Mucus Plugs and Emphysema in the Pathophysiology of Airflow Obstruction and Hypoxemia in Smokers (2024)

Keywords: COPD, computed tomography, FEV₁, mucus plugs, emphysema

At a Glance Commentary

SPIROMICS

A total of 2,973 participants have been recruited into the following four strata: never-smokers of cigarettes (stratum 1), current or former smokers without airflow obstruction (stratum 2), never-smokers of cigarettes with airflow obstruction (stratum 3) and current or former smokers with airflow obstruction (detailed in the online supplement). For the study reported here, 20 baseline multidetector CT (MDCT) lung scans were randomly selected from participants in stratum 1, 100 baseline scans were randomly selected from participants in each of strata 2 and 3, and 200 baseline scans were randomly selected from participants in stratum 4. The online supplement details methods for measuring peripheral oxygen saturation (Sp_O2) using pulse oximetry, the 6-minute-walk test (6MWT), the COPD Assessment Test (CAT) (9), mucus-related questions (10, 11), and sputum analysis.

Emphysema Percentage

SPIROMICS previously reported on quantitative CT lung assessment system (12) and methods for measuring emphysema and airway wall thickness (8). Emphysema percentage was calculated as the percentage of lung voxels below −950 Hounsfield units. To account for normal variation in emphysema percentage, derived reference equations for emphysema percentage were used that included terms for age; sex, race, and ethnicity; height; and body mass index category, as previously described (7). In this way, a binary variable of emphysema was generated in which “emphysema present” is an emphysema percentage value above the upper limit of normal and “emphysema absent” is an emphysema percentage value below the lower limit of normal. Additional details are provided in the online supplement.

MDCT Mucus Plug Score

For the study reported here, four radiologists (B.M.E, T.H., D.G., and M.L.S.) with subspecialty training in thoracic radiology scored mucus plugs on the MDCT scans using a scoring system based on bronchopulmonary segment anatomy, as described previously by a subset of the authors (4) and further detailed in the online supplement. Briefly, mucus plugs were identified as areas of opacification within the airway lumen, contiguous with patent airway lumen across sequential transverse CT slices. These opacities were less radiodense than adjacent blood vessels, and mucus plugs were defined as complete occlusion of a bronchus, irrespective of generation or size. When parallel to the scan plane, mucus plugs were recognized as tubular densities with or without branching. When oriented obliquely or perpendicularly to the scan plane, they were identified as oval or rounded opacities seen on sequential slices and were differentiated from blood vessels by their continuity with patent portions of the bronchial lumen and their position relative to adjacent blood vessels (see Supplemental Video 1 in Reference 4). The segments of each lobe were systematically examined for the presence or absence of mucus plugs and given a score of 1 or 0, accordingly. This generated a segment score ranging from 0 to 20 for each patient. The radiologists also examined the scans for the presence of bronchiectasis, as defined by a broncho:arteriole ratio of >1.5.

The 420 scans were randomly assigned to the four radiologists to be scored. The mucus plug score for each patient’s scan was generated by an individual radiologist. Radiologist scores were not combined or averaged in this paper. The radiologists analyzing the scans were blinded to any clinical details of the subjects and entered their mucus score data in real-time into Research Electronic Data Capture, a secure online study survey instrument. Interrater agreement was assessed in a subset of 100 scans. Each scan in the subset was assigned to be independently rescored by a second radiologist randomly selected from the three radiologists who did not score the initial scan. All radiologists participated in this concordance analysis with a relatively even representation of each radiologist pair in the analysis.

The SPIROMICS protocol was approved by the institutional review board at each participating institution, and all the participants provided written informed consent. The University of California, San Francisco, developed protocol for mucus plug scoring on MDCT lung scan analysis was reviewed and approved by the University of California, San Francisco, institutional review board.

