Ground glass opacity (GGO) is described as a "hazy increased attenuation of lung, with preservation of bronchial and vascular margins; it is caused by partial filling of air spaces, interstitial thickening, partial collapse of alveoli, normal expiration, or increased capillary blood volume."1 GGO is a nonspecific finding, and the differential diagnosis of the many causes of GGO can be lengthy. An "ABCs" approach and a "pattern" approach to the interpretation of GGO on HRCT scanning of the lungs have previously been described.2,3 This paper provides an abbreviated review of the physiologic correlates of the HRCT scan findings of GGO, focusing on infiltrative processes and their different GGO patterns of presentation.

GGO can be patchy, resulting in a mosaic pattern of lung attenuation. Such a pattern can be seen in infiltrative lung disease, airway abnormalities (e.g., asthma, bronchiolitis obliterans), and chronic pulmonary vascular disease (e.g., chronic thromboembolic disease).4 The distinction between these three entities can be made by observing the size of the pulmonary vessels in the area of increased lung attenuation (increased in both airway disease and vascular disease, but not in infiltrative disease), and by examining air trapping on expiratory scans (indicating airway disease) (figure 1).

Pitfalls in the interpretation of GGO on CT scanning

As recognition of GGO is based on a subjective assessment of lung attenuation, it is important to understand the parameters that can interfere with lung density and make attenuation measurements unreliable.5,6 Window widths and levels that are too narrow can erroneously create the appearance of GGO by artificially "blooming" small structures. In evaluating for GGO, collimation ideally should be 1.0 to 1.5 mm. True GGO can not always be visualized with a thicker collimation because of volume averaging, and a thicker collimation sometimes results in a pseudo-GGO pattern. GGO is therefore best imaged with high-resolution CT (HRCT).

Lung attenuation normally increases homogeneously with expiration. This increased attenuation can obscure underlying pathologic GGO. Furthermore, if the expiratory nature of the examination is not recognized, an erroneous interpretation of pathologic GGO can be made.

Cardiac and respiratory motion also can create pseudo-GGO, which can be distinguished from pathologic GGO by recognizing the blurring and double images of vessels and fissures. GGO in the gravity dependent portions of the lungs is often seen as a result of microatelectasis, which can be differentiated from pathologic GGO by re-scanning the area of question with the patient in the prone position.

Infiltrative processes resulting in GGO

Many patterns of distribution of ground glass opacity can be seen on HRCT of the lungs. It is important to emphasize that most such disease processes can and do result in more than one pattern, often simultaneously; the patterns change depending upon the acuity or chronicity of the disease process. We have categorized the etiologies of GGO according to the most commonly seen patterns of distribution:

Diffuse pattern of GGO -Disease processes commonly resulting in a diffuse pattern of GGO on CT scanning are listed in table 1. Acute rejection is common after lung transplantation. However, differentiating between reperfusion edema, infection, and rejection can be difficult both clinically and radiographically. HRCT is reported to be 65% sensitive and 85% specific in making the diagnosis of acute rejection in the lung transplant population.7 The only significant HRCT finding in acute rejection (seen in 65% of these patients) is GGO, which is patchy and localized in mild rejection and widespread in severe rejection (figure 2). The main differential diagnosis in this group of patients is cytomegalovirus pneumonia, which can have an identical radio-graphic appearance.

Adult respiratory distress syndrome (ARDS) is a form of nonhydrostatic pulmonary edema, characterized by leaky capillary membranes. These leaks lead to extravasation of protein-rich fluid into the interstitial and alveolar spaces of the lung. Among the common causes of ARDS are aspiration, contusion, smoke inhalation, and sepsis. CT scan findings of ARDS include bilateral and gravity-dependent lung opacities.8 Early in the course of ARDS, all patients demonstrate GGO on CT, which persists on follow-up CT in 50% of patients (figure 3).9

Both cardiogenic and non-cardiogenic edema occurs when the capacity of the lung lymphatics to drain capillary transudate is exceeded. Etiologies include venous and lymphatic obstruction, increased capillary permeability, and hypoproteinemia.5 HRCT scan findings in patients with hydrostatic pulmonary edema include areas of GGO, interlobular septal thickening, peribronchovascular interstitial thickening, increased vascular caliber, pleural effusion, and thickening of fissures.10

Extrinsic allergic alveolitis, also called hypersensitivity pneumonitis, is a complex immunologic reaction by the lung, primarily to inhaled organic antigens. The clinical presentation may be acute, subacute, or chronic. HRCT scan findings will vary with the stage of disease. In the acute and subacute phases, findings include GGO (in 82%), small nodules (55%), a reticular pattern (36%), and air trapping.11 GGO correlates histologically with mononuclear cell infiltration of the alveolar walls.11 The distribution of GGO can be diffuse, patchy, or centrilobular (figure 4) in this condition.

