Chapter 23. Diagnostic and Management Strategies for Acute Heart Failure in the Intensive Care Unit

• Patients are admitted to the intensive care unit (ICU) with manifestations of acute heart failure (AHF) that arise in three general contexts: decompensation of chronic heart failure, as a complication of a cardiac process such as ischemia or valve incompetence, and when cardiomyopathy complicates other critical illness.
• Determining which of the contexts of AHF is present is invaluable to guide therapy.
• Predominant signs and symptoms arise from venous congestion after elevation of ventricular end-diastolic pressure and hypoperfusion; in a given patient, different contributions of these processes may be present.
• Measurement of brain natriuretic peptide has been a useful addition to the diagnostic armamentarium to determine whether heart failure is a cause of acute dyspnea; this measurement is most useful when obtained acutely and before initiation of therapies and, hence, has a limited role for assessing patients after admission to the ICU.
• Echocardiography is an extremely useful tool to assess ventricular function and define cardiac anatomy; on occasion, information from invasive measurement of intravascular pressure or venous oxygen saturation is also required to guide therapy.
• Many patients will exhibit respiratory distress and different degrees of impaired oxygenation; whereas many, if not most, patients can be managed with oxygen for this component of AHF, ventilatory support in the form of continuous positive airway pressure or bilevel airway support (BiPAP) should be considered for patients not responding adequately to oxygen therapy alone or whose respiratory symptoms and findings are severe from the onset.
• Diuretics are useful for treating venous congestion and inotropes for inadequate perfusion, but each carry risk of excessive dosing; the mainstay of therapy for most patients should be afterload reduction and search for the underlying causes of ventricular dysfunction and decompensation.
• Vasoconstrictive agents should be used only when and as long as truly life-threatening hypotension is present

Acute heart failure (AHF) is the primary or an underlying diagnosis in many patients admitted to the intensive care unit (ICU), but its exact incidence is unknown. The cause of heart failure (HF) in 60% to 70% of hospitalized patients is ischemic heart disease,1–3 but many diagnoses, including arrhythmias, idiopathic dilated cardiomyopathy, systemic or pulmonary hypertension, congenital and valvular heart disease, or myocarditis, should be considered (see Chap. 24). HF is complicated by diabetes in 27% of hospitalized patients, renal dysfunction in 17%, and respiratory disease in 32%, and the vast majority of individuals hospitalized for AHF are older than 70 years.3

AHF may present with left or right HF or the combination of these conditions. The cardiac dysfunction may be systolic or “diastolic” (with preserved ejection fraction), and the underlying pathogenetic mechanism may be cardiac or extracardiac and may induce transient or permanent cardiac damage.4 Especially in the ICU, multiple extracardiac pathologies may result in AHF by changing preload, afterload, or contractility, including pericardial disease, renal failure, endocrinopathy, sepsis, thyrotoxicosis, anemia, end-stage liver disease, and central nervous system lesions. Some rare cardiac pathologies also may be responsible for AHF, such as tumors of the heart and cardiac contusion.

Mortality rate is high, with as many as 13.5% of patients succumbing in the first 3 months after an episode of AHF in some series.3 The incidence of hospitalizations for HF as a primary diagnosis have increased recently, and 24% of patients will be rehospitalized within 90 days. Most patients hospitalized for management of HF represent cases of decompensation of chronic HF.3,5–7

This chapter offers a definition of the clinical syndrome of HF and then describes the underlying pathophysiology. Diagnostic and therapeutic approaches to patients admitted to the ICU with HF are described. The definitions and diagnostic and management strategies described in this chapter have been recently formulated as Guidelines of the Task Force on Acute Heart Failure by the European Society of Cardiology.1

The clinical syndrome of AHF may occur with or without previous cardiac disease. The time needed to characterize HF as acute has not been defined. The different clinical presentations of AHF are:1,8

1. Pulmonary edema: AHF accompanied by severe respiratory distress and O2 saturation (SaO2) less than 90% with room air before treatment.

2. Cardiogenic shock: Tissue hypoperfusion, after correction of preload, induced by cardiac disease. There is no simple and precise definition for cardiogenic shock. It is characterized by decreased systolic blood pressure (<90 mm Hg; or a drop of mean arterial blood pressure >30 mm Hg) and/or low urine output (<0.5 mL/kg per hour) with a pulse rate faster than 60 beats/min with or without evidence of organ congestion.

3. Hypertensive AHF: Signs and symptoms of HF accompanied by extremely high blood pressure and relatively preserved left ventricular systolic function.

4. Acute decompensation of heart function, occurring de novo or as deterioration of chronic HF: Signs and symptoms of AHF that do not fulfill criteria for cardiogenic shock, pulmonary edema, or hypertensive HF.

5. High-output failure: Characterized by warm periphery, pulmonary congestion, and sometimes low blood pressure, as seen in thyrotoxicosis.

6. Right HF: Characterized by low-output syndrome with increased jugular venous pressure, peripheral edema, and sometimes hypotension. It is often associated with left HF but may occur in isolation as a result of lung or pulmonary vascular disease or ischemia largely limited to the right ventricle.

