Introduction
This chapter will focus on the pulmonary hypertension (PH) that results from chronic lung diseases. This is a common and significant clinical complication of lung disease, and it is a PH that differs from the other types both in causation and in therapeutic approach. The current (2013) “Updated Clinical Classification of Pulmonary Hypertension” ( Table 59-1 ) serves to define the terminology used in this chapter and in Chapter 58 . These groups are classified according to the underlying pathologic process that leads to PH. Note that the term pulmonary arterial hypertension (PAH) refers specifically to diseases in group 1, including idiopathic PAH. PH is used for the PH in groups 2, 3, 4, and 5. PH due to lung disease (PH-LD) is categorized in group 3. This is the PH that will be the topic of this chapter. Group 1 PH (PAH) is covered in Chapter 58 , and group 4 PH (chronic thromboembolic PH) is covered in Chapter 57 .
| GROUP 1: PULMONARY ARTERIAL HYPERTENSION (PAH) |
| GROUP 1′: PULMONARY VENO-OCCLUSIVE DISEASE (PVOD) AND/OR PULMONARY CAPILLARY HEMANGIOMATOSIS (PCH) |
| GROUP 1′′ PERSISTENT PULMONARY HYPERTENSION OF THE NEWBORN |
| GROUP 2: PULMONARY HYPERTENSION DUE TO LEFT HEART DISEASE |
| GROUP 3: PULMONARY HYPERTENSION DUE TO LUNG DISEASES AND/OR HYPOXIA |
| COPD |
| Interstitial lung disease |
| Other pulmonary diseases with mixed restrictive and obstructive pattern |
| Sleep-disordered breathing |
| Alveolar hypoventilation disorders |
| Chronic exposure to high altitude |
| Developmental lung diseases |
| GROUP 4: CHRONIC THROMBOEMBOLIC PULMONARY HYPERTENSION (CTEPH) |
| GROUP 5: PULMONARY HYPERTENSION WITH UNCLEAR MULTIFACTORIAL MECHANISMS |
* See Table 58-1 for complete classification.
In previous editions the title of this chapter was “Cor Pulmonale.” Historically, there has been no consensus about the definition of the term cor pulmonale . Today the term cor pulmonale generally refers to abnormalities of right heart structure and function that develop in the setting of lung disease and/or hypoxemia, including parenchymal lung disease, ventilatory impairment, or high-altitude hypoxemia. Cor pulmonale develops secondary to pulmonary hypertension, characterized by elevations in pulmonary vascular resistance (PVR) and in pulmonary artery pressure (P pa ), which causes increased right ventricular afterload and, in susceptible patients, eventually progresses to right ventricular failure.
The development of PH and subsequent cor pulmonale in patients with lung disease is clinically important because it is common and is associated with increased morbidity and mortality. In one large community-based series, PH-LD was the second most common cause of PH after left heart disease ( Fig. 59-1 ). Available data show that nearly all types of advanced lung disease can be complicated by PH and eventually progress to right heart failure. However, patients with chronic obstructive pulmonary disease (COPD) and idiopathic pulmonary fibrosis (IPF) will be the focus of this chapter. As in all patients with PH, a thorough and exhaustive diagnostic evaluation is critical to determine the primary cause of PH and to identify concomitant conditions that might worsen PH and right heart failure. The different groups of PH affect different areas of the pulmonary circulation ( Fig. 59-2 ); nonetheless, even those that affect the pulmonary arterioles have a different underlying pathologic process and response to treatment. Distinguishing between group 3 PH-LD and group 1 PAH is especially important because these diseases are pathologically and clinically distinct and respond differently to treatment. The best treatment of patients with PH-LD and subsequent cor pulmonale continues to be optimal management of the underlying lung disease, correction of hypoxemia, and timely consideration for lung transplantation.
Epidemiology of Pulmonary Hypertension Due to Lung Disease
Prevalence
The prevalence of PH among patients with known lung disease has been assessed in numerous studies. Available studies suffer from significant limitations, including inconsistent definitions of PH and frequent use of echocardiography rather than right heart catheterization to assess hemodynamics. Additionally, patient cohorts are often heterogeneous, and comprehensive assessments to exclude other important and common causes of PH (e.g., chronic thromboembolic PH, and left heart disease) generally have not been performed.
Chronic Obstructive Pulmonary Disease
The lung disease most frequently associated with PH is COPD (PH-COPD), which is by far the most common cause of cor pulmonale, accounting for more than 80% of all cases. However, the estimated prevalence of PH in patients with known COPD varies dramatically—from 2.7% to 90.8% —depending on the definition of PH and the study population.
The standard definition of PH in this population, as in other populations, is a mean P pa ( 
) greater than 25 mm Hg, although some studies have used different cutoff values (e.g., 20 mm Hg). The gold standard for this measurement is the right heart catheterization. A more convenient but less accurate measurement is by echocardiography which requires a tricuspid regurgitant jet to measure the pressure gradient across the valve and thus estimate pulmonary artery systolic pressure (PASP). Echocardiographic estimates of PASP are possible only in those patients with a measurable tricuspid regurgitant jet and are known to overestimate and underestimate true PASP.
