CT Findings in Diseases Associated with Pulmonary Hypertension: A Current Review

LEARNING OBJECTIVES FOR TEST 1

After reading this article and taking the test, the reader will be able to:

•.	List the various disorders associated with pulmonary hypertension.
•.	Discuss the pathophysiologic mechanisms leading to pulmonary hypertension.
•.	Describe the characteristic histologic features of pulmonary arterial and venous hypertension and the related morphologic changes seen at CT.

Introduction

Pulmonary hypertension is hemodynamically defined as a mean pulmonary artery pressure greater than 25 mm Hg at rest or greater than 30 mm Hg during exercise with an increased pulmonary vascular resistance (1,2). This condition may be caused by a wide variety of disease entities with overlapping radiologic-histologic features. The diagnosis is based on a clinical assessment of hemodynamic parameters, medical history, results of pulmonary function testing, and radiologic and histologic findings (1,3). Although pulmonary hypertension is associated with high morbidity and mortality, patient referral to a specialized center is often delayed because of the nonspecificity of clinical manifestations. High-resolution computed tomography (CT) and CT angiography play a crucial role in the diagnostic work-up of pulmonary hypertension and are particularly important for identifying patients with chronic or recurrent pulmonary thromboembolism and for assessing the feasibility of pulmonary thromboendarterectomy (4). Disease features seen on high-resolution CT scans and CT angiograms are helpful for diagnosing idiopathic pulmonary arterial hypertension and detecting and identifying disorders underlying secondary pulmonary hypertension (Table 1). The article presents an encyclopedic review of morphologic CT features and histologic findings of diseases associated with pulmonary hypertension, with reference to the underlying pathophysiologic mechanisms. Clinical features also are described.

Pulmonary Arterial Hypertension

Regardless of the underlying pathologic changes, the characteristic morphologic CT features of chronic pulmonary arterial hypertension are dilatation of the pulmonary artery trunk, the diameter of which frequently exceeds that of the ascending aorta; dilatation of the right and left main pulmonary arteries; abrupt narrowing and tapering of the peripheral pulmonary vessels; right ventricular hypertrophy; and right ventricular and atrial enlargement with inversion of the interventricular septum and dilatation of the tricuspid valve annulus (1). In adult patients, the presence of a distal main pulmonary artery with a CT-demonstrated diameter greater than or equal to 29 mm at its widest point has a positive predictive value of more than 95% and a specificity of 89%, and a CT-demonstrated distal main pulmonary artery diameter exceeding that of the ascending aorta has a positive predictive value of more than 95% and a specificity of more than 90%, for a diagnosis of pulmonary arterial hypertension (8,9). Pulmonary arterial hypertension can be reliably predicted when (a) the CT-demonstrated diameter of the distal main pulmonary artery is greater than or equal to 29 mm and the segmental artery-to-bronchus ratio is greater than 1:1 in three of four pulmonary lobes (specificity, 100%) or (b) the ratio of the CT-demonstrated distal main pulmonary artery diameter to the aortic diameter is greater than 1:1, particularly in patients younger than 50 years (8,9). In a recently published ECG-gated multidetector CT study, Revel et al (10) found that of all the ECG-gated CT parameters evaluated, right pulmonary artery wall distensibility, defined by the change in cross-sectional area between diastole and systole, was the most reliable parameter for identifying patients with pulmonary hypertension and showed the strongest correlation to mean pulmonary artery pressure. Distensibility was calculated by dividing the difference between the maximum cross-sectional area and the minimum cross-sectional area by the maximum cross-sectional area and multiplying the result by 100. The best pulmonary artery distensibility cutoff for distinguishing between patients with pulmonary hypertension and those without it was 16.5% (sensitivity, 86%; specificity, 96%) when a standard method of measurement was used (when the maximum and minimum right pulmonary artery cross-sectional areas were determined on the basis of 10 measurements, one measurement at every 10% of the cardiac cycle) and 13% when a simplified method was used (when the maximum and minimum right pulmonary artery cross-sectional areas were determined on the basis of two measurements, one in each of two predefined phases of the cardiac cycle). Systolic right ventricular outflow tract diameter and cross-sectional area were found to differ significantly between patients with pulmonary hypertension and patients without it, whereas the diastolic values were not significantly different (10). These findings suggest that the measurement of right pulmonary artery wall distensibility at ECG-gated CT may help improve the accuracy of CT-based diagnoses of pulmonary hypertension and that the measurement of diastolic right ventricular outflow tract wall thickness, which correlates with the mean pulmonary artery pressure and is potentially reversible, may be useful in patients referred for CT evaluation of pulmonary hypertension.

