Journal of Affective Disorders 206 (2016) 151–160

Contents lists available at ScienceDirect

Journal of Affective Disorders

journal homepage: www.elsevier.com/locate/jad

Research paper

Lisdexamfetamine dimesylate augmentation for adults with major

depressive disorder and inadequate response to antidepressant
monotherapy: Results from 2 phase 3, multicenter, randomized,
double-blind, placebo-controlled studies
Cynthia Richards a, Roger S. McIntyre b, Richard Weisler c, Angelo Sambunaris d,
Olga Brawman-Mintzer e, Joseph Gao f, Brooke Geibel a, Matthew Dauphin a,
Manisha Madhoo f,n
a
Formerly of Shire, Lexington, MA, USA
b
University of Toronto, Toronto, ON, Canada
c
Duke University Medical Center, Durham, and the University of North Carolina at Chapel Hill, Chapel Hill, NC, USA
d
Atlanta Institute of Medicine & Research, Atlanta, GA, USA
e
Medical University of South Carolina and Ralph H. Johnson VA Medical Center, Charleston, SC, USA
f
Shire, Lexington, MA, USA

art ic l e i nf o a b s t r a c t

Article history: Background: The efficacy, safety, and tolerability of lisdexamfetamine dimesylate (LDX) augmentation of
Received 9 March 2016 antidepressant monotherapy in adults with major depressive disorder (MDD) from two phase 3 studies
Received in revised form are reported.
14 June 2016
Methods: Across study 1 (placebo, n ¼201; LDX, n ¼201) and study 2 (placebo, n¼ 213; LDX, n ¼211),
Accepted 3 July 2016
Available online 5 July 2016
most participants (placebo and LDX) in the safety analysis set were female (study 1: 66.2% and 64.2%;
study 2: 67.1% and 66.8%); mean7 SD ages were 41.87 12.04 with placebo and 42.2 712.32 with LDX in
Keywords: study 1 and 42.6 711.41 with placebo and 42.0 7 11.63 with LDX in study 2. Participants (18–65 y) had
Augmentation DSM-IV-TR–diagnosed MDD and lead-in baseline Montgomery-Åsberg Depression Rating Scale (MADRS)
Lisdexamfetamine dimesylate total scores Z24. Eight-week antidepressant lead-in phases prospectively assessed antidepressant re-
Major depressive disorder
sponse. Then, 8 weeks of randomized (1:1), double-blind treatment with dose-optimized LDX (20–
Selective serotonin reuptake inhibitors
70 mg) or placebo in participants exhibiting inadequate antidepressant monotherapy responses (aug-
Serotonin-norepinephrine reuptake in-
hibitors
mentation baseline MADRS total scores Z18 and o 50% MADRS total score reductions from lead-in
Amphetamine baseline to augmentation baseline) was initiated. The primary endpoint was MADRS total score change
from augmentation baseline to week 16. Safety and tolerability measures included the occurrence of
treatment-emergent adverse events (TEAEs).
Results: Least squares mean (95% CI) treatment differences (LDX–placebo) for MADRS total score changes
from augmentation baseline to week 16 were not statistically significant in study 1 (0.1 [–1.7, 2.0],
P ¼0.883) or study 2 (–0.5 [–2.3, 1.3], P ¼0.583). The only TEAE reported by 45% of LDX participants at
twice the placebo rate in both studies was dry mouth.
Limitations: Limitations include the exclusion of participants with psychiatric comorbidities/active
medical disorders, the inability to assess specific MDD symptom domains (eg, anhedonia, cognition) or
subtypes, the use of telephone-based depression assessments, and the potential influence of placebo
response.
Conclusion: Contrary to expectations, LDX augmentation was not superior to placebo in reducing

Abbreviations: ABAC-A, Abbreviated Brief Assessment of Cognition in Affective Disorders; ACSA, Amphetamine Cessation Symptom Assessment; ADHD, attention-deficit/
hyperactivity disorder; BMI, body mass index; CGI-I, Clinical Global Impressions–Improvement; CMH, Cochran-Mantel-Haenszel; C-SSRS, Columbia-Suicide Severity Rating
Scale; DBP, diastolic blood pressure; ECG, electrocardiogram; EOS, end of study; FAS, full analysis set; LDX, lisdexamfetamine dimesylate; MADRS, Montgomery-Åsberg
Depression Rating Scale; MAF-GFI, global fatigue index of the Multidimensional Assessment of Fatigue; MDD, major depressive disorder; MMRM, mixed-effects model for
repeated measures; QIDS-SR, Quick Inventory of Depressive Symptomatology–Self-Report; QTcF, Fridericia's corrected QTC interval; SCID-CT, Structured Clinical Interview for
Diagnostic and Statistical Manual for Mental Disorders, Fourth Edition–Text Revision, Clinical Trial Version; SBP, systolic blood pressure; SDS, Sheehan Disability Scale; SNRIs,
serotonin-norepinephrine reuptake inhibitors; SSRIs, selective serotonin reuptake inhibitors; STAR*D, Sequenced Treatment Alternatives to Relieve Depression; TEAEs,
treatment-emergent adverse events
n
Correspondence to: Shire, 300 Shire Way, Lexington, MA 02421.
E-mail address: [email protected] (M. Madhoo).

http://dx.doi.org/10.1016/j.jad.2016.07.006
0165-0327/& 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
152 C. Richards et al. / Journal of Affective Disorders 206 (2016) 151–160

depressive symptoms in individuals with MDD exhibiting inadequate responses to antidepressant

monotherapy.
& 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction 2. Methods

