Introduction

Characteristic structural alterations in chronic obstructive pulmonary disease (COPD) include small airway remodeling and emphysema1, such as airway wall thickening, extracellular matrix deposition, and smooth muscle cell proliferation/hypertrophy2. These changes lead to irreversible airway narrowing and increased airflow resistance, contributing significantly to airflow limitation3. Small airway pathology typically precedes the development of emphysema and often occurs before detectable decline in pulmonary function. Structural abnormalities are already present in the early stages of the disease4,5. Small airway disease (SAD) and small airway dysfunction (SAdf) have garnered significant attention in recent years. However, it is important to clarify that SAdf refers solely to a single functional parameter, whereas SAD indicates that multiple functional components of airway dysfunction, under specific conditions, may demonstrate a pathophysiological phenotype characteristic of a clinically defined disease6. Small airway remodeling leads to air trapping and increased residual volume, further accelerating emphysematous progression. The interplay between progressive small airway remodeling and emphysema exacerbates airflow limitation and pulmonary function decline in COPD patients, ultimately worsening their quality of life and prognosis.

Exhaled nitric oxide (eNO) serves as a biomarker reflecting the degree of airway and alveolar inflammation, with studies demonstrating a significant correlation between the alveolar nitric oxide concentration (CaNO) and pulmonary function parameters, particularly the FEV1%pre in COPD patients7. CaNO is also negatively associated with specific lung function measures, including forced expiratory flow at 50% of vital capacity (FEF50) and the diffusing capacity of the lungs for carbon monoxide (DLCO)8. These findings suggest that elevated CaNO levels may be correlated with reduced pulmonary function, indicating the impact of airway inflammation on gas exchange efficiency. High-resolution computed tomography (HRCT) is a crucial clinical tool for quantitatively assessing structural changes in COPD patients, with a particular focus on small airway evaluation. The key measurable parameters include the lumen diameter (LD), inner luminal area (LA), airway wall thickness (WT), wall area (WA), percentage wall area (WA%), wall thickness-to-diameter ratio (TDR), WT-to-pulmonary artery ratio (WT/PA), and square root of the wall area for bronchi with a 10 mm perimeter (Pi10)9. These quantitative HRCT measurements provide detailed insights into the severity of airway remodeling and its impact on disease progression in COPD patients.

In pulmonary emphysema assessment, the percentage of the lung area with attenuation less than − 950 Hounsfield units (LAA− 950%) on inspiratory CT is a widely used method to quantify emphysema extent10. Additional parameters, such as the end-expiratory lung volume, volume difference, and volume ratio, further characterize emphysema severity11. These measures strongly correlate with impaired pulmonary function, aiding in the evaluation of COPD severity. Studies have demonstrated significant associations between LAA− 950% and FEV1/FVC in individuals over 65 years of age, with higher emphysema indices predicting greater airflow limitation12. COPD patients show correlations between LAA− 950% and multiple spirometric parameters (FEV1/FVC, FEV1%pre, FVC%pre)13. HRCT metrics enhance early COPD detection, as combined assessment of airway wall thickness (generations 3–8) and LAA− 950% improves diagnostic accuracy14. Importantly, both airway wall thickness and LAA− 950% correlate with acute exacerbation frequency and COPD assessment test (CAT) scores15, underscoring their clinical utility in COPD management.

HRCT enables crucial quantitative and locational evaluation of emphysema and small airway remodeling in COPD patients, greatly enhancing the understanding of COPD pathophysiology and guiding clinical treatment. For instance, CT-derived large airway metrics correlate with histological small airway indices, allowing indirect inference of small airway pathology even when isolated large airway wall thickening has minimal effects on pulmonary function. These HRCT-based structural parameters provide valuable diagnostic references, facilitating precise assessment of COPD-related lung structural changes and formulation of individualized treatment regimens.

Pulmonary function tests (PFTs), the traditional gold standard for COPD diagnosis, evaluate airflow limitation severity and disease progression while also enabling early identification of small airway dysfunction. High-resolution computed tomography (HRCT) provides complementary, objective lung tissue measurements with greater precision. Unlike PFTs—which inherently include nasal/oropharyngeal cavities, trachea, and airway dead space (-100 mL)—automated HRCT software excludes these regions from volumetric analysis, yielding more physiologically accurate data16. For critically ill or acutely exacerbated COPD patients who are unable to undergo PFTs, chest CT is indispensable because of its operational ease and safety. The current understanding of the relationships among exhaled nitric oxide (eNO), HRCT quantitative parameters, and pulmonary function remains limited, warranting investigations into their correlations to clarify their clinical utility and optimize personalized COPD management strategies.

