CENTER FOR DRUG EVALUATION AND
`RESEARCH
`
`
`APPLICATION NUMBER:
`
`209637Orig1s000
`
`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 D D E N D U M )
`CLINICAL STUDIES
`
`NDA/BLA #:
`
`NDA 209637
`
`Drug Name:
`Indication(s):
`
`Applicant:
`Date(s):
`
`Review Priority:
`
`Semaglutide Injection
`Improve Glycemic Control in Adults with Type 2 Diabetes
`Mellitus (T2DM)
`Novo Nordisk Inc.
`Date submitted: December 5, 2016
`Review due date: August 11, 2017
`PDUFA due date: December 5, 2017
`Standard
`
`Biometrics Division:
`Division of Biometrics II
`Statistical Reviewer:
`Jiwei He, Ph.D.
`Concurring Reviewers: Yun Wang, Ph.D. (Acting Team Leader)
`
`Medical Division:
`Clinical Team:
`
`Project Manager:
`
`Metabolism and Endocrinology Products (DMEP)
`Andreea Lungu, M.D.
`William Chong, M.D. (Team Leader)
`Jean-Marc Guettier, M.D. (Division Director)
`Peter Franks, M.S.
`
`Keywords: subgroup analysis
`
`Reference ID: 4189742
`
`

`

`We requested the applicant to examine the treatment effect of each of semaglutide 0.5 mg and
`semaglutide 1.0 mg on the change in HbA1c from baseline within each sex, age, race, and
`ethnicity subgroup and the differences in the treatment effect between the subgroups. The
`information request we sent on 11/21/2017 is included in the Appendix of this document.
`
`FINDINGS IN SUBGROUP POPULATIONS
`The results in this section were provided by the applicant. They are consistent with the results in
`the subgroup section of the statistical review for this NDA (documented in DARRTS on
`8/11/2017). In the statistical review, the placebo-controlled studies were not pooled, and the race
`subgroups were analyzed as White versus non-White.
`
`In general, the treatment effect of each of semaglutide 0.5 mg and semaglutide 1.0 mg was
`consistent in different subgroups. The few significant differences found between subgroups were
`quantitative instead of qualitative. An evaluation of whether differences in the treatment effect
`across subgroups is real should consider whether the finding has been reproduced and how
`strong is the evidence of a finding (which needs to consider how many possible comparisons can
`be done).
`
` For subgroup analyses by sex,
`o Each of semaglutide 0.5 mg and semaglutide 1.0 mg is superior to placebo with
`respect to the change in HbA1c from baseline within each sex.
`o In trial 3625 (SUSTAIN 4), a slightly higher effect was seen in females compared
`to males with the 0.5 mg dose (-0.45 versus -0.11, nominal p-value = 0.03), but
`there was no difference with the 1.0 mg dose.
`
` For subgroup analyses by age group (< 65, ≥ 65 years of age),
`o Each of semaglutide 0.5 mg and semaglutide 1.0 mg is superior to placebo with
`respect to the change in HbA1c from baseline within each age group.
`o In trial 3625 (SUSTAIN 4), a slightly higher effect was seen in younger patients
`as compared to older patients with the 1.0 mg dose (-0.7 versus -0.21, nominal p-
`value = 0.01), but no difference was observed with the 0.5 mg dose.
`
` For subgroup analyses by race (White, Black or African American, Asian, Other),
`o Each of semaglutide 0.5 mg and semaglutide 1.0 mg is superior to placebo with
`respect to the change in HbA1c from baseline within each race group.
`o In the placebo pool, a slightly higher effect was seen in Asians and Black/African
`Americans as compared to White patients with both doses (-1.42 and -2.23 versus
`-1.07, nominal p-value = 0.05 for semaglutide 0.5 mg; -1.92 and -2.23 versus -
`1.33, nominal p-value = 0.02 for semaglutide 1.0 mg).
`
` For subgroup analyses by ethnic group (Hispanic, Non-Hispanic),
`o Each of semaglutide 0.5 mg and semaglutide 1.0 mg is superior to placebo with
`respect to the change in HbA1c from baseline within each ethnic group.
`
`Reference ID: 4189742
`
`

`

`o In the placebo pool, a slightly lower effect was seen in Hispanic/Latino patients
`compared to non-Hispanic/non-Latino patients with both doses (-0.75 versus -
`1.28, nominal p-value = 0.04 for semaglutide 0.5 mg; -1.03 versus -1.62, nominal
`p-value = 0.04 for semaglutide 1.0 mg).
`
`Reference ID: 4189742
`
`

