`
`CENTER FOR DRUG EVALUATION AND
`
`RESEARCH
`
`APPLICA TION NUMBER:
`
`21-782
`
`MEDICAL REVIEW
`
`

`

`
`
`Food and Drug Administration
`«h
`Center for Drug Evaluation and Research
`Division of Anesthesia, Analgesia, and Rheumatology Products
`HFD—l70, Room 93—45, 5600 Fishers Lane, Rockville MD 20857 (301) 827—7410
`
`Addendum to Medical Officer Team Leader Memorandum
`
`
`
`Date:
`
`July 22, 2005
`
`To:
`
`File, NDA 21-782
`
`From:
`
`Rigoberto Roca, MD.
`Deputy Director
`Division of Anesthetic, Critical Care and Addiction Drug Products
`
`Re:
`
`NBA 21—782: Ramelteon (TAK—375)
`
`Takeda Global Research and Development, Inc.
`
`
`This is an addendum to my Medical Officer Team Leader memorandum dated June
`30‘”, 2005.
`It will articulate the reasons why 1 reached a different conclusion and
`recommendation from the primary clinical reviewer, Elizabeth McNeil, M.D.
`
`For the reader’s convenience, the original memorandum is reproduced below in its
`entirety, with the addendum clearly identified at the end of this document.
`
`Background
`Ramelteon (also known as TAK-375) is a melatonin receptor agonist with high affinity
`for the melatonin MT! and M1"; receptors. Melatonin receptors are found in various
`tissues throughout the body, and are classified into three subtypes: MTi, MTZ, and MT3.
`Rameltcon, and its active metabolite, Mil, have been shown through in vitro assays to
`
`have little affinity for MT3, other receptors, or enzymes.
`
`The applicant proposed that ramelteon’s interaction with the melatonin receptors is the
`basis of the mechanism of action, since it
`is believed that endogenous melatonin’s
`interaction with these receptors affects the maintenance of a normal circadian rhythm
`underlying the sleep-wake cycle.
`The applicant
`seeks
`the following indication:
`“[Ramelteon] is indicated for the treatment of insomnia. j;
`
`J
`
`
`
`

`

`
`
`
`
`NDA 21-782 Ramelteon (TAK—375)
`
`The clinical review of this supplement was performed by D. Elizabeth McNeil, MD. and
`the statistical review was performed by Dionne Price, PhD. David Lee, Ph.D., reviewed
`the pharrnacokinetic data and Adam Wasserman, PhD. reviewed the pharmacology and
`toxicology data. Pramoda Maturu, Ph.D., performed the CMC review and Katherine
`Benson, PhD. reviewed the abuse liability studies. A consultation response from the
`Division of Metabolic and Endocrine Drug Products was provided by Mary Parks, MD.
`This memorandum will summarize their
`findings, as well as my recommendation
`
`regarding the approvability of this application
`
`Regulatory History
`The applicant has performed numerous studies during the drug’s development, including
`pharmacokinetic studies, drug-drug interaction studies,
`food—interaction studies, abuse
`liability studies, and studies on the effect of ramelteon on human endocrine function
`Seven studies were specifically designed to evaluate ramelteon’s efficacy. Safety data
`were collected in all the studies.
`
`There were several interactions with the applicant prior to submission of the application,
`including an End-of—Phase 1 meeting, an End«of-Phase 2 meeting, a Pre-NDA meeting
`and several
`teleconferenccs. During these meetings, the number and types of clinical
`trials that would be required, as well as the study endpoints and statistical analyses that
`would support the indication of interest, were conveyed to the applicant.
`
`On February 1 1, 2004, during a teleconference held afier the End-of-Phase 2 meeting and
`before the Pre-NDA meeting, the applicant informed the Division that Study TL020 had
`failed in its primary efficacy endpoint, subjective sleep latency. They were informed that
`it might be possible to extrapolate efficacy to the younger population based on the results
`of Study 11025, which was then ongoing, but that this would depend on the results of the
`study. Although it is generally acknowledged that the ability to extrapolate data from one
`patient population to another involves multiple factors (pathophysiology, mechanism of
`action of the intervention, etc), part of this process also involves an assessment of the
`statistical robustness and clinical significance of the findings.
`
`At the Pre-NDA meeting the applicant informed the agency of their intention to utilize
`the following trials to support their proposed indication: Trials 01?, 02], 023, and 025.
`It
`is appropriate for the applicant
`to designate which trials they consider pivotal
`in
`support of their application.
`it
`is also appropriate for the reviewing division to request
`and review data from all trials which may be contain data that will allow the assessment
`of safety and/or efficacy, and to make its own determination of the appropriateness of the
`individual studies to provide information.
`
`
`
`Medical Officer Team Leader Memo
`
`2
`
`

