CENTER FOR DRUG EVALUATION AND
RESEARCH
APPLICATION NUMBER:

205641Orig1s000
STATISTICAL REVIEW(S)

 U.S. Department of Health and Human Services
Food and Drug Administration
Center for Drug Evaluation and Research
Office of Translational Science
Office of Biostatistics

Statistical Review
C LINICAL S TUDY
NDA / Sequence Number: NDA 205641 / Seq 0000
Drug Name:

Asmanex® HFA (mometasone furoate dihydride)

Proposed Indication:

Asthma

Applicant:

Review Priority:

Merck
Received:
06-27-2013
PDUFA Due Date: 06-27-2014
Standard

Biometrics Division:

Division of Biometrics II

Statistical Reviewer:

Robert Abugov, Ph.D.

Concurring Reviewer:

Ruthanna Davi, Ph.D. (Acting Team Leader)

Statistics Supervisor:

Thomas Permutt, Ph.D. (Division Director)

Medical Division:

Division of Pulmonary, Allergy, and Rheumatology Products
Kim Witzmann, M.D. (Medical Reviewer)
Anthony Durmowicz, M.D. (Team Leader)
Badrul A Chowdhury, M.D., Ph.D. (Medical Division Director)
Jessica Lee

Date(s):

Clinical Team:

Project Manager:

Keywords: NDA review, Clinical Studies

Reference ID: 3474419

 Contents
1

EXECUTIVE SUMMARY .................................................................................................... 4

2

INTRODUCTION................................................................................................................... 5
2.1

OVERVIEW .......................................................................................................................... 5
2.1.1 Drug Class and Indication ......................................................................................... 5
2.1.2 History of Drug Development .................................................................................... 5
2.1.3 Current Submission.................................................................................................... 6
2.2 DATA SOURCES .................................................................................................................. 7
3

STATISTICAL EVALUATION............................................................................................ 7
3.1
3.2

DATA AND ANALYSIS QUALITY .......................................................................................... 7
EVALUATION OF EFFICACY ................................................................................................. 7
3.2.1 Study Design and Endpoints ...................................................................................... 7
3.2.2 Statistical Methodologies ........................................................................................... 9
3.2.3 Patient Disposition, Demographic and Baseline Characteristics ........................... 11
3.2.4 Results and Conclusions .......................................................................................... 12
3.2.4.1 Study 34: M100 versus Placebo ........................................................................ 13
3.2.4.1.1
3.2.4.1.2
3.2.4.1.3

3.3
4

3.2.4.2 Study 31: M200 .................................................................................................. 16
EVALUATION OF SAFETY .................................................................................................. 16

FINDINGS IN SPECIAL/SUBGROUP POPULATIONS ................................................ 16
4.1
4.2

5

Primary Endpoint in Original Study: ΔFEV1 AUC0-12hr at Week 12 .................................................. 13
Primary Endpoint in Original Study: Exacerbation Incidence to Week 26 ......................................... 13
Exploratory Analyses of Secondary Endpoints .................................................................................. 15

GENDER, RACE, AGE, AND GEOGRAPHIC REGION ............................................................ 16
OTHER SPECIAL/SUBGROUP POPULATIONS....................................................................... 17

SUMMARY AND CONCLUSIONS ................................................................................... 17
5.1
5.2
5.3

STATISTICAL ISSUES ......................................................................................................... 17
COLLECTIVE EVIDENCE .................................................................................................... 18
CONCLUSIONS AND RECOMMENDATIONS ......................................................................... 18

2

Reference ID: 3474419

 Tables
Table 1. Phase 3 Studies in Current Submission. ........................................................................... 6
Table 2. Disposition of Randomized Subjects for Treatments on Proposed Product Label ......... 12
Table 3. ΔFEV1 AUC0-12hr (liter - hours), M100 versus Placebo, at Week 12 ............................. 13
Table 4. Kaplan Meier Incidence of Exacerbations, M100 versus Placebo, Week 26 ................ 13
Table 5. Kaplan Meier Incidence of Clinically Treated Exacerbations, M100 versus Placebo,
Week 26.................................................................................................................................. 14
Table 6. First Exacerbations, Study 34, by Cause ........................................................................ 14
Table 7. Other Secondary Endpoints, Study 34 ............................................................................ 15
Table 8. Change from Baseline Trough FEV1 at Week 12, Study 31 .......................................... 16