`

`Result Tables:
`
`Table 1. Effect of semaglutide on HbA1c by subgroup (Placebo-controlled trials)
`
`Semaglutide 0.5 mg
`
`Semaglutide 1.0 mg
`
`Control
`
`N
`
`Mean
`HbA1c at
`baseline
`
`LS Mean
`Change
`from
`Baseline
`
`N
`
`Mean
`HbA1c at
`baseline
`
`LS Mean
`Change
`from
`Baseline
`
`N
`
`Mean
`HbA1c at
`baseline
`
`LS Mean
`Change
`from
`Baseline
`
`Treatment Difference
`[95% Confidence
`Interval]
`(Semaglutide 0.5 mg
`minus Control)
`
`Test for
`Treatment
`by
`Subgroup
`Interactio
`n (p-
`value)
`0.60
`
`Treatment
`Difference [95%
`Confidence Interval]
`(Semaglutide 1.0 mg
`minus Control)
`
`Test for
`Treatment
`by
`Subgroup
`Interaction
`(p-value)
`
`134
`126
`
`195
`65
`
`191
`15
`45
`
`49
`211
`
`8.35
`8.1
`
`8.32
`7.94
`
`8.2
`8.39
`8.32
`
`8.3
`8.21
`
`-1.38
`-1.29
`
`-1.36
`-1.27
`
`-1.27
`-1.72
`-1.58
`
`-0.98
`-1.43
`
`157
`104
`
`212
`49
`
`186
`20
`48
`
`57
`204
`
`8.28
`8.13
`
`8.26
`8.05
`
`8.21
`8.27
`8.24
`
`8.22
`8.22
`
`-1.72
`-1.58
`
`-1.7
`-1.5
`
`-1.53
`-1.74
`-2.09
`
`-1.25
`-1.77
`
`141
`121
`
`191
`71
`
`179
`17
`56
`
`55
`207
`
`8.17
`8.22
`
`8.18
`8.22
`
`8.16
`8.51
`8.27
`
`8.16
`8.2
`
`-0.16 -1.21 [-1.515 ; -0.912]
`-0.17 -1.11 [-1.42 ; -0.791]
`
`-0.11 -1.25 [-1.501 ; -0.997]
`-0.29 -0.98 [-1.419 ; -0.546]
`
`-0.21 -1.07 [-1.326 ; -0.806]
`0.48 -2.23 [-3.258 ; -1.211]
`-0.16 -1.42 [-1.873 ; -0.975]
`
`-1.56 [-1.843 ; -1.287]
`-1.41 [-1.728 ; -1.084]
`
`-1.58 [-1.825 ; -1.345]
`-1.22 [-1.67 ; -0.777]
`
`-1.33 [-1.574 ; -1.076]
`-2.23 [-3.139 ; -1.315]
`-1.92 [-2.369 ; -1.47]
`
`0.32
`
`0.05
`
`0.04
`
`-0.24 -0.75 [-1.25 ; -0.246]
`-0.15 -1.28 [-1.519 ; -1.044]
`
`-1.03 [-1.509 ; -0.546]
`-1.62 [-1.852 ; -1.391]
`
`0.45
`
`0.13
`
`0.02
`
`0.04
`
`Demographic Parameters
`
`Sex
` Male
` Female
`Age Group
`below 65 years
`65 years and above
`Race
` White
` Black or African American
` Asian
` Other
`Ethnicity
` Hispanic or Latino
` Not Hispanic or Latino
`
`Reference ID: 4189742
`
`

`

`Table 2. Effect of semaglutide on HbAlc by subgroup (Trial 3624- vs exenatide)
`
`. emaglutide 1.0 mg
`
`'
`[95% Confidence
`
`Treatment
`Difference [95%
`
`Interval]
`
`Sex
`-——-—_-—————-m
`Male
`-__-ZE-EM-E—__
`Female
`-_—mm_--m-m————
`‘ e Grou-
`-——-—--—————-ma
`below 65 years
`-__umm-um-za-aa—— -0.49 [-0.708; -o.27sl_
`65 years and above
`-__m-M_-_-0-E——As [-0.844; -o.113]_
`Race
`-——-—_-—————-m
`White
`-_—mnm_m-m-m————
`Black or African American
`28
`8.59
`~0.86
`8.51
`-0.4
`-0.45 [-1.145 ; 0.236]
`
`Asian
`8
`9.34
`-1.26
`7.98
`-0.89
`-0.36 [-1.797; 1.076]
`Other
`-——-_W——_
`-——-—----—_—-E
`-_—-n-m_——m-m———
`-__‘E-E“-E—_—_
`
`Ethnicity
`Hispanic or Latino
`Not Hispanic or Latino
`
`Demographic Parameters
`
` Semaglutide 0.5 mg
`
`
`
`Reference ID: 41 89742
`
`

