27 Diaphragm, Chest Wall, Pleura
J. Verschakelen
J. Verschakelen, MD
Department of Radiology, University Hospitals, Herestraat 49, 3000 Leuven, Belgium
CONTENTS
27.1 Diaphragm 395 27.1.1 Introduction 395
27.1.2 Acquisition and Injection Techniques 396 27.1.3 Normal CT Appearance 397
27.1.4 Pathology of the Diaphragm 397 27.1.4.1 Diaphragmatic Hernia 397
27.1.4.2 Traumatic Diaphragmatic Rupture 398 27.1.4.3 Tumors of the Diaphragm 399 27.1.4.4 Eventration 400
27.1.4.5 Paralysis of the Diaphragm 400 27.1.5 Conclusions 400
27.2 Chest Wall and Pleura 400 27.2.1 Introduction 400
27.2.2 Acquisition and Injection Techniques 401 27.2.3 Pathology of the Chest Wall 401
27.2.3.1 Congenital Deformation of the Chest Wall 401 27.2.3.2 Trauma to the Chest Wall 401
27.2.3.3 Invasion of the Chest Wall by Tumor 401 27.2.4 Pleural Abnormalities 402
27.2.5 Tumors of the Breast 402 27.2.6 Conclusions 405
References 405
27.1 Diaphragm
27.1.1 Introduction
The diaphragm is a thin, fl at musculotendinous structure that separates the thoracic cavity from the abdominal cavity and that, being a respiratory muscle, has an important role in respiration.
Although diseases of the diaphragm itself are rela- tively infrequent, knowledge of radiographic appear- ance of the diaphragm and the peridiaphragmatic region is mandatory. It is important to differentiate between pathology originating from the diaphragm
itself and pathological processes located immediately above or below the diaphragm. Since diaphragmatic disease is also often benign, it is important to be able to differentiate between abnormalities that have no clinical relevance and abnormalities that need fur- ther exploration.
Unfortunately, radiology of the diaphragm is dif- fi cult (Panicek et al. 1988; Tarver et al. 1989), in part because many diaphragmatic and also peridiaphrag- matic abnormalities are obscure clinically, but also and mainly because there is no imaging technique that can clearly and entirely visualize the diaphragm. More- over, the radiological appearance of the diaphragm is variable since it depends on the function and integrity of the diaphragmatic muscle and is related to the tho- racic and abdominal volumes and contents and to the motion of rib cage and abdomen.
Although we usually speak of the top of the opaque abdominal mass (usually composed of liver, spleen, stomach, or colon) as being the diaphragm on a con- ventional chest fi lm, the diaphragmatic muscle as such is only visible when air is present above and below it.
Ultrasound (Lewandowski and Winsberg 1983;
Oyen et al. 1984; Pery et al. 1984; Verschakelen et al. 1989b), CT (Gale 1986; Kleinman and Rapto- poulos 1985; Shin and Berland 1985), and magnetic resonance imaging (MRI; Yamashita et al. 1993) are the only imaging modalities that can visualize the diaphragm itself. Although visualization is mostly partial and depends on the presence of pleural dis- ease in ultrasound (Verschakelen et al. 1989b) and subdiaphragmatic fat in CT (Fig. 1) and MRI (Gale 1986; Kleinman and Raptopoulos 1985). Both MR and CT have the ability to image the diaphragm in imaging planes other than the axial plane. Visualiza- tion of the diaphragm in the sagittal and frontal plane is indeed often helpful in the study of diaphragmatic and especially of peridiaphragmatic disease (Fig. 2;
Brink et al. 1994). Since the introduction of spiral CT and multidetector CT, it has become possible to obtain high-quality multiplanar and even 3D reconstructions of the diaphragmatic area (Fig. 2; Cassart et al. 1999;
Pettiaux et al. 1997). Moreover, the ability to perform
image acquisition during one breath hold allows elimi- nation of respiratory motion artifacts which improves image quality and also makes it possible to visualize the diaphragm at different levels of respiration. Study of the diaphragmatic function, initially only possible with fl uoroscopy and ultrasound, can now also be performed using dynamic MRI and spiral- and multi- detector CT (Cassart et al. 1997, 1999, 2001).
27.1.2 Acquisition and Injection Techniques
Since high detail is necessary to identify the dia- phragm correctly, a scanning protocol should be
chosen that offers the highest image quality (Brink et al. 1994). A thin detector set should be chosen to allow reconstruction of high-detail contiguous or overlapping thin sections and, when necessary, to make targeted reconstructions through the region of interest. A specifi c advantage of MDCT is that, choosing a thin detector set, detailed characterization of the diaphragm remains possible, even though it was initially planned to view the images with thicker slices. In this way a second high-resolution acquisi- tion through the region with thin collimation, which is usual with spiral CT, is obviated. In the setting of a trauma, when rapid and accurate diagnosis is critical and examination technique is limited due to patient positioning and clinical situation, MDCT allows a
Fig. 27.1. A MDCT of the diaphragm a frontal and b sagittal reconstructions. The costal part of the diaphragm is visible as a thin line between the fat below and the lung above; however, when there is no fat separating the diaphragm from the liver, the spleen, the stomach, or the heart, the diaphragm is not visible. The CT protocol: acquisition 4
×
1mm; pitch 1.75; slice thickness for frontal and sagittal reformations 1 ma
b
Fig. 27.2. A MDCT of the diaphragm a frontal and b 3D reconstructions. Patient with a large malignant tumor in the right lobe of the liver causing elevation of the right hemidiaphragm. Note that the diaphragmatic crura are visible as smooth linear opaci- ties originating from the central tendon and oriented downward, parallel to the vertebral column. The CT protocol: acquisition 16
×
0.75 mm; pitch 1.75; slice thickness for frontal reformation 1 mma b
rapid and comprehensive screening of the diaphragm and peridiaphragmatic regions.