Complications of sustained pulmonary arterial hypertension that are detectable with CT include central pulmonary artery thrombosis, premature atherosclerosis of the pulmonary arteries, pulmonary artery dissection, and right heart chamber hypertrophy and dilatation (1,5,11). Histologic hallmarks of pulmonary arterial hypertension are (a) intimal cellular proliferation and medial hypertrophy, primarily in the walls of muscular arteries, and (b) plexogenic arteriopathy, which manifests in small to medium-sized muscular pulmonary arteries as intimal cellular proliferation with focal disruption of the internal elastic lamina and media by “glomeruloid” small vascular channels, which ramify into alveolar septal capillaries (1–3). Plexogenic arteriopathy can occasionally be detected at CT, manifesting as small, tortuous peripheral arteries without a significant connection to pulmonary veins such as that seen in an arteriovenous shunt (5).

Idiopathic Pulmonary Arterial Hypertension

Idiopathic pulmonary arterial hypertension is a subtype of pulmonary arterial hypertension without an identifiable cause. Although occurrences have been reported in patients with widely varying ages, idiopathic pulmonary arterial hypertension is generally considered a disease of young adulthood, occurring most often in those between the ages of 20 and 45 years (7). Women are more commonly affected than men (12,13). The clinical manifestations are nonspecific and may include exertional dyspnea (60% of all cases), fatigue, angina, syncope, and cor pulmonale (1,14). The average delay between the onset of symptoms and the diagnosis of idiopathic pulmonary arterial hypertension is 2 years. The prognosis is poor, with a median survival of 2.8 years and a 5-year survival rate of only 34% (7). Factors contributing to the pathogenesis of idiopathic pulmonary arterial hypertension include genetic predisposition, endothelial cell dysfunction, abnormalities in vasomotor control, thrombotic obliteration of the vascular lumen, and vascular remodeling (1,2,7,13,15). Reported risk factors for development of the disease are autoimmune disorders such as the Raynaud phenomenon, a positive result of testing for antinuclear antibodies, pregnancy, and a history of appetite suppressant ingestion (1,7,13,14). Plexogenic arteriopathy is considered the histologic hallmark of idiopathic pulmonary arterial hypertension; it is seen in approximately 75% of cases (1). Acute and organizing intraluminal thrombi are recognized at histologic examination in more than 50% of cases (1,3,10,16). Although organizing thrombi associated with idiopathic pulmonary arterial hypertension may display histologic features similar to those of plexiform lesions (1–3), the pathologic differentiation of in situ thrombosis of peripheral vessels in patients with idiopathic pulmonary arterial hypertension from pulmonary artery occlusion due to chronic pulmonary thromboembolism may be difficult (1,10).

