D30-07 Longitudinal Changes in CT-based Tissue Heterogeneity (QtD) Is Associated With Lung Function Decline in COPD

Abstract Rationale Heterogeneity of the normal-appearing lung tissue on CT imaging is higher in COPD than in never-smokers and is associated with loss of lung function on follow-up at five years. Whether heterogeneity changes over time has not yet been assessed. To evaluate the longitudinal changes in lung heterogeneity measured by Quadtree Decomposition (QtD) between COPDGene Phase-1 and Phase-2, and to evaluate its association with lung function decline. Methods We analyzed Phase-1 and Phase-2 end-inspiratory CT imaging and spirometry data from 5,256 participants enrolled in the COPDGene cohort. Tissue heterogeneity was quantified using quadtree decomposition (QtD), where each axial slice of the segmented lung from CT is subdivided into quadrants until a level of homogeneity is reached. A high QtD value is indicative of higher tissue heterogeneity. A high QtD metric means higher tissue heterogeneity. QtD metrics in healthy-looking tissues with Hounsfield units ranging between -900 to -750 (QtDMD) were computed at each phase and compared between timepoints for groups comprising never-smokers, smokers without airflow obstruction, and GOLD stages 1-4. “Rapid decliners” were identified as participants within the lowest quartile of ΔFEV1 between Phase 1 and Phase 2. Multivariable regression models were used to investigate association between lung function decline and change in QtDMD. Predictive model performance was evaluated using area under the receiver-operating-characteristic curve (AUC). Results QtDMD values increased with GOLD stage severity in both Phase-1 and Phase-2 and were significantly different (p 0.001) from never-smokers. Among 2,028 COPD participants, a greater increase in ΔQtDMD was associated with decline in FEV1 (p = 0.008) after adjusting for baseline covariates such as age, sex, BMI, smoking status, emphysema% and FEV1. The multivariable logistic regression model, higher ΔQtDMD increased the odds of rapid FEV1 decline (AUC=0.71). Conclusions Longitudinal increase in CT-based tissue heterogeneity in healthy looking tissues tracks both structural and functional progression in COPD. ΔQtDMD remains shows to be a useful composite descriptor of parenchymal remodelling. The ability of ΔQtDMD to identify rapid decliners highlights its potential role in risk stratification and as an imaging biomarker for monitoring subtle disease progression beyond conventional emphysema quantification. This abstract is funded by: This work was supported by NHLBI grants U01 HL089897 and U01 HL089856 and by NIH contract 75N92023D00011.