‘ CENTER FOR DRUG EVALUATION AND
`
`RESEARCH
`
`APPLICA TION NUMBER:
`
`22-030
`
`STATISTICAL REVIEW! S)
`
`

`

`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:
`
`22-030
`
`Drug Name:
`
`Indication(s):
`
`Applicant:
`
`Date (s):
`
`ToviazTM (Fesoterodine Fumarate)
`
`Treatment of Overactive Bladder.
`
`Pfizer Global Pharmaceuticals.
`
`Submitted: 5/02/2008
`
`PDUFA: 11 /2/2008
`
`Review Priority:
`
`Standard
`
`Biometrics Division:
`
`Division of Biometrics III (HFD—725)
`
`Statistical Reviewer:
`
`Mahboob Sobhan, PhD. (HFD—725)
`
`Medical Division:
`
`Division of Reproductive and Urological Drug Products
`(HFD-580)
`
`Clinical Team:
`
`Mark Hirsch, M.D. (HFD-580)
`
`Project Manager:
`
`Ceilia Peacock (HFD—SSO)
`
`Harry Handlesman, M.D. (HFD-580)
`
`Keywords:
`
`NDA review, Clinical studies.
`
`

`

`NDA 20—030: Toviaz®
`Statistical Reviewer’s comment
`
`This submission pertains to revised labeling, in most part, the clinical pharmacology and adverse
`reactions section of the label. There was no new efficacy data submitted for our statistical review.
`We agreed with the revised efficacy results in the clinical trial Section. From a statistical perspective,
`there are no further efficacy comments pertaining to this submission.
`
`Appears This Way
`On Original
`
`

`

`This is a representation ofan electronic record that was signed electronically and
`this page is the manifestation of the electronic signature.
`
`Mahboob Sobhan
`
`10/20/2008 04:36:11 PM
`BIOMETRICS
`
`

`

`STATISTICAL EVALUATION OF NEW NDAs - FILEABILITY
`
`NDA:
`
`Drug Name:
`Sponsor:
`Indications:
`
`Medical Officer:
`Statistician:
`Project Manager:
`Submission Date:
`
`'
`
`22-030
`
`Fesoterodine
`Schwarz Pharma.
`Treatment of overactive bladder.
`
`Suresh Kaul, M.D., HFD-580
`Mahboob Sobhan, Ph.D., RFD—715
`'Jean Makie
`3/17/2006
`
`45 day Meeting Date:
`
`5/10/2006
`
`Two P3 controlled studies constitutes the main database to support the above indication. In
`addition, efficacy data from three ongoing open—label studies wee also submitted. The summary
`of the two P3 studies are as follows:
`
`Brief Summary of Controlled Trials
`
`No. of Patients
`
`Duration of
`
`Europe,
`Australia, New
`Zealand, South
`Africa
`
`Randomized/
`Treatments
`
`Total: 1 135
`
`Placebo: 285
`Feso 4mg; 272
`Feso 8mg: 288
`Tolt 4mg: 290
`
`Total: 836
`
`Placebo: 274
`Feso 4mg: 283
`Feso 8mg: 279
`
`0 Health outcomes
`
`Treatment,
`
`Endpoints (P-value*)
`V
`
`M:
`0 Change in micturitions
`0 Change in incontinence episodes
`
`Secondary
`0 Change in urge incontinence
`- Volume voided
`- Health outcomes
`
`My
`0 Change in micturitions
`0 Change in incontinence episodes
`'
`
`Secondag
`- Change in urge incontinence
`0 Volume voided
`'
`
`

`

`The following items were checked to determine the fileability conclusion.
`
`Check
`Yes, No, N/A
`Comments:
`Indechiemremtomcatewamem _—
`Original protocols and subsequent amendments
`includedin the submission.
`Designs utilized appropriate for the indications
`re uested.
`
`Endpoints and methods of analyses spelled out in the
`ro.tocols
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Effects of dro outs on rima
`anal ses investi;.ated
`
` Integrated summary of safety and efficacy included.
`
`
`
`
`
`Interim analyses (if present) plannedin the protocol
`and a uro -riate ad'ustmentsin Si-
`ificance level made
`
`Appropriate references included for novel statistical
`methodolog’
`if resent
`
`Sufficient data listings and intermediate analysis tables
`to nermit a statistical review
`
`Data from primary studies on diskettes and/or eCTD
`submitted
`
`Not planned
`
`Safety only
`
`
`
`
`
`
`Conclusion
`
`After the preliminary review ofthe submission, we have not identified any deficiencies that
`would be a reason for refuse-to-file. The sponsor provided the required information in this NDA
`to perform statistical evaluation and therefore, this NDA is fileable.
`
`,
`
`Mahboob Sobhan, Ph.D.
`Mathematical Statistician
`
`Division of Biometrics 3, HFD-725
`
`

