ORIGINAL ARTICLE
`
`Dextromethorphan Plus Ultra Low-Dose
`Quinidine Reduces Pseudobulbar Affect
`
`Erik P. Pioro, MD, PhD,1 Benjamin Rix Brooks, MD,2 Jeffrey Cummings, MD,3
`Randolph Schiffer, MD,1 Ronald A. Thisted, PhD,4 Daniel Wynn, MD,5
`Adrian Hepner, MD,6 and Randall Kaye, MD6 for the Safety, Tolerability, and Efficacy
`Results Trial of AVP 923 in PBA Investigators
`
`Objective: To evaluate dextromethorphan combined with ultra low dose quinidine (DMq) for treating pseudobulbar
`affect (PBA) in patients with amyotrophic lateral sclerosis (ALS) or multiple sclerosis (MS).
`Methods: In a 12 week randomized, double blind trial, ALS and MS patients with clinically significant PBA (a baseline
`score 13 on the Center for Neurologic Studies Lability Scale [CNS LS]) were maintained, twice daily, on placebo,
`DMq at 30/10mg (DMq 30), or DMq at 20/10mg (DMq 20).
`Results: In 326 randomized patients (of whom 283, or 86.8%, completed the study), the PBA episode daily rate was
`46.9% (p < 0.0001) lower for DMq 30 than for placebo and 49.0% (p < 0.0001) lower for DMq 20 than for placebo
`by longitudinal negative binomial regression, the prespecified primary analysis. Mean CNS LS scores decreased by
`8.2 points for DMq 30 and 8.2 for DMq 20, vs 5.7 for placebo (p¼ 0.0002 and p¼ 0.0113, respectively). Other
`endpoints showing statistically significant DMq benefit included, for both dosage levels, the likelihood of PBA
`remission during the final 14 days and, for the higher dosage, improvement on measures of social functioning and
`mental health. Both dosages were safe and well tolerated.
`Interpretation: DMq markedly reduced PBA frequency and severity, decreasing the condition’s detrimental impact
`on a patient’s life, with satisfactory safety and high tolerability. The findings expand the clinical evidence that DMq
`may be an important treatment for patients suffering from the socially debilitating symptoms of PBA.
`ANN NEUROL 2010;68:693 702
`
`Introduction
`
`Pseudobulbar affect (PBA) is a neurologic condition
`
`laughing
`characterized by involuntary outbursts of
`and/or crying incongruous or disproportionate to the
`patient’s emotional state.1 The condition, hypothesized to
`arise from disconnection of brainstem structures from
`cortical inhibition, is associated with underlying central
`including stroke,2 traumatic
`nervous system disorders,
`brain injury,3 Alzheimer disease,4 amyotrophic lateral scle
`rosis (ALS),5–7 and multiple sclerosis (MS).8 Prevalence
`studies have reported that it affects 11% of patients 1 year
`after a stroke,2 11% of patients during the first year after
`traumatic brain injury,9 18% of patients with Alzheimer
`disease,4 10% of patients with MS,8 and 49% of patients
`with ALS.10 In addition to the effects of the underlying
`disorder, PBA can have a severe impact on well being and
`
`social functioning and can be highly disabling, owing in
`part to the stigma attached to loss of emotional control.11
`Yet even with such a significant burden of illness, PBA
`appears to be poorly recognized and consequently is
`undertreated.11,12
`In settings of ALS or MS, dextromethorphan plus
`quinidine (DMQ) has been found to be beneficial in reduc
`ing PBA.13,14 Dextromethorphan (DM) is known to be a
`low affinity, noncompetitive antagonist of the N methyl d
`aspartate glutamate receptor,15 and also a sigma receptor
`agonist.16 To block its first pass metabolism, it was originally
`coadministered with low dose quinidine (Q), a potent cyto
`chrome P450 2D6 inhibitor,17 at DMQ dosage of 30/30mg
`in a capsule taken twice daily. Without such blockade, DM
`blood levels in some ALS patients have been undetectably
`low even following DM dosage as high as 750mg/day.17 In
`
`View this article online at wileyonlinelibrary.com. DOI: 10.1002/ana.22093
`
`Received Feb 24, 2010, and in revised form Apr 30, 2010. Accepted for publication May 20, 2010.
`
`Address correspondence to Dr Pioro, Director, Section of ALS and Related Disorders, Department of Neurology, Neurological Institute, Cleveland
`Clinic, Desk S90, 9500 Euclid Avenue, Cleveland, OH 44195. E-mail: PIOROE@ccf.org
`
`From the 1Cleveland Clinic, Cleveland, OH; 2Carolinas Medical Center, Charlotte, NC; 3David Geffen School of Medicine at University of California at Los
`Angeles, Los Angeles, CA; 4University of Chicago, Chicago IL; 5Consultants in Neurology, Northbrook, IL; and 6Avanir Pharmaceuticals, Aliso Viejo, CA.
`
`VC 2010 American Neurological Association 693
`
`Page 1 of 10
`
`Biogen Exhibit 2028
`Coalition v. Biogen
`IPR2015-01993
`
`

