Intrathoracic tracheal volume and collapsibility on inspiratory and end-expiratory ct scans correlations with lung volume and pulmonary function in 85 smokers.

Abstract

To evaluate the correlations of tracheal volume and collapsibility on inspiratory and end-expiratory computed tomography (CT) with lung volume and with lung function in smokers.
The institutional review board approved this study at each institution. 85 smokers (mean age 68, range 45-87 years; 40 females and 45 males) underwent pulmonary function tests and chest CT at full inspiration and end-expiration. On both scans, intrathoracic tracheal volume and lung volume were measured. Collapsibility of the trachea and the lung was expressed as expiratory/inspiratory (E/I) ratios of these volumes. Correlations of the tracheal measurements with the lung measurements and with lung function were evaluated by the linear regression analysis.
Tracheal volume showed moderate or strong, positive correlations with lung volume on both inspiratory (r = 0.661, P < .0001) and end-expiratory (r = 0.749, P < .0001) scans. The E/I ratio of tracheal volume showed a strong, positive correlation with the E/I ratio of lung volume (r = 0.711, P < .0001). A weak, negative correlation was found between the E/I ratio of tracheal volume and the ratio of forced expiratory volume in the first second to forced vital capacity (r = -0.436, P < .0001). Also, a weak, positive correlation was observed between the E/I ratio of tracheal volume and the ratio of residual volume to total lung capacity (r = 0.253, P = .02).
Tracheal volume and collapsibility, measured by inspiratory and end-expiratory CT scans, is related to lung volume and collapsibility. The highly collapsed trachea on end-expiratory CT does not indicate more severe airflow limitation or air-trapping in smokers.

Full text

Page 1

Intrathoracic Tracheal Volume
and Collapsibility on Inspiratory and
End-expiratory CT Scans:
Correlations with Lung Volume and Pulmonary Function in 85 Smokers
Tsuneo Yamashiro, MD, Ra? ul San Jos? e Est? epar, PhD, Shin Matsuoka, MD, PhD,
Brian J. Bartholmai, MD, James C. Ross, MS, Alejandro Diaz, MD, Sadayuki Murayama, MD, PhD,
Edwin K. Silverman, MD, PhD, Hiroto Hatabu, MD, PhD, George R. Washko, MD
Rationale and Objectives: To evaluate the correlations of tracheal volume and collapsibility on inspiratory and end-expiratory computed
tomography (CT) with lung volume and with lung function in smokers.
Materials and Methods: The institutional review board approved this study at each institution. 85 smokers (mean age 68, range 45–87
years; 40 females and 45 males) underwent pulmonary function tests and chest CT at full inspiration and end-expiration. On both scans,
intrathoracic tracheal volume and lung volume were measured. Collapsibility of the trachea and the lung was expressed as expiratory/
inspiratory (E/I) ratios of these volumes. Correlations of the tracheal measurements with the lung measurements and with lung function
were evaluated by the linear regression analysis.
Results: Trachealvolumeshowedmoderateor strong, positivecorrelationswithlung volumeonbothinspiratory (r = 0.661,P < .0001)and
end-expiratory (r = 0.749, P < .0001) scans. The E/I ratio of tracheal volume showed a strong, positive correlation with the E/I ratio of lung
volume (r = 0.711, P < .0001). A weak, negative correlation was found between the E/I ratio of tracheal volume and the ratio of forced
expiratory volume in the first second to forced vital capacity (r = ?0.436, P < .0001). Also, a weak, positive correlation was observed
between the E/I ratio of tracheal volume and the ratio of residual volume to total lung capacity (r = 0.253, P = .02).
Conclusions: Tracheal volume and collapsibility, measured by inspiratory and end-expiratory CT scans, is related to lung volume and
collapsibility.Thehighlycollapsedtracheaonend-expiratoryCTdoesnotindicatemoresevereairflowlimitationorair-trappinginsmokers.
Key Words: Tracheal volume; lung volume; chronic obstructive pulmonary disease; tracheomalacia.
ªAUR, 2011
T
he meaning behind the highly collapsed trachea,
observed on expiratory scans of computed tomog-
raphy (CT), is still controversial. Although it has
been acknowledged that the central airways markedly
decrease in size on dynamic or end-expiratory CT scans in
patients with tracheomalacia (TM) or tracheobronchomalacia
(TBM) (1–8), some studies have demonstrated that the highly
collapsed trachea can also be observed on these expiratory
scans in subjects with normal lung function (9,10). In
patients
withchronicobstructive
(COPD), published information on the size of the trachea is
morelimited.Although
demonstrated the presence of the highly collapsed central
airways in COPD using dynamic expiratory or end-
expiratory scans (11–14), it still remains unclear whether or
not the highly collapsed trachea indicates reduced lung
function.
In 2003, Ederle and colleagues demonstrated that, in both
normal subjects and subjects with obstructive pulmonary
disease, cross-sectional area of the trachea correlated with
mean lung density (MLD) and cross-sectional area of the lung
on both inspiratory and end-expiratory CT scans (11). They
also showed that the changes in tracheal cross-sectional area
pulmonarydisease
someinvestigations have
Acad Radiol 2011; 18:299–305
From the Center for Pulmonary Functional Imaging (T.Y., H.H.), Surgical
Planning Laboratory (R.S.J.E., J.C.R.), Pulmonary and Critical Care Division
(A.D., E.K.S., G.R.W.), and Channing Laboratory (E.K.S.), Brigham and
Women’s Hospital, 75 Francis Street, Boston, MA 02115; Department of
Radiology, University of the Ryukyus, 207 Uehara, Nishihara-cho, Okinawa
903-0215, Japan (T.Y., S.M.); Department of Radiology, St. Marianna
UniversitySchool ofMedicine,Kawasaki, Kanagawa,
Department of Radiology, Mayo Clinic, Rochester, MN (B.J.B.). Received
September 22, 2010; accepted November 3, 2010. This study used data
provided by the Lung Tissue Research Consortium (LTRC) supported by the
National Heart, Lung, and Blood Institute (NHLBI). Supported by NIH
K23HL089353-01A1 and a grant from the Parker B. Francis Foundation.
E.K.S. received an honorarium for a talk on COPD genetics in 2006 and
grant support and consulting fees from GlaxoSmithKline for two studies of
COPD genetics; he also received honoraria from Bayer in 2005 and from
AstraZeneca in 2007 and 2008. Address correspondence to: T.Y. e-mail:
clatsune@yahoo.co.jp
ªAUR, 2011
doi:10.1016/j.acra.2010.11.005
Japan (S.M.);
299

