Inspiratory high-resolution CT is typically obtained at the end of full inspiration using 1- to 2-mm collimation and 10-mm scan spacing. Expiratory high-resolution CT scans are obtained with thin collimation at the end of forced expiration. Usually, scans are obtained at two to six preselected levels or in a region of interest, depending on the reason for the study. Reconstruction with a high-frequency algorithm is mandatory. Because the lung may show unexpected air-trapping during exhalation, we recommend the routine use of expiratory scans at preselected levels even in patients with normal-appearing inspiratory scans.

Inhomogeneous lung attenuation is frequently encountered in patients with diffuse lung disease [1]. It may represent ground-glass attenuation, mosaic perfusion due to airway abnormality, or mosaic perfusion due to vascular abnormality. Expiratory scans may be valuable in distinguishing these causes.

Inspiratory scanning sometimes allows differentiation among the three common causes of inhomogeneous lung attenuation. For example, smaller vessels in the lower attenuation areas are seen in mosaic perfusion, regardless of cause (i.e., air-trapping or vascular obstruction), and the vessels are of comparable size in both high- and low-attenuation areas in ground-glass opacity (Figs. 1A,1B and 2A,2B). Small vessels in low-attenuation areas are present in 68-94% of cases of mosaic perfusion [1, 2]. The presence of bronchiectasis in low-attenuation areas favors mosaic perfusion due to air-trapping. However, “abnormalities” seen on inspiratory scanning do not always indicate a disorder and, in our experience, cannot be relied on for diagnosis [1] (Fig. 3A,3B).

When expiratory scans are available, mosaic perfusion due to air-trapping is often confidently discriminated from other causes of inhomogeneous opacity. The diagnostic value of expiratory scanning is further enhanced by the fact that normal-appearing lung parenchyma in patients with ground-glass attenuation or consolidation may in fact prove to represent areas of air-trapping. This may be seen in cases of bronchopneumonia, hypersensitivity pneumonitis, sarcoidosis, and concomitant airway obstructive and infiltrative lung diseases (Fig. 4A,4B).

Air-trapping may be seen in patients with normal inspiratory scans; it is reported in about 20% of patients clinically suspected of having chronic airways disease [3]. The differential diagnosis of this occurrence includes bronchitis (acute or chronic) (Fig. 5A,5B), bronchial asthma, bronchiolitis obliterans (Fig. 6A,6B), sarcoidosis (Fig. 7A,7B), hypersensitivity pneumonitis, and smoker's lung [4]. In patients with these diseases, pulmonary function test results are intermediate, falling between those of normal controls and those showing air-trapping and abnormalities on inspiratory scanning [4]. In selected cases, the expiratory scans reveal obstructive lung disease in an early stage, even when the pulmonary function test has normal findings.

Expiratory scanning is sensitive for the detection of air-trapping, which is a definitive sign of airway obstruction. Air-trapping is a frequent finding in bronchiectasis and often precedes the development of overt bronchiectasis (Figs. 8A,8B and 9A,9B). These areas of air-trapping correlate well with obstructive functional deficit. Air-trapping is seen more frequently in areas of mucoid impaction than in areas without.

In bronchiolitis obliterans, expiratory scanning enables early detection of air-trapping before any other noticeable abnormalities develop on inspiratory scans and is particularly useful in the early detection of this disease after lung transplantation. The extent of air-trapping is a good predictor of obstructive functional deficit.

Air-trapping and airway obstruction are seen not only in airway diseases but also in interstitial lung diseases, including hypersensitivity pneumonitis and sarcoidosis [5].

In hypersensitivity pneumonitis, chronic inflammatory infiltrates along small airways (cellular bronchiolitis) cause bronchiolar narrowing, and air-trapping is a common finding in many cases [6] (Fig. 10A,10B).

Sarcoidosis is a disease of the interstitium and usually shows a restrictive functional deficit. However, small airways obstruction is now considered an important feature [7]. Compression of airways by enlarged lymph nodes, the presence of endobronchial lesions, fibrotic scarring of endobronchial lesions, and bronchial distortion by peribronchial fibrosis and small airways abnormalities are considered to cause airway narrowing (Fig. 11A,11B).

The effectiveness of this technique depends on patient cooperation. Inadequate exhalation results in little increase in lung attenuation; this may be mistaken for diffuse air-trapping.

Expiratory scanning is considered more sensitive in detecting an obstructive deficit than the pulmonary function test, but this is not always true. Many patients who show obstructive functional deficit do not show air-trapping. This is partly because the pulmonary function test reflects overall lung function. On the contrary, air-trapping on the CT scan may reflect a more localized abnormality.

Mosaic perfusion due to vascular obstruction cannot always be discriminated from that due to air-trapping, even with expiratory scans. Some cases of pulmonary thromboembolism show evidence of air-trapping in the absence of an obvious airway abnormality (Fig. 12A,12B,12C). The mechanism of air-trapping in pulmonary thromboembolism is considered to be that recent embolus causes release of humoral mediators such as histamine and serotonin from circulating platelets, which in turn cause generalized and transient bronchoconstriction and asthmatic wheezing [8].

Expiratory high-resolution CT is useful in the differentiation of causes of inhomogeneous lung attenuation. Air-trapping in a patient with a normal-appearing inspiratory scan is a frequent finding. These patients generally show a mild obstructive functional deficit. Expiratory scans, revealing not only generalized but also localized air-trapping, may be more sensitive than the pulmonary function test in the diagnosis of obstructive lung disease. We recommend routine use of paired inspiratory and expiratory high-resolution CT in the diagnosis of diffuse lung diseases.

Presented at the annual meeting of the Radiological Society of North America, Chicago, November 1999.

Address correspondence to H. Arakawa.