An early study of right heart catheterization in 175 patients with severe COPD showed that 35.4% of patients had a 
greater than or equal to 20 mm Hg and 9.7% had a 
greater than 30 mm Hg. In another study, invasive hemodynamic values were assessed in 120 patients with severe airflow obstruction (mean FEV 1 27% of predicted) at the time of enrollment in the National Emphysema Treatment Trial. In this cohort, 90.8% of patients had a 
greater than 20 mm Hg, but only 5% had a 
greater than 35 mm Hg. Importantly, pulmonary capillary wedge pressure (P pw ) was greater than 12 mm Hg in 61.4% of patients and greater than 20 mm Hg in 6.4% of patients, suggesting that left-sided heart failure was an important contributor to PH in this population. To assess the prevalence of PH in an outpatient population with stable COPD, echocardiography was performed in 159 patients to estimate PASP. Tricuspid regurgitation was adequate to estimate PASP in 105 patients, and, of those, 60% had a PASP greater than or equal to 35 mm Hg and were classified as having PH. Patients with PASP greater than or equal to 35 mm Hg were older and had lower FEV 1 and D l CO values.
In one well-designed study, 998 patients underwent right heart catheterization at a referral center in France as part of evaluation for chronic respiratory failure; severe PH, defined as 
greater than or equal to 40 mm Hg, was diagnosed in 27 patients (2.7%) ( Fig. 59-3 ). Another cause of PH was found in 16 patients, resulting in a final diagnosis of PH secondary to COPD in only 11 patients (1.1%). Compared to other patients in the cohort, the patients with severe PH had lower D l CO and arterial P co 2 and P o 2 values. Patients with severe PH also had lower cardiac indices and higher right atrial pressures, suggesting reduced right ventricular function.
These studies demonstrate that mild PH is common among stable outpatients with COPD. Importantly, more severe hemodynamic abnormalities were observed only in a very small number of patients, and, among those patients, other causes of PH were common.
Examining cohorts of patients undergoing evaluation for lung transplantation is useful because they include a relatively homogenous patient population with complete hemodynamic data; however, they represent a cohort with particularly advanced lung disease in whom PH might be expected to be more frequent and more severe.
A recent retrospective study used right heart catheterization data from the Organ Procurement and Transplantation Network database to study PH in 4930 patients with COPD listed for lung transplantation. The prevalence of mild and moderate PH defined as a 
greater than or equal to 25 mm Hg and less than 35 mm Hg with a P pw less than or equal to 15 mm Hg was 30.4%, and the prevalence of severe PH defined as a 
greater than or equal to 35 mm Hg with a wedge pressure less than or equal to 15 mm Hg was 4.0%. A significant number of patients, 17.2%, had a 
greater than or equal to 25 mm Hg but also had a P pw greater than 15 mm Hg. Moreover, the P pw was greater than 15 mm Hg in about 50% of the patients with 
greater than or equal to 31 mm Hg, demonstrating that elevated left heart filling pressures are common in this population and may contribute significantly to PH. Findings were similar in another study of 409 patients with COPD undergoing evaluation for lung transplantation in Denmark. In this study 36% of patients had PH with a 
greater than or equal to 25 mm Hg, P pw less than or equal to 15 mm Hg, and 13% had a 
greater than or equal to 25 mm Hg but also had a P pw greater than 15 mm Hg. Only 6 (1.5%) patients had a 
greater than or equal to 40 mm Hg. In this population, PH was associated with the presence of more severe hypoxemia and lower FEV 1 values. Similar to the previously discussed studies, patients being evaluated for transplant frequently have elevations in 
; however, there is a low prevalence of severe PH, and a significant proportion of patients have elevated P pw , which suggests a contribution from left-sided heart failure.
There are limited longitudinal data concerning PH-COPD; however, the progression of PH in these patients appears to be slow. Among patients who are found to have normal P pa at rest, changes in pulmonary hemodynamics are frequently minimal over time. For example, in a group of 61 patients without initial PH, all of whom had arterial hypoxemia, a second right heart catheterization almost 8 years later revealed an average change in 
from 15.5 ± 2.4 to 19.6 ± 7.0 mm Hg. In a second group of 32 patients who had PH on their first catheterization, there was a nonsignificant rise in 
, from 27.7 ± 6.0 to 31.0 ± 9.3 mm Hg, after 5 years. An increase of 5 mm Hg or more was seen in approximately one third of patients and was clearly related to worsening hypoxemia.