`

`NDA 21—782
`Ramelteon (TAK-375)
`
`The table below, adapted from Dr. McNeil’s review, summarizes the studies which were
`reviewed to assess the efficacy and safety of ramelteon in patients.
`
`Study,
`Location,
`and Date
`
`Duration
`
`Type of
`Patient
`
`Population
`h.-
`
`Primary
`Efficacy
`Endpoint
`
`Treatment
`arms
`
`N0. of
`Patients
`
`placebo
`
`N = 375
`
`16 mg: 126
`64 mg: 126
`placebo: 123
`
`N = 289
`
`8 mg: 98
`16 mg: 94
`placebo: 97
`
`Healthy adults
`(35 — 60 yrs
`old) with
`transient
`insomnia
`
`Latency to
`persistent
`sleep (by
`PSG)
`
`16 mg
`64 mg
`placebo
`
`Healthy adults
`(18 — 64 yrs
`old) naive to a
`sleep
`laboratory
`environment
`
`Healthy adults
`(18-— 65 yrs
`old)
`with chronic
`insomnia
`
`Latency to
`persistent
`sleep (by
`
`[1.2133, Nu
`
`Latency to
`persistent
`sleep (by
`PSG)
`
`= 65 yrs old
`with
`chronic
`insomnia
`
`18-64 yrs old
`with chronic
`insomnia
`
`Latency to
`persistent
`sleep from
`nights I and 2
`of each
`treatment
`
`period
`
`Latency to
`persistent
`sleep (by
`PSG)
`
`8 mg
`16 mg
`placebo
`
`N =405
`
`8 mg
`16 mg
`placebo
`
`
`
`18; 64 yrs old
`with chronic
`insomnia
`
`Siibjectivew -
`sleep latency
`
`I
`
`8 mg
`16 mg
`placebo
`
`N = 848
`
`8 mg
`16 mg
`
`PNFP002
`14 centers in
`
`the U.S.;
`5/2000 —
`10/2000
`
`TL023
`15 centers in
`the US.;
`12/02 — 3/03
`
`Double-
`
`blind,
`randomized,
`placebo—
`controlled,
`sin_ le dose
`Double-
`blind,
`randomized,
`placebo-
`
`TL005
`13 centers in
`
`the US;
`9/01 — 2/02
`
`TL017
`17 centers in
`
`the U.S_;
`10/02— 7/03
`
`TL021
`29 centers in
`the U.S.;
`1/03 — 9/03
`
`iTLOZO
`79 centers in
`
`the 118.;
`1/03 — 9/03
`
`blind,
`randomized,
`placebo—
`controlled, 5-
`period
`crossover,
`dose
`
`response,
`safety and
`efficac
`Double-
`
`blind,
`randomized,
`placebo—
`controlled,
`crossmer,
`
`safety and
`
`placebo—
`controlled,
`fixed dose,
`PSG and
`
`outpatient
`safety and
`efficac
`
`Double-
`
`blind,
`randomized,
`placebo-
`controlled,
`fixed dose,
`
`Each
`
`period
`lasted 2
`
`days, with
`5 — 12
`
`days
`between
`
`periods
`
`Each
`
`period
`lasted 3
`
`days, with
`5 — 12
`
`days
`between
`teriods
`
`Medical Officer Team Leader Memo
`
`

