Bongard Frederic , Darryl Sue. Diagnosis and Treatment Critical Care

Подождите немного. Документ загружается.



IMAGING PROCEDURES

157

Radiographic Features

Septic pulmonary emboli appear radiographically as wedge-

shaped or rounded peripheral opacities. Septic emboli are

usually multiple and are more numerous in the lower lobes,

reflecting increased blood flow to the dependent lung. The

lesions may vary in size by virtue of variations in the timing

of embolization. Cavitation, typically thin-walled, is com-

mon, and necrotic debris may be identified within the cavity.

Hilar and mediastinal adenopathy can occur, and empyema

may occur.

The CT features of septic emboli have been described.

Peripheral nodules with identifiable feeding vessels, wedge-

shaped peripheral opacities, and cavitation are the most diag-

nostic features. Peripheral enhancement along the margins of

the wedge-shaped densities has been reported following

administration of intravenous contrast material (Figure 7–11).

It has been suggested that CT can detect disease earlier than the

plain radiograph and that it better characterizes the extent of

disease. Moreover, the cross-sectional perspective of CT affords

better identification of embolic lesions that may be obscured

on chest radiographs by edema or other diffuse consolidations.

Han D et al: Thrombotic and nonthrombotic pulmonary arterial

embolism: Spectrum of imaging findings. Radiographics

2003;23:1521–39. [PMID: 14615562]

Huang RM et al: Septic pulmonary emboli: CT-radiographic cor-

relation. AJR 1989;153:41–5. [PMID: 2735296]

Iwasaki Y, et al: Spiral CT findings in septic pulmonary emboli. Eur

J Radiol 2001;37:190–4. [PMID: 11274848]



Pulmonary Edema

ESSENTIALS OF RADIOLOGIC

DIAGNOSIS

Interstitial edema:



Kerley B (most common), A, and C lines.



Peribronchial cuffing.



Indistinct pulmonary vessels.



Hilar haze.

Alveolar edema:



Poorly marginated, coalescent opacities.



Air bronchograms.



“Butterfly” pattern.

General Considerations

Pulmonary edema—an excess of water in the extravascular space

of the lung—is a frequent cause of respiratory distress in the crit-

ically ill patient. The three main categories of pulmonary edema

are cardiac edema secondary to myocardial or endocardial dis-

ease, volume overloaded state due to renal failure or excess

administration of fluid, and increased capillary permeability,

which may result from a variety of insults to the microvascula-

ture of the lung. In the ICU patient, more than one mechanism

may contribute to the formation of edema, increasing the diffi-

culty of diagnostic interpretation on radiographs.

There are four principal mechanisms that result in the

development of edema: elevated capillary hydrostatic pressure,

decreased plasma oncotic pressure, increased capillary perme-

ability, and obstruction of lymphatic drainage. Decreased

plasma oncotic pressure and obstruction to lymphatic drainage

only rarely lead to pulmonary edema but may be contributing

factors in the setting of increased hydrostatic pressure. The



Figure 7–11. Young woman with septic emboli second-

ary to intravenous drug abuse. Blood cultures were positive

for Staphylococcus aureus. A. Peripheral nodular opacities

are present with evidence of cavitation (arrow). A feeding

vessel is identified leading to a pulmonary nodule, consis-

tent with hematogenous dissemination (arrowhead).

B. Wedge-shaped subpleural lesion is noted with periph-

eral enhancement after administration of intravenous contrast

material.



CHAPTER 7

158

most common cause of pulmonary edema is hydrostatic pres-

sure elevation due to cardiac disease. Acute myocardial infarc-

tion, acute volume overload of the left ventricle, and mitral

stenosis are common causes of cardiogenic edema.

Radiographic Features

The chest radiograph is the most commonly used noninva-

sive test in the evaluation of a patient with pulmonary

edema. Interstitial edema may be present radiographically in

the absence of clinical signs and symptoms, and the chest

radiograph may be the first indication of pulmonary edema.

