Dornase Alfa Reduces Air Trapping in

Children With Mild Cystic Fibrosis Lung

Disease*
A Quantitative Analysis
Terry E. Robinson, MD; Michael L. Goris, MD; Hongyun J. Zhu, MD;
Xiaorong Chen, MS; Prache Bhise, MD; Farzana Sheikh, MD; and
Richard B. Moss, MD, FCCP

Purpose: To evaluate quantitative air trapping measurements in children with mild cystic fibrosis
(CF) lung disease during a 1-year, double-blind, placebo-controlled, recombinant human deoxyri-
bonuclease (rhDNase) [dornase alfa] intervention trial and compare results from quantitative air
trapping with those from spirometry or visually scored high-resolution CT (HRCT) scans of the
chest.
Materials and methods: Twenty-five children with CF randomized to either daily rhDNase or
placebo aerosol were evaluated at baseline, and at 3 months and 12 months by spirometer-
triggered HRCT and spirometry. Outcome variables were percentage of predicted FVC, FEV1,
and forced expiratory flow, midexpiratory phase (FEF25–75%); total and subcomponent visual
HRCT scores; and quantitative air trapping measurements derived from chest HRCT images.
Results: At baseline, there were no statistical differences between groups in any of the variables
used as an outcome. After 3 months of treatment, both groups had improvements in percentage
of predicted FEV1 and FEF25–75%, and total HRCT visual scores. In contrast, the rhDNase group
had a 13% decrease in quantitative air trapping from baseline (severe air trapping [A3]),
compared to an increase of 48% in the placebo group (p ⴝ 0.023). After 12 months, both groups
had declines in percentage of predicted FVC and FEV1, but the rhDNase group retained
improvements in percentage of predicted FEF25–75% and quantitative air trapping. The mucus
plugging and total HRCT visual scores were also improved in the rhDNase group after 12 months
of treatment, with and without significant differences between groups (p ⴝ 0.026 and p ⴝ 0.676).
Quantitative air trapping (A3) remained improved in the rhDNase group (ⴚ 15.4%) and worsened
in the placebo group (ⴙ 61.3%) with nearly significant differences noted between groups
(p ⴝ 0.053) after 12 months of treatment.
Conclusions: Quantitative air trapping is a more consistent sensitive outcome measure than either
spirometry or total HRCT scores, and can discriminate differences in treatment effects in
children with minimal CF lung disease. (CHEST 2005; 128:2327–2335)

Key words: cystic fibrosis; dornase alfa; high-resolution CT; mucus plugging; quantitative air trapping

Abbreviations: A1 ⫽ mild, moderate, and severe air trapping; A2 ⫽ moderate and severe air trapping; A3 ⫽ severe air
trapping; CF ⫽ cystic fibrosis; FEF25–75% ⫽ forced expiratory flow, midexpiratory phase; HRCT ⫽ high-resolution CT;
HU ⫽ Hounsfield unit; PFT ⫽ pulmonary function test; rhDNase ⫽ recombinant human deoxyribonuclease (dornase
alfa)

T hein cystic
predominant cause of morbidity and mortality
fibrosis (CF) lung disease is progressive
obstructive lung disease resulting from reduced mu-
cociliary clearance, airway obstruction, recurrent en-
dobronchial infections, and persistent inflammation
and destruction of the airways. Early manifestations
of CF lung disease include submucosal gland hyper-
trophy, inspissated airway mucus, inflammatory in-
filtrates, and bronchiectasis.1,2 High-resolution CT
(HRCT) in infants and children with early and/or
mild CF lung disease has demonstrated evidence of
peripheral regional air trapping, bronchial/bronchio-
lar wall thickening, and bronchial/bronchiolar airway
dilatation.3–14 Using spirometer-triggered HRCT im-
aging and an automated approach for quantifying air
trapping defects, we previously demonstrated in-
creased quantitative air trapping in children with
mild CF lung disease compared to age-matched
normal children.9 Unlike percentage of predicted
FEV1 and forced expiratory flow, midexpiratory
phase (FEF25–75%), which did not discriminate sta-
tistical differences between these groups, the quan-
titative air trapping measures clearly showed robust
differences between groups with p values ⬍ 0.001.

www.chestjournal.org CHEST / 128 / 4 / OCTOBER, 2005 2327

Downloaded From: http://journal.publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/22032/ on 06/28/2017