Many other classifications for AHF have been used in the ICU. The most widely recognized are the Killip and the Forrester classifications originally designed to provide a clinical estimate of prognosis in acute myocardial infarction.9,10 The Forrester classification categorizes patients by clinical and hemodynamic criteria to four different subsets:

1. No evidence of pulmonary congestion or peripheral hypoperfusion (normal cardiac index and pulmonary capillary pressure).

2. Pulmonary congestion only (elevated pulmonary capillary wedge pressure [PCWP]).

3. Hypoperfusion only (depressed cardiac index).

4. Pulmonary congestion and hypoperfusion (depressed cardiac index and elevated PCWP).

These definitions have been used to classify AHF patients in an attempt to improve prognostic information and management. However, AHF is a continuum, and this should be taken into consideration during the diagnostic and therapeutic assessments.

AHF is characterized by a relatively sudden inability of the heart to maintain cardiac output adequate for the demands of the circulation. Cardiac output is reduced with associated tissue hypoperfusion, usually with increased cardiac filling pressures (as reflected by an elevated PCWP) and associated tissue congestion. The primary response to this state is neuroendocrine activation, such as the renin-angiotensin-aldosterone axis, sympathetic system activation, cytokine and endothelin release, and peptide release, including vasopressin and natriuretic peptides.11,12 Natriuretic peptides cause vasodilation, inhibit endothelin release, and counterbalance the effects of the sympathetic and renin-angiotensin-aldosterone systems. Moreover, they have diuretic, lusitropic (ventricular relaxation), and antifibrotic properties.13,14

Although activation of these pathways is no doubt an attempt to maintain homeostasis, the final result of neuroendocrine activation is vasoconstriction leading to retention of sodium and water, increased left ventricular filling pressures, and increased ventricular wall tension. This in turn causes myocardial oxygen demand to increase, unfortunately in a circumstance in which coronary blood flow cannot be increased in a commensurate fashion because of associated coronary artery disease or primary myocardial dysfunction and pump failure. In addition to renal and vascular effects, neurohormonal activation and hypoperfusion exert direct effects on the cardiac cells, leading to further deterioration of myocardial function within hours by induction of apoptosis and necrosis.15–17 The significance of myocardial cell regeneration, which has been recently described in humans, is not known.18 Large-dose catecholamine treatment can further induce cell death and desensitization of the β-adrenergic system,19 leading to tolerance to inotropic stimulation by therapeutic catecholamines.20 Oxygen free radical formation and endothelial injury have been implicated in AHF pathophysiology and may not only lead to abnormalities in the peripheral circulation but may also contribute to the pathogenesis of myocardial stunning and hibernation.

In patients with ischemic heart disease and acute coronary syndromes, the AHF syndrome may persist for hours to days after the restoration of coronary blood flow. This transient reduction in myocardial function, which may occur also after cardiac arrest, is called myocardial stunning.21 A mechanism of depressed myocardial function that is also important in the ICU is myocardial hibernation. This impaired myocardial function is a result of severe chronic hypoperfusion and decreased coronary blood flow due to coronary artery stenosis. It appears to be a protective mechanism aimed at limiting myocardial cell injury and death under conditions of excessive load and inadequate perfusion.22 These mechanisms may be particularly important in the ICU patient with a history of ischemic heart disease and myocardial dysfunction, even when AHF is not the primary admission diagnosis. Patients with stunned and hibernating myocardium can return to normal cardiac function when appropriately treated. Hibernating myocardium improves when coronary blood flow is restored, whereas stunned myocardium responds to inotropic stimulation.23,24 However, the exact role of treatment on the outcome of critically ill patients with coexisting coronary artery disease is unknown.

Etiologic diagnosis of AHF based on history and on clinical findings is difficult in the ICU. A 12-lead electrocardiogram is useful to identify acute myocardial infarction. It may also help in the diagnosis of acute right ventricular strain; ST elevation of the V4R lead is characteristic of right ventricular myocardial infarction, which is present in 30% to 50% of patients with an inferior myocardial infarction.

Chest radiography is necessary to evaluate cardiac size and shape and to assess pulmonary congestion. Blood count and biochemical tests, including calcium and thyroid function, are necessary in the initial evaluation of AHF patients. Troponin and creatinine kinase MB fractions have been used to evaluate acute myocardial injury, but these standard markers often lose sensitivity and specificity in complex ICU patients with confounding effects of renal failure, trauma, and recent surgery.25,26 Arterial blood gas analysis is extremely helpful to identify oxygenation abnormalities associated with pulmonary edema and metabolic acidosis resulting from hypoperfusion. Pulse oximetry and end-tidal CO2 are helpful monitoring tools but are less reliable in shock states.

