Sundheds- og Ældreudvalget 2015-16
SUU Alm.del Bilag 480
Offentligt
1613993_0001.png
Articles
Intravenous augmentation treatment and lung density in
severe α1 antitrypsin deficiency (RAPID): a randomised,
double-blind, placebo-controlled trial
Kenneth R Chapman, Jonathan G W Burdon, Eeva Piitulainen, Robert A Sandhaus, Niels Seersholm, James M Stocks, Berend C Stoel, Liping Huang,
Zhenling Yao, Jonathan M Edelman, Noel G McElvaney, on behalf of the RAPID Trial Study Group*
Summary
Background
The efficacy of α1 proteinase inhibitor (A1PI) augmentation treatment for α1 antitrypsin deficiency has
not been substantiated by a randomised, placebo-controlled trial. CT-measured lung density is a more sensitive
measure of disease progression in α1 antitrypsin deficiency emphysema than spirometry is, so we aimed to assess the
efficacy of augmentation treatment with this measure.
Methods
The RAPID study was a multicentre, double-blind, randomised, parallel-group, placebo-controlled trial of
A1PI treatment in patients with α1 antitrypsin deficiency. We recruited eligible non-smokers (aged 18–65 years) in
28 international study centres in 13 countries if they had severe α1 antitrypsin deficiency (serum concentration <11 μM)
with a forced expiratory volume in 1 s of 35–70% (predicted). We excluded patients if they had undergone, or were on
the waiting list to undergo, lung transplantation, lobectomy, or lung volume-reduction surgery, or had selective IgA
deficiency. We randomly assigned patients (1:1; done by Accovion) using a computerised pseudorandom number
generator (block size of four) with centre stratification to receive A1PI intravenously 60 mg/kg per week or placebo for
24 months. All patients and study investigators (including those assessing outcomes) were unaware of treatment
allocation throughout the study. Primary endpoints were CT lung density at total lung capacity (TLC) and functional
residual capacity (FRC) combined, and the two separately, at 0, 3, 12, 21, and 24 months, analysed by modified intention
to treat (patients needed at least one evaluable lung density measurement). This study is registered with
ClinicalTrials.gov, number NCT00261833. A 2-year open-label extension study was also completed (NCT00670007).
Findings
Between March 1, 2006, and Nov 3, 2010, we randomly allocated 93 (52%) patients A1PI and 87 (48%)
placebo, analysing 92 in the A1PI group and 85 in the placebo group. The annual rate of lung density loss at TLC and
FRC combined did not differ between groups (A1PI –1·50 g/L per year [SE 0·22]; placebo –2·12 g/L per year [0·24];
difference 0·62 g/L per year [95% CI –0·02 to 1·26], p=0·06). However, the annual rate of lung density loss at TLC
alone was significantly less in patients in the A1PI group (–1·45 g/L per year [SE 0·23]) than in the placebo group
(–2·19 g/L per year [0·25]; difference 0·74 g/L per year [95% CI 0·06–1·42], p=0·03), but was not at FRC alone
(A1PI –1·54 g/L per year [0·24]; placebo –2·02 g/L per year [0·26]; difference 0·48 g/L per year [–0·22 to 1·18],
p=0·18). Treatment-emergent adverse events were similar between groups, with 1298 occurring in 92 (99%) patients
in the A1PI group and 1068 occuring in 86 (99%) in the placebo group. 71 severe treatment-emergent adverse events
occurred in 25 (27%) patients in the A1PI group and 58 occurred in 27 (31%) in the placebo group. One treatment-
emergent adverse event leading to withdrawal from the study occurred in one patient (1%) in the A1PI group and
ten occurred in four (5%) in the placebo group. One death occurred in the A1PI group (respiratory failure) and
three occurred in the placebo group (sepsis, pneumonia, and metastatic breast cancer).
Interpretation
Measurement of lung density with CT at TLC alone provides evidence that purified A1PI augmentation
slows progression of emphysema, a finding that could not be substantiated by lung density measurement at FRC
alone or by the two measurements combined. These findings should prompt consideration of augmentation treatment
to preserve lung parenchyma in individuals with emphysema secondary to severe α1 antitrypsin deficiency.
Funding
CSL Behring.
Published
Online
May 28, 2015
http://dx.doi.org/10.1016/
S0140-6736(15)60860-1
See
Online/Comment
http://dx.doi.org/10.1016/
S0140-6736(15)60036-8
*Members listed at end of paper
Asthma and Airway Centre,
University Health Network,
Toronto Western Hospital, and
Division of Respiratory
Medicine, Department of
Medicine, University of
Toronto, Toronto, ON, Canada
(Prof K R Chapman MD);
St Vincent’s Hospital, Fitzroy,
Melbourne, VIC, Australia
(J G W Burdon MD);
Skåne
University Hospital, Lund
University, Malmö, Sweden
(E Piitulainen MD);
National
Jewish Health, Denver, CO,
USA
(Prof R A Sandhaus MD);
Gentofte Hospital, Hellerup,
Denmark
(N Seersholm MD);
University of Texas Health
Science Center at Tyler, Tyler,
TX, USA
(J M Stocks MD);
Division of Image Processing,
Radiology, Leiden University
Medical Center, Leiden,
Netherlands
(B C Stoel PhD);
CSL Behring, King Of Prussia,
PA, USA
(Z Yao MD,
J M Edelman MD, L Huang MD);
and Beaumont Hospital, Royal
College of Surgeons in Ireland,
Dublin, Ireland
(Prof
N G McElvaney MD)
Correspondence to:
Prof Kenneth R Chapman,
Asthma and Airway Centre,
University Health Network,
Toronto Western Hospital,
Toronto, ON M5T 2S8, Canada
[email protected]
Introduction
Severe deficiency of α1 antitrypsin, first described by
Laurell and Eriksson
1
in 1963, is associated with a strong
tendency for development of emphysema, often, but not
always, panlobular in character and basal in distribution.
This emphysema is thought to be the result of in-
adequate neutralisation of naturally occurring proteases,
such as neutrophil elastase, by α1 proteinase inhibitor
(A1PI), which normally serves as a protease inhibitor.
2
A1PI, purified from pooled human plasma and given as
an intravenous infusion once a week at a dose of
60 mg/kg, increases and maintains A1PI serum
concentrations at more than the accepted protective
threshold of 11 μM while producing measurable
increases in the antielastase activity of the epithelial
lining fluid of the lung.