A. Cardiogenic Pulmonary Edema—In the patient with heart

failure, pulmonary edema is preceded by pulmonary venous

hypertension. In patients with left ventricular failure, elevated left

ventricular end-diastolic pressure (pulmonary venous hyperten-

sion) is reflected in the pulmonary vasculature by dilation and

redistribution of pulmonary blood flow to the upper lobes. In the

normal erect patient, the upper zone vessels are smaller than the

lower zone vessels, and a significant fraction of the pulmonary

circulation, particularly to the upper lobes, is not perfused. In

conditions of increased pulmonary blood volume or left ventric-

ular failure, there is recruitment of these nonperfused reserve ves-

sels in the upper lobes, while reflex hypoxic vasoconstriction of

lower lobe vessels occurs. These and other pathophysiologic fac-

tors contribute to the phenomenon of upper lobe arterial and

venous redistribution. Vascular redistribution is often difficult to

observe on radiographs, particularly in critically ill patients imaged

in the semierect or supine position. As the pulmonary venous

pressure continues to increase, pulmonary edema develops.

Pulmonary edema may be present within the pulmonary

interstitium, the alveoli, or both. Radiographic evidence of inter-

stitial edema includes Kerley A, B, and C lines; peribronchial cuff-

ing; hilar haze; indistinct vascular markings; and subpleural

edema. Kerley lines represent thickened interlobular septa, with

Kerley B lines being the most easily and most frequently seen.

These lines are horizontal linear densities measuring 1–2 cm in

length and 1–2 mm in width. They are located peripherally,

extend to the pleural surface, and are best seen at the lung bases on

the frontal film (Figure 7–12). Kerley A lines are longer and more

randomly oriented and are best seen in the upper lobes, directed

toward the hila. Kerley C lines are presumably a superimposition

of many thickened interlobular septa and appear as a fine reticu-

lar pattern. Other signs of interstitial edema, including peri-

bronchial cuffing, hilar haze, and indistinct vascular markings,

result from accumulation of fluid in the perivascular and peri-

bronchial interstitium. Accumulation of fluid in the subpleural

interstitium is best demonstrated along the pleural fissures.

Alveolar edema occurs as fluid fills the air spaces of the

lungs (Figure 7–13). Although interstitial edema precedes alve-

olar edema and continues to be present in the alveolar filling

stage, the interstitial component is frequently obscured by

concomitant air space edema. With filling of the air spaces, the

lung becomes opaque, with poorly defined confluent opacity.

Air bronchograms are identified as tubular lucencies repre-

senting normal patent bronchi surrounded by fluid-filled air

spaces. The butterfly pattern, appearing as a dense perihilar

opacification, has been described in volume overloaded states

and cardiogenic edema.

In general, cardiogenic pulmonary edema is bilateral and

symmetric. Atypical edema patterns may be seen in patients

with underlying acute or chronic lung disease or as a conse-

quence of gravitational forces related to patient positioning.

Destruction of the lung due to emphysema may cause a

patchy, asymmetric distribution of edema that spares regions

of bullous disease. Gravitational forces also affect the distribu-

tion of edema, with increased edema in the dependent lung.

Shifting the patient’s position can change the appearance of



Figure 7–12. Interstitial edema. Kerley B lines are

identified at the lung bases (arrows).



Figure 7–13. Alveolar edema. Air space opacities

with vascular redistribution, perihilar haze, cardiomegaly,

and bilateral pleural effusions are secondary to cardio-

genic edema. A pulmonary artery catheter and nasogas-

tric tube are present.



IMAGING PROCEDURES

159

edema. Such maneuvers may help to distinguish atypical

edema from other air space processes such as pneumonia.

The temporal sequence of parenchymal opacification is also

crucial because the onset and resolution of hydrostatic

edema may be rapid, whereas in other conditions such as

pneumonia and ARDS, changes are more gradual.

The CT findings in heart failure have been studied and, as

predicted by the chest radiograph, include “ground glass”opac-

ities, interstitial and alveolar edema, and pleural effusions.

Small pulmonary nodules also have been described and likely

represent pulmonary vessels and regions of edema. Mediastinal

adenopathy may be present, with 35% of patients with chronic

heart failure demonstrating nodal enlargement on CT.