 This suggests that quantitative air trapping obtained
by HRCT imaging is a more sensitive outcome
measure than pulmonary function measurements in
discriminating early obstructive lung disease.9 The
sensitivity of HRCT measures being more sensitive
than PFT measures is further supported by de Jong
et al10 and Brody et al,11 who have recently demon-
strated that visual HRCT scores, including HRCT air
trapping scores, were more sensitive than pulmonary
function tests (PFTs) in detecting early and progres-
sive CF lung disease. The purpose of the present
study was to evaluate quantitative air trapping, spi-
rometric measurements, and chest HRCT scores as
outcome measures in therapeutic response to recom-
binant human deoxyribonuclease (rhDNase) [dor-
nase alfa] in 25 children with mild CF lung disease
over a 1-year, controlled intervention trial.
Materials and Methods
Study Design
We conducted a randomized, double-blind, placebo-con-
trolled, 1-year trial of dornase alfa in 25 children and adolescents
with mild CF lung disease. All children had a confirmed diagnosis
of CF by pilocarpine iontophoresis sweat chloride test and/or CF
gene mutation analysis. Inclusion criteria included routine med-
ical care in a CF clinic, age 6 to 18 years, percentage of predicted
FVC ⱖ 85% and a percentage of predicted FEV1 approximately
ⱖ 70%, and the ability to perform reproducible PFTs. Exclusion
criteria were inability to perform reproducible upright and supine
spirometry; inability to take the trial medication; acute asthma
attack; recent lower respiratory tract infection prior to enrollment
requiring a change in antibiotic, bronchodilator, or antiinflam-
matory therapy (inhaled or oral steroids; ibuprofen); or use of
dornase alfa within a previous 3-week period prior to enrollment
and testing. The patient cohort was described and evaluated in a
previous article12 with a different visual scoring system that
emphasized an outcome equation combining elements of HRCT
scores and PFT measures. Before enrollment into the study,
informed assent and consent were obtained from the patients and
*From the Departments of Pediatrics (Pulmonary) [Drs. Robin-
son, Bhise, Sheikh, and Moss], and Radiology (Nuclear Medi-
cine) [Drs. Goris and Zhu], Stanford University Medical Center,
Palo Alto; and Department of Statistics (Ms. Chen), Stanford
University, Stanford, CA.
Dr. Robinson received a research grant from Genentech, Inc. of
$3,700.00 representing the cost of pharmacist set-up and inven-
tory activities associated with Pulmozyme and placebo medica-
tion for this study.
Dr. Moss has received research grants from Genentech, Inc. for
studies related to Pulmozyme since 1993.
This study was funded by the Cystic Fibrosis Foundation and
Genentech, Inc.
Manuscript received November 26, 2004; revision accepted April
29, 2005.
Reproduction of this article is prohibited without written permission
from the American College of Chest Physicians (www.chestjournal.
org/misc/reprints.shtml).
Correspondence to: Terry E. Robinson, MD, Pediatric Pulmonary
Division, Stanford University Medical Center, 701 Welch Rd,
Whelan Building, #3328, Palo Alto, CA 94305-5786; e-mail:
ter@stanford.edu
2328
Downloaded From: http://journal.publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/22032/ on 06/28/2017
their parents respectively. The study protocol was approved by
the Stanford University Administrative Panel in Human Subjects.
Patients were studied at randomization, and at approximately 3
months and 12 months afterwards.
Eligibility was assessed at the initial pretreatment visit. Sub-
jects already receiving dornase alfa prior to the study discontin-
ued their treatment 3 weeks before screening and enrollment. At
the second visit, baseline testing was completed for all groups,
and CF patients were randomized to receive either 2.5 mg
dornase alfa or placebo aerosol once daily with a jet nebulizer
(Pari LC Plus; Pari; Richmond, VA) and compressor (Pulmo-
Aide; DeVilbiss; Somerset, PA; or Pari Pro Neb; Pari). Random-
ized treatment assignment was conducted by the Division of
Biostatistics, Department of Health Research and Policy, of
Stanford University in conjunction with the Lucile Packard
Children’s Hospital pharmacy department. All patients, investi-
gators, and study participants were blinded to the treatment
assignment until the study was completed.
The outcome variables for the study included spirometry
(FVC, FEV1, FEF25–75%), global and subcomponent HRCT
scores,15 and quantitative expiratory air trapping measurements
(mild, moderate, and severe air trapping [A1]; moderate and
severe air trapping [A2]; severe air trapping [A3]). During each
testing session, a brief medical history and physical examination
were performed, height and weight were measured, and spirom-
etry and spirometer-triggered inspiratory and expiratory CT
imaging were performed. Further details of the testing equip-
ment and protocols utilized have been described.12 Pulmonary
function measurements were expressed as percentages of pre-
dicted based on normal prediction equations derived from the
data of the Harvard Six Cities Study.16 Chest HRCT images were
obtained using a previously described protocol.12 Contiguous
1.5-mm images were obtained at ⱖ 95% slow vital capacity. To
allow for consistent matching of images from serial studies,
inspiratory scans were acquired volumetrically during a single
breath-hold from 1.5 cm above the aortic arch to 1 cm above the
right hemidiaphragm. Six thin-slice expiratory (1.5 mm per slice)
images that were equally spaced between 1.0 cm above the aortic
arch to 1 cm above the right hemidiaphragm were obtained at
near residual volume to evaluate the extent of air trapping. The
calculated total radiation exposure for the inspiratory and expi-
ratory CT scans for each testing session was 75 to 135 millirem
for children 6 to 18 years old, which is below the estimated
average annual radiation exposure in communities at high eleva-
tions such as Denver, CO, and is also below the average radon
exposure per year.17,18
HRCT Scoring and HRCT Image Analysis
Total (global) and component HRCT scores were determined
for each CT scan utilizing a scoring system similar to that of
Brody et al,6 with different scoring components and different
rating criteria.15 A total (global) score was calculated as the sum
of the seven component scores of extent and severity of bronchi-
ectasis, extent and severity of bronchial wall thickening, and
extent of mucus plugging, atelectasis/consolidation, and air trap-
ping. Each component feature was scored individually from 1 to
4 in each lobe (with the lingula considered as a separate lobe) at
each of the six image levels. A score of 4 for each component
would yield a total possible score of 168. The HRCT scans were
independently assessed by three radiologists, and the average of
the three readers was utilized for analysis. To evaluate regional
air trapping, chest HRCT images were postprocessed utilizing an
automated approach for lung segmentation and subsequent air
trapping defect determination as described.9 In short, air trap-
ping was based on the density distribution of individual voxel
densities (in Hounsfield units [HUs]) within the segmented lung
Clinical Investigations
 in the expiratory and inspiratory images, with a range defined by
the median HUs in the inspiratory vs expiratory images. Within
this defined range of HU densities, three thresholds were set.
Voxels with densities below the given thresholds were considered
to have air trapping. The threshold with the lowest density (in
HUs) defines those regions of segmented lung with A3, while the
threshold with the highest density corresponds to those regions of
segmented lung that include A1. An additional threshold was also
arbitrarily chosen midway between the lowest and highest density
threshold to represent those regions of segmented lung on CT
that represent A2. Global air trapping was therefore expressed as
a fraction of the number of involved voxels in the individual and
summated expiratory slices that corresponded to A1, A2, and A3
(Fig 1).9
Statistical Analysis
Group averages were compared at baseline using two-sample t
tests adjusted for unequal variances as well as the Fisher Exact
Test. Over the active treatment period, changes in air trapping,
spirometric measurements, and HRCT scores including the
HRCT score for the extent of air trapping were evaluated using
paired t tests for within-group comparisons and unpaired t tests
for between-group differences for percentage change from base-
line to 3 months: ([3-month value – baseline value]/baseline
value ⫻ 100) and baseline to 12 months ([12-month value –
baseline value]/baseline value ⫻ 100). Data met normality as-
sumption required for t tests. Results were considered statisti-
cally significant for p ⬍ 0.05.
Results
Twenty-five CF subjects were enrolled and ran-
domized, and 21 subjects completed the 1-year trial
with follow-up testing. All four noncompleters with-
drew for nonstudy drug-related reasons. For the

