CENTER FOR DRUG EVALUATION AND
`
`RESEARCH
`
`APPLICA TION NUMBER:
`
`NDA 22-253 & 22-254
`
`CROSS DISCIPLINE TEAM LEADER REVIEW
`
`

`

`Cross Discipline Team Leader Review
`
`Cross—Discipline” Team Leader Review
`
`Date
`
`
`From Norman Hershkowitz
`Subject Cross—Discipline Team Leader Review
`
`NDA/BLA #
`22—253, 22254, ‘="‘
`W”
`Supplement#
`(0)
`Applicant
`Schwarz Biosciences Inc. (UCB)
`Date of Submission
`9/28/07
`PDUFA Goal Date
`
`
`
`TBD/ Lacosamide
`
`Proprietary Name /
`Established (USAN) names
`Dosage forms / Strength
`
`W
`
`Tablets 50, lOO 150, 200, 250 and 300 mg (22253)
`_
`,
`iv solution (22254)
`
`
`.
`
`Prop_osed Indication(s)
`Recommended:
`
`Adjunctive Treatment of Partial Onset Seizures1n Adults
`Approval for tablets and1V solution
`“‘
`
`Cross Discipline Team Leader Review Template
`
`1.|nfloducflon
`
`Lacosamide has been developed for two separate indications, partial onset seizures and pain
`associated with diabetic peripheral neuropathy (DPN). This CDTL Division of Neurology
`Prodcuts (DNP) review will concentrate on efficacy results1n partial onset seizures That for
`DPN will be reviewed by Division of Anesthes1a Analgesia and Rheumatologic Products
`(DAARP) Safety data1n this application has been reviewed by both division, and while this
`review will concentrate on safety1n epilepsy, all data will be discussed. Because of specific
`interest in a potential cardiac signal the Division of Cardiovascular and Renal Products
`(DCRP) was asked to comment not only on the formal QT study butlssues of PR prolongation
`and general cardiac safety.
`
`2.Background
`
`According to the Sponsor Lacosamide, (R)-2-acetamido-N--benzyl—3--methoxypropionamide, is
`a member of a series of functional amino acids. From a mechanistic perspective lacosamide
`appears to act as a sodium channel blocker, an action shared by a number other anticonvulsants
`including phenytoin, carbamazepine, oxcarbazepine and lamictal. The Sponsor also notes that
`lacosamide’s anticonvulsant activity may also be related to its ability to bind to collapsin
`
`Page I of 23
`
`1
`
`

`

`Cross Discipline Team Leader Review
`
`response mediator protein-2 (CRMP-2), a phosphoprotein which is mainly expressed in the
`nervous system and is involved in neuronal differentiation and control of axonal outgrowth.
`This reviwer believes that this latter mechanism is highly speculative.
`
`3. CMC/Device
`
`In CMC review oftablets (22253) submitted to DFS, performed by Drs. Shiromani and Sood,
`a recommendation of “approvable” was made pending responses to a letter containing
`questions (3/20/08) and the final Compliance and Environmental Assessment
`recommendations. No phase 4 commitments were made. A later memo (7/16/08) submitted to
`DFS recommended approval based upon acceptable responses to question and an acceptable
`Compliance report. The environmental Assessment found no concerns. Off note the Sponsor
`agreed to the following (although these do not appear to be phase 4 commitments):
`
`Inspections and
`No issues CMC issues regarding the iv solution (22254) were identified.
`microbiology were also found adequate. Approval was recommended by the CMC reviwer.
`
`l
`
`l
`
`t
`
`4. Nonclinical Pharmacology/Toxicology
`
`The pharmacology/toxicology review found no nonclinical issues and is recommending
`approval. They are, however recommending, the following phase 4 commitment:
`
`‘I‘Further assessment of lacosamide’s effect on brain development is needed and that this
`assessment may be conducted postapproval. Such an assessment should certainly involve
`dosing in rat throughout the critical periods that correspond to the entire period of human fetal
`
`Page 2 of 23
`
`