Characteristic vascular features of idiopathic pulmonary arterial hypertension depicted at CT are central pulmonary artery dilatation, usually in the absence of detectable intraluminal thrombi; small tortuous peripheral vessels representing plexogenic arteriopathy; and an abrupt decrease in the caliber of segmental and subsegmental arteries. Wall-adherent apposition thrombi may form in the central pulmonary arteries in severe cases of idiopathic pulmonary arterial hypertension and usually are accompanied by massive enlargement of the pulmonary artery trunk and the right and left main pulmonary arteries (Fig 1) (11). Additional CT findings may include right heart enlargement, pericardial effusion, and a mosaic pattern of attenuation in lung parenchyma. The likelihood of a finding of pericardial effusion increases with the severity of pulmonary hypertension (17), and the presence of a pericardial effusion implies a worse prognosis (18). A mosaic pattern of attenuation caused by regional variations in lung perfusion is a frequent finding in patients with idiopathic pulmonary arterial hypertension. Unlike the segmental and subsegmental patterns of variable attenuation typically seen in chronic thromboembolic pulmonary hypertension (CTEPH) (11,19), the pattern seen in idiopathic pulmonary arterial hypertension often is characterized by focal perivascular hyperattenuating areas in a peripheral or perihilar distribution (Fig 2) or small, scattered, well-defined areas of low attenuation corresponding to the anatomic unit of a secondary pulmonary lobule with adjacent areas of increased attenuation in a patchy and diffuse distribution (11). The combination of mosaic lung attenuation with marked variations in the size of segmental vessels, which is common in CTEPH, is rarely seen in idiopathic pulmonary arterial hypertension (19,20). Peripheral parenchymal opacities due to previous infarction are also infrequently seen (19). Dilatation of bronchial and nonbronchial systemic arteries is more commonly seen in patients with CTEPH (73%) than in those with idiopathic pulmonary arterial hypertension (14%), and, in cases with inconclusive imaging findings, CT evidence of dilated bronchial arteries may help distinguish between these two disease entities (21).

Pulmonary Arterial Hypertension Due to Chronic Thrombotic and/or Embolic Disease

Pulmonary arterial hypertension may be caused by thrombotic or thromboembolic obstruction of the pulmonary arteries (as in CTEPH or sickle cell disease) (22) or by nonthrombotic embolic obstruction of the pulmonary vascular bed, which may be due to tumor emboli (originating from gastric carcinoma, carcinoma of the breast or prostate, hepatocellular carcinoma, malignant melanoma, choriocarcinoma, ovarian carcinoma, or right atrial myxoma), parasitic emboli (commonly produced by Schistosoma mansoni), or foreign material emboli. Fat emboli, amniotic fluid emboli, and septic pulmonary emboli rarely produce clinically significant pulmonary arterial hypertension (7). According to recently published epidemiologic data from several prospective studies, CTEPH occurs as a complication in 3.8% of cases of acute symptomatic pulmonary embolism (23–25). However, since the initial thromboembolic event is asymptomatic in most patients, CTEPH is likely more common than previously thought (23). CT features that are helpful for differentiating CTEPH from other disease entities are summarized in Table 2.

Pathogenesis of CTEPH

CTEPH is believed to result from obstruction of the pulmonary arteries by nonresolving thromboemboli that become dislodged during one or more pulmonary thromboembolic events from their venous sites of origin (23). In patients with CTEPH, pulmonary thromboemboli do not resolve but instead form endothelialized, sometimes recanalized fibrotic obstructions of the pulmonary vascular bed. In addition, secondary small-vessel arteriopathy of varying degrees of severity develops in response to factors such as shear stress, pressure, inflammation, and the release of cytokines and vasculotrophic mediators in unobstructed as well as partly or completely obstructed pulmonary arteries (23,26). Histologic examination reveals small-vessel arteriopathy (which is indistinguishable from the pulmonary vascular lesions found in any other form of pulmonary arterial hypertension) in combination with various stages of organized and recanalized thromboemboli in both the elastic and the muscular pulmonary arteries (1). After an asymptomatic period of months to several years, patients present with recurrent acute or progressive exertional dyspnea—most commonly, with New York Heart Association class III cardiac dysfunction (27)—which may be accompanied by a chronic nonproductive cough, atypical chest pain, tachycardia, syncope, or cor pulmonale (1,27–29). Although CTEPH is potentially curable with pulmonary thromboendarterectomy, the prognosis is poor, with a 5-year survival rate of only 30% if the mean pulmonary artery pressure exceeds 30 mm Hg (30,31). Splenectomy, ventriculoatrial shunt creation for treatment of hydrocephalus, placement of a permanent central intravenous line, and inflammatory bowel disease create conditions that predispose patients to CTEPH and are predictive of a poor prognosis (27).

