_..-.». ___.‘.a~..-.
`
`.
`
`.. .-
`
`..¢..’._.---
`
`' REST PossiéL‘E‘Co‘Pv'
`BEST POSSIBLE COPY
`
`‘ " '
`
`33
`
`REMSLEEPLATENCY
`
`The sponsor reports that the results of this analysis were similar to the findings for Average
`Sleep Latency for the modafinil 400 mg treatment group compared to placebo at Endpoint.
`However, the modafinil 200 mg group did not differ significantly from placebo. Review of the
`sponsor's Tables 8.2.0 through 8.2.9 (Item 8, Vol. 9, p. 03269-03288) confirms this
`conclusion.
`
`TOTAL SLEEP “ME
`
`The sponsor reports that on average, patients in the modafinil 400 mg treatment group
`exhibited lower Total Sleep Time (8.04 min.) than patients in the modafinil 200 mg treatment
`group (8.31 min.) or the placebo treatment group (9.74 min.) at Endpoint. However, the
`analyses were only statistically significant for the active treatment groups vs. placebo. The
`results for scheduled visits were similar to those seen at Endpoint. Review of the sponsor’s
`Tables 8.3.0 through 8.3.9 (Item 8, Vol. 9, p. 03289-03308) confirms these findings.
`
`PATIENT SUBJECTIVE EVALUATION OF SLEEP LATENC’Y
`
`3 -
`
`( '_
`
`,
`
`.
`
`Patients were asked to estimate how long they were able to stay awake at the end of each test
`period. The sponsor reports that both treatment arms and the combined treatment groups
`
`values were all significantly greater than the placebo group values on Tests 2, 3, 4. and on
`
`average (p-values <0.010) at Endpoint; and that no significant difference was found between the
`two active treatment groups at Endpoint. They also state that similar comparisons were made
`for each scheduled study visit. but do not summarize the results of these evaluations. Review of
`the sponsor’s Tables 8.4.0 through 8.4.9 (Item 8. Vol. 9, p. 03309-03328)
`confirms these
`findings for the Endpoint evaluations and documents significant drug effect at the other study
`visits with the following exceptions: 1) modafinil 200 mg compared to the placebo group for
`Test 3, Week 6 (11.7 min. vs. 9.01 min., respectively: p = 0.066), 2) combined active
`treatment groups compared to placebo group for Test 4. Week 6 (11.79 minJ12.18 min. [400
`mg/200 mg] vs. 9.01, respectively; p = 0.057) and, 3) modafinil 400 mg compared to placebo
`group for Test 4, Week 6 (11.79 min. vs. 9.01 min.. respectively; p = 0.237).
`
`The sponsor also reports that their analysis comparing the number of tests for which patients
`subjectively evaluated themselves as having stayed awake during the entire test revealed that
`the modafinil 400 mg, 200 mg and combined treatment groups all exhibited significantly more
`patients reporting staying awake at Endpoint than patients in the placebo group (all p-values <
`0.050). They also note that similar comparisons were made for all other scheduled visits. but
`do not summarize those results. However. review of the sponsor's Table 8.4.10 (Item 8. Vol.
`9. p. 03329-03331) documents a significant treatment effect in the 200 mg group vs. placebo
`group only for Week 3 (p = 0.008); the p-values for Weeks 6, 9 and Endpoint are 0.083.
`0.159 and 0.063. respectively. for this treatment group comparison.
`
`
`
`

`

`.A. .
`
`.
`
`.
`
`\
`
`.-_-._.«._, __.-.
`
`4
`
`--
`
`"BEs'T-‘POSSIBLE COPY!-
`BEST POSSIBLE COPY
`
`34
`
`QQtC;
`
`Patients in both active treatment groups had significantly greater improvement than patients in
`the placebo group at each visit. These results are summarized in the following table:
`
`Table 15.
`
`Patients Improved In CGI-c - Efficacy Evaluable Population
`
`w...
`
`Weeks
`
`Week9
`
`(_ ,
`
`‘
`
`.
`
`so (73%)
`
`so (64%)
`
`32 (37%
`
`" Modafinil 400 mg or 200 mg compared_to placebo
`
`[based on sponsor’s Table 75. Item 8, Vol. 8, p. 03067]
`
`Patients in the modafinil 400 mg treatment group had significantly greater improvement than
`the patients in the modafinil 200 mg treatment group only at Weeks 3 and 6 (p-values = 0.004
`and 0.049, respectively). No treatment group comparisons were significant for the CGI
`severity scores at baseline (all p-values > 0.200).
`
`MSLI;
`
`SLEEP LATENCY (16 sec)
`
`Patients in the modafinil 400 mg treatment group exhibited a longer average Sleep Latency (16
`sec). i.e. to the first 16 seconds of continuous sleep. time (5.15 min.) than patients in the
`modafinil 200 mg (4.70 min.) or placebo (3.29 min.) treatment groups at Endpoint. The
`sponsor’s analyses found statistically significant treatment effects for the 400 mg. 200 mg and
`combined active treatment groups when each is compared to the placebo group (p = 0.006.
`0.006 and 0.001. respectively). They also note statistically significant increases from
`Baseline in the active treatment groups of 1.86 min. in the 200 mg group and 1.85 min. in the
`400 mg group (p-values both <0.001). No significant differences were found between the two
`active treatment arms. These results are confirmed by review of the sponsor’s Tables 9.0.0 and
`9.0.1 (Item 8. Vol. 9. p. 03332-03333).
`
`I
`
`_
`
`