Pulmonary hemorrhage can be diffuse, patchy, or focal, depending on the underlying cause, of which there are many.12 In the acute phase, CT scans show consolidation or GGO (figure 5). In the subacute phase, CT shows 1- to 3-mm nodules that are distributed in a uniform fashion, commonly accompanied by GGO and interlobular septal thickening.13

Infectious pneumonia of any cause (e.g., bacterial, viral, mycobacterial, fungal, and parasitic) can cause GGO to appear on HRCT scans. A diffuse pattern of GGO in the absence of associated CT scan findings is a characteristic presentation for cytomegalovirus pneumonia (CMV) and Pneumocystis carinii pneumonia (PCP). CMV is the most common viral pathogen to cause substantial morbidity and mortality in patients with AIDS,14 and is a relatively common complication in organ transplant recipients. In patients with AIDS and CMV pneumonia, CT scanning will show GGO, dense consolidation, bronchial wall thickening or bronchiectasis, and interstitial reticulation without air-space disease (although GGO may occur in isolation).15 In organ transplant recipients with CMV pneumonia, CT scanning shows small nodules, consolidation, GGO, and irregular lines (figure 6). The presence of an isolated ground glass infiltrate without additional findings in patients with AIDS is highly suggestive of PCP (figure 7).16

Patchy GGO patterns -Many of the causes of a patchy distribution of GGO on HRCT scanning, listed in table 2, may also result in a diffuse pattern of GGO. Pulmonary alveolar proteinosis is a disease of the lung that results in filling in of the alveoli by a periodic acid-Schiff-positive proteinaceous material that is rich in lipid.17,18 HRCT scanning of this disorder shows GGO, with an overlying branching pattern of white linear structures forming geometric shapes and outlining polygonal, triangular, and square forms.19,20 This pattern is often referred to as "crazy paving," and is characteristic, but not pathognomonic, of the diagnosis of alveolar proteinosis (figure 8).21 Other processes that can show a crazy paving pattern at HRCT scanning include ARDS, lipoid pneumonia, and PCP.

Focal GGO patterns -There is overlap between causes of diffuse, patchy, and focal distributions of GGO (table 3) with chest CT scanning. When pulmonary hemorrhage is due to focal neoplasm, trauma, or pulmonary infarction, a focal pattern of opacity results. Certain infections, such as lobar pneumonia, also may result in a focal pattern of GGO.

Bronchoalveolar lavage is a procedure used to diagnose pulmonary diseases and to identify predictors of prognosis. The technique involves injection of normal saline through a bronchoscope that is generally wedged into the lingular or middle lobe bronchus. Most, but not all of the fluid is aspirated back into the scope and examined for inflammatory and immune mediator cells and specific proteins.22 The residual fluid demonstrates a segmental or lobar distribution of GGO on CT scanning, which should suggest the possibility of recent bronchoalveolar lavage, especially if the GGO is observed in the right middle lobe or lingula.5

The "halo" pattern of GGO -A "halo" of GGO occasionally can be seen around a nodule or focal area of lung consolidation. Table 4 lists the processes known to produce the halo sign. It was first reported as a sign of early invasive pulmonary aspergillosis in patients with leukemia.23 The GGO represents a peripheral ring of hemorrhage or hemorrhagic infarction surrounding target lesions of pulmonary aspergillosis (figure 9). Several infectious and noninfectious causes of the CT halo sign have since been reported.24 In most patients, hemorrhagic nodules can be distinguished from nonhemorrhagic nodules by the presence of a halo of GGO.

Another cause of focal GGO, or a nodule with a surrounding halo of GGO, is the post-biopsy pseudo nodule. These pseudo nodules have been described in patients who have undergone lung transplantation and transbronchial lung biopsy,25 but they may be seen in any patient after lung biopsy.

