Chapter 4. Cardiac Surgical Pharmacology

Clinical pharmacology associated with cardiac surgery is an important part of patient management. Patients in the perioperative period receive multiple agents that affect cardiovascular and pulmonary function. This chapter summarizes the pharmacology of the agents commonly used for treating the primary physiologic disturbances associated with cardiac surgery, hemodynamic instability, respiratory insufficiency, and alterations of hemostasis. For cardiovascular drugs, the common theme is that pharmacologic effects are produced by intracellular ion fluxes.

Several basic subcellular/molecular pathways are important in cardiovascular pharmacology, as shown in Fig. 4-1. The action potential in myocardial cells is a reflection of ion fluxes across the cell membrane, especially Na+, K+, and Ca2+.1,2 Numerous drugs used to control heart rate and rhythm act by altering Na+ (eg, lidocaine and procainamide), K+ (eg, amiodarone, ibutilide, and sotalol), or Ca2+ (eg, diltiazem) currents. Calcium also has a dominant effect on the inotropic state.3,4

Myocardial contractility is a manifestation of the interaction of actin and myosin, with conversion of chemical energy from ATP hydrolysis into mechanical energy. The interaction of actin and myosin in myocytes is inhibited by the associated protein tropomyosin. This inhibition is "disinhibited" by intracellular calcium. A similar situation occurs in vascular smooth muscle, where the interaction of actin and myosin (leading to vasoconstriction) is modulated by the protein calmodulin, which requires calcium as a cofactor. Thus intracellular calcium has a "tonic" effect in both the myocardium and vascular smooth muscle.

Numerous drugs used perioperatively alter intracellular calcium.3,4

Catecholamines (eg, norepinephrine, epinephrine, and dobutamine) with beta agonist activity regulate intramyocyte calcium levels via the nucleotide cyclic adenosine monophosphate (cyclic AMP) (Fig. 4-2). Beta agonists bind to receptors on the cell surface that are coupled to the intracellular enzyme adenylate cyclase via the stimulatory transmembrane GTP-binding protein. This leads to increased cyclic AMP synthesis, and cyclic AMP, in turn, acts as a second messenger for a series of intracellular reactions resulting in higher levels of intracellular calcium during systole. Less well known is that drugs with only alpha-adrenergic agonist activity also may increase intracellular Ca2+ levels, although by a different mechanism.5,6 Although under investigation, the probable basis for the inotropic effect of alpha-adrenergic drugs is the stimulation of phospholipase C, which catalyzes hydrolysis of phosphatidyl inositol to diacylglycerol and inositol triphosphate (see Fig. 4-2). Both of these compounds increase the sensitivity of the myofilament to calcium, whereas inositol triphosphate stimulates the release of calcium from its intracellular storage site, the sarcoplasmic reticulum. There is still some debate about the mechanism for the inotropic effect of alpha-adrenergic agonists and its significance for the acute pharmacologic manipulation of contractility, but there is little debate about the importance of this mechanism in vascular smooth muscle, where the increase in intracellular calcium stimulated by alpha-adrenergic agonists can increase smooth muscle tone significantly. However, intracellular calcium in vascular smooth muscle is also controlled by cyclic nucleotides.7,8 In contrast to the myocyte, in vascular smooth muscle, cyclic AMP has a primary effect of stimulating the uptake of calcium into intracellular storage sites, decreasing its availability (Fig. 4-3). Thus drugs that stimulate cyclic AMP production (beta agonists) or inhibit its breakdown (phosphodiesterase inhibitors) will cause vasodilation. In addition, cyclic guanosine monophosphate (cyclic GMP) also increases intracellular calcium storage (see Fig. 4-3), decreasing its availability for modulating the interaction of actin and myosin. Several commonly used pharmacologic agents act via cyclic GMP. For example, nitric oxide stimulates the enzyme guanylate cyclase, increasing cyclic GMP levels. Drugs such as nitroglycerin and sodium nitroprusside achieve their effect by producing nitric oxide as a metabolic product. Vasodilation is also produced by "cross-talk" between K+ and Ca2+ fluxes. Decreased levels of ATP, acidosis, and elevated tissue lactate levels increase the permeability of the ATP-sensitive K+ channel. This increased permeability results in hyperpolarization of the cell membrane that inhibits the entry of Ca2+ into the cell. This results in decreased vascular tone.

The simplistic overview of pathways of cardiac pharmacology as summarized in Figs. 4-1 through 4-3 also suggests the primary cause of difficulty in the clinical use of the drugs discussed in this chapter. The mechanisms of action for control of heart rate and rhythm, contractility, and vascular tone are interrelated. For example, beta-adrenergic agonists not only increase intracellular calcium to increase contractility, but they also alter K+ currents, leading to tachycardia. Catecholamines not only have beta-adrenergic agonist activity, with inotropic and chronotropic effects, but they also possess alpha-agonist activity, leading to increased intracellular calcium in vascular smooth muscle and vasoconstriction. Phosphodiesterase inhibitors not only may increase contractility by increasing cyclic AMP in the myocyte, but they also may cause excessive vasodilation by increasing cyclic AMP in the vasculature. The interplay of the various mechanisms means the clinical art of cardiac surgical pharmacology lies as much in selecting drugs for their side effects as for their primary therapeutic effects.

Arrhythmias are common in the cardiac surgical period. A stable cardiac rhythm requires depolarization and repolarization in a spatially and temporally coordinated manner, and dysrhythmias may occur when this coordination is disturbed. The mechanisms for arrhythmias can be divided into abnormal impulse initiation, abnormal impulse conduction, and combinations of both.9,10 Abnormal impulse initiation occurs as a result of increased automaticity (spontaneous depolarization of tissue that does not normally have pacemaking activity) or as a result of triggered activity from abnormal conduction after depolarizations during phase 3 or 4 of the action potential. Abnormal conduction often involves reentry phenomena, with recurrent depolarization around a circuit owing to unilateral conduction block in ischemic or damaged myocardium and retrograde activation by an alternate pathway through normal tissue. In this simplistic view, it is logical that dysrhythmias could be suppressed by slowing the conduction velocity of ectopic foci, allowing normal pacemaker cells to control heart rate, or by prolonging the action potential duration (and hence refractory period) to block conduction into a limb of a reentry circuit.

A scheme proposed originally by Vaughan Williams and modified subsequently11,12 is used often to classify antidysrhythmic agents, and although alternative schemes describing specific channel-blocking characteristics have been proposed and may be more logical,13 this discussion is organized using the Vaughan Williams system of four major drug categories. In this scheme, class I agents are those with local anesthetic properties that block Na+ channels, class II drugs are beta-blocking agents, class III drugs prolong action potential duration, and class IV drugs are calcium entry blockers. Amiodarone is discussed in detail owing to its expanding role in treating both supraventricular and ventricular arrhythmias and because its use has replaced many of the previously used agents. Because of the efficacy of intravenous amiodarone and its recommendations in Advanced Cardiac Life Support (ACLS) guidelines, many of the older drugs used in cardiac surgery have a historical perspective and are considered briefly.

Class I Agents

Although each of the class I agents blocks Na+ channels, they may be subclassified based on electrophysiologic differences. These differences can be explained, to some extent, by consideration of the kinetics of the interaction of the drug and the Na+ channel.14,15 Class I drugs bind most avidly to open (phase 0 of the action potential; see Fig. 4-1) or inactivated (phase 2) Na+ channels. Dissociation from the channel occurs during the resting (phase 4) state. If the time constant for dissociation is long in comparison with the diastolic interval (corresponding to phase 4), the drug will accumulate in the channel to reach a steady state, slowing conduction in normal tissue. This occurs with class Ia (eg, procainamide, quinidine, and disopyramide) and class Ic (eg, encainide, flecainide, and propafenone) drugs. In contrast, for the class Ib drugs (eg, lidocaine and mexiletine), the time constant for dissociation from the Na+ channel is short, drug does not accumulate in the channel, and conduction velocity is affected minimally. However, in ischemic tissue, the depolarized state is more persistent, leading to greater accumulation of agent in the Na+ channel and slowing of conduction in the damaged myocardium.

Procainamide is a class Ia drug that has various electrophysiologic effects.16 Administration may be limited by the side effects of hypotension and decreased cardiac output.17,18 The loading dose is 20 to 30 mg/min, up to 17 mg/kg, and should be followed by an intravenous infusion of 20 to 80 mg/kg per minute. Because procainamide prolongs action potential duration, widening of the QRS complex often heralds a potential overdose. The elimination of procainamide involves hepatic metabolism, acetylation to a metabolite with antiarrhythmic and toxic side effects, and renal elimination of this metabolite. Thus the infusion rate for patients with significant hepatic or renal disease should be at the lower end of this range.

Class Ib drugs include what is probably the best-known antiarrhythmic agent, lidocaine. As noted, lidocaine is a Na+ channel blocker that has little effect on conduction velocity in normal tissue but slows conduction in ischemic myocardium.14,15 Other electrophysiologic effects include a decrease in action potential duration but a small increase in the ratio of effective refractory period to action potential duration. The exact role of these electrophysiologic effects on arrhythmia suppression is unclear. Lidocaine has no significant effects on atrial tissue, and it is not recommended for therapy in shock-resistant ventricular tachycardia/fibrillation (VT/VF) in the recent Guidelines for Emergency Cardiovascular Care.19 After an initial bolus dose of 1 to 1.5 mg/kg of lidocaine, plasma levels decrease rapidly owing to redistribution to muscle, fat, etc. Effective plasma concentrations are maintained only by following the bolus dose with an infusion of 20 to 50 mg/kg per minute.20 Elimination occurs via hepatic metabolism to active metabolites that are cleared by the kidneys. Consequently, the dose should be reduced by approximately 50% in patients with liver or kidney disease. The primary toxic effects are associated with the central nervous system, and a lidocaine overdose may cause drowsiness, depressed level of consciousness, or seizures in very high doses. Negative inotropic or hypotensive effects are less pronounced than with most other antiarrhythmics. The other class Ib drugs likely to be encountered in the perioperative period are the oral agents tocainide and mexiletine, which have effects similar to lidocaine.15

The class Ic agents, including flecainide, encainide, and propafenone, markedly decrease conduction velocity.20,21 The Cardiac Arrhythmia Suppression Trial (CAST) study20,21 of moricizine found that although ventricular arrhythmias were suppressed, the incidence of sudden death was greater than with placebo with encainide and flecainide, and these drugs are not in wide use. Propafenone is available for oral use in the United States, but IV is available in Europe. The usual adult dose is 150 to 300 mg orally every 8 hours. It has beta-blocking (with resulting negative inotropic effects) as well as NA+ channel-blocking activity; lengthens the PR, QRS, and QT duration; and may be used to treat both atrial and ventricular dysrhythmias.15

Class II Agents

Beta-receptor–blocking agents are another important group of antiarrhythmic (denoted class II in the Vaughan Williams scheme). However, because of their use as antihypertensive as well as antiarrhythmic agents, they are discussed elsewhere in this chapter, and we will move on to consider bretylium, amiodarone, and sotalol, the class III agents in the Vaughan Williams scheme. These drugs have a number of complex ion channel-blocking effects, but possibly the most important activity is K+ channel blockade.22 Because the flux of K+ out of the myocyte is responsible for repolarization, an important electrophysiologic effect of class III drugs is prolongation of the action potential.23

Class III Agents

Ibutilide, dofetilide, sotalol, and bretylium are class III agents. Intravenous ibutilide and oral dofetilide are approved for the treatment of atrial fibrillation but carry the risk of torsades de pointes.24,25 Sotalol is a nonselective beta-blocker that also has K+ channel-blocking activity.26 In the United States, it is now available for IV and previously for oral administration. The approved indication is for treating life-threatening ventricular arrhythmias, although it is effective against atrial arrhythmias as well. Bretylium is not used currently nor recommended in the most recent AHA Guidelines for Emergency Cardiovascular Care.19

Class IV Agents

Calcium entry blockers (class IV in the Vaughan Williams scheme), including verapamil and diltiazem, are antiarrhythmics. In sinoatrial and atrioventricular nodal tissue, Ca2+ channels contribute significantly to phase 0 depolarization, and the atrioventricular (AV) nodal refractory period is prolonged by Ca2+ entry blockade.27,28

This explains the effectiveness of verapamil and diltiazem in treating supraventricular arrhythmias. It is also clear why these drugs are negative inotropes. Both verapamil and diltiazem are effective in slowing the ventricular response to atrial fibrillation, flutter, and paroxysmal supraventricular tachycardia and in converting to sinus rhythm.29–31 Verapamil has greater negative inotrope effects than diltiazem; therefore, it is used rarely for supraventricular arrhythmias. The intravenous dose of diltiazem is 0.25 mg/kg, with a second dose of 0.35 mg/kg if the response is inadequate after 15 minutes. The loading dose should be followed by an infusion of 5 to 15 mg/h. Intravenous diltiazem, although useful for rate control, has been replaced by intravenous amiodarone in clinical therapy of supraventricular tachycardia (SVT) and prophylaxis (see Amiodarone).