`

`This is a representation of an electronic record that was signed electronically and
`this page is the manifestation of the electronic signature.
`'
`
`Mahboob So’bhan
`7/26/2006 03 :53 :48 PM
`BIOMETRICS
`
`

`

`”(4)
`
`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:
`
`22-030
`
`Drug Name:
`
`Indication(s):
`Applicant:
`
`Date (3):
`
`\\ (Fesoterodine Furnarate)
`
`Treatment of Overactive Bladder.
`Schwarz Pharma.
`
`i
`
`Submitted: 3 / l 7 /2006
`
`PDUFA: 1/27/2007
`
`Review Priority:
`
`Standard
`
`Biometrics Division:
`
`Division of Biometrics III (HFD—725)
`
`Statistical Reviewer:
`
`Mahboob Sobhan, PhD. (HFD-725)
`
`Concurring Reviewer:
`
`Lisa Kammerrnan, PhD. (HFD—725)
`
`Medical Division:
`
`Division of Reproductive and Urological Drug Products
`(HFD-580)
`'
`
`Clinical Team:
`
`’
`
`Mark Hirsch, M.D. (HFD-580)
`
`Suresh Kaul, M.D. (HFD—SBO)
`
`Project Manager:
`
`jean Makie (HFD—580)
`
`Keywords:
`
`NDA review, Clinical studies, ANCOVA, Multiple
`comparisons/Multiplicities, Hierarchical Closed—testing
`Procedure.
`
`

`

`NDA 20—030:
`
`_ \.
`
`6!)
`
`M4}
`
`Table of Contents
`
`EXECUTIVE SUMMARY .......................... 4
`1.0
`1.1 Conclusion and Recommendations ................................................................................................... 4
`1.2 Brief Overview of Clinical Studies ..................................................................................................... 4
`
`1.3
`
`Statistical Issues and Principal Findings .................................................................... 5
`
`INTRODUCTION ............................................................................................................................ 6
`2.0
`Overview .............................................................................................................................................. 6
`2.1
`2.2 Data Sources ........................................................................................................................................ 6
`2.3
`Indication ....................................................................................................................... -...................... 6
`
`3.0
`
`STATISTICAL EVALUATION ...................................................................................................... 7
`
`Overview of Study SP583 and Study SP584 ................................................................................... 7
`3.1
`3.1.1
`Design and Objectives .............................................................................................................. 7
`3.1.2
`Reviewer’s Comments on the Design ...................................................................................... 9
`3.2 Results: Study SP583 ............................................................................................................................ 9
`3.2.1
`Subject Disposition ..................................................................................................................... 9
`3.2.2
`Patient demographics and baseline characteristics .............................................................. 10
`3.2.3
`Primary Efficacy ....... . ............................................................................................................... 10
`3.3
`Results: Study SP584......................................................................................................................... 14
`3.3.1
`Subject Disposition .................................................................................................................. 14
`3.3.2
`Patient demographics and baseline characteristics .............................................................. 14
`3.3.3
`Primary Efficacy....................................................................................................................... 15
`3.3.4
`Secondary Efficacy .................................................................................................................... 1 5
`3.3.5
`Efficacy at: Week 2 ................................................................................................................... 16
`3.3.6
`Adjustment for Multiple Comparisons/Multiplicities
`........................ 16
`3.3.7
`Revie'wer’s Comments on the Efficacy Results ................................................................... 16
`
`4.0
`
`SUMMARY AND CONCLUSIONS .............................................................................................17
`
`