Statistics

Study analysis details are in the online supplement. Linear regression models were informed by directed acyclic graphs (DAGs), and minimal adjustment sets required for regression modeling were identified using the web-based “DAGitty” platform (13). A negative binomial regression was used to model the association between mucus plugs and exacerbations adjusted for covariates. Wald tests were performed to assess the statistical significance of interaction terms, and statistically significant interactions were displayed using marginal-effect estimation (14). Marginal effects were calculated using the margins command in STATA (StataCorp). Emphysema was divided into tertiles for the interaction term, as it simplifies the model of interaction, allowing the use of the margins command (which does not accept noninteger values for factor variables) and making it easier to interpret and graph the interaction. Receiver operator characteristic (ROC) curve analyses with comparisons of the areas under multiple ROC curves were performed using the roccomp command in STATA. Statistical significance was accepted for two-sided P values of less than 0.05.

Human Subjects

Table 1 shows the clinical characteristics of the 300 patients with COPD, the 100 smokers without airflow obstruction, and the 20 nonsmoking subjects.

Airway Mucus Plugs and Emphysema Are Prevalent in Smokers

Mucus plugs were highly prevalent in smokers and appeared as focal or branching opacities, usually seen in subsegmental airways in the absence of bronchial dilatation. Examples of plugs occluding large and small airways are shown in Figures 1 and 2. The intraluminal plugs were only scored if they completely occluded the lumen; partial occlusions were not scored, as previously explained (4). Radiographically apparent intraluminal plugs comprise a mixture of mucins, plasma proteins, and inflammatory cells, and the term “mucus plugs” that we use here does not infer a predominance of any one of these components over the other. The intraclass correlation coefficient for interrater agreement in mucus plug score was 0.80.

Among the 400 ever-smokers, 229 (57%) had mucus plugs and 207 (52%) had emphysema. The median mucus plug score was 0 in the nonsmoking healthy control subjects, 0 in the ever-smokers with preserved lung function, and 3 in the ever-smokers with airflow obstruction (post-bronchodilator FEV₁/FVC of less than 0.70) (patients with COPD) (Figure 3A). Five of the 20 scans from the nonsmoking healthy control subjects had mucus scores between 1 and 5, a higher prevalence of mucus plugging in health than we have previously reported (4). However, the previous healthy control subjects we studied were nearly 30 years younger than the healthy control subjects studied here, and we suspect (minor) mucus plugging associated with older age. The 95% percentile value of the mucus plug scores in the nonsmokers was 5, a value used as the mucus plug high/low cutoff in subsequent analyses. Among all patients with COPD, 67% had a mucus plug score higher than 0 (Figure 3B), and this prevalence was similar in former smokers and current smokers (70% vs. 64%; P = 0.96). A diagnosis of asthma occurred with similar frequency in the low and high mucus subgroups in the ever-smoker subgroups (see Table E1 in the online supplement).

Among the current or former smokers, 48 of 400 (12%) had bronchiectasis, and the prevalence of bronchiectasis was higher in current or former smokers with a high mucus plug score than in those with a low score (19% vs. 9%; P < 0.001).

