Abstract Background Airway mucus plugging is a prevalent radiographic feature in chronic obstructive pulmonary disease (COPD) and is associated with airflow limitation, exacerbations, and mortality. However, the long-term natural history of mucus plugging and its relationship to COPD progression remain incompletely understood. Leveraging automated image analysis, we characterized 10-year trajectories of mucus burden in the COPDGene cohort. Methods We analyzed inspiratory CT scans from COPDGene at baseline, 5 years, and 10 years. An automated deep learning-based pipeline detected and segmented mucus plugs to derive four quantitative biomarkers: 1) Mucus plug score, 2) total mucus volume, 3) mean mucus volume (intra-subject), and 4) intra-subject mucus volume variability. Longitudinal change was estimated for each biomarker. Participants were stratified by baseline COPD status, baseline mucus plug status, and (c) clinical COPD progression defined by FEV1 decline categories. We compared rates of mucus progression across groups and determined the proportion of subjects demonstrating incident, persistent, or progressive mucus plugging. Results Among 5,296 participants with baseline and 5-year CT scans, and 2,565 with 10-year follow-up, mucus burden increased over time. Across the cohort, the mucus plug score rose by 0.63 ± 2.08 and total mucus volume increased by 26.7 ± 105.0 µL over 10 years. Mucus accumulation was common: from baseline to 5 years, ∼30% of participants without plugs developed new plugs, and ∼51% with severe burden (6) remained severe; transitions from 5 to 10 years were similar. Individuals with COPD experienced greater progression, with higher 10-year increases in mucus plug score (0.96 ± 2.74 vs 0.45 ± 1.58, p0.001) and total mucus volume (39.9 ± 139.4 vs 19.6 ± 80.1 µL, p0.001). Heterogeneity of mucus distribution increased more in COPD (p=0.019). Participants without baseline plugs showed greater mucus volume gain (35.4 ± 80.7 vs 12.0 ± 135.4 µL, p0.001) and increases in mean mucus volume and variability (p0.001), suggesting early acquisition in initially plug-free individuals. The greatest increases occurred among spirometric progressors, who demonstrated ΔMPScore 1.59 ± 2.98 and ΔTotal mucus 53.2 ± 143.0 µL, exceeding both stable COPD and controls (p0.001). Conclusion In COPDGene, mucus burden continues to increase for a decade, especially among participants with COPD, baseline mucus plugs, and spirometric progression. Progressive COPD showed 2-fold greater mucus accumulation than stable disease. These findings support mucus plugging as a progressive and clinically relevant airway phenotype and a potential therapeutic target and imaging biomarker in COPD. This abstract is funded by: This work was supported by NHLBI grants 1R01HL149877, U01 HL089897, and U01 HL089856 and by NIH contract 75N92023D00011
Estepar et al. (Fri,) studied this question.