Chapter 52. Cardiac Anesthesia

1. Initiation of cardiopulmonary bypass triggers an extremely complex and multifactorial response involving activation of complement, platelets, neutrophils, monocytes, and macrophages, thus initiating the coagulation, fibrinolytic, and kallikrein cascades. The systemic inflammatory response to cardiopulmonary bypass is further amplified by subsequent stimulated release of various endotoxins and cytokines, including interleukins and tumor necrosis factor, which further promote endothelial cell permeability.

2. Preparation for separation from cardiopulmonary bypass must be based on a clear understanding of the patient's preoperative condition and events of the operative course. Weaning from cardiopulmonary bypass is initiated after review and adjustment of numerous physiologic and technical variables, including temperature, laboratory data, heart rate and rhythm, myocardial contractility, and mechanical ventilation.

3. The anesthetic management of patients undergoing coronary artery bypass graft surgery requires an understanding of myocardial oxygen supply and demand, patient monitoring, and the anesthetic techniques that provide myocardial protection and favor oxygen supply over demand.

4. Patients with coronary artery disease presenting for coronary artery bypass graft surgery require special considerations in their anesthetic management. First and foremost are techniques that minimize myocardial oxygen demand while maximizing myocardial oxygen delivery. These considerations include preoperative preparation, intraoperative monitoring, and the use of anesthetic agents with hemodynamic effects that favor oxygen supply over demand and allow for myocardial protection. Postoperative management that provides particular attention to pain management, temperature control, and hemodynamic monitoring to avoid tachycardia, hypotension, and hypertension also must be considered.

5. Modern practices that focus on early extubation and "fast-tracking" cardiac surgical patients through the postoperative period use smaller narcotic doses, with supplementation by short-acting hypnotic agents.

6. Patients at risk for increased mortality after coronary artery bypass graft surgery are identified by preoperative factors. The most significant risk factors that increase mortality are older than 80 years of age, emergent surgery, prior cardiac surgery, and renal failure.

7. The unifying concept in all valve surgery includes the principles of preserving myocardial function and the influence of preload, afterload, inotropy, rate, rhythm, and diastolic function on myocardial performance and mechanics.

8. Intraoperative transesophageal echocardiography is an essential diagnostic tool and monitor of cardiac performance for patients undergoing heart valve procedures.

9. Cardiac anesthesia for heart valve surgery is associated with a number of special considerations not found in other aspects of cardiothoracic anesthesiology. Among these considerations is familiarity with the type of repair technique or prosthetic valve used.

10. The process of repairing or replacing a portion of the thoracic aorta typically requires the temporary or permanent interruption of blood flow through the aorta or its major branch vessels, creating the potential for ischemia or infarction of almost any major organ system. Techniques to protect organs during temporary interruption of blood flow in the thoracic aorta include deep hypothermic circulatory arrest, selective antegrade cerebral perfusion, retrograde cerebral perfusion, and partial left heart bypass for distal aortic perfusion. Intraoperative neurophysiologic monitoring and lumbar cerebrospinal fluid drainage are recognized techniques commonly used for repairs involving the descending thoracic or thoracoabdominal aorta to decrease the risk of spinal cord ischemia and infarction.

11. Preoperatively, the anesthesiologist must identify any neurologic deficits and ascertain the extent of major organ dysfunction, including renal or hepatic insufficiency, which are common in patients undergoing placement of a ventricular assist device. Any further deterioration in the perioperative period may prevent a full recovery and eliminate the possibility of a patient qualifying for subsequent heart transplantation at a later date.

12. Right heart failure is one of the most important causes of perioperative death in patients undergoing placement of a left ventricular assist device. Right ventricular dysfunction and failure may develop in up to 20% to 30% of patients implanted with an isolated left ventricular assist device. β-Adrenergic agonists, phosphodiesterase inhibitors, as well as nitric oxide should be initiated to improve right ventricular contractility and decrease right ventricular afterload in an attempt to improve transpulmonary blood flow. A supplemental temporary right ventricular assist device may be necessary if right ventricular dysfunction persists.

Despite the introduction and perseverance of off-pump coronary artery bypass graft surgery (OPCABG), cardiopulmonary bypass (CPB) remains an essential technique for most cardiac surgical patients.1 Although the past 50 years have brought improvements in extracorporeal technology, including improved gas exchange devices, venous reservoir construction, and heparin-coated circuits, the modern extracorporeal circuit is still remarkably similar to that developed a half-century ago.2 Despite such advances, fundamental questions remain regarding the management of CPB (pump flow, arterial blood pressure, temperature, acid-base strategy, hematocrit, or glucose level). A comprehensive understanding of the CPB circuit, pathophysiology, and important patient management issues is essential for anesthesiologists responsible for the perioperative care of these patients.

Fundamental Components of a CPB Circuit

Traditionally, the CPB circuit removes blood from the body via the venous circulation and returns blood to the body via the arterial circulation.3 Thus CPB must function "physiologically" as the patient's heart and lungs. Most commonly, blood is drained by gravity via cannulae in the superior and/or inferior vena cavae and then is similarly returned by a cannula in the ascending aorta. Although the CPB circuit consists of many important components, the 2 most important components are the pump, which functions as the patient's heart, and the oxygenator, which functions as the patient's lungs.

The 2 most common types of pumps used are roller pumps and centrifugal pumps. Roller pumps induce blood flow by compressing the tubing via rollers, thereby pushing the blood ahead through the tubing. Flow rate depends on the size of the tubing, length of the roller track, and rotation rate of the rollers (ie, revolutions per minute). Alternatively, centrifugal pumps induce blood flow by either a vaned impeller or a group of cones that reside inside a plastic housing. The impellers or cones are magnetically coupled with an electric motor and, when rotated rapidly, generate a pressure differential that may cause movement of blood through the tubing. Thus, unlike roller pumps, centrifugal pumps are afterload dependent, such that increases in resistance decreases forward flow delivered to the patient.

The 2 most common types of oxygenators used are membrane oxygenators and bubble oxygenators. Membrane oxygenators attempt to achieve separation between blood and gas in a manner analogous to the natural lung. Bubble oxygenators physically mix blood and gas, allowing sufficient time for adequate gas exchange to occur prior to defoaming and delivery to the patient. A heat exchanger usually is an integral part of the oxygenator. The type of oxygenator used influences the configuration of the CPB circuit. Membrane oxygenators typically are positioned after the pump because the resistance in most membrane oxygenators requires blood to be pumped through them. Thus a venous reservoir collects venous return from the patient before delivery to the pump, followed by the membrane oxygenator, then back to the patient. Alternatively, bubble oxygenators typically are positioned before the pump because they do not require blood to be pumped through them, and they contain a reservoir of oxygenated blood that supplies the pump before it directly delivers blood back to the patient.

Over the past 4 decades, CPB technology has undergone a dramatic metamorphosis and can no longer be simply viewed as "the pump."4,5 Additional vital components of the CPB circuitry include the arterial line filter, bubble detector, cardioplegia lines, suction lines, temperature monitor, arterial line pressure monitor, anesthetic vaporizer, and oxygen monitor, among others. Needless to say, the perfusionist plays an important role in intraoperative patient management during cardiac surgery.

Pathophysiology of CPB

Substantial controversy surrounds questions regarding "optimal" flow and pressure during CPB.6 Each patient must be individualized, with specific goals determined by numerous factors, including age, comorbidity, temperature, and body mass. However, most clinicians aim for flows in the range of 2.4 to 2.8 L/min/m2 and pressures in the range of 50 to 80 mm Hg. It has been known for many years that CPB induces a systemic inflammatory response syndrome (SIRS) in patients following cardiac surgery that can lead to major organ injury and postoperative morbidity.7,8 Initiation of CPB triggers an extremely complex and multifactorial response involving activation of complement, platelets, neutrophils, monocytes, and macrophages, thus initiating the coagulation, fibrinolytic, and kallikrein cascades. The SIRS response to CPB is further amplified by subsequent stimulated release of various endotoxins and cytokines, including interleukins (ILs) and tumor necrosis factor (TNF), which further promote endothelial cell permeability. Transvascular migration of activated leukocytes into tissues associated with SIRS results in the release of various proteases and neutrophil elastases, which cause additional vascular and parenchymal damage that can exacerbate ischemia-reperfusion injury associated with CPB and cardiac surgery.9 These physiologic changes associated with exposure to CPB are thought to initiate the detrimental changes seen in major organ systems (brain, heart, lung, kidneys) during the postoperative period.

The basic physiologic insults caused by CPB have been associated with major postoperative morbidity, including neurologic dysfunction, pulmonary dysfunction, renal dysfunction, and/or hematologic abnormalities. Additional clinical manifestations associated with SIRS include increased metabolism and fever, fluid retention, myocardial edema, and detrimental hemodynamic alterations. Recent controversies surrounding clinical management of patients exposed to CPB have focused on the potential detrimental physiologic effects of hemodilution,10,11 hyperglycemia,12,13 and use/non-use of colloids14,15 on postoperative morbidity. Over the years, a wide variety of anti-inflammatory treatment options have been used in patients subjected to CPB in hopes of attenuating SIRS, including leukocyte depletion techniques, neutrophil adhesion molecule blockade, heparin coating of CPB circuitry, and use of monoclonal antibodies directed specifically against various inflammatory mediators. Although results from animal work appear promising, the demonstration of a definite clinical benefit in humans has not been consistent.16-19

Management of Anticoagulation

Heparin Dosing

Anticoagulation is used during cardiac surgery to prevent thrombosis of the CPB circuit and to minimize excessive CPB-related activation of the hemostatic system. Heparin is used routinely because it is effective, immediately reversible, generally well tolerated, and inexpensive.20,21 Heparin induces anticoagulation primarily by potentiating the activity of antithrombin III. Most heparin preparations are "unfractionated," meaning that the heparin compound isolated from animal tissues, including porcine intestine and bovine lung, contains heparin molecules of various lengths. Because the relationship between mass (milligrams) and potency (units) varies among heparin preparations, most clinicians record heparin doses in units rather than milligrams. The anticoagulant effect of an intravenous dose of heparin occurs within minutes, and the intravenous heparin bolus dose administered prior to initiation of CPB may decrease arterial pressure and/or systemic vascular resistance.

Much controversy surrounds appropriate heparin dosing and monitoring of anticoagulation in patients exposed to CPB.22,23 Somewhat surprisingly, the literature does not consistently support the importance of anticoagulation monitoring techniques during CPB.24 However, bleeding and transfusion outcomes likely can be improved by refining heparin monitoring techniques, either by sustaining better anticoagulation during CPB or by optimizing reversal of anticoagulation with protamine. Most clinicians administer an initial dose of 200 to 400 U/kg heparin and maintenance doses are determined by the particular laboratory test used for assessment of anticoagulation. The activated clotting time (ACT) is perhaps the most widely used method, and most clinicians administer additional heparin periodically in order to maintain the ACT greater than 400 to 480 seconds. However, numerous other clinical conditions can affect the ACT, including hypothermia, hemodilution, and thrombocytopenia. Heparin concentration measurement techniques also are commonly used. Whatever laboratory test is used, most clinicians assess anticoagulation every 30 minutes during CPB.

Protamine

Protamine neutralizes heparin-induced anticoagulation. Salmon milt provides the pharmaceutical source of protamine, which neutralizes the antithrombin III effect of heparin by forming large complexes with heparin's sulfate groups. Protamine may exhibit antihemostatic properties by affecting platelets and by releasing tissue plasminogen activator from endothelial cells.

Much controversy surrounds the recommended dose of protamine to adequately neutralize heparin. Questions regarding the optimal ratio of milligrams of protamine to units of heparin, and how much heparin remains in the patient, are central to this controversy. Most clinicians use protocols that call for administration of protamine in slight excess doses of the usual ratio of 1 mg of protamine to 1 mg of heparin in order to ensure return of normal coagulation. Subsequent postoperative administration of protamine may be required to prevent "heparin rebound."25

A spectrum of adverse reactions may be associated with protamine use; the most important is hypotension and/or anaphylactoid/anaphylactic reaction.26 Hypotension following administration of protamine is fairly common and may be related to rate of injection, subsequent decreased systemic vascular resistance, and perhaps myocardial depression. True anaphylactoid/anaphylactic reactions to protamine are rare. Patients receiving protamine-containing insulins and/or previously exposed to protamine may be at slightly increased risk of developing anaphylactic reactions. Pulmonary vasoconstriction causing pulmonary hypertension may be associated with protamine use. Heparin–protamine complexes and complement activation may be the primary mediators of these adverse reactions. Slow administration of protamine may limit adverse reactions. If sudden hypotension and/or pulmonary hypertension occur, protamine administration should immediately cease. Heparin may be considered in order to reduce heparin–protamine complex load. Hemodynamic support with vasopressors and inotropes may be required in addition to reinitiation of CPB. Milder cases may resolve without intervention.

Heparin-Induced Thrombocytopenia

Unfractionated heparin given during CPB is remarkably immunogenic. In fact, 25% to 50% of cardiac surgical patients may develop heparin-dependent antibodies postoperatively over the first 5 to 10 days after exposure. This immunogenic response can strongly activate platelets and coagulation, causing the prothrombotic disorder known as heparin-induced thrombocytopenia (HIT).

HIT is defined as thrombocytopenia with a potential for associated thrombosis (HITT). A definitive diagnosis usually requires 1 or more positive tests for HIT antibodies.27 An otherwise unexplained perioperative platelet count fall of 50% or more from baseline or any thrombosis that occurs 5 to 14 days after cardiac surgery is suggestive of HIT, even when heparin is not still being given (ie, delayed-onset HIT). Regarding laboratory testing, commercially available enzyme immunoassays and washed platelet activation assays (ie, platelet serotonin release or heparin-induced platelet activation) are highly sensitive for detecting HIT antibodies such that a negative test result essentially rules out HIT. Antibody seroconversion of no clinical consequence is common following cardiac surgery. Thus the presence of HIT antibodies in the absence of thrombocytopenia or thrombosis does not indicate HIT. However, patients with acute HIT usually have strong positive HIT antibody results. In general, the greater the magnitude of a positive HIT antibody test result, the greater the likelihood the patient has HIT. Following cardiac surgery, the frequency of HIT in patients is approximately 2%. Although HIT antibodies are transient, at least half of patients with HIT develop thrombotic complications. Porcine intestine heparin has been associated with a lower risk of HIT than bovine lung heparin.

If high suspicion of HIT exists, all heparin should be discontinued. In general, patients with HIT or HITT who require urgent anticoagulation should be treated with 1 of the following alternative anticoagulants: lepirudin, argatroban, danaparoid, or bivalirudin.27 Warfarin and other oral anticoagulants are generally contraindicated during acute HIT and should be delayed pending substantial recovery of the platelet count. The alternative anticoagulant should be stopped only when platelet count recovery is complete and therapeutic oral anticoagulation is achieved. For patients strongly suspected of having HIT but without clinical evidence of thrombosis, an alternative anticoagulant in therapeutic doses is recommended because of the high risk of developing thrombosis.

Standard anticoagulation with heparin is recommended for cardiac surgical patients with previous HIT in whom HIT antibodies are no longer detectable (or only weakly detectable) by enzyme immunoassay. In patients with acute or subacute HIT who require cardiac surgery, and in whom the platelet count has recovered yet HIT antibodies remain detectable, 2 general approaches are available. One approach is administration of an alternative anticoagulant (ie, lepirudin, argatroban, danaparoid, bivalirudin) and avoidance of heparin exposure. The other is administration of standard heparin anticoagulation along with a platelet antagonist, including epoprostenol or tirofiban. Given the absence of prospective comparative studies, no single option can be generally recommended in these patients.