Objects and methods

Study objects

AECOPD group

A total of 115 patients with AECOPD were enrolled in this study. These patients were hospitalized in the Department of Respiratory and Critical Care Medicine at Fuyang Infectious Disease Clinical College of Anhui Medical University, between January and September 2023. The inclusion criteria were as follows: (1) Patients who met the 2023 Global Initiative for Chronic Obstructive Lung Disease (GOLD) diagnostic criteria17. COPD was defined as a postbronchodilator FEV1/FVC ratio < 0.70, whereas AECOPD was characterized by worsening dyspnea and/or cough with increased sputum production within 14 days, potentially accompanied by tachypnea and/or tachycardia, often triggered by respiratory infections, air pollution, or heightened inflammatory responses. The onset time of AECOPD is defined as the first day when the patient exhibits symptoms of acute exacerbation. On the basis of FEV1% predicted (FEV1%pre), lung function was categorized as GOLD 1 (FEV1%pre ≥ 80%), GOLD 2 (50% ≤ FEV1%pre < 80%), GOLD 3 (30% ≤ FEV1%pre < 50%), or GOLD 4 (FEV1%pre < 30%). (2) Complete medical records were required for all participants. The exclusion criteria were as follows: (1) Concomitant respiratory conditions, including pulmonary infections, cough-variant asthma, bronchiectasis, bronchial asthma, lung malignancies, or tuberculosis, as well as tracheal mucus plugs on CT. (2) Severe cardiocerebrovascular diseases, hepatic/renal dysfunction, or solid/hematologic tumors. (3) Oral/intravenous corticosteroid use within the past week or long-term (> 3 months). (4) Inability to complete eNO measurements, chest CT, or pulmonary function tests. (5) Patients with a history of acute exacerbation within 4 weeks before admission.

COPD group

This group consisted of 89 outpatients with stable COPD treated in the same department during the study period. The inclusion criterion was compliance with the aforementioned COPD diagnostic standards. The exclusion criterion was acute exacerbation within the previous 3 months, along with criteria (1)-(4) from the AECOPD group.

Healthy control group

This group comprised 70 age-matched healthy individuals undergoing routine health checkups. The inclusion criteria were as follows: aged ≥ 40 years, predominantly male, and had normal pulmonary function test results. The exclusion criterion was the same as criteria (1)-(4) for the AECOPD group.

Study population for analyses 3.3–3.9

Both the AECOPD and COPD groups were included

Research content and methods

Research content

General data

Demographic and clinical characteristics, including age, gender, height, weight, body mass index (BMI), body surface area (BSA), smoking history, and the use of inhaled corticosteroids (ICS), were recorded for all enrolled patients. The BSA was calculated via the following formula: BSA (m²) = 0.0061×height (cm) + 0.0128×weight (kg)–0.152.

Laboratory tests

Venous blood samples (5 mL × 2) were collected were collected on an empty stomach the morning after the patient’s admission for eosinophil count (Eos) and total Immunoglobulin E(IgE) analysis. Eos was performed using a SYSMEX automated hematology analyzer. Total IgE were quantified with a Hitachi biochemical analyzer.

eNO measurement

eNO levels were measured in AECOPD patients upon hospitalization, stable COPD patients during outpatient visits, and healthy controls during routine checkups. Testing followed the 2017 European Respiratory Society (ERS) technical standards for exhaled biomarker analysis18 and used a Sunvou-P100 electrochemical analyzer (Wuxi Sunvou, China). The device was factory-calibrated for both the electrochemical sensor and NO detector, with operational requirements including ambient temperature (5–35 °C), relative humidity (< 80%), and atmospheric pressure (700–1060 hPa). Testing was conducted in well-ventilated areas away from volatile organic compounds (e.g., alcohol). All tests were performed by board-certified pulmonologists who underwent standardized training covering exhalation protocols, device operation, and outlier interpretation. Pretest instructions: Participants refrained from eating, smoking, vigorous exercise, or pulmonary function tests 1 h prior and avoided nitrate-rich foods (e.g., broccoli, lettuce, celery, processed meats) for 3 h. The medication history (especially the use of corticosteroids/antibiotics) within 72 h was recorded. Disposable mouthpieces were used to prevent cross-contamination; tests were repeated if breath-holding, air leakage, or saliva interference occurred. Procedure: Seated participants wore a nasal clip, exhaled fully, and then inhaled rapidly before exhaling through the filter at 50 ml/s (for FeNO50) or 200 ml/s (for FeNO200) until the device indicated completion. CaNO was derived via the two-compartment model: FeNO = CaNOdual + Jaw-breath nitric oxide (JawNO)/VE + f, where VE = expiratory flow rate (ml/s) and f = correction factor from multiflow CaNO19.