`

`®
`, M
`NDA 20—0302‘
`1.0
`EXECUTIVE SUMMARY
`
`M4}
`
`1.1
`
`Conclusion and Recommendations
`
`Based on the efficacy data submitted from two Phase 3 studies, our analysis showed that at week 12,
`compared with placebo, both doses (4 and 8mg) of \\»/‘—‘ significantly
`(p<.05) reduced the average number of micturitions and number of urge incontinence episodes.
`“g“ 8mg was also statistically significant in the reduction of
`endpoints as early as week
`v
`*
`2,
`<“
`\
`
`L.
`
`'
`
`i
`
`2
`
`M4)
`
`From a statistical perspective, this application provided adequate data to support the efficacy of
`\Smg ,a and 4mg at week 12 in the treatment of overactive bladder
`symptoms.
`
`1.2
`
`Brief Overview of Clinical Studies
`
`The applicant, Schwarz BioSciences, GmbH, reports efficacy and safety data from two Phase 3
`clinical trials (studies SP583 and SP584) to support sustained—release ———in the treatment of
`overactive bladder symptoms (OAB). Study SP583 was conducted at 150 sites in Europe, South
`Africa, Australia and New Zealand while study SP584 was conducted at 50 sites in the United States.
`Both studies were parallel-group, randomized, double—blind, double—dummy, placebo—controlled
`studies, except for the inclusion of an additional active—controlled arm in study SP583. Subjects with
`signs and symptoms of OAB syndrome and with urinary urgency, but with or without incontinence
`
`for at least 6 months prior to enrollment, were randomized to receive either V": 4mg,
`,8
`mg/day, tolterodine 4mg/day (study SP583 only), or placebo during the 12—week treatment period.
`
`M4}
`
`The protocol—specified primary efficacy endpoints included two co-prirnary endpoints: change in the
`average number of micturitions and urge incontinence episodes per 24 hours from baseline to week
`12 of the treatment period. Both the outcomes were measured by a daily diary, where subjects
`recorded at least 3 consecutive days of number of micturitions, urge incontinence episodes, and the
`number of voidings per week during the course of the trial. The secondary efficacy endpoints were
`the change from baseline in the number of voidings, a responder analysis using a treatment benefit
`scale, change in the average number of micturitions during the day and sleeping time, change in
`severity of urinary urgency, and change in number of incontinent days.
`
`i The objective in both the studies was to demonstrate that sustained-release/ 4 and Sing/day
`is superior to placebo with respect to the two co—primary endpoints. Both studies were designed to
`detect a difference of Z 0.72/day in the mean change in micturitions and 20.57/day in the mean
`change in urge incontinence. A sample size of 270 per arm (adjusting for drop outs), for a total of
`1080 in study SP583 and 810 in study SP584 was determined to be adequate to test the superiority
`hypothesis with 90% power. At the completion of the trial, a total of 1135 subjects were treated in
`study SP583 and 836 were treated in study SP584, respectively.
`
`hi4?
`
`4
`
`

`

`NDA 20—030.-\n®_
`1.3
`Statistical Issues and Principal Findings
`
`‘ M4}
`
`Our review focused on several statistical issues: the impact of missing post—baseline diary data,
`adjustment for multiple comparisons (pair-wise comparison of each dose group versus placebo),
`multiplicity (multiple endpoints), and adequacy of study power with regards to all primary endpoints.
`Missing diaries were reported in less than 7% of the subjects (ranging from 3% to 7% across
`treatment groups) and did not appear to follow any missing pattern, i.e., missing either due to
`adverse events or lack of efficacy. The efficacy results using last—observation—carried—forward
`approach (LO'CF) and per protocol (completers at endpoint) analysis population were similar. The
`sponsor’s closed—testing procedure to control the false positive error rate'for multiple
`Comparison/multiplicity (co—primary endpoints) was acceptable. However, no such procedure was
`planned for evaluating efficacy at different time points (weeks) and for the secondary endpoints.
`Although, as per protocol we agreed that a closed—testing procedure is an appropriate method,
`alternative methods could be more intuitive and simpler to use. Under such an alternative method,
`there would be no need to order the sequence of family of hypotheses, which in some cases may not
`be appropriate from a‘clinical perspective, because efficacy must be demonstrated on both
`endpoints. Therefore, for exploratory and consistency purposes, we performed an alternative simple
`adjustment for multiple comparisons using Dunnett’s test to test for each endpoint separately.
`
`Based on the applicant’s data and our independent analysis, the efficacy results could be summarized
`as follows:
`
`(1)
`
`(2)
`
`(3)
`
`At week 12, compared with placebo, both doses of \ (4 and 8mg) treatment
`resulted in a reduction in both the co—primary endpoints: the change in the average number
`of micturitions and the average number of urge incontinence episodes. Our analysis showed
`that both doses were statistically significantly superior to placebo (p<.05, adjusting for
`multiple dose/multiple endpoints).
`
`its
`
`At week 12, compared with placebt , \. _ 4 and 8mg were also significantly superior
`to placebo in the improvement of the secondary endpoint:
`the voided volume per
`micturitions in study SP583, but not in study SP584, where the 4mg dose was not statistically
`significantly superior to placebo (p<.05, adjusting for multiple dose/multiple endpoints).
`
`At week 2, compared with placebo, <‘\~- 8mg was also effective in reducing f
`
`‘~\» incontinence episodes in both studies.
`4mg dose of
`
`was effective only in reducing the incontinence episodes (p<.01)
`,
`,
`.
`.
`
`m V
`
`.
`
`' 33(4)
`
`M4)
`
`