Figure 1. Example of the processing of slice No. 5 in a typical CF case. In the first column of the first
row (top), the inspiratory HRCT slice is shown next to the segmented lung at the same level. The
negative values are the median 90th percentile in HUs, a measure of CT density. In row 2 (center), the
slice image at near residual volume is in the first column, next to the segmented expiratory lung. In row
3 (bottom), the segmented expiratory lung is seen with three defects—A1, A2, and A3—in overlay in
white. The positive numbers are the percentage of voxels included in the defects. The negative values
are the HUs of the thresholds. The third image in the first row (top right) is the combined inspiratory
and expiratory histograms at that level. In row 2 (center, right), a composite of the expiratory slice with
the three defects (A1, A2, and A3) are presented in shades of gray.

www.chestjournal.org CHEST / 128 / 4 / OCTOBER, 2005 2329

Downloaded From: http://journal.publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/22032/ on 06/28/2017

 quantitative air trapping analysis, 25 subjects could
be evaluated at baseline and 3 months; however, only
19 subjects (3 fewer subjects from each group) could
be evaluated at baseline and 12 months due to
noncompleters and the unavailability of stored elec-
tronic imaging for postprocessed CT image analysis
in two subjects for which the expiratory scan digital
data were not saved. Adherence (average proportion
of medication taken based on returned used and
unused vials and diary sheets) over 12 months was
86.9% for the placebo group and 85.6% for the
dornase alfa group.
Baseline Assessment
The baseline characteristics for study subjects are
presented in Table 1. Between-reader reliability for
total HRCT scores and air trapping component at
test 1 for the three radiologists who scored the
HRCT scans were 0.83 and 0.62, respectively. This
interreader reliability for the total HRCT score was
comparable to the interreader reliability for two
readers (0.78) in our previous study15 in CF patients
with more severe disease using the same HRCT
scoring system. Although there were no statistically
significant differences in mean age, weight, height,
pulmonary function measurements, HRCT scores,
and quantitative air trapping indices between the
groups at baseline testing, the dornase alfa group
were older and had greater quantitative air trapping
values. Only 1 of 25 subjects (randomized to the
dornase alfa group) received rhDNase aerosol prior
to the study, but this subject did not receive rhD-
Nase for 3 weeks prior to enrollment in the study.
On morphologic analyses of segmented expiratory
chest HRCT scans in the majority of subjects in both
groups, there was a heterogeneous distribution of air
trapping defects as demonstrated in Figure 1. Uti-
lizing previous data defining normal standards (mean
⫹ 2 SD) for no air trapping in age-matched normal
subjects,9 73%, 64%, and 55% of the dornase alfa
subjects had air trapping at baseline, compared to
57%, 50%, and 43% of placebo subjects for A1, A2,
and A3 thresholds, respectively. Despite the higher
incidence of air trapping in the dornase alfa group,
there were no statistical differences between groups
for the number of subjects with air trapping at
baseline utilizing the two-tailed Fisher Exact Test
(p ⫽ 0.691, p ⫽ 0.689, and p ⫽ 0.677 for A1, A2, and
A3 thresholds, respectively).
Treatment Results
After 3 months of treatment, both groups had
improvements in percentage of predicted FEV1 and
FEF25–75% as well as total visual HRCT and mucus
plugging scores. The dornase alfa group, however,
had an increase (ie, worsening) in the visual air
trapping CT score, whereas the placebo group had a
decline (ie, improvement) in the visual air trapping