`

`
`
`NDA 21—782
`
`Ramelteon (TAK—375)
`
`Design
`
`Study.
`Location,
`and Date
`
`-—efficae
`
`TL025
`136 centers
`in the U.S.;
`12/02 — 1/04
`
`Double-
`blind,
`randomized,
`placebo-
`controlled,
`fixed dose '
`
`safety and
`efiicac
`
`Type of
`Patient
`Population
`
`Primary
`Efficacy
`Endpoint
`
`Treatment
`
`= 65 yrs old
`with
`chronic
`insomnia
`
`Subjective
`sleep latency
`
`insomnia
`
`TL022
`123 centers
`
`OpenJabel,
`long—term
`safety
`
`18 yrs old
`with chronic
`
`8 mg: 248
`l6 mg: 965
`
`Due to the number of studies involved, a detailed description of the designs of the study
`protocols (i.e., inclusion/exclusion criteria, assessments, efficacy parameters, and data
`analysis plans) will not be included in this memorandum;
`this can be found in Dr.
`McNeil’s review.
`
`Study Results — Transient
`
`insomnia
`
`Two studies were performed to evaluate ramelteon’s efficacy in a transient insomnia
`model, PNFP002 and TL023. Study PNFPOOZ utilized [6 and 64 mg of ramelteon,
`therefore the data derived from that study will not support
`the efficacy of 8 mg of
`ramelteon, the dose for which the applicant is seeking rmrketing approval. The data will
`be useful however, for evaluation of ramelteon’s safety.
`
`In Study TL023, analysis of the latency to persistent sleep (LPS) data for the intent—to—
`treat (ITT) population demonstrated statistically significant treatment effect overall when
`ramelteon was compared to placebo. However, when the doses were considered
`individually, the 8 mg treatment group maintained significance while the 16 mg treatment
`group did not. The table below, adapted from Dr. Price’s review, summarizes the results
`of the change in the mean latency to persistent sleep (in minutes).
`
`
`
`
`
`Placebo
`N : 97
`
`Ramelteon 8 mg
`N = 98
`
`Ramelteon
`16 mg
`N = 93
`
`LS mean (SE)
`LS mean difference
`from lacebo (SE
`95 % CI of difference __ (—101.03) —
`
`0-004
`0065 _
`
`19.7 (1.87
`
`12.2 (1.83)
`-7.6 (2.62)
`
`
`
`
`
`
`
`Overall
`
`
`p—value
`
`
`14.8 (1.93) _
`-4.9 (2.65)
`
`the treatment
`is worth noting that although the results are statistically significant,
`it
`effect, as represented by the mean difference from placebo,
`is less than 8 minutes.
`
`Medical Officer Team Leader Memo
`
`4
`
`

`

`
`
`
`
`NDA 21—782 Ramelteon (TAK-375)
`
`Whether this represents a treatment effect that is clinically significant is potentially up for
`debate.
`
`Other observations of Study TL023 included the following:
`- Gender analysis demonstrated a statistically significant difference for males at
`both doses, but not for females (at either dose).
`For persons < 40 years old, there was a statistically significant difference from
`placebo for those who were treated with the 8 mg dose, but not the 16 mg dose.
`- An evaluation by ethnic goup identified a statistically significant difference
`from placebo for Caucasians subjects treated with the 8 mg dose only.
`
`-
`
`Stuafil Results — Chronic insomnia
`Three studies evaluated the efficacy of ramelteon in chronic insomnia with LPS by
`polysomnography (PSG) as the primary efficacy parameter. The first two studies, Study
`TLOOS and Study TL017, utilized a multi—period crossover design. The third, Study
`TLOZ], utilized a fixed dose design. The table below, adapted from Dr. McNeil’s review,
`summarizes the results of the change in the mean latency to persistent sleep (in minutes).
`Ramelteon
`
`3 m
`
`g
`
`32 mg
`
`243*
`
`240*
`
`Overall
`—value
`<0 00!
`<0.001
`
`
`
`
`
`
`Study
`Visit
`TLOOS
`
`Placebo
`
`
`
`—__———
`TL021
`
`64
`.3
`——_———_
`———_-_—
`———_———
`
`
`
`_m
`* - denotes statistical significance
`'
`
`As with the trials in transient insomnia, although the mean change in LPS compared to
`placebo was statistically significant,
`the clinical significance is questionable,
`for the
`difference for the 8 mg treatment group was never greater than ~16 minutes (Study
`TL021, week 1).
`
`As noted by Dr. McNeil in her review, insomnia is different than other disorders in that
`both objective and subjective measurements are important, and it can be argued that from
`a clinical standpoint, the subjective parameters may even be more 50. Studies TLOOS,
`TLOl? and TLOZ] evaluated subjective sleep latency as one of the secondary efficacy
`parameters. The applicant also conducted two outpatient studies (Study TLOZO and
`Study TLOZS) where the primary efficacy endpoint was subjective sleep latency. The
`results on this endpoint are summarized in the table below, adapted from Dr. McNeil’s
`review.
`
`Ramelteon
`
`
`
`“m-
`
`
`
`
`Visit
`
`u-vaiue
`
`_————— -
`
`Medical Officer Team Leader Memo
`
`5
`
`