`

`Cross Discipline Team Leader Review
`
`development with, perhaps, direct dosing of the neonate, and, as Dr. Fisher notes, the use of
`sensitive methods for assessing neurobehavioral function and expanded histopathological
`examination of the brain. “
`
`Early animal studies indicated potential cardiac effects involving slowing of atrio—
`ventricular and ventricular conductivity as evidence by the lacosamide-induced increase in
`PR interval and QRS duration. However, obvious hERG channel effects or QTc changes
`were not apparent. For this reason clinical cardiac adverse events were closely monitored.
`A formal QTc study was, off course also performed.
`
`5. Clinical Pharmacology/Biopharmaceutics
`
`Dr. Fadiran and Zhang, clinical pharmacologists, performed the general clinical pharmacology
`review, while Dr Tandon reviewed the-i. 1V solution formulations.
`
`w)
`
`I GeneralPKProper/1a)";
`
`Lacosamide is a Biopharmaceutics Classificatidn System (BCS) class 1 drug. Lacosamide
`tablets bioavailability was approximated to be about 100%. It is absorbed with a Tmax of 0.4
`to 4 hours and a T1/2 of approximately 13 hours. This drug experiences <15% protein
`binding. The drug is eliminated by the kidneys. Most of the drug in the urine is in the form of
`lacosamide (40%) or its metabolites with the major metabolite (SPM12809) making up 30% of
`that which is recovered. Its major metabolite is believed to be inactive. The relative
`contribution of P450 isoforms in the oxidative metabolism of lacosamide is not clear. But, the
`Sponsor determined that formation of the major metabolite, SPM 12809, is through the
`CYP2C19 pathway. The clinical pharmacology reviwer, however, notes that the relative role
`of P450 isoforms in the oxidative metabolism of lacosamide is not clear.
`
`Drug-Drug fiz/é/dC/lb/Zf’. '
`
`In vitro studies indicate that lacosamide is not a significant inhibitor (1A1, 1A2, 2A6, 2B6,
`2C8, 2C9, 2D6, 2E1, 3A4, 3A5), although it inhibits CYP2C19 to some extent. In vitro studies
`also indicated some induction of CYP 2C9 and 2C19, but only a small effect was noted in
`interaction studies with omeprazole (see below).
`i
`
`No interactions becasue of protein bindings was anticipated. Lacosamide was not a substrate
`for p-glycoprotein.
`
`Definitive studies, in vivo, studies were performed on a number of potential concomitant drugs
`(anticonvulsants, oral birth control agents, hypoglycemic and cardio — active agents). The
`table below summarizes the conclusion, based upon CI of Cmax and AUC, drawn from these
`studies.
`
`Effect of lacosamide onpharmacokinetics of other drugs:
`
`Page 3 of23
`
`3
`
`

`

`Cross Discipline Team Leader Review
`
` arbamazepine
`
`
`Valproic acid
`
`
`
`None
`Di goxin
`
`
`
`Oral Contraceptive
`T Cmax of ethinylestradiol (~20%)
`
`
`
`None
`Om_eprazole
`
`Metformin
`
`
`
`effect controversial,
`one group showed increase and the other
`
`group showed decrease in exposure of
`
`metformin. PD not studies.
`
`Clinical relevance not clear
`
`
`
`
` Effect of other drugs on lacosamidgpharmAacokinetics'
`
`
`
`
`
`
` Valproic acid
`
`
`Omeprazole
`
`No effect on LCM, but t SPM12809 by 60%
`
`
`
`Metformin
`None
`
`Population PK drug—drug interactions were also examined, which indicated that LCM
`exposure is reduced by 15-20% when lacosamide is co—administered with carbamazepine,
`phenobarbital, or phenytoin. The finding on carbamazepine contradicts the above noted
`findings and according the clinical pharmacology reviwer is difficult to interpret becasue of
`this disparity, lack of statistical significance of this effect and confounding covariates.
`
`Jfiecz'a/Pquz/alz'oflyx
`
`The clinical pharmacology reviwer note that studies indicate that while no dose adjustments
`would be necessary for patients with mild to moderate renal impairment, patients with severe
`renal failure will require dose reductions. Studies indicate that similar adjustments would be
`necessary for patients with moderate hepatic impairment. Elderly patients experienced a 20-
`25% greater exposure when weight was taken into consideration. The clinical pharmacology
`reviewer felt that although this would not warrant dose adjustment on its own, becasue of
`increased incidence of impaired hepatic and renal function in this class of patients, some
`caution should be noted in this population. Although females experienced greater exposure,
`when weight was factored in this differnce disappeared. This led the clinical pharmacology
`reviwer to conclude that no adjustment is necessary. There were no racial differences in
`exposure when adjusted for body weight. Poor CYPC19 metabolizers were examined in a
`small study. No substantial differnce was noted in the plasma concentrations of the parent
`drug with extensive metabolizers. However, there was a significant differences (75 to 85%)
`observed in the SPM 12809 metabolite. Because of this metabolites low level, in comparison
`to the parent, this was not thought to be significant enough for a dose adjustment
`
`_/———\
`'
`‘
`
`1'12 So/u/z'o/zfawn/[4220125
`
`um
`
`Page 4 of 23
`
`4
`
`