In patients with COPD the presence of PH is clinically important because it is associated with worse exercise tolerance and survival compared to COPD patients without PH. In the retrospective organ procurement database study, 6-minute walk distance (6MWD) was on average 28 m less in patients with PH compared to those with normal hemodynamics; in addition, 
was an independent predictor of a low 6MWD in a multivariate model. In this study, adjusted risk for death on the transplant list was significantly increased in patients with PH (hazard ratio 1.27). Findings were similar in a study of 362 patients at a single center undergoing transplant evaluation. In multivariate analysis, higher 
was associated with a shorter 6MWD, including a reduction of 11 m for every 5 mm Hg increase in 
. In the Danish cohort, PH did not affect 6MWD but did affect the survival rate; the survival rate at 5 years was 37% for patients with PH compared to 63% in patients without PH ( P = 0.016). In this cohort the presence of PH did not affect outcomes following lung transplantation. Among the 11 patients with severe PH in the French COPD cohort, exertional dyspnea was significantly worse and survival was significantly shorter compared to patients without severe PH. In another study, the effects of PH on survival were assessed in a cohort of 84 patients who underwent hemodynamic evaluation before institution of long-term oxygen therapy. Adjusted 5-year survival rate was 62% for patients with an initial 
of 25 mm Hg or less, whereas it was only 36% in the 40 remaining patients who had an initial 
higher than 25 mm Hg. In this cohort, initial 
was a better prognostic indicator than either the FEV 1 , the degree of hypoxemia, or the level of hypercapnia. The results of a study of 101 patients with PH and COPD at a PH referral center came to similar conclusions. Survival at 3 years was 33% in patients with 
greater than or equal to 40 mm Hg versus 55% in patients with 
of 25 to 39 mm Hg. In a multivariate analysis, age, D l CO , mixed venous oxygen saturation, and World Health Organization functional classification were independent predictors of survival. These findings suggest that even mild PH may have a significant negative impact on exercise tolerance and survival and that the presence of PH may be a more important prognostic factor than the severity of lung disease.
Idiopathic Pulmonary Fibrosis
Most of the available data regarding PH associated with IPF (PH-IPF) comes from patients undergoing evaluation for lung transplantation. In a study of 79 patients with IPF referred for lung transplantation who had a right heart catheterization, 32% (25 of 79) had PH documented by a 
greater than 25 mm Hg. In another cohort of 124 patients undergoing evaluation for lung transplant at the Cleveland Clinic, 44% (54 of 124) had a 
greater than or equal to 25 mm Hg. A study of 101 Japanese patients with IPF who underwent right heart catheterization showed similar results with a prevalence of 
greater than 25 mm Hg of 15% (15 of 101). These studies show an overall high rate of PH-IPF in patients with advanced IPF.
The presence of PH-IPF is associated with worsening symptoms, functional impairment, and increased morbidity and mortality. In a retrospective study of 136 patients with IPF, the median survival of those with estimated PASP of greater than 50 mm Hg by echocardiography was less than 1 year (P = 0.009) compared to 4.8 years for patients without PH (PASP < 35 mm Hg), and 4.1 years for patients with mild PH (PASP 36 to 50 mm Hg). The 1-year survival rate in 79 patients undergoing lung transplant evaluation was lower (72%) in those patients with PH compared to IPF patients without PH (94.5%; P = 0.002). Furthermore, patients with PH who had lower D l CO values were more likely to require supplemental oxygen, to have shorter 6MWD, and to exhibit a lower arterial oxygen saturation nadir. In one study of 78 patients with PH-IPF, using a cutoff based on a receiver operating curve, the 5-year survival rate was 62% for the group ( n = 37) with 
values less than 17 mm Hg compared to 17% for the group (n = 24) with 
values greater than 17 mm Hg ( P < 0.001). The relative risk for death was significantly higher at 2.2 for the high- 
group. According to an analysis of 2972 patients with advanced IPF listed for lung transplantation between 1995 and 2004, those with PH were 1.6 times more likely to die after being listed for transplantation compared to those without PH. These data suggest that, similar to patients with PH-COPD, the consequences of PH-IPF are serious and that even mild increases in P pa are associated with increased mortality.
The presence of pretransplant PH may even have a significant effect on posttransplant outcomes. In a cohort of 126 patients, those who developed primary graft dysfunction had high pretransplant 
values, and each 10 mm Hg increase in 
was associated with an increase in the odds of primary graft dysfunction by 1.64.
Other Lung Diseases
Patients with the syndrome of combined emphysema and pulmonary fibrosis seem to be at particularly high risk for developing PH, with grave consequences. Prevalence of PH in these patients has been described to be between 30% and 50%; this group had markedly reduced survival. In one retrospective series of 40 patients with combined emphysema and pulmonary fibrosis, elevated 
above 40 mm Hg was present in 48% of patients. Patients had severe symptoms, 85% with New York Heart Association functional class III or IV, and an average 6MWD of 244 m. Outcomes in this population were poor. For those with a PVR above the median value, one year survival was 48% while, for those with a PVR below the median, one year survival was 100%.
In patients without parenchymal lung disease, hypoxemia from high altitude, obstructive sleep apnea (OSA), and obesity hypoventilation syndrome are important independent causes of PH by themselves, or they may contribute significantly to PH from other causes. Because of the large number of individuals who live and visit at high altitudes, environmental hypoxemia related to altitude may be one of the most common causes of PH worldwide. OSA appears to be an increasingly common cause of PH. In multiple studies of patients with OSA, the prevalence of PH ranges from 20% to 40% with mild elevation in 
(23-32 mm Hg). Interestingly, this PH may be reversible; a well-designed prospective trial showed that treatment with continuous positive airway pressure over 6 months reduced P pa in patients with OSA. Obesity hypoventilation syndrome, which is increasingly common worldwide, also causes PH. The prevalence of 
greater than 20 mm Hg among 29 patients with obesity hypoventilation syndrome was 59% in a cohort of untreated patients. In another study of 21 patients with obesity hypoventilation syndrome treated with noninvasive positive pressure ventilation, the prevalence of 
greater than 20 mm Hg was 81%, although only 3 patients (14%) had severe PH with 
greater than 35 mm Hg. In this study, 
correlated negatively with use of noninvasive positive pressure ventilation, and, by multivariate regression, the presence of PH was an independent predictor of poor physical functioning.
