Desvenlafaxine vs. placebo in the treatment of persistent depressive disorder

Desvenlafaxine vs. placebo in the treatment of persistent depressive disorder

Accepted Manuscript Desvenlafaxine vs. placebo in the treatment of persistent depressive disorder David J. Hellerstein M.D. , Jonathan W. Stewart M.D...

1MB Sizes 0 Downloads 54 Views

Accepted Manuscript

Desvenlafaxine vs. placebo in the treatment of persistent depressive disorder David J. Hellerstein M.D. , Jonathan W. Stewart M.D. , Ying Chen M.D. , Vinushini Arunagiri M.A. , Bradley S. Peterson M.D. , Patrick J. McGrath M.D. PII: DOI: Reference:

S0165-0327(17)32202-4 https://doi.org/10.1016/j.jad.2018.11.065 JAD 10283

To appear in:

Journal of Affective Disorders

Received date: Revised date: Accepted date:

24 October 2017 21 September 2018 3 November 2018

Please cite this article as: David J. Hellerstein M.D. , Jonathan W. Stewart M.D. , Ying Chen M.D. , Vinushini Arunagiri M.A. , Bradley S. Peterson M.D. , Patrick J. McGrath M.D. , Desvenlafaxine vs. placebo in the treatment of persistent depressive disorder, Journal of Affective Disorders (2018), doi: https://doi.org/10.1016/j.jad.2018.11.065

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Highlights



AC

CE

PT

ED

M

AN US



The SNRI medication desvenlafaxine did not differentiate from placebo on the primary outcome (Hamilton Depression Rating Scale) in non-major persistent depressive disorder. As a negative study, this does not support the use of DVLX for non-major PDD. Statistically significant secondary analyses suggest the overall negative result could be due to sample size or sampling, suggesting further studies of this medication might be appropriate in this population. Twelve week continuation phase of active medication was associated with additional improvement in symptoms and psychosocial functioning, and remission of depressive symptoms was associated with normalization of social functioning score on the SAS.

CR IP T



ACCEPTED MANUSCRIPT 2

revision 9/21/2018 3940 words 4 tables 1 Consort Diagram 1 figure

1,2

1,2

CR IP T

Title: Desvenlafaxine vs. placebo in the treatment of persistent depressive disorder

1,2

Authors: David J. Hellerstein, M.D. , Jonathan W. Stewart, M.D. , Ying Chen, M.D. , Vinushini 3 4 1,2 Arunagiri, M.A. , Bradley S. Peterson, M.D. , Patrick J. McGrath, M.D.

M

ED

Address for Correspondence: David J. Hellerstein, MD New York State Psychiatric Institute 1051 Riverside Drive, Unit #51 New York, NY 10032 tel. 646-774-8069 fax. 646-774-8034 email: [email protected]

AN US

1. New York State Psychiatric Institute, New York, NY 2. Columbia University College of Physicians and Surgeons, New York, NY 3. Teachers College, Columbia University, New York, NY 4. Institute for the Developing Mind at Children’s Hospital Los Angeles; Department of Psychiatry, Keck School of Medicine, University of Southern California, Los Angeles, CA

AC

CE

PT

Clinicaltrials.gov registration #: NCT01537068

ACCEPTED MANUSCRIPT 3

Abstract Introduction: Pharmacotherapy of non-major persistent depressive disorder (PDD) is little studied. We report a study of the serotonin-norepinephrine reuptake inhibitor (SNRI) desvenlafaxine (DVLX) for PDD.

CR IP T

Method: Non-psychotic, non-bipolar outpatients aged 20-65 having PDD without concurrent major depression (MDD) were randomized double-blind to desvenlafaxine or placebo for 12 weeks. All

had Hamilton Depression Rating Scale (HDRS-24) score >12. Open-label DVLX was offered for 12 weeks following the acute trial.

Results: Seventy-one subjects having mean baseline HDRS-24 20.27+4.77 were eligible, of whom

AN US

post-RZ data was available for all 59 randomized. The primary 12 week analysis did not

differentiate DVLX-treated subjects’ mean HDRS scores from those on placebo (6.53 + 3.98 vs. 8.24 + 4.96, F=3.33, df=1, p=.07). Several secondary analyses yielded statistically significant results, including Responder, CGI and QIDS.

M

Discussion: As the primary analysis did not reach statistical significance, this is a negative study which does not support the use of DVLX for non-major PDD. Nevertheless,

ED

statistically significant secondary analyses suggest the overall negative result could be due to sample size or sampling, suggesting further studies of this medication might be

PT

appropriate in this population.

CE

ClinicalTrials.gov identifier: NCT01537068

AC

Previous presentation: none

ACCEPTED MANUSCRIPT 4

Introduction

Chronic depression is common, affecting 1.5-5% of the population and produces significant 1-9

functional impairment, negative health outcomes, and social costs, psychosocial burden than acute depression.

10

DSM-5

11

resulting in greater

recognized the importance of chronicity by

CR IP T

consolidating dysthymic disorder (DD), chronic major depressive disorder (MDD), and residual MDD, into a single classification, Persistent Depressive Disorder (PDD).

Identifying effective treatments for PDD is clearly important to relieve symptoms and improve

AN US

psychosocial and health outcomes. Yet chronic depression remains under-studied: a 2000

Cochrane review, for example, found only 20 placebo-controlled studies of DD (chronic low-grade 12

depression), concluding that antidepressants are effective. A more recent review found 36 studies of pure dysthymia, and 22 studies of chronic major depression or dysthymia with current MDD, with 13

efficacy data on only 5,806 patients world-wide. . Few drugs have been adequately tested, e.g.

13

Other reviews

14-16

suggest these patients may require higher than

ED

greater efficacy than others

M

with 2 or more trials enrolling over 100 patients, though data suggests some medications may have

usual medication doses, combined medication and psychotherapy treatments, or concurrent dosing with two or more classes of medication to achieve optimal outcome, since residual symptoms and

PT

persistent functional impairment are common. An alternative to combining treatments has been dual-mechanism medications, such as serotonin-norepinephrine reuptake inhibitors (SNRIs),

CE

though SNRIs may have more side effects

17

in this population than other medications. We 18

previously reported efficacy of the SNRI duloxetine in chronic non-major depression.

AC

Desvenlafaxine (DVLX) (brand name Pristiq®),

19

like duloxetine, blocks serotonin and

norepinephrine reuptake, so may also effectively treat chronic depression. Thus, a study of the efficacy of DVLX for non-major PDD seems indicated.

Hypotheses: 1) Efficacy study: We expected that desvenlafaxine would be superior to placebo over a twelve-

ACCEPTED MANUSCRIPT 5

week period in the following outcomes: a.

Improved depression, as measured by HDRS-24 item total score at week 12 (primary outcome).

b.

The percentage of subjects classified as (a) Responders and (b) Remitters at week 12.

c.

Improved secondary measures of depression (Quick Inventory of Depressive

CR IP T

Symptomatology-Self-Rated (QIDS-SR), Beck Depression Inventory (BDI), Cornell

Dysthymia Rating Scale (CDRS), and overall severity of illness (Clinical Global ImpressionsSeverity (CGI-S)). d.

Improved psychosocial functioning (as measured by the Social Adjustment Scale (SAS) and

AN US

the Sheehan Disability Scale (SDS), global outcome (Global Assessment of Functioning (GAF)), and temperament (Temperament and Character Inventory (TCI)).

2) Open continuation study: In the continuation phase (weeks 13-24) we expected that patients continuing active DVLX would maintain improvement achieved at week 12 and would have additional functional improvements, and that patients initially treated with placebo would respond

AC

CE

PT

ED

M

similarly to DVLX-treated patients in the double-blind phase.