Mucus Plug Scores, Emphysema Percentage, and Airflow Obstruction

The Global Initiative for Chronic Obstructive Lung Disease (GOLD) system categorizes airflow limitation in COPD on the basis of post-bronchodilator FEV₁. An FEV₁ >80% predicted is GOLD stage 1, an FEV₁ between 50% and 80% is GOLD 2, an FEV₁ between 30% and 50% is GOLD 3, and an FEV₁ <30% is GOLD 4. We found that the mucus plug score was significantly higher in GOLD 3 and 4 patients than in smokers with preserved lung function (Figure 4A) and that 60% of GOLD 4 patients had a high mucus plug score (Figure 4B). We also found that the emphysema percentage was significantly higher in GOLD 2, 3, and 4 patients than in smokers with preserved lung function (Figure 4C) and that 96% of the GOLD 4 patients had emphysema percentage values higher than the upper limit of normal (Figure 4D). The relationship between mucus plug scores and emphysema percentage values in smokers was relatively weak (r_s = 0.42; P < 0.001) (Figure 4E), making it possible to explore the independent effects of these pathologies on airflow obstruction. A DAG identified emphysema, airway wall thickness expressed as the square root of wall area of a 10-mm lumen perimeter (Pi10), and smoking pack-years as the minimal sufficient adjustment set of covariates for estimating the total effect of mucus plug score on FEV₁ (Figure 4F). In linear regression models adjusting for these covariates, we found that both mucus plug score and emphysema percentage were independently associated with post-bronchodilator FEV₁ and FVC in smokers (Table 2). However, emphysema had a strong modifying effect on the relationship between mucus plug scores and FEV₁ in smokers, as revealed by an interaction term constructed between mucus plug score and tertiles of emphysema and the Wald test for interaction (P < 0.001) (Figure 5A). The slope of the line describing the inverse relationship between mucus plug score and FEV₁ was steeper in smokers in the lowest tertile of emphysema percentage values (emphysema percentage <2.4) than in the highest tertile (emphysema percentage >13.0) (Figure 5A). This strong modifying effect of emphysema was true for other spirometric outcomes as well (Table E2 and Figure E1). In other analyses, we explored the independent effects of emphysema and mucus plugs on airflow obstruction by stratifying smokers by presence or absence of emphysema. In these analyses, we found that smokers with mucus plugs have a low post-bronchodilator FEV₁ even when emphysema is absent (Figure 5B). Notably, the median FEV₁ in smokers with a high mucus plug score and no emphysema is similar to the median FEV₁ in smokers with emphysema and a low mucus plug score (45% vs. 44%) (Figure 5B).

Mucus Plug Scores, Emphysema Percentage, and Sp_O2

Among smokers, 28% were using supplemental oxygen, and oxygen use was higher in subgroups with a high mucus plug score and with emphysema (Table E1). Compared with smokers whose resting or postexercise Sp_O2 was >92%, those with resting or postexercise Sp_O2 <92% had higher mucus plug scores and higher emphysema percentage values (Table E3); this remained true when smokers using supplemental oxygen were excluded from the analyses (Table E3).

A DAG identified emphysema, airway wall thickness (Pi10), and smoking pack-years as the minimal sufficient adjustment set of covariates for estimating the total effect of mucus plug score on Sp_O2 (Figure 5F). Using regression models, we found that mucus plug score, emphysema percentage, and airway wall thickness were all independently related to resting and postexercise Sp_O2 in smokers, whereas smoking pack-years were not (Table 2). However, we found a strong modifying effect of emphysema percentage on the relationship between mucus plug scores and Sp_O2 (both resting and postexercise Sp_O2) in smokers, as revealed by the interaction term constructed between mucus plug score and tertiles of emphysema and the Wald test for interaction (P < 0.001) (Figure 5C). The slope of the line describing the inverse relationship between mucus plug score and resting Sp_O2 was steeper in smokers in the lowest tertile of emphysema percentage values than in those in the highest tertile (Figure 5C). We also found strong modifying effects of emphysema on the relationship between mucus plug scores and postexercise Sp_O2 (P value test for interaction, P < 0.001) (Figure 5E). In other analyses, we explored the independent effects of emphysema and mucus plugs on Sp_O2 by stratifying smokers by presence or absence of emphysema. In these analyses, we found that smokers with mucus plugs have low resting and postexercise Sp_O2 values even when emphysema is absent (Figures 5D and 5F).