3
1
www.thelancet.com
Published online May 28, 2015 http://dx.doi.org/10.1016/S0140-6736(15)60860-1
 SUU, Alm.del - 2015-16 - Bilag 480: Henvendelse af 23/3-16 fra Alfa-1 Danmark vedr. invitation til symposium om Alfa-1 antitrypsinmangel den 12. april 2016
1613993_0002.png
Articles
No randomised, placebo-controlled clinical trial has
been able to substantiate that progression of emphysema
is slowed by A1PI augmentation treatment as shown by
established disease variables such as forced expiratory
volume in 1 s (FEV
1
). Such trials were not regarded
as feasible when augmentation treatment was first
developed.
4,5
Changes in FEV
1
take place slowly for many
years, even in a rapidly progressive disease setting, so that
several hundred patients would need to be randomised to
augmentation treatment or placebo for 5 years to establish
the effect of augmentation treatment on emphysema.
4,5
In
a rare disease setting, to do such a trial was not thought
possible on the basis of several considerations—not just
the absence of a sufficiently large population of identified
patients available for study, but also the high costs of such
a study and ethical concerns raised by extended treatment
with placebo. Since the introduction of augmentation
treatment for clinical use in the USA, Germany, Canada,
and other nations, findings from observational and cohort
studies have shown that the rate of FEV
1
loss is slower in
individuals who receive augmentation treatment than in
those who do not.
6–8
The largest of these observational
studies, the National Institutes of Health registry study,
9
showed that augmentation treatment was associated with
reduced mortality in the most severely obstructed
patients. However, such non-randomised findings can be
confounded by other factors, such as differences in
socioeconomic status and health-care-seeking behaviour
between groups.
Investigators have sought more sensitive treatment
endpoints than FEV
1
that would make possible a
definitive randomised, placebo-controlled trial in fewer
patients for less time. One such outcome measure is
lung density as quantified by CT. In the setting of
emphysema related to α1 antitrypsin deficiency, CT lung
density seems to better show lung destruction and thus
disease severity than do traditional measurements of
lung function. CT lung density, for example, is a better
predictor of mortality in α1 antitrypsin deficiency
emphysema than FEV
1
is.
10
In 1999, Dirksen and
colleagues
11
examined both FEV
1
and CT lung density
endpoints in a randomised, placebo-controlled trial of
augmentation treatment, reporting slower rates of lung
density loss in patients given augmentation treatment
than in those given placebo, although the difference was
not significant. In a pilot study of new CT methods,
Dirksen and colleagues
12
reported similar findings.
Although the data from these two trials have been
pooled to show a highly significant preservation of lung
density with augmentation treatment,
13
no single,
randomised, placebo-controlled trial has been definitive
with respect to this endpoint. For this reason, we
undertook the RAPID trial to assess the effect on CT
lung density of intravenous A1PI augmentation treat-
ment compared with intravenous placebo in patients
with emphysema secondary to severe deficiency of α1
antitrypsin.
2
Methods
Patients and study design
In this multicentre, double-blind, randomised, parallel-
group, placebo-controlled trial, we recruited men and
women aged 18–65 years with emphysema secondary
to α1 antitrypsin deficiency (with a serum A1PI
concentration of ≤11 μM) and an FEV
1
of 35–70% of the
predicted normal value from 28 study centres in
13 countries. We excluded potential participants if they
had smoked tobacco within 6 months before recruit-
ment; had undergone or were on the waiting list to
undergo lung transplantation, lobectomy, or lung
volume-reduction surgery; or had selective IgA deficiency.
We did not allow concurrent augmentation treatment.
All patients provided written informed consent and we
obtained approval from local institutional review boards.
Randomisation and masking
We randomly allocated patients (1:1; done by Accovion,
Marburg, Germany) who completed a screening period
of 1–4 weeks treatment with A1PI or matching placebo. A
randomisation list containing the assignment of patient
number to treatment group (A1PI or placebo) was
generated by a computerised pseudorandom number
generator. We stratified patients by centre. Masked study
treatments were supplied to each site in blocks of four
containing sequential patient numbers. After a patient
met all qualifications for study participation, we assigned
them the next available patient number and the
appropriate study treatment was dispensed to give to the
patient. To achieve treatment concealment, A1PI and
placebo were packaged identically as lyophilised prep-
arations and individual packages were identified only by
patient number. Study drug material was suspended in
sterile water for injection and placed in an intravenous
bag that was covered with an opaque sleeve by a
designated study nurse or pharmacist who did not
interact with the patients. Clinical trial associates
monitored compliance with the masking procedure
throughout the trial.
All patients and study investigators were unaware of
treatment allocation throughout the study, including
those assessing outcomes. The randomisation codes
remained sealed until after data collection and cleaning,
and completion of a masked analysis. The data safety
monitoring board was unmasked.
Procedures
Patients randomly allocated A1PI received intravenous
A1PI (Zemaira; CSL Behring, PA, USA) 60 mg/kg per
week for 24 months. In non-US centres, patients
completing the double-blind portion (in both the A1PI
and placebo groups) of the protocol were eligible to
receive open-label augmentation treatment with A1PI
60 mg/kg per week for a further 2 years (non-US patients
were enrolled because of unavailability of A1PI treatment
in non-US countries).
www.thelancet.com
Published online May 28, 2015 http://dx.doi.org/10.1016/S0140-6736(15)60860-1
 SUU, Alm.del - 2015-16 - Bilag 480: Henvendelse af 23/3-16 fra Alfa-1 Danmark vedr. invitation til symposium om Alfa-1 antitrypsinmangel den 12. april 2016
1613993_0003.png
Articles
We did spiral CT scans at total lung capacity (TLC) and
functional residual capacity (FRC).
14
We transformed lung
density, measured in Hounsfield units, to g/L, and applied
a physiological volume correction to 15th percentile CT
lung density (PD15), as described previously.
12
We stored
CT scan electronic files on CDs in the DICOM-3.0 format,
and identified them by patient and visit number per
investigational site before sending them by courier to
the CT core laboratory for analysis (BioClinica, Leiden,
Netherlands), which used the PulmoCMS software package
(Medis Specials, Leiden, Netherlands).
Outcomes
The primary outcome variable was annual rate of decrease
in lung density calculated from the shift of the 15th
percentile of lung density measured by CT
12
at baseline, 3,
12, 21, and 24 months. Although previous studies have
focused exclusively on lung density at TLC, at the request of
the regulatory authorities, the primary outcome was a
combined assessment of CT lung density (PD15 values)
summing density values calculated at both TLC and FRC.