`

`db
`‘
`NDA 20 030-
`2.0
`I RTRODUCTION
`
`2.1
`
`Overview
`
`b(4)
`
`The applicant, Schwarz Pharma, is seeking approval of M, sustained—release (SR), for the
`treatment of overactive bladder syndrome (CAB). ' ~—/ has been developed as once-daily
`formulation with dosage strength of 4mg and 8mg.
`
`To support the safety and efficacy of --———- , clinical data from two Phase 3 pivotal studies were
`submitted. In addition, safety data from open—label extension studies and a QT study were also
`submitted to rule out any abnormal QT prolongation or other cardiac abnormality post-dose. This
`review will focus on the efficacy data from the two Phase 3 trials listed in Table 2.1 below.
`
`
`
`
`Table 2.1 Summa
`of Pivotal Studies
`Study#
`Study Site
`‘
`Study Design
`Number Randomized
`Duration of
`
`(number)
`/ Study Regimen
`Treatment
`
`W}
`
`M4)
`
`‘
`
`M4)
`
`
`
`
`
`
`
`
`SP583
`
`
`
`SP584
`
`
`
`Total Randomized: 1135
`
`Multi—center, double-
`Europe, Australia,
`
`blind, placebo and
`New Zealand,
`'
`
`
`active—controlled, Phase
`South Africa (150)
`Placebo: 285
`
`
`
`
`
`3.
`\7 4mg: 272
`_
`
`
`
`\- 8mg: 288
`
`Tolterodine 4m: 290
`
`
`US (83)
`Multi—center, double—
`Total Randomized: 836
`
`
`blind, placebo—
`
`
`controlled, Phase 3.
`Placebo: 274
`
`
`
`
`‘_ 4mg: 283
`
`w 8m
`: 279
`
`
`12 weeks
`-
`
`2.2
`
`Data Sources
`
`The submission was in hard copy and partially'electronic. Submitted data were stored in folder
`
`
`Cdsesub] 1122030 N 000 2006—03—17 crt datasets in FDA’S Electronic Document Room
`(EDR). The data quality of the submission was within acceptable limits.
`
`2.3
`
`Indication
`
`\fimmmz‘e is. indicatedfor 1/96 treatment ofoveractive bladder 91772117107723 with mge windy intontineflce,
`zzggmgl, and urinaglflequmgl.
`
`W4)
`
`

