Micro-Computed Tomography Comparison of Preterminal Bronchioles in Centrilobular and Panlobular Emphysema

Abstract

Rationale: Very little is known about airways that are too small to be visible on thoracic multidetector computed tomography but larger than the terminal bronchioles.

Objectives: To examine the structure of preterminal bronchioles located one generation proximal to terminal bronchioles in centrilobular and panlobular emphysema.

Methods: Preterminal bronchioles were identified by backtracking from the terminal bronchioles, and their centerlines were established along the entire length of their lumens. Multiple cross-sectional images perpendicular to the centerline were reconstructed to evaluate the bronchiolar wall and lumen, and the alveolar attachments to the outer airway walls in relation to emphysematous destruction in 28 lung samples from six patients with centrilobular emphysema, 20 lung samples from seven patients with panlobular emphysema associated with alpha-1 antitrypsin deficiency, and 47 samples from seven control (donor) lungs.

Measurements and main results: The preterminal bronchiolar length, wall volume, total volume (wall + lumen), lumen circularity, and number of alveolar attachments were reduced in both centrilobular and panlobular emphysema compared with control lungs. In contrast, thickening of the wall and narrowing of the lumen were more severe and heterogeneous in centrilobular than in panlobular emphysema. The bronchiolar lumen was narrower in the middle than at both ends, and the decreased number of alveolar attachments was associated with increased wall thickness in centrilobular emphysema.

Conclusions: These results provide new information about small airways pathology in centrilobular and panlobular emphysema and show that these changes affect airways that are not visible with thoracic multidetector computed tomography scans but located proximal to the terminal bronchioles in chronic obstructive pulmonary disease.

Keywords: chronic obstructive pulmonary disease; computed tomography; imaging; small airway.

Figure 1.

Schematic explanation of preterminal bronchiolar assessment. (A) Three-dimensionally segmented preterminal bronchiole (gray scale) with centerline identified (blue). The circle and triangle indicate the proximal and peripheral ends of the centerline of the preterminal bronchiole, respectively. PreTB = preterminal bronchiole; TB = terminal bronchiole. (B) Cross-sectional image generated perpendicular to the centerline. (C) Image near a branch point, which was excluded from this study. (D and E) Segmentation of the airway wall and examples of alveolar attachments to the outer wall (arrowheads), respectively.

Figure 2.

Representative original micro–computed tomography image (left) and 10 reconstructed cross-sectional images from one preterminal bronchiole (right) from a control lung (A) and a lung affected by either centrilobular (B) or panlobular (C) emphysema. Each arrow indicates the location of the preterminal bronchiole. Scale bars indicate 1 mm.

Figure 3.

Micro–computed tomography measurements obtained using the methods illustrated in Figures 1 and 2. (A) Mean luminal area, total airway area (i.e., lumen + wall) were reduced in centrilobular emphysema (CLE) compared with panlobular emphysema (PLE) and control subjects, but mean wall area did not differ among the three groups. (B) Mean wall thickness was increased in CLE compared with control subjects and PLE. (C) Wall volume and total airway volume were reduced in CLE and PLE compared with control subjects. (D and E) Coefficients of variation (CV%) of luminal area and wall thickness within and between bronchioles in which greater variation was found in CLE than control subjects and PLE. Boxplots indicate the median value (horizontal line) and the interquartile range (box). *Adjusted P value < 0.05.

Figure 4.

Compares preterminal bronchiolar lumen shape and numbers of alveolar attachments to the outer walls of the airways. (A) Circularity of the lumen, (B) numbers of alveolar attachments, and (C) reduction in the numbers of attachments per length of outer airway wall circumference in centrilobular (CLE) and panlobular (PLE) emphysema compared with control subjects. (D) Reduction in number of attachments is associated with reduced internal diameter in CLE (solid circles, standardized beta [β*] = 0.59, P = 0.0006), PLE (open circles, β* = 0.62, P = 0.002) and control subjects (open triangles, β* = 0.72, P = 0.000004). (E) Reduction in numbers of attachments is associated with the greater wall thickness only in CLE (β* = −0.60; P = 0.0008). Boxplots indicate the median value (horizontal line) and the interquartile range (box). *Adjusted P value < 0.05.