Case reports and small series describing PH as a complication of a variety of other lung diseases have been published. These conditions will not be specifically discussed here but may include sarcoidosis, Langerhans cell histiocytosis, lymphangioleiomyomatosis, adult bronchopulmonary dysplasia, and cystic fibrosis. Limited published data suggest that development of PH in all of these conditions is associated with worse outcomes.
Pathologic Changes and Pathogenesis
Although the mechanisms underlying the development and progression of PH in patients with lung disease are incompletely understood, they appear to be multifactorial and to vary with the underlying lung disease. It is likely that early vascular endothelial injury is caused by factors such as hypoxemia and inflammation leading to endothelial dysfunction and subsequently the development of structural vascular changes. These abnormalities lead to an increase in PVR and a subsequent increase in P pa .
Pulmonary Vascular Remodeling
Many of the individual elements of pulmonary vascular remodeling seen in PH-LD are similar to those seen in PAH. The primary differentiating pathologic feature is that patients with PH-LD generally do not have the plexiform lesions that are seen in patients with PAH. Multiple mechanisms most likely contribute to the development of the vascular pathologic features observed in PH-LD, which, once established, contribute to elevated P pa .
There are a variety of pathologic vascular changes described in PH-LD that vary with the type of lung disease and overlap significantly with the changes seen in PAH ( Fig. 59-4 ). Knowledge about pathologic changes in PH-LD comes from autopsy series and studies of explanted lungs at the time of transplant. A characteristic feature often seen in PH-LD that is also seen in PAH is the extension of smooth muscle into small pulmonary arterioles less than 80 µm, where it is not found in healthy lungs. This is termed “muscularization” and is characterized by circularly oriented smooth muscle cells between the two layers of elastic lamina. Muscularization may result from hypertrophy and proliferation of existing smooth muscle and from development of new smooth muscle cells. Another common finding in lungs from patients with PH-LD that overlaps with PAH is proliferation of the intimal layers of the pulmonary arterioles. Inflammation and thrombosis in situ may also be seen.
Pathologic changes that are unique to PH-LD are those that are associated with the underlying lung disease. In patients with PH-COPD there is destruction of alveoli and concomitant destruction of the associated pulmonary vasculature, which causes a further reduction in the pulmonary vascular cross-sectional area, leading to an increase in PVR. A unique finding in patients with PH-IPF is that vascular changes are seen both in areas of lung fibrosis and in areas of normal lung, though the vascular changes in areas of normal lung are less severe and found in a smaller proportion of vessels. Also, occlusive pathologic changes of pulmonary veins have been observed much more frequently in PH-IPF than in PAH. Ultimately, in patients with IPF, large portions of the pulmonary vascular bed may be destroyed or obliterated from progressive parenchymal fibrosis, inflammation, perivascular fibrosis, and/or thrombotic angiopathy.
The severity and extent of structural changes correlate variably with the degree of hemodynamic abnormalities seen in PH-LD. A recent study compared the structural changes observed in pulmonary arterioles from patients with COPD with and without PH undergoing transplant. Even without PH, patients with COPD frequently had muscularization and medial thickening of the pulmonary arterioles. Severity of pathologic changes worsened as the hemodynamic abnormalities worsened. In another study of tissue from COPD patients who died during the National Institutes of Health Nocturnal Oxygen Therapy Trial, there was no correlation between the histologic appearance of pulmonary arterioles and PH severity or change in P pa .
Several studies suggest that vascular remodeling in patients with mild COPD without concomitant hypoxemia is a component of early-stage disease as evidenced by increased inflammatory cell density in the adventitia and increased thickness of the intima of pulmonary arteries. Furthermore, endothelial lesions have been shown to be present in cigarette smokers without chronic airflow obstruction. These findings suggest that, although the pathologic changes in the pulmonary vasculature predominate, other abnormalities appear to contribute to the development of clinically relevant PH.
Pathogenesis
The pathogenesis of PH-LD may have multiple mechanisms, including hypoxic pulmonary vasoconstriction, vasoconstrictive neurohormones, and inflammation.
Hypoxic Pulmonary Vasoconstriction
Among all the mechanisms leading to PH in lung disease, the most potent and most important is alveolar hypoxia. Hypoxic pulmonary vasoconstriction is a normal physiologic response to alveolar hypoxia that was first demonstrated in isolated cat lungs and subsequently confirmed in healthy human volunteers. This effect is unique to the pulmonary circulation. Whereas acute hypoxia in the systemic circulation induces vasodilation, acute hypoxia in the pulmonary circulation causes constriction. The constriction of precapillary pulmonary arterioles effectively shunts blood away from poorly ventilated lung units to preserve optimal ventilation-perfusion matching.
Hypoxic pulmonary vasoconstriction results from hypoxia-mediated inhibition of voltage-gated potassium channels in pulmonary artery smooth muscle cells. Hypoxia inhibits the outward flow of potassium through these channels, resulting in depolarization of the membrane and entry of calcium, which causes smooth muscle cell contraction and sustained vasoconstriction. When activated, this mechanism results in smooth muscle contraction within seconds of exposure to hypoxia. Smooth muscle contraction is augmented by the activity of RhoA and Rho-associated kinase. Hypoxia signals through the G protein RhoA stimulating Rho-associated kinase to increase phosphorylation of smooth muscle cell myosin light chain and to augment smooth muscle cell contraction, regardless of the level of intracellular calcium.
