HRCT findings of collagen vascular disease-related interstitial pneumonia (CVD-IP): a comparative study among individual underlying diseases.

AIM: To identify characteristic high-resolution computed tomography (CT) findings for individual collagen vascular disease (CVD)-related interstitial pneumonias (IPs).

MATERIALS AND METHODS: The HRCT findings of 187 patients with CVD, including 55 patients with rheumatoid arthritis (RA), 50 with systemic sclerosis (SSc), 46 with polymyositis/dermatomyositis (PM/DM), 15 with mixed connective tissue disease, 11 with primary Sjögren's syndrome, and 10 with systemic lupus erythematosus, were evaluated. Lung parenchymal abnormalities were compared among CVDs using χ2 test, Kruskal-Wallis test, and multiple logistic regression analysis. A CT-pathology correlation was performed in 23 patients.

RESULTS: In RA-IP, honeycombing was identified as the significant indicator based on multiple logistic regression analyses. Traction bronchiectasis (81.8%) was further identified as the most frequent finding based on χ2 test. In SSc IP, lymph node enlargement and oesophageal dilatation were identified as the indicators based on multiple logistic regression analyses, and ground-glass opacity (GGO) was the most extensive based on Kruskal-Wallis test, which reflects the higher frequency of the pathological nonspecific interstitial pneumonia (NSIP) pattern present in the CT-pathology correlation. In PM/DM IP, airspace consolidation and the absence of honeycombing were identified as the indicators based on multiple logistic regression analyses, and predominance of consolidation over GGO (32.6%) and predominant subpleural distribution of GGO/consolidation (41.3%) were further identified as the most frequent findings based on χ2 test, which reflects the higher frequency of the pathological NSIP and/or the organising pneumonia patterns present in the CT-pathology correlation.

CONCLUSION: Several characteristic high-resolution CT findings with utility for estimating underlying CVD were identified.