Expert Opinion on Pharmacotherapy
`
`ISSN: 1465-6566 (Print) 1744-7666 (Online) Journal homepage: http://www.tandfonline.com/loi/ieop20
`
`A review of the clinical efficacy, safety and
`tolerability of the antidepressants vilazodone,
`levomilnacipran and vortioxetine
`
`William James Deardorff & George T Grossberg
`
`To cite this article: William James Deardorff & George T Grossberg (2014) A review of the clinical
`efficacy, safety and tolerability of the antidepressants vilazodone, levomilnacipran and vortioxetine,
`Expert Opinion on Pharmacotherapy, 15:17, 2525-2542, DOI: 10.1517/14656566.2014.960842
`To link to this article: https://doi.org/10.1517/14656566.2014.960842
`
`Published online: 16 Sep 2014.
`
`Submit your article to this journal
`
`Article views: 938
`
`View related articles
`
`View Crossmark data
`
`Citing articles: 16 View citing articles
`
`Full Terms & Conditions of access and use can be found at
`http://www.tandfonline.com/action/journalInformation?journalCode=ieop20
`
`Argentum EX1037
`
`Download by: [Mount Sinai Health System Libraries]
`
`Date: 05 December 2017, At: 10:11
`
`Page 1
`
`

`

`1.
`
`Introduction
`
`2. New generation of
`antidepressants
`
`3. Mechanisms of action
`
`4.
`
`Clinical efficacy
`
`5. Onset of action
`
`6.
`
`7.
`
`8.
`
`9.
`
`10.
`
`11.
`
`Functional impairment in
`MDD
`
`Cognitive dysfunction in MDD
`
`Long-term treatment studies
`
`Safety and tolerability
`
`Conclusion
`
`Expert opinion
`
`Review
`
`A review of the clinical efficacy,
`safety and tolerability of the
`antidepressants vilazodone,
`levomilnacipran and vortioxetine
`
`& George T Grossberg
`William James Deardorff
`†St. Louis University School of Medicine, St Louis, MO, USA
`
`†
`
`Introduction: As a leading cause of disability, major depressive disorder (MDD)
`is characterized by reduced quality of life and altered functioning. Current
`pharmaceutical treatment options are limited in their success by modest
`effects and adverse events that often lead to discontinuation. One current
`trend in antidepressant development is to combine inhibition of the seroto-
`nin transporter with other pharmacological targets, including the norepi-
`nephrine transporter or different serotonin receptors.
`Areas covered: In a span of < 3 years, the FDA approved three new antidepres-
`sants for the treatment of MDD: vilazodone in January 2011, levomilnacipran
`in July 2013 and vortioxetine in September 2013. This article reviews the effi-
`cacy, safety and tolerability of these three drugs mainly from the Phase III
`trial data.
`Expert opinion: All three drugs are effective in the treatment of MDD, but
`data comparing them to other antidepressants is currently lacking. Vilazo-
`done was proposed to produce a more rapid onset and have fewer sexual
`side effects but neither effect has been conclusively shown. Levomilnacipran
`appears to be effective in improving functional impairment, including both
`social and work functioning. Vortioxetine is currently the only drug of the
`three with proven efficacy in elderly patients. It also appears to have cognitive
`enhancing properties which are largely independent of improved depressive
`symptoms. Overall, these drugs represent a promising step forward in antide-
`pressant drug development.
`
`Keywords: antidepressant, levomilnacipran, major depressive disorder, vilazodone, vortioxetine
`
`Expert Opin. Pharmacother. (2014) 15(17):2525-2542
`
`1. Introduction
`
`Major depressive disorder (MDD) is one of the leading causes of disability and is
`predicted to be the second-leading cause of burden of disease worldwide by
`2030 [1,2]. Although antidepressants play an important role in treating MDD, the
`first decade of the twenty-first century saw a slowing in the development of novel
`antidepressants. A select sample of the FDA’s approvals shows that many were for
`repackaged drugs: an enantiomer of the racemic selective serotonin reuptake inhib-
`itor (SSRI) citalopram (escitalopram), a transdermal patch of the monoamine
`oxidase inhibitor selegiline, a major metabolite of venlafaxine (desvenlafaxine), a
`different formulation of bupropion (bupropion hydrobromide) and an extended-
`release version of trazodone [3]. Some reasons that might account for the shortcom-
`ings of antidepressant drug discovery include inadequate dosing in clinical trials,
`limited time period of medications under patent protection and the financial risk
`of drugs with novel mechanisms [4]. The second decade appears more promising
`thus far with the FDA’s approval of vilazodone in January 2011, levomilnacipran
`
`10.1517/14656566.2014.960842 © 2014 Informa UK, Ltd. ISSN 1465-6566, e-ISSN 1744-7666
`All rights reserved: reproduction in whole or in part not permitted
`
`2525
`
`Downloaded by [Mount Sinai Health System Libraries] at 10:11 05 December 2017
`
`Page 2
`
`

`

`W. J. Deardorff & G. T. Grossberg
`
`Article highlights.