ACCEPTED MANUSCRIPT 6

Methods

Study procedures This study was conducted at New York State Psychiatric Institute’s Depression Evaluation Service (DES) and approved by its Institutional Review Board (IRB). The goal was to enroll subjects with

CR IP T

PDD without a current MDD episode. We conducted a power analysis, based on the results of previous double-blind placebo controlled studies in dysthymic disorder. Assuming a

moderate effect size, with an 80% likelihood of finding a difference at significance p<.05, we calculated that a sample of 30 per arm was sufficient. Potential subjects were evaluated by

20

AN US

research psychiatrists using the Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I) , 21

and 24-item Hamilton Depression Rating Scale (HDRS-24) . Appropriate patients signed IRBapproved consent. A physical examination was performed and blood and urine samples were collected, including urine for pregnancy and toxicology. This study was registered at ClinicalTrials.gov (identifier: NCT01537068). It was funded by Pfizer, Inc. as an investigator-initiated

ED

M

trial.

Inclusion criteria were: age 20 to 65 years; baseline HDRS-24 score >=12, chronic (>= 2 years) depression. Exclusion criteria included: DSM-5 diagnosis of current major depression, bipolar

PT

disorder, schizophrenia or other psychotic disorder; dementia or other cognitive disorder; current (past 6 months) drug or alcohol use disorder; current psychoactive medication use; serious risk for

CE

suicide during the course of the study; unstable medical conditions; current or planned pregnancy;

AC

lack of capacity to consent to study participation.

Conduct of study: The study had two phases: 1) a 12-week double blind placebo controlled phase; and 2) a 12-week open treatment phase. Phase 1 RCT: randomization occurred approximately 1 week after intake assessment, with doubleblind treatment assigned based on a random number table generated by an independent

ACCEPTED MANUSCRIPT 7

biostatistician using SAS software (“Random = ranuni (seed)” procedure with a block of 4). Visits at weeks 1, 2, 4, 6, 8, and 12 assessed longitudinal course of presenting symptoms and functioning and emergence of new symptoms. Phase 2, Continuation Treatment: Patients initially treated with DVLX continued effective medication; non-responders were switched to other antidepressants. Patients initially treated with

up, patients were seen every 4 weeks for 12 weeks.

CR IP T

placebo began active DVLX treatment, up to a maximum of 100 mg/day. During open-label follow-

Clinicians administered the HDRS-24, Cornell Depression Rating Scale (CDRS)

22,23

, Clinical Global 24

Impression (CGI), and the Systemic Assessment for Treatment Emergent Events (SAFTEE) , while 25

(TCI)

26,27

28

, Social Adjustment Scale (SAS),

AN US

patients completed the Beck Depression Inventory (BDI) , Temperament and Character Inventory 29

and Ruminative Responses Scale (RRS) . Study

measures assessed depressive symptoms (BDI, HDRS, CDRS), general and social functioning (SAS, GAF, CGI), and ruminations (RRS). Tolerability was assessed at each visit by self-rated scales and clinician assessment. Clinicians rated adverse events (AEs) using the SAFTEE, and

receiving CGI-Improvement

30

M

rated each AE by duration and severity. Participants verbally re-consented at week 6. Participants scores of 6 (“much worse”) or 7 (“very much worse”) were withdrawn

PT

Drug Administration:

ED

from double-blind treatment.

DVLX (50 mg capsules) or matching placebo was begun at visit 2 (week 0); dose could be

CE

increased after 4 weeks to 100 mg q AM, if tolerated and still depressed (i.e. CGI-Improvement score >2). The blind was broken at Week 12, at which point DVLX was continued in those already

AC

taking it, or begun in those initially treated with placebo. Those still depressed on DVLX could be switched to another antidepressant.

Statistical Analyses The primary outcome measure was a priori defined as the HDRS-24 item total score. Efficacy was evaluated using chi square tests for response and remission. Paired T-test and repeated measures

ACCEPTED MANUSCRIPT 8

analysis of variance (ANOVA) were applied for changes in scale scores (HDRS, CDRS, GAF, CGI, QIDS-SR, PGI, SAS, SDS, TCI, BDI and RRS). Response was defined as: >50% decrease from baseline in HDRS-24 score and CGI-I score of 1 ("very much improved") or 2 ("much improved"). Remission was defined as: HDRS-17 score < 4 and HDRS item #1 (depressed mood) score = 0. (In this study, as in other PDD studies by our group

18,31

, we defined remission more rigorously than is

CR IP T

usually done in MDD trials (HDRS score<=6). Because patients with PDD often present with lower symptom severity, inclusion criteria allow a baseline minimum Hamilton Depression Rating Scale-24 item score>=12. If remission was defined as a score <=6 there would be no difference between response and remission for such patients.) Frequencies of side effects are reported. The last

AN US

observation was carried forward (LOCF) as an endpoint score for those who had missing data. Data analyses used IBM SPSS version 24 and SAS (9.4). Week 24 analyses followed a similar pattern. Analyses were run with both raw data and LOCF. Results did not change significantly. Therefore,

AC

CE

PT

ED

M

we report results from raw data.

ACCEPTED MANUSCRIPT 9

Results Sample (see Consort Diagram) Seventy-one subjects signed study consent, of whom 61 were assigned to receive blinded medication or placebo. Twelve eligible subjects did not start study medication for reasons described in Consort Diagram. Our intention-to-treat (ITT) group included 61 patients, of whom post-randomization data were available for 59 (two patients were randomized but

CR IP T

did not receive study medication). Of these 59 patients beginning treatment, 1 placebo-treated

patient dropped out after wk 2, and 2 after wk 6. Of desvenlafaxine-treated patients, 2 dropped out after wk 2, and 2 after wk 8. Table 1 shows demographic characteristics of desvenlafaxine (N=31) and placebo (N=30) treated subjects. Subjects were predominantly (54.1%) female, with mean age

AN US

37.9+13.1 years (range, 20-63); mostly Caucasian (62.3%, 38/61), and most (48/61 = 80%) had

graduated college. Half (30/60) were employed full time. Most (52/61 = 85%) were unmarried. Two thirds of subjects (35/55) had early-onset (before age 21) PDD. Forty-four per cent (27/61) had a lifetime anxiety disorder diagnosis, including GAD (3.2%; 2/61), social phobia (29.5%; 18/61), obsessive-compulsive disorder (9.8%; 6/61), and anxiety disorder NOS (3.3%; 2/61). Prior alcohol

M

use disorder was present in 9.8% (6/61) and prior drug abuse was present in 8.2% (5/61). History of eating disorders were found in 11.5% (7/61) of subjects (4 binge eating disorder, 3 bulimia). Prior

ED

MDD episodes were reported by 47.5% (29/61), with 19.7% (12/61) reporting one prior episode, and 27.9% (17/61) reporting two or more prior episodes. Of 29 patients who had previous MDD

PT

episodes, 22 had treatment history information; there was no difference between assignment to

CE

treatment and placebo groups (Pearson chi-square=0.029, df=1, p=0.864).

Mean DVLX or placebo (equivalent) dose at Week 12 was 96.5 mg/day (sd=12), with 7 taking 50

AC

mg/day, and 52 taking 100 mg/day. The active medication group had 95.5% compliance and mean final dose of 95 (sd = 15.3) mg/day of DVLX, with 3 people taking 50 mg/day and 27 taking 100 mg/day. The placebo group had 95.4% compliance (sd = 14.5); average final dose equivalence (50 mg/capsule) was 93 mg/day (sd = 17.6) of placebo, with 4 people taking 50 and 25 taking 100 mg/day. There was no association between treatment group and dose level and compliance (t=0.02, df=44, p=0.98).

