Pos1249 lung ultrasound changes compared with automated quantitative computed tomography for detecting systemic sclerosis-associated interstitial lung disease

Background B-lines and pleural line (PL) changes represent lung ultrasound (LUS) findings for detection of systemic sclerosis-associated interstitial lung disease (SSc-ILD). The ≥10 B-lines cut-off was found to be closely related to moderate ILD on high-resolution chest tomography (CT). Recently proposed Fairchild’s criteria for PL evaluation resulted in high accuracy and reproducibility and a low spending time. Semi-quantitative PL irregularity score by Pinal-Fernandez proved to be superior to total B-lines number to detect SSc-ILD. Although several evidence-based LUS scores have been presented, there is currently a lack of studies involving a comprehensive LUS assessment and comparing different LUS scores with CT, particularly with automated quantitative CT (qCT) assessment. Objectives To evaluate the accuracy of quantitative and qualitative LUS B-lines and PL changes for SSc-ILD detection compared to CT with qCT assessment. Methods Consecutive SSc patients according to 2013 ACR/EULAR classification criteria underwent LUS by two certified blinded operators using a 14-scans method. The ≥10 B-lines cut-off and Fairchild’s criteria fulfilment were selected as qualitative findings. From quantitative point of view, total B-lines number and the sum of the PL score adapted from Pinal-Fernandez were collected. CT scans performed over a ±6 months period were recorded and evaluated by two thoracic radiologists, with further processing by automated texture analysis software. Results The study population consisted of 29 SSc patients (Table 1). Agreement between the two operators was almost perfect for Fairchild’s criteria [Cohen’s kappa (k) 0.84] and substantial for ≥10 B-lines cut-off (k 0.78). Both qualitative LUS scores were predictive of ILD presence on CT, with Fairchild’s criteria resulting in slightly more accuracy (Figure 1A). Results were confirmed on multivariate analysis, introducing confounders like age, disease duration, ongoing immunosuppressant therapy, and current/ever smoking (Fairchild’s criteria: p 0.0003 and B-line cut-off: p 0.03). All qualitative and quantitative LUS findings were found to be significantly associated with ILD extension on qCT (p<0.05). Total B-lines number correlated with extension of reticulations and PL quantitative score correlated with extension of both ground-glass and reticulations (Figure 1B and C). Lung bases PL quantitative score, but not B-lines number of the same site, was found to correlate with basal ILD extension on qCT (Figure 1D). Conclusion This study highlights the reliability and good accuracy of a comprehensive and systematic LUS assessment for SSc-ILD detection, compared to automated qCT. LUS is increasingly emerging as an accurate, feasible, low-cost, and radiation-free pre-CT screening tool. For these reasons, standardisation of LUS appears to be required, in order to achieve its wider use in clinical practice. References [1]Gutierrez M et al. Radiol Med , 2019 [2]Xie, H.Q et al. Arthritis Res Ther , 2019 [3]Tardella M et al. Medicine (Baltimore), 2018 [4]Pinal-Fernandez I et al. Clin Exp Rheumatol, 2015 [5]Fairchild R et al. Arthritis Care Res, 2021 Table 1 Data of study population, LUS assessment and qCT software analysis. Female/male, N° (%) 26/3 (89.6/10.4) Median age (years) [quartiles] 59 [49;70] Median disease duration (years) [quartiles] 8 [4;12] Diffuse/limited cutaneous disease, N° (%) 20/9 (69/31) Immunosuppressive therapy, N° (%) 15 (51.7) Interstitial Lung Disease (ILD), N° (%) 22 (75.9) Fairchild’s criteria for PL fulfilling, N° (%) 21 (72.4) ≥10 cumulative B-lines, N° (%) 21 (72.4) Total B-lines (median [quartiles]) 24 [8;57] Total PL score sum (median [quartiles]) 13 [4;22] Ground-glass (cm 3 / %) (median [quartiles]) 179 [30;578]/ 3.8 [0.65;15] Reticulations (cm 3 / %) (median [quartiles]) 47 [28;103]/ 1.2 [0.6;5.2] Total ILD (cm 3 / %) (median [quartiles]) 323 [70;652]/ 8.2 [1.6;19] Basal ILD (cm 3 ) (median [quartiles]) 191 [51;305] LUS, lung ultrasound; qCT, quantitative chest tomography; PL, pleural line. Acknowledgements: NIL. Disclosure of Interests None Declared.