The Core Curriculum: Cardiopulmonary Imaging, 1st Edition (2004)

Chapter 17. Radiology of the Pleura

Evaluation of the pleura is in several important ways the evil twin of evaluation of the mediastinum. Whereas plain film assessment of the mediastinum is extremely limited, chest radiograph (CXR) evaluation of the pleura is generally efficacious. Although computed tomography (CT) is a phenomenal tool for assessing mediastinal anatomy and abnormality, pleural CT is often requested for indications of questionable validity. In this chapter we discuss typical pleural abnormalities, their differential diagnosis, and applications of CT.

Pleural Abnormalities

The major pleural findings of abnormality are effusion, pneumothorax, focal or multifocal pleural implants or plaques, and solitary or multiple pleural masses.

Pleural Effusion

As discussed in Chapter 3, pleural effusion is typically gravity dependent. It accumulates in the most dependent portion of the chest, varying with changes in patient position. In the presence of pleural adhesions or scars, loculated effusion may occur. Pleural fluid may track into pleural fissures, even when not loculated. Although this is often easily recognizable based on its tapering cigar-shaped appearance (Fig. 17.1), it sometimes results in confusing appearances.

On standard frontal and lateral radiographs, the posterior costophrenic sulcus is the place to look for small pleural effusions. As effusion enlarges, it may blunt the lateral costophrenic angles, or it may accumulate in a subpulmonic location. Neophyte interpreters of radiographs seem to overcall pleural effusions on frontal radiographs more than any other finding. They are quick to assume that any basilar opacity on a portable radiograph indicates effusion. Most such “effusions” actually represent basilar atelectasis or epicardial fat or cardiomegaly on a hypoventilatory anteroposterior CXR. Our advice is to suggest pleural effusion if there are findings of effusion but to recognize that we cannot expect to diagnose everything with 100% accuracy on limited portable radiographs. In other words, try not to envision something simply because you are afraid of missing it.

As pleural effusions are usually gravity-dependent, the posterior costophrenic sulcus is the place to look for small effusions.

Generally, the biggest challenge in the diagnosis of pleural effusion is to recognize its presence on supine or semisupine radiographs. In that situation, fluid layers posterior to the lung. Because it is visualized en face, there is no sharp edge to outline the margin of effusion against normal pleura or lung. As a consequence, vague opacity is present, typically worst at the bases of the hemithoraces and becoming gradually less obvious as the eye travels cephalad. In the presence of large supine effusion, fluid will typically “cap” over the apex of the hemithorax (Fig. 17.2).

Figure 17.1 Pseudotumors of fluid in fissures. A. Posteroanterior chest radiograph: right pleural effusion tracks into major and minor fissures, with typical cigar-shaped appearance of fluid in minor fissure (arrows)B. Lateral chest radiograph: similar cigar-shaped appearance of fluid in major fissure (arrowheads) is better appreciated (arrows indicate fluid in minor fissure).

If it is important to know whether there really is pleural effusion in such a patient, solutions are available. As previously discussed inChapter 3, the first step is to obtain lateral decubitus radiography; the side of suspected effusion should be the down side. Pleural ultrasonography is an alternative, particularly when a small effusion is suspected, because ultrasonography can be used to guide thoracentesis as well. CT is not a reasonable alternative simply for detection of pleural effusion.

Figure 17.2 Supine left pleural effusion. Increased opacity throughout the left hemithorax with continued visualization of left lung vessels. Note fluid capping left apex (arrows).

A potentially important effect of large pleural effusions is diaphragmatic inversion. In the presence of diaphragmatic inversion, the two hemidiaphragms move in opposite directions during respiration. As a consequence, air may move back and forth between the two lungs (pendelluft, or pendulum respiration) (1). The net effect is markedly increased dead space and significant dyspnea. Thoracentesis with removal of enough fluid to restore the affected hemidiaphragm to its normal position may result in marked relief of symptoms. Left hemidiaphragmatic inversion is more easily diagnosed because it results in mass effect in the left upper quadrant, displacing the gas-filled stomach (Fig. 17.3).

Large effusions may result in hemidiaphragmatic inversion, with significant dyspnea.

The other potentially important effect of diaphragmatic inversion is that it changes expected anatomic relationships that assist localization of peridiaphragmatic fluid at CT (2). In a patient with a normal convex-upward hemidiaphragm, fluid visualized outside the diaphragmatic outline is supradiaphragmatic (pleural) and fluid visualized inside the diaphragmatic outline is subdiaphragmatic (peritoneal) (2). In the presence of diaphragmatic inversion, these relationships are reversed.

In hospitalized patients, pleural effusions are most commonly seen with congestive heart failure or as a consequence of surgery. The stigmata of recent adjacent surgery (in the chest or in the upper abdomen) will usually be evident or might (rarely) be indicated in the provided clinical history for performance of the radiograph. If effusions are a result of congestive heart failure, they are typically bilateral and reasonably symmetric. Congestive heart failure effusions are occasionally unilateral and not infrequently somewhat asymmetric; in most such patients, right pleural effusion is larger.