`

`
`
`
`
`NDA 21—782 Ramelteon (TAK-375)
`
` Ramelteon
`8 mg
`32 mg
`
`Placebo
`
`Overail
`Study
`i value
`Visit
`
`——__———
`_———_——
`_———_—_
`—__—_——
`
`———————
`—_—————
`__————_
`__“I-n—
`
`_————__
`_——__—_
`———————
`
`
`
`
`___—___
`"‘ -denotes statistical significance
`
`
`
`
`
`
`
`Dr. Price confirmed the applicant’s analyses, and due to concerns about the imputation
`scheme for lost data used by the applicant, specifically a last—observatiorrearricd- forward
`(LOCF) method, she re—analyzed the data using a baseline observation carried forward
`technique. The results of both imputation techniques were comparable.
`
`The results for 8 mg in Study TL025 are statistically significant, but a similar observation
`is made regarding the clinical significance of the result, since the maximum mean
`difference compared to placebo is ~13 minutes.
`
`Additional analyses performed by the applicant included a responder analysis, where a
`responder was defined as a participant having latency to persistent sleep less than or
`equal to 30 minutes- The results did not support the primary analysis at Week I. Dr.
`Price reanalyze the data altering the responder definition to include only those patients
`who completed the study; the results were comparable to what the applicant reported.
`
`Secondary endpoints included subjective total sleep time (sTST), sleep quality, and
`clinical global impression (CGI) of the change of condition. There were no significant
`treatment differences at any of the timepoints for sleep quality or CGI. A significant
`difference was seen at Weeks 1 and 3 for sTST for the 4 mg treatment group, but not the
`
`8 mg treatment group.
`
`Safety
`The number of patients that were exposed to a particular dosage, and the duration of that
`exposure, is summarized in the table below.
`It is apparent from the table that although
`approximately a fifth of the patients on 8 mg had some amount of data extending to 6
`months, the substantial amount of the data for the 8 mg dose are in the 7 — 35 day range.
`
`
`
`Medical Officer Team Leader Memo
`
`6
`
`

`

`64 mg
`32 mg
`16 mg
`8 mg
`4 mg
`< 4 mg
`Placebo
`N=209
`N= 169
`N=1961
`N=1250
`N=511
`N320
`N=1370
`0
`%
`"/
`%)
`%)
`(%)
`%)
`%)
`--——--
`WI-
`mum-.-
`—-_-I-l-l-_-_
`M“...-
`“MIMI-"Imm-
`
`
`
`
`
`
`NDA 21-782
`Ramelteon (TAK—375)
`
`Exposure
`(days)
`
`
`
`duration
`
`
`
`
`
`——-_—‘--§I--_
`
`Adverse events
`
`Deaths
`
`There were two deaths reported in the application, both on the 16 mg treatment arm in
`Study TL022. The first fatality was a 57—year old woman who died on study day 159
`after having been struck by a motor vehicle while walking down a highway at 2:30 in the
`morning; her autopsy revealed a blood ethanol level of 0.238 gm/dl. Based on her diary
`entries,
`the applicant deduced that the patient’s last dose was approximately 6 weeks
`prior to her accident. Although it is not possible to completely rule out an association
`
`with the study drug, there is not a clear causal connection.
`
`The second fatality was 58—year old man, who was on study day 227' when he was struck
`
`by a motor vehicle while crossing a parking lot. His last dose of medication was on the
`night before his accident.
`It was also not possible in this case to completely rule out an
`association with the study drug, and the case report
`form did not contain enough
`information to permit a clear causal connection.
`
`Serious adverse events
`
`There were 56 serious adverse events (SAEs) identified in the database, 18 of which
`resulted in patient discontinuation. The adverse events that resulted in discontinuation
`were in the 8 mg and 16 mg ramelteon treatment groups, and there was no obvious
`pattern to the SAEs with respect to the system organ class affected.
`
`Most commonly reported adverse event
`The most commonly reported adverse events for 8 mg of ramelteon were headache,
`somnolence, fatigue and dizziness, as summarized in the table below, adapted from Dr.
`McNeil‘s review.
`
`Ramelteun
`
`
`
`
`
`Term
`
`N:l370
`
`N:Sll
`
`- 4mg
`
`(“/0)
`
`cm
`
`17 8-1
`
`N:1250
`
`8mg
`
`(%)
`
`N=l96l
`
`16mg
`
`%)
`
`N=169
`
`32mg
`
`%)
`
`58(4.6)
`
`20mm)
`20410.4)
`
`N=209
`
`64mg
`
`“/o)
`74 35-4)
`15(72)
`
`Medical Officer Team Leader Memo
`
`7
`
`