`

`Cross Discipline Team Leader Review
`
`As all efficacy studies were performed using a tablet formulation it was necessary to establish
`equivalent bioavailability between this formulation and the
`‘
`\‘ ‘ .1V solution, P—
`
`6(4)
`
`./,/ 1/
`
`/
`
`The sponsor conducted two bioequivalent studies in healthy subjects evaluating the
`bioequivalence of solution for infusion at different infusion rates versus the oral tablets (Study
`SP645 and SP65 8). Dr Tandon, the clinical pharmacology reviewer notes that such studies
`demonstrated:
`
`0
`
`0
`
`0
`
`15 minute IV infusion of 200 mg versus tablets (2x100 mg):
`BE with respect to AUC(O-t)
`Not BE with respect to Cmax
`30 minute IV infusion of 200 mg versus tablets (2x100 mg):
`BE with respect to both AUC(O-t) and Cmax
`60 minute IV infusion of 200 mg versus tablets (2x100 mg):
`BE with respect to both AUC(O—t) and Cmax
`
`In addition to the above definitive bioequivalence studies the sponsor performed two studies in
`epilepsy patients already on a presumed therapeutic dose of lacosamide tablets. The
`intravenous formulation was substituted for tablets for a period of up to 5 days. The
`intravenous formulation infused over various times (10, 15 and 30 minutes). Minimal
`differences in the Ctrough and Cmax values for the 10, 15 and 30 minute infusion were
`observed. Dr Tandon concluded that it could be concluded, “the. 10, 15, 30 and 60 minute
`infusions at a given dose give comparable plasma concentrations of LCM.”
`
`OCPA’ecommezzda/z’omfi '
`
`0 OCP found the application acceptable “provided that amutually satisfactory agreement
`can be reached between the sponsor and the Agency regarding the language in the
`package insert.”
`0 A phase 1V commitment is recommended to “determine which enzymes may be
`involved in the metabolism of lacosamide in addition to CYP2C19. ”
`
`6. Clinical Microbiology
`
`No issues were identified (see CMC).
`
`Page 5 of 23
`
`5
`
`

