Mucus Plugs in Asthma at CT Associated with Regional Ventilation Defects at 3He MRI

Abstract

Background Airway mucus plugs in asthma are associated with exacerbation frequency, increased eosinophilia, and reduced lung function. The relationship between mucus plugs and spatially overlapping ventilation abnormalities observed at hyperpolarized gas MRI has not been assessed quantitatively. Purpose To assess regional associations between CT mucus plugs scored by individual bronchopulmonary segment and corresponding measurements of segmental ventilation defect percentage (VDP) at hyperpolarized helium 3 (³He) MRI. Materials and Methods In this secondary analysis of a Health Insurance Portability and Accountability Act-compliant prospective observational cohort, participants in the Severe Asthma Research Program (SARP) III (NCT01760915) between December 2012 and August 2015 underwent hyperpolarized ³He MRI to determine segmental VDP. Segmental mucus plugs at CT were scored by two readers, with segments scored as plugged only if both readers agreed independently. A linear mixed-effects model controlling for interpatient variability was then used to assess differences in VDP in plugged versus plug-free segments. Results Forty-four participants with asthma were assessed (mean age ± standard deviation, 47 years ± 15; 29 women): 19 with mild-to-moderate asthma and 25 with severe asthma. Mucus plugs were observed in 49 total bronchopulmonary segments across eight of 44 patients. Segments containing mucus plugs had a median segmental VDP of 25.9% (25th-75th percentile, 7.3%-38.3%) versus 1.4% (25th-75th percentile, 0.1%-5.2%; P < .001) in plug-free segments. Similarly, the model estimated a segmental VDP of 18.9% (95% CI: 15.7, 22.2) for mucus-plugged segments versus 5.1% (95% CI: 3.3, 7.0) for plug-free segments (P < .001). Participants with one or more mucus plugs had a median whole-lung VDP of 11.1% (25th-75th percentile, 7.1%-18.9%) versus 3.1% (25th-75th percentile, 1.1%-4.4%) in those without plugs (P < .001). Conclusion Airway mucus plugging at CT was associated with reduced ventilation in the same bronchopulmonary segment at hyperpolarized helium 3 MRI, suggesting that mucus plugging may be an important cause of ventilation defects in asthma. © RSNA, 2021 Online supplemental material is available for this article.

Graphical abstract

Figure 1:

Flowchart of study sample and image processing steps with number and percentage of total at each stage. All imaging was performed after administration of a bronchodilator. Segmental mucus plugs were scored by a team of radiologists as part of a larger study that required 3-year follow-up CT. Thus, patients who were lost to follow-up because they moved out of the area or otherwise withdrew from the study (n = 5) were not scored and were excluded from our final sample. SARPIII = Severe Asthma Research Program III, VDP = ventilation defect percentage.

Figure 2:

Representative images of (A) segmental ventilation defect percentage (VDP) affecting 31% of the segmental lobe volume at hyperpolarized helium 3 MRI (arrow) and (B) a spatially overlapping mucus plug visualized at CT (arrowheads).

Figure 3:

Box plot shows segmental ventilation defect percentage (VDP) in segments with no mucus plug (left) and with at least one mucus plug (right). Horizontal lines represent the median, 25th percentile, and 75th percentile values. Each dot represents one individual bronchopulmonary segment. Of the 825 total bronchopulmonary segments evaluated, 49 (5.9%) were identified as having a mucus plug. Segments with mucus plugs had a median segmental VDP of 25.9% (25th–75th percentile, 7.3%–38.3%) versus 1.4% (25th–75th percentile, 0.1%–5.2%) in segments without a mucus plug (P < .001).

Figure 4:

Plots show segmental ventilation defect percentage (VDP) in plugged versus nonplugged segments in each of the eight individual participants (participants A–H) with identified mucus plugs. Horizontal lines represent the median, 25th percentile, and 75th percentile values. Each dot represents one individual bronchopulmonary segment. The number of plugs identified in each participant ranged from 1 to 11, as indicated in each subplot title, with a median value of 5.5 plugged segments. Median segmental VDP in plugged segments was higher than that in the unplugged segments in all but one participant (participant H, with only one mucus plug identified). In specific participants, there is considerable overlap in values for plugged versus unplugged segments, as in participant B, for whom nearly all segments have substantial segmental VDP, and in participant C, for whom all segments have relatively low segmental VDP.