ATS 2025: Prediction of 5-Year Mortality in COPD Using Visual Measures of Structural Changes on Chest CT

Introduction:

Globally, chronic obstructive pulmonary disease (COPD) ranks third as a leading cause of death. Various indices, markers, and scans are routinely used to assess COPD severity and predict related mortality. The Body mass index, airflow Obstruction, Dyspnea, and Exercise capacity (BODE) index has been shown to predict mortality; however, its utilization in routine practice is limited due to the need for multiple measurements and calculations. Chest computed tomography (CT) scans are widely performed and may offer a simpler tool to predict mortality in COPD patients.

Aim:

To evaluate whether structural changes in CT scans accurately predict all-cause mortality in COPD.

Methods:

1. Study Design: Multicentre analysis of 2000 chest CT scans from COPDGene participants
2. Baseline Characteristics: Mean age: 60.4 years (SD 8.2), 50% males (991 participants)
3. Visual CT Measures Evaluated at Full Inspiration:
   1. Coronary artery calcification (CAC) using the Weston score (threshold CAC ≥ 7)
   2. Interstitial lung abnormalities (ILA) (definitely present)
   3. Pulmonary artery to the aorta diameter ratio (PA: A ≥ 1.1)
   4. Centrilobular emphysema severity ≥ moderate
   5. Paraseptal emphysema per the Fleischner Society classification (definitely present)
   6. Bronchial wall thickening (BWT) (definitely present)
   7. Presence of mucus plug (MP score ≥ 2)
4. Statistical Analysis: A Cox model was used to predict 5-year mortality, with performance evaluated using Harrell’s c-index (95% CI via 1000 bootstraps) and compared to the BODE index on a test dataset split from an 80/20 training-validation setup.

Results:

• Visual CT assessments were performed for 2,000 participants.
• The c-index for the visual CT model was 0.78 (95% CI: 0.77–0.78), while the BODE index showed a slightly higher c-index of 0.81 (95% CI: 0.80–0.81).

Conclusion:

Visual assessment of simple structural changes on chest CT scans may accurately predict 5-year all-cause mortality in COPD patients with accuracy comparable to the more complex BODE index.

Am J Respir Crit Care Med 2025; 211: A7687

American Thoracic Society 2025 International Conference, May 18-21, San Francisco