CENTER FOR DRUG EVALUATION AND
`
`RESEARCH
`
`
`
`
`APPLICATION NUMBER:
`
`
`213051Orig1s000
`
`
`STATISTICAL REVIEW(S)
`
`
`
`
`
`
`
`
`

`

`
`
`
`
` U.S. Department of Health and Human Services
` Food and Drug Administration
`
`
` Center for Drug Evaluation and Research
`
` Office of Translational Sciences
`
`
` Office of Biostatistics
`
`
`
`
`
`
` S T A T I S T I C A L R E V I E W A N D E VA L U A T I O N
`
`
` CLIN IC A L STUDIE S
`
`
`
` NDA/BLA:
` Sequence #:
`
`
` Drug Name:
`
` Indication(s):
`
` Applicant:
`
` Date(s):
`
`
`
` Review Priority:
`
`
` Biometrics Division:
`
` Statistical Reviewer:
` Concurring Reviewers:
`
`
`
` Medical Division:
`
` Clinical Team:
`
` NDA 213051
`
`
`0001
`
` Semaglutide
` Type 2 Diabetes Mellitus (T2DM)
`
` Novo Nordisk
`
`
`
`
` Received: March 20, 2019
`Primary Review: August 15, 2019
`
` PDUFA Due Date: September 20, 2019
`
` Expedited
`
` Division of Biometrics II
`
`
`
`
`
`
`
`
`
`
`
`
` Robert Abugov, Ph.D.
`
` Yun Wang, Ph.D. (Team Leader)
`
` Mark Rothmann, Ph.D. (Acting Division Director)
`
`
` Division of Metabolism and Endocrinology Products
` Andrea Lungu, M.D. (Medical Officer)
`
`
`
`
` Mitra Rauschecker, M.D. (Acting Team Leader)
`
` Lisa Yanoff, M.D., Ph.D. (Acting Division Director)
`
`
`
`
`
`
`
`
`
` Project Manager:
`
`
`
`
`
` Peter Franks
`
`
`
` Keywords: NDA Review, Clinical Studies
`
`
`
`
`
`Reference ID: 4469585Reference ID: 4497378
`
`
`
`
`
`

`

`
`Contents
`
`
`
` 1 EXECUTIVE SUMMARY .................................................................................................... 5
`
`
`
`2 INTRODUCTION................................................................................................................... 6
`
`
`
` OVERVIEW.......................................................................................................................... 6
`
`
`
`
`2.1.1 Drug Class and Indication ......................................................................................... 6
`
`
`
`2.1.2 History of Drug Development.................................................................................... 6
`
`
`
`
`2.1.3 Data Sources.............................................................................................................. 6
`
`
`
`
`3 STATISTICAL EVALUATION............................................................................................ 7
`
`
`
` DATA AND ANALYSIS QUALITY.......................................................................................... 7
`
`
`
`
`
`
` EVALUATION OF EFFICACY................................................................................................. 7
`
`
`
`
`
` 3.2.1 Study Design and Endpoints ...................................................................................... 7
`
`
`
`
` 3.2.2 Statistical Methodologies......................................................................................... 10
`
`
`
`
` 3.2.3 Patient Disposition, Demographic and Baseline Characteristics ........................... 14
`
`
`
`
` 3.2.4 Results and Conclusions .......................................................................................... 20
`
`
`
`
` Primary Endpoint: Week 26 Change from Baseline Percent HbA1c................ 20
`
`
`
`
`
`
`
`
`
`Key Secondary Endpoint: Week 26 Change from Baseline Body Weight ......... 21
`
`
`
`
`
`Supportive Endpoint: Week 26 Percent HbA1c Less Than 7% ......................... 23
`
`
`
`
`
`Supportive Endpoint: Week 26 Change from Baseline Fasting Blood Glucose 25
`
`
`
`
`Efficacy Summary .............................................................................................. 25
`
`
`
` EVALUATION OF SAFETY .................................................................................................. 27
`
`
`
`
`
`
`4 FINDINGS IN SPECIAL/SUBGROUP POPULATIONS................................................ 30
`
`
`
`
`5 SUMMARY AND CONCLUSIONS ................................................................................... 42
`
`
`
` STATISTICAL ISSUES ......................................................................................................... 42
`
`
`
`
` COLLECTIVE EVIDENCE .................................................................................................... 42
`
`
`
`
` CONCLUSIONS AND RECOMMENDATIONS ......................................................................... 43
`
`
`
`
` LABELING RECOMMENDATIONS ....................................................................................... 44
`
`
`
`
`
`
`6 APPENDIX: SHRINKAGE ESTIMATES FOR SUBGROUP ANALYSES.................. 45
`
`
`
`
`
`
`
`
`
`
`
`2
`
`
`
`
`
`Reference ID: 4469585Reference ID: 4497378
`
`