When there is unilateral left pleural effusion (or left effusion is significantly larger than right effusion) in a nonsurgical patient, other etiologies should at least be considered (Box 17.1). For differential diagnostic purposes, in this situation you might run through categories of disease. First, because it is the category you do not want to forget, is neoplasm. Pleural metastases are far more common than pleural primary neoplasms. Many primary neoplasms metastasize to the pleura. Those closest to the pleura (particularly lung and breast cancers) are especially likely, but in a given patient the most likely source of pleural metastases is that patient’s neoplasm. In other words, even a primary neoplasm that is not frequently associated with pleural metastases deserves serious consideration if a patient with that neoplasm develops pleural effusion. Lymphoma is also in the differential diagnosis, especially if there are multiple enlarged thoracic lymph nodes.

Figure 17.3 Effusion inverting left hemidiaphragm. A. Posteroanterior chest radiograph: typical left pleural effusion (E) with apparent left upper quadrant abdominal mass displacing stomach inferomedially (arrows)B. Posteroanterior chest radiograph after thoracentesis: stomach (S) in a more normal position, with no left upper quadrant mass.

Box 17.1: Causes of Pleural Effusion

Congestive heart failure


Neoplasm—metastatic, primary

Inflammatory disease—infection, autoimmune

Abdominal disease—trauma, infection, inflammation, liver failure

Vascular—pulmonary embolus

Trauma—motor vehicle accident, stab or bullet wound, iatrogenic (catheter or feeding tube malposition)

Among abdominopelvic neoplasms, ovarian carcinoma deserves special mention. It frequently spreads to the peritoneum, and pleural effusion in a patient with malignant ascites sometimes reflects leakage of fluid through the diaphragm. Pleural metastases from ovarian carcinoma do occur; they are commonly referred to as Meig syndrome. In fact, the original description of Meig syndrome referred to the association of pleural effusion with benign ovarian fibroma (Fig. 17.4). Over time, this term has become less precise and is now usually applied to any combination of pleural effusion and ovarian mass. Patients with pleural metastases from any primary may have unilateral or bilateral effusions.

Malignant mesothelioma is the most common pleural primary neoplasm associated with pleural effusion, and most patients with malignant mesothelioma have pleural effusion (Fig. 17.5). Effusion is not usually an isolated finding in this setting. Instead, there is usually also lobulated pleural mass. A history of asbestos exposure can be a helpful clue, but well-documented asbestos exposure is only established in about half of patients with malignant mesothelioma. In some patients lack of mediastinal shift away from the side of extensive pleural abnormality is a clue to the diagnosis of malignant mesothelioma (Fig. 17.6). Metastatic adenocarcinoma may closely mimic the appearance of malignant mesothelioma. Pleural effusions in patients with malignant mesothelioma are often unilateral.

Most patients with malignant mesothelioma have pleural effusion.

Figure 17.4 Meig syndrome. A. Right lateral decubitus radiograph: layering right pleural effusion (E). B. Pelvic computed tomography: large solid pelvic mass (F) represents an ovarian fibrothecoma.

Figure 17.5 Malignant mesothelioma. A. Scout chest radiograph: large left pleural effusion (E) without right shift of the mediastinum.B. Computed tomography: effusion (E) is accompanied by solid pleural mass (M) and implants (arrows)C. Computed tomography more cephalad than B: additional pleural mass (M) adjacent to aortopulmonary window.

Inflammatory disease is the next category to consider. Under this heading, infection is an important consideration. Empyema can occur as a consequence of any pneumonia. Some organisms, such as M. tuberculosis and fungi, have a particular predilection for the pleural space (Fig. 17.7). The differentiation of empyema from lung abscess can generally be accomplished by CXR alone. Empyema typically has right or obtuse angles with the adjacent chest wall, whereas lung abscess usually demonstrates acute angles. Empyema is usually lenticular in shape and therefore is much larger on one of two right-angle projections, whereas lung abscess is more spherical in shape and more similar in size on right-angle projections. In selected instances, CT can aid in this distinction by demonstrating that empyema is more mass-like, deflecting vessels and bronchi in its path, whereas lung abscess is more destructive of lung structures but less mass-like. Smoother margins of pleural abnormality and the CT “split pleura sign” of enhancing visceral and parietal pleura around an empyema can also aid in the differential diagnosis (3) (Fig. 17.8).

CT signs of empyema include mass-effect on adjacent lung structures, smooth margins, lenticular shape, obtuse or right angles with lung parenchyma, and the “split pleura sign.”

Figure 17.6 Malignant mesothelioma. Despite extensive right pleural abnormality (M), mediastinum is actually shifted to the right.

However, CXR and CT alone cannot establish the presence or absence of infection in the lung or pleural space. Clinical findings (such as fever and elevated white blood cell count) are important in raising the possibility of infection. The role of imaging is to localize the infection to lung or pleura. The final etiology of pleural effusion is sometimes established on clinical grounds alone. In difficult cases, diagnosis generally relies on thoracentesis and analysis of pleural fluid; patients with lung infection commonly have uninfected pleural fluid (a so-called sympathetic effusion). The need for thoracentesis applies equally well to pleural neoplasm and to most other etiologies of pleural effusion.

Thoracentesis is often required to establish an etiology for pleural effusion, whether effusion is benign or malignant.

