2. Journal of Affective Disorders
Official Journal of the International Society for Affective Disorders
ā¢ Publishing since 1979
ā¢ Editors-in-chief: P. Brambilla,
PhD, MD & J. C. Soares, MD,
PhD
ā¢ Impact Factor: 6.533
ā¢ Peer reviewed, English language
ā¢ Frequency: Biweekly
3. Title
ā¢ Acceptability of escitalopram versus duloxetine in
outpatients with depression who did not respond to initial
second-generation antidepressants: A randomized,
parallel-group, non-inferiority trial
4.
5. ā¢ Received 8 July 2020;
ā¢ Received in revised form 30 November 2020;
ā¢ Accepted 23 December 2020
ā¢ Available online 5 January 2021
6. Registration number and name of trial registry
ā¢ Trial registration: UMINāCTR(UMIN000012367),
registered on December 1st, 2013 and last updated on
April 4th, 2017.
7. Introduction
ā¢ Only 50% of patients treated with antidepressants respond, and only one
third achieve remission (Kato et al., 2018) .
ā¢ Study shows that escitalopram was safe, well tolerated, and reduced the risk
of relapse in long-term treatments (Rapaport et al., 2004)
ā¢ The efficacy and safety of duloxetine in the acute treatment of MDD have
also been established in several placebo-controlled studies (Detke et al.,
2002a, 2002b; Nemeroff et al., 2002)
8. Introduction
ā¢ Acceptability, namely, the continuity of treatment, is a
potential outcome measure of clinical effectiveness.
ā¢ Research suggests that a continuous treatment with an
antidepressant for more than 6 months reduces the risk of
recurrence two years after recovery.(Kim et al., 2011;
Melfi et al., 1998)
9. Introduction
ā¢ It is unclear which specific antidepressant should be used
first to the patients with MDD who failed to respond to the
initial drug. Available studies do not yet compare
escitalopram or duloxetine directly as second step
treatment
10. Aim
ā¢ To compare acceptability between escitalopram, and
duloxetine, among patients with MDD whose responses to
initial antidepressant therapy were suboptimal in a
pragmatic design.
11. hypothesis
ā¢ escitalopram is non-inferior in acceptability to duloxetine
as a second-line treatment among patients with MDD.
12. Methods
ā¢ study design : Randomized, flexible-dose, parallel-group,
multi-center trial.
ā¢ Stratified block randomization strategy to obtain an
equal distribution between Group A and Group B in all
study sites using a centralized computer assignment
program.
14. Inclusion Criteria
ā¢ Fulfill criteria for major depressive disorder, as defined by DSM-IV criteria for single or recurrent
without psychotic features, as determined by clinical assessment by treating psychiatrist and
confirmed by the Mini-International Neuropsychiatric Interview (M.I.N.I.).
ā¢ Aged 20-65 years at screening.
ā¢ Participants who have been treated with therapeutic dose of SSRI (sertraline, paroxetine or
fluvoxamine) or SNRI (milnacipran) or NaSSA (mirtazapine) for at least 3 weeks.
ā¢ Depression symptoms of at least moderate severity based on Clinical Global Impression of
Severity (CGI-S) score ā„ 4, meaning moderate to severe depression.
ā¢ Major depressive disorder is the primary diagnosis for the treatment and treating psychiatrist has
judged study drug (i.e. escitalopram or duloxetine) to be appropriate for prescribing.
ā¢ Competent and able to understand the meaning of the observation, evaluation and clinical
examination in the judgment of the treating psychiatrist.
ā¢ Competent and able to give their own informed consent.
ā¢ Available on the phone for assessments.
15. Exclusion Criteria
ā¢ Did not respond to two or more adequate antidepressants (each for at least 4
weeks with therapeutic dose) during current depressive episode judged by the
study physician.
ā¢ Comorbid psychiatric condition (DSM-IV axis I) other than major depressive
disorder that is considered as the primary diagnosis within one year of screening.
ā¢ History of bipolar disorder, schizophrenia, or other psychotic disorder at screening
by the study physician.
ā¢ History of substance abuse/dependence within one year of screening, except
caffeine and nicotine.
ā¢ Have an Axis II disorder that, judged by study physician, would interfere with
compliance with the study protocol.
ā¢ Did not respond to escitalopram or duloxetine on maximum dose for at least four
weeks during the past depressive episode.
ā¢ Women who are currently pregnant or breastfeeding.
16. Exclusion Criteria
ā¢ Participants who are judged by the study physician to be at serious risk for harm
to self or others.