3

Reference ID: 3474419

 1 EXECUTIVE SUMMARY
Merck proposes Asmanex HFA metered dose inhaler, mometasone furoate 100 mcg (M100) or
200 mcg (M200) two actuations per twice daily dosing (bid), for the maintenance treatment of
asthma as prophylactic therapy in patients 12 years of age and older. Efficacy and safety of this
inhaled corticosteroid (ICS) were examined in three phase 3 clinical trials.
The submission provides one adequately controlled phase 3 trial to evaluate the efficacy of M100
in asthma patients. That study shows statistically significant differences between M100 and
placebo for change from baseline to week 26 incidence of exacerbations. Further exploratory
analyses also show nominally significant differences between M100 and placebo for Δ trough
FEV1, ΔAQLQ (S), ΔACQ, ΔAM PEFR and Δ proportion of nights awakened requiring use of
short-acting beta agonists (SABA).
The submission also provides a study purporting to demonstrate efficacy of M200. However that
study was conducted without a control arm to distinguish between treatment effects and placebo
effects. An assessment of M200 relative to M100 in the presence of F5 suggested a numerical
difference between the M200/5 and M100/5 groups which was not statistically significant.

4

Reference ID: 3474419

 2 INTRODUCTION
2.1 Overview
2.1.1 Drug Class and Indication
Asmanex is an inhaled corticosteroid proposed for the maintenance treatment of asthma as
prophylactic therapy in patients 12 years of age and older.

2.1.2 History of Drug Development
Asmanex is the mometasone furoate component of Dulera®, a combination product approved
June 22, 2010 for the treatment of asthma in patients 12 years of age and older. Dulera is
comprised of mometasone furoate 100 mcg plus formoterol fumarate dihydrate 5 mcg (M100/F5)
or mometasone furoate 200 mcg plus formoterol fumarate dihydrate 5 mcg (M200/F5),
administered by pressurized metered dose inhaler using a hydrofluoroalkane (HFA-227)
propellent. At the labeling meeting on June 4, 2010, FDA recommended development of a
corresponding mometasone monotherapy, presumably using the same HFA-227 propellant and
metered dose inhaler, for patients at risk for complications associated with the LABA component
formoterol. On June 11, 2010, the sponsor confirmed plans to develop such a monotherapy.
In a preliminary written response on September 9, 2011 to a briefing package submitted by the
sponsor under PIND 112669 as background for a planned Type C meeting, the Agency agreed
that studies completed under the Dulera program and an earlier mometasone monotherapy
program would adequately characterize efficacy and safety for a mometasone monotherapy
NDA. However the Agency anticipated difficulty in describing efficacy on the label because
none of the trials compared mometasone 200 mcg (M200) to placebo (P).
In a written response, requested by the sponsor after cancelling the planned September 2011
Type C meeting, the Agency confirmed on September 14, 2011 that concerns for the NDA
submission would center on labeling rather than approvability, and that inclusion of efficacy data
from an older mometasone formulation would be problematic.

5

Reference ID: 3474419

 2.1.3

Current Submission

To support indications for M100 and M200 monotherapies, the submission included two phase 3,
parallel arm, double-blind trials, P04334 and P04331, hereafter referred to as studies 34 and 31
(Table 1). Study 34 randomized 781 patients 1:1:1:1 to M100/F5 or M100 using an HFA-227
propellant, formoterol fumarate dihydrate 5 mcg using HFA 134 as a propellant (F5), or placebo
(P), and study 31 randomized 728 patients 1:1:1:1 to M200/F5, M100/F5, or M200 using
HFA-227 as a propellant. Designs for these studies are further discussed in section 3.2.1 below.
Study P04334 was conducted from 17 November 2006 to 10 October 2008, and study P04431
was conducted from 17 November 2006 to 10 October 2008.