`

`
`List of Tables
`
`
`
` Table 1. Randomized Confirmatory Trials Reviewed for Effectiveness........................................ 7
`
`
`
`Table 2. Randomized Confirmatory Trials Omitted from Efficacy Review .................................. 8
`
`
`
`Table 3. Criteria for Provision of Rescue Medications................................................................... 9
`
`
`Table 4. ANCOVA Statistical Models for Primary and Secondary Endpoints ............................ 10
`
`
`
`
`Table 5. Patient Disposition at Week 26, Placebo Controlled Studies 4233 and 4224 ................ 14
`
`
`Table 6. Patient Disposition at Week 26, Active Controlled Studies 4222 and 4223 .................. 15
`
`
`Table 7. Patient Disposition at Week 26, Placebo Controlled Studies 4234 and 4280 ................ 16
`
`
`Table 8. Patient Demographics, Placebo Controlled Studies 4233 and 4224, Full Analysis Set . 17
`
`
`
`Table 9. Patient Demographics, Active Controlled Studies 4222 and 4223, Full Analysis Set ... 18
`
`
`
` Table 10. Patient Demographics, Placebo Controlled Studies 4234 and 4280, Full Analysis Set19
`
`
`
`Table 11. Change from Baseline Percent HbA1c Compared to Placebo, Week 26 ..................... 20
`
`
`Table 12. Change from Baseline Percent HbA1c Compared to Active Controls, Week 26 ......... 21
`
`
` Table 13. Change from Baseline Body Weight Compared to Placebo, Week 26. (Measurement is
`
`
`in kilograms) .......................................................................................................................... 22
`
`
`
`Table 14. Change from Baseline Body Weight Compared to Active Controls, Week 26.
`
`(Measurement is in kilograms)............................................................................................... 22
`
`
`
`Table 15. Raw Proportion of HbA1c < 7%, Week 26 .................................................................. 23
`
`
`
`Table 16. Adjusted Proportion of HbA1c < 7% Compared to Placebo, Week 26 ....................... 24
`
`
`Table 17. Adjusted Proportion of HbA1c < 7% Compared to Active Controls, Week 26 ........... 25
`
`
`Table 18. Change from Baseline Fasting Plasma Glucose Compared to Placebo, Week 26 ....... 26
`
`
`
`Table 19. Change from Baseline Fasting Plasma Glucose Compared to Active Controls, Week 26
`
`
`................................................................................................................................................ 26
`
`
`Table 20. Hypoglycemia Percent Incidence While on Randomized Treatment........................... 27
`
`
`
`
`Table 21. Hypoglycemia Event Rate While on Randomized Treatment...................................... 29
`
`
`
`
`
`Table 22. Sample Sizes for Subgroup Analyses on Change from Baseline HbA1c..................... 39
`
`
`
`
`
`
`
`
`Table 23. Sample Sizes for Subgroup Analyses on Change from Baseline HbA1c, Total for
`
`Placebo-controlled Studies 4224, 42333, 4234, and 4280 ..................................................... 42
`
`
`
`
`
`
`
`
`
`
`3
`
`
`
`
`
`Reference ID: 4469585Reference ID: 4497378
`
`

`

`
`
` List of Figures
`
`
`
` Figure 1. Control of Type 1 Error, Trial 4222 .............................................................................. 12
`
`
`
`Figure 2. Control of Type 1 Error, Trial 4223 .............................................................................. 12
`
`
`
`Figure 3. Control of Type 1 Error, Study 4224............................................................................. 13
`
`
`Figure 4. Control of Type 1 Error, Trials 4233 and 4280 ............................................................. 13
`
`
`Figure 5. Subgroup Analyses, Semaglutide 14 mg vs Sitagliptin, Study 4222 ............................ 32
`
`
`Figure 6. Subgroup Analyses, Semaglutide 14 mg vs Empagliflozin, Study 4223 ...................... 33
`
`
`
`Figure 7. Subgroup Analyses, Semaglutide 14 mg vs Placebo, Study 4224 ................................ 34
`
`
`
`Figure 8. Subgroup Analyses, Semaglutide 14 mg vs Placebo, Study 4233 ................................ 35
`
`
`
`Figure 9. Subgroup Analyses, Semaglutide 14 mg vs Placebo, Study 4234 ................................ 36
`
`
`
`Figure 10. Subgroup Analyses, Semaglutide 14 mg vs Placebo, Study 4280 .............................. 37
`
`
`
`Figure 11. Subgroup Meta-analyses, Semaglutide 14 mg vs Placebo, Studies 4224, 4234, 4233,
`
`and 4280 ................................................................................................................................. 38
`
`
`
`
`
`4
`
`
`
`
`
`Reference ID: 4469585Reference ID: 4497378
`
`

`

`
`
`1 EXECUTIVE SUMMARY
`
`
`
` Six randomized trials demonstrate effectiveness of semaglutide 7 or 14 mg administered as oral
`
`
`
` tablets (po) once daily (qd) for the improvement of glycemic control in patients with type 2
` diabetes mellitus (T2DM). Four of the trials, 4224, 4233, 4234, and 4280, were conducted with a
`
`
`
` placebo control. One of the trials, 4224, included both a placebo and an active (liraglutide
` injection 1.8 mg qd) control, and two trials, 4222 and 4223, were conducted using respective
`
`
` active controls empagliflozin 25 mg po qd and sitagliptin 100 mg po qd.
`
`
`Compared to the control, five of the six trials showed statistically significant effects of 7 and/or
`
`14 mg semaglutide for the primary endpoint, week 26 change from baseline HbA1c. In trial
`
`
`4224, however, no significant difference was seen between semaglutide 14 mg and liraglutide
`
`
`
`1.8 mg, another drug in the same class. Of the five remaining trials testing statistical efficacy for
`
`the key secondary endpoint, week 26 change from baseline body weight, five showed
`
`
`statistically significant effects of both 7 and 14 mg semaglutide against control, and one trial
`
`
`4223 showed superiority of the 14 mg, but not the 7 mg dose, to placebo. Supportive endpoints,
`
`
`proportion of patients with HbA1c less than 7%, and week 26 change from baseline fasting
`
`plasma glucose, all trended in the direction expected if semaglutide is efficacious. Please refer to
`Section 3.2.4 for detailed efficacy results.
`
`
`
`
`No major statistical issues have been identified in this submission. The collective evidence from
`
`
`primary, secondary, and supportive endpoints in these studies consistently supports effectiveness
`
`
`
`
`of oral semaglutide for the treatment of type 2 diabetes. Safety risks associated with use of oral
`
`
`
`semaglutide are acceptable. Therefore, I recommend approval of oral semaglutide for glycemic
`
`
`control in patients with type 2 diabetes mellitus. See Section 5.3 for further details.
`
`
`
`
`
`5
`
`
`
`
`
`Reference ID: 4469585Reference ID: 4497378
`
`