Further primary outcomes were separate measurements of
PD15 density measures at FRC and TLC alone. Secondary
endpoints, measured at ten clinic visits scheduled at
intervals through the trial, were the number of exacerbations
as defined by the Anthonisen criteria,
15
exacerbation
duration and severity, FEV
1
, single-breath diffusion capacity,
baseline and achieved A1PI concentrations (functional and
antigenic assays), incremental shuttle walk test results,
health status established with the St George’s Respiratory
Questionnaire (for which high scores represent increased
disability), body-mass index, mortality, and safety.
We deemed any untoward medical event occurring
during the trial as an adverse event, and they were
assessed by the investigators as being not related,
possibly related, probably related, or related to the trial
treatment, and classified as mild, moderate, or severe.
We deemed adverse events resulting in death, judged
life-threatening, or resulting in admission to hospital
serious adverse events.
date, treatment, and treatment-by-time interaction were
fixed effects of independent variables. Patient and
patient-by-time interaction (ie—annual rate of decrease
at an individual level) were random coefficients. We
calculated percentage reduction in the rate of lung
density decrease relative to placebo for all three lung
density outcome measures. We analysed the primary and
secondary endpoints for both the modified intention-to-
treat population, excluding patients for whom no lung
density measurements were available, and the per-
protocol population, excluding patients with a major
protocol violation.
We did a planned interim descriptive analysis of the
patients in the extension study when at least 50% of them
had at least two valid CT lung density measurements at
different timepoints and repeated this analysis at the
request of the regulatory authorities when approximately
75% of patients met this criterion. Using data from the
208 patients assessed for eligibility
28 ineligible
180 randomised
93 assigned A1PI
(intention-to-treat)
1 excluded because
no scans available
92 with scans (modified
intention-to-treat)
87 assigned placebo
(intention-to-treat)
2 excluded because
no scans available
85 with scans (modified
intention-to-treat)
9 withdrew
1 death
1 adverse event
5 withdrew consent
1 missing reason
1 other reason
1 lung
transplantation
Statistical analysis
We calculated the sample size using findings from a
previous randomised, controlled trial by Dirksen and
colleagues,
11
in which the treatment effect—the difference
in the rate of lung density decline between the treatment
group and placebo—of 1·07 g/L per year had a common
SD of 2·17 g/L per year. After accounting for a dropout
proportion of 25%, we calculated that 180 patients recruited
and randomly assigned evenly to the two groups would
provide at least 80% power against a two-sided α of 0·05.
We applied a mixed-effect model to the primary
endpoint using SAS PROC MIXED. In this model, the
value of adjusted PD15 measured at baseline, 3, 12, 21,
and 24 months was the dependent variable. An indicator
of whether the value of adjusted PD15 was measured at
TLC or FRC, country, time elapsed since randomisation
18 withdrew
3 deaths
4 adverse events
7 withdrew
consent
1 protocol
violation
3 other reasons
1 suspicion of
pulmonary
cancer
1 lung
transplantation
1 no longer
wanted to
participate
84 completed treatment
(per protocol)
8 US patients
ineligible
76 assigned A1PI in
extension study
3 withdrew
50 completed treatment
(interim analysis)
69 completed treatment
(per protocol)
4 US patients
ineligible
1 non-US patient
declined
64 assigned A1PI in
extension study
3 withdrew
47 completed treatment
(interim analysis)
Figure 1:
Trial profile
A1PI=α1 proteinase inhibitor.
www.thelancet.com
Published online May 28, 2015 http://dx.doi.org/10.1016/S0140-6736(15)60860-1
3
 SUU, Alm.del - 2015-16 - Bilag 480: Henvendelse af 23/3-16 fra Alfa-1 Danmark vedr. invitation til symposium om Alfa-1 antitrypsinmangel den 12. april 2016
1613993_0004.png
Articles
double-blind portion of this trial, we also did a post-hoc
stepwise regression analysis to establish factors that
affected trough serum concentration of A1PI achieved and
the relation between concentration achieved and efficacy.
A data safety monitoring board (consisting of a
statistician and clinician independent of the funder and
study) monitored the study for safety on the basis of
adverse events and possible occurrence of anti-A1PI
serum antibodies.
This study is registered with ClinicalTrials.gov, number
NCT00261833 (extension study NCT00670007).
A1PI (n=93)
Mean age (years)
Sex
Male
Female
Race
White
FEV
1
predicted (%)
Baseline antigenic A1PI serum
concentration (μM)
Baseline CT lung density (g/L)
TLC
FRC
Combined
45·5 (15·8)
47·6 (15·7)
46·6 (15·6)
48·9 (15·5)
50·7 (15·0)
49·8 (15·1)
93 (100%)
47·4% (12·1)
6·38 (4·62)
87 (100%)
47·2% (11·1)
5·94 (2·42)
48 (52%)
45 (48%)
50 (57%)
37 (43%)
53·8 (6·9)
Placebo (n=87)
52·4 (7·8)
Role of the funding source
The funder had a role in oversight and management of
data collection. JME, LH, and ZY, who are employees of
the funder, participated in data analysis, data inter-
pretation, and writing of the report. Both placebo and
A1PI treatments were provided by the funder. The funder
paid Accovion to do the randomisation. The corresponding
author had full access to all the data in the study and had
final responsibility for the decision to submit for
publication.
Results
Between March 1, 2006, and Nov 3, 2010, we screened
208 patients, randomly assigning 180 to active treatment
(93 [52%] patients) or placebo (87 [48%] patients),
completing data collection on Sept 26, 2012 (figure 1,
table 1). Of these 180 patients, 168 (93%) were
ZZ genotype; the remainder were other variants with α1
antitrypsin serum concentrations of less than 11 μM.
16 (9%) patients had previously received augmentation
treatment, but none within 3 months before random-
isation. Assessable lung density data for at least two
timepoints were available for 92 patients in the A1PI
group and 85 in the placebo group. Fewer patients
receiving augmentation treatment (nine [10%]) withdrew
from the trial prematurely than did those receiving
placebo (18 [21%]; p=0·04). For both the modified
intention-to-treat and per-protocol populations, active
and placebo groups were well matched. The charac-
teristics of the patients who continued into the open-
label extension study were similar to those of the overall
population in this trial (appendix).
