Background: Quantitative computed tomography (CT) analysis has been proposed as a means of objectively assessing fibrotic interstitial pneumonia (IP) including idiopathic pulmonary fibrosis (IPF). We investigated whether percentages of high-attenuation areas (HAA%) and cystic areas (CA%) quantified from CT images were useful as indices of fibrotic IP.
Methods: CT images of 74 patients with fibrotic idiopathic interstitial pneumonias (IPF, 36; non-specific interstitial pneumonia, 9; unclassifiable idiopathic interstitial pneumonia, 29) were analyzed via in-house computer software, which automatically calculated HAA%, CA%, mean lung density (MLD), standard deviation of lung density (SD-LD), kurtosis, and skewness from CT attenuation histograms. These indices were compared in each instance with physiologic measures, visual fibrosis score, clinical diagnosis, radiologic CT pattern, and prognosis.
Results: HAA% correlated significantly with physiologic measures and visual fibrosis score to a moderate extent (%forced vital capacity, rs = -0.59; % carbon monoxide diffusion capacity, rs = -0.43; fibrosis score, rs = 0.23). Densitometric parameters (MLD, SD-LD, kurtosis, and skewness) correlated significantly with physiologic measures and fibrosis score (|rs| = 0.28-0.59). CA% showed no association with pulmonary functions but differed significantly between IPF and other interstitial pneumonias (IPs) (1.50 ± 2.41% vs. 0.41 ± 0.80%; P < 0.01) and between the definite usual interstitial pneumonia (UIP) pattern and other patterns (1.48 ± 2.38% vs. 0.55 ± 1.19%; P < 0.01). On univariate analysis, HAA%, MLD, SD-LD, kurtosis, skewness, fibrosis score, and definite UIP pattern all correlated with survival, with kurtosis alone identified as a significant predictor of mortality on multivariate analysis (hazard ratio = 0.67; 95% CI, 0.44-0.96; P = 0.03).
Conclusion: CA% and HAA% are novel quantitative CT indices with differing properties in fibrotic IP evaluations. HAA% largely reflects physiologic impairments, whereas CA% corresponds with diagnosis and HRCT pattern. Of the CT indices examined, kurtosis constituted the strongest predictor of mortality.