Computed tomography of asbestos-related pleural abnormalities

To study the observer agreement in several asbestos-related pleural abnormalities and to define criteria to discriminate between pleural changes in workers with occupational disease, and those in controls.

Pleural abnormalities in spiral computed tomography of 602 construction workers with asbestosis or bilateral pleural plaques and 49 controls were reviewed by three radiologists using structured forms.

Intra- and inter-observer agreement (weighted kappa) was 0.4 or better with regard to the calcification, extent and thickness of pleural disease. These factors all correlated positively with the duration of asbestos exposure. There were significant differences in these pleural changes between the workers (mean extent per side 83 cm(2)) and controls (mean extent per side 40 cm(2)). Of the controls, 84% showed pleural lesions with an estimated extent of 10 cm(2) or more, bilateral in 64%. The extent of 45 cm(2) in pleural disease was the best value for discriminating between the controls and diseased workers, with a sensitivity of 82% and a specificity of 66%. The degree of pleural calcification, however, was the best discriminator between these groups, but quantitative methods are necessary for its use in the diagnostics of individuals.

The extent, calcification and thickness were well-repeatable indicators of benign pleural pathology and thus their use in future classification systems in computed tomography is recommended. In our material, the extent of 45 cm(2) and the degree of calcification were helpful in discriminating between pleural changes in workers with occupational disease, and those in controls who also presented marked pleural pathology.