Chapter 26. Acute Right Heart Syndromes

• Right heart syndromes (RHS) as a cause of shock are less common than left heart dysfunction, but recognizing them requires a high level of vigilance.
• Clues to recognizing RHS as a cause of shock include a history of a condition that is associated with pulmonary hypertension, elevated neck veins, peripheral edema greater than pulmonary edema, or a right-sided third heart sound, in addition to electrocardiographic, radiographic, and echocardiographic findings.
• Echocardiography is extremely valuable, not only for demonstrating the presence of RHS, but also for guiding hemodynamic management.
• Progressive right heart shock can be worsened by excessive fluid infusion, concomitant left ventricular failure, inappropriate application of extrinsic positive end-expiratory pressure (PEEP) and hypoxia.
• The drug of choice for resuscitation of patients with acute RHS is dobutamine, initially infused at 5 μg/kg per minute. Systemically-active vasoconstrictors may provide additional benefit.
• Prostacyclin and nitric oxide are often beneficial in improving pulmonary hemodynamics and oxygenation, but may not improve survival.

In the majority of patients with shock due to “pump failure,” assessment is focused appropriately on the left ventricle. However, in a substantial minority of patients, right heart dysfunction is the cause of shock. Examples include acute pulmonary embolism (PE), other causes of acute right heart pressure overload (e.g., acute respiratory distress syndrome [ARDS] treated with positive pressure ventilation), acute deterioration in patients with chronic pulmonary hypertension, and right ventricular infarction. Although right ventricular infarction differs from the other right heart syndromes (RHS) in that the pulmonary artery pressure is not high, in many other regards right ventricular infarction resembles the other syndromes, so we will consider them together. Failure to consider the right heart in the differential diagnosis of shock risks incomplete or inappropriate treatment of the shock. It would be hard to overemphasize the importance of echocardiography, both in aiding the recognition of the right heart syndromes and in guiding management. In this chapter we review the notable features that distinguish the right heart from the left, describe the themes that unify the acute RHS and allow their recognition, discuss the pathophysiology and differential diagnosis of RHS, and review their management.

The right ventricle (RV) has long been considered the “forgotten ventricle,” because under normal pressure and volume loading conditions the RV is thought to function as a passive conduit for systemic venous return. When the pulmonary vasculature is normal, right ventricular performance has little impact on the maintenance of cardiac output. In animal models, complete ablation of the right ventricular free wall has little effect on venous pressures.

Despite the requirement for an equal, average cardiac output between the left and right ventricles, the bioenergetic requirement for RV ejection is approximately one fifth of the left ventricle (LV). This is in large part accounted for by the significant difference in downstream vascular resistance between the systemic and pulmonary circulations. In comparison with the LV, the RV ejects into a low-resistance circuit (normally only one tenth the resistance of the systemic arteries).

The pressure-volume relationship of the normal RV differs significantly from that of the LV. In contrast to the LV ejection, the RV ejects into the pulmonary outflow tract early during systole, continuing even after the maximal development of RV systolic pressure.1 This exaggerated “hang out” period (ventricular outflow between the onset of right ventricular pressure decline and pulmonary valve closure) optimizes pump efficiency and results in a triangular pressure-volume relationship compared with the square wave pump of the LV.

Under conditions of increased RV impedance (e.g., pulmonary stenosis or pulmonary embolism) the RV pressure-volume relationship assumes a square wave appearance similar to that of the LV.2 Unlike the LV, however, even modest acute increases in RV afterload may precipitate ventricular failure. This is not the case if volume and pressure loading develop more chronically. Significant contractile reserve is supported by RV myocyte hypertrophy and is regulated in part by increased expression of angiotensin II, insulin-like growth factor-I, and endothelin-1.3 Ventricular hypertrophy is not uniform and is frequently associated with regional diastolic and systolic dysfunction.4 Increased cardiac output is accommodated by recruitment of previously unperfused pulmonary vessels and by distention of vessels.

However, when pulmonary vascular resistance and pulmonary artery (PA) pressure rise, right ventricular systolic function deteriorates more readily than that of the left ventricle. Right ventricular ejection fraction falls as mean PA pressure rises and RV end-systolic and end-diastolic pressures rise. During acute PA hypertension, RV preload, afterload, and contractile state rise at the same time that heart rate rises. These features join to raise the RV myocardial oxygen consumption. At the same time, when an acute RHS is sufficiently severe to cause systemic hypotension, coronary perfusion of the RV may fall. The combination of rising oxygen demand and falling coronary oxygen supply subjects the RV to ischemia sufficient to reduce RV contractility and reduce systolic ejection against the increased PA pressure afterload (Fig. 26-1). The close anatomic approximation between the right and left ventricles confers a mechanical and functional interdependence in the face of right ventricular dysfunction.