`

`r .:.--1;
`
`BEST POSSIBLE COPY
`BESTi‘POSSIBLE-COPY-I
`
`35
`
`The sponsor reports that the 400 mg, 200 mg and combined active treatment group Sleep
`Latency (16 sec) times were all significantly increased compared to the placebo group results
`for Tests 2, 3. 4, and the four test average (all p-values < 0.050). Review of the sponsor's
`Tables 9.0.0 through 9.0.9 (Item 8. Vol. 9. p. 03332-03341) confirms these findings, with
`the exception that the modafinil 400 mg to placebo comparison for Week 9. Test 4 resulted in a
`p-value of 0.053.
`
`The sponsor reports that there were no significant differences between treatment groups for the
`Number of Patients Staying Awake for 0 to 4 Tests at Week 9 or Endpoint with all p-values >
`0.050. This is confirmed by review of the sponsor’s Table 9.0.10 (Item 8, Vol. 9, p. 03342-
`03343L
`
`REM SLEEP LATENCY
`
`There were no significant differences between treatment groups in Rem Sleep Latency at Week 9
`or Endpoint. All three treatment groups (including placebo) exhibited significant increases in
`mean Rem Sleep Latency time from the Baseline results (all p-values <0.050). When the
`individual Tests were analyzed, the 400 mg. 200 mg and combined active treatment groups
`were significantly better than placebo only for Test 2 (p = 0.002) at Week 9 and Endpoint.
`There were no significant differences between the active treatment groups for any individual
`test or on average. These findings are confirmed by review of the sponsor’s Tables 9.1.0
`through 9.1.9 (Item 8. Vol. 9. p. 03344-03353).
`
`FIRST CONTINUED SLEEP LATENCY
`
`The sponsor reports that these results are similar to those for the Sleep Latency (16 sec). That
`is confirmed by review of the sponsor's Tables 9.2.0 through 9.2.9 (Item 8, Vol. 9, p. 03354-
`03363)
`
`STAGE 2 AND STAGE 3 SLEEP LA‘IBthY
`
`The sponsor reports that there were no significant differences between treatment groups for
`any test period or on average (all p-values > 0.100) at Week 9 or Endpoint. This conclusion is
`confirmed by review of the sponsor’s Tables 9.3.0 through 9.3.9 (Item 8, Vol. 9. p. 03364-
`03374).
`
`Data was also collected regarding Stage 1" sleep. Similar results were reportedly seen in each
`treatment group. Twenty percent or less of patients did not achieve Stage III sleep during one or
`more tests. No significant differences were found in any treatment group comparisons (all p-
`values > 0.100) at Week 9 or Endpoint. These results are continued by review of the sponsors
`Table 9.4.0 (Item 8. Vol. 9. p. 03374-03375).
`
`PATimT SUBJECTIVE EVALUATION OF SLEEP LATENCY
`
`The sponsor reports that the two active treatment arms showed significant (p-values < 0.001)
`mean Patient Subjective Evaluation of Sleep Latency value increases from baseline at Endpoint.
`The 400 mg treatment group and combined active treatment groups were significantly better
`
`'
`
`