`

`ANNALS of Neurology
`
`atrial fibrillation and flutter, Q is utilized for conversion as
`well as reduction in frequency of relapse. However, Q doses
`are often in excess of 1,000 to 1,600mg/day and may affect
`cardiac function in ways that include prolongation of the
`QTc interval,18 which in turn may be associated with risk of
`ventricular arrhythmias.19 In the treatment of PBA, a formal
`pharmacokinetic/pharmacodynamic analysis has predicted
`that the Q dosage can be reduced to 10mg per capsule (ultra
`low dosage, q), as a treatment referred to as DMq, with
`maintained efficacy and a decreased potential for proarrhyth
`mic risk.20 The present 12 week trial was designed to evalu
`ate DMq at 30/10mg and at 20/10mg twice daily versus pla
`cebo for treating PBA in patients with ALS or MS. An
`additional objective was to determine the pharmacokinetic
`parameters of each DMq formulation in a subset of the
`study population, as will be reported separately.
`
`Patients and Methods
`Design
`This was a 12-week, randomized, double-blind, placebo-con-
`trolled, 3-arm, parallel-group study conducted at 60 centers in the
`United States and South America between December 2007 and
`March 2009. Patients completed screening procedures 1 to 4
`weeks before their baseline visit. At the screening visit, those meet-
`ing all inclusion/exclusion criteria (see below) were randomized
`(1:1:1) to receive placebo, DM 30mg þ Q 10mg (DMq-30), or
`DM 20mg þ Q 10mg (DMq-20). For the first treatment week,
`patients took a single capsule of study drug in the morning. Dur-
`ing weeks 2 through 12, they took study drug once in the morn-
`ing and once in the evening. Follow-up visits occurred at 2, 4, 8,
`and 12 weeks. In addition, for 1 week prior to baseline and
`throughout the trial, patients were required to maintain a diary
`recording the daily number of laughing and/or crying episodes
`experienced, the medications they took, and any adverse experien-
`ces. Patients completing the study were eligible to continue treat-
`ment in a 12-week open-label phase with DMq-30 twice daily.
`The study protocol was approved by local institutional
`review boards or independent ethics committees and was con-
`ducted in accordance with Good Clinical Practice Consolidated
`Guidance, as approved by the International Conference on Har-
`monization (1997), and also with local or national
`laws and
`regulations. Prior to entry, study procedures and risks were
`explained to each subject, and written informed consent was
`obtained. The study’s randomization code (blocked by center
`and by underlying neurological disorder) was computer-gener-
`ated, and study drug was supplied in blister packs of identical-
`looking capsules. The sponsor, all patients, and all investigators
`were blind to treatment identification and allocation.
`
`Patients
`For entry, men or women 18 to 80 years old were required to
`have clinically significant PBA, with a score 13 on the Center
`for Neurologic Study Lability Scale (CNS-LS),21 and a diagno-
`sis either of ALS (by El Escorial criteria22) within the past 30
`
`FIGURE 1: Subjects’ disposition. DMq 30 5 dextromethorphan
`combined with ultra low dose quinidine at 30/10mg; DMq 20 5
`DMq at 20/10mg.
`
`months or of MS or probable MS (by McDonald criteria23).
`Patients were excluded for any evidence of clinically significant
`abnormality on screening electrocardiogram, a family history of
`congenital QT-interval prolongation syndrome, a resting respi-
`ratory rate outside the range of 12 to 20/min, or a resting diur-
`nal oxygen saturation <95%. Patients were also excluded for
`any presence or history of major psychiatric disturbance, includ-
`ing current symptoms of a depressive disorder (or a score >19
`on the Beck Depression Inventory II [BDI-II]24); major sys-
`temic disease or organ dysfunction capable of interfering with
`study assessments or putting the patient at risk; and exacerba-
`tion of the patient’s underlying ALS or MS within the previous
`2 months. Women with childbearing potential were required to
`use a medically acceptable form of birth control; pregnant or
`lactating women were excluded.
`
`Efficacy Assessments
`The primary efficacy outcome was a patient’s change from base-
`line in the number of PBA episodes (laughing and/or crying)
`per day, as recorded in the patient’s diary. Diary data also
`yielded, as secondary outcomes, a responder analysis (the pro-
`portion of patients with an improvement from baseline PBA
`rate, assessed across all degrees of improvement); number of
`episode-free days; and occurrence of
`remission from PBA
`(defined by absence of episodes during the study’s final 14
`days). Additional secondary outcomes were a patient’s change
`from baseline on CNS-LS, which was administered at baseline
`and at each follow-up visit, and on BDI-II, the Neuropsychiat-
`ric Inventory (NPI),25 and the Medical Outcomes Study 36-
`Item Short-Form Health Survey Version 1.0 (SF-36),26 which
`were administered at baseline and at 12 weeks.
`The CNS-LS is a 7-item self-assessment of PBA severity,
`validated for measuring PBA in ALS21 and MS.27 Total scores
`range from 7 to 35; a score 13 is the instrument’s range for
`clinical PBA. The BDI-II is a 21-item self-assessment of symp-
`toms of depression. A total score of 14 19 is considered mild,
`20 28 is moderate, and 29 63 is severe. The NPI is a question-
`naire
`covering 12 neuropsychiatric
`symptom domains;
`it
`
`694
`
`Volume 68, No. 5
`
`Page 2 of 10
`
`