`
`. The FDA approved three antidepressants for the
`treatment of major depressive disorder (MDD)
`in < 3 years: vilazodone in January 2011,
`levomilnacipran in July 2013 and vortioxetine in
`September 2013.
`. All three antidepressants showed significant clinical
`efficacy compared to placebo and were generally well
`tolerated, with adverse events being mostly
`gastrointestinal in nature.
`. Although vilazodone was proposed to have an
`accelerated onset of action and fewer sexual side
`effects, neither effect has been conclusively shown so
`far. Vilazodone might play a role in the treatment of
`patients with MDD and secondary anxiety symptoms.
`. Levomilnacipran significantly improved functional
`impairment across work/school, social and family life
`settings.
`. Vortioxetine is efficacious in the elderly and has
`cognitive-enhancing properties.
`. Thus far, no conclusion can be made in regard to the
`effect of vilazodone or levomilnacipran on sexual
`functioning. Vortioxetine’s effect on sexual functioning
`was more extensively studied, but the results have been
`mixed. No significant weight gain was observed with
`any of the three drugs.
`. Future trials that compare the efficacy of these drugs to
`other common antidepressants are needed to further
`clarify their role as possible first-line treatments for
`MDD.
`
`This box summarizes key points contained in the article.
`
`in July 2013 and vortioxetine in September 2013. Vortioxe-
`tine was also approved by the European Medicines Agency
`in December 2013. The approval of three new antidepressants
`within a period of < 3 years is an exciting development in the
`treatment of a disorder with an estimated lifetime prevalence
`of 16% of the US population [5]. This review will analyze
`the clinical efficacy, safety and tolerability of these three
`drugs, primarily using data from select Phase III trials, and
`will emphasize their unique properties.
`
`2. New generation of antidepressants
`
`The second generation of antidepressants includes SSRIs,
`serotonin-norepinephrine reuptake inhibitors (SNRIs) and
`drugs that exhibit multiple receptor profiles [6]. The second-
`generation antidepressants likely do not differ much from
`each other in terms of efficacy, meaning that expected side
`effects and cost play a large role in treatment selection [7,8].
`A general trend in the development of newer drugs has been
`to target multiple receptors in addition to the serotonin trans-
`porter (SERT). This type of ‘intramolecular polypharmacy’
`affects multiple neurotransmitter pathways, potentially lead-
`ing to greater efficacy on a broader range of symptoms [6].
`Another trend is the exploration of drugs affecting systems
`other than the aminergic system, such as different glutamate
`
`modulators and agomelatine, a melatonergic MT1/MT2
`agonist and serotonin (5-hydroxytryptamine)
`(5-HT)2C
`antagonist licensed in Europe.
`
`3. Mechanisms of action
`
`Figure 1 reviews the mechanisms of action of the three antide-
`pressants. Vilazodone acts as a partial 5-HT1A receptor
`agonist and serotonin reuptake inhibitor [9]. Vilazodone inhib-
`its serotonin reuptake with an IC50 of 1.6 nM and binds to the
`human 5-HT1A binding sites with an IC50 of 2.1 nM [10].
`Since the release of serotonin is regulated by presynaptic
`5-HT1A autoreceptors, vilazodone was proposed to augment
`the effect of SERT inhibition [11]. Data from preclinical studies
`were hopeful that vilazodone could produce a rapid onset of
`action by increasing 5-HT output either through inhibiting
`or rapidly desensitizing 5-HT1A autoreceptors [12]. Early stud-
`ies showed a greater elevation of extracellular 5-HT levels in
`the rat brain frontal cortex compared to SSRIs and a 30-fold
`greater potency at inhibiting serotonin reuptake compared to
`fluoxetine [13-15]. Rapid onset would be advantageous because
`of the delayed onset of antidepressants, which is generally
`attributed to the need for long-term adaptations such as recep-
`tor desensitization and signaling pathway alterations [16].
`Levomilnacipran is an SNRI and the more active enantio-
`mer of milnacipran, currently used in the USA for treating
`fibromyalgia and neuropathic pain and available in Europe
`for treating depression. Although levomilnacipran is not a
`novel antidepressant, its selectivity is slightly different com-
`pared to other SNRIs. The SNRIs duloxetine, venlafaxine
`and desvenlafaxine all preferentially inhibit 5-HT reuptake
`compared to norepinephrine (NE) reuptake. However, levo-
`milnacipran shows an approximately twofold greater potency
`at inhibiting NE relative to 5-HT reuptake and a NE:5-HT
`transporter selectivity ratio of 27 and 17 times that of dulox-
`etine and venlafaxine, respectively [17]. The IC50 values of
`levomilnacipran at human NE transporter and SERT recep-
`tors stably expressed in Chinese hamster ovary cells were
`10.5 and 19.0 nM, respectively [17]. This is potentially advan-
`tageous because the noradrenergic system might be involved
`in mental and physical slowing, decreased concentration,
`social functioning and loss of energy [18]. In fact, treatment
`with milnacipran appears
`to enhance social
`functioning
`throughout a variety of studies [18].
