Articles

Brexanolone (SAGE-547 injection) in post-partum

depression: a randomised controlled trial
Stephen Kanes, Helen Colquhoun, Handan Gunduz-Bruce, Shane Raines, Ryan Arnold, Amy Schacterle, James Doherty, C Neill Epperson,
Kristina M Deligiannidis, Robert Riesenberg, Ethan Hoffmann, David Rubinow, Jeffrey Jonas, Steven Paul, Samantha Meltzer-Brody

Summary
Background Post-partum depression is a serious mood disorder in women that might be triggered by peripartum Published Online
fluctuations in reproductive hormones. This phase 2 study investigated brexanolone (USAN; formerly SAGE-547 June 12, 2017
http://dx.doi.org/10.1016/
injection), an intravenous formulation of allopregnanolone, a positive allosteric modulator of γ-aminobutyric acid S0140-6736(17)31264-3
(GABAA) receptors, for the treatment of post-partum depression.
See Online/Comment
http://dx.doi.org/10.1016/
Methods For this double-blind, randomised, placebo-controlled trial, we enrolled self-referred or physician-referred S0140-6736(17)31546-5
female inpatients (≤6 months post partum) with severe post-partum depression (Hamilton Rating Scale for Sage Therapeutics Inc,
Depression [HAM-D] total score ≥26) in four hospitals in the USA. Eligible women were randomly assigned (1:1), via Cambridge, MA, USA
(S Kanes MD, H Colquhoun MD,
a computer-generated randomisation program, to receive either a single, continuous intravenous dose of brexanolone
H Gunduz-Bruce MD,
or placebo for 60 h. Patients and investigators were masked to treatment assignments. The primary efficacy endpoint R Arnold DO, A Schacterle PhD,
was the change from baseline in the 17-item HAM-D total score at 60 h, assessed in all randomised patients who J Doherty PhD, E Hoffmann,
started infusion of study drug or placebo and who had a completed baseline HAM-D assessment and at least one J Jonas MD, S Paul MD);
2b Analytics, Wallingford, PA,
post-baseline HAM-D assessment. Patients were followed up until day 30. This trial is registered with ClinicalTrials.gov,
USA (S Raines); Department of
number NCT02614547. Psychiatry, Perelman School of
Medicine, University of
Findings This trial was done between Dec 15, 2015 (first enrolment), and May 19, 2016 (final visit of the last enrolled Pennsylvania, Philadelphia, PA,
USA (Prof C N Epperson MD);
patient). 21 women were randomly assigned to the brexanolone (n=10) and placebo (n=11) groups. At 60 h, mean
University of Massachusetts
reduction in HAM-D total score from baseline was 21·0 points (SE 2·9) in the brexanolone group compared with Medical School, Worcester, MA,
8·8 points (SE 2·8) in the placebo group (difference –12·2, 95% CI –20·77 to –3·67; p=0·0075; effect size 1·2). No USA (K M Deligiannidis MD);
deaths, serious adverse events, or discontinuations because of adverse events were reported in either group. Four of Women’s Behavioral Health,
Zucker Hillside Hospital, New
ten patients in the brexanolone group had adverse events compared with eight of 11 in the placebo group. The most
York, NY, USA
frequently reported adverse events in the brexanolone group were dizziness (two patients in the brexanolone group vs (K M Deligiannidis); Atlanta
three patients in the placebo group) and somnolence (two vs none). Moderate treatment-emergent adverse events Center for Medical Research,
were reported in two patients in the brexanolone group (sinus tachycardia, n=1; somnolence, n=1) and in two patients Atlanta, GA, USA
(R Riesenberg MD); and
in the placebo group (infusion site pain, n=1; tension headache, n=1); one patient in the placebo group had a severe Department of Psychiatry, UNC
treatment-emergent adverse event (insomnia). School of Medicine, Chapel Hill,
NC, USA (Prof D Rubinow MD,
Interpretation In women with severe post-partum depression, infusion of brexanolone resulted in a significant and S Meltzer-Brody MD)

clinically meaningful reduction in HAM-D total score, compared with placebo. Our results support the rationale for Correspondence to:
Dr Samantha Meltzer-Brody,
targeting synaptic and extrasynaptic GABAA receptors in the development of therapies for patients with post-partum
Department of Psychiatry, UNC
depression. A pivotal clinical programme for the investigation of brexanolone in patients with post-partum depression School of Medicine, Chapel Hill,
is in progress. NC 27514, USA
samantha_meltzer-brody@
med.unc.edu
Funding Sage Therapeutics, Inc.

