CENTER FOR DRUG EVALUATION AND
`
`-
`
`RESEARCH -
`
`'
`
`APPLICA TION NUMBER:
`
`21-995
`
`STATISTICAL REVIEW! S!
`
`

`

`STATISTICAL REVIEW AND EVALUATION
`
`Biometrics Division: VI
`
`’N’bA’NOg
`
`.
`
`.7
`
`..
`
`7
`
`2.1__9.975
`
`W
`V
`iSERIAL No.:
`,DATERFICEIVED BY THE CENTER:
`
`iAmendment
`September 21, 2006"
`
`ii
`
`I
`
`M
`
`DRUGNAME
`
`7 ,7,
`
`7,
`
`, MKO4‘3‘1
`
`..
`
`..
`
`
`
`DOSAGE FORM
`'
`'
`'
`'
`sponsors“
`DOCUMENTS REVIEWED A
`
`i
`
`..
`
`I
`
`‘
`I
`
`' VTablets
`Merck 1m.
`1 Electronic Copy Dated September 19, 2006
`
`3NAMESDESTATIISTICALARSEVIEWERS: ,Yi Tsong, PhD.
`7
`Meiyu Shen, Ph.D.
`
`NAME OF CHEMISTRY REVIEWER:
`
`Shah, Vibhakar'J, Ph.D.
`
`:
`
`'
`
`- Christine Moore, Ph.D.
`
`' Jinglin Zhong, Ph.D.
`
`Yi Tsong, Ph.D.
`Deputy Director
`
`_Meiyu Shen, Ph.D., Mathematical Statistician
`
`
`
`Jinglin Zhong, Ph.D., Mathematical Statistician
`
`Concur:
`
`Stella G. Machado
`
`Director, DBVI
`
`Distribution:
`
`NDA 21-995
`
`OBVI/Y. Tsong, PhD.
`OBVI/S.Machado PhD.
`OB/Bob O’Neill. Ph.. D.
`OB/ Patrician. Lillian. MS
`
`

`

`Statistical Review of NDA21-995, 10/11/2006
`
`ONDQA/DPE/ Shah, Vibhakar J, PhD.
`ONDQA/DPE/ Christine Moore, PhD.
`
`APPEARS ““3 WAY
`0N ORIGINAL
`
`APPEARS THlS WAY
`0N ORifimAL
`
`Page 2 of 7
`
`

`

`5 Page(s) Withheld
`
`4 § 552(b)(4) Trade Secret / Confidential
`
`§ 552(1b)(5)Deliberative Process .
`
`_;__ § 552(b)(5) Draft Labeling
`
`

`

`This is a representation of an electronic record that was signed electronically and
`this page is the manifestation of the electronic signature.
`
`MeiYu Shen
`
`10/11/2006 02:42:02 PM
`BIOMETRICS
`
`Yi Tsong
`10/12/2006 11:07:18 AM
`BIOMETRICS
`
`Jinglin Zhong
`10/12/2006 02:54:03 PM
`BIOMETRICS
`
`

`

`Office of Biostatistics
`
`US. Department of Health and Human Services
`Food and Drug Administration
`Center for Drug Evaluation and Research
`Office of Pharmacoepidemiology and Statistical Science
`
`STATISTICAL REVIEW AND EVALUATION
`
`CLINICAL STUDIES
`
`NDA/ Serial Number:
`
`21995
`
`Drug Name:
`
`Indication(s):
`
`Applicant:
`
`Date(s): .
`
`Januvia (Sitagliptin Phosphate) 25 qd, 50 qd and 100 mg qd
`
`An adjunct to diet and exercise to improve glycemic control in patients
`with type 2 diabetes mellitus as monotherapy and as combination therapy
`with metformin or a PPAR—y agonist (e.g. thiazolidinedione) when diet and
`exercise plus the single agent do not provide adequate glycemic control
`
`Merck & Co., Inc.
`
`Received 12/16/05; user fee (10 months) 10/16/06
`
`Review Priority:
`
`Standard
`
`Biometrics Division:
`
`Division of Biometrics 2
`
`Statistical Revievver:
`Concurring Reviewers:
`
`Lee—Ping Pian, Ph.D.
`Todd Sahlroot, Ph.D. Biometrics Team Leader
`Thomas Permutt. Ph.D. Division Director
`
`Medical Division:
`
`Division of Metabolism and Endocrinology Products (DMEP)
`
`Clinical Team:
`
`Project Manager:
`
`Ilan Irony, M.D.
`Mary Parks, M.D. Division Director
`Lina Aljuburi, Pharm D.
`
`Keywords: New Drug Application (NDA) review, clinical studies
`
`

