need for intraoperative pacing and cardioversion, decreasing the length of postoperative mechanical ventilation, and decreasing the length of ICU stay.

In a series of studies over the past 30 years, Khuri and colleagues have increased our understanding of the potential importance of intramyocardial pH management. These investigators have developed a system using electrodes implanted in the ventricular wall that allows continuous pH measurements to monitor for regional myocardial ischemia and decreased coronary perfusion. Their retrospective analyses conclude that low myocardial pH can predict clinically relevant outcomes ranging from an increased need for inotropic support, to an increased risk of 30-day adverse events, and finally to decreased long-term survival. On the basis of their findings, they have developed a series of recommendations aimed at keeping myocardial pH in a safe range throughout all aspects of any cardiac surgery procedure. Though compelling, these data and recommendations require prospective, randomized examination, and little external validation has ever been done; therefore, widespread adoption of intramyocardial pH monitoring has not become widely accepted in clinical practice.

Blood glucose

Diabetes mellitus has long been considered an established risk factor both for the development of cardiovascular disease and for significant perioperative morbidity and mortality associated with cardiac surgery. Even after adjusting for other confounding risk factors, such as age, hypertension, hypercholesterolemia, and smoking, diabetes has been shown in numerous studies to be a significant independent predictor of both short- and long-term survival after CABG. The data are not all consistent, as some studies have not identified diabetes as an independent predictor of mortality. Nevertheless, because diabetes now affects almost one-third of patients undergoing bypass surgery, optimal perioperative glucose management must be a priority for all cardiac surgeons.

Although the mechanism of diabetes-related cardiac pathophysiology is multifactorial, patients with more severe forms of the disease (i.e., those who require preoperative insulin therapy), and by extension those with higher blood glucose levels (poor control), have a poorer prognosis. Both acute and chronic hyperglycemia increase the risk of ischemic myocardial injury through a number of mechanisms, all of which may play a role around the time of cardiac surgery. These include a decrease in coronary collateral blood flow, endothelial dysfunction, and attenuation of the protective effects of inhaled anesthetics and other pharmacologic preconditioning agents. The association of higher blood glucose with increased morbidity and mortality has shifted the focus of research in this area away from characterizing the risk from diabetes per se to the role of elevated blood glucose in cardiac pathophysiology. In the absence of intervention, serum glucose concentrations in the intraoperative and perioperative periods often become elevated far above the normal range, even in nondiabetic patients. The cause of this elevation, which is similar to that which occurs in other forms of surgery and is in response to stress such as trauma or infection, reflects a combination of acute glucose intolerance in the form of insulin suppression, stress-hormone–induced gluconeogenesis, and impaired glucose excretion as a consequence of enhanced renal tubular resorption. The metabolic effects of diabetes and elevated blood glucose have been shown to be similarly wide ranging and include a higher incidence of left ventricular dysfunction, more diffuse coronary artery disease, altered endothelial function, and abnormal fibrinolytic and platelet function.

Maintenance of normal glucose levels during the intraoperative and perioperative periods is difficult, even in nondiabetic patients, and carries with it the risk for potentially life-threatening iatrogenic hypoglycemia. Nevertheless, Furnary and colleagues performed a series of studies investigating the feasibility of tight perioperative glycemic control and its effects on the important outcomes of sternal wound infection and death. An initial study of 1585 diabetic patients undergoing cardiac surgery demonstrated that elevated blood glucose levels (> 200 mg/dL) on the first and second postoperative days were associated with a higher incidence of deep sternal wound infection, and the average blood glucose level over those 2 days was the strongest predictor of deep sternal wound infection in a diabetic patient. On the basis of these findings, these investigators hypothesized that tight glycemic control would decrease the incidence of postoperative sternal wound infections. A prospective study of 2467 diabetic patients undergoing cardiac surgery was performed in which maintaining serum glucose at a level of less than 200 mg/dL was the goal. The control group (968 patients) was treated with intermittent doses of subcutaneous insulin, with administration based on a sliding scale; the study group (1499 patients) was treated with a continuous intravenous insulin infusion in an attempt to maintain a blood glucose level of less than 200 mg/dL. Continuous intravenous insulin infusion resulted in better glycemic control and a significant reduction in the incidence of deep sternal wound infection (0.8%) compared with the intermittent subcutaneous insulin injection group (2.0%; P < .01). A subsequent retrospective review of 3554 diabetic patients undergoing isolated CABG demonstrated that continuous insulin infusion resulted in better glycemic control. Furthermore, tight glycemic control led to a 57% reduction in mortality, with this reduction being accounted for by cardiac-related deaths. On the basis of these results, the authors concluded that diabetes mellitus per se is not a true risk factor for death after CABG and that continuous insulin infusion should become the standard of care for glucose control in diabetic patients undergoing CABG. After review of the available data, other investigators have also reached similar conclusions, namely that poor glycemic control, not a diagnosis of diabetes per se, significantly increases the risk of adverse clinical outcomes, prolonged hospitalizations, and increased health-care costs following cardiac surgery. Additional evidence comes from a prospective study involving 1548 critically ill patients in the surgical ICU, in which even tighter control (a serum glucose level goal of between 80 and 110 mg/dL versus 180 to 200 mg/dL) was associated with significantly improved mortality (4.6% versus 8.0%; P < .04). On the basis of these data, it is reasonable to conclude that tight glycemic control is likely beneficial for all patients undergoing cardiac surgery, and this has become a current standard of practice.