`

`
`
`NDA 21—782
`
`Ramelteon (TAK-375)
`
`1‘9""
`
`(%)
`
`Ramelteon
`
`4 mg
`N=511
`
`%)
`
`8 mg
`N=1250
`
`16 mg
`N=1961
`
`(%)
`
`(”M
`
`32 mg
`N=169
`
`(“2)
`
`64 mg
`N=209
`
`(%)
`
`“ml-Imm-
`
`m 0‘
`
`5
`
`
`
`exacerbated
`
`Upper respiratory
`tract infection
`NOS
`
`Diarrhea nos
`
`26 (1.9)
`
`4 (0.8)
`
`33 (2.6)
`
`62 (3.2)
`
`3 (1.8)
`
`2 (l 0)
`
`24 1.8
`12 0.9
`
`15 2.9
`
`24 1.9
`21
`1.7
`
`l 0.6)
`37 L9
`18 0.9 Iran-n
`
`I.
`
`Additional considerations
`
`Pharmacology/toxicology
`The non-clinical data submitted by the applicant has identified a positive finding in one in
`vitro chromosome aberration genetic toxicology study.
`It was negative in an in vitro
`bacterial reverse mutation (Ames) assay using Salmonella typhinurium and Escherichia
`coli, an in vitro mammalian cell gene mutation assay using the mouse lymphoma TK ”'
`cell line, an in viva/in vitro unscheduled DNA synthesis assay in rat hepatocytes, and in
`the in viva micronucleus assays conducted in mouse and rat. Based on these results, Dr.
`Wasserman‘s conclusion is that ramelteon does not have a mutagenic or direct DNA
`effect, but did demonstrate clastogenecity.
`
`The carcinogenicity assessment identified dose-dependent development of hepatic tumors
`in mice, including adenoma, carcinoma, and hepatoblastoma. Although the occurrence of
`hepatic tumors
`in rodent carcinogenicity studies is not uncommon,
`the Executive
`Carcinogenicity Assessment Committee (eCAC) concluded that the clinical relevance of
`these findings could not be excluded.
`
`Rats treated with TAK—375 also manifested an increase in the development of hepatic
`tumors that was dose-dependent, but an increase in Leydig cell tumors compared to
`control-treated males was noted as well. The eCAC once more concluded that the
`
`clinical significance of these tumors could not be excluded.
`
`Administration of TAK—375 to pregnant rats during organogenesis resulted in teratogenic
`effects: dose-dependent fetal malformations; specifically diaphragmatic hernia, cysts on
`the external genitalia, and irregularly shaped scapula and ribs. Although the dose of
`ramelteon that were required to produce the teratogenic effects were many multiples the
`maximum recommended human dose based on a body surface area comparison, these
`data require rameltcon to be designated a Pregnancy Category C.
`
`Potential interaction in patients who are active smokers
`Ramclteon was not formally assessed in patients who smoke. Since in vitro studies
`indicate that ramelteon is primarily metabolized by CYP1A2, and it is well known that
`
`Medical Officer Team Leader Memo
`
`8
`
`