A peripheral pattern of GGO -Processes that are known to result in a peripheral lung distribution of GGO with HRCT scanning are listed in table 5. This particular distribution pattern can be very helpful in narrowing the differential diagnosis, especially when combined with other clinical data and associated CT scan findings.

Bronchiolitis obliterans organizing pneumonia (BOOP) is a disease characterized histologically by the presence of granulation tissue plugs within respiratory bronchioles and alveolar ducts, and organizing pneumonia extending into the surrounding alveoli.26 CT scans show patchy GGO (in 8 to 75% of patients), nodules, or areas of consolidation with a predominantly peripheral (50% of patients), bilateral, and nonsegmental distribution (figure 10).27-29

Collagen vascular diseases are multisystem disorders characterized by vascular changes, fibrosis, and inflammation of connective tissue. Specific diseases include progressive systemic sclerosis (scleroderma), systemic lupus erythematosus, polymyositis/dermatomyositis, rheumatoid arthritis, and Sjogren's syndrome. GGO is seen on CT scanning in 63 to 100% of these patients,30 and is a sign of active inflammation in the absence of significant honeycombing, bronchiectasis, or other signs of lung fibrosis.31

Pulmonary contusion results from trauma to the chest wall and lung, with bleeding into the air spaces and lung interstitium. Generally, the cause is a compression injury with significant kinetic energy absorption adjacent to the site of chest wall injury. The CT scan appearance of lung contusion is that of ill-defined areas of GGO, consolidation, or both, usually with a peripheral, non-anatomic distribution (figure 11).32,33

Desquamative interstitial pneumonitis is characterized by alveolar filling with macrophages. The HRCT scan findings consist of GGO with a lower lung zone (73%) and a peripheral (59%) predominant distribution (figure 12). Usual interstitial pneumonitis, or idiopathic pulmonary fibrosis, results in a similar distribution of GGO on CT scanning but typically with more areas of honeycombing and traction bronchiectasis (figure 13).

Pulmonary toxicity has been associated with numerous drugs and a variety of radiographic and CT patterns. CT scanning shows nodular areas of GGO and consolidation, often with a peripheral distribution.35,36

Pulmonary eosinophilia occurs with a variety of conditions or diseases, or can be idiopathic. Chronic idiopathic eosinophilic pneumonia is characterized by multiple dense areas of opacity on chest radiographs and CT scans. In one study of patients with chronic eosinophilic pneumonia, the most common HRCT finding was GGO, usually adjacent to areas of consolidation, with a peripheral distribution.37 Acute idiopathic eosinophilic pneumonia is characterized by diffuse GGO and micronodules on chest radiographs and CT scans, often in a bronchovascular distribution.38

Sarcoidosis is a multisystemic disorder of unknown cause characterized by the presence of noncaseating granulomatous inflammation affecting various sites of the body, with a propensity to involve the respiratory tract. The most common HRCT scan findings of pulmonary sarcoidosis are irregularly thickened bronchovascular bundles (88%) and small nodules along vessels (50%).39 GGO is present in 75% of patients with sarcoidosis, which corresponds histologically with many granulomatous lesions, with or without perigranulomatous fibrosis, in the interstitium and alveolar septa around small vessels. Sarcoidosis can result in a predominantly peripheral distribution of GGO and/or consolidation, or a diffuse or patchy pattern of only GGO (figures 14 and 15).

Bronchovascular and centrilobular patterns of GGO -Processes that can result in GGO in a predominantly bronchovascular distribution include eosinophilic pneumonia and sarcoidosis. A predominantly centrilobular distribution of GGO has been described with both extrinsic allergic alveolitis and respiratory bronchiolitis. All reported cases of respiratory bronchiolitis have occurred in cigarette smokers.40-42 In the majority of these patients, HRCT shows GGO, which is often extensive, as the predominant finding.43 Pigmented macrophages within respiratory bronchioles and adjacent alveolar ducts and alveoli lead to the bronchovascular distribution of GGO.

