Ab0898 lung ultrasonography for the screening and prognostic stratification of interstitial lung disease in systemic sclerosis: results from a systematic literature review

Background Interstitial lung disease (ILD) affects almost all systemic sclerosis (SSc) patients and is the major cause of morbidity and mortality. Chest high resolution computed tomography (HRCT) is today the gold standard imaging method to diagnose and disclose the characteristic SSc ILD features. More than a decade ago, ultrasonography (US) has been shown to be a less invasive and radiation-free methodology to screen and show the lung involvement in SSc. Objectives To evaluate and confirm the reliability of lung US as a screening tool for SSc ILD. Methods A systematic literature search in several databases was conducted following the PICO framework on July 2022. Two researchers independently screened abstracts and titles and full texts were subsequently reviewed to determine eligibility (original research articles enrolling adult SSc patients undergoing lung US and HRCT). Data from eligible articles were extracted and risk of bias (RoB) was assessed with validated tools. Owing to extensive interstudy heterogeneity, narrative summaries had to be used to present the data. Results Out of 1656 retrieved articles, only 95 were selected for full text evaluation, and only 15 were considered eligible for inclusion. In these15 studies, published between 2014 and 2022, only 3 of these enrolled a control group of normal subjects. 14 studies had a cross-sectional design and 1 study was a prospective longitudinal study. In 10 studies, information about patient characteristics (SSc subset and/or autoantibody positivity) were lacking (67%). The studies were highly heterogeneous in terms of type of US probe used (2,5-3,5 MHz vs 7-10 MHz), number of scanned sites (ranging between 18 to 72) and US score/threshold to classify the US findings as ILD. The sensitivity of lung US ranged between 73.58 and 100% while the specificity ranged between 55 to 92.3%: 7 studies did not report either sensitivity or specificity or both, and it was not possible to calculate it on the basis of data provided. In 7 studies, feasibility was explored, showing that the scanning of up to 72 sites always took less than 10 minutes. In 3 studies, a good to very good reproducibility (intraclass correlation coefficients=0.95-0.96 and Cohen’s k=0.72) was detected. The concordance between US and HRCT in identifying ILD pathological findings was around 80%. The only longitudinal study revealed that the number of B-lines at baseline was correlated with the change of the diffusing capacity for carbon monoxide (DLCO) after 12 months. Conclusion The results of our Systematic Literature Review confirm that lung US is a useful screening tool for SScILD. However, the data on the prognostic value of lung US are still too few to draw any definite conclusions. The implementation of routine lung US might allow an early identification and prompt treatment of SSc-ILD patients. However, the lung US procedure still need to be unified and harmonised (e.g. type of probe, number of scanned sites) together with the scoring systems/threshold to allow a comparison of research studies and a broad implementation of lung US in clinical practice. REFERENCES: NIL. Acknowledgements: NIL. Disclosure of Interests None Declared.