Antifibrinolytics

Antifibrinolytic agents, including the lysine analogues ϵ-aminocaproic acid and tranexamic acid, and the serine protease inhibitor aprotinin have become mainstay prophylactic therapies for reducing bleeding and blood product transfusions for cardiac surgical patients.28,29 Although many studies have demonstrated beneficial effects of these agents in reducing perioperative bleeding and morbidity,30 others have reported adverse side effects associated with multiorgan dysfunction and prothrombotic outcomes,31 especially among patients with a genetic predisposition to a hypercoagulable state.32 Aprotinin, in fact, was taken off the market because of these detrimental effects.33 Thus, similar to all pharmacologic agents, the decision to administer any antifibrinolytic should include a thorough risk-to-benefit consideration that is individualized to each patient.34

Management of Patients during Weaning and Separation from CPB

Weaning from CPB should represent a smooth transition from the mechanical pump back to the patient's heart and lungs as the source of blood flow and gas exchange. The process should always be conducted in a coordinated fashion with all members of the team, including the surgeon, anesthesiologist, and perfusionist. Preparation for separation from CPB must be based on a clear understanding of the patient's preoperative condition and events of the operative course. Weaning is initiated after review and adjustment of numerous physiologic and technical variables, including temperature, laboratory data, heart rate and rhythm, myocardial contractility, and mechanical ventilation, among others. The patient should be normothermic and demonstrate normal laboratory values for hemoglobin, potassium, calcium, and acid–base balance. Myocardial function, including heart rate, cardiac rhythm, myocardial contractility, and preload, should be optimized, which may require administration of intravenous medications, depending on the desired goal (Table 52-1). Many patients usually require medications that increase myocardial contractility during separation from CPB. In extreme cases, mechanical devices may be required (eg, intra-aortic balloon pump [IABP], ventricular assist device [VAD]). It now is clear that information obtained from transesophageal echocardiography (TEE) is far superior in quality and quantity to that obtained from a pulmonary artery catheter (PAC) and may be very helpful in guiding therapy toward optimizing hemodynamics and guiding de-airing procedures during weaning from CPB. Furthermore, TEE has proved invaluable in assessing the quality of the surgical procedure following separation from CPB, especially among patients with significant ventricular dysfunction and those undergoing valve procedures, congenital heart surgery, and aortic surgery.35-39

The anesthetic management of patients undergoing coronary artery bypass graft surgery (CABG) requires an understanding of myocardial oxygen supply and demand, patient monitoring, and the anesthetic techniques that provide myocardial protection and favor oxygen delivery over consumption. Myocardial ischemia results when there is an imbalance between the oxygen supply of the coronary circulation and the metabolic demand of myocardial tissue. Ischemia initially leads to contractile dysfunction. However, if ischemia is severe or prolonged, it can lead to cell death, tissue necrosis, and permanent loss of contractile function of the affected myocardial region. This section reviews the basic pathophysiology of myocardial ischemia, the anesthetic management of patients at risk for developing myocardial ischemia, and anesthetic considerations during CABG surgery.

Coronary Artery Anatomy

Coronary artery anatomy is particularly relevant to the anesthesiologist caring for patients with coronary artery disease. The severity of the coronary obstruction correlates with the margin of reserve for tolerating tachycardia and hypotension. Patients with coronary lesions obstructing 99% of the lumen (subtotal occlusion) may not tolerate even mild degrees of tachycardia and hypotension, whereas patients with lesions in the 70% to 75% range may tolerate some degree of hemodynamic compromise before developing ischemia. Knowledge of specific coronary lesions also allows for focused monitoring of targeted myocardial regions at increased risk for myocardial ischemia.

The 2 major coronary arteries are the first arterial branches of the aorta, which arise from 2 of the 3 sinuses of Valsalva in the aortic root (Fig. 52-1). The right coronary sinus is anteriorly located, and the left coronary sinus is located laterally and slightly posterior. The left coronary artery (LCA) divides into the left anterior descending coronary artery (LAD) and left circumflex artery (LCx). The LAD gives rise to the diagonal branches and supplies the anterior wall of the right ventricle (RV), the anterior two-thirds of the interventricular septum, the anterior wall of the left ventricle (LV), and the ventricular apex. The LCx gives rise to the obtuse marginal branches and supplies the left atrium (LA), and the posterior and lateral walls of the LV. Patients are described as having "a left main" if they have a significant lesion in the LCA. These patients are at particular risk for developing myocardial ischemia that affects a large portion of the LV, which would cause rapid hemodynamic compromise and cardiac arrest. Patients described as having a "left main equivalent" have high-grade obstructions in both the LAD and LCx arteries. These patients potentially have the same risk of coronary ischemia and rapid hemodynamic compromise as patients with left main disease.

The right coronary artery (RCA) supplies blood to the lateral and posterior walls of the RV, the inferior wall of the LV, and the posterior third of the interventricular septum. The RCA terminates as the posterior descending artery (PDA) in 85% to 90% of the population. The blood supply of the PDA determines the pattern of coronary dominance: RCA for right dominant and LCx for left dominant. Most patients have an RCA-dominant or balanced pattern of blood supply to the PDA. A balanced pattern is used to describe coronary anatomy with no particular dominance in terms of the blood supply to the PDA. The presence of a right-dominant or balanced system during OPCABG frequently leads to bradycardia and hypotension during RCA occlusion because it supplies blood to the sinoatrial node in 55% of patients, whereas the PDA supplies the atrioventricular node in right-dominant patients.

Myocardial blood collects in the coronary veins, which drain into the coronary sinus, and subsequently the right atrium (RA). The coronary sinus is often used as a conduit for the delivery of cardioplegia to the myocardium in a retrograde fashion. This is possible because the coronary veins lack valves, allowing blood to flow in either direction. Retrograde cardioplegia is used in patients with aortic insufficiency (AI), during aortic valve (AV) surgery, in the presence of high-grade coronary obstructions, and in patients with previous CABG who have a patent internal mammary artery graft. All of these situations limit the antegrade delivery of cardioplegia to the myocardium from the aortic root. Retrograde delivery of cardioplegia via the coronary sinus also has limitations. A coronary sinus catheter inserted beyond the small cardiac vein will prevent delivery of cardioplegia to the RV. Subsequently, the right heart may be poorly protected from myocardial ischemia during aortic cross-clamping.40 Coronary sinus catheters that are directed into the small or middle cardiac vein may cause coronary sinus rupture.

The LAD is positioned on the anterior aspect of the heart. The obtuse marginal branches of the LCx and diagonal branches of the LAD are located on the lateral or posterior aspect of the heart (see Fig. 52-1). The surgeon must rotate or lift the heart to gain access to these vessels, causing hemodynamic compromise if the cardiac chambers are compressed. The pulmonary outflow tract also may be compressed with cardiac rotation, again resulting in severe hypotension by dramatically reducing preload. Additionally, the electrocardiographic (ECG) axis changes during cardiac manipulation and may change the ECG–coronary artery anatomic relationship, limiting the ability to use ECG for ischemia monitoring. Use of TEE for ischemia and ventricular function monitoring may be compromised when the heart is lifted or manipulated within the surgical field. The PAC may be malpositioned during these maneuvers, providing erroneous data regarding cardiac output and filling pressures.

Coronary Blood Flow Physiology

The physiology of coronary blood flow is based on the assumption that coronary arteries are nondistensible tubes, and that blood is a homogeneous fluid. Blood flows from a region of higher pressure (aorta) to one of lower pressure (capillaries). The rate of flow is dependent on the pressure gradient that moves red blood cells through the coronary arteries in a laminar flow pattern. Flow is most dependent on the radius of the blood vessel, which is why coronary arterioles maximally dilate in response to coronary arterial stenosis.41 Coronary stenosis causes the vessel to maximally dilate distal to the stenosis, creating a vessel with a fixed radius. Manipulation of the coronary perfusion pressure then becomes the most important factor that determines coronary blood flow and the most important physiologic parameter manipulated by the anesthesia provider in the setting of coronary ischemia. Exercise-induced ischemia causes a compensatory increase in heart rate, α1-adrenergic–induced vasoconstriction, and increased LV filling pressure, all of which reduce coronary blood flow.42 Adequate coronary perfusion pressure is the most important factor in the prevention and treatment of myocardial ischemia, both in the operating room and during exercise.

A second important factor is blood viscosity, determined rheologically as the concentration or suspension of erythrocytes within the blood. Patients with coronary artery disease (CAD) have disturbed blood flow patterns that create an increased tendency for coronary arterial thrombosis.43 These abnormal flow patterns underscore the importance of aspirin therapy in the medical management of patients with CAD by reducing platelet adhesion and aggregation at the site of coronary stenoses.

Coronary blood flow has a characteristic phasic perfusion pattern, 70% to 80% of which occurs during the diastolic phase of the cardiac cycle. Cardiac contraction during systole impedes myocardial perfusion by increasing intraventricular cavitary pressure and coronary artery resistance, thus producing a nonlinear relationship between heart rate and diastolic time.44 β-Blockade is a very effective medical therapy for patients with CAD by preventing even small increases in heart rate during the perioperative period, reducing mortality with heart rate reduction, and improving outcome.45

Heart rate reduction improves subendocardial coronary artery blood flow,46,47 allowing for better matching of myocardial oxygen supply and demand (myocardial perfusion–contraction coupling), thus preserving regional myocardial contractility.48 Recovery of stunned or hibernating myocardium in patients with ischemic heart disease occurs with restoration of myocardial perfusion–contraction coupling.

Myocardial Oxygen Delivery

Myocardial oxygen delivery depends on the oxygen content in the blood and is composed of hemoglobin-bound oxygen and dissolved oxygen. Hemoglobin-bound oxygen composes most of the blood's carrying capacity. However, delivery of oxygen to myocardial cells is dependent on release of oxygen from hemoglobin and is represented by the oxygen–hemoglobin dissociation curve. A leftward shift of this curve from normal indicates a greater affinity of oxygen by hemoglobin, which has the effect of drawing more oxygen into the blood as it passes through the lungs but reducing oxygen release at the cellular level. A leftward shift is caused by alkalosis (both metabolic and respiratory), hypothermia, carboxyhemoglobinemia, methemoglobinemia, and decreased red blood cell 2,3-diphosphoglycerate (DPG), which may be observed after transfusion of a large volume of old blood stored in acid-citrate-dextrose. A rightward shift indicates less affinity of the red cells for oxygen, which has the effect of greater oxygen release to the tissues but at the expense of drawing less oxygen into the blood as it passes through the lungs. A rightward shift is caused by acidosis (both metabolic and respiratory), hyperthermia, and increased 2,3-DPG in the red blood cells.

Although anemia clearly reduces the oxygen-carrying capacity of blood, clinical studies have not determined the lowest acceptable level of anemia that does not produce myocardial ischemia. The degree of anemia that produces myocardial ischemia is dependent on factors that are specific for each patient and the loading conditions of the ventricle. Factors include the severity of CAD, myocardial wall thickness and tension, heart rate, and perfusion pressure. Isovolemic reduction in hemoglobin to 4.6 to 5.3 mg/dL in healthy volunteers produced ST-segment changes on Holter monitoring in 2 of 11 subjects, whereas hemoglobin levels of 5.0 mg/dL led to myocardial ischemia infrequently.49 In another investigation, ECG ST-segment changes suggestive of ischemia occurred in only 3 of 55 subjects during acute reduction in hemoglobin concentrations to 5 g/dL. These authors attributed the imbalance of myocardial oxygen supply and demand to tachycardia.50 Although excessive hemodilution (median lowest hematocrit below 25%) during CPB is a risk factor for major morbidity even in the absence of blood transfusion, preoperative anemia may not be associated with an increased risk of morbidity, provided that the lowest hematocrit during cardiopulmonary bypass is maintained above 28%.12

Patients with ischemic heart disease require a higher hemoglobin concentration to minimize perioperative complications. A higher incidence of postoperative mortality was found in cardiac surgical patients older than 75 years whose preoperative systemic oxygen delivery was less than 320 mL/min/m2 and who had anemia on the second postoperative day.51 Bracey et al found no increase in morbidity, mortality, or patient self-assessment of fatigue when the hemoglobin threshold for red cell transfusion was lowered to 8.0 g/dL after CABG.52 Therefore, the lower limit of hemoglobin concentration depends on multiple factors, such as the patient's age, heart rate, perfusion pressure, clinical evidence of myocardial ischemia, and success of coronary revascularization. Moderate exposure to allogenic blood products is not associated with reduced long-term survival after coronary artery bypass surgery.53

Myocardial Oxygen Demand

Myocardial oxygen demand is primarily determined by heart rate, ventricular wall tension, and myocardial contractility. Acting on the myocardial β receptors, β-blockers decrease heart rate and reduce contractility and thus are the primary treatment for patients with CAD at risk for myocardial ischemia. Although treatment with atenolol in the perioperative period does not significantly alter the neuroendocrine stress response,54 perioperative β-blockade reduces mortality and myocardial infarction (MI) among patients undergoing noncardiac surgery.55-57 Patients treated with metoprolol for whom ventricular filling pressures were unchanged had an up to 40% reduction in myocardial oxygen consumption.45 Metoprolol also improves survival, patient well-being, and New York Heart Association (NYHA) functional class in patients with congestive heart failure (CHF).58,59 β-Blockade and subsequent heart rate reduction clearly decrease myocardial oxygen consumption and increase oxygen supply in patients with CHF. Although most anesthesiologists recognize the benefits of perioperative β-blockade, few institutions have formalized protocols for administering β-blockers to surgical patients.60

Preload

Preload is the ventricular volume at end-diastole that determines myocardial fiber length, which in part determines the force of ventricular contraction. Manipulation of preload is an important therapeutic option in the care of patients with myocardial ischemia. Nitroglycerin is commonly used for treatment of myocardial ischemia, primarily exhibiting its antianginal effect by preload reduction through venodilation and coronary vasodilation. Morphine is useful for treating myocardial ischemia by causing vasodilation (preload reduction) and providing pain relief, thereby leading to reduced heart rate. Furosemide reduces preload through both its diuretic action and venodilation.61

Determination of end-diastolic volume is problematic in the clinical arena. Pulmonary artery occlusion pressure (PAOP) is commonly used for approximating LV end-diastolic volume (LVEDV), but multiple assumptions must be made in order to use PAOP to estimate preload. Use of a pressure measurement to estimate volume must take into account LV compliance, which is the change in unit pressure for each change of unit volume. Both LVEDV and the compliance of the myocardium determine the LV end-diastolic pressure (LVEDP). Myocardial ischemia decreases ventricular compliance, thereby increasing LVEDP for the same LVEDV. This change may be reflected in PAOP measurements, allowing the use of PAOP as an ischemia monitor.

Other factors that affect the relationship between PAOP and LVEDP are mitral stenosis (MS), LA compliance, and intrathoracic pressure. Although the severity of MS does not change over the course of an anesthetic or surgical procedure, the pressure gradient between the LA and LV is dynamic, dependent on the cardiac output, heart rate, and flow though the mitral valve (MV) during the diastolic phase of the cardiac cycle. Tachycardia decreases diastolic time, thereby increasing flow through the MV during diastole and thus increasing the pressure gradient between the LA and LV. An increase in PAOP in this circumstance is due to an increase in the pressure gradient across the MV rather than an increase in preload. Actual LV preload may be reduced because tachycardia impedes LV filling in patients with MS. When PAOP is used for ischemia monitoring, trends in pressure changes must be taken in the context of these other hemodynamic variables in patients with MS.

Figure 52-2 shows a normal central venous pressure waveform and the relationship of the waveform with the ECG. PAOP appears similar but reflects LA pressure rather than RA pressure. The presence of large, prominent V waves is indicative of mitral regurgitation (MR), which can occur with ischemic papillary muscle dysfunction. This acute increase in LA volume decreases the compliance of the LA and pulmonary veins.

Large changes in intrathoracic pressure also affect PAOP.62 During spontaneous inspiration, mean PAOP declines because of the decrease in intrathoracic pressure. Positive-pressure ventilation causes increased intrathoracic pressure, which is reflected in the pulmonary venous pressure. Measurement of PAOP is made at the end of expiration to minimize the effects of inspiration on intrathoracic pressure and the pulmonary vasculature.

Afterload

Afterload is impedance to ventricular ejection and is best described with pressure–volume loops. The ratio of end-systolic pressure to stroke volume defines the elastance of the arterial tree. In the absence of aortic stenosis (AS), this is primarily determined by arterial vasculature tone. Afterload conditions that allow more fiber shortening allow greater metabolic efficiency and reduced oxygen consumption.63 The clinician can manipulate afterload by changing the size or radius of the LV through preload manipulation or more commonly by affecting systemic vascular resistance or blood viscosity. Although systemic vascular resistance is only 1 component of afterload, it is the only factor that is easily measured and readily changed clinically.