`

`A“)
`NDA 20-030: _
`
`31(4);
`
`3.0
`
`3.1
`
`STATISTICAL EVALUATION
`
`Overview of Study SP583 and Study SP584
`
`3.1.1
`
`Design and Objectives
`
`Studies SP583 and SP584 were identical in design except that study SP583 included one additional
`active—controlled arm and was conducted using a separate protocol in several countries, while study
`SP584 was conducted at US sites. The methodologies used in both trials were the same; therefore,
`the study descriptions are applicable to both studies, unless otherwise indicated:
`
`Design: Both studies SP583 and SP584 were multi-center, randomized, and placebo—controlled, and
`were conducted at 150 sites across Europe, South Africa, Australia and New Zealand, while study
`SP584 was conducted at 83 sites in the United States. In addition to a placebo treatment arm, study
`SP583 contained an active control treatment arm (tolterodine SR 4mg/day). The objectives of both
`
`studies were to evaluate the efficacy, tolerability and safety of
`as compared to placebo in
`subjects with OAB.
`
`23(4)
`
`Following enrollment and a two week placebo run—in, subjects with a known history of OAB
`symptoms with at least 8 micturitions per 24 hours for at least 6 months were randomized to one of
`the following treatment groups: ——-’~ . 4mg/day, \i 8mg/day, tolterodine SR 4mg/day
`(study SP583) or placebo.
`
`Wt
`
`The planned duration of the trial was approximately 16 weeks: 2 weeks of run—in, 12 weeks of
`treannent, and 2 weeks of safety follow-up. Treatment compliance was assessed by instructing
`subjects to return all unused medication and micturition diaries at each applicable trial visit. For
`subjects taking less than 75% or more than 125% of the given dosage, a decision was to be made as
`to whether the subject should continue or withdraw from the study.
`
`Primary Efficacy Endpoints: As per protocol, the following endpoints were considered co—
`primary:
`
`1) Change in the average number of micturitions (frequency) per 24 hours (from baseline to week 12
`of treatment period).
`’
`
`Number of micturitions was defined as the number of times a subject passed urine per day (not
`including incontinence episodes). Subjects were to record the number of micturitions using a diary
`for 3 consecutive days during the week immediately prior to scheduled viSits. A time had to be
`recorded in the diary for the data to be included.
`
`2) Change in average number of urge incontinence episodgper 24 hours.
`
`Urge incontinence episode was defined as the complaint of a sudden compelling desire to pass urine,
`
`7
`
`

`

`(.71)
`
`a desire which is difficult to defer.
`
`Secondary Efficacy Endpoints: The following endpoints were Considered secondary in this study:
`
`1) Change in average voided volume per micturition measured using the urine c'up provided
`during 1 day collection period.
`2) Treatment response (yes/no), derived from a 4—grade treatment benefit scale assessing
`subject condition: 1: greatly improved, 2=improved, 3=not changed, and 4=worsened.
`Treatment response was dichotOmized as yes for category 1 and 2, no otherwise.
`3) Change in number of micturitions during daytime.
`4) Change in number of micturitions during sleeping time.
`5) Change in number of urgency episodes per 24 hours defined as number of times a subject
`recorded an urgency episode with or without incontinence per day within the 3—day
`collection period.
`6) Change in severity of urinary urgency based on 4—grade scale: 1=none, 2=mild, 3=moderate,
`and 4=severe, and
`7) Change in health outcome parameters based on King’s Health Questionnaire.
`
`Determination of Sample Size: The sample size was calculated to test the superiority hypothesis
`for both co—primary endpoints. Using a clinically meaningful difference of 20.72 in daily number of
`mictu'ritions between 'M\and placebo with a mean square error of approximately 2.5 (based on
`Phase 2 study SP582), the protocol called for a planned sample size of 249 per group, to test the null
`hypothesis of no difference assuming a type-I error (2-sided) of 5% and a power of 90%. For the
`change in urge incontinence, at least 205 per arm subjects would be needed to detect a difference of
`equal or greater than 0.57 per 24 hours with 80% power at the type-I error of 5%.
`
`Definition of Analysis Sets (Population): For efficacy analysis, two analysis sets were used: Full
`Analysis Set (FAS), and Per Protocol Set (PPS). FAS included subjects who were randomized using
`intent—to—treat principle, i.e., all subjects randomized regardless of actual treatment received. Subjects
`who did not obtain any close of the medication or who did not have rnicturition measurements at
`baseline or under double—blind treatment period, were excluded from the FAS. PPS excluded
`subjects with major protocol violations and/or with duration of double—blind treatment shorter than
`2 weeks.
`
`Handling of Missing Data: Missing diary data on micturitions and urge incontinent episodes
`from the double—blind period of the treatment were imputed by LOCF method from the last
`available post-baseline diary data. For the missing treatment response variable, in addition to LOCF,
`a ‘non—response’ was set to subjects without post-baseline measurement data.
`
`Pooling of Sites: Because of the small numbers of subjects per site, sites were pooled within each
`' country and incorporated into statistical analyses to adjust for site variability by treatment.
`
`Statistical Methods: For comparison of treatment groups with respect to both co—primary
`endpoints, the statistical methods included ANOVA models including country, treatment, and
`baseline by treatment interactions as factors. Pair—wise comparisons were reported as least square
`
`8 .
`
`