CT Features of CTEPH

Directly visualized intraluminal thrombi in the pulmonary arteries are the vascular CT finding with the highest specificity for a diagnosis of CTEPH (32,33). The spectrum of CT findings that are due to partial vascular obstruction includes thickening of the pulmonary arterial wall; an irregular contour of the intimal surface; and intraluminal bands, webs, and wall-adherent soft-tissue formations, which may be masslike, interposed between the intimal surface and the column of contrast-enhanced blood (Figs 3–5) (34,35). At conventional angiography, complete vascular obstruction appears as a convex margin of the contrast material bolus. This feature, which has been described as a “pouch defect,” is difficult to detect on axial CT sections (Fig 4b); an abrupt decrease in vessel diameter and the absence of contrast material in the vessel segments distal to the obstruction are easier to identify (36). CT scans viewed at lung window settings demonstrate segmental and subsegmental vessels with diameters that are abnormally narrowed in comparison with the diameters of the accompanying bronchi, abrupt vessel cut-offs, and marked variation in the size of segmental vessels (35). Findings of disparity in the size of segmental vessels and mosaic lung attenuation reliably distinguish CTEPH from nonthromboembolic pulmonary arterial hypertension (19). Mild pericardial thickening or pericardial effusion may be seen in patients with severe CTEPH (17). Lymph node enlargement is common and corresponds histologically to vascular transformation of the lymph node sinus with various degrees of sclerosis (37).

Owing to the increased bronchial artery blood flow secondary to chronic obstruction of the pulmonary arteries, dilated bronchial arteries (those with a diameter ≥ 1.5 mm) are seen in 47%–77% of cases (30,38) and nonbronchial systemic collaterals (notably, inferior phrenic, intercostal, and internal mammary arteries) are visible in as many as 45% of cases of CTEPH (Fig 5) (21). In a study by Remy-Jardin et al (21), enlarged bronchial and nonbronchial systemic arteries were found in 73% of patients with CTEPH and in only 14% of patients with idiopathic pulmonary arterial hypertension. According to two other reports (30,38), postoperative pulmonary vascular resistance and mortality after pulmonary thromboendarterectomy were significantly lower among patients with dilatation of bronchial arteries than among those without it. The finding of higher postoperative pulmonary vascular resistance suggests that patients without dilated bronchial arteries may have a higher degree of distal vascular disease because of distal thromboembolic occlusion or secondary small-vessel arteriopathy (30,38).

The parenchymal findings in patients with CTEPH are nonspecific but often helpful toward achieving a definitive diagnosis in the appropriate clinical setting. Mosaic lung perfusion is the key imaging feature produced by CTEPH-related parenchymal changes. Mosaic lung perfusion is characterized by sharply demarcated regions of hypoattenuation with reduced vessel size and without air trapping, interspersed with adjacent areas of normal attenuation or relative hyperattenuation (33). Hypoattenuation is produced by hypoperfusion in lung areas distal to occluded vessels or by small-vessel arteriopathy in nonobstructed lung areas, whereas hyperattenuation is produced by a compensatory increase in blood flow to open vessels because of blood flow redistribution and collateral blood flow (20,21,33). Mosaic lung perfusion is seen in 77%–100% of patients with CTEPH (19,33) and is found significantly more often among those with pulmonary arterial hypertension due to vascular disease than among those with pulmonary arterial hypertension due to cardiac or lung disease (74% [17 of 23] versus 8% [three of 38] of the patients in one series) (20). Regions of CTEPH-related hypo- and hyperattenuation are typically segmental or subsegmental in distribution (Fig 6). A pattern of exclusively perihilar hyperattenuating areas and peripheral perfusion defects also may be seen but is relatively uncommon in CTEPH (Fig 7). Although evidence of air trapping in hyperfused lung areas on expiratory CT scans obtained in patients with CTEPH has been described (39), typically all lung areas in CTEPH-related mosaic lung perfusion change equally in attenuation with inspiration and expiration, showing no air trapping on expiratory CT scans. Among the various CT parenchymal and vascular findings of CTEPH, the number of abnormally perfused lobes correlates most closely with hemodynamic measurements of pulmonary artery pressure and pulmonary vascular resistance, standard indicators of disease severity (30). Although a CT appearance of normal perfusion of the lung parenchyma does not exclude the presence of one or more proximal intravascular thromboemboli in the supplying artery, a mosaic pattern of attenuation is indicative of chronic pulmonary thromboembolism even without direct visualization of intraluminal thrombi (11).