`

`
`8
`g _I [95%Confidence
`Difference[95%
`Treagment
`
`
`
`
`Treatment
`
`Test for
`
`Interval]
`Subgroup Confidence Interval] S b y
`
`Mean
`(Semaglutide 0.5 mg lnteractio (Semaglutide 1.0 mg
`u ”of",
`
`
`HbAlc at
`Interaction
`minus Control)
`minus Control)
`
`baseline
`.
`(p-value)
`'
`.
`-__-__-__—-EE—-EE
`WWW——
`WEE--
`-__-—_-__—-M—-E
`m-E_mfl__”fl-E -027 {-0.474 ; -0.065]— -0-7 {-0.906; -0.495]_
`m__--IE_-_——-E__
`-__-__-__—-Ifl—_
`mam-mam -o.3[-o.49;-o.1021_—_
`12-52 -1-17mm 4-65 Izm— -o.os {-0.614; 0.4511— -o.56 {-1.105; mm_
`mnmmm——
`mun—m—
`-__-__-__—_—-M
`-_-__-__——-E__
`301
`-1.15mm -1.49
`281mm -o.21[-o.4o4;-o.oos]_ -o.55[-o.751;-o.35]_
`
`Table 3. Effect of semaglutide on HbA1c by subgroup (Tr'al 3625 - vs lGlar)
`
`Sema Iutide 0.5 m
`
`. ema lutide 1.0 m
`
`
`
`Demographic Parameters
`
`Sex
`Male
`Female
`
`Age Group
`below 65 years
`65 years and above
`Race
`White
`Black or African American
`Asian
`Other
`
`Ethnicity
`Hispanic or Latino
`Not Hispanic or Latino
`
`Reference ID: 41 89742
`
`

`

`
`
`Table 4. Effect of semaglutide on HbA1c by subgroup (Trial 3626 - vs sitagliptin)
`
`Semaglutide 0.5 mg
`
`Semaglutide 1.0 mg
`
`Control
`
`N
`
`Mean
`HbA1c at
`baseline
`
`LS Mean
`Change
`from
`Baseline
`
`N
`
`Mean
`HbA1c at
`baseline
`
`LS Mean
`Change
`from
`Baseline
`
`N
`
`Mean
`HbA1c at
`baseline
`
`LS Mean
`Change
`from
`Baseline
`
`Treatment Difference
`[95% Confidence
`Interval]
`(Semaglutide 0.5 mg
`minus Control)
`
`Test for
`Treatment
`by
`Subgroup
`Interactio
`n (p-
`value)
`0.78
`
`Treatment
`Difference [95%
`Confidence Interval]
`(Semaglutide 1.0 mg
`minus Control)
`
`Test for
`Treatment
`by
`Subgroup
`Interaction
`(p-value)
`
`207
`202
`
`333
`76
`
`279
`18
`106
`6
`
`69
`340
`
`8.01
`8.02
`
`8.02
`8
`
`7.98
`8.54
`7.99
`8.42
`
`8
`8.02
`
`-1.31
`-1.27
`
`-1.3
`-1.23
`
`-1.27
`-1.41
`-1.31
`-1.48
`
`-1.47
`-1.25
`
`205
`204
`
`332
`77
`
`279
`24
`99
`7
`
`67
`342
`
`8.16
`7.91
`
`8.08
`7.89
`
`7.99
`8.06
`8.13
`8.56
`
`8.15
`8.02
`
`-1.47
`-1.49
`
`-1.49
`-1.44
`
`-1.47
`-1.56
`-1.44
`-2.35
`
`-1.57
`-1.46
`
`208
`199
`
`328
`79
`
`281
`17
`102
`7
`
`73
`334
`
`8.12
`8.22
`
`8.2
`8.06
`
`8.12
`8.31
`8.29
`7.97
`
`8.28
`8.15
`
`-0.76 -0.55 [-0.761 ; -0.337]
`-0.67 -0.59 [-0.812 ; -0.373]
`
`-0.69 -0.61 [-0.786 ; -0.444]
`-0.85 -0.39 [-0.736 ; -0.035]
`
`-0.72 -0.55 [-0.735 ; -0.368]
`-0.86
`-0.56 [-1.39 ; 0.279]
`-0.7 -0.61 [-0.907 ; -0.315]
`-0.79 -0.69 [-1.806 ; 0.422]
`
`-0.74 -0.72 [-1.074 ; -0.37]
`-0.71 -0.54 [-0.709 ; -0.367]
`
`0.24
`
`0.99
`
`0.35
`
`-0.71 [-0.914 ; -0.503]
`-0.82 [-1.037 ; -0.601]
`
`-0.8 [-0.972 ; -0.636]
`-0.59 [-0.934 ; -0.253]
`
`-0.75 [-0.933 ; -0.57]
`-0.7 [-1.429 ; 0.028]
`-0.74 [-1.038 ; -0.45]
`-1.56 [-2.628 ; -0.485]
`
`-0.83 [-1.177 ; -0.48]
`-0.75 [-0.916 ; -0.58]
`
`0.46
`
`0.27
`
`0.58
`
`0.68
`
`Demographic Parameters
`
`Sex
` Male
` Female
`Age Group
`below 65 years
`65 years and above
`Race
` White
` Black or African American
` Asian
` Other
`Ethnicity
` Hispanic or Latino
` Not Hispanic or Latino
`
`Reference ID: 4189742
`
`