Table 1. Phase 3 Studies in Current Submission.
Study1

Design

P04334
(Trial 1)

M100/F5
M100
F5
P

Population
Adults and adolescents (≥ 12 yr)
Persistent asthma (≥ 1 yr)
On medium dose ICS
Trough FEV1 at baseline
(≥60%, ≤85% predicted)

Supplemental Claims
Incidence exacerbations W26
ΔTrough FEV1 W12
Nocturnal awakenings to W26
ΔAQLQ(S) W26
AM peak expiratory flow W26

Parallel arm
DB
N=781 1:1:1:1
P to W26
P04431
(Trial 2)

M200/F5
M100/F5
M200
Parallel arm
DB
To W12

Adults and adolescents (≥ 12 yr)
Persistent asthma (≥ 1 yr)
On high dose ICS
Trough FEV1 at baseline
(≥50%, ≤85% predicted)

ΔTrough FEV1 W12
Nocturnal awakenings to W12
AM peak expiratory flow W12
ΔAQLQ(S) W12

At least one exacerbation 2 to 12
months prior to screening

N=728 1:1:1
1, Designation in parentheses corresponds to trial number on proposed label.
DB double blind

6

Reference ID: 3474419

 2.2

Data Sources

Data for both studies was provided by the sponsor in the original Dulera submission and is
currently located at:
\\Cdsesub1\evsprod\NDA022518\0000\m5\datasets .

3 STATISTICAL EVALUATION
3.1

Data and Analysis Quality

Data analysis and quality was adequate in the original Dulera submission. For further details, see
the Biometrics review submitted to DARRTS on May 19, 2010.

3.2

Evaluation of Efficacy

3.2.1 Study Design and Endpoints
Analyzed data was from two double blind, parallel group, multicenter phase 3 efficacy studies
(Table 1) which were conducted on patients with persistent asthma to demonstrate effectiveness
of Dulera, a combination drug containing mometasone and formoterol. Study 34 randomized 781
patients with persistent asthma previously treated with medium doses of ICS among four
treatment arms; M100/F5 (N=191), M 100 (N=192), F 5 (N=202), and placebo (N=196).
Treatment continued for 26 weeks. Study 31 randomized 728 patients with persistent asthma
previously treated with high doses of ICS among three treatment arms; M200/F5 (N=233),
M100/F5 (N=255), and M 200 (N=240). Treatment continued for 12 weeks.
Serial pulmonary function tests (PFTs) were performed for each patient during clinical visits at
baseline, week 1, and the final visit (week 26 of study 31, week 12 of study 34) beginning 30
minutes and immediately before (0 hour) the subject’s morning dose of study medication, and
then at 5 minutes, 15 minutes, 30 minutes, and 1, 2, 3, 4, 6, 8, 10, 11, and 12 hours post-dose.
PFTs at all visits were to be performed prior to the morning dose of study medication, and at
least 12 hours after the previous evening’s dose. For clinical visits at baseline, week 4, and week
8, FEV1, forced expiratory flow rate between 25% and 75% of forced vital capacity (FEFR), and
FVC were measured. Subjects were also instructed to perform triplicate peak expiratory flow
(PEF) measurements twice daily before administration of the study medication. Subjects also
recorded, in daily diaries, short term beta-2 agonist (SABA) and oral prednisone use, number of
nocturnal awakenings requiring SABA use, AM and PM asthma symptom scores, and scores for
the asthma symptoms diary scale after daytime and overnight.
7

Reference ID: 3474419

 In the original Dulera submission, the sponsor planned to conclude efficacy in study 34
according to the following co-primary efficacy endpoints :
1. time-to-first asthma exacerbation over the 26-week treatment period
and
2. FEV1 AUC0-12 hr after 12 weeks of treatment compared to mometasone alone.
For primary endpoint 1 above, an exacerbation was defined as
1.a. decrease in FEV1 (absolute value) below the treatment period stability limit at any
visit during the treatment period, defined as 80% of the average of the two predose FEV1
measurements 0 and 30 minutes prior to the first dose of randomized study medication,
or
1.b. decrease in AM or PM peak flow below the treatment period stability limits on any 2
consecutive days during the treatment period, defined as 70% of the respective mean AM
or PM PEF obtained over the last 7 days immediately prior to receiving the first dose of
randomized study medication,
or
1.c. any clinical deterioration of asthma resulting in emergency treatment, hospitalization
due to asthma, or treatment with additional, excluded asthma medication (other than
SABA) as judged by the clinical investigator.