Contractility

Inotropy describes the contractile state of the ventricle and is measured using either the ejection or isovolumic phase of ventricular contraction. Pressure–volume loops consisting of ejection, relaxation, and isovolemic pressure are drawn under different loading conditions. The slope of the serial end-systolic pressure–volume relationship describes the myofibril contractile state independently from preload or afterload (Fig. 52-3). Simple clinical tools for measuring contractility independent of preload and afterload do not exist; however, some investigators have used simultaneous pressure and echocardiography area relationships to provide measurements of contractility.64

Anesthetic Considerations for Patients Undergoing CABG

Patients with CAD presenting for CABG require special considerations in their anesthetic management. First and foremost are techniques that minimize myocardial oxygen demand while maximizing myocardial oxygen delivery, as described in previous sections. These considerations include preoperative preparation, intraoperative monitoring, and the use of anesthetic agents with hemodynamic effects that favor oxygen supply over demand and allow for myocardial protection. Postoperative management that provides particular attention to pain management, temperature control, and attentive hemodynamic monitoring to prevent tachycardia, hypotension, and hypertension, must also be considered. Many of these considerations apply to patients undergoing CABG, with or without CPB. However, certain considerations apply to patients who undergo CABG with CPB, although others apply if coronary revascularization is performed without CPB (ie, OPCABG).

Preoperative Evaluation

Most patients presenting for CABG undergo extensive preoperative testing of their cardiac disease and other medical problems. Medical conditions that predispose patients to the development of CAD also affect other organ systems. They include smoking, hypertension, hypercholesterolemia, diabetes, obesity, and advanced age. It is particularly important to identify comorbidities that may prolong or complicate the patient's postoperative course. One comorbidity that deserves particular attention is respiratory insufficiency. Patients with CAD often complain of dyspnea due to ventricular dysfunction caused by myocardial ischemia. Dyspnea related to myocardial ischemia should resolve with adequate revascularization. However, patients with a long-standing smoking history who develop underlying pulmonary disease may not derive the same benefit after CABG, but instead experience exacerbation of their pulmonary disease and require prolonged postoperative mechanical ventilation and suffer pulmonary complications after surgery.

Diabetes is a progressive metabolic disorder that leads to the development of CAD, cerebral vascular disease, neuropathy, nephropathy, and retinopathy. An important aspect of diabetes in the patient with CAD is that myocardial ischemia may occur in the absence of classic symptoms. Patients may suffer from ischemic episodes without therapy, leading to potentially irreversible cardiac damage. They may have prolonged recovery from CABG surgery because of the increased need for inotropic support. Tight control of glycemic blood levels during the perioperative period is important for reducing the incidence of wound infection following cardiac surgery.

Hypertension is another medical condition common in patients with CAD. It also is a risk factor for cerebral vascular disease, renal failure, and CHF. This condition may be asymptomatic, yet it carries the risk of end-organ damage if untreated for prolonged periods. Patients with uncontrolled hypertension often are more difficult to manage because of wide swings in blood pressure associated with events such as sternotomy and pericardotomy. Their vasculature is more responsive to catecholamines causing an exaggerated response, and the relatively volume contracted state leads to exaggerated hypotension in response to vasodilator therapy. Intraoperative blood pressure variability is associated with 30-day postoperative mortality in patients undergoing coronary bypass surgery,65 and an increased perioperative pulse pressure is associated with poor long-term survival after CABG.66

Renal disease is associated with cardiac surgical patients, especially in those with hypertension and diabetes. Development of renal failure after cardiac surgery is a concern, which occurs in 0.9% of CABG patients and 2.0% of patients undergoing valve procedures. More importantly, operative mortality has been reported to be 63.7% in patients who develop acute renal failure versus 4.3% for patients who do not develop renal failure.67 The risk of postoperative MI, reoperation for bleeding, and mediastinitis is higher in patients who develop postoperative renal failure.68

Monitoring

Continuous ECG monitoring, along with blood pressure measurement, pulse oximetry, and end-tidal carbon dioxide (CO2) analysis are standard monitors for all anesthetized patients. It is preferred that cardiac surgical patients have a display monitor that allows viewing of 2 ECG leads simultaneously with automated ST-segment analysis. ECG limb leads II and V5 allow for monitoring of myocardial regions supplied by the RCA and LAD coronary arteries, respectively. Automated ST-segment trending has only moderate sensitivity and specificity (75% for both) in detecting changes found by off-line Holter monitoring69 but is a marked improvement over mere observation by the clinician, whose attention must also be directed at providing anesthetic care. Monitoring 2 ECG leads simultaneously increases the sensitivity of detecting ischemia to 80% if leads II and V5 are monitored, 82% if leads II and V4 are monitored, and 90% if leads V4 and V5 are monitored (Fig. 52-4).70 Factors such as LV hypertrophy, cardiac conduction changes, electrolytes, and drugs such as digitalis all can affect the interpretation of ST segments,71 but the primary concern is acute ST-segment changes that occur during the perioperative period (Fig. 52-5).

All cardiac surgical patients require invasive arterial blood pressure monitoring. Although the radial artery is used at many institutions, consideration must be given to the possibility of harvesting the radial artery as a conduit for CABG. When used for this purpose, the radial artery is harvested from the patient's nondominant hand, so radial artery cannulation should be performed in the dominant hand. Femoral artery cannulation can be used and is preferred by some clinicians and institutions. The benefits of using femoral artery cannulation include a better correlation with mean arterial pressure in the immediate post-CPB period and access to the femoral artery if an IABP must be inserted. The intra-arterial catheter can be placed prior to or immediately after induction, but preinduction placement allows the clinician to respond more rapidly to the changing hemodynamic conditions that often occur during induction of anesthesia. Preinduction placement of the arterial catheter is preferred in the setting of a potentially difficult airway and for patients with a greater risk of rapidly changing hemodynamics (left main disease or severe ventricular dysfunction).

All cardiac surgical patients should have central venous access for the purpose of administering important vasoactive medications into the central circulation and for assessing volume status. The issue of whether to place a central venous pressure (CVP) catheter versus a PAC is controversial. A PAC measures pulmonary artery (PA) pressure and provides a means for sampling mixed venous oxygen saturation (SvO2). PA pressure is not easily determined with TEE, and SvO2 cannot be obtained with a CVP catheter or by TEE. Some studies have not demonstrated improved cardiac surgical outcomes with the use of a PAC.72 However, prolonged pre-CPB pulmonary hypertension and post-CPB elevation of PA diastolic pressure are predictors for the development of perioperative MI.73 Early treatment of pulmonary hypertension is presumed to improve outcome.

The American Society of Anesthesiologists (ASA) Task Force on Pulmonary Artery Catheterization has provided practice guidelines for PAC insertion (Table 52-2).74 PA pressure monitoring is indicated based on the medical condition of the patient or the nature of the surgery (Table 52-3) but is not recommended when the patient, procedure, and practice setting each poses a low risk for hemodynamic complications.74 Cardiac surgical patients with unstable angina, recent MI, active CHF, severe CAD or valvular heart disease, and severe pulmonary or renal disease should have PA pressure monitoring. However, routine PA catheterization is not necessary in all patients undergoing CABG. Outcome after CABG is not influenced by routine use of a PAC, suggesting its use can be delayed until a clinical need develops.72

Right bundle-branch block occurs in approximately 3% of patients undergoing PA catheterization.75 For this reason, placement of a PAC in patients with a previous left bundle-branch block is not recommended unless precautions are taken for managing the patient should complete heart block occur. A decision-making algorithm for inserting a PAC in patients with a preexisting left bundle-branch block is shown in Fig. 52-6.

TEE can be used to assess global and regional myocardial contractility, volume status, and valvular function during cardiac surgery. Although TEE has revolutionized the intraoperative assessment and management of patients undergoing cardiac surgery, few data suggest that routine use of TEE for all patients with normal ventricular function undergoing elective CABG improves outcome. Regional wall-motion assessment by qualitative inspection of radial shortening and wall thickening is subjective. Accurate diagnosis is dependent on observer experience and having the ischemic myocardial segment in the imaging plane during the ischemic episode. More sophisticated techniques such as computerized digitations, color kinesis, and tissue Doppler imaging may allow for better assessment, but these techniques are not readily available or familiar to all users. Nonetheless, information obtained from TEE is far superior in quality and quantity to that obtained from a PAC and may be very helpful in guiding therapy toward optimizing hemodynamics in the perioperative period. Furthermore, TEE has proven invaluable for assessing the quality of the surgical procedure following separation from CPB, especially among patients undergoing CABG surgery who have significant ventricular dysfunction and those undergoing concurrent valve procedures, congenital heart surgery, and/or aortic surgery.35-39

Near-infrared reflectance spectroscopy has been used to demonstrate a correlation between low bifrontal regional cortical oxygen saturation and cognitive dysfunction, prolonged hospital stay, and perioperative stroke. Intraoperative cerebral oxygen desaturation is associated with an increased risk of cognitive decline and prolonged hospital stay after coronary artery bypass surgery.76 Thus some clinicians have advocated monitoring of cerebral oximetry in CABG patients as a technique for preventing profound cerebral desaturation and associated morbidity.77

Anesthetic Induction

Induction of general anesthesia for CABG patients can be accomplished using a variety of medications, provided the goals of preventing tachycardia, hypotension, and hypertension are achieved. The worst combination of hemodynamic perturbations is tachycardia with hypotension. Although hypertension increases ventricular wall tension and therefore myocardial oxygen demand, it also increases coronary perfusion pressure. Myocardial depression and the vasodilating effects of the anesthetic agents are the most important considerations in patients with severely impaired ventricular function or valvular heart disease. Laryngoscopy and intubation are the first intraoperative events after induction of anesthesia that test the effects of the anesthetic technique used. Coincident with these events is the institution of positive-pressure ventilation, which may result in hypotension among hypovolemic patients or in patients with air trapping due to pulmonary disease.

Patients with CAD benefit from a narcotic-based technique because of the lack of myocardial depression, a tendency for decreased heart rate, and an attenuated response to laryngoscopy and intubation. Fentanyl probably is the most frequently used opioid in cardiac surgery, although sufentanil is also a reasonable choice. The pre-bypass addition of remifentanil to a fentanil and propofol anesthetic reduces the release of biochemical markers of myocardial damage to patients undergoing elective on-pump coronary artery bypass grafting. This benefit may be attributed to either remifentanil itself or to an overall increased opioid dose.78 Use of remifentanil and sufentanil during induction has been associated with a high incidence of bradycardic/asystolic complications in patients who have been treated with β-blockers and calcium channel blockers.79,80 Many patients with CAD are likely to be taking 1 or both types of drugs. A nondepolarizing muscle relaxant devoid of cardiovascular side effects is used to facilitate intubation. Pancuronium can be used with a large dose of narcotic because the tachycardic side effects of pancuronium are balanced by the bradycardic side effects of the large-dose narcotic. Use of vecuronium or rocuronium is prudent when the initial narcotic dose is smaller because the bradycardic effects of the narcotic are not as pronounced. Pancuronium use with only a small narcotic bolus may result in undesirable tachycardia.

Modern practices that focus on early extubation and "fast-tracking" the patients through the postoperative period use smaller narcotic doses, with supplementation by short-acting hypnotic agents such as midazolam. Dexmedetomidine can also be used as an adjunct to anesthetic induction to attenuate the hemodynamic response to tracheal intubation, which is particularly important when low-dose fentanyl is used for induction.81 An intraoperative awareness protocol should be used to minimize the incidence of intraoperative recall, which is more of a concern with a low-dose narcotic-based anesthetic technique. A low concentration of a potent inhalational agent or intravenous propofol is used for hypnosis and amnesia. Reduced doses of opioids often are administered with small doses of hypnotic drugs such as benzodiazepines, thiopental, propofol, or etomidate to promote early postoperative extubation. Care must be taken when hypnotics are administered concomitantly with opioids. Mean arterial pressure can drop precipitously in hypovolemic patients. Midazolam significantly decreases blood pressure and increases heart rate when used as an induction agent.82 Propofol causes venodilation and can profoundly decrease blood pressure.83 A vasoactive drug such as ephedrine or phenylephrine should be readily available to treat hypotensive episodes during the induction of anesthesia. Although regional anesthesia for cardiac surgery may provide superior postoperative analgesia, shorter postoperative ventilation, reduced incidence of supraventricular dysrhythmias, and lower rates of perioperative MI, most clinicians use general anesthesia during CABG procedures.84

The same hemodynamic goals must be achieved for the patient with a difficult airway who requires an awake/sedated fiberoptic intubation prior to general anesthesia. The patient's airway should be well anesthetized with local anesthetics and the patient sedated with short-acting intravenous agents that blunt the response to laryngoscopy and intubation. An inhalational induction also might be considered in some patients. Alternatively, an intubating laryngeal mask airway can be used to secure the difficult airway under general anesthesia. The particular technique or approach is individualized to the patient's anatomy and medical condition, and the experience of the clinician in using 1 or more of these techniques.

Pre-CPB Management

Once the airway is secured and general anesthesia induced, anesthesia care is directed toward maintaining a stable blood pressure and heart rate. If access to the central circulation was not obtained before induction of anesthesia, cannulation of the internal jugular or subclavian vein is performed, and a PAC is inserted if indicated. The right internal jugular vein is often used because of its predictable location, accessibility, and direct route into the RA. The patient is placed in the Trendelenburg position during central venous cannulation, which increases preload and helps maintain blood pressure at an acceptable range following the induction of anesthesia, in the absence of surgical stimulation. Use of surface ultrasound for guiding the insertion and placement of a central venous cannula has become more popular.85 Blood pressure frequently declines as the operating room table is leveled following central venous cannulation. A fluid bolus or small doses of vasoactive drugs may be required.

Additional tasks for the anesthesia provider before surgical incision include placement of a TEE probe, antibiotic administration, determination of a baseline ACT, and infusion of antifibrinolytic drugs. Although many studies have demonstrated the beneficial effects of antifibrinolytic agents in reducing perioperative bleeding and morbidity,30 others have reported adverse side effects associated with multiorgan dysfunction and prothrombotic outcomes,31,32 especially among patients with a genetic predisposition to a hypercoagulable state.33 Thus, similar to all pharmacologic agents, the decision to administer any antifibrinolytic should include a thorough risk-to-benefit consideration that is individualized to each patient.

Patients must be anesthetized to a depth that avoids a tachycardic and hypertensive response to surgical incision. Small doses of short-acting agents such as esmolol (50-100 mg) or nitroglycerin (50-100 mcg) can be administered, or the anesthetic depth can be deepened with additional narcotic or potent inhalational agent. The patient's response to skin incision is a gauge to his or her subsequent response to the more stimulating event of median sternotomy. If the anesthetic depth was not adequate for skin incision, then additional narcotic or a bolus of esmolol should be administered before sternotomy. The lungs typically are deflated during sternotomy to prevent the sternal saw from cutting the lung parenchyma. Another complication that can occur during sternotomy is accidental tearing of the innominate vein or the RV. Sternotomy is especially hazardous during repeat CABG, where adhesions from previous surgery placed the heart immediately posterior to the sternum. Banked packed red blood cells must be immediately available at the time of sternotomy, particularly for patients at increased risk for this complication.