`

`Forest Plots:
`
`Reference ID: 4189742
`
`

`

`SUSTNN P {comparator- slaguifll
`5“
`
`5" I15 a.
`
`u... .,.... .....,.....
`
`9......“ m...-
`
`5-!- ‘I my
`
`FH
`
`..u_.—--; m...”
`
`H_.
`
`H-I
`
`H—{
`
`Mlle
`
`Fulfill
`
`Po- m
`¢ 55
`
`a: 55
`
`Rm
`
`While
`BIICI at Mac.“ mm
`
`Mum
`fllh-r
`
`Emmy
`
`HIIPII: a Latino
`Nu Haw-c or um
`
` 2.0 —1.5
`
`
`H-l
`F———Q—-——i
`
`l—O—l
`i—Q—tf
`
`H—I
`0-H
`
`-'|.fl
`
`415 MI
`
`0.5
`
`1!!
`
`1i
`
`P-mflu unwell” mm Ihllhlrllu new Illlcl all" um: mumps
`{Ia p-vubrlfit “WW Madam) furSUST-NN 2
`in: an Ann 021 Run an; Emmy OJSmrSmflSwu-rwmmmr
`S.- I”! Act 02? Pin 05! Ethnicity Duhrmiwwmcwflw
`
`mm.
`
`q..- ,. _ .A mun-Ir. .‘._-..
`. Iu.~._.....‘ m... _._..
`
`Figure 2
`
`Forest Plat for SUSTAIN 2 (Trial NN9535—3 626)
`
`SUSTAIN 3 (cmur Hm ER!
`SI:
`
`Sm Hm
`
`an -. 1.... a...“
`m.-. . “mm—.4.
`
`Mu-
`
`Fume
`Aw m
`« $5
`
`=-= 6E
`
`R‘-
`
`W'hle
`Blad- u Mac-1 Murman-
`
`“In
`Olhfl
`
`Emit!!!
`
`Him in Lima
`Nu Hm: or Lama
`
`
`
`I—H
`
`I‘-+-4
`
`FH
`
`|—-—9—-—l
`
`PH
`i—O—l
`
`l—O—l
`t—f—‘l
`
`r—o—+
`H—4
`
`.10
`
`.15
`
`—1.0
`
`.05
`
`on
`
`:15
`
`In
`
`15
`
`
`
`Wk- “m1mmmmm Huh-Welland m 41mm
`in :2 Huh: test of umwm mum-hr BUST-ml 3
`$1: a :1 HI 031mm 1 an ElllwimmI 09hr Sun ‘I my win-«comic:-
`n..._.- I: _ _-. _,,.,__,_,. ...._...... H
`.- --...-- .—.....-......-. r
`Efi' MmWI
`
`
`Figure 3
`
`Forest Plat for SUSTAIN 3 (Trial NN9535—3624)
`
`Reference ID: 4189742
`Reference ID: 4189742
`
`

`

`APPENDIX: FDA’s Information Request
`Please complete the shell table for HbA1c results by subgroup using the retrieved dropout
`analysis method in Section 14 of the proposed product label based on analyses in each of the
`following studies or combinations of studies:
`
` Placebo-controlled key efficacy studies 3623 and 3627 combined
`
` Active-controlled key efficacy studies 3626, 3624, 3625 individually
`
`For the individual active-controlled studies, estimate the treatment effect of each of semaglutide
`0.5 mg and semaglutide 1.0 mg relative to the active comparator within subgroups and test for
`the difference in overall treatment effect across subgroups. For the combination of placebo-
`controlled studies, estimate the treatment effect of each of semaglutide 0.5 mg and semaglutide
`1.0 mg relative to the placebo within subgroups by combining the estimates for individual
`studies inversely weighted by their variances.
`
`With respect to the interaction tests of the treatment effect by subgroup factor (e.g., race), for an
`individual study the ANCOVA model should include the factors/terms:
`
`
`
`
`
`race (as a categorical factor)
`
`treatment
`
`Reference ID: 4189742
`
`