Transfusion strategy

Despite the development of national consensus guidelines for blood transfusion in the 1980s, in 2002 it was estimated that some 20% of all allogeneic blood transfusions in the United States were associated with cardiac surgery. Despite established guidelines, a number of studies have demonstrated that transfusion practices vary dramatically across institutions, with some centers transfusing less than 5% of patients and others transfusing nearly all patients. Different transfusion practices even within the same institution highlight the lack of widespread accepted transfusion thresholds. Furthermore, a recent analysis of cardiac surgery patients undergoing cardiac surgery found that red blood cell transfusion appears to be associated more closely with morbidity and mortality than with preoperative anemia; therefore, efforts to minimize unnecessary blood transfusion are justified.

The myocardium relies on either increased blood flow or increased oxygen content to satisfy increased oxygen demand. One of the primary rationales for blood transfusions in the setting of cardiac ischemia, therefore, is to increase oxygen delivery to the stressed myocardium. Unfortunately, very little evidence exists to support this rationale. On the contrary, some degree of anemia is required during hypothermic CPB to reduce blood viscosity and allow adequate flow without excessive arterial blood pressure. Furthermore, a number of large studies have concluded that blood transfusion is associated with increased short- and long-term mortality, including transfusions in the setting of CABG.

Decreased hematocrit is one of the prime drivers of the decision to transfuse, but management of hematocrit during CPB is controversial. Numerous studies have demonstrated that normovolemic anemia is well tolerated in cardiac patients, even at levels as low as 14%. Spiess and colleagues analyzed more than 2200 bypass patients and found that high hematocrit (34% or greater) upon entry to the ICU was associated with a significantly higher rate of MI than was a low hematocrit (less than 24%), leading the authors to conclude that low hematocrit might be protective against perioperative MI. In contrast, Klass and colleagues performed a study of 500 CABG patients and found no association between perioperative MI rate and hematocrit value on entry into the ICU. Habib and colleagues examined 5000 operations using CPB and found that a number of clinically significant outcomes, including stroke, MI, cardiac failure, renal failure, pulmonary failure, and mortality, were all increased if the intraoperative hematocrit nadir was less than 22%. Similarly, DeFoe and colleagues demonstrated that low hematocrit during CABG is associated with perioperative cardiac failure and increased in-hospital mortality. Each of these studies is limited by its retrospective design, and significant differences in patient populations and other key factors make direct comparisons difficult. Until a well-designed prospective study is performed, the optimal hematocrit value for CPB will remain undetermined, leaving the evaluation of this key indicator of the need to transfuse to the discretion of the physician.

Although the optimal hematocrit during CABG surgery is the subject of continued debate, data are accumulating on the deleterious effects of blood transfusion. Several studies have demonstrated the proinflammatory properties of transfused blood. In addition, the immunomodulatory effects of transfusion have been known for more than 2 decades, and blood transfusion has been associated with increased risk of bacterial as well as viral infection. A number of blood conservation strategies have been designed with the specific intent of decreasing the need for transfusion, including technical modifications to the bypass circuit and the use of various cell salvage techniques, such as autologous blood transfusion and Cell-Saver use. For those patients in whom transfusion cannot be avoided, the use of leukoreduced blood is gaining favor as a method of minimizing the detrimental effects of transfused blood. A national universal leukoreduction program in Canada has been credited with decreasing mortality and antibiotic use in high-risk patients. In the setting of CABG, the role of transfusing leukoreduced blood is unsettled. At least two studies have shown that leukoreduction is not associated with a decrease in postoperative infections. However, in a well-conducted, prospective trial, Furnary and colleagues showed that transfusing leukoreduced blood confers a survival advantage that is present at 1 month and persists up to 1 year.

In sum, despite the regular occurrence of blood transfusion in cardiac surgery patients, the indications, goals, effectiveness, and safety of this common clinical practice remain uncertain. Clinician preference and habit therefore continue to be the prime determinants of many blood transfusion strategies. Data on transfusion practices in cardiac surgery are mixed, but as one prominent expert in the field has concluded, the predominance of data regarding red blood cell transfusion does not support the premise that it improves outcome. Thus, until the appropriate patients and circumstances of transfusion are better defined, it is a practice to be used judiciously.