Maintenance of anesthesia during the pre-CPB period can be accomplished with a variety of techniques. Isoflurane–fentanyl anesthesia and propofol–fentanyl anesthesia both are acceptable techniques for maintaining anesthesia during CABG.86 Postoperative troponin release, cardiac morbidity, and mortality were similar between anesthetic regimens consisting of propofol-opioid versus isoflurane-opioid in patients undergoing CABG surgery.87 Use of sevoflurane and desflurane has been shown to result in shorter intensive care unit (ICU) and hospital length of stay.88 This finding seemed to be related to better preservation of early postoperative myocardial function. Considerable research has recently focused on the myocardial protective effects of potent inhalational anesthetics. Isoflurane protected the myocardium during ischemic episodes in an experimental model.89,90 Isoflurane and other volatile anesthetics may mimic the protective effects of a process called ischemic preconditioning.91 Brief periods of ischemia activate the protein kinase C–mediated pathway that confers cardioprotection during subsequently longer periods of ischemia.92 Sevoflurane decreases the inflammatory response after CPB, as measured by the release of IL-6, neutrophil β-integrins (CD11b/CD18), and TNF-α. Myocardial function after CPB, as assessed by regional wall motion and LV stroke work index, also is improved with sevoflurane.93 Both halothane and isoflurane have been shown to provide significant preservation of adenosine triphosphate (ATP) levels during ischemia, but this preservation did not improve hemodynamic recovery.94 In a comparative study in which patients received either propofol or inhalational anesthesia during CABG surgery, patients who received inhalational anesthesia (sevoflurane or desflurane) but not propofol had preserved LV function after CPB, with less evidence of postoperative myocardial injury.95 The cardioprotective effects are clinically most apparent when the inhalational agent is administered throughout the operation.96 Although recent evidence suggests that remifentanil may play a similar role in protective preconditioning, the effect may be attributable to an overall increased opioid dose rather than to remifentanil itself.79

Important activities that occur during the pre-CPB period include surgical dissection of the left internal mammary artery, opening of the pericardium, and placement of the aortic and venous cannulas in preparation for CPB. Sternal retraction that exposes the internal mammary artery may affect the function of intravenous and arterial catheters placed in the ipsilateral arm. This most commonly affects the left arm with harvesting of the left internal mammary artery. Papaverine is commonly injected into the ligated internal mammary artery to prevent vasospasm but may cause hypotension. This decrease in blood pressure usually is brief but may require treatment with a small dose of phenylephrine. Hypertension may develop during surgical manipulation of the pericardium and aorta as a result of sympathetic activation. A small bolus of narcotic or esmolol can blunt this hypertensive response.

Heparin is administered prior to ligation of the internal mammary artery (if used) and cannulation of the aorta to prevent thrombus formation. Heparin is a polyanionic mucopolysaccharide that increases the rate of anticoagulant effect of antithrombin III on factors II, X, XI, XII, and XIII. It usually is given in a dose range of 300 to 400 U/kg to achieve ACT greater than 400 seconds. The adequacy of anticoagulation must be determined prior to initiating CPB. For patients undergoing OPCABG, some surgeons administer a smaller dose of heparin, with the goals of achieving ACT greater than 300 seconds.

CABG Without CPB

CABG without CPB (OPCABG) was first reported during the 1960s and 1970s.97,98 Advancements in the safety of CPB with better equipment and techniques allowed surgical access to more distal coronary target sites and a quiet surgical field, thereby eliminating the need for CABG without CPB.1,2 Developments in mechanical stabilization devices during the 1990s renewed the interest among surgeons to return to this technique in order to avoid the deleterious effects of CPB, particularly as the age of the surgical population increased to include more elderly patients with calcified aortas. OPCABG also allows faster recovery and fewer ICU days, providing an economic incentive to using this technique.

The anesthetic management of patients undergoing OPCABG encompasses the same hemodynamic goals as the pre-CPB management of patients undergoing CABG with CPB, but the goals are more difficult to achieve, particularly when distal coronary anastomoses are being performed. Myocardial protection with the use of inhalational agents remains an important consideration. Compared to patients who receive propofol, patients receiving sevoflurane have less myocardial injury in the first 24 postoperative hours.99 Similarly, patients who received isoflurane during OPCABG had lower troponin-T leakage than patients who received a propofol infusion.100 OPCABG requires more attention, vigilance, and intervention on the part of the anesthesia provider while the surgeon is performing the distal coronary anastomoses. Maintenance of perfusion pressure, cardiac output, and normothermia, while avoiding profound myocardial ischemia, is challenging during surgical manipulation that produces ventricular compression and temporary coronary occlusion. Ischemia monitoring is compromised by cardiac displacement that alters ECG polarization and affects the anatomic relation with a TEE probe. Vasoactive medications and the Trendelenburg position are used to maintain blood pressure and cardiac output. Despite the best efforts, however, decreases in cardiac output with elevations in central venous pressure and PAOP are commonly observed during surgical manipulation. Cardiac dysrhythmias are not uncommon and are treated by increasing the perfusion pressure and using medications such as lidocaine, amiodarone, and magnesium. Sevoflurane anesthesia is associated with less atrial fibrillation or supraventricular arrhythmias after OPCABG surgery than an equivalent dose of desflurane.101 Malignant dysrhythmias that are not corrected by medications or electrical cardioversion may prompt the need for CPB. Temporary cardiac pacing may be required in patients with right coronary dominant anatomy because of bradycardia or cardiac arrest during right coronary occlusion. Upon successful completion of all distal coronary anastomoses, the blood pressure is lowered before application of the partial aortic clamp for the proximal anastomoses. Release of this clamp after all of the proximal grafts are completed may produce reperfusion dysrhythmias or send air through the coronary grafts, if they were not adequately de-aired by the surgeon.

Anticoagulation is reversed with protamine after the surgeon is satisfied with the grafts and the absence of surgical bleeding. The surgical wound is closed, and the patient is transported to the ICU. Some patients may be candidates for extubation immediately after surgery, but this should be individualized according to the patient's temperature, need for inotropic and mechanical support of the circulation, coexisting medical problems, and degree of mediastinal bleeding.102 Although extubation within 2 to 4 hours after surgery is a reasonable goal for most patients, immediate extubation is possible after OPCABG using either opioid-based or thoracic epidural-based anesthesia.103 Thoracic epidural analgesia may be of particular benefit in obese patients (>30 kg/m2 body mass index) by providing early tracheal extubation and shorter ICU stays.104 The benefits of thoracic epidural analgesia must obviously be weighed heavily against the risks associated with anticoagulation and appropriate management of the epidural catheter.

CABG with CPB

For patients undergoing CABG with CPB, a 2-stage cannula is placed in the RA to direct blood away from the patient to the CPB circuit, and a cannula is placed in the aorta to return oxygenated blood to the patient's circulatory system. The patient must be fully heparinized before cannulation to avoid thrombus formation. The arterial blood pressure is lowered to 85 to 90 mm Hg systolic pressure prior to aortic cannulation to reduce aortic wall tension and minimize the risk of aortic dissection as the aortic wall is punctured. The surgeon may request hand-bag ventilation to provide better visualization of the RA for insertion of the venous cannula. Atrial fibrillation may develop during placement of the venous or coronary sinus catheters. Cardioversion with internal paddles may be required if the patient becomes hypotensive prior to CPB.

Some institutions "retrograde prime" the CPB circuit by allowing the patient's blood to displace the clear fluid priming volume from the aortic and venous lines to the CPB machine. This process nearly always causes hypotension due to volume depletion and requires vigilance and treatment by the anesthesia provider to avoid profound hypotension. Small boluses of phenylephrine are effective for raising blood pressure during this process. Once the retrograde priming is complete, pump volume is infused through the aortic cannula and the venous cannula is unclamped, thereby initiating CPB.

Cardiopulmonary Bypass

Mechanical ventilation of the lungs is no longer necessary for oxygenation and ventilation during CPB because the bypass machine provides these physiologic functions. Mechanical ventilation of the lungs is usually discontinued when a calculated circulation flow through the bypass machine is achieved. Although some have suggested that continued pulmonary ventilation during CPB might mitigate reperfusion injury to the lungs, recent evidence in humans did not demonstrate a significant difference in pulmonary vascular resistance between patients mechanically ventilated during CPB versus those who were not.105 Dexamethasone may have a beneficial effect on A-a oxygen gradient, respiratory index, and PaO2/FiO2 ratio at 12 and 24 hours, but no effect on extubation time or lung compliance following CPB.106

The surgical field is observed for complications related to cannulation such as venous obstruction, aortic dissection, bleeding, and cardiac distension. Anesthesia is maintained by volatile anesthetics administered through a vaporizer on the CPB machine or by continuous infusion of hypnotic drugs such as propofol, titrated according to mean arterial blood pressure and readings on the awareness monitor. Some clinicians routinely administer hypnotics with initiation of CPB and during rewarming to reduce the incidence of intraoperative awareness. Additional muscle relaxation helps reduce oxygen consumption by minimizing shivering as the patient is cooled. ECG, systemic and pulmonary pressures, urine output, and temperature are monitored. PA pressure may increase with cardiac distension or because the PAC has migrated to more peripheral regions in the lung during cardiac manipulation. In the absence of cardiac distension, the PAC should be withdrawn until the pressure measured in the distal port decreases.

The clinician should use the CPB time to prepare for post-CPB events. Most patients with good ventricular function do not require inotropic support following successful coronary revascularization. However, patients with poor preoperative ventricular function and those who undergo complicated surgical procedures and prolonged prebypass ischemia often require inotropic or IABP support in the postbypass period. Prior to termination of CPB, the surgeon places epicardial atrial and ventricular pacing leads, to be used as needed to establish an adequate cardiac rhythm and rate.

Post-CPB Management

The patient is rewarmed to normothermia with CPB as the surgeon completes the coronary anastomoses. The rewarming process may take longer in patients with a higher body mass index and more profound hypothermia. Mechanical ventilation is resumed upon the surgeon's request, after inspection of the anastomotic sites reveals the absence of a surgical cause for bleeding. If inhaled anesthetics are being used in the CPB circuit, their delivery should be continued via the anesthesia machine. The dose of the potent inhalational agent used should be minimal to avoid myocardial depression. Vasoactive infusions should be started or maintained. Communication between the surgeon and the anesthesiologist is paramount to confirm that both are ready to begin the process to separate the patient from CPB. Venous drainage to the pump is reduced, and the patient's heart begins to receive more blood from the circulation. Preload is adjusted by observing mean arterial pressure, cardiac distension, and PA or central venous pressure. The physician performing TEE provides information to the surgeon regarding volume status and contractility, which also aids the separation from CPB. A cardiac index is determined if a PAC is used. The venous cannula is removed from the RA with satisfactory separation from CPB. The surgeon, perfusionist, and anesthesiologist all must be aware when protamine is administered to reverse the anticoagulant effects of heparin. CPB must not be reinitiated once protamine administration has started without administering another dose of heparin. Once it is determined that the patient will not return to cardiopulmonary bypass, the administration of protamine is completed. With adequate neutralization of the heparin, the surgeon inspects the surgical field for bleeding and closes the surgical wound with adequate hemostasis. The patient is transported to the ICU for postoperative care.

Postoperative Outcome after CABG

Postoperative myocardial ischemia occurs in up to 48% of cardiac surgical patients and is associated with adverse cardiac outcomes.107 Thirty-eight percent of ischemic episodes occur during the first 2 postoperative days and peak within the first 2 hours of revascularization. These findings have important implications for monitoring, diagnosis, and treatment. Treatment of myocardial ischemia is directed toward adjusting the factors that determine myocardial oxygen supply and demand. Ischemia that develops immediately upon termination of CPB usually is related to air or particulate emboli in the bypass grafts. Elevation of mean arterial pressure and incremental increases in boluses of nitroglycerin are effective for treating ischemia caused by air in the venous grafts. The surgeon should confirm graft patency with Doppler flow probes or palpation. Persistent myocardial ischemia is treated with placement of an IABP. "Stunned myocardium" may be the result of reperfusion injury or intraoperative MI. These patients will require support of cardiac function until myocardial function improves.

Patients at risk for increased mortality after CABG are identified by several factors that evolve and are more prevalent or less over time. For example, patients undergoing myocardial revascularization between 1999 and 2004 were older and more likely to have metabolic syndrome or diabetes and peripheral vascular disease but fewer were smokers, compared with patients undergoing myocardial revascularization between 1993 and 1998. In terms of complications, the latter cohort had a higher rate of postoperative infarction and renal insufficiency, but a lower incidence of stroke and shorter duration of mechanical ventilation and hospital stays.108 Preoperative risk factors that increase mortality include age older than 80 years, emergent surgery, prior cardiac surgery, and renal failure.68,108 Mediastinitis is more common among patients with chronic obstructive lung disease, severe obesity, diabetes, renal failure, emergent surgery, and preoperative ejection fraction less than 40%.68,109 Renal failure is associated with advanced age, history of CHF, and preexisting renal disease.68,109

Growing areas of development are genomic predictors and other proteins that can be used to identify patients at risk for complications. For example, noncoding single-nucleotide polymorphisms within the chromosome 4q25 region are independently associated with atrial fibrillation after CABG.110 Preoperative C-reactive protein levels as low as 3 mg/L are associated with increased long-term mortality and extended hospital length of stay in patients undergoing primary CABG.111 Similarly, heart-type fatty acid–binding protein peaks earlier and is a superior independent predictor of postoperative mortality and ventricular dysfunction after CABG.112

Overview

Isolated valve surgery and valve surgery in combination with CABG are increasingly more common than CABG alone. The overall incidence of all valvular disease is unknown, but the progression and manifestation of symptoms is more common as the population ages.113 This combination of an aging population, expanding indications for surgery, improvements in prosthetic valve construction, and reduced patient morbidity and mortality after surgical intervention all suggest that valve surgery will continue to increase in coming years.

The 2 most common valves requiring surgical intervention are the MV and AV. Over 5 million persons have moderate to severe valvular regurgitation in the United States.114 Myxomatous MV disease is the most common cause of pure MR.115 The defect is usually associated with advancing age and degenerative changes. The pathophysiology includes elongated and/or ruptured chordae with generous leaflet tissue, a dilated annulus, and severe regurgitation. The defect is associated with increased mortality, even in asymptomatic patients.116 Most of these patients are amenable to valve repair rather than valve replacement (Fig. 52-7). Aortic valve disease is a common indication for surgery. Calcific aortic stenosis (AS) is seen in elderly patients, and in younger patients as well, especially with a bicuspid AV. Bicuspid AVs are found in approximately 0.5% of the population.117 The bicuspid AV is a heritable condition, often associated with a defect in fibrillin and matrix metalloproteinase-2, which leads to weakening and aneurysm formation (Figs. 52-8 and 52-9).118,119 These patients often present for replacement of both the AV and ascending aorta, requiring deep hypothermic circulatory arrest.

The spectrum of valve surgery extends well beyond this limited review. There are both regurgitant and stenotic lesions of all heart valves. The etiology varies widely from congenital birth defects, senile calcific stenosis, infective endocarditis, rheumatic disease, trauma, and other causes. However, the unifying concepts in all valve surgery include the principles of myocardial function and the influence of preload, afterload, inotropy, heart rate/rhythm, and diastolic function on myocardial performance and mechanics (Table 52-4). In valve defects that present acutely (ie, endocarditis, traumatic rupture), the compensatory ability of the heart and vascular system is limited. In chronic conditions (myxomatous MR, senile calcific AS), the ability to compensate can be remarkable. In caring for these patients, the clinician must understand these principles of myocardial performance, the acuity of the defect, the extent and mechanism of compensation, and the interplay of anesthetic drugs and surgery on the patient's condition.

Valve surgery will remain a mainstay of the contemporary cardiothoracic anesthesiology. This section reviews the preoperative evaluation of patients presenting for heart valve surgery, perioperative management techniques, and special considerations relating to the perioperative care of these complex patients.

Preoperative Evaluation

In heart valve disease, there are a number of questions that must be answered that will directly influence anesthetic management. Typically, procedures are posted on the operating room schedule as "Aortic Valve Replacement" or "Mitral Valve Replacement." There is no information on whether this is for AS, aortic insufficiency (AI), or both; MS, MR, or both. Understanding the specific pathology of the valve preoperatively is essential in preparing hemodynamic goals for the induction, maintenance, and postoperative care of these patients. Obtaining a focused history and performing a physical examination are critical in preparing for the care of these patients. Patient outcome is dependent on a number of variables; however, understanding baseline status is the first step in ensuring the best possible outcome.

In addition to the usual anesthetic concerns (ie, airway, "NPO" status, comorbidities, etc), much additional information is required that may be unrelated to the valve pathology. For example, will there be other procedures during this repair including a carotid endarterectomy, CABG, septal myomectomy, a MAZE procedure, or other intervention? What is the clinical condition of the patient? Is the patient in a compensated or uncompensated state? For example, with AS, the timing of surgery will have a great impact on outcome depending on the patient's condition. A patient in the early to mid-stages of AS may have a hyperdynamic, hypertrophied heart that will likely perform well after CPB. In contrast, a patient in the late stages of AS may be in CHF, have a reduced ejection fraction (EF), and will likely require extensive inotropic support after CPB.