Introduction of all cases of post-partum depression, depending on the

Post-partum depression is a serious mood disorder
consistently observed in an estimated 10–20% of all
mothers who give birth in high-income and low-income
countries worldwide.1–3 Following delivery, post-partum
depression is characterised by clinically significant
depressive symptoms, often co-occurring with anxiety.1–6
Severe post-partum depression is defined as a major
depressive episode in the post-partum period with
marked impairment in functioning in both the Inter­
national Classification of Diseases (ICD)-10 and
Diagnostic and Statistical Manual of Mental Disorders
(DSM)-5.7,8 Estimates of the point prevalence of severe
post-partum depression generally range from 5% to 10%
setting.1–3 Furthermore, post-partum depression is a
leading cause of maternal mortality9,10 and, by affecting
maternal functioning, poses serious risks to the
emotional, cognitive, behavioural, and physical
development of the infant and siblings.11–13 Findings
from several studies implicate peripartum fluctuations
in reproductive hormones (in particular, the major
progesterone metabolite allopregnanolone) having
pivotal pathophysio­ logical roles in post-partum
depression.14–17
Allopregnanolone, a potent positive allosteric
modulator of synaptic and extrasynaptic GABAA
receptors,17,18 has been shown to have profound effects on

www.thelancet.com Published online June 12, 2017 http://dx.doi.org/10.1016/S0140-6736(17)31264-3 1

 Articles
Research in context
Evidence before this study
Preclinical and clinical studies have shown that neuroactive
steroids might have an important role in the pathophysiology
of post-partum depression. During pregnancy in mice,
expression of the γ-aminobutyric acid type A (GABAA) receptor δ
subunit is downregulated as allopregnanolone concentrations
increase, and at parturition, expression of the GABAA receptor δ
subunit is recovered as allopregnanolone concentrations drop
steeply. GABAA receptor δ-subunit-deficient mice do not adapt
to the substantial changes in allopregnanolone concentration
during pregnancy and parturition, and showed depression-like
and anxiety-like behaviours and abnormal maternal behaviours
that were reversed by administration of allopregnanolone.
These findings lend support to the hypothesis that changes in
neuroactive steroid concentrations during pregnancy and post
partum are capable of provoking affective dysregulation.
Neuroactive steroids such as allopregnanolone might function
as behavioural switches, suggesting a potentially important role
in treatment of reproductive and endocrine-related mood
disorders such as post-partum depression.
We searched PubMed (all indexed dates up to Feb 1, 2017) for
clinical trials with the terms “allopregnanolone”, “neuroactive
steroid”, “GABAA positive allosteric modulator”, and “postpartum
depression”. This search retrieved no trials examining the
neuroactive steroid or GABAA-receptor mechanism in
post-partum depression. A previous report by several of this
study’s investigators describes an open-label, exploratory study of
brexanolone in four women with severe post-partum depression.
Added value of this study
To our knowledge, this study is the first randomised,
double-blind, placebo-controlled trial of a therapeutic
anxiety and depression in animal models.17,19–21 Plasma
allopreg­nanolone concentrations rise in concert with
progesterone throughout pregnancy, reaching the
highest physiological concentrations in the third
trimester.22 After childbirth, these concentrations
decrease abruptly.23 Failure of GABAA receptors to adapt
to these changes at parturition has been postulated to
have a role in triggering post-partum depression.24,25
Alterations in concentrations or ratios of serum
allopregnanolone and other neuroactive steroids have
been reported in some,23,26 but not all,27 women at risk for
or who develop post-partum depression. Moreover,
symptoms of post-partum depression are precipitated in
at-risk women by recreating hormonal fluctuations
associated with pregnancy and delivery.14 For some
women, onset of mood symptoms occurs in the third
trimester of pregnancy, and these symptoms can
substantially worsen in the immediate post-partum
period.6,28,29 Whether this group of women are similar or
different from women who have onset after childbirth
is not known; however, women with onset in the
2 www.thelancet.com Published online June 12, 2017 http://dx.doi.org/10.1016/S0140-6736(17)31264-3
formulation of the neuroactive steroid allopregnanolone in
patients with post-partum depression. The current standard
of care for post-partum depression includes psychotherapy
and pharmacological therapies. However, no pharmacological
therapies are specifically indicated for the treatment of
post-partum depression. Antidepressant medications used to
treat depressive disorders outside of the perinatal period,
such as selective serotonin reuptake inhibitors and tricyclic
antidepressants, are commonly used in post-partum
depression. However, these therapies are not directly linked
with existing hypotheses regarding the causes of
post-partum depression, their onset of efficacy can be delayed
by several weeks or months, and their overall remission rate
in post-partum depression is low. In particular, rapid onset of
action is desirable in severe post-partum depression to
quickly mitigate the serious, negative effects of the disorder
on the mother, infant, and family. Our study suggests the
potential for the development of a GABAA-positive allosteric
modulator, such as the neuroactive steroid brexanolone, as a
new mechanism for treatment of post-partum depression
that is related to the underlying pathophysiology.
Implications of all the available data
Together with preclinical and clinical studies suggesting a role
for neuroactive steroids and GABAA receptor regulation in the
pathophysiology of post-partum depression, our findings
support the rationale for further examining brexanolone in
patients with post-partum depression. Several pivotal clinical
trials are currently examining the efficacy and safety of
brexanolone in post-partum depression.
third trimester are an important group to consider in
terms of differential sensitivity to alterations in changing
concentrations of neurosteroids during the peripartum
period. Although the cause of post-partum depression is
not entirely understood, this collective body of work
supports the rationale for exploring the potential
treatment of women with post-partum depression with
doses of allopregnanolone that result in serum
concentrations equivalent to those present during the
third trimester, including women with onset of
symptoms in the third trimester of pregnancy and
throughout the early post-partum period.
Allopregnanolone has low aqueous solubility, poor
oral bioavailability, and is rapidly metabolised; however,
a soluble, proprietary, β-cyclodextrin-based formulation
of allopregnanolone—brexanolone (USAN; formerly
SAGE-547 injection)—can be administered intravenously
to produce stable physiological serum concentrations.
Here, we report the results of a randomised, placebo-
controlled study of brexanolone in women with severe
post-partum depression.