Relationships between Mucus Plug Scores, Symptoms of Chronic Mucus Hypersecretion, COPD Control Outcomes, and Sputum Granulocytes

Among all smokers, 23% had symptoms of chronic mucus hypersecretion (CMH) using World Health Organization criteria (15), and the prevalence of CMH was higher in current smokers than in former smokers (35% vs. 16%; P < 0.001). In current smokers, the prevalence of CMH was similar in patients with a high mucus score and those with a low score (44% vs. 31%; P = 0.18), but the prevalence of CMH was higher in former smokers with a high mucus score than in those with a low score (27.9% vs. 11.2%; P = 0.001). Overall, the sensitivity of CMH symptoms for a high mucus plug score was low (33%) and the specificity was moderate (81%). Similarly, the sensitivity of sputum symptoms captured using the St. George’s Respiratory Questionnaire data for a high mucus plug score in former smokers was also low (34%), as was the specificity (75%).

CAT scores were higher in patients with a high mucus plug score than in those with a low score (17.5 ± 7.6 vs. 14.3 ± 8.5; P < 0.001), and in a linear regression model that controlled for age, sex, race, and smoking status, the mucus plug score was significantly associated with CAT score (β = 3.77; 95% CI, 2.0–5.5; P < 0.001). In addition, the number of COPD exacerbations per person per year was higher in patients with a high mucus plug score than in those with a low score, and this was true when the exacerbation number was for the year before enrollment (0.75 ± 1.1 vs. 0.43 ± 0.85; Kruskal-Wallis test; P = 0.004) or for the year after enrollment (0.89 ± 1.3 vs. 0.43 ± 1.0; Kruskal-Wallis test; P = 0.004). In a negative binomial regression, the mucus plug score was associated with both the preenrollment COPD exacerbation number (β = 0.63; 95% CI, 0.27–1.0; P = 0.001) and the Year 1 postenrollment COPD exacerbation number (β = 0.75; 95% CI, 0.35–1.2; P < 0.001). The mucus plug score and the CAT score predicted exacerbations (≥1 treatments with steroids in the 12 months from baseline to Year 1) and hospitalizations (≥1 hospitalization for COPD in the 12 months from baseline to Year 1) similarly, as evidenced by similar areas under the curve in ROC curve analyses (exacerbations: 0.62 vs. 0.67; P = 0.2; hospitalizations: 0.59 vs. 0.66; P = 0.18). Finally, the 6MWT distance was lower in patients with a high mucus plug score than in those with a low score (329 ± 115 m vs. 392 ± 117 m; P < 0.001), and in a linear regression model controlled as above, the mucus plug score was significantly inversely associated with 6MWT distance (β = −59.3; 95% CI −85.4 to −33.3; P < 0.001).

Among the 400 smokers, 127 had data for neutrophil percentage and eosinophil percentage in induced sputum. The median [interquartile range] sputum neutrophil percentage was higher in patients with a high mucus plug score than in those with a low score (86% [71–94%] vs. 72% [52–83%]; P < 0.001). The median [interquartile range] sputum neutrophil number (×10⁶/ml) was also higher in patients with a high mucus plug score than in those with a low score (1.12 [0.27–3.62] vs. 0.24 [0.09–0.59]; P = 0.0002). The sputum eosinophil percentage was similar in both subgroups (0.1 [0–1.1] vs. 0.3% [0–1.0]; P = 0.55), and the sputum eosinophil number was also similar (data not shown).

Stability of the Mucus Plug Phenotype in Smokers with COPD

To determine the stability of the mucus plug phenotype in smokers, we focused on analysis of mucus plugs in CT lung scans from a subset of the smokers with COPD (SPIROMICS stratum 3 and 4), and we took advantage of the fact that CT lung scans were repeated at 1 year of follow up in SPIROMICS participants. Specifically, we randomly selected 100 scan pairs from patients with COPD stratified by tertiles of segment score to ensure a broad representation of mucus plug scores. The scan reads were divided among the four radiologists. Scan pairs from each patient were read by the same radiologist to eliminate any between-reader error (within-reader variability was minimal, as previously described) (4). Radiologists did not score the same subject sequentially or have access to both scans at the same time to compare scores. In this way, we found that the mucus plug phenotype among patients with COPD was remarkably stable (Figure 6A). Patients with COPD with low mucus plug scores (0–1) at baseline tended to have low scores at Year 1, and patients with high mucus plug scores (7–20) at baseline tended to have high scores at Year 1 (Figure 6A). We also did analyses based on bronchopulmonary segments. Among segments in which mucus plugs were absent at baseline, the vast majority of segments remained free of mucus plugs 1 year later (Figure 6B); among segments in which mucus plugs were present at baseline, the majority of segments still had mucus plugs 1 year later (Figure 6C).