Interdependence is influenced by (1) the cardiac fibroskeleton that limits acute annular distension, (2) the interventricular septum, and (3) the pericardium. As right heart pressures rise, the interventricular septum shifts progressively to the left, causing left ventricular diastolic dysfunction, further reducing systemic cardiac output and coronary perfusion pressure. Additionally, the pericardium restricts excessive acute ventricular distension while impairing diastolic filling of both the left and right heart. A vicious cycle ensues in which RV ischemia impairs right ventricular ejection, which leads to progressive dilation of the RV and septal displacement that causes more LV diastolic dysfunction, progressive systemic hypotension, and further impairment of RV perfusion.

This cycle has long been recognized in the acute inability of the RV to sustain a mean pulmonary artery pressure greater than about 40 mm Hg, based on studies of pulmonary hemodynamics in patients with acute PE without prior cardiopulmonary disease.1,5

There is significant evidence that RV ischemia underlies acute RV failure in settings of acute pulmonary hypertension. Indirect indications include the significantly increased load tolerance of the right ventricle when aortic pressure is raised,6 and a beneficial hemodynamic response to infusion of norepinephrine.7 These findings suggest, but do not establish, that greater coronary flow driven by the higher aortic pressure enhances RV function by relieving ischemia. More direct evidence comes from a biochemical analysis of the RV during PA constriction8 (Table 26-1). In this experiment, constriction of the pulmonary artery led to both hemodynamic failure and to biochemical evidence of ischemia. Moreover, the infusion of a vasoconstrictor reversed the hemodynamic deterioration and reversed the biochemical evidence of ischemia. Additional support for the role of RV ischemia comes from the occasional patient with electrocardiographic evidence of RV infarction or elevated myocardium-derived enzymes (MB fractions of serum creatine phosphokinase and troponins). Significant troponin elevation may be an early and reliable marker of right ventricular dysfunction in acute pulmonary embolism, and has been shown to predict an adverse outcome.9 Significant elevations of serum cardiac troponins T and I are thought to result from RV microinfarction.10

Recognizing the Right Heart Syndromes: Clinical Clues

In the hypoperfused patient, several clinical features should suggest the possibility of an acute right heart syndrome (Table 26-2). First, any history of pulmonary hypertension raises the possibility that the new shock state represents a (potentially minor) precipitant on top of preexisting right heart compromise (acute-on-chronic pulmonary hypertension; Table 26-3). When there is no antecedent history of pulmonary hypertension, elevated neck veins, a pulsatile liver, peripheral edema out of proportion to pulmonary edema, a right-sided third heart sound, or tricuspid regurgitation, these factors should alert the intensivist that she or he may be dealing with an RHS. The pulmonic component of the second heart sound may be loud, and the time interval between the aortic (A2) and the pulmonary (P2) components of the second heart sound (A2-P2 splitting) is increased in the presence of pulmonary hypertension. However, these findings are appreciable with a binaural stethoscope in only a minority of patients with acute pulmonary embolism,11 and are probably too subjective to be useful. More sophisticated acoustic processing of digitally acquired heart sounds may provide an accurate estimation of pulmonary arterial pressures.12

Despite the insensitivity of individual clinical signs to detect and diagnose acute right heart syndromes, a combination of clinical features (symptoms of deep venous thrombosis [DVT]; an alternative diagnosis is less likely than PE; heart rate >100 bpm; immobilization or surgery in the previous 4 weeks; previous DVT or PE; hemoptysis; and cancer, being treated currently or within the previous 6 months) and laboratory results, especially serum D-dimer level, can be useful in excluding pulmonary embolism as a likely cause.13

Electrocardiographic (ECG) evidence of pulmonary hypertension includes right axis deviation or a rightward shift in axis, right atrial enlargement, right ventricular hypertrophy, right bundle-branch block (RBBB), right precordial T-wave inversions, and the S1Q3T3 pattern. In the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) trial, fewer than 1 patient in 17 with proven pulmonary embolism had any of the patterns of acute right heart strain.11 By contrast, the experience at a single Italian center of 160 cases of proven PE resulted in a higher rate of electrocardiographic abnormalities. Of all patients, 76% had at least one abnormality; RBBB in 47%, S1Q3T3 in 37%, T-wave inversion in 32%, and an inferior distribution “pseudonecrosis pattern” in 11%.14 This variation in detection between studies suggests relative insensitivity in the performance characteristics of the 12-lead ECG in broad groups of mixed severity right heart syndrome patients. However, in patients with hemodynamically significant pulmonary embolism, the likelihood of suggestive electrocardiographic findings is probably much higher. For example, among 49 patients with PE (all of whom had RV dilation and tricuspid regurgitation by echocardiography), 37 (76%) had electrocardiographic abnormalities strongly suggestive of PE, including at least three of the following: incomplete or complete RBBB; S waves greater than 1.5 mm in leads I and aVL; shift of the precordial transition zone to V5; Q waves in leads III and aVF, but not lead II; right axis deviation or an indeterminate axis; low QRS voltage in the limb leads; or T-wave inversion in leads III and aVR or in leads V1 to V4.15 The electrocardiographic signs of right ventricular infarction are described below in the section on right ventricular infarction.