Table 1—Baseline Characteristics of the Study Population (n ⴝ 25)*

Characteristics Dornase Alfa (n ⫽ 11) Placebo (n ⫽ 14)

Demographics and anthropometrics

Age, yr 11.8 (3.7; 6.4 to 17.9) 9.9 (3.4; 6.6 to 17.3)
Male/female gender 7 (64)/4 (36) 8 (57)/6 (43)
Weight, kg 42.0 (18.2; 21.5 to 74.6) 32.4 (15.0; 18.6 to 66.8)
Height, cm 143.9 (19.4; 116 to 176) 134.6 (21.2; 110 to 168)
Genotyping
Homozygous ⌬F508 7 (64) 9 (64)
Compound heterozygous ⌬F508 2 (18) 5 (36)
Compound heterozygous S492F 1 (9)
Unknown/unknown 1 (9)
Prior dornase alfa use 1 (9)
Spirometry
FVC, % predicted 114 (20; 87 to 143) 114 (19; 90 to 149)
FEV1, % predicted 105 (20; 79 to 136) 104 (17; 69 to 128)
FEF25–75%, % predicted 91 (34; 40 to 151) 87 (21; 35 to 117)
Chest HRCT scores
Total HRCT score 67 (13.4; 47 to 89) 68 (16.7; 42 to 93)
Bronchiectasis extent score 9 (2.2; 7 to 13) 9 (2.8; 6 to 13)
Bronchial wall thickness extent score 11 (3.2; 7 to 16) 11 (3.6; 6 to 16)
Mucus plugging score 8 (1.9; 6 to 12) 7 (2.2; 6 to 13)
Quantitative air trapping
A1 32 (20.5; 7 to 65) 25 (19.8; 2 to 68)
A2 20 (14.9; 3 to 46) 15 (13.8; 1 to 45)
A3 9 (7.4; 1 to 25) 7 ( 6.9; 0 to 21)
*Data are presented as mean (SD; range) or No. (%).

2330 Clinical Investigations

Downloaded From: http://journal.publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/22032/ on 06/28/2017

 CT score (Table 2). For the visual air trapping score,
the interreader reliability for the three radiologists
was 0.65. At 3 months, the dornase alfa group had a
fourfold larger increase in mean percentage of pre-
dicted FEV1 and FEF25–75% than the placebo group,
resulting in a nearly statistical difference for im-
provement in percentage of predicted FEF25–75%
(baseline, 91 ⫾ 33.8%; 3 months, 112 ⫾ 26.7%
[p ⫽ 0.07]) when paired t tests were utilized within
groups. Additional improvements for the dornase
alfa group were also noted in the visual extent of
bronchiectasis and bronchial wall thickness scores.
No statistical differences, however, were noted be-
tween groups for changes after 3 months. Unlike the
visual HRCT air trapping score at 3 months, the
quantitative air trapping measurements in the dor-
nase alfa group had mean reductions in air trapping
of 2 to 13% for A1 to A3, compared to mean increases
of 34 to 48% for A1 to A3 in the placebo group (Table
2). Figure 2 demonstrates changes in quantitative
HRCT air trapping from a subject in the placebo
group compared to a subject in the dornase alfa
group after 3 months of treatment. These differences
between groups for percentage change from baseline
for A2 and A3 after 3 months of treatment were
statistically significant.
After 12 months of treatment, the placebo group
declined in all spirometric measurements but
showed continued improvement in the visual total
HRCT score and minimal improvement in the mean
visual air trapping score (Table 2). The dornase alfa
group had a similar decline in the percentage of
predicted FEV1, a greater decline in percentage of
predicted FVC, but continued improvement in per-
centage of predicted FEF25–75% and all visual HRCT
scoring components except for the mean air trapping
score. When paired t tests were utilized for within-
group analysis, only percentage of predicted FVC
and FEV1 showed significant declines in both groups
(placebo group [n ⫽ 11]: baseline FVC, 116 ⫾ 31.2%;
12 months, 102 ⫾ 8.9% [p ⫽ 0.034]; baseline FEV1,
104 ⫾ 19.6%; 12 months, 92 ⫾ 11.8% [p ⫽ 0.051]; and
dornase alfa group [n ⫽ 8]: baseline FVC, 122 ⫾ 15.2%;
12 months, 102 ⫾ 15.2% [p ⫽ 0.003]; baseline FEV1,
107 ⫾ 21.8%; 12 months, 97 ⫾ 17.9% [p ⫽ 0.046]). At 12
months, the dornase alfa group had a significant increase
in the visual air trapping HRCT score compared to the
mean baseline score in the eight dornase alfa subjects
evaluated (baseline air trapping score, 12.3 ⫾ 3.66%; 12
months, 13.7 ⫾ 3.59% [p ⫽ 0.036]). However, the inter-
reader reliability for the three-reader visual air trapping
score at 12 months was only 0.49. Despite worsening
visual air trapping HRCT scores, there was no significant
difference between groups for changes in visual air trap-
ping scores after 12 months of the intervention. When
other changes in visual HRCT scores were compared
between groups after 12 months of treatment, the dornase
alfa group had a 6% decline in the mean total mucus
plugging score, while the placebo group had a 12%
increase (p ⫽ 0.026). When quantitative air trapping mea-
surements were evaluated after 12 months of treatment,
the dornase alfa group continued to show reduction in
mean air trapping (A1 to A3, 14 to 16%), while the placebo
group had a 38 to 61% increase in mean air trapping (A1,
A2, and A3: p ⫽ 0.091, p ⫽ 0.061, and p ⫽ 0.053, respec-
tively; Table 2).
Discussion
In this 1-year intervention study evaluating the
effect of dornase alfa in children with mild CF lung
disease, only the visual mucus plugging CT score at