`

`
`
` 2
`
`
`
` INTRODUCTION
`
`Overview
`
`
` 2.1.1 Drug Class and Indication
`
`
`
`
`
`
`
`
`Semaglutide is a glucagon-like peptide-1 receptor agonist (GLP-1 RA) proposed as an adjunct to
`
`diet and exercise to improve glycemic control in adults with T2DM. It is a GLP-1 analogue
`
`
`modified for resistance to metabolic degradation through albumin binding and additionally
`
`
`
`modified for resistance to enzymatic degradation by dipeptidyl peptidase-4 (DPP-4). Tablets for
`
`
`
`
`
`oral administration are co-formulated with the absorption enhancer salcaprozate sodium (aka
`
`
`
`sodium N-8-[(2-hydroxybenzoyl) amino] caprylate, abbreviated as SNAC).
`
`
`
` 2.1.2 History of Drug Development
`
`
`
`
` Semaglutide for subcutaneous injection was originally approved on December 5, 2017 as an
`
`
`
` adjunct to diet and exercise to improve glycemic control in adults with T2DM. The present
` submission proposes this same indication for an orally administered version of this product.
`
`
`
`
`
` IND 114464 for semaglutide po was opened on September 9, 2013. In minutes from the
`
`
` pre-filing meeting, submitted to DARRTS on December 28, 2018, the applicant agreed to use of
`
`
`
`
` the treatment policy estimand for all primary analyses of efficacy. The applicant further agreed
` that analyses for the Summary of Clinical Efficacy and Integrated Summary of Effectiveness
`
`
`
` would be conducted without pooling of data across trials. In addition, the Agency agreed that the
` Integrated Study of Safety need only evaluate studies on oral semaglutide, with a phase 3a pool,
`
`
` placebo pool, and a placebo-dose pool. To facilitate separate analyses, FDA requested that
`
` studies conducted only in Japan be flagged in the safety datasets.
`
`FDA further requested that the NDA include the analysis programs, each indexed with a table
`
` including the input datasets, macros used, and the output file names, and that core variable names
` be retained in the analysis datasets across the different studies in the submission.
`
`
`
`
`
`
`
`
`
`
`
` 2.1.3 Data Sources
`
`
`
`
`
`
`
`
`
`
` Data sources for the current review are located at
`
`\\cdsesub1\evsprod\NDA213051\0001\m5\datasets .
`
`
`6
`
`
`
`
`
`Reference ID: 4469585Reference ID: 4497378
`
`

`

` 3 STATISTICAL EVALUATION
`
`
`
`
`
`
`
`
`Data and Analysis Quality
`
`
`
`
`
`
`
` Data and analysis programs provided by the applicant were consistently well organized.
` Adequacy of trial design and analyses are further discussed in this review.
`
`
`
`
`
`
` Evaluation of Efficacy
`
`
`
`
` 3.2.1 Study Design and Endpoints
`
`
`
`
`
`
`This review will focus on six randomized controlled confirmatory trials conducted in adult
`
`
`
`T2DM patients with a variety of background medications and concentrations of baseline
`
`hemoglobin A1c (Table 1).
`
`
`Table 1. Randomized Confirmatory Trials Reviewed for Effectiveness
` Inclusion
` Trial Background
`
`
`Blinding
`Treatments
`
`
`
`
` HbA1c
`
` 7.0-9.5% DB
`
`
`
` Met±SU
`
` SU±Met
`
` BasIns
`
` BasIns±Met
`
` Ins±Met
`
`Japan: Met
`
` w/BasIns only
`
`
` source: reviewer
` Met metformin, SU sulfonylurea, SGLT2i sodium-glucose co-transporter 2 inhibitor, BasIns basal insulin, Ins basal
`
`
`
`
`
`
`
`
`
` with or without bolus, or premixed insulin, OAD oral anti-diabetic drug using any of the above therapies, plus
` thiazolidinedione with or without metformin, renal imp moderate renal impairment, DB double-blind, OL open-
`
`
`
`
`
` label, DD double-dummy, S3 S7 S14 semaglutide maintenance doses 3 mg, 7 mg, or 14 mg qd po, Pbo placebo,
`
`
` DPP-4i dipeptidyl peptidase-4 inhibitor, Si sitagliptin 100 mg qd po, E empagliflozin 25 mg qd po, L1.8 liraglutide
`
`
`
`
`
` injection qd escalated to 1.8 mg, DPP-4i (Si)
`7
`
`
`
` S3
`
` S7
`S14
`
` Pbo
`
` S14
` SGLT2i (E)
`
` S3
`
` S7
`S14
` DPP-4i (Si)
`
` S14
` GLP1 RA (L1.8)
`
` Pbo
`
` S14
`Pbo
`
`
`
`
`
`
`
`
`
` N
`
` 175
`
`175
`175
`
` 178
`
` 411
`
` 410
`
` 466
`465
`465
`
` 467
`
` 284
`283
`
` 142
`
` 163
`161
`
`
` S3
`
` S7
`S14
`
` Pbo
`
` 184
`
`182
`181
`
` 184
`
`
` Duration
`
` (Weeks)
`
` 26
`
`
`
` 52
`
`
` 78
`
`
`
`
`
` 52
`
`
`
` 26
`
`
`
` 52
`
`
`
`
`
`P1
` (4233)
`
`
`
`
` P2
`
` (4223)
`
`
` P3
` (4222)
`
`
`
` P4
`
` (4224)
`
`
`
`
` Met
`
` Met
`
` SU±Met
`
`
`
`
`
`
`
` 7.0-10.5% OL
`
`
`
` 7.0-10.5% DB
`
`
`
`
`
` Met±SGLT2i
`
`
`
` 7.0-9.5% DB
`
` DD
`
`
` 7.0-9.5%
`
` renal imp
`
`
`
`
` DB
`
`
`
`
` 7.0-9.5% DB
`
`
`
` P5
`
` (4234)
`
`
` P8
`
` (4280)
`
`
`
`
`Reference ID: 4469585Reference ID: 4497378
`
`

