Mucus plugs in patients with asthma linked to eosinophilia and airflow obstruction

Abstract

Background: The link between mucus plugs and airflow obstruction has not been established in chronic severe asthma, and the role of eosinophils and their products in mucus plug formation is unknown.

Methods: In clinical studies, we developed and applied a bronchopulmonary segment-based scoring system to quantify mucus plugs on multidetector computed tomography (MDCT) lung scans from 146 subjects with asthma and 22 controls, and analyzed relationships among mucus plug scores, forced expiratory volume in 1 second (FEV1), and airway eosinophils. Additionally, we used airway mucus gel models to explore whether oxidants generated by eosinophil peroxidase (EPO) oxidize cysteine thiol groups to promote mucus plug formation.

Results: Mucus plugs occurred in at least 1 of 20 lung segments in 58% of subjects with asthma and in only 4.5% of controls, and the plugs in subjects with asthma persisted in the same segment for years. A high mucus score (plugs in ≥ 4 segments) occurred in 67% of subjects with asthma with FEV1 of less than 60% of predicted volume, 19% with FEV1 of 60%-80%, and 6% with FEV1 greater than 80% (P < 0.001) and was associated with marked increases in sputum eosinophils and EPO. EPO catalyzed oxidation of thiocyanate and bromide by H2O2 to generate oxidants that crosslink cysteine thiol groups and stiffen thiolated hydrogels.

Conclusion: Mucus plugs are a plausible mechanism of chronic airflow obstruction in severe asthma, and EPO-generated oxidants may mediate mucus plug formation. We propose an approach for quantifying airway mucus plugging using MDCT lung scans and suggest that treating mucus plugs may improve airflow in chronic severe asthma.

Trial registration: Clinicaltrials.gov NCT01718197, NCT01606826, NCT01750411, NCT01761058, NCT01761630, NCT01759186, NCT01716494, and NCT01760915.

Funding: NIH grants P01 HL107201, R01 HL080414, U10 HL109146, U10 HL109164, U10 HL109172, U10 HL109086, U10 HL109250, U10 HL109168, U10 HL109257, U10 HL109152, and P01 HL107202 and National Center for Advancing Translational Sciences grants UL1TR0000427, UL1TR000448, and KL2TR000428.

Keywords: Asthma; Cytokines; Pulmonology; Th2 response.

Figure 1. Consort diagram of CT substudy.

Flow chart shows the number of asthma patients who were screened, enrolled, and included in the final analyses.

Figure 2. Development and distribution of the CT mucus score in asthma patients and healthy subjects.

(A) Mucus plug with branching (yellow arrow) seen in longitudinal section is identified as a tubular opacification (frontal plane). (B) Mucus plug (yellow arrow) with extensive branching seen in longitudinal section (transverse plane). (C) Mucus plug (yellow arrow) seen in cross-section is identified as rounded opacification (transverse plane). (D) Schematic representation showing how MDCTs were evaluated to generate the mucus score. Airways within the 2 cm peripheral zone on MDCT (shown in red) or airways that were partially occluded were excluded from assessment. Mucus plugs were defined as complete occlusion of an airway. Each bronchopulmonary segment was assessed and scored for the presence or absence of 1 or more mucus plug(s), and the segment scores were summed to generate the mucus score. (E) Segment score in healthy patients and patients with asthma. (F) Frequency distribution of segment score in patients with asthma. The color code above the x axis defines 3 mucus groups: green indicates patients with a mucus score of 0 (zero mucus group); blue indicates patients with mucus scores between 0.5 and 3.5 (low mucus group); and orange indicates patients with mucus scores of 4.0 or more (high mucus group). (G) Sankey bar graph showing the change in mucus score in 25 asthmatic subjects from SARP 1/2 to SARP 3. (H) Pie chart of segments with mucus plugging on baseline scan; 65% of these segments had mucus plugging on rescan. Pie chart of segments with no mucus plugging on baseline scan; 79% of these segments had no mucus plugging on rescan. (I) MDCTs showing a mucus plug occluding the airway (yellow arrow) of the right lower lobe in 2010 and a mucus plug occluding the same airway, visible more proximally (yellow arrow) and branching into the adjacent airway, in 2013. ***P < 0.001, unequal variances t test.

Figure 3. Relationship between bronchiectasis and mucus plugging.

(A) Frequency distribution of bronchiectasis score in patients with asthma. (B) Prevalence of bronchiectasis versus mucus plugging in each lung lobe. The prevalence of mucus plugging is 4 to 5 times higher than the prevalence of bronchiectasis in each lobe. There is no significant difference in prevalence of bronchiectasis or mucus plugging across individual lobes. (C) Mucus plugging is present in 35% of lobes that have no bronchiectasis present and 58% of lobes that have bronchiectasis present (P = 0.001). (D) Bronchiectasis is present in 5% of lobes that have no mucus plugging present and only 12% of segments that have mucus plugging present (P = 0.001). There is a positive association between mucus plugging and bronchiectasis, but mucus plugging usually occurs in the absence of bronchiectasis. (E) Pie charts illustrating the prevalence of bronchiectasis in repeat CT scans in 25 patients. The data show that 83% of lung lobes with mucus plugs visible on the first scan had mucus plugging visible on the second scan; in contrast, 99% of lung lobes with no mucus plugs visible in the first scan also had no mucus plugs visible on the second scan.