Figure 6

Download as PowerPointOpen in Image Viewer

Other frequent parenchymal findings in CTEPH include peripheral opacities caused by previous infarction, which are seen in 72%–87% of cases (30,34) (Figs 6, 8, 9), and cylindrical bronchial dilatation, which is seen in as many as 64% of cases (40) (Fig 10). Bronchial dilatation occurs predominantly in the lower lung zones, at the level of the segmental or subsegmental bronchi, in areas of severely stenosed or completely occluded pulmonary arteries. The hypothesized causal mechanisms of bronchial dilatation include local bronchovascular mechanical interactions associated with a reduction in vascular volume because of chronic occlusion of the pulmonary arteries; traction bronchiectasis caused by postthrombotic fibrovascular organization at the level of severely narrowed arteries; and a steal phenomenon at the expense of the bronchial wall, which might be weakened by the increased demand imposed on the bronchial circulation by the hypoperfused lung parenchyma (40–42).

Figure 9

Download as PowerPointOpen in Image Viewer

Pulmonary Arterial Hypertension Associated with Lung Disease

Lung disease is the most common cause of pulmonary hypertension, and the presence of pulmonary hypertension in this setting is an unfavorable prognostic sign (Figs 3, 4, 11–13). Patients typically present with signs and symptoms related to the specific underlying disease. Restrictive lung diseases associated with pulmonary hypertension include idiopathic interstitial pneumonias (eg, idiopathic pulmonary fibrosis, in which pulmonary hypertension is reported to occur with a prevalence as high as 46%) (43–46); secondary interstitial pneumonias due to connective tissue disease, sarcoidosis (Fig 12), vasculitis, drug toxicity, and exposure to various environmental toxins; and a heterogeneous group of conditions such as thoracic cage deformities (due to kyphoscoliosis, thoracoplasty, or restrictive pleural disease), diaphragmatic disorders, neuromuscular diseases, and spinal cord injuries. The pathophysiologic mechanisms of pulmonary hypertension in obstructive and restrictive lung diseases include acute hypoxic vasoconstriction, vascular remodeling due to sustained alveolar hypoxia, loss of cross-sectional lung area caused by destruction of the alveolar capillary septa, compression of the alveolar vessels by increased intraalveolar pressures (chronic obstructive pulmonary disease), and fibrotic compression or obliteration of the pulmonary vessels (interstitial lung disease) (7). At CT, the characteristic features of pulmonary arterial hypertension are seen in combination with pathologic changes caused by the underlying lung disease. Coexistent CTEPH may be difficult to identify, especially in patients with chronic obstructive pulmonary disease, because CTEPH-related vascular changes may be obscured by vascular narrowing and vessel wall irregularities due to pulmonary emphysema (Fig 11).

Figure 12

Download as PowerPointOpen in Image Viewer

On one hand, the lack of correlation between the mean pulmonary artery pressure and the extent of fibrosis at CT suggests that pulmonary hypertension in patients with fibrotic lung disease is not simply a consequence of fibrotic obliteration of the pulmonary vasculature but also reflects pathologic processes such as small vessel remodeling (44,47–49). On the other hand, this finding indicates that a hypothetical traction effect of pulmonary fibrosis–associated reduced lung volumes and increased lung recoil on vascular structures cannot be the sole mechanism causing pulmonary artery dilatation in the absence of pulmonary hypertension in patients with fibrotic lung disease (8,47,50).