Other Drugs

One of the difficulties of classifying antiarrhythmics by the Vaughan Williams classification is that not all drugs can be incorporated into this scheme. Three examples are digoxin, adenosine, and magnesium, each of which has important uses in the perioperative period.

Digoxin inhibits the Na+, K+-ATPase pump, leading to decreased intracellular K+, a less negative resting membrane potential, increased slope of phase 4 depolarization, and decreased conduction velocity. These direct effects, however, usually are dominated by indirect effects, including inhibition of reflex responses to congestive heart failure and a vagotonic effect.10,32 The net effect is greatest at the AV node, where conduction is slowed and the refractory period is increased, explaining the effectiveness of digoxin in slowing the ventricular response to atrial fibrillation. The major disadvantages of digoxin are the relatively slow onset of action and many side effects, including proarrhythmia effects, and it is now used rarely for rate control in acute atrial fibrillation because of the advent of IV amiodarone and diltiazem.

Adenosine is an endogenous nucleoside that has an electrophysiologic effect similar to that of acetylcholine. Adenosine decreases AV node conductivity, and its primary antiarrhythmic effect is to break AV nodal reentrant tachycardia.33 An intravenous dose of 100 to 200 μg/kg is the treatment of choice for paroxysmal supraventricular tachycardia. Adverse effects, such as bronchospasm, are short-lived because its plasma half-life is so short (1 to 2 seconds). This short half-life makes it ideal for treating reentry dysrhythmia, in which transient interruption can fully suppress the dysrhythmia.

Appropriate acid-base status and electrolyte balance are important because electrolyte imbalance can perturb the membrane potential, leading to arrhythmia generation, as can altered acid-base status, by effects on K+ concentrations and sympathetic tone. Therapy for dysrhythmia should include correction of acid-base and electrolyte imbalances. Magnesium supplementation should be considered.34 Magnesium deficiency is common in the perioperative period, and magnesium administration has been shown to decrease the incidence of postoperative dysrhythmia.35

Intravenous amiodarone has become one of the most administered intravenous antiarrhythmics used in cardiac surgery because of its broad spectrum of efficacy. Amiodarone was developed originally as an antianginal agent because of its vasodilating effects, including coronary vasodilation.36 It has various ion channel-blocking activities.10,29,36 The resulting electrophysiologic effects are complex, and there are differences in acute intravenous and chronic oral administration. Acute intravenous administration can produce decreases in heart rate and blood pressure, but there are minimal changes in QRS duration or QT interval. After chronic use, there may be significant bradycardia and increases in action potential duration in AV nodal and ventricular tissue, with increased QRS duration and QT interval.37–39

Pharmacokinetics

Amiodarone is a complex highly lipophilic drug that undergoes variable absorption (35 to 65%) after oral administration and is taken up extensively by multiple tissues with interindividual variation and complex pharmacokinetics.38–40 The short initial context-sensitive half-life after intravenous administration represents drug redistribution. The true elimination half-life for amiodarone is extremely long, up to 40 to 60 days. Because of the huge volume of distribution (~60 L/kg) and a long duration of action, an active metabolite loading period of several months may be required before reaching steady-state tissue concentrations. Further, in life-threatening arrhythmias, intravenous loading often is starting to establish initial plasma levels. Measuring amiodarone plasma concentrations is not useful owing to the complex pharmacokinetics and the metabolites of the parent drug. Plasma concentrations greater than 2.5 mg/L have been associated with an increased risk of toxicity. The optimal dose of amiodarone has not been well characterized and may vary depending on the specific arrhythmias treated. Further, there may be differences in dose requirements for therapy of supraventricular and ventricular arrhythmias.37–40

Because of these distinctive pharmacokinetic properties, steady-state plasma levels are achieved slowly. Oral administration for a typical adult consists of a loading regimen of 80 to 1600 mg/d (in two or three doses) for 10 days, 600 to 800 mg/d for 4 to 6 weeks, and then maintenance doses of 200 to 600 mg/d. For intravenous loading, specific studies will be reviewed, but recommended dosing is 150 mg given over 10 minutes for acute therapy in an adult, followed first by a secondary loading infusion of 60 mg/h for 6 hours and then by a maintenance infusion of 30 mg/h to achieve a 1000 mg/d dosing.37–40

Electrophysiology

The electrophysiologic actions of amiodarone are complex and incompletely understood. Amiodarone produces all four effects according to the Vaughan Williams classification. It also has been shown to have use-dependent class I activity, inhibition of the inward sodium currents, and class II activity.10 The antiadrenergic effect of amiodarone, however, is different from that of beta-blocker drugs because it is noncompetitive and additive to the effect of beta-blockers. Amiodarone depresses sinoatrial (SA) node automaticity, which slows the heart rate and conduction and increases refractoriness of the AV node, properties useful in managing supraventricular arrhythmia. Its class III activity results in increases in atrial and ventricular refractoriness and prolongation of the QTc interval. The effects of oral amiodarone on SA and AV nodal function are maximal within 2 weeks, whereas the effects on ventricular tachycardia (VT) and ventricular refractoriness emerge more gradually during oral therapy, becoming maximal after 10 weeks or more.

Indications

The primary indication for amiodarone is ventricular tachycardia or fibrillation.40–48 It is also effective for the treatment of atrial dysrhythmias and in treating atrial fibrillation (see Atrial Fibrillation).

Side Effects

Although there are numerous adverse reactions to amiodarone, they occur with long-term oral administration and are rarely associated with acute intravenous administration. The most serious is pulmonary toxicity, which has not been reported with acute administration in a perioperative setting. Some case series have reported an increased risk of marked bradycardia and hypotension immediately after cardiac surgery in patients already on amiodarone at the time of surgery.49,50 Other case-control studies, however, have not reproduced this finding.51 None of the placebo-controlled trials of prophylactic amiodarone for perioperative atrial fibrillation prevention report adverse cardiovascular effects, although bradycardia and hypotension are known side effects.52–56 Case reports and case series of postoperative acute pulmonary toxicity are similarly lacking in the rigor of randomized, controlled methodology.

Ventricular Tachyarrhythmias

Intravenous amiodarone is approved for rapid control of recurrent VT or VF. Three randomized, controlled trials of patients with recurrent in-hospital, hemodynamically unstable VT or VF with two or more episodes within the past 24 hours who failed to respond to or were intolerant of lidocaine, procainamide, and (in two of the trials) bretylium have been reported.42,44,46 Patients were critically ill with ischemic cardiovascular disease, 25% were on a mechanical ventilator or intra-aortic balloon pump before enrollment, and 10% were undergoing cardiopulmonary resuscitation at the time of enrollment. One study compared three doses of IV amiodarone: 525, 1050, and 2100 mg/d.44 Because of the use of investigator-initiated intermittent open-label amiodarone boluses for recurrent VT, the actual mean amiodarone doses received by the three groups were 742, 1175, and 1921 mg/d. There was no statistically significant difference in the number of patients without VT/VF recurrence during the 1-day study period: 32 of 86 (41%), 36 of 92 (45%), and 42 of 92 (53%) for the low-, medium-, and high-dose groups, respectively. The number of supplemental 150-mg bolus infusions of amiodarone given by blinded investigators was statistically significantly less in those randomized to higher doses of amiodarone.

A wider range of amiodarone doses (125, 500, and 1000 mg/d) was evaluated by Sheinman and colleagues, including a low dose that was expected to be subtherapeutic.46 This stronger study design, however, also was confounded by open-label bolus amiodarone injections given by study investigators. There was, however, a trend toward a relationship between intended amiodarone dose and VT/VF recurrence rate (p =.067). After adjustment for baseline imbalances, the median 24-hour recurrence rates of VT/VF, from lowest to highest doses, were 1.68, 0.96, and 0.48 events per 24 hours.

The third study compared two intravenous amiodarone doses (125 and 1000 mg/d) with bretylium (2500 mg/d).42 Once again, the target amiodarone dose ratio of 8:1 was compressed to 1.8:1 because of open-label boluses. There was no significant difference in the primary outcome, which was median VT/VF recurrence rate over 24 hours. For low-dose amiodarone, high-dose amiodarone, and bretylium, these rates were 1.68, 0.48, and 0.96 events per 24 hours, respectively (p =.237). There was no difference between high-dose amiodarone and bretylium; however, more than 50% of patients had crossed over from bretylium to amiodarone by 16 hours.

The failure of these studies to provide clear evidence of amiodarone efficacy may be related to the "active-control study design" used, a lack of adequate statistical power, high rates of supplemental amiodarone boluses, and high crossover rates. Nonetheless, these studies provide some evidence that IV amiodarone (1 g/d) is moderately effective during a 24-hour period against VT and VF.

Sustained Monomorphic Ventricular Tachycardia and Wide QRS Tachycardia

Although the most effective and rapid treatment of any hemodynamically unstable sustained ventricular tachyarrhythmia is electrical cardioversion or defibrillation, intravenous antiarrhythmic drugs can be used for arrhythmia termination if the VT is hemodynamically stable. The Guidelines for Emergency Cardiovascular Care19 has removed former recommendation of lidocaine and adenosine use in stable wide QRS tachycardia, now labeled as "acceptable" but not primarily recommended (lidocaine) or not recommended (adenosine). Intravenous procainamide and sotalol are effective, based on randomized but small studies;10 amiodarone is also considered acceptable.19

Shock-Resistant Ventricular Fibrillation

The Guidelines for Emergency Cardiovascular Care recommends at least three shocks and epinephrine or vasopressin before any antiarrhythmic drug is administered.10,19 No large-scale controlled, randomized studies have demonstrated efficacy for lidocaine, bretylium, or procainamide in shock-resistant VF,10,19 and lidocaine and bretylium are no longer recommended in this setting.19 Two pivotal studies have been reported recently studying the efficacy of agents in acute shock-resistant cardiac arrest.

The Amiodarone in the Out-of-Hospital Resuscitation of Refractory Sustained Ventricular Tachycardia (ARREST) study was randomized, double-blind, and placebo-controlled. The ARREST study in 504 patients showed that amiodarone 300 mg administered in a single intravenous bolus significantly improves survival to hospital admission in cardiac arrest still in VT or VF after three direct-current shocks (44% versus 34%; p <.03).43 Although the highest survival rate to hospital admission (79%) was achieved when the amiodarone was given within 4 to 16 minutes of dispatch, there was no significant difference in the proportional improvement in the amiodarone group compared with the placebo group when drug administration was delayed (up to 55 minutes). Amiodarone also had the highest efficacy in patients (21% of all study patients) who had a return of spontaneous circulation before drug administration (survival to hospital admission increased to 64% from 41% in the placebo group). Among patients with no return of spontaneous circulation, amiodarone only slightly improved outcome (38% versus 33%).

