Background
§§ Two large clinical studies (TONADO®
1 + 2) demonstrated
the benefits of treatment with a combination of once-daily
tiotropium (T), a long-acting muscarinic antagonist, and
olodaterol (O), a long-acting 2
-agonist, compared to
treatment with the monocomponents in patients with
moderate to very severe chronic obstructive pulmonary
disease (COPD) over 52 weeks.1
§§ Until recently, the focus of treatment had been on
improving the single outcome of lung function. However,
it is becoming clear that many factors come into play as
COPD deteriorates, and it is of interest to investigate
whether these clinically significant events can also be
delayed with optimized bronchodilation, which, in turn,
could lead to longer-term disease stabilization.
§§ A post hoc analysis of the UPLIFT study demonstrated that
composite end points, which included clinically significant
events for patients with Global initiative for chronic
Obstructive Lung Disease (GOLD) 2 and GOLD stage B
COPD, can be sensitive to treatment effects, with T
delaying clinically significant events compared to placebo.2
§§ The two composite end points in the UPLIFT analysis
evaluated the time to first clinically important deterioration
in: (1) trough forced expiratory volume in 1 second (FEV1
),
St. George’s Respiratory Questionnaire (SGRQ) score,
severe exacerbation, or death; or (2) trough FEV1
,
SGRQ score, or moderate or severe exacerbation.2
The first composite end point showed limitations, with the
severe exacerbations and death components not of value
in patients with moderate COPD.
§§ Using the same methodology as for the UPLIFT
analysis, we investigated whether T+O is more effective
than T at delaying clinically significant events in patients
with GOLD stage B COPD (symptomatic COPD and a
low risk of exacerbations) in a post hoc analysis of the
TONADO®
studies.
Methods
Study design
§§ A total of 5162 patients were randomized to
T+O 2.5/5 µg or 5/5 µg, T 2.5 µg or 5 µg, or O 5 µg
(delivered via Respimat®
inhaler) in two 52-week,
parallel-group, double-blind studies (TONADO®
1
[NCT01431274] and TONADO®
2 [NCT01431287])
(Figure 1).
§§ Assessment of trough FEV1
was performed on Day 1
and at Weeks 2, 6, 12, 18, 24, 32, 40, and 52.
§§ SGRQ was completed on Day 1 and after 12, 24, and
52 weeks.
Analyses
§§ Post hoc analysis of time to first clinically important
deterioration in patients classified as GOLD stage B
using combined TONADO®
study data.
§§ Clinically important deterioration was defined according
to two composite end points.
–– Composite end point 1: time to first decrease in
trough FEV1
from baseline of 100 mL, time to
first increase in SGRQ total score from baseline of
4 units, time to first severe (hospitalized)
exacerbation, or time to death.
–– Composite end point 2: time to first decrease in trough
FEV1
from baseline of 100 mL, time to first increase
in SGRQ total score from baseline of 4 units, or time
to first moderate or severe exacerbation.
§§ The time to first occurrence of one of these events was
recorded as the time to clinically important deterioration.
§§ Data are presented for comparisons of the licensed
doses of T+O 5/5 µg and T 5 µg.
Statistical analyses
§§ Time to first clinically significant event (individual
components of composite end points) and time to
first clinical deterioration (composite end points) were
calculated in days and reported for the 25th percentile
(median time not reached for most events) for each
treatment group.
§§ Hazard ratios (HRs) for treatment comparisons were
obtained from fitting a Cox proportional hazard
regression model with treatment as the only covariate.
§§ Kaplan–Meier estimates of probability of clinical
deterioration based on the composite end points were
generated for each treatment group.
Results
§§ Patient demographics and disease characteristics
were comparable for GOLD stage B patients in the
T+O 5/5 µg and T 5 µg treatment groups included in
this analysis (Table 1).
§§ Using composite end point 1 (time to first trough FEV1
decline, SGRQ total score increase, severe
exacerbation, or death), time to clinically important
deterioration was significantly longer with T+O 5/5 µg
than T 5 µg (HR [95% confidence interval (CI)] 0.65
[0.52, 0.81]; p0.0001) (Figure 2).
§§ For the individual clinically significant events included in
composite end point 1, time to trough FEV1
decline (HR
[95% CI] 0.66 [0.51, 0.86]; p=0.0016) and time to SGRQ
score increase (HR [95% CI] 0.71 [0.52, 0.96]; p=0.0271)
were significantly longer with T+O 5/5 µg than T 5 µg
(Table 2).