Currently approved treatments for major depressive disorder 2.1. Study design and treatment regimens
(MDD), including selective serotonin reuptake inhibitors (SSRIs)
and serotonin-norepinephrine reuptake inhibitors (SNRIs), effec- Study 1 (ClinicalTrials.gov identifier: NCT01436149) was con-
tively reduce MDD symptoms (Arroll et al., 2005; Cipriani et al., ducted at 76 sites across Canada, Croatia, Mexico, Spain, and the
2009). Nevertheless, more than half of patients with MDD do not United States. Study 2 (ClinicalTrials.gov identifier: NCT01436162)
experience full remission with antidepressant monotherapy or was conducted at 94 sites across the Czech Republic, Estonia,
Finland, Germany, Hungary, Poland, Romania, South Africa, Swe-
with augmentation therapy (Rush et al., 2006; Trivedi et al., 2006a,
den, and the United States. Both studies were conducted in ac-
2006b). For example, in the Sequenced Treatment Alternatives to
cordance with guidelines from the International Conference on
Relieve Depression (STAR*D) study, low remission rates (27.5%
Harmonization Good Clinical Practice and the principles of the
based on the 17-item Hamilton Depression Rating Scale, 32.9%
Declaration of Helsinki. All study protocols were approved by
based on the 16-item Quick Inventory of Depressive Symptoma- ethics committees and regulatory agencies before initiating the
tology–Self-Report [QIDS-SR]) were observed with citalopram studies, and written informed consent was obtained before per-
monotherapy (Trivedi et al., 2006b). In addition, only 29.7% to forming study-related procedures.
39.0% of individuals receiving second-step augmentation with Each study employed a multicenter, randomized, double-blind,
sustained-release bupropion or buspirone achieved remission parallel-group, placebo-controlled, flexible-dose design. With the
(Trivedi et al., 2006a). exception of differences in some of the secondary endpoints, the
Despite the relatively high partial remission rates following studies were identically designed and consisted of 4 phases
antidepressant therapy, there are few medications approved by (Fig. 1): a screening period of 1–4 weeks, an 8-week anti-
the US Food and Drug Administration for use as augmentation depressant lead-in phase (to prospectively confirm inadequate
agents in MDD. Psychostimulants have been investigated for use as response to antidepressant monotherapy), an 8-week double-
augmentation agents in MDD since the 1950s (Robin and Wise- blind treatment phase (3 weeks of dose optimization followed by
berg, 1958). However, in 2 meta-analyses (Candy et al., 2008; 5 weeks of dose maintenance), and a 7- to 9-day follow-up period.
Fleurence et al., 2009), mixed results were reported with respect After screening, eligible participants entered the 8-week sin-
to the efficacy of psychostimulants or eugorics (modafinil) as gle-blind antidepressant lead-in phase, which prospectively as-
sessed response to antidepressant monotherapy and blinded par-
augmentation therapy in MDD.
ticipants to the time of randomization. During the antidepressant
Lisdexamfetamine dimesylate (LDX) is approved in the United
lead-in phase, participants were assigned by investigators to 1 of
States and in other countries for use in patients 6 years and older
4 commercially available antidepressant medications: an SSRI
with attention-deficit/hyperactivity disorder (ADHD) and for use
(escitalopram oxalate [10 or 20 mg] or sertraline hydrochloride
in adults with moderate to severe binge eating disorder only in the
[50–200 mg]) or an SNRI (venlafaxine hydrochloride extended
United States Vyvanse® [lisdexamfetamine dimesylate] 2015. The release [37.5–375 mg] or duloxetine hydrochloride [30–120 mg]).
efficacy of LDX in MDD has been investigated in small, phase 2, Antidepressant assignment was unblinded to both investigators
randomized, double-blind, placebo-controlled studies in in- and participants and was based on clinical factors, including prior
dividuals with MDD who exhibited inadequate response to anti- antidepressant use, response, and tolerability. Each individual in-
depressant monotherapy (Madhoo et al., 2014; Trivedi et al., 2013). vestigator was asked to distribute antidepressant treatments
In one study, LDX augmentation of escitalopram met signal-de- equally at their site, attempting to not assign any 1 antidepressant
tection criteria (prespecified 2-sided significance level α¼0.10) for to 440% of participants. Investigators managed antidepressant
significant reduction in mean Montgomery-Åsberg Depression distribution at the site level and the study sponsor provided on-
Rating Scale (MADRS) total score versus placebo in escitalopram going internal review of the overall distribution. No action from
nonremitters (Trivedi et al., 2013). In a separate study that focused the sponsor was taken to ensure that the protocol-stated anti-
on treating executive dysfunction in individuals with fully or depressant distribution was attained. The initial antidepressant
partially remitted MDD (defined as MADRS total scores r18), LDX dose was taken after the lead-in baseline visit (week 0), with
augmentation of SSRI monotherapy produced significantly greater treatment titrated to the maximum tolerated approved dose by
week 4; dose adjustments were not permitted after week 4.
reductions in MADRS total score than placebo (Madhoo et al.,
Weekly dose increases were made by the investigator according to
2014).
product label guidelines. Antidepressant strength and doses varied
This report presents the findings of two phase 3 studies de-
based on commercial availability and country and were adminis-
signed to assess the efficacy, safety, and tolerability of LDX aug-
tered according to local guidelines. All participants also received
mentation in adults with MDD who exhibit inadequate response
LDX-matched placebo during this phase.
to an 8-week prospective course of antidepressant therapy. The Following the lead-in phase, participants exhibiting an in-
primary objective of each study was to assess the efficacy of LDX adequate response to antidepressant monotherapy entered the
augmentation, as measured by change in MADRS total score. double-blind treatment phase and were randomized (1:1) to LDX
Secondary objectives included assessment of LDX augmentation or placebo. Randomization criteria included having a MADRS total
effects on the Sheehan Disability Scale (SDS; the key secondary score of Z18 at augmentation baseline (week 8), having a MADRS
endpoint), on other secondary efficacy endpoints, and on safety total score reduction of o50% from lead-in baseline to augmen-
and tolerability. tation baseline, exhibiting improvement in depressive symptoms
 C. Richards et al. / Journal of Affective Disorders 206 (2016) 151–160 153

Fig. 1. Study designs for randomized participants (A) and non-randomized participants (B). LDX ¼ lisdexamfetamine; V ¼ visit; Wk¼ week.

from antidepressant lead-in baseline to augmentation baseline 2.2. Study populations

measured by MADRS total score, and having no changes since
lead-in baseline in physical examination, clinical laboratories, Both studies enrolled adult (18 [or age of majority if 418 y by
electrocardiograms (ECGs), or vital signs that preclude LDX treat- local regulation]-65 y at the time of consent) males or non-
ment. Participants whose depressive symptoms improved but who pregnant females. Eligible participants were also required to have
did not meet randomization criteria were not randomized (non- a primary nonpsychotic MDD diagnosis (single or recurrent), as
randomized participants); these participants continued on a defined by the Structured Clinical Interview for Diagnostic and Sta-
tistical Manual for Mental Disorders, Fourth Edition–Text Revision,
modified schedule (visit 10 [week 10], 12 [week 12], and 14 [week
Clinical Trial Version (SCID-CT), that had lasted for Z8 weeks be-
16]) and received single-blind placebo. Participants whose de-
fore screening and to have a MADRS total score of Z24 at lead-in
pressive symptoms did not improve or worsened (based on
baseline.
MADRS total score) on their assigned antidepressant were dis-
Participants were excluded if they had a current MDD episode
continued because LDX was not being assessed for efficacy as and did not sufficiently respond to adequate treatment ( Z6 weeks
monotherapy and continuing a treatment with no demonstrated at the maximum tolerated dose) with Z2 approved anti-
efficacy for the background antidepressant was considered to have depressants or to an approved augmentation therapy; a lifetime
an unfavorable risk/benefit balance. An interactive voice/web re- history of treatment-resistant depression, defined as unresponsive
sponse system was used to generate a randomization number, to adequate treatment (Z8 weeks at the maximum tolerated
stratified by sex and antidepressant type (SSRI vs SNRI), at aug- dose) with Z2 treatments, including distinct classes of approved
mentation baseline. antidepressant monotherapy and adjunctive treatment; had been
The double-blind treatment phase consisted of 3 weeks of dose hospitalized within the past 12 months for the current MDD epi-
optimization (weeks 8–11) followed by 5 weeks of dose main- sode; had current comorbid psychiatric disorders (any significant
tenance (weeks 11–16). Study drug was taken each morning Axis II disorder, ADHD, bipolar disorder, current or lifetime psy-
(7:00 a.m. [ 72 h]); antidepressant treatment schedules were chosis, posttraumatic stress disorder, obsessive compulsive dis-
maintained from the lead-in phase. To maintain blinding, LDX order, pervasive development disorder, anorexia nervosa, bulimia
capsules were over-encapsulated and appeared identical to pla- nervosa) established by the SCID-CT and controlled with pro-
cebo. During dose optimization, treatment was initiated with hibited medications or uncontrolled and associated with sig-
nificant symptoms; had symptoms (eg, agitated states) that con-
30 mg LDX (or placebo). During weeks 9 and 10, the LDX dosage
traindicated LDX treatment or could confound study assessments;
could be maintained, increased to 50 or 70 mg, or down-titrated to
or had a first-degree relative with bipolar I disorder (to ensure
20 mg (Fig. 1). Dosage could be reduced once by a single dose level
depressive symptoms were not related to undiagnosed bipolar
at any time during dose optimization. Once reduced, the dosage
disorder in a participant who had not yet experienced a manic
was maintained for the remainder of the study. During dose episode).
maintenance, the dose being taken at week 11 was maintained Additional exclusion criteria included suspected substance
until week 16 unless there was an emergent safety concern, in abuse or dependence (except nicotine) within the past 6 months;
which case a single dose reduction was permitted. A vital signs considered a suicide risk in the opinion of the investigator, had
assessment was to be taken at the time of this dose reduction. All made a suicide attempt within the past 3 years, or currently de-
participants returned for a safety follow-up visit approximately monstrating active suicidal ideation; a history of symptomatic
7 to 9 days after the last study drug dose. cardiovascular disease or other cardiovascular medical conditions;
154 C. Richards et al. / Journal of Affective Disorders 206 (2016) 151–160

a history of moderate to severe hypertension; systolic blood 2.4. Statistical analyses