`

`Table of Contents
`
`1. EXECUTIVE SUMMARY ............................................................................................................. 2
`
`1.1 CONCLUSIONS AND RECOMMENDATIONS ......................................................................................................... 2
`1.2 OVERVIEW OF CLINICAL PROGRAM AND STUDIES REVIEWED ..................................................................... 3
`
`2.
`
`INTRODUCTION ..................................................................................................... 4
`
`2.1
`2.2
`
`OVERVIEW ........................................................................................................................................................... 4
`
`DATA SOURCES ........................................................................................................................................... 8
`
`3.
`
`STATISTICAL EVALUATION ............................................................................................. 8
`
`3.1
`
`3.1.1
`3.1.2
`3.1.3
`3.1.4
`3.1.5
`
`EVALUATION OF EFFICACY ............................................................................................................................. 8
`
`MON OTHERAPY — 021 (24—WEEK) ..................................................
`MONOTHERAPY — 023 (1 8—WEEK) ............................................................................................................ 19
`ADD ON TO PIOGLITAZONE - 019 ........................................................................................................... 27
`ADD ON To METFORMIN — 020 ................................................................................................................ 32 »
`PHASE 2 STUDIES AND RENAL INSUFFICIENCY STUDY (1 Z—WEEK) .................................................... 39
`
`FINDINGS IN SPECIAL/SUBGROUP POPULATIONS ...................................................... 41
`
`SAFETY .................................................................................................................................... 46
`
`SUMMARY AND CONCLUSIONS ......................................................................................... 47
`
`CONCLUSIONS AND RECOMMENDATIONS ...................................................................................................... 47
`
`4.
`
`5.
`
`6.
`
`7.
`
`LABELING COMMENTS ....................................................................................................... 48
`
`1. EXECUTIVE SUMMARY
`
`1.1_C0nclusions and Recommendations
`
`This NDA includes clinical trial data for MK—043l (Januvia), a dipeptidyl peptidase 4 (DPP— 4) inhibitor
`for the treatment of Type 2 diabetes mellitus. The submission includes:
`
`0
`
`0
`
`Two 12—week Phase 2b studies in patients with Type 2 diabetes
`
`Four Phase 3 studies (18 to 24 week durations) in patients with Type 2 diabetes
`
`0 One 12—week Phase 3 study in Type 2 diabetics with chronic renal insufficiency. (This drug is
`substantially excreted by the kidney.)
`
`See Table l for study summaries including doses studied. . Of the four longer—term Phase 3 controlled
`efficacy and safety Studies, two studies evaluated MK—0431 as monotherapy (one 18—week study (023)
`and one 24—week study (021)) and two studies evaluated MK—0431 as add—on therapy to metformin (020)
`or pioglitazone (019).
`
`The Phase 2 studies (010 and 014) were well designed with sufficient power on the primary endpoint,
`HbAlc change from baseline, to assess dose response.
`
`

`

`MK—0431 doses studied in Phase 2 ranged from 10 mg to 100 mg daily doses given once a day (qd) or as
`split doses (bid). 100 mg and 200 mg qd (the latter as monotherapy only) were studied in the Phase 3
`trials in diabetic patients with normal renal function. The Phase 3 renal safety study assessed 25 mg and
`50 mg qd. The sponsor has proposed a (single) MK—0431 dose of 100 mg qd in patients with Type 2
`diabetes and normal renal function and reduced doses of 25 or 50 mg qd in patients with Type 2 diabetes
`complicated by renal insufficiency.
`
`The phase 2 data showed that efficacy did not appear to be substantially affected if the drug was given as
`a single daily dose or as split (bid) doses. The 25, 50 and 100 mg doses were consistently superior to
`placebo in both Phase 2 studies. Maximal efficacy of the drug was achieved at 50 mg with no clear
`additional clinical benefit from 100 mg.
`
`The proposed doses of 25 mg and 50 mg qd were shown to be effective in the renal—impaired population
`(study 028).
`i
`
`In the four Phase 3 studies, MK—0431 100 mg qd and 200 mg qd were statistically superior to placebo in ‘
`HbA]c change from baseline in patients with Type 2 diabetes. In the 24-week monotherapy study, the
`Least Squared Mean (LSM) differences from placebo (95% confidence intervals) in HbA1C change from
`baseline were —0.79% (-O.96, —0.62) and —0.94°/o (1.11, —0.77), respectively, for the 100 mg and 200 mg
`doses, respectively. In the 18—week monotherapy study, the differences were —0.60% (—0.82, —0.39) and —
`0.48% (—0.70, —0.26), respectively. Therefore the efficacy of the two doses overlapped in the two
`monotherapy studies. In the pioglitazone and the metformin add—on studies, the LSM differences
`between 100 mg and placebo in l-le1C change from baseline were —0.70 (—0.85, —0.54), and —0.65 (-O.77, —
`0.53), respectively (Fig. 1). At the End—of—Phase 2 meeting, the Agency had questioned the sponsor’s
`proposed MK—0431 daily dose of 200 mg in the Phase 3 studies. In retrospect, the MK—0431 daily 50 mg
`dose should have been included in the Phase 3 studies in patients with normal renal function.
`
`In summary, 100 mg was shown to be efficacious as add—on therapy to metformin or pioglitazone. I
`recommend that, based on the Phase 2 efficacy results showing no clear lessening of clinical benefit for
`50 mg compared to the proposed 100 mg dose, daily doses of 50 mg and 100 mg should both be made
`available to patients with Type 2 diabetes and normal renal function as monotherapy. For diabetic '
`patients with renal insufficiency, the efficacy data suggest that doses of 25 mg and 50 mg are efficacious.
`
`1.2 Overview of Clinical Program and Studies Reviewed
`
`Table 1 Summary of Clinical Studies (Phase 2 and Phase 3)
`
`Total Sample Size
`
`Patient population
`
`_
`
`# of
`Study -
`Center,
`ID
`Period
`Country
`(Phase 2 or
`
`Phase 3)
`021
`7/047/05
`h
`7)
`(P “56 )
`
`Duration
`
`Phase A: 24 wks
`Phase 13:80 \ka
`
`Phase A:18 \vks
`Phase 13:36 \vks
`
`111,
`56 U51 &
`PR3
`16 E 3
`.
`u.
`39 RD“ 4
`114,
`60 us
`37 E
`u
`_
`'
`17 ROW"
`
`100 mg 238
`200 mg 250
`1
`b 253
`P ace 0
`‘
`”"50 1111]
`Total
`741
`100 mg 205
`200 mg 206
`l
`b “0
`P RFC O
`ratio 2:211
`total
`521
`
`023
`10/04—8/05
`h
`3
`(P ase )
`
`218 to 575 years of age
`(1) “0t 0“ AHA5 (Off for 33 weeks}
`(2) on a single AHA
`(3) on a dual oral combination ( 550% maximal dose of
`both components)
`Rescue: metformin
`218 to 575 years of age
`(1) not on AHA (off for 38 “’eekS)
`(2) on a single AHA
`(3) on a dual oral combination ( 550% maximal dose of
`both components)
`Rescue: metformin
`
`.
`
`019
`
`71,
`
`100 mg+pio(30/45 mg) 175
`
`218 years of age
`
`24 wks
`
`