Placebo (n=87)
24 months
Baseline
24 months
A1PI
vs
placebo
Least-square mean difference
Data are n (%) or mean (SD). A1PI=α1 proteinase inhibitor. FEV
1
=forced expiratory
volume in 1 s. TLC=total lung capacity. FRC=functional residual capacity.
See
Online
for appendix
Table 1:
Baseline demographic and clinical characteristics (intention-to-
treat patients)
A1PI (n=93)
Baseline
Spirometry
Predicted FEV
1
(%)
D
LCO
(mL/mm Hg per min; %)
SGRQ score
Total
Symptoms
Activity
Impact
Shuttle walk distance (m)
A1PI concentration (μM)
Antigenic
Functional
Exacerbations‡
Annual number
Relative duration (days)
··
··
1·70 (1·51–1·89)
13·8 (15·0)
6·38 (4·62)
2·88 (3·65)
10·12 (3·52)
7·30 (2·50)
44·3 (17·1)
46·5 (22·7)
62·1 (18·6)
33·6 (18·4)
424·5 (183·0)
1·4 (11·1)
–1·4 (16·7)
1·7 (12·4)
2·1 (14·8)
10·8 (139·8)
47·4% (12·1)
13·6% (5·3)
–3·1% (10·7)
–2·2% (18·2)
47·2% (11·1)
15·0% (5·6)
42·4 (18·0)
44·1 (24·8)
60·1 (21·4)
31·4 (17·6)
435·1 (199·7)
5·94 (2·42)
2·30 (1·34)
··
··
–2·3% (13·1)
–1·5% (19·5)
2·2 (11·7)
2·0 (20·1)
2·6 (13·5)
1·8 (12·5)
16·1 (101·6)
–0·07 (1·32)
0·12 (0·96)
1·42 (1·23–1·61)
10·8 (11·6)
–2·26%* (p=0·21)
–1·31%* (p=0·64)
–0·19* (p=0·91)
–1·11* (p=0·67)
–0·16* (p=0·94)
0·74* (p=0·72)
–13·09* (p=0·48)
10·05† (p=0·02)
7·18† (p=0·02)
1·26§ (0·92–1·74)
0·56 (p=0·18)
Data are mean (SD) or n (95% CI), unless otherwise stated. A1PI=α1 proteinase inhibitor. FEV
1
=forced expiratory volume in 1 s. D
LCO
=diffusion capacity. SGRQ=St George’s
Respiratory Questionnaire. *Adjusted for country, treatment group, and baseline values. †Based on a post-hoc analysis and are the results from t tests. ‡Exacerbations
occurring in the first 2 years. §Presented as an adjusted risk ratio from a negative binomial regression model in which country and treatment were fixed effects, and
adjustment was made for treatment duration.
Table 2:
Summary of other efficacy variables (intention-to-treat patients)
4
www.thelancet.com
Published online May 28, 2015 http://dx.doi.org/10.1016/S0140-6736(15)60860-1
 SUU, Alm.del - 2015-16 - Bilag 480: Henvendelse af 23/3-16 fra Alfa-1 Danmark vedr. invitation til symposium om Alfa-1 antitrypsinmangel den 12. april 2016
1613993_0005.png
Articles
When measured at TLC and FRC combined, the
absolute difference in lung density between the
augmentation treatment group and placebo group was
0·62 g/L per year (95% CI –0·02 to 1·26, p=0·06;
A1PI –1·50 g/L per year [SE 0·22]; placebo –2·12 g/L per
year [0·24]), corresponding to a relative reduction of 29%
(0·93–76·4), but the difference was not significant. At
TLC alone, mean annual rate of lung density loss was
significantly lower in the augmentation treatment group
(–1·45 g/L per year [SE 0·23]) than in the placebo group
(–2·19 g/L per year [0·25]; p=0·03), with an absolute
difference of 0·74 g/L per year (95% CI 0·06–1·42),
corresponding to a relative reduction of 34% (2·2–84·5)
in favour of augmentation treatment. However, the
difference was also not significant at FRC alone: 0·48 g/L
per year (95% CI –0·22 to 1·18, p=0·18; A1PI –1·54 g/L
per year [SE 0·24]
vs
placebo –2·02 g/L per year [0·26]).
SDs for unadjusted PD15 values were lower at TLC (A1PI
2·23; placebo 2·38) than at FRC (A1PI 2·31, placebo
2·73; TLC and FRC combined: A1PI 2·11; placebo 2·20).
One (1%) patient in the active treatment group died
during the trial (respiratory failure) and three (3%) died
in the placebo group (sepsis, pneumonia, and metastatic
breast cancer). Secondary outcome variables are shown
in table 2 and did not differ significantly between the two
groups, except for A1PI concentration. Reported adverse
events of treatment were similar between active and
placebo groups, with 1298 treatment-emergent adverse
events occurring in 92 (99%) patients in the A1PI group
and 1068 events occurring in 86 (99%) patients in the
placebo group (table 3). 71 severe treatment-emergent
adverse events occurred in 25 (27%) patients in the A1PI
group and 58 events occurred in 27 (31%) patients in the
placebo group (table 4). One treatment-emergent adverse
event leading to withdrawal from the study occurred in
one patient (1%) in the A1PI group and ten events
occurred in four (5%) patients in the placebo group. The
time to first Anthonisen exacerbation did not differ
between groups (appendix). However, a post-hoc per-
protocol analysis done in the overall study population
showed that lung density correlated significantly with
pulmonary function and clinical variables at baseline and
study completion. For example, at study end (24 months),
the Pearson correlation coefficients were low to moderate:
0·31 (p<0·001) for predicted FEV
1
, 0·44 (p<0·001) for
diffusion capacity, 0·26 (p=0·002) for incremental
shuttle walk test, and –0·22 (p=0·02) for St George’s
Respiratory Questionnaire total score.
Trough serum A1PI concentrations achieved during
active treatment during the double-blind portion of the
trial tended to be higher in patients of higher bodyweight
and higher pretreatment serum A1PI concentrations
(data not shown). A post-hoc pharmacometric analysis
showed that annual rate of lung density loss was inversely
proportional to the trough serum A1PI concentrations
achieved, with no evidence of a plateau during the
measured range (p=0·03; figure 2).