ACCEPTED MANUSCRIPT 10

At week 24, 6 patients were not receiving DVLX. Three initially DVLX-treated patients switched to other antidepressants (two to bupropion, and one to duloxetine); two of the three initially-placebotreated patients did not take antidepressant medication in the follow-up phase, and one received escitalopram. Mean dose for patients remaining on DVLX was 94.0 (sd=22.5) mg/day and mean

CR IP T

dose for patients who switched from placebo to DVLX was 86.5 (sd=25) mg/day.

Average baseline scores for the rating scales used as outcome measures appear in Table 2. There

AN US

were no significant baseline differences between treatments.

Efficacy analyses Week 12 results Change on rating scales over time

M

Primary analysis. As shown in Table 2, HDRS-24 at 12-weeks, covaried for baseline HDRS24, did not significantly differentiate DVLX from placebo (6.53 + 3.98 vs. 8.24 + 4.96, F=3.33,

ED

df=1, p=.07).

Secondary analyses. Also, in Table 2, are results of analyses of each secondary outcome

PT

measure (HDRS-17, CDRS, BDI, GAF, CGI-Severity, QIDS-SR, RRS, TCI, SDS, and SAS) with randomization group (active drug or placebo) as the between subjects factor and time (Baseline and

CE

week 12 or LOCF ratings) as the within subjects factor. All measures of depressive symptoms and psychiatric functioning showed a significant main effect of Time, indicating that on average, subjects

AC

improved over time, regardless of treatment group. Of the various clinician and patient ratings, the QIDS-SR and the CGI-I-Pt showed significant treatment by time differences; and HDRS-17 scores (Figure 1) favored medication at week 12 at the trend level, while analyses of other measures did not differentiate the treatments.

Treatment response and remission at week 12 Of 59 subjects in the ITT sample who received study medication, response rate at week 12 was

ACCEPTED MANUSCRIPT 11

61.5% (16/26) for DVLX vs. 26.9% (7/26) for placebo (chi sq (df=1, n=52) = 6.32, p=.025); remitter rate was 26.7% (8/30) for DVLX vs. 17.2% (5/29) for placebo (chi sq (df=1,n=59)=.76, p=.38). All responders to active drug (n=16) were taking 100 mg. Responders to placebo (n=7) included 2 taking 50 mg equivalent, and 5 taking 100 mg equivalent. Non-responders to placebo (n=19) included taking 50 mg equivalent and 18 taking 100 mg equivalent; active drug non-responders

CR IP T

(n=9) were all taking 100 mg/d. Of 27 patients who had previous MDD episodes, 20 had treatment history information and there was no difference between treatment and placebo groups in terms of response at week 12 (Fisher’s exact test p=0.370 (2 sided)). Since most of the cells had counts less

AN US

than 5, we did not further evaluate this variable with other analyses.

Social adjustment and global functioning

At week 12, both treatment groups showed improvements in SAS, SDS and GAF but there was no drug-by-time effect, indicating there was no greater improvement in social functioning with DVLX

M

than placebo (Table 2).

Temperament

ED

The TCI includes four factors of temperament: Harm Avoidance (HA), Novelty Seeking (NS), Reward Dependence (RD), and Persistence (PS). At baseline, HA scores (23.50+7.47 for DVLX

PT

subjects; 21.39+7.20 for placebo subjects) were approximately 2 SD above community norms of 32

33

10.6+6.0 for men and 12.9+6.1 for women, similar to our prior findings.

After 12 weeks of

CE

treatment, there were no significant changes by group, time or group by time in HA, NS or RD,

AC

though there was a time effect for PS (Table 2).

Tolerability

There were no significant differences between placebo- and DVLX-treated subjects on the measures of cognitive functioning (MOTCS), sexual functioning (ASEX), or patient-perceived drugrelated cognitive impairment (ABNAS) and pain (BPI); (data not reported). Adverse events (AEs) as collected on the SAFTEE were reported by 19 of the 29 subjects on placebo (65.5%) and 29 of the

ACCEPTED MANUSCRIPT 12

30 (96.7%) subjects on DVLX (Table 3). The mean number of AEs for the DVLX group was 3.32 (sd=2.54) and for the placebo group 1.94 (sd=2.29) with the medication-treated group showing a trend for more AEs (t = -1.80, df = 42, p = 0.08). Most common side effects in DVLX-treated subjects were nausea (30%), dry mouth (23.3%), and dizziness, sweating, agitation and anorgasmia (each 16.7%). Most common side effects in placebo-treated subjects were decreased

CR IP T

libido (13.8%), GI upset, cognitive impairment, delayed orgasm, and nausea (each 10.3%). Side effects were generally mild to moderate. Only 1 subject (on DVLX) discontinued due to adverse events.

AN US

Week 24 results Continued DVLX

During the continuation period (Table 4, Figure 1), ratings comparing week 12 to week 24 showed significant changes in several patient-rated inventories, including depression (QIDS-SR and BDI), and ruminations (Ruminations Responses Scale (RRS)). Response and remission rates also

M

increased by Week 24 (Table 4). Of 8 DVLX remitters by Week 12, 5 (62.5%) remained in remission at week 24 (2 Week 12 remitters declined to continue medication, and 1 previous remitter

ED

was a nonresponder by Week 24).

PT

Placebo treated patients beginning DVLX

Patients initially treated with placebo who began open label DVLX at Week 12, generally showed

CE

significant improvement (Table 4) between weeks 12 and 24 in clinician-rated scales but not in selfrated scales. Compared to baseline, they had significant improvement in nearly all scales, with the

AC

exception of three temperament measures, RD, NS and PS. Unlike the patients continuing DVLX, there was not a significant improvement in HA with 12 weeks of DVLX treatment in this group.

Response and remission rates at week 24 Patients who continued DVLX had response rate of 82.6% and remission rate of 47.8% at 24 weeks Patients beginning DVLX in continuation phase had response rate of 83.3% and remission rate of

ACCEPTED MANUSCRIPT 13

50.0% at 24 weeks.

Temperament and social functioning (Table 4) At 24 weeks, the only temperamental factor that showed significant change from baseline was Harm Avoidance (HA) in the DVLXDVLX group. No temperamental factor changed significantly in

CR IP T

the PLACDVLX group. Functional impairment (SDS, SAS, GAF) continued to improve during the continuation phase for the DVLXDVLX group (Table 4). Fewer than half of subjects were within 1SD of the normal mean score for the SAS and HA, suggesting that continued impairment was

common. Remission was associated with normalization of SAS scores: mean SAS score at week

AN US

24 differed significantly between remitters (1.86 ± 0.32) and non-remitters (2.44 ± 0.64), p=0.001.

Remitters' SAS mean scores were within the normal community range. In contrast, responders who

AC

CE

PT

ED

M

were not remitters (N=16) had mean SAS scores of 2.36+0.56) suggesting continued impairment.

ACCEPTED MANUSCRIPT 14

Discussion

The DSM-5 recognized the importance of chronicity by creating the category of persistent depressive disorder (PDD).