`

`
`
`# of
`Center,
`Country
`
`Total Sample Size
`
`Patient population
`
`Study
`ID
`Period
`(Phase 2 or
`
`Phase 3)
`7/04»9/05
`(Phase 3)
`
`Placebo+pio(30/45 mg) 178
`28 US
`Ratio 1:1
`11 Eu
`11 ROW Total
`
`-
`
`353
`
`020
`7/04-7/05
`(Phase 3)
`
`028
`12/04_
`10/05
`(Phase 2)
`
`100 mg -- MP 464
`100
`
`Placebo-- MP 237
`46 US
`__
`_
`__
`ratio 2‘1
`25 bu
`29 ROW total
`
`701
`
`57 U88:
`PR
`5 Eu
`..
`13 R0“
`
`65
`MK—0431
`Stratum 1 (not 0") dialysis)
`50 mg
`'
`,
`Stratum 2 (on dialysis)
`25 mg
`Placebo
`
`26
`
`Ratio 2:1:i\'lK»O431:placebo
`
`91
`Total
`4 doses ofMK-0431 (5, 12.5,
`83 US
`46 ROW 25, 50 mg bid), Placébor
`gllPIZlde 5 mg (CJCCFVClY
`titrated to 10 mg bid)
`.
`_
`743 randomized
`4 doses of MK—0431 25, 50,100
`(
`65 ROW ms ad or 50 bid), Placebo
`555 randomized
`
`59 US
`
`'
`
`010
`7/03—8/04
`(Phase 2b
`‘
`dose
`ranging)
`014
`
`9/03.7/04
`(Phase 2b
`close
`
`
`ranging)
`
`U54: minim,
`Hyperglycemic Agent
`
`PuroRico,
`
`(1) not on AHA
`(2) on a single AHA (PPAR or non—PPAR)
`(3) on dual oral combination (PPAR +AHA)
`.
`.
`Rescue. metformin
`
`218 to S78 years of age
`8 not 04} AJHAWA (
`f
`.
`.
`h )
`on a sing e
`’ met ormin or or er
`(3) on dual oral combination (metformin+AHA)
`R
`_
`.
`.
`escue. pioglitazone
`218 years of age
`Type 2 diabetes and chronic renal insufficiency
`(1)
`not-on oral AHA (off 28 weeks); or
`(2)
`on insulin monotherapy
`(3)
`on a single on] AHA; or
`on a low dose dual oral combination agent ( 350%
`maximal dose of both components)
`
`21 -75 years of age
`(1) not on AHA with HbA1c 26.5 to <10°/o
`(2) on monotherapv HbA1C 26 to 59%
`‘
`
`2175 years ofa e
`.
`g
`(1) not on AHA with HbArc 26.5 to <10%
`(2) on monotherapy I-le1c 26 to 39%
`
`Duration
`
`Phase Az24 wks
`Phase 13:80 wks
`
`Phase A212 wks
`Phase 13:42 \vks
`
`p
`12 weeks
`
`12 weeks
`
`
`Euopre, ROW4: Rest of the od, Wind
`
`The proposed indications for MK-0431 100 mg qd were: (1) as an adjunct to diet and exercise to
`improve glycemic control in patients with type 2 diabetes mellitus as monotherapy and (2) as
`combination therapy with metformin or a PPAR-y agonist (e.g. thiazolidinedione) when diet and
`exercise plus the single agent do not provide adequate glycemic control.
`
`2. INTRODUCTION
`
`2.1 Overview
`
`JANUVIA (sitagliptin phosphate; MK-0431) belongs to a new therapeutic class called dipeptidyl
`peptidase 4 (DPP—4) inhibitor which prolongs the half life of incretin hormones that regulate glucose
`homeostasis. The postulated effect of incretin hormones is to regulate glucose—dependent insulin
`secretion following their release into circulation from the gut in response to food intake.
`
`For the four Phase 3 studies, two MK—O431 dOSes, 100 mg and 200 mg, were tested in the monotherapy
`trials and only one dose, 100 mg was tested in the add—on trials.
`
`