Inflammatory pleural effusion also includes a variety of autoimmune and related disorders. Among collagen vascular diseases, pleural effusion is particularly likely in systemic lupus erythematosus (Fig. 17.9) and rheumatoid arthritis. In fact, pleural effusion is the most common thoracic manifestation of rheumatoid arthritis. As with all extraarticular manifestations of rheumatoid arthritis, pleural effusion is far more common in men than in women. Effusion is also seen in patients with an autoimmune response after myocardial infarction (Dressler syndrome) or cardiac surgery. Pleural effusions in this category may be unilateral or bilateral, symmetric or asymmetric.

Figure 17.7 Tuberculous empyema (E).

Figure 17.8 Empyema or lung abscess? A. Abnormality (E) is compressing and displacing vessels and bronchi in the right lung, typical of empyema. B. Abnormality (A) extends right up to undeviated vessels (arrows), typical of abscess. C. Abnormality (A) has thick irregular wall, typical of abscess. D. Bilateral abnormalities (E) with smooth thin walls and separation of enhancing visceral and parietal pleura (“split pleura sign”) (arrows), typical of empyemas.

Next for consideration are abdominal etiologies of pleural effusion. Adjacent upper abdominal abnormalities can cause effusion. Common etiologies in this category are splenic trauma, subdiaphragmatic abscess (Fig. 17.10), pancreatitis, and ascites. History is obviously very helpful, and thoracentesis can be very revealing (e.g., if there is elevated pleural fluid amylase in a patient with pancreatitis). Patients with ascites, such as those with liver failure, develop pleural effusions for several reasons, including leak of ascitic fluid through the diaphragm and anasarca as a result of decreased serum albumin. When right subdiaphragmatic abscess is an important concern, ultrasound can be very helpful. Because referring clinicians remain uninformed about ultrasound’s capabilities even at this relatively late date, they instead virtually always order CT, which is in fact better than ultrasound for left upper quadrant fluid collections. Most effusions resulting from abdominal disease are unilateral.

Figure 17.9 Systemic lupus erythematosus with right pleural effusion (E).

Another category to consider is vascular disease. In particular, pulmonary embolus is extremely difficult to diagnose via CXR. Most patients with pulmonary embolus have a normal CXR. When abnormality is present, it is often nonspecific; pleural effusion and atelectasis are the most common findings. Pleural effusion in pulmonary embolus can be unilateral or bilateral, because emboli are commonly multiple.

Trauma can result in pleural effusion. There may be obvious trauma, as in patients with multiple rib fractures. A somewhat less obvious example of trauma would be a patient with aortic laceration in the absence of fractures. An even less obvious example would be a malpositioned vascular catheter or feeding tube (4) (Fig. 17.11). Inadvertent placement of a feeding tube into the lung is usually benign as long as it is discovered before a feeding is administered. However, pleural placement is not similarly benign. Tension pneumothorax and empyema are among its important consequences. Traumatic pleural effusion is more commonly unilateral than bilateral.

Figure 17.10 Left subphrenic abscess. A. Posteroanterior chest radiograph: left pleural effusion (E) with extraintestinal left upper quadrant gas (arrows)B. Computed tomography: large subdiaphragmatic gas and fluid collection is subphrenic abscess (A).

Figure 17.11 Abdominal radiograph to show misplacement of feeding tube. A. Abdominal radiograph: Feeding tube tip (T) not where it should be. B. Anteroposterior chest radiograph after removal of feeding tube from right pleura: lung edge (arrows) displaced by large pneumothorax. C. Right lateral decubitus radiograph: air and fluid in right pleura, indicating hydropneumothorax.

Congenital and miscellaneous causes probably account for some pleural effusions (such as in lymphangiectasia and in lymphangiomyomatosis). They are sufficiently uncommon causes of pleural effusion that they are not discussed further here.


Pleural air tends to accumulate in the nondependent portion of the pleural space. Upright frontal radiography is generally the preferred technique for detecting pneumothorax. As discussed in Chapter 3, lateral decubitus radiography with the suspected side of pneumothorax up is a perfectly acceptable alternative; in experiments with cadavers (5) it was even sometimes better than upright views for detecting pleural air.

Because air accumulates in the non-dependent pleural space, upright or lateral decubitus radiographs are preferred for demonstration of pneumothorax.

The diagnosis of pneumothorax is best made by visualization of the lung edge outlined by pleural air (Fig. 17.12). Apical lucency and absence of vessels are far less reliable criteria that can be produced by bullae, for example. As with diagnosis of pleural effusion, detection of pneumothorax becomes trickier when the patient is supine. In that setting the nondependent portion of the pleural space is near the hemidiaphragm. Lucency near the lung bases raises concern for pneumothorax, particularly in the presence of the deep sulcus sign (Fig. 17.13), even when no lung edge is visualized. Air in the minor fissure has been reported as another sign of supine pneumothorax (Fig. 17.14). Just as pleural effusion can track into fissures, pneumothorax can be visualized in the major fissure (Fig. 17.12) or even in accessory fissures (Fig. 17.15).

Figure 17.12 Moderate left pneumothorax. Lung edge is visualized (arrows), and gas extends into the major fissure (arrowheads).