B. Distinguishing Cardiogenic from Noncardiogenic

Pulmonary Edema—Three principal features have been pro-

posed to distinguish cardiogenic from noncardiogenic pul-

monary edema radiographically: distribution of pulmonary

flow, distribution of pulmonary edema, and width of the vas-

cular pedicle. Ancillary features include pulmonary blood vol-

ume, peribronchial cuffing, septal lines, pleural effusions, air

bronchograms, lung volume, and cardiac size. The vascular

pedicle width is defined as the width of the mediastinum just

above the aortic arch, with normal width ranging from 43 to

53 mm in an erect patient. The vascular pedicle is enlarged in

60% of patients with cardiac failure and in 85% of patients

with renal failure or volume overload. This is in contrast to

patients with noncardiogenic capillary permeability edema,

who have a normal or narrowed vascular pedicle in 70% of

cases. The distribution of flow is also a discriminating feature

in that patients with hydrostatic edema more typically have bal-

anced flow or vascular redistribution. In contrast, patients with

capillary permeability edema usually demonstrate a normal or

balanced distribution of flow. Finally, the distribution of edema

is symmetric and perihilar or basilar in patients with cardio-

genic edema or volume overloaded states, whereas capillary

permeability edema appears patchy and peripheral.

Heart size and the presence or absence of septal lines also

may be useful criteria for differentiating cardiogenic from per-

meability edema with an accuracy of 83%. Thus, if the heart is

enlarged or of normal size and septal lines are present, cardio-

genic edema is likely, but if the heart size is of normal and sep-

tal lines are absent, permeability edema is more likely. There

may be considerable overlap. In one study, a classic hydro-

static pattern occurred in 90% of patients with hydrostatic

edema, but 40% of patients with increased permeability

edema had radiographic features consistent with hydrostatic

edema. A peripheral or patchy air space pattern was relatively

specific for capillary permeability edema. Overlapping features

may arise from differences in patient populations, including

differences in the severity of edema, underlying heart or lung

disease, and radiologic technique and patient positioning.

The radiographic diagnosis of edema may be complicated

by several factors. However, general guidelines can be suggested.

In general, noncardiogenic edema typically demonstrates nor-

mal cardiac size with air space opacities (Figure 7–14) and



Figure 7–14. Noncardiogenic pulmonary edema secondary to near-drowning. A. Anteroposterior chest radiograph

demonstrates asymmetric air space opacities bilaterally. Heart size is normal, and there are no pleural effusions.

Endotracheal tube is high in position, and a nasogastric tube is present. B. Radiograph 48 hours after admission shows het-

erogeneous parenchymal opacification with worsening at the lung bases. A left thoracostomy tube and pulmonary artery

catheter are now present, and the endotracheal tube is in satisfactory position. There is evidence of barotrauma with pneu-

momediastinum (arrow).



CHAPTER 7

160

infrequent Kerley lines, peribronchial cuffing, or pleural effu-

sions. In contrast, hydrostatic edema is associated with cardiac

enlargement, septal lines, and frequent pleural effusions. The

accuracy of chest radiographic diagnosis depends on the inte-

gration of all available clinical and physiologic data.

Aberle DR et al: Hydrostatic versus increased permeability pul-

monary edema: Diagnosis based on radiographic criteria in crit-

ically ill patients. Radiology 1988;168:73–9. [PMID: 3380985]

Gluecker T et al: Clinical and radiologic features of pulmonary

edema. Radiographics 1999;19:1507–31. [PMID: 10555672]

Lewin S, Goldberg L, Dec GW: The spectrum of pulmonary abnor-

malities on computed chest tomographic imaging in patients

with advanced heart failure. Am J Cardiol 2000;86:98–100.

[PMID: 10867103]

Martin GS et al: Findings on the portable chest radiograph corre-

late with fluid balance in critically ill patients. Chest

2002;122:2087–95. [PMID: 12475852]

Miller RR, Ely EW: Radiographic measures of intravascular vol-

ume status: The role of vascular pedicle width. Curr Opin Crit

Care 2006;12:255–62. [PMID: 16672786]

Thomason JW et al: Appraising pulmonary edema using supine

chest roentgenograms in ventilated patients. Am J Respir Crit

Care Med 1998;157:1600–8. [PMID: 9603144]



Acute Respiratory Distress Syndrome

ESSENTIALS OF RADIOLOGIC

DIAGNOSIS



Early ARDS: Decrease in lung volumes, but lungs are

generally clear. If ARDS is caused by aspiration or pneu-

monia, parenchymal opacifications may be present.



Later: Air space opacification is usually bilateral but may

be asymmetric and patchy and may progress later to

more uniform consolidation. Air bronchograms are usu-

ally present.



Late ARDS associated with collagen deposition shows

less dense parenchymal consolidations with interstitial

or “ground glass” opacities.