`

`
`
`
`
`
`
`DB:
`
` Pbo, S
`
` not dose
`
`
` S14
`
`Pbo
`
`
`
`
`
` N Duration
`
` (Weeks)
`
`581
`
` 1591
`
`1592
`
`
`
`
`
` Trials omitted from efficacy evaluation in the present review (Table 2) include trial 4221 for
`
` cardiovascular outcomes, trial 4257 in which ramped dosing occurred at eight-week intervals,
`
`
` and two randomized trials, 4281 and 4282, conducted solely in Japan.
`
`
`
`
`Table 2. Randomized Confirmatory Trials Omitted from Efficacy Review
`Blinding Treatments
` Trial Background
`
`
` Inclusion
`
`
`
`
` HbA1c
`
` SOC - Incretins not
` applicable
`
` CVOT
`
` 7.5-9.5%
`
` P6
`
` (4221)
`
`
`
` P7
`
` (4257)
`
`
` P9
` (4281)
`
`
` Japan
`
`
`
`
`
` All
`
`
`
`
`
` OL
`
`
`
` Monotherapy
`
`
`
` Met+other
`
`
`
`DB:
` on OAD
`
`
` Pbo, S
` 6.5-9.5%
`
`
`
`
` OL:
`
` no OAD
` L .9
`
`
` 7.0-10%
`
` 7.0-10.5% OL
`
`
`
`
`
` S_flex
`
`
` DPP-4i (Si)
`
` S3
`
` S7
`S14
` GLP1 RA (L.9)
`
` Pbo
`
` S3
`
` S7
`S14
` GLP1 RA (D)
`
`
`
`
`
`
`
` 253
`
` 251
`
` 49
`49
`49
`48
`
` 49
` 131
`
`132
`130
`
` 165
`
`
`
` 52
`
`
` 52
`
`
`
`
` 52
`
`
`
`
`
`
` P10
`
` (4282)
`
`
` Japan
`
` source: reviewer
`
` S_flex semaglutide flexible dose ramped at eight week intervals, L.9 liraglutide escalated to .9 mg, D dulaglutide
`
`
`
`
` 0.75 mg q7d, , CVOT cardiovascular outcomes trial
` 1 Mean time on trial product in event driven cardiovascular outcomes trial
`
`
`For subjects randomized to semaglutide, the semaglutide dose was escalated every four weeks,
`
` from 3 to 7 to 14 mg, until the randomized maintenance dose was reached.
`
` The primary efficacy endpoint in all of the trials was week 26 change from baseline percent
`
`
`
`
`
` concentration of HbA1c (ΔHbA1c). The secondary efficacy endpoint was week 26 change from
` baseline body weight (ΔBW). Endpoints proposed for inclusion on the product label, but not
`
`included in the analysis hierarchy, included descriptive week 26 proportion of individuals with
`
`
`
`
`HbA1c < 7% and week 26 change from baseline fasting plasma glucose (ΔFPG).
`
` During the trials, patients with persistent and unacceptable hyperglycemia were offered
`
`
`
`
`
`
` intensifications of treatment as rescue (Table 3) according to local or American Diabetes
` Association/ European Association for the Study of Diabetes guidelines. GLP-1 receptor
`
`
` agonists, DPP-4 inhibitors and amylin analogues were not allowed as rescue medications.
`
`
`
`
`
`
`
`
`
`
`
`
`8
`
`
`
`
`
`Reference ID: 4469585Reference ID: 4497378
`
`

`

`Table 3. Criteria for Provision of Rescue Medications
`
`
` Rescue Criteria
` Time
`Trial
`
`
`
`
`
`
`
` Investigator discretion W8 to end-of-trial
`
`
`
`P1
`
`
` (4233)
`
` P2
`
` (4223)
`
` P3
`
` (4222)
`
`
` P4
`
` (4224)
`
`
` P5
`
`
` (4234)
`
` P8
` (4280)
`
`
` Investigator discretion W8 to end-of-trial
`
`
`
`
`
`
`
`
`
` W8 to W13
`
`
`
` FPG>14.4
` W14 to W25
`
` FPG>13.3
`
`
` FPG>11.1
`
` W26 to end-of-trial
` HbA1c>8.5%
`
` W26 to end-of-trial
`
` FPG>13.3
`
`
` W8 to W13
`
` FPG>11.1
`
` W14 to end-of-trial
` HbA1c>8.5%
`
` W26 to end-of-trial
`
`
` Investigator discretion W12 to end-of-trial
`
`
`
`
`
` HbA1c>8.5%
`
`
`
` W26 to end-of-trial
`
` source: reviewer
`
` FPG fasting plasma glucose mmol/L, W week
`
`
`
`
`
`
`
`
`
`
`
`
`
`9
`
`
`
`
`
`Reference ID: 4469585Reference ID: 4497378
`
`