`

`Mei
`
`NDA 20—030: x
`
`(LS) means. To examine the robustness of the results, non-parametric analyses were also performed.
`
`Multiple Comparisons/Multiplicities: To preserve the false positive error (alpha) rate for co—
`primary endpoints at multiple doses, the protocol—specified plan was to use a closed—testing
`procedure in a hierarchical sequentially rejective manner. In this method, the plan was to test for’
`statistical significance at 0.05 (two-sided) for the comparisOn of mean change in micturitions
`between ‘ L---—»8rng and placebo first, and if the p—value for this test was < .05, then the test would
`proceed for the next lower dose, i.e., testing f0 \— 4mg and placebo comparison in the second
`step and so on for testing change in urge incontinence. If the test result Was not statistically
`significant at any step, then all remaining tests would be considered statistically non-significant.
`
`53(4)
`
`3.1.2 Reviewer’s Comments on the Design
`
`The sample rirze war adequatefor terting the mperiorzy hypotherz'rfor hoth co-prnnagl enclbointr in hath stadiec. The
`cloyed—tei‘tingprocedure to preserve thefaltepositive error war alto acceptahle, hat it war not clearél indicated in the
`protocol why the testfor mictnritionr would he conductedfirst, when the Divirion require; hoth primagy enapointrfir
`approval. Thereflire, ”\wart demonstrate reduction in hath primagl enapoint: compared to placeho. The are ofa
`hierarchical elated-testingprocedure was appropriatefor controlling ype‘I error rate with regard: to co-prirnagy
`enclbointr. However, no Inchplan war in the protocolfor teetz'ng .recona’agl enclvoint: or even to tertfiir the co-prz'rnagl
`endpoints at clifi‘erent weeks. In thir review, we will use other method: while evaluating the secondagl enapointi‘.
`
`M4)
`
`3.2
`
`Results: Study SP583
`
`3.2.1
`
`Subject Disposition
`
`At 150 sites, a total of 1135 subjects were randomized approximately equally to the treatment groups
`as shown in Table 3.2.1. Subject enrollment was similar across sites. No single site was predominant
`in terms of subject enrollment. For analysis, sites were clustered together by country. A toral of 147
`(13%) subjects discontinued the study prematurely. The major reasons for discontinuation were
`adverse event (3%) and withdrawal of consent (3.5%), followed by protocol deviation 2%. The
`discontinuation rates were similar across treatment groups, and did not appear to impact the efficacy
`results. The full analysis (ITT—LOCF) population of 1103 subjects is well over the required 1070
`subjects, while the per protocol analysis (completers at endpoint) population of 1027 is also in the
`acceptable range. '
`
`

`

`a
`NDA 20-0301 \‘__ _
`
`33(4)
`
`
`
`
`Total Randomized
`
`Completed study
`Discontinued (°/o):
`Adverse Event
`
`Table 3.2.1 Disposition of Subjects: Study SP583
`
`rou . s
`Treatment
`
`
`
`
` 1135
`
`
`
`
`
`
`
`33(12)
`41(15)
`36(13)
`37(13)
`147(13)
`
`
`6 (2)
`9 (3)
`14 (5)
`10 (3)
`39 (3)
`
`
`
`
`
`
`
`Lack of Efficacy
`1 (<1)
`2 (1)
`2 (1)
`3 (1)
`8 (<1)
`
`
`
`
`
`
`
`Withdrawn Consent
`12 (4)
`9 (3)
`9 (3)
`10 (3)
`40 (3.5)
`
`
`
`
`
`
`
`Protocol deviation
`6 (2)
`9 (3)
`4(1)
`5 (2)
`24 (2)
`
`
`
`
`
`
`
`Compliance
`3 (1)
`0
`2 (1)
`1 (<1)
`6 (<1)
`
`
`
`
`
`
`
`Lost to follow-up
`0
`2 (1)
`1 (<1)
`4 (1)
`7 (<1)
`
`
`
`
`
`
`
`4 (1)
`Other Reasons
`6 (2)
`9 (3)
`4 (l)
`23 (2)
`
`
`
`
`
`
`
`
`279
`265
`276
`283
`1103
`
`Full Analysis population
`
`(ITT-LOCP)
`-
`
`
`262
`246
`253
`266
`1027
`Per Protocol Population
`
`
`
`
`* I'IT population included all randomized subjects who received treatments and had diary response for
`at least 3 consecutive das.
`
`3.2.2
`
`Patient demographics and baseline characteristics
`
`The baseline characteristics such as age, race, gender, body mass index were similar across treatment
`groups. Concomitant medication use and prior drug treatment for OAB were also similar between
`treatment groups.
`'
`<
`
`3.2.3
`
`Primary Efficacy
`
`Two endpoints were considered primary in this study: the change in micturitions and change in urge
`incontinence episodes per 24 hours from baseline‘ to week 12. As per protocol, a hierarchical closed—
`testing procedure was used to control false positive error rate (type-I) for multiplicity. To use this
`method, a family of hypotheses with respect to multiple endpoints and doses are hierarchically
`ordered and the hypotheses are tested in a sequence. In this protocol, the sponsor ordered the
`sequence starting with the micturition hypothesis at the highest dose of -~—/
`followed by lower
`dose and so on for the urge incontinence hypotheses. For the secondary endpoints, no testing was
`plannedin a hierarchical order. Therefore, we used Dunnett’3 test to adjust for multiple dose
`comparisons.
`
`To evaluate the treatment difference between \— doses and placebo, we also performed a
`statistical analysis similar to the sponsor’s analysis using 3 analysis of covariance (ANCOVA) model
`10
`
`