`

`
`
`
`
`treatment by race interaction term
`
`the covariates used in the primary analysis
`
`When performing an interaction test of the treatment effect by subgroup factor (e.g., race) for a
`combination of studies, additionally include the factors/terms:
`
`
`
`
`
`
`
`race by study interaction term
`
`treatment by study interaction term
`
`interaction terms with study for each covariate used in the primary analysis
`
`In addition, please provide a forest plot for each set of subgroup analysis. An example forest plot
`may be found at: https://www.fda.gov/Drugs/InformationOnDrugs/ucm532714.htm under the
`MORE INFO section of the question addressing whether there were any differences in how well
`the drug worked in clinical trials among sex, race and age.
`
`Please provide the code and a description of the statistical methods used to generate these
`analyses.
`
`Reference ID: 4189742
`
`

`

`---------------------------------------------------------------------------------------------------------
`This is a representation of an electronic record that was signed
`electronically and this page is the manifestation of the electronic
`signature.
`---------------------------------------------------------------------------------------------------------
`/s/
`----------------------------------------------------
`
`JIWEI HE
`12/04/2017
`
`YUN WANG
`12/04/2017
`
`Reference ID: 4189742
`
`

`

`
`
`
`
`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 and Evaluation
` CARCINOGENICITY STUDIES
`
`IND/NDA Number:
`Drug Name:
`Indication:
`Studies:
`Applicant:
`
`NDA 209637 (IND 79754)
`Semaglutide
`Treatment of Type 2 Diabetes Mellitus (T2D)
`104 Weeks of Carcinogenicity Studies in Rats and Mice
`Sponsor:
`
`Novo Nordisk A/S,
`Novo Nordisk Park,
`DK-2760 Måløv,
`Denmark.
`Testing Facility:
`
`Review Priority:
`Biometrics Division:
`Statistical Reviewer:
`Concurring Reviewer:
`Medical Division:
`Reviewing Pharmacologist:
`Keywords:
`
`
`
`
`
`Standard
`Division of Biometrics - VI
`Hepei Chen
`Karl Lin, Ph.D.
`Division of Metabolism and Endocrinology Products
`Federica Basso, Ph.D.
`Carcinogenicity, Dose response
`
`Reference ID: 4176069
`
`(b) (4)
`
`

`

`NDA 209637 (Semaglutide)
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Page 2 of 43
`
`Table of Contents
`
`Background ..........................................................................................................................3
`Rat Study ..............................................................................................................................3
`2.1.
`Sponsor's analyses ....................................................................................................4
`2.1.1. Survival analysis ..........................................................................................4
`Sponsor’s findings .......................................................................................4
`2.1.2. Tumor data analysis .....................................................................................4
`Adjustment for multiple testing ...................................................................5
`Sponsor’s findings .......................................................................................5
`2.2. Reviewer's analyses .................................................................................................6
`2.2.1. Survival analysis ..........................................................................................6
`Reviewer’s findings .....................................................................................6
`2.2.2. Tumor data analysis .....................................................................................7
`Multiple testing adjustment..........................................................................8
`Reviewer’s findings .....................................................................................8
`Mouse Study ......................................................................................................................13
`3.1.
`Sponsor's analyses ..................................................................................................14
`3.1.1. Survival analysis ........................................................................................14
`Sponsor’s findings .....................................................................................14
`3.1.2. Tumor data analysis ...................................................................................14
`Sponsor’s findings .....................................................................................14
`3.2. Reviewer's analyses ...............................................................................................15
`3.2.1. Survival analysis ........................................................................................15
`Reviewer’s findings ...................................................................................15
`3.2.2. Tumor data analysis ...................................................................................16
`Reviewer’s findings ...................................................................................16
`Summary ............................................................................................................................17
`Appendix ............................................................................................................................20
`Table 1A: Intercurrent mortality rate in male rats
`Table 1B: Intercurrent mortality rate in female rats
`Table 2A: Tumor rates and p-values for trend and pairwise comparisons in male rats
`Table 2B: Tumor rates and p-values for trend and pairwise comparisons in female rats
`Table 3A: Intercurrent mortality rate in male mice
`Table 3B: Intercurrent mortality rate in female mice
`Table 4A: Tumor rates and p-values for trend and pairwise comparisons in male mice
`Table 4B: Tumor rates and p-values for trend and pairwise comparisons in female mice
`Figure 1A: Kaplan-Meier survival functions for male rats
`Figure 1B: Kaplan-Meier survival functions for female rats
`Figure 2A: Kaplan-Meier survival functions for male mice
`Figure 2B: Kaplan-Meier survival functions for female mice
`References ..........................................................................................................................43
`
`
`
`1.
`2.
`
`3.
`
`4.
`5.
`
`6.
`
`
`
`Reference ID: 4176069
`
`