Supplements

Author disclosures

Acknowledgment

The authors thank the SPIROMICS participants and participating physicians, investigators, and staff for making this research possible. More information about the study and how to access SPIROMICS data is at www.spiromics.org. The authors acknowledge the following current and former investigators of the SPIROMICS sites and reading centers: Neil E. Alexis, M.D.; Mehrdad Arjomandi, M.D.; Lori A. Bateman, M.Sc.; Surya P. Bhatt, M.D.; Ronald G. Crystal, M.D.; Jeffrey L. Curtis, M.D.; Christine M. Freeman, Ph.D.; Craig Galban, Ph.D.; Yvonne Huang, M.D.; Robert J. Kaner, M.D.; Richard E. Kanner, M.D.; Eric C. Kleerup, M.D.; Lisa M. LaVange, Ph.D.; Deborah A. Meyers, Ph.D.; Wendy C. Moore, M.D.; John D. Newell, Jr., M.D.; Laura Paulin, M.D., M.H.S.; Cheryl Pirozzi, M.D.; Nirupama Putcha, M.D., M.H.S.; Elizabeth C. Oelsner, M.D., M.P.H.; Sanjeev Raman, M.B. B.S., M.D.; Stephen I. Rennard, M.D.; Donald P. Tashkin, M.D.; J. Michael Wells, M.D.; and Robert A. Wise, M.D. The project officers from the Lung Division of the NHLBI were Lisa Postow, Ph.D., and Lisa Viviano, B.S.N.