The differential diagnosis for pneumothorax is far less extensive than that for pleural effusion (Box 17.2). Most cases are clearly related to some form of trauma or else occur secondary to ruptured blebs (Fig. 17.16). The latter is the usual explanation for most spontaneous pneumothoraces and typically occurs in asthenic young males. Interstitial lung disease is another potential explanation for the development of pneumothorax. Any patient with honeycombing or lung cysts is at risk for the development of pneumothorax, but it is particularly likely in males with eosinophilic granuloma (Fig. 17.17) and in females with lymphangiomyomatosis.

Figure 17.13 Supine left pneumothorax. A. Anteroposterior chest radiograph: deep sulcus sign (arrows)B. Anteroposterior chest radiograph the next day: larger left pneumothorax outlines lung edge (arrows).

Figure 17.14 Supine right pneumothorax. Air tracks into minor fissure (arrows).

Figure 17.15 Right pneumothorax with air in azygos fissure (arrows).

Box 17.2: Causes of Pneumothorax

Ruptured bleb or bulla

Trauma (penetrating or with rib fracture)

Interstitial lung disease


Cavitary metastases (squamous, sarcomas)

Increased intrathoracic pressure

Figure 17.16 Ruptured bleb (B) outlined by resultant pneumothorax (P).

Pneumothorax can develop in any patient with chronic interstitial lung disease, but is particularly likely in males with eosinophilic granuloma and in females with lymphangiomyomatosis.

In patients without interstitial lung disease, trauma, or ruptured blebs, there are other potential causes to consider, many with normal or nearly normal CXRs. Subpleural abnormalities that tend to cavitate are one group of causes, including tuberculosis and cavitary metastases. Overall, squamous primary neoplasms account for most cavitary metastases, but sarcomas actually have a higher predilection to cavitate (Fig. 17.18). Increased intrathoracic pressure (such as in asthmatics and in pregnant patients) is another potential explanation for pneumothorax with an otherwise normal or nearly normal CXR. A rare cause is pleural endometriosis, which may explain cyclical pneumothoraces or hemothoraces. Knowledge of this rare entity may allow the enlightened visiting professor to score a major coup when presented with an unknown radiograph; sometimes things do not work out quite as well (Fig. 17.19).

Figure 17.17 Eosinophilic granuloma with upper lobe honeycombing causing right apical pneumothorax. Lung edge outlined by arrows.

Figure 17.18 Cavitary metastatic osteosarcoma with resultant pneumothorax. A. Computed tomography of cavitary left upper lobe nodule (M). B. Computed tomography more caudal than A: other lung nodules (M) and moderate left pneumothorax (P).

Figure 17.19 Cyclical hemothorax. Based on the history I immediately suggested pleural endometriosis. A. Posteroanterior chest radiograph: moderate left pleural effusion (E). B. Computed tomography: abnormal left posterior chest wall vessels (arrows) with pleural extension (arrowheads), indicating chest wall hemangioma as cause of hemothoraces.

Clinicians are frequently interested in knowing the percentage of the hemithorax occupied by a pneumothorax. It is hard to imagine that anyone can estimate such a percentage accurately. The chest is a complex three-dimensional structure, and many patients with suspected pneumothorax only have frontal radiographic evaluation. Instead, it is preferable to measure the displacement of the lung edge from the inner margin of the adjacent rib, often (but not always) at the lung apex. For persistent clinicians, a gestalt of small, moderate, or large pneumothorax may be helpful. Serial CXR assessment to determine whether pneumothorax is stable, resolving, or increasing is probably of even greater value.

Pleural Implants or Plaques

CT is seldom necessary to demonstrate pleural fluid or air. Small amounts of fluid or air may be better detected by CT when positional views cannot be obtained, but this is an unusual circumstance. However, pleural implants or plaques are often far better seen with CT than on CXR. Rarely, the situation is reversed. In one memorable instance, a pleural metastasis from malignant thymoma was questioned twice from CXRs and overlooked twice on resultant CTs. The pleural implant was finally seen by the interpreter of the patient’s third CT, and it was also present in retrospect on the previous scans.

Figure 17.20 Asbestos-related pleural plaques. A. Posteroanterior chest radiograph: peculiar opacities, especially in left hemithorax lateral to left hilum. Calcified right lung granuloma (G) incidentally noted. B. Lateral chest radiograph: plaques are often difficult to visualize on the lateral radiograph, although calcified diaphragmatic plaque (arrow) is classic for asbestos exposure. C. Computed tomography in a different patient: multiple calcified and noncalcified pleural plaques (arrows).

Pleural plaques are common in patients exposed to asbestos (Fig. 17.20). The skilled interpreter knows not to refer to this as asbestosis. Diagnosis of asbestosis implies the presence of interstitial lung disease; pleural plaques alone only indicate asbestos exposure. Plaques are usually bilateral and are sometimes visibly calcified. In fact, a very characteristic appearance of asbestos exposure is calcified plaques along the diaphragmatic pleura (Fig. 17.20B). Because many plaques are visualized en face, the resultant radiographic appearance is often that of irregularly shaped opacities that are poorly marginated. One of our pulmonologists who was a surprisingly skilled interpreter of chest radiographs used to say that the more bizarre the CXR appearance, the more likely it indicated pleural plaques.

Unilateral pleural plaques can also be seen with asbestos exposure. However, in this setting the alternative possibilities of old empyema and old hemothorax should be raised. Tuberculosis in particular may create a characteristic appearance of a small hemithorax with extensive pleural calcification, a so-called fibrothorax (Fig. 17.21).