Even brief periods of hypoxia may result in persistent hemodynamic abnormalities. In six healthy volunteers exposed to 8 hours of hypoxia, there was an increase in PVR, which did not return to normal after 2 hours of normoxia.
Long-term hypoxia–induced structural changes in the pulmonary arterioles were demonstrated in an invasive hemodynamic study of 11 high-altitude natives, which showed that hemodynamics did not normalize even after 2 years at low altitude. Structural changes in the pulmonary vasculature similar to those observed in COPD and IPF patients are present in the lungs of animals and humans exposed to hypoxia. In an autopsy series of tissues from Andean subjects born and living at high altitude who died without cardiac or pulmonary disease, medial hypertrophy of the pulmonary arteries was seen.
The degree of increase in P pa secondary to hypoxic pulmonary vasoconstriction varies by species. Pigs, horses, and cows respond with brisk increases in P pa , whereas dogs, yaks, and llamas have a minimal response; humans and rodents have an intermediate response. In addition, the response in humans varies widely among individuals from absent to very intense, with a rise to a 
of 40 mm Hg in 1% to 2% of healthy individuals. One explanation for the observation that some patients with lung disease develop PH whereas others do not is that there are heritable differences in ventilatory sensitivity either to hypoxia and carbon dioxide or in pulmonary vascular reactivity, or both.
Although hypoxia may be a significant contributing factor in the development of PH in chronic lung disease, hypoxia is not the sole factor in generating PH. This conclusion is supported by the finding that oxygen therapy in patients with lung disease has an inconsistent effect and does not normalize P pa . In addition, hypoxia is not necessary to produce vascular changes; pulmonary vascular structural changes have been observed in patients with mild COPD who do not have hypoxemia, and structural changes seen in the pulmonary vasculature in patients with IPF are more extensive than can be explained by hypoxia alone.
The development of hypercapnia may be one reason why patients with lung disease develop more PH than individuals exposed to environmental hypoxia, such as those living at high altitudes. In patients with PH-COPD, hypoxia is often accompanied by hypoventilation and hypercapnia, causing acidosis, which worsens hypoxic pulmonary vasoconstriction and PVR. This is in contrast to high-altitude dwellers such as healthy Andeans who have chronic hypoxia but generally have hyperventilation with resultant hypocapnia. Despite the presence of hypoxia, the high-altitude dwellers generally do not develop PH, suggesting that hypercapnia and acidemia are important contributing elements.
Neurohormones
Neurohormonal abnormalities and increased sympathetic nervous system activation are observed in patients with PH-LD and cor pulmonale and contribute to development of hemodynamic abnormalities. Plasma catecholamine levels generally rise in patients with decompensated right heart disease in a manner comparable to that seen in patients with heart failure secondary to primary myocardial disorders. It is clear that increased sympathetic activity is present with decompensation from cor pulmonale, resulting in high plasma levels of circulating catecholamines and stimulation of the renin-angiotensin-aldosterone systems. In the measurements by Anand and associates, vasopressin actually rose to a higher level in patients with heart failure due to PH-COPD than was seen in comparable patients with heart failure not associated with COPD. Because cardiac output is often normal or may even be increased in decompensated cor pulmonale, it is likely that the decrease in systemic vascular resistance (SVR) resulting from both hypercapnia and hypoxia leads to a reflex increase in circulating catecholamines and other neural hormones. Central sympathetic stimulation may also be encouraged by the direct effect of the increased P co 2 on the central nervous system.
Angiotensin.
Angiotensin II is a potent vasoconstrictor of the pulmonary vascular bed. The pulmonary vasculature appears to be even more sensitive to the vasoconstrictor effects of angiotensin II than is the systemic vascular bed. This is noteworthy because increased circulating levels of angiotensin II and aldosterone are found in patients with cor pulmonale secondary to COPD who have hypoxia and hypercapnia. Activation of the renin-angiotensin-aldosterone systems, together with hypoxemia, is probably an underlying pathophysiologic mechanism responsible for the elevation in PVR that is observed in patients with decompensated cor pulmonale.
Angiotensin induces a dose-response increase in PVR in normal subjects. Angiotensin II inhibitors have been shown to lower PVR and to be beneficial in patients with cor pulmonale, especially in patients with IPF. Angiotensin II inhibitors have also been demonstrated to improve survival in patients with systolic heart failure caused by both coronary disease and idiopathic dilated cardiomyopathy. It is encouraging to note that the angiotensin-converting enzyme inhibitor lisinopril has been demonstrated experimentally to attenuate the pulmonary pressor response to hypoxic pulmonary vasoconstriction in healthy human volunteers. In addition, angiotensin II receptor blockade has been shown to produce a similar effect in hypoxic pulmonary vasoconstriction in humans. However, studies performed to date have not clarified a role for angiotensin-converting enzyme inhibitors or angiotensin II receptor antagonists in the treatment of cor pulmonale.
Endothelin.