`

`NDA 20-030:
`
`A
`
`_
`
`We)
`
`,
`
`with factors for baseline, treatment, country (sites pooled by country), and baseline by treatment
`interactions. We do not disagree with the sponsor’s approach, but need to look at other methods for
`consistency of the results, although it is highly unlikely to differ in conclusions. Our analysis was also
`based on the ITT population using last observation carried forward (LOCF) for missing post
`baseline data. We used LOCF because the percentages of subjects with post baseline missing diary
`ranged from 3°/o-7°/o. It was similar across treatment groups and did notappear to follow any
`systematic pattern that could either be considered as missing not at random or otherwise.
`
`Change in micturitions: Results of our analyses are shown in Table 3.2.3. At week 12, the average
`reductions from baseline in micturitions were —1.9 and —2.0, for A (4mg and 8mg) doses,
`compared with —1.1 for placebo. The reductions for both doses 0 ~* were statistically
`significant (p<.05, after adjusting for multiple comparisons) compared to placebo. However, there
`were no statistically significant differences between the \doses and both doses appeared
`equally effective in improving rnicturitiOns per 24 hours.
`
`Change in Urge Incontinence: Similar effects were also noted for average reductions in urge
`incontinence at week 12 of 4.94 and —2.2 for K—N doses, compared with 4.14 for plaCebo. The
`reductions were again statistically significantly different from placebo after adjusting for multiple
`comparisons.
`-
`
`Results using from the completers (not shown here), similar to the sponsor’s definition of per
`protocol set, were similar to I'IT using the LOCF analysis population. Both analysis population sets
`showed consistent efficacy results with respect to both endpointsIn support of. /—~ compared to
`placebo.
`
`
`Table 3.2.3
`
`Change+ from Baseline to Week 12in the Mean Number of Micturitions and Urge Incontinence
`E-isodes net 24 hours. ITT-LOCF Po-ulation Stud SP583
`
`1 on s M
`
`Mean
`
`Mean
`
`Difference
`
`unad'usted
`
`Number of Micmritions
`per 24 hours
`
`Urge Incontinence
`
`Placebo (279)
`.\4mg (265)
`—— 8mg (276)
`Tolt 4mg (283)
`
`Placebo (211)
`' \4mg (199)
`L—e’ 8mg (223)
`Tolt 4mg (223)
`
`V
`
`
`
`+ Chan e frombaselnaebbas
`treatme t, coun
`ase
`“Unad‘usted
`
`liLS mean difference from ANCOVA model with factors for baseline values,
`e bytreatment interaction.
`
`11
`
`m4)
`
`W)
`
`all)
`
`wet "
`
`