There are a number of essential imaging and laboratory studies that must be reviewed preoperatively. Imaging of the heart will provide the diagnosis and degree of the structural abnormality. Preoperative echocardiography will establish the type of lesion (stenosis versus regurgitant), valve gradients, and global heart function. Knowledge of the valve area is of critical importance in assessing any stenotic valve. In patients with normal LV function, AV gradients in severely stenotic lesions may be very high (Figs. 52-10 and 52-11). Alternatively, in patients with poor ventricular performance, there may be anatomically critical lesions (AV area <0.7 cm2) and a low gradient. These patients are especially prone to intraoperative instability with the induction of general anesthesia and may require significant inotropic support post CPB. Other important imaging modalities include coronary angiography to rule out concurrent coronary artery disease, a chest radiograph, and perhaps thoracic computed tomography (CT) or magnetic resonance imaging (MRI) to evaluate thoracic anatomy. A chest x-ray is essential in "redo" operations involving the heart and mediastinum. The heart may be adherent to the posterior margin of the sternum, rendering sternotomy quite hazardous (Fig. 52-12). All precautions including adequate venous access, immediate availability of blood products, and intraoperative awareness and vigilance are required in these situations.

Monitoring

Monitoring during heart valve surgery varies little from monitoring required during other types of heart surgery including CABG. Outside of the American Society of Anesthesiologists standard monitors, an arterial line and a central venous line are essential. Many institutions routinely use PACs to monitor cardiac output, mixed venous oxygenation, pulmonary artery pressures, and PAOPs. Although this is standard practice by many clinicians, outcome data supporting pulmonary artery catheterization are lacking.

Intraoperative TEE is an essential component of anesthesia for all heart valve procedures. Extensive preoperative evaluation by TEE often reveals previously unknown structural or functional defects (Fig. 52-13). In addition, TEE can provide instant assessment of the status of a valve repair procedure. TEE may also diagnose perivalvular leaks requiring prompt intervention (Fig. 52-14). The assessment of air in the left side of the heart is important after an open-heart procedure. TEE may assist in locating the air, thereby facilitating direct venting procedures. TEE is helpful in guiding post-CPB hemodynamic intervention and treatment. In recognition of this tool's utility in cardiac surgical procedures, the American Society of Echocardiography and the American Heart Association provide a Class 1 recommendation for TEE use in all heart valve surgical procedures.120

The assessment of diastolic dysfunction by TEE is important in all patients presenting for cardiac surgical procedures. Dourvas et al reported 40 consecutive cases of right ventricular (RV) diastolic dysfunction in patients with significant AI, LVEF greater than 55%, LV end-diastolic pressure less than 15 mm Hg, RV systolic pressure less than 30 mm Hg, and normal coronary arteries.121 The pathophysiology of this defect may be related to rapid dilation of the LV in early diastole, a bulging septum, and interference with RV filling. On examination, these patients show a small RV and a large right atrium.122 An awareness of diastolic dysfunction may guide intraoperative and postoperative hemodynamic therapy.

In addition to newer TEE imaging modalities including 3-dimensional echocardiography,123,124 other special monitoring devices in heart valve surgery may include processed EEG monitoring for either patient awareness or for EEG activity during procedures in which it is desirable to induce an isoelectric state (ie, hypothermic circulatory arrest). In addition, many clinicians find the data provided by cerebral oxymetry to be useful in preventing perioperative neurological injury.

Induction of Anesthesia

Heart valve disease poses an additional burden on the anesthesiologist when planning for the induction of general anesthesia. The clinician should have a clear understanding of how preload, afterload, inotropy, heart rate and rhythm, and diastolic function all interacted in the patient during the time of induction and maintenance of anesthesia. Table 52-4 outlines several basic considerations regarding these parameters. On occasion, a patient will present with a constellation of disease processes. For example, the patient may have severe AS with moderate to severe AI. This same patient may also have obstructive CAD requiring a CABG procedure, as well as moderate MR. How should the information in Table 52-4 be applied to such a patient? There are 2 approaches. One approach identifies the lesion that poses the greatest proximate risk. For example, in a patient with AS and tricuspid regurgitation (TR), it is most likely that the patient will be symptomatic from the effects of the AS rather than the low pressure, high volume lesion of TR. When caring far a patient like this, it is sometimes best to apply the hemodynamic recommendations supplied by the AS algorithm. Another approach is to identify those hemodynamic parameters at which the patient appears compensated. For example, in a patient who presents with compensated AI, it is wise to try to keep the patient in an unaltered hemodynamic condition, that is, keep preload, afterload, inotropy, and rate and rhythm unchanged from the conscious, compensated condition.

Much has been written on specific induction agents, contrasting dose schemes, and medication management for induction of anesthesia. Despite these data, no specific induction agents are indicated in any of the heart valve conditions. Careful titration, attention to the hemodynamic alterations, and prompt intervention when necessary are the key elements in induction of anesthesia for valvular surgery. More important than a specific agent is the adherence to identifying and maintaining clear hemodynamic goals. In certain disease states, abrupt alteration of hemodynamics can lead to catastrophic results. For example, in patients with severe LV hypertrophy, AS, and CAD, a decrease in arterial blood pressure may lead to significant myocardial ischemia and cardiovascular collapse. These patients are unable to alter ejection due to the outflow obstruction, and therefore any drop in preload and afterload may reduce both stroke volume and coronary perfusion. In a hypertrophied heart subject to increased myocardial oxygen demands, the resulting ischemia will be magnified in severity and hastened in onset.

Maintenance of Anesthesia

No agents or techniques (ie, "high-dose" narcotic technique vs inhalation-based anesthesia) are specifically indicated for maintenance of anesthesia in any specific valve disease condition. Again, adherence to the hemodynamic goals outlined in Table 52-4 will provide appropriate guidance when maintaining anesthetic depth. The sound principles of anesthesia apply, that is, loss of consciousness, muscle relaxation, and hemodynamic stability (blunting of the autonomic nervous system).

Patient awareness during cardiac surgical procedures requires special attention. Awareness, especially auditory recollection, may occur in up to 6% of patients requiring cardiopulmonary bypass.125 Sebel et al reported an incidence of operative awareness of 0.13% in a broad population of 19 575 surgical patients.126 Multivariate logistic regression identified increasing ASA status and type of procedure as risk factors for recall. For cardiac surgical procedures the odds ratio for recall was 3.58 with a 95% confidence interval of 0.72 to 17.9. As with any procedure, attention to hemodynamic signs, depth of anesthesia, the use of benzodiazepines, and at least one-half MAC of an inhalational agent will likely reduce the risk of operative recall.

There is good evidence to support the concept of "fast-track" or early extubation protocols in the care of patients presenting for CABG.127 These benefits include reduced hospital length of stay, intensive care unit length of stay, and cost of care. Data on early extubation protocols in valve surgery are less common, although one might expect similar benefits to those observed in CABG.128 One should be aware, however, that many valvular procedures are complex in nature, in elderly populations, requiring extended CPB time, potential deep hypothermic circulatory arrest, and increased requirements for blood and blood product transfusion. All of these factors may limit the opportunity for early extubation.

Weaning from CPB and the Post-CPB Period

Once the valve is repaired or replaced, the hemodynamic goals of the patient may be fundamentally different from the pre-CPB period. A more normal hemodynamic profile is usually immediately observed while in the operating room (Fig. 52-15). The anticipation is that the heart will now perform normally; however, this is frequently not the case. The heart may suffer from an acute insult associated with the surgical intervention, aortic cross-clamping, and the myocardial depressant effects of the various anesthetic agents. As with any cardiac surgical procedure, there may be evidence of new regional wall motion abnormalities, global cardiac dysfunction, and peripheral vascular dysfunction. In valvular heart surgery, the heart often requires time to remodel before significant improvement is observed.

In some patients, hemodynamics may be greatly improved post CPB. In simple, compensated AS, the patient may be hyperdynamic and hypertensive after weaning from CPB. It may also be possible, however, that due to the LV hypertrophy, cardioplegia may have been incomplete for total myocardial protection, and the patient may emerge from CPB with transient myocardial dysfunction. In complex repairs, or surgery on more than 1 valve, an extended CPB run may be required, thereby increasing the risk of immediate postoperative cardiac dysfunction and the requirement of inotropic support. Ejection fraction and LV function may decline after surgery for MR. Once thought to be due to the new loading conditions imposed on the heart after removal of the "pop-off" associated with MR outflow, emerging data from MV repair procedures, in which there is sparing of the MV apparatus, show little alteration in EF.129 This preservation of mitral annular integrity stabilizes LV shape and contractility, thereby preserving function. The diminished performance is likely secondary to other factors including extended cross-clamp time, inadequate myocardial protection, and myocardial stunning. Combined CABG and valvular heart surgery is associated with inotrope use during separation from CPB. In a retrospective study of 1009 consecutive patients undergoing either isolated CABG or CABG and valve surgery requiring CPB, a multivariate risk analysis revealed extended cross-clamp time, worsening wall motion score index, reoperation, combined procedures, severe MR, and low EF as predictors of inotrope use during separation from CPB.130

TEE is especially useful in the immediate postbypass period. TEE will facilitate the assessment of myocardial function and filling parameters, rendering important information affecting decisions about inotropes and volume management. TEE can also assess the condition of the valve repair or replacement. On every case in which TEE is used, a complete evaluation of the heart and valve is required. This examination includes an assessment of structure and function. In a valve repair, the degree of regurgitation must be documented. A gradient across the valve is required, and images of the valve movement are necessary. In a valve replacement, all of the above are essential and a further evaluation for a perivalvular leak is required. After mitral valve repair, an assessment for systolic anterior motion (SAM) of the mitral valve is required. These examinations should be recorded and a report generated for the patient's record.

The post-CPB period may be a time of unstable hemodynamics. The heart is adjusting to new conditions of preload, afterload, and inotropy. There is usually ongoing bleeding that, in many patients, and especially repeat procedures, may be severe. Constant vigilance regarding hemodynamics, blood volume, blood gas chemistry, and metabolic state (ie, ionized calcium, glucose concentrations, and urine output) are required. There may be new rhythm disturbances. Patients with preoperative atrial fibrillation may now be converted to sinus rhythm after mitral valve repair and a Maze procedure. These patients may be dramatically improved. Other patients, however, may suffer from third-degree heart block, bradycardia, tachycardia, or other rate and rhythm abnormalities, making management difficult.

In the ICU, the usual considerations of temperature management, metabolic and hemodynamic management, vigilance for excessive chest tube bleeding, and evidence of cardiac tamponade must be performed. Elderly patients and those with additional risk factors may require additional support.

Special Considerations and Long-Term Prognosis

Cardiac anesthesia for heart valve surgery carries numerous special considerations not found in other aspects of cardiothoracic anesthesiology. Among these considerations is familiarity with the type of valve used in the operative repair. There are a number of prosthetic valve types including mechanical and bioprosthetic valves. The choice of valve is up to the surgeon; however, the anesthesiologist may be asked to provide TEE measurements that may guide the surgeon in valve selection. For example, the measurement of the AV annulus size is important when determining the size of a prosthetic valve. In the assessment for a possible Ross procedure, the pulmonary annulus size and the degree of pulmonary insufficiency are important factors before committing to the procedure. After the repair, it is important to assess the new valve function. Bioprosthetic valves are accompanied with a complete hemodynamic profile in their packaging, which should be consulted when rendering a decision on the function of a new valve. The expected gradients are sometimes higher, and the calculated valve area may be less than one might expect. The impact of valve prosthesis-patient mismatch (PPM) on mortality is as high as 25%.131 In a follow-up study of 1563 patients undergoing aortic valve replacement (AVR), the adjusted hazard ratio (95% confidence interval) at 5 years for heart failure with a PPM mismatch was 1.64 (1.01-2.56), p = 0.047, and for heart failure deaths was 2.09 (1.03-4.27), p = 0.043.132 Reference to the packaging insert will guide the clinician as to whether intraoperative gradients and calculated valve areas are acceptable. In valve repair procedures, one must not only look for the cessation of regurgitation, but also ensure that there is no new stenosis. Measurement of valve gradients and orifice areas will help in this regard.

Valve repair for patients with endocarditis presents all of the aforementioned hemodynamic considerations in addition to the frequently septic condition of the patient (Fig. 52-16). Patients with endocarditis usually have acute regurgitant lesions and present in heart failure. The in-hospital mortality for acute endocarditis is 20%, and 82% of patients have endocarditis of a native heart valve.133 The early predictors of in-hospital mortality include embolic events, diabetes mellitus, Staphylococcus aureus, and APACHE II score. The use of intraoperative TEE is especially helpful in the diagnosis of endocarditis and quantification of the valve defect. Valve surgery reduces 6-month mortality. In a study by Iung et al, 513 patients with complicated left-sided infective endocarditis were treated with antibiotics; 45% underwent valve surgery and 55% received medical therapy alone.134 The 6-month mortality was 16% in the surgery group compared to 33% in the medical group, p < 0.01. Patients with the most severe heart failure demonstrated the greatest reduction in mortality with surgery (14% vs 51%, p = 0.001).134

There are significant data to support valve repair and replacement in patients with heart valve disease over medical management.115,135 After surgery for AS, initial remodeling of the heart is associated with subsequent LV mass reduction.136-138 Between 1986 and 2001, 1410 patients had surgery for severe AI, of whom 160 (11%) had valve repair.139 There was 1 operative death (0.6%), and 2 patients required early re-repair. At 7 years after repair, the survival was 89% and the reoperation rate on the AV was 15%.139

Patients with a history of MR often experience symptomatic relief and improvement in LVEF postoperatively.140 Although there is known early mortality with heart valve surgery, the long-term benefit remains superior to medical management. As the population ages, more patients will present for valve procedures. Although advancing age is associated with poor outcome, there is no upper limit of age at which time valve surgery is contraindicated. In summary, these data and others indicate that surgery offers an opportunity for improved mortality, symptomatic relief, and hemodynamic improvement in nearly all patients with valve dysfunction.

Percutaneous approaches to valve repair and replacement are becoming more popular, especially in higher-risk populations.141-144 It is unknown what impact percutaneous valves and other nonconventional valve repair procedures will have on the volume of heart valve surgeries. Anesthetic management is complicated by the duration of the procedures, the interruption of flow during deployments, and rhythm disturbances.145 In such cases, all of the hemodynamic principles described above apply. In addition, advanced competency in perioperative TEE is required for both percutaneous AV replacement procedures (Fig. 52-17) and percutaneous mitral annular reduction for mitral regurgitation, as well as endovascular mitral valve repair.146

Thoracic aortic diseases are generally surgical problems (Table 52-5).147 The most common operations performed on the thoracic aorta are repair of aortic dissection, aortic aneurysm, traumatic aortic injury, and aortic coarctation. Operative repairs involving the aortic root, ascending aorta, or transverse aortic arch typically are approached through a median sternotomy. Operative repairs involving the distal aortic arch or descending thoracic and thoracoabdominal aorta typically are approached through a left thoracotomy or thoracoabdominal incision. Endovascular stent graft techniques for repair of descending thoracic aortic aneurysm (TAA) and dissection have been developed that are accomplished by access through the femoral or iliac arteries.148,149 Hybrid repairs that combine both endovascular and open techniques with extra-anatomic bypass of aortic branch vessels have also been developed for the treatment of aortic arch aneurysms.150

Specialized anesthetic management of patients undergoing thoracic aortic operations has contributed to the overall success of these operations and requires knowledge of several unique techniques that are practiced routinely in few other areas of medicine. Clinical application of TEE and ultrasound imaging performed by the anesthesiologist provides a means to emergently diagnose acute aortic syndromes, identify associated life-threatening complications, and detect cerebral malperfusion, permitting early surgical intervention and treatment.147,151 The process of repairing or replacing a portion of the thoracic aorta typically requires the temporary or permanent interruption of blood flow through the aorta or its major branch vessels, creating the potential for ischemia or infarction of almost any major organ system. Techniques to protect organs during temporary interruption of blood flow in the thoracic aorta include deep hypothermic circulatory arrest (DHCA), selective antegrade cerebral perfusion, retrograde cerebral perfusion, and partial left heart bypass for distal aortic perfusion. Intraoperative neurophysiologic monitoring and lumbar cerebrospinal fluid (CSF) drainage are recognized techniques commonly used for repairs involving the descending thoracic or thoracoabdominal aorta to decrease the risk of spinal cord ischemia and infarction.