The more bizarre the CXR appearance, the more likely it represents pleural disease, especially pleural plaques.

Figure 17.21 Fibrothorax resulting from tuberculosis. A. Posteroanterior chest radiograph: small left hemithorax with calcified pleural abnormality (F). B. Computed tomography viewed at soft tissue window settings: heavily calcified pleural abnormality (F) with adjacent left lung bronchiectasis (arrows)C. Bone window photography of B: better demonstration of internal composition of pleural abnormality.

Pleural implants may be seen with metastatic disease. More commonly, metastatic disease causes pleural effusion. However, even in that case CT may reveal soft tissue implants, suggesting the malignant nature of the effusion. Pleural hematoma may simulate the CT appearance of malignant effusion. Malignant thymoma often causes unilateral multifocal pleural implants, generally not associated with effusion.

An unusual cause of left pleural implants (except at the Radiological Society of North America [RSNA] film panel, where it is surprisingly common) is splenosis (Fig. 17.22). After left upper quadrant trauma, pieces of spleen may implant along the left pleural surface. In such a situation, surgical clips in the left upper quadrant and evidence of old left lower rib fractures constitute important diagnostic clues. Radionuclide scanning with damaged red blood cells may clinch the diagnosis.

Figure 17.22 Thoracic splenosis. A. Posteroanterior chest radiograph: left infrahilar mass (S). B. Lateral chest radiograph: mass (S) is extraparenchymal, with right angles at the chest wall (arrows)C. Computed tomography: extraparenchymal shape of mass (S) confirmed. D. Computed tomography of upper abdomen: key additional finding is absence of spleen after trauma.

Pleural Mass(es)

Many pleural masses are obscured on the CXR (and even at CT) by surrounding pleural effusion. CT is generally better able to distinguish solid elements of extensive pleural mass from effusion (Fig. 17.5), as in a typical malignant mesothelioma. However, not all pleural masses are accompanied by effusion.

Benign fibrous tumor of the pleura (previously known as benign mesothelioma) often presents as a single pleural mass. Tumor size runs the gamut from small to enormous. Lesions may be pedunculated, and as a consequence the tumor may change position in the thorax by twisting around the pedicle (Fig. 17.23). Even quite large benign fibrous tumors are generally distinct from malignant mesothelioma in their focal nature (malignant mesothelioma is typically relatively widespread in the pleura) and in the lack of associated pleural effusion. Clinical clues are sometimes also useful in diagnosing benign fibrous tumor. Lesions may present with associated hypertrophic pulmonary osteoarthropathy, and hypoglycemia is also sometimes associated with these tumors.

Benign fibrous tumors are generally easily recognized as pleural lesions by their shape. The situation is more confusing when a benign fibrous tumor occurs in a pleural fissure (6) (Fig. 17.24). In that circumstance the lesion will often appear to be round and will be sharply marginated on all sides. Such a lesion will appear to be a noncalcified lung nodule, in which case malignancy would be a significant concern. CT localization of the “nodule” to the pleural fissure makes benign fibrous tumor the overwhelmingly likely diagnosis.

Benign fibrous tumor of pleura is generally easily recognizable as a pleural mass, but intrafissural fibrous tumors may closely simulate lung nodules.

Another benign etiology is pleural lipoma (Fig. 17.25). Intuitively, it would seem that a fatty lesion would obviously be less opaque on the CXR than a solid lesion. Except with enormous lesions, this turns out not to be true. Only at CT is the fatty nature of such a lesion usually evident.

Multiple pleural masses may reflect growth of any of the lesions described above in multiple implants. In addition, it may be very difficult to distinguish pleural from extrapleural masses (Box 17.3). A mass in either location often demonstrates sharp margination but only for part of its circumference (Figs. 17.26 and 17.27). The shape of the lesion is not always conclusive in the determination of its location. Demonstrable bone destruction indicates extrapleural masses, but otherwise they should be lumped together as extraparenchymal lesions. Common causes of extrapleural masses include metastatic disease, multiple myeloma, and trauma. This brief list is far from exhaustive (Fig. 17.28).

Figure 17.23 Mobile benign fibrous tumor of pleura. A. Posteroanterior chest radiograph in 1985: large mass (P) in upper right hemithorax. B. Posteroanterior chest radiograph in 1986: even larger mass (P) is now in lower right hemithorax. Lung torsion is another uncommon cause of thoracic lesions that change position. (Courtesy of Dr. Phil Templeton, Baltimore, MD.)

Figure 17.24 Fissural benign fibrous tumor of pleura. A. Posteroanterior chest radiograph: small right upper thoracic nodule (N). B.Computed tomography: lesion (N) appears to be round and smooth, suggesting that it is in the lung, but closer inspection shows that it is centered at the avascular plane of the major fissure (arrows).

Figure 17.25 Pleural lipoma (L).

Box 17.3: Common Extrapleural Abnormalities


Multiple myeloma


Fibrous dysplasia

Primary neoplasm (especially sarcoma)

Figure 17.26 Extrapleural metastatic melanoma. A. Posteroanterior chest radiograph: mass (M) with sharp superomedial border(arrows) that fades inferiorly. B. Lateral chest radiograph: mass (M) with sharp posterior border (arrows) but less distinct margin elsewhere. C. Computed tomography: lesion is a destructive rib metastasis (M).