Complications include pulmonary interstitial emphy-

sema, pneumomediastinum, and pneumothorax.

General Considerations

ARDS is a catastrophic consequence of acute lung injury,

with damage to the alveolar epithelium and pulmonary vas-

culature resulting in increased capillary permeability edema.

Despite numerous attempts at clarification in the literature,

there is still disagreement about the best way to describe this

disorder. It is usually characterized clinically by refractory

hypoxemia, decreased lung compliance, severe acute respi-

ratory distress, and pulmonary parenchymal consolidations

on chest radiographs. A number of disorders are associated

with ARDS, including both direct insults to the lungs and

nonpulmonary systemic conditions.

Radiographic Features

A. Chest Radiographs—The radiographic manifestations

correlate with the pathologic changes seen in the lungs and

vary with the stage of lung injury. Three stages have been

described in ARDS. Stage I (also known as the acute exuda-

tive phase) is the earliest and most transient stage of lung

injury and occurs during the first hours after the insult.

Pathologically, this stage is characterized by pulmonary cap-

illary congestion, endothelial cell swelling, and extensive

microatelectasis. Fluid leakage is confined to the intersti-

tium and is limited. Clinically, respiratory distress with

tachypnea and hypoxemia is present. In patients with ARDS

secondary to systemic insults, diffuse microatelectasis and

diminished lung compliance may result in a decrease in lung

volumes, but the lungs are generally clear. Interstitial fluid is

usually too mild to be radiographically apparent (Figure 7–15).

In primary pulmonary insults causing ARDS, such as aspira-

tion or pneumonia, parenchymal opacifications may be pres-

ent (Figure 7–16). Physiologic changes due to therapy are also

reflected on the radiograph, including volume overload and

barotrauma. The use of positive end-expiratory pressure

(PEEP) may cause improvement in aeration on the chest

radiograph without physiologic or clinical improvement. In

fact, occasionally there is paradoxical worsening of oxygena-

tion from alveolar overdistention with subsequent diversion

of pulmonary flow to poorly ventilated regions.

In stage II (also referred to as the fibroproliferative phase),

the pathologic features of hemorrhagic fluid leakage, fibrin

deposition, and hyaline membrane formation result in radi-

ographic consolidation. Air space opacification is usually bilat-

eral but may be asymmetric and patchy and may progress later

to more uniform consolidation. Air bronchograms are usually

present and become more conspicuous with severe consolida-

tion. The transition to stage II may occur 1–5 days following

the pulmonary insult depending on its type and severity. More

severe injuries result in a more rapid transition. Pleural effu-

sions are uncommon and, when present, are small.

Stage III (also referred as the fibrotic or recovery phase) is

characterized by hyperplasia of type II alveolar epithelial cells

and collagen deposition. Decreased lung compliance,

ventilation-perfusion imbalance, diffusion impairment, and

destruction of the microvascular bed result in abnormal gas

exchange and lung mechanics. Radiographically, parenchy-

mal consolidations become less dense and confluent.

Interstitial or “ground glass” opacities develop as fluid is

replaced by the deposition of collagen. Subpleural lucencies

may develop in regions of peripheral ischemia and ischemic

necrosis. The treatment of ARDS, including positive-pressure

ventilation, sometimes results in barotrauma that is mani-

fested as pulmonary interstitial emphysema, pneumomedi-

astinum, and pneumothorax (Figure 7–17).

Long-term sequelae of ARDS are variable. The overall mor-

tality rate is approximately 50%. Although long-term survivors



IMAGING PROCEDURES

161

may have complete recovery of pulmonary function, respiratory

impairment may result from pulmonary fibrosis and microvas-

cular damage. Improvement in lung function is relatively rapid

during the first 3–6 months, reaching maximum recovery

within 6–12 months following the onset of ARDS. The chest

radiograph may continue to show hyperinflation and some

residual lung opacities, but most often it returns to normal.

B. CT Scans—The CT appearance of ARDS has been

described by numerous investigators. In general, CT demon-

strates a variable and patchy distribution, with most marked

involvement in the dependent lung regions. These opacities

probably represent severe diffuse microatelectasis as well as

edema fluid and have been observed to migrate under the

influence of gravity. Air bronchograms are frequent, and

pleural effusions, typically small, occur in approximately

one-half of patients. The distribution of consolidation may

depend on the stage of ARDS. Early changes may show

patchy areas of “ground glass” opacity or consolidation dif-

fusely but not uniformly, without central or gravity depend-

ence. Later changes show more homogeneity as the lung

becomes more edematous, and gravity-dependent atelectasis

increases. On CT, barotraumatic lung cysts and infectious

complications such as cavitation or empyema are better

identified than on projectional radiographs (Figure 7–18).