pressure (SBP) 4139 mmHg or diastolic blood pressure
(DBP) 489 mmHg or a body mass index (BMI) of o18.5 or 440 at Sample sizes in both studies were based on the primary efficacy
screening or lead-in baseline; current use (within 30 days of endpoint (MADRS total score change from the augmentation
screening) of any centrally acting medication that could affect the baseline to week 16). Assuming a 3-point mean between-group
condition being studied or affect the action, absorption, or dis- difference (SD ¼ 8.1), it was estimated that 155 participants per
position of LDX; or participation in an LDX clinical trial or previous treatment would provide 90% power (2-sided 5% significance le-
use of commercially available LDX. vel). Based on an estimated 20% discontinuation rate during the
double-blind treatment phase, approximately 388 participants
2.3. Study endpoints (194 per treatment) were to be randomized.
MADRS total score change in randomized participants was as-
2.3.1. Efficacy sessed in the full analysis set (FAS: all participants taking Z1
The primary efficacy endpoint was MADRS total score change randomized study drug dose and having Z1 post-augmentation
from augmentation baseline to week 16. On the 10-item, clinician- baseline MADRS assessment). The primary efficacy analysis was
rated MADRS (Montgomery and Asberg, 1979) each item is scored conducted using a mixed-effects model for repeated measures
on a 7-point scale (0–6; higher scores indicate more severe (MMRM) analysis with treatment, visit, the treatment-by-visit
symptoms). The MADRS was assessed at screening, antidepressant interaction, sex, and antidepressant class included as factors, and
lead-in baseline (week 0), week 6, and augmentation baseline augmentation baseline score as a covariate; the interaction of
(week 8) for all participants; and at all double-blind treatment augmentation baseline score with visit was adjusted for in the
visits (weeks 8 through 16) for randomized participants. Although model. Hypothesis testing was performed at the 2-sided 0.05 level
all participants visited the study site for each visit, the MADRS was of significance. Analysis of SDS total score change (the key sec-
completed by a central rater via telephone (MedAvante, Inc; Ha- ondary endpoint) was conducted using the same MMRM model
milton, NJ) who was blinded to study visit and entry criteria. described for the primary efficacy endpoint. The between-treat-
The prespecified key secondary endpoint was SDS total score ment comparisons at week 16 were of main interest for both the
change from augmentation baseline to week 16. The SDS (Sheehan primary and key secondary endpoints.
et al., 1996) measures impairment of work/school, social life/lei- Additional secondary efficacy endpoints were also assessed at
sure activity, and family life/home responsibilities on scales ran- week 16/EOS in the FAS. Percentages of participants demonstrat-
ging from 0 to 10 (higher scores indicate more impairment; total ing MADRS total score reductions from augmentation baseline of
score range: 0–30). The SDS was assessed at antidepressant lead-in 25% or 50% and the percentage of participants achieving MADRS
baseline and augmentation baseline all participants and at all remission (defined by a MADRS total score r10) were assessed
double-blind treatment visits (weeks 8 through 16/end of study using Cochran-Mantel-Haenszel (CMH) tests stratified by sex and
[EOS]) for randomized participants. antidepressant class. Scores on the CGI-I were dichotomized as
Other secondary efficacy endpoints were the Clinical Global improved (very much improved or much improved) or not im-
Impressions-Improvement (CGI-I) scale (Guy, 1976), the QIDS-SR proved (minimally improved through very much worse) and as-
(study 1 only) (Rush et al., 2003), the global fatigue index of the sessed using CMH tests stratified by sex and antidepressant class.
Multidimensional Assessment of Fatigue (MAF-GFI; study 2 only) Changes from augmentation baseline for QIDS-SR total score
scale (Basia, 2014), and the Abbreviated Brief Assessment of Cog- (study 1 only), ABAC-A composite T-scores (study 2 only), and the
nition in Affective Disorders (ABAC-A; study 2 only), also referred MAF-GFI were assessed using an analysis of covariance model that
to as the Brief Assessment of Cognition in Schizophrenia (Kaneda included treatment, sex, and antidepressant class as factors and
and Keefe, 2015). In both studies, CGI-I was assessed at all post- augmentation baseline score as a covariate.
antidepressant lead-in baseline visits. The QIDS-SR (study 1 only), Safety and tolerability were assessed in the safety analysis set,
MAF (study 2 only), and the ABAC-A (study 2 only) were assessed which included all randomized participants taking Z1 rando-
at lead-in baseline, augmentation baseline, and week 16/EOS. mized study drug dose and having Z 1 safety assessment during
double-blind treatment. All safety and tolerability endpoints are
2.3.2. Safety and tolerability presented using descriptive statistics.
Safety and tolerability included examination of adverse events
(AEs), vital signs, clinical laboratory and ECG results, and responses
on the Columbia-Suicide Severity Rating Scale (C-SSRS) and the 3. Results
Amphetamine Cessation Symptom Assessment (ACSA) scale. AEs
were assessed at all visits from the time of informed consent 3.1. Participant disposition and demographics
through follow-up and were categorized based on severity, ser-
iousness, and relatedness to treatment; treatment-emergent AEs Participant disposition and participant demographics and
(TEAEs) were defined as AEs that began or deteriorated on or after clinical characteristics are summarized in Fig. 2 and Table 1, re-
the date of the first randomized treatment dose and no later than spectively. Across studies, most randomized participants were fe-
3 days after the final randomized treatment dose. Clinical labora- male and were white (Table 1). Higher percentages of participants
tory tests were assessed at screening, lead-in and augmentation were assigned to SSRIs than SNRIs (Table 1). The most commonly
baselines, week 12, and week 16/EOS. Vital signs and weight were assigned agent was escitalopram; the distribution of final lead-in
assessed at screening and all study visits through follow-up. antidepressant doses is summarized in Table 1. Mean 7SD MADRS
Twelve-lead ECGs were recorded at screening, lead-in and aug- total scores at antidepressant lead-in baseline and at augmenta-
mentation baselines, and at week 16/EOS. The ACSA, a self-com- tion baseline were comparable between treatment groups in both
pleted scale assessing amphetamine-related withdrawal symp- studies.
toms (McGregor et al., 2008), was completed at follow-up. The
C-SSRS, a semistructured interview measuring the occurrence, 3.2. Prior MDD medication use
severity, and frequency of suicide-related thoughts and behaviors
(Posner et al., 2009), was assessed at screening, all study visits, and Before participation in the current studies, use of an MDD
follow-up. medication was captured in the safety analysis set prior to
 C. Richards et al. / Journal of Affective Disorders 206 (2016) 151–160 155

Fig. 2. Participant disposition. DBP¼ diastolic blood pressure; ECG ¼ electrocardiogram; LDX¼ lisdexamfetamine dimesylate; MADRS ¼ Montgomery-Åsberg Depression
Rating Scale; SBP ¼systolic blood pressure. *Participants could have had multiple reasons for not being randomized to double-blind treatment. †Sustained elevations in
average (of 3 readings) sitting SBP (an increase of Z 10 mmHg from lead-in baseline and an average value Z 140 mmHg on 2 consecutive visits), or sitting DBP (an increase
of Z 10 mmHg from lead-in baseline and an average value Z 90 mmHg on 2 consecutive visits), or pulse rate (an increase of Z 20 bpm from lead-in baseline and an average
value Z100 bpm on 2 consecutive visits).

