• Chiara Romei, Laura M Tavanti, Alessandro Taliani, Annalisa De Liperi, Ronald Karwoski, Alessandro Celi, Antonio Palla, Brian J Bartholmai, and Fabio Falaschi.
• Azienda Ospedaliero-Universitaria Pisana, Dipartimento Diagnostica e Immagini, UO Radiodiagnostica 2, Via Paradisa 2, 56126, Pisa, Italy. Electronic address: ch.romei@ao-pisa.toscana.it.
• Eur J Radiol. 2020 Mar 1; 124: 108852.

PurposeTo investigate the role of a quantitative analysis software (CALIPER) in identifying HRCT thresholds predicting IPF patients' survival and lung function decline and its role in detecting changes of HRCT abnormalities related to treatment and their correlation with Forced Vital Capacity (FVC).MethodsThis retrospective study included 105 patients with a multidisciplinary diagnosis of IPF for whom one HRCT at baseline and concomitant FVC were available. HRCTs were evaluated with CALIPER and the correlation between FVC and radiological features were assessed. Radiological thresholds for survival prediction and functional decline were calculated for all patients. Fifty-nine patients with at least 2 serial HRCTs were classified into two groups based on treatment. For patients for whom a FVC within 3 months of the HRCT was available (n = 44), the correlation of radiological and clinical progression was evaluated.ResultsThe correlation between FVC and CALIPER-derived features at baseline was significant and strong. A baseline CALIPER-derived interstitial lung disease (ILD%) extent higher than 20 % and pulmonary vascular related structures (PVRS%) score greater than 5 % defined a worse prognosis. A significant progression of CALIPER-derived features in all patients was found with a faster increase in untreated patients. ILD% and PVRS% changes during follow-up demonstrated strong correlations with FVC changes.ConclusionsCALIPER quantification of fibrosis and vascular involvement could distinguish disease progression in treated versus untreated patients and predict the survival. The changes in CALIPER-derived variables over time were significantly correlated to changes in FVC.Copyright © 2020 Elsevier B.V. All rights reserved.

21 keywords...

hide…