`

`J--—-§-?3
`
`A, "Q-BESTPOSSIBLE COPY. .. -.
`BEST POSSIBLE COPY
`
`.
`
`36
`
`than placebo for Test 3 only (p 5 0.006) at Endpoint. No significant difference was found
`between the 200 mg group and the placebo group or between the two active medication groups
`for any individual test or on average (all p-values > 0.050). These results are confirmed by
`review of the sponsor’s Tables 9.5.0 through 9.5.10 (Item 8, Vol. 9, p 3376-03387); with
`the additional observation from this review that there were no significant differences between
`any treatment groups in the observed values for the Patient Subjective Evaluation of Sleep
`Latency at Week 9 or Endpoint.
`
`Efi:
`ESS scores from patients in the modafinil 400 mg and 200 mg treatment groups were
`significantly lower than scores from patients in the placebo group. This indicates less
`likelihood of falling asleep or dozing during the listed activities. These results are summarized
`in the following table:
`
`Table 16. Observed Values for E85 Score by Visit
`Population
`
`- Efficacy Evaluable
`
`Visit
`
`Week 3
`
`Week 6
`
`Week 9
`
`Statistic Modafinil 400 m Modafinll 200 m—
`
`17.1 i 4.2
`
`17.9 i 3.3
`
`18.3 1 3.3
`
`mean i s.d.
`
`12.6 :t 5.6ab
`
`14.0 :I: 5.4813
`
`16.8 i 4.7b
`
`meanis.d.
`
`12.6 :l: 5.6ab
`
`13.9 :l: 6.0“
`
`16.8 i 4.8b
`
`meanis.d.
`
`13.0 :I: 5.7813
`
`14.4 :I: 5.7ab
`
`17.1 :l:5.0b
`
`
`
`
`EIII
`
`
`
`
`
`meanis.d.
`
`12.8 i 5.8“
`
`I
`
`14.3 i 5.7“?
`
`I
`
`18 significantly different from placebo (p < 0.001)
`
`significantly different from baseline (p < 0.001)
`
`17.0 :t 4.9b
`
`
`
`
`
`
`
`
`
`[based on sponsor's Table 76, Item 8, Vol. 8. p. 03073]
`
`Comparisons between treatment groups showed modatinil 400 mg, modafinil 200 mg and
`modafinil combined treatment group ESS scores were all significantly lower than placebo for all
`scheduled visits and Endpoint (all p-values < 0.001). No significant difference was found
`between the two active treatment groups for any scheduled visit or at Endpoint. These findings
`are confirmed by review of the sponsor's Table 10.0.0 (Item 8, Vol. 8., p. 03388-03389).
`
`‘
`
`
`
`

`

`' “
`
`BEST POSSIBLE COPY
`_-. -- "“‘VBESTPOSSIBL'E‘COPY'
`
`37
`
`EQEE:
`
`On average the patients in the modafinil 400 mg treatment group hit a smaller percentage of
`obstacles at Endpoint (5.9%) than either the modafinil 200 mg treatment group (7.5%) or the
`placebo treatment group (7.9%). However. none of the pairwise comparisons indicated
`significance between any of the treatment groups (all p-values > 0.100). At Weeks 3 and 6
`each modafinil treatment group showed a significantly lower percentage than the placebo
`treatment group (all p-values < 0.050).
`Improvement from baseline was significant for
`modafinil 400 mg at Week 6 (p = 0.020) and for modafinil 200 mg at Weeks 3 and 6 (both p-
`values < 0.010). These findings are confirmed by review of the sponsor's Table 7H (Item 8,
`Vol. 8, p. 03076) and Tables 11.1.0 and 11.1.1 (Item 8. Vol. 8, p. 03396-03399).
`
`WW
`
`There was a reduction from Baseline to Endpoint in periodic leg movements of sleep for the
`modafinil 400 mg treatment group only. The number of periodic leg movements accompanied by
`arousals were similarly reduced for this group. Differences from baseline and between
`treatment groups for other parameters were not significant.
`
`W T
`
`he sponsor reports that, on average. patients in the modafinil 400 mg and 200 mg treatment
`groups reported, at Endpoint,
`fewer minutes of sleep during the day, fewer episodes of
`unwanted sleep during the day and fewer episodes of desire for sleep during the day. The
`actively treated patients also reported more cataplectic attacks per day than did placebo treated
`patients at Endpoint. However. this was also true at Baseline. The incidence of daily reported
`cataplectic attacks decreased during the study for both modafinil treated groups but not for the
`placebo group; similar results were observed at Weeks 3. 6 and 9. The ratings of the ability to
`resist sleep and of general alertness were similar among all three treatment groups throughout
`the study.
`
`Reportedly, there were no negative effects of modafinil treatment on nightly sleep. The active
`treatment groups reported fewer awakenings, fewer episodes of hypnagogic hallucinations and
`fewer episodes of sleep paralysis.
`
`The sponsor did not report the results of any statistical analyses of this data.
`
`QQlJN;
`
`Higher numbers of patients in the two active treatment groups. compared to patients in the
`placebo group. responded positively to questions regarding 'feelings about life as a whole”,
`'quality of life during the past week”. ‘general health',
`'social functioning”.
`'productivity',
`“bodily pain”. and ‘driving capability.‘
`
`However, the sponsor did no report the results of any statistical analyses of this data.
`
`/".\,
`
`

