Quantitative CT and COPD: cluster analysis reveals five distinct subtypes with varying exacerbation risks

Area of Science:

• Pulmonary Medicine
• Radiology
• Data Science

Background:

• Chronic Obstructive Pulmonary Disease (COPD) heterogeneity is increasingly recognized.
• Understanding COPD subtypes is crucial for personalized treatment strategies.
• Quantitative CT imaging offers detailed insights into lung structure.

Purpose of the Study:

• To characterize COPD heterogeneity using quantitative CT imaging.
• To identify distinct COPD subtypes through principal component analysis (PCA) and cluster analysis.
• To evaluate the clinical relevance of identified subtypes regarding acute exacerbation risk.

Main Methods:

• Utilized data from 1879 participants in the SPIROMICS study.
• Applied PCA and k-means clustering to quantitative CT parameters.
• Validated cluster stability using normalized mutual information (NMI) across 10 sampling cycles.

Main Results:

• Identified five COPD subtypes based on quantitative CT features:
• 1. Male-dominated low disease impact
• 2. Obesity with high symptom burden
• 3. Airway wall lesions
• 4. Upper lobe predominant emphysema
• 5. Severe emphysema
• Significant differences in acute exacerbation risk were observed among the five clusters.

Conclusions:

• Cluster analysis successfully identified five distinct COPD subtypes based on quantitative CT.
• These subtypes exhibit significant variations in baseline characteristics and acute exacerbation risk.
• The identified clustering is stable and reproducible, supporting its potential clinical utility.