Pulmonary function testing (PFTs)

PFTs were performed for AECOPD patients after symptom stabilization (3–7 days after the acute exacerbation), stable COPD patients during outpatient visits, and healthy controls during routine exams, following the 2019 American Thoracic Society (ATS)/ERS guidelines20. Measurements were conducted via a MasterScreen PFT spirometer (CareFusion, Germany) by certified technicians. The quality control protocols included daily leakage checks and volume calibration with a standard syringe. Before testing, technicians documented participants’ medical histories and recent medication use, excluding corticosteroids, bronchodilators, or other drugs affecting the results. The participants abstained from tea, coffee, or cola for ≥ 2 h. Disposable mouthpieces were used; tests were repeated if breath-holding, air leakage, or saliva interference occurred. Procedure: Seated subjects wore a nasal clip, sealed their lips around the mouthpiece, and performed breathing maneuvers under technician guidance. Tests were repeated twice, with the best of three trials recorded. Bronchodilator testing was added for obstructive patterns: subjects inhaled 200 µg salbutamol (GSK) via a metered-dose inhaler (MDI), with FEV1 remeasured after 15–30 min. Safety monitoring included 30 min of posttest observation for adverse reactions. Calculate z-scores for pulmonary function parameters (FEV1, FEV1/FVC, FEF75, MMEF, DLco/VA, etc.) adjusted for age, sex, height, and ethnicity according to the 2012 Global Initiative for Chronic Obstructive Lung Disease (GLI-2012).

HRCT quantitative analysis

Perform HRCT scanning on the day of admission or the following day for AECOPD patients, with PFT and HRCT conducted on separate days. Healthy controls underwent HRCT during routine examinations. Scans were acquired via a GE Revolution 256-slice CT scanner with standardized protocols: (1) Hardware: 120 kV tube voltage, automatic tube current modulation (100–500 mA), 256 × 1.25 mm detector collimation, 0.6 s/rotation gantry speed; (2) Scanning: pitch 0.992, table feed 132.29 mm/s, contiguous 1.25 mm slices; (3) Reconstruction: 40 cm display field-of-view, 1024 × 1024 matrix with standard algorithm, 0.625 mm thin-slice reconstruction. The participants practiced end-inspiration breath-holding before scanning in the supine position (arms elevated, midline alignment). Images were analyzed on an AW 4.7 Workstation (GE Healthcare). Airway quantification followed the consensus bifurcation classification (generation 1: main bronchus; generations 6–8: subsegmental bronchi). Seed-point tracking identified orthogonal airway cross-sections, excluding oblique branches. Adaptive threshold segmentation (-450 HU window width, 1200 HU level) with edge detection defined lumen-outer wall boundaries, measuring WT, lumen diameter, and wall area percentage (WA%). Curvature correction was used to standardize WT measurements to minimize tilt/ellipticity artifacts. Two radiologists independently measured RB1 (0.5 cm from the bifurcation) and RB1a (0.5 cm from the RB1 bifurcation) for WT, WA, and the WA-to-diameter ratio (TDR), alongside the adjacent pulmonary artery diameter (PA) and WT/PA ratio. Emphysema quantification uses global thresholding (-500 to -1024 HU) with vascular/tissular artifact removal (morphological opening) and semimanual diaphragm motion correction. The low-attenuation area percentage (LAA− 950%) was calculated at -950 HU, with apex-base density gradients normalized via 3D interpolation. Inspiratory-phase data (≥ 75% lung field inclusion) were selected using anatomical landmarks (aortic arch/carina) to exclude expiratory collapse. Body surface area (BSA)-normalized metrics (WT/BSA, WA/BSA) were reviewed by two associates and one chief radiologist. To evaluate the reliability of HRCT airway parameter measurements, this study conducted intra - and inter - observer consistency analyses using a double - blind method. Two radiologists (Radiologist A and Radiologist B) with more than 5 years of experience in chest imaging diagnosis served as independent observers, and they randomly selected HRCT imaging data of 20 COPD patients (accounting for approximately 10% of the total sample size) for measuring airway parameters. For the inter - observer measurement: the two radiologists independently completed the measurement of HRCT airway parameters in 20 patients without knowing the patients’ clinical information. For the intra - observer measurement: Radiologist A performed a second repeated measurement on the HRCT images of the same 20 patients after an interval of 2 weeks, with the measurement conditions consistent with the first time. The intraclass correlation coefficient (ICC) was used to assess the measurement consistency. The two - way mixed model was selected for the inter - observer consistency analysis, and the one - way random model was adopted for the intra - observer consistency analysis, both using the absolute agreement type; an ICC > 0.8 was considered to indicate good measurement consistency. All statistical analyses were performed using SPSS 21.0 software, with a test level of ɑ = 0.05. The reliability analysis of HRCT airway parameter measurements showed that the intra - observer consistency ICC value was 0.92 (95%CI: 0.87–0.96), and the inter - observer consistency ICC value was 0.89 (95%CI: 0.82–0.94). Both values were greater than 0.8, suggesting that the intra - and inter - observer consistency of HRCT airway measurements in this study was good.