The primary endpoint in study 31 was change from baseline FEV1 AUC0-12 hr after 12 weeks of
treatment.
In both studies, key secondary endpoints were change from baseline to week 26 in AQLQ(S),
change from baseline to week 26 ACQ total score, and change from baseline proportion of nights
across the treatment period without awakenings due to asthma that required use of short-acting
beta 2-agonists (SABA).
.

8

Reference ID: 3474419

 For studies 31 and 34, other secondary endpoints included:
1. Change from baseline in AM FEV1 pre-dose assessment, or trough FEV1, at each
visit and at study endpoint,
2. Daytime and nighttime SABA usage, including time to first SABA usage,
3. Proportion of subjects with 2 consecutive nights with nocturnal awakenings due to
asthma which require use of SABA rescue medication during the treatment period,
4. Proportion of subjects with at least 2 consecutive days of more than eight inhalations
of SABA or two or more nebulized treatments, during the treatment period,
5, Change from baseline in pulmonary function tests (FEF 25% to 75%, FVC, and %
predicted FEV1) at each visit and endpoint,
6. Change from baseline in AM and PM PEF, AM and PM symptom scores, and daytime
and nocturnal assessments from e-diaries at each week and endpoint (last week of
diary data for each subject),
7. Change from baseline in proportion of days with no symptoms of asthma during the
treatment period, and
8. Change from baseline to endpoint (last week for each subject) in mean number of
nocturnal awakenings due to asthma which required use of SABA.

3.2.2 Statistical Methodologies
Two analysis populations were defined: safety and efficacy. The safety population included all
randomized patients who received double-blind study medication during the trial and/or open
label mometasone during the run-in period, a population often referred to as the modified intentto-treat (mITT) population in other programs. The efficacy population included all individuals in
the safety population managed per protocol. The sponsor’s use of ‘safety’ and ‘efficacy’ for
these populations may be considered misnomers because the primary efficacy statistical analyses
were conducted on the safety population.
For studies 31 and 34, primary analyses for FEV1 and other continuous endpoints were
conducted using analyses of covariance (ANCOVA) with independent factors treatment, study
site, and baseline. For trough FEV1 in study 34, an amendment was implemented prior to
unblinding which specified the primary analysis as a mixed effect repeated measures model
(MMRM) to guage treatment effect at week 12, with fixed effects treatment, time, baseline, and
visit by treatment interaction, and with random effect subject.
Proportions in both studies were analyzed using Cochran-Mantel-Haenzel (CMH) tests of
proportions controlled for study site.
Time to first asthma exacerbation in study 34 was analyzed using the log-rank test.
Results from all analyses were evaluated at the 0.05 level of significance.
9

Reference ID: 3474419

 No new protocols were developed for the study of Asmanex. Therefore, control of type 1 in the
submitted protocols followed that for Dulera.
In study 34, to control overall type 1 error, analyses were conducted in the following hierarchical
order:
1. Week 12 ΔFEV1 AUC0-12hr
a. M100/F5 vs M100
b. F5 vs P
2. Time to first asthma exacerbation over 26 weeks
a. MF 100/5 vs F5
b. M 100 vs P
If all four of the above co-primary endpoint comparisons were significant at the 0.05 level, then
the following four endpoints were to be tested in the following sequense.
3. Week 12 trough ΔFEV1
a. MF 100/5 vs F5 and MF 100/5 vs P ('simultaneously')
4. ΔAQLQ(S) Week 26
a. MF 100/5 vs P
5. ΔACQ Week 26
a. MF 100/5 vs P
6. ΔProportion of nights awakened requiring use of SABA
a. MF 100/5 vs P
Other study 34 endpoints were not controlled for multiplicity, and should therefore be considered
exploratory, including comparisons proposed for the Asmanex label between M100 and P for
Δtrough FEV1, Δproportion of night awakenings requiring use of SABA, ΔAQLQ(S), proportion
of patients with AQLQ(S) response, and ΔPEFR
In study 31, to control overall type 1 error, analyses were conducted in the following hierarchical
order:
1. MF 200/5 vs M 200
a. Week 12 ΔFEV1 AUC0-12hr
b. ΔACQ Week 12
c. ΔAQLQ(S) Week 12
d. ΔProportion of nights awakened requiring use of SABA
10