`

`Pioro et aI: Pseudobulbar Affect
`
`TABLE 1: Patients’ Baseline Characteristis (ITT Population)
`
`Characteristic
`
`No.
`
`Age, mean yr (SD)
`
`Females. No. (%)
`
`Ethnic origin, No. (%)
`White
`
`Hispanic
`
`Black
`
`Other
`
`Diagnosis, No. (%)
`
`A15
`
`MS
`
`Time since ALS diagnosis, mean mon (SD)
`
`PBA episodeslday, mean (SD)
`
`CNS LS smre, mean (SD)
`
`BDI H score, mean (SD)
`
`NPI score, mean (SD)
`
`Frequency
`
`Severity
`
`SF 36 score, mean (SD)
`
`Mental Summary
`
`Physical Summary
`
`DMq-30
`
`DMq—20
`
`Placebo
`
`110
`
`53.1 (11.0)
`
`64 (58.2)
`
`80 (72.7)
`
`21 (19.1)
`
`6 (5.5)
`
`3 (2.7)
`
`65 (59.1)
`
`45 (40.9)
`
`22.7 (29.8)
`
`4.7 (9.5)
`
`1.7 (3.4)
`
`3.0 (6.7)
`
`19.8 (4.9)
`
`9.4 (6.1)
`
`6.2 (6.3)
`
`5.8 (3.9)
`
`107
`
`50.8 (11.1)
`
`54 (50.5)
`
`80 (74.8)
`
`21 (19.6)
`
`2 (1.9)
`
`4 (3.7)
`
`68 ((33.6)
`
`39 (36.4)
`
`16.3 (22.9)
`
`6.8 (12.9)
`
`4.1 (11.8)
`
`2.8 (4.2)
`
`21.0 (5.0)
`
`10.9 (5.8)
`
`7.8 (6.7)
`
`7.0 (4.5)
`
`109
`
`50.3 (11.9)
`
`59 (54.1)
`
`83 (76.1)
`
`21 (19.3)
`
`4 (3.7)
`
`1 (0.9)
`
`64 (58.7)
`
`45 (41.3)
`
`13.4 (18.0)
`
`4.5 (7.6)
`
`2.5 (7.4)
`
`2.0 (20)
`
`19.9 (4.7)
`
`10.5 (5.4)
`
`7.0 (6.7)
`
`6.3 (4.5)
`
`44.0 (10.9)
`
`40.1 (10.1)
`
`44.6 (11.2)
`
`37.0 (10.4)
`
`44.9 (10.6)
`
`38.5 (9.8)
`
`1'IT = intent-to-treat; DMq—30 = dexttomethorphan combined with ultra low-dose quinidine at 30/101115 DMq-20 = DMq at
`20I10mg: SD = standard deviation; ALS = amyotrophic lateral sclerosis; MS = multiple scletosk; PBA = pseudobulbar affect;
`CNS-I.S = Center for Neurologic Study Lability Scale; BDI-II = Beck Depression Inventory Second Edition; NP1 = Neuro-
`psychiatric Inventory; SF-36 = Medical Outcomes Study 36—Itun Short-Form Health Survey.
`
`provides a brief, inlbrmant-based assesment of neutopsychiatric
`symptoms and caregiver distres. The SF-36 "B a 36-item
`health-status assessment, with subdomains for Physical Func-
`tioning, Role-Physical, Bodily Pain, General Hmlth, Vitality,
`Social Ftmctioning, Role-Emotional, and Mental Health. Each
`of 2 summary scores (Mental Component and Physical Com-
`ponent) is standardized so that 50 represents the US general
`population norm (for 1998).
`
`Safety/Tolerability Assessments
`At all visits, vital
`were measured, 12-lead electrocardiography
`was performed, and reports of adverse events (AE5) were obtained
`Serious AEs were defined as final,
`life-threatening, significantly
`disabling, or requiring hospitalimtion. Resting diurnal oxygn sat-
`
`uration and nocturnal oxygen saturation were measured (with
`pulse oximetry) at screening and at 2 weelcs.
`diumal oxy-
`gensaturarionwasalsomeasuredat12weJs.Clini<allaboratory
`testingwasperfi1rmedatsa‘eeningandat4and12weelts.
`
`Statistical Analyses
`In the intent-to-treat population, comprising all randomized
`patients, change from baseline in
`epkode rate
`was analymd using longitudinal negative binomial regression,”
`with adjustment for baseline rate, diagnosis, and study site. As
`a sensitivity analysis, change in epkode rate was also assessed by
`a nonlongitudinal negative binomial model. In addition, 12-
`week change in episode rate was analyzed using the Wilamon
`rank sum test. For number of episode-free days, a 2-sample t
`
`November, 2010
`
`Page 3 of 10
`
`695
`
`