`

`Cross Discipline Team Leader Review
`
`7. Clinical/Statistical- Efficacy
`
`The clinical efficacy review was performed by this CDTL, Dr. Norman Hershkowitz.
`
`The Sponsor submitted 3 adequate and well controlled trials for review. Supportive studies
`were also included. The adequate well controlled trial consists of a phase 2b, dose finding
`study (SP667) and two phase 3 trials (SP754 and SP755). All three trials were of similar
`design (see below). The table below presents a summary of dose, time and numbers of patients
`studied in these protocols.
`
`Trial numberffflixsicai development
`phasef’i‘fial design
`
`Nanak) er of
`
`Number of
`
`sub§eets
`randumized t9
`receive LC)???
`
`:3an eats
`randomized
`to receive
`
`placeba'“
`
`S96§7fphase Ez’muiticenter, doubte-
`bfind: randomized, piacebo—commiied,
`para Rel-group trial to investigate the
`efficacy and safety ofLCM (2&0; 480,
`and Gflflnlg/day)’
`
`SP7'S4fPi1ase Simulticenter: doubte—
`bfind, randomized, placebo—controlfeci,
`paraliel—group trial'to investigate the
`efficacy and safety DfLCM (430 anti
`600mgfday)
`
`Egflmgg’day: E0?
`
`400mgfday: 1108
`
`fififlfligtday: I306
`
`403mgfdayt 204
`
`600mgr’day: 97
`
`Elaxizmtm
`dam {ion cf
`trea Em 13th
`
`2% weeks
`
`’2. ,E weeks ‘
`
`i8 weeks
`
`Total: 944
`
`SPFSSKPMSe Br’multicenter, doubte-
`ijiind, randomfized, placebo—controfied,
`333m gel-group triai to hivestigate the
`efficacy and safety ofLCM (200 am}.
`
`430mgfda3’)
`
`200mg‘day: 1453
`
`400mm}: 159
`
`_
`
`
`
`Total
`
`Zflflm‘gfday 27a
`400mgmay: 4?:
`609mg‘day: 2G3
`
`LCM=1a 6033111162
`
`
`
`3'
`
`’0.
`
`Because of audit findings suggesting noncompliance with the 81365? pmtoeol: an 3 randomized am!
`treated subjeete at Site 12 were removed from the Safety Set (33). As a result, 418 subjects were
`included in the SS.
`All 3 Maia had a 12-week Maintemmee Phase.
`
`The Sponsor describes 4 additional trials as supportive for the claim of efficacy. All
`supportive trials were uncontrolled and open-label studies whose data principally contributed
`to the safety database.
`
`Page 6 of 23
`
`6
`
`

`

`Cross Discipline Team Leader Review
`
`As noted above all three studies were of a similar design. They were all multi-institutional,
`double-blind, placebo-control, parallel cohort, adjunctive treatment studies in adults (>16 years
`old) with partial epilepsy (simple partial, complex partial and partial secondarily generalized)
`Trials were rather similar in design. The schedule of evaluations was similar across studies.
`Initial screening was performed on the first day of the baseline period. Seizure diaries were
`provided at this time and patients were. instructed in their use. Patients then entered an 8 week
`baseline phase. They were randomized following this period ifthey continued to fulfill
`-
`inclusion/exclusion criteria (there was a requirement for a minimal seizure frequency during
`this period). Inclusion/exclusion criteria were relatively routine for this class of study. Patients
`entered the treatment phase following randomization which consisted of a titration and a
`maintenance period. The titration period in SP 667 and SP754 were of 6 weeks duration and
`that of SP 755 were of 4 weeks in duration. All titrations proceeded at the rate of 100 mg qD
`(in a BID divided dose) every week. All doses were administered in an evenly divided BID
`regimen. Subjects who could not tolerate their final dose were permitted one back step of 100
`mg/day during the titration period. The titration period was followed by a 12 week
`maintenance period in all studies. No back titration was permitted during this period. After the
`study was completed the patients were given a choice to continue on lacosamide in an open
`label study at a dose of 200 mg/day. If they so decided, they would undergo a blind transition
`period where they were titrated to a dose of 200 mg/day. If they declined they would undergo
`a down-titration that would proceed at a rate of 200 mg/day every week.
`
`The primary endpoint required by the FDA and EMEA where different, but were based upon
`the standards typically used for those agencies. These different primary endpoints were agreed
`upon by the FDA in an end of phase 2 meeting. The FDA assigned endpoint Was the change
`in partial seizure frequency per 28 days from baseline to the maintenance period. Seizure
`frequency (SF) wascalculated by the formula: SF = (Number of Seizures) x (28 / D), where, D
`is the number of days. The manner that baseline seizure frequency was calculated was
`different between the initial dose ranging study, SP667 and the two phase 3 studies, SP754 and
`SP755. These differences were protocol driven. Thus, for SP667 baseline values were based
`upon the complete 8 week baseline period, but for SP754 and SP755 baseline value was based
`upon the last 56 days of the baseline period. For patients who discontinued during
`maintenance phase an LOCF frequency value was calculated. Ifthe patient dropped out prior
`to entering the maintenance period an LOCF value for the titration period was caléulated.
`
`Statistical analysis ofthe seizure frequency change was performed on the log-transformed
`seizure frequency1 based on an ANCOVA model with terms for treatment and pooled site.
`Log-transformed average seizure frequency during the Baseline Phase was used as the
`covariate. This maneuver is rather commonly used in these studies to normalize such data. The
`seizure frequency between treatment and placebo was compared using LS means. Percentage
`reduction over placebo was calculated by: 100 x (1 - exp[LSM Treatment — LSM Placebo]),
`where LSM is the least squares mean from the analysis. This analysis was previously
`described in the Sponsor’s statistical analysis plans. The log transformation allows a
`normalization of data. Criteria for statistical significance were P 5 0.05.
`
`
`
`1 Log transformation was based upon the formula 1n(x+]), where x is equal to the seizure frequency.
`
`Page 7 of23
`
`’7
`
`

