Early Recognition of Progressive Lung Fibrosis in Systemic Rheumatic Diseases: a Characterization of the Pulmonary Environment Through Extracellular Vesicles, Advanced and Functional Imaging

Background and rationale. Connective tissue diseases (CTDs) cover a broad range of systemic rheumatic disorders characterized by abnormal immune activation, chronic inflammatory response, and fibrosis of internal organs. The most prevalent is interstitial lung disease (ILD), a severe pulmonary complication seen in 10 to 50% of CTDs and a major determinant of disability and death. Prevalence and clinical course of CTD-ILDs vary widely and seem to be independent of treatment. Current screening and prognosis prediction strategies based on clinical variables and auto-antibodies are inadequate, and disease biomarkers are lacking. Extracellular vesicles (EVs) are lipid bilayer-bound particles secreted by most living cells and found in all body fluid. EVs have been proposed as biomarkers, therapy targets, or carriers due to unique properties including the possibility to be associated with producing cells and reflect their biology or cross biological barriers. The proteomic and transcriptomic analysis of EVs represents a very challenging procedure because EVs can circulate and cross biological barriers to reach specific target organs allowing a continuous cross-talk between serum and specific biologic niches, like lungs. Some promising data on EVs are available in idiopathic pulmonary fibrosis (IPF) and other chronic lung diseases, but data about their characteristics in CTD-ILD is still lacking.High-resolution computed tomography (HRCT) is the standard diagnostic tool in lung fibrosis, although interpreting the biological meaning of some abnormalities is often challenging. Inflammatory changes, metabolically active fibrosis, or established fibrosis could be difficult or impossible to distinguish. The use of artificial intelligence (AI) and functional chest imaging can overcome the intrinsic limitations of conventional imaging. AI-mediated evaluation of HRCT has the potential to quantitatively characterize lung texture, airways, and vessels, offering a non-invasive approach to comprehensively characterize pulmonary anatomy. Currently, accurate biological and clinical biomarkers for the early diagnosis and prognostic outcomes of CTD-ILDs are absent, posing a significant clinical challenge in managing these patients, including in therapeutic decision-making.In this scenario, our study will combine baseline serum EV-derived biomarkers and automated evaluation of HRCT scans via deep-learning AI, to characterize patients with CTD at high risk to develop ILD and patients that present a functional progression during 12 months of follow-up.

Objectives. The primary objective of this study is to compare the proteomic and transcriptomic profiles of serum EVs in CTD patients with and without ILD.

The secondary objective of this study is to compare the baseline proteomic and transcriptomic profiles of serum EVs in CTD-ILD and AI-detected difference in HRCT features between progressive and no progressive ILD patients at the 12-month follow-up.

Endpoints. The primary endpoint will be the differences in serum EV single-protein quantity and RNA expression between CTD patients with and without ILD, expressed as fold change.

The secondary endpoints will be the differences in serum EV single-protein quantity and RNA expression between progressive and non-progressive CTD-ILD patients, expressed as fold change, and AI-collected HRCT quantitative measures. The latter will include total lung volume, lung texture analysis (percentage of ground-glass opacities, reticulation, consolidation, and honeycombing), airways analysis (volume and wall thickness), and vessel analysis (vascular volume and vessel density).