Introduction

• Quantitative measurements of the extent of ILD or lung fibrosis on CT have been shown to be predictive of decline in FVC and short-term mortality in patients with various ILDs.
• More information is needed on the prognostic potential of quantitative CT measurements for clinical outcomes in patients with PPF.

Methods

The INBUILD trial and its sub-study

• Patients in the INBUILD trial had an ILD other than idiopathic pulmonary fibrosis, an extent of fibrosis on HRCT >10%, and met criteria for ILD progression within the prior 24 months.
• Patients were randomized to receive nintedanib or placebo, stratified by fibrotic pattern on HRCT (UIP-like fibrotic pattern or other fibrotic patterns).
• In a sub-study of the INBUILD trial, HRCT scans were taken at baseline, week 24 and week 52. Non-contrast volumetric HRCT was performed with contiguous slices with ≤1 mm thickness in supine position at full inspiration, with no iterative reconstruction.

Quantitative CT parameters

• The UCLA algorithm classifies abnormal lung tissue based on pixel density or texture.
  • Quantitative lung fibrosis (QLF) score: extent of reticular patterns with architectural distortion due to fibrosis
  • Quantitative honeycomb (QHC) score: extent of honeycomb cysts
  • Quantitative ground glass opacity (QGGO) score: extent of ground glass opacities
  • The total of these scores is the quantitative ILD (QILD) score.
• e-Lung is an artificial intelligence-developed image processing module. Features analyzed include:
  • Weighted Reticulovascular Score (WRVS): Measure of fibrosis combining reticular abnormalities and vascular structures
  • Total Disease Extent (TDE): Combines ground glass opacification and reticulovascular structures to provide a total ILD extent.

Analyses

• In the placebo group, we evaluated:
  • Associations between baseline QILD score, QLF score, e-Lung TDE and e-Lung WRVS (analyzed as continuous variables) and rate of decline in FVC over 52 weeks, using random coefficient regression.
  • Rate of decline in FVC over 52 weeks by baseline QILD score, QLF score, e-Lung TDE and e-Lung WRVS ≤ vs > median, using random coefficient regression.
  • Associations between baseline QILD score, e-Lung TDE, QLF score and e-Lung WRVS ≤ vs > median and time to the following, using Cox regression:
    • Absolute decline in FVC % predicted >10% or death over 52 weeks
    • Death over the whole trial

Results

A summary of the results can be found in the PDF version of the poster.

Conclusions

In the INBUILD trial in patients with PPF, a greater extent of ILD (QILD score or e-Lung TDE) or lung fibrosis (QLF score or WRVS) at baseline was associated with a greater rate of decline in FVC over 52 weeks and a greater risk of death over the whole trial.

These findings suggest that quantitative CT methods can facilitate the prediction of clinically relevant outcomes in patients with PPF.