Methods
Study design and participants
This multicentre, randomised, double-blind, parallel-
group, placebo-controlled trial received institutional
review board approval for 11 sites in the USA and
recruited patients in four sites (University of
Pennsylvania, University of Massachusetts, Atlanta
Center for Medical Research, and the University of North
Carolina, Chapel Hill; site principal investigators CNE,
KMD, RR, and SM-B). Participants were enrolled by the
trial site investigator and gave written informed consent.
The trial was completed upon reaching the enrolment
target. Additional details of the study design are provided
in the appendix.
Participants were accrued through both self-referral
and physician referrals to clinical research and specialised
psychiatric units. A summary of diagnoses of post-
partum depression and time of enrolment is available in
the appendix. Participants met the following complete
inclusion criteria for eligibility: signed informed consent
form before any study-specific procedures were
performed; ambulatory female aged between 18 and
45 years of age; good physical health and no clinically
significant findings as determined by the investigator on
physical examination, 12-lead electro­cardiograph (ECG),
or clinical laboratory tests; agreed to adhere to the study
requirements; had a negative pregnancy test at screening
and day 1 before the start of study drug infusion; had a
major depressive episode that began no earlier than the
third trimester and no later than the first 4 weeks
following delivery as diagnosed by the Structured Clinical
Interview for DSM-IV Axis I Disorders (SCID-I);30 had a
17-item Hamilton Rating Scale for Depression (HAM-D)
total score of 26 or higher at screening and day 1 (before
randomisation); was within 6 months post partum at the
time of enrolment; and had no detectable hepatitis B
surface antigen, anti-hepatitis C virus, or HIV antibody
at screening. Patients could be on stable antidepressants,
but they had to be willing to delay the start of
other antidepressant or anxiety medication and any
new pharmacotherapy regimens, including as-needed
benzodiazepine anxiolytics, until the study drug infusion
and 72 h assessments were completed. Additionally, the
initial inclusion and exclusion requirements included
the provision that patients either must have ceased
lactating at screening or if still lactating at screening,
must have already fully and permanently weaned their
infant(s) from breastmilk, or if still actively breastfeeding
at screening, must have agreed to cease giving breastmilk
to their infant(s) before receiving study drug. A protocol
amendment allowed for resumption of breastfeeding
after day 12 of the study. The HAM-D cutoff was chosen
on the basis of previous studies of severe cases of post-
partum depression (defined as perinatal major depressive
disorders by both ICD-10 and DSM-5),2 studies of severe
major depressive disorder defined by HAM-D score,31
and the principal investigators’ previous experiences in
www.thelancet.com Published online June 12, 2017 http://dx.doi.org/10.1016/S0140-6736(17)31264-3 3
Articles
an open-label trial of brexanolone in severe post-partum
depression.32
Exclusion criteria included active psychosis; attempted
suicide associated with an index case of post-partum
depression; history of seizures, bipolar disorder, schizo­
phrenia, or schizoaffective disorder; and history of
alcoholism or drug addiction (including benzodiazepines)
in the 12 months before screening. Additional details
about the inclusion and exclusion criteria are provided
in the appendix.
Randomisation and masking
Eligible patients were randomly assigned (1:1), according
to a simple computer-generated randomisation schedule, See Online for appendix
with no blocking, to brexanolone or placebo. SAS was used
to generate random numbers for randomisation. The
randomisation schedule was produced by an independent
statistician at Applied Statistics and Consulting (Spruce
Pine, NC, USA), a contract research organisation entirely
independent from any trial investigator.
Patients, clinicians, and study teams were masked to
treatment allocation. Patients in the placebo group
received equivalent infusion rates, and the infusion bags
for both treatments were identical in appearance. There
were no differences in the characteristics of any of the
infusion bags or the solutions within, including smell or
colour. The pharmacist at each site who prepared the
infusion bags according to the randomisation schedule,
and a monitor who performed drug accountability during
the study, were not masked to treatment assignment. No
other study personnel were unmasked until after formal
locking of the study database. The study database was
locked when the final visit of the last enrolled patient was
completed, data entry into the clinical database was
completed, and the database for all patients was deemed
clean with all queries resolved (June 22, 2016). Allocation
concealment was facilitated by a secure, web-based
method. Only the clinic pharmacist was given a copy of
the randomisation schedule. In the event of a medical
emergency, the pharmacist was to reveal actual infusion
contents to the primary investigator, who was to alert the
sponsor of the emergency. In all cases, if the study drug
allocation for a patient had been unmasked, pertinent
information (including the reason for unmasking) was to
be documented in the patient’s records and on the
electronic case report form. If the patient or study centre
personnel were unmasked, the patient was to be
terminated from the study. No such unmasking occurred
during the study.
Procedures
Brexanolone is a sterile solution of allopregnanolone
5 mg/mL in 250 mg/mL sulfobutylether-β-cyclodextrin
(SBECD) buffered with citrate, which is diluted with
sterile water for injection to render it isotonic for
intravenous infusion. Each patient received a single
continuous intravenous infusion of masked study drug
 Articles