Page 2

betweeninspiratoryandexpiratoryscanssignificantlycorrelated
to the changes in MLD. Further, more recent studies have
demonstrated significant correlations between MLD and lung
volume (LV) (15,16). Based on these observations, it can be
predicted that the observations of Ederle and colleagues
would be reproduced using volumetric measurements, such as
LVand tracheal volume.
We thereforehypothesized thattrachealvolume andchanges
in tracheal volume on inspiratoryand end-expiratory CT scans
would be correlated to LVand changes in LVamong smokers.
Although CT-based volumetric analysis of the whole lung has
been gradually performed in COPD (15–19), published
information on tracheal volumetric measurements is still
limited (20). With the development of imaging technology, it
has become easier to advance the methodology for measuring
tracheal size from dimensional to volumetric indices. Further-
more,itwouldbeofinteresttoevaluatewhetheror nottracheal
volume and collapsibility is correlated with lung function, and
whether or not volumetric measurements of the trachea are
equivalent to dimensional measurements.
Thus, the aims of this studyare: 1) toverify the correlations
between tracheal volumetric measures, including tracheal
collapsibility, and LV measures on inspiratory and end-
expiratory CT scans; 2) to evaluate the relationship between
tracheal collapsibility and lung function; and 3) to confirm
the correlations between tracheal volumetric measures and
dimensional measures.
MATERIALS AND METHODS
This retrospective study was approved by the institutional
reviewboardateachinstitution.Informedconsentwasobtained
from all subjects. The subjects’ CTand clinical data used in this
studywerepartiallyanalyzedwithadifferentobjectiveforother
research (15).
Subjects
Between March 2005 and February 2008, 184 subjects were
enrolled in the Lung Tissue Research Consortium (LTRC),
a multicenter trial for pulmonary diseases, at four institutions,
including Mayo Clinic at Rochester, University of Michigan,
UniversityofPittsburgh,and UniversityofColorado. CT scans
and patients’ clinical information provided for this study were
collected in accordance to the protocols outlined by the
LTRC. Further information on the LTRC, including its
mission and inclusion criteria, is available on its website
(www.ltrcpublic.com). From the 184 subjects, 85, who under-
wentchestCTwithtwostandardizedprotocols,wereeventually
selected and analyzed in this study. Exclusion criteria for this
study included: 1) use of different CT scanners that were not
selectedforthisstudy(n=8),2)CTscanswithcontrastmedium
or different slice thickness (n = 31), 3) different tube current or
voltage(n=15),4)noavailableend-expiratoryCTscan(n=6),
5)nohistoryofcigarettesmoking(n=1),6)previoushistoryof
lung resection (n = 5), 7) severe artifacts on CT images (n = 4),
and 8) presence of pneumothorax, lobar atelectasis, large bullae
(>3 cm in diameter), severe fibrosis, or a mass (>3 cm in diam-
eter) (n = 29).
The characteristics of the 85 subjects are summarized in
Table 1. Forty females and 45 males were included (mean
age 68 years, age range 45–87 years). All subjects were current
or former smokers and their mean smoking index was 48.4
pack-years ? 32.9.
Pulmonary Function Tests
All 85 subjects performed prebronchodilator spirometry,
including forced expiratory volume in the first second
(FEV1) and forced vital capacity (FVC), according to Amer-
ican Thoracic Society standards, as described previously (21).
FEV1wasalsoexpressedasthepercentagesofpredictedvalues.
Residual volume (RV) and total lung capacity (TLC) were
measured by plethysmography (22). The results of pulmonary
function tests (PFT) are summarized in Table 1.
According to the Global Initiative for Chronic Obstructive
Lung Disease (GOLD) staging (23), 85 subjects were classified
as follows: smokers with normal lung function, n = 14;
GOLDstage1,n=14;stage2,n=38;stage3,n=11;andstage
4, n = 8.
Thin-section CT
All subjects were scanned at full inspiration and full expiration
(end-expiration), without receiving a contrast medium.
Before CT scanning, subjects were coached to hold their
breath at full inspiration and full expiration. Among the 85
subjects, 46 subjects were scanned with 16-detector CT
(LightSpeed 16 or LightSpeed Pro16, GE Medical Systems,
Milwaukee, WI) Images were obtained using 140 kV and
300 mA. Rotation time was 0.53 seconds and the matrix size
was 512 ? 512 pixels. Images were reconstructed with
a1.25-mm-slicethickness(with0.625mmoverlapping),using
the ‘‘Bone’’ algorithm. The other 39 subjects were scanned
with 64-detector CT (Sensation 64, Siemens Healthcare,
Munich,Germany).Tubevoltagewas140kVandtubecurrent
varied by automatic regulation (approximately 75–450 mA),
which was based on the slice location and the subject’s body
TABLE 1. Clinical Characteristics of the 85 Subjects
Mean ? SD
67.8 ? 8.0
48.4 ? 32.9
0.56 ? 0.14
60.8 ? 23.6
0.50 ? 0.10
6.45 ? 1.34
3.24 ? 0.97
(Range)
Age (y)
Smoking index (pack-years)
FEV1/FVC
FEV1(%predicted)
RV/TLC
TLC (L)
RV (L)
(45?87)
(2?180)
(0.17?0.83)
(14?126)
(0.32?0.73)
(4.10?10.20)
(1.50?7.00)
FEV1, forced expiratory volume in 1 second; FVC, forced vital
capacity; RV, residual volume; SD, standard deviation; TLC, total
lung capacity.
YAMASHIRO ET ALAcademic Radiology, Vol 18, No 3, March 2011
300

