Extracellular vesicles from bronchoalveolar lavage fluid provide insights into the inhaled corticosteroids treatment response in COPD

Inhaled corticosteroids (ICS) are widely used to treat chronic obstructive pulmonary disease (COPD), but treatment responses vary among individuals. Identifying biomarkers that can improve our understanding of disease mechanisms and help predict ICS responsiveness is urgently needed. Extracellular vesicles (EVs), key mediators of intercellular communication, may offer novel insights and serve as a potential biomarker source. 34 COPD patients participated were treated for 6 months with either placebo or ICS (500 µg fluticasone ± 50 µg salmeterol). Lung function (FEV1% predicted, FEV1/FVC%, and RV/TLC% predicted) was assessed at baseline and 6 months. Proteins from BALF-derived EVs were analyzed at both time points using label-free quantitative proteomics. Weighted gene co-expression network analysis was applied to identify EV protein modules associated with lung function (n = 24) at baseline. Baseline EV protein levels were further correlated with changes in lung function after ICS treatment. Statistical analyses were performed in R (v4.3.2), using Mann-Whitney U (two groups) or Kruskal-Wallis with Dunn's post hoc (multiple groups), Student's t-test for paired data, and Pearson correlation. Statistical significance was set at p < 0.05. Thirteen EV protein co-expression modules were identified. Each module was assigned a color-based label. The red and salmon modules showed significant associations with baseline lung function: FEV1% predicted (r = - 0.46, p = 0.02) and FEV1/FVC% (r = 0.43, p = 0.04), respectively. Furthermore, 25 proteins from the red and 11 from the salmon module were significantly correlated with lung function improvements post-ICS treatment. Members of the cystatin (CST) superfamily, particularly CST1, showed strong correlations with ΔFEV1% predicted (r = 0.61, p = 0.003) and ΔFEV1/FVC% (r = 0.46, p = 0.035). These proteins also exhibited contrasting expression patterns between ICS responders and non-responders, suggesting a potential role in treatment sensitivity and links to type 2 inflammation. Our findings highlight the potential of BALF-derived EVs as a biomarker source for predicting ICS responsiveness in COPD. The CST family, especially CST1, potentially serves as a valuable indicator for identifying patients who are more likely to benefit from ICS treatment. ClinicalTrials.gov: NCT00158847, pre-registered April 2000.