`

`
`
` 3.2.2 Statistical Methodologies
`
`
`
`
`
`
`
`
`
` Week 26 change from baseline
`
`
`
`
`
`
`
`
`
`The primary and secondary endpoints were analyzed using ANCOVA with dependent variable
`change from baseline and with independent variables treatment, region, baseline value, and
`
`
`
` stratification factors used in the randomization (Table 4). Adjusted means were calculated at
`
` mean baseline values, and used coefficients proportional to the frequencies in the data of the
`
`
`
`
`
` various categorical variables.
`
`
`
`
`Table 4. ANCOVA Statistical Models for Primary and Secondary Endpoints
` Independent Variables
`Trial
`
`
` Dependent Variable
`
`
`
` Week 26 change from baseline
`
`
` treatment
`P1
`
`region1
` (4233)
`
` baseline
`
`
` treatment
`
`region1
` baseline
`
`
` treatment
`
`region1
` baseline
`
`background medication
`
`descent1
`
` treatment
`
`region1
` baseline
`
`background medication
`
`descent2
`
` treatment
`
`region1
` baseline
`
`background medication
`
`renal function3
`
`interaction4
`
` treatment
`
`region1
` baseline
`
`background medication5
`
`descent2
`
`interaction4
`
` P2
`
` (4223)
`
`
` P3
`
` (4222)
`
`
` P4
`
` (4224)
`
`
` P5
`
` (4234)
`
`
` P8
`
` (4280)
`
`
`
`
` Week 26 change from baseline
`
`
`
`
`
` Week 26 change from baseline
`
`
`
`
`
` Week 26 change from baseline
`
`
`
`
`
` Week 26 change from baseline
`
`
`
`
`
` source: reviewer
`
`
`
` 1 Categories include North America, South America, Africa, Asia, Europe,
`
` 2 Japanese vs non-Japanese
`
`
`
` 3 eGFR 30-44 or 45-59 mL/min/1.73 m2
`
`
` 4 interaction between stratification factors
`
`
` 5 (metformin, no metformin) and (basal insulin, basal + bolus insulin, premixed insulin)
`
`
`
`
`
`
`
`
`10
`
`
`
`
`
`Reference ID: 4469585Reference ID: 4497378
`
`

`

`The treatment policy estimand was used to evaluate efficacy, with missing data at week 26
`
`imputed using pattern-mixture models. For each study, imputation groups were defined
`
`
`according to randomized treatment and treatment status (remain on treatment without rescue vs
`
`discontinued treatment or rescued). Then, for each of these groups, missing data were imputed
`
`using a regression based on non-missing data collected from other individuals in the same group,
`
`
`
`with dependent variable change from baseline and independent variables as specified in Table 4
`
`
`
`
`
`above, with the exception of treatment, which not included since it was used to define each
`
`group, and region, which was defined as North American vs not North American. The regression
`
`results, with associated covariances, were used to simulate 1000 complete datasets, and each of
`
`these datasets was analyzed using the primary ANCOVA analysis, with the 1000 analysis results
`
`
`
`
`combined according to Rubin's rule.
`
`
`To evaluate the sensitivity of statistically significant results to the missing-at-random
`
`
`
`pattern-mixture imputations described immediately above, tipping point analysis were conducted
`
`for week 26 change from baseline percent HbA1c and change from baseline body weight. The
`
`
`tipping point analyses imposed penalties which reduced the treatment effect in imputed data, and
`the penalties were increased until the primary analyses combining results from the 1000
`
`
`
`
`simulated datasets 'tipped,' i.e. failed to reject the null hypothesis. The applicant then discussed
`
`
`
`the plausibility of imposed penalties at these tipping points.
`
`
`As implied by the treatment policy estimand, efficacy was analyzed using the full analysis set
`
`
`(FAS), comprised of all randomized patients regardless of initiation of correct treatment, rescue,
`
`dropout, or protocol violations.
`
`
`To control the probability of overall type 1 error at the two-sided .05 level of significance,
`
`
`
`weighted Bonferroni-based closed testing procedures were used. In all trials evaluating
`
`
`non-inferiority hypotheses, upper 95% confidence limits were compared to a 0.4 margin. For
`
`
`trials with multiple doses of semaglutide, results from the 14 mg dose were tested first, followed
`
`
`
`by results from the 7 mg and 3 mg doses (Figure 1, Figure 4). For single dose study 4223, after
`
`
`
`
`testing HbA1c for non-inferiority, alpha was allocated equally to tests on HbA1c and body
`
`weight (Figure 2). In study 4224, the primary and secondary endpoints were tested sequentially
`
`
`
`against placebo and then tested against liraglutide (Figure 3).
`
`
`
`
`For study 4234, the superiority test comparing S14 to placebo for change from baseline HbA1c
`
`
`was followed by the superiority test for change from baseline body weight, each tested at the .05
`
`
`level of significance.
`
`
`
`
`11
`
`
`
`
`
`Reference ID: 4469585Reference ID: 4497378
`
`

`

`Figure 1. Control of Type 1 Error, Trial 4222
`
`
`
`
`
` source: CSR studies P3
`
`
`
`
`Figure 2. Control of Type 1 Error, Trial 4223
`
`
`
`
`
`
`
`
` source: Figure 2-1 of SAP
`
`
`
`
`
`
`
`
`
`12
`
`
`
`
`
`Reference ID: 4469585
`Reference ID: 4497378
`
`
`

`

`Figure 3. Control of Type 1 Error, Study 4224
`
`
`
`
`
` source: Figure 17-1 of protocol
`
`
`Figure 4. Control of Type 1 Error, Trials 4233 and 4280
`
`
`
`
`
`
`
`
`
`
` source: CSR studies P1 and P8
`
`
`
`
`
`
`13
`
`
`
`
`
`Reference ID: 4469585
`Reference ID: 4497378
`
`
`