Caironi P et al: Radiological imaging in acute lung injury and acute

respiratory distress syndrome. Semin Respir Crit Care Med

2006;XX:404–15. [PMID 16909374]

Desai SR et al: Acute respiratory distress syndrome caused by pul-

monary and extrapulmonary injury: A comparative CT study.

Radiology 2001;218:689–93. [PMID: 11230641]

Gattinoni L et al: What has computed tomography taught us about

the acute respiratory distress syndrome? Am J Respir Crit Care

Med 2001;164:1701–11. [PMID: 11719313]

IMAGING IN PLEURAL DISORDERS



Pleural Effusions

ESSENTIALS OF RADIOLOGIC

DIAGNOSIS



Blunting of the lateral costophrenic angle (meniscus sign).



Elevation of the apparent level of the diaphragm. Increased

separation between the lung and the stomach bubble.



Homogeneous increased density of the involved

hemithorax.



Fluid capping the lung apex.



Decreased visibility of pulmonary vessels below the

diaphragm.



Increased density within the pleural fissures (“pseudo-

tumor”).

General Considerations

Pleural fluid is primarily formed on the parietal pleural sur-

face and absorbed on the visceral pleural surface, with

approximately 25 mL of fluid present normally in the pleural

space. Pleural effusion is an excess accumulation of

intrapleural fluid. A wide variety of disorders result in excess

pleural fluid. Although the chest radiograph is useful for

detecting and estimating the amount of pleural effusion, the

differentiation between transudate, exudate, empyema, and

hemorrhagic pleural effusion requires a thoracentesis.

Congestive heart failure is the most common cause of pleu-

ral effusion in the ICU population.



Figure 7–15. ARDS secondary to sepsis in an immunocompromised patient following bone marrow transplantation.

A. Stage I ARDS. The lungs are clear, despite marked dyspnea and hypoxemia. Lung volumes are slightly decreased.

B. Stage II ARDS. Within 24 hours, the chest radiograph shows diffuse parenchymal opacification consistent with ARDS.



CHAPTER 7

162

Radiographic Features

The distribution of fluid within the pleural space is greatly

affected by lung elastic recoil and gravity. On erect frontal and

lateral radiographs, free pleural effusions typically have a con-

cave, upward-sloping contour (the meniscus appearance). Since

the posterior costophrenic angles are usually deeper than the

lateral costophrenic angles, small pleural effusions are typically

best seen on the lateral view. Blunting of the lateral costophrenic

angle—detectable on an erect posteroanterior chest radi-

ograph—may occur with as little as 175 mL of fluid, although in

some cases as much as 525 mL will be present before blunting is

noted. Pleural effusion also may accumulate in a subpulmonary

location between the lung base and diaphragm without causing



Figure 7–16. ARDS secondary to pneumococcal pneumonia in a patient with a history of Hodgkin’s disease and

splenectomy several years earlier. A. Initial chest radiograph demonstrates patchy bilateral consolidation. B. Within

12 hours of admission, dense air space consolidation is present, necessitating intubation. Clinical course was consistent

with ARDS. C. Follow-up radiograph 5 weeks after admission to the ICU shows a coarse reticular pattern bilaterally.

Lung volumes are slightly decreased in comparison with the admission radiograph.



IMAGING PROCEDURES

163

blunting of the lateral costophrenic sulcus. These subpul-

monary collections simulate elevation of the diaphragm; on the

left, the distance between the gastric air bubble and the “pseu-

dodiaphragm” will be increased. The pulmonary vessels are not

seen through the basilar pulmonary parenchyma. The pseudo-

diaphragm is elevated and flattened, with the dome appearing

more lateral than normal.

Pleural effusions may extend into the fissures, with the

radiographic appearance depending on the shape and orientation

of the fissure, the location of the fluid, and the direction of the

radiographic beam. Collections of fluid in the fissures may

mimic a mass, resulting in a “pseudotumor” appearance.

