Systemic Inflammatory Biomarkers Define Specific Clusters in Patients with Bronchiectasis: A Large-Cohort Study

Area of Science:

• Pulmonology
• Biostatistics
• Data Science

Background:

• Bronchiectasis is a complex lung disease with significant heterogeneity.
• Understanding patient phenotypes is crucial for effective management and treatment strategies.

Purpose of the Study:

• To identify distinct patient phenotypes in bronchiectasis using data mining techniques.
• To correlate these phenotypes with clinical characteristics, biomarkers, and disease severity.

Main Methods:

• Utilized hierarchical clustering (scikit-learn) on systemic biomarkers (neutrophils, eosinophils, lymphocytes, CRP, hemoglobin) from 1092 patients.
• Assessed disease severity using scores like FACED, EFACED, and Bronchiectasis Severity Index (BSI).
• Employed multivariate ordinal logistic regression for confirmation.

Main Results:

• Three clusters (mild, moderate, severe) were identified based on biomarkers.
• The severe cluster (Cluster #3) exhibited significantly higher disease severity scores (FEV1, age, colonization, extension, dyspnea, exacerbations, BSI).
• Cluster #3 also showed increased exacerbations, hospitalizations, Charlson index, inflammatory markers, Pseudomonas aeruginosa colonization, and COPD prevalence, with reduced airflow and diffusion capacity.

Conclusions:

• Clustering analysis effectively characterizes bronchiectasis patient heterogeneity.
• Identified phenotypes have significant clinical implications for managing complex bronchiectasis cases.
• Findings support tailored management strategies based on distinct patient phenotypes.