Figure 17.27 Extrapleural metastatic esophageal carcinoma. A. Posteroanterior chest radiograph: mass (M) with sharp inferior border but otherwise indistinct margins. B. Computed tomography: metastasis (M) to right first rib.

Figure 17.28 Chest wall chondrosarcoma. A. Posteroanterior chest radiograph: mass (C) shows hilum overlay sign and questionable internal matrix. B. Lateral chest radiograph: matrix is better displayed (arrows). Margins are typical of extraparenchymal location. C.Computed tomography: obvious chest wall extent and abundantly calcified matrix (M).


Selected Topics in Radiology of the Pleura and Chest Wall

CT has been used to image a wide variety of pleural and chest wall abnormalities. Because of its superior contrast resolution and the advantages of cross-sectional display of anatomy, it is quite clear that CT can image such lesions. The more important issue of whether CTshould image various pleural and chest wall abnormalities is generally not addressed. In unselected cases CT makes little or no real contribution to the diagnostic workup. The CT appearance is often not specific for a given abnormality. Pleural abnormalities are often well seen on conventional radiographs (throughout this chapter) and are frequently diagnosed via thoracentesis or pleural biopsy. Chest wall abnormalities are often easily accessible to physical examination and, if necessary, biopsy.

In this section we discuss normal anatomy of the pleura and chest wall and address specific issues in pleural and chest wall CT. In particular, areas where CT is especially useful are highlighted, as well as those where it does not make a major contribution.

Normal Anatomy and Anatomic Variants

Major, Minor, and Variant Fissures

CT identifies the major fissures with a frequency approaching 100%. Detection of the minor fissure is variable, ranging from 50% to 100% (7,8). More correctly, it is the parafissural lung that allows localization of the fissures with routine CT. Subpleural lung is composed primarily of secondary pulmonary lobules containing sparse and small-caliber vasculature. This relatively avascular lung is imaged at CT as a lucent band, which is the most common CT fissural appearance (Fig. 17.24B). When imaged perpendicular to the scan plane, the fissure is a thin line. Visualization as a line, which occurs more commonly on the left because that major fissure is more vertical, is frequent with thin section CT, where there is less volume averaging. Occasionally, the fissure appears as a dense band, probably because of partial volume effect or respiratory motion. A double fissure occurs with motion artifact, particularly cardiac motion at the left lung base.

The minor fissure is horizontal between the right upper and middle lobes, about 3 to 4 cm below the origin of bronchus intermedius. Given a parallel minor fissure and axial scanning plane, this fissure is seldom a line, even with thin section CT. Most often the minor fissure is a triangular avascular area, with its apex at the hilum. When the superior right middle lobe is convex, the superior minor fissure appears as a round or oval avascular zone. On scans through a convex superior middle lobe, lung anterior to the fissure is anterior segment of right upper lobe, and lung posterior to the fissure is superior segment of lower lobe. Bronchopulmonary segmental anatomy aids the differentiation of right upper and middle lobes. The upper lobe bronchi course lateral to their arteries, whereas the middle lobe bronchi course medial to their arteries (9). This relationship is constant even with volume loss.

Accessory fissures are more commonly visualized anatomically than radiographically. They can be demonstrated on CXR and CT. Most common are the superior and inferior accessory fissures. Other variants include azygos, hemiazygos, and left minor fissures.

The superior accessory fissure is between the superior segment of a lower lobe and the remaining lower lobe segments. It is more common on the right and is found in 5% to 30% of autopsy specimens (10). Because it is usually horizontal, the superior accessory fissure appears to be similar to minor fissure on frontal CXR, except that it is more caudal. On the lateral projection, this fissure extends posteriorly from the major fissure (Fig. 17.29). It was recognized by Proto and Speckman (11) in 6% of lateral radiographs. On CT, the superior accessory fissure is more caudal and posterior than the minor fissure. This appearance can be simulated by a horizontal major fissure, resulting from lower lobe volume loss (10). When slightly oblique, the superior accessory fissure can resemble the major fissure.

Figure 17.29 Superior accessory fissure (arrows) is posterior and caudal to minor fissure (arrowheads).

The inferior accessory fissure separates the medial basal segment from the remaining lower lobe basilar segments. It also occurs more commonly on the right. Although this fissure is present in 30% to 50% of autopsy specimens, it is radiographically demonstrable in only 5% to 10% of patients (10). The inferior accessory fissure appears on frontal and lateral radiographs as a thin vertical linear opacity that originates near the medial diaphragm, is posterior to the major fissure, and courses obliquely toward the hilum. The CT appearance depends on whether routine or thin sections are obtained, but when visible the fissure extends from the inferior pulmonary ligament to the major fissure (10).

The inferior accessory fissure is present in 30%-50% of autopsy specimens, but is demonstrable radiographically in only 5%-10% of patients.

A left minor fissure, which is found in 8% to 18% of anatomic specimens, separates the lingula from the remaining left upper lobe (12). Radiographs demonstrate the left minor fissure in less than 2% of patients (Fig. 17.30). Its CT appearance is analogous to the right minor fissure, although it occasionally occurs more cephalad (11).