`

`
`
` 3.2.3 Patient Disposition, Demographic and Baseline Characteristics
`
`
`
`
`
`The number and percent of patients with missing HbA1c at week 26 was consistently less than
`10% (Table 5, Table 6, and Table 7)1. In all studies, semaglutide exhibited a numerically
`
`
`
`
`monotonic dose-response trend for treatment discontinuation due to adverse events.
`
`
`
`
`Table 5. Patient Disposition at Week 26, Placebo Controlled Studies 4233 and 4224
`
` 4224
`
` Trial
` 4233
`
`
` S3
`
` Lira
`
`
` Pbo
` S7
`
` 175 178
`175
`284
`
`
` (100)
` (100)
` (100)
`
`
`
` (100)
`18
`12
`19
`28
` (10.3)
`
`
` (6.9)
`
` (10.7)
`
` (9.9)
`
`7
`4
`4
`21
` (4)
`
`
` (2.3)
`
` (2.2)
`
` (7.4)
`1
`2
`0
`0
` (0.6)
`
`
` (1.1)
`
` (0)
`
` (0)
`
`5
`0
`3
`3
` (2.9)
`
`
` (0)
`
` (1.7)
`
` (1.1)
`5
`6
`12
`4
` (2.9)
`
`
` (3.4)
`
` (6.7)
`
` (1.4)
`4
`13
`27
`9
` (2.3)
`
`
` (7.4)
`
` (15.2)
`
` (3.2)
`0
`0
`0
`2
`
`
` (0)
` (0)
`
` (0)
`
` (0.7)
`
`
` 160 167
`168
`272
`
` (91.4)
` (95.4)
`
`
` (94.4)
`
` (95.8)
`15
`8
`10
`12
` (8.6)
`
`
` (4.6)
`
` (5.6)
`
` (4.2)
`8
`16
`34
`26
` (4.6)
`
`
` (9.1)
`
` (19.1)
`
` (9.2)
`
`
`
`
` Randomized (FAS)
`
`
`
` Discontinued Treatment
`
`
`
` Adverse Event
`
`
`
` Protocol Violation
`
`
`
` Study Withdrawal
`
`
`
` Other
`
`
`
` Rescue Medication
`
`
`
` Death
`
`
`
` HbA1c Measured
`
`
`
` HbA1c Not Measured
`
`
`
` Retrieved Data
`
`
`
` source: disposit.sas
`
` S14
`
`175
` (100)
`
`24
` (13.7)
`
`13
` (7.4)
`
`0
`
` (0)
`5
`
` (2.9)
`6
`
` (3.4)
`2
`
` (1.1)
`1
`
` (0.6)
`160
` (91.4)
`
`15
` (8.6)
`
`11
` (6.3)
`
`
` S14
`
`285
`
` (100)
`38
`
` (13.3)
`30
`
` (10.5)
`1
`
` (0.4)
`2
`
` (0.7)
`5
`
` (1.8)
`10
`
` (3.5)
`2
`
` (0.7)
`278
`
` (97.5)
`7
`
` (2.5)
`41
`
` (14.4)
`
`
`
` Pbo
`
`142
` (100)
`
`14
`
` (9.9)
`6
`
` (4.2)
`0
`
` (0)
`2
`
` (1.4)
`6
`
` (4.2)
`11
`
` (7.7)
`1
`
` (0.7)
`134
` (94.4)
`
`8
`
` (5.6)
`17
`
` (12)
`
`
`
` 1 Disposition given as number of patients and, in parentheses, percent of full analysis set
`
`
`
`14
`
`
`
`
`
`
`Reference ID: 4469585Reference ID: 4497378
`
`

`

`Table 6. Patient Disposition at Week 26, Active Controlled Studies 4222 and 4223
`
`
`
` Trial
` 4223
`
` 4222
` S14 Empa
`
`
` S3
`
`
` S7
` Sita
`
`465
`411
`466
`410
`467
` (100)
`
`
` (100)
`
` (100)
`
` (100)
` (100)
`
`
`44
`48
`32
`53
`28
` (12.9)
` (9.5)
`
`
` (10.3)
`
` (6.9)
`
`
` (6.8)
`20
`15
`13
`36
`10
` (4.3)
`
`
` (3.2)
`
` (2.8)
`
` (8.8)
`
` (2.4)
`5
`4
`4
`0
`3
` (1.1)
`
`
` (0.9)
`
` (0.9)
`
` (0)
`
` (0.7)
`0
`1
`0
`0
`0
` (0)
`
`
` (0.2)
`
` (0)
`
` (0)
`
` (0)
`3
`11
`1
`5
`6
` (0.6)
`
`
` (2.4)
`
` (0.2)
`
` (1.2)
`
` (1.5)
`16
`17
`14
`12
`9
`
` (3.4)
`
`
` (3.6)
` (3)
`
` (2.9)
`
` (2.2)
`13
`11
`25
`8
`5
` (2.4)
`
`
` (5.4)
`
` (2.8)
`
` (1.9)
`
` (1.2)
`
`2
`1
`0
`0
` 0
` (0)
` (0.4)
`
`
` (0.2)
`
` (0)
`
` (0)
`
`438
`435
`446
`392
`395
` (94.2)
`
`
` (93.3)
`
` (95.5)
`
` (95.4)
` (96.3)
`
`27
`31
`21
`19
`15
` (5.8)
`
`
` (6.7)
`
` (4.5)
`
` (4.6)
`
` (3.7)
`31
`43
`25
`44
`18
` (6.7)
`
`
` (9.2)
`
` (5.4)
`
` (10.7)
`
` (4.4)
`
`
`
` S14
`
`465
` (100)
`
`58
` (12.5)
`
`39
` (8.4)
`
`3
`
` (0.6)
`0
`
` (0)
`4
`
` (0.9)
`12
` (2.6)
`
`5
`
` (1.1)
`0
`
` (0)
`436
` (93.8)
`
`29
` (6.2)
`
`38
` (8.2)
`
`
`
`
` Randomized (FAS)
`
`
`
` Discontinued Treatment
`
`
`
` Adverse Event
`
`
`
` Protocol Violation
`
`
`
` Pregnancy
`
`
`
` Study Withdrawal
`
`
`
` Other
`
`
`
` Rescue Medication
`
`
`
` Death
`
`
`
` HbA1c Measured
`
`
`
` HbA1c Not Measured
`
`
`
` Retrieved Data
`
`
`
` source: disposit.sas
`
`
`
`
`15
`
`
`
`
`
`Reference ID: 4469585Reference ID: 4497378
`
`

