Miranda Kirby PhD, Presenter: Nothing to Disclose

Wan C. Tan, Abstract Co-Author: Nothing to Disclose

Jean Bourbeau, Abstract Co-Author: Nothing to Disclose

James Hogg MD, Abstract Co-Author: Nothing to Disclose

Youbing Yin, Abstract Co-Author: Nothing to Disclose

Harvey Owen Coxson PhD, Abstract Co-Author: Research Grant, GlaxoSmithKline plc Contract, GlaxoSmithKline plc Contract, Olympus Corporation Steering Committee, GlaxoSmithKline plc

Chronic obstructive pulmonary disease (COPD) poses a substantial and under-appreciated burden on society. The advent of novel computed tomography (CT) image analysis tools can facilitate clinical phenotyping and provide quantitative information that may have prognostic value. It has been proposed that registration of CT images acquired at end-inspiration to end-expiration will allow the quantification of emphysema and small airway disease. One analysis method uses fixed CT density thresholds to assess lung structure (Galbán Nat Med. 2012). In contrast, another emerging method, the disease probability measure (DPM), uses continuous, voxel-by-voxel assessment of lung density to determine the extent of emphysema and air trapping. Our objective was to compare measurements derived from the threshold-based approach to DPM measurements in subjects with and those at risk of COPD.  

We evaluated 150 subjects between 45-90yrs from the population-based Canadian Cohort of Obstructive Lung Disease (CanCOLD) study. At Risk subjects were smokers without COPD; COPD subjects were classified according to GOLD criteria. All subjects underwent inspiratory/expiratory CT, spirometry and body plethysmography. CT emphysema (Emph) and air trapping (AirTrap) measurements were generated for both the threshold-based and DPM methods (Apollo software, VIDA Diagnostics Inc).

Emphysema measurements were associated with measurements of airflow obstruction (FEV1) (Emph: r=-0.48, p<0.0001; DPMEmph: r=-0.57, p=0.0001) and the diffusing capacity of the lung (DLCO) (Emph:r=-0.45, p<0.0001; DPMEmph: r=-0.53, p=0.0005). Air trapping measurements were associated with FEV1 (AirTrap: r=-0.20, p=0.03; DPMAirTrap: r=-0.23, p=0.01) and RV/TLC (AirTrap: r=0.25, p=0.008; DPMAirTrap: r=0.40, p<0.0001). Although the threshold-based and DPM measurements distinguished between At Risk and severely obstructed subjects (GOLD III, p<0.0001), only DPMAirTrap distinguished subjects with early or mild disease from moderate COPD (At Risk vs. GOLD II: p=0.0007; GOLD I vs. GOLD II: p=0.03).

While both DPM and the threshold-based measurements were associated with functional measurements, only the DPM approach provided a way to differentiate COPD subjects with early or mild disease.

DPM may provide a way to sensitively phenotype COPD patients with mild disease which may improve patient outcomes by preventing disease progression.

Kirby, M, Tan, W, Bourbeau, J, Hogg, J, Yin, Y, Coxson, H, Novel Computed Tomography Disease Probability Measure Analysis Technique Provides Sensitive Imaging Biomarkers of Chronic Obstructive Pulmonary Disease Phenotypes in Subjects with Mild-to-Moderate Disease.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14008171.html