In general, the administration of contrast is not necessary in every patient. Contrast is given in order to better locate or identify peridiaphragmatic masses and abnormalities. The diaphragm itself does not show marked enhancement and its visualization depends more on the presence or absence of subdia- phragmatic fat.
In order to have good opacifi cation of atelectatic lung tissue or lung tumor adjacent to the diaphragm, and in order to have good enhancement of the liver, a long scan delay (45–60 s) is preferred.
The administration of oral contrast can be neces- sary when there is suspicion of herniated bowel or stomach.
27.1.3 Normal CT Appearance
On CT scans the costal part of the diaphragm can be visible as a soft tissue stripe between the fat below and the aerated lung above. The diaphragm is not visible when it is tangential to the scanning plane or where there is no fat separating it from soft tissue structures such as the liver, spleen, stomach, or colon (Fig. 1; Gale 1986; Kleinman and Raptopoulos 1985). The MDCT-generated multiplanar recon- structions can resolve to a certain degree the fi rst problem, but a good visualization of the diaphragm is only achieved when enough fat separates it from adjacent tissue even when high-detail reformations are available. In some cases it is possible with CT to differentiate between the muscular part of the dia- phragm and the central tendon. The muscular part presents as a double line representing the muscle layers, whereas the central tendon presents as a single line. Unlike in ultrasound, pleural disease needs not to be present in order to differentiate these two parts of the diaphragm (Verschakelen et al. 1989b). The diaphragm can appear nodular because of the visu- alization of hypertrophic muscular bundles. Using high-detail acquisition techniques small anatomic structures, such as the inferior phrenic arteries, can become visible (Smith 2002; Ujita et al. 1993).
In most patients the diaphragmatic crura can easily be recognized both on axial and on reformat- ted images (Fig. 2a). They appear mostly as smooth linear opacities originating from the central tendon oriented downward parallel and lateral to the aorta.
They usually appear smooth but can also have a nodular appearance (Caskey et al. 1989). These pseu-
dotumors increase in number and severity with age.
Defects in the crura can also be present and are seen more in older patients and patients with emphysema (Caskey et al. 1989).
27.1.4 Pathology of the Diaphragm
27.1.4.1 Diaphragmatic Hernia
Bochdalek hernia, Morgagni hernia and hiatus hernia are the most frequently occurring herniations of the diaphragm.
Large Bochdalek hernias usually become evident in the neonatal period because they cause respiratory symptoms (Snyder and Greany 1965). Small Boch- dalek hernias, however, only rarely have symptoms and are often an incidental fi nding in adults (Gale 1985; Wilbur et al. 1994). They are usually without any clinical importance, but they should be differen- tiated from diaphragmatic and peridiaphragmatic masses. These hernias often fi rst present in the same way as tumors do: a single focal bulge on the diaphragmatic contour. However, this fi nding is very suggestive for Bochdalek hernia when the patient has no symptoms and when the bulge is centered approximately 4–5 cm anterior to either posterior diaphragmatic insertion. The CT can make the diag- nosis in case of doubt (Gale 1985). The hernia pres- ents as a soft tissue or fatty mass abutting the surface of the posteromedial aspect of either hemidiaphragm (Fig. 3). This mass is in continuity with subdiaphrag- matic structures through a diaphragmatic defect presenting as a discontinuity of the soft tissue line of the diaphragm. The fact that the defect is located in the posteromedial aspect of the diaphragm makes it usually visible on axial scans. In some cases, sagittal reformations give additional information because the diaphragmatic stripe is better identifi ed (Yamana and Ohba 1994). Also 3D imaging can be useful for stereographic perception of the hernia (Yamana and Ohba 1994).
The incidence of Morgagni hernia detected in the
neonatal period because of symptoms is low. In older
children and adults, Morgagni hernia is often an inci-
dental fi nding. Although the weak area at the fi bro-
tendinous elements between the costal and the crural
part of the diaphragm is congenital, Morgagni hernia
can be acquired. Increase in abdominal pressure due
to severe effort, trauma or obesity is probably respon-
sible (Paris et al. 1973; Thomas and Clitherow 1977;
Wilbur et al. 1994). When bowel is herniated to the chest, the diagnosis can usually be made with conven- tional chest-fi lm and barium studies. Herniated liver or fat can be identifi ed with CT. Because there is very often fat above and below the diaphragm in that area, the diaphragmatic defect is frequently visible on axial scans or on sagittal and frontal reformations.
The diagnosis of hiatus hernia can be made on con- ventional posteroanterior chest fi lm, but its presence is usually confi rmed by a barium study. Hiatus hernias are often an incidental fi nding on CT. Multiplanar reconstructions can be helpful in some selected cases in order to better demonstrate the exact position of the diaphragm (Bogaert et al. 1995).
Multiplanar reconstructions can also be helpful in determining the nature, relationship of the herniated organs, the precise side and size of the diaphrag- matic defect in patients with non-traumatic acquired defects (Coulier et al. 1999).