`

`Patients in the monotherapy studies were 218 and E75 years of age with type 2 diabetes who either (1)
`not on AHA (Anti—Hyperglycemic Agent) (off for 28 weeks) or (2) on a single AHA; or (3) on low
`doses of dual oral combination agents (i.e., at 550% of maximal dose of both components).
`
`For combination studies, patients were either: (1) not on AHA or (2) on a single AHA (background
`AHA or non background AHA) or (3) on dual oral combination treatment with background AHA and
`another AHA.
`
`After screening (visit 1) patients either entered directly into the single—blind placebo run—in period (2
`weeks, visit 2/ 3) or washout/dose titration/ stabilization period (6—12 weeks, visit 2).
`
`The randomization criterion for HbA1c was 27% and 510% at, or within the 2 weeks prior to, the
`single-blind placebo run—in period, visit 3 /week —2.
`
`For patients not meeting specific glycemic goals during the treatment period, rescue medication was
`initiated. The rescue criteria were:
`
`1. >270 mg/dL after Visit 4/Day 1 through Visit 6/Week 6,
`2. >240 mg/dL after Visit 6/Week 6 through Visit 7/\Week 12, and
`3. >200 mg/dL after Visit 7/Week 12 up to (but not including) the final visit
`
`The primary efficacy variable was HbA1c change from baseline. The primary analysis population was the
`all—patients—treated (APT) population which included all randomized patients who have a baseline
`HbA1c and at least one post—randomization HbA1c. The imputation method to handle missing values
`was last observation carried forward (LOCF). For rescued patients the last HbA1c prior to rescue was
`carried forward in the analysis. The secondary analysis was on the completers who had a baseline and an
`end of study HbAk.
`
`The analysis used a covariance model which included treatment and AHA status as fixed effects and
`baseline HbA]C as covariate. Although the prior AHA status was not a stratification factor, the sponsor
`included it in the ANCOVA model because it is a strong predictor of change in HbAk.
`
`Figure 1 displays the placebo subtracted differences for MK—O431 by study. Figure 2 and figure 3 display
`HbA1c changes from baseline and absolute HbA1c over time, respectively for the APT population. The
`placebo groups had small increases in HbA1c mean change from baseline (0.17, 0.16) for the
`monotherapy studies and small decreases (—0.18, —0.08) in the combination studies. 200 mg MK—0431
`outperformed 100 mg MK—0431 in the 24—week monotherapy study (-0.94% vs. —0.79%) but not in the
`18—week monotherapy study (—0.48% vs. —0.60%) in LSM difference from placebo in HbA1c change. For
`the 100 mg MK-0431 proposed dose, the least squared mean differences from placebo were —0.79°/o, and
`—0.6% for the 24—week and 18—week monotherapy studies, respectively, and —0.7% and —0.65% for the
`pioglitazone and metformin add—on studies, respectively (Table 2). Similar to the APT analysis, the
`completers analyses were all statistically significant; however, the between group differences from the
`completers analysis were consistently less favorable than the APT analysis. The LSM differences
`between 100 mg and placebo were —0.65% and —0.48°/o for the 24—week, and 18—week monotherapy
`studies, respectively, and —0.63% and -0.55% for the pioglitazone add—on and metformin add—on studies,
`respectively, in the completers analysis (Table 3).
`
`Table 2_HbA1c change from baseline to study end — APT LOCF* .
`_
`Mean (SD)
`LSM (SE)
`LSM (CI)
`
`Treatment
`N Baseline
`Endpoint
`Change
`Change
`Difference from placebo
`100 mg
`i 229
`8.01 (0.88)
`7.39 (1.15)
`-0.62 (1.02)
`—0.61 (0.06)
`-0.79 (—0.96, -0.62)
`200 mg
`238
`8.08 (0.94) '
`7.31 (1.14)
`-0.78 (0.91)
`0.76 (0.06)
`-0.94 (-1.11, —O.77)
`
`' Monorherapy 24 wks
`
`

`

`WWW—WM
`
`LSM (Cl)
`LSM (SE)
`'Mean (SD)
`Treatment
`N Baseline
`Endpoint
`Change
`Change
`Difference from placebo
`
`Placebo
`244
`8.03 (0.82)
`8.2 (1.37)
`+0.17 (1.06)
`+0.18 (0.06)
`
`Monotherapy 18 wks
`
`-0.60 (-0.82, -0.39)
`-0.48 (0.07)
`~0.46 (0.85)
`7.58 (1.15)
`8.04 (0.82)
`193
`100 mg
`200 mg
`199
`8.14 (0.91)
`7.81 (1.31)
`—0.34 (0.95)
`0.36 (0.06)
`-0.48 (-0.70, —0.26)
`
`Placebo
`103
`8.05 (0%))
`8.21J1.35)
`+0.1fl0.93)
`+0.1fl0.09L
`Add on Pioglitazone
`100 mg
`163
`8.05 (0.81)
`7.17 (0.91)
`0.88 (0.70)
`0.85 (0.07)
`-0.70 (-0.85, -0.54)
`
`Placebo
`174
`8.00 (m 7.82 (1.19L -0.18 (0.7_9L
`-0.15 (0.06)
`100 mg
`453
`7.96 (0.81)
`7.26 (0.97)
`0.70 (0.72)
`0.67 (0.05)
`Placebo
`_ 224
`8.03 (0.82)
`7.95 (1.10
`—0.08 (0.89)
`—0.02 (0.06
`
`
`
`
`*values prior to rescue medication were carried forward
`
`Add on Metformin
`
`-0.65 (-0.77, -0.53)
`
`
`
`Table 3 HbA1c change from baseline to study end — Completers*
`WW“
`Mean (SD) LSM (CI) LSM (SE)
`7
`Treatment N
`Baseline
`Endpoint
`Change
`Change
`Difference from placebo
`Monotherapy 24 wks
`100 mg
`189
`7.92 (0.86)
`7.13 (0.87)
`~0.79 (0.89)
`0.76 (0.06) —0.65 (-0.82, —0.49)
`200 mg
`198
`8.04 (0.87)
`7.14 (0.95) —0.90 (0.86) —0.86 (0.06)
`—O.75
`(—0.91, —0.58)
`
`Placebo
`176
`7.88 (0E)
`7.76 (1.09) —0.12 (0.93)
`0.11 (0.06)
`I
`
`
`
`Monotherapy 18 wks
`
`—0.48 (0.71, -0.26)
`—0.59 (0.06)
`—0.57 (0.77)
`7.4 (0.98)
`7.96 (0.77)
`168
`100 mg
`»O.45 (0.07) —0.34 (0.57, —0.12)
`0.43 (0.90)
`7.58 (1.19)
`8.02 (0.86)
`161
`200 mg
`
`Placebo
`74
`7.9 (0.86L 7.5m118) —0.0fl0.87) —0.10 (0.10)
`
`—0.63 (0.79, —0.47)
`—0.91 (0.07)
`6.98 (0.75) —0.95 (0.70)
`7.94 (0.73)
`131
`100 mg
`Add on Pioglitazone
`
`Placebo
`136
`7.86 (07—7)
`7.57 (0.87)
`-0.29 (0.73)
`-0.28 (0.07L
`Add on Metformin
`100 mg
`399
`7.91 (0.76)
`7.12 (0.79)
`—0.80 (0.66)
`—0.81 (0.05)
`-0.55 (—0.67,-0.44)
`Placebo
`171
`7.92 0.76)
`7.66(0.89
`-0.26 0.84
`0.26 0.06
`
`
`
`* panstit withbaseline and sdy end HbA1
`'
`'
`' "
`...,
`H
`H
`'
`"
`
`Figure 1 LSM difference V(C.I.) between MK—0431 and placebo by study - APT
`
`Add on PIO
`
`100 mg 1
`Add on MF
`
`100 mg
`Mono 24 wks
`
`—I—
`
`—I—
`
`100 mg —I—
`
`200 mg 1 —-I—
`Mono 18 wks
`‘
`
`
`
`
`
`
`
`100 mg j
`
`. —.—.
`
`200 mg f
`l
`
`—I—’—-
`
`Figure 2 HbA1C change from baseline over time — APT, LOCF
`
`