Dorian performed a randomized trial comparing intravenous lidocaine with intravenous amiodarone as an adjunct to defibrillation in victims of out-of-hospital cardiac arrest.48 Patients were enrolled if they had out-of-hospital ventricular fibrillation resistant to three shocks, intravenous epinephrine, and a further shock or if they had recurrent ventricular fibrillation after initially successful defibrillation. They were randomly assigned in a double-blind manner to receive intravenous amiodarone plus lidocaine placebo or intravenous lidocaine plus amiodarone placebo. The primary end point was the proportion of patients who survived to be admitted to the hospital. In total, 347 patients (mean age 67 ± 14 years) were enrolled. The mean interval between the time at which paramedics were dispatched to the scene of the cardiac arrest and the time of their arrival was 7 ± 3 minutes, and the mean interval from dispatch to drug administration was 25 ± 8 minutes. After treatment with amiodarone, 22.8% of 180 patients survived to hospital admission compared with 12.0% percent of 167 patients treated with lidocaine (p = .009). Among patients for whom the time from dispatch to the administration of the drug was equal to or less than the median time (24 minutes), 27.7% of those given amiodarone and 15.3% of those given lidocaine survived to hospital admission (p = .05). The authors concluded that compared with lidocaine, amiodarone leads to substantially higher rates of survival to hospital admission in patients with shock-resistant out-of-hospital ventricular fibrillation.

Supraventricular Arrhythmias

A supraventricular arrhythmia is any tachyarrhythmia that requires atrial or atrioventricular junctional tissue for initiation and maintenance. It may arise from reentry caused by unidirectional conduction block in one region of the heart and slow conduction in another, from enhanced automaticity akin to that seen in normal pacemaker cells of the sinus node and in latent pacemaker cells elsewhere in the heart, or from triggered activity, a novel type of abnormally enhanced impulse initiation caused by membrane currents that can be activated and inactivated by premature stimulation or rapid pacing.56–58 Pharmacologic approaches to treating supraventricular arrhythmias, including atrial fibrillation, atrial flutter, atrial tachycardia, AV re-entrant tachycardia, and AV nodal re-entrant tachycardia, continue to evolve.56–60 Because atrial fibrillation is perhaps the most common arrhythmia after cardiac surgery, this condition is emphasized in detail.

Atrial Fibrillation

Atrial fibrillation (AF) is a common complication of cardiac surgery that increases the length of stay in the hospital with resulting increases in health-care resource utilization.56–61 Advanced age, previous AF, and valvular heart operations are the most consistently identified risk factors for this arrhythmia. Because efforts to terminate AF after its initiation are problematic, current interests are directed at therapies to prevent postoperative AF. Most studies suggest that prophylaxis with antiarrhythmic compounds can decrease the incidence of AF, length of hospital stay, and cost significantly. Class III antiarrhythmic drugs (eg, sotalol and ibutilide) also may be effective but potentially pose the risk of drug-induced polymorphic ventricular tachycardia (torsades de pointes). Newer promising intravenous agents including vernakalant are also being investigated. Defining which subpopulations benefit most from such therapy is important as older and more critically ill patients undergo surgery. Intravenous sotalol is currently available in the United States.

Amiodarone is also an effective approach for prophylactic therapy of AF. Intravenous amiodarone is an important consideration because loading with oral therapy is often not feasible in part owing to time required. There also may be added benefits of prophylactic therapies in high-risk patients, especially those prone to ventricular arrhythmias (ie, patients with pre-existing heart failure).

Two studies deserve mention regarding prophylaxis with amiodarone. To determine if IV amiodarone would prevent atrial fibrillation and decrease hospital stay after cardiac surgery, Daoud and colleagues assessed preoperative prophylaxis in 124 patients who were given either oral amiodarone (64 patients) or placebo (60 patients) for a minimum of 7 days before elective cardiac surgery.62 Therapy consisted of 600-mg amiodarone per day for 7 days and then 200 mg/d until the day of discharge from the hospital. The preoperative total dose of amiodarone was 4.8 ± 0.96 g over 13 ± 7 days. Postoperative atrial fibrillation occurred in 16 of the 64 patients in the amiodarone group (25%) and 32 of the 60 patients in the placebo group (53%). Patients in the amiodarone group were hospitalized for significantly fewer days than were patients in the placebo group (6.5 ± 2.6 versus 7.9 ± 4.3 days; p = .04). Total hospitalization costs were significantly less for the amiodarone group than for the placebo group ($18,375 ± $13,863 versus $26,491 ± $23,837; p = .03). Guarnieri and colleagues evaluated 300 patients randomized in a double-blind fashion to IV amiodarone (1 g/d for 2 days) versus placebo immediately after open-heart surgery.54 The primary end points of the trial were incidence of atrial fibrillation and length of hospital stay. Atrial fibrillation occurred in 67 of 142 (47%) patients on placebo versus 56 of 158 (35%) on amiodarone (p = .01). Length of hospital stay for the placebo group was 8.2 ± 6.2 days, and 7.6 ± 5.9 days for the amiodarone group. Low-dose IV amiodarone was safe and effective in reducing the incidence of atrial fibrillation after heart surgery but did not significantly alter length of hospital stay.

In summary, AF is a frequent complication of cardiac surgery. Many cases can be prevented with appropriate prophylactic therapy. Beta-adrenergic blockers should be administered to most patients without contraindication. Prophylactic amiodarone should be considered in patients at high risk for postoperative AF. The lack of data on cost benefits and cost efficiency in some studies may reflect the lack of higher-risk patients in the study. Patients who are poor candidates for beta blockade may not tolerate sotalol, whereas amiodarone does not have this limitation. Additional studies also need to be performed to better assess the role of prophylactic therapy in off-pump cardiac surgery.

Some depression of myocardial function is common after cardiac surgery.63–65 The etiology is multifactorial—preexisting disease, incomplete repair or revascularization, myocardial edema, postischemic dysfunction, reperfusion injury, etc—and usually is reversible. Adequate cardiac output usually can be maintained by exploiting the Starling curve with higher preload, but often the cardiac function curve is flattened, and it is necessary to use inotropic agents to maintain adequate organ perfusion.

The molecular basis for the contractile property of the heart is the interaction of the proteins actin and myosin, in which chemical energy (in the form of ATP) is converted into mechanical energy. In the relaxed state (diastole), the interaction of actin and myosin is inhibited by tropomyosin, a protein associated with the actin-myosin complex. With the onset of systole, Ca2+ enters the myocyte (during phase 1 of the action potential). This influx of Ca2+ triggers the release of much larger amounts of Ca2+ from the sarcoplasmic reticulum. The binding of Ca2+ to the C subunit of the protein troponin interrupts inhibition of the actin-myosin interaction by tropomyosin, facilitating the hydrolysis of ATP with the generation of a mechanical force. With repolarization of the myocyte and completion of systole, Ca2+ is taken back up into the sarcoplasmic reticulum, allowing tropomyosin to again inhibit the interaction of actin and myosin with consequent relaxation of contractile force. Thus inotropic action is mediated by intracellular Ca2+.66 A novel drug, levosimendan, currently under clinical development in the United States but approved in several countries, increases the sensitivity of the contractile apparatus to Ca2+,67 whereas the positive inotropic agents available for clinical use achieve their end by increasing intracellular Ca2+ levels.

The first drug to be considered is simply Ca2+ itself. In general, administration of calcium will increase the inotropic state of the myocardium when measured by load-independent methods, but it also will increase vascular tone (afterload) and impair diastolic function. In addition, the effects of calcium on myocardial performance depend on the plasma Ca2+ concentration. Ca2+ plays important roles in cellular function, and the intracellular Ca2+ concentration is highly regulated by membrane ion channels and intracellular organelles.68,69 If the extracellular Ca2+ concentration is normal, administration of Ca2+ will have little effect on the intracellular level and less pronounced hemodynamic effects. On the other hand, if the ionized plasma calcium concentration is low, exogenous calcium administration may increase cardiac output and blood pressure.70 It also should be realized that even with normal plasma Ca2+ concentrations, administration of Ca2+ may increase vascular tone, leading to increased blood pressure but no change in cardiac output. This increased afterload, as well as the deleterious effects on diastolic function, may be the basis of the observation that Ca2+ administration can blunt the response to epinephrine.71 Routine use of Ca2+ at the end of bypass should be tempered by the realization that Ca2+ may have little effect on cardiac output while increasing systemic vascular resistance, although this in itself may be of importance. If there is evidence of myocardial ischemia, Ca2+ administration may be deleterious because it may exacerbate both coronary spasm and the pathways leading to cellular injury.72,73

Digoxin, although not effective as acute therapy for low-cardiac-output syndrome in the perioperative period, nevertheless well illustrates the role of intracellular Ca2+. Digoxin functions by inhibiting Na+, K+-ATPase, which is responsible for the exchange of intracellular Na+ with extracellular K+.3,4 It is thus responsible for maintaining the intracellular/extracellular K+ and Na+ gradients. When it is inhibited, intracellular Na+ levels increase. The increased intracellular Na+ is an increased chemical potential for driving the Ca2+/Na+ exchanger, an ion exchange mechanism in which intracellular Na+ is removed from the cell in exchange for Ca2+. The net effect is an increase in intracellular Ca2+ with an enhancement of the inotropic state.

The most commonly used positive inotropic agents are the beta-adrenergic agonists. The beta1 receptor is part of a complex consisting of the receptor on the outer surface of the cell membrane and membrane-spanning G-proteins (so named because they bind GTP), which in turn stimulate adenylate cyclase on the inner surface of the membrane, catalyzing the formation cyclic adenosine monophosphate (cyclic AMP). The inotropic state is modulated by cyclic AMP via its catalysis of phosphorylation reactions by protein kinase A. These phosphorylation reactions "open" Ca2+ channels on the cell membrane and lead to greater release and uptake of Ca2+ from the sarcoplasmic reticulum.3,4

There are many drugs that stimulate beta1 receptors and have a positive inotropic effect, including epinephrine, norepinephrine, dopamine, isoproterenol, and dobutamine, the most commonly used catecholamines in the perioperative period. Although there are differences in their binding at the beta1 receptor, the most important differences between the various catecholamines are their relative effects on alpha- and beta2-adrenergic receptors. In general, alpha stimulation of receptors on the peripheral vasculature causes vasoconstriction, whereas beta2 stimulation leads to vasodilation (see the discussion elsewhere in this chapter). For some time it was believed that beta2 and alpha receptors were found only in the peripheral vasculature, as well as a few other organs, but not in the myocardium. However, alpha receptors are found in the myocardium and mediate a positive inotropic effect.5,6 The mechanism for this positive inotropic effect is probably the stimulation of phospholipase C, leading to hydrolysis of phosphatidyl inositol to diacylglycerol and inositol triphosphate, compounds that increase Ca2+ release from the sarcoplasmic reticulum and increase myofilament sensitivity to Ca2+. It is also possible that alpha-adrenergic agents increase intracellular Ca2+ by prolonging action potential duration by inhibition of outward K+ currents during repolarization or by activating the Na+/H+ exchange mechanism, increasing intracellular pH and increasing myofilament sensitivity to Ca2+. Just as the exact mechanism is uncertain, the exact role of alpha-adrenergic stimulation in control of the inotropic state is unclear, although it is apparent that onset of the effect is slower than that of beta1 stimulation.

Besides the discovery of alpha receptors in the myocardium, beta2 receptors are present in the myocardium.74 The fraction of beta2 receptors (compared with beta1 receptors) is increased in chronic heart failure, possibly explaining the efficacy of drugs with beta2 activity in this setting. This phenomenon is part of the general observation of beta1-receptor downregulation (decrease in receptor density) and desensitization (uncoupling of effect from receptor binding) that is observed in chronic heart failure.75 Interestingly, it has been demonstrated in a dog model that this same phenomenon occurs with cardiopulmonary bypass (CPB).76 In this situation, a newer class of drugs, the phosphodiesterase inhibitors, may be of benefit. These drugs, typified by the agents available in the United States, amrinone and milrinone, increase cyclic AMP levels independently of the beta receptor by selectively inhibiting the myocardial enzyme responsible for the breakdown of cyclic AMP.3,4

In clinical use, selection of a particular inotropic agent usually is based more on its side effects than its direct inotropic properties. Of the commonly used catecholamines, norepinephrine has alpha and beta1 but little beta2 activity and is both an inotrope and a vasopressor. Epinephrine and dopamine are mixed agonists with alpha, beta1, and beta2 activity. At lower doses, they are primarily inotropes and not vasopressors, although vasopressor effects become more pronounced at higher doses. This is especially true for dopamine, which achieves effects at higher doses by stimulating the release of norepinephrine.77 Dobutamine is a more selective beta1 agonist, in contrast to isoproterenol, which is a mixed beta agonist. Selection of a drug depends on the particular hemodynamic problem at hand. For example, a patient with depressed myocardial function in the presence of profound vasodilation may require a drug with both positive inotropic and vasopressor effects, whereas a patient who is vasoconstricted may benefit from some other choice. We recommend an empiric approach to selecting inotropic agents with careful monitoring of the response to the drug and selection of the agent that achieves the desired effect.