The terminal event for progressive emphysema is either
lung transplantation or death, which occurred in five
patients. Average lung density at study exit for these patients
was less than 19·0 g/L (95% CI 3·5–29·5), and at baseline
for enrolled patients (n=180) was 47·1 g/L (23·0–76·1). With
these two lung density values and the rates of annual lung
density decrease at TLC in the two groups, the time to
terminal respiratory function can be extrapolated. In the
augmentation treatment group, we estimated time to
terminal respiratory failure to be 18·1 years (12·2–30·1); for
patients receiving placebo, the estimate was 12·3 years
(8·1–19·9).
Annual rate of lung density decrease during both the
double-blind and open-label portions of the trial is shown
in figure 3 for all patients who had completed the
open-label extension at the time of the second interim
analysis. The rate of lung density loss was greater in
patients who were taking placebo during the double-
blind portion of the trial than in those given A1PI, but
slowed to parallel that of patients who had received active
treatment throughout in the extension study.
A1PI (n=93)
Patients
Any TEAE
Infections and infestations
Bronchitis
Influenza
Nasopharyngitis
Pneumonia
Sinusitis
Upper respiratory
Lower respiratory
Viral*
Respiratory disorders
Chronic obstructive pulmonary disease
Cough
Dyspnoea
Oropharyngeal pain
Gastrointestinal disorders
Nausea
General and administration site disorders
Condition aggravated
Fatigue
Pyrexia
Nervous system
Headache
Musculoskeletal and connective tissue disorders
Back pain
92 (99%)
77 (83%)
12 (13%)
14 (15%)
30 (32%)
11 (12%)
12 (13%)
14 (15%)
18 (19%)
3 (3%)
63 (68%)
30 (32%)
20 (22%)
17 (18%)
22 (24%)
46 (49%)
15 (16%)
48 (52%)
20 (22%)
8 (9%)
13 (14%)
46 (49%)
37 (40%)
35 (38%)
12 (13%)
Events
1298 (7·58)
334 (1·95)
26 (0·15)
14 (0·08)
53 (0·31)
15 (0·09)
17 (0·10)
26 (0·15)
88 (0·51)
5 (0·03)
249 (1·45)
107 (0·63)
31 (0·18)
29 (0·17)
36 (0·21)
104 (0·61)
23 (0·13)
144 (0·84)
62 (0·36)
14 (0·08)
15 (0·09)
194 (1·13)
98 (0·57)
68 (0·40)
12 (0·07)
Placebo (n=87)
Patients
86 (99%)
76 (87%)
11 (13%)
10 (11%)
26 (30%)
12 (14%)
10 (11%)
14 (16%)
17 (20%)
4 (5%)
49 (56%)
20 (23%)
7 (8%)
10 (11%)
10 (11%)
47 (54%)
8 (9%)
42 (48%)
14 (16%)
10 (11%)
6 (7%)
43 (49%)
33 (38%)
37 (43%)
10 (11%)
Events
1068 (7·23)
369 (2·50)
16 (0·11)
12 (0·08)
58 (0·39)
25 (0·17)
18 (0·12)
25 (0·17)
72 (0·49)
6 (0·04)
127 (0·86)
53 (0·36)
7 (0·05)
11 (0·07)
13 (0·09)
92 (0·62)
11 (0·07)
101 (0·68)
41 (0·28)
12 (0·08)
8 (0·05)
134 (0·91)
105 (0·71)
75 (0·51)
12 (0·08)
Data are n (%) or n (annualised incidence rate). The annualised incidence rate is based on exposures of 171·14 A1PI subject
years and 147·75 placebo patient years. Each patient could have more than one adverse event. A1PI=α1 proteinase
inhibitor. TEAE=treatment-emergent adverse event. *Experienced by less than 10% of patients in either treatment group.
Table 3:
Reported TEAEs and exposure-adjusted incidence rates organised by selected system organ
classifications and preferred terms experienced by at least 10% of patients in either treatment group
(safety population)
www.thelancet.com
Published online May 28, 2015 http://dx.doi.org/10.1016/S0140-6736(15)60860-1
5
 SUU, Alm.del - 2015-16 - Bilag 480: Henvendelse af 23/3-16 fra Alfa-1 Danmark vedr. invitation til symposium om Alfa-1 antitrypsinmangel den 12. april 2016
1613993_0006.png
Articles
A1PI (n=93)
Patients
Any TEAE
Mild
Moderate
Severe
Any related TEAE
Any TEAE within 24 h
Any related TEAE within 24 h
Any AR†
Occurring within 72 h
Related
Any serious TEAE
Any related serious TEAE
Any TEAE leading to withdrawal from study
Any related TEAE leading to withdrawal from study
92 (99%)
13 (14%)
54 (58%)
25 (27%)
21 (23%)
78 (84%)
15 (16%)
86 (92%)
85 (91%)
21 (23%)
28 (30%)
1 (1%)
1 (1%)
1 (1%)
Events
1298 (7·58)
780 (4·56)
447 (2·61)
71 (0·41)
91 (0·53)
373 (2·18)
51 (0·30)
702 (4·10)
677 (3·96)
91 (0·53)
57 (0·33)
1 (0·01)
1 (0·01)
1 (0·01)
Placebo (n=87)
Patients
86 (99%)
16 (18%)
43 (49%)
27 (31%)
21 (24%)
78 (90%)
18 (21%)
83 (95%)
83 (95%)
21 (24%)
28 (32%)
1 (1%)
4 (5%)
1 (1%)
Events
1068* (7·23)
666 (4·51)
343 (2·32)
58 (0·39)
50 (0·34)
328 (2·22)
35 (0·24)
560 (3·79)
549 (3·72)
50 (0·34)
45 (0·30)
1 (0·01)
10 (0·07)
4 (0·03)
Discussion
Although the primary statistical endpoint of PD15 lung
density at TLC and FRC combined was non-significant
(along with the primary endpoint of FRC alone), this
finding can be accounted for by the fact that measure-
ment error for unadjusted PD15 is highest for CT scans
obtained at lowest lung volumes (eg, FRC) and lowest
for those acquired at highest volumes (eg, TLC).
16
The
combination of CT data obtained at TLC and FRC
results in a measurement error intermediate to that at
either TLC alone or FRC alone. CT lung density
measurement at TLC alone (a primary endpoint) did
show a significant difference between the rate of lung
parenchymal loss in patients with α1 antitrypsin
deficiency emphysema who received infusions of pur-
ified A1PI and those who did not—about a third slower
in those that received A1PI than in those who did not.