`Vortioxetine functions as a SERT inhibitor (K
`i = 1.6 nM,
`
`IC50 = 5.4 nM), 5-HT1A receptor agonist (Ki = 15 nM),
`5-HT3 receptor antagonist (K
`i = 3.7 nM), 5-HT7 receptor
`(K
`i = 19 nM), 5-HT1D receptor antagonist
`antagonist
`(K
`i = 54 nM) and 5-HT1B receptor partial agonist
`(K
`i = 33 nM) [19,20]. This mechanism is thought to be advanta-
`geous for several reasons. 5-HT1A receptor agonists could
`desensitize somatodendritic and activate postsynaptic 5-HT1A
`receptors to elevate serotonin. 5-HT3 and 5-HT7 antagonists
`have been shown to regulate various neurotransmitter systems,
`fully understood [21,22].
`although the mechanism is not
`
`2526
`
`Expert Opin. Pharmacother. (2014) 15(17)
`
`Downloaded by [Mount Sinai Health System Libraries] at 10:11 05 December 2017
`
`Page 3
`
`

`

`A review of the clinical safety, efficacy and tolerability
`
`Serotonergic
`
`Noradrenergic
`
`5-HT1A
`
`Vilazodone
`
`5-HT1B
`
`Vortioxetine
`
`Levomilnacipran
`
`NET
`
`Vortioxetine
`vilazodone
`
`Vortioxetine
`
`SERT
`
`Levomilnacipran,
`vortioxetine,
`vilazodone
`
`Vortioxetine
`
`NER
`
`5-HT7
`5-HT3
`
`5-HT1A
`5-HT1B
`5-HT1D
`Post-synaptic neuron
`
`SERT
`inhibitor
`
`NET
`inhibitor
`
`5-HT1A
`agonist
`
`5-HT1B
`agonist
`
`5-HT1D
`antagonist
`
`5-HT3
`antagonist
`
`5-HT7
`antagonist
`
`Vilazodone
`
`Levomilnacipran
`
`Vortioxetine
`
`Partial
`
`Partial
`
`Figure 1. Schematic representation of the mechanisms of action of vilazodone, levomilnacipran and vortioxetine. 5-HT1A,
`5-HT1B, 5-HT1D and 5-HT7 receptors are G-protein-coupled receptors. The 5-HT3 receptor is ionotropic. The receptor locations
`are simplified, and the postsynaptic neuron is not meant to represent any one type of neuron. Lines with arrows indicate
`agonist properties. Lines with bars indicate antagonism or inhibition.
`NER: Norepinephrine receptor; NET: Norepinephrine transporter; SERT: Serotonin transporter.
`
`Vortioxetine does appear to modulate glutamate and GABA
`transmission based on preclinical data involving serotonin
`depleted and progesterone-withdrawal studies in rats [23]. This
`could be attributed to its 5-HT3 antagonism given that
`5-HT3 antagonists have been shown to enhance glutamate
`transmission via reductions in GABA-mediated inhibition [24].
`In rat microdialysis studies, vortioxetine showed significant
`increases in 5-HT, dopamine and noradrenaline levels in the
`ventral hippocampus and medial prefrontal cortex [19,25]. Vorti-
`oxetine increased extracellular levels of 5-HT in the rat ventral
`hippocampus more than SSRIs and SNRIs [26]. Combined,
`the preclinical data show that vortioxetine’s multiple receptor
`targets are important for its antidepressant activity, indicating
`a novel mechanism distinct from SSRIs and SNRIs [27].
`
`4. Clinical efficacy
`
`gov. Additional data reviewed by the FDA were obtained
`through http://www.accessdata.fda.gov/scripts/cder/drugsatfda/
`index.cfm. Poster presentations at conferences such as the
`American Psychiatric Association and American Society of
`Clinical Psychopharmacology were also reviewed.
`Several measures are commonly used to evaluate the clinical
`efficacy of antidepressants. Based on a primary efficacy mea-
`sure, such as the Montgomery-A˚ sberg Depression Rating Scale
`(MADRS) or Hamilton Depression Rating Scale (HAM-D/
`HDRS), 2-point differences from placebo at end point are
`[28]. Differences
`generally considered clinically relevant
`> 10% in response rate between placebo and treatment and a
`number needed to treat (NNT) < 10 are also considered clin-
`ically advantageous. Based on Cohen’s d effect size, values of
`0.2 and 0.5 are considered small and medium, respectively [29].