`

`ANNALS of Neurology
`
`FIGURE 2: Twelve week time course of pseudobulbar affect
`weekly episode rate and Center for Neurologic Study
`Lability Scale (CNS LS) score (intent to treat population).
`Weekly rates (top chart) are shown as change from baseline
`at each visit in mean daily rates 3 7. CNS LS scores (bottom
`chart) are the means at each visit. DMq 30 5 dextromethor
`phan combined with ultra low dose quinidine at 30/10mg;
`DMq 20 5 DMq at 20/10mg.
`
`test was used. Changes on CNS-LS, SF-36, BDI-II, and NPI
`were analyzed with analysis of covariance, using the method of
`Frison and Pocock.29 Baseline value, study site, and diagnosis
`were covariates. Observed cases were used in the sensitivity
`analyses, with no imputation for missing data. All analyses were
`2-sided hypothesis tests at the 0.05 significance level.
`The safety population comprised all patients who took at
`least 1 dose of study medication. Their AE rates, for types reported
`by 5% of patients in any treatment group, were compared among
`groups, and mean change in resting nocturnal oxygen saturation
`(from baseline to day 15) was assessed by 2-sample t test.
`
`Sample Size Calculation
`Based on PBA episode rates in previous studies of DMq for
`PBA in ALS13 and in MS,14 a sample size of approximately 90
`
`patients (60 with ALS and 30 with MS) per treatment group
`was planned. This size was expected to be sufficient to detect a
`36% reduction in mean episode rate for DMq-30 vs placebo
`with at least 90% power. The study was not powered to test a
`difference between DMq-30 and DMq-20.
`
`Results
`Subjects
`In all, 332 patients were screened, and among them 326
`were randomized, 110 to DMq 30, 107 to DMq 20, and
`109 to placebo (Fig 1). The main reasons for screening
`failure were unwillingness to discontinue disallowed medi
`cations, CNS LS score <13, and BDI II score >19. In all,
`283 patients (86.8% of 326) completed the study, includ
`ing 101 (91.8% of 110)
`in the DMq 30 group, 88
`(82.2% of 107) in the DMq 20 group, and 94 (86.2% of
`109) in the placebo group. Demographically and in base
`line PBA characteristics, the treatment groups were well
`matched (Table 1), except for a higher baseline PBA epi
`sode rate in the DMq 20 group than in the other groups,
`and a longer time since ALS diagnosis in the DMq 30
`group. At entry, no patient had clinical depression.
`
`Efficacy
`Over the course of the study, all 3 groups showed sub
`stantial reduction in daily PBA episode rates relative to
`baseline. However, the reduction in daily PBA episode
`rate was significantly greater in each of the DMq groups
`than in the placebo group. By longitudinal negative bino
`mial model (predefined primary efficacy analysis), the
`
`FIGURE 3: Responder analysis by treatment group (intent
`to treat population). Each curve graphs the proportion of
`patients improved with a given degree of improvement
`from baseline pseudobulbar affect (PBA) rate at endpoint.
`DMq 30 5 dextromethorphan combined with ultra low dose
`quinidine at 30/10mg; DMq 20 5 DMq at 20/10mg.
`
`696
`
`Volume 68, No. 5
`
`Page 4 of 10
`
`

