The purpose of this study was to assess the value of HRCT for determining the cause of subpleural opacities.

We evaluated 49 cases of subpleural opacities on HRCT scan, among with the patients with subpleural opacities seen on the conventional chest radiographs. Two "blinded" reviewers retrospectively analyzed the CT scans by working in consensus.

The patients consisted of COP (n=14), NSIP (n=13), UIP (n=10), fibrosis associated with connective tissue disease or drug toxicity (n=4), CEP (n=4), Churg-Strauss syndrome (n=2), DIP (n=1) and AIP (n=1). The predominant findings were consolidation (57%) with a peribronchovascular distribution (57%) in the COP patients, GGO (69%) and the associated focal reticular densities (61%) in the NSIP patients, and reticular or reticulonodular densities with a paucity of GGO in the UIP patients (100%). For the diagnosis of COP, NSIP and UIP, the use of HRCT demonstrated a high sensitivity (86%, 85% and 90%, respectively), specificity (97%, 86% and 95%) and accuracy (94%, 86% and 94%).

Although an overlap of CT findings is seen for diseases showing subpleural opacities, consolidation with a subpleural and peribronchovascular distribution is highly suggestive for COP, subpleural GGO is highly suggestive of NSIP, subpleural reticular or reticulonodular densities with a paucity of GGO is highly suggestive of UIP, and subpleural consolidation accompanied by reticular densities is suggestive of fibrosis.