ā¢ Participants who are judged by the study physician to have serious and/or
unstable illness including problems in liver, kidney, respiratory system,
hematological, endocrine system, or CNS, or traumatic brain injury.
ā¢ Have a serious or unstable cardiovascular illness (including severe arrhythmia
with bradycardia, prescribed with drugs known to cause QTc prolongation,
congestive heart failure, hypokalemia), or clinically significant ECG abnormality
(male: QTc>450ms, female: QTc>470ms).
ā¢ Ongoing treatment with MAO inhibitors within two weeks of cessation of
treatment.
ā¢ Have uncontrolled closed-angle glaucoma.
ā¢ Ongoing treatment with pimozide .
ā¢ Participants who are judged by the study physician to be inappropriate to
17. interim analysis
ā¢ As per initial protocol of the study :
ā¢ The primary endpoint was the rate of discontinuation for any
reason at week 52 & sample size 242 patients.
ā¢ As enrollment was far behind schedule.
ā¢ In the interim analysis when 100 cases had been enrolled, the
discontinuation rates of the pooled data at week 8, 16 and 52
were 15%, 21% and 45%, respectively.
18. interim analysis
ā¢ The researchers were not allowed to access the data and the pooled
discontinuation rates were calculated by the data manager, who was blinded
to drug allocation of each individual data.
ā¢ An independent statistician, who was also blind to drug allocation, proposed
that the statistical power is maintained while reducing the sample size by
changing the primary endpoint to discontinuation rate at week 8
19. interim analysis
Proposal of test design change submitted to the data safety
monitoring committee (DSMC), which is a third-party organization independent
from the researchers.
Assuming the discontinuation rates at week 8 in the two groups were 15% and
20%, respectively, and the noninferiority margin was set to 10% and the
significance level was one-sided 5%, the number of cases with 80% power was
calculated as 80 in each group.
Considering both feasibility and scientific validity, the DSMC and finally the
ethics committee approved the study design change
20. Subjects
ā¢ Recruitment from
ā¢ Keio and Showa university hospitals,with their branch hospitals and clinics,
ā¢ National Center of Neurology and Psychiatry (NCNP).
ā¢ The study protocol was approved by an independent ethics committee at
each study site.
ā¢ All patients provided written informed consent prior to any study procedure
21. Drug allocation and switching
ā¢ Subjects was randomly allocated to one of the study
medications, and the study drug gradually replaced their
current antidepressant by week 4 in Step 1.
ā¢ The drug switching scheme was not specified and was left
to the treating psychiatristās discretion throughout the trial.
22. Screening, and Baseline assessment
ā¢ At the baseline within 2 weeks from screening,
ā¢ onsite :
ā¢ Severity of depression : 16 items Quick Inventory of Depressive Symptomatology ā Self-
Reported (QIDS-SR16) and CGI-S
ā¢ Quality of life : European Quality of Life Questionnaire-5 Dimension (EQ-5D).
ā¢ Telephonically Central raters blinded to the treatment allocation evaluated:
ā¢ Depression severity : Patient Healthcare Questionnaire ā 9 items (PHQ-9)
ā¢ Functional impairment level : Sheehan Disability Scale (SDS)
ā¢ The inter-rater reliability of the trained central raters had been established .
23. trial schedule
ā¢ The protocol encourages all patients to receive an
adequate dose of the allocated medication using a flexible
dose schedule.
ā¢ Escitalopram 10ā20 mg/day starting from 10 mg/day
ā¢ Duloxetine 40ā60 mg/day starting from 20 mg/day
ā¢ Drug dose level was in accordance with the approved
dosage in Japan.
24. trial schedule
ā¢ End of Step 1:
Patients with CGI-S score ā„4 and CGI-I score ā„4
Intolerant to the drug
ā¢ Discontinued from the currently allocated medication and
switched to the other study medication for the subsequent
treatment.
ā¢ Those who did not fulfill the switching criteria were
encouraged to maintain the current medication.
25. trial schedule
ā¢ Patients continued another 8-week treatment (Step 2) except for
those with clear intolerance or significant clinical worsening based
on CGI-I score ā„6 at two consecutive visits during Step 1, which
was treated as discontinuation of the trial.
ā¢ After week 16, patients were followed up to week 52, while
adherence, efficacy, adverse events, and health outcomes were
assessed
26. Primary outcome measures
ā¢ Primary endpoint: rate of discontinuation for any reason during Step 1.
ā¢ The discontinuation divided into 7 subgroups based on its reason.
1. Intolerance verified by the scores of FIBSER (Frequency, Intensity, and Burden of Side
Effects Rating).
2 . Voluntarily: complaints from the patientās side.