`

`TABLE 2: Twelve-Week Mean Change on CNS-LS, NPI, and BDI-ll
`
`Pioro et aI: Pseudobulbar Affect
`
`DMq—30
`
`DMq—20
`
`Phcebo
`
`Endpoint
`
`CNS LS
`
`No.
`
`Mean change (p vs placebo)
`BDI II
`
`No.
`
`103
`
`-8.2 (0.0002)
`
`103
`
`96
`
`-8.2 (0.01 13)
`
`97
`
`-1.0 (0.2707)
`
`79
`
`-2.6 (0.0938)
`
`54
`
`— 1.6 (0.207)
`
`101
`
`-5.7
`
`101
`
`0.02
`
`66
`
`-1.3
`
`48
`
`— 1.0
`
`Mean change (p vs placebo)
`
`-1.6 (0.0368)
`
`NP1 (fiequency)
`No.
`
`74
`
`Mean change (p vs placebo)
`
`-1.6 (0.6558)
`
`NP1 (severity)
`No.
`
`46
`
`Mean change (p vs placebo)
`
`-0.7 (0.510)
`
`psychianic Inventory; BDI-11 = Beck Depression Invumory
`CNS-IS = Center fin Ncurologic Study-Iabiliry Sale; NPI = Ncuro
`Second Edition; DMq-30 = dextromethorphan combined with ultra low-dose quinidine at 30/10mg; DMq-20 = DMq at
`20/10mg.
`
`treatment effect in each DMq group over that seen in
`the placebo group was an incremental reduction in PBA
`episode rate of 46.9% (p < 0.0001) for DMq 30 com
`pared to placebo and 49.0% (p < 0.0001) for DMq 20
`compared to placebo. By nonlongitudinal negative bino
`rnial model with constant dispersion (predefined efficacy
`sensitivity analysis), the additional improvement over pla
`
`cebo at both dosage levels was also statistically significant
`(p < 0.0001 and p= 0.0370, respectively). The 12 week
`mean change in
`episode rate was -4.1 for DMq 30
`and -3.9 for DMq 20, vs -3.0 for placebo (p= 0.0099
`and p= 0.0048, respectively). Weekly rates (daily rates X
`7) showed signifiarnt decrease at all time points assessed,
`beginning with day 15 (Fig 2, top).
`the 12 week mean
`Among secondary outcomes,
`reduction from baseline CNS LS score was significantly
`greater at both DMq dosage levels than for placebo
`(Table 2), and for DMq 30, the man reduction was sig
`nificant at all time points assessed, beginning with day
`15 (see Fig 2, bottom). Among secondary outcomes
`derived from
`data, the proportion of patients with
`
`an improvement from their baseline PBA rate was higher
`for both DMq 30 and DMq 20 than for placebo, across
`all degrees of improvement (Fig 3). The proportion of
`patients’ episode free days was significantly greater for
`DMq 30 than fin placebo at all time points assessed, and
`for DMq 20 vs placebo at all time points except day 15
`
`(Fig 4, left). lastly, the proportion of patients reporting
`remission of PBA was significantly greater at both DMq
`dosage levels than for placebo (see Fig 4, right).
`
`November, 2010
`
`Page 5 of 10
`
`On BDI II, mean improvement was significantly
`greater for DMq 30 than for placebo (see Table 2). On
`NPI, total scores showed no significant change for either
`dosage vs placebo (see Table 2). On SF 36, improvement
`was significant for DMq 30 vs placebo on the Mental
`Summary score and on its subdomains for social func
`
`tioning and mental health (Table 3).
`
`Safety and Tolerability
`The proportion of patients reporting at least 1 AE was
`similar in all treatment groups, at 82.7% of DMq 30
`recipimts, 79.4% of DMq 20 recipients, and 82.6% of
`
`100
`
`90
`__ ao
`5 7.
`E 60
`°
`'5
`so
`-
`so
`30
`20
`10
`O
`
`100
`
`so
`.. so
`i 7.
`3 60
`0
`*5
`so
`E
`.
`so
`30
`20
`10
`D
`
`
`
`Visit 5
`Vlsvt 4
`V51! 3
`VISII 2
`(DQY 35)
`(Dav-'17)
`(DEV 29)
`(Dav 15)
`-DMq-3D(N=H0) IDMQ-20(N=107) C]P|aoebo(N=10m
`
`"""'”'°"
`
`Flnal 14 Days
`'p<0D5vsD|uct-be
`“ptooosvspioocho
`
`FIGURE 4: Decrease of pseudobulbar affect. as assessed by
`freedom from episodes and by remission (intent to treat
`population). Freedom from episodes (left) was defined as
`the percentage of episodefree days since the preceding
`visit. Remission (right) was defined by absence of episodes
`throughout the study's final 14 days. DMq 30 = dextromethor
`phan corrbined wi1h ultra Iowdose qiriciie at 30/1Gng;
`DMq 20 = DMq at 20/1 Gng.
`
`697
`
`