Clinical experience suggests that phosphodiesterase inhibitors can be effective when catecholamines do not produce an acceptable cardiac output.78–80 There are few differences in the hemodynamic effects of the two drugs available for use in the United States, amrinone and milrinone. Both agents increase contractility with little effect on heart rate, and both are vasodilators. There is significant venodilation, as well as arteriodilation, and maintaining adequate preload is important in avoiding significant hypotension.81,82 If amrinone is used, the bolus dose recommended in the product insert, 0.75 mg/mL, is inadequate to maintain therapeutic plasma levels, and a loading dose of 1.5 to 2.0 mg/mL should be used.83 With either drug, administering the loading dose over 15 to 30 minutes may attenuate possible hypotension. Plasma levels drop rapidly after a loading dose because of redistribution, and the loading dose should be followed immediately by a continuous infusion.83,84 Because of their longer half-lives, it is rather more difficult to readily titrate plasma levels than with catecholamines (which have plasma half-lives of a few minutes).

Phosphodiesterase inhibitors, specifically milrinone, facilitate separation from CPB with biventricular dysfunction and are used for treating low-cardiac-output syndrome after cardiac surgery.82,85–87 Doolan and colleagues also demonstrated that milrinone, in comparison with placebo, significantly facilitated separation of high-risk patients from CPB.88

Levosimendan

Levosimendan is a class of drugs known as calcium sensitizers. The molecule is a pyridazinone-dinitrile derivative with additional action on adenosine triphosphate (ATP)–sensitive potassium channels.67,89,90 Levosimendan is used intravenously for treating decompensated heart failure (HF) because it increases contractility and produces antistunning effects without increasing myocardial intracellular calcium concentrations or prolonging myocardial relaxation. Levosimendan also causes coronary and systemic vasodilation. In patients with HF, IV levosimendan reduced worsening HF or death significantly. IV levosimendan significantly also increased cardiac output and decreased filling pressure in decompensated HF in large, double-blind, randomized trials and after cardiac surgery in smaller trials. Levosimendan is well tolerated without arrhythmogenicity. In addition to sensitizing troponin to intracellular calcium, levosimendan inhibits phosphodiesterase III and open ATP-sensitive potassium channels (KATP), which may produce vasodilation. Unlike currently available intravenous inotropes, levosimendan does not increase myocardial oxygen use, and has been used effectively in beta-blocked patients. Levosimendan does not impair ventricular relaxation. Clinical studies have demonstrated short-term hemodynamic benefits of levosimendan over both placebo and dobutamine. Although large-scale, long-term morbidity and mortality data are scarce, the Levosimendan Infusion versus Dobutamine in Severe Low-Output Heart Failure (LIDO) study suggested a mortality benefit of levosimendan over dobutamine. Clinical studies comparing levosimendan with other positive inotropes, namely, milrinone, are lacking, and this agent is not available in North America.

Clinical Trials

Despite their common use after cardiac surgery, there have been relatively few comparative studies of inotropic agents in the perioperative period. In 1978, Steen and colleagues reported the hemodynamic effects of epinephrine, dobutamine, and epinephrine immediately after separation from CPB.91 The largest mean increase in cardiac index was achieved with dopamine at 15 g/kg per minute. However, it should be noted that the only epinephrine dose studied was 0.04 g/kg per minute. In a later comparison of dopamine and dobutamine, Salomon and colleagues concluded that dobutamine produced more consistent increases in cardiac index, although the hemodynamic differences were small, and all patients had good cardiac indices at the onset of the study.92 Fowler and colleagues also found insignificant differences in the hemodynamic effects of dobutamine and dopamine, although they reported that coronary flow increased more in proportion to myocardial oxygen consumption with dobutamine.93 Although neither of these groups reported significant increases in heart rate for either dopamine or dobutamine, clinical experience has been otherwise. This is supported by a study by Sethna and colleagues, who found that the increase in cardiac index with dobutamine occurs simply because of increased heart rate, although they found that myocardial oxygen was maintained.94 Butterworth and colleagues subsequently demonstrated that the older and much cheaper agent, epinephrine, effectively increased stroke volume without as great an increase in heart rate as dobutamine.95 More recently, Feneck and colleagues compared dobutamine and milrinone and found them to be equally effective in treating low-cardiac-output syndrome after cardiac surgery.96 This study was a comparison of two drugs, and the investigators emphasized that the most efficacious therapy is probably a combination of drugs. In particular, phosphodiesterase inhibitors require the synthesis of cyclic AMP to be effective, and thus use of a combination of a beta1-adrenergic agonist and a phosphodiesterase inhibitor would be predicted to be more effective than either agent alone.

Finally, while global hemodynamic goals (ie, heart rate, blood pressure, filling pressures, and cardiac output) may be achieved with inotropic agents, this does not guarantee adequate regional perfusion, in particular renal and mesenteric perfusion. So far there have been few investigations of regional perfusion after cardiac surgery. There has been more interest in regional (especially mesenteric) perfusion in the critical care medicine literature, and some of the studies may be relevant to postoperative care of the cardiac surgical patient. Two studies have indicated that epinephrine may impair splanchnic perfusion, especially in comparison with combining norepinephrine and dobutamine.97,98 Norepinephrine alone has variable effects on splanchnic blood flow in septic shock,99 although adding dobutamine can improve splanchnic perfusion significantly when blood pressure is supported with norepinephrine.98 Low-dose dopamine improves splanchnic blood flow,100 but there is evidence that dopamine in higher doses impairs gastric perfusion.101 The relevance of these studies of septic patients for the cardiac surgical patient is unclear, although there are similarities between the inflammatory responses to CPB and to sepsis.

CPB often is characterized by derangements of vascular tone. Sometimes CPB induces elevations in endogenous catecholamines, as well as other mediators, such as serotonin and arginine vasopressin (AVP), leading to vasoconstriction. However, often CPB is characterized by endothelial injury and a systemic inflammatory response, with a cascade of cytokine and inflammatory mediator release and profound vasodilation. The pathophysiology has a striking resemblance to that of sepsis or an anaphylactic reaction. Further, vasodilation after cardiac surgery may be exacerbated by the preoperative use of angiotensin-converting enzyme (ACE) inhibitors and post-CPB use of milrinone.

The mechanisms of vasodilatory shock have been reviewed recently.102 Vascular tone is modulated by intracellular Ca2+, which binds calmodulin. The Ca2+-calmodulin complex activates myosin light-chain kinase, which catalyzes the phosphorylation of myosin to facilitate the interaction with actin. Conversely, intracellular cyclic GMP activates myosin phosphatase (also via a kinase-mediated phosphorylation of myosin phosphatase), which dephosphorylates myosin and inhibits the interaction of actin and myosin. A primary mediator of vasodilatory shock is nitric oxide (NO), which is induced by cytokine cascades. NO activates guanylate cyclase, with resulting loss of vascular tone. Another mechanism of vasodilation that may be particularly relevant to prolonged CPB is activation of ATP-sensitive potassium (KATP) channels. These channels are activated by decreases in cellular ATP or increases in hydrogen ion or lactate. All these could result from the abnormal perfusion associated with CPB and/or hypothermia. Increases in potassium channel conductance result in hyperpolarization of the vascular smooth muscle membrane, which decreases Ca2+ flux into the cell, leading to decreased vascular tone. A third mechanism of vasodilatory shock that also may be particularly relevant to cardiac surgery is deficiency of vasopressin. As noted earlier, CPB often induces the release of vasopressin, and this may contribute to the excessive vasoconstriction sometimes seen after CPB. However, it has been observed in several experimental models of shock that the initially high levels of vasopressin decrease as shock persists, leading some investigators to suggest that vasopressin stores are limited and are depleted by the initial response to hypotension.

Excessive vasodilation during shock usually is treated with catecholamines, most typically phenylephrine, dopamine, epinephrine, or norepinephrine.103 Although catecholamines produce both alpha- and beta-adrenergic effects, alpha1-adrenergic receptor stimulation produces vasoconstriction. As noted, stimulation of these receptors activates membrane phospholipase C, which in turn hydrolyzes phosphatidylinositol 4,55 diphosphate.7 This leads to the subsequent generation of two second messengers, including diacyl glycerol and inositol triphosphate. Both these second messengers increase cytosolic Ca2+ by different mechanisms, which include facilitating release of calcium from the sarcoplasmic reticulum and potentially increasing the calcium sensitivity of the contractile proteins in vascular smooth muscle.

Mediator-induced vasodilation often is poorly responsive to catecholamines,103 and the most potent pressor among catecholamines, norepinephrine, is required frequently. Some clinicians are concerned about renal, hepatic, and mesenteric function during norepinephrine administration. However, in septic patients, norepinephrine can improve renal function,102–107 and there is evidence that it may improve mesenteric perfusion as well.108 Given the hemodynamic similarities between septic patients and some patients at the end of CPB, these results often are extrapolated to the cardiac surgical patient but have not been confirmed by a systematic study. In some cases of profound vasodilatory shock, even norepinephrine is inadequate to restore systemic blood pressure. In this situation, low doses of vasopressin may be useful. Argenziano and colleagues109 studied 40 patients with vasodilatory shock (defined as a mean arterial blood pressure of less than 70 mm Hg with a cardiac index greater than 2.5 L/m2 per minute) after cardiac surgery. Arginine vasopressin levels were inappropriately low in this group of patients, and low-dose vasopressin infusion (≤0.1 U/min) effectively restored blood pressure and reduced norepinephrine requirements without significantly changing cardiac index. These observations were similar to an earlier report of the use of vasopressin in vasodilatory septic shock.110 Vasopressin also has been reported to be useful in treating milrinone-induced hypotension.111 In this latter report, vasopressin was reported to increase urine output, presumably via glomerular efferent arteriole constriction. However, the overall effects on renal function are unclear. In addition, there are still important unanswered questions about vasopressin and mesenteric perfusion. Although vasopressin effectively may restore blood pressure in vasodilatory shock, it must be remembered that in physiologic concentrations it is a mesenteric vasoconstrictor, and mesenteric hypoperfusion may be a factor in developing sepsis and multiorgan dysfunction syndrome.

Different pharmacologic approaches are available to produce vasodilation (Table 4-1). Potential therapeutic approaches include: (1) blockade of alpha1-adrenergic receptors, ganglionic transmission, and calcium channel receptors; (2) stimulation of central alpha2-adrenergic receptors or vascular guanylate cyclase and adenylate cyclase; and (3) inhibition of phosphodiesterase enzymes and angiotensin-converting enzymes.112 Adenosine in low concentrations is also a potent vasodilator with a short half-life, but it is used, as noted earlier, for its ability to inhibit atrioventricular conduction. Losartan, a novel angiotensin II (AII) antagonist, has just been released for treating hypertension but is not available for intravenous use.

