Comparison of airway remodelling assessed by computed tomography in asthma and COPD.

Abstract

Few studies have directly compared airway remodelling assessed by computed tomography (CT) between asthma and chronic obstructive pulmonary disease (COPD). The present study was conducted to determine whether there are any differences between the two diseases with similar levels of airflow limitation under clinically stable conditions.
Subjects included older male asthmatic patients (n = 19) showing FEV(1)/FVC <70% with smoking history less than 5-pack/year. Age- and sex-matched COPD patients (n = 28) who demonstrated similar airflow limitation as asthmatic patients and age-matched healthy non-smokers (n = 13) were recruited. Using proprietary software, eight airways were selected in the right lung, and wall area percent (WA%) and airway luminal area (Ai) were measured at the mid-portion of the 3rd to 6th generation of each airway. For comparison, the average of eight measurements per generation was recorded.
FEV(1)% predicted and FEV(1)/FVC was similar between asthma and COPD (82.3 ± 3.3% vs. 77.6 ± 1.8% and 57.7 ± 1.6% vs. 57.9 ± 1.4%). At any generation, WA% was larger and Ai was smaller in asthma, both followed by COPD and then controls. Significant differences were observed between asthma and controls in WA% of the 3rd to 5th generation and Ai of any generation, while no differences were seen between COPD and controls. There were significant differences in Ai of any generation between asthma and COPD.
Airway remodelling assessed by CT is more prominent in asthma compared with age- and sex-matched COPD subjects in the 3rd- to 6th generation airways when airflow limitations were similar under stable clinical conditions.

Full text

Page 1

Instructions for use
Title
Comparison of airway remodelling assessed by computed
tomography in asthma and COPD
Author(s)
Shimizu, Kaoruko; Hasegawa, Masaru; Makita, Hironi;
Nasuhara, Yasuyuki; Konno, Satoshi; Nishimura, Masaharu
Citation Respiratory Medicine, 105(9): 1275-1283
Issue Date2011-09
Doc URLhttp://hdl.handle.net/2115/47046
Right
Type article (author version)
Additional
Information
Hokkaido University Collection of Scholarly and Academic Papers : HUSCAP

Page 2

Text (Introduction through Discussion): 3055 words
Abstract: 246 words

Comparison of Airway Remodelling Assessed by Computed Tomography
in Asthma and COPD



Kaoruko Shimizu M.D., Masaru Hasegawa PhD, Hironi Makita PhD, Yasuyuki Nasuhara
PhD, Satoshi Konno PhD, and Masaharu Nishimura, PhD

First Department of Medicine, Hokkaido University School of Medicine, Sapporo, Japan

Correspondence to: Masaharu Nishimura, M.D.
First Department of Medicine, Hokkaido University School of Medicine
N-15 W-7, Kita-ku, Sapporo 060-8638, Japan
Tel: +81-11-706-5911; Fax: +81-11-706-7899
E-mail: ma-nishi@med.hokudai.ac.jp

Running title: Airway Remodelling assessed by CT in Asthma and COPD

Key words
Airway remodelling, Asthma, COPD, three-dimensional computed tomography

Page 3

ABSTRACT (246/250words)
Background: Few studies have directly compared airway remodelling assessed by computed
tomography (CT) between asthma and chronic obstructive pulmonary disease (COPD). The
present study was conducted to determine whether there are any differences between the two
diseases with similar levels of airflow limitation under clinically stable conditions.
Methods: Subjects included older male asthmatic patients (n=19) showing FEV1/FVC <70%
with smoking history less than 5-pack/year. Age- and sex-matched COPD patients (n=28)
who demonstrated similar airflow limitation as asthmatic patients and age-matched healthy
non-smokers (n=13) were recruited. Using proprietary software, eight airways were selected
in the right lung, and wall area percent (WA%) and airway luminal area (Ai) were measured
at the mid-portion of the 3rd to 6th generation of each airway. For comparison, the average of
eight measurements per generation was recorded.
Results: FEV1 % predicted and FEV1/FVC was similar between asthma and COPD
(82.3±3.3% vs. 77.6±1.8% and 57.7±1.6% vs. 57.9±1.4%). At any generation, WA% was
larger and Ai was smaller in asthma, both followed by COPD and then controls. Significant
differences were observed between asthma and controls in WA% of the 3rd to 5th generation
and Ai of any generation airway, while no differences were seen between COPD and controls.
There were significant differences in Ai of any generation between asthma and COPD.
Conclusions: Airway remodelling assessed by CT is more prominent in asthma compared
with age- and sex-matched COPD subjects in the 3rd- to 6th-generation airways when airflow
limitations were similar under stable clinical conditions.