for 60 h during inpatient care under the following

23 women screened schedule: 30 µg/kg per h (0–4 h); 60 µg/kg per h (4–24 h);
90 µg/kg per h (24–52 h); 60 µg/kg per h (52–56 h);
2 excluded for ineligibility*
30 µg/kg per h (56–60 h). Infusion was initiated between
0800 h and 1000 h, apart from one patient whose infusion
was initiated in the afternoon (full details provided in the
21 enrolled and randomly assigned appendix). Infusion rate adjustments were allowed on the
basis of tolerability, side-effects, and predetermined
protocol rules. The dose and infusion rate were based on
the protocol of the previous open-label exploratory trial in
10 assigned to brexanolone 11 assigned to placebo
patients with severe post-partum depression32 and on
pharmacokinetic modelling. Participants remained as
10 received brexanolone and 11 received placebo and inpatients during the 60 h study infusion period. After
were followed up were followed up dosing was complete, patients were followed up until
day 30, with clinical and safety assessments done at days 7
10 included in safety population 11 included in safety population and 30.
10 included in efficacy population 11 included in efficacy population
Outcomes
The primary outcome measure was the change from
Figure 1: Trial profile
baseline in HAM-D total score at the end of the treatment
*One patient demonstrated continuing illicit drug use (positive screen for period (60 h). Secondary analyses included changes in
amphetamines, barbiturates, and opiates) and the other did not meet the HAM-D score from baseline at 2 h up to 30 days. HAM-D
symptom severity criteria (Hamilton Rating Scale for Depression score <26). measurements were obtained frequently to monitor for
rapid onset of improvement of symptoms and were
started within a window of tolerance of 25 min before or
after the designated timepoint in the first 24 h. Secondary
Brexanolone (n=10) Placebo (n=11)
HAM-D endpoints were the proportion of patients
achieving remission (HAM-D total score ≤7), the
Characteristics
proportion of patients achieving response (≥50% reduction
Age (years) 27·4 (5·3); 27 (20–40) 28·8 (4·6); 28 (22–36)
in HAM-D total score), change from baseline in the
Ethnicity
Bech-6 subscore,33 which assesses the core symptoms of
Hispanic or Latino 0 0 Montgomery-Åsberg Depression Rating Scale (MADRS)
Not Hispanic or Latino 10 (100%) 11 (100%) total score,34 Clinical Global Impression-Global Improve­
Race ment,35 major depression, and changes in the HAM-D
Black or African-American 7 (70%) 6 (55%) depressed mood item score. Additional prespecified
White 3 (30%) 5 (45%) secondary and exploratory endpoints are detailed in
Height (cm) 162·4 (7·1); 163·5 (153–175) 161·7 (6·7); 162·0 (151–174) the appendix, including the Generalised Anxiety Disorder
Weight (kg) 86·7 (28·8); 76·5 (49·7–130·7) 77·0 (22·3); 73·5 (53·3–122·6) Questionnaire,36 Edinburgh Postnatal Depression Scale,37
Body-mass index (kg/m2) 32·7 (9·9); 30·5 (20·4–47·1) 29·3 (7·8); 28·2 (21·0–45·0) Patient Health Questionnaire-9,38 and Barkin Index of
Personal history Maternal Function.39 To facilitate consistent scoring of the
Psychiatric disorder HAM-D, investigators received training on use of the
Depression (non-PPD) 6 (60%) 6 (55%) scale from MedAvante (Hamilton, NJ, USA), and each
Anxiety 2 (20%) 5 (45%) patient was assessed by the same investigator throughout
Other 1 (10%) 2 (18%) the study.
Previous PPD episodes 7 (70%) 4 (36%) The safety and tolerability of brexanolone were assessed
Antidepressant medication 3 (30%) 3 (27%) by recording and summarising adverse events, clinical
Baseline HAM-D 28·1 (27·0–30·0) 28·8 (26·0–32·0) laboratory measurements, vital signs, and ECGs (including
Family history changes from baseline); concomitant medication was also
Perinatal psychiatric disorders assessed. Emergent suicidal ideation and behaviours were
Mother 2 (20%) 2 (18%) assessed with the Columbia-Suicide Severity Rating
Sister(s) 1 (10%) 1 (9%) Scale;40 patient-reported sedation or sleepiness was
assessed with the Stanford Sleepiness Scale.41
Data are n (%) or mean (SD); median (range). Age was derived from the birth date and screening date. Bodyweight and
height were measured at screening. Body-mass index was programmatically calculated in the electronic case report
form. Medical histories were coded according to the Medical Dictionary for Regulatory Activities (MedDRA) version 17 Statistical analysis
or later. PPD=post-partum depression. HAM-D=Hamilton Rating Scale for Depression. On the assumption of a two-sided test at an α level of 0·10,
a sample size of ten evaluable patients per group provided
Table 1: Baseline demographics and characteristics
80% power to detect an effect size of 1·2 between the

4 www.thelancet.com Published online June 12, 2017 http://dx.doi.org/10.1016/S0140-6736(17)31264-3

 Articles

brexanolone and placebo groups with regard to the primary

A
outcome variable. An effect size of 1·2 corresponds to a 30 Placebo (n=11)
placebo-adjusted difference of 12 points in the change Brexanalone (n=10)
from baseline in HAM-D total score at 60 h with an 25
assumed standard deviation of 10 points. By including

