Type 2 Inflammation Is Associated with Emphysema on Computed Tomography

To the Editor:

Between 20% and 40% of patients with chronic obstructive pulmonary disease (COPD) have evidence of predominant type 2 inflammation, often indicated by high peripheral blood eosinophil count (BEC). This endotype is associated with frequent exacerbations and with an airway phenotype, as evidenced by chronic bronchitis, and has been identified as a potential target for therapy (1, 2). Whether high BEC is also associated with the emphysema phenotype is not known. IL-13 (interleukin-13) is a key driver of type 2 inflammation and promotes MMP-12 (matrix metalloproteinase-12) production in alveolar macrophages, and it may contribute to emphysema formation (3). A recent study of murine emphysema suggested that eosinophil-derived cathepsin L may contribute to the degradation of the extracellular matrix (4). Prior epidemiologic studies of the relationship between BEC and emphysema were limited by exclusion of patients with mild disease or by the evaluation of emphysema by lobar quantification (5, 6). We hypothesized that high BEC would be associated with greater emphysema on computed tomography (CT) regardless of the severity of airflow obstruction.

We analyzed CT scans from participants enrolled in visit 2 of the COPDGene (Genetic Epidemiology of COPD) study who had available BECs (7). COPDGene included current and former smokers aged 45–80 years, with at least a 10 pack-year smoking history. High-resolution CT images were acquired at full inspiration (total lung capacity) and end-expiration. Inspiratory and expiratory CT scans were coregistered voxel to voxel using LungQ (Thirona) (8). Voxels with density < −950 Hounsfield units (HU) on inspiration and <−856 HU on expiration were categorized as emphysematous. Voxels with density > −950 HU on inspiration but <−856 HU on expiration were deemed to have functional small airway disease (fSAD) (9). The square root of the wall area of a theoretical airway with an internal perimeter of 10 mm (Pi10) was calculated to estimate airway wall thickening. BEC ≥ 300 cells/μl was considered indicative of high type 2 inflammation. The COPDGene study was approved by the institutional review boards of all 21 participating centers, and participants provided written informed consent before enrollment in the study. Multivariable generalized linear regression models were evaluated to test the associations between high BEC and CT emphysema, CT fSAD, and Pi10 in separate models, with adjustment for age, sex, race, body mass index, smoking status, pack-years of smoking, and CT scanner type. All analyses were performed using R statistical package version 4.2.2 (https://www.R-project.org/). A two-sided α value of 0.05 was deemed to indicate statistical significance.

Of the 6,284 participants in visit 2 of COPDGene, we included 4,866 participants after excluding 1,418 individuals without good-quality expiratory CT images. Of these, 904 (18.6%) had high BECs and 3,009 (61.8%) were active smokers. No airflow obstruction was noted in 2,131 (43.8%), and 557 (11.4%) had preserved ratio impaired spirometry. Of the 2,103 (43.2%) with airflow obstruction, 455 (9.4%), 980 (20.1%), 486 (10.0%), and 182 (3.7%) had Global Initiative for Chronic Obstructive Lung Disease stages 1–4, respectively. CT emphysema ranged from 0% to 60.8%, with a mean (standard deviation) of 4.6% (9.0%). There was no significant correlation between percentage emphysema and BEC. A comparison of participants with and without high BECs is shown in Table 1. On bivariable analyses, compared with those with low BECs, those with high BECs had a greater degree of percentage emphysema (mean ± standard deviation, 5.2 ± 9.5 vs. 4.4 ± 8.9; P = 0.026), percentage fSAD (17.0 ± 13.3 vs. 16.2 ± 13.0; P = 0.081), and Pi10 (2.39 ± 0.58 vs. 2.23 ± 0.56 mm; P < 0.001). On multivariable analyses, compared with low BEC, high BEC was associated with a greater degree of percentage emphysema (least square mean [LSM] ± standard error, 4.8 ± 0.3 vs. 4.2 ± 0.1; P = 0.048) and Pi10 (LSM, 2.42 ± 0.02 vs. 2.28 ± 0.01 mm; P < 0.001) but not percentage fSAD (LSM, 16.9 ± 0.4 vs. 16.4 ± 0.2; P = 0.308).

Table 1.