Data from this trial substantiates previous reports
16,17
that CT density measured at TLC has smaller variation
than does that measured at FRC, and thus CT data
acquired at TLC are deemed more reliable than those
acquired at FRC. These findings are consistent with the
understood biological mechanisms of α1 antitrypsin
protein and the reported results of observational and
cohort studies
7–9
showing reduced rates of FEV
1
decrease
and mortality with augmentation treatment (panel).
Moreover, our estimates of lung density decrease are
consistent with those reported for treated and untreated
patients in previous studies using CT densitometry.
11,12
The rate of lung density decrease in this trial was similar
to that noted in a randomised controlled trial by Dirksen
and colleagues
12
(0·86 g/L per year [95% CI −0·08 to
1·78]). In another randomised controlled trial with some
methodological differences,
11
the rate was 1·07 g/L per
year (SE 0·58).
Our analyses provide two further arguments to suggest
that augmentation treatment has a disease-modifying
effect in patients with α1 antitrypsin deficiency emphy-
sema. First, although our study was not designed to
study the effect of different treatment doses across the
range of post-treatment serum concentrations achieved
with active and placebo treatment, the effect of treatment
was dose-related such that patients with the highest
trough serum concentrations tended to have the slowest
annual rates of lung density loss. Second, our analysis of
the open-label treatment extension makes an artifactual
effect of augmentation treatment unlikely. If deposition
of exogenous protein in the epithelial lining fluid of the
lung could lead to lung density overestimation by CT
techniques, we would expect delayed introduction of
augmentation treatment to patients previously given
placebo to return their estimated lung density to that of
Figure 2:
Rates of lung density decrease at total lung capacity versus trough
A1PI serum concentrations achieved
(A) All datapoints for patients across the entire range of observed lung density
decrease. (B) Response-exposure curve. Shaded area represents 90% CIs.
A1PI=α1 proteinase inhibitor.
Data are n (%) or n (annualised incidence rate). The annualised incidence rate is based on exposures of 171·14 A1PI
patient years and 147·75 placebo patient years. Each patient could have more than one adverse event. A patient with
more than one TEAE was counted in the severity category associated with the most severe TEAE. TEAE=treatment-
emergent adverse event. AR=adverse reaction. *One TEAE (panic attack) was not classified by severity. †ARs could
occur both within 72 h and be related to treatment.
Table 4:
Reported TEAEs and exposure-adjusted incidence rates organised by severity (safety population)
A
15
Lung density decrease rate (g/L per year)
Placebo
A1PI
10
5
0
–5
–10
–15
0
5
10
15
20
25
30
Median postbaseline serum A1PI concentration (μM)
B
0
Lung density decrease rate (g/L per year)
Placebo (10th, 50th, 90th percentile)
A1PI (10th, 50th, 90th percentile)
–0·5
–1·0
–1·5
–2·0
–2·5
–3·0
5
10
15
20
25
30
Median postbaseline serum A1PI concentration (μM)
6
www.thelancet.com
Published online May 28, 2015 http://dx.doi.org/10.1016/S0140-6736(15)60860-1
 SUU, Alm.del - 2015-16 - Bilag 480: Henvendelse af 23/3-16 fra Alfa-1 Danmark vedr. invitation til symposium om Alfa-1 antitrypsinmangel den 12. april 2016
1613993_0007.png
Articles
continuously treated patients. Instead, their rate of lung
density loss slowed to match that of continuously
augmented patients, but the density lost was not
recovered.
Unsurprisingly, findings from our study did not show
significant differences between active and placebo
treatment in conventional pulmonary function and
clinical endpoints; the study was not designed with
sufficient power to detect such changes. Small numerical
differences between groups in rate of FEV
1
change and
exacerbations favouring placebo were non-significant,
but could have been affected by the different withdrawal
between groups. We believe that estimation of lung
density with CT is not just a more sensitive outcome
measure than those used conventionally, but is more
appropriate for this patient population. In typical non-α1
antitrypsin-deficient chronic obstructive pulmonary
disease, the degree of emphysema present on CT scans
can be discordant with clinical severity, a finding that
shows the heterogeneous nature of the disease.
17
By
contrast, emphysema in individuals with α1 antitrypsin
deficiency is more homogeneous than is chronic
obstructive pulmonary disease. Estimates of lung density
for this form of emphysema correlate well with
conventional measures of lung function and disease
outcome, but lung density estimates have greater
sensitivity and prognostic value than do conventional
measures.
10
In a post-hoc analysis, we noted an inverse relation
between α1 antitrypsin serum concentration achieved
and clinical efficacy as measured by rate of lung density
decrease. We did not note a plateau to this dose–response
relation, raising the possibility that the dose of 60 mg/kg
per week is not the optimum augmentation treatment
dose for all patients. This possibility has been considered
previously because the present dose of treatment was
based on achievement of serum concentrations at the
lower limit of the range seen in mildly deficient
genotypes, individuals thought to have no increased risk
of emphysema and now understood to have a high risk
of Global Initiative on Obstructive Lung Disease Stage
II chronic obstructive pulmonary disease (odds ratio of
more than 1·2).
18–21
Authors of preliminary studies have
noted that infusions of 120 mg/kg per week are well
tolerated,
22
and efficacy studies have been planned
(NCT01669421 and NCT01983241).
In a further post-hoc analysis, we estimated that
patients receiving purified A1PI would be expected to
take more time to reach terminal respiratory function
(transplantation or death) when compared with those
not receiving active treatment, and the results pointed
to the potential clinical effect of a reduction of the rate
of lung density decrease in patients with emphysema.
However, the precise numbers should be interpreted
with caution as they are based on a very small number
of patients who reached terminal respiratory failure or
death—further investigations are needed.
A
0
–1·45 g/L per year
Lung density change from baseline (g/L)
–1
–2
–3
–4
–5
–6
–7
Double-blind trial*
Open-label extension trial†
–1·31 g/L per year
–1·08 g/L per year
–2·19 g/L per year
A1PI
Placebo
B
0
Lung density change from baseline (g/L)
–1
–2
–3
–2·06 g/L per year
–4
–5
–1·31 g/L per year
–6
–7
0
12
Double-blind trial†
Month
24
36
Open-label extension trial†
48
–1·37 g/L per year
–1·08 g/L per year
Figure 3:
Rates of lung density decrease at TLC during the double-blind and open-label portions of the trial in
(A) all patients and (B) patients completing the open-label study only
Values on graph are annual rates of decrease. A1PI=α1 proteinase inhibitor. *A1PI n=92; placebo n=85. †A1PI n=50;
placebo n=47.