Statistical analysis

Statistical analyses were performed via SPSS 21.0 (IBM, USA), GraphPad Prism 5.0 (GraphPad Software, USA), and Origin 2024 (OriginLab, USA). Categorical variables are expressed as proportions (%) and were compared via chi-square tests. Normally distributed continuous data are presented as the mean ± standard deviation (SD), with group differences assessed by Student’s t test or ANOVA. Nonnormally distributed data are reported as medians (M) with interquartile ranges (IQRs, P25–P75) and were analyzed via nonparametric tests (Mann‒Whitney U tests, Wilcoxon rank‒sum tests, or Kruskal‒Wallis H tests). For multiple-group comparisons, Dunnett’s T3 post hoc test was applied with Bonferroni correction. Partial correlation analysis was used to examine the relationship between eNO and pulmonary function parameters as well as quantitative HRCT indicators, exploring linear relationships among continuous variables after controlling for potential confounding factors. Pearson correlation analysis was applied to indicators following a normal distribution, while Spearman correlation analysis was used for those not following a normal distribution. Both analyses were corrected using the False Discovery Rate (FDR) method. All tests were two-tailed, with statistical significance set at ɑ = 0.05 (P < 0.05).

Results

Comparison of general characteristics, laboratory Tests, and eNO levels among three groups

There were no statistically significant differences in general characteristics (gender, age, height, weight, BMI, BSA, smoking status), ICS use or laboratory parameters (eosinophil count, total IgE) among the three groups (P > 0.5, Table 1).

Table 1 Comparison of the general characteristics of the patients in the three groups.
Full size table

Comparison of HRCT quantitative parameters among the three groups

The results of the quantitative HRCT analysis revealed no significant difference in the LAA− 950% between the AECOPD group and the COPD group (P = 0.302). However, compared with the COPD group, the AECOPD group presented a significantly greater airway wall thickness-to-body surface area ratio (WT/BSA) and airway wall thickness-to-accompany pulmonary artery diameter ratio (WT/PA) in the RB1 segment (P < 0.05), indicating more pronounced small airway structural remodeling during acute exacerbations. Compared with healthy controls, the AECOPD group presented significantly greater WT/BSA values, airway wall area-to-body surface area (WA/BSA) ratios, wall thickness-to-diameter (TDR) ratios, and WT/PA ratios in both the RB1 and RB1a segments (all P < 0.05). Compared with the control group, the COPD group presented greater TDR in only the RB1 segment and greater WT/BSA and TDR in the RB1a segment (P < 0.05), confirming that small airway remodeling persists during stable disease phases. Notably, no significant differences in WA/BSA or TDR in the RB1 segment or any parameters in the RB1a segment were detected between the AECOPD and COPD groups (all P > 0.05, Table 2).

Table 2 Comparison of quantitative HRCT indicators in patients in the AECOPD group, COPD group, and healthy control population.
Full size table

Correlations of FeNO50, FeNO200, and CaNO with quantitative HRCT parameters of RB1 in COPD patients (AECOPD and COPD groups)

The associations between eNO concentrations at different flow rates and small airway remodeling exhibited distinct patterns. At the conventional 50 mL/s flow rate (FeNO50), no significant correlations were observed with WT/BSA, WA/BSA, TDR, or WT/PA (P > 0.05). In contrast, FeNO200 was significantly positively correlated with WT/BSA, WA/BSA, and WT/PA (P < 0.05). Additionally, CaNO was positively correlated with all four parameters: WT/BSA, WA/BSA, TDR, and WT/PA(P < 0.05), indicating that peripheral airway and alveolar inflammation may be more pronounced than in proximal airways, with stronger links to structural changes. The broader association of CaNO likely reflects its heightened sensitivity to small airways, alveolar tissue, and pulmonary vascular damage (Table 3).

Table 3 Correlations between FeNO50, FeNO200, and CaNO with quantitative HRCT indices of RB1 in COPD patients.
Full size table

Correlations of FeNO50, FeNO200, and CaNO with HRCT quantitative parameters of RB1a in COPD patients

FeNO50 was not significantly correlated with the WT/BSA, WA/BSA, TDR, or WT/PA of RB1a (P > 0.05). In contrast, FeNO200 and CaNO was positively correlated with WT/BSA, WA/BSA, and WT/PA (P < 0.05) but not with TDR (P > 0.5). The strong links between FeNO200/CaNO and HRCT parameters support a synergistic progression between peripheral inflammation and airway remodeling. The notable CaNO-WT/PA correlation suggests a potential interaction between pulmonary arterial remodeling and peripheral inflammation, suggesting that airway wall thickening may be inflammation-driven. Dynamic monitoring of peripheral eNO combined with high-resolution computed tomography (HRCT) quantification could be used to assess the efficacy of anti-inflammatory treatment in delaying airway remodeling (Table 4).