Figure 4. Mucus plugging is associated with low lung function.

(A) Spirometric measures of lung function (FEV1, FVC, and FEV1/FVC) in the subjects with a high mucus score were significantly lower than in subjects with a low mucus score and subjects with a zero mucus score. (B) High mucus plug scores were much more common in patients with severe airflow obstruction. ***P < 0.001; **P < 0.01, Kruskal-Wallis test with Dunn’s correction.

Figure 5. Persistent airflow obstruction is seen in subjects with high mucus scores after treatment with bronchodilators and steroids.

(A) The absolute change in FEV1 percentage of predicted volume after bronchodilator treatment did not differ across mucus groups. (B) The FEV1 percentage of predicted volume after bronchodilator treatment was significantly lower in the high mucus group than in the zero mucus group. (C) Residual postbronchodilator abnormalities in FEV1 (FEV1 < 80%) occur more commonly in subjects with a high mucus score than in those with a zero mucus score. (D) The absolute change in FEV1 percentage of predicted volume after steroid treatment did not differ across mucus groups. (E) The FEV1 percentage of predicted volume after steroid treatment was significantly lower in the high and low mucus groups than in the zero mucus group. (F) Residual poststeroid abnormalities in FEV1 (FEV1 < 80%) occurred more commonly in subjects with a high mucus score than in those with a zero mucus score. (G) The absolute change in FEV1 percentage of predicted volume after bronchodilator and steroid treatment was significantly higher in the high mucus group than in the zero mucus group. (H) The FEV1 percentage of predicted volume after bronchodilator and steroid treatment was significantly lower in the high mucus group than in the zero mucus group. (I) Residual postbronchodilator and poststeroid abnormalities in FEV1 (FEV1 < 80%) occurred more commonly in subjects with a high mucus score than in those with a zero mucus score. ***P < 0.001; **P < 0.01, Kruskal-Wallis test with Dunn’s correction.

Figure 6. High mucus score is associated with markers of type 2 inflammation.

(A) Sputum eosinophil percentage is significantly increased in patients with a high mucus score and remains significantly increased in patients with a high mucus score following treatment with intramuscular steroid (triamcinolone acetonide). (B) The sputum eosinophil percentage is significantly and positively associated with the mucus score. (C) Gene expression for IL-13 is significantly increased in patients with a high mucus score and remains significantly increased in patients with a high mucus score following treatment with intramuscular steroid. (D) Gene expression for IL-5 is significantly increased in patients with a high mucus score and remains significantly increased in patients with a high mucus score following treatment with intramuscular steroid. (E) The MUC5AC/MUC5B ratio is significantly increased in patients with high mucus scores. ***P < 0.001; **P < 0.01; *P < 0.05. P values were determined by Kruskal-Wallis test with Dunn’s correction unless otherwise indicated.

Figure 7. Eosinophil products are associated with mucus plugging.

(A) The sputum eosinophil percentage is positively associated with sputum EPO levels. (B) Sputum EPO is higher in the high mucus group (n = 32) than in the zero mucus group (n = 45) and healthy controls (n = 39). (C) Schematic representation of the cysteine-linking assay: 2 cysteines labeled with BODIPY FL fluoresce green as monomers but quench when oxidized to form a cystine dimer. (D) Effect of EPO and H₂O₂ on cysteine crosslinking in the presence of chloride, bromide, or thiocyanate. Cysteines do not undergo significant crosslinking with EPO and H₂O₂ in the presence of chloride, but cysteines exposed to EPO and H₂O₂ in the presence of bromide, and especially thiocyanate, undergo much more oxidation and crosslinking. RFU, relative fluorescent units. (E) Effect of HOSCN, the product of EPO-catalyzed reaction of H₂O₂ and thiocyanate, on the viscoelastic properties of a thiolated hydrogel measured by rheology. A large increase in the elastic modulus (G′) of the hydrogel was seen following exposure of the hydrogel to EPO with H₂O₂ and KSCN. There was no significant increase in G′ in the hydrogel in the absence of EPO. (F) Conceptual model for how type 2 inflammation promotes airway mucus plug formation in asthma. IL-13 increases thiocyanate transfer into the airway lumen. Once in the airway lumen, it is oxidized by H₂O₂ to form HOSCN, a reaction catalyzed by EPO. HOSCN targets cysteine thiol groups in secreted mucin polymers to generate covalent disulfide mucin crosslinks. Crosslinked mucins have a high elasticity that decreases their clearance by the mucociliary escalator and results in mucus plug formation. The data are presented as mean ± SD of 3 replicates in D and 4 replicates in E. *P < 0.05; **P < 0.01; ^†P < 0.01; ^‡P < 0.001. P value was determined by ANOVA with Bonferroni’s correction.