administration of the first randomized treatment dose (excluding 96.2% (205/213), LDX: 98.1% (207/211)]) had compliance rates
the assigned background antidepressant in the current studies). (defined as [total number of capsules taken  100]/total planned
Prior MDD medication use was reported by 57.0% (229/402) and days of dosing) in the range of 80% to 120%.
64.2% (272/424) of participants in studies 1 and 2, respectively.
The most frequently used MDD medications (Z5% of participants) 3.4. Efficacy
were escitalopram (16.4% [66/402]), sertraline (13.2% [53/402]),
duloxetine (12.4% [50/402]), fluoxetine (11.7% [47/402]), citalo- 3.4.1. Primary endpoint – change in MADRS total score
pram (11.4% [46/402]), bupropion (10.9% [44/402]), paroxetine In both studies, least squares (LS) mean MADRS total score
(9.0% [36/402]), and venlafaxine (7.7% [31/402]) in study 1 and decreases from augmentation baseline to week 16 were observed
citalopram (18.9% [80/424]), sertraline (18.6% [79/424]), venlafax- with placebo and LDX (Fig. 3). As reported in Table 2, the between-
ine (17.7% [75/424]), escitalopram (16.0% [68/424]), fluoxetine treatment difference (LDX – placebo) for mean change from aug-
(11.6% [49/424]), bupropion (8.0% [34/424]), paroxetine (8.0% [34/ mentation baseline to week 16 in MADRS total score in the entire
424]), duloxetine (7.8% [33/424]), and any investigational drug FAS was not statistically significant in either study (both P4 0.05).
(5.9% [25/424]) in study 2. Findings were similar when assessed by sex or antidepressant
class (Supplemental Fig. 1).
3.3. Extent of exposure and treatment compliance
3.4.2. Secondary efficacy endpoints
During double-blind treatment, mean 7SD exposure days for Findings for the SDS (key secondary endpoint) and other sec-
placebo and LDX, respectively, were 49.5 7 15.54 and 50.1 714.76 ondary efficacy endpoints are presented in Table 2. In both studies,
in study 1 and 52.5 712.33 and 51.57 13.17 in study 2. During the there were no statistically significant or clinically meaningful
double-blind treatment periods in studies 1 and 2, respectively, treatment differences for the change from augmentation baseline
74.1% (149/201) and 62.6% (132/211) of participants had their LDX to week 16 between placebo and LDX for SDS total score (Table 2)
dose increased to 50 mg and 39.3% (79/201) and 33.6% (71/211) or for the individual SDS domain scores (data not shown). Across
had their dose increased to 70 mg; 6.0% (12/201) and 9.5% (20/211) most of the other secondary efficacy endpoints (dichotomized CGI-
had their dose down-titrated to 20 mg. The mean 7SD daily LDX I, QIDS-SR, ABAC-A, MADRS remission rate, MADRS 50% response
dose during double-blind treatment was 46.5 713.74 mg in study rate) similar findings were observed, with few nominally statisti-
1 and 43.4 714.35 mg in study 2. Most participants (study 1 cally significant treatment effects being observed between placebo
[placebo: 88.1% (177/201), LDX: 91.5% (184/201)]; study 2 [placebo: and LDX at week 16/EOS (Table 2). However, because the primary
156 C. Richards et al. / Journal of Affective Disorders 206 (2016) 151–160

Table 1
Demographic and baseline clinical characteristics, safety analysis set.

Study 1 Study 2

Placebo (n¼ 201) LDX (n¼ 201) Placebo (n ¼213) LDX (n¼ 211)

Mean 7 SD age, years 41.8 712.04 42.2 7 12.32 42.6 7 11.41 42.0 7 11.63
Sex, n (%)
Male 68 (33.8) 72 (35.8) 70 (32.9) 70 (33.2)
Female 133 (66.2) 129 (64.2) 143 (67.1) 141 (66.8)
Race, n (%)
White 160 (79.6) 154 (76.6) 183 (85.9) 174 (82.5)
Black/African American 36 (17.9) 39 (19.4) 24 (11.3) 34 (16.1)
Native Hawaiian/Pacific Islander 0 0 1 (0.5) 1 (0.5)
Asian 5 (2.5) 5 (2.5) 3 (1.4) 1 (0.5)
American Indian/Alaska Native 0 1 (0.5) 0 0
Other 0 2 (1.0) 2 (0.9) 1 (0.5)
Mean 7 SD weight, kg 82.4 717.90 81.3 718.34 80.8 7 18.52 82.2 7 17.66
Mean 7 SD body mass index, kg/m2 28.9 75.43 28.5 7 5.66 28.1 75.47 28.4 7 5.45
Mean 7 SD MADRS total score
Lead-in baseline 33.27 4.63 33.87 5.01 34.1 75.21 33.47 4.39
Augmentation baseline 25.2 75.03 25.4 7 4.75 25.7 7 5.29 26.0 7 5.15
Mean 7 SD SDS scores at augmentation baselinea
Total score 15.6 7 5.65 15.9 7 5.99 16.8 7 5.75 16.8 7 5.34
Disrupted family life score 5.2 72.05 5.2 7 2.19 5.5 7 2.18 5.6 7 1.98
Disrupted social life score 5.6 72.16 5.7 7 2.09 5.8 7 2.14 5.8 7 1.95
Disrupted work/school score 4.7 72.26 5.0 7 2.36 5.5 7 2.23 5.3 7 2.27
Lead-in baseline antidepressant, n (%)
Venlafaxine HCl extended-release (SNRI) 25 (12.4) 21 (10.4) 52 (24.4) 48 (22.7)
Duloxetine HCl (SNRI) 62 (30.8) 65 (32.3) 46 (21.6) 48 (22.7)
Escitalopram oxalate (SSRI) 67 (33.3) 80 (39.8) 66 (31.0) 63 (29.9)
Sertraline HCl (SSRI) 47 (23.4) 35 (17.4) 49 (23.0) 52 (24.6)
Final antidepressant dose, n (%)
Venlafaxine HCl extended-release (SNRI)
37.5 mg 1 (0.5) 0 – –
75 mg 0 1 (0.5) 7 (3.3) 7 (3.3)
150 mg 3 (1.5) 4 (2.0) 18 (8.5) 17 (8.1)
225 mg 21 (10.4) 16 (8.0) 22 (10.3) 20 (9.5)
300 mg – – 3 (1.4) 3 (1.4)
375 mg – – 2 (0.9) 1 (0.5)
Duloxetine HCl (SNRI)
30 mg 1 (0.5) 2 (1.0) 0 2 (0.9)
40 mg 5 (2.5) 5 (2.5) 1 (0.5) 1 (0.5)
60 mg 48 (23.9) 44 (21.9) 31 (14.6) 37 (17.5)
90 mg 3 (1.5) 6 (3.0) 3 (1.4) 2 (0.9)
120 mg 5 (2.5) 8 (4.0) 11 (5.2) 6 (2.8)
Escitalopram oxalate (SSRI)
10 mg 7 (3.5) 13 (6.5) 12 (5.6) 11 (5.2)
20 mg 60 (29.9) 67 (33.3) 54 (25.4) 52 (24.6)
Sertraline HCl (SSRI)
50 mg 3 (1.5) 2 (1.0) 4 (1.9) 8 (3.8)
100 mg 8 (4.0) 7 (3.5) 13 (6.1) 18 (8.5)
150 mg 14 (7.0) 12 (6.0) 4 (1.9) 15 (7.1)
200 mg 22 (10.9) 14 (7.0) 28 (13.1) 11 (5.2)

LDX¼ lisdexamfetamine dimesylate; MADRS ¼ Montgomery-Åsberg Depression Rating Scale; SDS ¼Sheehan Disability Scale; SNRI ¼ serotonin-norepinephrine reuptake in-
hibitor; SSRI¼ selective serotonin reuptake inhibitor.
a
Based on Full Analysis set: study 1 (total score [placebo, n ¼198; LDX, n ¼198], disrupted family life score [placebo, n ¼199; LDX, n ¼199], disrupted social life score
[placebo, n ¼ 199; LDX, n ¼ 198], disrupted work/school score [placebo, n ¼ 198; LDX, n¼ 199]); study 2 (placebo, n ¼213; LDX, n ¼ 209).