11

Independent of cross-sectional severity, chronicity is a significant risk

factor for poor outcome, both in terms of continuing symptomatology and impaired functioning. Few

analyses

13,17

CR IP T

studies address the treatment specifically of the various forms of chronic depression. Recent

find efficacy data only on 5806 patients, data for acceptability of pharmacotherapy, 13

17

psychotherapy or combined treatment on 5348 , and safety data on 4769 . Furthermore, while antidepressant medications have widely been shown to be more effective in more severe

AN US

depression (and less effective in milder depression), there are limited data in non-major persistent depression. This situation is likely to continue now that the various forms of chronic depression (more severe chronic major depression, less severe dysthymic disorder, etc.) have been collapsed into the single DSM-5 PDD category. Yet even non-major PDD is associated with high levels of 10,34

psychosocial morbidity

. Hence, it is valuable to define the efficacy of ADMs in less severe PDD.

M

To our knowledge, this is the second reported double-blind trial for an SNRI medication in treatment 18

ED

of non-major persistent depressive disorder (PDD).

Efficacy phase: Our primary analysis of the 12-week double-blind phase of this DVLX study

PT

reached only trend significance (.07), so this was a negative study that does not support the use of DVLX as treatment for non-major PDD. This negative result could have occurred

CE

because DVLX is not effective in this population. Alternatively, relative to our other studies, this study had a relatively high improvement in patients assigned placebo,

18,31,35

power may

AC

have been insufficient or the negative result may have occurred due to sampling. Several positive secondary analyses suggest the value of this study may be in generating hypotheses for further studies. A moderate ES (.436) on the primary outcome measure suggests adequate power (>.8) to demonstrate efficacy with a sample size of 60 per treatment arm; the effect size of studies with DVLX in MDD approximate 0.3

36-39

so this is comparable. Of note, the patient-rated QIDS-SR

and patient-rated CGI-Improvement did differentiate medication from placebo in our study. Contrary

ACCEPTED MANUSCRIPT 15

to expectations, secondary measures of social functioning (SAS, SDS) and global outcome (GAF) did not show significant differences at week 12, nor did temperamental abnormalities (HA, RD, NS on TCI), though ruminations (RRS) showed a time effect.

In contrast, our duloxetine study showed more robust response in the initial (10 week) phase, with

40

below).

CR IP T

less additional improvement during the 12 week continuation period (see Continuation Phase

This potentially greater initial response to duloxetine might result from differing relative Ki

profiles for these two SNRI medications. Ki, or the inhibitory constant, is related to the

concentration of a drug needed to decrease the activity of a receptor or enzyme by half; a low Ki

AN US

level represents a high level of binding affinity, whereas a high Ki level implies that a relatively high concentration of a drug is required before the binding site is maximally occupied. Duloxetine potently inhibits binding to the human NE and 5-HT transporters, with Ki values of 7.5 and 0.8 nM, 41

respectively, and a Ki ratio of 9.

In comparison, desvenlafaxine has Ki values of 2953 for NE and

61.4 for 5-HT, and a Ki ratio of 48 (values similar to its parent compound venlafaxine: Ki 2480 for 19

M

NE and 82 for 5HT receptors, respectively). Whereas duloxetine appears to have both significant NRI and SRI activity, DVLX appears to have predominantly SRI activity, and more limited NRI

ED

activity. DVLX’s relatively lower NE reuptake effect might have contributed to this study’s less

PT

robust findings.

Another possible explanation for our study’s negative result relates to dosage. Whereas DVLX has

CE

been studied in doses up to 400 mg/day, our study respected the FDA approved maximum of 100 mg/day. Three subjects received doses up to 150 mg/day during open continuation treatment

AC

following the study due to nonresponse at 100 mg/day dose, two reporting further improvement of depressive symptoms subsequent to above-PDR-recommended dosing. In summarizing DVLX efficacy studies, Perry et al.

42

concluded that higher dose did not produce additional benefit;

however, we are unaware of studies of non-responders to DVLX 100 mg/day to assess benefit of increases above the currently FDA-approved range. Andrade

36

has suggested that DVLX might be

essentially an SSRI at lower doses, as has been found for venlafaxine, which requires dosing >150

ACCEPTED MANUSCRIPT 16 43

mg/day to elicit noradrenergic effects .

Finally, some

44,45

46

but not all studies suggest lower efficacy of antidepressants in milder major

depression. However most studies of non-major persistent depressive disorder (previously called 13,47

dysthymic disorder) suggest that medication is more effective thanplacebo

. The negative result

CR IP T

of this study does not strongly support either hypothesis; a larger sample, allowing for an analysis of the moderating effect of initial severity of PDD would be helpful in addressing this issue.

DVLX side effects were generally mild, only once resulting in discontinuation. Self-reports

AN US

(MOTCS, ABNAS, ASEX) confirmed the tolerability of DVLX. There was also high compliance, 95.5% for the medication group, perhaps another measure of tolerability.

Significant psychosocial morbidity and costs are associated with PDD regardless of subtype or 10,48

cross-sectional severity.

In this study, duration of illness averaged 12.8+13.5 years, with

M

significant impairment by numerous measures. Despite 67.8% of our participants having college or graduate degrees, most were unemployed (38.5%; 15/39) or working part-time (52.5%; 31/59).

ED

Most subjects (84.8%) were currently unmarried, suggesting impairment in interpersonal relationships. Mean score on the GAF of 63.3+7.3 (n=59), indicates a moderate degree of functional

PT

impairment, as do baseline mean scores on SAS of 2.50+0.5,12 weeks of over 2 SD above the 49

community mean (1.4+0.5, ), which remained 1.6 SD above this mean following DVLX (2.22+0.4

CE

LOCF) or placebo (2.29+0.5 LOCF) treatment. Similarly, SDS scores remained elevated posttreatment (10.32+7.27 for placebo and 11.80+7.56 for DVLX) compared to community norms 50

The literature

AC

(6.1+7.0).

51-54

about continued improvement in psychosocial functioning in PDD is

mixed. Of note, in this study, remission at week 24 was associated with normalization of SAS scores, underscoring its importance as a treatment goal.

Social functioning and Harm Avoidance (which has been associated with functional impairment) did show further improvement between weeks 12 and 24 for patients continuing DVLX, but most

ACCEPTED MANUSCRIPT 17

patients continued to have scores in the abnormal range. Since continued antidepressant 40

monotherapy alone generally does not lead to normalization of psychosocial functioning in PDD , alternative approaches should be investigated more thoroughly, including medication augmentation, combined medication-psychotherapy strategies, particularly including cognitive behavioral 55,56

therapy

and behavioral activation therapy

57,58

. Future studies should aim both to enhance

CR IP T

remission rates and improve psychosocial outcomes.

Open Continuation Phase: At week 24, most patients (62.5%) who achieved remission by week 12 on DVLX continued to be in remission. Additional improvements were noted in the continuation

AN US

DVLX sample in self-rated scales, including depression (the QIDS-SR and BDI), as well as the

RRS, suggesting benefits of continuation treatment. Findings for continuation treatment at week 24 were consistent with broad improvement—in core depressive scales, in measures of psychosocial impairment, and in the harm avoidance component of temperament as well as in ruminations. This suggests that additional improvement in these areas, at least with DVLX, may require longer

M

medication exposure. In contrast our recent study with duloxetine

40

showed little additional

improvement between acute treatment and continuation treatment (10 and 22 weeks respectively in 51-54

about continued improvement in psychosocial functioning in persistent

ED

that study). The literature

depression is mixed. The reasons for discrepant findings for two SNRI medications are unclear; it is

PT

possible that the maximum dose of each (120 mg/day for duloxetine vs. 100 mg/day for desvenlafaxine) may have differing potency in affecting neurotransmitter reuptake, duloxetine

CE

possibly having higher relative reuptake inhibition or other effects at doses used in these 19,41

AC

studies.