Reference ID: 3474419

 Other study 31 endpoints were not controlled for multiplicity, and should be considered
exploratory, including all comparisons of M200/5 and M100/5 proposed for description of
efficacy of the M200 mono-component in the Asmanex label.
In both studies 31 and 34, the sponsor imputed missing FEV1 AUC0-12 hr data within visits using
last observation carried forward (LOCF) for patients terminating FEV1 measurements two or
more hours post-dose during a visit. No FEV1 AUC0-12 hr was recorded for visits in which
patients missed more than three consecutive evaluation times or terminated measurement prior to
two hours post-dose. The sponsor replaced missing FEV1 values using linear interpolation of
measurements prior to and following the missing data if patients missed at most three
consecutive evaluation times.
Pulmonary function test results for missing visits was imputed using LOCF.
Baseline for pulmonary function tests were defined as the average of the two pre-dose values 0
and 30 minutes prior to initial dosing. Baseline ACQ and AQLQ were obtained from a single
measurement prior to initial dosing. For proportion of nightime awakenings requiring use of
SABA, asthma symptoms, and PEF, baseline was derived from diary data during the 7 days prior
to initial dosing.
To summarize, except for time to first exacerbation in study 34, all endpoints proposed for
inclusion on the Asmanex product label are exploratory as studies 34 and 31 were designed to
assess the efficacy of the combination product rather than the mometasone furoate component
alone. Significance levels for these comparisons should therefore be considered nominal, with pvalues underestimating the true type 1 error. Further, study 31 did not include a placebo control
arm; a true treatment effect of M200 monotherapy therefore cannot be distinguished from a
placebo effect.

3.2.3 Patient Disposition, Demographic and Baseline Characteristics
The earlier review for Dulera NDA established that there were no obvious differences between
treatment groups in demographic baseline characteristics in these studies. In both studies,
approximately 80% of patients in the mometasone arm and 60% of the patients in the placebo
arm completed treatment to week 26 (Table 2). Patterns of treatment discontinuation in study 34
were consistent with efficacy, with 23% of patients in the placebo arm and 7% of the patients in
the mometasone arm discontinuing due to treatment failure.

11

Reference ID: 3474419

 Table 2. Disposition of Randomized Subjects for Treatments on Proposed Product Label
Disposition
Randomized
Discontinued
Adverse event
Treatment failure
Lost to followup
Withdrew, reasons unrelated
Withdrew, reasonss related
Noncompliance
Did not meet protocol eligibility
Administrative
Completed treatment period

Study 34
M100
P
192
196
(100)
(100)
33 (17) 77 (39)
6 (3)
7 (4)
13 (7)
46 (23)
0 (0)
2 (1)
3 (2)
8 (4)
1 (1)
5 (3)
5 (3)
6 (2)
4 (2)
3 (2)
1 (1)
0 (0)
159
119
(83)
(61)

Study 31
M200
240
(100)
33 (14)
5 (2)
13 (5)
1 (<1)
4 (2)
1 (<1)
3 (1)
5 (2)
1 (<1)
207
(86)

Source: Table 3 in clinical study reports

3.2.4 Results and Conclusions
Sponsor calculations were generally confirmed in this review and in the earlier review for the
Dulera combination product.
For both studies, all comparisons of treatment arms involved post-hoc exploratory analyses, and
the calculated p-values therefore underestimated true type I error rates. As a consequence, results
from all statistical tests should be viewed with skepticism because they were conducted at a
significance level larger than the standard value of 0.05.
Study 31 purported to demonstrate effect of M200 monotherapy by showing that patients treated
with M200 improved from baseline after 12 weeks of treatment. However, because no control
arm was included, the study could not clarify whether observed improvements from baseline
were the result of treatment with M200 rather than the placebo effect. An assessment of M200
relative to M100, only in the presence of F5, is possible in that study because it includes the
M200/F5 and M100/F5 treatment arms.