`

`NDA 209637 (Semaglutide)
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Page 3 of 43
`
`1. Background
`
`
`In this submission the sponsor included reports of two animal carcinogenicity studies, one in rats
`and one in mice. These studies were to assess the carcinogenic potential of NNC 0113-0217 (a
`GLP-1 agonist being developed for the treatment of Type 2 diabetes), when administered to CD
`rats and CD-1 mice by daily subcutaneous administration over a period of 104 weeks.
`
`In this review the phrase "dose response relationship" refers to the linear component (trend) of the
`effect of treatment, and not necessarily to a strictly increasing or decreasing mortality or tumor
`incidence rate as dose increases.
`
`
`2. Rat Study
`
`
`Two separate experiments, one in male rats and one in female rats were conducted. As indicated
`in Table 1, in each of these two experiments there were four treated groups and one vehicle
`control group. Three hundred fifty CD rats of each sex were assigned randomly in size of 70 rats
`per group. The dose levels for the four treated groups were 0.0025, 0.01, 0.025, and 0.1
`mg/kg/day for both male and female rats. In this review these dose groups were referred to as the
`low (Group 2), mid (Group 3), mid-high (Group 4), and high (Group 5) dose groups,
`respectively. The rats in the vehicle control group were administrated with the vehicle [1.42
`mg/mL disodium
` phosphate, dihydrate; 14.0 mg/mL propylene glycol; 5.50 mg/mL
`phenol in water for injection, adjusted to pH 7.4], and handled for the same duration and in the
`same manner as the treated groups.
`
`
`Table 1: Experimental Design in Rat Study
`
`Test Material
`
`Group
`No.
`1
`2
`3
`4
`5
`
`No. of Animals
`Male
`Female
`70
`70
`70
`70
`70
`70
`70
`70
`70
`70
`
`Vehicle Control
`NNC 0113-0217 Low
`NNC 0113-0217 Mid
`NNC 0113-0217 Mid-High
`NNC 0113-0217 High
`
`Dosage Level (mg/kg/day)
`Male
`Female
`0
`0
`0.0025
`0.0025
`0.01
`0.01
`0.025
`0.025
`0.1
`0.1
`
`
`Animals were inspected visually at least twice daily for evidence of ill-health or reaction to
`treatment. In addition, a more detailed weekly physical examination, which included palpation,
`was performed on each animal to monitor general health. Particular attention was paid to any
`superficial or palpable swellings, for which the location, size, consistency, time of first
`observation and subsequent history were recorded. Animals were killed for reasons of animal
`welfare where necessary. Animals killed during the study and those surviving until the end of the
`scheduled treatment period were killed by carbon dioxide asphyxiation. All animals were subject
`to a detailed necropsy. After a review of the history of each animal, a full macroscopic
`examination of the tissues was performed. All external features and orifices were examined
`visually including the parenteral sites. Any abnormal position, morphology or interaction was
`recorded. The organs and tissues were examined as appropriate. Any abnormality in the
`appearance or size of any organ and tissue was recorded and the required tissue samples
`preserved in appropriate fixative. All tissues preserved for examination (as specified above) were
`
`Reference ID: 4176069
`
`(b) (4)
`
`