27.1.4.2 Traumatic Diaphragmatic Rupture
Blunt trauma and penetrating wounds of the chest are the most frequent causes of traumatic diaphrag- matic rupture. In blunt trauma the tear is left sided in 70–90% of all cases (Dee 1992). This is probably due to the protective function of the liver (Fataar et al. 1979; van Daele et al. 1987). Injuries from penetrating wounds may be found in any area of the diaphragm (Cotter and Tyndal 1986).
Although in the majority of patients with dia- phragmatic rupture, abnormalities are demonstrated on the chest radiograph at the time of the injury, diag- nosis is often delayed. Aspecifi c clinical symptoms and radiographic fi ndings are responsible for the fact that some patients are fi rst seen within 3 years of the time the injury occurred with symptoms of bowel herniation (Carter et al. 1951; Heiberg et al. 1980).
Most hemodynamically stable patients with blunt diaphragmatic injury as a result of severe polytrauma require an admission CT examination to evaluate the extent and anatomical sites of coexisting thoracoab- dominal injuries (Linsenmaier et al. 2002; Shanmu- ganathan et al. 2000); however, there is disagreement in the literature about the use of CT in the diagnosis of traumatic diaphragmatic rupture. Several case reports have demonstrated that traumatic diaphrag- matic rupture can be identifi ed at CT (Demos et al.
1989; Gurney et al. 1985; Heiberg et al. 1980; Hol- land and Quint 1991); however, in a series of seven patients reported by Gelman et al. (1991), CT made the correct diagnosis in only one patient. On the other hand, Worthy et al. (1995) found diagnostic features with CT in 9 of 11 patients. This contradiction is not surprising since, as discussed previously, it is often dif- fi cult to identify the diaphragm on CT scan, especially when it is immediately adjacent to abdominal organs.
The CT is generally more diagnostic when herniated abdominal organs or bowel can be demonstrated but is less diagnostic when there is a small tear without herniation of abdominal content. In a retrospective
Fig. 27.3. A MDCT of the diaphragm a axial and b sagittal reconstructions. Small Bochdalek hernia presenting as a fatty mass in continuity with subdiaphragmatic structures through a diaphragmatic defect presenting as a discontinuity of the soft tissue line of the diaphragm. The CT protocol: acquisition 4
×
2.5 mm; pitch 1.38; slice thickness for axial slice and sagittal reforma- tion 5 mma b
study of 35 patients with surgically confi rmed dia- phragmatic rupture, CT was able to demonstrate dia- phragmatic rupture in all cases where thoracic hernia- tion of the abdominal organs was present; however, CT fi ndings were questionable in 25% of the cases when no herniation was seen (Scaglione et al. 2000). Simi- lar fi ndings were reported by Killeen et al. (1999). In their study CT had a sensitivity of 78% and a specifi c- ity of 100% for the detection of left-sided diaphrag- matic rupture, whereas these fi gures were respectively 50 and 100% for right-sided diaphragmatic rupture.
The CT signs of diaphragmatic rupture include: dis- continuity of the diaphragm with direct visualization of the diaphragmatic injury, herniation of abdominal organs with liver, bowel, or stomach in contact with the posterior ribs (“dependent viscera sign”; Bergin et al. 2001) thickening of the crus (“thick crus sign”;
Leung et al. 1999) constriction of the stomach or bowel (“collar sign”; Gurney et al. 1985; Naidich et al. 1991), active arterial extravasation of contrast mate- rial near the diaphragm and in case of a penetrating diaphragmatic injury depiction of a missile or punc- turing instrument trajectory. In a recent study Larici et al. (2002) found that the dependent viscera sign had the highest sensitivity, whereas the collar sign, extrava- sation of contrast material and a trajectory were the most specifi c manifestations of diaphragmatic injury.
In this study coronal and sagittal reconstructions were of limited use in establishing or refuting the diagnosis of acute diaphragmatic injury. This is in contradic- tion with the fi ndings of others who have shown that multiplanar reformations can increase diagnostic accuracy (Killeen et al. 1999). The detection of small diaphragmatic defects requires high detail and it can be expected that MDCT giving these detailed images can provide additional information. This is especially true for small tears in the dome and at the musculoten- dinous junction where spiral CT was not successful in one study (Scaglione et al. 2000).
When the rupture is missed at the time of the trauma and patient has recovered from the associated lesions a “latent “phase of diaphragmatic rupture can occur. Symptoms and signs are then caused by recur- rent herniation of abdominal structures through the diaphragmatic defect and are again often aspecifi c.
The MDCT might again be helpful in detecting these missed tears.
27.1.4.3 Tumors of the Diaphragm
Tumors of the diaphragm are rare lesions that are often diffi cult to assess and classify both clinically and
roentgenographically (Anderson and Forrest 1973).
Malignant tumors are more frequent than benign tumors and can be primary or secondary (Anderson and Forrest 1973; Tarver et al. 1989). Benign tumors (lipomas, fi bromas, angiofi bromas, neurofi bromas, and neurolemmomas) are mostly asymptomatic and often found at post-mortem examination (Anderson and Forrest 1973; Schwartz and Wechsler 1989).
