A stepwise multi-disciplinary algorithm for diagnosis of fibrosing lung diseases contributing MDCT, MRI, and PET/CT: a study on 250 patients using significance and validation analyses

Abstract

Abstract Background The new guidelines limited the use of lung biopsy in the evaluation of lung fibrosis because of its hazards. The differential diagnosis of interstitial pulmonary fibrosis (IPF) or usual interstitial pneumonia (UIP) is challenging because of overlapping multi-detector computed tomography (MDCT) morphologic features between interstitial and non-interstitial fibrosing lung diseases. Scar carcinoma is a serious complication that needs to be excluded in certain conditions. Aim of the work: To achieve a multi-disciplinary algorithm for the diagnosis of fibrosing lung diseases to limit the need for lung biopsy by combining the clinico-laboratory and radiological roles. Results This study included two major steps. The first step (prevalence/significance analysis of the contributing parameters for the diagnosis of fibrosing lung diseases) was retrospectively conducted on 150 patients pathologically proved with fibrosing lung disease during the period between January/2016 and April/2018. Based on a P-value < 0.001, honeycombing bronchiectasis was significant to IPF. Basal traction bronchiectasis/bronchiolectasis was relevant to fibrosing non-specific interstitial pneumonia (NSIP). "Head cheese" CT-sign, history of allergen exposure, blood eosinophilia, and broncho-alveolar lavage (BAL) lymphocytosis were relevant to chronic hypersensitivity pneumonitis (HP). Upper peripheral lung fibrosis was significant to pulmonary tuberculosis (TB) and pleuroparenchymal fibroelastosis (PPFE). Cavitations, tree-in-bud, and calcific nodules were relevant to TB, while the "platy-thorax" CT-sign was relevant to PPFE. The upper peribronchovascular fibrosis was relevant to sarcoidosis and progressive massive fibrosis (PMF); additionally, calcific changes were relevant to PMF. Bright T2-signal, diffusion weighted-image (DWI) restriction in magnetic-resonance imaging (MRI), and high standardized uptake value (SUV) in positron emission tomography (PET-CT) were significant to scar carcinoma. Eventually, an algorithm was created. The second step (validation analysis) prospectively targeted 100 patients initially diagnosed with lung fibrosis during the period from June/2018 to June/2022. It revealed 83.3–100% sensitivity, 96.3–100% specificity, 85.7–100% PPV, 96.4–100% NPV, and 96–100% accuracy, with balanced accuracy = 0.91–1. Four consulting radiologists and two consulting pulmonologists participated in this study. Conclusions A valid stepwise multi-disciplinary algorithm was proposed for the diagnosis of interstitial and non-interstitial fibrosing lung diseases to limit the need and hazards of lung biopsy. It contributed significant clinico-laboratory data, MDCT features, T2-WI and DWI-MRI findings as well as PET/CT results.

Keywords

• Algorithm
• Fibrosing lung diseases
• CT
• MRI
• PET/CT