Stimulation of Adenylate Cyclase (Cyclic AMP)

Prostacyclin, prostaglandin E1, and isoproterenol increase cyclic nucleotide formation (eg, adenosine-3′,5′ -monophosphate and cyclic AMP) in vascular smooth muscle to produce calcium mobilization out of vascular smooth muscle. Inhibiting the breakdown of cyclic AMP by phosphodiesterase also will increase cyclic AMP.112 Increasing cyclic AMP in vascular smooth muscle facilitates calcium uptake by intracellular storage sites, thus decreasing calcium available for contraction. The net effect of increasing calcium uptake is to produce vascular smooth muscle relaxation and hence vasodilation. However, most catecholamines with beta2-adrenergic activity (eg, isoproterenol) and phosphodiesterase inhibitors have positive inotropic and other side effects that include tachycardia, glycogenolysis, and kaluresis.113 Prostaglandins (ie, prostacyclin and prostaglandin E1) are potent inhibitors of platelet aggregation and activation. Catecholamines with beta2-adrenergic activity, phosphodiesterase inhibitors, and prostaglandin E1 and prostacyclin have been used to vasodilate the pulmonary circulation in patients with pulmonary hypertension and right ventricular failure (see Table 4-1).113

Nitrates, Nitrovasodilators, and Stimulation of Guanylyl Cyclase (Cyclic GMP)

The vascular endothelium modulates vascular relaxation by releasing both nitric oxide and prostacyclin.114–116 Inflammatory mediators also can stimulate the vascular endothelium to release excessive amounts of endothelium-derived relaxing factor (EDRF, or nitric oxide), which activates guanylyl cyclase to generate cyclic GMP.89,90 Nitrates and sodium nitroprusside, however, generate nitric oxide directly, independent of vascular endothelium.115,116 The active form of any nitrovasodilator is nitric oxide (NO), in which the nitrogen is in a +2 oxidation state. For any nitrovasodilator to be active, it first must be converted to NO. For nitroprusside, this is easily accomplished because nitrogen is in a +3 oxidation state, with the nitric oxide molecule bound to the charged iron molecule in an unstable manner, allowing nitroprusside to readily donate its nitric oxide moiety. For nitroglycerin, nitrogen molecules exist in a +5 oxidation state, and thus they must undergo significant metabolic transformations before they are converted to an active molecule. Nitroglycerin is a selective coronary vasodilator and does not produce coronary steal compared with nitroprusside because the small intracoronary resistance vessels, those less than 100 μm thick, lack whatever metabolic transformation pathway is required to convert nitroglycerin into its active form of nitric oxide.115,116 Chronic nitrate therapy can produce tolerance through different mechanisms.114–118 Sodium nitroprusside and nitroglycerin are effective vasodilators that produce venodilation that contributes significantly to the labile hemodynamic state.114 Intravenous volume administration often is required with nitroprusside owing to the relative intravascular hypovolemia.

Dihydropyridine Calcium Channel Blockers

Dihydropyridine calcium channel blockers are direct arterial vasodilators.119 Nifedipine was the first dihydropyridine calcium channel blocker, and intravenous forms studied in cardiac surgery include clevidipine, isradipine, and nicardipine. These agents are selective arterial vasodilators that have no effects on the vascular capacitance bed, atrioventricular nodal conduction, or negative inotropic effects.120–125 Clevidipine and nicardipine are available in the United States, and offer novel and important therapeutic options to treat perioperative hypertension following cardiac surgery. Also, intravenous dihydropyridines can be used to treat acute hypertension that occurs during the perioperative period (ie, intubation, extubation, cardiopulmonary bypass–induced hypertension, and aortic cross-clamping) and postoperative hypertension.

Phosphodiesterase Inhibitors

The phosphodiesterase inhibitors currently available for use produce both positive inotropic effects and vasodilation.126 When administered to patients with ventricular dysfunction, they increase cardiac output while decreasing pulmonary artery occlusion pressure, systemic vascular resistance, and pulmonary vascular resistance. Because of their unique mechanisms of vasodilation, they are especially useful for patients with acute pulmonary vasoconstriction and right ventricular dysfunction. Multiple forms of the drug are currently under investigation. The bipyridines (eg, amrinone and milrinone), imidazolones (eg, enoximone), and methylxanthines (eg, aminophylline) are the ones most widely available. Papaverine, a benzyl isoquinolinium derivative isolated from opium, is a nonspecific phosphodiesterase inhibitor and vasodilator used by cardiac surgeons for its ability to dilate the internal mammary artery.126

Angiotensin-Converting Enzyme Inhibitors

ACE inhibitors have growing use in managing heart failure, and more patients are receiving these drugs. The ACE inhibitors prevent the conversion of angiotensin I to angiotensin II by inhibiting an enzyme called kininase in the pulmonary and systemic vascular endothelium. This enzyme is also important for the metabolism of bradykinin, a potent endogenous vasodilator, and for release of EDRF. Although there are little data in the literature regarding the preoperative management of patients receiving these drugs, withholding them on the day of surgery has been our clinical practice based on their potential to produce excessive vasodilation during CPB. Although Tuman was unable to find any difference in blood pressure during CPB in patients receiving ACE inhibitors, contact activation during CPB has the ability to generate bradykinin and thus amplify the potential for vasodilation. The vasoconstrictor requirements were increased after bypass in his study.

Angiotensin II–Receptor Blockers

ACE inhibitors may not be tolerated in some patients owing to cough (common) and angioedema (rare). Inhibition of kininase II by ACE inhibitors leads to bradykinin accumulation in the lungs and vasculature, causing cough and vasodilation. Alternative treatment with angiotensin II–receptor blockers (ARBs) may be associated less frequently with these side effects because ARBs do not affect kinin metabolism. Six ARBs are currently available for antihypertensive therapy in the United States: losartan (Cozaar), valsartan (Diovan), irbesartan (Avapro), candesartan (Atacand), eprosartan (Teveten), and telmisartan (Micardis). Mortality in chronic heart failure is related to activation of the autonomic nervous and renin-angiotensin systems, and ACE inhibitor therapy seems to attenuate progression of myocardial dysfunction and remodeling. ACE inhibitors do not completely block angiotensin II (A-II) production,127 and may even increase circulating A-II levels in patients with heart failure. It was thought initially that ARBs might offer advantages over ACE inhibitors for heart failure therapy in terms of tolerability and more complete A-II blockade. Although ARBs were better tolerated, all-cause mortality and the number of sudden deaths or resuscitated cardiac arrests were not different when losartan (Cozaar) and captopril (Capoten) were compared in patients (>60 years old, NYHA classes II to IV, LVEF < 40%).128 Perioperative hypotension may be encountered in ARB-treated patients as well as ACE inhibitor–treated patients, and increased inotropic support may be required.

Numerous pharmacologic approaches to manipulate the hemostatic system in cardiac surgery include attenuating hemostatic system activation, preserving platelet function, and decreasing the need for transfusion of allogeneic blood products. Pharmacologic approaches to reduce bleeding and transfusion requirements in cardiac surgical patients are based on either preventing or reversing the defects associated with the CPB-induced coagulopathy.

Not surprisingly, most of the effects observed after administration of a beta-adrenergic receptor blocker reflect the reduced responsiveness of tissues containing beta-adrenergic receptors to catecholamines present in the vicinity of those receptors. Hence the intensity of the effects of beta-blockers depends on both the dose of the blocker and the receptor concentrations of catecholamines, primarily epinephrine and norepinephrine. In fact, a purely competitive interaction of beta-blockers and catecholamines can be demonstrated in normal human volunteers as well as in isolated tissues studied in the laboratory. The presence of disease and other types of drugs modifies the responses to beta-blockers observed in patients, but the underlying competitive interaction is still operative. The key to successful use of beta-adrenergic receptor blockers is to titrate the dose to the desired degree of effect and to remember that excessive effects from larger than necessary doses of beta-adrenergic receptor blockers can be overcome by: (1) administering a catecholamine to compete at the blocked receptors; and/or (2) administering other types of drugs to reduce the activity of counterbalancing autonomic mechanisms that are unopposed in the presence of beta-receptor blockade. An example of the latter is propranolol-induced bradycardia, which reflects the increased dominance of the vagal cholinergic mechanism on cardiac nodal tissue. Excessive bradycardia may be relieved by administering atropine to block the cholinergic receptors, which are also located in the sinoatrial (SA) and atrioventricular (AV) nodes (see Table 4-2).

Knowledge of the type, location, and action of beta receptor is fundamental to understanding and predicting effects of beta-adrenergic receptor–blocking drugs129 (see Table 4-2). Beta-adrenergic receptor blockers are competitive inhibitors; hence, the intensity of blockade depends on both the dose of the drug and the receptor concentrations of catecholamines, primarily epinephrine and norepinephrine.

Beta-adrenergic receptor antagonists (blockers) include many drugs (Table 4-3) that typically are classified by their relative selectivity for beta1 and beta2 receptors (ie, cardioselective or nonselective), the presence or absence of agonistic activity, membrane-stabilizing properties, alpha-receptor–blocking efficacy, and various pharmacokinetic features (eg, lipid solubility, oral bioavailability, and elimination half-time).129 The practitioner must realize that the selectivity of individual drugs for beta1 and beta2 receptors is relative, not absolute. For example, the risk of inducing bronchospasm with a beta1-adrenergic (cardioselective) blocker (eg, esmolol or metoprolol) may be relatively less than with a nonselective blockers (eg, propranolol); however, the risk is still present.

Acute Myocardial Infarction

Earlier clinical trials of intravenous beta-adrenergic blockers in the early phases of acute myocardial infarction suggest they decrease mortality. Following myocardial infarction, chronic oral beta-blocking agents reduce the incidence of recurrent myocardial infarction (see Table 4-3).130

Supraventricular Tachycardias and Ventricular Dysrhythmias

Beta-adrenergic blocking agents are Vaughan Williams class II antidysrhythmics that primarily block cardiac responses to catecholamines. Propranolol (Inderal), esmolol (Brevibloc), and acebutolol (Sectral) are used commonly for this indication. Beta-blocking agents decrease spontaneous depolarization in the SA and AV nodes, decrease automaticity in Purkinje fibers, increase AV nodal refractoriness, increase threshold for fibrillation (but not for depolarization), and decrease ventricular slow responses that depend on catecholamines. Amiodarone, a class III agent, also exerts noncompetitive alpha- and beta-adrenergic blockade, which may contribute its antidysrhythmic and antihypertensive actions. Sotalol is another class III antidysrhythmic with nonselective beta-blocking action. There is evidence that beta-blocking agents also decrease intramyocardial conduction in ischemic tissue and reduce the risks of dysrhythmias to the extent that they decrease myocardial ischemia. Beta-adrenergic blockers are not particularly effective in controlling dysrhythmias that are not induced or maintained by catecholamines.131

Hypertension

Hypertension is a major risk factor for developing heart failure and other end-organ damage. Beta-blockers, along with diuretics, are considered to be the initial drug of choice for uncomplicated hypertension in patients aged less than 65 years.

During the early phases of therapy, there is a decrease in cardiac output, a rise in systemic vascular resistance (SVR), and relatively little change in mean arterial blood pressure. Within hours to days, SVR normalizes, and blood pressure declines. In addition, the release of renin from the juxtaglomerular apparatus in the kidney is inhibited (beta1 blockade). Beta-blocking agents with intrinsic agonistic activity reduce systemic vascular resistance below pretreatment levels presumably by activating beta2 receptors in vascular smooth muscle. Most beta-adrenergic blockers are used with other agents in treating chronic hypertension. When combined with a vasodilator, beta-blockers limit reflex tachycardia. For example, when propranolol is combined with intravenous nitroprusside (a potent arterial dilator), it prevents reflex release of renin and reflex tachycardia induced by nitroprusside.132

Acute Dissecting Aortic Aneurysm

The primary goal in managing dissecting aneurysms is to reduce stress on the dissected aortic wall by reducing the systolic acceleration of blood flow. Beta-blockers reduce cardiac inotropy and ventricular ejection fraction. Beta-blockers also may limit reflex sympathetic responses to vasodilators that are used to control systemic arterial pressure.

Pheochromocytoma

The presence of catecholamine-secreting tissue is tantamount to the continuous or intermittent infusion of a varying mixture of norepinephrine and epinephrine. It is absolutely essential that virtually complete alpha-adrenergic receptor blockade be established prior to administering the beta-blocker to prevent exacerbation of hypertensive episodes by unopposed alpha-adrenergic receptor activity in vascular smooth muscle.

Chronic Heart Failure

It is now understood that activation of the autonomic nervous system (ANS) and renin-angiotensin system (RAS) as compensatory mechanisms for the failing heart actually may contribute to deterioration of myocardial function. Mortality in chronic heart failure seems related to activation of ANS and RAS. Progression of myocardial dysfunction and remodeling may be attenuated by the use of beta-blocking agents and ACE inhibitors. Carvedilol (Coreg) is a beta-blocker approved by the FDA to treat patients with heart failure. It has an alpha1- and non-selective beta-blocking activity (alpha:beta = 1:10). It is contra-indicated in severe decompensated heart failure and asthma. In patients with atrial fibrillation and left-sided heart failure treated with carvedilol, improved ejection fraction, and a trend toward a decreased incidence of death and chronic heart failure hospitalization were observed in a retrospective analysis of a U.S. carvedilol study. There are several ongoing clinical trials with carvedilol, metoprolol (Toprol), and bisoprolol (Zebeta). The results of these studies may provide answers as to which beta-blocking agent would be most successful in the treatment of specific patient populations.