Page 3

habitus.Rotationtimewas0.5secondsandthematrixsizewas
512 ? 512 pixels. Images were reconstructed with a 1.0-mm-
slice thickness (with 0.5 mm overlapping), using the ‘‘B46f’’
algorithm.
Measurements of the Trachea
Thecross-sectionalareaofthetracheaandintrathoracictracheal
volume were measured by a thoracic radiologist (T.Y., with 9
years’ experience in interpreting thoracic CT) using open-
source software (Image J, National Institutes of Health,
Bethesda, MD; http://rsb.info.nih.gov.ij), without prior
knowledge of clinical information. On an axial CT image, the
software automatically segmented the lumen of the trachea
from the soft tissues in the mediastinum, using thresholds of
-1024 and -500 Hounsfield units, which was followed by the
measuring of the cross-sectional area of the trachea (Fig 1). To
measure the cross-sectional area of the trachea, two anatomic
levels were selected on inspiratory scans (1 cm above the aortic
arch and 1 cm above the carina), according to the most recent
investigation (9). Onend-expiratory scans,cross-sectional areas
ofthetracheawerealsomeasuredatthesamelevels,determined
byinterpretingallanatomicalobservations.CTimageswereset
with lung window to perform these measurements.
Intrathoracic tracheal volume was also measured using the
same software on both inspiratory and end-expiratory scans.
Inbrief,thefollowingprocesswasperformed.1)Theradiologist
selected all CT images that include the trachea, from the top of
the apex to the carina. 2) The software showed the images for
every two slices because the original scan had a 50% overlap.
3) Extratracheal air regions (the lungs, bronchi, and esophagus)
were automatically segmented by the software and manually
excluded on each image. 4) A relatively large measuring area
was set in the center of the chest, which included the trachea
throughout the selected images. 5) The software measured
cross-sectional areas of the trachea on the all selected images.
6)Thetotalcross-sectionalareawascalculated,andintrathoracic
tracheal volume was obtained by multiplying the total cross-
sectional area by slice thickness. It took approximately 20
minutes to complete the whole process per subject.
Collapsibility of the tracheal cross-sectional area or tracheal
volume was calculated by the expiratory/inspiratory (E/I)
ratios of these two indices.
Measurements of the Lung Volume
LV was full-automatically measured by different open-source
software (Airway Inspector, Brigham and Women’s Hospital,
Boston, MA; www.airwayinspector.org), as described previ-
ously (15). In brief, the software 1) segmented the lung
parenchyma (?1024 to ?500 Hounsfield units) from the chest
wall and the hilum, 2) created a density histogram of the lung
parenchyma, and 3) measured LV by summing all pixels
included in the histogram and multiplying them by slice thick-
ness.Ineachsubject,itwasconfirmedbyanoperator(T.Y.)that
thesoftwareproperlyexcludedthetracheaandthemainbronchi
from the measured lung field. The E/I ratios of LV were
calculated by the operator.
Reproducibility of Tracheal Volume Measurements
Intraobserver error was tested by having one observer (T.Y.)
measure tracheal volume twice on inspiratory scans in 20
subjects,whowererandomlyselectedfromatotalof85subjects.
Thesecondmeasurementwasperformed2monthsafterthefirst
session. To evaluate interobserver error, two observers (T.Y.,
A.D.) independently measured tracheal volume in the 20
subjects. Analysis of intra- and interobserver reproducibility
was conducted using the Bland-Altman analysis (24).
Statistical Analysis
All statistical analyses were performed using JMP 7.0 software
(SAS Institute, Cary, NC). Data were expressed as mean ?
standard deviation. The linear regression analysis was used to
estimate the relationships among measured CT indices and
between CT indices and PFT values. P values less than .05
were considered statistically significant.
RESULTS
Reproducibility of Tracheal Volume Measurements
The intra-and interobserver reproducibilityoftracheal volume
measurementsisshowninTable2.Themeandifferencedidnot
appreciablydeviatefromzero,andthelimitsof agreementwere
small (Fig 2).
Figure 1. A 69-year-old man with chronic
obstructive pulmonary disease (GOLD stage 2).
An example of the measurement for tracheal
cross-sectional area onhis inspiratorycomputed
tomographyscanisdemonstrated.Theboundary
ofthelowertracheaistracedautomaticallybythe
software (yellow line).Cross-sectional area of the
trachea is calculated as 350.6 mm2.
Academic Radiology, Vol 18, No 3, March 2011TRACHEAL COLLAPSIBILITY AND LUNG VOLUME ON CT
301