`

`
`
`
`
`
`
`
`
`
`
`
`
`
`
`NDA 209637 (Semaglutide)
`
`examined for all animals sacrificed on completion of the scheduled treatment period and for all
`animals killed or dying during the study. Tissues reported at macroscopic examination as being
`grossly abnormal were examined for all animals in line with current practice.
`
`
`Page 4 of 43
`
`2.1.
`
`Sponsor's analyses
`
`
`2.1.1. Survival analysis
`
`In the sponsor’s analysis, the numbers of animal deaths during the study, up to terminal sacrifice,
`were analyzed by logrank tests for a trend across the groups. The numbers of animal deaths
`during the study were presented as life-tables and Kaplan-Meier survival curves. The following
`statistical tests were carried out: 1) a two-tailed test for a trend with dose level, and 2) a two-
`tailed pairwise comparison test of each treatment group against the control group. Where the test
`for trend was statistically significant, the highest dose group was excluded and the trend test was
`repeated (using a one-tailed test), until the test was no longer statistically significant.
`
`As a check, tests for non-linearity (not presented) were carried out. In this study, the non-
`linearity test was statistically significant for the females (p=0.006) and thus the results of the
`pairwise tests are to be preferred to the trend test. For the males the non-linearity test was not
`statistically significant at the 1% level and thus the result of the trend test is to be preferred.
`P-values were calculated using χ2 tests and adjusted with a continuity correction where there was
`only one degree of freedom.
`
`Sponsor’s findings:
`
`The sponsor’s analysis showed that the numbers of rats surviving to their terminal necropsy were
`35 (50%), 32 (46%), 28 (40%), 40 (57%), and 33 (47%) in Groups 1, 2, 3, 4, and 5 for male rats,
`respectively, and 34 (49%), 31 (44%), 20 (29%), 19 (27%), and 31 (44%) for female rats
`respectively. For male rats, the trend test was not statistically significant when all groups were
`included in the analysis (p=0.699), without any statistically significant pairwise comparisons.
`For female rats, statistically significant pairwise comparisons of the control group with the mid
`and mid-high dose groups were noted (p=0.028 and p=0.019 respectively), while the trend test
`was not statistically significant (p=0.649).
`
`2.1.2. Tumor data analysis
`
`In the sponsor’s report, the analyses were carried out for benign, malignant and benign and
`malignant tumors combined. If an animal had a benign and a malignant non-palpable tumor then
`only the malignant tumor was included in the analysis of both tumors together. Also, if an animal
`had more than one palpable tumor of the same category, benign or malignant, then only the first
`observed tumor was included in the analysis. Tumor types were selected for full statistical
`analysis where at least two tumors were observed in treated groups for which all animals were
`examined.
`
`For the life-table analysis the study was divided into time strata. For non-incidental tumors, the
`strata are defined as those weeks during which there were deaths. For incidental tumors, the
`
`Reference ID: 4176069
`
`

`

`
`
`
`
`
`
`
`
`
`
`
`
`Page 5 of 43
`
`NDA 209637 (Semaglutide)
`
`following fixed time intervals were used to adjust for differential mortality between the treatment
`groups: 1-52, 53-78, 79-92, 93-104 weeks and terminal sacrifice (FDA 2001).
`
`Log-rank methods were used to analyze the number of animals with tumors across treatment
`groups. The following χ2 statistical tests were carried out: 1) a one-tailed test for a trend using
`nominal dose levels, with the control group; 2) a one-tailed pairwise comparison test of each
`treatment group against the control group. Where the test for trend was statistically significant,
`the highest dose group was excluded and the trend test was repeated, using a one-tailed test until
`the test was no longer statistically significant. The p-values were adjusted using a continuity
`correction where there was one degree of freedom.
`
`As a check, tests for non-linearity were carried out (but not presented in this report). For benign
`C-cell adenoma and benign C-cell adenoma and malignant C-cell carcinoma combined (thyroids)
`in males and females (p<0.001) and for benign adenoma (mammary areas) in females (p=0.007)
`the non-linearity tests were significant, therefore the pairwise tests are to be preferred to the trend
`test. For all other analyses the non-linearity tests were not statistically significant at the 1% level,
`hence the trend tests are to be preferred.
`
`Where there were fewer than ten observed tumor bearing animals across all treatment groups,
`exact one-tailed p-values were calculated using permutation tests for stratified contingency
`tables, to test for trend and for pairwise comparisons of each treatment group against the control
`group.
`
`Adjustment for multiple testing:
`
`In the sponsor’s report, no text can be located in terms of the criteria used for the significance
`adjustment for multiple testing. Based on the sponsor’s discussion of the tumor findings, no
`adjustment for multiple testing in the context of rare or common tumors was considered in the
`sponsor’s analysis.
`
`Sponsor’s findings:
`
`In the sponsor’s analysis, for benign C-cell adenoma of thyroids in male rats, the trend test was
`statistically significant (p<0.001), along with statistically significant pairwise comparisons of the
`control group versus the low, mid, mid-high, and high groups (p=0.012, p<0.001, p<0.001 and
`p<0.001 respectively). For malignant C-cell carcinoma, the trend test was statistically significant
`(p<0.001), along with statistically significant pairwise comparisons of the control group versus
`the mid, mid-high, and high groups (p=0.014, p=0.008 and p<0.001 respectively). For benign C-
`cell adenoma and malignant C-cell carcinoma combined, the trend test was statistically
`significant (p<0.001), along with statistically significant pairwise comparisons of the control
`group versus the low, mid, mid-high, and high groups (p=0.020, p<0.001, p<0.001 and p<0.001
`respectively).
`
`For malignant large granular cell lymphoma of haematopoietic in male rats, the trend test was
`statistically significant (p=0.049), while no statistically significant pairwise comparisons was
`noted.
`
`
`
`Reference ID: 4176069
`
`