`

`7' ' BEST-POSSIBLE COPY
`BEST POSSIBLE COPY
`
`-.
`
`.~.
`
`38
`
`Section 7.2.1.6 Reviewer's Efficacy Discussion:
`
`For this study the sponsor chose two primary efficacy variables, average sleep latency at
`Endpoint on the MWT and improvement of the CGl-C score. Both variables compared modafinii
`400 mg to placebo. By their analyses, modafinii 400 mg was proven effective for the treatment
`of excessive daytime sleepiness in the narcoleptic patient, with p < 0.001 for each variable.
`
`There were numerous secondary efficacy variables“ Sleep latency and REM latency results on
`the MWT supported the efficacy of modafinii over placebo, but did not support a difference
`between the two modafinii doses, 200 mg and 400 mg. The Total Sleep Time documented
`increases in improvement in ability to stay awake in a dose-dependent manner. However, the
`analyses were again only statistically significant for the active groups versus placebo.
`Subjective evaluation by patients and CGl-C confirmed the above findings.
`
`MSLT sleep latencies also noted improvement in sleep latency (lengthening) in both dose groups
`and the combined dose groups compared to placebo; and no significant difference between the two
`active treatment arms. The MSLT REM latencies were not as supportive, and there were no
`significant differences between the active groups at any point. Modafinil did not appear to affect
`Stage ll or ill sleep latency. Subjective patient evaluations of sleep latency on the MSLT did not
`indicate consistent improvement with either the combined active treatment groups compared to
`placebo, either individual dose of modafinii compared to placebo, or between the active
`treatment groups.
`
`Scores on the E88 again showed significant improvement with either the combined treatment
`group versus placebo or the individual dose groups versus placebo. No significant difference
`was found between the two active treatment groups. SCPT results noted a trend towards
`improvement with treatment, but no significant pairwise comparisons were observed. Patient
`Sleep Log results and QOLIN results did appear to find trends towards improvement with either
`dose compared to placebo, but the results are difficult to interpret.
`
`Discussion with Dr. David Hoberman. statistical reviewer for this NDA, confirmed the
`sponsor’s conclusions regarding the statistical analyses of the primary efficacy variable data.
`
`APPEARS THIS WAY 0N ORIGINAL
`APPEARS THIS WAY ON ORIGINAL
`
`
`
`

`

`‘-
`
`>
`
`BEST POSSIBLE COPY
`
`' “"BEST'POS’SIBL‘E‘COPY" --
`
`'
`
`3 9
`
`SECTION 7.2.2
`
`STUDY 01538a/302/NA/US:
`
`Section 7.2.2.1
`
`Protocol Synopsis:
`
`line; A Nine-Week Placebo-Controlled, Double-Blind, Randomized. Parallel-Group Study of
`the Safety and Efficacy of Two Fixed Doses (200 mg, 400 mg) of Oral Modafinil in
`Patients with Narcolepsy Followed by a 2-Week Discontinuation Segment, Followed by a
`40 Week, Open-Label. Flexible-Dose Continuation Study.
`
`'The purpose of the double-blind treatment segment (segment I) of this study is
`Qbiefljxes;
`to compare the safety and efficacy of two fixed doses of modafinil and placebo in the treatment of
`patients with narcolepsy...
`
`'The purpose of Segment II (the 2 week double.blind discontinuation segment) is to determine
`the effect of abrupt, double-blind discontinuation of modafinil on subsequent selected efficacy
`and safety assessments. The purpose of the open label segment of this study is to collect
`additional information regarding the safety and persistence of effect of modafinil during extended
`exposure. Efficacy and safety data are collected for hypothesis generation."
`
`[item 8, Vol. 18, p. 07549]
`
`W T
`
`he double blind segment I portion of this study is a multicenter, randomized, parallel group,
`placebo controlled, fixed dose study of modafinil in patients with narcolepsy. Segment II is a
`double blind discontinuation phase which will follow the treatment phase. The open label phase
`is a 40 week, flexible dose study. The double blind phase will begin with a screening period
`followed by randomization to either placebo or one of two dosage levels of modafinil for a period
`of nine weeks. The protocol calls for three groups of 95 patients each to be randomly assigned to
`one of the three treatment arms. Approximately 15 patients are to be randomized at each of the
`twenty sites during a six month enrollment period. The number of patients entering Segment II
`will be determined by the number of patients completing Segment I. Eligible patients will
`receive a specified number of tablets to be taken daily for nine consecutive weeks in one of the
`following three treatment groups:
`
`Group I
`Group II
`Group III
`
`placebo
`modafinil, 200 mg/day
`modafinil, 400 mg/day
`
`Both Group II and III patients will be closed 100 mg/day of modafinil for the first week of
`Segment l.
`
`APPEARS THIS WAY ON ORIGINAL
`APPEARS THIS WAY ON ORIGINAL
`
`
`
`