Radiographic signs include an enlarged pulmonary artery or right ventricle, oligemia of a lobe or lung (Westermark's sign), and a distended azygos (or other central) vein (Figs. 26-2 and 26-3). Contrast-enhanced computed tomography of the pulmonary vasculature (helical CT angiography) has evolved as a central diagnostic tool in the evaluation of acute right heart syndromes, particularly pulmonary thromboembolism, as discussed in Chap. 27.16–19 The range from 53% to 89%, and specificities from 78% to 100% for single-slice helical CT diagnosis of acute PE.20 Newer multi-row detector scanners should increase sensitivity to more than 90%.21 In addition to detecting the presence of a pulmonary vascular clot, CT is able to detect RV dilatation and septal shift. In a small series of patients with acute PE, CT sensitivity was 78% for detecting RV dysfunction when compared with transthoracic echocardiography (TTE).22

Furthermore, TTE can be used to quantify pulmonary artery pressures and assess right ventricular function, thereby allowing for rapid initiation of appropriate therapeutic interventions. Echocardiography may also provide indirect evidence of pulmonary embolism by demonstrating a specific pattern of right ventricular dysfunction characterized by free-wall hypokinesis with apical sparing, a finding possibly useful in differentiating pulmonary embolism from other causes of right ventricular dysfunction.23–25 Finally, echocardiography can be used to visualize massive pulmonary embolism directly in many patients.26 Therefore we believe that echocardiography is a practical and readily available diagnostic tool that should be considered in the evaluation of patients with suspected pulmonary embolism.

The typical echocardiographic findings include a normally contracting left ventricle, often with end-systolic obliteration of the LV cavity; a thin-walled, dilated, poorly contracting RV; right atrial enlargement; tricuspid insufficiency with a high-velocity regurgitant jet; increased estimated PA pressures; leftward shift of the interventricular septum causing the typical “D” shape of the LV on the short-axis view (Fig. 26-4); right PA dilation; or loss of respirophasic variation in the inferior vena cava.27 RV infarction can usually be readily distinguished from acute pulmonary hypertension in that high PA pressures are lacking. Right ventricular diastolic dimensions can be obtained by measuring right ventricular end-diastolic area in the long axis, from an apical four-chamber view, or by a transesophageal approach in the volume-repleted patient.28

Echocardiography is of great utility in the detection of RHS and should be obtained early in the hypoperfused patient whenever one of the previously mentioned clinical indicators is present.24 Of course, most of these signs are not specific for RHS, but their recognition is important because the treatment of RHS is unique in several regards (Table 26-4). TTE is particularly useful in differentiating at the bedside right ventricular pressure overload from myocardial infarction, aortic dissection, or pericardial tamponade, all of which may be clinically indistinguishable from PE.29 Identification of a patent foramen ovale and free-floating right-heart thrombus are echocardiographic markers of particularly grave prognosis, including recurrent PE and death.26,29,30

Pulmonary artery catheterization can estimate pulmonary arterial pressures more accurately than echocardiography. However, interpretation of mean pulmonary pressures and measurement of tricuspid regurgitation by thermodilution are confounded by technical limitations. A pulmonary artery catheter (PAC) with a fast-response thermistor has been advocated for accurate measurement of right ventricular volume and hemodynamic parameters by thermodilution in the presence of tricuspid regurgitation. However, as is the case for PAC use in the management of ARDS and left heart shock, the fast-response thermistor PAC has not been demonstrated to confer an improvement in survival.

Severity of RV systolic failure is an independent determinant of serum levels of brain natriuretic peptide (BNP) in patients with severe heart failure.31 Furthermore, BNP elevations have been demonstrated to predict RV dysfunction in patients with acute PE. In one study, the relative risk ratio for RV dysfunction was 28.4 (95% CI, 3.22 to 251.12) if the serum BNP >90 pg/mL.32 A lower cut-off of 50 pg/mL might improve the specificity of BNP as a predictor of favorable outcome, but it remains unclear whether measurement of BNP has any role in these patients.10,33

