Supplementary file 1_Elevated plasma eosinophil-derived neurotoxin is associated with echocardiographic pulmonary vascular involvement in stable COPD.docx

The primary analysis estimated the adjusted EDN difference between COPD groups using ordinary least-squares regression with HC3 robust standard errors. Covariates were age, sex, forced expiratory volume in 1 s, diffusing capacity for carbon monoxide, and room-air arterial oxygen tension. Secondary models examined adjusted associations with tricuspid regurgitation (TR) pressure gradient and echocardiographic pulmonary artery-to-aorta ratio (PA/Ao).</p>Results<p>Median plasma EDN was 181.58 ng/mL (interquartile range [IQR], 164.98–202.53) in participants with echocardiographic pulmonary vascular involvement and 125.66 ng/mL (IQR, 113.14–140.54) in those without such involvement.

The adjusted group difference was 48.93 ng/mL (95% CI, 30.04–67.81; P < 0.001). Secondary slopes were positive for TR pressure gradient (0.86 ng/mL per mmHg; 95% CI, 0.07–1.65; nominal P = 0.033) and PA/Ao (93.40 ng/mL per unit; 95% CI, 9.71–177.09; nominal P = 0.030); Holm-adjusted P = 0.059 for both. Leave-one-case-out estimates ranged from 44.59 to 53.37 ng/mL, with every lower 95% confidence bound above zero.</p>Conclusion<p>After adjustment for measured respiratory variables, plasma EDN was higher in stable COPD with echocardiographic pulmonary vascular involvement.

The corresponding patterns observed for TR pressure gradient and PA/Ao provide additional rationale for evaluating EDN in external cohorts with standardized echocardiography, paired eosinophil measurements, and invasive hemodynamic assessment.</p>