Natriuretic peptides and especially B-type natriuretic peptide (BNP) play a significant role as adaptive mechanisms in AHF. BNP is released from the cardiac ventricles in response to increased wall stretch resulting from pressure and volume overload.27 BNP and NT-BNP (the product of the posttranslational processing of the pro-BNP molecule) have been used for the diagnosis of decompensated HF in patients with dyspnea in the emergency department1,28 and increased levels of plasma BNP and NT-BNP carry important prognostic information in patients with decompensated HF.29,30 Moreover, it has been suggested that increased BNP level is a strong predictor for cardiac dysfunction in ICU patients.31 However, the exact time of BNP rise in patients with AHF is not known, and BNP levels may be normal at the time of admission. BNP has a half-life of 22 minutes and NT-BNP has a half-life of 120 minutes, suggesting that meaningful hemodynamic changes could be reflected after 2 to 12 hours.32 Many of the interventions used to treat AHF no doubt rapidly influence BNP levels, thus making use of this test questionable once the patient has had treatment initiated. Moreover, BNP concentration increases in patients with chronic renal failure, pulmonary embolism, pulmonary hypertension, and chronic obstructive pulmonary disease with right heart strain. Thus, an elevated level cannot be taken as unequivocal evidence of left HF, and patients with findings consistent with acute right heart strain require further diagnostic workup.33–38 The effect of acute lung injury or ARDS with or without right HF upon BNP serum levels has not been evaluated. Plasma BNP is also increased after acute myocardial infarction39 and is a predictor of major adverse events independent of ejection fraction.40,41

Although it has been suggested that increased plasma BNP concentrations are a useful marker for predicting cardiac rupture after acute myocardial infarction, the exact role of BNP in patients with cardiogenic shock or pulmonary edema from a mechanical complication after an acute myocardial infarction has not been extensively evaluated.42

Echocardiography is essential for the evaluation of the patient with AHF. Transthoracic echocardiography can be used to evaluate left and right ventricular functions, valvular integrity, pericardial pathology, and mechanical complications of acute myocardial infarction and is useful to identify intracavitary lesions. An echo Doppler study can be used for the hemodynamic evaluation of the HF patient, although in the ICU transesophageal echocardiography is often used because of the limited quality of the transthoracic images, particularly in patients undergoing mechanical ventilation.43–45 Echocardiography has not been compared with right heart catheterization in patients with AHF, but the information from these two tools would likely be complementary by providing pressure and volume assessments of ventricular function.

Insertion of a pulmonary artery catheter for the diagnosis of AHF is usually unnecessary but can be helpful in complex patients with concurrent cardiac and pulmonary disease.45 The measurement of mixed venous saturation or even of superior vena caval blood saturation as a surrogate measure (obtainable from a standard central venous catheter) is very useful to determine the adequacy of the cardiac output; saturations less than 60% strongly suggest inadequate tissue perfusion apart from the cardiac output measured by other means.

There are no prospective randomized clinical trials assessing the efficacy of different monitoring devices in AHF patients. However, as in all critically ill patients, temperature, respiratory rate, heart rate, electrocardiogram, and blood pressure should be monitored. The frequency with which these observations need to be taken depends on the severity of decompensation. Blood pressure should be checked regularly until the patient has demonstrated stability on a given dosage of vasodilators, diuretics, or inotropes. Noninvasive blood pressure monitoring is usually sufficient. An arterial catheter is sometimes used in shock states or when continuous analysis of arterial blood pressure or multiple arterial blood analyses are necessary. The SaO2 of hemoglobin should be checked by pulse oximetry, continuously, or at regular intervals in any patient receiving oxygen therapy for AHF. The SaO2 estimate by pulse oximetry is of limited value when the patient is in cardiogenic shock.

Many different noninvasive hemodynamic monitoring devices have been used in the ICU, but their exact role in the management of an AHF patient has not been definitively determined. Central venous lines can be used for the delivery of fluids and drugs to the central venous circulation and for monitoring right atrial pressure. However, right atrial pressure correlates poorly to preload in HF patients.

Pulmonary artery catheters (PACs) can measure intracardiac pressures, cardiac output, right ventricular end-diastolic volume and ejection fraction, and mixed venous saturation (see Chap. 13). Several retrospective studies have demonstrated an increased mortality rate associated with the use of this technology, including in patients with acute myocardial infarction.46,47 These observations have been ascribed to limitations of the retrospective study, but the consistency of this association in other patient groups raises concern regarding the value of this mode of monitoring.48–50 Recently, a prospective study randomized critically ill patients to a PAC group or to treatment without the use of data from a PAC. In this study, management with the PAC led to increased fluid resuscitation within the first 24 hours but no long-term differences in patient outcome.51 Other studies evaluating the utility of PAC monitoring in patients with AHF and other causes of critical illness are currently underway.

AHF is a complex and heterogeneous syndrome, with only a few randomized controlled trials that have assessed treatment approaches. The comorbidities that are common in ICU patients make management more difficult and complicated. The immediate goals of treatment include improvement of tissue perfusion and oxygenation, correction of underlying hemodynamic abnormalities and the causes of cardiac decompensation, and control of symptoms. These short-term goals should then be linked to longer-term management strategies that ideally improve morbidity and mortality rates.