Acute Pulmonary Hypertension

Acute pulmonary hypertension is caused by an abrupt increase in pulmonary vascular resistance due to vascular obstruction or surgical resection. The prototype of acute pulmonary hypertension is acute pulmonary embolism (PE; see Chap. 27), but other forms of embolism (e.g., air or fat), microvascular injury (e.g., ARDS), drug effect, and inflammation can acutely raise pulmonary vascular resistance (see Table 26-3). For example, Zapol and Snider described the pulmonary hemodynamics in 30 patients with severe ARDS (20% survival; 8 had received extracorporeal membrane oxygenation).34 Following correction of hypoxemia, the mean pulmonary artery pressure was about 30 mm Hg and was abnormally elevated in all patients. Similar levels of pulmonary artery pressure were seen in a group of postsurgical patients with ARDS who were treated with nitric oxide (NO).35 Mean pulmonary artery pressure fell from 33 ± 2 mm Hg to 28 ± 1 mm Hg when NO was inhaled at 18 ppm. Contributors to PA hypertension in ARDS include hypoxic pulmonary vasoconstriction, mediator release, high alveolar pressure during mechanical ventilation, and microthrombi.

The frequency of significant pulmonary hypertension in ARDS has not been clearly defined. In the patients studied by Zapol and Snider, right heart dysfunction was clinically significant, even when the change in PA pressures and pulmonary vascular resistance was small.34 The large prospective European ARDS Collaborative Study evaluated pulmonary hemodynamic variables in 424 of 586 ARDS patients.36 In most patients, PA pressure was modestly elevated on admission (26.1 ± 8.5 mm Hg) and was persistently elevated at 48 hours in nonsurvivors compared with survivors (28.4 ± 8.5 mm Hg versus 24.1 ± 6.7 mm Hg). The ratio of RV to LV stroke work was also significantly elevated in all patients, and along with the ratio of partial oxygen pressure to the fraction of inspired oxygen (OR 0.96 to 0.98), was identified as an independent predictor of survival (OR 20 to 85; p = 0.0001).

These findings would suggest an aggressive approach to lowering RV afterload in patients with ARDS by reducing alveolar pressures and administering inhaled nitric oxide or prostacyclin. However, despite reproducible reductions in PA pressure and improvements in oxygenation indices, randomized controlled studies using this approach have repeatedly failed to demonstrate a survival benefit, as discussed below.

Sepsis itself is probably capable of causing pulmonary hypertension, even in the absence of acute lung injury, based on animal models37 and limited human studies.38,39 Although common in patients with severe sepsis, it is our experience that acute pulmonary hypertension is only of clinical importance when ARDS (or another clear precipitant) is present. It seems likely that the systemic hypotension of septic shock makes the RV more vulnerable to ischemic systolic dysfunction when combined with modest increases in afterload.40 It has been argued that this right ventricular perfusion gradient accounts for the differentially impaired perfusion and contractility of the RV compared with the LV in sepsis.39 The sepsis-associated proinflammatory cytokines tumor necrosis factor-α and interleukin-1β have been demonstrated to have negative inotropic effects on the ventricular myocardium.

A notable insight into the complex role of endogenous nitric oxide in regulating pulmonary vascular tone in septic shock patients was derived from a randomized, placebo-controlled, double-blind study of the nitric oxide synthase 546C88.41 Patients who were randomized to the treatment arm had a 10% absolute higher mortality rate at 28 days than patients in the placebo arm. 546C88-treated patients had a greater incidence of pulmonary hypertension, with an initial increase in the pulmonary vascular resistance and a sustained reduction in the pulmonary venous admixture, possibly through augmented hypoxic pulmonary vasoconstriction. Three patients in the treatment arm developed right heart failure. It has been suggested that sepsis-associated NO production may have a partially protective effect on the pulmonary vasculature by optimizing pulmonary ventilation-perfusion relationships.

Acute pulmonary hypertension in sickle chest syndrome results from pulmonary microvascular in-situ thrombosis, pulmonary fat embolism from infarcted long bone marrow, and hypoxic vasoconstriction. Recurrent episodes result in secondary chronic pulmonary hypertension and cor pulmonale. Inhaled NO, in addition to supplemental oxygen, blood transfusions, and bronchodilators, may provide some additional benefit for generalized vasoocclusive crises,42 but has not been systematically studied for sickle-associated pulmonary hypertension.43

Acute right heart failure following cardiac surgery, especially in patients operated on for severe mitral valve disease, some congenital cardiac defects, acute pulmonary embolism, or following heart transplantation or institution of left ventricular mechanical assistance, continues to vex cardiac surgeons and surgical intensivists. The mechanisms underlying this are multifactorial44 and include cardiopulmonary bypass–induced activation of pulmonary inflammatory pathways,45 and impairment of nitric oxide by pulmonary endothelial cells. A favorable response to inhaled NO has been demonstrated when used postoperatively46,47 or perioperatively.48 More recently, inhaled prostacyclin has been shown to improve PA hypertension and RV dysfunction.49