`

`ANNALS ofNeurology
`
`TABLE 3: Twelve-Week Mean Changes on SF-36 (I'lT Population)
`
`SF-36 Domain,
`
`Mean Change
`(p vs placebo)
`
`Mental Summary
`
`Vitality
`
`Social Functioning
`Role Emotional
`
`Mental Health
`
`Physiarl Summary
`
`Physical Functioning
`
`Role Physical
`
`Bodily Pain
`
`General Health
`
`DMq-30
`(n = 110)
`
`4.5 (00193)
`
`-0.9 (0.2972)
`
`9.3 (0.0033)
`
`11.6 (03658)
`
`5.5 (00028)
`
`-0.8 (05877)
`
`-0.9 (0.2972)
`
`3.5 (0.3063)
`
`4.1 (0.0740)
`
`-1.5 (0.8703)
`
`DMq-20
`(n = 107)
`
`1.8 (0.6792)
`
`-5.3 (0.75l0)
`
`1.4 (05544)
`
`-1.8 (0.6838)
`
`3.1 (0.4457)
`
`-1.0 (0.9967)
`
`-5.3 (0.7Sl0)
`
`-4.3 (0.2292)
`
`5.8 (0.0678)
`
`-3.0 (0.3583)
`
`Placebo
`
`(n = 109)
`
`0.3
`
`-4.1
`
`-3.1
`
`2.4
`
`-0.3
`
`-1.3
`
`-4.1
`
`-1.8
`
`-1.1
`
`-1.3
`
`SF-36 = Medical Outcomes Study 36-10:-.m Short-Form Halrh Survey; I'IT = intent-to-treat; DMq-30 = dcxtromethorphan
`combined with ultra low—dose quinidine at 30/10mg; DMq-20 = DMq at 20/ 10mg.
`
`placebo recipients. Overall, AE incidence was distributed
`evenly throughout
`the study, except for slightly higher
`rates in the DMq 30 and placebo groups during the ini
`tial treatment week. The proportion of patients reporting
`serious AES was also similar across groups, at 7.3% (8
`patients) in the DMq 30 group, 8.4% (9 patients) in the
`DMq 20 group, and 9.2% (10 patients) in the placebo
`group. Two serious AEs, both in the DMq 20 group,
`were reported as possibly treatment related. In 1 of these
`patients, the event was reported as respiratory depression
`and ALS progresion. The other patient had Iorsening
`musde spastrcrty. Seven deaths were reported, a
`In ALS
`patients: 3 In the DMq 30 group, 3 m the DMq 20
`group, and 1 in the placebo group. All deaths were classi
`fied by an independent mortality adjudication committee as
`having a respiratory cause likely to be the result of progres
`sion of the underlying neurologic disease. No acute decom
`pensation of respiratory function after iniliation of study
`drugwasobserved,andnodeathswereascribedtoacardiac
`muse. Discontinuations due to AEs were more frequent in
`11.: DMq 20 group, at 9.3% (10 patients), tlnn an the
`DMq 30 group, at 5.5% (6 patients), or the placebo group,
`at 1-8% (2 1miems>- Among freqtrndy reported AEs
`(r‘*“‘’ 4)’ ‘‘‘”"‘= “W """"‘°"~ ""1 “""“'>’ ‘W
`infection were more frequent for DMq 30 than for placebo,
`whereas falls, headadre, somnolence, fatigue, and other AES
`occurred at rates resembling those for placebo.
`resting
`Vital signs, physical examination findings,
`diurnal oxygen saturation, and clinical laboratory values
`showed no significant changes from their baseline means
`in any treatment group. For resting nocturnal oxygen sat
`uration, Table 5 compares baseline and day 15 findings.
`
`698
`
`Page 6 of 10
`
`At day 15, the mean change was -0.2 percentage points
`in the DMq 30 group and -0.7 percentage points in
`the DMq 20 group, vs -0.1 for placebo (p= 0.794 and
`
`TABLE 4: Adverse Events Reported by 25% of
`Any Group (Safety Population)‘
`
`five.“ Type’
`0' 0%)
`Fall
`Dizziness
`Headache
`
`Nausea
`Diarrhea
`Sornnolence
`
`Dlulrlailo DNEFIZO plafblo
`(n _
`) (n _ 07) (n _ 09)
`22 (20_())
`14 (13_1)
`22 (203)
`20 (182)
`11 (103) 6 (55)
`15 (13.6)
`15 (14.0)
`17 (15-6)
`
`14 (127)
`11 (10.0)
`11 (10.0)
`
`10 (92)
`8 (75)
`14 (13.1) 7 (6.4)
`9 (8.4)
`10 (9.2)
`
`Fatigue
`
`9 (8.2)
`
`11 (10.3)
`
`10 (9.2)
`
`Nasopharyngitis 9 (8.2)
`
`6 (5.6)
`
`8 (7.3)
`
`8 (7-3)
`
`Urinary tract
`inf°°ti°"
`7 (6-4)
`Constipation
`7 (6.4)
`Muscle spasms
`Muscle weakness 6 (5.5)
`_
`5 (45)
`Dysplugla
`Pain in extremity 5 (4.5)
`
`4 (3-7)
`
`3 (2-3)
`
`7 (6-5)
`4 (3.7)
`5 (4.7)
`
`6 6'6)
`2 (1.9)
`
`9 (8-3)
`10 (9.2)
`4 (3.7)
`
`4 6'7)
`8 (7.3)
`
`6 (5.5)
`1 (0.9)
`0
`Depression
`=3’. M¢.«1DRA Prcfinad term, lkred by fiequency in the
`DMq—30 group.
`DMq-30 = dcxuomcthorphan combined widr ultra low-
`dme quinidine at 30/10mg; DMq—20 = DMq at 20Il0rng.
`
`Volume 68, No. 5
`
`