Pulmonary Arterial Hypertension Associated with Cardiac Disease

Pulmonary arterial hypertension due to sustained cardiac left-to-right shunt and shunt reversal (Eisenmenger syndrome) can be seen in patients with congenital cardiac abnormalities such as ventricular septal defect, atrial septal defect (Fig 14), and patent ductus arteriosus. Patients with transposition of the great arteries are predisposed to early onset of pulmonary arterial hypertension (Fig 15) (1). It is calculated that 10% of patients with a ventricular septal defect of any size that is older than 2 years may develop Eisenmenger syndrome, compared with 4%–6% of subjects with an atrial septal defect. Among patients with a large defect, 50% of those with a ventricular septal defect and 10% of those with an atrial septal defect develop pulmonary arterial hypertension. Among patients with a small defect (a ventricular septal defect < 1 cm or an atrial septal defect < 2 cm in effective diameter assessed with echocardiography), the exact pathophysiologic role of the heart defect in the development of pulmonary arterial hypertension is unknown (51).

The severity of structural changes induced in the pulmonary arteries by congenital heart disease appears to correspond to the magnitude of pulmonary artery pressure alone, having no correlation with either the patient’s age or the location of the intracardiac defect (1), and is traditionally described by using the Heath and Edwards grading system (52) (Table 3). Because congenital cardiac abnormalities are generally detected and repaired at an early age, the development of severe pulmonary hypertension secondary to congenital cardiac lesions is rare, and in these cases a lung biopsy may be helpful for determining operability and prognosis by assessing the potential reversibility of pulmonary hypertension (1,3).

Pulmonary Hypertension Associated with Arteriovenous Shunts

Pulmonary arteriovenous malformations may occur sporadically but are most commonly seen in association with Osler-Weber-Rendu disease (hereditary hemorrhagic telangiectasia), an autosomal dominant disorder that is characterized by cutaneous and mucosal telangiectasia and arteriovenous malformations in other organs (5). Diffuse pulmonary arteriovenous shunting also is seen in pregnancy, polysplenia syndrome, liver cirrhosis, and complex cardiac malformations (53). Untreated or occult arteriovenous malformations may increase in size over time, causing a hyperdynamic circulatory status with severe hypoxemia due to high shunt volumes, with the eventual development of pulmonary hypertension. In patients with pulmonary hypertension due to (pulmonary) arteriovenous malformations, the characteristic vascular features of pulmonary arterial hypertension may be seen at CT in combination with a mosaic pattern of lung perfusion (Fig 16), an appearance produced by secondary small-vessel arteriopathy or by thromboembolic obstruction of the pulmonary arteries because of emboli originating in or passing through arteriovenous connections. Other frequent CT findings include small, rounded parenchymal opacities in the lung periphery with a dilated, tortuous feeding artery and an enlarged draining vein, features that represent CT-detectable macroscopic arteriovenous shunts (5). Diffuse macroscopic arteriovenous shunting typically manifests at CT as dilatations of the pulmonary arteries and veins with a weblike, reticular vascular pattern in the lung periphery and centrilobular vessel-associated micronodules connected by arcadelike, dilated vascular branches (4), whereas diffuse telangiectatic microshunts may be visible only at microscopic analysis.

Hepatopulmonary Syndrome and Portopulmonary Hypertension

In patients with liver cirrhosis, two different causes of pulmonary hypertension may be distinguished: hepatopulmonary syndrome and portopulmonary hypertension. Hepatopulmonary syndrome develops in 15%–20% of patients with liver cirrhosis (54). It is characterized by an increased alveolar-arterial oxygen gradient during normal breathing of ambient room air and by pulmonary vascular dilatation, especially of the small vessels and capillaries. The presumed underlying pathophysiologic mechanism is excessive vascular production of vasodilators (particularly, nitric oxide). At histologic analysis, intrapulmonary vascular dilatation manifests as dilated precapillaries, direct arteriovenous shunts, and dilated pleural vessels (55,56).

Portopulmonary hypertension is reported to occur in 2%–5% of patients with liver cirrhosis and portal hypertension (55,57). Three mechanisms are thought to play a causal role: First, vasoactive substances such as serotonin, interleukin 1, endothelin 1, and thromboxane, which may cause vasoconstriction and mitogenesis in the pulmonary arteries, escape hepatic detoxifying mechanisms through portosystemic shunts or are not cleared as effectively by the diseased liver. Second, venous thromboemboli arising from the portal vein or other systemic sources may reach the pulmonary circulation through portosystemic shunts, causing pulmonary hypertension (55,58). Third, high cardiac output in patients with liver cirrhosis leads to increased shear stress on the pulmonary vascular bed, with resultant vasoconstriction, vascular remodeling, and fibrotic narrowing of the pulmonary vessels (55,59) (Fig 17).