Table 2—Percentage Change in PFT Results, HRCT Scores, and Air Trapping After 3 Months and 12 Months of
Treatment in CF Patients Randomized to rhDNase and Placebo Groups*

3 mo 12 mo

Dornase Alfa Placebo Dornase Alfa Placebo

Outcomes† (n ⫽ 11) (n ⫽ 14) p Value (n ⫽ 8) (n ⫽ 11) p Values
Pulmonary function measurements and HRCT scores
FVC, % predicted ⫺ 0.3 ⫾ 7.9 0.3 ⫾ 18.8 0.914 ⫺ 15.4 ⫾ 9.0 ⫺ 9.7 ⫾ 13.6 0.286
FEV1, % predicted 4.8 ⫾ 11.8 1.2 ⫾ 25.5 0.649 ⫺ 8.9 ⫾ 9.5 ⫺ 9.6 ⫾ 16.9 0.913
FEF25–75%, % predicted 37.9 ⫾ 56.7 9.4 ⫾ 40.2 0.176 17.1 ⫾ 39.2 ⫺ 4.9 ⫾ 28.5 0.202
Total HRCT score ⫺ 1.4 ⫾ 8.6 ⫺ 2.1 ⫾ 12.9 0.880 ⫺ 4.8 ⫾ 10.0 ⫺ 2.2 ⫾ 17.4 0.676
Bronchiectasis extent score ⫺ 0.9 ⫾ 14.8 4.2 ⫾ 23.6 0.520 ⫺ 4.5 ⫾ 17.7 5.0 ⫾ 19.9 0.286
Bronchial wall thickness extent score ⫺ 3.2 ⫾ 18.1 ⫺ 0.1 ⫾ 21.1 0.705 ⫺ 3.9 ⫾ 19.1 1.3 ⫾ 19.0 0.561
Mucus plugging score ⫺ 4.0 ⫾ 13.4 ⫺ 4.3 ⫾ 13.2 0.952 ⫺ 6.0 ⫾ 11.0 12.4 ⫾ 17.4 0.026
Air trapping score 7.6 ⫾ 26.9 ⫺ 4.7 ⫾ 16.9 0.206 14.2 ⫾ 20.1 ⫺ 0.5 ⫾ 16.4 0.111
Air trapping measurements
A1 ⫺ 2.1 ⫾ 44.8 ⫹ 34.4 ⫾ 62.1 0.102 ⫺ 13.6 ⫾ 44.8 38.3 ⫾ 80.2 0.091
A2 ⫺ 9.3 ⫾ 42.9 ⫹ 43.4 ⫾ 73.2 0.035 ⫺ 16.2 ⫾ 52.3 50.1 ⫾ 90.4 0.061
A3 ⫺ 13.1 ⫾ 40.5 ⫹ 48.2 ⫾ 81.2 0.023 ⫺ 15.4 ⫾ 55.5 61.3 ⫾ 102.8 0.053
*Data are presented as % change from baseline.
†Outcome measures are determined as (final measure ⫺ baseline measure)/baseline measure ⫻ 100.

www.chestjournal.org CHEST / 128 / 4 / OCTOBER, 2005 2331

Downloaded From: http://journal.publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/22032/ on 06/28/2017

 Figure 2. Quantitative air trapping (percentage of air trapping for A1, A2, and A3) obtained from
matched apical HRCT images in a placebo subject before (top, A) and after 3 months (center, B) of
treatment; and quantitative air trapping obtained from matched upper-lobe HRCT images in a dornase
alfa subject before (center, C) and after 3 months (bottom, D) of treatment. T1 ⫽ time 1 (baseline);
T2 ⫽ time 2 (3 months).