`

`,.
`NDA 20—030: k—~‘
`
`'W
`
`3.2.4
`
`Secondary Efficacy
`
`Several outcomes, as noted in section 3.1.1, were considered secondary in this study. Among them,
`the clinical team considered the changes in voided volume per nnctutiitions as one of the important
`secondary endpoints and, therefore, we have performed an analysis of voided volume using the
`same ANOVA models. The results of our analysis are shown in Table 3.2.4. Relative to placebo,
`, \‘\ (4mg and 8mg) improved the mean voided volume per micturition from baseline to
`endpoint by 271111. and 33 mL, respectively. The improvements for both doses were statistically
`significant compared to placebo.
`
`Table 3.2.4
`
`Change“ from Baseline to Week 12 in the Mean Volume Voided:
`ITT-LOCF Po nulation Stud SP583
`
`_m——Grou- s M
`
`47(4)
`
`32(4)
`
`Difference
`
`—-
`18.0
`24.0
`15.0
`
`
`
`Voided Volume per
`micturition
`
`Placebo (278)
`\‘4mg (265)
`__..~.8mg (275)
`Tolt 4mg (282)
`
`Mean
`
`'8 Mean
`9 0
`.
`27.0
`33.0
`24.0
`
`sitChange from baseline based on LS mean difference from ANCOVA model with factors for treatment and
`+“P-values ad'usted for 'multile com-arisons with lacebo b Dunnett’s Test.
`
`3.2.5
`
`Efficacy at Week 2
`
`At week 2, z. \j both doses of \Adid reduce the average 1 ‘~’—“_. urge
`inconfinence,W "
`
`3.2.6 _
`
`Adjustment for Multiple Comparisons /Multiplicities
`
`As mentioned in previous sections, a hierarchical closed—testing procedure was used to control the
`false positive error rate (type-I) for multiple doses and multiple endpoints. To use this method, a
`family of hypotheses with respect to multiple endpoints and doses are hierarchically ordered and the
`hypotheses are tested in a sequence. In this method, the plan was to test for statistical significance at
`0.05 (two—sided) for the comparison of mean change in micturitions between '\ 8mg and
`placebo first, and if the p—value for this test was <05, then the test would proceed for the next lower
`dose, i.e., testing for \4mg and placebo comparison in the second step and so on for testing
`change in urge incontinence. If the test result was not statistically significant at any step, then all
`remaining tests would be considered statistically non—significant. Table 3.2.6 shows the ordering of
`the hypotheses and the significance level at each step of the test. At each step, the p-value for the
`12
`
`M42
`
`

`

`NDA 20-03 0:
`
`(uh
`
`7
`
`treatment difference was less than the pre-specified two—sided p—value of 0.05, meaning after
`adjusting for both multiple dose and c07primary endpoints, .7
`\greatment improved in the
`number of micturitions and urge incontinence episodes in a statistically significant manner compared
`to placebo.
`
`For consistency purpose, we also performed an alternative adjustment for multiple comparisons for
`doses by Dunnett’s test for each co—primary endpoint separately, because we thought efficacy must
`be demonstrated for both co—prirnary endpoints without a need for ordering the hypotheses. The
`results and the conclusions by bbth adjustment methods remained the same,
`
`
`
`Table 3.2.6
`
`Statistical Significance of Primary Efficacy Endpoints (Week 12) with adjustment for
`
`Multiple Comparisons /Mu1tiplicities: Study SP583
`
`
`
`
`
`
`Sponsor’s adjustment using Closed-Testing
`Procedure
`
`Adjustment using Dunnett’s Test
`
`' Testing Steps
`
`I Comparison
`
`P-value++
`
`
`
`
`
`
`“‘1 4mg vs placebo
`/- 8mg vs placebo
`
`
`Step 2: Number of Mictuu'tions
`Tolt vs placebo
`
`
`
`.\ 4mg vs placebo)
`
`
`
`Urge Incontinence
`"\ A 4mg vs placebo
`
`
`
`Step 3: Number of Urge Incontin
`—/
`L 8mg vs placebo
`
`
`
`\(Smg vs placebo)
`Tolt vs placebo
`
`
`
`
`Step 1: Number of Micrudtions
`-~"L 8mg vs placebo)
`
`be» I
`
`'
`
` Step 4: Number of Urge Incontin
`\Amg vs placebo)
`
`
`
`
`+ Chan e from baseljiaebbas
`0 LS mean difference from ANCOVA model with factors for baseline values,
`
`treanne t, country, an
`ase
`e y treatment interaction.
`‘
`
`
`++P—value ad'usted for nair wise com-arisons b Dunnett’s test.
`
`
`
`3.2.7