Respiratory Therapy and Ventilatory Support

A priority in treating patients with acute cardiac failure is the maintenance of arterial oxygen saturation (SaO2) within the normal range to improve tissue oxygenation. This can be achieved by using supplemental oxygen in most patients and by the use of noninvasive (see Chap. 33) or invasive (see Chap. 36) ventilatory strategies for patients with greater degrees of respiratory compromise.

Several trials have suggested a benefit of continuous positive airway pressure (CPAP) in patients with AHF, in particular a reduction in the need for endotracheal intubation.52–56 These investigations have been criticized for their small size and a failure to demonstrate a difference in mortality rate, but a recent meta-analysis showed a trend to decreased hospital mortality rate when CPAP was used as a routine measure for patients with AHF.57

Other forms of noninvasive ventilatory support have been studied, most often bilevel support delivered by nasal or full-face mask. Three randomized trials compared this approach with conventional oxygen therapy, large-dose nitrate therapy, or CPAP,58–60 with mixed results. A recent multicenter study of treatment of patients in pulmonary edema secondary to AHF showed no benefit for the general population of patients but did show a trend toward improvement by subset analysis of the more severely compromised patients.61

Analysis of these studies suggests that benefit is likely to be seen in more severely ill patients not stabilized with oxygen therapy alone. One approach is to use CPAP or other forms of support in patients with pulmonary edema, an elevated arterial partial pressure of CO2 (suggesting impending ventilatory failure) and a low pH even when an arterial pressure of CO2 is normal (suggesting more severe hypoperfusion).

Medical Treatment

Vasodilators are the first line therapy in patients with pulmonary congestion and adequate blood pressure. Nitrates relieve pulmonary congestion by reducing left ventricular preload and afterload. The highest tolerable dose of nitrates with low dose furosemide should be administered since it is more effective than furosemide treatment alone in acute pulmonary edema.62 However, close monitoring during intitiation of therapy is necessary because excessive vasodilation may induce a steep reduction in blood pressure, ischemia, sometimes shock and renal failure. Moreover, rapid development of tolerance to nitrates, especially when they are given intravenously, limits their effectiveness after 16 to 24 hours.

Sodium nitroprusside (SNP) has been also used for patients with acute heart failure and has found its widest application in patients with hypertensive crisis and ventricular decompensation. Controlled trials of its use are lacking in heart failure per se. Use of SNP requires close blood pressure monitoring. Prolonged administration can be associated with thiocyanate and/or cyanide toxicity and patients should be converted early to alternative agents to avoid these problems, particularly when renal and/or hepatic dysfunction are present.

Diuretics are indicated in patients with symptoms secondary to fluid retention. Diuretics increase the urine volume by enhancing the excretion of water and sodium chloride leading to decreased plasma and extracellular fluid volume and a decrease in peripheral and pulmonary edema. The acute hemodynamic effect of the intravenous administration of loop diuretics is vasodilation manifested by a reduction in ventricular filling pressures and in pulmonary vascular resistance.63 With high doses reflex vasoconstriction has been reported. In acutely decompensated heart failure the use of diuretics normalizes loading conditions and may reduce neurohormonal activation in the short term.64,65

Intravenous administration of loop diuretics (furosemide, bumetanide, torasemide) is preferred in acute heart failure. The dose should be titrated according to the diuretic response and relief of congestive symptoms. Administration of a loading dose followed by continuous infusion has been shown to be more effective than bolus alone with a lower incidence of toxicity.65–69 Diuretic resistance has been described and when present alternative means of fluid removal by dialysis or ultrafiltration should be considered.

Morphine can be used in the treatment of patients with pulmonary edema although its effectiveness has not been studied in clinical trials. Morphine induces venodilation and mild arterial dilation, reduces heart rate70 and relieves breathlessness in patients with acute pulmonary edema.

There is no evidence from clinical trials to support the use of anticoagulants in critically ill patients with AHF without myocardial ischemia or atrial flutter/fibrillation. However, the morbidity and mortality from thromboembolic disease in the ICU warrants the consideration of prophylactic anticoagulation in patients with AHF. A placebo controlled study evaluating enoxaparine in acutely ill patients, which included patients with AHF, demonstrated a reduction in the incidence of complicating venous thromboembolism but no improvement in survival or duration of hospitalization.71

Nesiritide, a recombinant human BNP, has been used for the treatment of AHF. Nesiritide reduces right atrial pressure, pulmonary capillary wedge pressure and increases stroke volume. Additional beneficial actions include coronary artery vasodilation, diuresis, natriuresis and neurohormonal suppression.72 The drug has been shown to be more efficacious than placebo in symptom improvement without a pro-arrhythmic effect. When compared to intravenous nitroglycerin neseritide results in a more rapid improvement in hemodynamic profile in patients with AHF.73 In an open-label randomized trial of nesiritide and standard care versus standard care alone, patients treated with nesiritide exhibited a trend toward decreased readmissions and six-month mortality.74 Nesiritide is administered as a 2 μg/kg bolus followed by a fixed dose infusion of 0.01 μg/kg per minute. The dose can be increased every three hours, to a maximum of 0.03 μg/kg/min, if the pulmonary capillary wedge pressure is more than 20 mm Hg and systolic blood pressure is more than 100 mm Hg (using a 1 μg/kg bolus followed by an increase of 0.005 μg/kg/min over the previous infusion rate). If hypotension occurs, the infusion should be discontinued and restarted after blood pressure stabilization, at a 30% lower infusion rate without bolus. Hypotension is a significant adverse effect of the drug, and may persist despite immediate cessation of infusion. Accordingly, nesiritide should not be employed in patients with extremely labile blood pressure dipping below a level of 90 mm Hg systolic.