Mean HAM-D total score

two treatment groups and using a randomisation ratio of 20
1:1, a total of 20 evaluable patients was required.
15
After 16 participants had completed HAM-D efficacy
assessments at 60 h, a sample size re-estimation was *
10 *
done. For this analysis, an independent statistician from * * * *
*
the Datatrial (Newcastle upon Tyne, UK) contract research 5
organisation calculated the unmasked mean percentage End of infusion
change from baseline in HAM-D total score at 60 h for 0
the placebo group. On the basis of this information only,
the independent statistician communicated one of the B
40
following messages back to the sponsor: if placebo
response was less than 35%, no adjustment to the sample 35

size is required; or if placebo response was 35% or more, 30

Mean MADRS total score

increase the sample size by five subjects per group. No 25

personnel involved in the conduct of the study had access
to the interim results generated by the independent 20

statistician. Since the sponsor was kept masked to the 15

data and uninformed of the interim results (ie, response
rates) until final database lock, no statistical adjustment
was made to the level of significance for hypothesis
testing at the end of the study. No increase in sample size
was needed in this trial. On the basis of the results of the
sample size re-estimation, the sample size could have
been increased to a maximum of 32 randomised patients.
This adjustment to the sample size would have allowed
for an effect size of 1·0 to be detected.
The safety population included all randomised patients
who started infusion of study drug or placebo. The
efficacy population included the subset of the safety
population who had a completed baseline HAM-D
assessment and at least one post-baseline HAM-D
assessment. The change from baseline in HAM-D and
MADRS total score was analysed with a mixed effects
model for repeated measures. This linear model included
centre, treatment, baseline HAM-D total score,
assessment timepoint, and timepoint-by-treatment as
explanatory variables. Centre was treated as a random
effect, while all other explanatory variables were treated
as fixed effects. This analysis model was prespecified in
the trial protocol and statistical analysis plan. Model-
based point estimates (ie, least squares means, 95% CIs,
and p values) were reported for each timepoint. The
primary comparison was between brexanolone and
placebo at the 60 h timepoint. Other changes from
baseline endpoints were analysed with similar methods.
The HAM-D response and remission rates at each
timepoint were analysed with Fisher’s exact test. Point
estimates (ie, odds ratios [ORs]), 95% CIs, and p values
are reported. Additional details of the statistical analysis
plan and statistical methods are provided in the appendix.
Analyses were done with SAS 9.2. This study is registered
with ClinicalTrials.gov, number NCT02614547.
* *
* *
*
10 *
5 End of infusion
0
0 12 24 36 48 60 72 168 720
(day 7) (day 30)
Timepoint (h)
Figure 2: Mean HAM-D and MADRS total scores
(A) The Hamilton Rating Scale for Depression (HAM-D) is a 17-item diagnostic
questionnaire used to measure the severity of depressive episodes in patients
with mood disorders. It consists of individual ratings related to the following
symptoms: depressed mood (sadness, hopeless, helpless, worthless), feelings of
guilt, suicide, insomnia (early, middle, late), work and activities, retardation
(slowness of thought and speech; impaired ability to concentrate; decreased
motor activity), agitation, anxiety (psychic and somatic), somatic symptoms
(gastrointestinal and general), genital symptoms, hypochondriasis, loss of
weight, and insight. Higher HAM-D scores indicate more severe depression.
(B) The Montgomery-Åsberg Depression Rating Scale (MADRS) is a ten-item
diagnostic questionnaire used to measure the severity of depressive episodes in
patients with mood disorders. Higher MADRS scores indicate more severe
depression, and each item yields a score of 0–6, producing total score ranges
from 0 to 60. p values were calculated by two-sided t test. *Denotes statistical
significance versus placebo, defined as p≤0·01.
Role of the funding source
The funding source assisted in the study design, data
collection, data analysis, data interpretation, and writing
of the report. All authors had full access to all the data in
the study and the corresponding author had final
responsibility for the decision to submit for publication.
Results
This trial was done between Dec 15, 2015 (first enrolment),
and May 19, 2016 (final visit of the last enrolled patient).
23 patients were screened, of whom 21 were eligible for
enrolment (figure 1). All randomised patients completed
the 60 h inpatient dosing protocol and completed the
30 day follow-up period. There were little to no missing
data in the study. One patient missed an assessment of

www.thelancet.com Published online June 12, 2017 http://dx.doi.org/10.1016/S0140-6736(17)31264-3 5