`

`W .
`
`.-.._;.;-',.‘-__-‘._--.”BEST-POSSIBLE.copv’.
`BEST POSSIBLE COPY
`
`Table 17. Dosing Schedule
`
`40
`
`Weeks 1
`
`through '
`
`
`
`Number of Tablets
`
`Modal‘inll Dos
`
`
`
`
`MM
`400 mg modailnil
`
`Weeks 9 through
`1 1
`
`Week 2-9: 4 x 100 mg modaiinil'
`
`400 mg/day
`
`Number of Tablets
`
`Modaflnil Dos-
`
`
`
`
`
`
`
`
`
`
`
`
`— 4 x placebo
`E
`
`0 mg/day
`
`o
`
`Wig.)
`4 x 100 mg modafinil
`_[gupn 9 (29°19)
`' Group III a and b patients will be titrated to 200 mg Day 8, 400 mg Day 9
`
`400 mg/day
`
`[based on sponsor's Dosing Schedule Tables, Item 8, Vol. 18, p. 07570]
`
`Patients eligible for the double blind phase will be male or female outpatients, 18 to 65 years
`of age, inclusive. Females must be either surgically sterile, two years postmenopausal, or,
`if of
`child bearing potential, using an acceptable birth control method. Patients must have a current
`diagnosis of narcolepsy. Nocturnal polysomnography and a Multiple Sleep Latency Test (MSLT)
`will be done at the Screen Visit unless they have been perion'ned within the five years prior to
`screening. The diagnosis of narcolepsy must include the following characteristics: recurrent
`daytime naps or lapses into sleep that have occurred almost daily for at least three months, and
`a history of loss of postural muscle tone in association with intense emotion. i.e. cataplexy.
`Associated features may include: sleep paralysis. hypnogogic hallucinations, automatic
`behaviors, and nocturnal sleep disruption. The MSLT must document a mean sleep latency of s 8
`minutes and two sleep onset REM periods. Eligible patients must demonstrate an absence of any
`medical or psychiatric disorders that could account for the excessive daytime sleepiness.
`Patients requiring routine use of anticataplectic medication will be excluded. All drugs or
`substances with psychotropic effects are prohibited during the study. although deviation from
`this criterion may be approved by the 'study medical monitor.
`
`r .
`
`