12

Reference ID: 3474419

 3.2.4.1 Study 34: M100 versus Placebo
3.2.4.1.1 Primary Endpoint in Original Study: ΔFEV1 AUC0-12hr at Week 12
The difference between M100 and P for change from baseline FEV1 AUC0-12hr at Week 12 was
not statistically significant in in study 34 (Table 3). This endpoint was included in the Dulera
submission to evaluate the efficacy of the formoterol component in the combination product by
comparing M/F to M alone.
Table 3. ΔFEV1 AUC0-12hr (liter - hours), M100 versus Placebo, at Week 12
Study
34

FEV1 AUC0-12hr
M100
P
1.30 (169)
0.57 (128)

Difference
M(x) - P
P-Value
0.73
0.140

Source: auc.sas, see also Table 7, clinical study report

3.2.4.1.2 Primary Endpoint in Original Study: Exacerbation Incidence to Week 26
In study 34, the incidence of exacerbations at week 26 differed significantly between
mometasone and placebo (Table 4).
Table 4. Kaplan Meier Incidence of Exacerbations, M100 versus Placebo, Week 26
Study
34

Incidence of Exacerbations
M100
P
0.65
0.41

Difference
M100 – P P-Value
0.23
<.0001

Source: main.sas

The difference between mometasone and placebo for incidence of 'clinically treated
exacerbations' at week 26, exacerbations involving emergency room treatment, hospitalization,
or use of prohibited medications, did not differ significantly between placebo and mometasone
(Table 5). The low rate of such exacerbations may explain the lack of statistical significance.

13

Reference ID: 3474419

 Table 5. Kaplan Meier Incidence of Clinically Treated Exacerbations, M100 versus Placebo,
Week 26
Study

Incidence of Exacerbations
M100
P
0.96
0.94

34

Difference
M100 – P
P-Value
0.02
0.2144

Source: main.sas

The proportion of patients with exacerbations, broken out by cause, is given in (Table 6).

Table 6. First Exacerbations, Study 34, by Cause
Cause
All Exacerbations
FEV1
NONE
PEFR
Prohibited Meds
Hospitalization
Multiple Causes

N (%)
M100
P
65
109
(34)
(56)
19
39
(10)
(20)
127
87
(66)
(44)
41
59
(21)
(30)
5
7
(3)
(4)
0
0
(0)
(0)
0
4
(0)
(2)

source: main.sas

14

Reference ID: 3474419

 3.2.4.1.3 Exploratory Analyses of Secondary Endpoints
Because the protocol did not predefine comparisons between M100 and P, the following analyses
of secondary endpoints, many of which are proposed for inclusion in the label, are only
exploratory. While the measures of statistical significance are therefore only nominal, with
p-values underestimating true probabilities of type 1 error, each of the p-values associated with
these endpoints are highly significant, making it unlikely that there is no true treatment effect.
Differences between M100 and placebo were nominally significant for change from baseline in
trough FEV1 at week 12 (Table 7), and for changes at study endpoint from baseline of AQLQ(S)
AQC, proportion of of nights with nocturnal awakening requiring SABA, and PEFR. Proportion
AQLQ (S) response, defined by the sponsor as ΔAQLQ (S) ≥ 0.5, was also greater among
patients randomized to M100 than among patients randomized to placebo.

Table 7. Other Secondary Endpoints, Study 34
Week
12

Var
Δ Trough FEV1 (mL)

26

ΔAQLQ (S)

26

ΔACQ

26

ΔProportion Noct Awakening

26

ΔPEFR (L/sec)

26

ΔAQLQ (S) Response

M100
71
(190)
0.37
(189)
-0.23
(186)
-0.05
(191)
1.75
(188)
0.42
(189)

P
-52
(192)
-0.01
(189)
0.14
(187)
0.00
(194)
-28.44
(193)
0.23
(189)

Diff
123

P-Value
0.001

0.38

<.0001

-0.37

<.001

-0.05

0.003

30.18

<.001

0.19

<.001

source main.sas, aq.sas,, propn.sas

AQLQ higher values correspond to improvement, ACQ lower values correspond to improvement