Page 4

Tracheal Measurements
Table 3 demonstrates the measurements of the tracheal cross-
sectional and tracheal volume. The mean cross-sectional area
of the upper-trachea decreased from 273.3 mm2? 63.7 at
inspiration to 236.4 mm2? 60.2 at end-expiration, and the
mean E/I ratio was 0.87 ? 0.12. Similar observations were
found for the lower trachea (273.1 mm2? 66.7 at inspiration,
234.4 mm2? 62.1 at end-expiration, and E/I ratio of 0.86 ?
0.11).Themeantrachealvolumealsodecreasedfrom24.6mL
? 6.8 at inspiration to 19.5 ? 6.1 mL at end-expiration, with
the mean E/I ratio of 0.79 ? 0.13. Strong, positive correla-
tions (P < .0001) were found between measurements of the
tracheal cross-sectional area and tracheal volume on both
inspiratory and expiratory scans (upper trachea, r = 0.855 at
inspiration, r = 0.879 at expiration; lower trachea, r = 0.816
at inspiration, r = 0.818 at expiration, respectively). The E/I
ratios of tracheal cross-sectional areas also showed strong,
positivecorrelations(P<.0001)withtheE/Iratiosoftracheal
volume (upper trachea, r = 0.841; lower trachea, r = 0.853,
respectively).
CT-based Lung Volume
CT-basedLVwas5.44L?1.35atinspirationand3.83L?1.17
atexpiration(Table3).Inspiratory LVshowedastrong, positive
correlation withTLC(r= 0.867, P <.0001)and expiratory LV
with RV (r = 0.815, P < .0001), respectively.
Correlations between Tracheal Volume and CT-based
Lung Volume
Table 4 shows correlations between tracheal volume measures
and LV measures. Moderate or strong, positive correlations
werefoundbetweeninspiratorytrachealvolumeandinspiratory
LV (r = 0.661, P < .0001), and between expiratory tracheal
volumeandexpiratoryLV(r=0.749,P<.0001,Fig3).Further,
theE/Iratiooftrachealvolumedemonstrated astrong,positive
correlation with the E/I ratio of LV
P < .0001, Fig 3).
(r = 0.711,
Correlations between Tracheal Volume and Lung
Function
Correlations of tracheal volume measurements and PFTresults
are shown in Table 4. A moderate, negative correlation
was found between end-expiratory tracheal volume and
FEV1/FVC (r = ?0.445, P < .0001). Also, the E/I ratio of
tracheal volume showed moderate or weak, negative correla-
tions with FEV1/FVC (r = ?0.436, P < .0001) and FEV1%
TABLE 2. Reproducibility of Tracheal Volume Measurements
Mean
Difference ? SD
Limit of
Agreement
Tracheal volume (mL)
Intraobserver error
Interobserver error
0.05 ? 0.13
0.04 ? 0.15
?0.20 to 0.30
?0.22 to 0.29
SD, standard deviation.
Figure 2. Intra- and inter-observer reproducibility for tracheal
volume measurements. Plots show intraobserver (a) and interob-
server (b) error for measurement of tracheal volume according to
the Bland-Altman analysis. The mean of the two measurements
and the difference between them are plotted. The mean difference
does not appreciably deviate from zero, and limits of agreement
are small.
TABLE 3. Measurements of the Trachea and Lung Volume on
Chest CT
InspirationExpirationE/I ratio
Cross-sectional
area (mm2)
Upper trachea
Lower trachea
Tracheal volume (mL)
Lung volume (l)
273.3 ? 63.7 236.4 ? 60.2 0.87 ? 0.12
273.1 ? 66.7 234.4 ? 62.1 0.86 ? 0.11
24.6 ? 6.8
5.44 ? 1.35
19.5 ? 6.1
3.83 ? 1.17 0.71 ? 0.14
0.79 ? 0.13
CT, computed tomography; E/I, expiratory/inspiratory.
YAMASHIRO ET AL Academic Radiology, Vol 18, No 3, March 2011
302