• 1.Celli BR, Wedzicha JA. Update on clinical aspects of chronic obstructive pulmonary disease. N Engl J Med. 2019;381:1257–1266. doi: 10.1056/NEJMra1900500. [DOI] [PubMed] [Google Scholar]
• 2.Barnes PJ, Burney PG, Silverman EK, Celli BR, Vestbo J, Wedzicha JA, et al. Chronic obstructive pulmonary disease. Nat Rev Dis Primers. 2015;1:15076. doi: 10.1038/nrdp.2015.76. [DOI] [PubMed] [Google Scholar]
• 3.Burgel P-R, Martin C. Mucus hypersecretion in COPD: should we only rely on symptoms? Eur Respir Rev. 2010;19:94–96. doi: 10.1183/09059180.00004410. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 4.Dunican EM, Elicker BM, Gierada DS, Nagle SK, Schiebler ML, Newell JD, et al. National Heart Lung and Blood Institute (NHLBI) Severe Asthma Research Program (SARP) Mucus plugs in patients with asthma linked to eosinophilia and airflow obstruction. J Clin Invest. 2018;128:997–1009. doi: 10.1172/JCI95693. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 5.Smith BM, Austin JH, Newell JD, Jr, D’Souza BM, Rozenshtein A, Hoffman EA, et al. Pulmonary emphysema subtypes on computed tomography: the MESA COPD study. Am J Med. 2014;127:94.e7–94.e23. doi: 10.1016/j.amjmed.2013.09.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 6.Lynch DA, Moore CM, Wilson C, Nevrekar D, Jennermann T, Humphries SM, et al. Genetic Epidemiology of COPD (COPDGene) Investigators. CT-based visual classification of emphysema: association with mortality in the COPDGene study. Radiology. 2018;288:859–866. doi: 10.1148/radiol.2018172294. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 7.Hoffman EA, Ahmed FS, Baumhauer H, Budoff M, Carr JJ, Kronmal R, et al. Variation in the percent of emphysema-like lung in a healthy, nonsmoking multiethnic sample: the MESA lung study. Ann Am Thorac Soc. 2014;11:898–907. doi: 10.1513/AnnalsATS.201310-364OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 8.Couper D, LaVange LM, Han M, Barr RG, Bleecker E, Hoffman EA, et al. SPIROMICS Research Group. Design of the subpopulations and intermediate outcomes in COPD study (SPIROMICS) Thorax. 2014;69:491–494. doi: 10.1136/thoraxjnl-2013-203897. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 9.Jones PW, Harding G, Berry P, Wiklund I, Chen WH, Kline Leidy N. Development and first validation of the COPD assessment test. Eur Respir J. 2009;34:648–654. doi: 10.1183/09031936.00102509. [DOI] [PubMed] [Google Scholar]
• 10.Kim V, Crapo J, Zhao H, Jones PW, Silverman EK, Comellas A, et al. COPDGene Investigators. Comparison between an alternative and the classic definition of chronic bronchitis in COPDGene. Ann Am Thorac Soc. 2015;12:332–339. doi: 10.1513/AnnalsATS.201411-518OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 11.Kim V, Zhao H, Regan E, Han MK, Make BJ, Crapo JD, et al. COPDGene Investigators. The St. George’s respiratory questionnaire definition of chronic bronchitis may be a better predictor of COPD exacerbations compared with the classic definition. Chest. 2019;156:685–695. doi: 10.1016/j.chest.2019.03.041. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 12.Sieren JP, Newell JD, Jr, Barr RG, Bleecker ER, Burnette N, Carretta EE, et al. SPIROMICS Research Group. SPIROMICS protocol for multicenter quantitative computed tomography to phenotype the lungs. Am J Respir Crit Care Med. 2016;194:794–806. doi: 10.1164/rccm.201506-1208PP. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 13.Textor J, van der Zander B, Gilthorpe MS, Liskiewicz M, Ellison GT. Robust causal inference using directed acyclic graphs: the R package ‘dagitty’. Int J Epidemiol. 2016;45:1887–1894. doi: 10.1093/ije/dyw341. [DOI] [PubMed] [Google Scholar]
• 14.Vittinghoff E, Glidden DV, Shiboski SC, McCulloch CE. Regression methods in biostatistics: linear, logistic, survival, and repeated measures models. New York, NY: Springer-Verlag; 2012. [Google Scholar]
• 15.American Thoracic Society. Definitions and classification of chronic bronchitis, asthma and pulmonary emphysema. Am Rev Respir Dis. 1962;85:762–768. [Google Scholar]
• 16.Sullivan J, Pravosud V, Mannino DM, Siegel K, Choate R, Sullivan T. National and state estimates of COPD morbidity and mortality — United States, 2014-2015. Chronic Obstr Pulm Dis (Miami) 2018;5:324–333. doi: 10.15326/jcopdf.5.4.2018.0157. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 17.Okajima Y, Come CE, Nardelli P, Sonavane SK, Yen A, Nath HP, et al. Luminal plugging on chest CT scan: association with lung function, quality of life, and COPD clinical phenotypes. Chest. 2020;158:121–130. doi: 10.1016/j.chest.2019.12.046. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 18.Wagner PD, Dantzker DR, Dueck R, Clausen JL, West JB. Ventilation-perfusion inequality in chronic obstructive pulmonary disease. J Clin Invest. 1977;59:203–216. doi: 10.1172/JCI108630. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 19.Petersson J, Glenny RW. Gas exchange and ventilation-perfusion relationships in the lung. Eur Respir J. 2014;44:1023–1041. doi: 10.1183/09031936.00037014. [DOI] [PubMed] [Google Scholar]
• 20.Boucher RC. Muco-obstructive lung diseases. N Engl J Med. 2019;380:1941–1953. doi: 10.1056/NEJMra1813799. [DOI] [PubMed] [Google Scholar]

Supplementary Materials