Other Indications

The other clinical applications of beta-adrenergic receptor blockers listed in Table 4-4 are based on largely symptomatic treatment or empirical trials of beta-adrenergic antagonists.

Side Effects and Toxicity

The most obvious and immediate signs of a toxic overdose of a beta-adrenergic receptor blocker are hypotension, bradycardia, congestive heart failure, decreased AV conduction, and a widened QRS complex on the electrocardiogram. Treatment is aimed at blocking the cholinergic receptor responses to vagal nerve activity (eg, atropine) and administering a sympathomimetic to compete with the beta-blockers at adrenergic receptors. In patients with asthma and chronic obstructive pulmonary disease (COPD), beta-blockers may cause bronchospasm. Beta blockers may increase levels of plasma triglycerides and reduce levels of high-density lipoprotein (HDL) cholesterol. Rarely, beta-blockers may mask the symptoms of hypoglycemia in diabetic patients. Other side effects include mental depression, physical fatigue, altered sleep patterns, sexual dysfunction, and gastrointestinal symptoms, including indigestion, constipation, and diarrhea (see Table 4-4).

Drug Interactions

Pharmacokinetic drug interactions include reduced gastrointestinal absorption of the beta-blocker (eg, aluminum-containing antacids and cholestyramine), increased biotrans-formation of the beta-blocker (eg, phenytoin, phenobarbital, rifampin, and smoking), and increased bioavailability owing to decreased biotransformation (eg, cimetidine and hydralazine). Pharmacodynamic interactions include an additive effect with calcium channel blockers to decrease conduction in the heart and a reduced antihypertensive effect of beta-blockers when administered with some of the nonsteroidal anti-inflammatory drugs (NSAIDs).

Diuretics are drugs that act directly on the kidneys to increase urine volume and produce a net loss of solute (principally sodium and other electrolytes) and water. Diuretics and beta-blockers are initial drugs of choice for uncomplicated hypertension in patients younger than 65 years.132 The currently available diuretic drugs have a number of other uses in medicine (eg, glaucoma and increased intracranial pressure). The principal indications for the use of diuretics by intravenous administration in the perioperative period are to: (1) increase urine flow in oliguria; (2) reduce intravascular volume in patients at risk for acute heart failure from excessive fluid administration or acute heart failure; and (3) mobilize edema.

Renal function depends on adequate renal perfusion to maintain the integrity of renal cells and provide the hydrostatic pressure that produces glomerular filtration. There are no drugs that act directly on the renal glomerulus to affect glomerular filtration rate (GFR). In the normal adult human of average size, GFR averages 125 mL/min, and urine production approximates 1 mL/min. In other words, 99% of the glomerular filtrate is reabsorbed. Diuretics act primarily on specific segments of the renal tubule to alter reabsorption of electrolytes, principally sodium, and water.

There are two basic mechanisms behind the renal tubular reabsorption of sodium. First, sodium is extruded from the tubular cell into peritubular fluid primarily by active transport of the sodium ion, which reflects the action of the Na+, K+-ATPase pump as well as the bicarbonate reabsorption mechanism (see the following). This extrusion of sodium creates an electrochemical gradient that causes diffusion of sodium from the tubular lumen into the tubular cell. Second, sodium moves from the glomerular filtrate in the tubular fluid into the peritubular fluid by several different mechanisms. The most important quantitatively is the sodium electrochemical gradient created by the active extrusion of sodium from the tubular cell into the peritubular fluid. In addition, sodium is coupled with organic solutes and phosphate ions, exchanged for hydrogen ions diffusing from the tubular cell into the tubular lumen, and coupled to the transfer of a chloride ion or a combination of potassium and two chloride ions (Na+-K+-2Cl− cotransport) from the tubular fluid into the tubular cell. Diuretics are classified by their principal site of action in the nephron and by the primary mechanism of their naturetic effect (Table 4-5).

Osmotic Diuretics

Mannitol is the principal example of this type of diuretic, which is used for two primary indications: (1) prophylaxis and early treatment of acute renal failure that is characterized by a decrease in GFR leading to a decreased urine volume and an increase in the concentration of toxic substances in the renal tubular fluid; and (2) enhancing the actions of other diuretics by retaining water and solutes in the tubular lumen, thereby providing the substrate for the action of other types of diuretics. Normally, 80% of the glomerular filtrate is reabsorbed isosmotically in the proximal tubules. By its osmotic effect, mannitol limits the reabsorption of water and dilutes the proximal tubular fluid. This reduces the electrochemical gradient for sodium and limits its reabsorption so that more is delivered to the distal portions of the nephron. Mannitol produces a prostaglandin-mediated increase in renal blood flow that partially washes out the medullary hypertonicity, which is essential for the countercurrent mechanism promoting the reabsorption of water in the late distal tubules and collecting system under the influence of antidiuretic hormone (ADH). Mannitol is used often (25 to 50 g) as part of the priming solution of cardiopulmonary bypass for the above-mentioned indications. The principal toxicity of mannitol is acute expansion of the extracellular fluid volume leading to HF in the patient with compromised cardiac function (see Table 4-5).

High-Ceiling (Loop) Diuretics

Furosemide (Lasix), bumetanide (Bumex), and ethacrynic acid (Edecrin) are three chemically dissimilar compounds that have the same primary diuretic mechanism of action. They act on the tubular epithelial cell in the thick ascending loop of Henle to inhibit the Na+-K+-2Cl− cotransport mechanism. Their peak diuretic effect is far greater than that of the other diuretics currently available. Administered intravenously, they have a rapid onset and relatively short duration of action, the latter reflecting both the pharmacokinetics of the drugs and the body's compensatory mechanisms to the consequences of diuresis. These three diuretics increase renal blood flow without increasing GFR and redistribute blood flow from the medulla to the cortex and within the renal cortex. These changes in renal blood flow are also short-lived, reflecting the reduced extracellular fluid volume resulting from diuresis. Minor actions, including carbonic anhydrase inhibition by furosemide and bumetanide and actions on the proximal tubule and on sites distal to the ascending limb, remain controversial. All three of the loop diuretics increase the release of renin and prostaglandin, and indomethacin blunts the release as well as the augmentation in renal blood flow and natruresis. All three of the loop diuretics produce an acute increase in venous capacitance for a brief period of time after the first intravenous dose is administered, and this effect is also blocked by indomethacin.

Potassium, magnesium, and calcium excretion is increased in proportion to the increase in sodium excretion. In addition, there is augmentation of titratable acid and ammonia excretion by the distal tubules leading to metabolic alkalosis, which is also produced by contraction of the extracellular volume. Hyperuricemia can occur but usually is of little physiologic significance. The nephrotoxicity of cephaloridine, and possibly other cephalosporins, is increased. A rare but serious side effect of the loop diuretics is deafness, which may reflect electrolyte changes in the endolymph.

Because of their high degree of efficacy, prompt onset, and relatively short duration of action, the high-ceiling or loop diuretics are favored for intravenous administration in the perioperative period to treat the three principal problems cited earlier. Dosage requirements vary considerably among patients. Some may only require furosemide 3 to 5 mg IV to produce a good diuresis. And for some patients, the less potent benzothiazides may be sufficient.

Benzothiazides

Hydrochlorothiazide (HCTZ) is the prototype of more than a dozen currently available diuretics in this class. Although the drugs differ in potency, they all act by the same mechanism of action and have the same maximum efficacy. All are actively secreted into the tubular lumen by tubular cells and act in the early distal tubules to decrease the electroneutral Na+-Cl− cotransport reabsorption of sodium. Their moderate efficacy probably reflects the fact that more than 90% of the filtrated sodium is reabsorbed before reaching the distal tubules. Their action is enhanced by their combined administration with an osmotic diuretic such as mannitol. The benziothiazides increase urine volume and the excretion of sodium, chloride, and potassium. The decreased reabsorption of potassium reflects the higher rate of urine flow through the distal tubule (diminished reabsorption time).

This class of diuretics produces the least disturbance of extracellular fluid composition, reflecting their moderate efficacy as diuretics and perhaps suggesting their usefulness when a moderate degree of diuretic effect is indicated. Their principal side effects include hyperuricemia, decreased calcium excretion, and enhanced magnesium loss. Hyperglycemia can occur and reflects multiple variables. With prolonged use and development of a contracted extracellular fluid volume, urine formation decreases (ie, tolerance develops to their diuretic actions). These agents also have a direct action on the renal vasculature to decrease GFR.

Carbonic Anhydrase Inhibitors

Acetazolamide (Diamox) is the only diuretic of this class available for intravenous administration. Its use is directed primarily toward alkalinization of urine in the presence of metabolic alkalosis, which is a common consequence of prolonged diuretic therapy. It acts in the proximal convoluted tubule to inhibit carbonic anhydrase in the brush border of the tubular epithelium, thereby reducing the destruction of bicarbonate ions (ie, conversion to CO2 that diffuses into the tubular cell). The carbonic anhydrase enzyme in the cytoplasm of the tubular cell is also inhibited, and as a consequence, conversion of CO2 to carbonic acid is reduced markedly, as is the availability of hydrogen ions for the Na-H exchange mechanism. Hence the reabsorption of both sodium and bicarbonate in the proximal tubules is diminished. However, more than half the bicarbonate is reabsorbed in more distal segments of the nephron, thereby limiting the overall efficacy of this class of diuretics.

Potassium-Sparing Diuretics

Spironolactone (Aldactone) is a competitive antagonist of aldosterone. Spironolactone binds to the cytoplasmic aldosterone receptor and prevents its conformational change to the active form, thereby aborting the synthesis of active transport proteins in the late distal tubules and collecting system in which the reabsorption of sodium and secretion of potassium are reduced.

Triamterene (Dyrenium) and amiloride (Midamor) are potassium-sparing diuretics with a mechanism of action independent of the mineralocorticoids. They have a moderate natriuretic effect leading to an increased excretion of sodium and chloride with little change or a slight increase in potassium excretion when the latter is low. When potassium secretion is high, they produce a sharp reduction in the electrogenic entry of sodium ions into the distal tubular cells and thereby reduce the electrical potential that is the driving force for potassium secretion.

Both types of potassium-sparing diuretics are used primarily in combination with other diuretics to reduce potassium loss. Their principal side effect is hyperkalemia. It is appropriate to limit the intake of potassium when using this type of diuretic. It is also appropriate to use this type of diuretic cautiously in patients taking ACE inhibitors, which decrease aldosterone formation and consequently increase serum potassium concentrations.

Other Measures to Enhance Urine Output and Mobilization of Edema Fluid

The infusion of albumin (5 to 25% solutions) or other plasma volume expanders (eg, hetastarch) is often employed in an attempt to draw water and its accompanying electrolytes (ie, edema fluid) osmotically from the tissues into the circulating blood and thereby enhance their delivery to the kidneys for excretion. In the presence of a reduced circulating blood volume, this approach seems to be a logical method to increase the circulating blood volume and renal perfusion. The limiting feature of this approach to enhancing diuresis relates to the fact that the osmotic effect of albumin and plasma expanders is transient because they can diffuse (at a rate slower than water) from blood through capillary membranes into tissue. The albumin or plasma expander then tends to hold water and its accompanying electrolytes in tissue (ie, rebound edema). The same limiting feature applies to osmotic diuretics such as mannitol, which may transiently draw water and its accompanying electrolytes from tissues into the circulating blood for delivery to the kidneys, where the mannitol passes through the glomerulus and delays the reabsorption of water and its accompanying electrolytes from the proximal tubular fluid. Although this mechanism may enhance the actions of other diuretics, it is a transient effect that is limited by the diffusion of mannitol from blood into tissues with the production of rebound edema.