`

`24-wk Mono (021)
`
`
`
`
`
`Add on Pic (019)18-wk Mono (023) Add on MF(020)
`
`PROTOCOL
`
`
`
`WEEK
`
`
`
`e
`
`1218 24
`WEEK
`
`o
`
`6
`
`1218 24
`WEEK
`
`'
`
`121824
`
`APPEARS‘THIS WAY
`0N ORIGINAL
`
`

`

`Figure 3 HbA;c (”/o) over time — APT, LOCF
`PROTOCOL
`
`
`
`24-wk Mono (021) Add on MF(020) 18-wk Mono (023) Add on Pic (019)
`
`
`
`
`
`
`
`HbA1c("/a)
`
`2.2 Data Sources
`
`The eCTD is located at: \ \ CDSESUBl \ E\r"SPR()D\ 33021995. The location for datasets is
`datasets\
`atf‘x\cdsesubl\evs rod‘xNOZl995‘\0(’)()0\m§‘
`. The data folders for the 4 Phase 3 trials were
`
`
`
`
`
`p019, p020v1, p021vl and p023vl.
`
`'
`
`3. STATISTICAL EVALUATION
`
`3.1 Evaluation of Efficacy
`
`Study Design and Endpoints
`
`All 4 efficacy studies were multinational, double-blind, randomized and placebo—controlled to demonstrate
`efficacy of MK—0431 compared to placebo in glycemic control in patients with type 2 diabetes on diet and
`exercise. Patients not reaching specific glycemic goals during the placebo—controlled treatment period were to
`have rescue therapy initiated. The primary efficacy variable was l-le1c change from baseline to endpoint.
`
`The sample size was greater than planned since more patients were eligible for randomization after the long
`run—in period. The planned samples size was sufficient to detect with 298% power, a treatment mean
`difference of 0.5% (SD 1.0) in HbA1c change from baseline with an oc=0.05 and assuming a <10% dropout
`rate.
`
`Table 4 displays the planned sample size and randomized patients for the 4 studies.
`
`-
`
`Table 4 Summary of sample size
`
`.
`
`021
`
`Planned
`
`
`
`CA
`00
`
`
`
`0U1 O
`
`
`
`DJ
`00
`Add-0n n-iolitazone
`019
`
`
`
`Add—on metformin
`701
`2:1
`
`
`
`
`N N
`
`The efficacy analysis only used data before rescue medication. The sponsor did not perform sensitivity
`analysis using post rescue data.
`
`3.1.1 Monotherapy — 021 (24-week)
`
`

`

`The 2 monotherapy studies 021 and 023 included patients 18 to 75 years of age who were not currently on an
`AHA or on monothe'rapy (or low dose combination therapy at 550% of maximum dose of either agent).
`Patients with HbA1c 27% and 510% were randomized in a ratio of 1:1:1 for P021 and 1:22 for P023 to
`placebo:100 mg: 200 mg MK—0431 qd The double-blind treatment period (phase A) was 24 weeks for Study
`021 and 18 weeks for Study 023. The phase B study (not reported in this submission) was 80 weeks for Study
`P021 and 36 weeks for Study P023.
`
`The design of the study included a screening diet/exercise run—in period of up to 15 weeks (including a 1—
`week screening period [Visits 1 to 2], up to a 12-week diet/exercise run—in period and antihyperglycemic agent
`(AHA) “wash—out” for patients on AHAs [Visits 2 to 3] and a 2—week single—blind placebo run—in period
`[Visits 3 to 4]) prior to randomization.
`
`The monotherapy studies screened type 2 diabetes who were either (1) not on AHA (off for 28 weeks); or (2)
`on a single AHA; or (3) on low—dose dual oral combination therapy (i.e., at 550% of maximum labeled dose
`of both components). Eligible patients had HbA1c 27% and 510% at or within 2 weeks prior to Visit
`3/Week —2. The exclusion criteria for fasting glucose was >260 mg/dL. Patients not on AHA with HbA1c
`27% and 510% and met all other enrollment criteria were to directly enter the 2—week single—blind placebo
`run—in period at a combined Visit 2/ 3. Table 5 displays guidelines for run—in period management (Sponsor’s
`table 9—1).
`
`Table 5 Guidelines for Run-in Period Management
`Only patients at Screening Visit/Visit I considered by the investigator as likely to meet Visit 3 HbA1c inclusion
`criteria (based upon evaluation of patient’s current diet and exercise regimen, medication regimen, and the
`
`Visit 1 HbA1c level) may have been continued in study.
`Patient’s Medication and Dose at
`Visit l/Screening
`Screening
`Visit HbA1c level
`Patient not on antihyper—
`27% and 510%
`glycemic agent therapy (off for
`>10%
`28 weeks)
`
`Visit 2 to Visit 3 Duration
`
`Go directly to combined Visit 2/3
`Up to 6 weeks
`6 weeks
`
`On antihyperglycemic therapy
`(monotherapy or low dose
`combination)
`
`
`'
`
`27% and Sl0%
`
`26% and <7%
`
`
`Note: Patients on a TZD at Screening Visit/Visit I
`must have 8 weeks between Visit 2 and Visit 3
`6 to 10 weeks
`Note: Patients on a TZD at Screening Visit/Visit 1
`may have up to l2 weeks with a minimum of
`8 weeks between Visit 2 and Visit 3
`
`Patient Disposition, Demographic and Baseline Characteristics —— 24-wk monotherapy
`
`A total of 1807 patients were screened and 741 patients were randomized at 104 sites worldwide. This
`exceeded the 600 protocol planned sample size. The US randomized 42% of the patients, Costa Rica, 9% and
`<5% for the rest 16 countries. Of the 310 patients randomized in the U.S. sites, 66% were White, 19% were
`Hispanic and 12% were Black.
`
`