3 . Worsening of symptoms: defined as CGI-I score ā„6 in two consecutive visits.
4 . Withdrawal of consent
5 . Other safety issues: such as gestation, accidents, or admission to hospital with a physical
illness.
6 . Unable to continue treatment: inevitable reasons such as moving.
7 . Other reasons
27. Secondary outcome measures
ā¢ The rates of discontinuation of the trial when both study
drugs failed at week 16 and week 52.
28. Secondary outcome measures
ā¢ Efficacy
ā¢ Change in the severity of depression measured by PHQ-9, CGI-S, and QIDS-
SR16 at week 8, 16 and 52.
ā¢ The proportion of responders at week 8, 16 and 52. Response is defined as a
50% or greater reduction from the baseline PHQ-9 score.
ā¢ The proportion of remitters at week 8, 16 and 52. Remission is defined as the
PHQ-9 score ā¤4.
29. Secondary outcome measures
ā¢ Health Outcomes
ā¢ Change in quality of life measured by the EQ-5D at week
8, 16 and 52.
ā¢ Change in functional impairment measured by the SDS at
week 8, 16 and 52.
30. Secondary outcome measures
ā¢ Safety
ā¢ The global burden of side effects measured by the FIBSER Scale
at week 8, 16 and 52.
ā¢ Adverse Events (AEs) and Serious Adverse Events (SAEs) at all
such times.
ā¢ Electrocardiography and blood tests as needed.
31. Statistical analysis
ā¢ Primary analysis:
ā¢ Primary analysis population : Intention-to-treat (ITT). All patients enrolled included in the
analysis, except those who did not take the study medication at all.
ā¢ Primary endpoint : Proportion of discontinuation of the allocated treatment for any reason at
week 8.
ā¢ The non-inferiority of escitalopram to duloxetine tested using the one-sided 95% confidence
interval (CI) for the between-group difference in the proportion of discontinuation
(escitalopramā duloxetine).
ā¢ The CI is estimated with the normal approximation. If the upper limit of the one-sided 95% CI
was less than the non-inferiority margin (Ī = 10%), escitalopram was considered to be non-
inferior to duloxetine.
32. Statistical analysis
ā¢ Secondary analyses :
ā¢ The proportion of discontinuation of the trial for any reason at week 16 and 52 between the
allocated treatment were compared using the two-sided 95% Wald CI.
ā¢ For repeated measurements of continuous endpoints (PHQ-9, QIDS-SR16, EQ-5D) at week 8,
16 and 52, a mixed-effects model for repeated measurements (MMRM) with a unstructured
covariance matrix was used to compare mean changes from the baseline between the two
groups. Degrees of freedom for errors were adjusted with the Kenward-Roger method.
ā¢ The 95% CI for the proportion in each binary endpoint (response rate, remission rate) was
estimated using the Clopper Pearson method, and proportions were compared with chi-
squared test.
33. Statistical analysis
ā¢ Secondary analyses :
ā¢ Change from baseline ordinal variables (CGI-S, SDS):one sample Wilcoxon
signed-rank test used
ā¢ FIBSER scores at weeks 8, 16 and 52 were compared between groups using
Wilcoxon rank-sum test
ā¢ Kaplan-Meier method was used for analysis with time-to-event data.
ā¢ significance level :
ā¢ primary non-inferiority hypothesis : one-sided 0.05,
ā¢ secondary endpoints : two-sided 0.05
38. Primary endpoint
ā¢ The primary endpoint of non-inferiority of escitalopram to duloxetine in
discontinuation rate at week 8 was established as the one-sided 95% upper
confidence limit (ā 0.06) fell below the non-inferiority margin of 0.10 and in
fact escitalopram (4.9%) was significantly below duloxetine (19.2%) in
discontinuation rate (difference=ā 0.14, 90% CI: ā 0.23 to ā 0.06, P = 0.007)
39. Secondary endpoints
ā¢ Discontinuation rate at week 16 and 52,
ā¢ Group A (9.8%) was significantly below Group B (26.9%)
at week 16 (difference=0.17, 95% CI: ā 0.27 to ā 0.07, P
= 0.005),
ā¢ At week 52 the between-group difference (Group A:
31.7%; Group B: 43.6%) was no longer significant
(difference=0.12, 95% CI: ā 0.22 to 0.01, P = 0.12)
40.
41.
42.
43.
44.