In about 1.5% of CPB patients, reversal of heparin anticoagulation with protamine is accompanied by transient, at times intense, pulmonary vasoconstriction.50 This phenomenon is thought to result from thromboxane B2 generation.50 Effective prophylaxis for this syndrome has not yet been reported. Inhibition of poly(ADP-ribose) polymerase, a terminal effector of oxidative stress injury, may offer future therapeutic opportunities for this syndrome.51

Acute‐on-Chronic Pulmonary Hypertension

Many patients with acute RHS have preexisting pulmonary vascular disease, at times with clinically recognized pulmonary hypertension, but often without (see Table 26-3). In such patients, intercurrent critical illness may unmask pulmonary vascular disease when a higher-than-normal cardiac output is needed. For example, in 12 men with moderate to severe but stable chronic obstructive pulmonary disease (COPD), mean PA pressure was normal at rest (17 ± 6 mm Hg), and the systolic PA pressure ranged from 21 to 27 mm Hg.52 During exercise (25 to 50 Watts), mean PA pressure rose significantly to 31 ± 11 mm Hg and systolic PA pressure to 20 to 55 mm Hg. At the same time, right ventricular end-diastolic volume increased, and right ventricular ejection fraction failed to rise.

When pulmonary hypertension is diagnosed during the course of critical illness, the potential for underlying chronic pulmonary vascular disease should be considered, especially when the history suggests chronic disease, the mean PA pressure is higher than 40 mm Hg, or echocardiography shows evidence of RV hypertrophy.

Right Ventricular Infarction

Right ventricular infarction is a well-recognized and fatal feature of inferior myocardial infarction.53,54 It is also seen in anterior infarcts. Meta-analysis of six studies involving 1198 patients with RV myocardial involvement suggested an increased probability of death (OR 3.2; 95% CI 2.4 to 4.1) compared with non RV MI.55 In most cases RV free wall infarction or ischemia is accompanied by varying degrees of septal and posteroinferior left ventricular injury, but relatively isolated RV injury is occasionally seen. RV myocardial injury and dysfunction representing noninfarcted hibernating myocardium may be able to sustain long periods of low coronary oxygen delivery and ultimately recover substantial contractile function.56

RV dilation accompanies significant myocardial injury. Concomitant LV infarction involving the interventricular septum may lead to further hemodynamic deterioration in patients with RV infarction because of the loss of LV septal contraction which can assist RV ejection. Elevation of right atrial pressure on physical examination or direct measurement in a patient with an inferior myocardial infarction and clear lungs by exam and chest x-ray should lead to suspicion of RV infarction. When these features occur in a critically ill patient, the distinction is between RHS resulting from acute PA hypertension and RHS resulting from RV infarction. Confirmatory evidence includes a right precordial electrocardiogram or echocardiographic evidence of RV injury (see Chap. 25). The focus of management in RV infarction is on maintenance of optimal RV preloading to avoid worsened RV distension, preservation of RV synchrony, reduction in RV afterload (particularly when LV dysfunction is present), and inotropic and mechanical support of the RV.57

Early reperfusion with fibrinolytics or direct coronary intervention may have a role in many patients. Echocardiography can be highly useful in confirming RV infarction and in determining the response to therapeutic interventions.

Some patients with acute RHS may benefit from specific therapies, such as thrombolysis for acute pulmonary embolism (see Chap. 27). In most patients, however, the two basic aims of treatment are supportive: to reduce systemic oxygen demand while improving oxygen delivery (Table 26-5). Oxygen demand can be lowered by treating fever, sedating the patient, instituting mechanical ventilation, and in severe cases, using therapeutic muscle relaxation. Oxygen delivery can be enhanced by correcting hypovolemia, transfusing red blood cells, relieving alveolar hypoxia, infusing vasoactive drugs, and avoiding detrimental ventilator settings. The goals of oxygen therapy in RHS are to enhance arterial saturation (SaO2) and to block alveolar hypoxic vasoconstriction (AHV). Using a sufficient oxygen concentration to achieve 88% SaO2 is advocated in ARDS and other alveolar flooding diseases (see Chap. 38), but in RHS not associated with intrapulmonary shunt, we target SaO2 to >96% to ensure alveolar oxygen values sufficient to block AHV (PaO2 >55 mm Hg). It may be useful to correct anemia with red blood cell transfusion, raising the arterial oxygen content, and reducing the necessary cardiac output. The resulting increased blood viscosity (and its tendency to raise pulmonary vascular resistance) probably does not outweigh the reduced demand for forward flow.

Fluid therapy, ventilator management, and vasoactive drug infusion are discussed below and have been the subject of a recent review.58

Fluid Therapy

In most patients with shock it is appropriate to administer fluid, often in massive quantities, to restore left ventricular diastolic filling and boost cardiac output. Despite the recognition that the right heart becomes extremely preload dependent during ischemia and infarction,56 excessive fluid administration is likely to worsen hemodynamic stability. In many of these patients the right-sided pressures are already well above normal, signaled by neck vein distention. Data from animal models of pulmonary embolism, as well as from studies of patients with right ventricular infarction, demonstrate that fluid therapy may be unhelpful or even detrimental.