Table 4 Correlations between FeNO50, FeNO200, and CaNO with quantitative HRCT indices of RB1a in patients in the COPD patients.
Full size table

Correlations of FeNO50, FeNO200, and CaNO with LAA− 950% in COPD patients

After controlling for confounding factors including age, smoking volume, and ICS usage, a correlation analysis revealed that neither FeNO50 nor FeNO200 showed a significant correlation with LAA− 950% (P > 0.05), whereas CaNO was positively correlated with LAA-950% (r = 0.176, P = 0.012, 95%CI [0.013, 0.347]). These findings suggest that the inflammatory microenvironment during COPD progression exhibits distinct anatomical partitioning-localized airway inflammation (reflected by FeNO50/FeNO200) that may not be directly coupled with panlobar emphysema progression, whereas distal alveolar NO (CaNO) potentially participates in lung parenchyma destruction through direct or indirect regulatory mechanisms (Fig. 1).

Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

Correlation between CaNO and LAA− 950% in COPD patients.

Correlations of FeNO50, FeNO200, and CaNO with pulmonary function parameters in COPD patients

FeNO50 was not associated with lung function parameters (P > 0.05), suggesting that eosinophilic inflammation pathways are shared between large and small airways, with a limited impact of large-airway inflammation on airflow limitation. FeNO200, and CaNO were negatively correlated with FEV1, z−scores, FEV1/FVC, z−scores, FEF75, z−scores, and MMEF, z−scores (P < 0.05), further supporting the role of small airways and alveolar inflammation in accelerating airway remodeling and impairing ventilation. Additionally, CaNO correlated with patients’ DLCO/VA, z−scores, suggesting that diffusion impairment may influence CaNO levels (Table 5).

Table 5 Correlations of FeNO50, FeNO200, and CaNO with lung function in COPD patients.
Full size table

Correlation of HRCT quantitative parameters in RB1 with pulmonary function in COPD patients

The WT/BSA, and TDR were negatively correlated with FEV1, z−scores, FEV1/FVC, z−scores, FEF75, z−scores, and MMEF, z−scores (P < 0.05), while the WT/PA was negatively correlated with FEV1/FVC, z−scores, FEF75, z−scores, and MMEF, z−scores (P < 0.05). In contrast, WA/BSA was not significantly associated with lung function (P > 0.05). This differential correlation suggests microstructure-specific effects of airway remodeling on pulmonary function, where WT/BSA and morphometric indices (TDR, WT/PA) better reflect inflammation-mediated bronchial narrowing. WA/BSA, however, likely fails to capture pathological heterogeneity due to the exclusion of perivascular inflammatory infiltration (Fig. 2).

Fig. 2
Fig. 2The alternative text for this image may have been generated using AI.
Full size image

Heatmap of the correlation between quantitative HRCT indices of RB1 and lung function in COPD patients.

Correlation of HRCT quantitative parameters in RB1a with pulmonary function in COPD patients

The WT/BSA was negatively correlated with FEV1, z−scores, FEV1/FVC, z−scores, FEF75, z−scores, and MMEF, z−scores (P < 0.05), suggesting that wall thickness changes in this terminal bronchiolar region predominantly affect distal air trapping. WA/BSA correlated negatively with FEV1, z−scores, and MMEF, z−scores (P < 0.05) but not with FEV1/FVC, z−scores, and FEF75, z−scores (P > 0.05), indicating functional complementarity between airway morphological parameters in assessing dynamic ventilation impairment in COPD patients. Notably, TDR was negatively correlated with FEV1, z−scores, and FEF75, z−scores, while WT/PA was significantly negatively correlated with FEV1/FVC, z−scores, FEF75, z−scores, and MMEF, z−scores (P < 0.05), highlighting the pronounced impact of combined airway and vascular remodeling on pulmonary function (Fig. 3).

Fig. 3
Fig. 3The alternative text for this image may have been generated using AI.
Full size image

Heatmap of the correlation between quantitative HRCT indices of RB1a and lung function in patients with COPD.

Correlations between LAA− 950% and pulmonary function parameters in COPD patients

LAA− 950% was significantly negatively correlated with FEV1, z−scores, FEV1/FVC, z−scores, FEF75, z−scores, and MMEF, z−scores (all P < 0.05), indicating that parenchymal destruction preferentially affects distal alveolar‒capillary units. In early expiration, preserved airway elastic recoil maintains relatively intact airflow, whereas progressive loss of tissue support during later respiratory phases worsens distal airway collapse due to alveolar fusion, leading to a marked deterioration in mid-to-late expiratory flow parameters (Fig. 4).