Page 5

predicted (r = ?0.333, P < .001), and a weak, positive correla-
tion with RV/TLC (r = 0.253, P = .02), suggesting that higher
collapsibilityofthetracheaindicateslesssevereairflowlimitation
or air-trapping in the 85 subjects.
DISCUSSION
In the current study, we first demonstrate tracheal volumetric
measurements and their correlations with LV measurements.
In addition to significant correlations between LVand tracheal
volume both on inspiratoryand end-expiratory scans, collaps-
ibilityof thetracheaisstrongly correlatedwithcollapsibilityof
the lung. The E/I ratio of the trachea negatively correlates
with FEV1/FVC and FEV1%predicted, suggesting that the
highly collapsed trachea on end-expiratory scans indicates
less severe airflow limitation in the subjects analyzed in the
study. We believe therefore that the highly collapsed trachea
on end-expiratory scans, which is included in the highly
collapsed thorax, is not a morbid finding and should be distin-
guished from the abnormally collapsed trachea in TM/TBM.
These observations would be of importance for future radio-
logical and physiological studies of tracheal collapse in COPD
or TM/TBM subjects, as well as for the accurate interpreta-
tion of tracheal collapse on clinical CT scans.
Althoughthecorrelationbetweentrachealcollapsibilityand
lung collapsibility has not been assessed, the relationships
between tracheal sizes, measured by CTor acoustic reflection
techniques,andplethysmographiclungvalues(TLC,RV)have
been investigated in early physiological studies using normal
subjects (25–28). Also, it has also been reported that, on both
inspiratoryand end-expiratory CT scans, cross-sectional areas
of the trachea are correlated with cross-sectional areas of the
lung in normal subjects and subjects with obstructive disease
(11).Thecurrentstudycanbeconsideredtoreplicateprevious
observations using CT volumetric indices, such as tracheal
volume and LV. Furthermore, this study first demonstrates
the correlation between tracheal collapsibility and lung
collapsibility, which can also be useful for future radiological
and physiological studies of the central airways.
Although tracheal collapse in COPD has been investigated
in some studies (11–14,29), it still remains ambiguous as to
whether it is directly correlated with lung function. To the
best of our knowledge, only a limited number of studies has
discussed the relationship between tracheal collapsibility and
PFT results (13,14), which leads to the conclusion that high
collapsibility of the trachea does not solely indicate
compromised lung function. It has also been reported that
no significant difference is found in the presence of airway
malacia among GOLD stages (13). Furthermore, similar to
an early study by Thiriet and colleagues (10), Boiselle and
colleagues have recently demonstrated that the highly
collapsed trachea (>50% in cross-sectional area) on dynamic-
expiratory CT scans, which is concordant with the current
diagnostic criteria of TM/TBM, is frequently observed in
subjects with normal lung function (9). These results suggest
that it still remains difficult to distinguish a severe tracheal
collapse in normal subjects from an abnormal collapse in
TM/TBM, and that it is useful to understand the mechanism
of tracheal collapsibility using some other factors outside the
trachea. The current study suggests that the highly collapsed
trachea, observed on end-expiratory CT scans, indicates less
severe airflow limitation or air-trapping in smokers. These
observations can be partially explained by the prediction that
LV would be maintained on end-expiratory scans in more
severe COPD patients, because of severe air-trapping or expi-
ratory airflowobstruction (15), resulting in smaller changes in
intrathoracic pressure and in the tracheal size. However, it has
alsobeenacknowledgedthatsmoking,severeemphysema,and
chronic bronchitis are contributors to the development of
TM/TBM (4). Further, it can be predicted that the trachea
can collapse more highly on dynamic expiratory scans
(1,2,4,5). Although published information is limited on the
frequency of TM/TBM in the general COPD population,
further research using dynamic expiratory scans is needed to
comprehend the prevalence and clinical impact of the highly
collapsed trachea in subjects with COPD.
There are some important limitations in the current study.
First, this study did not use dynamic expiratory CT scans. It
has already been acknowledged that end-expiratory CT scans
TABLE 4. Correlations of Tracheal Volume with CT-based Lung Volume and with Lung Function
CT-based Lung VolumeLung Function
Inspiration ExpirationE/I RatioFEV1/FVC
FEV1
(% predicted) RV/TLC
Tracheal volume
Inspiration
Expiration
E/I ratio
0.661*
0.677*
0.135
0.472*
0.749*
0.551*
?0.126
0.293z
0.711*
?0.221{
?0.445*
?0.436*
?0.110
?0.179
?0.333y
?0.182
?0.020
0.253{
CT,computedtomography;E/I,expiratory/inspiratory;FEV1,forcedexpiratoryvolumein1second;FVC,forcedvitalcapacity;RV/TLC,ratioof
residual volume to total lung capacity.
*P < .0001.
yP < .001.
zP < .01.
{P < .05.
Academic Radiology, Vol 18, No 3, March 2011TRACHEAL COLLAPSIBILITY AND LUNG VOLUME ON CT
303