Conclusion

GGO from infiltrative lung disease is a nonspecific finding on HRCT scans of the chest, correlating histologically with partial filling of air spaces, inflammatory or fibrotic interstitial thickening, or increased capillary

blood volume. GGO also can represent mosaic perfusion secondary to chronic vascular disease, or air trapping from small airways disease. It is important to correlate the HRCT pattern of GGO with clinical history and associated HRCT scan findings in developing a differential diagnosis. AR

References

1. Austin JHM, Müller NL, Friedman PJ, et al: Glossary of terms for CT of the lungs: Recommendations of the Nomenclature Committee of the Fleischner Society. Radiology 200:327-331, 1996.

2. Collins J, Stern EJ: Ground-glass opacity at CT: The ABCs. AJR 169:355-367, 1997.

3. Collins J, Stern EJ: Patterns of ground glass opacity on CT scanning of the chest. Postgraduate Radiology. In press.

4. Stern EJ, Swensen SJ, Hartman TE, Frank MS: CT mosaic pattern of lung attenuation: Distinguishing different causes. AJR 165:813-816, 1995.

5. Remy-Jardin M, Remy J, Giraud F, et al: Computed tomography assessment of ground-glass opacity: Semiology and significance. J Thorac Imaging 8:249-264, 1993.

6. Primack SL, Remy-Jardin M, Remy J, Müller NL: High-resolution CT of the lung: Pitfalls in the diagnosis of infiltrative lung disease. AJR 167:413-418, 1996.

7. Loubeyre P, Revel D, Delignette A, et al: High-resolution computed tomographic findings associated with histologically diagnosed acute lung rejection in heart/lung transplant recipients. Chest 107:132-138, 1995.

8. Tagliabue M, Casella TC, Zincone GE, et al: CT and chest radiography in the evaluation of adult respiratory distress syndrome. Acta Radiol 35:230- 234, 1994.

9. Owens CM, Evans TW, Keogh BF, Hansell DM: Computed tomography in established adult respiratory distress syndrome: Correlation with lung injury score. Chest 106:1815-1821, 1994.

10. Storto ML, Kee ST, Golden JA, Webb WR: Hydrostatic pulmonary edema: High-resolution CT findings. AJR 165:817-820, 1995.

11. Hansell DM, Wells AU, Padley SPG, Müller NL: Hypersensitivity pneumonitis: Correlation of individual CT patterns with functional abnormalities. Radiology 199:123-128, 1996.

12. Albelda SM, Gefter WB, Epstein DM, Miller WT: Diffuse pulmonary hemorrhage: A review and classification. Radiology 154:289-297, 1985.

13. Cheah FK, Sheppard MN, Hansell DM: Computed tomography of diffuse pulmonary hemorrhage with pathological correlation. Clin Radiol 48:89-93, 1993.

14. Wallace MJ, Hannah J: Cytomegalovirus pneumonitis in patients with AIDS: Findings in an autopsy series. Chest 92:198-203, 1987.

15. McGuinness G, Scholes JV, Garay SM, et al: Cytomegalovirus pneumonitis: Spectrum of parenchymal CT findings with pathologic correlation in 21 AIDS patients. Radiology 192:451-459, 1994.

16. Sider L, Gabriel H, Curry DR, Pham MS: Pattern recognition of the pulmonary manifestations of AIDS on CT scans. Radiographics 13:771-784, 1993.

17. Rosen SH, Castleman B, Liebow AA: Pulmonary alveolar proteinosis. N Engl J Med 258: 1123-1144, 1958.

18. Nhieu JTV, Vojtek AM, Bernaudin JF, et al: Pulmonary alveolar proteinosis associated with Pneumocystis carinii: Ultrastructural identification in bronchoalveolar lavage in AIDS and immunocompromised non-AIDS patients. Chest 98:801-805, 1990.

19. Godwin JD, Müller NL, Takasugi JE: Pulmonary alveolar proteinosis: CT findings. Radiology 169:609-613, 1988.

20. Murch CR, Carr DH: Computed tomography appearances of pulmonary alveolar proteinosis. Clin Radiol 40:240-243, 1989.

21. Franquet T, Giménez A, Bordes R, et al: The crazy-paving pattern in exogenous lipoid pneumonia: CT-pathologic correlation. AJR 170:315, 1998.

22. Fraser RS, Pare JAP, Fraser RG, Pare PD, (eds): Methods of clinical, laboratory, and functional investigation. In: Synopsis of diseases of the chest, ed 2, pp 141-164. Philadelphia, WB Saunders, 1994.