`Most antidepressants show effect sizes between 0.2 and 0.5 [30].
`
`The following is a selective review of published and non-
`published results
`from clinical
`trials
`that were obtained
`through online searches of http://www.ncbi.nlm.nih.gov/
`pubmed, http://www.clinicaltrials.gov, and http://www.fda.
`
`4.1 Vilazodone
`A total of five 8-week Phase II trials involving vilazodone were
`performed with two using fixed doses (5, 10 and 20 mg/day)
`and three using a flexible titration design [31]. None of these
`
`Expert Opin. Pharmacother. (2014) 15(17)
`
`2527
`
`Downloaded by [Mount Sinai Health System Libraries] at 10:11 05 December 2017
`
`Page 4
`
`

`

`W. J. Deardorff & G. T. Grossberg
`
`Table 1. Summary of MADRS total score data for vilazodone from four short-term placebo-controlled trials.
`
`Study
`
`Vilazodone dose
`(daily)
`
`Baseline mean MADRS
`
`LSM change in MADRS,
`end point
`
`LSMD
`
`Rickels et al. (2009) [33]
`8-week Phase III study
`
`Khan et al. (2011) [35]
`8-week Phase III study
`
`Gommoll et al. (2014) [43]
`(NCT01473381) 10-week
`Phase IV study
`
`NCT01473394 [39]
`8-week Phase IV study
`
`Titration design,
`20 or 40 mg vilazodone
`
`Titration design,
`40 mg vilazodone
`
`20 mg vilazodone
`40 mg vilazodone
`40 mg citalopram
`
`40 mg vilazodone
`
`Placebo
`
`Vilazodone
`
`Placebo
`
`Vilazodone
`
`30.7
`
`32.0
`
`30.8
`
`31.9
`
`-9.7
`
`-12.9
`
`-10.8
`
`-13.3
`
`-11.0
`
`-16.1
`
`-3.2*
`
`-2.5*
`
`-2.57*
`-2.82*
`-2.74*
`
`-5.1*
`
`*Indicates statistically significant difference from placebo. Last-observation-carried-forward approach was used for the two Phase III trials. Mixed model, repeated
`measures was used for NCT01473394.
`LSMD: Least squares mean difference; MADRS: Montgomery-A˚ sberg depression rating scale.
`
`five trials showed a statistically significant difference from
`placebo on the HAM-D. Three of the five trials used an active
`comparator (fluoxetine or citalopram), which also did not
`show significant differences from placebo. A review of vilazo-
`done cites flaws in methodology, use of low doses, a large num-
`ber of investigative sites and use of the HAM-D as opposed to
`the MADRS as potential reasons for the failed studies [32].
`The FDA’s approval of vilazodone was largely based on two
`Phase III 8-week trials and one 52-week trial assessing long-
`term safety (Table 1). The first Phase III trial enrolled patients
`aged 18 -- 65 years with MDD [33]. Doses were increased from
`10 to 40 mg over a course of 2 weeks, with patients able to
`receive 20 mg if they could not tolerate the 40 mg dose. Vila-
`zodone showed a statistically significant difference from pla-
`cebo at week 8, with a least-squares mean difference (LSMD)
`of -3.2 points on the MADRS total score (d = 0.30) [34]. Signif-
`icant difference from placebo occurred at week 1 for patients
`on vilazodone (LSMD = -1.7, p = 0.0002).
`The second Phase III trial involved patients aged 18 -- 70
`with MDD [35]. Doses of vilazodone were increased from
`10 to 40 mg over 2 weeks. Reducing the dose to 20 mg was
`not allowed. The mean difference on MADRS total score at
`week 8 was -2.5 between vilazodone and placebo (d = 0.23)
`[34]. Significant difference from placebo occurred starting at
`week 6 on vilazodone, which is considerably later than the
`previous trial.
`Both trials showed > 10% difference in response rates (‡ 50%
`decrease in MADRS total score from baseline) between vilazo-
`done and placebo, which is considered clinically meaningful [28].
`MADRS response rates were 40.4% for vilazodone and 28.1%
`for placebo in one trial and 43.7% for vilazodone and 30.3% for
`placebo in the other [34]. This is similar to a meta-analysis of five
`trials involving escitalopram, which showed response rates of
`52.9% after 8 weeks on the MADRS for escitalopram and
`37.3% for placebo [36]. The NNT for response and remission
`
`on MADRS with vilazodone was 8 and 12 in the pooled data,
`respectively [37]. Whereas both Phase III trials did not show sig-
`nificant differences in remission, data from a recent Phase IV
`trial using 40 mg/day vilazodone showed a significant difference
`after 8 weeks in remission rates (MADRS £ 10) from placebo
`(34 vs 22%, NNT = 9) [38,39]. In the Phase IV trial, the
`LSMD for vilazodone 40 mg versus placebo was -5.1 on
`MADRS total score, with categorical improvements across all
`10 MADRS items [40].