45. Adverse events n (%) Group A (n=82) Group B (n=78) P value*
All events 58 (70.73) 46 (58.97) 0.14
Headache 11 (13.41) 2 (2.56) 0.02
Cold 9 (10.98) 7 (8.97) 0.79
Sleepiness 9 (10.98) 7 (8.97) 0.79
Nausea 8 (9.76) 8 (10.26) 1.00
Flu 8 (9.76) 4 (5.13) 0.37
Diarrhea 7 (8.54) 4 (5.13) 0.54
Insomnia 2 (2.44) 7 (8.97) 0.09
Constipation 6 (7.32) 6 (7.69) 1.00
Dry mouth 3 (3.66) 4 (5.13) 0.71
Supplementary Table 2. Incidence of patients with adverse events during Step 1 (a) and the total
period (b) of treatment (> 5% in either group)
Adverse events n (%) Group A (n=82) Group B (n=78) P value*
All events 39 (47.56) 31 (39.74) 0.34
Cold 9 (10.98) 7 (8.97) 0.79
Sleepiness 7 (8.54) 3 (3.85) 0.33
Diarrhea 6 (7.32) 3 (3.85) 0.50
Headache 6 (7.32) 1 (1.28) 0.12
Nausea 5 (6.10) 5 (6.41) 1.00
Constipation 2 (2.44) 5 (6.41) 0.27
Flu 5 (6.10) 1 (1.28) 0.21
* Fisherās exact test
(a)
(b)
46. Discussion
ā¢ primary endpoint, non-inferiority of escitalopram to
duloxetine in discontinuation rates within 8 weeks of
treatment was established. In fact, the discontinuation
rate for any reason was statistically larger in duloxetine
group than in escitalopram group
ā¢ āwithdrawal of consentā and āvoluntarilyā, which appears
to reflect the patientās intent, contributed substantially to
the between-group difference.
47. Discussion
ā¢ no significant between-group difference in either efficacy or safety measures
during Step 1.
ā¢ Acceptability appears to rely heavily on the patientās intent . Neither efficacy
or safety measures alone may not be sensitive enough to reflect
acceptability.
ā¢ Nearly half of the patients continued their second-line antidepressant whether
escitalopram or duloxetine for 52 weeks, which appeared better than those
reported in previous studies, which are nearly 50% in six months (Furukawa
et al., 2013; Sansone and Sansone, 2012; Sawada et al., 2009)
49. limitations
ā¢ withdrawal of consent the major reason for the discontinuation in
GroupB not fully explored.
ā¢ In Step 2 the medication had been switched to the other drug in
some patients and thus, the outcome could not be attributed to
acceptability of drugs per se.
ā¢ Due to stratified block randomization per center strategy, was
small block in some sites,concealment of allocation was not
assured beforehand, which may have caused a selection bias.
50. limitations
ā¢ The drug switching scheme was notspecified . which might have influenced
the occurrence of adverse effects.
ā¢ The patients who discontinued due to clearintolerance (n = 4) or clinical
worsening (n = 1) during Step 1 dropped out of the study, which removed
those patients from Step 2 causing potential selection bias in findings
regarding Step 2.
ā¢ The change of the primary endpoint during the trial may have caused
potential bias to the discontinuation rate.
51. conclusion
ā¢ Escitalopram was not inferior to duloxetine in acceptability by the patients
who had inadequate response to the firstline antidepressant drug, which
were mostly SSRIs.
ā¢ The discontinuation rate at 8-week timepoint was significantly lower in
patients allocated to escitalopram than duloxetine as a second-line
antidepressant.
ā¢ When clinicians meet patients with inadequate response or intolerability to an
SSRI, switching to a non-SSRI may not be necessarily more beneficial than
switching to another SSRI.
52. Funding/support
ā¢ Mitsubishi Tanabe Pharma Corporation,Mochida
Pharmaceutical Co., Ltd.
ā¢ The Japan Foundation for Neuroscience and Mental
Health. whose funding sources were Shionogi & Co.,Ltd.
and Mochida Pharmaceutical Co., Ltd.
53. Funding/support
The funders had no role in:
ā¢ the design of the study ,its execution, analyses, data
interpretation,
ā¢ Decision to submit the results obtained,
ā¢ but had the opportunity to review the results for medical
and scientific accuracy.