`

`
`
`
`
` Randomized (FAS)
`
`
`
` Discontinued Treatment
`
`
`
` Adverse Event
`
`
`
` Pregnancy
`
`
`
` Protocol Violation
`
`
`
` Study Withdrawal
`
`
`
` Other
`
`
` Table 7. Patient Disposition at Week 26, Placebo Controlled Studies 4234 and 4280
`
` Trial
`
`
` 4234
` 4280
` Pbo
` S3
`
`
` S14
`
` S7
`
` S14
`
`163
`182
`161
`184
`181
` (100)
`
` (100)
`
`
` (100)
`
` (100)
`
` (100)
`29
`20
`24
`15
`
`31
` (17.8)
`
` (12.4)
`
`
` (13.2)
`
` (8.2)
` (17.1)
`
`24
`10
`12
`9
`22
` (14.7)
`
` (6.2)
`
`
` (6.6)
`
` (4.9)
`
` (12.2)
`
`0
`0
`5
`2
`2
`
` (2.7)
` (1.1)
`
` (0)
` (0)
`
`
`
` (1.1)
`1
`3
`1
`0
`0
`
` (0.6)
` (1.9)
`
`
` (0.5)
`
` (0)
`
` (0)
`0
`2
`2
`0
`2
`
` (0)
` (1.2)
`
`
` (1.1)
`
` (0)
`
` (1.1)
`4
`5
`4
`4
`5
`
` (2.5)
` (3.1)
`
`
` (2.2)
`
` (2.2)
`
` (2.8)
`7
`16
`2
`5
`4
`
` (4.3)
` (9.9)
`
`
` (1.1)
`
` (2.7)
`
` (2.2)
`1
`2
`0
`0
`
`2
`
` (0.6)
` (1.2)
`
`
` (0)
`
` (0)
` (1.1)
`
`154
`155
`174
`176
`173
` (94.5)
`
` (96.3)
`
`
` (95.6)
`
` (95.6)
`
` (95.7)
`
`9
`6
`8
`8
`8
` (4.4)
`
` (5.5)
` (3.7)
`
`
` (4.4)
`
`
` (4.3)
`28
`28
`27
`18
`14
` (17.2)
`
` (17.4)
`
`
` (14.9)
`
` (9.9)
`
` (7.6)
`
`
`
` Pbo
`
`184
` (100)
`
`13
`
` (7.1)
`3
`
` (1.6)
`2
`
` (1.1)
`0
`
` (0)
`2
`
` (1.1)
`6
`
` (3.3)
`9
`
` (4.9)
`0
`
` (0)
`176
` (95.7)
`
`8
`
` (4.3)
`13
`
` (7.1)
`
`
`
` Rescue Medication
`
`
`
` Death
`
`
`
` HbA1c Measured
`
`
`
` HbA1c Not Measured
`
`
`
` Retrieved Data
`
`
`
` source: disposit.sas
`
`
`
`
`
`
`16
`
`
`
`
`
`Reference ID: 4469585Reference ID: 4497378
`
`