endpoint was not statistically significant, any nominal differences 1 (syncope in an LDX participant; suicide attempt in a placebo
in secondary endpoints should not be considered statistically participant; both resolved) and 2 cases in study 2 (anxiety in an
significant. LDX participant; suicidal ideation in a placebo participant). The
most frequently occurring TEAEs are summarized in Table 3. The
3.5. Safety and tolerability TEAEs reported by Z5% of participants treated with LDX and at
twice the rate of placebo were dry mouth and nasopharyngitis in
3.5.1. Adverse events study 1 and dry mouth and decreased appetite in study 2. There
Treatment-emergent AE (TEAEs) are summarized in Table 3. In were no instances of psychosis/mania or aggression reported as
both studies, higher percentages of TEAEs were reported with LDX TEAEs or deaths in either study.
than with placebo. Most TEAEs were mild or moderate in in- Few serious TEAEs were reported in either study (Table 3). In
tensity; the frequency of severe TEAEs was low in both studies (see study 1, there were 3 serious TEAEs in the LDX group (nonsuicidal
footnote to Table 3 for a complete listing). TEAEs leading to dis- overdose of diphenhydramine, syncope, and appendicitis; none
continuation were generally of mild to moderate intensity in both were considered treatment related by the investigator) and 7 ser-
studies (see footnote to Table 3 for a complete listing of TEAEs ious TEAEs in the placebo group (suicide attempt, major depres-
leading to discontinuation), with the exception of 2 cases in study sion, vertigo, syncope, appendectomy, laceration, and skin
 C. Richards et al. / Journal of Affective Disorders 206 (2016) 151–160 157

treated participants (0.9%) in study 2 had a Z 7% increase in body

weight from the augmentation baseline; no LDX participants met
this criterion in either study. There were no differences of clinical
concern between treatment groups regarding clinical laboratory
findings in either study.
On the C-SSRS, Z1 positive suicidal ideation occurred during
double-blind treatment in 14 placebo (7.0%) and 14 LDX (7.0%)
participants in study 1 and in 20 placebo (9.4%) and 19 LDX (9.0%)
participants in study 2. At least 1 suicide attempt was reported on
the C-SSRS during double-blind treatment in 1 (0.5%) placebo
participant in each study; the suicide attempt in study 2 was also
recorded as a TEAE of mild intensity.
At follow-up, mean 7SD ACSA total scores were 15.1 710.71
with placebo (n ¼178) and 14.7 710.94 with LDX (n ¼183) in study
1 and 17.2710.56 with placebo (n¼ 199) and 17.0 710.80 (n¼ 194)
with LDX in study 2.

4. Discussion

The primary finding of the current studies was that augmen-

tation of antidepressant therapy with LDX was not associated with
significantly different reductions in MADRS total score (the pri-
mary efficacy endpoint) than placebo augmentation in adults with
inadequate responses to 8 weeks of standard antidepressant
therapy. Similar findings were observed for the prespecified key
secondary endpoint (SDS total score) and the other secondary ef-
ficacy endpoints.
These findings do not support previously published findings of
LDX augmentation for MDD based on 2 small phase 2 studies
(Madhoo et al., 2014; Trivedi et al., 2013), but they are generally
consistent with published findings for psychostimulant (Patkar
Fig. 3. Change in MADRS total score during double-blind treatment (full analysis et al., 2006; Postolache et al., 1999; Ravindran et al., 2008) and
set) for Study 1 (A) and Study 2 (B). LDX¼ lisdexamfetamine; dimesylate modafinil augmentation studies for MDD (DeBattista et al., 2003;
MADRS ¼ Montgomery-Åsberg Depression Rating Scale. Dunlop et al., 2007; Fava et al., 2007) and with the conclusions of
published meta-analyses (Candy et al., 2008; Fleurence et al.,
infection). The suicide attempt and case of major depression were 2009). Discrepancies between the current phase 3 studies and
considered to be treatment related by the investigator. In study 2, previously published phase 2 studies might be attributed to study
the 2 serious TEAEs (suicidal ideation in a placebo participant and design differences. The phase 3 studies required that participants
accelerated hypertension in an LDX participant) were both con- have MADRS total scores of Z18 at augmentation baseline and
sidered related to treatment; the suicidal ideation event resulted MADRS total score reductions of o50% from the antidepressant
in study withdrawal but the case of accelerated hypertension did lead-in baseline to augmentation baseline, but the phase 2 studies
not. included participants with 17-item Hamilton Depression Rating
Scale scores Z4 (Trivedi et al., 2013) or with executive dysfunction
3.5.2. Vital signs and electrocardiogram and MADRS total scores r 18 (Madhoo et al., 2014). Second, both
At week 16/EOS, mean SBP and DBP decreased with placebo phase 2 studies included smaller study populations and used only
and increased with LDX in study 1. In study 2, increases in SBP and SSRIs as the antidepressant medication (Madhoo et al., 2014; Tri-
DBP at week 16/EOS were numerically larger with LDX than with vedi et al., 2013). It has been suggested that using small, under-
placebo (Table 3). In both studies, mean increases in pulse with powered, proof-of-concept studies in the design of larger, rando-
LDX were numerically larger than those observed for placebo mized clinical trials may be problematic (Kraemer and Kupfer,
(Table 3). The proportion of participants with potentially clinically 2006). Despite these potential explanations, it is important to note
important changes in SBP, DBP, and pulse was generally higher that the phase 3 studies described in this report were more rig-
with LDX than with placebo (Table 3). Based on the ECG, mean orously designed and better powered than the published phase
heart rate increases and QT duration decreases were numerically 2 studies (Madhoo et al., 2014; Trivedi et al., 2013). As such, these
larger with LDX than with placebo in both studies (Table 3). Fri- findings indicate that LDX was not superior to placebo in reducing
dericia-corrected QT interval decreases were comparable with the depressive symptoms of MDD.
placebo and LDX in study 1, but were numerically larger with LDX The lack of efficacy observed in the current study may indicate
in study 2. that psychostimulants as a class are ineffective in treating un-
differentiated residual depressive symptoms in individuals who
3.5.3. Other safety and tolerability endpoints exhibit an inadequate response to antidepressant monotherapy.
Mean weight and BMI increased with placebo and decreased Alternatively, it is possible that clinician-rated scales, such as
with LDX in both studies (Table 3). Six LDX-treated participants MADRS, do not capture the true effects of psychostimulants or that
(3.0%) in study 1 and 8 LDX-treated participants (3.8%) in study psychostimulants do not provide multidimensional symptom re-
2 exhibited Z 7% decreases in body weight from the augmentation lief. Rather, they may provide dimensional efficacy for individuals
baseline; no placebo participants met this criterion in either study. with symptoms related to anhedonia, fatigue, or cognitive dys-
Four placebo-treated participants (2.0%) in study 1 and 2 placebo- function. The possibility that the failure to observe an effect of LDX
158 C. Richards et al. / Journal of Affective Disorders 206 (2016) 151–160

Table 2
Efficacy assessments at end of double-blind treatment, full analysis set.