Some limitations to our study should be noted. First,
although we have attempted to estimate the clinical
effect of lung density changes on clinical outcomes with
post-hoc analyses, our study does not allow us to
establish the effect of lung density preservation on
typical clinical outcomes of lung function, exacerbations,
and survival. Second, although we have provided some
evidence of efficacy of augmentation treatment at the
currently recommended dose of 60 mg/kg per week, we
have not established that this is the optimum dose.
Third, we do not know whether preservation of lung
density or structure is uniform across all severely
deficient patients or stages of the disease. In non-α1
antitrypsin deficiency chronic obstructive pulmonary
disease, for example, lung function changes occur more
rapidly in mild than in severe disease.
23
Additionally, we
do not know the duration of the protective effect with
treatment continued beyond 4 years. Finally, the higher
7
www.thelancet.com
Published online May 28, 2015 http://dx.doi.org/10.1016/S0140-6736(15)60860-1
 SUU, Alm.del - 2015-16 - Bilag 480: Henvendelse af 23/3-16 fra Alfa-1 Danmark vedr. invitation til symposium om Alfa-1 antitrypsinmangel den 12. april 2016
1613993_0008.png
Articles
Panel:
Research in context
Systematic review
We searched PubMed up to April 24, 2015, for randomised, controlled trials of CT lung
scanning to assess emphysema progression in patients with α1 antitrypsin deficiency who
received intravenous α1 proteinase inhibitor (A1PI) augmentation treatment or control.
We used the following search terms (with no language restrictions): “alpha-1 antitrypsin”,
“augmentation therapy”, “computed tomography”, and “randomized controlled trial”.
We identified two relevant randomised controlled trials. In 1999, Dirksen and colleagues
11
reported data from 56 ex-smokers with α1 antitrypsin deficiency (all PI*ZZ) and moderate
emphysema who received monthly infusions of either intravenous A1PI or 2% human
albumin for at least 3 years. No significant difference was noted between the two groups
in terms of forced expiratory volume in 1 s, but a trend towards slower annual loss of lung
density, as measured by CT densitometry, was apparent in the treated group. In a
subsequent trial, Dirksen and colleagues
12
took lung CT measurements of 77 patients with
α1 antitrypsin deficiency who received A1PI or 2% human albumin every week for
2−2·5 years, reporting a trend towards treatment benefit. Authors of a pooled analysis of
these two randomised controlled trials
13
concluded that intravenous A1PI augmentation
treatment significantly reduces the rate of CT-measured lung density decrease in patients
with α1 antitrypsin deficiency-related emphysema.
Interpretation
Our data suggest that intravenous augmentation of serum α1 antitrypsin in individuals
with α1 antitrypsin deficiency-related emphysema can slow loss of lung parenchyma as
ascertained by CT-measured lung density at TLC, but this finding was not significant when
measured at TLC and FRC combined, or FRC alone. In an exploratory post-hoc analysis, the
finding was underscored by an apparent dose–response relation such that the higher the
serum concentration achieved with infusion, the slower the resulting loss of lung density.
This estimate was based on the small variation in serum concentrations seen after
administration of the single standard recommended dose and raises the question of
whether present dosing recommendations are the best possible. In a planned interim
analysis of an open-label extension to the present randomised trial, we noted that delayed
introduction of augmentation treatment slowed lung density loss, but lung parenchyma
lost during the delay was not recovered. These findings should encourage early
introduction of augmentation treatment in those with emphysema secondary to severe
α1 antitrypsin deficiency and should stimulate further research into optimum dosing.
Finland
R Mäkitaro.
Sweden
E Piitulainen.
Poland
M Sanak,
A Szczeklik, W Z Tomkowski.
Denmark
N Seersholm, T Skjold.
Romania
P I Stoicescu.
RAPID Trial Data Safety Monitoring Board
Netherlands
J Stolk (chair).
Germany
C Vogelmeier, F Schindel
(independent statistician).
Declaration of interests
KRC has received compensation for consulting with AstraZeneca, Baxter,
Boehringer Ingelheim, CSL Behring, GlaxoSmithKline, Grifols, Kamada,
Novartis, Nycomed, Roche, and Telacris; has done research funded by
Amgen, AstraZeneca, Baxter, Boehringer Ingelheim, CSL Behring, Forest
Labs, GlaxoSmithKline, Grifols, Novartis, Roche, and Takeda; and has
participated in continuing medical education activities sponsored in
whole or part by AstraZeneca, Boehringer Ingelheim, GlaxoSmithKline,
Grifols, Merck Frosst, Novartis, Pfizer, and Takeda. He is participating in
research funded by the Canadian Institutes of Health Research operating
grant entitled Canadian Cohort Obstructive Lung Disease. He holds the
GlaxoSmithKline-Canadian Institutes of Health Research Chair in
Respiratory Health Care Delivery at the University Health Network, ON,
Canada. JGWB has received research funding from CSL Behring and
consulted with Baxter. EP has received research funding for this study
from CSL Behring. RAS reports grants and personal fees from CSL
Behring during this study, other grants and personal fees from CSL
Behring and Grifols, personal fees from Baxter, and non-financial
support from the Alpha-1 Project (venture philanthropy) outside of the
submitted work, and is employed by AlphaNet, a not-for-profit
organisation providing disease management services for patients with α1
antitrypsin deficiency. All other authors declare no competing interests.
References
1
Laurell CB, Eriksson S. The electrophoretic alpha1-globulin pattern
of serum in alpha1-antitrypsin deficiency.
Scand J Clin Lab Invest
1963;
115:
132–40.
2
Gadek JE, Fells GA, Zimmerman RL, Rennard SI, Crystal RG.
Antielastases of the human alveolar structures. Implications for the
protease–antiprotease theory of emphysema.
J Clin Invest
1981;
68:
889–98.
3
Wewers MD, Casolaro MA, Sellers SE, et al. Replacement therapy
for alpha 1-antitrypsin deficiency associated with emphysema.
N Engl J Med
1987;
316:
1055–62.
4
Burrows B. A clinical trial of efficacy of antiproteolytic therapy:
can it be done?
Am Rev Respir Dis
1983;
127:
S42–43.
5
Idell S, Cohen AB. Alpha-1-antitrypsin deficiency.
Clin Chest Med
1983;
4:
359–75.
6
Chapman KR, Stockley RA, Dawkins C, Wilkes MM, Navickis RJ.