Page 6

are less sensitive in evaluating tracheal collapse than dynamic
expiratory scans, particularly in TM/TBM (1,2,4,5). In fact,
compared with a study for COPD subjects using dynamic
expiratory scans (29), collapsibility of the trachea was much
smaller and might have been underestimated in this study.
However, based on the fact that end-expiratory scans have
been used more frequently and proved to be effective in
COPD analysis (15,30,31), we believe that it is still
meaningful to demonstrate tracheal collapsibility using end-
expiratory scans in general COPD subjects.
Second,inspiratoryandend-expiratoryscanswereobtained
without the spirometry-gated technique in our study, which
may lead to questioning of whether these scans were truly
obtained at subjects’ full inspiration and full expiration. In
this study, inspiratory LV was smaller than plethysmographic
TLC and expiratory LV larger than RV, which is similar to
previous studies (17,18,32,33). These disturbances between
CT-based LVand plethysmographic measures can be partially
explained by a lack of patient cooperation, the difference of
the body position, or the larger effort of breath-holding at
expiratory CT scanning than at PFT.
Third,trachealvolumemeasuredinthestudyrequiresfurther
discussion. We measured intrathoracic tracheal volume, which
wasmeasuredfromthetopoftheapextothecarina,suggesting
thattheapicalpointwouldhavebeeninfluencedbythelocation
and size of the lung in each subject. However, it is known that
the intrathoracic part of the trachea behaves differently from
the extrathoracic part of the trachea (20). Further, from the
perspective of programming automatic measurements, we
believe that this apical point is a reasonable and the easiest
anatomical location for the standardized method.
Fourth,thenumberof subjectsinthecurrentstudy wasrela-
tivelysmall.Thismaybeoneofthereasonsthatwerarelyfound
subjectswithseveretrachealcollapse,comparedtoalargestudy,
which enrolled more than 1000 smokers and found some cases
with the highly collapsed trachea using similar end-expiratory
CT scans (12).
Fifth, two different scanning protocols were used in this
study, which may lead to a questioning of our original CT
data and the reliability of measured parameters. Compared
with other CT indices, such as MLD or low attenuation
area percent, it is predictable that LV or tracheal volume are
less sensitive to the differences in scanning protocols, because
the upper threshold of ?500 Housfield units is the only point
to be influenced by the protocols. We believe therefore that
the difference in the protocols did not severely impact the
measurements, and that the major observations in the current
study were not amplified.
Insummary,wefirstmeasuredtrachealvolumeandanalyzed
volumetric measurements of the trachea with lung indices
using inspiratory and end-expiratory CT scans. Significant
correlations were found between tracheal volume and LV, as
well as correlations between collapsibility of the trachea and
collapsibility of the lung. Further, higher collapsibility of the
trachea did not indicate lower FEV1%predicted, lower
FEV1/FVC, or higher RV/TLC in this study.
Figure 3. Correlations between measurements of tracheal volume
andlung volume.Strongor moderate,positivecorrelations are found
between inspiratory tracheal volume and inspiratory lung volume (a,
r = 0.661, P < .0001), between expiratory tracheal volume and
expiratory lung volume (b, r = 0.749, P < .0001), and between the
expiratory/inspiratory (E/I) ratios of tracheal volume and lung volume
(c, r = 0.711, P < .0001).
YAMASHIRO ET ALAcademic Radiology, Vol 18, No 3, March 2011
304