`

`Pioro et aI: Pseudobulbar Affect
`
`TABLE 5: Summary of Nocturnal Oxygen-Saturation Data (Safety Population)
`
`Data
`
`DMq—30
`(n = 108 or 106’),
`n = 110
`
`DMq-20
`(n = 102 or 100’),
`n = 107
`
`Placebo
`(n = 108 or 109‘),
`n = 109
`
`94.1 (5.4)
`
`94.4 (2.1)
`
`94.9 (2.0)
`
`94.1 (2.5)
`
`-0.2 (2.0) (0.794)
`
`-0.7 (2.0) (0.039)
`
`94.6 (2.2)
`
`94.4 (2.2)
`
`-0.1 (2.1)
`
`Saturation, mean % (SD)
`
`At baseline
`
`At day 15
`
`Mean change
`(p vs placebo)
`Number of events
`<88%. mean (SD)
`
`At baseline
`
`At day 15
`Total time in minutes
`<88%, mean (SD)
`
`At baseline
`
`At day 15
`
`11.2 (39.2)
`
`4.1 (19.7)
`
`9.9 (39.1)
`
`11.0 (41.7)
`
`9.1 (23.1)
`
`11.2 (20.9)
`
`‘For saturation and desaturation analyses, respectively.
`DMq—30 = dextromethorphan combined with ultra low-dose quinidine at 30/10 mg; DMq-20 = DMq at 20/10 mg; SD =
`standard deviation.
`
`p = 0.039, respectively). The differences between groups
`were not clinically signifimnt. Descriptive analyses of desa
`
`izcd in Table 6. At all time points assessed, no DMq recipi
`ent had a QTC interval absolute value >480 milliseconds
`
`turation data identified no substantial differences between
`groups (see Table 5). QTc interval changes are summar
`
`(with Fridericia correction) or a change from baseline >60
`milliseconds.
`
`TABLE 6: Summary of O.Tc-Interval Data (Safety Population)
`
`Data
`
`DMq-30 (n = 110)
`
`DMq-20 (n = 107)
`
`Placebo (n = 109)
`
`QTcB/QTcF at baseline, mean ms
`
`418.2/406.6
`
`QTcB/QTcF at day 84, mean ms
`
`420.6/411.8
`
`QTcB/QTcF change from baseline,
`H1631! HIS
`
`3.0/4.8
`
`Proportion of postbaseline ECGs with
`absolute QTCB/QTCF
`
`>450 ms
`
`>480 ms
`
`>500 ms
`
`Proportion of postbaseline ECGs with
`change from baseline QTCB/QTCF
`30 60 ms
`
`>60 ms
`
`>90 ms
`
`6.3%I1.9%
`
`0.2%/0.0%
`
`0.0%/0.0%
`
`7.0%/7.2%
`
`0.5%/0.0%
`
`0.0%I0.0%
`
`416.4/404.2
`
`413.8/405.1
`
`-1.9/1.0
`
`4.9%/ 1.2%
`
`0.0%/0.0%
`
`0.0%/0.0%
`
`3.9%/2.9%
`
`0.2%/0.0%
`
`0.0%/0.0%
`
`416.1/404.7
`
`416.8/405.8
`
`1.6/1.0
`
`6.1%/2.4%
`
`0.9%/0.0%
`
`0.2%/0.0%
`
`6.6%/3.5%
`
`0.5%/0.5%
`
`0.0%/0.0%
`
`DMq-30 = datromerhorphan combined with ultra low-dose quinidine at 30/10mg; DMq-20 = DMq at 20/10mg; QTcB = QT
`interval corrected for heart rate (Bamett’s formula); QTcF = QT interval corrected for heart rate (Fridericia’s formula); ECG =
`electrocardiogram.
`
`November, 2010
`
`Page 7 of 10
`
`699
`
`