`

`
`
`
`
`NDA 21-782 Ramelteon (TAK-375)
`
`there is the possibility that smokers may have
`smoking will induce CYP1A2 activity,
`lower levels of ramelteon. What impact this could have on the efficacy of ramelteon is
`unknown.
`
`Potentialfor drug-drug interactions
`Ramelteon’s metabolism is significantly hindered by CYP1A2 inhibition. An in viva
`pharmacokinetie study assessing the interaction of fluvoxamine and ramelteon revealed
`that ramelteon’s AUCm 3 was increased 190—fold, and the Cm“ was increased 7'0- fold. A
`study evaluating the
`eo—administration of a CYP1A2 substrate
`(theophylline)
`demonstrated an increase in AUCO? g of approximately 40% and in increase in Cnax of
`approximately 35%.
`
`
`Large inherent i_n viva variability in absolute bioavailability
`The absolute bioavailability of ramelteon is approximately 2%, with a range of 0.5% to
`12%. This property can potentially increase the clinical implications of coadministration
`of ramelteon with CYP1A2 inhibitors.
`
`Interactions with the human endocrine system
`The potential effects of ramelteon on the endocrine system were evaluated in three
`studies: TL03l
`(a 4week study), TL032 (a Gmonth study), and TL022 (a long-term
`safety study still underway at the time of the application’s submission). However, due to
`the short duration of Study TL031,
`the results observed need to be interpreted with
`caution, since it is unlikely that an effect on the endocrine system would be detectabb in
`this time period. Further, although Study TL022 offered the possibility of following
`patients for a longer term (12 months), its lack of a control group will also limit its ability
`to permit any definitive conclusions to be made. As noted in Dr. Park’s consultation
`response, any differences noted in the elderly group compared to the younger group in
`this study may be reflecting the underlying risks of the older age group to develop
`endocrine abnormalities, and not be related to drug therapy.
`
`Study TL03l
`TL03l was a 4-week, randomized, double—blind, placebo—controlled, parallel— group study
`in healthy adult volunteers. There was a total of 99 patients randomized to either placebo
`or 16 mg of ramelteon (49 placebo; 50 ramelteon); 96 patients completed the study (47
`placebo; 49 ramelteon). There were no significant differences reported in the mean
`changes from baseline in the endocrine parameters assessing thyroid function, the adrenal
`axis, or the reproductive axis between the treatment groups. However, as noted above,
`the short duration of the study limits its ability to detect any effect by ramelteon on the
`endocrine system.
`
`Study TL032
`TL032 was a 6-month, randomized, double-blind, placebo—controlled, parallel—group
`study in healthy adults with chronic insomnia.
`Patients were randomized to either
`placebo or 16 mg of ramelteon. A total of l22 patients were randomized (65 placebo; 57
`ramelteon). The number of patients completing the study was low (63% in the placebo
`and 44% in the ramelteon group). The most common reason cited for study withdrawal
`
`
`
`Medical Officer Team Leader Memo
`
`9
`
`

`

`
`
`NDA 21—782 Ramelteon (TAK-375)
`
`was withdrawal of informed consent and adverse events, and seemed to occur early in the
`course of the study.
`
`There were no statistically significant differences noted between ramelteon and placebo
`for the endocrine parameters assessing thyroid function and the adrenal axis. There was a
`statistically significant difference in the overall mean change of prolactin levels from
`baseline to the end of treatment (-0.6 pig/L change in the placebo group compared to 2.9
`ug/L in the ramelteon group). A higher percentage of patients on the ramelteon group
`had an increase in prolactin levels documented from a normal value at baseline (31.5% in
`the ramelteon group, 18.5% in the placebo group). Although most of these were in the
`range of 20 — 30 ML, five patients in the ramelteon group had an increase > 40 ML,
`compared to one patient in the placebo group. Based on these data alone, causality is
`difficult to definitively establish; however,
`there is published literature indicating an
`association between melatonin levels and prolactin elevations. Due to this possible
`association, continued evaluation of ramelteon’s effect on prolactin levels, and its long-
`term consequences on bone metabolism and reproductive health should be considered.
`
`Study TL022
`TL022 is a12— month, open-label, uncontrolled, fixed-dose study. Patients were assigned
`to either 8 mg of ramelteon (=65 years of age), or 16 mg of ramelteon (18 — 64 years of
`age). For purpose of data analyses, they were categorized into one of the following:
`-
`24—week compliant: subjects who had taken an average of = 3 doses/week
`during the first 24 weeks of the study
`48—week compliant: subjects who had taken an average of : 3 doses/week
`during the first 48 weeks of the study
`
`-
`
`it is important to note that due to a high dropout rate, the majority of the patients had
`study medication exposures of < 32 weeks; only 77 patients had a total drug exposure of
`48 weeks or greater.
`
`to the findings, the incidence of abnormal thyroid function studies was
`With respect
`comparable to what was observed in the other two studies, and may be reflective of the
`background rate of thyroid dysfunction. There were two patients (0.16%) with abnormal
`morning cortisol levels who subsequently were evaluated with ACTH stimulation testing
`and were found to be abnormal. There were no patients in the two controlled studies who
`had abnormal ACTH stimulation tests. There was a decrease in the mean Total and Free
`
`testosterone levels noted in the 8 mg dose goup from baseline to Months 4 and 8, while
`the 16 mg group had a slight increase in mean testosterone levels over time. Without a
`placebo group, it is not possible to discern the significance of this finding.
`
`The overall conclusion based on the data available to date is that the number of patients,
`and the duration of exposure are insufficient to exclude the possibility that ramelteon is
`associated with chronic hyperprolactinemia. However, due to the fact
`that prolactin
`levels can increase for a variety of reasons, routine monitoring of prolactin levels is not
`recommended while on ramelteon therapy, but should instead be considered as part of the
`
`Medical Officer Team Leader Memo
`
`10
`
`