Study 1 Study 2

Placebo LDX Placebo LDX

MADRS total score (primary endpoint)a

Mean 7 SD total score at augmentation baseline 25.2 7 5.03 25.4 7 4.77 25.7 7 5.29 26.0 7 5.16
Mean 7 SD total score at week 16 18.9 7 9.99 19.17 9.10 18.9 78.85 18.3 79.58
LS mean (95% CI) change from augmentation baseline at week 16  6.3 (  7.6,  4.9)  6.1 (  7.5,  4.8)  6.8 (  8.1,  5.5)  7.3 (  8.6,  6.0)
LS mean (95% CI) treatment difference at week 16 0.1 (  1.7, 2.0); P o0.883  0.5 (  2.3, 1.3); P o0.583
SDS total score (key secondary endpoint)b
Mean 7 SD total score at augmentation baseline 15.6 7 5.65 15.9 75.99 16.8 7 5.75 16.8 7 5.34
Mean 7 SD total score at week 16 11.5 7 7.30 11.0 7 6.86 12.17 7.03 11.4 7 7.23
LS mean (95% CI) change from augmentation baseline at week 16  4.3 (  5.3,  3.3)  4.7 (  5.6,  3.7)  4.3 (  5.2,  3.4)  4.9 (  5.8,  4.0)
LS mean (95% CI) treatment difference at week 16  0.4 (  1.8, 1.0); P¼ 0.576  0.6 (  1.9, 0.7); P¼ 0.354
MAF-GFI (secondary endpoint)c
Mean 7 SD total score at augmentation baseline – – 32.3 7 9.36 32.6 7 8.77
Mean 7 SD total score at week 16 – – 27.8 7 11.78 25.6 7 11.16
LS mean (95% CI) change from augmentation baseline at week 16 – –  4.4 (  5.8,  3.0)  6.6 (  8.0,  5.1)
LS mean (95% CI) treatment difference at week 16 –  2.2 (  4.1,  0.2); nominal P¼ 0.031
QIDS-SR total score (secondary endpoint)d
Mean 7 SD total score at augmentation baseline 11.9 7 3.74 11.8 7 3.97 – –
Mean 7 SD total score at week 16 9.6 7 4.66 9.6 7 4.79 – –
LS mean (95% CI) change from augmentation baseline at week 16  2.6 (  3.2,  1.9)  2.3 (  2.9,  1.7) – –
LS mean (95% CI) treatment difference at week 16 0.3 (  0.6, 1.1); nominal P¼ 0.500 –
ABAC-A T-composite score (secondary endpoint)e
Mean 7 SD total score at augmentation baseline – – 46.1 713.18 45.0 7 13.35
Mean 7 SD total score at week 16 – – 48.27 12.96 47.7 712.70
LS mean (95% CI) change from augmentation baseline at week 16 – – 2.5 (1.5, 3.5) 3.0 (2.1, 4.0)
LS mean (95% CI) treatment difference at week 16 – 0.5 (  0.8, 1.9); nominal P ¼0.435
25% MADRS response at week 16/EOS (secondary endpoint)
n/N (%) 130/200 (65.0) 149/200 (74.5) 158/213 (74.2) 144/209 (68.9)
Nominal P value 0.039 0.234
50% MADRS response at week 16/EOS (secondary endpoint)
n/N (%) 77/200 (38.5) 82/200 (41.0) 79/213 (37.1) 87/209 (41.6)
Nominal P value 0.591 0.342
MADRS remission at week 16/EOS (secondary endpoint)
n/N (%) 45/200 (22.5) 37/200 (18.5) 38/213 (17.8) 48/209 (23.0)
Nominal P value 0.335 0.195
CGI-I Improved at week 16/EOS (secondary endpoint)
n/N (%) 106/199 (53.3) 110/199 (55.3) 114/213 (53.5) 119/209 (56.9)
Nominal P value 0.729 0.486

ABAC-A ¼ Abbreviated Brief Assessment of Cognition in Affective Disorders; CGI-I ¼Clinical Global Impressions-Improvement; EOS ¼ end of study; MADRS ¼ Montgomery-
Åsberg Depression Rating Scale; MAF-GFI=global fatigue index of the Multidimensional Assessment of Fatigue; QIDS-SR=Quick Inventory of Depressive Symptomatology–
Self-Report; SDS ¼ Sheehan Disability Scale.
a
MADRS sample sizes: augmentation baseline (study 1 [placebo (n ¼200), LDX (n¼ 200)]; study 2 [placebo (n¼ 213), LDX (n ¼ 209)]); week 16, LS mean change from
augmentation baseline, and LS mean treatment difference at week 16 (study 1 [placebo (n ¼165), LDX (n ¼ 163)]; study 2 [placebo (n ¼189), LDX (n ¼181)]).
b
SDS sample sizes: augmentation baseline (study 1 [placebo (n¼ 198), LDX (n ¼198)]; study 2 [placebo (n ¼ 213), LDX (n ¼209)]); week 16 (study 1 [placebo (n¼ 165),
LDX (n ¼163)]; study 2 [placebo (n¼ 189), LDX (n ¼ 180)]); LS mean change from augmentation baseline and LS mean treatment difference at week 16 (study 1 [placebo
(n¼ 164), LDX (n ¼ 162)]), study 2 (placebo [n¼ 189], LDX [n ¼180]).
c
MAF-GFI sample sizes: augmentation baseline (study 2 [placebo (n¼ 212), LDX (n ¼209)]); week 16, LS mean change from augmentation baseline and LS mean
treatment difference at week 16 (study 2 [placebo (n ¼205), LDX (n¼ 197)]).
d
QIDS-SR sample sizes: augmentation baseline (study 1 [placebo (n¼ 197), LDX (n¼ 199)]); week 16 (study 1 [placebo (n ¼ 189), LDX (n¼ 186)]); LS mean change from
augmentation baseline and LS mean treatment difference at week 16 (study 1 [placebo (n ¼186), LDX (n ¼185)]).
e
ABAC-A sample sizes: augmentation baseline (study 2 [placebo (n ¼213), LDX (n ¼209)]); week 16, LS mean change from augmentation baseline and LS mean
treatment difference at week 16 (study 2 [placebo (n ¼198), LDX (n ¼ 194)]).

was related to the use of the MADRS seems unlikely because of the are needed to understand the nuanced effects of psychostimulants
consistent lack of effect observed across all study assessments, and to further delineate the biological differences between
including the QIDS-SR. In support of the latter hypothesis, me- symptoms of apathy and mood enhancement in individuals with
thylphenidate has been reported to be efficacious in reducing MDD.
apathy in individuals with Alzheimer's disease (Rosenberg et al., The overall safety and tolerability findings observed in these
2013) and apathy and fatigue in individuals with MDD (Ravindran studies are consistent with previously published studies of LDX
et al., 2008). Additionally, a previously published study reported augmentation for MDD with other clinical studies of LDX for which
LDX augmentation significantly improved self-reported and in- LDX is approved for use (Adler et al., 2008; Madhoo et al., 2014;
formant-reported executive dysfunction, as measured by the Be- McElroy et al., 2016; Trivedi et al., 2013 Wigal et al., 2010). In-
havior Rating Inventory of Executive Function-Adult Version, in somnia, dry mouth, headache, decreased appetite, and nasophar-
participants with mild MDD (Madhoo et al., 2014). However, given yngitis were among the most frequently reported TEAEs with LDX.
the lack of effect of LDX on the ABAC-A in the current studies, this LDX was associated with greater mean increases in pulse rate and
finding also needs to be cautiously considered. Nevertheless, the blood pressure than placebo and with decreases in weight and
possibility that psychostimulants may effectively reduce apathy, BMI.
inattention, or indifference (symptoms that may not be effectively These findings should be considered in light of several limita-
captured by the MADRS or ABAC-A) remains. Additional studies tions. First, these studies used remote telephone assessment of the
 C. Richards et al. / Journal of Affective Disorders 206 (2016) 151–160 159

Table 3
Summary of safety and tolerability during double-blind treatment, safety analysis set.

Study 1 Study 2

Placebo (n¼ 201) LDX (n¼ 201) Placebo (n ¼213) LDX (n¼ 211)