In the near future, therapies directed at potentiating the beneficial effects of endogenous natriuretic peptides by reducing their degradation may find clinical application. A target under current investigation is neutral endopeptidase, the primary enzyme involved in their degradation. However, preliminary studies of inhibitors of neutral endopeptidase showed increased neurohormonal activation with increased levels of angiotensin II and endothelin-1 after enzyme inhibition.75 Thus, vasopeptidase inhibitors, which could provide simultaneous inhibition of angiotensin converting enzyme and neutral endopeptidase have been developed. Early use of such agents has been confounded by a high incidence of angioedema.76,77

Beta-blocking agents are contraindicated in patients with AHF, at least if it is clear that systolic dysfunction of the left ventricle is significant. Among patients in whom ischemia and tachycardia per se are felt to contribute significantly to their deterioration, beta blockade may be required; under such circumstances a short acting agent such as esmolol or metoprolol may given intravenously to permit cessation of therapy if adverse effects are encountered. Patients admitted to the ICU with AHF already receiving beta blockers should be continued on this therapy unless inotropic support is needed or excessive dosages is suspected.

In the initial management of the patient with acute heart failure, ACE inhibitors are not recommended and calcium antagonists are contraindicated. ACE inhibitors have an important role in the patient with systolic heart failure early after stabilization. Some new compounds with diuretic effect and vasodilating properties are under investigation, including vasopressin receptor antagonists.

Inotropic agents are indicated in the presence of hypotension and low cardiac output, with or without pulmonary congestion, refractory to optimal doses of diuretics and vasodilators (see Table 23-1). Inotropes are potentially harmful in chronic HF as they increase oxygen demand and calcium loading and should be used with caution.78–80 However, in patients with AHF, symptom relief and prognosis appear dependent on correction of grossly deranged hemodynamics. Only a few controlled trials assessing specific inotropic regimens have been performed and thus rigid guidelines for drug administration do not exist.80

Dopamine is an endogenous cathecolamine and a precursor of norepinpehrine. Its effects differ according to dosage. At low doses (≤2 to 3 μg/kg/min) activation of dopaminergic DA1 and DA2 receptors predominates with vasodilation in renal, mesenteric, coronary and cerebral vascular beds. The activation of DA2 receptors inhibits norepinephrine release from nerve terminals, activates the emetic center, decreases prolactin release and inhibits angiotensin II mediated aldosterone secretion. In the past, dopamine was widely used in the ICU to improve renal blood flow and diuresis in AHF patients with hypotension and low urine output. However, administration of low-dose dopamine to critically ill patients at risk of renal failure does not confer clinically significant protection from renal dysfunction.81,82 Small trials suggested some benefit of dopamine at low doses in patients with AHF but this general practice is not strongly supported by existing literature.83 In AHF with hypotension dopamine can be used at higher doses for both inotropic and vasoconstrictive purposes but adverse effects of increased afterload or mismatch of oxygen supply to myocardial demand should be anticipated.

Dobutamine is a synthetic, sympathomimetic agent with substantial beta agonist activity resulting in significant inotropic and chronotropic effects on the heart.84 It is usually initiated with a 2 to 3 μg/kg/min infusion rate without a loading dose. The infusion rate may be progressively titrated according to symptoms, diuretic response or assessments of hemodynamics (cardiac output, mixed venous oxygen saturation). The dose may be initially increased to 20 μg/kg/min as required. When undesirable side effects are encountered (tachycardia, hypotension, chest pain, arrhythmia), discontinuation of the drug is usually associated with reversal of effect due to short half life. Dobutamine may have an additive effect when given concurrently with phosphodiesterase inhibitors (see below). Prolonged infusion of dobutamine (above 48 hours) is associated with tolerance effects.85,86

Type III phosphodiesterase inhibitors (PDEI) are drugs with hemodynamic activity intermediate between that of a pure vasodilator and that of an inotrope.85 They are mainly used in patients with AHF and peripheral hypoperfusion, refractory to optimal diuretic and vasodilating therapy. Milrinone and enoximone are the PDEIs used in clinical practice. They act by inhibiting the breakdown of cyclic-AMP (cAMP) into AMP in both the vascular and myocardial smooth muscle. As a result, contractility improves in cardiomyocytes and relaxation in the smooth muscle cells. Because their site of action is distal to the beta-adrenergic receptors PDEIs maintain their effects even during concomitant beta-blocker therapy. When administered to patients with advanced HF, milrinone improves hemodynamics without tolerance development.86 Hypotension may be caused by excessive peripheral venodilation and is observed mainly in the patients with low intravascular volume. Therefore PDEIs are used in patients with preserved systemic blood pressure. Outcome data regarding PDE-I administration in patients with acute HF are limited but not encouraging especially in patients with ischemic cardiomyopathy. Based on hemodynamic data alone, the combination of milrinone and dobutamine has a greater positive inotropic effect than each drug alone. However, the pharmacokinetic profile of milrinone is not optimal for ICU patients because individual responses to milrinone vary considerably.