Thoracic Aortic Aneurysm

Aortic aneurysm is a dilation of the aorta containing all 3 layers of the vessel wall that has a diameter at least 1.5 times that of the normal diameter of the corresponding aortic segment. TAAs are common. They are detected in 10% of autopsies, have an incidence of 5.9 per 100000 person-years, and are the most common reason for thoracic aortic operation.152 TAAs can be characterized by etiology, location, diameter, and extent of aortic involvement. Aortic pseudoaneurysms are caused by a contained rupture of the aorta or arise from intimal disruptions, penetrating atherosclerotic ulcers, or partial dehiscence of the suture line at the site of a previous aortic vascular graft. Aortic aneurysms can be associated with dissection of the vessel wall.

Surgical repair of TAAs is performed for acute rupture, AI, refractory pain, or to prevent eventual rupture. Rupture of the aortic root or proximal ascending aorta will cause cardiac tamponade because the first several centimeters of the ascending aorta lies within the pericardial sac. Dilation of the proximal ascending aorta may cause AI by distortion of the aortic root and outward tethering of the AV cusps.152,153 Because the risk of aortic rupture or dissection is associated with aneurysm size, aortic diameter is commonly used as a clinical indication for elective repair (Table 52-6). In general, indications for operative repair are ascending aortic aneurysm diameter 4.5 cm or greater in patients with Marfan syndrome, collagen vascular disease, bicuspid AV, or a family history of aneurysm; ascending aortic aneurysm diameter 4.5 cm or greater in patients requiring CABG or valve procedures; ascending aortic aneurysm diameter 5.5 cm or greater in any patient; symptomatic aneurysms; or aneurysm growth rate greater than 0.5 cm per year.147,154-156 Because the operative risks associated with open repair of the descending thoracic aorta is greater, endovascular repairs should be considered, and repair is usually not performed until an aneurysm size of 5.5 to 6.0 cm is reached.147,155,156 Very large aneurysms of the ascending aorta, aortic arch, or descending thoracic aorta may produce a mediastinal mass effect, causing extrinsic compression of the trachea, left mainstem bronchus, right PA, RV outflow tract, or esophagus (Fig. 52-18). In contrast to patients with collagen vascular or familial aortic aneurysms, patients with atherosclerotic aneurysms generally are elderly and have more comorbid conditions and peripheral vascular disease.

Ascending Aorta and Aortic Arch Aneurysm

Aortic aneurysms limited to the aortic root and proximal ascending aorta can be repaired using standard CPB with cannulation of the aneurysm, ascending aorta, or femoral artery and cross-clamping of the distal ascending aorta. Based on assessment of the AV, the AV may be replaced, resuspended, or reimplanted within the prosthetic vascular graft. Replacement of the aortic root requires reimplantation of the right and left coronary arteries. Intraoperative TEE is useful for evaluating the native AV prior to repair and for determining the presence of residual AI after valve-sparing operations.

Operative repair of aortic aneurysms that extend into or involve the aortic arch require temporary interruption of cerebral perfusion. The primary technique used to protect the brain from ischemic injury in the absence of cerebral blood flow is DHCA. In the conduct of DHCA, achieving a satisfactory level of hypothermia is considered the most important intervention for brain protection, but the optimal temperature for DHCA, best site for temperature measurement, and safe duration of DHCA have not been established. EEG during the conduct of DHCA has demonstrated that the average nasopharyngeal temperature required to produce electrocortical silence was approximately 18°C, but a nasopharyngeal temperature of 12.1°C or cooling on CPB for at least 50 minutes was necessary to achieve electrocortical silence in 95% of patients (Fig. 52-19).157 Clinical studies indicate that the cerebral metabolic rate decreases by a factor of approximately 2.6 for each 10°C decrease in temperature (Q10 ratio for adults).158 Assuming the brain can tolerate an ischemic period of approximately 3 to 5 minutes at 37°C, the reduction in cerebral metabolism at a temperature of 17°C would predict that the brain should tolerate approximately 20 to 34 minutes of ischemia. Clinical studies support this prediction and have detected the onset of neuronal ischemia at approximately 18 minutes after DHCA (Fig. 52-20).159 Other studies have indicated that postoperative neurocognitive dysfunction was more frequent when DHCA duration exceeded 25 minutes.160

Strategies for improving the safety of DHCA to facilitate operations requiring temporary interruption of blood flow in the aortic arch include techniques to provide retrograde cerebral perfusion (RCP) or selective antegrade cerebral perfusion. Retrograde cerebral perfusion can be provided immediately after initiation of DHCA by infusing cold oxygenated blood into the superior vena cava at flow rates averaging 150 to 250 mL/min.161 During RCP, the pressure within the superior vena cava is generally maintained at or below 25 mm Hg to decrease the risk of cerebral edema, the aortic arch is opened to atmospheric pressure to prevent pressurization of the arterial system, and the patient is positioned in an 8- to 10-degree Trendelenburg to prevent air entry into the aortic arch branch vessels. Experimental and clinical studies have demonstrated that RCP via the superior vena cava provides perfusion to the brain, but existing evidence indicated that the flow achieved with RCP was insufficient to prevent cerebral ischemia.161 Nevertheless, advocates of RCP argue that even some substrate delivery increases the margin of safety of DHCA, that cerebral hypothermia is maintained with RCP, and that RCP decreases the risk of cerebral embolization by flushing out particulate matter in the cerebral arteries prior to resumption of antegrade cerebral perfusion.

Selective antegrade cerebral perfusion can be accomplished by arterial cannulation and perfusion of the axillary artery, innominate artery, subclavian artery, or even the internal carotid arteries.162,163 Antegrade perfusion into a single arch branch vessel provides flow to the contralateral cerebral hemisphere through a functional circle of Willis, but contralateral perfusion pressures may vary among patients.163 Selective antegrade cerebral perfusion is generally performed in combination with deep hypothermia for operations where the duration of DHCA is anticipated to exceed 30 minutes. Flow rates for antegrade cerebral perfusion typically range from 400 to 1000 mL/min at radial artery pressures of 50 to 80 mm Hg. Clinical studies supporting the efficacy of antegrade cerebral perfusion demonstrated acceptable outcomes despite antegrade cerebral perfusion duration times greater than 45 minutes at a temperature of 25°C.163

Attempts to further improve the safety of DHCA by manipulating pH, blood viscosity, or hemoglobin concentration or by administration of pharmacologic agents often are practiced, but no clinical evidence has supported the efficacy of these other interventions. Barbiturates or other central nervous system depressants are sometimes administered in an effort to decrease cerebral oxygen demands. Glucocorticoids, magnesium sulfate, lidocaine, and mannitol are sometimes administered in an effort to protect against cerebral and end-organ injury. In general, the existing evidence suggests that pharmacologic neuroprotection is unproven and should not be considered a substitute for hypothermia or selective antegrade cerebral perfusion to protect against cerebral ischemia during operative repairs of the aortic arch.

Descending Thoracic and Thoracoabdominal Aortic Aneurysm

Classification of TAAs and thoracoabdominal aortic aneurysms (TAAAs) according to anatomic extent provides a useful guide to surgical approaches, anesthetic management, and estimate of perioperative mortality and postoperative paraplegia associated with operative repair (Table 52-7).147,164,165 In the Crawford classification, extent I TAAA involves the entire descending thoracic aorta from the origin of the left subclavian artery to the level of the diaphragm; extent II TAAA involves the entire descending thoracic aorta with extension across the diaphragm through the abdominal aorta to the aortic bifurcation; extent III TAAA involves the distal half of the descending thoracic aorta and most of the abdominal aorta; and extent IV TAAAs are confined to the upper abdominal aorta (Fig. 52-21). Isolated TAAs are those confined to the descending thoracic aorta between the origin of the left subclavian artery and the diaphragm. TAAAs can be further distinguished into those with dissection and those without dissection.

Operative repair of TAAA with an interposition tube graft is approached through a left thoracoabdominal incision and requires single-lung ventilation. Repairs originally were accomplished by cross-clamping the thoracic aorta proximal to the aneurysm. Modifications of this technique included passive arterial shunting (Gott shunt) or partial left heart bypass using extracorporeal circulation to provide distal aortic perfusion while the proximal descending thoracic aorta was cross-clamped.166 CPB with DHCA may be necessary for aneurysms that extend into the distal aortic arch,167,168 Endovascular stent graft repair has become an option for repair of isolated TAA and TAAA.148,149,165 One of the major complications of operative repair is spinal cord ischemia or infarction resulting in postoperative paraplegia. The risk of spinal cord ischemia is a consequence of the interruption of distal aortic perfusion or the sacrifice of intercostal, lumbar, and sacral artery branches that provide collateral flow supplying the spinal cord. Strategies to prevent, detect, and treat spinal cord ischemia are important aspects in the anesthetic management of patients undergoing TAA and TAAA repair. Spinal cord protection strategies include deliberate hypothermia, surgical techniques to minimize ischemia time, arterial pressure augmentation, avoiding hypotension, lumbar CSF drainage, and perioperative neurophysiologic monitoring of spinal cord function.147 Deliberate hypothermia is an established technique to protect against neuronal ischemia and can be accomplished by passive cooling to 32°C, active cooling using extracorporeal circulation, or selective cooling by infusion of cold saline into the epidural space (Table 52-8).147,168,169 Passive arterial shunting using an ascending aorta to distal aortic shunt or partial left heart bypass using extracorporeal circulation to direct blood from the LA to the femoral artery can be used to provide distal aortic perfusion during construction of the proximal aortic anastomosis and decreases the duration of lower body ischemia. While the descending aorta is cross-clamped, spinal cord perfusion via the vertebral arteries can be augmented by maintaining the proximal aortic pressure at or above 90 mm Hg. Intraoperative neurophysiologic monitoring of somatosensory evoked or motor evoked potentials from the posterior tibial nerves has been advocated for detection of spinal cord ischemia during operation in the anesthetized patient, to permit early intervention or to prompt the reattachment of intercostal artery branches.170,171 Pharmacologic agents such as glucocorticoids or even naloxone have been administered for spinal cord protection, but their efficacy remains unproven. Patients with delayed-onset paraplegia or paraparesis after TAA, TAAA, or endovascular stent graft repair often respond to strategies to improve spinal cord perfusion by increasing the arterial pressure and drainage of CSF.168,169 One rationale for the routine use of lumbar CSF drainage is that the lumbar CSF pressure may increase during repair as a consequence of aortic cross-clamping or spinal cord edema during reperfusion. Hypotension associated with spinal cord ischemia may be a consequence of general anesthesia, regional anesthesia, blood loss, vasodilation, or even neurogenic shock from spinal cord ischemia and requires prompt treatment. The effectiveness of lumbar CSF drainage for treatment of spinal cord ischemia may be improved when combined with arterial pressure augmentation. Interventions to increase spinal cord perfusion pressure appear to be most effective when applied immediately upon detection of spinal cord ischemia. The effectiveness of lumbar CSF drainage as a therapeutic adjunct to decrease the risk of spinal cord ischemia is supported by several randomized controlled trials, case series, and meta-analyses.172-174 Complications of lumbar CSF drainage are uncommon and include intracranial hypotension, subdural hematoma, catheter fracture, meningitis, and hemorrhagic complications.175-177 Precise monitoring of CSF pressure, controlled drainage of CSF to maintain a lumbar CSF pressure of at least 10 mm Hg, and supervised insertion and removal of catheters may decrease the risk of complications. Considering the morbidity associated with paraplegia, the application of lumbar CSF drainage, arterial pressure augmentation, and neurophysiologic monitoring can be justified in patients at risk for postoperative spinal cord ischemia.147 Patients at risk for spinal cord ischemia include those undergoing Crawford extent I, II, and III open TAAA repair and endovascular stent graft repairs with prior distal aortic operations.165,169,172

Aortic Dissection

Aortic dissection evolves from a tear in the intima that allows blood to enter the medial layer of the vessel, causing the intima to separate or dissect circumferentially and longitudinally within the aorta. Aortic dissection is diagnosed by imaging studies demonstrating a true and a false lumen separated by an intimal flap within the aorta (Fig. 52-22).178,179 Aortic intramural hematoma is a variant of aortic dissection characterized by hemorrhage into the medial layer producing circumferential thickening of the aortic wall.180 Aortic dissection affects a wide demographic population and age range. The most common associated diseases are hypertension, atherosclerosis, and cystic medial degeneration. Other associated conditions include collagen vascular disease, familial aortic dissection, bicuspid AV, aortic coarctation, pregnancy, cocaine abuse, arteritis, and aortic trauma. Early complications of aortic dissection include dissection into aortic branch vessels causing malperfusion, resulting in myocardial, cerebral, mesenteric, or limb ischemia (Table 52-9). Dissection into the aortic root may cause acute aortic regurgitation. Rupture of the proximal ascending aorta causes cardiac tamponade. Dilation and expansion in the diameter of the aorta over time as a consequence of a weakened vessel wall is a long-term complication of aortic dissection.

Aortic dissection is classified according to location. Dissections involving the ascending aorta and aortic arch are classified as Stanford type A or DeBakey type I or II. Dissections confined to the descending aorta are classified as Stanford type B or DeBakey type III (Fig. 52-23). Aortic dissections involving the ascending aorta (Stanford type A or DeBakey type I or II) are considered surgical emergencies. According to an international registry for acute aortic dissection, mortality rates for patients with Stanford type A aortic dissection managed without surgery was approximately 1% to 2% per hour following the initial symptom onset for the first 48 hours, 60% by day 6, 74% by 2 weeks, and 91% by 6 months.181 When managed with surgery, the mortality rate for type A aortic dissection was 26% (Fig. 52-24).181 Aortic dissections confined to the descending thoracic aorta (Stanford type B) are considered surgical emergencies only if there is evidence of a life-threatening complication such as malperfusion or aortic rupture. Mortality in patients with acute type B aortic dissection managed surgically was 31.4% at 30 days compared to a mortality rate of 10.7% at 30 days in medically managed patients (see Fig. 52-24).181

Acute management of patients with suspected aortic dissections requires establishing the diagnosis, distinguishing Stanford type A from Stanford type B aortic dissection, and detecting malperfusion or rupture. In an international registry of aortic dissection, the initial diagnosis was established by computed tomography in 61% of patients.179 Echocardiography was used as the initial diagnostic technique in 33% of cases but was used as a secondary diagnostic technique in 56% of cases.179 TEE is useful for intraoperative diagnosis and classification of dissection in unstable patients. Intraoperative TEE also is useful for detecting AI, cardiac tamponade, and myocardial ischemia. Surgery for type A aortic dissection involves replacement of the ascending aorta or aortic arch together with AV repair or replacement. Patients with type B aortic dissection are preferentially managed medically unless they have evidence of aortic rupture or malperfusion. Endovascular stent graft repair has become an alternative to open repair in patients with type B aortic dissection complicated by malperfusion or rupture.149 The anesthetic management of operations for the repair of type A aortic dissection is similar to the management of patients undergoing repair of ascending aortic and arch aneurysms, and typically requires temporary interruption of cerebral perfusion and DHCA. A rare but lethal complication of CPB in patients with aortic dissection is acute cerebral malperfusion caused by inadvertent cannulation of the false lumen of the aorta or compression of the true lumen by expansion of the false lumen.182 This complication can be detected by intraoperative TEE evaluation of blood flow in the aorta and ultrasound confirmation of carotid artery flow during initiation of CPB. Treatment consists of immediate discontinuation of CPB, cannulation of the contralateral femoral artery for CPB, or open fenestration of the intimal flap within the aorta. The anesthetic management of operations for type B aortic dissections requires the same strategies and techniques used for patients undergoing TAA and TAAA repair.

