QCT has enabled the development of diverse imaging biomarkers to characterize structural changes in COPD [11,12]. Emphysema burden is commonly quantified using the 15th percentile point of lung density (PD15) and the mean lung density, both of which are reproducible and sensitive to longitudinal change [13,14]. Changes in lung density-based metrics have been validated as endpoints in interventional trials, particularly in alpha1-antitrypsin deficiency, where augmentation therapy with alpha1-proteinase inhibitor has been shown to attenuate emphysema progression [15-17]. In addition, low attenuation area (LAA%) [18], reflecting emphysema burden, is associated with symptoms, lung function decline, exacerbations, and mortality [19-22]. Although the CT quantification of emphysema is affected by different scanning conditions and reconstruction methods [23], a threshold of ≥5% is considered as clinically significant disease and identifying individuals at risk for COPD even without airflow obstruction [24,25]. The spatial distribution of emphysema also influences lung function [26,27]. Fractal geometry-derived exponent D, which quantify the size distribution of emphysematous regions, detects parenchymal destruction in early-stage COPD even when LAA% remains unchanged [20]. The combination of D with LAA% improves histologic severity estimation and predicts outcomes, with longitudinal decreases in D and increases in LAA% indicating disease progression [28].

Small airway disease can be assessed by parametric response mapping (PRM), which classifies lung voxels into emphysema, functional small airway disease (fSAD), or normal tissue through voxel-wise registration of inspiratory and expiratory CT scans. fSAD is defined as lung regions that appear normal on inspiratory CT (attenuation value >-950 Hounsfield unit [HU]) but fail to show the expected increase in attenuation on expiratory CT (<-856 HU) (Figure 1), reflecting gas trapping due to small airway dysfunction [10]. In patients with airflow obstruction (GOLD 1-4 COPD) in the COPDGene study, both PRM fSAD and PRM emphysema predicted subsequent forced expiratory volume in 1 second (FEV₁) decline, whereas in GOLD 0, only PRM fSAD was predictive [29]. In the COPDGene cohort, the ratio of the mean lung density at end expiration (E) to that at end inspiration (I) was calculated in normal density (ND) lung regions [30]. Smokers exhibited higher ND-E/I ratios than nonsmokers, and 26.3% of smokers without airflow obstruction exceeded the 90th percentile value noted in nonsmokers. Elevated ND-E/I was associated with worse lung function, poorer health status, reduced functional capacity, and faster FEV₁ decline, indicating that subthreshold gas trapping, reflecting mild small airway disease is common and clinically relevant in this population.

Jacobian determinant mapping extends this voxel-based approach by quantifying regional lung volume change between inspiration and expiration, values greater than 1 indicate regional expansion, whereas values less than 1 indicate contraction, serving a crude estimate of lung elasticity. Mechanically affected lung (MAL) refers to morphologically normal-appearing regions that are mechanically influenced by adjacent emphysematous clusters. MAL is defined as normal voxels located within a specified distance from emphysematous areas (Figures 2, 3). Higher MAL values, particularly MAL₂ (within 2 mm), are associated with accelerated FEV₁ decline, indicating that morphologically normal-lung tissue can be mechanically compromised by adjacent emphysema [31]. Three-dimensional modeling using Gaussian smoothing further demonstrates that most new emphysema develops in areas of high (60.5%) or intermediate MAL (37.1%), with only small fraction arising from fSAD regions (4.8%) or zero MAL regions (2.4%) [32]. These findings suggest that mechanical stress exerted on surrounding lung tissue by existing emphysema is a dominant driver of disease progression. Consistently, higher MAL has also been linked to increased all-cause mortality [31,32].

Airway remodeling is another important hallmark of COPD pathology that can be quantified using several CT-derived metrics. Pi10 represents the square root of the wall area for a standardized airway with a 10 mm internal perimeter [33,34]. In cigarette smokers, airway wall thickening reflected by higher Pi10 is thought to result from inflammatory changes and airway remodeling, and is associated with reduced lung function, diminished functional capacity, and worse quality of life across all GOLD stages [34]. Pi10 may be modifiable. Longitudinal studies have shown that Pi10 decreases after smoking cessation and increases with resumed smoking, suggesting that Pi10 is a potentially reversible marker of airway inflammation. Additionally, in individuals with severe asthma, treatment with dupilumab has been shown to reduce the wall area percentage, even in a small study population [35]. PiSlope is a novel metric that quantifies the slope of the luminal perimeter versus wall thickness across the entire airway size spectrum, overcoming the size-specific limitation of Pi10 [36]. PiSlope summarizes airway wall remodeling across a range of airway sizes by describing how wall thickness changes relative to airway perimeter. PiSlope is calculated as the slope of the luminal perimeter-wall thickness relationship using linear regression, normalized by the participants’ height squared (Figure 4). Lower PiSlope values indicate greater airway remodeling and are associated with worse clinical outcomes. PiSlope is moderately correlated with Pi10 (r=-0.57) but decreases with COPD severity and is independently associated with dyspnea, functional limitation, exacerbations, FEV₁ decline, and mortality, even after adjusting for Pi10.

Several measures of airway luminal remodeling are also been described. In COPDGene participants, airway surface area-to-volume ratio (SA/V) distinguished airway narrowing from airway loss [37]. Higher SA/V correlated with better function and quality of life, while lower SA/V predicted faster lung function decline and worse survival over 5 years in patients with predominant airway loss. These findings support SA/V as a clinically relevant imaging biomarker for COPD phenotyping and prognosis. Additional airway metrics such as the airway lumen tapering slope (T-Slope) and airway fractal dimension (AFD) provide complementary prognostic insights. T-Slope is the airway lumen tapering index, measured as slope of cumulative lumen area over airway generations (Figure 5) [38]. Lower T-Slope values reflect lumen tapering and greater remodeling, and are linked to worse lung function, lower functional capacity, worse quality of life, faster FEV₁ decline, higher exacerbation frequency, and higher mortality. AFD indicates loss of airway branching complexity (Figure 6), and is also associated with impairments in lung function, respiratory symptoms, functional capacity, exacerbation, lung function decline, and mortality [39]. When combined with higher peri-bronchial emphysema, lower AFD identifies individuals at substantially higher risk of death, offering prognostic value beyond conventional airway metrics and spirometry. Expiratory central airway collapse (ECAC) is characterized by excessive airway collapse during expiration either from cartilaginous weakening or redundancy of the posterior membranous wall, which includes tracheobronchomalacia and excessive dynamic airway collapse (Figure 7) [40]. ECAC is defined as ≥50% reduction in central airway luminal area during expiration, occurs in approximately 5% of COPDGene participants and is independently associated with worse respiratory symptoms, increased exacerbations, and higher hospitalization rates, even in individuals without spirometric obstruction. Lastly, mucus plugs on CT are defined as the complete occlusion of medium- to large-sized airways [41]. In a retrospective COPDGene analysis with a median follow-up of 9.5 years, mucus plug burden, categorized as 0, 1-2, or ≥3 lung segments, was identified as an independent predictor of all-cause mortality in COPD. Moreover, clearance of mucus plugs has been used as a surrogate outcome for assessing treatment response to biologics. Notably, dupilumab [35] and tezepelumab [42] have demonstrated efficacy in patients with severe asthma for the reduction of mucus plugs, and tozorakimab [43] in patients with COPD.