Endothelin-1 (ET1) is a 21–amino acid peptide secreted by vascular endothelial cells in response to stimuli, including pulsatile stretch, sheer stress, neurohormones, cytokines, growth factors, and thrombin. The secretion of ET1 has been shown to be increased by hypoxemia in humans. The effects of ET1 are mediated by both endothelin A (ET A ) and endothelin B (ET B ) receptors. ET A is localized on vascular smooth muscle cells, and ET B is expressed on vascular smooth muscle cells, endothelial cells, and fibroblasts. The effects of ET1 include vasoconstriction, hyperplasia, hypertrophy, fibrosis, and increased vascular permeability. Activation of ET B receptors on endothelial cells mediates release of prostacyclin (prostaglandin I 2 ) and nitric oxide (NO), which exert vasodilatory and antiproliferative effects while also inhibiting ET1 production by endothelial cells. Furthermore, the pulmonary endothelial ET B receptors are responsible for the pulmonary clearance of up to 50% of circulating ET1. Endothelin-1 is a potent vasoconstrictor and mitogen, and NO is a pulmonary vasodilator and inhibitor of fibrosis.
Increased ET1 production has been described in patients with PAH, and both increased ET1 levels and increased expression of receptors have been shown to be present in plexiform lesions of the lung in patients with PAH. Furthermore, the high plasma levels of ET1 correlate with disease severity and adverse prognosis. ET1 concentration is also elevated in the sputum and urine of patients with COPD compared to normal subjects; moreover, urinary levels increase further during COPD exacerbations. In addition, plasma ET1 levels have been shown to be increased in subjects who exhibit worsening oxygen saturation with exercise or at night. In another study, patients with PH-COPD were shown to have elevated transpulmonary ET1 levels. Increased plasma levels and lung tissue expression of ET1 have also been identified in patients with IPF with or without PH.
Inflammation
Inflammation has been hypothesized to play a significant role in the development of the vascular changes seen in PH-LD because increased markers of inflammation and inflammatory mediators have been observed in patients with PH-LD. However, the exact role of inflammation in PH-LD remains controversial. Increased numbers of CD8 + T lymphocytes have been observed in the walls of pulmonary vessels of patients with COPD. Furthermore, the presence of these cells was correlated to the enlargement of the intimal layer. Similar findings have been observed in the lungs of smokers without COPD or PH, suggesting that cigarette smoke might induce these inflammatory markers. In another study, elevated serum levels of inflammatory cytokines, including C-reactive protein and tumor necrosis factor-α, were seen in patients with PH-COPD. In patients with PH-IPF, elevated levels of multiple inflammatory mediators have been seen, including thromboxane A 2 , tumor necrosis factor-α, platelet-derived growth factor, transforming growth factor-β, and fibroblast growth factor. Studies of gene expression suggest that inflammatory mediators are overexpressed in patients with PH-IPF. Further studies are needed to determine the causal nature of these changes and whether antagonizing inflammation could be of therapeutic value.
Right Ventricle
The healthy right ventricle (RV) is a thin-walled structure with a complex shape that appears crescentic when viewed in cross section and triangular when viewed from the side ( Fig. 59-5 ). Superficial RV muscle fibers are arranged circumferentially, and deep fibers are arranged longitudinally from apex to base so that the RV contracts by three separate mechanisms: (1) shortening along the long axis, drawing the base to the apex; (2) inward movement of the free wall, which produces a bellows effect with the RV squeezing against the thick wall of the left ventricle (LV); and (3) traction on the RV free wall at the point of attachment with the LV. Because the RV is in series with the LV, cardiac output from both chambers is equal. The RV, however, is coupled to the normally low-pressure and compliant pulmonary arterial tree so that RV stroke work is significantly less and the RV functions as a volume pump as opposed to a pressure pump like the LV. Under normal physiologic conditions, the RV handles increased volume easily by increasing RV stroke volume. Regulation of RV contractility in the healthy RV is similar to the LV and is dependent on factors such as heart rate, the Frank-Starling mechanism (stroke volume increases as preload increases), and autonomic neural input.
The pattern of development of RV structural and functional abnormalities is similar in patients with PH-LD and PAH and is secondary to the increased afterload on the RV that results from the increase in PVR (see Fig. 59-5 ). The changes in the RV from pressure overload are distinct from changes in the RV secondary to volume overload as from conditions such as severe tricuspid regurgitation. RV afterload generally increases slowly in patients with chronic lung disease as P pa and PVR rise, leading to progressive RV hypertrophy, which minimizes wall stress. The RV eventually dilates so that the normal crescent shape of the RV is progressively transformed into a more spherical structure that is better able to generate an increased stroke work (see Fig. 59-5 ). RV dilation and wall thinning result in increased RV wall stress that, along with increased heart rate, leads to further increases in myocardial oxygen consumption, decreased myocardial perfusion, and RV ischemia. As the RV dilates, severe tricuspid regurgitation may develop, further compromising RV cardiac output and LV filling. RV dilation in the setting of an intact pericardium compromises LV filling by shifting the interventricular septum toward the LV and reducing LV filling and cardiac output. Nevertheless, several studies of end-diastolic pressure-volume relationships suggested well-preserved RV contractility in patients with COPD. Only the additional presence of acidemia or infection precipitates RV failure. When chronic hypercapnia with acidosis is present in patients with alveolar hypoventilation, the ability of the RV to increase its work appears to be significantly impaired, and RV end-diastolic pressure increases.