Dopamine, at doses from 1 to 3 μg/kg per minute, has been used conventionally to support mesenteric and renal perfusion. Its vascular action is mediated via vascular dopamine 1 (D1) receptors in coronary, mesenteric, and renal vascular beds. By activating adenyl cyclase and raising intracellular concentrations of cyclic AMP, D1-receptor agonists cause vasodilatation. There are also dopamine 2 (D2) receptors that antagonize D1-receptor stimulation. Fenoldopam (Corlopam), a parenteral D1-receptor–specific agonist, was approved by the FDA recently. The Joint National Commissions VII recommendations include this drug for hypertensive emergencies.132,133 Infusion of fenoldopam (0.1 to 0.3 μg/kg per minute) causes an increase in GFR, renal blood flow, and Na+ excretion.

Clinical trials of dopamine failed to show improvement in renal function, which probably is a result of the nonspecificity of dopamine. As a catecholamine and a precursor in the metabolic synthesis of norepinephrine and epinephrine, dopamine has inotropic and chronotropic effects on the heart. The inotropic effect is mediated by beta1-adrenergic receptors and usually requires infusion rates higher than those able to produce enhanced renal perfusion and diuresis. However, there are varied pharmacokinetic responses to dopamine infusion even in healthy subjects; therefore, the use of a "renal dose" dopamine regimen may not always result in the desirable effects. Stimulation of catecholamine receptors and D2 receptors antagonizes the effects of D1-receptor stimulation. Current data do not consistently demonstrate improved renal outcomes with use of the D1-receptor–specific agonist fenoldopam.

A large number of Americans take herbal remedies for their health. Most of these herbal therapies are not supported by clear scientific evidence and are not under rigorous control by the FDA.134–136 Patients who take alternative remedies may not necessarily disclose this information to their physicians.135 There are increasing concerns regarding serious drug interactions between herbal therapy and prescribed medication. Some of the most common herbal remedies and drug interactions are summarized in Table 4-6.

Airway Management

Airway management in cardiovascular surgical patients is important because patients often present with coexisting conditions that may complicate endotracheal intubation. For example, a patient with morbid obesity and sleep apnea may require awake intubation with a fiberoptic bronchoscope, or a history of smoking and COPD may make the patient susceptible to rapid desaturation and/or bronchospasm. Airway management in the perioperative period is a primary responsibility of the anesthesiologist, but the surgeon becomes involved in the absence of the anesthesiologist or in assisting the anesthesiologist in difficult situations. Airway management involves instrumentation and mechanics (not discussed here) and employs pharmacologic approaches to overcome pathophysiologic problems contributing to airway obstruction and to facilitate manipulation and instrumentation of the airway. Pharmacologic agents are considered at the end of this section.

Five major challenges may be encountered in airway management. Each of these is described succinctly below to facilitate understanding of the roles that drugs play in meeting the challenges. The five challenges are (1) overcoming airway obstruction, (2) preventing pulmonary aspiration, (3) performing endotracheal intubation, (4) maintaining intermittent positive-pressure ventilation (IPPV), and (5) reestablishing spontaneous ventilation and airway protective reflexes.

Airway Obstruction

Obstruction to gas flow can occur from the entry of a foreign object (including food) into the airway and as a result of pathophysiologic processes involving airway structures (eg, trauma and edema). In the anesthetized or comatose patient, the loss of muscle tone can allow otherwise normal tissues (eg, tongue and epiglottis) to collapse into the airway and cause obstruction. The first measure in relieving such obstructions involves manipulation of the head and jaw, insertion of an artificial nasal or oral airway device, and evacuation of obstructing objects and substances (eg, blood, secretions, or food particles). Except for drugs used to facilitate endotracheal intubation (see the following), the only drug useful to improve gas flow through a narrowed airway is a mixture of helium and oxygen (Heliox), which has a much reduced viscosity resulting in reduced resistance to gas flow.

Aspiration

The upper airway (above the larynx/epiglottis) is a shared porthole to the lungs (gas exchange) and gastrointestinal tract (fluids and nutrition). Passive regurgitation or active vomiting resulting in accumulation of gastric contents in the pharynx places the patient at risk of pulmonary aspiration, especially under circumstances in which airway reflexes (eg, glottic closure and coughing) and voluntary avoidance maneuvers are suppressed (eg, anesthesia or coma). Particulate matter can obstruct the tracheobronchial tree, and acidic fluid (pH < 2.5) can injure the lung parenchyma. The resulting pneumonitis can cause significant morbidity (eg, acute respiratory distress syndrome) and has a high mortality rate. Preoperative restriction of fluids and food (NPO status) does not guarantee the absence of aspiration risks. Similarly, the advance placement of a nasogastric or orogastric tube may serve to reduce intragastric pressure but does not guarantee complete removal of gastric contents. Nevertheless, both NPO orders and the insertion of a nasogastric or orogastric tube under some circumstances are worthwhile measures to reduce the risks of pulmonary aspiration. In some circumstances, the deliberate induction of vomiting in a conscious patient may be indicated, but this is done rarely and almost never involves the use of an emetic drug. In fact, more often antiemetic drugs are employed to reduce the risks of vomiting during airway manipulation and induction of anesthesia.

Drug therapy to reduce the risks of pulmonary aspiration is focused on decreasing the quantity and acidity of gastric contents and on facilitating endotracheal intubation (see the following). Nonparticulate antacids (eg, sodium citrate [Bicitra]) are used to neutralize the acidity of gastric fluids. Drugs to reduce gastric acid production include H2-receptor blockers (eg, cimetidine [Tagamet], ranitidine [Zantac], famotidine [Pepcid]) and inhibitors of gastric parietal cell hydrogen-potassium ATPase (proton pump inhibitors, eg, omeprazole [Prilosec], lansoprazole [Prevacid], and esomeprazole [Nexium]). Metoclopramide [Reglan] enhances gastric emptying and increases gastroesophageal sphincter tone. Cisapride [Propulsid] also increases gastrointestinal motility via the release of acetylcholine at the myenteric plexus.

Antiemetic drugs are used more commonly in the postoperative period and include several different drug classes: anticholinergics (eg, scopolamine [Transderm Scop]), antihistamines (eg, hydroxyzine [Vistaril] and, promethazine [Phenergan]), and antidopaminergics (eg, droperidol [Inapsine] and prochlorperazine [Compazine]). Antidopaminergic agents may cause extrapyramidal side effects in elderly patients. More costly but effective alternatives include the use of antiserotinergics (eg, ondansetron [Zofran] and dolasetron [Anzemet]).

Endotracheal Intubation

Drugs are employed for three purposes in facilitating endotracheal intubation: (1) to improve visualization of the larynx during laryngoscopy; (2) to prevent closure of the larynx; and (3) to facilitate manipulation of the head and jaw.

For bronchoscopy, laryngoscopy, or fiberoptic endotracheal intubation, the reflex responses to airway manipulation can be suppressed by several different methods alone or in combination. Topical anesthesia (2 or 4% lidocaine spray) can be used to anesthetize the mucosal surfaces of the nose, oral cavity, pharynx, and epiglottis. Atomized local anesthetic can be inhaled to anesthetize the mucosa below the vocal cords. The subglottic mucosa also can be anesthetized topically by injecting local anesthetic into the tracheal lumen through the cricothyroid membrane. A bilateral superior laryngeal nerve block eliminates sensory input from mechanical contact or irritation of the larynx above the vocal cords. It must be remembered that anesthesia of the mucosal surfaces to obtund airway reflexes compromises the reflex protective mechanisms of the airway and increases the patient's vulnerability to aspiration of substances from the pharynx. Improvement of visualization of the larynx includes decreasing salivation and tracheal bronchial secretions by administration of an anticholinergic drug (eg, glycopyrrolate), reducing mucosal swelling by topical administration of a vasoconstrictor (eg, phenylephrine), and minimizing bleeding owing to mucosal erosion by instrumentation, which also is minimized by topical vasoconstrictors. The use of steroids in minimizing acute inflammatory responses in the airway may have some delayed benefit, but steroids usually are not indicated just before intubation.

Systemic drugs, usually administered intravenously, can be used to obtund the cough reflex. Intravenous lidocaine (1 to 2 mg/kg) transiently obtunds the cough reflex without affecting spontaneous ventilation to any significant degree. The risks of central nervous system (CNS) stimulation and seizure-like activity have to be kept in mind and can be reduced by the prior administration of an intravenous barbiturate or benzodiazepine in small doses. Intravenous opioids are effective in suppressing cough reflexes, but the doses required impair spontaneous ventilation to the point of apnea. A combination of an intravenous opioid and a major tranquilizer (eg, neuroleptic analgesia) allows the patient to tolerate an endotracheal tube with much smaller doses of the opioid and less embarrassment of spontaneous ventilation. Small doses of opioids are also useful in obtunding airway reflexes during general anesthesia provided either by intravenous (eg, thiopental) or inhaled anesthetics (eg, isoflurane). Not only do the opioids obtund the cough reflex that results in closure of the larynx, but they also are useful in limiting the autonomic sympathetic response to endotracheal intubation that typically leads to hypertension and tachycardia.

Skeletal muscle relaxants are used most commonly in conjunction with a general anesthetic to allow manipulation of the head and jaw and prevent reflex closure of the larynx. Of course, they also render the patient apneic, and two procedures are used commonly to maintain oxygenation of the patient's blood. First, the patient breathes 100% oxygen by mask while still awake to eliminate nitrogen from the lungs, and then a rapid-sequence administration of an intravenous anesthetic (eg, thiopental) is followed immediately by a rapid-acting neuromuscular blocker [eg, succinylcholine or rocuronium], and cricoid pressure is applied (Sellick maneuver). As soon as the muscle relaxation is apparent (30 to 90 seconds), laryngoscopy is performed, an endotracheal tube is inserted, the tracheal tube cuff is inflated, and the position of the tube in the trachea is verified. Second, when there is minimal risk of pulmonary aspiration (eg, presumed empty stomach); the patient is anesthetized and paralyzed while ventilation is supported by intermittent positive pressure delivered via a face mask. At the appropriate time, laryngoscopy is performed, and the endotracheal tube is inserted.

Normalizing Pulmonary Function during Positive-Pressure Ventilation

Once an endotracheal tube is in place, it is common practice in the operating room to maintain general anesthesia and partial muscular paralysis in order to facilitate positive-pressure ventilation and continued toleration of the endotracheal tube by the patient. Postoperatively, in the post-anesthesia care unit (PACU) and intensive care unit (ICU), general anesthesia and partial muscular paralysis may be continued if prolonged positive-pressure ventilation is anticipated, or sedatives may be administered by intravenous infusion to allow toleration of the endotracheal tube in anticipation of recovery of spontaneous ventilation and tracheal extubation.

Three other problems are encountered in the patient whose ventilation is supported mechanically by an endotracheal tube: (1) poor ventilatory compliance; (2) bronchoconstriction; and (3) impaired gas exchange. Poor ventilatory compliance can reflect limited compliance of the chest wall and diaphragm, limited compliance of the lungs per se, or both. Deepening general anesthesia and administration of a skeletal muscle relaxant can be used to reduce intercostal and diaphragmatic muscle tone, but they obviously cannot improve the chest cavity compliance that is fixed by disease (eg, scoliosis or emphysema).

Poor lung compliance may reflect pulmonary interstitial edema, consolidation, bronchial obstruction (eg, mucus plugs), bronchoconstriction, or compression of the lung by intrathoracic substances (eg, pneumothorax, hemothorax, or tumor mass). Treatment of these involves drug therapy of heart failure and infection and procedures such as bronchoscopy, thoracentesis, etc.

Bronchoconstriction may exist chronically (eg, asthma or reactive airways disease), and these conditions can be exacerbated by the collection of tracheobronchial secretions in the presence of an endotracheal tube, which reduces the effectiveness of coughing in clearing the airway. Occasionally bronchoconstriction can be induced by mechanical stimulation of the airway by an endotracheal tube or other object in an otherwise normal patient. Drug treatment is focused on reducing bronchial smooth muscle tone (eg, beta2 sympathomimetic or anticholinergic agents), minimizing tracheal bronchial secretions, and decreasing sensory input from the tracheal bronchial tree (eg, topical anesthetic, deeper general anesthesia, intravenous lidocaine, or an opioid). Acute treatment of bronchoconstriction may involve any combination of the following: (1) an aerosolized beta2 sympathomimetic and/or anticholinergic agent; and (2) systemic intravenous administration of a beta2 sympathomimetic agent, a phosphodiesterase inhibitor (eg, theophylline salts [aminophylline]), and/or an anticholinergic agent.

