CT Quantified Central Pulmonary Arterial Enlargement Is an Independent Predictor of Worse Airflow Obstruction in Chronic Obstructive Pulmonary Disease

Abstract

RATIONALEThe role of the central pulmonary vasculature in chronic obstructive pulmonary disease (COPD) is not well understood. We investigated a central pulmonary arterial biomarker automatically extracted from computed tomography (CT) and its relation to airflow obstruction in COPD. METHODS Baseline inspiratory CT scans on a random selection of 40 subjects per GOLD stage including PRISm (preserved ratio impaired spirometry) and never smoker controls were selected from the COPDGene study, leading to a total data set of 280 subjects. Lung quantification software (LungQ, Thirona, Nijmegen, the Netherlands) was used to identify the lung fields and all central arterial branches from CT, defined as the segmental, sub-segmental, and sub-subsegmental arterial branches. The arterial diameter was calculated for each branch and averaged to compute the mean central arterial (MCA) diameter. Emphysema was defined as the percentage of low attenuation areas ≤ -950HU (LAA -950 ). CT derived total lung capacity (TLC CT ) was calculated from the extracted lung fields. Spirometry measurements were used to define airflow obstruction, including FEV 1 %-predicted and FEV 1 /FVC. Quality of life was defined by the SGRQ-score. Univariate Pearson's correlation was extracted between MCA and LAA -950 , FEV 1 %-predicted, FEV 1 /FVC, and SGRQ-score. Multiple linear regression analysis was performed to investigate the relative contribution of MCA for each of the spirometry and quality of life measurements. All models were adjusted for gender, age, weight, height, pack years, TLC CT , smoking status and LAA -950 . Results are expressed as coefficients, standard error (se), and p-values. RESULTS MCA had a significant (p<0.001) Pearson correlation with all investigated measurements ( r=0.32 for LAA -950 , r=-0.45 for FEV 1 %-predicted, r=-0.42 for FEV 1 /FVC, and r=0.34 for SGRQ-score). In the multiple regression analysis, MCA showed to be a significant independent predictor for FEV 1 %-predicted (B=-10.9, se=2.79, p<0.001), FEV 1 /FVC (B=-0.06, se=0.014, p<0.001), and SGRQ-score (B=-8.02, se=2.85, p=0.005). CONCLUSIONS Larger central arterial diameters in COPD were independently associated with worse spirometry and quality of life, even when adjusted for the extent of emphysema. Enlarged central pulmonary arteries may be an indication for pulmonary hypertension and could therefore be a clinically relevant CT derived marker.