 Articles

–12·2 points (95% CI –20·77 to –3·67; two-tailed t test,

HAM-D MADRS
p=0·0075). The effect size for the clinical efficacy at 60 h
Brexanolone Placebo LSmeans p value Brexanolone Placebo Mean p value was 1·2. Prespecified secondary analyses showed a
(n=10) (n=11) difference (n=10) (n=11) difference
–11·3 point (95% CI –18·86 to –3·65) mean difference
2h –6·27 –4·12 –2·2 (2·3) 0·3685 ·· ·· ·· ··
between groups at 24 h (two-tailed t test, p=0·0059), with
4h –8·67 –5·21 –3·5 (2·9) 0·2476 ·· ·· ·· ·· significant improvements seen for the brexanolone group
8h –10·57 –5·94 –4·6 (3·1) 0·1549 ·· ·· ·· ·· at 36, 48, 60, and 72 h, as well as days 7 and 30 (figure 2).
12 h –12·67 –6·66 –6·0 (3·7) 0·1164 ·· ·· ·· ·· The primary endpoint data (HAM-D) were also analysed
24 h –19·37 –8·12 –11·3 (3·6) 0·0059 –26·36 –8·83 –17·5 (5·4) 0·0042 post-hoc by a Wilcoxon Signed Rank Test, a non-parametric
36 h –20·17 –8·21 –12·0 (4·0) 0·0078 ·· ·· ·· ·· method, which reached the same conclusions as the
48 h –21·87 –9·21 –12·7 (4·0) 0·0047 –29·86 –11·46 –18·4 (5·3) 0·0026 MMRM method (see appendix). Change from baseline in
60 h –20·97 –8·75 –12·2 (4·1) 0·0075 –27·96 –12·10 –15·9 (5·5) 0·0104 MADRS total score showed similar results to those
72 h –21·07 –8·39 –12·7 (4·3) 0·0078 –28·66 –12·46 –16·2 (5·5) 0·0090 obtained with the HAM-D score (figure 2 and table 2).
7 days –20·97 –8·06 –12·9 (3·9) 0·0038 –27·56 –11·60 –16·0 (5·4) 0·0091 Remission from depression (HAM-D total score ≤7) was
30 days –20·77 –8·84 –11·9 (4·1) 0·0095 –26·26 –11·19 –15·1 (5·2) 0·0100 seen in seven of ten patients in the brexanolone group and
in one of 11 patients in the placebo group at 60 h
Data are LSmeans (SE). The change from baseline in the Hamilton Rating Scale for Depression (HAM-D) mean total score
and Montgomery-Åsberg Depression Rating Scale (MADRS) mean total score was analysed using a mixed effects model (OR –23·33, 95% CI –1·56 to 1152·71; Z test from log
for repeated measures. MADRS was not assessed at 2, 4, 8, 12, or 36 h. p values were calculated by two-sided t test. transformation of the OR, p=0·0364; figure 3). This
difference was seen at 24 h (six patients in the brexanolone
Table 2: Mean HAM-D and MADRS total scores
group vs one patient in the placebo group; OR 15·00,
95% CI 1·07–756·72; Z test, p=0·0561) and a difference
80 Placebo (n=11) was maintained until the 30-day follow-up (seven vs two;
Brexanolone (n=10)
OR 10·50, 95% CI 1·01–140·57; Z test, p=0·0499). More
70 patients demonstrated a 50% or greater reduction in
HAM-D total score in the brexanolone group than in the
Patients with remission of symptoms (%)

60
50
40
30
20
10
0
Baseline 2 4 8 12 24 36 48 60 72
Timepoint (h)
Figure 3: Patients with remission of symptoms
Remission was defined as a Hamilton Rating Scale for Depression (HAM-D) total score of 7 or lower. A larger
proportion of patients in the brexanolone group than in the placebo group achieved HAM-D remission at each
timepoint after 2 h. The difference was significant at 24 h (p=0·0561), 48 h (p=0·0897), 60 h (p=0·0449), and 72 h
(p=0·0364), as well as at day 7 (p=0·0449) and day 30 (p=0·0449). p values were calculated by Z test from log
transformation of the odds ratio.
HAM-D and MADRS at day 7 and had an additional,
unscheduled assessment at day 17.
The proportion of patients with a previous history
of psychiatric disorders was similar between treatment
groups, apart from history of anxiety (table 1). The
proportion of patients with at least one previous episode
of post-partum depression was higher in the brexanolone
group than in the placebo group. Use of antidepressant
medication was balanced between the groups.
At the end of the 60 h infusion, mean reduction in
HAM-D total score was 21·0 points (SE 2·9) in the
brexanolone group, compared with 8·8 points (2·8) in the
placebo group; mean difference between groups
placebo group across all timepoints. Although this finding
was not significant at 60 h (70% [seven patients] in the
brexanolone group vs 36% [ four patients] in the placebo
group; p=0·1450), it did achieve significance at both 72 h
(80% [eight patients] brexanolone vs 27% [three patients]
placebo; p=0·0374) and day 7 (80% [eight patients]
brexanolone vs 20% [two patients] placebo; p=0·0335;
appendix). The observed improvement in symptoms of
post-partum depression following brexanolone
administration also extended beyond core depressive
symptoms, as evidenced by the significant treatment
168 720 difference observed for CGI-I response.
(day 7) (day 30) Brexanolone was generally well tolerated. There were no
deaths, serious adverse events, or discontinuations in
either group. Overall, fewer patients who received
brexanolone reported adverse events compared with
patients who received placebo (four of ten patients in the
brexanolone group vs eight of 11 in the placebo group;
table 3). The most frequently reported adverse events in
the brexanolone group were dizziness (two patients in the
brexanolone group vs three patients in the placebo group)
and somnolence (two vs none). Sedation was reported in
one patient in the brexanolone group and in no patients in
the placebo group. Moderate treatment-emergent adverse
events in the brexanolone group were sinus tachycardia
(one [10%] patient) and somnolence (one [10%] patient).
One (9%) patient in the placebo group had a severe
treatment-emergent adverse event of insomnia. Moderate
treatment-emergent adverse events in the placebo group
were infusion site pain (one [9%] patient) and tension
headache (one [9%] patient). All other treatment-emergent
adverse events in both groups were mild.