Any TEAE, n (%) 118 (58.7) 131 (65.2) Any TEAE, n (%) 108 (50.7) 139 (65.9)
Serious TEAEs 5 (2.5) 3 (1.5) Serious TEAEs 1 (0.5) 1 (0.5)
Severe TEAEsa 6 (3.0) 4 (2.0) Severe TEAEs a
3 (1.4) 7 (3.3)
Treatment-related TEAEs 56 (27.9) 75 (37.3) Treatment-related TEAEs 43 (20.2) 85 (40.3)
TEAEs leading to discontinuationb 7 (3.5) 8 (4.0) TEAEs leading to 1 (0.5) 2 (0.9)
discontinuationb
TEAEs reported by 45% of participants in either treatment group TEAEs reported by 45% of participants in either treatment group
Insomnia 15 (7.5) 19 (9.5) Headache 16 (7.5) 25 (11.8)
Dry mouth 6 (3.0) 19 (9.5) Dry mouth 6 (2.8) 25 (11.8)
Decreased appetite 8 (4.0) 15 (7.5) Nasopharyngitis 19 (8.9) 14 (6.6)
Headache 21 (10.4) 13 (6.5) Decreased appetite 5 (2.3) 13 (6.2)
Nausea 10 (5.0) 13 (6.5) Insomnia 7 (3.3) 11 (5.2)
Nasopharyngitis 4 (2.0) 11 (5.5) Hyperhidrosis 1 (0.5) 11 (5.2)
Dizziness 10 (5.0) 9 (4.5)
Changes (mean 7 SD) from augmentation baseline at week 16/EOS Changes (mean 7 SD) from augmentation baseline at week 16/EOS
c
Diastolic blood pressure, mmHg  0.4 77.69 1.5 78.60 Diastolic blood pressure, mmHgc 0.1 76.55 2.6 7 7.17
c
Systolic blood pressure, mmHg  0.4 710.04 1.2 79.70 Systolic blood pressure, mmHgc 0.3 7 8.39 2.6 7 9.79
c c
Pulse, bpm 0.7 7 9.75 5.9 7 10.64 Pulse, bpm 0.5 7 8.31 5.2 7 10.34
ECG heart rate, bpmd 0.3 7 8.84 6.17 10.29 ECG heart rate, bpmd 0.5 7 8.01 5.2 7 10.67
ECG QT duration, msd  1.8 7 21.67  11.4 7 23.57 ECG QT duration, msd  1.6 7 19.80  10.7 7 25.23
ECG QTcF, msd  0.8 711.30  0.8 7 12.80 ECG QTcF, msd 0.0 7 12.27  1.7 7 13.03
Weight, kgc 0.3 7 1.90  1.0 7 2.16 Weight, kgc 0.2 7 1.64  1.0 7 2.21
2c 2c
BMI, kg/m 0.17 0.67  0.4 7 0.77 BMI, kg/m 0.1 70.58  0.3 70.77
Potentially clinically important vital signs at 2 consecutive visits (of which 1 was the Potentially clinically important vital signs at 2 consecutive visits (of which 1 was
last visit)c the last visit)c
Systolic blood pressure 2 (1.0) 0 Systolic blood pressure 0 3 (1.4)
Z 140 mmHgc Z 140 mmHgc
Diastolic blood pressure 2 (1.0) 0 Diastolic blood pressure 0 5 (2.4)
Z 90 mmHgc Z 90 mmHgc
c c
Pulse Z100 bpm 0 0 Pulse Z 100 bpm 0 3 (1.4)

BMI¼ body mass index; ECG ¼electrocardiogram; EOS ¼end of study; QTcF ¼Fridericia's corrected QTC interval; TEAE ¼treatment-emergent adverse event.
a
Severe TEAEs: study 1 (placebo: gingivitis, wound infection, laceration, headache, insomnia, suicide attempt [n¼ 1 for all]; LDX: fatigue, appendicitis, migraine, syncope,
apathy [n ¼1 for all]); study 2 (placebo: seasonal allergy, sleep disorder, suicidal ideation [n¼ 1 for all]; LDX: dental caries, dry mouth, dyspepsia, nasopharyngitis, anxiety,
insomnia, restlessness, hot flush [n¼ 1 for all]).
b
TEAEs leading to discontinuation: study 1 (placebo: diarrhea, disturbance in attention, liver function test abnormal, syncope, suicide attempt, major depression, and
rash [n¼ 1 for all]; LDX: accidental overdose, sedation, systolic blood pressure increased, liver function test abnormal, musculoskeletal pain, pain in extremity, syncope, self-
injurious behavior, and palpitations [n¼ 1 for all]); study 2 (placebo: suicidal ideation [n¼ 1]; LDX: headache and anxiety [both n ¼1]).
c
Study 1 (placebo, n ¼ 200; LDX, n¼ 201); study 2 (placebo, n ¼ 213; LDX, n ¼ 209).
d
Study 1 (placebo, n ¼ 189; LDX, n ¼ 189); study 2 (placebo, n ¼202; LDX, n ¼ 196).

primary efficacy endpoint by a small pool of calibrated raters to subtypes. It should also be noted that the treatment duration in
reduce the overall number of raters and the potential for varia- the current studies was relatively short. A longer duration anti-
bility across raters. As inter-rater reliability was not assessed in the depressant lead-in phase may have increased the number of par-
current studies, it is a limitation that variability across raters is ticipants exhibiting antidepressant responses, and a longer dura-
unknown. The use of telephone measures is a method now being tion double-blind treatment phase may have increased the mag-
used in clinical trials to reduce bias (Kobak et al., 2008). Even
nitude of LDX responses. Finally, clinical trials for MDD are sus-
though it may offer less specificity and more variability than a
ceptible to having high placebo responses (Walsh et al., 2002),
face-to-face clinical assessment, due in part to the inability to vi-
which could mask a potential therapeutic effect.
sually observe the participant, it is important to note that the
In conclusion, in the 2 largest, randomized clinical trials of
validity of using standardized central rating via telephone corre-
lates well with face-to-face interviews (Kobak et al., 2008) and stimulant augmentation of antidepressant therapy completed to
that site- and participant-rated secondary assessments were con- date, LDX was not superior to placebo in reducing the depressive
sistent with the primary efficacy assessment. In addition, as al- symptoms of MDD in individuals with inadequate responses to
luded to above, it is possible that psychostimulants provide di- standard antidepressant monotherapy. Although these studies did
mensional efficacy and that assessment of specific MADRS items or not show a treatment difference, the potential exists that stimu-
clusters may have revealed effects of LDX that were not apparent lant augmentation may be appropriate in subgroups of individuals
in the total score. However, a detailed assessment of MADRS do- with MDD (eg, treatment-resistant MDD) or with a different sti-
mains was not performed because the primary and key secondary mulant augmentation.
endpoints consistently failed in both studies, making such as-
sessments statistically inappropriate and unreliable. Also, partici-
pants in the current studies lacked diversity in terms of race (most
were White) and gender (most were female). Those with psy- Appendix A. Supporting information
chiatric comorbidities (including ADHD) and/or active medical
disorders were excluded and executive functioning was not fully Supplementary data associated with this article can be found in
examined in these studies. Therefore, these findings are limited to the online version at http://dx.doi:10.1016/j.jad.2016.07.006.
a specific MDD population and do not extend to potential MDD
160 C. Richards et al. / Journal of Affective Disorders 206 (2016) 151–160

References sensitive to change. Br. J. Psychiatry 134, 382–389.