The majority of the primary malignant tumors are of fi brous origin (fi brosarcoma, fi bromyosarcoma, fi broangioendothelioma) or are undifferentiated sarcomas (Anderson and Forrest 1973; Schwartz and Wechsler 1989). In contrast to benign tumors, malignant tumors mostly induce symptoms (pleu- ritic chest pain, pain referred to the epigastrium) and are often associated with pleural effusion. Sec- ondary malignant tumors are mostly due to direct invasion from adjacent lesions originating from the lung, the stomach, the pancreas, the adrenals, the colon, and the liver (Fig. 4; Anderson and Forrest 1973; Schwartz and Wechsler 1989). Metastatic implants are rare and only found in cases of widely disseminated disease. Thin sections MDCT performed during one breath hold and multiplanar reformations generated from MDCT data allow a better delineation of the diaphragm and of the relationship between the mass and the diaphragmatic muscle. Even when the diaphragmatic muscle as such is not visible (because of the absence of fat between the diaphragm and the subdiaphragmatic organ), contact between a mass and the diaphragm or diaphragmatic invasion by a tumor can often be diagnosed by looking at the expected contour of the diaphragm.
Fig. 27.4. Malignant lung tumor invading in the diaphragm.
Coronal reconstruction shows close contact between the tumor and the diaphragm and the extension of the tumor into the subdiaphragmatic fat. The CT protocol: acquisition 16
×
0.75 mm; pitch 0.950; slice thickness 3 mm27.1.4.4 Eventration
Eventration of the diaphragm is defi ned as an abnormally high or elevated position of one leaf of the intact diaphragm as a result of paralysis, aplasia, or atrophy of varying degrees of muscle fi bers (Bis- gard 1947). In the area of eventration the normal diaphragmatic muscle fi bers are replaced by a thin layer of connective tissue and a few scattered muscle fi bers. Eventration can be congenital resulting from congenital failure of proper muscularization of a part or of the entire diaphragmatic leaf (Lindstrom and Allen 1966; Tarver et al. 1989). Eventration can also be acquired and is then the result of long- lasting paralysis causing atrophy and scarring of the diaphragmatic muscle (Bovornkitti et al. 1960;
McNamara et al. 1968; Michelson 1961). Total eventration of one hemidiaphragm is more often seen on the left side. Partial eventrations are usually right sided with a predilection for the anteromedial portion; however, diaphragmatic eventration may occur almost anywhere along the diaphragmatic surface and is commonly multifocal.
Since eventration usually is asymptomatic, the role of CT is mostly limited to the differentiation of this entity from a tumoral mass in a patient present- ing with a focal bulge on the diaphragmatic contour.
In contrast to a herniation, where there is a defect in the diaphragm and where abdominal fat or an abdominal organ is protruding into the chest, in eventration the diaphragm, although thin and only consisting of a thin layer of connective tissue, is not interrupted. Differential diagnosis is often easy because of typical localization of both congenital hernias and eventrations, but can be more diffi cult when herniation is the result of a posttraumatic dia- phragmatic tear. Differential diagnosis with focal eventually reversible paralysis is, however, mostly impossible.
27.1.4.5 Paralysis of the Diaphragm
The role of CT in the diagnosis of diaphragmatic paralysis is limited. The radiological evaluation of paralysis requires chest radiographs, adequate fl uoroscopic tests or when not possible ultrasound together with a good knowledge of the clinical his- tory of the patient. The CT can be helpful in the diag- nosis of intrathoracic causes of phrenic nerve injury (Ujita et al. 1993). The CT and especially MDCT can also be valuable in the differential diagnosis of
a paralyzed (hemi)diaphragm and peridiaphrag- matic pathology (subpulmonary pleural effusion, ascites, lung atelectasis, lung or liver mass adjacent to the diaphragm). As mentioned previously, the dif- ferential diagnosis with eventration can be diffi cult and is often impossible.
27.1.5 Conclusions
Because of its capability to scan the diaphragm and the peridiaphragmatic region during one breath hold, and especially because of its ability to perform high-detail multiplanar reformations, MDCT can become very important in the study of the diaphragm and of its adjacent structures.
Especially sagittal and coronal reformations can be very important to localize abnormalities to the diaphragm itself, to the intraabdominal viscera, the cardiophrenic space, the pleura, the lung, or the pericardium.
27.2 Chest Wall and Pleura
27.2.1 Introduction
The use of CT in the study of diseases of the chest wall is usually limited to the evaluation of tumor extension: detection of pleural and chest wall invasion of peripheral bronchogenic carcinoma or determination of tumor extension in patients with tumor of breast or mesothelioma. However, it is well known that 2D CT scans generally have a low sensitivity and specifi city when invasion of the parietal pleura and chest wall by a lung or pleura tumor is examined (Epstein et al. 1986; Grenier et al. 1989; Pearlberg et al. 1987; Pennes et al. 1985;
Webb et al. 1991). Studies have shown that 2D or 3D reformations obtained with spiral CT can add some valuable information (Kuriyama et al. 1994a).
Moreover, the ability offered by spiral CT to observe
a tumor during its different phases of contrast
enhancement has been used to study tumors of the
breast (Teifke et al. 1994). It can be expected that
MDCT, because of its ability to produce high-detail
images in different imaging planes, will further
increase the role of CT in the study of pleura and
chest wall abnormalities.
27.2.2 Acquisition and Injection Techniques
Since the added value of MDCT is related to its ability to perform reformations, and since high detail is nec- essary, scans should be performed with a thin detector set and reconstructed with overlap. Of course these parameters have to be adapted to the scan volume and to the possibility of the patient to stop breathing.