`

`Cross Discipline Team Leader Review
`
`The primary outcome described above and its method of analysis is similar to those used for
`the approval of a number of drugs. The single difference is the fact that only the maintenance
`period as opposed to the full treatment (titration plus maintenance period) was used to
`calculate post-treatment seizures. More Commonly the titration and maintenance are included
`in this calculation. Off note, this analysis was performed as a secondary endpoint analysis.
`_
`
`A number of secondary analyses were preformedincluding, but not limited to: 50 percent
`responder to Maintenance Phases (the EMEA primary endpoint), change in partial seizure
`frequency per 28 days from Baseline to the Treatment Phase (ie, Titration + Maintenance
`Phases: a more typical for the primary endpoint as noted above), other responder rates(375%,
`350% and 325%), Proportion of seizure-free days during the Maintenance Phase for subjects
`who entered-the Maintenance Phase, proportion of subjects who achieved “seizure-free status”
`during the Maintenance Phase for subjects who completed the Maintenance Phase, Response
`to treatment by seizure type, Clinical Global Impression of Change, Quality of Life in
`Epilepsy—3 1.
`
`All 3 studies underwent changes in sample size during their implementation. One had a
`decrease in sample size because of unexpectedly fewer dropouts and 2 had an increase in
`sample size becasue a repeat calculation indicated that the original determination of standard
`deviation and effect size, based upon another anticonvulsant study, was incorrect. These
`changes were made without unblinding and, according to the statistics reviewer, Dr. Massie,
`are justified.
`
`Drop out rate during the trial differed slightly between placebo and the 200 mg/day dose, with
`the ranges in trials being 11% tol4% and 17 to 21% for placebo and lacosamide (200 mg/day),
`respectively. That for the 400 mg dose showed a larger difference with 11% to 14% versus 21
`to 26% for placebo versus drug, respectively. High drop out rates where observed for the 600
`mg/day with a range of 11 to 13% versus 33% to 42% for placebo and drug, respectively.
`Most drop outs in the drug treatment groups resulted from adverse events (see safety).
`
`Subject demographics were comparable across treatment groups. The mean age amongst all
`studies was approximately 40 years old. Most patients were categorized as Caucasian with
`“black” making up only 2 to 6 percent of the studied population. Seizure types were also well
`distributed across treatment groups in all studies. Complex partial and partial secondary
`generalized were more common then simple partial seizures. The most common concomitant
`AED were carbamazepine (35.2% subjects), followed by 1amotrigine (31.2%) and
`levetiracetam (29.0% subjects). The majority of patients were on 2 concomitant medications.
`
`The results of the primary endpoint (percent change from baseline to maintenance) over
`placebo is presented for all three trials in the table below. The percent reduction from placebo
`is based upon logarithmically transformed data, but is actually very close to arithmetic percent
`changes. From these data it is apparent that both the 400 and 600 mg daily dose resulted in a
`significant reduction in seizures from placebo. This was also the conclusion of the Pharmaco-
`metrics reviewer, by Dr. Zhu, who noted that in a nonlinear regression least squares modeling
`response curve started to flatten out beyond the median exposure of 400 mg dose. From the
`data below, and as per Dr Zhu’s analysis, there is no obvious additional therapeutic benefit
`
`Page 8 of 23
`
`8
`
`