Pulmonary Hypertension Associated with Connective Tissue Disease

Pulmonary hypertension associated with connective tissue disease is most commonly observed in systemic sclerosis, notably in its limited variant previously defined as CREST syndrome (calcinosis, Raynaud disease, esophageal dysmotility, sclerodactyly, and telangiectasia). A registry study of pulmonary hypertension in 722 patients with systemic sclerosis in the United Kingdom showed a prevalence of approximately 12%, and in several pulmonary hypertension centers more than 10% of patients with severe pulmonary arterial hypertension have an associated connective tissue disease, most often the CREST variant of scleroderma (51). Pulmonary hypertension in this patient group is related to interstitial fibrosis, which is most commonly manifested histologically as nonspecific interstitial pneumonia (60–63), pulmonary vascular disease, or both. Vasculopathy may represent a complication of pulmonary fibrosis but also frequently occurs in the absence of significant interstitial lung fibrosis, and, characteristically, the severity of pulmonary hypertension is not correlated with the degree of interstitial fibrosis depicted at high-resolution CT (61,64,65). Histopathologic changes in pulmonary arterial hypertension associated with connective tissue disease are generally indistinguishable from those of idiopathic pulmonary arterial hypertension, except for scleroderma-associated vasculopathy, which is characterized histologically by medial and intimal hyperplasia with a marked absence of plexiform lesions (61). Pulmonary hypertension is less frequently seen in mixed connective tissue disease, polymyositis, dermatomyositis, and systemic lupus erythematosus (60). By contrast, in a study by Tanaka et al (66), rheumatoid arthritis–associated pulmonary arterial hypertension was found in 26 (41.3%) of 63 patients with rheumatoid arthritis–associated lung disease.

In comparison with patients with idiopathic pulmonary arterial hypertension, patients with pulmonary arterial hypertension associated with connective tissue disease are mainly women, are older, have a significantly lower cardiac output, and show a trend toward a shorter survival (51).

Pulmonary Arterial Hypertension in HIV Infection

HIV-associated pulmonary arterial hypertension is a life-threatening complication of HIV infection (Fig 18). The estimated prevalence of pulmonary arterial hypertension among HIV-infected patients—0.46%, according to recently published prospective epidemiologic data (67)—is higher than that of idiopathic pulmonary arterial hypertension in the general population (0.02%) (7). The mechanism leading to the development of pulmonary arterial hypertension in the setting of HIV infection is unknown. Indirect action of the virus through second messengers such as cytokines and growth factors is strongly suspected because of the absence of viral DNA in pulmonary endothelial cells (7,51,68). In addition, genetic predisposition may be invoked, because this complication affects only a minority of HIV-infected patients. At diagnosis, most patients have CD4+ lymphocyte counts of more than 200 cells per microliter and New York Heart Association class III cardiac dysfunction (69). The overall prognosis is poor, mainly because of the lowered CD4+ lymphocyte cell count and decreased cardiac function (69,70). In a multivariate analysis, a cardiac index of more than 2.8 L/min/m² and a CD4+ lymphocyte count of more than 200 cells per microliter were independent predictors of survival (69). Although the effect of highly active antiretroviral therapy (HAART) on the outcome of HIV-associated pulmonary hypertension is debated, a combination of HAART and pulmonary arterial hypertension–specific therapy has been shown to significantly improve hemodynamic parameters and survival (69,70). Because of considerable overlap between HIV-associated and idiopathic pulmonary arterial hypertension with regard to their clinical, pathologic, and radiologic manifestations, HIV testing should be considered in patients with CT evidence of pulmonary arterial hypertension without another identifiable cause (7).