Page 4

Ai = inner luminal area
Ao= outer area of the bronchus
COPD = chronic obstructive pulmonary disease
CT = computed tomography
Di= inner diameter
DLCO= carbon monoxide diffusing capacity
FEV1= forced expiratory volume in 1 s
FRC= functional residual capacity
FVC= forced vital capacity
ICS = inhaled corticosteroid
IgE= Immunoglobulin E
LAV= low attenuation volume
MMF= maximum mid-expiratory flow rate
RV=residual volume
SEM=standard error of measurement
TLC= total lung capacity
VA= alveolar volume
VC= vital capacity
WA = airway wall area
3D = three-dimensional

Page 5

INTRODUCTION
Bronchial asthma is characterized by reversible airflow limitation and airway
hyper-responsiveness to constricting stimuli. Some asthmatic patients have irreversible
airflow limitations despite treatment, possibly caused by airway remodelling1-3. In contrast,
airflow limitation observed in chronic obstructive pulmonary disease (COPD) is by definition
not fully reversible. In both diseases, airway inflammation is present, although the
characteristics of the inflammation are different. Airway remodelling is also a common
feature of both diseases, but the characteristics differ in nature as well as in severity.
The inflammation and remodelling in these two diseases have been described with regards
to physiology4-9, pathology10-13 and biology14-16. Differences in airway inflammation have
been well characterized between bronchial asthma and COPD4,13.
However, few studies have directly compared airway remodelling assessed by computed
tomography (CT) in asthma and COPD17,18 despite the increasing use of this modality for the
assessment of airway dimensions in these diseases19,20. Some investigators have speculated
that airway wall area is increased without a decrease in luminal area in asthma, whereas
increased airway wall area is associated with a decrease in luminal area in COPD21,22.
However, this speculation is based on results of two independent studies in subjects with
different levels of airflow limitation.
In the present study, possible differences in airway dimensions between the two diseases
were investigated in clinically stable patients with similar levels of airflow limitation.
Specifically, we assessed proximal airway remodelling by CT and compared between subjects
with asthma and COPD. We hypothesized that airway remodelling estimated by CT scans
would be more prominent in bronchial asthma, because the airways we could measure were
located in the proximal, but not in the distal, by definition. Airflow limitation in bronchial
asthma is characterized by airway remodelling in the proximal airways while that of COPD is

Page 6

characterized by mixture of emphysema and airway remodelling in the small airways.
Proprietary computer software enabling precise analysis of the short-axis image of the
airways perpendicular to the long axis at the 3rd-to 6th-generation airways was used23,24.

Page 7

METHODS
Subjects
Subjects were recruited from the outpatient clinic of Hokkaido University Hospital from
October 2006 through June 2009. First, asthmatic male patients who were older than 55 years
and had poorly reversible airflow limitation of FEV1/ FVC <70% despite appropriate drug
therapy were recruited. Diagnosis of bronchial asthma was made based on the definition of
American Thoracic Society, “Asthma is a chronic inflammatory disorder of the airways. The
chronic inflammation causes an associated increase in airway hyperresponsiveness that leads
to recurrent episodes of wheezing, breathlessness, chest tightness, and coughing, particularly
at night or in the early morning. These episodes are usually associated with widespread but
variable airflow obstruction that in often reversible either spontaneously or with treatment”.25.
Additional entry criteria included: i) clinically stable (no asthma attacks or major changes in
medication for more than 6 weeks before study entry); ii) life-long non-smokers or smokers
with a lifetime smoking history less than 5-pack/year; and iii) no apparent emphysema on CT
by visual assessment. Male patients with COPD were selected from subjects who participated
in the Hokkaido COPD cohort study26. Subjects were chosen based on age and values of
FEV1/FVC, FEV1% and all other data were blinded. Subjects in the COPD group who had
apparent giant bulla and/or bronchiectasis which might have anatomically affected bronchial
structure were excluded. Also recruited were age-matched male healthy volunteers as controls
with normal pulmonary function who had no history of respiratory diseases or respiratory
symptoms.
All the subjects underwent CT and pulmonary function tests sequentially on the same day.
Asthmatic patients had taken their regular medications on the examination day. Patients with
COPD had refrained from taking respiratory medications for 1-2 days, according to the
protocol of the Hokkaido COPD cohort study26.