`Given that the anxiolytic drug buspirone and vilazodone act
`at the 5-HT1A receptor, vilazodone was hypothesized to help
`patients with MDD and high levels of anxiety [11]. In a post-
`hoc analysis of both Phase III trials, a subgroup of patients
`classified as anxious based on HAM-D17 anxiety/somatization
`subscale scores showed significant difference from placebo on
`the MADRS (LSMD = -3.6) and the Hamilton Anxiety Scale
`(HAM-A) total score (LSMD = -1.82, d = 0.25) [41]. This is
`important clinically, given the poorer treatment outcomes
`and lower rates of remission in patients with anxious depres-
`sion [42]. However, the authors explain that the retrospective
`nature limits the results since neither trial was designed to com-
`pare patients with anxious versus non-anxious depression.
`A recent unpublished multicenter, randomized, double-
`blind, 10-week trial compared 20 mg vilazodone (n = 288)
`and 40 mg vilazodone (n = 287) with placebo (n = 281).
`The SSRI citalopram was used for assay sensitivity (40 mg,
`n = 282) [43]. The results showed a significant difference on
`MADRS total score (primary efficacy outcome) at week
`10 for all groups compared to placebo. The LSMDs from pla-
`cebo were -2.57 for vilazodone 20 mg, -2.82 for vilazodone
`40 mg and -2.74 for citalopram 40 mg.
`
`4.2 Levomilnacipran
`Five short-term trials ranging from 8 to 10 weeks and one
`24-week long-term maintenance trial were evaluated by the
`
`2528
`
`Expert Opin. Pharmacother. (2014) 15(17)
`
`Downloaded by [Mount Sinai Health System Libraries] at 10:11 05 December 2017
`
`Page 5
`
`

`

`Table 2. Summary of MADRS and SDS total score data for levomilnacipran from five short-term, placebo-
`controlled trials.
`
`A review of the clinical safety, efficacy and tolerability
`
`Measure
`
`Baseline mean
`
`LSM change, end point
`
`LSMD
`
`Placebo Levomilnacipran Placebo Levomilnacipran
`
`MADRS
`SDS
`
`30.5
`20.8
`
`30.9
`21.3
`
`-14.5
`-7.7
`
`-18.7
`-11.1
`
`-4.2*
`-3.4*
`
`-3.2*
`-1.4
`-4.0*
`-2.5*
`-4.9*
`-2.6*
`
`-3.3*
`-1.8*
`-3.1*
`-2.7*
`-3.1*
`
`-2.6*
`
`-1.5
`-0.6
`
`Levomilnacipran
`dose (daily)
`
`75 -- 100 mg
`
`Study
`
`Montgomery et al.
`(2013) [45]
`10-week, positive,
`Phase II study
`
`Asnis et al. (2013) [47]
`8-week, positive,
`Phase III study
`
`Bakish et al. (2014) [48]
`8-week, positive,
`Phase III study
`
`Sambunaris et al.
`(2014) [49]
`8-week, positive,
`Phase III study
`
`Gommoll et al.
`(2014) [50]
`8-week, negative,
`Phase III study
`
`Pooled population
`Montgomery et al. (2014) [51]
`5 Phase II/III studies
`
`Pooled population
`Sambunaris et al. (2014) [78]
`5 Phase II/III studies
`
`40 mg
`
`80 mg
`
`120 mg
`
`40 mg
`
`80 mg
`
`40 -- 120 mg
`
`MADRS
`SDS
`MADRS
`SDS
`MADRS
`SDS
`
`MADRS
`SDS
`MADRS
`SDS
`
`MADRS
`SDS
`
`35.6
`21.5
`35.6
`21.5
`35.6
`21.5
`
`31.0
`16.4
`31.0
`16.4
`
`35.2
`19.7
`
`40 -- 120 mg
`
`MADRS
`SDS
`
`35.5
`20.8
`
`36.0
`21.1
`36.1
`21.4
`36.0
`21.3
`
`30.8
`16.7
`31.2
`17.6
`
`35.0
`20.1
`
`35.9
`21.7
`
`-11.6
`-7.2
`-11.6
`-7.2
`-11.6
`-7.2
`
`-11.3
`-5.4
`-11.3
`-5.4
`
`-12.2
`-5.4
`
`-14.8
`-8.6
`-15.6
`-9.7
`-16.5
`-9.7
`
`-14.6
`-7.3
`-14.4
`-8.2
`
`-15.3
`-8.0
`
`-14.2
`-8.2
`
`-15.7
`-8.8
`
`MADRS
`
`33.3
`
`33.8
`
`-12.9
`
`-15.8
`
`-3.0*
`
`SDS
`
`20.1
`
`20.4
`
`-2.2*
`
`*Indicates statistically significant difference from placebo. Mixed model repeated measures was used for all calculations.