1 year demonstrated significant differences between
groups, with a continued decline in the dornase alfa
group compared to an increase in the placebo group.
However, this discrimination only occurred at 1 year
and not at 3 months. Overall, much more consistent
differences between groups at 3 months and 12
months were demonstrated in the quantitative air
trapping measurements compared to spirometry or
visual HRCT scores. This was especially true for the
quantitative A2 and A3, which demonstrated signifi-
cant differences between groups after 3 months of
treatment and nearly statistical differences between
groups after 12 months of intervention. It is sus-
pected that if there had been comparable subject
numbers evaluated at 12 months similar to 3-month
testing, the quantitative air trapping measures would
have demonstrated significant differences between
groups after 12 months of treatment as well. This
suggests that quantitative CT air trapping measure-
ments better discriminate differences in treatment
effects in children with minimal CF lung disease
(baseline FEV1 ⱖ 70%).
In this study, there was a marked difference
between the results of the visual HRCT air trapping
scores and the quantitative air trapping measures (A1
to A3). In contrast to all HRCT scoring components
in Table 2 and quantitative air trapping measures
that improved in the dornase alfa group with treat-

2332 Clinical Investigations

Downloaded From: http://journal.publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/22032/ on 06/28/2017

 ment, the visual air trapping scores worsened after 3
months and 12 months of therapy. In a similar
manner, the visual air trapping scores in the placebo
group improved after 3 months and 12 months
compared to a worsening in quantitative air trapping
scores. This surprising and counterintuitive discrep-
ancy between the visual HRCT air trapping score
and quantitative air trapping measurements may be
explained by three factors. First, subjective differ-
ences among the CT readers contributed since the
interreader reliability at baseline, 3 months, and 12
months were only 0.62, 0.65, and 0.49, respectively.
This illustrates the intrinsic weakness of reader-
dependent intersubjectivity in visual CT scoring
systems. Second, the visual HRCT air trapping scor-
ing system (scaled from 6 to 24 for the total of six
lung zones) reaches an asymptote at the higher end
of the scale (corresponding to scores ⱖ 17) com-
pared to the quantitative air trapping measures.
When regression analysis was done to evaluate the
relationship between visual CT air trapping scores
and quantitative air trapping measures, the relation-
ship was linear only for visual scores between 6 and
12 (r ⫽ 0.84, slope ⫽ 0.69, with visual scores as
independent variable for scores 6 to 12), becoming
alinear for scores ⬎ 12 (r ⫽ 0.72, slope ⫽ 3.2, with
visual scores as independent variable for scores 13 to
24). This demonstrates that for visual scores ⬎ 12, a
small increase in the visual HRCT score corresponds
to a larger increase in A3 measurements. Per hap-
penstance, the saline solution placebo group started
with higher visual HRCT air trapping scores (me-
dian, 15.67 vs 12.50 in the rhDNase group), and
higher scores were more likely to decrease since they
were at the higher (asymptotic) values. In contrast,
the range of quantitative air trapping measures (A2
and A3) were from 2.5 to 46.1% and 0.9 to 24.9%
(median, 15.9% and 6.4%) in the rhDNase group,
and 0.7 to 44.6% and 0.4 to 21.1% (median, 13.2%
and 5.1%) in the saline solution placebo group,
respectively. Finally, the visual HRCT air trapping
score (1 ⫽ absent; 2 ⫽ ⬍ 25% of lobar surface area;
3 ⫽ 25 to 50% of lobar surface area; 4 ⫽ ⬎ 50% of
lobar service area) did not have enough gradations in
scoring (ie, an additional numeric score to account
for milder air trapping corresponding to 7 to 20% of
the lobar surface area) to pick up the milder extent of
regional air trapping seen in CF patients with mild
lung disease.9
After 3 months of treatment, both groups demon-
strated improvements in percentage of predicted
FEV1 and FEF25–75% as well as total HRCT scores.
Improvements seen at 3 months in the placebo
group as well as the dornase alfa group may have
occurred due to the well-known finding that subjects
participating in research studies often have better
www.chestjournal.org
Downloaded From: http://journal.publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/22032/ on 06/28/2017
follow-up and adherence with medications during
the initial phases of a clinical study. At 12 months of
treatment, despite continued therapy, both groups
demonstrated declines in percentage of predicted
FVC and FEV1 but continued improvements in the
total visual HRCT score. In part, this may have been
due to the smaller sample sizes in each group tested
at 1 year, who had initial higher PFT values at
baseline in each of the groups compared to the larger