Although the preceding radiographic appearances of

pleural effusion are well known and easily recognized on pos-

teroanterior and lateral chest radiographs, these projections

are infrequently obtained in the ICU patient, and recognition

of pleural effusion in the supine patient may be difficult. In

supine patients, the most dependent regions of the pleural



Figure 7–17. Barotrauma in ARDS. A. Chest radiograph demonstrates diffuse lung consolidation secondary to ARDS.

Parenchymal stippling is present with lucent perivascular halos secondary to pulmonary interstitial emphysema. B. On

chest radiograph 4 days later, pneumomediastinum is now identified with extensive subcutaneous emphysema. C. In

another patient with ARDS, subpleural cysts (arrow) and parenchymal stippling due to pulmonary interstitial emphy-

sema are present.



CHAPTER 7

164

space are the posterior aspects of the bases and the lung apex.

Free pleural effusions layer posteriorly, resulting in a homoge-

neous increased density of the lower involved hemithorax.

Fluid also may accumulate at the apex of the thorax, resulting

in apical capping. These findings, however, are seen frequently

only in moderate or large pleural effusions, and small effu-

sions may not be detected on supine radiographs. Although

very small accumulations of pleural effusion can be detected

on lateral decubitus views, this projection is logistically diffi-

cult to obtain in the ICU patient.

Atelectasis and lung consolidation may be difficult to dis-

tinguish from a pleural effusion because they too may result

in elevation of the hemidiaphragm and decreased visibility of

lower lobe vessels. Cross-sectional imaging using ultrasound or

CT is very helpful in detecting small amounts of pleural effu-

sion and in distinguishing complicated pleural and parenchy-

mal processes. These imaging methods are also used frequently

to guide interventional procedures, including diagnostic thora-

centesis, drainage of empyema or malignant pleural effusions,

intracavitary fibrinolytic therapy, and sclerotherapy.

Ultrasound can be performed at the bedside and can easily

detect both free pleural effusions and loculated collections

(Figure 7–19). In most situations, ultrasound is the imaging

method of choice for guiding thoracentesis and may decrease

the incidence of iatrogenic pneumothorax. The percutaneous

drainage of pleural fluid collections with small catheters

instead of large-bore thoracostomy tubes has been shown to be

effective in treating both sterile and infected effusions.

Intracavitary fibrinolytic therapy, the installation of fibrinolytic

enzymes into the pleural space, has greatly improved the effec-

tiveness of pleural fluid drainage with smaller catheters.

CT is extremely sensitive in detecting even small amounts

of free pleural effusion, demonstrating loculations, and evalu-

ating the underlying lung parenchyma. The excellent contrast

resolution of CT allows demonstrations of regions of high atten-

uation secondary to blood or proteinaceous collections and

shows calcifications that are not apparent on chest radiographs.

By virtue of the cross-sectional perspective, air-fluid levels are

easily identified. In complicated cases, intravenous contrast

administration will help to differentiate pulmonary and pleural

processes in that perfused, consolidated lung will be enhanced,

whereas pleural processes will not (see Figure 7–7). The disad-

vantages of CT are its relatively high cost and the need for trans-

porting the critically ill patient to the radiology department.

Emamian SA et al: Accuracy of the diagnosis of pleural effusion on

supine chest x-ray. Eur Radiol 1997;7:57–60. [PMID: 9000398]

Ruskin JA et al: Detection of pleural effusions on supine chest radi-

ographs. AJR 1987;148:681–3. [PMID: 3493648]

Moulton JS: Image-guided management of complicated pleural

fluid collections. Radiol Clin North Am 2000;38:345–74. [PMID:

10765394]

Qureshi NR, Gleeson FV: Imaging of pleural disease. Clin Chest

Med 2006;27:193–213. [PMID: 16716813]



Pneumothorax

ESSENTIALS OF RADIOLOGIC

DIAGNOSIS



Identification of a visceral pleural line.



Absence of pulmonary vessels peripheral to visceral

pleural line.



Basilar hyperlucency in the supine patient.



Deep sulcus sign (supine patient).



Figure 7–18. Adult respiratory distress syndrome. CT

shows heterogeneous consolidation with subpleural air

cyst secondary to barotrauma.



Figure 7–19. Pleural effusion on ultrasound. Right

pleural effusion is seen as a region of low echogenicity

(asterisk) above the hyperechoic diaphragm (arrow).