Study limitations and future directions This was a negative study and positive secondary analyses should not be overinterpreted. For one reason, the multiple secondary analyses presented were not corrected for multiple comparisons. Therefore, some significant findings could be due to chance. Further, the sample size of 59 limits power to detect differences smaller than “moderate”. The exclusion of acutely suicidal patients, and

ACCEPTED MANUSCRIPT 18

those with medical and psychiatric comorbidities, limits the generalizability of findings. Further study with a larger and more varied sample of individuals with PDD would allow analyses of moderating effects of baseline severity. Higher doses of DVLX could be studied. Future directions in PDD psychopharmacology should include comparative and long-term studies, perhaps following the STAR*D design

59,60

, in which a large cohort of PDD patients are followed for an extended period of

CR IP T

time and offered various treatments, including the commonly used medication classes such as

SSRIs, SNRIs, bupropion, and other antidepressant agents, as well as psychotherapy, including medication switches, augmentation, and other strategies to enhance response and remission,

particularly to improve social functioning. Because some patients with PDD might respond (or remit)

AN US

with doses above current FDA approved maximum of 100 mg/day, further study should address further treatment of patients with PDD who do not remit with conventional dosing.

Author disclosure: This study received funding from Pfizer, Inc., New York, NY. The study was performed at the New York State Psychiatric Institute/Columbia University Department of Psychiatry, New York, NY

ED

M

Role of funding: Funding from Pfizer, Inc. supported conduct of the Investigator Initiated Study (desvenlafaxine vs. placebo clinical trial and open-label continuation treatment) described in this paper, as well as funding for analysis of results. The paper was written without input from Pfizer, Inc. or its employees.

AC

CE

PT

Contributors: The authors had roles in this paper as follows: David J. Hellerstein, MD: design of study as PI, obtaining funding, treating patients, supervision of conduct of study and analyses of results, writing paper. Jonathan W. Stewart, M.D.: co-investigator in study, treating patients, assisting in interpretation of results, editing manuscript Bradley S. Peterson, M.D.: co-investigator in study, treating patients, assisting in interpretation of results, editing manuscript Vinushini Arunagiri, M.A.: literature review, drafting manuscript, assisting in interpretation of results Ying Chen, M.D.: data management, statistical analyses, describing statistical methods in paper Patrick J. McGrath, M.D.: co-investigator in study, treating patients, assisting in interpretation of results, editing manuscript

Acknowledgements: The authors would like to thank Deborah Deliyannides, MD, Mark Rodriguez, BA, and Donna O’Shea, RN, MS, for assistance in conduct of the study

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

M

AN US

CR IP T

19

ACCEPTED MANUSCRIPT 20

REFERENCES

8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.

AC

20.

CR IP T

7.

AN US

5. 6.

M

4.

ED

3.

PT

2.

Kessler RC, Chiu WT, Demler O, Merikangas KR, Walters EE. Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry. 2005;62(6):617-627. Berndt ER, Koran LM, Finkelstein SN, et al. Lost human capital from early-onset chronic depression. American Journal of Psychiatry. 2000;157:940-947. Rappaport MH, Clary C, Fayyad R, Endicott J. Quality of life impairment in depressive and anxiety disorders. American Journal of Psychiatry. 2005;162:1171-1178. Keller MB. Dysthymia in clinical practice:course, outcome and impact on the community. Acta Psychiatrica Scandinavica, Supplementum. 1994;89(s383):24-34. Friedman RA. Social impairment in dysthymia. Psychiatric Annals. 1993;23:632-637. Hays RD, Wells KB, Sherbourne CD, Rogers W, Spritzer K. Functioning and well-being outcomes of patients with depression compared with chronic general medical illnesses. Archives of general psychiatry. 1995;52(1):11-19. Kocsis JH, Zisook S, Davidson J, et al. Double-blind comparison of sertraline, imipramine, and placebo in the treatment of dysthymia: psychosocial outcomes. Am J Psychiatry. 1997;154(3):390-395. Klein DN, Shankman SA, Rose S. Ten-Year Prospective Follow-Up Study of the Naturalistic Course of Dysthymic Disorder and Double Depression. American Journal of Psychiatry. 2006;163(5):872-880. Klein DN, Schwartz JE, Rose S, Leader JB. Five-Year Course and Outcome of Dysthymic Disorder: A Prospective, Naturalistic Follow-Up Study. American Journal of Psychiatry. 2000;157(6):931-939. Hellerstein DJ, Agosti V, Bosi M, Black SR. Impairment in psychosocial functioning associated with dysthymic disorder in the NESARC study. J Affect Disord. 2010;127(1-3):84-88. Association AP. Diagnostic and statistical manual of mental disorders (DSM-5®). American Psychiatric Pub; 2013. Lima MS, Moncrieff J. Drugs versus placebo for dysthymia. Cochrane Database Syst Rev. 2000(4):CD001130. Kriston L, Wolff A, Westphal A, Hölzel LP, Härter M. Efficacy and acceptability of acute treatments for persistent depressive disorder: A network meta‐analysis. Depression and anxiety. 2014;31(8):621-630. Pampallona S, Bollini P, Tibaldi G, Kupelnick B, Munizza C. Combined pharmacotherapy and psychological treatment for depression: a systematic review. Archives of general psychiatry. 2004;61(7):714719. Dunner DL. Acute and maintenance treatment of chronic depression. The Journal of clinical psychiatry. 2000;62:10-16. Geddes JR, Carney SM, Davies C, et al. Relapse prevention with antidepressant drug treatment in depressive disorders: a systematic review. The Lancet. 2003;361(9358):653-661. Meister R, von Wolff A, Mohr H, et al. Comparative safety of pharmacologic treatments for persistent depressive disorder: a systematic review and network meta-analysis. PloS one. 2016;11(5):e0153380. Hellerstein DJ, Stewart JW, McGrath PJ, et al. A Randomized Controlled Trial of Duloxetine Versus Placebo in the Treatment of Nonmajor Chronic Depression. Journal of Clinical Psychiatry. 2012;73:984-991. Deecher DC, Beyer CE, Johnston G, et al. Desvenlafaxine succinate: a new serotonin and norepinephrine reuptake inhibitor. Journal of Pharmacology and Experimental Therapeutics. 2006;318(2):657665. First MB, Spitzer RL, Gibbon M, Williams JB. User's guide for the Structured clinical interview for DSM-IV axis I disorders SCID-I: clinician version. American Psychiatric Pub; 1997. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry. 1960;23:56-62. Mason BJ, Kocsis JH, Leon AC, et al. Measurement of severity and treatment response in dysthymia. Psychiatric Annals. 1993;23(11):625-631. Hellerstein DJ, Batchelder ST, Lee A, Borisovskaya M. Rating dysthymia: an assessment of the construct and content validity of the Cornell Dysthymia Rating Scale. J Affect Disord. 2002;71(1-3):85-96. Rabkin JG, Markowitz JS, Ocepek-Welikson K, Wager SS. General versus systematic inquiry about emergent clinical events with SAFTEE: implications for clinical research. LWW; 1992. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression. Arch Gen Psychiatry. 1961;4:561-571. Cloninger CR, Bayon C, Svrakic DM. Measurement of temperament and character in mood disorders: a model of fundamental states as personality types. J Affective Disorders. 1998;51:21-32.

CE

1.

21. 22. 23. 24. 25. 26.

ACCEPTED MANUSCRIPT 21

33. 34. 35. 36. 37.

38. 39. 40. 41. 42. 43. 44. 45.

AC

46.