There may be subtle changes before elevations in P pa can be detected. For example, even before the development of significant elevations in P pa , RV hypertrophy and pathologic remodeling has been observed in patients with chronic lung disease. This was studied by echocardiographic and invasive hemodynamic evaluation of the RV in a group of 98 patients with stable moderate to severe COPD and no known heart disease. Compared to 34 healthy controls, patients with COPD but without PH had increased RV wall thickness, RV size, and outflow tract dimension, as well as functional abnormalities assessed by myocardial performance index, RV isovolumic acceleration, and RV strain. In another study of patients with COPD, pulmonary artery compliance was reduced before PVR was significantly elevated, suggesting that decreases in compliance may be an early marker of hemodynamic compromise. In patients with lung disease, it is likely that early destruction of the pulmonary vascular bed, though not sufficient to increase PVR, significantly reduces pulmonary artery compliance and contributes to early increases in RV afterload and subsequent RV hypertrophy.
RV myocardial ischemia may also play a role in RV failure. RV coronary perfusion decreases with the increase in wall thickness even as the hypertrophy and dilation of the RV produces a significant increase in RV myocardial oxygen consumption. RV coronary perfusion also decreases with the increase in end-diastolic pressure as RV stiffness increases during diastole. Taken together, these factors result in an imbalance between RV myocardial oxygen demand and supply.
This impairment of myocardial contractility in the presence of hypercapnia probably plays a significant role in producing decompensated pulmonary heart disease in response to acute increases in arterial P co 2 associated with exacerbations of COPD and accompanying decreases in alveolar ventilation. The development of RV volume overload with ventricular dilation results in a decreased ejection fraction, because stroke volume tends to be maintained close to the normal range in decompensated cor pulmonale. With exercise, patients with COPD significantly raise their RV afterload, which causes further increases in RV end-diastolic volume and decreases in ejection fraction. It is likely that this deterioration in hemodynamic performance with exercise is a major factor limiting the ability of such patients to exercise normally.
In the presence of RV failure and elevated central venous pressure, the patient can stand up without a decrease in RV end-diastolic volume or stroke volume; consequently, the heart rate does not change. This lack of postural reflex compensation is attributable to failure of any incremental gravitational pooling of blood in the venous system, because of increased plasma volume, increased tissue pressure from edema with decreasing venous dispensability, and increased venomotor tone.
Left Ventricle
Although LV ejection performance is unimpaired, cardiac catheterization studies have revealed abnormal LV end-diastolic pressure-volume relationships. Echocardiographic studies have also shown progressive impairment of LV diastolic function that correlates with the severity of PH. It is likely that this results, in large part, from bulging of the interventricular septum from the hypertrophied and dilated RV into the cavity of the LV, and ventricular interdependence exerted by pericardial constraint. As a result, LV diastolic geometry becomes distorted, and filling characteristics may be altered so that a higher filling pressure is required to accomplish the same end-diastolic fiber stretch needed for a given stroke work, in accordance with the Frank-Starling mechanism. With severe RV failure and marked elevation in right atrial pressure, coronary venous pressure increases significantly; this increase in coronary venous pressure can result in an increase in LV wall dimension limiting LV distensibility. This mechanism leading to reduced LV preload appears to act independently of diastolic ventricular interaction caused by RV enlargement as previously described.
Lung Mechanics
In healthy individuals, ordinary inspiration causes a small reduction of 3 to 5 cm H 2 0 in pleural pressure that is adequate to generate a normal tidal volume with only a small transmission of intrathoracic pressure affecting the heart and pulmonary vasculature. Increased stiffness of the chest wall or the lung parenchyma means that patients with obesity or obstructive or restrictive lung diseases must generate more negative pleural pressures to achieve an adequate tidal volume. Additionally, patients with COPD often have lung hyperinflation secondary to airflow obstruction, loss of elastic lung parenchyma, or dynamic hyperinflation. These abnormalities in pulmonary mechanics may contribute significantly to the pathophysiologic features of PH.
Lung hyperinflation affects RV function through changes in RV preload and afterload. Increased lung volumes in diseases such as COPD can passively compress alveolar vessels directly, increasing PVR and RV afterload, and in some cases, hyperexpanded lungs may even directly compress the heart with negative effects on cardiac performance. Exaggerated swings in intrathoracic pressure such as those seen in patients with OSA can result in increased right-sided venous return, which causes acute RV enlargement and in turn impedes LV filling and cardiac output. At the same time, reductions in intrathoracic pressure increase LV afterload, which may be transmitted backwards and result in transient increases in P pa . In patients with acute bronchospasm, echocardiographic studies have shown acute inspiratory RV dilation and simultaneous reduction in LV cavity size that reverse during expiration, resulting in increased PVR from hyperinflation and increased RV wall tension from exaggerated negative pleural pressure. Importantly, the clinical significance of these effects may depend largely on the cardiovascular reserve of the specific patient.
Clinical Presentation
There are many causes of PH, as noted in Table 59-1 . At times the underlying cause may be difficult to establish, and often the cause of PH is multifactorial. In one study of 998 patients with COPD undergoing right heart catheterization, 27 were found to have 
greater than or equal to 40 mm Hg; after complete evaluation, 16 (59%) were found to have another cause of PH, including administration of appetite suppressants, left ventricular dysfunction, chronic thromboembolic PH, collagen vascular disease, portal hypertension, and OSA. When evaluating a patient with a new diagnosis of PH, an exhaustive evaluation is required to identify all possible causes of increased P pa . The prognosis and treatment of PH vary dramatically with the cause so that the significance of PH and optimal treatment require a complete and accurate understanding of the underlying cause or causes.