A new class of medications which enhance calcium sensitization of the contractile proteins has found application in Europe and agents are in survival clinical trials in Europe and North America. Levosimendan is such an agent, and has been used in patients with symptomatic low cardiac output secondary to cardiac systolic dysfunction without severe hypotension. Levosimendan improves cardiac contractility by increasing the sensitivity of troponin C to calcium although is also type III PDE inhibitor.87–91 Activation of adenosine triphosphate (ATP) sensitive potassium channels are responsible for its vasodilatory properties. Levosimendan has an active, potent, acetylated metabolite with half-life of approximately 80 hours, which probably explain the prolonged hemodynamic effects of a 24-hours levosimendan infusion.88–91 Levosimendan has not been used extensively in patients with AHF. In patients with decompensated chronic systolic HF increases cardiac output, heart rate and stroke volume and decreases the pulmonary capillary wedge pressure and systemic and pulmonary vascular resistance.88–91 An improvement in symptoms and decreased mortality has been shown in trials comparing levosimendan with dobutamine in patients with advanced chronic HF.91 The hemodynamic effect of levosimendan is maintained in heart failure patients on concomitant beta-blocker therapy. Tachycardia and hypotension are its main side effects and levosimendan is not recommended in patients with low systolic blood pressure.

When the combination of inotropic agent and fluid challenge fails to restore organ perfusion and blood pressure, therapy with vasoconstrictors may be needed. Since this treatment may result in excessive afterload applied to a failing ventricle, it should be used as an emergency measure to sustain life and maintain perfusion in the face of life-threatening hypotension. Recent data indicate that if a vasopressor is needed, norepinephrine may be used because epinephrine has more negative effects on splanchnic circulation and on lactate metabolism.92,93

Cardiac glycosides inhibit sodium-potassium ATPase, thereby increasing calcium-sodium exchange with resulting positive inotropic effects. In AHF, cardiac glycosides produce a small increase in cardiac output and a reduction of filling pressures.94,95 However, in patients with heart failure following myocardial infarction, increase of creatine kinase was more pronounced in patients receiving cardiac glycosides,96,97 and in some studies a proarrhythmic effect and an adverse effect on outcome have been described. Accordingly, cardiac glycosides are not recommended for the acute management of AHF especially after acute myocardial infarction.

Mechanical Assist Devices

Mechanical circulatory support may be used in patients with AHF without permanent end-organ dysfunction, non-responding to conventional therapy, as a bridge to recovery or as a bridge to heart transplant. In clinical trials left ventricular assist devices have been used as destination therapy for patients with advanced chronic heart failure but device related complications limit their use for this indication.

Intraaortic Balloon Pump (IABP)

IABP is most frequently used in patients with cardiogenic shock or ongoing severe myocardial ischemia, refractory to initial medical therapy, in preparation for revascularization. It is also useful in patients with acute myocardial infarction complicated by significant mitral regurgitation or rupture of the interventricular septum, to obtain hemodynamic stabilization for definitive diagnostic studies and treatment or after cardiac surgical procedures for patients with low cardiac output states.98 The IABP has obvious advantages over other cardiac assist devices because of the rapidity and ease of insertion, accomplished through a femoral artery and performed at the bedside of the ICU patient. However, IABP does not augment forward flow to the extent achieved with ventricular assist devices (see below).

IABP operation is based on the principle of counterpulsation. A balloon is positioned in the descending aorta just distal to the left sublavian artery. The balloon deflates at the onset of systole immediate prior to the ejection of blood from the left ventricle. Thus reduces afterload, thereby decreasing left ventricular work and myocardial oxygen consumption. It inflates in diastole, increasing diastolic pressure resulting in increased coronary artery perfusion. In the setting of non-ischemic heart failure the afterload reduction may be still useful but the benefits of increased coronary flow are not clear. Vascular complications are common and careful monitoring of lower extremity perfusion is necessary. Intraabdominal (mesenteric ischemia, pancreatitis, splenic infarction), neurologic (stroke, paraplegia) and infectious complications have been described. Balloon rupture with helium embolization and balloon entrapment has been also reported. IABP is contraindicated in patients with aortic dissection or significant aortic insufficiency and its use should be restricted to patients with a correctable underlying condition.