IMAGING PROCEDURES

165

General Considerations

Pneumothorax is a frequent and serious complication in

the ICU. Iatrogenic pneumothorax may develop as a

sequela of invasive diagnostic or therapeutic procedures,

including central venous catheterization, endotracheal

intubation, tracheostomy, thoracentesis, pleural biopsy,

percutaneous lung biopsy, bronchoscopy, cardiothoracic or

abdominal surgery, and interventional abdominal proce-

dures to the liver and upper abdominal viscera.

Pneumothorax also may result from blunt chest trauma or

underlying lung diseases such as COPD, asthma, cystic fibro-

sis, and interstitial lung disease. Pneumothorax can compli-

cate the course of cavitary pneumonias due to infections

with M. tuberculosis, staphylococci, Klebsiella and other

gram-negative organisms, or fungi; similarly, there is an

increased incidence of pneumothorax in patients with AIDS

who develop Pneumocystis pneumonia. Finally, in patients

receiving positive-pressure mechanical ventilation, pneu-

mothorax may result from pulmonary interstitial emphy-

sema due to barotrauma.

In a recent study of pneumothorax in ICU patients, 35 of 60

patients (58%) who developed a pneumothorax during the

study period had procedure-related pneumothoraces. Patients

with pneumothoraces due to barotrauma or who had concur-

rent septic shock or a tension pneumothorax had a higher risk of

mortality than patients with postprocedural pneumothoraces.

Radiographic Features

A. Simple Pneumothorax—As with fluid in the pleural

space, the distribution of a pneumothorax is influenced by

gravity, lung elastic recoil, potential adhesions in the pleural

space, and the anatomy of the pleural recesses. In the upright

patient, air accumulates in the nondependent region of the

pleural space, the apex. Radiographically, a pneumothorax is

identified by separation of the visceral pleural surface from the

chest wall and the absence of pulmonary vessels peripheral

to the pleural line. A pneumothorax typically is better seen

on expiratory images because of a relative decrease in lung

volumes compared with the air in the pleural space.

Imaging in the supine position alters the radiographic

appearance of pneumothorax. In this position, the least

dependent regions of the pleural space are the anteromedial

and subpulmonary regions. Pleural air in the anteromedial

space results in sharp delineation of mediastinal contours,

including the superior vena cava, the azygos vein, the heart

border, the inferior vena cava, and the left subclavian artery.

The accumulation of air in the subpulmonary region is seen as

a hyperlucent upper quadrant of the abdomen; a deep, hyper-

lucent lateral costophrenic sulcus (“deep sulcus sign”); sharp

delineation of the ipsilateral diaphragm; and visualization of

the inferior surface of the lung (Figure 7–20). Air can accumu-

late in the apicolateral pleural space in the supine patient just

as in the erect patient, especially when a large pneumothorax

is present. In the presence of lower lobe collapse, air can

accumulate in the posteromedial pleural recess. This results in

a sharp delineation of the posterior mediastinal structures,

including the descending aorta and the costovertebral sulcus.

Subtle pneumothoraces may require other projections for

detection, such as decubitus or cross-table lateral views. CT

is an excellent method for diagnosing a pneumothorax not

demonstrated on plain chest radiographs.

Several conditions may be confused with a pneumothorax.

Pneumoperitoneum may result in a hyperlucent upper

abdomen, mimicking pneumothorax. Skin folds can be con-

fused with apicolateral pneumothorax but should be recog-

nized when they extend outside the bony thorax or are traced

bilaterally. Pneumomediastinum may simulate medial pneu-

mothorax, but pneumomediastinum may cross the midline

and extend into the retroperitoneum.

B. Tension Pneumothorax—Recognition of even small

pneumothoraces is crucial to prevention of progressive accu-

mulation of pleural air collections, particularly in patients

being maintained on mechanical ventilation. Tension pneu-

mothorax occurs when the pressure of air in the pleural space

exceeds ambient pressure during the respiratory cycle. With

this pressure gradient, air enters the pleural space on inspi-

ration but is prevented from exiting the pleural space during

expiration due to a check-valve mechanism. A tension



Figure 7–20. Pneumothorax in a supine patient with

ARDS. Chest radiograph demonstrates a large right pneu-

mothorax with intrapleural air adjacent to the diaphragm

and evidence of a deep sulcus (arrow). The margin of the

right hemidiaphragm is obliterated by adjacent adhesions.