Symptoms and Signs
The first step in the evaluation is a comprehensive history and physical examination to identify any conditions that might cause PH. Specifically the patients should be questioned regarding use of appetite suppressants or other toxic substances, a history of liver disease and portal hypertension and diagnosis or symptoms of systemic lupus erythematous, scleroderma, or other collagen vascular disease, or a history of venous thromboembolism. In patients already diagnosed with lung disease, it is important that the lung disease be well characterized and that definitive diagnostic studies have been performed and are reviewed. Studies to exclude daytime hypoxemia with exercise and nocturnal hypoxemia from OSA should be performed. Left-sided heart disease is the most common cause of PH and must be identified if it is contributing to PH. Although echocardiography is useful to detect LV systolic dysfunction and valvular abnormalities that may cause PH, heart failure with preserved ejection fraction (formerly called diastolic heart failure) can be missed and may be detected only during comprehensive invasive hemodynamic assessment.
Mild PH-LD reflected in small chronic elevations in RV pressure generally causes minimal, if any, clinical, radiologic, or electrocardiographic findings. When moderate or severe PH-LD develops ( 
> 40 mm Hg), symptoms are often similar to those associated with the underlying pulmonary disease. Most commonly, these are dyspnea on exertion, chronic cough productive of mucoid sputum, wheezing, and occasional cyanosis. Clubbing of the fingers may be present. In addition to exertional dyspnea and fatigue, some patients experience dizziness or exertional syncope, attributable to the inability to increase cardiac output during exercise in the face of a marked increase in PVR. Furthermore, these patients may have chest pain owing to RV ischemia or stretching of the main pulmonary artery.
When resting P pa is sufficiently elevated, patients may eventually reach a point at which the RV cannot meet the need for increased stroke work without a significant increase in right heart filling pressures. The resultant increase in central venous pressure is associated with developing symptoms of right-sided heart failure, such as peripheral edema, right upper quadrant discomfort, nocturia, and easy fatigability.
On examination, the patient is often cyanotic and sitting upright with tachypnea, with prominent use of the accessory muscles of breathing, and arms extended holding on to the edges of the mattress. In COPD, pulsus paradoxus may be present, the chest is often hyperinflated, and otherwise mild wheezing may be audible. Sinus tachycardia is often present; however, atrial and ventricular arrhythmias are also common. Evidence of fluid retention may include dependent edema and ascites. The liver may be enlarged and tender to palpation and may be pulsatile, reflecting the presence of severe tricuspid regurgitation. Similarly, the neck veins may be distended and, when tricuspid regurgitation is present, show a large c-v wave with rapid y descent. Signs of volume overload secondary to RV dysfunction must be distinguished from sympathetically mediated renal salt and water retention without RV dysfunction, which can also develop in patients with lung disease.
On examination of the chest, there may be a left parasternal systolic lift, owing to the overactivity of the enlarged RV, and a thud felt over the pulmonary area as the pulmonary valve closes. The heart sounds are often difficult to hear if the patient has underlying COPD. The pulmonic component of the second heart sound (S 2 ) may be accentuated and be heard earlier than usual, so the normal splitting may be abolished and a single loud S 2 heard. Normally not heard at the apex, the pulmonic component of S 2 may be clearly heard. A high-pitched systolic ejection click may be heard in the second and third left intercostal spaces next to the sternum. It is often followed by a soft, localized systolic ejection murmur produced by ejection of the stroke volume into a dilated pulmonary artery. An S 3 gallop arising from the right side of the heart may be heard in the fourth and fifth interspaces immediately to the left of the sternum or even next to the xiphoid process. A presystolic S 4 gallop may also be heard, reflecting the increased forcible contraction of the right atrium with expulsion of blood into the hypertrophied and dilated RV. Often tricuspid regurgitation is present, and this results in a prominent blowing pansystolic murmur with respiratory variation in the same location. When prominent PH is present, a diastolic murmur of pulmonic valve regurgitation may be heard; this murmur, known as a Graham Steell murmur, is a soft, blowing decrescendo diastolic murmur, usually well localized to the second and third left intercostal spaces next to the sternum.
Electrocardiography
Characteristically, the P wave of the electrocardiogram has a “p pulmonale” pattern with right-axis deviation resulting in an increase in its amplitude in leads II, III, and aVF to more than 2.5 mm ( Fig. 59-6 ). The P wave may also be tall in the right precordial leads. The QRS vector in the frontal plane often shifts to the right in cor pulmonale, and a low-voltage QRS complex is common if lung hyperinflation is present. Prominent S waves are seen in leads I, II, and III. An incomplete right bundle-branch block pattern is also frequently observed. When PH is moderate or severe, the more classic findings of RV hypertrophy may dominate the electrocardiogram, including tall R waves in V 1 with an R/S ratio of more than 1, and a prominent S wave in V 5 and V 6 with an R/S ratio of less than 1. The presence of electrocardiographic evidence of cor pulmonale in patients with COPD is a poor prognostic sign.