Brief History of Ventricular Assist Devices

The first left ventricular assist device (LVAD) was implanted by DeBakey in 1966 as a "bridge to recovery" in a patient suffering from postcardiotomy cardiogenic shock.183 The decompressed and rested native ventricle recovered after 6 days of support and the device was explanted. Barnard performed the first orthotopic cardiac transplantation a year later. During the 1970s the development of heart transplantation as a viable therapy for end-stage heart failure was hindered by the use of nonspecific immunosuppressive agents, which resulted in patient deaths from opportunistic infections. By the early 1980s, with the introduction of cyclosporine, a specific T-cell inhibitor, heart transplantation became a viable therapy for end-stage heart failure with risk-adjusted survival rates equaling 50% at 10 years.184 Ventricular assist devices subsequently assumed a new role as a "bridge to transplant." End-stage heart failure patients on transplant waiting lists, if supported by an assist device prior to transplantation, could wait longer, recover secondary organ function, be discharged home, and enjoy improved survival after transplantation.185 Over 500000 Americans are diagnosed each year with heart failure, but only 2200 donor hearts become available each year worldwide for transplantation.186 In the late 1990s the REMATCH (Randomized Evaluation of Mechanical Assistance for the Treatment of Congestive Heart Failure) trial demonstrated that the HeartMate pulsatile pneumatic (Thoratec Corporation, Pleasantville, California) LVADs could be used as an alternative to transplantation (ie, "destination therapy") in end-stage heart failure patients who were not candidates for transplantation. Use of the device conferred a survival and quality-of-life benefit when compared to optimal medical therapy. Over the next decade LVAD technology continued to evolve and the next-generation continuous-flow LVADs were smaller and more durable than their pulsatile predecessors. An original randomized study (Clinical Trials.gov NCT00121485) by Slaughter et al demonstrated superiority of continuous-flow over pneumatic LVADs at 2 years in patients receiving an LVAD for destination therapy.187 Fifty-eight percent of end-stage heart failure patients receiving the continuous-flow HeartMate II LVAD were alive after 2 years versus 23% for pulsatile pneumatic (HeartMate XVE) versus 8% for medical therapy. Actuarial–survival analysis performed by means of the Kaplan-Meier method demonstrated a 2-fold increase in survival with continuous-flow devices over pulsatile pneumonic, and a 4-fold increase in survival over optimal medical therapy.

In 2009 the PROTECT I trial demonstrated yet another application of LVAD technology. The Impella RECOVER LP 2.5 System (Abiomed Inc, Danvers, Massachusetts) could be delivered via a catheter-based technique and provide temporary support (averaging 2 h) during high-risk percutaneous coronary intervention (PCI) in patients with unprotected left main coronary artery disease or last patent coronary conduit.188 The VAD is used to maintain a mean arterial blood pressure above 60 mm Hg during occlusion of the target vessel. The PROTECT II trial is designed to compare the use of the Impella 2.5 LVAD versus IABP for high-risk PCI. However, the true benefit of left main PCI has yet to be validated using adequately powered studies.

Most VADs placed in community hospitals are implanted for rapid stabilization of patients in acute cardiogenic shock who, despite multiple inotrope therapy and insertion of intra-aortic balloon pump, continue to deteriorate. These VADs are generally implanted as a "bridge to a bridge" or a "bridge to decision," or simply a "bridge to explantation" in that they are meant to serve as a short-term solution to cardiogenic shock and resultant multisystem organ failure until decisions can be made regarding possible cardiac recovery, neurologic status, feasibility of transfer to a tertiary facility for long-term assist device implantation, and possible heart transplant.189 A disadvantage of temporary support is that multiple surgeries and the need for anticoagulation are required. An advantage, at least in patients with multisystem organ failure or neurologic status that does not recover, is that permanent LVAD implantation is avoided. The cost of LVAD implantation and follow-up is estimated to be between $500,000 and $1.4 million per patient.190 The Bio-Medicus systems (Medtronic Inc, Minneapolis, Minnesota), CentriMag (Levitronix LLC, Waltham, Massachusetts), and ABIOMED devices have all demonstrated relative successes with 30-day survival rates ranging from 31% to 75%.189,191,192 Extracorproal membrane oxygenation (ECMO) may also be used to support patients in acute cardiogenic shock. Hoefer et al reported a 50% mortality rate in patients with ECMO support who were bridged to permanent VAD implantation.193 Which device is best for short-term support and which patients will benefit from the technology continue to be matters for debate.

General Principles

In general VADs are pumps that collect blood returning to the heart and then eject it downstream of the failing ventricle. They differ from CPB and ECMO in that they do not provide any direct respiratory function. The goal of VAD implantation is to decompress the acutely ischemic failing or the chronically failing ventricle, thereby diminishing radius, wall tension, and oxygen consumption. The ventricle may recover while perfusion to the body is maintained by the assist device. For left ventricular support the inflow cannula is generally placed in either the left atrium (LA) or LV and drains into the device. Better drainage and decompression occur when the cannula is placed in the ventricular apex; however, postinfarction ventricular tissue may not be durable enough to hold sutures. Furthermore, LA inflow cannulation may result in a higher incidence of LV clot.186,194 The outflow cannula is generally placed in the ascending aorta. Notable exceptions include the Jarvik 2000 in which the outflow cannula can be sutured to the descending aorta,195 and the TandemHeart (CardiacAssist, Inc, Pittsburgh, Pennsylvania) in which the inflow cannula is in the LA and the outflow cannula is in the iliac artery (Table 52-10).196 The Impella 2.5 catheter crosses the AV and aspirates blood from the LV and ejects it into the ascending aorta with the inflow and outflow orifices being contained in a single catheter. For RV support, blood is drained from the right atrium (RA) or ventricle to the pump and returned to the main pulmonary artery. The pumps come in 2 basic varieties: pulsatile and axial (continuous) flow.

How VADs Are Selected

The duration of support, patient body habitus, ventricle(s) to be supported, surgical preference, and reliability are the main determinants for choosing a particular VAD (Table 52-10). Some devices, such as the pVAD and the Impella 2.5 device, are meant for short-term LV support and can be implanted percutaneously in the cardiac catheterization laboratory or operating room. These devices provide time to transfer the patient to a tertiary center, or may be used for support during high-risk interventions in the cardiac cath laboratory. The ABIOMED BVS 5000, CentriMag, and BioMedicus VADs provide an intermediate duration of support, from 7 to 14 days. They can support the right, left, or both ventricles. Because only a slit is made in the ventricle or atrium and a purse string suture is used to secure the inflow cannula, insertion of these VAD cannulas is less traumatic to the tissue, and explantation is much easier compared to the Thoratec or HeartMate devices in which the LV apex must be cored out in order to place the inflow cannula and upon explantation a patch repair of the apex made. The Thoratec (Thoratec Laboratories, Pleasantville, California) is the only Food and Drug Administration (FDA)–approved device that can provide long-term (months to years) biventricular support to serve as a bridge to transplant. It can be used to support the right, left, or both ventricles. The pneumatic HeartMate XVE provides only left ventricular support and provides long-term support as either a bridge to transplant or as destination therapy. Because the HeartMate XVE is implanted intracorporeally, in the left upper quadrant of the abdomen in the preperitoneal or intra-abdominal compartment, it is only suitable for patients with a body surface area greater than 1.8 m2. The HeartMate II continuous-flow device may be used for long-term support in patients awaiting heart transplant185 or as an alternative to the HeartMate XVE for destination therapy.187 The advantages of the HeartMate II are that it can be implanted in patients whose body surface area is less than 1.8 m2. It requires less dissection to implant than the HeartMate XVE. This may translate into lower infection rates. It has no bearings, and this design has translated into less mechanical failure, less need for reoperation, and higher 2-year survival rates when compared to the HeartMate XVE. In a study by Slaughter et al, 21 out of 59 HeartMate XVE pneumatic LVADs implanted required replacements because of bearing wear, valve malfunction, or infection.187 In the same study, 13 out of 133 HeartMate II continuous-flow LVADs required replacement owing to breakage of the percutaneous lead (n = 11), pump thrombosis (n = 2), and outflow elbow disconnection (n = 1).187 Pump thrombosis and thromboembolism are issues for both devices, but the HeartMate II requires higher levels of anticoagulation and is associated with an increased risk of bleeding secondary to anticoagulation. Continuous-flow pumps with diminished pulsatile flow have not been shown to have unfavorable effects on major organ function.197

Modes of Operation

Most pneumatic VADs operate in a "volume mode" in which the VAD ejects as soon as the chamber is filled. VADs function best with slightly increased intravascular volume and slightly decreased intravascular resistance. Hypovolemia slows the rate of filling, thereby decreasing the number of pump cycles per minute and overall VAD output. Increased vascular resistance may impede VAD ejection, resulting in prolonged ejection or decreased chamber emptying and decreased pump output. Incomplete chamber emptying may result in blood stasis in the pump chamber with resultant thromboembolic risk. In the case of the LVAD, systemic vascular resistance is the main component of afterload, and for RVAD, pulmonary vascular resistance is the major determinant of afterload.

The ABIOMED pump fills by continuous gravity drainage. The pump will automatically eject when full. It is preprogrammed to deliver approximately 5 L/min of flow. There are no dials to adjust. Anticoagulation with heparin is mandatory to prevent thrombus formation on the mechanical valves in the pump. The goal activated clotting time (ACT) is between 180 and 200 seconds. ABIOMED filling is independent of the underlying cardiac rhythm. Patients with malignant arrhythmias with biventricular support will remain hemodynamically stable.

The Thoratec pump is filled by vacuum-assisted drainage. The pump can operate in 1 of 3 operating modes. In "volume mode" the pump will eject as soon as it is full-similar to the ABIOMED. In "asynchronous mode" the device operates in a fixed manner according to defined variables. The "external synchronization mode" is used to provide counterpulsation during device weaning. Anticoagulation is initially achieved with heparin with a goal ACT of between 180 and 200 seconds, but once the patient is tolerating oral medications, anticoagulation can be maintained with coumadin with a goal INR of 2.5 to 3.5.198

The HeartMate XVE pneumatic LVAD has an extremely sophisticated control algorithm that adjusts device output to the patient's needs. As long as intravascular volume and the RV are functioning properly, there is no need to adjust the control settings. The CentriMag continuous-flow VAD is based on the "bearingless motor" technology that combines the drive, magnetic bearing, and rotor function in a single unit. The motor generates the magnetic bearing force that levitates the impeller in the pump housing while also generating the torque necessary for unidirectional flow. The maximal flow rate is 10 L/min. The Impella 2.5 is a continuous-flow device that can deliver 2.5 L/min of flow.

The HeartMate II and Jarvik 2000 are axial-flow devices using an impeller rotating at 6000 to 15000 rpm (revolution per minute) with maximum outputs of 10 L/min. The impeller of the device continues at its set speed regardless of inflow to the ventricular cannula. Delivery of blood to the inflow cannula is dependent on adequate intravascular volume, RV function, and a properly positioned cannula.

Preanesthetic Considerations

Most patients presenting for VAD implantation are critically ill whether they are arriving from intensive care units, emergency rooms, or the cardiac catheterization laboratory, or have developed postcardiotomy cardiogenic shock (PCCS) in the operating room. When indicated, VAD insertion should be performed early in the setting of PCCS. If a patient is weaned from CPB on 2 high-dose inotropes, hospital mortality is 42% and increases to 80% if 3 high-dose inotropes are used to separate from CPB.191,192 If these patients undergo VAD insertion within 3 hours of the first attempt to wean from CPB, a VAD wean rate of 60% and a hospital discharge rate of 43% can be achieved.191,192 If VAD insertion is delayed, a VAD wean rate of 27% and hospital discharge rate of 7% are to be expected. The delay in VAD insertion results in end-organ damage and decreased survival.191,192

The anesthesiologist must identify any neurologic deficits and ascertain the extent of major organ dysfunction (renal or hepatic insufficiency), which is very common in this patient population. Any further deterioration in the perioperative period may prevent a full recovery and eliminate the possibility of a patient qualifying for heart transplantation at a later date.198

Strict sterile technique for all invasive procedures and appropriate antibiotic prophylaxis is mandatory in this high-risk population. Antibiotic therapy is ineffective in treating assist device infection once it occurs.

Central venous pressure monitoring may be of questionable accuracy in patients with an RVAD. However, central access may still be useful for drug and volume infusions. Pulmonary artery catheters provide little useful information in the patient with an LVAD (that displays cardiac output). However, they may be of some use in patients with pulmonary hypertension at risk for developing RV failure post LVAD implantation. PA catheters should not be placed in patients with a functioning RVAD. Transesophageal echocardiography is the intraoperative monitor of choice.

The Pre-CPB Intraoperative Tee Examination

The pre-CPB transesophageal echocardiographic assessment occurs in the operating room prior to assist device implantation. The goal of this assessment is to identify anatomic and physiologic abnormalities that could lead to postoperative complications or VAD malfunction.199,200

Aortic Insufficiency (AI)

Over 20% of patients scheduled for LVAD insertion have some degree of AI.199,200 Aortic insufficiency status post LVAD implantation results in high pump-flow rates secondary to increased LVAD preload from the regurgitant volume but decreased systemic blood flow. Hypotension, organ hypoperfusion, and metabolic acidosis may occur despite high VAD output. Some centers report oversewing the valve shut, or repairing or replacing the aortic valve if AI is at least mild to moderate. Oversewing the AV is considered if the LVAD is being implanted as a bridge to transplant. However, the outcome of LVAD failure in the setting of an oversewn AV is ominous. If the valve cannot be repaired and is to be replaced, bioprosthetic valves are preferred to mechanical valves because of their lower thrombotic potential. Perioperative TEE is used to identify and quantify the severity of AI. General anesthesia, blood loss, and low cardiac output can result in an underestimation of load-dependent echocardiographic measurements. Therefore, if AI is suspected or mild, an LV vent may be used instead of echocardiography to evaluate the severity of AI. Vent flow rates exceeding 1.5 L/min indicate the need for surgical correction.201

There are a number of reasons why AI can worsen after LVAD implantation. After LVAD implantation the aorta to left ventricular pressure gradient is increased because the device ejects blood into the aorta at arterial pressures, but during device diastole, subphysiologic pressures are recorded in the LV. Additionally a properly functioning LVAD unloads the LV. Aortic valve opening is minimized and commissural fusion is accompanied by valve distortion, and AI may result.

Patent Foramen Ovale and Interatrial Septal Aneurysm

A patent foramen ovale (PFO) is found in up to 9% of all prepump TEEs for VAD placement.199,200 Under normal physiologic conditions, LA pressure is greater than RA pressure, and flow is left to right through the defect. Arterial saturation remains normal. However, with initiation of LVAD inflow, LA and LV pressures fall dramatically and frequently become lower than right-sided pressures. Right-to-left shunting through the PFO occurs. Undiagnosed or unrepaired PFOs may result in significant arterial hypoxemia presenting during separation from CPB and LVAD device activation (Fig. 52-25).202 The presence of a PFO presents an additional risk of paradoxical embolism, particularly in LVAD patients.

Interatrial septal aneurysm (IASA) is defined echocardiographically as protrusion of the aneurysm 10 mm beyond the plane of the atrial septum. According to 1 study, as many as 85% of patients with IASA have a coexisting PFO.203 IASA is associated with PFO and may be a cardiogenic source of thromboembolus.204 A bubble test in which 10 cc of agitated saline is injected into the RA immediately after release of a Valsalva maneuver is considered the gold standard for PFO identification. However, this test occasionally fails to identify a PFO, and it is possible that a PFO may only be identified after VAD insertion and the associated changes in hemodynamics.199,200 Repair of a PFO or interatrial septal aneurysm requires dual cannulation and necessitates prompt identification and communication with the surgical team. A PFO identified post CPB necessitates a return to CPB for repair.

Right Heart Failure

There is a 20% to 30% incidence of RV failure in patients implanted with an isolated LVAD.206 Decompression of the left side of the heart after LVAD activation can lead to altered RV geometry. A leftward shift of the interventricular septum after LV decompression results in increased RV compliance and decreased RV contractility.205 Additionally, although an LVAD should decrease RV afterload and thereby improve RV function in patients with normal PVR, patients with fixed PVR may actually experience an increase in RV afterload as increased left-sided output from the VAD increases venous return and flow through the pulmonary vasculature.