Fig. 4
Fig. 4The alternative text for this image may have been generated using AI.
Full size image

(a) Correlation of LAA− 950% with FEV1, z−scores in COPD patients. (b) Correlation of LAA− 950% with FEV1/FVC, z−scores. (c) Correlation of LAA− 950% with FEF75, z−scores. (d) Correlation of LAA− 950% with MMEF, z−scores.

Comparison of LAA− 950% across different GOLD stages in COPD patients

COPD patients with GOLD 1 had significantly lower LAA− 950% than those with GOLD 3 and 4 [median (IQR): 1.65 (0.36–8.58) % vs. 6.89 (3.31–14.85) % vs. 10.58 (7.15–20.82) %, all P < 0.01], demonstrating concordance between emphysema severity and ventilatory impairment. Compared with GOLD 4 patients, GOLD 2 patients presented a lower LAA− 950% [5.44 (2.54–9.47) % vs. 10.58 (7.15–20.82) %, P = 0.021], whereas no significant differences were detected between GOLD 1 and 2 patients (P = 0.208), GOLD 2 and 3 patients (P = 0.275), or GOLD 3 and 4 patients (P = 0.684). This finding suggests the absence of a critical threshold for alveolar collapse in early-stage disease (GOLD 1–2) but reveals a compensatory plateau in late-stage disease (GOLD 3–4) due to progressive parenchymal destruction (Fig. 5).

Fig. 5
Fig. 5The alternative text for this image may have been generated using AI.
Full size image

Comparison of LAA− 950% in COPD patients with different lung function classifications, *P < 0.05, **P < 0.01, N.S. not significant.

Discussion

During the progression of chronic obstructive pulmonary disease (COPD), pathological changes such as airway mucosal metaplasia, increased mucus secretion causing luminal occlusion, and thickening of the airway epithelium, smooth muscle, and submucosal layers lead to extensive small airway injury, resulting in structural and functional alterations21. These structural changes primarily manifest as abnormal airway wall thickening and increased airway wall area, which may further contribute to emphysema development. Comparative analysis of terminal and transitional bronchioles (first-generation airways) between smokers with normal lung function and COPD patients at different GOLD stages revealed significantly reduced numbers of these airways in COPD patients, with a progressive decline corresponding to disease severity. These findings suggest that small airway structural remodeling precedes emphysema development22. With advancements in CT technology and intelligent measurement software, HRCT now enables precise quantification of airway parameters and automated calculation of emphysema indices in clinical practice.

This study selected the right upper lobe bronchus as the primary measurement site because of its minimal cardiac motion interference and near-perpendicular orientation (90°) to the CT imaging plane, ensuring measurement accuracy. To mitigate software-related measurement variability, both RB1 and RB1a were analyzed. All WT and WA measurements were normalized to BSA to control for anthropometric variations. Comparative HRCT analysis revealed significantly elevated RB1 WT/BSA and WT/PA values in AECOPD patients compared with those in stable COPD patients. Compared with healthy subjects, the AECOPD group presented greater RB1 and RB1a WT/BSA, WA/BSA, TDR, and WT/PA values. Stable COPD patients presented higher RB1 TDR and RB1a WT/BSA, and TDR values than did the healthy individuals.

Weikert et al. demonstrated significantly greater small airway wall thickness in COPD patients than in healthy controls, with progressive thickening correlating with GOLD stage severity14. Gender-stratified analyses further confirmed elevated Pi10 airway wall thickness in both male and female COPD patients compared with non-COPD individuals23. Micro-CT evaluation of four explanted GOLD-4 COPD lungs versus four unused donor lungs revealed a 10-fold reduction in the terminal bronchiolar count and a 100-fold decrease in the minimal luminal cross-sectional area in COPD tissue (both emphysematous and nonemphysematous regions)24. These structural alterations likely stem from aberrant tissue repair/remodeling and impaired mucociliary clearance, leading to persistent inflammatory exudate accumulation25. Consequently, bronchial wall thickening in patients with AECOPD and stable COPD may result from mucosal inflammatory infiltration and hypersecretion, whereas a reduced airway caliber reflects the combined effects of wall thickening and bronchospasm.