`

`-. .
`
`1 - N‘- .- ~ -.._BEST POSSIBLECOPY j‘
`BEST POSSIBLE COPY
`
`41
`
`Figure 2.
`
`Study Schemata - Segment
`
`I/Segment
`
`ll
`
`VISIT NUMBER:
`
`1
`
`z
`
`3
`
`5
`
`5
`
`g
`
`z
`
`a
`
`STUDY DAY:
`
`.
`
`.
`
`.
`
`{48° visit)
`
`,
`{48° visit}
`
`z]
`
`n
`
`in addition to examination and laboratory studies, the patient will
`At the Baseline Visit,
`complete: an Epworth Sleepiness Scale; ‘Steer Clear” Perlonnance Test training session; two
`nocturnal polysomnography recordings within 48 hours, one followed by a MSLT and the other
`followed by a Maintenance of Wakefulness Test (MWI'); one “Steer Clear” Performance Test
`(SOFT); Patient’s Daily Sleep Log; Baseline Signs and Symptoms; Quality of Life in Narcolepsy
`(QOLIN) patient inventory; and, Clinical Global Impression of Severity (CGl-S).
`
`.
`( 7.
`
`_
`
`-
`
`On-Study Visits during Segment l(end of Weeks 1, 3 and 6 following the Baseline Visit) will
`include, in addition to vital signs and laboratory studies: HLA typing at Week 1; a nocturnal
`polysomnography recording; a MWT; a SCPT; the Clinical Global Impression of Change (CGl-C)
`Scale; the Epworth Sleepiness Scale; the Patient’s Daily Sleep Log; and, adverse experience and
`concomitant drug review.
`
`in addition to physical examination and laboratory studies:
`The Termination Visit will include,
`two nocturnal polysomnography recordings within 48 hours. one followed by a MSLT and the
`other by a MWT; one SCPT; the CGl-C; the Epworth Sleepiness Scale; the Patient’s Daily Sleep
`Log; the QOLIN inventory; and adverse experience and concomitant drug review.
`
`Discontinuation evaluations for Segment ii at end of Weeks 10 and 11 will include: physical
`exam, clinical laboratory tests. vital signs, 12-lead ECG. recording of concomitant medications
`and AE's. At end of Week 11 patients will also complete: MWT. ESS. SCPT. Patient Daily Sleep
`Log, and, as baseline data for the Open Label Phase, urine drug screen, modafinil plasma level,
`QOLIN patient inventory. and CGI-S.
`
`Patients entering the open label phase must have completed the double blind phase or at least
`two efficacy evaluations post Baseline of the double blind phase and terminated for reasons other
`than noncompliance or a study drug related adverse experience.
`
`
`
`Qualifying patients will begin the open label phase taking 200 mg/day of modafinil for one week.-
`
`~
`
`C -
`
`

`

`._
`
`L
`
`‘
`
`.1
`
`.rg.;..;i:;;‘;43ggtsoissrBLEcopy:
`BEST POSSIBLE COPY
`
`Section 7.2.2.2 Statistical Analysis:
`
`'
`
`There will be two primary efficacy variables, the MWT and the CGI-C in the double blind
`segment. The primary hypothesis will determine if treatment with modafinil (the combined
`200 mg/day and 400 mg/day treatment groups), when compared to placebo, can modify the EDS
`of narcolepsy in patients receiving medication for s 9 weeks. as reflected by the MWT and the
`CGl-C assessed by an independent clinician. Endpoint analysis will be the primary analysis,
`performed to include data from all patients with post-baseline evaluations, and Endpoint defined
`as the last patient evaluation post-baseline. Both the MWT and the CGl-C analyses must result
`in statistically significant efficacy being shown for study drug (p s 0.05, two-tailed) in order
`to support the primary objective of the study. Segment l and Segment II data will be analyzed
`separately. A secondary hypothesis will determine independently. for each of the two modafinil
`doses, whether or not those doses produce beneficial effects measured by the MWT and CGl-C
`which may be attributable to both doses of modafinil.
`If the primary hypothesis is not
`significant, the pairwise comparisons of dose level will utilize the Dunnett test procedure to
`adjust the level of significance.
`
`“Parametric analyses adjusting for investigator effect will be performed if the assumptions of
`normality are met. Parametric analyses will be performed as supportive evidence of treatment
`by investigator homogeneity. Two-tailed tests will be used to test study hypotheses. Mantel-
`Haenszel tests, having investigator as the strata, will also be performed for the CGl-C.” [Item
`8, Vol. 18, p. 07596]
`
`,'
`("
`
`Secondary efficacy variables will include the MSLT. SCPT, ESS, Patient’s Daily Sleep Log, and
`QOLIN patient inventory. Changes from baseline will be compared between treatment groups
`when applicable.
`
`‘ *
`
`For Segment ll. within group statistical comparisons of Baseline and Week 9 MWI' results with
`MWT results obtained at the end of the discontinuation segment will be conducted. Persistence of
`effect will also be examined during the open label portion of the study.
`
`Section 7.2.2.3
`
`Protocol Amendments:
`
`Amendmenu;
`
`This amendment was dated 3/3/95.
`
`It consists of the following features:
`
`( -
`
`1 )
`Exclusion criteria have been expanded to include prior responses to stimulant
`medication such as chest pain. ischemic ECG changes or clinically significant cardiac
`arrhythmia; also excludes clinically significant manifestations of mitral valve prolapse.
`
`
`
`