`

`Figure 4 Number of patients randomized by country
`
`
`
`
`
`
`
`
`
`
`
`Approximately 14% randomized patients withdrew from the study (Table 6). The most frequent reason for
`withdrawal was patient consent (5.1%).
`’
`
`Country
`
`Table 6 Patient disposition - Monotherapy Study 021
`
`
`K0M—41 100 mg 4K0431 00 Plcebo '
`’ Total
`'
`'
`
`n=238
`n=250
`n=253
`n=741
`
`639 (86.2%)
`216 (85.4%)
`214 (85.6%)
`209 (87.8%)
`Completed Phase A*
`102 (13.8%)
`37 (14.6%)
`36 (14.4%)
`29 (12.1%)
`Discontinued
`13 (1.8%)
`4 (1.6%)
`4 (1 .6%)
`5 (2.10 0)
`Clinical AB
`1 (0.1%)
`1 (0.4%)
`0
`0
`Lab AB
`17 (2.3%)
`9 (3.6%)
`5 (2.0%)
`3 (1.3%)
`Lack efficacy
`11 (1.5%)
`2 (0.8%)
`4 (1.6%)
`5 (2.1%)
`Lost to follow—up
`10 (1.3%)
`5 (2.0%)
`3 (1.2%)
`2 (0.8%)
`Other
`5 (0.7%)
`1 (0.4%)
`1 (0.4%)
`3 (1.3%)
`Pat. Moved
`38 (5.1%)
`11 (4.4%)
`17 (6.8%)
`10 (4.2%)
`Pat.\X/ithdrew consent
`
`Protocol dev 7 (0.9%) 1 (0.4%) 2 (0.8%) 4 (1.6%)
`
`
`
`* Including patients rescued but not discontinued
`
`Percentages of patients rescued were 9% (21 /237), 5% (12/250), and 21% (52/253), for 100 mg, 200 mg, and
`placebo, respectively.
`
`Table 7 (from sponsor’s Table 10—3) displays the accounting of patients in the efficacy analysis populations
`for HbA1c.
`
`
`
`Table 7 Patients accounting in efficacy analysis population
`
`
`i
`' MK—0431 "
`'
`MK3—041
`"
`" Placebo
`‘
`'
`Total
`
`100 mg
`200 mg_
`
`238
`250
`253
`741
`
`Total Randoiezd
`
`'
`
`‘
`
`i
`
`‘
`
`‘
`
`71 I (96.0)
`244 (96.4)
`238 (95.2 )
`229 (96.2 )
`Included in the APTT Analysis
`563 (76.0)
`176 (69.6)
`198 (79.2 )
`189 (79.4 )
`Included in the Completers Analysis
`30 ( 4.0)
`9 ( 3.6)
`12 ( 4.8)
`9 ( 3.8)
`Excluded from the APT}L Analysis
`2 (0.3)
`0 ( 0.0)
`0 ( 0.0)
`2 ( 0.8)
`'
`No Baseline Data
`
`0"."ea‘mem
`. 7(1)
`9(3-5. .....
`. 2803-8)
`.
`..
`.
`
`10
`
`

`

`
`Total Randomized
`MK—0431
`MK-0431
`Placebo
`Total
`
`100 mg
`200 mg__
`
`238
`250
`253
`741
`
`148 (20.0)
`68 (26.9)
`40 (16.0)
`40 (16.8)
`Excluded from the Completers Analysisi
`Rescued Prior to Week 24§ No Data at
`72(9.7)
`45(17.8)
`10(4.0)
`17(7.1)
`
`Week 24” 76(10.3) 23(9.7) 30(12.0) 23(9.1)
`
`
`
`1' APT: All—Patients-Treated.
`:1; The completers population is a subset of the APT population including all patients with \‘(r’eek 24 data.
`§ Efficacy data obtained on a patient after initiation of rescue therapy are treated as missing.
`ll For patients not on rescue medication.
`
`Table 8 (from sponsor’s Table 10-4) displays patient demographics.
`
`The mean age of patients was 54.2 in years. More males were in the 100 mg group (57%) than in the 200 mg
`group (47%). Approximately 50% were Caucasian, 24% were Hispanic, 14% were Asian and 5% were Black.
`The mean baseline weight was 84.6 kg.
`
`Table 8 Patient demographics
`
`Age (years)
`Treatment
`MK-0431 100 mg
`MK-0431 200 mg
`Placebo
`All
`> Gender
`
`N
`238
`250
`253
`741
`
`Mean
`53.4
`54.9
`54.3
`54.2
`
`Sd
`9.5
`10.1
`10.1
`9.9
`
`Median
`55.0
`55.0
`54.0
`54.0
`
`Range
`24.0
`18.0
`23.0
`18.0
`
`to 75.0
`to 75.0
`to 75.0
`to 75.0
`
`Treatment
`MK—0431 100 mg
`MK—O431 200 mg
`Placebo
`All
`Race
`
`Treatment
`MK—0431 100 mg
`MK—O431 200 mg
`Placebo
`A11
`
`Male
`
`. N (%)
`136 (57.1)
`117 (46.8)
`130 (51.4)
`383 (51.7)
`
`Female
`
`N (%)
`102 (42.9)
`133 (53.2)
`123 (48.6)
`358 (48.3)
`
`Total
`
`N
`238
`250
`253
`741
`
`White
`
`N (%)
`122 (51.3)
`132 (52.8)
`127 (50.2)
`381 (51.4)
`
`Black
`
`N (%)
`10 (4.2)
`12 (4.8)
`16 (6.3)
`38(5.1)
`
`Hispanic
`
`Asian
`
`N (%)
`58 (24.4)
`53 (21.2)
`64 (25.3)
`175 (23.6)
`
`N (%)
`32 (13.4)
`37 (14.8)
`34 (13.4)
`103(13.9)
`
`Other
`
`N (%)
`16 (6.7)
`16 (6.4)
`12 (4.7)
`44 (5.9)
`
`Total
`N
`238
`250
`253
`741
`
`Baseline Body Weight (kg)
`Treatment
`N
`238
`MK—0431 100 mg
`250
`MK-0431 200 mg .
`Placebo
`253
`Al]
`741
`
`Mean
`85.0
`83.7
`85.0
`84.6
`
`SD
`18.4
`19.2
`18.1
`18.5
`
`Median
`83.3
`83.8
`83.3
`83.4
`
`44.5
`45.0
`49.9
`44.5
`
`Range
`to 138.7
`to 145.4
`to 137.1
`to 145.4
`
`Baseline Body Mass Index (kg/m2)
`Treatment
`N
`Mean
`Range
`SD Median
`237
`30.3
`to 43.3
`20.4
`29.7
`MK-0431 100 mg
`250
`30.3
`to 43.0
`19.1
`29.8
`MK—O431 200 mg
`Placebo
`252
`30.8
`30.1
`20.2
`to 44.7
`All
`739 30.5
`
`29.8
`19.1
`to 44.7
`
`11
`
`