Figure 17.30 Left minor fissure (arrows).

Figure 17.31 Drawing of typical and sagittal minor fissures. (From 

Gross BH, Spizarny DL, Granke DS. Sagittal orientation of the anterior minor fissure: radiography and computed tomography. Radiology 1988;166:717–719

, with permission.)

The azygos fissure is formed by four layers of pleura invaginating into the lung apex and extending to a laterally displaced azygos vein. It occurs in 1% of anatomic specimens and is demonstrated radiographically (Fig. 17.15) in 0.5% of patients (13). Almost all cases occur on the right, with a few reports of left hemiazygos fissures connecting to left superior intercostal vein. Given its vertical orientation, this fissure is often recognized on CT as an arcuate linear opacity extending from the posterolateral aspect of an upper thoracic vertebral body to the superior vena cava or right brachiocephalic vein. A nodular posterior azygos fissure can occur secondary to an intraparenchymal azygos vein.

A sagittal anterior minor fissure is a normal variant that can result in misleading appearances of right upper and middle lobe disease (14). Occurring as an oblique vertical linear opacity in the inferomedial right lung, the sagittal minor fissure results in inferomedial extension of right upper lobe (Fig. 17.31). As a result, the anterior segment of upper lobe occupies a paracardiac location and can border the diaphragm (14). Right upper lobe disease with a sagittal minor fissure can silhouette the heart (Fig. 17.32), whereas medial segment right middle lobe disease can spare the cardiac border. In either event, erroneous localization of disease may occur.

Figure 17.32 Confusing appearance of right upper lobe airspace disease (A) in patient with sagittal minor fissure. Caudal extent of abnormality and silhouetting of right heart border erroneously suggest middle lobe disease.

Figure 17.33 Vertebral body types for counting ribs. A. Lowest thoracic vertebra, characterized by medial ribs (arrows)B. Highest lumbar vertebra, with transverse processes (arrows) rather than ribs.

Counting Ribs

Counting ribs (15) seems like a trivial subject until a CT is presented for evaluation of a bone scan abnormality of the eighth rib in a patient with normal rib radiographs. The value of precise anatomic localization of that rib becomes readily apparent. A step-by-step approach is as follows:

·      Step 1: The first rib is located on image with mid-clavicles.

·      Step 2: The next two or three ribs are counted on same slice.

·      Step 3: Progressively lower ribs are located at the costovertebral junctions.

·      Step 4: With bilateral lesions, each side is counted separately.

For lesions in the lower ribs, an alternative is to find the first lumbar-type vertebra (Fig. 17.33) and then count upward at the costovertebral junctions.

Issues in Pleural and Chest Wall Computed Tomography

Chest Wall Invasion in Lung Cancer

CT evaluation of chest wall invasion by lung masses is not very accurate. CT in 33 patients with peripheral pulmonary malignancies contiguous with a pleural surface was reviewed for useful signs in diagnosing chest wall invasion (16). Criteria that were evaluated included pleural thickening adjacent to the tumor, encroached upon or high attenuation extrapleural fat, asymmetric extrapleural soft tissues, apparent mass invading chest wall, and rib destruction. Individual CT criteria were either very sensitive but not specific (local pleural thickening) or very specific but not sensitive (all other criteria). Overall sensitivity was 38%, with a specificity of 40%. In one series (17), focal chest pain was much more specific than any CT signs. Magnetic resonance imaging may be more accurate than CT (18).

CT evaluation of chest wall invasion by lung masses is not very accurate; focal chest pain may be more specific than any CT signs.

It should be noted that preoperative staging of chest wall involvement in lung cancer is not as important as assessment of mediastinal involvement. Local chest wall invasion does not contraindicate surgical resection, although it does necessitate en bloc resection of the contiguous chest wall. In light of the high morbidity of this procedure, we suggest chest wall invasion whenever we see local pleural thickening, thereby alerting the surgeon that en bloc resection may be necessary.

Lung Cancer Crossing Fissures

The double layer of fissural pleura creates a barrier to the spread of inflammation and neoplasm. This is especially true when lobar separation by fissures is complete. However, anatomic studies demonstrate fusion across part of right major fissure in approximately 70% of specimens, across minor fissure in more than 90% of specimens, and across left major fissure in 40% to 50% of specimens (19). Lobar fusion allows for disease spread that would otherwise be confined to one lobe or segment. When non-neoplastic disease crosses an intact fissure, it is commonly an atypical infection. Actinomycosis is especially likely to ignore anatomic barriers (Fig. 17.34), but blastomycosis, nocardiosis, and other fungal infections are also possible. Neoplasm is less likely to cross an intact fissure, but lymphoma has a predilection for crossing anatomic barriers (20).

In lung cancer patients who are candidates for lobectomy, extension of neoplasm anywhere across the left major fissure or across the right major fissure cephalad to the minor fissure necessitates pneumonectomy. CT assessment of transfissural crossing is up to 100% specific when correlated with pathologic findings but only 50% sensitive when standard scanning is performed (21). Thin section CT probably increases sensitivity.

Thin section CT scanning increases sensitivity for detecting extension of lung cancer across fissures, a finding that requires more extensive surgery if curative resection is attempted.