To investigate the relationships between eNO, HRCT-derived small airway parameters, and pulmonary function in COPD patients, this study analyzed AECOPD patients and stable COPD patients while excluding those with concurrent pulmonary infections or mucus plugging to prevent confounding effects on quantitative CT measurements. To avoid confounding effects of age, smoking status, ICS use, and emphysema on eNO, this study employed partial correlation analysis to control for these potential confounders. Given that prior studies have not examined eNO-HRCT correlations, we hypothesized that FeNO200 and CaNO, which reflect distal airway/alveolar inflammation, are associated with CT-quantified small airway remodeling and emphysema severity (LAA− 950%). The analytical results confirmed positive correlations for both FeNO200 and CaNO with RB1/RB1a WT/BSA, WA/BSA, and WT/PA, whereas CaNO additionally correlated with LAA− 950% and RB1 TDR. Notably, proximal airway-derived FeNO50 showed no significant associations, suggesting preferential linkages between peripheral airway inflammation and structural pathology. The flow-dependent eNO measurement gradient implies distinct pathological processes across airway generations, supporting the clinical utility of partitioned eNO analysis in COPD phenotyping.

Prior studies analyzing asthma patients demonstrated spearman correlations between the FeNO50 and WA/BSA across generations 3–6 bronchi, with partial rank analysis specifically identifying generation 6 WAs/BSAs as significantly associated26. This finding likely reflects the strong correlation of FeNO50 with the predominant type 2 inflammation pathway in asthma. In contrast, COPD involves predominantly small airway inflammation, explaining why FeNO200 and CaNO (reflecting distal inflammation) show greater clinical relevance than does FeNO50 in COPD and correlate more strongly with small airway structural changes on high-resolution computed tomography (HRCT), as confirmed in our study. Three potential mechanisms may underlie these associations: (1) impaired NO diffusion across thickened airway walls may increase eNO levels27; (2) increased NO production via L-arginine metabolism generates nitrites/nitrates, with elevated airway NO concentrations in asthma/COPD patients promoting hyperreactivity, which is driven by wall thickening28; and (3) pathological NO elevation induces eosinophil accumulation and Th2 overactivation, exacerbating IgE-mediated responses that drive remodeling29.

Our findings confirm a positive correlation between CaNO and LAA− 950%, suggesting that airway NO production contributes to emphysema progression. Supporting evidence comes from murine emphysema models in which triple nitric oxide synthase (NOS) knockout (nNOS/iNOS/eNOS), but not single or double-knockout, induced characteristic emphysematous changes (enlarged alveolar spaces, alveolar wall degradation, and reduced elastic fibers) alongside genomic alterations30. Elevated CaNO in AECOPD patients indicates more severe small airway inflammation, which is often accompanied by diffuse alveolar destruction. When airway thickening and luminal narrowing occur, expiratory airflow limitation exacerbates air trapping, reducing gas exchange and promoting progressive emphysema formation due to impaired alveolar mechanics31. Additionally, emphysema leads to destruction of the alveolar-capillary bed, impairing NO diffusion and clearance, which in turn causes elevated CaNO levels.

COPD patients exhibit persistent small airway inflammation, leading to increased airway resistance and obstructive ventilatory dysfunction. To assess whether eNO correlates with these functional parameters, this study also conducted partial correlation analyses between eNO and these indicators after controlling for potential confounding factors. To eliminate confounding effects of age, gender, and ethnicity on pulmonary function parameters and enhance comparability across studies, this research converted pulmonary function parameters into z-scores for analysis. The results revealed no significant association between FeNO50 and any spirometric measure, whereas FeNO200, and CaNO was negatively correlated with FEV1, z−scores, FEV1/FVC, z−scores, FEF75, z−scores, and MMEF, z−scores. These findings highlight that the spatial heterogeneity of inflammation in COPD-peripheral airway dysfunction is inversely related to small airway small airway airflow limitation but synergizes with distal alveolar inflammation (reflected by CaNO), suggesting that proximal inflammatory mediators may propagate distal structural damage. Combining FeNO200 and CaNO measurements may thus aid in delineating inflammation distribution and stratifying patients with “small airway–alveolar comorbidity” phenotypes, offering clinical utility for targeted management. Additionally, this study found that CaNO correlates with patients’ DLCO/VA, z−scores, consistent with previous research findings7, suggesting that impaired diffusion capacity may lead to reduced ability of alveolar nitric oxide to diffuse into pulmonary blood vessels.

An international study involving COPD patients, asthmatics, and healthy controls demonstrated inverse correlations of CaNO and FeNO50 with FEV1 and MMEF, with CaNO exhibiting stronger associations32. Brindicci et al. reported a significant negative correlation between CaNO and FEV1%pre in smokers and COPD patients (r = -0.6, P < 0.001)33. A recent domestic study revealed that FeNO50 was negatively correlated with FEF50%, whereas both FeNO200 and CaNO were negatively associated with FVC, FEV1, PEF, FEF75%, and MMEF%pre34, which is consistent with our findings. Fan et al. compared eNO levels among 59 stable COPD patients and 39 AECOPD patients stratified by GOLD stage (1–2, 3, and 4) and reported significantly greater CaNO levels in GOLD 1–2 patients than in GOLD 4 patients, whereas FeNO50 and FeNO200 levels showed no intergroup differences, which aligns with our results19. Current evidence on CaNO-lung function relationships is more established in asthma, where CaNO negatively correlates with FEV1%pre, FEV1/FVC, MMEF%pre, and FEF50%pre more robustly than does FeNO50, highlighting its superior specificity as an airway obstruction marker35.