6 www.thelancet.com Published online June 12, 2017 http://dx.doi.org/10.1016/S0140-6736(17)31264-3

 Articles

At baseline, mean Stanford Sleepiness Scale scores

Brexanolone (n=10) Placebo (n=11)
were similar in the brexanolone and placebo groups
(2·7 vs 2·6, respectively); these scores were also similar Any adverse event 4 8
after treatment, suggesting that there were no differences Dizziness 2 3
in sleepiness between groups (appendix). One patient in Nausea 1 3
the brexanolone group, who was taking clonazepam Abnormal dreams 0 2
(6 mg), reported sleepiness; this patient required a dose Headache 0 2
reduction of the study drug and, after dose reduction, Infusion site pain 0 2
completed participation in the study. Insomnia 0 2
Improvements in Columbia-Suicide Severity Rating Rash 1 1
Scale suicidal ideation items were noted in both treatment Somnolence 2 0
groups (appendix). Two patients in the brexanolone group Abdominal pain 0 1
reported active suicidal ideation with a specific plan and Anxiety 0 1
intent at baseline but not at the post-treatment assessment. Dizziness postural 1 0
No individuals experienced worsening of suicidal ideation Dry mouth 1 0
or behaviour during treatment or follow-up. Hot flush 1 0
Infusion site extravasation 0 1
Discussion Localised oedema 0 1
Severe post-partum depression is a serious and disabling Pain in extremity 0 1
mood disorder that often requires admission to hospital Pyrexia 1 0
and is a major risk factor for maternal mortality from Sedation 1 0
suicide. In addition to psychotherapy, the standard of Sinus tachycardia 1 0
care for moderate to severe post-partum depression Skin abrasion 0 1
includes pharmacological therapies, used judiciously, Tension headache 0 1
particularly in nursing mothers.42 Evidence for the Vertigo 1 0
efficacy of antidepressants in treatment of post-partum
depression is based on their empirical use in the general Data are number of patients reporting treatment-emergent adverse event,
defined as an adverse event with onset after the start of study drug, or any
population and limited data from several randomised, worsening of a pre-existing medical condition or adverse event with onset after
placebo-controlled trials.42–44 Some data suggest that time the start of study drug and until the follow-up visit on day 7 (ie, approximately
to response is longer and need for polypharmacy is 4 days after the end of the infusion). Adverse events were coded according to the
Medical Dictionary for Regulatory Activities version 18.0.
greater when patients with post-partum depression are
treated with antidepressants, compared with non- Table 3: Treatment-emergent adverse events
perinatal women with major depressive disorder.13,45,46
Moreover, many women treated for post-partum
depression do not achieve full remission of symptoms.47 study contrasts with the antidepressant effects reported
Therefore, there remains a great need for improved in a study of a single infusion of ketamine in patients
pharmacological treatment options. with major depressive disorder, which were maintained
This study is the initial placebo-controlled trial of in most patients in the study for less than a week.50,51 The
brexanolone in a clearly defined population of women observed improvement in symptoms of post-partum
with severe post-partum depression. Our findings depression following brexanolone administration also
provide the first placebo-controlled clinical support for extended beyond core depressive symptoms, as evidenced
the role of extrasynaptic GABAA receptors in the by the significant treatment difference observed for CGI-I
modulation of mood and affective states in any clinical response. Finally, brexanolone was well tolerated, and
population. The large effect size (1·2) seen in this trial fewer patients receiving brexanolone had adverse events
contrasts with that observed in studies of currently than did patients in the placebo group, and sedation did
available and widely used antidepressants, including not differ between the groups, as assessed by the Stanford
selective serotonin reuptake inhibitors (SSRIs), Sleepiness Scale.
serotonin–noradrenaline reuptake inhibitors (SNRIs), Although this multisite phase 2 study had a modest
and tricyclic antidepressants.48,49 For example, a pooled number of study participants, the study was designed to
meta-analysis of placebo-controlled studies of fluoxetine find as representative a sample as possible. To that end,
showed an effect size of –0·30 in favour of fluoxetine we recruited study participants from a wide geographic
(95% CI –0·39 to –0·21).50 range that included urban, suburban, and rural settings
A treatment modality with rapid onset of action is in the USA. Travel and logistic support was provided for
desirable in severe post-partum depression in view of the patients who did not live near our open research sites.
extensive adverse impact of the disease on the mother, Thus, the study participants reflect a diverse group in
infant, and family. Moreover, the 30 day maintenance of terms of race, geography, and health insurance payer
treatment effect seen in the brexanolone group in our mix. The study sites themselves were also diverse and

www.thelancet.com Published online June 12, 2017 http://dx.doi.org/10.1016/S0140-6736(17)31264-3 7