CHAPTER 7

166

pneumothorax may result in acute respiratory distress and, if

untreated, cardiopulmonary arrest and death. The diagnosis

of tension pneumothorax is made clinically, reflecting the

hemodynamic sequelae of impaired venous return to the

right side of the heart. Radiographic signs include displace-

ment of the mediastinum toward the contralateral thorax,

inferior displacement or inversion of the diaphragm, and

total lung collapse (Figure 7–21). However, significant hemo-

dynamic compromise can exist in the absence of these find-

ings. Adhesions may prevent mediastinal shift, and lung

collapse may not occur in patients with stiff lungs such as

those with ARDS. A small pneumothorax may convert to a

tension pneumothorax, particularly in patients receiving

mechanical ventilatory support. In patients with ARDS,

poorly compliant lungs and pleural adhesions may result in

difficulty identifying a pneumothorax on portable chest

radiographs, and CT may be particularly useful in the diag-

nosis of loculated pneumothorax and in guiding appropriate

chest tube placement.

Chen KY et al: Pneumothorax in the ICU: Patient outcomes and

prognostic factors. Chest 2002;122:678–83. [PMID: 12171850]

Kong A: The deep sulcus sign. Radiology 2003;228:415–6. [PMID:

12893899]

Moss HA, Roe PG, Flower CDR: Clinical deterioration in ARDS:

An unchanged chest radiograph and functioning chest drains

do not exclude an acute tension pneumothorax. Clin Radiol

2000;55:637–51. [PMID: 10964737]

Rankine JJ, Thomas AN, Fluechter D: Diagnosis of pneumothorax

in critically ill adults. Postgrad Med J 2000;76:399–404. [PMID:

10878196]

Woodside KJ et al: Pneumothorax in patients with acute respira-

tory distress syndrome: Pathophysiology, detection, and treat-

ment. J Intensive Care Med 2003;18:9–20. [PMID: 15189663]



Pulmonary Interstitial Emphysema

& Pneumomediastinum

ESSENTIALS OF RADIOLOGIC

DIAGNOSIS



Pulmonary interstitial emphysema: Perivascular “halo”

(air surrounding pulmonary vessels seen on end), linear

radiolucencies radiating toward the hila, irregular radi-

olucent mottling, parenchymal cysts, or collections of

air along visceral pleural surface.



Pneumomediastinum: Linear lucencies adjacent to the

heart and aortic arch, descending aorta, and great ves-

sels. May have subcutaneous emphysema with linear

radiolucencies extending along tissue planes in the

chest wall and neck.

General Considerations

Barotrauma is a serious and frequent complication in the

ICU patient. Defined as damage secondary to the presence

of extraalveolar or extraluminal air, the incidence is highest

in patients being supported by mechanical ventilation.

Alveolar overdistention and an increased intraalveolar pres-

sure gradient from alveolus to vascular sheath allow rupture

of air into the interstitial space along the perivascular

sheaths, resulting in pulmonary interstitial emphysema.

Reduction in the caliber of pulmonary vessels—as well as

general and local alveolar overinflation—contributes to the

pressure gradient, causing alveolar rupture. Although com-

monly associated with mechanical ventilation, barotrauma

may also result from coughing, straining, trauma, pneumo-

nia, a Valsalva maneuver, anesthesia or resuscitation, partu-

rition, positive-pressure breathing, and asthma. Other

manifestations of barotraumas develop because air from

ruptured alveoli follows the path of least resistance. Air dis-

sects centrally to cause pneumomediastinum and dissects

via the cervical fascial planes, resulting in subcutaneous

emphysema in the neck and chest wall. Air also can dissect

from the mediastinum into the abdomen, leading to

retroperitoneal air and pneumoperitoneum or into the

pleural space resulting in a pneumothorax.

Barotrauma has a high incidence in patients with ARDS. In

one study of 15 patients with ARDS—all requiring positive-

pressure ventilation—radiographic evidence of pulmonary inter-

stitial emphysema was found in 87%. Although there was no

correlation with positive end-expiratory pressure or mean airway



Figure 7–21. Spontaneous tension pneumothorax. The

left lung is completely collapsed, with visualization of a vis-

ceral pleural line and hyperlucency of the thorax. The medi-

astinum is shifted to the right, and there is depression of the

left hemidiaphragm consistent with tension pneumothorax.