Figure 18

Download as PowerPointOpen in Image Viewer

Primary Sarcoma of the Pulmonary Arteries

In patients with undifferentiated sarcoma or leiomyosarcoma of the pulmonary arteries, pulmonary hypertension may result from direct occlusion of the arterial lumen by a tumor mass or apposition thrombus or from obliteration of the pulmonary vessels by secondary pulmonary artery thrombosis. Secondary thromboembolic events are common and may be the only clinical evidence of a tumor at presentation, with clinical features similar to those of recurrent pulmonary embolism. Pulmonary hypertension in such cases is often misdiagnosed as CTEPH. Pulmonary artery sarcoma generally develops in the main or central pulmonary arteries, often in close relationship to the pulmonary valve. On contrast-enhanced CT scans, the tumor appears as an enhancing intraluminal soft-tissue mass, unlike apposition thrombi, which are nonenhancing (4). Pulmonary artery sarcomas often fill the vessel lumen completely. They can extend into the contralateral pulmonary artery or beyond the pulmonary valve into the right ventricle. Local pulmonary and bronchial invasion can occur, and the lung is frequently affected by metastasis (Fig 19).

Pulmonary Venous Hypertension

Pulmonary Venous Hypertension Associated with Cardiac Disease

Pulmonary venous hypertension is most frequently caused by left-sided cardiac disorders such as left ventricular failure, its most common cause both in adults and in children. Less frequent causes are left atrial thrombus, left atrial neoplasm (myxoma, sarcoma, or metastasis), mitral stenosis, and congenital cardiac anomalies (1,7). Left-sided cardiac disease may display imaging features similar to those of pulmonary veno-occlusive disease. The distinguishing feature is left atrial enlargement, which is frequently seen in left-sided cardiac disease but not in pulmonary veno-occlusive disease (1,71).

Pulmonary Veno-occlusive Disease

Pulmonary veno-occlusive disease is histologically characterized by organized and recanalized thrombi and eccentric intimal fibrosis in the pulmonary veins and venules (1,3,72). The condition primarily affects children, among whom it is distributed equally between the sexes, and young adults, among whom it is distributed with a slight male predominance (2,13). Its etiology is unknown. Clinical associations with high estrogen levels during pregnancy, oral contraceptive use, viral infection, bone marrow transplantation, and drug (eg, bleomycin, mitomycin) toxicity have been described (1,4,73). Patients present with slowly progressive dyspnea and episodes of acute pulmonary edema, sometimes complicated by hemoptysis (4). Hemodynamic measurements reveal normal or variably elevated capillary wedge pressures combined with the characteristic hemodynamic derangements of pulmonary arterial hypertension, without increased left atrial and left ventricular pressures.

For clinicians, the distinction between pulmonary veno-occlusive disease and pulmonary arterial hypertension is particularly important because the administration of vasodilator therapy for presumed pulmonary arterial hypertension may induce potentially fatal pulmonary edema (1). At CT, the combination of features of pulmonary arterial hypertension with interstitial and alveolar edema is virtually diagnostic of pulmonary veno-occlusive disease. CT scans show markedly small central pulmonary veins, interlobular septal thickening, and patchy centrilobular ground-glass opacities representing interstitial and alveolar edema (Fig 20). Additional CT findings include dilatation of the central pulmonary arteries, right ventricular enlargement, a normal-sized left atrium, pleural effusion, and mediastinal lymphadenopathy (1,71,74–76). The diagnosis is based on a combination of anatomic findings at CT and hemodynamic findings at cardiac catheterization with conventional angiographic evidence of delayed filling of small pulmonary veins; however, histologic analysis may be necessary to confirm the diagnosis (13,71–73).

Conclusions

Pulmonary hypertension is a frequent clinical diagnosis associated with high patient morbidity and mortality. Diseases that can induce pulmonary hypertension display a wide spectrum of partially overlapping CT features, and definitive diagnosis may require correlation of CT imaging findings with clinical, histopathologic, and angiographic findings. To allow appropriate therapeutic management, awareness of the various disease entities associated with pulmonary hypertension and knowledge of the entire spectrum of their imaging features are essential.