`

`Table 9 (Table 10—5 from the sponsor) displays baseline efficacy characteristics. Baseline mean HbA1c (SD)
`was 8.0% (0.9). The range was 6.3 to 10.9%. Approximately 50% of patients used AHA at baseline. Median
`duration of type 2 diabetes was 3 years (range 0 to 38 years).
`Table 9 Baseline characteristics
`WWW
`Baseline HbAlc (%)
`
`Treatment
`MK-0431 100 mg
`
`MK-0431 200 mg
`
`N
`236
`
`250
`
`Mean
`8.0
`
`8.1
`
`8.0
`253
`Placebo
`8.0
`739
`All
`Distribution of HbAlc at Baseline
`
`SD
`0.9
`
`0.9
`
`0.8
`0.9
`
`Median
`7.8
`
`7.9
`
`7.9
`7.9
`
`Range
`6.610 10.9
`
`' 6.3 to 10.5
`
`6.6 to 10.7
`6.3 to 10.9
`
`Treatment
`
`MK-0431 100 mg
`
`MK-0431 200 mg
`
`Placebo
`
`N
`
`236
`
`250
`
`253
`
`739
`All
`Baseline FPG (mg/dL)
`
`Number (%) of Patients with Baseline l-leI c
`28 and <9%
`
`29%
`
`<8%
`
`135 (572)
`
`129 (51.6)
`
`132 (52.2)
`
`396 (53.6)
`
`62(26.3)
`
`69 (27.6)
`
`85 (33.6)
`
`216 (29.2)
`
`'
`
`39 (16.5)
`
`52 (20.8)
`
`36 (14.2)
`
`127 (17.2)
`
`Treatment
`MK-0431 100 mg
`
`N
`238
`
`Mean
`170.7
`
`SD'
`43.0
`
`46.2
`
`Median
`165.0
`
`171.0
`
`Range
`94.0 to 427.0
`
`86.0 to 409.0
`
`MK-O431 200 mg
`
`249
`
`7174.2
`
`176.1
`253
`Placebo
`173.7
`740
`All
`Baseline Fasting Insulin (microIU/mL)
`
`Treatment
`MK-0431 100 mg
`
`MK-0431 200 mg
`
`N
`233
`
`247
`
`Mean
`14.5
`
`14.1
`
`41.8
`43.7
`
`SD
`13.1
`
`11.8
`
`10.8
`14.9
`252
`Placebo
`11.9
`14.5
`732
`All
`Duration of Type 2 Diabetes Mellitus (years)
`
`Treatment
`MK-0431 100 mg
`
`MK-0431 200 mg
`
`N
`238
`
`250
`
`Mean
`4.3
`
`4.3
`
`SD
`4.9
`
`4.7
`
`4.7
`4.6
`252
`Placebo
`4.8
`4.4
`740
`All
`Use of Anti-Hyperglycemic Medication at Screening
`Present
`Absent
`
`Treatment
`MK-O431 100 mg
`
`N (%)
`114 (47.9)
`
`N (%)
`124 (52.1)
`
`MK-0431 200 mg
`
`125 (50.0)
`
`125 (50.0)
`
`172.0
`168.5
`
`Median
`11.8
`
`10.7
`
`12.2
`11.5
`
`Median
`3.0
`
`3.0
`
`3.0
`3.0
`
`Total
`
`N
`238
`
`250
`
`253
`129 (51.0)
`124 (49.0)
`Placebo
`A11 741mm363 (49.0) 378 (51.0)
`
`
`
`
`
`
`
`73.0 to 283.0
`73.0 to 427.0
`
`Range
`0.1 to 138.9
`
`0.5 to 89.9
`
`1.5 to 76.0
`0.1 to 138.9
`
`7 Range
`0.0 to 38.0
`
`0.0 to 30.0
`
`0.0 to 35.0
`0.0 to 38.0
`
`12
`
`