When intravenous contrast is given, an appropri- ate delay between the start of the injection and the scan should be chosen in order to allow enhancement of the tumor, but also, when present, of the thickened pleura and atelectatic lung tissue.
27.2.3 Pathology of the Chest Wall
27.2.3.1 Congenital Deformation of the Chest Wall
In evaluating congenital disease of the chest wall, one needs to use CT only occasionally (Gouliamos et al.
1980; Toombs et al. 1981); however, conditions such as congenital absence of a pectoral muscle, deformi- ties of the sternum, the ribs, and the vertebrae can sometimes be assessed to advantage in transaxial images (Laufer et al. 1999). In addition, when surgi- cal correction is considered, 3D reconstruction of the bony chest can probably help the surgeon to better understand the deformation of the chest and to chose the best surgical procedure to correct it (Haller et al. 1989; Hurwitz et al. 1994).
27.2.3.2 Trauma to the Chest Wall
Most osseous injuries are easily evaluated on plain fi lms, but sometimes CT imaging adds signifi cant information (Dee 1992; Mirvis and Templeton 1992; Stark and Jaramillo 1986). Although the presence of life-support devices and orthopedic devices can make it diffi cult to position a trau- matized patient, and although cooperation of the patient is often limited, MDCT has the advantage over conventional and spiral CT that scanning time can be further reduced. In this way it can eventually help to detect and locate multiple frag- ments of bones, hematomas in soft tissues, foreign bodies, subluxations, damage to the spinal canal, and extrapleural air collections (Fig. 5; Kurihara et al. 1997).
27.2.3.3 Invasion of the Chest Wall by Tumor
Assessment of pleural and chest wall invasion is an important component of lung cancer staging; how- ever, the accuracy of 2D CT in those cases where the tumor is adjacent to the chest wall without any bone destruction is low (Epstein et al. 1986; Grenier et al. 1989; Pearlberg et al. 1987; Pennes et al. 1985;
Webb et al. 1991). A fi rst reason for this is the axial format not allowing to evaluate lesions in contact with the apex of the chest and the diaphragm. A second reason is the fact that features such as a large contact (>3 cm) between the mass and the pleura, an obtuse angle between the tumor and the chest wall, an associated pleural thickening, and the pres- ence of pleural tags, mostly considered as signs of chest wall invasion, also occur with benign lesions.
Three-dimensional techniques have been used suc- cessfully to study lung tumors, peripheral pulmo- nary vessels, and pleural surface (Kuriyama et al.
1994a,b). In a study where they reviewed 2D and 3D images obtained with spiral CT in 42 patients with peripheral bronchogenic carcinoma, Kuriyama and co-workers (1994b) found that 3D-recon- struction imaging was superior to 2D CT in the assessment of pleural invasion. Three-dimensional reconstructions allowed them to correctly predict pleural involvement and to differentiate between visceral pleura and parietal pleural involvement or chest wall invasion. According to their results, it was possible to differentiate between simple pleural
Fig. 27.5. A MDCT of the chest. Three-dimensional reformat- ted image very clearly shows the osteolytic lesion with a pathological fracture of the rib
tags (i.e., fi brotic bands extending from the lesion to the visceral pleura) from visceral pleural invasion (i.e., pleural puckering associated with an indrawn locally thickened pleura). Another way of evaluating parietal pleural invasion of lung masses was pro- posed by Shirakawa and co-workers (1994). These investigators performed spiral CT scans of the chest during deep inspiration and expiration in patients with peripheral lung tumors in contact with the chest wall. They found the presence of respiratory phase shift to be a reliable indicator of the lack of parietal pleural invasion for tumors in middle and lower lung lobes.
In cases where tumor invasion in the chest wall or the diaphragm is obvious, 2D sagittal or frontal reformatted images can be helpful in studying the extent of the mass (Fig. 6; Deschildere et al. 1994).
It is not determined yet whether these reformatted images give more information than sagittal or fron- tal magnetic resonance images. The MRI is at this moment still considered as the image modality of choice in studying superior sulcus tumors and their extension to the chest wall (Takasugi et al. 1989).
It can be expected that MDCT, due to the lack of motion artifacts and the better evaluation of inva- sion in the bony cortex of the ribs, will play a more important role than either CT and spiral CT. Three- dimensional image reconstruction methods can also be used in selected cases to clarify a complex relationship between a tumor invading the chest wall and vascular structures of the thoracic inlet (Fig. 7; Tello et al. 1993).
27.2.4 Pleural Abnormalities
The combination of axial scans and 2D and 3D reformations can be helpful in selected cases to study the extent of pleural disease and to estimate the total surface that is involved (Fig. 8; Meier et al. 1993). The MDCT can also be helpful in cases where it is diffi cult to differentiate pleural disease from lung or diaphragmatic abnormalities (Fig. 9).
Sagittal and coronal reformations can better show the curvation of bronchovascular structures toward a pleural-based area of lung atelectasis (comet tail sign): an additional diagnostic feature of rounded atelectasis that previously could only be shown by sagittal conventional tomograms or MRI (Fig. 10;
Verschakelen et al. 1989a).
27.2.5 Tumors of the Breast
Computed tomography is often performed in patients with a tumor of the breast when tumor infi ltration in the chest wall is suspected or to look for axillary, parasternal, and mediastinal adenopathy.
