The higher incidence of smoking in patients with pleural plaques may also contribute to the higher incidence of pulmonary pathologic findings seen in those patients.
Each form of pulmonary fibrosis associated with pleural plaques can be expected to have a distinct appearance on HRCT; for example, peribronchiolar fibrosis should be manifested by increased radiodensity in the centrilobular areas, and scar-related emphysema should be seen by an area of high attenuation adjacent to an area of low attenuation.
Although a strong association was found between these forms of fibrosis and pleural plaques, no single morphologic finding was exclusively present in the patients with pleural plaques; for example, peribronchiolar fibrosis was present in 39 percent (36/93) of the patients without plaques. This result is not entirely surprising, as peribronchiolar fibrosis, although associated with asbestosis, can be caused by smoking. These results suggest that radiologists should not seek a single “sign” by which to diagnose asbestosis, but rather should use radiography to evaluate the extent of pulmonary parenchymal disease, with the understanding that several disease processes may cause the same findings. More info
This conclusion is supported by recent case reports in which overreading of roentgenograms and scans from CT led to an erroneous diagnosis of asbestosis. Collins et al, in a review of 45 scans from CT in asbestos workers, presented a case in which pulmonary sarcoidosis was diagnosed as asbestosis. Similarly, Pilate et al have reported the subpleural curvilinear shadow on CT in patients with no known history of pulmonary disease.
While this study suggests that the diagnosis of asbestosis should not be based solely on radiographic findings, the possibility that focal nonspecific pathologic findings might combine to form patterns unique to asbestosis cannot be ruled out. Detailed pathologic-radiographic correlations, such as can be established with inflation-fixed lungs, would be the best method for detecting these patterns.