Comparison of characteristics of participants with and without high type 2 inflammation

	Overall (n = 4,866)	BEC < 300 cells/μl (n = 3,962)	BEC ≥ 300 cells/μl (n = 904)
Age, yr	65.4 (8.6)	65.2 (8.6)	66.1 (8.6)
Female sex, n (%)	2,399 (49.3)	1,974 (51.0)	375 (41.5)
African American race, n (%)	1,372 (28.2)	1,115 (28.8)	207 (22.9)
Body mass index, kg/m2	29.1 (6.4)	28.9 (6.3)	29.6 (6.4)
Current smokers, n (%)	3,009 (61.8)	2,380 (61.5)	583 (64.5)
Pack-years of smoking	43.9 (23.6)	43.2 (23.5)	46.3 (23.9)
FEV1, L	2.16 (0.83)	2.18 (0.83)	2.08 (0.84)
FEV1, % predicted	78.8 (24.2)	79.9 (24.2)	74.0 (23.9)
FVC, L	3.15 (0.94)	3.16 (0.94)	3.13 (0.95)
FVC, % predicted	87.4 (17.4)	88.2 (17.4)	84.5 (17.3)
FEV1:FVC	0.68 (0.15)	0.68 (0.14)	0.65 (0.15)
BEC, cells/μl	185 (147)	133 (67)	409 (179)
% Emphysema on CT	4.6 (9.0)	4.4 (8.9)	5.2 (9.5)
% fSAD on CT	16.3 (13.1)	16.2 (13.0)	17.0 (13.3)
Pi10, mm	2.27 (0.56)	2.23 (0.56)	2.39 (0.58)

Definition of abbreviations: BEC = blood eosinophil count; CT = computed tomography; FEV₁ = forced expiratory volume in 1 second; fSAD = functional small airway disease; FVC = forced vital capacity; Pi10 = square root of the wall area of a theoretical airway with an internal perimeter of 10 mm.

Data are expressed as mean (standard deviation) unless otherwise indicated.

In a cohort of current and former smokers with and without airflow obstruction, we demonstrated that type 2 inflammation, indicated by high BEC, is associated not only with greater airway wall thickness but also with emphysema. Predominant type 2 inflammation is increasingly recognized as a treatable endotype in COPD. Individuals with this endotype have been noted to experience frequent exacerbations and a faster decline in lung function. Recent studies have focused mostly on airway disease, given the known relationships between the prime drivers of type 2 inflammation (IL-4, IL-5, and IL-13) and epithelial barrier disruption, goblet cell hyperplasia, mucus production, and smooth muscle contractility. Two recent murine studies have shed light on the possible involvement of eosinophil-mediated pathways in emphysema development. Xu and colleagues demonstrated in a murine model of emphysema induced by porcine pancreatic elastase that the initial inflammation was predominantly neutrophilic, but repeated instillations of elastase resulted in alveolar eosinophilia, which was associated with emphysema (4). When eosinophils were depleted via IL-5 blockade, elastase instillation resulted in a significantly lower amount of emphysema. There was a parallel increase in eosinophil-derived cathepsin L, and its blockade resulted in lower degrees of emphysema formation. Doyle and colleagues showed in a mouse model that chronic type 2 inflammation of the lungs was associated with emphysema that was in turn associated with elevated MMP-12 concentrations mediated by eosinophil-derived IL-4/13 (3). Other proteases associated with emphysema, MMP-2 and MMP-9, were also elevated (3). IL-13 may be especially critical, given that it can polarize macrophages to an M2 phenotype and induce MMP-12 production (10). Both these studies also demonstrated in small numbers of human subjects that high BEC was associated with higher degree of emphysema (3, 4).

Previous human epidemiologic studies, however, have not shown these relationships between high BEC and emphysema on CT. In the COPDGene and ECLIPSE (Evaluation of COPD Longitudinally to Identify Predictive Surrogate Endpoints) cohorts, Yun and colleagues found no difference in CT emphysema between those with high and low BECs at thresholds of 200 and 300 cells/μl (5). That study, however, included only those with Global Initiative for Chronic Obstructive Lung Disease stages 2–4 in both cohorts. In the SPIROMICS (Subpopulations and Intermediate Outcome Measures in COPD Study) cohort, Hastie and colleagues stratified subjects by BEC threshold of 200 cells/μl and sputum eosinophil threshold of 1.25% and found no differences in CT emphysema by BEC but did find differences in some lobes by sputum eosinophilia (6). That study did not compare overall lung emphysema by BEC. We did not find a linear relationship between emphysema and BEC and therefore selected a BEC threshold of 300 cells/μl as indicative of high type 2 inflammation. This threshold is also associated with a significant increase in exacerbation risk (5). Exacerbations, in turn, are associated with faster emphysema progression (11). There are no known therapies for emphysema. Our findings should be confirmed in other cohorts and also in clinical trials of pharmacological therapies to test whether lowering type 2 inflammation can decrease the rate of progression of emphysema.

Strengths and Limitations

The strengths of the study include quality-controlled CT examinations and the inclusion of a large proportion of African Americans. The study has a few limitations. BEC was measured at only one time point, and therefore we may have missed some individuals with fluctuating BECs. Other markers of type 2 inflammation, such as fractional exhaled nitric oxide, immunoglobulin E, and periostin, were not measured. CT scans were not respiration gated, but participants were coached to full inspiration to acquire scans at total lung capacity.

Conclusions

Our results indicate that type 2 inflammation, indicated by high BEC, is associated with not only airway disease but also emphysema. This may be a potential target for therapy.