The development of spiral CT in the early 1990s induced a renewed interest to use this technique not only for staging but also for detection of breast car- cinoma (Teifke et al. 1994; Sardanelli et al. 1995).
Spiral CT has the advantage that it can examine the breasts and axillary areas in a short time with high
Fig. 27.6. A MDCT of the chest and chest wall a sagittal and b 3D reformations. Reformatted image shows tumor extension in the anterior and posterior chest wall. Bony destruction is best demonstrated on the 3D reconstruction
a b
Fig. 27.7a–c. Superior sulcus tumor invading the chest wall.
a Axial slice, b coronal, and c 3D reconstructions show the tumor extending into the apical soft tissues and surrounding the left subclavian artery which is narrowed. The CT protocol:
acquisition 16
×
0.75 mm; pitch 0.950; slice thickness, axial slice 5 mm; coronal reformation 3 mma
c b
a
b Fig. 27.8a, b. Malignant mesothelioma. a Coronal and b 3D reconstruction. The MDCT is very helpful in studying the extent of pleural involvement and in estimating the total pleural surface that is involved. The CT protocol: coronal slice acquisition 4
×
2.5 mm; pitch 5 mm; 3D view acquisition 16×
0.75 mm; pitch 0.950Fig. 27.9. Bilateral pleural effusion with extension into the fi ssure. Coronal reformation shows the extent of the pleural effusion and helps in differentiating this pleural effusion from the atelectatic lung tissue in both lower lobes. The CT protocol:
acquisition 16
×
0.75 mm; pitch 0.950; slice thickness 3 mmdetail during optimal enhancement of the tumor.
Teifke and co-workers (1994) found spiral CT very helpful for elucidating problems in the diagnosis of breast lesions. They appreciated especially the speed of the method, the comfort for the patient, the absence of movement artifacts, the easy standard- ization, and the wide applicability of the method making it a good alternative for MRI when this tech- nique is not available. These authors compared spiral scans performed before and 50 s after injection of contrast. A mass showing increase in density of less than 30 Hounsfi eld units (HU) is very likely to be a benign lesion (fi broadenoma), whereas lesions with an increase in density of more than 60 HU are very
suggestive for malignant tumors. For lesions with an enhancement between 30 and 60 HU the differential diagnosis between benign and malignant disease was diffi cult and should be based on other criteria such as delineation of the tumor. Presently, however, spiral CT is only occasionally used and most institutions use MRI to explore breast tumors after a mammogram and an ultrasound have been performed. It is not sure whether MDCT will change this, but this technique has some advantages over MRI. Examination time is shorter, reducing motion artifacts, invasion in the bony chest wall is better seen, and the examination can be performed in the same body position as the surgical procedure (Fig. 11).
Fig. 27.10a–d. Rounded atelectasis of the lung. b, c Coronal and d sagittal reformations add information to a the axial slice by showing additional curvilinear opacities extending towards the pleural based mass. The CT protocol: acquisition 16
×
0.75 mm;pitch 0.950; slice thickness, coronal and sagittal reformations 3 mm
a b
c d
Fig. 27.11a, b. Malignant tumor of the breast. a Sagittal reformation shows the mass which is separated from the pectoralis muscle by a small amount of fat. On this sagittal image and also on the b coronal view an important pleural effusion and sev- eral pleural metastases are seen (arrows). The CT protocol: acquisition 4
×
2.5 mm; pitch 1.38; slice thickness for coronal and sagittal reformations 3 mm27.2.6 Conclusions
The advantage of MDCT in the study of chest wall and pleura is related to its ability to perform high- detail images in a very short time. In this way motion artifacts and respiratory motion artifacts, often occurring with conventional CT of these areas, can be avoided. The value of 2D and 3D reformations is not yet fully determined.
References
Anderson LS, Forrest JV (1973) Tumors of the diaphragm. Am J Roentgenol 119:259–265
Bergin D, Ennis R, Keogh C et al. (2001) The”dependent viscera” sign in CT diagnosis of blunt traumatic diaphrag- matic rupture. Am J Roentgenol 177:1137–1140
Bisgard JD (1947) Congenital eventration of the diaphragm. J Thorac Surg 16:484–488
Bogaert J, Weemaes K, Verschakelen JA et al. (1995) Spiral CT fi ndings in a postoperative intrathoracic gastric herniation:
a case report. Eur Radiol 5:192–195
Bovornkitti S, Kangsadal P, Sandvichien S et al. (1960) Neuro- genic muscular aplasia (eventration) of the diaphragm. Am Rev Respir Dis 82:876–880
Brink JA, Heiken JP, Semenkovich J et al. (1994) Abnormalities of the diaphragm and adjacent structures: fi ndings on mul- tiplanar spiral CT scans. Am J Roentgenol 163:307–310
Carter B, Giuseffi J, Felson B (1951) Traumatic diaphragmatic hernia. Am J Roentgenol 65:56–71
Caskey CI, Zerhouni EA, Fishman EK et al. (1989) Aging of the diaphragm: a CT study. Radiology 171:385–389
Cassart M, Pettiaux N, Gevenois PA et al. (1997) Effect of chronic hyperinfl ation on diaphragm length and surface area. Am J Respir Crit Care Med 156:504–508
Cassart M, Verbandt Y, de Francquen P et al. (1999) Diaphragm dimensions after single-lung transplantation for emphy- sema. Am J Respir Crit Care Med 159:1992–1997
Cassart M, Hamacher J, Verbandt Y et al. (2001) Effects of lung volume reduction surgery for emphysema on diaphragm dimensions and confi guration. Am J Respir Crit Care Med 163:1042–1043
Coulier B, Ramboux A, Mailleux P et al. (1999) Diagnosis of non-hiatal diaphragmatic hernia using helical computed tomography. JBR BTR 82:1–5
Cotter CP, Tyndal EC (1986) Traumatic diaphragmatic injuries.
