High resolution computed tomography as a predictor of lung histology in systemic sclerosis.

Abstract

BACKGROUND: The relative proportions of fibrosis and inflammation seen by open lung biopsy examination is a predictor of disease outcome in fibrosing alveolitis. This study was designed to assess the ability of high resolution computed tomography to predict the histological appearance of open lung biopsy specimens from patients with systemic sclerosis. METHODS: Twenty abnormal biopsy specimens from 12 patients were assessed; abnormalities were categorised as fibrotic (fibrosis exceeding inflammation) or inflammatory (inflammation equal to or exceeding fibrosis). Computed tomography appearances were scored for the lobe from which the biopsy specimen was taken; scans were graded from parenchymal opacification alone through to a reticular pattern alone. RESULTS: Two lobar appearances were identified on computed tomograms: amorphous parenchymal opacification equal in extent to reticulation (grade 3) and a predominantly reticular pattern (grade 4). There was a significant association between a fibrotic histological appearance and a grade 4 computed tomogram, and between an inflammatory histological appearance and a grade 3 computed tomogram. Computed tomography grade 4 was associated with a fibrotic histological appearance in 12 out of 13 lobes, and grade 3 with an inflammatory histological appearance in four out of seven lobes. CONCLUSION: Computed tomography discriminated between biopsy specimens that were predominantly fibrotic and a smaller group with a larger amount of inflammation.