CR IP T

32.

AN US

31.

M

30.

ED

29.

PT

28.

Cloninger CR, Svrakic DM, Przybeck TR. A psychobiological model of temperament and character. 50. 1993:975-990. Weissman MM, Bothwell S. Assessment of social adjustment by patient self-report. Arch Gen Psychiatry. 1976;33(9):1111-1115. Treynor W, Gonzalez, R., Nolen-Hoeksema, S. Rumination Reconsidered: A Psychometric Analysis. Cognitive Therapy and Research. 2003;27:247-259. Guy We. ECDEU Assessment Manual for Psychopharmacology: Publication ADM 76-338. Rockville, MD, US Department of Health, Education, and Welfare. 1976. Hellerstein DJ, Batchelder ST, Hyler S, et al. Escitalopram versus placebo in the treatment of dysthymic disorder. International Clinical Psychopharmacology. 2010;25:143-148. Cloninger CR, Przybeck TR, Svrakic DM. The Tridimensional Personality Questionnaire: U.S. normative data. Psychological Reports. 1991;69:1047-1057. Hellerstein DJ, Kocsis JH, Chapman D, Stewart JW, Harrison W. Double-blind comparison of sertraline, imipramine, and placebo in the treatment of dysthymia: effects on personality. Am J Psychiatry. 2000;157(9):1436-1444. Blanco C, Okuda M, Markowitz JC, Liu SM, Grant BF, Hasin DS. The epidemiology of chronic major depressive disorder and dysthymic disorder: results from the National Epidemiologic Survey on Alcohol and Related Conditions. J Clin Psychiatry. 2010;71(12):1645-1656. Hellerstein DJ, Yanowitch P, Rosenthal J, et al. A randomized double-blind study of fluoxetine versus placebo in the treatment of dysthymia. Am J Psychiatry. 1993;150(8):1169-1175. Andrade C. Desvenlafaxine. Indian journal of psychiatry. 2009;51(4):320. Tourian KA, Padmanabhan SK, Groark J, Brisard C, Farrington D. Desvenlafaxine 50 and 100 mg/d in the treatment of major depressive disorder: an 8-week, phase III, multicenter, randomized, double-blind, placebo-controlled, parallel-group trial and a post hoc pooled analysis of three studies. Clinical therapeutics. 2009;31:1405-1423. Thase ME, Kornstein SG, Germain J-M, Jiang Q, Guico-Pabia C, Ninan PT. An integrated analysis of the efficacy of desvenlafaxine compared with placebo in patients with major depressive disorder. CNS spectrums. 2009;14(03):144-154. Lieberman DZ, Montgomery SA, Tourian KA, et al. A pooled analysis of two placebo-controlled trials of desvenlafaxine in major depressive disorder. International clinical psychopharmacology. 2008;23(4):188-197. Hellerstein DJ, Hunnicutt-Ferguson K, Stewart JW, et al. Do social functioning and symptoms improve with continuation antidepressant treatment of persistent depressive disorder? An observational study. Journal of Affective Disorders. 2017;210:258-264. Bymaster FP, Dreshfield-Ahmad LJ, Threlkeld PG, et al. Comparative affinity of duloxetine and venlafaxine for serotonin and norepinephrine transporters in vitro and in vivo, human serotonin receptor subtypes, and other neuronal receptors. Neuropsychopharmacology. 2001;25(6):871-880. Perry R, Cassagnol M. Desvenlafaxine: a new serotonin-norepinephrine reuptake inhibitor for the treatment of adults with major depressive disorder. Clinical therapeutics. 2009;31:1374-1404. Preskorn SH. Two in one: The venlafaxine story. J Pract Psychiatry Behav Health. 1999;5(6):346-350. Fournier JC, DeRubeis RJ, Hollon SD, et al. Antidepressant drug effects and depression severity: a patient-level meta-analysis. Jama. 2010;303(1):47-53. Kirsch I, Deacon BJ, Huedo-Medina TB, Scoboria A, Moore TJ, Johnson BT. Initial severity and antidepressant benefits: a meta-analysis of data submitted to the Food and Drug Administration. PLoS medicine. 2008;5(2):e45. Stewart JA, Deliyannides DA, Hellerstein DJ, McGrath PJ, Stewart JW. Can people with nonsevere major depression benefit from antidepressant medication? The Journal of clinical psychiatry. 2012;73(4):518525. de Lima MS, Hotoph M, Wessely S. The efficacy of drug treatments for dysthymia: a systematic review and meta-analysis. Psychol Med. 1999;29(6):1273-1289. Agosti V, Hellerstein, DJ. Chronic major depression in NESARC: employment status compared to Acute major depression and general population. unpublished. 2012. Weissman MM, Prusoff BA, Thompson WD, Harding PS, Myers JK. Social adjustment by self-report in a community sample and in psychiatric outpatients. The Journal of Nervous and Mental Disease. 1978;166(5):317-326. Leon AC, Olfson M, Portera L, Farber L, Sheehan DV. Assessing psychiatric impairment in primary care with the Sheehan Disability Scale. The international journal of psychiatry in medicine. 1997;27(2):93-105. Friedman RA, Markowitz JC, Parides M, Gniwesch L, Kocsis JH. Six months of desipramine for

CE

27.

47. 48. 49. 50. 51.

ACCEPTED MANUSCRIPT 22

53. 54. 55. 56. 57. 58.

AC

CE

PT

ED

M

AN US

59. 60.

dysthymia: can dysthymic patients achieve normal social functioning? J Affect Disord. 1999;54(3):283-286. Kocsis J, Schatzberg A, Rush A, et al. Psychosocial outcomes following long-term, double-blind treatment of chronic depression with sertraline vs placebo. Archives if General Psychiatry. 2002;59:723-728. Kocsis JH, Friedman RA, Markowitz JC, et al. Maintenance therapy for chronic depression. A controlled clinical trial of desipramine. Arch Gen Psychiatry. 1996;53(9):769-774; discussion 775-766. Rhebergen D, Beekman A, de Graaf R, et al. Trajectories of recovery of social and physical functioning in major depression, dysthymic disorder and double depression: a 3-year follow-up. Journal of Affective Disorders. 2010;124:148-156. McCullough JP, Jr. Treatment for chronic depression using Cognitive Behavioral Analysis System of Psychotherapy (CBASP). J Clin Psychol. 2003;59(8):833-846. Markowitz JC. Interpersonal psychotherapy for dysthymic disorder. American Psychiatric Pub; 2004. Erickson G, Hellerstein DJ. Behavioral activation therapy for remediating persistent social deficits in medication-responsive chronic depression. J Psychiatr Pract. 2011;17(3):161-169. Hellerstein DJ, Erickson G, Stewart J, et al. Behavioral activation therapy for return to work in medication-responsive chronic depression with persistent psychosocial dysfunction. Comprehensive Psychiatry. 2015;57:140-147. Rush AJ. STAR* D: what have we learned? The American journal of psychiatry. 2007;164(2):201. Gilmer WS, Gollan JK, Wisniewski SR, et al. Does the duration of index episode affect the treatment outcome of major depressive disorder? A STAR* D report. The Journal of clinical psychiatry. 2008;69(8):1,478-1256.

CR IP T

52.