In a canine autologous clot model of pulmonary embolism, the effects of fluid loading were studied before embolism, then following embolism.59 Before embolism, fluid loading significantly raised the right atrial pressure, the transmural left ventricular end-diastolic pressure (LVEDP), and the left ventricular end-diastolic area index (a measure of left ventricular volume using sonomicrometry). Following multiple emboli, fluid loading raised right atrial pressure, but transmural LVEDP fell significantly as did the left ventricular end-diastolic area index. These findings indicate that fluid loading following embolism causes further leftward displacement of the interventricular septum, further compounding LV diastolic dysfunction. In a canine glass bead embolization model, fluid loading was found to precipitate right ventricular failure, even when relatively small volumes were infused.60

Similar results have been shown in human right ventricular infarction.61,62 Despite raising the right atrial and wedge pressures, fluid loading failed to increase the cardiac index, blood pressure, or left and right ventricular stroke work. These findings should serve as a caution regarding fluid administration to patients with shock due to acute RHS. Since some patients may be volume depleted at presentation, a fluid challenge is reasonable, especially if the neck veins are flat or right heart filling pressures are low. Nevertheless, fluid should be given with a healthy degree of skepticism and careful attention to the consequences. We recommend that a discrete crystalloid fluid bolus of no more than 250 mL be administered while assessing relevant indicators of perfusion such as blood pressure, heart rate, pulsus paradoxus, cardiac output, central venous oxyhemoglobin saturation, or urine output. If no benefit can be detected, further fluids should not be given, and attention should shift to vasoactive drugs.

Vasoactive Drug Therapy

A wide variety of vasoactive drugs has been tried in patients or animal models for the treatment of acute RHS due to pulmonary embolism, ARDS, or right ventricular infarction, with variable success. These include nonspecific vasodilators (hydralazine 63 and nitroprusside 61,64,65), vasoconstrictors (norepinephrine,7,60,66  epinephrine,67  phenylephrine,8,68  dopamine,69 and vasopressin 70,71), inotropes (dobutamine,61,62,72–74 amrinone,75  milrinone,76  isoproterenol,7 epinephrine,67 and levosimendan 77), and pulmonary vasodilators (prostaglandin E1,78,79 prostaglandin I2,35 and nitric oxide 35,65,70,80–86). Predicting the response to any of these drugs a priori is complicated by their tendency toward opposing effects. Conflicting data from studies of an agent in different animal models suggest that the interspecies variation and prevailing pulmonary vascular tone are important in determining if a particular agent has a predominantly pulmonary vasodilatory or vasoconstricting effect.73,74 Thus the choice of vasoactive drugs cannot be based solely on the presumed pathophysiology, but also must be based on the results of human and animal studies summarized below. We contend that a vasoactive drug is effective in RHS when it significantly raises cardiac output without significantly worsening systemic hypotension, SaO2, or RV ischemia. Dobutamine is our preferred positive inotrope, inhaled NO (and perhaps aerosolized prostacyclin) have salutary short-term physiologic effects as pulmonary vasodilators, and norepinephrine may provide added benefit as a systemic vasoconstrictor and positive inotrope by raising coronary perfusion pressure to an ischemic RV.

Catecholamines

In massive pulmonary embolism, dobutamine and norepinephrine appear superior to other vasoactive drugs.7,72 In human PE, dobutamine has been most intensively studied. For example, of 10 patients with shock due to massive PE treated with dobutamine, 1 rapidly died, but 9 showed impressive hemodynamic improvement (Table 26-6). These results show that dobutamine improves cardiac output by improving right ventricular function or reducing pulmonary vascular resistance. Although fewer data are available regarding norepinephrine in human embolism, animal studies and limited human data support its use.7,6672 In a canine model of pulmonary embolism, dobutamine and dopamine had essentially identical hemodynamic effects.69 Data from a separate canine study suggest that at doses less than 10 μg/kg per minute, dobutamine-induced pulmonary circulatory changes are exclusively flow dependent.74 At higher doses, changes in pulmonary vascular resistance are variable and may depend on the prevailing pulmonary vascular tone. These drugs should be titrated according to clinical measures of the adequacy of perfusion, such as renal function, mentation, thermodilution cardiac output, or central venous oxyhemoglobin saturation, rather than to blood pressure alone. We begin dobutamine at 5 μg/kg per minute, raising the dose in increments of 5 μg/kg per minute every 10 minutes. If the patient fails to respond to dobutamine (or the response is incomplete), we substitute (or add) norepinephrine infused at 0.4 to 4 μg/kg per minute. In patients with hypoperfusion due to right ventricular infarction, dobutamine is superior to nitroprusside 61 (and to fluid infusion61,62), significantly improving right ventricular ejection fraction and cardiac output. Therefore dobutamine is the drug of first choice in all cases of RHS. We avoid the use of dopamine because of its highly variable pharmacokinetics and concern for disproportionate splanchnic vasoconstriction, even in relatively low doses.