groups comprising the 25 subjects, or the observed
greater decline in lung function that occurs in CF
children with mild obstructive lung disease,19 –21
given that our subjects had essentially normal spiro-
metric measurements at initial enrollment with
mean percentage of predicted values for FVC and
FEV1 ⬎ 100%. In this study, the dornase alfa group
had 38% and 17% mean increases in percentage of
predicted FEF25–75% at 3 months and 12 months,
respectively, compared to a mean increase of 9% at
3 months and a mean decline of 5% at 12 months in
the placebo group. Despite the large differences in
group mean results, there was no significant differ-
ences noted between groups probably due to large
among-subject variances in FEF25–75% measure-
ments (Table 2), which have also been observed by
others.19,21–23 Although there was a decline in lung
function at 1 year, continued improvement was seen
in the average visual total HRCT scores in each
group with no significant difference in means be-
tween the groups (Table 2). These results suggest
that changes in spirometric airflow measurements
are not tracking changes in structural differences
noted by HRCT scoring in children with mild CF
lung disease. This has also been recently confirmed
by de Jong et al10 and Brody et al.11 When individual
component HRCT scores were evaluated at 3
months and 12 months, only the average total mucus
plugging score at 12 months was significantly differ-
ent between groups, with the dornase alfa group
having an improvement of 6% compared to worsen-
ing of 12% in the placebo group (p ⫽ 0.026). This
suggests that dornase alfa therapy may effect mucus
clearance in children with mild CF lung disease.
Quantitative CT air trapping measurements are
one of several new CT postprocessing techniques
that appear to be promising new methods for pro-
viding standardized quantitative CT measures simi-
lar to quantitative PFT measurements.9,13,14,24 –27
After 3 months of treatment, there were significant
differences noted between treatment groups for
percentage change in quantitative air trapping deter-
mined using A2 or A3 thresholds (Table 2). After 12
months of treatment, there were nearly significant
differences noted between the dornase alfa and
placebo groups for these outcome measures, de-
spite a smaller sample size. In each treatment
CHEST / 128 / 4 / OCTOBER, 2005 2333
 period, the dornase alfa group had improvement in
quantitative HRCT air trapping measures, while
the placebo group demonstrated worsening re-
sults, suggesting that dornase alfa improved global
HRCT air trapping measures. In addition, after 12
months of dornase alfa therapy, there was contin-
ued decreases in total visual mucus plugging
HRCT scores as well as continued improvement
from baseline in percentage of predicted FEF25–
75%, suggesting enhanced mucus clearance and
improvement and perhaps prevention of further
small airway obstruction with dornase alfa therapy
in children with mild CF lung disease. These
findings also suggest a possible association be-
tween decreased mucus plugging and improve-
ments in air trapping and small airway flow rates
(FEF25–75%) in children with mild CF lung dis-
ease.
We have previously demonstrated improvements
in the composite CT/PFT score after dornase alfa
therapy in this same subject population.12 This score
incorporates aspects of visual CT scores and pulmo-
nary function measurements to evaluate changes in
CF lung disease with treatment. The findings pre-
sented here corroborate the beneficial effects of
dornase alfa therapy in this group. Since these results
were obtained in a small sample (n ⫽ 25), they
should not be generalized without further studies in
larger groups of patients. However, in our view, this
study demonstrates the potential utility of quantita-
tive CT measures during therapeutic interventions in
subjects with early or mild obstructive airways dis-
ease.
In conclusion, we found that quantitative air trapping
measurements were more consistent sensitive outcome
measures at 3 months and 12 months of treatment than
spirometric pulmonary function measurements or visual
HRCT scores, discriminating differences in treatment
effects in children with minimal CF lung disease. These
findings suggest a potential advantage of using quantita-
tive air trapping measurements for understanding the
pathogenesis of CF lung disease and as outcome mea-
sures in clinical trials in subjects with mild CF lung
disease.
ACKNOWLEDGMENT: The authors thank Tyson Holmes from
the Division of Biostatistics, Health Research and Policy, Stan-
ford University Medical Center, for statistical analysis; Malayattil
Vijayalakshmi, Anne S. Bonnel, Krishnaveni and Kesavaraju for
technical support; and Glenn Hodge (Pediatric Pulmonary Func-
tion Laboratory) and Lisa McClennan and Diane Holmes (Pedi-
atric Radiology Section- Ultrafast CT imaging) for their partici-
pation in the study.
References