Arch Fr Pediatr 33:197–203
Dee PM (1992) The radiology of chest trauma. Radiol Clin North Am 30:291–306
Demos TC, Solomon C, Posniak HV et al. (1989) Computed tomography in traumatic defects of the diaphragm. Clin Imaging 13:62–67
Deschildre F, Petyt L, Rémy-Jardin M et al. (1994) Evaluation de la TDM par balayage spirale volumique (BSV) vs IRM dans le bilan d’extension pariétal des masses thoraciques.
Rev Im Med 6:188
Epstein DM, Stephenson LW, Gefter WB et al. (1986) Value of CT in the preoperative assessment of lung cancer: a survey of thoracic surgeons. Radiology 161:423–427
Fataar S, Rad FF, Schulman A (1979) Diagnosis of diaphrag- matic tears. Br J Radiol 52:375–381
a b
Gale ME (1985) Bochdalek hernia: prevalence and CT charac- teristics. Radiology 156:449–452
Gale ME (1986) Anterior diaphragm: variations in the CT appearance. Radiology 161:635–639
Gelman R, Mirvis SE, Gens D (1991) Diaphragmatic rupture due to blunt trauma: sensitivity of plain chest radiographs.
Am J Roentgenol 156:51–57
Gouliamos AD, Carter BL, Enami B (1980) Computed tomog- raphy of the chest wall. Radiology 134:433–436
Grenier P, Dubray B, Carette MF et al. (1989) Preoperative tho- racic staging of lung cancer: CT and MR evaluation. Diagn Intervent Radiol 1:23–28
Gurney J, Harrison NL, Anderson JC (1985) Omental fat simu- lating pleural fl uid in traumatic diaphragmatic hernia: CT characteristics. J Comput Assist Tomogr 9:1112–1114 Haller JA, Scherer LR, Turner CS et al. (1989) Evolving manage-
ment of pectus excavatum based on a single institutional experience of 664 patients. Ann Surg 209:578–582 Heiberg E, Wolverson MK, Hurd RN et al. (1980) CT recogni-
tion of traumatic rupture of the diaphragm. Am J Roent- genol 135:369–372
Holland DG, Quint LE (1991) Traumatic rupture of the dia- phragm without visceral herniation: CT diagnosis. Am J Roentgenol 157:17–18
Hurwitz DJ, Stofman G, Curtin H (1994) Three-dimensional imaging of Poland’s syndrome. Plast Reconstr Surg 94:
719–723
Killeen KL, Mirvis SE, Shanmuganathan K (1999) Helical CT of diaphragmatic rupture caused by blunt trauma. Am J Roentgenol 173:1611–1616
Kleinman PK, Raptopoulos V (1985) The anterior diaphrag- matic attachments: an anatomic and radiologic study with clinical correlates. Radiology 155:289–293
Kuriyama K, Hosomi N, Sawai Y et al. (1994a) Three-dimen- sional imaging of focal lung disease using spiral volumetric CT. Jpn J Clin Radiol 39:9–13
Kuriyama K, Tateishi R, Kumatani T et al. (1994b) Pleural inva- sion by peripheral bronchogenic carcinoma: assessment with three-dimensional helical CT. Radiology 191:365–369 Kurihara Y, Nakajima Y, Niimi H et al. (1997) Extrapleural air
collections mimicking pneumothorax: helical CT fi nding. J Comput Assist Tomogr 21:771–772
Larici AR, Gotway MB, Litt HI et al. (2002) Helical CT with sag- ittal and coronal reconstructions: accuracy for detection of diaphragmatic injury. Am J Roentgenol 179:451–457 Laufer L, Schulman H, Hertzanu Y (1999) Intrathoracic rib
demonstrated by helical CT with three-dimensional recon- struction. Eur Radiol 9:60–61
Leung JC, Nance ML, Schwab CW et al. (1999) Thickening of the diaphragm: a new computed tomography sign of dia- phragm injury. J Thorac Imaging 14:126–129
Lewandowski B, Winsberg F (1983) Echographic appearance of the right hemidiaphragm. J Ultrasound Med 2:243–249 Lindstrom CH, Allen RP (1966) Bilateral congenital eventra-
tion of the diaphragm. Am J Roentgenol 97:216–221 Linsenmaier U, Krotz M, Hauser H et al. (2002) Whole-body
computed tomography in polytrauma: techniques and management. Eur Radiol 12:1728–1740
McNamara JJ, Paulson DL, Urschel HC et al. (1968) Eventration of the diaphragm. Surgery 64:1013–1021
Meier RA, Marianacci EB, Costello P et al. (1993) 3D image reconstruction of right subclavian artery aneurysms. J Comput Assist Tomogr 17:887–890
Michelson E (1961) Eventration of the diaphragm. Surgery 49:410–422
Mirvis SE, Templeton P (1992) Imaging in acute thoracic trauma. Semin Roentgenol 27:184–210
Naidich DP, Zerhouni EA, Siegelman SS (1991) Computed tomography and magnetic resonance of the thorax. Raven Press, New York
Oyen R, Marchal G, Verschakelen J et al. (1984) Sonographic aspect of hypertrophic diaphragmatic muscular bundles. J Clin Ultrasound 12:121–123
Panicek DM, Benson CB, Gottlieb RH et al. (1988) The dia- phragm: anatomic, pathologic and radiologic consider- ations. Radiographics 8:385–425
Paris F, Tarazona V, Casillas M et al. (1973) Hernia of Morgagni.