Vasopressin

The role of vasopressin (and its longer acting congener, terlipressin) remains controversial and incompletely evaluated. Vasopressin clearly functions as a systemic vasoconstrictor at high doses. In patients with septic shock, replacement of acutely depleted endogenous vasopressin with a low-dose infusion (0.04 U/min) is thought to improve catecholamine sensitivity via the functionally vasoconstricting V1 receptor. The pulmonary vasculature has been shown by some investigators to express V1 receptors, but that vasopressinergic stimuli may paradoxically mediate pulmonary vasodilation.87,88 This might suggest a salutary potential for vasopressin therapy in acute right heart syndromes. In a canine model, however, vasopressin caused both systemic and pulmonary vasoconstriction while impairing RV contractility.71 Our present practice is to avoid vasopressin for acute right heart syndromes unless catecholamine-dependent septic shock is present.

Prostaglandins

Prostaglandin E1 (PGE1) is a potent pulmonary vasodilator that exhibited promise in the treatment of ARDS. When infused at a dose of 0.02 to 0.04 μg/kg per minute to patients with severe ARDS and mean PA pressure greater than 20 mm Hg, PA pressure fell 15% despite an increase in cardiac output. At the same time, however, systemic blood pressure fell to a similar degree, and intrapulmonary shunting rose significantly.78 In an oleic acid model of porcine ARDS, PGE1 lowered pulmonary artery pressure, but stroke volume and stroke work did not improve significantly.79 In patients with ARDS given prostacyclin (4 ng/kg per minute), pulmonary artery pressure fell, RV ejection fraction rose, and cardiac output increased significantly.35 A small series of patients with chronic pulmonary hypertension have been given aerosolized prostacyclin, and they demonstrated pulmonary vasodilation, increased cardiac output, and improved arterial oxyhemoglobin saturation.89 Systemic blood pressure fell somewhat, but to a much lesser degree than when prostacyclin was infused intravenously (for similar degrees of pulmonary vascular effect). When compared for acute hemodynamic effects in patients with primary pulmonary hypertension (PPH), aerosolized prostacyclin (approximately 14 ng/kg per minute over 15 minutes) was demonstrated to be a pharmacologically more potent acute vasodilator than inhaled NO (NO 40 ppm for 15 minutes).90

In a similar comparison in ARDS patients, gas exchange parameters were comparably improved when inhaled prostacyclin (7.5 ± 2.5 ng/kg per minute) was compared with inhaled NO at a dose lower than that in the PPH study (17.8 ± 2.7 ppm).91 This may suggest that in patients with right heart syndromes and long-standing pulmonary hypertension, inhaled prostacyclin may afford greater efficacy.

Although not conclusively demonstrated, inhaled prostacyclin has been used with some success in perioperative acute RHS.

Adenosine

Adenosine is an endogenous vasodilator that has a very short half-life (less than 10 seconds) due to rapid metabolism by adenosine deaminase. When used following cardiac surgery, adenosine lowered pulmonary artery pressure, raised cardiac output, and did not cause hypotension.46,92 Adenosine was infused centrally at a dose of 50 μg/kg per minute.

Phosphodiesterase Inhibitors: Amrinone, Milrinone, Dipyridamole, and Sildenafil

Amrinone is an inotrope and vasodilator with potential in the acute right heart syndromes. In a canine model of massive embolism, amrinone (0.75 mg/kg bolus followed by 7.5 μg/kg per minute) lowered pulmonary artery pressure, raised cardiac output, and raised systemic blood pressure.75 Limited data are available for the use of milrinone in acute RHS and its use is limited by a long half-life and limited ability of titration.76 Additionally, milrinone has been shown to be less efficacious than inhaled NO in treating pulmonary hypertension post–cardiac surgery.82 Another phosphodiesterase inhibitor, dipyridamole, has been evaluated as an adjunct to NO in pediatric patients with acute RHF, and shown to have some additional pulmonary vasodilatory effects.93,94

Significant interest has arisen in the therapeutic potential of the selective type 5 PDE inhibitor sildenafil, presently approved for male erectile dysfunction. Impressive acute reductions in pulmonary arterial pressures have been demonstrated with oral and intravenous administration in animal models of acute lung injury95 and RHS, in patients with established pulmonary hypertension,96 and in 93 patients with pulmonary hypertension complicating pulmonary fibrosis.97 Additionally, synergistic effects of selective PDE inhibitors in combination with inhaled and intravenous vasodilators has been demonstrated in acute lung injury–associated right heart syndromes.98–100

