Correlation between CT-based phenotypes and serum biomarker in interstitial lung diseases.

Area of Science:

• Pulmonary Medicine
• Radiology
• Biomarkers

Background:

• Quantitative CT analysis and serum KL-6 levels are crucial for interstitial lung disease (ILD) management.
• Understanding the association between CT findings and KL-6 is vital for improved ILD prognostication.

Purpose of the Study:

• To investigate the correlation between quantitative CT-based phenotypes and serum KL-6 levels in ILD patients.
• To evaluate the predictive performance of combined CT features and KL-6 for respiratory hospitalizations in ILD.

Main Methods:

• Retrospective observational study (June 2020-March 2022) at a tertiary hospital.
• Deep learning-based quantitative CT analysis of reticulation, GGO, honeycombing, and consolidation.
• Correlation analysis between CT phenotypes and serum KL-6; evaluation of predictive performance for hospitalization.

Main Results:

• 131 ILD patients (median age 67) were analyzed.
• Significant positive correlations found between reticulation, GGO, consolidation, honeycombing extents and KL-6 levels (p < 0.05 for all).
• Combined reticulation and KL-6 achieved an AUC of 0.810 (p < 0.001) for predicting respiratory hospitalizations.

Conclusions:

• Quantitative CT features and serum KL-6 levels are positively correlated in ILD.
• The combination of reticulation and KL-6 effectively predicts respiratory hospitalizations in ILD patients.
• This combined approach aids in monitoring and prognosticating diverse ILD types.