Ventricular Assist Devices

AHF patients not responding to conventional treatment including intra-aortic balloon pumping are candidates for mechanical circulatory support with ventricular assist devices (VAD).99–101 This situation should be understood as one of bridging the patient ventricular function is expected to improve or transplantation is planned. VAD has been used after acute myocardial infarction, in post cardiac surgery low cardiac output syndrome and in acute myocarditis. Specific hemodynamic criteria have been used for patient selection in decompensated chronic heart failure but patient selection is more difficult in AHF. Patients with right heart failure are candidates for biventricular support, patients with aortic valve prosthesis are at increased risk for thromboembolism with some systems and aortic regurgitation is exaggerated after left VAD insertion because of the lowered left ventricular end diastolic pressure. An incompetent aortic valve may need repair or replacement before VAD insertion.

Many VAD are currently available and can be categorized in several different ways including: (a) left versus right VAD; (b) pulsatile versus non pulsatile flow; (c) internal versus external placement. No current single VAD is appropriate for the management of all AHF patients. Selection of the device is determined by the expected duration of support, the specific cardiac pathology, type of support needed (right, left or biventricular), the device availability and surgical team experience. It is not possible for all cardiac surgery centers to use all available systems. Ideally, a center should have access to a short term support system for AHF, a long term support system and a system capable of providing support to the right ventricle.101

The patient with a VAD is a challenge for the ICU team because thromboembolism, bleeding, infection, hemolysis and device malfunction are common complications associated with the use of these systems.

Acute Myocardial Infarction

In all patients with acute myocardial infarction and signs and symptoms of heart failure echocardiography is necessary for assessment of ventricular function and mechanical complications (see Table 23-2). In acute myocardial infarction complicated by cardiogenic shock emergency percutaneous intervention or surgery improves prognosis and should be considered at an early state. If neither of these are available, or can be only provided after long delay, fibrinolytic therapy should be considered.102,103 Temporary stabilization of the patient can be achieved by adequate fluid and pharmacological support, intra aortic balloon pump and mechanical ventilation.

Acute right heart failure after myocardial infarction is usually related to acute right ventricular infarction with a typical electrocardiogram and echocardiogram. Early revascularization of the culprit coronary artery may improve prognosis.

Free Wall Rupture

The incidence of cardiac wall rupture after an acute myocardial infarction is 0.8–6.2%.104–107 Recent studies have shown that percutaneous intervention decreased the incidence of cardiac rupture, compared with the thrombolytic era. Older age, female gender, anterior location of the infarct and longer time to reperfusion carry an increased risk of cardiac rupture. Usually, death occurs within minutes due to cardiac tamponade and the diagnosis is established in autopsy. However, in around 30% of cases thrombus seals the rupture giving an opportunity for intervention, if the condition is recognized. Most of these patients have signs of hemodynamic collapse and in some patients rupture is preceded by chest pain, nausea and T wave changes or new ST segment elevation in the infarct related leads. All these patients should undergo immediate echocardiography. The clinical presentation, pericardial effusions more than 10 mm and intrapericardial echo densities consistent with thrombus confirm the diagnosis. Pericardiocentesis, fluids and positive inotropes have been used for temporary hemodynamic stabilization. However, the patients should be immediately transferred to the operating room for surgical repair of the rupture. Free wall rupture has been also described during dobutamine stress echocardiography after acute myocardial infarction.107

Post-Infarction Ventricular Septal Rupture (VSR)

Acute rupture of the interventricular septum occurs in 1–2% of all infarcts. VSR usually occurs in the first 5 days after an acute myocardial infarction. However, in the thrombolytic era the incidence has decreased and the median time of VSR is 1 day.108–111 Septal rupture is more frequent in patients with transmural anterior myocardial infarction but in these patients VSR is generally apical and simple. In patients with an inferior myocardial infarction ruptures are in the basal inferoposterior septum and are more complex. The VSR is typically presented as acute pulmonary edema or cardiogenic shock in a patient with right-sided ventricular failure and a harsh, loud, holosystolic murmur usually at the left lower sternal border. With low output state the murmur is difficult to identify. Echocardiography is generally diagnostic. It will identify the site of the ventricular rupture, the size of the left to right shunt, left and right ventricular function and the existence of mitral incompetence. Severe mitral regurgitation coexists in 20 percent of patients with VSR. Pulmonary artery catheterization can be helpful if echocardiography is not diagnostic. It will document the increase in oxygen saturation in the right ventricle and the pulmonary to systemic blood flow ratio (usually more than 2). Temporary medical therapy consists of diuretics, vasodilators, inotropes, intraaortic balloon pump, oxygen and in some patients positive pressure ventilation. However, surgical repair should be performed soon after the diagnosis in most patients.108 The timing of the surgical repair has been an issue of debate. It was believed that in the acute phase of the myocardial infarction the myocardium was too fragile for the safe repair of the VSR. However, current Guidelines recommend immediate surgical repair, regardless of the clinical status of the patient, because the rupture can abruptly expand even in stable patients.1,111

In some small, retrospective clinical studies has been supported that concomitant coronary artery bypass grafting improves late functional status and survival.109 Thus, revascularization is usually performed although the number of patients in these studies was too small to draw any firm conclusions.109 Percutaneous closure of the VSR has been used to stabilize selected patients but more experience is needed before it can be recommended. VSR has been also described in patients with normal coronary arteries and myocardial bridge.