`

`
`
`Prevalence of Metabolic SyndromeT
`Present
`
`Treatment
`MK-0431 100 mg
`
`N (%)
`139 (58.4)
`
`.
`
`Absent
`
`N (%)
`99 (41.6)
`
`MK-0431 200 mg
`
`150 (60.0)
`
`100 (40.0)
`
`Total
`
`N
`238
`
`250
`
`I
`
`253
`84 (33.2)
`169 (66.8)
`Placebo
`741
`283 (38.2)
`458 (61.8)
`All
`Using the definition of the National Cholesterol Education Program
`
`Results and Conclusions
`
`The primary efficacy variable was I-le]C change from baseline at Week 24. The secondary efficacy variables
`were change from baseline in Fasting Plasma Glucose (FPG) and change from baseline in 2—hour post—meal
`glucose at Week 24. For HbA1c the hierarchical‘testing procedure compared the 100 mg group first then the
`200 mg group to the placebo group. The testing procedure for the secondary efficacy variables proceeded in
`the order of FPG then 2-hour post-meal glucose conditioned on the prior test showed statistically
`significance at 0:20.05. Both treatment groups were statistically significantly better than placebo in HbA1c
`change from baseline (Table 10, Fig. 5). The placebo subtracted least squared mean difference in HbA1c
`change from baseline was —0.79 0/o for the 100 mg group and —0.94% for the 200 mg group. For completers,
`the differences from placebo were —O.65°/o and —0.75°/o for 100 mg and 200 mg. respectively (Table 1 1 and Fig.
`6).
`
`Table 10 ANCOVA results of HbA1c (%) change from baseline at week 24 - 24—week Monotherapy
`
`Treatment N
`
`LSM change (SE) Difference from placebo (CI)
`
`p—value
`
`Week 24
`Baseline
`Mean (SD) Mean (SD)
`—0.61 (0.06)
`8.01 (0.88)
`7.39 (1.15)
`229
`100 mg
`—O.76 (0.06)
`8.08 (0.94)
`7.31 (1.14)
`238
`200mg
`
`Placebo
`244
`8.03 (0.82)
`8.20 (1.37)
`0.18 (0.06)
`ANCOVA model included factors treatment and prior AHA and baseline HbAlc as covariate
`Table 11 ANCOVA for- HbA1c (%) change from baseline at week 24 — Completers
`
`
`
`
`L%SE)
`
`
`
`_- '
`Week 18
`Change
`Difference from flcebo (Cl)
`p—value
`
`
`
`
`100 m
`7.92 (0.86)
`7.13 (0.87)
`-0.76 (0.06) —0.65 (-0.82, —0.49)
`<0.001
`
`—O.86 (0.06)
`200m
`_ E4191, -0.58)
`<0.001
`
`
`
`
` Placebo
`7.76 (1.09)
`-0.11(0.06)
`176
`7.88 (0.75)
`
`—O.79(—0.96,—0.62)
`_0.94(-1.11,_0.77)
`
`<o.001
`<0.001
`
`
`
`189
`
`Figure 5 HbA1C (%) change from baseline over time — APT
`
`APPEARS 711:3 WAY
`00, GRiGlNAL
`
`13
`
`

`

`PROTOCOL
`021
`
`---- Placebo
`
`TREATMNT:
`100 mg
`—-—- 200 mg
`
`WEEK
`
`Figure 6 HbA1c (°/o) change from baseline by completion status
`Corn pleter
`
`No
`Yes
`
`
`
`
`
`
`
`
` 24-weekmonotherapy
`
`TREATMNT:
`MK-0431 100 mg '
`MK-0431 200 mg
`— - Placebo
`
`24
`
`
`
`0
`
`6
`
`12
`
`18
`
`Treatment—by—prior AHA stratum interaction was significant (p20.05). The LSM difference between MK—
`0431 group and placebo for each stratum is in Table 12. Figure 7 displays the HbA1c change from baseline by
`stratum over time. The LSM increased over time only in placebo patients (0.55%) with their AHA washed out
`before randomization (Prior treated). Figure 8 displays that the interaction was quantitative but not qualitative
`in nature.
`
`Table 12 HbA1c change from baseline by prior medication stratum
`
`TREATMNT
`100 n_ig_
`Yes
`No
`Yes
`No
`PRIORMED
`(n=117)
`(n=121)
`(n=108)
`(n=121)
`n
`-0.61 (0.09)
`—0.92 (0.09)
`—0.38 (0.09)
`—0.85 (0.09)
`LSM (StdErr)
`.1.15(0.12)
`—0.74 (0.12)
`-0.93(0.13)
`—0.66 (0.12)
`LSMdiff(SE)from placebo
`
`(—0.97,_0.50)
`(-1.18,-0.68)
`(_0.90,—0.42)
`LSMdiffCl.
`.1.40,»0.91)
`
`
`
`Placebo
`
`No
`(n=126)
`-0.18 (0.08)
`
`Yes
`(n=118)
`0.55 (0.09)
`
`200 mg
`
`Figure 7 HbA]C change from baseline over time stratified by prior AHA use
`
`_
`
`14
`
`

`

`PROTOCOL
`021
`
`PRIORM ED
`
`
`
`
`
`TREATMNT:
`100 mg
`— ‘ — - 200 mg
`
`— — - - Placebo
`
`
`
`WEE(
`
`APPEARS THlS WAY
`0N ORIGINAL
`
`15
`
`

`

`Figure 8 Mean (95% CI) in HbA1c change from baseline at week 24 by treatment and prior AHA use
`PROTOCOL
`021
`
`ac
`
`o
`
`n=126 0
`
`°
`
`”
`
`n=108
`n=lZl
`
`c
`
`”
`
`”
`
`. n= 1 l 7
`n=121
`
`n= 1 18
`
`T
`
`
`
`TMNT:
`100mg
`200 mg
`Placebo
`
`-1
`
`u
`
`N
`
`B 100 mg: Yes
`:2:
`z
`lOOmg:No
`
`9z
`
`.
`7
`-00 mg. Yes
`92
`E 200mg:No
`[—
`2
`E Placebo: Yes
`"‘ Placebo: No
`
`95% t-lnlervals for mean BASECHNG
`
`Treatment—by—baseline HbA1c was significant (p<0.001). Figu