1 Bedrossian CW, Greenberg SD, Singer DB, et al. The lung in
cystic fibrosis: a quantitative study including prevalence of
2334
Downloaded From: http://journal.publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/22032/ on 06/28/2017
pathologic findings among different age groups. Hum Pathol
1976; 7:195–204
2 Esterly JR, Oppenheimer EH. Observations in cystic fibrosis
of the pancreas: 3. Pulmonary lesions. Johns Hopkins Med J
1968; 122:94 –101
3 Lynch D, Brasch R, Hardy K, et al. Pediatric pulmonary
disease: assessment with high-resolution ultrafast CT. Radi-
ology 1990; 176:243–248
4 Long FR, Castile RG, Brody AS, et al. Lungs in infants and
young children: improved thin-section CT with a noninvasive
controlled-ventilation technique: initial experience. Radiology
1999; 212:588 –593
5 Helbich TH, Heinz-Peer G, Eichler I, et al. Cystic fibrosis:
CT assessment of lung involvement in children and adults.
Radiology 1999; 213:537–544
6 Brody AS, Molina PL, Klein JS, et al. High-resolution
computed tomography of the chest in children with cystic
fibrosis: support for use as an outcome surrogate. Pediatr
Radiol 1999; 29:731–735
7 Marchant JM, Masel JP, Dickinson FL, et al. Application of
chest high-resolution computer tomography in young chil-
dren with cystic fibrosis. Pediatr Pulmonol 2001; 31:24 –29
8 Long FR. High-resolution CT of the lungs in infants and
young children. J Thorac Imaging 2001; 16:251–258
9 Goris ML, Zhu HJ, Blankenberg FG, et al. An automated
approach to quantitative air trapping measurements in mild
CF lung disease. Chest 2003; 123:1655–1663
10 de Jong, Nakano Y, Lequin MH, et al. Progressive damage on
high resolution computed tomography despite stable lung
function in cystic fibrosis. Eur Respir J 2004; 23:93–97
11 Brody AS, Klein JS, Molina PL, et al. High-resolution
computed tomography in young patients with cystic fibrosis:
distribution of abnormalities and correlation with pulmonary
function tests. J Pediatr 2004; 145:32–38
12 Robinson TE, Leung AN, Northway WH, et al. Composite
spirometric-computed tomography outcome measure in early
cystic fibrosis lung disease. Am J Respir Crit Care Med 2003;
168:588 –593
13 Bonnel AS, Song SM, Kesavaraju K, et al. Quantitative air
trapping analysis in children with mild cystic fibrosis pulmo-
nary disease. Pediatr Pulmonol 2004; 38:396 – 405
14 Long FR, Williams BS, Castile RG. Structural airway abnor-
malities in infants and young children with cystic fibrosis.
J Pediatr 2004; 144:154 –161
15 Robinson TE, Leung AN, Northway WH, et al. Spirometer-
triggered high resolution CT and pulmonary function mea-
surements during an acute exacerbation in patients with cystic
fibrosis. J Pediatr 2001 Apr; 138:553–559
16 Wang X, Dockery DW, Wypij D, et al. Pulmonary function
between 6 and 18 years of age. Pediatr Pulmonol 1993;
15:75– 88
17 Huda W. Radiation risk: CT doses are ALARA; doses and
risks are small and acceptable. Presented at: CT Conference,
North American Cystic Fibrosis Conference, Anaheim, CA,
October 17, 2003
18 McNitt-Gray M. Radiation dose from CT scanning of CF
patients: risks, benefits, and alternatives. Presented at: CT
Conference, North American Cystic Fibrosis Conference,
Anaheim, CA, October 17, 2003
19 Corey M, Levison H, Crozier D. Five- to seven-year course of
pulmonary function in cystic fibrosis. Am Rev Respir Dis
1976; 114:1085–1092
20 Davis PB, Byard PJ, Konstan MW. Identifying treatments
that halt progression of pulmonary disease in cystic fibrosis.
Pediatr Res 1996; 41:161–165
Clinical Investigations
 21 Corey M, Edwards L, Levison H, et al. Longitudinal analysis
of pulmonary function decline in patients with cystic fibrosis.
J Pediatr 1997; 131:809 – 814
22 Quan JM, Tiddens HA, Sy JP, et al. A two-year randomized,
placebo-controlled trial of dornase alfa in young patients with
cystic fibrosis with mild lung function abnormalities. J Pediatr
2001; 139:813– 820
23 Ramsey BW, Boat TF. Outcome measures for clinical trials in
CF: summary of a Cystic Fibrosis Conference. J Pediatr 1994;
124:177–192
24 King GG, Muller NL, Whittall KP, et al. An analysis algo-
rithm for measuring airway lumen and wall areas from
www.chestjournal.org
Downloaded From: http://journal.publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/22032/ on 06/28/2017
high-resolution computed tomography data. Am J Respir Crit
Care Med 2000; 161:574 –580
25 Nakano Y, Whittall KP, Kalloger SE, et al. Development and
validation of human airway analysis algorithm using multide-
tector row CT. SPIE 2002; 4683:460 – 469
26 de Jong PA, Ottink MD, Robben SGF, et al. Pulmonary
disease assessment in cystic fibrosis: comparison of CT
scoring systems and value of bronchial and arterial dimension
measurements. Radiology 2004; 231:434 – 439
27 de Jong PA, Nakano Y, Hop WC, et al. Changes in airway
dimension on computed tomography scans of children with
cystic fibrosis. Am J Respir Crit Care Med 2005; 172:218 –224
CHEST / 128 / 4 / OCTOBER, 2005 2335