Volumetric CT analysis of normal lung as a predictor of mortality in idiopathic pulmonary fibrosis

Abstract

Background: Although semi-quantification methods of chest computed tomography (CT) have been utilized for prediction of prognosis of idiopathic pulmonary fibrosis (IPF), whether volumetric quantification of them is useful for this purpose is unknown. Objective: The aim of this study was to investigate the ability of normal lung quantification by threshold-based volumetric CT analysis for predicting mortality in IPF. Methods: A total of 179 IPF patients who were referred to Tosei General Hospital were retrospectively enrolled. CT analysis was performed by using a commercially available CT imaging analysis workstation (Fujifilim, Japan), and the volumes of whole lung (WL) as defined by CT value between -1000 Hounsfield unit (HU) and 0 HU, the volumes of normally attenuated lung as defined by CT value between -950 HU to −701 HU (NL), and its percentage of WL (NL%), were calculated. Univariate analysis were performed and hazard ratio (HR) and 95% confidence intervals (CI) were calculated to clarify the ability of predicting mortality. Multivariable logistic regression analysis was performed for comparing NL% and GAP (gender, age, and physiology) stage. (Ley B, et al. Ann Intern Med 2012; 156: 684-691.) Results: Univariate analysis showed that NL% and GAP stage were predictors of mortality (NL% [HR, 0.93, 95% CI, 0.92 to 0.96], GAP stage [HR, 3.81, 95% CI, 2.50 – 5.83]). Multivariable logistic regression analyses showed that NL% was approximately equivalent to GAP stage (Likelihood ratio chi-square, NL% : 25.5, GAP index : 20.7; p Conclusion: The measurement of NL% by threshold-based volumetric CT analysis is useful for predicting mortality of IPF.