`LSMD: Least squares mean difference; MADRS: Montgomery-A˚ sberg depression rating scale; SDS: Sheehan Disability Scale.
`
`FDA, prior to approval [44]. In all these studies, levomilnaci-
`pran was used on adult outpatients with MDD at a range of
`40 -- 120 mg/day. Two fixed-dose studies and two flexible-
`dose studies were considered supportive. One flexible-dose
`study and a long-term maintenance study did not achieve
`statistically significant results. The LSMD on the MADRS
`total score from baseline to the end of trial ranged from -3.1
`to -4.9 for the four positive studies, compared to placebo,
`with statistically significant difference from placebo occurring
`during week 3 or week 4. Efficacy data for the five short-term
`trials are displayed in Table 2.
`In a 10-week, Phase II non-US study, patients receiving
`flexible doses (75 -- 100 mg/day) saw significant improve-
`ments in MADRS total score compared to placebo starting
`from week 3 onward, with a LSMD of -4.2 at week 10 [45].
`In a post-hoc analysis of this trial, the authors noted that every
`item on the MADRS achieved statistically significant differen-
`ces, indicating improvement across a broad range of symp-
`toms [46].
`
`A Phase III, USA, 8-week outpatient study compared three
`fixed doses of levomilnacipran (40, 80 and 120 mg/day) with
`placebo [47]. Significant improvements in MADRS total score
`were seen for all three doses compared to placebo, with the
`80 and 120 mg groups achieving statistical significance at
`week 4. The authors concluded that dosing up to 120 mg
`was effective without decreased tolerability.
`A Phase III, 8-week outpatient trial in the USA and Canada
`compared fixed doses of levomilnacipran (40 and 80 mg/day)
`with placebo [48]. This study was unique in that it only
`included patients with recurrent depression. There was a
`statistically significant difference in change in MADRS as
`early as week 4 for both treatment groups.
`A Phase III, USA, 8-week trial compared flexible dose
`levomilnacipran (40 -- 120 mg/day) with placebo [49]. There
`was a statistically significant difference in change in MADRS
`as early as week 4. The Motivation and Energy Inventory
`Short Form (MEI-SF) was used to evaluate patient motivation
`and energy. The levomilnacipran group showed a statistically
`
`Expert Opin. Pharmacother. (2014) 15(17)
`
`2529
`
`Downloaded by [Mount Sinai Health System Libraries] at 10:11 05 December 2017
`
`Page 6
`
`

`

`W. J. Deardorff & G. T. Grossberg
`
`significant difference from placebo on the MEI-SF total score
`and the cognitive and social subscales. This was promising
`given that deficits in the noradrenergic symptom cluster are
`thought to involve lack of energy.
`The failed short-term Phase III trial was an 8-week outpa-
`tient
`study comparing flexible doses of
`levomilnacipran
`(40 -- 120 mg/day) with placebo [50]. Although a numerically
`greater improvement in MADRS was seen (LSMD = -1.5),
`the results were not statistically significant. The authors
`explain that one reason for the lack of statistical significance
`was that the placebo response was 2 -- 3 points greater than
`in other comparable studies. The reason for this was unknown
`given the similarities to other trials.
`A post-hoc pooled analysis of all
`five short-term trials
`showed a significantly greater improvement in MADRS in
`the levomilnacipran groups compared to placebo (LSMD =
`-3.0) [51]. All subgroups, except those with MDD duration
`of < 2 years and current episode duration of > 12 months,
`showed statistically significant improvements. The authors
`noted that patients with current episodes of > 12 months are
`more likely to have refractory MDD which might not respond
`to therapy. The group with MDD of < 2 years was relatively
`small and had a high response to placebo. The highest differ-
`ences in response rates from placebo occurred in patients
`‡ 60 years of age (17.9% difference), having ‡ 5 prior depres-
`sive episodes (14.6% difference) and having baseline MADRS
`total score of < 30 (14.1%). The subgroup with the smallest
`NNT for response (NNT = 6) and highest LSMD from
`placebo in MADRS total score (LSMD = -4.4) occurred in
`patients aged > 60. This is interesting given the small and
`inconsistent effect
`sizes
`seen in trials
`involving elderly
`patients [52]. However, the group size for patients ‡ 60 years
`was relatively small compared to the other subgroups. The
`overall pooled NNT for response and remission was 10 and
`17, respectively.