Moreover, a new systolic murmur and cardiogenic shock have been described in some patients with anterior myocardial infarction.112 In these patients with apical anterior myocardial infarction compensatory hyperkinesis of the basal segments of the heart induces left ventricular outflow tract (LVOT) obstruction. The syndrome persists until appropriate therapy decreases the LVOT obstruction.

Acute Mitral Regurgitation (MR)

Mild MR is common after an acute myocardial infarction and is not usually associated with hemodynamic compromise. However, acute severe MR is the cause of shock in approximately 10% of patients with cardiogenic shock after an acute myocardial infarction.113,114 It occurs 1–10 days after the infarction. Successful reperfusion by thrombolytic therapy or percutaneous intervention probably reduces the incidence of acute MR and is associated with an earlier occurrence (median 1 day). Several different mechanisms are responsible for acute MR after an acute myocardial infarction. The most dramatic is complete papillary muscle rupture where most of the non operated patients die in the first 24 hours. Cordae tendineae rupture, partial papillary muscle rupture or dysfunction of the papillary muscle are more common and with better survival. In most patients posteromedial papillary muscle is involved because of its single blood supply from the posterior descending coronary artery.

Acute severe MR is manifested by flash pulmonary edema and shock. The characteristic apical systolic murmur may be absent, due to the acute elevation of pressure in the non compliant left atrium and the rapid pressure equalization between left atrium and left ventricle. Moreover, auscultatory finding of pulmonary edema may mask the MR murmur. Chest radiograph shows pulmonary edema (might be unilateral). Echocardiography is the standard diagnostic tool for detecting the presence and severity of mitral regurgitation. The left atrium is usually small or slightly enlarged. Transesophageal echocardiography may be needed to establish the diagnosis in some patients with not adequate transthoracic images. However, when the mitral valve is not well visualized, the demonstration of preserved systolic function in a patient with acute myocardial infarction and pulmonary edema is helpful. A pulmonary artery catheter is rarely needed for diagnostic purposes but it may be useful in monitoring the effects of therapy.

The patients with severe acute MR and pulmonary edema or shock from papillary muscle or chordal rupture require emergency surgical repair. Even if medical therapy and intra aortic balloon pumping result in hemodynamic stabilization operation should be done early because many patients deteriorate suddenly.

Management of Acute Heart Failure from Prosthetic Valve Thrombosis

The incidence of prosthetic valve thrombosis depends on valve location and type and the adequacy of anticoagulation. It is from 0.2–6% per patient year in patients with mitral or aortic valve prostheses up to an overall 20% in patients with prostheses in the tricuspid valve.115–121 Clinical presentation of patients with prosthetic valve thrombosis (PVT) includes patients with clinically silent PVT, patients with stroke or peripheral embolism and patients with hemodynamic compromise and signs and symptoms of heart failure. All patients with a prosthetic valve and heart failure symptoms should undergo immediately a transthoracic and/or transesophageal echocardiographic study. Thrombolytic therapy is considered the treatment of choice for tricuspid valve thrombosis. The ideal therapy for left sided PVT remains controversial.116–118 Surgical mortality is high for emergency operations in critically ill patients with hemodynamic instability (New York Heart Association functional class III or IV symptoms, pulmonary edema, hypotension). Thus, thrombolytic therapy was recommended for high risk surgical candidates.116 However, thrombolysis takes several hours to be effective and this delay may lead to further deterioration in critically ill patients. The recommended dose of rtPA is 10 mg intravenous bolus followed by 90 mg infused over 90 min or 250–500,000 IU of streptokinase infused over 30 min followed by 100.000 IU per hour for 10 hours. Urokinase can be also used in a dose of 4,400 IU/kg/h for 12 hours without heparin or 2000 IU/kg/h with heparin for 24 hours. Echocardiography should be repeated in all patients after thrombolytic therapy. Thrombolysis is ineffective in approximately 18% and in these patients surgical intervention is indicated, although repeated infusions of thrombolytic therapy have been used in some studies.118,119 Moreover, mortality is 6% and the risk of thromboembolism is 12%. Patients with aortic PVT and patients with large or mobile thrombi have a higher risk for thromboembolic complications after thrombolytic therapy and surgical intervention should be considered in these patients. Moreover, thrombolysis is not effective when fibrous tissue ingrowth (pannus) is implicated in the obstruction of the prosthetic valve. Clinical history, presentation and transesophageal echocardiography are helpful in the differential diagnosis but sometimes the distinction is still difficult.120 Patients who are in stable clinical condition usually are not admitted in the ICU. Their management is also controversial. Small retrospective studies suggest that fibrinolytic therapy may be equally effective and in these patients.116,118

In conclusion, the clinical presentation of the individual patient, the characteristics of the PVT and the expertise at each center should be used to arrive at a therapeutic decision.