`

`_. _- ._
`
`.-__.. ._._H_--,.-..--~.- BEST POSSIBLE COPY '
`BEST POSSIBLE COPY
`
`43
`
`A change in the Inclusion Criteria which provides that the diagnosis of narcolepsy is
`2 )
`based on the criteria established by the American Sleep Disorders Association, published in
`'The International Classification of Sleep Disorders” in 1990. Based on this document the
`patient must meet one of two minimal diagnostic criteria:
`
`Criteria A: recurrent daytime naps or lapses into sleep occurring almost daily for at
`least three months; plus sudden bilateral loss of postural muscle tone in association with
`intense emotion.
`
`Criteria B: a complaint of excessive sleepiness or sudden muscle weakness; plus
`associated features such as sleep paralysis, hypnagogic hallucinations, automatic
`behaviors, and disrupted major sleep episode; plus polysomnography demonstrating
`either 1) sleep latency less than 10 minutes, or 2) REM sleep latency less than 20
`minutes, and 3) an MSLT that demonstrates a mean sleep latency less than 5 minutes,
`and 4) two or more sleep onset REM periods; plus absence of any medical or psychiatric
`disorder that could account for the symptoms.
`
`Eligible patients diagnosed under Criteria A must have a MSLT with a mean sleep latency of s 8
`minutes. Eligible patients diagnosed under Criteria B must have a MSLT with a mean sleep
`latency of s 5 minutes. Eligible patients must have two sleep onset REM periods documented
`within the MSLT.
`
`Allows for the collection of additional modatinil plasma samples to facilitate population
`3 )
`pharmacokinetic analyses.
`
`Addition of a third phase to the protocol which represents an additional 48 week extended
`4 )
`openlabel period.
`
`5 )
`
`6 )
`
`Clarification of the drug supply in the Study Drug section.
`
`A change in clinical monitor.
`
`Amendmentz;
`
`This amendment was dated 8/2/95.
`
`It consists of two features:
`
`Allows patients to take concomitant medications used for the treatment of cataplexy
`1 )
`during the open label phase of the study.
`
`2)
`
`Provides minor editorial revisions.
`
`Amendment;
`
`The exact date of this amendment is not clear from the submission. However it appears to have
`been entered into the protocol at the saine time as amendments 2 and 4, which would make the
`date 3/5/96.
`It consists of two features:
`
`

`

`W7.L.-.'
`."..'.'-...‘.-i—;~~ '---BEs‘1—‘ POSSIBLE COPY‘
`1 ;
`.'.::.'
`_:
`BEST POSSIBLE COPY
`
`44
`
`1 )
`
`Changes one clinical monitor and adds one new clinical monitor.
`
`Allows for the use of 200 mg tablets of modafinil containing a logo during the extended
`2 )
`open label phase of the protocol.
`
`AmendmentA;
`
`This amendment was dated 3/5/95.
`follows:
`
`It revises the statistical analysis section of the protocol as
`
`There are two primary measures of efficacy: The MWT and the CGl-C score. The
`primary hypothesis will be that treatment with modafinil at 400 mg/day for up to nine weeks
`will result in a statistically significant (5% level of significance, two sided test) increase in
`sleep latency compared to placebo on the MWT and a statistically significant improvement in the
`CGl-C score compared to placebo. The analysis population will include all randomized patients
`who receive study medication and have at least one post-baseline measurement for both MWT
`and CGl-C. The primary analysis endpoint will be the last double blind measurement for each
`patient. Measures of sleep latency will be analyzed using a generalized ANCOVA model including
`effects for treatment group. study site and baseline sleep latency.
`
`The CGl-C score will be analyzed with a generalized Cochran-Mantel-Haenszel (CMH) test
`(Mantel’s Test) for ordinal categorical data, including effects for treatment and baseline CGI
`severity. Severity will be measured with three strata: normal or borderline ill or slightly ill;
`moderately ill; markedly ill or extremely ill.
`
`The following variables will be analyzed as secondary measures of efficacy:
`
`1 )
`
`2 )
`
`MWT
`a)
`b)
`
`c)
`d)
`e)
`
`MSLT
`a)
`b)
`
`c)
`d)
`e)
`f)
`9)
`
`sleep latency (min.) (by test)
`sleep latency (percentage of patients remaining awake 20 minute {complete
`success} for O, 1, 2. 3 or 4 tests)
`sleep latency to 10 seconds (average of 4 tests, by test)
`total sleep time (average of 4 tests, time of day)
`patient subjective evaluation of sleep latency (average of 4 tests, by test)
`
`sleep latency (min.)(average of 4 tests. by test)
`sleep latency (categorization of patients remaining awake for the entire 20
`minutes {complete success) for 0. 1, 2. 3 or 4 tests)
`sleep latency to 16 seconds (average of 4 tests, by test)
`latency to REM sleep (average of 4 tests, by test)
`Stage II latency (average of 4 tests, by test)
`Stage Ill latency (percentage of patients with ‘NO” for O. 1. 2. 3 or 4 tests)
`patient subjective evaluation of sleep latency (average of 4 tests, by test)
`
`3 )
`
`Epworth Sleepiness Score (ESS)
`
`