`
`4.3 Vortioxetine
`Ten short-term trials and one relapse prevention trial were
`evaluated by the FDA to assess vortioxetine’s efficacy [53]. Of
`the 10 short-term trials, 6 and the relapse prevention trial
`were positive. One study was considered failure since the
`active comparator did not statistically differ from placebo.
`The final three were negative. Unlike studies with levomilna-
`cipran, many of the studies included an active comparator
`(duloxetine or venlafaxine). Also unique to vortioxetine was
`a study involving elderly patients. Among the successful trials,
`differences in MADRS from placebo varied from -2.8 to -7.1,
`with significant difference from placebo occurring between
`weeks 2 and 4. A recent meta-analysis of nine short-term trials
`found that the mean difference in MADRS total score at end
`point for vortioxetine compared to placebo was -2.6 (5 mg), -
`3.5 (10 mg), -2.6 (15 mg) and -4.5 (20 mg) points [54]. Dose-
`related improvements on all MADRS single items were
`observed for vortioxetine. A summary of the efficacy data is
`shown in Table 3 [55].
`
`The first study was a non-USA, 6-week Phase II trial using
`fixed doses (5 and 10 mg) of vortioxetine versus placebo, with
`venlafaxine as an active comparator [56]. Both vortioxetine
`groups showed statistically significant differences in change
`from baseline in MADRS compared to placebo, with corre-
`sponding effect sizes of d = 0.56 (5 mg) and d = 0.54
`(10 mg). Statistical difference from placebo occurred during
`week 2 for 10 mg and during week 3 for the 5 mg group.
`Both doses showed significant improvements across 9 out of
`10 MADRS items compared to placebo, indicating broad
`range efficacy.
`A Phase III, non-USA, 8-week trial compared 1, 5 and
`10 mg vortioxetine with placebo [57]. There was a statistically
`significant reduction in HDRS-24 total score at week 8 for
`only the 10 mg group, with difference occurring at week 2.
`The authors note that the lower placebo response rates in
`this study might have helped in detecting an earlier significant
`response in the vortioxetine group.
`A non-USA, Phase III, 8-week trial included inpatients and
`outpatients aged 18 -- 75 with a current major depressive
`episode of > 3 months [58]. This trial used higher doses
`(15 and 20 mg) than previous trials. This was expected to
`result in a broader clinical profile since higher doses of vorti-
`oxetine result in increased 5-HT receptor occupancy [59].
`Duloxetine was used as an active comparator. A significant
`change in MADRS from baseline to week 8 was seen for
`both doses of vortioxetine, with significant difference occur-
`ring at week 2 for the 20 mg group and at week 4 for the
`15 mg group. All 10 MADRS single items showed significant
`separation from placebo. Even patients with a baseline
`HAM-A of ‡ 20 showed significant differences from placebo
`in MADRS (-5.2 for 15 mg and -6.4 for 20 mg). Significant
`improvements on the Quality of Life Enjoyment and Satisfac-
`tion Questionnaire Short-Form were also seen with effect
`sizes of 0.38 (15 mg) and 0.52 (20 mg),
`indicating
`clinically relevant increases in enjoyment and satisfaction in
`daily life.
`In two USA, 8-week trials involving adults with MDD,
`only vortioxetine 20 mg showed statistically significant differ-
`ence from placebo on the MADRS total score (LSMD of -2.8
`and -3.6) [60,61]. The 15 and 10 mg vortioxetine groups were
`not significant at end point. Interestingly, doses < 20 mg
`demonstrated efficacy only in non-US studies for reasons
`that were not clear to the FDA reviewers [53].
`Vortioxetine is unique among the three antidepressants in
`that a trial on elderly patients was performed [62]. Finding
`antidepressants that are suitable for the elderly is important
`given the small effect sizes seen in controlled trials involving
`elderly patients [52]. In this 8-week USA and non-US study,
`vortioxetine 5 mg was compared with placebo in patients
`with a mean age of 71 years. Vortioxetine 5 mg significantly
`separated from placebo at week 6, with a mean difference
`on the HAM-D24 (primary efficacy measure) of -3.32 at
`week 8. Duloxetine showed a mean difference of -5.48 at
`week 8 on the HAM-D24. When analyzed by country, the
`
`2530
`
`Expert Opin. Pharmacother. (2014) 15(17)
`
`Downloaded by [Mount Sinai Health System Libraries] at 10:11 05 December 2017
`
`Page 7
`
`

`

`A review of the clinical safety, efficacy and tolerability
`
`Table 3. Summary of MADRS and HAM-D24 total score data for vortioxetine from six successful short-term,
`placebo-controlled trials.