Patkar, A.A., Masand, P.S., Pae, C.U., Peindl, K., Hooper-Wood, C., Mannelli, P., Cic-
cone, P., 2006. A randomized, double-blind, placebo-controlled trial of aug-
Adler, L.A., Goodman, D.W., Kollins, S.H., Weisler, R.H., Krishnan, S., Zhang, Y., Bie- mentation with an extended release formulation of methylphenidate in out-
derman, J., 303 Study Group, 2008. Double-blind, placebo-controlled study of patients with treatment-resistant depression. J. Clin. Psychopharmacol. 26,
the efficacy and safety of lisdexamfetamine dimesylate in adults with atten- 653–656.
tion-deficit/hyperactivity disorder. J. Clin. Psychiatry 69, 1364–1373. Posner, K., Brent, D., Lucas, C., Gould, M., Stanley, B., Brown, G., Fisher, P., Zelazny, J.,
Arroll, B., Macgillivray, S., Ogston, S., Reid, I., Sullivan, F., Williams, B., Crombie, I., Burke, A., Oquendo, M., Mann, J., 2009. Columbia-Suicide Severity Rating Scale.
2005. Efficacy and tolerability of tricyclic antidepressants and SSRIs compared
[Online]. Available from: 〈http://www.cssrs.columbia.edu/scale_versions.html〉
with placebo for treatment of depression in primary care: a meta-analysis. Ann.
(cited January 23, 2015).
Fam. Med. 3, 449–456.
Postolache, T.T., Rosenthal, R.N., Hellerstein, D.J., Aromin, R., Kelton, G.M., Muran, J.
Basia, B., Multidimensional Assessment of Fatigue. [Online]. Available from: 〈http://
C., Londono, J.H., 1999. Early augmentation of sertraline with methylphenidate.
www.son.washington.edu/research/maf/〉 (cited January 9, 2014).
J. Clin. Psychiatry 60, 123–124.
Candy, M., Jones, L., Williams, R., Tookman, A., King, M., 2008. Psychostimulants for
Ravindran, A.V., Kennedy, S.H., O’Donovan, M.C., Fallu, A., Camacho, F., Binder, C.E.,
depression. Cochrane Database Syst. Rev. CD006722.
2008. Osmotic-release oral system methylphenidate augmentation of anti-
Cipriani, A., Furukawa, T.A., Salanti, G., Geddes, J.R., Higgins, J.P., Churchill, R., Wa-
depressant monotherapy in major depressive disorder: results of a double-
tanabe, N., Nakagawa, A., Omori, I.M., McGuire, H., Tansella, M., Barbui, C., 2009.
blind, randomized, placebo-controlled trial. J. Clin. Psychiatry 69, 87-–94.
Comparative efficacy and acceptability of 12 new-generation antidepressants: a
Robin, A.A., Wiseberg, S., 1958. A controlled trial of methyl phenidate (ritalin) in the
multiple-treatments meta-analysis. Lancet 373, 746–758.
treatment of depressive states. J. Neurol. Neurosurg. Psychiatry 21, 55–57.
DeBattista, C., Doghramji, K., Menza, M.A., Rosenthal, M.H., Fieve, R.R., Modafinil in
Rosenberg, P.B., Lanctot, K.L., Drye, L.T., Herrmann, N., Scherer, R.W., Bachman, D.L.,
Depression Study Group, 2003. Adjunct modafinil for the short-term treatment
Mintzer, J.E., ADMET Investigators, 2013. Safety and efficacy of methylphenidate
of fatigue and sleepiness in patients with major depressive disorder: a pre-
for apathy in Alzheimer’s disease: a randomized, placebo-controlled trial. J.
liminary double-blind, placebo-controlled study. J. Clin. Psychiatry 64,
Clin. Psychiatry 74, 810–816.
1057–1064.
Rush, A.J., Trivedi, M.H., Ibrahim, H.M., Carmody, T.J., Arnow, B., Klein, D.N., Mar-
Dunlop, B.W., Crits-Christoph, P., Evans, D.L., Hirschowitz, J., Solvason, H.B., Rickels,
K., Garlow, S.J., Gallop, R.J., Ninan, P.T., 2007. Coadministration of modafinil and kowitz, J.C., Ninan, P.T., Kornstein, S., Manber, R., Thase, M.E., Kocsis, J.H., Keller,
a selective serotonin reuptake inhibitor from the initiation of treatment of M.B., 2003. The 16-item Quick Inventory of Depressive Symptomatology (QIDS),
major depressive disorder with fatigue and sleepiness: a double-blind, placebo- clinician rating (QIDS-C), and self-report (QIDS-SR): a psychometric evaluation
controlled study. J. Clin. Psychopharmacol. 27, 614–619. in patients with chronic major depression. Biol. Psychiatry 54, 573–583.
Fava, M., Thase, M.E., DeBattista, C., Doghramji, K., Arora, S., Hughes, R.J., 2007. Rush, A.J., Trivedi, M.H., Wisniewski, S.R., Nierenberg, A.A., Stewart, J.W., Warden,
Modafinil augmentation of selective serotonin reuptake inhibitor therapy in D., Niederehe, G., Thase, M.E., Lavori, P.W., Lebowitz, B.D., McGrath, P.J., Ro-
MDD partial responders with persistent fatigue and sleepiness. Ann. Clin. senbaum, J.F., Sackeim, H.A., Kupfer, D.J., Luther, J., Fava, M., 2006. Acute and
Psychiatry 19, 153–159. longer-term outcomes in depressed outpatients requiring one or several
Fleurence, R., Williamson, R., Jing, Y., Kim, E., Tran, Q.V., Pikalov, A.S., Thase, M.E., treatment steps: a STAR*D report. Am. J. Psychiatry 163, 1905–1917.
2009. A systematic review of augmentation strategies for patients with major Sheehan, D.V., Harnett-Sheehan, K., Raj, B.A., 1996. The measurement of disability.
depressive disorder. Psychopharmacol. Bull. 42, 57–90. Int. Clin. Psychopharmacol. 11, 89–95.
Guy, W., 1976. Clinical Global Impressions. US Department of Health, Education, and Trivedi, M.H., Cutler, A.J., Richards, C., Lasser, R., Geibel, B.B., Gao, J., Sambunaris, A.,
Welfare. Public Health Service, Alcohol, Drug Abuse and Mental Health Ad- Patkar, A.A., 2013. A randomized controlled trial of the efficacy and safety of
ministration. NIMH Psychopharmacology Research Branch, Rockville, MD. lisdexamfetamine dimesylate as augmentation therapy in adults with residual
Kaneda, Y., Keefe, R.S., 2015. An abbreviated version of the brief assessment of symptoms of major depressive disorder after treatment with escitalopram. J.
cognition in schizophrenia (BACS). Eur. J. Psychiatry 29, 131–134. Clin. Psychiatry 74, 802–809.
Kobak, K.A., Williams, J.B., Jeglic, E., Salvucci, D., Sharp, I.R., 2008. Face-to-face Trivedi, M.H., Fava, M., Wisniewski, S.R., Thase, M.E., Quitkin, F., Warden, D., Ritz, L.,
versus remote administration of the Montgomery-Asberg Depression Rating Nierenberg, A.A., Lebowitz, B.D., Biggs, M.M., Luther, J.F., Shores-Wilson, K.,
Scale using videoconference and telephone. Depress. Anxiety 25, 913–919. Rush, A.J., 2006a. Medication augmentation after the failure of SSRIs for de-
Kraemer, H.C., Kupfer, D.J., 2006. Size of treatment effects and their importance to pression. N. Engl. J. Med. 354, 1243–1252.
clinical research and practice. Biol. Psychiatry 59, 990–996. Trivedi, M.H., Rush, A.J., Wisniewski, S.R., Nierenberg, A.A., Warden, D., Ritz, L.,
Madhoo, M., Keefe, R.S., Roth, R.M., Sambunaris, A., Wu, J., Trivedi, M.H., Anderson, Norquist, G., Howland, R.H., Lebowitz, B., McGrath, P.J., Shores-Wilson, K., Biggs,
C.S., Lasser, R., 2014. Lisdexamfetamine dimesylate augmentation in adults with M.M., Balasubramani, G.K., Fava, M., 2006b. Evaluation of outcomes with cita-
persistent executive dysfunction after partial or full remission of major de- lopram for depression using measurement-based care in STAR*D: implications
pressive disorder. Neuropsychopharmacology 39, 1388–1398. for clinical practice. Am. J. Psychiatry 163, 28–40.
McElroy, S.L., Hudson, J.I., Ferreira-Cornwell, M.C., Radewonuk, J., Whitaker, T., Vyvanses (lisdexamfetamine dimesylate), 2015. Shire US Inc. Wayne, PA.
Gasior, M., 2016. Lisdexamfetamine dimesylate for adults with moderate to Walsh, B.T., Seidman, S.N., Sysko, R., Gould, M., 2002. Placebo response in studies of
severe binge eating disorder: results of two pivotal phase 3 randomized con- major depression: variable, substantial, and growing. JAMA 287, 1840–1847.
trolled trials. Neuropsychopharmacology 41, 1251–1260 (In press). Wigal, T., Brams, M., Gasior, M., Gao, J., Squires, L., Giblin, J., for the 316 Study Group,
McGregor, C., Srisurapanont, M., Mitchell, A., Longo, M.C., Cahill, S., White, J.M., 2010. Randomized, double-blind, placebo-controlled, crossover study of the
2008. Psychometric evaluation of the amphetamine cessation symptom as- efficacy and safety of lisdexamfetamine dimesylate in adults with attention-
sessment. J. Subst. Abus. Treat. 34, 443–449. deficit/hyperactivity disorder: novel findings using a simulated adult workplace
Montgomery, S.A., Asberg, M., 1979. A new depression scale designed to be environment design. Behav. Brain Funct. 6, 34.