The aim of this study was to establish the cut-off point of ultrasound (US) B-lines number for detecting the presence of significant interstitial lung disease (ILD) in patients with systemic sclerosis (SSc) (SSc-ILD) in relation to high-resolution computed tomography (HRCT) findings.
Consecutive SSc-ILD patients underwent chest HRCT, lung US (LUS), pulmonary function test, and clinical assessment. Exclusion criteria were represented by the presence of a coexisting congestive heart failure and a clinical history suggestive of lung or pleural diseases. HRCT images were scored for the presence of ILD by 2 readers, in accordance with the Warrick scoring system. US assessment was performed by a US skilled rheumatologist, blinded to HRCT results and clinical data, and included the bilateral evaluation of 14 lung intercostal spaces (LIS). In each LIS, the number of B-lines was recorded and summed. To test discriminant validity, we used the receiver operating characteristic (ROC) curve analysis applying a Warrick score of 7 as external criterion for the presence of SSc-ILD.
Forty patients completed the study. The US B-lines number and the Warrick score confirmed excellent correlation (Spearman rho: 0.958,
P
= .0001). The ROC curve analysis revealed that the presence of 10 US B-lines is the cut-off point with the greatest positive likelihood ratio (12.52) for the presence of significant SSc-ILD.
The detection of 10 B-lines is highly predictive for the HRCT presence of significant SSc-ILD. In SSc patients, the LUS assessment as first imaging tool may represent an effective model to improve the correct timing of chest HRCT.