Page 8

The study protocols were approved by the Clinical Research Ethics Committee of
Hokkaido University Hospital on 6th July 2006. Written informed consent was obtained from
all the subjects.

CT Data Acquisition and Analysis
CT was performed using a multidetector-row spiral CT scanner with a 64 detector array
(Aquilion Multi, TSX-101A/6A; Toshiba Medical Systems, Gunma, Japan). Data were
acquired with the following parameters: 120 kVp, 300 mA, 64 detector  0.5-mm collimation,
slice thickness 0.5 mm, 0.5 s/rotation, helical pitch 41. While subjects were in the supine
position, holding their breath at deep inspiration, the entire lung was scanned. The data were
transferred to a workstation and then reconstructed into three-dimensional (3D) images (AZE
Ltd., Tokyo, Japan). The detailed process of CT data acquisition and reconstruction has been
described previously27.
First, a three-dimensional bronchial skeleton was automatically reconstructed using a certain
threshold level, determined on an individual basis (-950HU to -980 HU) to obtain airway
images as distal as possible. Any portions of lung parenchyma remaining with the skeleton
were manually removed to prevent analysis error. Finally, we obtained a bronchial skeleton
and were able to identify any bronchus in the source images of axial, sagittal, and coronal
slices. The selected bronchial pathway was automatically converted to a curved multiplanar
reconstruction image. As the automatically obtained bronchial skeleton often contained only
up to the 3rd- (segmental) to 5th-generation airway, depending on the quality of the images,
airways were selected and then extended to the 6th generation. Identification of the generation
of bronchi was made by careful inspection while simultaneously using the longitudinal and
short axis images and searching for any bifurcation in the entire circumference.
The bronchial long-axis image appeared as a straight pathway, and short-axis images from

Page 9

the 3rd to the 6th generation, were identified (Figure 1). For measuring airway dimensions, the
software used the full-width at half-maximum principle for defining the wall area. However,
as the automatically obtained outline of airway walls was often out of contour, manual
corrections were made as follows. Based on manual plotting at several points, the software
used cubic spline interpolation and built a new circle. Finally, values were obtained for the
inner luminal area (Ai) and the outer area of the bronchus (Ao) and wall area (WA) was
calculated as (Ao – Ai). Wall area percent (WA%) was defined as (Ao - Ai) / Ao  100. A total
of eight airways per subject were chosen for the measurements, one airway from each of the
following bronchi: the apical (B1), posterior (B2), and anterior (B3) of the upper lobe, the
lateral (B4) and medial (B5) of the middle lobe, and the anterior basal (B8), lateral basal (B9),
and posterior basal (B10) of the lower lobe in the right lung. The measurement site was
generally at the midpoint between the bifurcations. All these measurements were done by one
of the authors (K. S.) who was blinded to any of the subjects’ characteristics and pulmonary
function data.
Total lung volume and volume-based severity of emphysema were measured. In short,
whole lung containing airways (A) were extracted from the 3D image of the thorax, resulting
in deletion of the heart and major vessels in the lungs. Then the bronchial skeleton (B) was
extracted from the whole lung, resulting in the lung consisting of parenchyma without either
major vessels or proximal bronchial trees. Total lung volume was defined as (A) – (B).
Severity of emphysema was assessed as low attenuation volume (LAV), based on the
threshold value of -950 HU. LAV% was defined as lung low attenuation volume divided by
total lung volume. All CT measurements were done by one author (KS) who was blinded to
the subjects’ information.
Pulmonary Function Tests