Malignant Effusion and Mesothelioma

The CT findings of malignant mesothelioma are well known. The typical diffuse, multifocal, irregular pleural mass is generally also well demonstrated by CXR. Final diagnosis typically requires biopsy of the lesion. Therefore, CT is not crucial for diagnosis in most patients. As for staging, the prognosis remains so poor for patients with malignant mesothelioma and for those with pleural metastases that the usefulness of CT staging remains doubtful. In fact, extent of pleural mesothelioma may be underestimated at CT because portions of tumor may be difficult to distinguish from pleural effusion.


Clinicians will often request chest CT in patients with a large pleural effusion of unknown etiology. CT is unfortunately as limited in value as CXR in this setting. Although clinicians believe that CT will better evaluate the compressed underlying lung for lung cancer, this is seldom true. Clinicians occasionally offer to perform thoracentesis to improve the yield of CT; they should be reminded that CXR after thoracentesis may be more than enough for diagnostic purposes.

Figure 17.34 Actinomycosis crossing anatomic barriers. A. Posteroanterior chest radiograph: left upper lobe airspace disease (A) silhouettes left heart border. B. Computed tomography: gross extension of abnormality into left anterior chest wall (A), with underlying air bronchograms.

The important bottom line in this setting, as in many others, is how this study will change the patient’s workup and management. In many circumstances it is clear that the proposed study (often CT) is superfluous. If that is the case, an alternative approach should be considered.

It is still possible (and preferable) to conclude on a positive note. Pleural abnormality is sometimes responsible for unusual radiographic appearances. CXR alone may be confusing in some such patients. In that setting, CT often renders a final diagnosis (or at least explains the reason for CXR confusion) with startling clarity. It is a helpful problem-solving tool in appropriately selected patients, even in occasional patients with findings that do not typically necessitate CT.


1. Felson B. Chest roentgenology. Philadelphia: WB Saunders, 1973.

2. Alexander ES, Proto AV, Clark RA. CT differentiation of subphrenic abscess and pleural effusion. AJR Am J Roentgenol 1983;140:47–51.

3. Stark DD, Federle MP, Goodman PC, et al. Differentiating lung abscess and empyema: radiography and computed tomography. AJR Am J Roentgenol 1983;141:163–167.

4. Sheffner SE, Gross BH, Birnberg FA, et al. Iatrogenic bronchopleural fistula caused by feeding tube insertion. J Can Assoc Radiol1985;36:52–55.

5. Carr JJ, Reed JC, Choplin RH, et al. Plain and computed radiography for detecting experimentally induced pneumothorax in cadavers: implications for detection in patients. Radiology 1992;183:193–199.

6. Spizarny DL, Gross BH, Shepard JO. CT findings in localized fibrous mesothelioma of the pleural fissure. J Comput Assist Tomogr1986;10:942–944.

7. Frija J, Schmit P, Katz M, et al. Computed tomography of the pulmonary fissures: normal anatomy. J Comput Assist Tomogr1982;6:1069–1074.

8. Proto AV, Ball JB. Computed tomography of the major and minor fissures. AJR Am J Roentgenol 1983;140:439–448.

9. Otsuji H, Hatakeyama M, Kitamura I. Right upper lobe versus right middle lobe. Differentiation with thin-section, high-resolution CT.Radiology 1989;172:653–656.

10. Godwin JD, Tarver RD. Accessory fissures of the lung. AJR Am J Roentgenol 1985;144:39–47.

11. Proto AV, Speckman JM. The left lateral radiograph of the chest. Med Radiogr Photogr 1979;55:30–74.

12. Austin JH. The left minor fissure. Radiology 1986;161:433–436.

13. Speckman JM, Gamsu G, Webb WR. Alterations in CT mediastinal anatomy produced by an azygous lobe. AJR Am J Roentgenol1981;137:47–50.

14. Gross BH, Spizarny DL, Granke DS. Sagittal orientation of the anterior minor fissure: radiography and CT. Radiology 1988;166:717–719.

15. Bhalla M, McCauley DI, Golimbu C, et al. Counting ribs on chest CT. J Comput Assist Tomogr 1990;14:590–594.

16. Pennes DR, Glazer GM, Wimbish KJ, et al. Chest wall invasion by lung cancer: limitations of CT evaluation. AJR Am J Roentgenol1985;144:507–511.

17. Glazer HS, Duncan-Meyer J, Aronberg DJ, et al. Pleural and chest wall invasion in bronchogenic carcinoma: CT evaluation. Radiology1985;157:191–194.

18. Haggar AM, Pearlberg JL, Froelich JW, et al. Chest-wall invasion by carcinoma of the lung: detection by MR imaging. AJR Am J Roentgenol 1987;148:1075–1078.

19. Raasch BN, Carsky EW, Lane EJ. Radiographic anatomy of the interlobar fissures: a study of 100 specimens. AJR Am J Roentgenol1982;138:1043–1049.

20. Shuman LS, Libshitz HI. Solid pleural manifestations of lymphoma. AJR Am J Roentgenol 1984;142:269–273.

21. Quint LE, Glazer GM, Orringer MB. Central lung masses: prediction with CT of need for pneumonectomy versus lobectomy. Radiology1987;165:735–738.