54. conflict of interest
ā¢ All authors have declared a potential conflict of interest
beacause having recived lecture fee by various pharma
companies
56. Title and abstract
ā¢ 1a Identification as a randomised trial in the title
ā¢ yes , the title shows its a RCT
ā¢ 1b Structured summary of trial design, methods, results,
and conclusions
ā¢ yes, abstracts shows gives adequate information
58. background and objectives
ā¢ 2a Scientific background and explanation of rationale
ā¢ stated
ā¢ 2b Specific objectives or hypotheses
ā¢ presented
60. Trial design
ā¢ 3a Description of trial design (such as parallel, factorial)
including allocation ratio
ā¢ Trial design given in detail. with equal allocation
ā¢ 3b Important changes to methods after trial
commencement (such as eligibility criteria), with reasons
ā¢ changes made after interim analysis stated
61. Participants
ā¢ 4a Eligibility criteria for participants
ā¢ adequately presented
ā¢ 4b Settings and locations where the data were collected
ā¢ mentioned
62. Interventions
ā¢ The interventions for each group with sufficient details to
allow replication, including how and when they were
actually administered
ā¢ yes details given,
63. Outcomes
ā¢ 6a Completely defined pre-specified primary and
secondary outcome measures, including how and when
they were assessed
ā¢ stated but were changed after interim analysis
ā¢ 6b Any changes to trial outcomes after the trial
commenced, with reasons
ā¢ yes done , the reason for which was stated
64. Sample size
ā¢ 7a How sample size was determined
ā¢ determination of sample size inthe initial protocol not
given
ā¢ 7b When applicable, explanation of any interim analyses
and stopping guidelines
ā¢ changes to sample size after interim analysis made and
explained
65. Randomisation: Sequence generation
ā¢ 8a Method used to generate the random allocation
sequence
ā¢ centralized computer assignment program used
ā¢ 8b Type of randomisation; details of any restriction (such
as blocking and block size)
ā¢ stratified block randomization strategy used no evidence
of any restriction
66. Allocation concealment mechanism
ā¢ 9 Mechanism used to implement the random allocation
sequence (such as sequentially numbered containers),
describing any steps taken to conceal the sequence until
interventions were assigned
details of allocation concealment not provided. Also at some
small centers study as noted it may have not been effective
67. Implementation
ā¢ 10 Who generated the random allocation sequence, who
enrolled participants, and who assigned participants to
interventions
ā¢ clear details have not been mentioned. expect that it is
central computer generated
68. Blinding
ā¢ 11a If done, who was blinded after assignment to
interventions (for example, participants, care providers,
those assessing outcomes) and how
ā¢ only raters were blinded
ā¢ 11b If relevant, description of the similarity of interventions
ā¢ except for the drugs ,the interventions were very much the
same
69. Statistical methods
ā¢ 12a Statistical methods used to compare groups for
primary and secondary outcomes
ā¢ have been mentioned adequately
ā¢ 12b Methods for additional analyses, such as subgroup
analyses and adjusted analyses
ā¢ have not been carried out
71. Participant flow (a diagram is strongly recommended)
ā¢ 13a For each group, the numbers of participants who
were randomly assigned, received intended treatment,
and were analysed for the primary outcome
ā¢ Participant flow diagram given
ā¢ 13b For each group, losses and exclusions after
randomisation, together with reasons
ā¢ have been mentioned
72. Recruitment
ā¢ 14a Dates defining the periods of recruitment and follow-
up
ā¢ not mentioned
ā¢ 14b Why the trial ended or was stopped
ā¢ not clearly mentioned, it is assumed that after fulfillment of
sample size as per interim analysis
73. Baseline data
ā¢ 15 A table showing baseline demographic and clinical
characteristics for each group
ā¢ given
74. Numbers analysed
ā¢ 16 For each group, number of participants (denominator)
included in each analysis and whether the analysis was
by original assigned groups
yes it was
75. Outcomes and estimation
ā¢ 17a For each primary and secondary outcome, results for
each group, and the estimated effect size and its
precision (such as 95% confidence interval)
ā¢ yes done
ā¢ 17b For binary outcomes, presentation of both absolute
and relative effect sizes is recommended
ā¢ not done
76. Ancillary analyses
ā¢ 18 Results of any other analyses performed, including
subgroup analyses and adjusted analyses, distinguishing
pre-specified from exploratory
ā¢ not done
77. Harms
ā¢ 19 All important harms or unintended effects in each
group (for specific guidance see CONSORT for harms)
ā¢ have been mentioned in the supplement to the article
79. Limitations
ā¢ 20 Trial limitations, addressing sources of potential bias,
imprecision, and, if relevant, multiplicity of analyses
ā¢ major limitations have been mentioned
80. Generalisability
ā¢ 21 Generalisability (external validity, applicability) of the
trial findings
ā¢ limited by need to understand the reason for
discontinuation among participants before it can be
generalized
81. Interpretation
ā¢ 22 Interpretation consistent with results, balancing
benefits and harms, and considering other relevant
evidence
ā¢ interpretation appear balanced