`

`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Page 6 of 43
`
`NDA 209637 (Semaglutide)
`
`
`For benign C-cell adenoma of thyroids in female rats, the trend test was statistically significant
`(p<0.001), along with statistically significant pairwise comparisons of the control group versus
`the low, mid, mid-high, and high groups (p=0.009, p<0.001, p<0.001 and p<0.001 respectively).
`For malignant C-cell carcinoma, the trend test was not statistically significant (p=0.057), while a
`pairwise comparison of the control group versus the high dose group was statistically significant
`(p=0.043). For benign C-cell adenoma and malignant C-cell carcinoma combined, the trend test
`was statistically significant (p<0.001), along with statistically significant pairwise comparisons
`of the control group versus the low, mid, mid-high, and high groups (p=0.011, p<0.001, p<0.001
`and p<0.001 respectively).
`
`For benign adenoma of Pituitary (pars distalis) in female rats, the trend test was not statistically
`significant (p=0.803), along with a statistically significant pairwise comparison of the control
`group versus the mid-high dose group (p=0.029). For benign adenoma and malignant carcinoma
`combined, the trend test was not statistically significant (p=0.818), while the pairwise
`comparison of the control group versus the mid-high dose group was statistically significant
`(p=0.029).
`
`For malignant adenocarcinoma in mammary areas of female rats, the trend test was not
`statistically significant (p=0.875), while the pairwise comparison of the control group versus the
`mid dose group was statistically significant (p=0.038).
`
`For benign squamous papilloma and malignant squamous cell carcinoma combined in uterus of
`female rats, the trend test was statistically significant when all groups were included in the
`analysis (p=0.032), while no statistically significant pairwise comparisons was noted.
`
`
`2.2. Reviewer's analyses
`
`
`To verify the sponsor’s analyses and to perform additional analyses suggested by the reviewing
`toxicologist, this reviewer independently performed the survival and tumor data analyses using
`the data provided by the sponsor electronically.
`
`
`2.2.1. Survival analysis
`
`In the reviewer’s analysis, the survival distributions of rats in all five groups (Groups 1, 2, 3, 4, and
`5) were estimated using the Kaplan-Meier product limit method. The dose response relationship was
`tested across Groups 1, 2, 3, 4, and 5 using the likelihood ratio test, and the homogeneity of survival
`distributions was tested using the log-rank test. The Kaplan-Meier curves for survival rates are
`given in Figures 1A and 1B in the appendix for all five groups in male and female rats, respectively.
`The intercurrent mortality data of all five groups, and the results of the tests for dose response
`relationship and homogeneity of survivals for Groups 1, 2, 3, 4, and 5 are given in Tables 1A and
`1B in the appendix for male and female rats, respectively.
`
`Reviewer’s findings:
`
`The reviewer’s analysis showed that the numbers of rats surviving to their terminal necropsy
`
`Reference ID: 4176069
`
`

`

`
`
`
`
`
`
`
`
`
`
`
`
`Page 7 of 43
`
`NDA 209637 (Semaglutide)
`
`were 33 (47%), 32 (46%), 28 (40%), 40 (57%), and 33 (47%) in Groups 1, 2, 3, 4, and 5 for male
`rats, respectively, and 34 (49%), 31 (44%), 20 (29%), 19 (27%), and 31 (44%) for female rats
`respectively. No statistically significant dose response relationship and pairwise comparisons in
`mortality was noted for male rats; while for female rats, statistically significant pairwise
`comparisons in mortality were noted when comparing the control group versus the mid and mid-
`high dose groups (p=0.0227 and p=0.0151 respectively), but without the corresponding
`statistically significant dose response relationship.
`
`For the control group in male rats, there is a discrepancy of the number of animal surviving to
`their terminal sacrifice between the sponsor’s and reviewer’s report (35 vs. 33). The reason is
`that, two animals (#46 and #60) in the control group of male rats that died during the Week 105,
`were labeled by the sponsor as “Natural Death” instead of “Terminal Sacrifice” in the tumor.xpt
`dataset. Therefore the reviewer considered them as the natural death, whereas the sponsor
`considered them as the surviving animals.
`
`2.2.2. Tumor data analysis
`
`The tumor data were analyzed for dose response relationships across Groups 1, 2, 3, 4, and 5, and
`pairwise comparisons of each of the four treated groups (Groups 2, 3, 4, and 5) against the vehicle
`control group (Group 1), using the Poly-k method described