–– Event rates for time to severe exacerbation and time
to death were very low (one death and 16 patients
with severe exacerbations across both treatment
groups) and 25th percentiles were non-estimable.
§§ For time to moderate or severe exacerbation, the HR
(95% CI) was 0.94 (0.67, 1.32) and 25th percentiles were
non-estimable (Table 2). However, event rates were
greater than for severe exacerbations: T+O 5/5 µg,
67 patients (21.6%); T 5 µg, 69 patients (22.5%).
§§ Using composite end point 2 (including time to first trough
FEV1
decline, SGRQ total score increase, or moderate or
severe exacerbation, and excluding severe exacerbation
or death), time to clinically important deterioration was
significantly longer with T+O 5/5 µg than T 5 µg (HR [95%
CI] 0.68 [0.56, 0.83]; p=0.0002) (Figure 3)
Limitations
§§ In contrast to the UPLIFT study analysis,2
a clinically
significant event did not have to be confirmed at a
second clinic visit in order to be included in the
analysis. This was a consequence of the length of
the TONADO®
studies (52 weeks) and the number of
assessments occurring during this period (only three
SGRQ assessments).
§§ The temporal relationship between clinically significant
events is not known, as only the time to first individual
event is included in the analysis.
Benefits of tiotropium + olodaterol over tiotropium at delaying clinically
significant events in patients with COPD classified as GOLD B
Roland Buhl,1
Lorcan McGarvey,2
Stephanie Korn,1
Gary T. Ferguson,3
Lars Grönke,4
Christoph Hallmann,4
Florian Voß,4
Klaus F. Rabe,5
François Maltais6
1
Pulmonary Department, Mainz University Hospital, Mainz, Germany; 2
Centre for Infection and Immunity, School of Medicine, Dentistry and Biomedical Sciences, Queen’s University Belfast, Belfast, UK; 3
Pulmonary Research Institute of Southeast Michigan,
Farmington Hills, Michigan, USA; 4
Boehringer Ingelheim Pharma GmbH & Co. KG, Ingelheim, Germany; 5
Lung Clinic Grosshansdorf, Grosshansdorf, Germany; 6
Centre de Recherche, Institut Universitaire de Cardiologie et de Pneumologie de Québec, Québec, Canada
http://goo.gl/7Eya01
Poster presented at the American Thoracic Society International Conference, San Francisco, California, USA, May 13–18, 2016
Poster P9
References
1.	 Buhl R et al. Eur Respir J 2015; 45: 969–979.
2.	 Rabe K et al. Abstract presented at the American Thoracic Society
Annual Meeting, San Francisco, California, USA, May 13–18, 2016.
Acknowledgments
This study was supported by Boehringer Ingelheim. Under the authors’
conceptual direction, medical writing assistance for this poster was provided
by Gail Rickard at Complete HealthVizion; medical writing assistance was
supported by Boehringer Ingelheim.
Disclosures
RB reports financial support to his institution from Boehringer Ingelheim
during the conduct of the study, and outside of the submitted work,
financial support from Boehringer Ingelheim, AstraZeneca, Chiesi, Novartis,
TEVA, Roche, and GlaxoSmithKline for attendance at advisory boards and
speaking activities, and research support to his institution from Boehringer
Ingelheim, GlaxoSmithKline, Novartis, and Roche. FM reports financial
support from Boehringer Ingelheim during the conduct of the study, and,
outside of the submitted work, financial support from Boehringer Ingelheim
and GlaxoSmithKline for attendance at advisory boards, financial support
from Boehringer Ingelheim, GlaxoSmithKline, AstraZeneca, and Novartis
for speaking activities, and research support from Boehringer Ingelheim,
GlaxoSmithKline, AstraZeneca, Nycomed, and Novartis.
Conclusions
§§ In the TONADO®
studies, T+O increased time to
clinically important deterioration compared to
T alone in patients with GOLD stage B COPD,
suggesting that T+O is more effective than T in
preventing these significant events in this
patient population.
§§ Based on these results, T+O not only significantly
improves lung function but may also lead to a
slower clinical deterioration of the disease
through its effects beyond lung function.
§§ For this patient population with less severe
disease, and in which severe exacerbations
and deaths occur infrequently, it appears
that comparing the effectiveness of different
treatments in delaying clinically important
deterioration can be achieved using a composite
end point that includes time to first trough FEV1
decline or SGRQ score increase, or moderate
or severe exacerbation.
§§ Further studies are warranted to prospectively
study this effect.