Intravenous steroids are indicated in severe bronchoconstriction, especially in asthmatic patients, for whom they have been effective in the past. With the administration of 100% oxygen, blood oxygenation usually is not the main problem in patients with bronchoconstriction; it is the progressive development of hypercarbia and the trapping of air in lung parenchyma that reduces ventilatory compliance and increase intrathoracic pressure. These, in turn, reduce venous return and may cause a tamponade-like impairment of cardiac function.

Impaired alveolar-capillary membrane gas exchange can result from alveolar pulmonary edema (treated by diuretics, inotropes, and vasodilators), decreased pulmonary perfusion (treated by inotropes and vasodilators), and lung consolidation (antibiotic therapy for infection).

Restoration of Spontaneous Ventilation and Airway Protective Mechanisms

The anesthesiologist attempts to tailor the anesthetic plan according to postoperative expectations for the patient. In the relatively healthy patient for whom tracheal extubation can be anticipated in the operating room, the goal is to have the patient breathing spontaneously with airway reflexes intact and the patient arousable to command immediately on completion of the operation. The challenge for the anesthesiologist is to maintain satisfactory general anesthesia through the entire course of the operation and yet have the patient sufficiently recovered from anesthetic drugs, including hypnotics and opioids, shortly after conclusion of the operation. If this is not possible, then the patient is transferred to the PACU to allow additional time for elimination of drugs that depress spontaneous ventilation and cough reflexes. Another possibility is to administer antagonists to opioids (eg, naloxone) and benzodiazepines (eg, flumazenil), but this approach risks sudden awakening, pain, and uncontrolled autonomic sympathetic activity leading to undesirable hemodynamic changes. And there is the risk of recurrent ventilatory depression because it is difficult to match the doses of the antagonists to the residual amounts of anesthetic drugs. On the other hand, it is fairly routine for the effects of neuromuscular blockers to be antagonized by administration of an anticholinesterase (eg, neostigmine) in combination with an anticholinergic agent (eg, atropine) to limit the autonomic cholinergic side effects of the anticholinesterase.

When the expectation is for maintenance of mechanical ventilation for some time in the postoperative period, then the patient's tolerance of the endotracheal tube is facilitated by the persistent effect of residual anesthetic drugs subsequently supplemented by administration of intravenous hypnotics (eg, propofol) and opioids (eg, fentanyl or morphine). These agents can be associated with side effects, including respiratory depression, especially when they are used concurrently. Dexmedetomidine (Precedex), an alpha2-adrenergic agonist, may offer advantages for sedation during weaning from mechanical ventilation because it provides sedation, pain relief, anxiety reduction, stable respiratory rates, and predictable cardiovascular responses. Dexmedetomidine facilitates patient comfort, compliance, and comprehension by offering sedation with the ability to rouse patients. This "arousability" allows patients to remain sedated yet communicate with health-care workers.

When the appropriate time comes to have the patient take over his or her own ventilation completely, these sedative and analgesic drugs are weaned to a level allowing satisfactory maintenance of blood oxygenation and carbon dioxide removal, easy arousal of the patient, and at least partial restoration of airway reflex mechanisms.

Pharmacology Related to Airway and Lung Management

Cardiac surgical patients are usually kept intubated and ventilated at the end of surgery, and ventilator support is withdrawn in the intensive care unit. Sedative drugs are required to facilitate this transfer from operating room to intensive care, and while the patient is ventilated. Some patients require continued ventilation because of primary pulmonary disease, underlying cardiac disease and/or perioperative complications. Others may require medications to treat bronchospasm or airway obstruction, or to facilitate re-intubation if they develop respiratory failure once extubated.

Sedative Medications to Facilitate Mechanical Ventilation

Propofol

The medications used in cardiac surgery patients are the same as those used in other mechanically ventilated patients. A common approach is to use propofol, an intravenous anesthetic/sedative agent that is rapidly eliminated by the liver and therefore facilitates rapid awakening. This drug is not water soluble so is supplied in a lipid emulsion; for short-term infusions (less than a day or two) lipid accumulation is not an issue, but triglyceride levels should be measured if the drug is given in high doses, especially at the same time as intravenous feeding. The hemodynamic effects of propofol are mostly a result of vasodilation, both arterial and venous, as well as the usual "withdrawal of sympathetic tone," which occurs when anxious, stressed patients are sedated. In addition there is a mild bradycardia and blunted heart rate response to hypotension, and a mild negative inotropic effect usually not relevant at clinical doses. The clinical dose range for sedation is 25 to 75 ug/kg per minute.

Benzodiazepines

Benzodiazepines are also used to facilitate mechanical ventilation, commonly midazolam (Versed) or lorazepam (Ativan). These drugs do not have the vasodilating effects of propofol, are good amnestic agents, but have a major disadvantage in that awakening is less predictable and more often prolonged, especially with lorazepam. In addition, these drugs appear to be more associated with postoperative delirium than other sedative drugs. Despite these problems, benzodiazepines are useful in the sedation of severely hypotensive patients requiring vasoconstrictor support. They have minimal if any direct cardiovascular effects, but withdrawal of sympathetic tone can lead to hypotension. Midazolam is usually administered as a continuous infusion of 1 to 4 mg/h, while lorazepam is most often given by intermittent dosing of 1 to 2 mg every 4 to 6 hours.

Dexmedetomidine

A relatively new addition to the sedative drug family is dexmedetomidine (Precedex), a centrally acting alpha2 agonist. This drug is the same class as clonidine, but has a much greater affinity for the alpha receptor. Not surprisingly, the most common hemodynamic effect, which sometimes limits its use, is bradycardia and hypotension. This drug can be used as a single agent to facilitate transfer from the OR to the ICU and the period of ventilator weaning and extubation. The major advantage of dexmedetomidine is that it does not affect ventilatory drive and patients can usually be aroused while receiving the drug. It also has a significant analgesic component, different from propofol or the benzodiazepines. The usual infusion rate of dexmedetomidine is 0.2 to 0.7 ug/kg per hour, preceded by a loading dose of 1 ug/kg over 20 to 30 minutes. The loading dose can be reduced or not given if hemodynamics warrant.

Opioids

The main reason patients need sedation during mechanical ventilation is to tolerate the presence of a transoral tracheal tube. In addition to general discomfort, there is continuous airway stimulation by the foreign object. To blunt airway reflexes as well as to treat postoperative pain, it is useful to add an opioid drug to whichever sedative agent the patient is receiving. For early postoperative awakening (ie, in the first hours) this is less of an issue as there is usually residual opioid effect from drugs given intraoperatively; however, patients will need additional opioid as they awaken and experience pain. For patients ventilated overnight or longer, use of opioid drugs will have these same benefits and likely result in a reduced need for the sedative drug. Fentanyl is a commonly used drug in this setting, partly because it has a rapid onset of action and relatively rapid recovery if dosing is not prolonged. Infusions of 1 to 4 ug/kg per hour are used. Alternatively intermittent dosing with 1 to 4 mg of morphine or 0.5- to 1-mg hydromorphone (Dilaudid) every 1 to 4 hours may be employed.

Medications Used to Facilitate Urgent Tracheal Intubation

Patients in extremis (eg, during cardiac or respiratory arrest) rarely require medications to facilitate intubation. In decompensating patients who are fully or partially conscious, challenges may include agitation, resistance to opening the mouth, and/or closed vocal cords when intubation is attempted. In some patients small doses of sedative/opioid medications with a rapid onset such as midazolam and fentanyl, respectively, are adequate; in others an anesthetic induction agent such as propofol or etomidate can be used, with or without a paralytic drug. The disadvantage to propofol is hypotension caused by vasodilation, as referred to in the preceding. This drug should be used in incremental doses, 10 to 20 mg at time, until a dose of 1 to 2 mg/kg has been given. Etomidate has no direct cardiovascular effects and is usually given in a dose of 0.15 to 0.3 mg/kg. This drug has an unusual side effect of impairing steroid synthesis and has been associated with adrenal insufficiency in critically ill patients when given by infusion. In addition, many patients experience unusual movements as this drug is administered. In general, the use of paralytic drugs is discouraged unless there is an anesthesiology provider or someone skilled in the use of these drugs and airway management present.

Succinylcholine is the most rapidly acting paralytic agent with an onset of 30 to 60 seconds, and is given in a dose of 1 mg/kg. This depolarizing agent causes potassium release and must be avoided in hyperkalemic patients. It causes stimulation of muscarinic and nicotinic receptors, an increase in serum catecholamines, and many different, usually mild dysrhythmias have been reported. Alternatively, rocuronium in a dose of 0.5 mg/kg can be given, but 60 to 90 seconds are required for adequate paralysis. Rocuronium in this dose has no cardiovascular effect. If an anesthetic induction agent and/or paralytic drug are being considered, an important issue is the possibility of aspiration of stomach contents. In an urgent situation the risk of this is reduced if the drugs are administered in rapid sequence with an assistant providing cricoid pressure (to occlude the esophagus), and manual ventilation is minimized until the tube is in place.

Drugs to Treat Airway Pathology: Stridor/Edema, Bronchospasm, Secretions

Fluid overload, prolonged positioning with the head dependent, or superior vena cava obstruction may all lead to edema of the upper airway. Alternatively, trauma from intubation or other devices in the oropharynx (eg, TEE probe) may cause glottis edema. There may be edema below the cords resulting from presence of the endotracheal tube/cuff. The common treatments for airway narrowing resulting from edema are: (1) use of a mixture of helium/oxygen; (2) inhaled racemic epinephrine; (3) diuretics and head-up position; and (4) a brief course of dexamethasone (IV).

Helium/Oxygen

Helium with oxygen (Heliox) is supplied as an 80%:20% mix. Helium is less dense than nitrogen, facilitating laminar rather than turbulent flow through a narrowed airway. If a patient can not tolerate the relatively low concentration of oxygen (20%), then additional oxygen will need to be administered. Although the greatest benefit is provided by 80% helium, concentrations of 40 to 50% provide some benefit.

Racemic Epinephrine

Racemic epinephrine is supplied as 1% solution and administered through an aerosol mask in a dose of 2.5 mL to cause vasoconstriction and reduced airway edema. Absorption of epinephrine can cause tachycardia and hypertension.

Diuretics and Dexamethasone

Specific diuretic drugs are discussed in the preceding; when a rapid effect is desired then usually a loop diuretic such as furosemide is employed. Although this may not specifically treat airway edema, the combination of elevated head and diuresis will reduce edema in the upper body. Dexamethasone is used as part of the "steroids for anything that swells" philosophy. There is little evidence of efficacy of corticosteroids in this setting; dexamethasone is used to avoid the mineralocorticoid effect of other potent intravenous steroid preparations. The usual dose is 8 mg followed by 4 mg every 6 hours for four to eight doses.

Bronchodilators and Mucolytics

A summary of bronchodilator and mucolytic drugs is given in Table 4-7. In the cardiac surgical patient, wheezing is more likely because of fluid overload or left ventricular failure than primary bronchospastic disease, but inhaled bronchodilators may give some relief while treatment of the primary disorder is being instituted (eg, diuresis or administration of an inotropic drug). In order of preference, inhaled beta agonists, anticholinergic agents, and intravenous steroids are used.137 All three may be indicated. Beta agonists relax smooth muscle in the airways and usually have the most rapid therapeutic effect. In the acute setting aerosolized drug via a facemask is more effective than metered-dose inhalers. It should be noted that management of chronic asthma focuses on inhaled anti-inflammatory agents, but these drugs are not helpful in the acute setting such as postoperatively. The presence of tenacious secretions should prompt consideration of an inhaled mucolytic such as dornase/DNAse (Pulmozyme). N-acetylcysteine (Mucomyst) is another mucolytic but may cause airway irritation so should not be given in the setting of acute bronchospasm. See the table for dosages of these drugs (see Table 4-7).