15

Reference ID: 3474419

 3.2.4.2 Study 31: M200
Study 31 purported to demonstrate effect of M200 monotherapy by showing that patients treated
with M200 improved from baseline after 12 weeks of treatment. However, because no control
arm was included, it remains unclear whether such improvements were the result of treatment
with M200 rather than the placebo effect.
If and only if there is clinical information that the effect of mometasone is independent of
coadministered formoterol, efficacy of M200 may be estimated by showing superiority of
M200/F5 over M100/F5, and therefore this comparison is provided herein for trough FEV1 at
week 12. There was no statistically significant difference between M200/F5 and M100/F5 for
change from baseline in trough FEV1 at week 12 (Table 8). Similarly, no statistically significant
differences between M200/F5 and M100/F5 were seen for week 12 ΔFEV1 AUC0-12hr,
ΔAQLQ(S), ΔAQC, or Δproportion of nights with awakenings requiring SABA (data not
shown).

Table 8. Change from Baseline Trough FEV1 at Week 12, Study 31
Var
ΔTrough FEV1 (mL)

M200/F5
178
(255)

M100/F5
139
(232)

Diff
39

P-Value (95% CI)
0.182
(-18, 97)

Source: Main study 31.sas

3.3

Evaluation of Safety

Safety evaluations for this submission were conducted by the Medical Reviewer, Kimberly
Witzmann, M.D. and are provided in her review.

4 FINDINGS IN SPECIAL/SUBGROUP POPULATIONS
4.1

Gender, Race, Age, and Geographic Region

Effects of gender, age, and geographic region on treatment efficacy were discussed in the
statistical review of Dulera. Statistical tests involving all treatments revealed no significant
interactions between these subgroups and treatment with mometasone.

16

Reference ID: 3474419

 For this submission, additional subgroup analyses were conducted for statistically significant
results from study 34 by adding subgroup, treatment by subgroup interaction, and treatment by
subgroup by study week interaction terms to the original MMRM for ΔFEV1 using data from
only the placebo and M100 arms. While gender and age were not seen to affect treatment
differences, there was a nominally significant interaction between treatment, week, and region
(p=0.0057), with point estimates of differences between M100 and P for ΔFEV1 at week 12
equal to 222 mL in the USA and 80 mL outside of the USA.
For study 31, similar analyses were conducted with only M200/F5 and M100/F5 data, replacing
site with site nested within country, and with subgroup and subgroup by treatment interactions.
While region and race were not seen to affect treatment differences, there was a nominally
significant (p=0.028) interaction between treatment and age category (≤17 yr, >17 yr and <64,
>64 yr). Point estimates for ΔFEV1 at study endpoint were, for MF 100/5, -26, 125, and 242 mL
respectively for the ≤17 yr, >17 yr and <64, >64 yr age categories, and for MF 200/5 were 254,
158, and 140 mL respectively for the ≤17 yr, >17 yr and <64, >64 yr age categories.
Overall then, results from the subgroup analyses are not consistent between studies, and any
nominally significant interactions may represent type 1 error associated the large number of
comparisons made. The interaction with the smallest nominal p-value, for country by treatment
at week 12 in study 34 for ΔFEV1, was not qualitatively important because numerical
improvements were seen regardless of region.

4.2

Other Special/Subgroup Populations

No other subgroup populations were analyzed for this submission.

5 SUMMARY AND CONCLUSIONS
5.1

Statistical issues

In study 34, except for incidence of exacerbations, comparisons between M100 and P were not
preplanned. Therefore, while tests for statistical significance were only nominal, with the
p-values underestimating true probabilities of type 1 error, each of the p-values associated with
these endpoints was highly significant, making it unlikely that there is no true treatment effect.
Study 31 purported to demonstrate effect of M200 monotherapy by showing that patients treated
with M200 improved from baseline after 12 weeks of treatment. However, because no control
arm was included, it remains unclear whether observed improvements resulted from treatment
with M200 or instead stemmed from the placebo effect. An assessment of M200 relative to
M100, only in the presence of F5, is possible using data from Study 31; however, no significant
difference between M200/F5 and M100/F5 were observed.
17

Reference ID: 3474419

 5.2 Collective evidence
The collective evidence suggests that, compared to placebo, M100 does provide some
improvement for asthma patients. However the evidence is somewhat undermined by lack of
control for type 1 error in the face of multiple unplanned endpoints.
The submission also included a study purporting to directly demonstrate efficacy of M200.
However that study was conducted without a control arm to distinguish between treatment
effects and placebo effects. An assessment of M200 relative to M100 in the presence of F5
suggested a numerical differences between the MF 200/5 and MF 100/5 groups which was not
statistically significant.