Nitric Oxide

Nitric oxide brings together the potential for hemodynamic as well as gas exchange improvement. When patients with ARDS and pulmonary hypertension were given NO via endotracheal inhalation at a dose of 18 ppm, PA pressure fell, right ventricular ejection fraction rose, and RV end-systolic and end-diastolic volumes fell.35 There was no detectable change in mean arterial pressure, and arterial oxygen pressure rose significantly. Increasing the dose of NO to 36 ppm had no incremental effect. These findings have been confirmed in similar patients with ARDS who were managed with permissive hypercapnia (mean arterial carbon dioxide pressure = 71 mm Hg) and given a lower dose of NO (5 ppm), although the effect was more modest.80 Disappointingly, survival has not been improved in four large randomized controlled studies of NO in ARDS patients.83–86

Antiproliferative Agents

Both prostacyclin and the newer nonselective endothelin receptor antagonists (ETRA) have been demonstrated to have antiproliferative activity on the pulmonary vasculature. This mechanism has been suggested to account for the modest functional improvement in patients with chronic pulmonary hypertension.101,102 Although single case reports suggest beneficial effects of the orally administered nonselective ETRA bosentan, this agent has not been subjected to rigorous evaluation in patients with acute right heart syndromes, and it may have limited potential in critically ill patients because of significant associated hepatic toxicity.

Ventilator Management

Ventilator manipulation has the potential to dramatically affect the circulation in patients with shock, including those with acute RHS. For example, in animal models of shock, institution of mechanical ventilation significantly prolongs survival, an effect much greater than that seen with fluid therapy or vasoactive drugs. Of particular interest in patients with RHS is the maintenance of oxygenation, the role of hypercapnia (including permissive hypercapnia), and the effects of tidal volume and positive end-expiratory pressure (PEEP).

Hypercapnia increases pulmonary artery pressure. In patients with ARDS, reducing minute ventilation as part of the strategy of permissive hypercapnia leads to small but real increases in mean pulmonary artery pressure.103–105 In most patients with ARDS who do not exhibit right heart limitation, this effect of hypercapnia is probably unimportant. However, in the subset of patients with severe pulmonary hypertension, permissive hypercapnia may lead to unacceptable hemodynamic deterioration.

The effect of PEEP on right ventricular function is complex, controversial, and highly variable from patient to patient.106,107 Many studies are limited by the failure to correlate hemodynamic pressures to juxtacardiac pressure. The effect of PEEP can be expected to differ depending on whether atelectatic or flooded lung is recruited, or whether relatively normal lung is overdistended. In a study of patients with ARDS, PEEP had little effect on RV function when given in amounts up to that associated with improving respiratory system compliance.106 At higher levels of PEEP, the dominant effect was to impair RV systolic function.

The dominant effect of mechanical ventilation is related to its effect on preload. Sustained airway pressure increases in volume-repleted patients with normal RV function result in a mild increase in right atrial pressure that is offset by increases in abdominal pressures that sustain venous return. However, it remains to be determined if this is true for patients with acute RHS and elevated right heart pressures.108 Large-tidal-volume breathing impairs RV systolic function, presumably by increasing pulmonary vascular resistance in alveolar vessels. In a canine model with normal lungs, raising the tidal volume above 10 mL/kg caused a detectable rightward and downward shift of the RV function curve.109

These effects of mechanical ventilation on right ventricular function suggest the following strategy in patients with critical compromise of the RV: (1) give sufficient oxygen to reverse any hypoxic vasoconstriction; (2) avoid hypercapnia; (3) keep PEEP at or below a level at which continued alveolar recruitment can be demonstrated and seek to minimize self-controlled PEEP (Auto-PEEP); and (4) use the lowest tidal volume necessary to effect adequate elimination of carbon dioxide. Of course, the acute effects of each intervention should be measured to confirm that cardiac output increases. These principles are consonant with the goals of ventilation in most patients with ARDS, except that when there is an RHS, hypercapnia should be avoided if it leads to further hemodynamic deterioration.

Mechanical Therapy

In contrast to the now well-defined role for mechanical assist devices in decompensated left heart failure,110 there remains relatively little experience with mechanical therapy for the failing right heart. Notably, progressive right ventricular dysfunction complicates left ventricular assist device implantation or orthotopic heart transplantation for decompensated left heart failure111,112 and is associated with progressive end-organ dysfunction.113 The presently available approaches include extracorporeal and paracorporeal pulsatile and centrifugal pump ventricular assist systems.75,76 An alternative approach uses a right atrial catheter to draw blood into a centrifugal pump and a percutaneously placed pulmonary artery catheter as the outflow cannula.114 Small implantable centrifugal pumps are under development.