Predicting Outcomes in Idiopathic Pulmonary Fibrosis Using Automated Computed Tomographic Analysis

Area of Science:

• Pulmonary Medicine
• Radiology
• Medical Imaging

Background:

• Idiopathic pulmonary fibrosis (IPF) is a progressive lung disease with significant mortality.
• Quantitative computed tomographic (CT) measures may predict disease severity and mortality risk in IPF patients.
• Identifying reliable biomarkers is crucial for effective IPF drug trial design.

Purpose of the Study:

• To evaluate computer-derived CT measures, specifically vessel-related structures (VRS), for predicting functional decline and survival in IPF.
• To determine if VRS measures can serve as a cohort enrichment tool in IPF drug trials.
• To assess the potential of VRS to reduce required sample sizes in clinical trials.

Main Methods:

• Volumetric noncontrast CT imaging was performed on IPF patients meeting drug trial entry criteria.
• Computer-Aided Lung Informatics for Pathology Evaluation and Rating (CALIPER) software was used to derive CT features.
• Vessel-related structures (VRS) scores were analyzed for their predictive value regarding mortality and forced vital capacity (FVC) decline in discovery and validation cohorts.

Main Results:

• CALIPER-derived VRS scores were strong predictors of survival and FVC decline in both cohorts.
• VRS scores outperformed traditional pulmonary function measures, especially in patients with less extensive disease.
• Utilizing a VRS score threshold of 4.4% reduced the required sample size for an IPF drug trial by 26%.

Conclusions:

• A validated quantitative CT measure, the VRS score, effectively predicts outcomes in IPF patients.
• VRS scores can significantly reduce IPF drug trial population size, thereby lowering costs.
• VRS scores identify patients who benefit from antifibrotic medications, prolonging life and improving FVC.