This study analyzed HRCT-derived small airway metrics and revealed that RB1 WT/BSA, and TDR as well as RB1a WT/BSA, were negatively correlated with FEV1, z−scores, FEV1/FVC, z−scores, FEF75, z−scores, and MMEF, z−scores, whereas WT/PA of RB1, and RB1a were inversely associated with FEV1/FVC, z−scores, FEF75, z−scores, and MMEF, z−scores, and WA/BSA of RB1a was negatively linked to FEV1, z−scores, and MMEF, z−scores, TDR was negatively linked to FEV1, z−scores, and FEF75, z−scores. No significant correlations were detected for RB1 WA/BSA. The sensitivity of WT/BSA and WT/PA underscores that airway wall thickening better captures airflow limitation than does luminal narrowing alone, challenging the “homogeneous airway model” in COPD. These findings highlight the need for multilevel bronchial quantification to guide personalized management strategies.

Kumar et al. demonstrated significant negative correlations between FEV1%pre and three HRCT indices (LAA− 950%, WA%, and Pi10) in 50 COPD patients, with all the parameters showing high sensitivity and specificity in predicting FEV1%pre < 50%36. Hasegawa et al. analyzed 52 COPD patients and revealed that LA and WA% in RB1 and RB8 (3rd-6th generations) correlated more strongly with FEV1%pre as airway generation increased37. Koo et al.’s analysis of 370 COPD patients stratified by GOLD stage revealed greater airway wall thickening in GOLD 3–4 patients than in GOLD 1–2 patients, with Pi10 specifically associated with FEV1/FVC in GOLD 3 patients38, corroborated by another study linking higher Pi10 to lower FEV1%pre (P = 0.002) and FVC%pred (P < 0.001)39. The mechanism of COPD-related small-airway remodeling, driven by epithelial/stromal hyperplasia and nonspecific hyperresponsiveness, results in airway wall thickening that disrupts airflow dynamics and ultimately leads to a decrease in FEV125.

Studies indicate that the severity of emphysema in COPD patients is correlated with lung function impairment and predicts a decrease in FEV140. This study further revealed that LAA− 950% was negatively correlated with FEV1, z−scores, FEV1/FVC, z−scores, FEF75, z−scores, and MMEF, z−scores. Notably, LAA− 950% was greater in GOLD 3–4 patients than in GOLD 1–2 patients, with GOLD 1 < GOLD 3–4 and GOLD 2 < GOLD 4 (all P < 0.05). Similarly, Lu et al. (n = 52) reported significant inverse associations between LAA− 950% and FEV1%pre, FEV1/FVC, and MMEF%pre41. A meta-analysis confirmed strong negative correlations (pooled CC: -0.49 for inspiratory CT, -0.56 for expiratory CT) between LAA− 950% and FEV1%pre42. Koo et al.‘s study compared 10 healthy smokers and 20 COPD patients categorized into GOLD stages 1–4, revealing a significantly greater emphysema proportion in GOLD stage 4 patients than in other groups (P < 0.05)22, reinforcing that worsening lung function is correlated with greater emphysema severity.

Based on the above findings, patients with elevated CaNO and FeNO200 levels exhibit more severe small airway structural abnormalities and impaired lung function, necessitating early intervention to control type 2 airway inflammation. However, as this study did not obtain dynamic eNO data from patients, future research will further investigate the impact of dynamic eNO changes on airway structure and lung function.

Conclusion

This study demonstrated significant alterations in chest HRCT quantitative indices in COPD patients compared with healthy controls, with more pronounced increases observed during AECOPD than during stable COPD. Notably, CaNO and FeNO200 were correlated with both HRCT indices and lung function parameters, with CaNO showing stronger associations, suggesting its potential as a sensitive biomarker for airway inflammation and remodeling in COPD patients. The negative correlations between HRCT airway indices and pulmonary function testing (spirometry) further support the critical role of airway inflammation and structural remodeling in COPD pathophysiology. Additionally, the inverse relationship between LAA− 950% and pulmonary function, along with its variation across GOLD stages, underscores the impact of emphysema severity on functional decline. These findings provide new insights into COPD pathobiology, potentially assisting in biomarker discovery and therapeutic strategies for improved disease monitoring and management. In conclusion, this study highlights significant interrelationships among HRCT airway indices, exhaled NO, and lung function in COPD patients, warranting further research to optimize their clinical application in personalized COPD management.