Figure 2. Kaplan–Meier estimates of probability of decline
from baseline in trough FEV1
of 100 mL, increase from baseline
of 4 units in SGRQ score, severe exacerbation, or death
(composite end point 1) in patients with GOLD stage B COPD.a
1.00
Estimatedprobabilityofnoevent
0.95
0.90
0.85
0.80
0.75
0.70
0.65
0.60
0.55
0.50
0.45
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
0 50 100 150 200 250 300 350
Patients at risk
T+O 5/5 µg
T 5 µg
278
239
310
306
229
187
220
172
192
141
177
128
163
112
166
114
Study day
380
Patients with event, n (%) 188 (61.4) 150 (48.4)
HR (95% CI) 0.65 (0.52, 0.81)
p value <0.0001
Time to event (25th percentile), days 85 128
T+OT
T+O 5/5 µg
T 5 µg
a
Full analysis set
Figure 3. Kaplan–Meier estimates of probability of decline from
baseline in trough FEV1
of 100 mL, increase from baseline of
4 units in SGRQ score, or moderate or severe exacerbation
(composite end point 2) in patients with GOLD stage B COPD.a
1.00
Estimatedprobabilityofnoevent
0.95
0.90
0.85
0.80
0.75
0.70
0.65
0.60
0.55
0.50
0.45
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
0 50 100 150 200 250 300 350 380
Patients at risk
T+O 5/5 µg
T 5 µg
267
229
310
306
217
179
201
161
175
128
156
112
136
94
143
98
Study day
Patients with event, n (%) 210 (68.6) 174 (56.1)
HR (95% CI) 0.68 (0.56, 0.83)
p value 0.0002
Time to event (25th percentile), days 78 89
T+OT
T+O 5/5 µg
T 5 µg
a
Full analysis set
Figure 1. TONADO®
1 + 2 study designs.
-2 0 2 52 +324a
Weeks
Run-in
R
Treatment
T+O Respimat®
5/5 µg once daily
T+O Respimat®
2.5/5 µg once daily
O Respimat®
5 µg once daily
T Respimat®
5 µg once daily
T Respimat®
2.5 µg once daily
Follow-up
S
C
R
E
E
N
I
N
G
a
Primary end-point assessment
R, randomization
Table 1. Baseline demographic and disease characteristics for
GOLD B patients with COPD: treated set.
T 5 µg T+O 5/5 µg
Patients, n 307 311
Male, n (%) 199 (64.8) 210 (67.5)
Mean ± SD age, years 63.4 ± 9.3 63.8 ± 8.4
Mean ± SD body mass index, kg/m2
26.8 ± 5.4 26.9 ± 5.4
Smoking status, n (%)
Ex-smoker 166 (54.1) 166 (53.4)
Current smoker 141 (45.9) 145 (46.6)
Mean ± SD smoking history, pack-years 45.0 ± 26.0 47.2 ± 23.4
Mean ± SD pre-bronchodilator FEV1
,a
L 1.53 ± 0.47 1.51 ± 0.46
Mean ± SD % of predicted FEV1
a
55.3 ± 10.2 55.4 ± 10.5
Medication use, n (%) 230 (74.9) 238 (76.5)
LAMA 92 (30.0) 113 (36.3)
LABA 121 (39.4) 122 (39.2)
ICS 113 (36.8) 125 (40.2)
a
At screening.
SD, standard deviation; LAMA, long-acting muscarinic antagonist; LABA, long-acting 2
-agonist;
ICS, inhaled corticosteroid.
Table 2. Individual components of composite end points: event rates and time to first event (25th percentile).
Event, n (%)
Time to first event
(25th percentile), days
Time to first event treatment
comparison (T+O – T)
T 5 µg
(n=306)
T+O 5/5 µg
(n=310)
T 5 µg
(n=306)
T+O 5/5 µg
(n=310)
HR
(95% CI) p value
Trough FEV1
decline from baseline 100 mLa
135 (44.1) 103 (33.3)b
91 226b
0.66 (0.51, 0.86) 0.0016
SGRQ score increase from baseline 4 unitsa
95 (32.5)c
73 (24.4)d
175c
369d
0.71 (0.52, 0.96) 0.0271
Moderate or severe exacerbation 69 (22.5) 67 (21.6) NE NE 0.94 (0.67, 1.32) NS
Severe exacerbation 6 (2.0) 10 (3.2) NE NE 1.64 (0.60, 4.51) NS
Death 0 (0.0) 1 (0.3) NE NE – –
a
Only patients with baseline and post-baseline trough FEV1
and SGRQ score values are included; b
n=309; c
n=292; d
n=299.
NE, non-estimable, NS, not significant.