Vessel-related structures predict UIP pathology in those with a non-IPF pattern on CT

Area of Science:

• Radiology
• Pulmonary Medicine
• Pathology

Background:

• Interstitial lung diseases (ILDs) encompass a heterogeneous group of disorders.
• Distinguishing usual interstitial pneumonia (UIP) from other ILDs is crucial for prognosis and treatment.
• Current diagnostic methods rely on a combination of imaging and histopathology, which can have limitations.

Purpose of the Study:

• To evaluate if quantitative imaging of vessel-related structures (VRS) can identify patients with a non-IPF diagnosis CT pattern who are likely to have UIP histologically.
• To assess the diagnostic performance of VRS compared to standard qualitative CT assessment for pathological UIP.

Main Methods:

• Retrospective analysis of chest CT scans from patients with ILD and multidisciplinary diagnosis.
• Utilized the Computer-Aided Lung Informatics for Pathology Evaluation and Rating (CALIPER) program for quantitative analysis of CT patterns.
• Correlated quantitative data (reticulation, honeycombing, VRS) with pathological diagnosis and qualitative CT patterns.

Main Results:

• Quantitative VRS, reticulation, and honeycombing volumes were associated with UIP on univariate analysis.
• Only VRS remained significantly associated with UIP pathology on multivariable analysis (p = 0.013).
• A VRS cut-off of 173 cm³ demonstrated sensitivity (55.9%) and specificity (80.4%) for pathological UIP, comparable to qualitative CT assessment.

Conclusions:

• Volume of vessel-related structures (VRS) is associated with UIP pathology.
• VRS effectively differentiates UIP in patients with a non-IPF diagnosis CT pattern.
• High VRS values may obviate the need for invasive pathology in suspected UIP cases, serving as an adjunct to CT imaging.