ACCEPTED MANUSCRIPT 23

Table 1. Demographic characteristics overall and by treatment group (n=61) Treatment Group

Variables

n

DVLX to DVLX (n=31)

Mean (SD) or %

n

Demographics Gender

Mean (SD) or %

Placebo to DVLX (n=30) n

Mean (SD) or %

Male

28

45.9%

12

38.7%

16

53.3%

Female

33

54.1%

19

61.3%

14

46.7%

Marital 43

70.5%

Married

9

14.8%

Separated

2

3.3%

Divorced

7

11.5%

Race

Caucasian

38

Asian

6

Other

7

Ethnicity/Hispanic

54

PT

Non-Hispanic Hispanic

7

CE

Employment Status

16.4%

71.0%

21

70.0%

6

19.4%

3

10.0%

0

0.0%

2

6.7%

3

9.7%

4

13.3%

p-value

0.252

0.369

0.161

3

9.7%

7

23.3%

62.3%

19

61.3%

19

63.3%

9.8%

3

9.7%

3

10.0%

11.5%

6

19.4%

1

3.3%

M

10

ED

Black/AA

22

AN US

Single

Diff between groups

CR IP T

Total Sample (n=61)

0.246

88.5%

26

83.9%

28

93.3%

11.5%

5

16.1%

2

6.7% 0.067

30

50.0%

21

67.7%

9

31.0%

Employed part time for pay

15

25.0%

6

19.4%

9

31.0%

Student

3

5.0%

1

3.2%

2

6.9%

Unemployed (< 6months)

5

8.3%

1

3.2%

4

13.8%

Unemployed (> 6months)

7

11.7%

2

6.5%

5

17.2%

Age

61

38.4 (13.0)

31

40.1 (14.1)

30

36.7 (11.7)

0.305

Education

61

15.2 (3.7)

31

15.7 (3.9)

30

14.7 (3.4)

0.270

AC

Employed full time for pay

a

Baseline differences are assessed using t-tests for continuous measures and chi-square test for categorical measures

a

ACCEPTED MANUSCRIPT 24

Table 2. Primary and secondary outcome measures for subjects receiving desvenlafaxine (DVLX; n = 31) versus those receiving placebo (n = 30), with results of repeated measurement ANOVA comparing Week 0 and Week12. Measu res

Week 0 Placebo n=30

SAS RRS TCIRD TCINS TCI-PS

(SD )

Me an

(SD )

Me an

(SD )

20. 03 33. 83 --

4.4 9 11. 04 --

20. 29 36. 52 --

5.0 3 10. 41 --

11. 58 18. 65 2.8 8

6.9 6 11. 41 0.9 5

9.3 5 19. 62 2.2 3

5.8 0 12. 71 0.9 5

0.007

11. 63 19. 17 2.5 1 53. 57 15. 59 19. 72 4.3 5 20. 86

4.0 3 7.5 0 .45

13. 19 21. 07 2.4 7 55. 28 14. 84 18. 21 4.2 9 23. 71

4.0 8 8.3 5 .45

8.0 8 12. 82 2.2 9 41. 78 15. 0 19. 96 4.1 9 20. 27

4.9 2 9.2 0 .46

7.3 5 11. 64 2.2 2 43. 11 15. 92 18. 15 4.6 5 20. 35

4.4 2 6.4 7 .41

0.005

11. 49 4.1 8 4.5 0 1.8 8 8.2 7

0.014

12. 90 4.3 6 5.0 2 2.0 4 7.6 3

8.4 9 4.3 9 4.7 0 2.0 4 7.4 4

0.013

0.6 96 0.6 10 --

5 0 5 0 5 0

0.046

0.5 81 0.5 20 0.2 39 0.5 34 0.0 03 0.0 29 0.0 99 0.1 32

5 0 5 1 4 8 4 6 4 8 4 8 4 8 4 8

0.091

AN US

M 13. 39 3.9 2 5.2 5 1.6 7 6.6 8

P

CR IP T

Me an

AC

TCIHA

DVLX n=26

(SD )

ED

Patient Rating s QIDSSR BDI

Placebo n=26

Effect Size (Partial Eta Squared) * Treatm Tim D Time x ent e F Treatm ent

Me an

PT

CGI-I

DVLX n=31

CE

Clinicia n Rating s HDRS24 CDRS

Week 12

0.109

0.003 0.003

0.003 0.022 0.001 0.002

0.009

--

0.034 0.004 <0.001 0.078 0.031 0.041 0.024

0.1 12 0.5 13 0.0 17

0.0 30 0.1 87 0.6 63 0.9 81 0.0 49 0.2 20 0.1 58 0.2 82

Key to abbreviations: HDRS: Hamilton Depression Rating Scale, CDRS: Cornell Dysthymia Rating Scale, BDI: Beck Depression Inventory, SAS: Social Adjustment Scale, RRS: Ruminative Responses Scale, TCI: Temperament and Character Inventory, subscales: HA: Harm Avoidance; RD: Reward Dependence; NS: Novelty Seeking; PS: Persistence; CGI-I-PT: CGI-Improvement, patient-rated. * The Effect sizes reported in this table are Partial Eta Squared: 0.01 (small), 0.06 (medium) 0.14 (large)

ACCEPTED MANUSCRIPT 25 Table 3. Adverse events observed with desvenlafaxine (DVLX) or initial placebo followed with DVLX treatment

AC

CE

PT

ED

M

AN US

CR IP T

*Pa tien DVLX, n (%) Placebo, n (%) ts AE Week 12 Week 24 Week 12 Week 24* initi N=30 N=24 N=29 N=23 ally rec Daytime sleepiness 11 (36.7) 5 (20.8) 6 (17.2) 1 (4.4) eivi Decreased sleep 10 (33.3) 9 (37.5) 4 (13.8) 3 (13.0) ng pla Headache 9 (30.0) 4 (16.7) 4 (13.8) 1 (4.4) ceb Nausea 9 (30.0) 8 (33.3) 3 (10.3) 3 (13.0) o Dry mouth 7 (23.3) 5 (20.8) 2 (6.9) 1 (4.4) (we ek Agitation 5 (16.7) 2 (8.4) 0 0 0Dizziness 5 (16.7) 4 (16.7) 0 1 (4.4) 12) Sweating 5 (16.7) 2 (8.4) 2 (6.9) 2 (8.8) wer e Anorgasmia 4 (13.3) 3 (12.5) 2 (6.9) 1 (4.4) trea Decreased appetite 4 (13.3) 4 (16.7) 0 0 ted with Gastrointestinal upset 4 (13.3) 2 (8.4) 3 (10.3) 3 (13.0) des Constipation 3 (10.0) 3 (12.5) 2 (6.9) 2 (8.8) ven Vivid dreams 3 (10.0) 3 (12.5) 0 0 lafa xin Cognitive impairment 2 (6.7) 2 (8.4) 3 (10.3) 5 (21.7) e Decreased concentration 2 (6.7) 1 (4.2) 1 (3.5) 0 (DV Decreased libido 2 (6.7) 2 (8.4) 4 (13.8) 6 (26.1) LX) fro Fatigue 2 (6.7) 2 (8.4) 1 (3.5) 1 (4.4) m Increase of appetite 2 (6.7) 1 (4.2) 1 (3.5) 0 we Palpitations 2 (6.7) 2 (8.4) 0 0 eks 13Yawning 2 (6.7) 2 (8.4) 0 0 24. Anxiety 1 (3.3) 3 (12.5) 1 (3.5) 1 (4.4) Not Delayed orgasm 1 (3.3) 1 (4.2) 3 (10.3) 0 e: In Muscle spasms 1 (3.3) 5 (20.8) 1 (3.5) 2 (8.8) the Spacey feeling 1 (3.3) 1 (4.2) 3 (10.3) 0 foll Tingling 1 (3.3) 1 (4.2) 1 (3.5) 1 (4.4) owSeizure 0 0 1 (3.5) 1 (4.4) up peri od, 9 patients reported no side effects. Among them, 1 was in the continued DVLX treatment group and 8 were in the placeboDVLX treatment group.