`

`- ‘- -‘- *“BEs’TPOSSIBLE COPY -- _.
`BEST POSSIBLE COPY
`
`4 5
`
`-
`
`total ESS score (0-24) from eight questions
`
`4 )
`
`SCPT
`-
`
`number of obstacles hit, percentage hit and obstacles passed, percentage passed in
`the 30 minute test period
`
`5)
`
`Nocturnal Polysomnography
`
`a)
`
`b)
`
`c)
`
`QOUN
`a)
`b)
`
`time variables (time in bed, time awake after sleep onset, total REM sleep, sleep
`latency, REM latency, etc.)
`duration variables (sleep Stages I,
`percentages, etc.)
`counts (numbers of awakenings, awakenings > 2 min., arousals, periodic leg
`movements in sleep [PLMS], PLMS with arousals, PLMS with awakening,
`respiratory disturbance index, etc.)
`
`ll, Ill, IV, time in bed, minutes and
`
`qualitative questions scored on an ordinal scale of worst to best
`Visual Analog Scale (VAS) with consecutive ovals representing scale from worst
`to best
`
`-
`Patient Daily Sleep Log
`a)
`time variables (total time asleep, time to fall asleep, etc.)
`b)
`counts (number of episodes of unwanted sleep, number of episodes of desire for
`sleep, number of episodes of sleep paralysis, etc.)
`qualitative variables (feeling of tension/anxiety when getting up in the morning,
`feeling of sleepiness when getting up in the morning, etc.)
`
`c)
`
`6 )
`
`7 )
`
`Analysis of treatment efficacy will be performed after 3, 6 and 9 weeks of double blind
`treatment except for the QOLIN results for which efficacy at 9 weeks will be compared to
`baseline and the Patient Daily Sleep Log for which the average profiles of response overtime in
`the double blind phase will be compared among treatment groups. The patient population for the
`secondary efficacy analyses will be the same as that for the primary analyses.
`
`In addition. as the CGl-C, SCPT, ESS and Patient Daily Sleep Log will be measured after one week
`of treatment when all patients randomized to study drug will be receiving modafinil 100 mg, an
`analysis of treatment effect will be performed for these measures for that timepoint.
`
`Following the Treatment Withdrawal Period (Weeks 9 through 11), at Week 11, the MWT,
`CGl-S, SCPT, ESS. QOLIN and Patient Daily Sleep Log will be performed. The effect of
`withdrawal of modafinil treatment will be assessed by analyzing the change from Week 9 to
`Week 11 within modafinil dose groups, and by comparing modafinil/placebo patients to
`placebo/placebo patients at Week 11.
`
`Continuous secondary efficacy variable will be analyzed with a generalized ANCOVA model.
`Categorical responses will be analyzed with a logistic regression model. Both models will
`include effects for treatment group, study site and baseline value of the variable as a covariate.
`
`-
`
`

`

`-¥~~' -4“.-- »——«--~¥BEST POSSIBLE COPY '
`BEST POSSIBLE COPY
`
`46
`
`Exploratory hypotheses will be performed for the primary and secondary efficacy variables. A
`multivariate ANCOVA and a step-wise selection procedure will be utilized to identify important
`covariates with a statistical criterion for inclusion of p < 0.10.
`
`“For the primary efficacy analyses of MWT and CGl-C, the primary comparison of interest is
`modafinil 400 mg versus placebo. Each comparison will be a two-sided test at the 5% level of
`significance. Since modafinil 400 mg is expected to be more superior to placebo than modafinil
`200 mg. the comparison of modafinil 200 mg versus placebo will be tested as secondary
`hypothesis.
`In addition, a comparison of the combined doses of modafinil (200 mg, 400 mg)
`will be tested versus placebo as a secondary hypothesis.
`
`‘For all continuous analyses, pairwise comparisons among the treatment groups...will be
`performed with variance estimated from the model mean square error (MSE) without
`adjustment. Dose response will be tested by partitioning the treatment sum of squares into
`single degree of freedom tests for linear and quadratic trend. For CGl-C, painivise comparisons
`among treatment groups..