`

`Cross Discipline Team Leader Review
`
`In the 2 studies that examined the
`observed for the 600 mg/day as compared to 400 mg/day.
`200 mg/day dose a therapeutic trend was noted. This effect, however, was statistically
`significant for only one study. This reviewer believes that the 200 mg dose is therapeutic in
`some patients but may on average have a smaller effect resulting in an inconsistent statistical
`finding between both studies.
`
`TfialfCampm‘ison 6f
`LCM! to pfacabo-
`595457
`’
`
`We reduction over
`pmcebo
`
`P—Wflue
`
`95%: CI for 4%": refineries}
`aver placebo
`
`{61), 34.1)
`
`@3023“
`
`0.13034“
`
`{113,422}
`
`mafia.)
`
`LCM 4filflxngfday 135218?)
`
`Law 5001;1ng (Nzlos)
`
`2 13%
`
`ELM 400mgt‘ck1y (Nana
`LCM eongfday {New}
`
`21.5%
`24.6%
`
`0.0073“
`c-0051 M
`
`{5.3, 3&5}
`{7.8, 38.3)
`
`a.
` LCM gammy (N=16fi))
`14.4%
`020272;?“
`’22, 25.1)
`{1.4, 26-8)
`0:03:25“
`
`Emit 408mgfday G~E=153)
`
`As noted above, the change in frequency from baseline to maintenance phase is not a typical
`endpoint. The more conventional endpoint of change from baseline to the experimental period
`(titration + maintenance) was examined as a secondary endpoint. Data from thisanalysis is
`presented below, and differs little from the primary endpoint. This serves as an excellent
`sensitivity analysis tothe Sponsor’s endpoint.
`
`APPEARS THIS WAY
`0N ORlGlNAL
`
`Page 9 on3
`
`-
`
`9
`
`

`

`
`
`
`‘12.} reduction over
`glint-she
`
`Famine
`
`
`
`95% CE for “3.93: reduction
`aver placebo
`
`30.3%
`
`20.3%
`
`0. 3530
`
`i41.9, 24.3}
`
`{53. 32.9}
`
`
`
`
`
`LCM 200111gfday (34:10?)
`
`
`
`LCM 4Qrimge‘dzty 1:24:10?)
`
`
`
`LCM 500mgr‘day (Haas)
`{17.3, 32.3}
`0.9333”
`23.3%
`
`SP?54
`
`
`
`
`
`
`
`
`19.0%
`0.0043“
`LCM dBMfday {24:201)
`
`
`
`
`LCM amalgam @429?)
`
`
`
`
`
`
`LCM EUGmQrday {amen}
`1.23%
`8.20294*
`LCM 400mgfday (24:1 53}
`
`
`
`{It-03164alt
`13.1%
`
`
`Cross Discipline Team Leader Review
`
`
`
`Tfiaifiifemimrisen of
`LCM in placebo
`SPfifi?’
`
`19.9%
`
`sense“
`
`{5.5, 32.1}
`
`{1-3, 2'21)
`
`{3.0, 25-?)
`
`The statistical significance of secondary endpoint, 50% responder rate (the-EMEA primary
`analysis), exhibited results identical, in terms of which doses were statically significant from
`placebo, to the primary endpoints in the FDA analysis. Other secondary endpoints, dealing
`with numerical alterations is seizure rates exhibited statistical significant effects as compared
`to placebo or trended in the correct direction. The Global evaluations trended toward
`improvement in the 400 and 600mg closes. Effects of quality of life measures were small and
`inconsistent.
`
`Another secondary endpoint was the reduction in seizUres by seizure type (i.e. simple partial,
`complex partial and partial secondarily generalized). These data were only presented using
`descriptive statistics. There was likely insufficient power to draw definitive conclusions. In
`general both complex partial seizures and partial secondarily generalized all trended in a
`direction that suggested a therapeutic effect. The effect on simple partial was more
`inconsistent. No definitive trend was observed, with some studies showing decreases and
`others increases in seizure activity of drug over placebo. Nothing can be definitively drawn
`from these data as these seizures were the least frequently observed and the data would be
`prone to a sampling error.
`
`Dr Massie, the statistical reviewer, confirmed the Sponsor’s analysis for all performed studies.
`Dr Massie also noted that “overall, there was no compelling evidence that the treatment effect
`varied by gender.” He also determined that there was no obvious age dependency for the age
`range studied (16 to 71 years of age). Considering the limitation of the small size of the non-
`Caucasian sample size, it was concluded that no obvious racial differences in effect was
`observed.
`
`This reviwer concludes that both the 200, 400 mg/day close (divided bid) impart a therapeutic
`effect in adjunctive treatment of partial seizures. The 600mg/day dose does not appear, on
`average, to be superior to the 400 mg dose. The 200 mg dose may, on average, appears to
`
`Page 10 on3
`
`10
`
`