Page 7

REFERENCES
1. Gilkeson RC, Ciancibello LM, Hejal RB, et al. Tracheobronchomalacia:
dynamic airway evaluation with multidetector CT. Am J Roentgenol
2001; 176:205–210.
2. Aquino SL, Shepard JA, Ginns LC, et al. Acquired tracheomalacia: detec-
tion by expiratory CT scan. J Comput Assist Tomogr 2001; 25:394–399.
3. Zhang J, Hasegawa I, Feller-Kopman D, et al. Dynamic expiratory volu-
metric CT imaging of the central airways: comparison of standard-dose
and low-dose techniques. Acad Radiol 2003; 10:719–724.
4. Carden KA, Boiselle PM, Waltz DA, et al. Tracheomalacia and tracheo-
bronchomalacia in children and adults: an in-depth review. Chest 2005;
127:984–1005.
5. Baroni RH, Feller-Kopman D, Nishino M, et al. Tracheobronchomalacia:
comparison between end-expiratory and dynamic expiratory CT for
evaluation of central airway collapse. Radiology 2005; 235:635–641.
6. Lee KS, Sun ME, Ernst A, et al. Comparison of dynamic expiratory CT with
bronchoscopy in diagnosing airway malacia. Chest 2007; 131:758–764.
7. Ferretti GR, Jankowski A, Perrin MA, et al. Multi-detector CT evaluation in
patients suspected of tracheobronchomalacia: comparison of end-
expiratory with dynamic expiratory volumetric acquisitions. Eur J Radiol
2008; 68:340–346.
8. Boiselle PM, Ernst A. Tracheal morphology in patients with tracheomala-
cia: prevalence of inspiratory ‘‘lunate’’ and expiratory ‘‘frown’’ shapes.
J Thorac Imaging 2006; 21:190–196.
9. Boiselle PM, O’Donnell CR, Bankier AA, et al. Tracheal collapsibility in
healthy volunteers during forced expiration: assessment with multidetec-
tor CT. Radiology 2009; 252:255–262.
10. Thiriet M, Maarek JM, Chartrand DA, et al. Transverse images of the
human thoracic trachea during forced expiration. J Appl Phsiol 1989; 67:
1032–1040.
11. EderleJR, Heussel CP,Hast J,et al. Evaluation ofchanges incentral airway
dimensions,lungareaandmeanlungdensityatpairedinspiratory/expiratory
high-resolution computed tomography. Eur Radiol 2003; 13:2454–2461.
12. Ochs RA, Petkovska I, Kim HJ, et al. Prevalence of tracheal collapse in an
emphysema cohort as measured with end-expiration CT. Acad Radiol
2009; 16:46–53.
13. Sverzellati N, Rastelli A, Chetta A, et al. Airway malacia in chronic obstruc-
tive pulmonary disease: prevalence, morphology and relationship with
emphysema, bronchiectasis and bronchial wall thickening. Eur Radiol
2009; 19:1669–1678.
14. Lee CJ, Lee JH, Song JW, et al. Correlation of tracheal cross-sectional
area with parameters of pulmonary function test in COPD. Tuberc Respir
Dis 1999; 46:628–635.
15. Yamashiro T, Matsuoka S, Bartholmai BJ, et al. Collapsibility of lung
volume by paired inspiratory and expiratory CT scans: correlations with
lung function and mean lung density. Acad Radiol 2010; 17:489–495.
16. Camiciottoli G, Cavigli E, Grassi L, et al. Prevalence and correlates of
pulmonary emphysema in smokers and former smokers. A densitometric
study of participants in the ITALUNG trial. Eur Radiol 2009; 19:58–66.
17. Kauczor HU, Heussel CP, Fischer B, et al. Assessment of lung volumes