Page 10

Spirometry, carbon monoxide diffusion capacity, and lung volume assessed by the helium
closed-circuit method (CHESTAC-33, CHEST M. I., Inc., Tokyo, Japan) were measured.
Spirometric measurements included forced vital capacity (FVC), forced expiratory volume in
1 s (FEV1), and maximal mid-expiratory flow (MMF). Lung volume measurements included
total lung capacity (TLC), functional residual capacity (FRC), and residual volume (RV). We
used the rolling seal type of spirometer, the CHESTAC-33 (CHEST MI, Inc., Tokyo, Japan).
The procedures and results of pulmonary function tests met the requirements of the
pulmonary function test guidelines of the Japanese Respiratory Society Guidelines28, which
are similar to those of the American Thoracic Society. The diffusing capacity of the lung for
carbon monoxide (DLco), based on the single-breath method, was also measured in all
subjects according to the pulmonary function test guidelines of the Japanese Respiratory
Society. DLco divided by alveolar volume (VA) was expressed as percentage of predicted
values according to the prediction equations of Burrows29. Lung volumes (total lung capacity
(TLC), functional residual capacity (FRC), and residual volume (RV)) were measured by the
helium closed-circuit method. Lung volumes were expressed as a percentage of predicted
values according to the prediction equations of Nishida30. Spirometry was repeated 30 min
after inhalation of salbutamol (200 μg for bronchial asthma, 400 μg for COPD, as determined
by the Hokkaido COPD cohort study). Reversibility was defined as (post-bronchodilator
FEV1 – pre-bronchodilator FEV1) / post-bronchodilator FEV1  100. There were not
significant differences between the asthma and COPD groups in any indices of
pre-bronchodilator FEV1, post-bronchodilator FEV1, reversibility.

Statistical Analysis
SPSS was used for all statistical analyses (SPSS, Tokyo, Japan). All data are expressed as
mean ± standard error of measurement (SEM) for comparison. For comparison of lung

Page 11

volume, diffusing capacity, and %LAV between the asthmatic patients and the COPD patients
the t-test was used, and analysis of variance was used for comparison of parameters between
the three groups (age, pulmonary function tests, and airway dimensions). A p value <0.05 was
considered statistically significant.

Page 12

RESULTS
Characteristics in the three groups are summarized in Table 1. Nineteen asthmatic patients
and twenty-eight patients with COPD and thirteen healthy controls fulfilled the criteria.
Classification of asthma was based on the Global Initiative for Asthma 200531. One subject
was classified as mild persistent, 11 subjects were classified as moderate persistent and 7
subjects were classified as severe persistent. All subjects except one were taking inhaled
corticosteroids. In asthmatic patients, 14 subjects were life-long non-smokers and 5 subjects
were ex-smokers. Among the 28 COPD patients, 17 were classified as mild and 11 were
classified as moderate, according to the GOLD stage32. Airflow limitation indices such as
FEV1% predicted (82.3±3.3 vs. 77.6±1.8) and FEV1/FVC (57.7±1.6 vs. 57.9±1.4) were
similar between the two disease groups. There were no significant differences either in the
spirometric data or the data of lung volumes. The results of the reversibility test were similar
in these two groups, because the asthmatic patients had taken regular medications and the
COPD patients had refrained from taking respiratory medications for 1-2 days on the
day(Table 2). DLco% predicted and DLco/VA% predicted were lower in the COPD group
compared to the asthma group. Similarly, the severity of emphysema assessed as %LAV was
significantly greater in COPD patients compared to asthmatic patients and controls (Fig. 1).
The representative airway images from two typical cases with similar airflow limitation are
shown (Fig. 2) With regard to airway dimensions, WA% was larger and Ai was smaller at any
generation of bronchi in the asthmatic group, both followed by the COPD group and then the
controls (Fig. 3, 4). Significant differences were observed between the asthma and control
groups in WA% of the 3rd to 5th generation (p<0.01) and Ai of any generation airway (p<0.01
for 4th, p<0.05 for 3rd, 5th, 6th ), while no differences were seen between the COPD and control
groups. There were significant differences in Ai of any generation between asthma and COPD
(p<0.01 for 4th, 5th, p<0.05 for 3rd, 6th ). The means of calculated calibers from Ai in the 3rd to