`

`NDA 21—782
`
`Ramelteon (TAK—375)
`
`focused clinical evaluation in someone who presents with amenorrhea or sexual
`dysfunction.
`
`Appears This Way
`On Original
`
`Medical Officer Team Leader Memo
`
`1 l
`
`

`

`
`
`NDA 21-782 Ramelteon (TAK—375)
`
`Pediatric patient population
`The applicant had originally requested a deferral of pediatric studies during the Pre-NDA
`meeting. These studies should be deferred until ramelteon’s impact on the endocrine
`system is better evaluated.
`
`Scheduling recommendation
`Based on review of the data from abuse liability studies submitted by the applicant, the
`Controlled Substances Staff is proposing that ramelteon not be controlled under the
`Controlled Substances Act. This recommendation is usually not incorporated into the
`decisionmaking process regarding the approvability of a product; however,
`it
`is
`important
`to be cognizant of the potential ramifications that,
`if approved, ramelteon
`would represent
`the first unscheduled hypnotic.
`It
`is highly probable that such a
`classification would result in different prescribing patterns, with the potential for greater
`patient exposures to ramelteon than other hypnotics.
`
`Recommendations
`
`The applicant has conducted a significant number of studies in the course of the
`development of ramelteon.
`They have been interactive with the Division at
`the
`appropriate junctures in their application.
`llowever, after approximately 3500 patients
`being exposed to ramelteon in various studies, the final assessment is that ramelteon has a
`statistically significant treatment effect that is of marginal clinical significance.
`
`In addition to the findings that the treatment effect does not seem robust, either in the
`form of additional analyses, or in the case of some of the secondary efficacy endpoints,
`there is the observation that that ramelteOn fails to demonstrate a treatment effect in the
`
`ramelteon’s unique
`The applicant proposes that
`subjective efficacy parameters.
`mechanism of action makes it difficult for patients to appreciate the shortened LPS and
`increased TST provided, and the efficacy of ramelteon may be more vulnerable to the
`effects of poor sleep hygiene than benzodiazepine receptor agoinist. Although the
`applicant’s proposal may be true, at this point it appears to be more speculative and not
`supported by any data. Furthermore, even if the applicant is correct, the end result is the
`same in that the patients who are currently being targeted by the proposed indication do
`not seem to recognize any benefit from treatment with ramelteon.
`
`Ordinarily, a marginally clinically significant treatment effect would not preclude an
`approval of a product. However, the ability to approve such a product would then focus
`even more on the safety profile, as the riskzbenefit assessment is being made.
`
`In the case of ramelteon, there are several issues in the safety profile that are of concern.
`First
`is the observation that a significant portion of patients experienced one type of
`adverse event or another, highlighting that ramelteon is not an entirely benign product.
`Secondly, there is the observation that there appeared to be a number of patients who
`experienced hyperprolactinemia. Due to the number of patients exposed and the duration
`of exposure, it
`is not possible to determine whether there was a true causal relationship;
`however,
`it
`is also not possible to definitively exclude a relationship between the
`hyperprolactinemia and ramelteon therapy. Third, there is the positive result in one of the
`
`Medical Officer Team Leader Memo
`
`12
`
`