`

`Cross Discipline Team Leader Review ‘
`
`have a smaller therapeutic benefit. However, on an individual basis, dosing will have to be
`adjusted not only based upon therapeutic benefit but also on tolerability. As will be discussed
`in the safety section, the 600 mg dose was poorly tolerated.
`
`The Sponsor intends to market
`‘7, formulations of lacosamide: tablets, intravenous solution
`:‘h All pivotal studies were performed using a tablet formulation. Conclusions for
`efficacy for other formulations are based upon studies demonstrating equivalent bioavailability
`between those formulations and the tablet formulation. w
`M— ‘
`Bioequivalence was also demonstrated with the iv infiision
`solution when such infusions were performed over 30 and 60 minutes. Shorter infusions
`resulted in higher Cmax values in the formal bioequivalence studies (see Pharmacokinetic
`section above).
`
`“‘43
`
`8. Safety
`
`Two separate major safety reviews were perfomed because of the two independent proposed
`indications: one by Dr. Villalba, for its anticonvulsant indication, and the other by Dr
`Pokrovnichka, for the neuropathic pain indication.
`In addition cardiology was consulted, not
`to only comment on QT studies, but also on other cardiac issues described below. A CSS
`review is also included with scheduling recommendations (see below). Although the Sponsor
`has simultaneously submitted an application to the European Union for approval, there is no
`foreign marketing experience.
`
`[)(4)
`
`Dr Villalba principally reviewed phase 1 to phase 3 studies relevant for the epilepsy indication.
`Dr Vilalba also reviewed safety data from the studies using the \ iv as well as the
`tablet formualtions. Dr Pokrovnichka’s review concentrated on the tablet as it applied to the
`indication for diabetic neuropathic pain (DNP). The application includes a total of 4012
`unique adult subjects exposed to LCM (including all routes of administration, all indications
`and healthy volunteers). Of these, 1338 subjects were in the partial-onset seizure studies (1327
`subjects from studies with the oral tablet) and 2001 subjects in the neuropathic pain studies.
`The exposures in the epilepsy studies where of sufficient dosage and duration and met ICH
`guidelines. The database included both double-blinded placebo-controlled and open label
`studies. Of the subjects with partial-onset seizures exposed to oral LCM, 199 subjects also
`received IV LCM in Phase 2/3 trials. Intravenous studies were generally shorter in duration
`and either open label or were designed for comparison to the tablet formulation. Assuming
`similar PK and no obvious local issues of irritation, while these studies use a much smaller
`database, they should be considered sufficient for a determination of additional risks over the
`oral formulations.
`
`In her review, Dr. Villalba distinguishes two phase 2/3 safety pools: EP 81 which includes
`patients from al/Jp/aceéo-cwz/m/[eaf doué/ef-é/z'zm’ea’studies and EP 82 which included all
`
`Page 1] of23
`
`1]
`
`