Page 13

the 6th generation airways were 5.08mm, 3.94, 3.15, 2.57 in COPD, and 4.56, 3.39, 2.67, 2.26
in bronchial asthma.There were no differences in total lung volume determined by CT among
the asthma, COPD, and control groups (4239.9±204.5, 4252.6±199.3, and 4223.8±251.7 cm3,
respectively).

Page 14

DISCUSSION
In the present study, changes in airway dimensions at the 3rd to 6th generation airways
assessed by 3D-CT were more prominent in older asthmatic patients compared with
age-matched male COPD patients with clinically stable disease and similar levels of airflow
limitation. At any generation, WA% was larger and Ai was smaller in the asthma group,
followed by the COPD group and then the controls. Significant differences were observed
only between the asthma and control groups in WA% of the 3rd to 5th generation airways and
Ai of any generation airway, but not in any variable between the COPD and control groups.
There were significant differences in Ai of any generation airway between the asthma and
COPD groups.
Despite similar levels of FEV1/FVC, FEV1%, and lung volumes (including RV/TLC), there
was a marked difference in DLco/VA between the asthma and COPD groups. In
addition, %LAV assessed by CT showed a marked difference between the asthma and COPD
groups. These data suggest that the presence of emphysema was significant and contributed to
airflow limitation in COPD. Moreover, considering previous reports13 on the airway
pathology of asthma and COPD, it is likely that, under similar airflow limitations, airway
inflammation and/or remodeling are more prominent in the proximal airways in asthma
compared with in COPD. In COPD, the small airways and emphysema are likely to be
significant contributing factors to airflow limitation. Thus, the present study clearly
demonstrated that when airflow limitation is mild to moderate, 3D-CT has the capability to
demonstrate differences in airway dimensions between asthma and COPD patients.
In the present study, subjects were selected so that patients with asthma and COPD were
matched for airflow limitation parameters. To the best of our knowledge, only two previous
studies have attempted to compare airway dimensions between asthma and COPD using CT
data. In one study17, subjects were not matched for age or sex, which could have caused

Page 15

significant biases32-35. Furthermore, the asthma group included a substantial number of
smokers. In the other study18, all parameters of airway dimensions in HRCT analysis,
including airway wall area, airway wall area %, inner luminal area, airway luminal diameter
and wall thickness, were reported to be similar in the subjects with the two diseases and
similar airflow limitations. However, only the airways in which the ratio of the diameter of
the long axis to that of the short axis was less than 1.2 were measured, thus allowing inclusion
of some airways cut obliquely on the CT slice. Additionally, random selection of airways
might have led to comparison of airways of different generations and different sizes. Another
explanation might be that the degree of airflow limitations in bronchial asthma was milder
compared with that of COPD although not statistically significant, in other words, the two
groups were not so exactly matched for airflow limitation indices as in our study, which might
have obscured the difference. In the present study, lung volume when the CT was taken was
considered as a possible confounding factor because it might be vitally important in
measuring airway dimensions. Lung volume assessed by CT volumetric data demonstrated no
significant differences between asthma and COPD, which again strengthened the results of the
present study.
We have previously reported the relationships of airway dimensions, WA% and Ai, with
airflow limitation indices such as FEV1, % predicted in older patients with clinically-stable
asthma36 and also patients with COPD23, in both of which the subjects displayed variable
levels of airflow limitation. We found significant correlations of airflow limitation indices
with airway dimensions in the 3rd to 6th generations with similar correlation coefficients in
patients with bronchial asthma, and on the other hand, in patients with COPD, the correlation
coefficients that we found similarly significant improved better as the airways became smaller
in size from the 3rd to 6th generations. These study prompted us to examine the two diseases
with similar levels of airflow limitation under stable conditions and we hypothesized that