using helical CT at inspiration and expiration: comparison with pulmonary
function tests. Am J Roentgenol 1998; 171:1091–1095.
18. Zaporozhan J, Ley S, Eberhardt R, et al. Paired inspiratory/expiratory
volumetric thin-section CT scan for emphysema analysis: comparison of
different quantitative evaluations and pulmonary function test. Chest
2005; 128:3212–3220.
19. Iwano S, Okada T, Satake H, et al. 3D-CT volumetry of the lung using
multidetector row CT: comparison with pulmonary function tests. Acad
Radiol 2009; 16:250–256.
20. Griscom NT, Wohl ME. Tracheal size and shape: effects of change in intra-
luminal pressure. Radiology 1983; 149:27–30.
21. Miller MR, Hankinson J, Brusasco V, et al. ATS/ERS task force: standard-
isation of spirometry. Eur Respir J 2005; 26:319–338.
22. Stocks J, Quanjer PH. Reference values for residual volume, functional
residual capacity and total lung capacity: ATS workshop on lung volume
measurements/official statement of the European Respiratory Society.
Eur Respir J 1995; 8:492–506.
23. Rabe KF, Hurd S, Anzueto A, et al. Global strategy for the diagnosis,
management, and prevention of chronic obstructive pulmonary disease:
GOLD executive summary. Am J Respir Crit Care Med 2007; 176:
532–555.
24. Bland JM, Altman DG. Statistical methods for assessing agreement
between two methods of clinical measurement. Lancet 1986; 1:307–310.
25. Hoffstein V. Relationship between lung volume, maximal expiratory flow,
forced expiratory volume in one second, and tracheal area in normal
men and women. Am Rev Respir Dis 1986; 134:956–961.
26. Hoffstein V, Castile RG, O’Donnell CR, et al. In vivo estimation of tracheal
distensibility and hysteresis in normal adults. J Appl Physiol 1987; 63:
2482–2489.
27. Martin TR, Castile RG, Fredberg JJ, et al. Airway size is related to sex but
not lung size in normal adults. J Appl Physiol 1987; 63:2042–2047.
28. Dolyniuk MV,FaheyPJ.Relationshipoftracheal sizetomaximalexpiratory
airflow and density dependence. J Appl Physiol 1986; 60:501–515.
29. Heussel CP, LeyS,Biedermann A,etal.Respiratorylumenal changeof the
pharynx and trachea in normal subjects and COPD patients: assessment
by cine-MRI. Eur Radiol 2004; 14:2188–2197.
30. Akira M, Toyokawa K, Inoue Y, et al. Quantitative CT in chronic obstructive
pulmonary disease: inspiratory and expiratory assessment. Am J Roent-
genol 2009; 192:267–272.
31. Matsuoka S, Kurihara Y, Yagihashi K, et al. Airway dimensions at
inspiratory and expiratory multisection CT in chronic obstructive
pulmonary disease: correlation with airflow limitation. Radiology 2008;
248:1042–1049.
32. O’Donnell CR, Loring SH. Comparison of plethysmographic, helium dilu-
tion and CT-derive total lung capacity. Am J Respir Crit Care Med 2005;
171. A293 (abstract).
33. Garfield JL, Marchetti N, Gaughan JP, et al. Lung volume by plethsmogra-
phy and CT in advanced COPD. Am J Respir Crit Care Med 2009; 179.
A2902 (abstract).
Academic Radiology, Vol 18, No 3, March 2011TRACHEAL COLLAPSIBILITY AND LUNG VOLUME ON CT
305