23. Kuhlman JE, Fishman EK, Siegelman SS: Invasive pulmonary aspergillosis in acute leukemia: Characteristic findings on CT, the CT halo sign, and the role of CT in early diagnosis. Radiology 157:611-614, 1985.

24. Primack SL, Hartman TE, Lee KS, Müller NL: Pulmonary nodules and the CT halo sign. Radiology 190:513-515, 1994.

25. Kazerooni EA, Cascade PN, Gross BH: Transplanted lungs: Nodules following transbronchial biopsy. Radiology 194:209-212, 1995.

26. Epler GR, Colby TV: The spectrum of bronchiolitis obliterans. Chest 83:161-162, 1983.

27. Müller NL, Miller RR: State of the art: Computed tomography of chronic diffuse infiltrative lung disease, part 2. Am Rev Respir Dis 142:1440-1448, 1990.

28. Corcoran HL, Renner WR, Milstein MJ: Review of high-resolution CT of the lung. Radiographics 12:917-939, 1992.

29. Bouchardy LM, Kuhlman JE, Ball WC, et al: CT findings in bronchiolitis obliterans organizing pneumonia (BOOP) with radiographic, clinical, and histologic correlation. J Comput Assist Tomogr 17:352-357, 1993.

30. Johkoh T, Ikezoe J, Kohno N, et al: High-resolution CT and pulmonary function tests in collagen vascular disease: Comparison with idiopathic pulmonary fibrosis. Eur J Radiol 18:113-121, 1994.

31. Remy-Jardin M, Remy J, Cortet B, et al: Lung changes in rheumatoid arthritis: CT findings. Radiology 193:375-382, 1994.

32. Wagner RB, Crawford WO, Schimpf PP: Classification of parenchymal injuries of the lung. Radiology 167:77-82, 1988.

33. Schnyder P, Gamsu G, Essinger A, Duvoisin B: Trauma. In: Moss AA, Gamsu G, Genant HK (eds): Computed tomography of the body, pp 311-323. Philadelphia, WB Saunders, 1992.

34. Hartman TE, Primack SL, Swensen SJ, et al: Desquamative interstitial pneumonia: Thin-section CT findings in 22 patients. Radiology 187:787-790, 1993.

35. Aronchick JM, Gefter WB: Drug-induced pulmonary disorders. Semin Roentgenol 30:18-34, 1995.

36. Patz EF, Peters WP, Goodman PC: Pulmonary drug toxicity following high-dose chemotherapy with autologous bone marrow transplantation: CT findings in 20 cases. J Thorac Imaging 9:129-134, 1994.

37. Ebara H, Ikezoe J, Johkoh T, et al: Chronic eosinophilic pneumonia: Evolution of chest radiograms and CT features. J Comput Assist Tomogr 18:737-744, 1994.

38. Tsunemi K, Kanayama I, Kondo T, et al: Acute eosinophilic pneumonia evaluated with high-resolution computed tomography. Intern Med 32:891-894, 1993.

39. Nishimura K, Itoh H, Kitaichi M, et al: Pulmonary sarcoidosis: Correlation of CT and histopathologic findings. Radiology 189:105-109, 1993.

40. Myers JL, Veal CFJ, Shin MS, Katzenstein A-LA: Respiratory bronchiolitis causing interstitial lung disease: A clinicopathologic study of six cases. Am Rev Respir Dis 135:880-884, 1987.

41. Yousem SA, Colby TV, Gaensler EA: Respiratory bronchiolitis-associated interstitial lung disease and its relationship to desquamative interstitial pneumonia. Mayo Clin Proc 64:1373-1380, 1989.

42. Myers J: Respiratory bronchiolitis with interstitial lung disease. Semin Respir Med 13:134-139, 1992.

43. Holt RM, Schmidt RA, Godwin JD, Raghu G: High resolution CT in respiratory bronchiolitis-associated interstitial lung disease. J Comput Assist Tomogr 17:46-50, 1993.

Dr. Collins is in the Department of Radiology at the University of Wisconsin Hospital and Clinics, in Madison, WI. Dr. Stern is in the Department of Radiology at Harborview Medical Center, University of Washington in Seattle, WA.