`
`Study
`
`Treatment dose
`(daily)
`
`Measure
`
`Baseline mean
`
`LSM change, end
`point
`
`LSMD
`
`Placebo
`
`Treatment
`
`Placebo
`
`Treatment
`
`Alvarez et al. (2012) [56]
`6-week, non-US Phase II;
`LOCF
`
`Boulenger et al. (2014) [58]
`8-week, non-US Phase III;
`MMRM
`
`Mahableshwarkar
`et al. (2013) [61] 8-week,
`US Phase III; MMRM
`
`Vortioxetine 5 mg
`Vortioxetine 10 mg
`Venlafaxine 225 mg
`
`Vortioxetine 15 mg
`Vortioxetine 20 mg
`Duloxetine 60 mg
`
`Vortioxetine 15 mg
`Vortioxetine 20 mg
`Duloxetine 60 mg
`
`Jacobsen et al. (2013) [60]
`8-week, US Phase III; MMRM
`
`Vortioxetine 10 mg
`Vortioxetine 20 mg
`
`MADRS
`MADRS
`MADRS
`
`MADRS
`MADRS
`MADRS
`
`MADRS
`MADRS
`MADRS
`
`MADRS
`MADRS
`
`Henigsberg et al. (2012) [57]
`8-week, non-US Phase III;
`MMRM
`
`Katona et al. (2012) [62]
`8-week, US and non-US
`Phase III study in elderly
`patients; LOCF
`
`Pooled data Thase et al.
`(2014) [54]
`9 Phase II/III trials
`MMRM
`
`Vortioxetine 1 mg
`Vortioxetine 5 mg
`Vortioxetine 10 mg
`
`Vortioxetine 5 mg
`Duloxetine 60 mg
`
`HAM-D24
`HAM-D24
`HAM-D24
`
`HAM-D24
`HAM-D24
`
`Vortioxetine 5 mg
`Vortioxetine 10 mg
`Vortioxetine 15 mg
`Vortioxetine 20 mg
`
`MADRS
`MADRS
`MADRS
`MADRS
`
`33.9
`33.9
`33.9
`
`31.5
`31.5
`31.5
`
`31.5
`31.5
`31.5
`
`32.0
`32.0
`
`32.7
`32.7
`32.7
`
`29.4
`29.4
`
`34.1
`34.0
`34.2
`
`31.8
`31.2
`31.2
`
`31.9
`32.0
`32.8
`
`32.2
`32.5
`
`32.5
`32.1
`33.1
`
`29.2
`28.5
`
`-14.5
`-14.5
`-14.5
`
`-11.7
`-11.7
`-11.7
`
`-12.8
`-12.8
`-12.8
`
`-10.8
`-10.8
`
`-11.3
`-11.3
`-11.3
`
`-10.3
`-10.3
`
`-20.4
`-20.2
`-20.9
`
`-17.2
`-18.8
`-21.2
`
`-14.3
`-15.6
`-16.9
`
`-13.0
`-14.4
`
`-14.8
`-15.4
`-16.2
`
`-13.7
`-15.8
`
`-5.9*
`-5.7*
`-6.4*
`
`-5.5*
`-7.1*
`-9.5*
`
`-1.5
`-2.8*
`-4.1*
`
`-2.2
`-3.6*
`
`-3.5
`-4.1
`-4.9*
`
`-3.3*
`-5.5*
`
`-2.6*
`-3.5*
`-2.6
`-4.5*
`
`*Indicates statistically significant difference from placebo.
`HAM-D24: Hamilton depression rating scale 24 item; LOCF: Last-observation-carried-forward; LSMD: Least squares mean difference; MADRS: Montgomery-A˚ sberg
`depression rating scale; MMRM: Mixed model repeated measures.
`
`US subgroup (n = 171) showed a LSMD of -0.7 for vortiox-
`etine and -2.8 for duloxetine (both not significant) compared
`to placebo [53]. The non-US subgroup (n = 277) showed
`significant differences from placebo of -4.9 and -7.1 for vorti-
`oxetine and duloxetine, respectively.
`One recent 12-week non-US trial not included in the
`FDA’s
`review randomized
`patients
`to
`vortioxetine
`(10 -- 20 mg) or agomelatine (25 -- 50 mg) after inadequate
`response to SSRI/SNRI monotherapy [63].Eligibility criteria
`included patients with inadequate response for at least 6 weeks
`to an SSRI (citalopram, escitalopram, paroxetine and sertra-
`line) or SNRI (duloxetine and venlafaxine). Vortioxetine
`was statistically superior to agomelatine on the MADRS
`starting at week 4, with a difference at week 8 of 2.2 points.
`Vortioxetine was also significantly superior in response and
`remission rates. At week 12, response rates were 69.8% for
`vortioxetine and 56.0% for agomelatine, and remission rates
`were 55.2% for vortioxetine and 39.4% for agomelatine.
`Switching to vortioxetine from either an SSRI or SNRI was
`numerically superior to switchin