5.3

Conclusions and Recommendations

The submission provides one adequately controlled phase 3 trial to evaluate the efficacy of M100
in asthma patients. That study shows statistically significant differences between M100 and
placebo for change from baseline to week 26 incidence of exacerbations. Further exploratory
analyses also show nominally significant differences between M100 and placebo for Δ trough
FEV1, ΔAQLQ (S), ΔACQ, ΔAM PEFR and Δ proportion of nights awakened requiring use of
short-acting beta agonists (SABA).
The submission also provides a study purporting to demonstrate efficacy of M200. However that
study was conducted without a control arm to distinguish between treatment effects and placebo
effects. An assessment of M200 relative to M100 in the presence of F5 suggested a numerical
difference between the M200/5 and M100/5 groups which was not statistically significant.

18

Reference ID: 3474419

 --------------------------------------------------------------------------------------------------------This is a representation of an electronic record that was signed
electronically and this page is the manifestation of the electronic
signature.
--------------------------------------------------------------------------------------------------------/s/
---------------------------------------------------ROBERT ABUGOV
03/20/2014
RUTHANNA C DAVI
03/20/2014

Reference ID: 3474419

 STATISTICS FILING CHECKLIST FOR A NEW 

NDA Number: 205641 Applicant: Merck Stamp Date: 06-27?2013
Drug Name: Asmanex Type: NDA (Standard)

On initial overview of the application for RTF:

 

 

 

Content Parameter Yes No NA Comments
1 Index is suf?cient to locate necessary reports, tables, data, 
etc.
2 [88, ISE, and complete study reports are available 

(including original protocols, subsequent amendments, etc.)

 

3 Safety and ef?cacy were investigated for gender, racial, 
and geriatric subgroups investigated.

 

4 Data sets in EDR are accessible and conform to applicable 
guidances existence of de?nepdf ?le for data sets).

 

 

 

 

 

 

 

 

IS THE STATISTICAL SECTION OF THE APPLICATION 

If the is not ?leable from the statistical perspective, state the reasons and provide
comments to be sent to the Applicant.

Please identify and list any potential review issues to be forwarded to the Applicant for the 74?

 

 

 

 

 

 

 

day letter.

Content Parameter (possible review concerns for 74- Yes No NA Comment
day letter)

Designs utilized are appropriate for the indications requested. 
Endpoints and methods of analysis are speci?ed in the 
protocols/statistical analysis plans.

Interim analyses (if present) were pre?speci?ed in the protocol 
and appropriate adjustments in signi?cance level made.

DSMB meeting minutes and data are available.

Appropriate references for novel statistical methodology (if 
present) are included.

Safety data organized to permit analyses across clinical trials 
in the 

Investigation of effect of dropouts on statistical analyses as 
described by applicant appears adequate.

 

 

 

 

 

 

 

File name: 5_Statistics Filing Checklist for a New 10207

Reference ID: 3364903
Reference ID: 3501120

STATISTICS FILING CHECKLIST FOR A NEW 

Comment:

This application included clinical trials that were already reviewed by Robert Abugov under
NDA 22518. Therefore, a full statistical review is not necessary.

 

David Hoberman 8/28/13
Reviewing Statistician Date
Supervisor/Team Leader Date

File name: 5_Statistics Filing Checklist for a New 

Reference ID: 3364903
Reference ID: 3501120

 

This is a representation of an electronic record that was signed
electronically and this page is the manifestation of the electronic
signature.

 

DAVID HOBERMAN
08/28/2013

Reference ID: 3364903
Reference ID: 3501120