Augmentation therapy for alpha1 antitrypsin deficiency: a
meta-analysis.
COPD
2009;
6:
177–84.
7
Seersholm N, Wencker M, Banik N, et al. Does alpha1-antitrypsin
augmentation therapy slow the annual decline in FEV
1
in patients with
severe hereditary alpha1-antitrypsin deficiency? Wissenschaftliche
Arbeitsgemeinschaft zur Therapie von Lungenerkrankungen (WATL)
alpha1-AT study group.
Eur Respir J
1997;
10:
2260–63.
8
Wencker M, Fuhrmann B, Banik N, Konietzko N, for
Wissenschaftliche Arbeitsgemeinschaft zur Therapie von
Lungenerkrankungen. Longitudinal follow-up of patients with
alpha(1)-protease inhibitor deficiency before and during therapy
with IV alpha(1)-protease inhibitor.
Chest
2001;
119:
737–44.
9
Survival and FEV
1
decline in individuals with severe deficiency of
alpha1-antitrypsin. The Alpha-1-Antitrypsin Deficiency Registry
Study Group.
Am J Respir Crit Care Med
1998;
158:
49–59.
10 Dawkins PA, Dowson LJ, Guest PJ, Stockley RA. Predictors of
mortality in alpha1-antitrypsin deficiency.
Thorax
2003;
58:
1020–26.
11 Dirksen A, Dijkman JH, Madsen F, et al. A randomized clinical trial
of alpha(1)-antitrypsin augmentation therapy.
Am J Respir Crit Care Med
1999;
160:
1468–72.
12 Dirksen A, Piitulainen E, Parr DG, et al. Exploring the role of CT
densitometry: a randomised study of augmentation therapy in
alpha1-antitrypsin deficiency.
Eur Respir J
2009;
33:
1345–53.
13 Stockley RA, Parr DG, Piitulainen E, Stolk J, Stoel BC, Dirksen A.
Therapeutic efficacy of alpha-1 antitrypsin augmentation therapy
on the loss of lung tissue: an integrated analysis of 2 randomised
clinical trials using computed tomography densitometry.
Respir Res
2010;
11:
136.
number of withdrawals in the patients given placebo
than in those in the treatment group remains un-
explained, but is concordant with the greater loss of lung
density in this group and the relative insensitivity of
conventional clinical variables to such deterioration.
Differential withdrawal in other large trials of treatment
in chronic obstructive pulmonary disease unrelated to
α1 antitrypsin deficiency has been regarded as an
indicator of treatment efficacy.
24
Contributors
KRC, JGWB, EP, RAS, NS, JMS, and NGM were study investigators and
planned the study, collected the data, reviewed the analyses, and wrote
the report. KRC prepared the first draft of the report. NGM was the
principal investigator. BCS, LH, ZY, and JME designed the study and
analysed the data. All authors reviewed and approved the final version
of the report.
RAPID Trial Study Group
Canada
R T Abboud, K R Chapman, P Hernandez.
Estonia
A Altraja.
Australia
J G W Burdon, R Edwards, A Glanville, M Holmes,
P Thompson, P A Wark.
USA
M Campos, T J Craig, R A Sandhaus,
J M Stocks.
Czech Republic
J Chlumsky.
Germany
J Ficker, J F Herth,
K Schulze, H Teschler.
Russia
T Martynenko.
Ireland
N G McElvaney.
8
www.thelancet.com
Published online May 28, 2015 http://dx.doi.org/10.1016/S0140-6736(15)60860-1
 SUU, Alm.del - 2015-16 - Bilag 480: Henvendelse af 23/3-16 fra Alfa-1 Danmark vedr. invitation til symposium om Alfa-1 antitrypsinmangel den 12. april 2016
1613993_0009.png
Articles
14
15
16
17
18
19
Stoel BC, Putter H, Bakker ME, et al. Volume correction in
computed tomography densitometry for follow-up studies on
pulmonary emphysema.
Proc Am Thorac Soc
2008;
5:
919–24.
Anthonisen NR, Manfreda J, Warren CP, Hershfield ES, Harding GK,
Nelson NA. Antibiotic therapy in exacerbations of chronic obstructive
pulmonary disease.
Ann Intern Med
1987;
106:
196–204.
Parr DG, Dirksen A, Piitulainen E, Deng C, Wencker M,
Stockley RA. Exploring the optimum approach to the use of CT
densitometry in a randomised placebo-controlled study of
augmentation therapy in alpha 1-antitrypsin deficiency.
Respir Res
2009;
10:
75.
de-Torres JP, Blanco D, Alcaide AB, et al. Smokers with CT detected
emphysema and no airway obstruction have decreased plasma
levels of EGF, IL-15, IL-8 and IL-1ra.
PLoS One
2013;
8:
e60260.
Molloy K, Hersh CP, Morris VB, et al. Clarification of the risk of
chronic obstructive pulmonary disease in alpha1-antitrypsin
deficiency PiMZ heterozygotes.
Am J Respir Crit Care Med
2014;
189:
419–27.
Dahl M. Genetic and biochemical markers of obstructive lung
disease in the general population.
Clin Respir J
2009;
3:
121–22.
20
21
22
23
24
Seersholm N, Kok-Jensen A. Intermediate alpha 1-antitrypsin
deficiency PiSZ: a risk factor for pulmonary emphysema?
Respir Med
1998;
92:
241–45.
Sorheim IC, Bakke P, Gulsvik A, et al. Alpha-antitrypsin protease
inhibitor MZ heterozygosity is associated with airflow obstruction
in two large cohorts.
Chest
2010;
138:
1125–32.
Campos MA, Kueppers F, Stocks JM, et al. Safety and
pharmacokinetics of 120 mg/kg versus 60 mg/kg weekly
intravenous infusions of alpha-1 proteinase inhibitor in alpha-1
antitrypsin deficiency: a multicenter, randomized, double-blind,
crossover study (SPARK).
COPD
2013;
10:
687–95.
Vestbo J, Edwards LD, Scanlon PD, et al. Changes in forced
expiratory volume in 1 second over time in COPD.
N Engl J Med
2011;
365:
1184–92.
Calverley PM, Anderson JA, Celli B, et al. Salmeterol and
fluticasone propionate and survival in chronic obstructive
pulmonary disease.
N Engl J Med
2007;
356:
775–89.
www.thelancet.com
Published online May 28, 2015 http://dx.doi.org/10.1016/S0140-6736(15)60860-1
9