Right heart failure is one of the most important causes of perioperative death in LVAD patients.206,207 Blood flow into the LVAD is dependent on adequate pressure gradients to drive blood flow between the LV and device reservoir/pump. Right heart failure leads to diminished transpulmonary flow and decreased left-sided heart pressures, thus reducing the driving force for LVAD filling. A preoperative RV fractional area change less than 20% is associated with RV failure upon LVAD device activation.199,200 Low preoperative mean PA pressures and RV stroke work index may also be used to identify patients at risk for developing RV failure post LVAD implantation.208

Previous studies have demonstrated that female gender, nonischemic etiology of LV dysfunction, elevated central venous pressure, low mean pulmonary pressure, and low RV stroke work index are independent predictors of RV failure or need for RV assist device (RVAD) support after LVAD implantation. Preoperative evaluation of TR and RV geometry may help select patients who would benefit from biventricular support even if it is temporary. In a case series by Loforte et al, the authors reported using the CentriMag VAD for RVAD support in patients receiving a HeartMate II for LV support.209 In cases of preoperative alteration of RV function as assessed by echocardiography, RVADs should be placed even prophylactically to avoid RV dilation and irreversible cardiomyocyte stretching that occur with a delayed RVAD insertion. RVADs were successfully explanted in these study patients on an average of 17 days postoperatively following weaning from nitric oxide in a stepwise fashion on postoperative days 1 to 2, administration of sildenafil, and weaning from CentriMag starting on postoperative day 13 by reducing RVAD flow by about 10% every 12 hours.

Thrombus

In addition to sepsis and device malfunction, neurologic dysfunction including stroke and transient ischemic attacks is one of the most common adverse events following assist device implantation.210 Intracavitary thrombus has been identified as a thromboembolic risk factor in patients with assist devices (Fig. 52-26).211 Ventricular or LA thrombus was found in 9.4% and 3%, respectively, of patients prior to LVAD insertion.199,200

Left ventricular thrombus may develop under conditions of blood stasis or low-velocity flow. Apical akinesis, ventricular aneurysm, and diffuse ventricular dysfunction (EF <20%) are conditions associated with blood flow stasis and thrombus formation. The sensitivity of echocardiography for detecting LV thrombi is operator dependent. Even when an LV thrombus is not identified on echocardiographic examination, the likelihood of thrombus formation remains high in the aforementioned patient population. Evidence of apical flow stasis or of continuous swirling of flow around the apex identifies patients at risk for apical thrombus. Certain clues should help the operator identify apical thrombus and distinguish it from apical trabeculations. A thrombus is somewhat more echogenic than myocardium, has a distinct contour from the endocardial border, and is located in a region of wall motion abnormality. Intracavitary thrombus can also lead to VAD inflow cannula obstruction and mechanical pump failure.212 Thus it is important to communicate with the surgical team regarding the intended site of cannula insertion and to inspect that area carefully.

Miscellaneous Valvular Pathology

Tricuspid regurgitation is commonly found in patients prior to VAD insertion. However, unless ascites is present, no benefit has been found in repairing this lesion.207 As LV failure improves with device support, restoration of RV function and resolution of TR often follow. Mitral stenosis can compromise device inflow and should be corrected at the time of implant. Of note, for patients temporarily supported by an Impella 2.5 device, a catheter-induced incidence of MS has been reported. Impella catheter migration can lead to the prevention of normal anterior mitral leaflet excursion and result in MS. Mitral regurgitation often improves with LVAD activation and with prolonged support.213

Coronary Artery Disease

Coronary artery disease is common in LVAD candidates. Adequate evaluation of CAD is important to ensure the best outcome after VAD implantation. Right-sided bypass grafts may be particulary necessary to support RV function after LVAD implantation. If a CABG is to be performed, placement of the proximal anastamosis site should take into account the LVAD outflow cannula anastomosis site. The lesser curvature of the aorta has been suggested as a good location for the proximal anastomosis site if the VAD outflow cannula is to be sutured to the anterolateral aspect of the ascending aorta.207

Ascending Aorta

The ascending aorta should be evaluated for atheromatous plaque as well as aneurysm. Grade 3 plaque or higher has been identified as increasing the incidence of postoperative stroke from embolization associated with manipulation of the ascending aorta. Sites with the lowest plaque burden should be identified for LVAD outflow cannula anastomosis.

Weaning from CPB and Initiating VAD Operation

De-Airing

Before activation of the LVAD, the outflow graft is slowly unclamped and the pump is hand cranked. A modified TEE AV long-axis view at 130 degrees is used to monitor air efflux from the outflow graft into the ascending aorta (Fig. 52-27). If large quantities of air are observed, the hand cranking is stopped, the outflow graft is reclamped, and the aorta is vented. This process is repeated until minimal air is observed. The pump is then activated at a low rate, approximately 4 L/min.

Assessing RV Function

In the early minutes after LVAD pump activation, the critical issue is RV function. RV failure is often associated with significant dilation and is often accompanied by acute TR. If transpulmonary blood flow does not improve and the device continues to empty the left-sided chambers, these 2 chambers will collapse, which may lead to obstruction of the inflow cannula. If the device pump continues to operate, air may be entrained from the sewing ring or the inlet cannula connections and ejected in large quantities into the aorta.199,200 If these conditions exist, the device should be turned off and the patient returned to CPB. β-Adrenergic agonists, phosphodiesterase inhibitors, and nitric oxide should be initiated to improve RV contractility and decrease RV afterload in an attempt to improve transpulmonary blood flow. After several minutes, the LVAD can be restarted. If the same hemodynamic conditions persist, a supplemental temporary RVAD should be considered.

Inflow Cannula Position

Inflow cannula patency is essential to adequate stroke volume and device output. The LVAD inflow cannula orifice should be directed at the mitral opening and be located in the center of the apex (Fig. 52-28).195 In practice, the inflow cannula is often misdirected anteroseptally (Fig. 52-29). Apical trabeculations, thrombi, papillary muscles, and surgical misplacement can lead to cannula obstruction and LVAD stroke volume reduction.

Inflow cannula obstruction can be assessed using color-flow Doppler and looking for turbulent flow. Continuous-wave Doppler in the midesophageal TEE 4-chamber view may be used to identify cannula obstruction. Using continuous-wave Doppler and orienting the ultrasound beam down the center of the cannula, peak velocities greater than 2.5 m/s are consistent with obstruction and require repositioning of the cannula.199,200

Patent Foramen Ovale

PFOs may go undetected in the preoperative examination, as discussed previously. Arterial oxygen desaturation following LVAD initiation should raise suspicion of an undiagnosed PFO. Under these conditions, color-flow Doppler and contrast studies should be performed again, and if a PFO is identified, CPB should be reinstituted and the defect should be repaired.

Troubleshooting

Hypotension and Low Pump Flow Rates

Hypovolemia, RV failure, inflow cannula obstruction, pulmonary embolus, and cardiac tamponade are the most frequent causes of hypotension associated with low pump flow rates. TEE clues to pericardial tamponade include pericardial effusion, systolic collapse of the RA, and diastolic collapse of the RV. These echocardiographic findings should be correlated with low QRS voltage on the ECG and equalization of chamber pressures by catheterization.214 Pulmonary embolus may be difficult to distinguish from acute RV failure because both result in RV dilation and severe TR. However, in pulmonary embolus it may be possible to identify an echogenic density in the main or proximal branch right pulmonary artery. Hypovolemia is best assessed by TEE. The end-diastolic area is a more accurate measure of LV preload than the PA catheter occlusion pressure. The range of normal values for the end-diastolic area is wide. Therefore, a fluid challenge may be needed.

Device malfunction is the third most common cause of death in long-term HeartMate pulsatile pneumatic VAD-assisted patients. Two fairly uncommon device malfunctions may also lead to hypotension with decreased pump flow rates: inflow graft obstruction and outflow graft kinking. Clues to inflow conduit obstruction can be gained by assessing color-flow Doppler of the normal laminar diastolic filling pattern into the apical cannula. In cases of cannula obstruction this flow pattern is intermittently interrupted. Pulsed Doppler of the inflow cannula may also indicate periods of obstruction to flow. The most common cause of inflow cannula obstruction is deviation of the inflow cannula toward the interventricular septum (see Fig. 52-29). Outflow graft obstruction may occur and is best diagnosed using contrast angiography to demonstrate an acute angle bend or "kinking."

Hypotension and High Pump Flow Rates

Hypotension and high pump flow rates may present in either immediately post-CPB or after patients have been discharged from the hospital when the natural course of assist device decline and malfunction occurs. In LVAD patients who are both hypotensive and demonstrate high VAD flow rates, it is important to determine whether or not the right- and left-sided cardiac outputs are equal. This may mean placement of a PA catheter to determine the right-sided cardiac output, or performance of a TEE or transthoracic echocardiographic (TTE) examination with calculation of right-sided cardiac output. High biventricular flow rates suggest sepsis. If left-sided is greater than right-sided cardiac output, 1 of 3 etiologies must be ruled out: AI (Fig. 52-30), inflow valve regurgitation (IVR), or outflow valve regurgitation (OVR). In 1 prospective study, 26% of HeartMate pulsatile pneumatic LVAD patients discharged from the hospital developed inflow valve regurgitation, 32% developed new aortic insufficiency, and 0% were identified as having outflow valve incompetence.215

As mentioned earlier, AI may develop in LVAD patients secondary to native valve distortion, after LV decompression and diminished ejection through the native valve. Inflow valve incompetence may be caused by endocarditis, a torn cusp, or commissural dehiscence of the prosthetic. Although the inflow valve cannot be directly viewed by TEE, clues to incompetency may be gained by placing the color Doppler window over the orifice of the inflow cannula during device systole. Backflow into the ventricle during device systole suggests inflow valve incompetence (Fig. 52-31).216 Aortic valve opening may be a clue that inflow valve regurgitation exists. Sixty-five percent of LVAD patients with IVR had frequent AV opening compared with only 19% in patients without IVR.215 The outflow graft velocity, velocity time integral, and stroke volume are all significantly diminished in patients with IVR.

Of special note is a phenomenon with the continuous-flow Jarvik 2000 (Jarvik Heart, New York, New York) coined the "suckdown effect."217 This effect may occur in any continuous-flow devices, including the CentriMag and the HeartMate II. If the revolutions per minute of the device are increased too quickly, without adequate preload a "suck-down" effect occurs in which the ventricular walls collapse and contact the LVAD inflow, leading to obstruction. The ventricular septum bows to the left and RV assumes an unfavorable geometry, leading to RV dysfunction further worsening LV preload. Although the natural tendency may be to increase the revolutions per minute on a continuous-flow device with a patient who is hemodynamically unstable, it will actually worsen the clinical picture if the cause is inadequate volume. The correct maneuver is to decrease the revolutions per minute and restore intravascular volume.217

Another unusual complication of a continuous-flow device was reported by Badiwala et al.218 Three days post implantation of a HeartMate II device, the patient became profoundly hypotensive despite the LVAD at 9800 rpm. Echocardiography showed a poorly decompressed LV with paradoxical motion of the interventricular septum. The LVAD was adjusted to increase flow with decompression of the LV and septum returning to midline. Despite these maneuvers, the patient continued to deteriorate over the next several days and was started on pressors. Echocardiographic examination again showed a dilated LV with a dyskinetic septum and mild to moderate RV hypokenisis with a CVP of 19 mm Hg. Shortly thereafter the patient became profoundly hypotensive and expired. On autopsy the cause of death was determined to be an acute ventricular septal defect caused by erosion of the LVAD's inflow cannula into the interventricular septum.

Anesthetic Management of VAD Patients for Noncardiac Surgery

As the population of assist-device patients grows, more of them will be requiring anesthesia for percutaneous insertion219 and for noncardiac surgery. These patients can be managed using basic anesthetic principles with a few minor adjustments. Because infection and thromboembolic events are the 2 most common causes of morbidity in long-term VAD-assisted patients, every effort should be made to correctly manage antibiotic prophylaxis, ensure the strictest sterile technique for invasive procedures, and manage anticoagulation therapy appropriately.

Patients with long-term intracorporeal devices, such as the HeartMate XVE, suffer from delayed gastric emptying and early satiety because of the proximity of the device to the stomach. Cricoid pressure and a rapid sequence induction should be performed for these individuals if general anesthesia is desired. Except for the HeartMate XVE, which has a titanium-covered pump chamber and requires only aspirin for anticoagulation, most other VADs will require the maintenance of some level of anticoagulation to avoid thromboembolic complications. Increased intraoperative bleeding should be anticipated. A discussion should occur between the surgeon and the anesthesiologist regarding coagulation management prior to surgery. Giving small amounts of fresh frozen plasma to achieve the lower limits of anticoagulation recommended by the device manufacturer may be helpful in controlling excessive blood loss. Regional techniques are usually contraindicated secondary to the need to maintain anticoagulation.

Basic anesthetic principles regarding decisions to intubate and extubate patients apply in VAD-assisted patients. Although positive pressure ventilation may affect venous return to the left heart and diminish LVAD filling, in most circumstances this is only a minor consideration and mechanical ventilation should be implemented if needed. In general, invasive monitoring is not necessary because the VAD control console continuously displays the device output. An arterial line may be inserted under strict sterile precautions if frequent blood gases will be needed or if large alterations in blood pressure are anticipated. Otherwise, a noninvasive blood pressure cuff is usually adequate. Finally, because the VAD-assisted patient likely was transported to the operating room on battery-assisted power, it is important to remember to reconnect the VAD to the power supply in the operating room to avoid battery failure.

• Borer JS, Bonow RO. Contemporary approach to aortic and mitral regurgitation. Circulation. 2003;108:2432-2438.
• De Hert SG, Van der Linden PJ, Cromheecke S, et al. Choice of primary anesthetic regimen can influence intensive care unit length of stay after coronary surgery with cardiopulmonary bypass. Anesthesiology. 2004;101:9-20.
• Duncan AE, Abd-Elsayed A, Maheshwari A, et al. Role of intraoperative and postoperative blood glucose concentrations in predicting outcomes after cardiac surgery. Anesthesiology. 2010;112:860-871.
• Eltzschig H, Rosenberger P, Löffler M, et al. Impact of intraoperative transesophageal echocardiography on surgical decision making in 12,566 cardiac surgical patients. Ann Thorac Surg. 2008;85:845-852.
• Estrera AL, Sheinbaum R, Miller CC, et al. Cerebrospinal fluid drainage during thoracic aortic repair: safety and current management. Ann Thorac Surg. 2009;88:9-15.
• Fukamachi K, McCarthy PM Smedira NG, et al. Preoperative risk factors for right ventricular failure after implantable left ventricular assist device insertion. Ann Thorac Surg. 1999;68:2181-2184.
• Hiratzka LF, Bakris GL, Beckman JA, et al. 2010 ACCF/AHA/AATS/ACR/ASA/SCA/SCAI/SIR/STS/SVM guidelines for the diagnosis and management of patients with thoracic aortic disease: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines, American Association for Thoracic Surgery, American College of Radiology, American Stroke Association, Society of Cardiovascular Anesthesiologists, Society for Cardiovascular Angiography and Interventions, Society of Interventional Radiology, Society of Thoracic Surgeons, and Society for Vascular Medicine. Circulation. 2010;121:e266-e369.
• Levy J, Dutton R, Hemphill J, et al. Multidisciplinary approach to the challenge of hemostasis. Anesth Analg. 2010;110:354-364.
• Mckinlay KH, Schinderle DB, Swaminathan M, et al. Predictors of inotrope use during separation from cardiopulmonary bypass. J Cardiothorac Vasc Anesth. 2004;18:404-408.
• Practice guidelines for perioperative transesophageal echocardiography. An updated report by the American Society of Anesthesiologists and the Society of Cardiovascular Anesthesiologists Task Force on Transesophageal Echocardiography. American Society of Anesthesiologists and Society of Cardiovascular Anesthesiologists Task Force on Transesophageal Echocardiography. Anesthesiology. 2010;112:1084-1096.
• Rady MY, Ryan T, Starr NJ. Perioperative determinants of morbidity and mortality in elderly patients undergoing cardiac surgery. Crit Care Med. 1998;26:225-235.
• Smith RC, Leung JM, Mangano DT. Postoperative myocardial ischemia in patients undergoing coronary artery bypass graft surgery. S.P.I. Research Group. Anesthesiology. 1991;74:464-473.
• Stone ME, Soong W, Krol M, Reich DL. The anesthetic considerations in patients with ventricular assist devices presenting for noncardiac surgery: a review of eight cases. Anesth Analg. 2002;95:42-49.