ACCEPTED MANUSCRIPT 26

Table 4. Primary and secondary outcome measures for patients treated with desvenlafaxine (DVLX) in phase 1 and phase 2 and patients treated with placebo in phase 1 and DVLX in phase 2 (n = 47).

DVLX  DVLX Group (n=24) WK 12

WK24

Me an (SD )

Me an (SD )

Mean (SD)

Phase II: Open Label WK12 vs WK24 Eff d p ect f Siz e*

Overall Changes Pre-tx vs Wk24 Eff ect Siz e

d f

RRs

TCIRD

Eff ect Siz e

d f

p

<0.

<0.

33.7

20.35(1

12.57(

0.5

2

<0.

0.7

2

<0.

001

(11.38)

0.92)

8.99)

13

2

001

84

2

001

0.0

2

0.1

0.8

2

<0.

20.17(

12.52(1

6.17(1.

0.5

2

<0.

0.8

2

67

4

(1.14)

90

3

45

54

3

001

1.07)

.28)

16)

66

2

001

35

2

001

(1.0

(1.2

5)

5)

37.

19.

15.54(

0.1

2

0.0

0.7

2

12.06)

45

3

60

52

3

08

33

(11.

(11.

64)

82)

4.0

2.6

2.21

0.1

2

0.0

0.8

2

4

7

(0.83)

63

3

46

07

3

(0.6

(1.0

9)

1)

7.5

5.92

0.2

2

0.0

0.8

2

25

4

(4.35)

32

3

15

62

(4.4

(4.4 6) 11.

54

96

(8.9

(6.4

<0.

4.22

3.04

2.35

0.3

2

0.0

0.8

2

<0.

001

(0.74)

(0.88)

0(.94)

60

2

01

05

2

001

<0.

12.73

8.68

7.39

0.0

2

0.2

0.6

2

<0.

3

001

(3.88)

(4.75)

(4.27)

68

2

17

20

1

001

ED

5) 21.

PT

13.

7.58

0.3

2

0.0

0.7

2

<0.

19.57

12.91(8

10.17(

0.1

2

0.1

0.5

2

<0.

(6.08)

74

3

01

53

3

001

(8.37)

.95)

6.99)

13

2

09

68

2

001

1)

8)

2.4

2.2

2.08

0.1

2

0.0

0.3

2

0.0

2.49

2.30

2.26

0.0

2

0.9

0.2

2

0.0

2

1

(0.46)

47

3

58

43

3

02

(0.48)

(0.50)

(0.69)

01

2

13

71

2

14

AC

SAS

p

Mean (SD)

7.63

CE

BDI

d f

Mean (SD)

9.2

Patient Ratings QID SSR

Eff ect Siz e*

p

AN US

CGI -S

Mean (SD)

WK24

M

CD RS

20.

Overall Changes Pre-tx vs Wk24

WK12

Clinician Ratings HRS D24

Phase II: Open Label WK12 vs WK24

Pre-tx

CR IP T

Pre -tx

Placebo  DVLX Group (n=23)

(0.3

(0.4

9)

3)

54.

41.

37.14

0.2

2

0.0

0.7

2

<0.

52.76(

39.559(

37.14(

0.1

1

0.0

0.6

2

<0.

(8.73)

10

1

30

66

1

001

13.26)

11.90)

10.65)

68

9

70

56

0

001

50

64

(8.2

(10.

3)

36)

15.

16.

15.92

0.0

2

0.1

<0.

2

1.0

15.61

14.52

15.19

0.0

2

0.3

0.0

2

0.5

92

71

(4.44)

90

3

45

001

3

00

(4.82)

(3.99)

(4.32)

67

2

03

02

2

04

(3.9

(4.0

ACCEPTED MANUSCRIPT 27

TCIPS

TCIHA

4)

17.

17.

18.20

0.0

2

0.6

0.0

2

0.6

18.62

19.87

20.48

0.0

2

0.4

0.2

2

0.0

86

93

(4.47)

11

3

15

08

3

65

(4.55)

(5.08)

(5.76)

05

2

58

63

2

30

(3.8

(4.3

1)

5)

4.6

4.8

5.0

0.0

2

0.5

0.0

2

0.1

4.13

4.13

3.91

0.0

2

0.2

0.0

2

0.4

3

3

(1.91)

15

3

58

87

3

53

(1.96)

(1.60)

(1.78)

63

2

85

12

2

47

(2.0

(1.8

2)

3)

21.

19.

18.79

0.0

2

0.3

0.2

2

0.0

21.87

21.04

(8.96)

40

3

36

42

3

12

(7.81)

(6.95)

96

71

(7.3

(8.2

2)

4)

CR IP T

TCINS

3)

20.62

0.0

2

0.5

0.2

2

0.2

(6.89)

06

2

37

66

2

38

AN US

Key to abbreviations: Pre-tx: Pre-treatment HDRS: Hamilton Depression Rating Scale, CDRS: Cornell Dysthymia Rating Scale, BDI: Beck Depression Inventory, CGI-I,CGI-S: clinician-rated Clinical Global Impression, Severity of Illness, SAS: Social Adjustment Scale, TCI: Temperament and Character Inventory, subscales: HA: Harm Avoidance; RD: Reward Dependence; NS: Novelty Seeking; PS: Persistence

AC

CE

PT

ED

M

* The Effect sizes reported in this table are Cohen’s D: 0.2 (small), 0.5 (medium) and 0.8 (large)

ACCEPTED MANUSCRIPT 28

Figures Consort Diagram: Desvenlafaxine (DVLX) study consort diagram (medication)

Enrollment

CR IP T

Assessed for eligibility (n=350)

Not eligible (n= 279)

Excluded (n= 10)  No longer met inclusion criteria (n= 1)  Declined to participate (n= 5)  Other reasons (n= 4) 3 lost to follow up, 1 began a new job

Eligible (n= 71)

AN US

Randomized (n= 61)

Allocation (wk RZ) Allocated to Placebo (n= 30) Received allocated intervention (n=29) Did not take allocated intervention (n=1)

M

Allocated to Active DVLX (n=31) Received allocated intervention (n=30) Did not take allocated intervention (n=1)

CE

PT

Lost to follow-up (n= 2) Discontinued intervention Patient request (n= 1)

ED

End of Acute Phase (wk 12)

AC

Analyzed (n= 26) Excluded from analysis (n=0)

Lost to follow-up (n= 1) Discontinued intervention Patient request (n= 1) Side effects (n= 1) No improvement (n=1)

Analysis Analyzed (n= 26) Excluded from analysis (n=0)

End of Continuation Phase (wk 24)

Did not finish follow up (n=3) Two placebo responders One placebo non-responder

Did not finish follow up (n=2) Both responded at week16 and never came back for week24 evaluation.

Continuation Phase Analysis Analyzed (n= 23) Placebo  DVLX (n=23)

Excluded from analysis (n=0)

Analyzed (n= 24) DVLX  DVLX (n=24) Excluded from analysis (n=0)

ACCEPTED MANUSCRIPT 29

Figure 1. Weekly HDRS-24 scores for desvenlafaxine (DVLX) vs. placebo treated subjects 30

Placebo treated patients began active medication at week12

CR IP T

20

15

10

AN US

Hamd24 Mean ± SD

25

5

0 6

AC

CE

PT

ED

Treatment

12

M

0

18

Week

PLAC -> DVLX

DVLX -> DVLX

24