`

`
`
`
`
`NDA 21—782 Ramelteon (TAK—375)
`
`is acknowledged that
`It
`in vitro chromosome aberration genetic toxicology studies.
`several other assays were negative, and it may possible that this result actually represents
`an erroneous finding, however,
`this still needs to be addressed to determine whether
`ramelteon is truly a genotoxic carcinogen Lastly,
`the pharmacokinetic findings that
`indicate a large inherent in viva variability and potential for drug-drug interaction portend
`potential difficulties in the general pepulation.
`
`These concerns could potentially be handled in the labeling of the product, with
`appropriate information, advice, and/or warning language that would help the prescriber
`use ramelteon most appropriately. However,
`that presupposes that ramelteon offers
`something to the patient population being proposed by the applicant. The applicant has
`not submitted sufficient data to support that position.
`
`My recommendation is that the current application be deemed “Approvable.” In order
`for this application to be approved,
`the applicant will be required to either identify a
`patient population in which the treatment effect demonstrated by ramelteon is not only
`statistically significant, but also of significant clinical significance to outweigh the
`currently known risks of ramelteon. Alternatively, the applicant can provide sufficient
`information to put the currently known risks of ramelteon into perspective. This would
`include
`further
`elucidation
`of
`the
`relationship
`of
`ramelteon
`therapy
`and
`hyperprolactincmia, and re—assessmcnt of the positive result in the genotoxicity assay.
`
`Depending on the additional information submitted, a decision can then be made as to
`whether the riskzbenefit profile would support approval of ramelteon.
`
`ADDENDUM:
`
`The primary reviewer, D. Elizabeth McNeil, M.D., recommended an approval action
`based on the applicant successfully being able demonstrate that ramelteon C
`J by' demonstrating a decrease in the latency to persistent sleep for up to
`35 days of therapy, utilizing objective measurements (i.e., polysomnography). She noted
`that the evidence was inconsistent when subjective measurements were used to assess
`ramelteon’s effect on the latency to persistent sleep endpoint. Her final assessment was
`that ramelteon has an immediate hypnotic effect and may appropriately be used in the
`short—term treatment of insomnia.
`
`As noted in my original memorandum, even though the applicant was able to
`demonstrate a statistically significant difference between ramelteon and placebo,
`it was
`my opinion that
`this statistically significant difference was not clinically meaningful.
`When this observation was combined with inconsistent
`results
`in the subjective
`measurements, which would presumably reflect what the benefit the patients felt they
`were obtaining from treatment with ramelteon, and the potential for an association with
`hyperprolactinemia,
`it was my opinion that
`the
`applicant
`had
`not
`adequately
`demonstrated a favorable risk:bencfit ratio for the patient population in which they had
`
`expressed an interest for marketing.
`
`
`
`Medical Officer Team Leader Memo
`
`13
`
`

`

`
`
`NDA 21—782
`
`'
`
`Ramelteon (TAK—375)
`
`My recommendation of an “approvabie” action on this application is intended to reflect
`my opinion that ramelteon does appear to possess a certain amount efficacy, however, the
`applicant would need to conduct studies to identify the patient population in whom the
`benefit of ramelteon therapy would outweigh the currently known risks. Conversely, the
`applicant could perform additional studies to further elucidate ramelteons interaction with
`the human endocrine system, so that the ramelteon’s risks could be evaluated in view of
`the currently known clinical benefit.
`
`AppeOi’S This Way
`On Original
`
`
`
`Medical Officer Team Leader Memo
`
`14
`
`
`
`

`

`
`
`This is a representation of an electronic record that was signed electronically and
`
`this page is the manifestation of the electronic signature.
`
`/S/
`
`Rigoberto Roca
`7/22/05 01:51:41 PM
`MEDICAL OFFICER
`NBA 21 - 782
`
`

`

` (J
`
`1
`
`o"
`*¢
`..4.-
`5 C
`5
`(5%h FDA CENTER FOR DRUG EVALUATION AND RESEARCH
`
`1'
`
`DIVISION OF ANESTHESIA, ANALGESIA AND RHEUMATOLOGY PRODUCTS
`HFD-l70, Room 98-45, 5600 Fishers Lane, Rockville MD 20857
`
`Tel:(301) 827-7410
`
`DIVISION DIRECTOR SUMMARY REVIEW AND RECOMMENDATION FOR!
`APPROVAL
`
`DATE:
`
`DRUG:
`
`NDA:
`
`July 18, 2005
`
`Rozerem (ramelteon, 8—mg tablets)
`
`21—782
`
`NDA Code:
`
`Type IS NDA
`
`SPONSOR:
`
`Takeda Global Research & Development Center Inc.
`
`INDICATION:
`
`For the treatment of insomnia
`
`
`
`Takeda submitted NDA 2 I -782 in support of marketing approval for Rozerem, 8—mg
`tablets, on September 2l , 2004.
`
`Review of the CMC portion of this application was completed by Pramoda Maturu, PhD.
`Review of the general pharmacology and toxicology data presented in this application was
`completed by Adam M. Wasserman, PhD. Supervisory reviews were provided by Daniel
`
`Mellon, Ph.D., Supervisory Pharmacologist in this division and by Kenneth L. Hastings,
`Ph.D., Associate Director for Pharmacology and Toxicology, Office of Drug Evaluation
`II. Review of the clinicai pharmacology and biopharmaceutics data in the application was
`completed by David Lee, PhD. A statistical review and eva