`

`ANNALS of Neurology
`
`Discussion
`In this
`large, double blind, placebo controlled study,
`both dosage levels of DMq were significantly superior to
`placebo for
`reducing PBA episode frequency among
`patients with underlying ALS or MS, as assessed by lon
`gitudinal and nonlongitudinal statistical models and also
`by mean change in daily PBA episode rate. At both dos
`age levels, DMq also significantly reduced the severity of
`PBA, as represented by CNS LS score. For reduction in
`episode rate, a prespecified responder analysis showed, at
`both dosage levels, a substantial difference from placebo
`across all degrees of improvement, despite a strong pla
`cebo effect (resembling those seen in previous studies13).
`The differences between DMq and placebo included, at
`both dosage levels, a significantly higher likelihood of
`PBA remission on DMq than on placebo, suggesting that
`for large proportions of patients, the active treatment’s
`amelioration of PBA may be marked.
`Numerically, the responses to the higher DMq dos
`age were more robust than those to the lower dosage in
`several ways, including an earlier improvement in CNS
`LS score, an earlier time to significant difference vs pla
`cebo in number of episode free days, and a slightly
`greater 12 week mean change vs placebo in PBA daily
`episode rate. At the higher dosage, DMq was also associ
`ated with significant
`improvement
`in mental health
`measures, by BDI II and SF 36. Because none of the
`subjects
`in this
`study was clinically depressed, and
`because PBA can result in substantial reduction of quality
`of life,11 this improvement may have been in well being.
`Specific improvements on SF 36 subdomains for social
`functioning and mental health are further evidence that
`the social and psychological disability associated with
`PBA may have been reduced. However, the possibility
`that DMq may have direct antidepressant properties can
`not be excluded, and would require further study. Over
`all, the efficacy reported for DMq containing Q at ultra
`low dosage
`10mg per capsule
`resembled the benefits
`reported by measures including CNS LS scores and PBA
`episode counts in studies of DMq in its original formula
`tion, in which the Q content per capsule was 30mg.13,14
`In the present study, both dosage levels were safe.
`In particular, cardiovascular safety was satisfactory, with
`mild QTc prolongation and no proarrhythmic events.
`Respiratory findings appeared to be consistent with ALS
`progression. However, physicians should always exercise
`caution in managing a patient population that has com
`promised respiratory function. Both dosage levels were
`also well tolerated, with only 13% of DMq recipients
`discontinuing during 12 weeks of double blind treatment.
`The overall discontinuation rate was lower for DMq 30, at
`8%, than for DMq 20, at 18%. In studies of DMQ as
`
`originally formulated, with DM at 30mg per capsule and
`Q at 30mg, the 12 week discontinuation rate had been
`much higher, at 28% in ALS patients13 and 25% in MS
`patients.14 The implication is that the improved tolerability
`demonstrated in the present study may reflect ultra low
`dosing of Q. Usage of dose escalation (with once daily
`dosing during week 1) may also have contributed.
`A body of published evidence suggests that PBA
`may be ameliorated pharmacologically,1 but
`the trials
`assessing current agents, all of which are being utilized
`off label, have limitations. In 1979, dopaminergic treat
`ment, specifically L dopa, was reported to be effective for
`incontinence,’’30 but in a follow up uncon
`‘‘emotional
`trolled study of L dopa or amantadine, only 10 of 25
`recipients responded.31 Since then, reports have centered
`on antidepressants, notably tricyclic agents (eg, amitripty
`line32 or nortriptyline33) and selective serotonin reuptake
`inhibitors (eg, fluoxetine,34 citalopram,35,36 paroxetine,36
`or sertraline37). Overall, the trials have been hampered
`by small size (12 to 28 subjects, among those referenced
`above) and by methodological problems, such as their
`definitions of PBA improvement. Substantial placebo
`effects, as demonstrated in the present
`study, make
`uncontrolled findings all the more difficult to interpret.
`In brief, well controlled data to support current options
`are scarce, and no option is currently approved by the
`US Food and Drug Administration. In addition, antide
`pressants are associated with incompletely elucidated AE
`profiles (including QT interval prolongation19).
`In interpreting the present study’s findings, the trial’s
`limitations should be taken into account. Because its subjects
`were carefully selected, the findings should be generalized to
`a broader
`spectrum of PBA with caution. The study
`required, for instance, a baseline CNS LS score of at least
`13. Accordingly, the effects of DMq 30 or DMq 20 on
`milder forms of PBA are unknown. In addition, the study
`enrolled only patients with underlying ALS or MS. Because
`the pathophysiologic mechanisms causing PBA are probably
`similar regardless of the underlying CNS pathology, DMq
`will likely be effective in reducing symptoms of PBA arising
`in various brain disease or injury states, much in the same
`way that antispastic medication reduces spasticity, irrespective
`of the underlying condition. Even so, additional studies of
`the effect of DMq on PBA in various neurological disorders
`could provide enhanced safety, efficacy, and health outcome
`insight. Hence, further clinical studies of PBA are warranted.
`Nevertheless, the present study represents the largest
`and longest double blind, randomized, placebo controlled
`trial of DMq conducted to date in PBA, and also the
`first to test DMq in PBA patients at ultra low Q dosage.
`Its
`findings expand the clinical evidence that with
`
`700
`
`Volume 68, No. 5
`
`Page 8 of 10
`
`

`

`satisfactory safety and high tolerability, DMq markedly
`reduces PBA frequency and severity, decreasing the con
`dition’s detrimental impact on a patient’s life.
`
`Acknowledgment
`This study was supported by Avanir Pharmaceuticals.
`
`Potential Conflicts of Interest
`E.P.P. has received research support and compensation for
`consulting from Avanir Pharmaceuticals. B.R.B. has
`received compensation for consulting from Avanir Phar
`maceuticals, Bayer Healthcare Pharmaceuticals, Biogen
`Idec, Genentech, and Teva Neuroscience, and has received
`research support
`from Avanir Pharmaceuticals, Biogen
`Idec, National Institutes of Neurological Disorders and
`Stroke Clinical Research Consortium, Novartis, and Teva
`Neuroscience. J.C. has received compensation for consult
`ing from Abbott, Acadia, Accera, ADAMAS, Astellas,
`Avanir Pharmaceuticals, Bristol Myers Squibb, CoMentis,
`Eisai, Elan, EnVivo, Forest, GlaxoSmithKline, Janssen,
`Lilly, Lundbeck, Medivation, Merck, Merz, Myriad,
`Neuren, Novartis, Pfizer, Prana, Schering Plough, Sonexa,
`Takeda, Toyama, and Wyeth, and owns the copyright of
`the Neuropsychiatric Inventory. A.H. and R.K. are
`employees of and own stock options in Avanir Pharma
`ceuticals. R.S. has received compensation for consulting
`from Teva and Avanir Pharmaceuticals and owns 300
`shares of common stock in Johnson & Johnson Corpora
`tion.
`R.A.T.
`is a statistical consultant
`to Avanir
`Pharmaceutical