Thorax 28:631
Pearlberg JL, Sandler MA, Beute GH et al. (1987) Limitations of CT in evaluation of neoplasms involving chest wall. J Comput Assist Tomogr 11:290–293
Pennes DR, Glazer GM, Wimbish KJ et al. (1985) Chest wall invasion by lung cancer: limitations of CT evaluation. Am J Roentgenol 144:507–511
Pery M, Kaftori J, Rosenberger A (1984) Causes of abnormal right diaphragmatic position diagnosed by ultrasound. J Clin Ultrasound 12:121–123
Pettiaux N, Cassart M, Paiva M et al. (1997) Three-dimensional reconstruction of human diaphragm with the use of spiral computed tomography. Appl Physiol 82:998–1002 Sardanelli F, Melani E, Calabrese M et al. (1995) Dynamic spiral
CT of suspected breast tumors. Eur Radiol 5 (Suppl):280 Scaglione M, Pinto F, Grassi R et al. (2000) Diagnostic sensitiv-
ity of computerized tomography in closed trauma of the diaphragm. Retrospective study of 35 consecutive cases.
Radiol Med (Torino) 99:46–50
Schwartz EE, Wechsler RJ (1989) Diaphragmatic and paradia- phragmatic tumors and pseudotumors. J Thorac Imaging 4:19–28
Shanmuganathan K, Killeen K, Mirvis SE et al. (2000) Imaging of diaphragmatic injuries. J Thorac Imaging 15:104–111 Shin MS, Berland LL (1985) Computed tomography of retro-
crural spaces: normal, anatomic variants and pathologic conditions. Am J Roentgenol 145:81–86
Shirakawa T, Fukuda K, Miyamoto Y et al. (1994) Parietal pleural invasion of lung masses: evaluation with CT per- formed during deep inspiration and expiration. Radiology 192:809–811
Smith TR (2002) Demonstration of inferior phrenic arteries on spiral CT. Clin Imaging 26:27–29
Snyder WH, Greany EM (1965) Congenital diaphragmatic hernia; 77 consecutive cases. Surgery 57:576–588
Stark P, Jaramillo D (1986) CT of the sternum. Am J Roent- genol 147:72–77
Takasugi JE, Rapoport S, Shaw C (1989) Superior sulcus tumors: the role of imaging. J Thorac Imaging 4:41–48 Tarver RD, Cones DJ, Cory DA et al. (1989) Imaging the dia-
phragm and its disorders. J Thorac Imaging 4:1–18 Teifke A, Schweden F, Cagil H et al. (1994) Spiral-Comput-
ertomographie der Mamma. Fortschr Roentgenstr 161:
495–500
Tello R, Scholz E, Finn JP et al. (1993) Subclavian vein throm- bosis detected with spiral CT and three-dimensional reconstruction. Am J Roentgenol 160:33–34
Thomas GG, Clitherow NR (1977) Herniation through the foramen of Morgagni in children. Br J Surg 64:215–217
Toombs BD, Sandler CM, Lester RG (1981) Computed tomog- raphy of chest trauma. Radiology 140:733–738
Ujita M, Ojiri H, Ariizumi M et al. (1993) Appearance of the inferior phrenic artery and vein on CT scans of the chest: a CT and cadaveric study. Am J Roentgenol 160:745–747 Van Daele G, Joris L, Eyskens E (1987) Traumatische diafrag-
maruptuur. Tijdschr Geneeskd 43:1649–1653
Verschakelen JA, Demaerel P, Coolen J et al. (1989a) Rounded atelactasis of the lung: MR appearance. Am J Roentgenol 152:965–966
Verschakelen JA, Marchal G, Verbeken E et al. (1989b) Sono- graphic appearance of the diaphragm in the presence of pleural disease: a cadaver study. J Clin Ultrasound 17:
222–227
Webb WR, Gatsonis C, Zerhouni EA et al. (1991) CT and MR imaging in staging non-small cell bronchogenic carci- noma: report of the radiologic diagnostic oncology group.
Radiology 178:705–713
Wilbur AC, Gorodetsky A, Hibbeln JF (1994) Imaging fi ndings of adult Bochdalek hernias. Clin Imaging 18:224–229 Worthy SA, Kang EY, Hartman TE et al. (1995) Diaphragmatic
rupture: CT fi ndings in 11 patients. Radiology 194:885–888 Yamana D, Ohba S (1994) Three-dimensional image of Boch-
dalek diaphragmatic hernia; a case report. Radiat Med 12:
39–41
Yamashita K, Minemori K, Matsuda H et al. (1993) MR imaging in the diagnosis of partial eventration of the diaphragm.
Chest 104:328
