Lung CT density correlates with measurements of airflow limitation and the diffusing capacity

Abstract

We studied 80 subjects (63 M, 17 F; 23-82 yrs) and related lung computerized tomography (CT) density with age, height, spirometry, lung volumes, diffusing capacity and arterial blood gas tensions. These subjects demonstrated a wide range of physiological impairment (forced expiratory volume in one second (FEV1) 8-116% predicted; diffusing capacity (Kco) 15-139% predicted; arterial oxygen tension (Pao2) 38-91 mmHg). They ranged from normal subjects to patients with chronic respiratory failure. Lung density was derived from CT density histograms measured as both mean Emergency Medical Information (EMI) number (EMI scale: 0 = water, -500 = air, EMI number of normal lung tissue range approximately -200 to -450) and the lowest 5th percentile EMI number, the latter value being more likely to represent the density of lung parenchyma. Lung CT density correlated most strongly with airflow obstruction (EMI 5th percentile versus FEV1/forced vital capacity (FVC) % predicted, r = 0.73, p less than 0.001) and diffusing capacity (EMI 5th percentile versus Kco, r = 0.77, p less than 0.001). This suggests that reduction in lung density, which reflects loss of the surface area of the distal airspaces, is a major index of respiratory function in patients with smoking related chronic obstructive pulmonary disease (COPD). These data provide no indication of other factors such as small and large airways disease, and loss of elastic recoil, which may contribute to airflow limitation, or disruption of the pulmonary vascular bed which may also affect CT lung density.

Parenchymal emphysema measured by CT lung density correlates with lung function in patients with bullous disease

“…The latter value should relate better to the density of lung parenchyma, since the mean EMI value of the density histogram is affected by the presence within the lung field of high density pixels representing blood vessels, large airway walls, and small areas of atelectasis. This was confirmed in our two previous studies [14,15]. Measurements of respiratory function were correlated with the following er variables: I) Mean EMI number of non-bullous lung; 2) lowest 5th percentile EMI number of non-bullous lung; and 3) percentage area of the lungs occupied by bullae.…”

“…These subjects had a very wide range of age (63 males and 17 females; 23-82 yrs) and respiratory function (FEV 1 8-116% predicted; Dr.r.oN A 15-139% predicted). That study suggested that lung physical density (which relates to the severity of emphysema) is a major determinant of lung function [15]. We therefore compared the results obtained in those subjects with COPD [15] to the results obtained in the group of subjects in the present study with bullous disease, to determine whether the same relationships apply.…”

“…Correlation coefficients showing the relationships between respiratory function and er measurements are shown in tables 2-4. CT lung density measurements in non-bullous lung were correlated with lung function for the group of patients with bullae alone, and then compared to those values from our previous data in 80 patients with COPD (15] (tables 3 and 4). The data from the two groups were then combined and plotted graphically to permit visual assessment of the overlap and similarity between the two populations.…”

“…Measurements of respiratory function were correlated with the following er variables: I) Mean EMI number of non-bullous lung; 2) lowest 5th percentile EMI number of non-bullous lung; and 3) percentage area of the lungs occupied by bullae. We have previously shown significant correlations between lung er density and measurements of airflow limitation, hyperinflation and diffusing capacity, in a group of smokers with chronic obstructive pulmonary disease (COPD) but without obvious bullae [15]. These subjects had a very wide range of age (63 males and 17 females; 23-82 yrs) and respiratory function (FEV 1 8-116% predicted; Dr.r.oN A 15-139% predicted).…”

“…derived from morphometric measurements of distal airspace size [14]. In a group of 80 cigarette smokers, er lung density correlated both with the severity of airflow limitation and impairment of the diffusing capacity of the lung for carbon monoxide (DLCo) [15]. In this study, we have applied CT scanning, as a measure of emphysema.…”