`

`Cross Discipline Team Leader Review
`
`patients receiving drug product in a//phase 2/3 studies. These will be referred to below.
`
`Deaths
`
`No deaths were observed in phase 1 trials. A total of 9 deaths were observed in the epilepsy '
`phase 2/3 studies. Eight of these nine occurred during open label studies. No deaths were
`observed in the placebo group. This leaves a Comparison of 1 in drug Vs 0 in placebo in the
`EP Sl population. It should be remembered that the placebo population was third the size of
`those who received drug in the EP 81 population. These numbers are insufficient to draw any
`conclusions regarding an excess of drug-induced deaths. Four deaths were believed to be
`result of Sudden Unexpected Death in Epilepsy (SUDEP). Calculations by Dr. Villalba
`revealed no excess over that which would be expected in the studied population. On death, in
`a patient with a history of depression, was attributed to a completed suicide. None of the other
`deaths followed a particular pattern that can be easily attributed to a common cause.
`
`There were a total of 15 deaths in patients on lacosamide in the DNP population. Four of these
`(4/1023) were in the controlled studies with none (0/291) in the placebo group. Of the 15 total
`deaths a majority (8) were cardiac-related (ventricular fibrillation, myocardial infarction, heart
`failure (n=2), myocarditis, cardiac arrest (n=2) and sudden death). Such a number is not
`unexpected for a patient population with diabetes and with many patients also having a history
`of hypertension, coronary artery disease, cerebrovascular disease, and/or peripheral vascular
`disease. One of the cardiac deaths was noted to include myocarditis/toxic hepatitis which
`occurred following completion of LCM treatment. This case may represent a suspected case
`of multiorgan hypersensitivity and will be discussed below. Three of the cardiac deaths
`however were observed in the placebo control studies, which may be suggestive of a potential
`cardiac related signal. However, the numbers of patients exposed in the placebo population is
`substantially-lower then that in the drug population. These data however must be viewed
`against the background of other cardiac events, which will be discussed below. All but two of
`the remainder of deaths (5) was from a variety of cancers. No one type stood out. One case
`of a completed suicide was observed. Suicide and suicide ideation will be discussed below.
`
`Other Serious Adverse Events
`
`Comparison by Dr Villalba of rates of serious adverse events in the EP Sl epilepsy population
`revealed a higher rate amongst patients on drug then on placebo: :i.e. 6.5% and 3.8%,
`respectively. No obvious dose response was observed for these grouped rates. The most
`frequent reported serious adverse events, classified by system organ class (SOC), were
`Nervous Systems Disorders (1.6% in placebo and 2.1% in lacosamide in the EP S] pool). The
`most frequent single preferred term was “convulsions” with 0.8 in placebo and 0.8 in the
`lacosamide group. While it may be unexpected that these rates are the same, when you lump
`all other epilepsy preferred terms (e.g. epilepsy, complex partial seizures, etc) you observe a
`comparison of 1.7% Vs 1.3%, in placebo and drug. Other, non-convulsive serious CNS
`adverse events observed which were more common in drug as compared to placebo groups,
`were dizziness, nystagmus, coordination abnormal, loss of consciousness and tremor. _No
`
`Page 12 of23
`
`12
`
`

`

`Cross Discipline Team Leader Review
`
`placebo patients exhibited these vents. Except for dizziness and nystagmus which were
`observed in 0.3% and 0.2% of patients, respectively, all events occurred in only 0.1% of
`patients (1 patients). Although the numbers of some of these events are low, many of these
`events are common with other anticonvulsants, with CNS adverse events limiting the dose that
`can be used. These are very common adverse events reactions associated with this class of
`anticonvulsants.
`'
`
`The next most frequent SAEs in the EP 81 population for patients with epilepsy were in the
`Psychiatric disorders SOC (0.7% for LCM and 0 for placebo-treated patients). Psychiatric
`events included preferred terms such as hallucinations, epileptic psychosis, psychotic
`disorders, completed suicide (see above), suicide attempt and insomnia. As Dr Villalba points
`out, the risk of such events are commonly seen in patients with epilepsy and although occurred
`in small numbers were only observed in patients receiving lacosamide. Dr. Villalba
`consequently reviewed each case, many of which an alternative explanation could be found
`(e.g. previous history of similar behavior). Dr Villalba suggested that the low numbers and
`perhaps other explainable cases undermine a casual attribution to drug use. This CDTL
`agrees.
`
`The next most common SAEs in the EP 81 population for patients with epilepsy included GI
`disorders systems (0.6% for LCM and 0.3% for placebo—treated patients, respectively) and
`infections (0.5% for LCM and 0.3% for placebo-treated patients, respectively). These events
`were not thought to be related to treatment.
`
`Examination of SAEs in the EP 82 pool for the epilepsy population did not reveal much
`additional information. A high number of injuries from fracture were noted (16 patients) and
`were possibly thought to be related to dizziness and ataxia, which appear to be drug related.
`
`Serious adverse events in the DPN controlled population Were similar to that reported for the
`epilepsy population: i.e. 6.6% (68/1023) of subjects who received lacosamide and 4.8%
`(14/291) of subjects who received placebo. The highest rate in the controlled database, by
`SOC, of serous adverse events was under the classification of cardiac disorders. There was
`actually a greater percen