`

`
`
`
`
` Randomized (FAS)
`
`
`
` Sex Female
`
`
`
` N (%)
`
`
`
`
` Age mean
`
`
` range
`
` < 65 N (%)
`
`
`
`
`
` ≥ 65 N (%)
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`There were no obvious differences between treatments for baseline characteristics in the
`submitted studies (Table 8, Table 9, Table 10).
`
`
` Table 8. Patient Demographics, Placebo Controlled Studies 4233 and 4224, Full Analysis Set
`
` 4224
`
` Trial
`
`4233
` Pbo
`
` S14
`
` Pbo
`
` S3
`
` Lira
`
` S7
`
` S14
`
`142
`285
`178
`284
`175
`175
`175
` (100)
`
`
` (100)
`
` (100)
`
` (100)
`
` (100)
`
` (100)
` (100)
`
`68
`138
`89
`86
`135
`82
`89
`
` (47.9)
`
` (48.4)
`
` (49.1)
` (46.9)
` (50.9)
`
`
` (50)
`
` (47.5)
`
` 56.5
`
` 56.2
`
`
`
`
` 53.8
`
` 54
` 55.5
` 54.9
` 56.4
` 29-83
`
` 27-81
`
`
`
` 22-78 31-79 23-84
`
` 28-79
` 35-78
`
`
`109
`232
`
`144
`135
`136
` 145
`220
` (76.8)
`
`
` (81.4)
`
` (77.7)
` (81.5)
` (82.3)
`
` (77.1)
`
`
` (77.5)
`33
`53
`31
`40
`39
`33
`64
`
` (23.2)
`
` (18.6)
` (17.7)
`
` (22.9)
`
` (22.3)
`
` (18.5)
`
` (22.5)
`99
`208
`130
`131
`135
`132
`212
`
` (69.7)
`
` (73)
` (74.3)
`
` (74.9)
`
` (77.1)
`
` (74.2)
`
` (74.6)
`19
`39
`
`29
`30
`31
`31
`36
`
` (13.4)
`
`
` (17.4)
` (13.7)
` (16.6)
`
` (17.1)
`
` (17.7)
`
` (12.7)
`8
`10
`11
`6
`10
`12
`9
`
` (5.6)
` (5.7)
`
`
` (6.3)
`
` (3.4)
`
` (5.6)
`
` (4.2)
`
` (3.2)
`1
`
`1
`1
`1
`1
` 0
`1
`
` (0.7)
`
` (0)
` (0.6)
`
` (0.6)
`
` (0.6)
`
` (0.6)
`
` (0.4)
`
` 0
`
`
`
`
`1
` 0
` 0
` 0
` 0
`1
` (0)
`
` (0)
` (0)
` (0)
` (0)
` (0.6)
`
`
`
`
`
` (0.4)
`12
`23
`4
`2
`2
`4
`17
`
` (8.5)
` (2.3)
`
` (1.1)
`
` (1.1)
`
` (2.2)
`
` (8.1)
`
` (6.0)
`
`3
`46
`31
`52
`51
`3
`8
`
` (2.1)
` (26.3)
`
` (17.7)
`
` (29.7)
`
` (28.7)
`
` (1.1)
`
` (2.8)
`5
`122
`133
`116
`121
`17
`18
`
`
`
`
` (69.7)
`
` (76)
`
` (66.3)
` (68)
` (6)
`
` (6.3)
` (3.5)
`137
`268
`
` 266
`7
`11
`7
`6
`
`
` (93.7)
` (96.5)
`
` (4)
`
` (6.3)
`
` (4)
`
` (3.4)
` (94)
`
`47
`53
`53
`60
`75
`76
`28
` (26.9)
`
` (30.3)
`
` (30.3)
`
` (33.7)
`
` (26.3)
`
` (26.8)
`
` (19.7)
`
`
`
`
`
`
`
`
`
` Race White N (%)
`
`
`
`
`
` Asian
`
`
`
` African American
`
`
`
`
`
` Native American
`
`
`
`
`
` Pacific Islander
`
`
`
`
`
` Other
`
`
`
`
`
` Ethnicity Hispanic
`
`
`
` N (%)
`
`
`
`
`
`
`
` Not Hispanic
`
`
`
`
`
` N/A
`
`
`
` Country USA N (%)
`
`
`
` source: disposit.sas
`
`
`
`
`17
`
`
`
`
`
`Reference ID: 4469585Reference ID: 4497378
`
`

`

`
`
` Randomized (FAS)
`
`
`
`
`
` Sex Female N (%)
`
`
`
`
` Age mean
`
`
` range
`
` < 65 N (%)
`
`
`
`
`
`
`
`
`
`Table 9. Patient Demographics, Active Controlled Studies 4222 and 4223, Full Analysis Set
`
`
` Trial
`
` 4222
`
`4223
` S3
` S14 Emp
`
`
`
` S7
`
` Sita
`411
`466
`410
`465
`467
`
` (100)
`
` (100)
` (100)
`
`
` (100)
`
` (100)
`212
`205
`201
`220
`229
`
` (49)
`
` (49.9)
`
` (49)
`
` (45.5)
`
` (47.3)
`
` 57.8
`
`
` 58.2
`
`
` 57.7
` 58.2
` 57.4
` 30-84 27-82
` 26-79 27-84 18-84
`
`
`
`335
`339
`346
`306
`300
`
` (72)
`
` (72.7)
`
` (74.1)
`
` (74.5)
` (73.2)
`
`130
`127
`121
`105
`110
`
` (28)
`
` (27.3)
`
` (25.9)
`
` (25.5)
` (26.8)
`
`330
`344
`333
`355
`353
`
` (71)
`
` (73.8)
`
` (71.3)
`
` (86.4)
` (86.1)
`
`69
`56
`59
`28
`21
`
` (14.8)
`
` (12)
`
` (12.6)
`
` (6.8)
`
` (5.1)
`38
`38
`39
`26
`33
`
` (8)
`
` (8.2)
`
` (8.2)
`
` (8.4)
`
` (6.3)
`
`
`
`3
`4
`6
` 0
` 0
`
` (0)
`
` (0.6)
`
` (0)
`
` (0.9)
`
` (1.3)
`
`
` 0
`
`
` 0
`1
` 0
` 0
`
` (0)
`
`
` (0)
`
` (0)
`
` (0.2)
` (0)
`
`14
`
`18
`10
` 0
` 0
` (0)
`
` (0)
` (3)
`
` (2.1)
`
` (3.9)
`
`
`11
`13
`12
`2
`3
`
` (2.4)
`
` (2.8)
`
` (2.6)
`
` (0.5)
`
` (0.7)
`108
`77
`76
`93
`91
`
` (16.6)
`
` (16.3)
`
` (19.9)
`
` (22.1)
` (26.3)
`
`378
`385
`366
`320
`302
`
` (81.3)
`
` (82.6)
`
` (78.4)
`
` (77.9)
`
` (73.7)
`
`
`10
`5
`8
` 0
` 0
` (0)
` (0)
`
` (2.2)
`
` (1.1)
`
` (1.7)
`
`
`115
`127
`13