 Articles
included academic medical centres and clinical research
organisations. Therefore, we believe our findings to
be representative of women with severe post-partum
depression, defined as a HAM-D score of 26 or higher.
Several forms of advertising were used to ensure broad
representation of eligible study participants, including
recruiting directly from study sites and local physicians
in the form of print media (brochures, flyers, and
physician referral letters) and social media (Facebook,
Google, Bing/Yahoo, and Swoop Native advertising).
Although all approaches yielded potential participants,
the use of social media substantially increased our ability
to reach patients more broadly, reflecting the high use of
social media in women of childbearing age.
As with the complex causes of mood disorders in
general, multiple biological and non-biological factors
undoubtedly have a role in the pathophysiology of
post-partum depression.14 However, we postulated that
women with onset of severe depressive symptoms in
the third trimester or triggered by childbirth—often
associated with rapid fluctuations in gonadal steroids—
are more likely to have a hormone-responsive form of
post-partum depression that could respond to therapeutic
doses of neuroactive steroids, such as allopregnanolone.20
The results of this trial of brexanolone support this
hypothesis and are consistent with preclinical literature
on the central role of neuroactive steroids and extra­
synaptic GABAA receptors in modulating affective
behaviour and anxiety. Notably, this trial addresses
the treatment of severe post-partum depression with
brexanolone, and additional trials will be required to see
if the observed effects are generalisable for varying
degrees of severity of post-partum depression and if
brexanolone treatment can be applied for depression.
The precise pathophysiology underlying the pre­
disposition to development of post-partum depression is
unknown; however, preclinical work has implicated
alterations in GABAA receptors, especially extrasynaptic
GABAA receptors.24 Dynamic changes in GABAA receptor
expression in response to changing neuroactive steroid
concentrations might be necessary for maintaining
homeostasis of neurotransmission during pregnancy and
the post-partum period.24 During the course of pregnancy,
plasma concentrations of the endogenous neuroactive
steroid allopregnanolone, a potent positive modulator of
synaptic and extrasynaptic GABAA receptors, increase
from less than 5 nM before pregnancy to 157 nM in the
third trimester.22 Enhanced GABAergic inhibition
triggered by progressively increased allopregnanolone
concentrations throughout pregnancy is thought to be
offset by a simultaneous alteration in GABAA receptor
activity or expression, changes in activation kinetics, or a
combination of both, to prevent a substantial shift in the
excitatory and inhibitory balance in the brain.52
Importantly, these changes must then be quickly reversed
at the time of parturition, because allopregnanolone
concentrations begin to drop steeply within hours, and
8 www.thelancet.com Published online June 12, 2017 http://dx.doi.org/10.1016/S0140-6736(17)31264-3
return back to pre-pregnancy concentrations within 5 days
post partum.53 Although GABAA receptor expression and
tonic inhibitory currents normally return to pre-pregnancy
concentrations following parturition, the inability to adapt
to rapid changes in neuroactive steroid concentrations
during pregnancy and parturition might contribute to the
pathophysiology of post-partum depression in some
women, resulting in affective symptoms upon the rapid
decline of allopregnanolone concentrations post partum
or possibly beginning in the third trimester of pregnancy.14
This trial also demonstrates that a study of women with
post-partum depression is feasible and that complex trial
designs are not necessarily required to overcome
presumed placebo responses, especially with the large
effect size we observed with brexanolone. Furthermore,
trials in such a clearly defined and previously
understudied population are crucial to develop novel
treatments for post-partum depression. This trial is
limited by a small sample size, although because of the
large effect size and rapid response, the study was
adequately powered. Notably, because of the observed
response, the data were not distributed normally, but in
addition to the prespecified mixed effects model for
repeated measures used for analysis, a non-parametric
analysis (Wilcoxon signed rank test) also showed
statistical significance.
This trial is also limited by the use of a very strict
definition of severe post-partum depression (HAM-D ≥26),
which was defined on the basis of existing studies of
severe major depressive disorder assessed with HAM-D31
and our previous experience in an open-label trial of
brexanolone in severe post-partum depression.32 This
restriction raises two questions that will need to be
addressed in future trials: first, what overall percentage of
women with severe post-partum depression would
ultimately respond to brexanolone; and second, how to
generalise our results to the broader group of all patients
with post-partum depression. Up to 20% of women giving
birth will have post-partum depression (including
both minor and major depression), and approximately
5–10% of these women will have severe post-partum
depression (as defined by a major depressive episode).1–3
Although this population of women with severe
post-partum depression is undoubtedly an important
population to study, it represents only a fraction of women
with post-partum depression. A phase 3 programme is
currently investigating the potential use of brexanolone in
post-partum depression of varying degrees of severity.
Additional limitations of our trial are the 30 day follow-up
period, possible respondent fatigue in HAM-D assessment
from frequent administration, and the inability (because
of the small study size) to stratify patients on the basis of
previous history of post-partum depression.
These findings replicate our recent open-label
exploratory study in four women with severe post-partum
depression32 and demonstrate a substantial treatment
effect of brexanolone in the trial population of patients

with severe post-partum depression, an indication for
which there are no currently approved pharmacological
therapies. The rapid and marked antidepressant response
associated with brexanolone administration contrasts
with the 4–6 weeks needed (and low remission rates) that
have been reported with SSRIs and other antidepressants
in patients with post-partum depression.44 A rapid onset
of antidepressant action is crucial because speed of onset
is a strong determinant of the likelihood of near-term
recovery, and full remission is the objective of any
treatment.54 Moreover, rapid onset of action could
decrease potential risks to the mother from self-harm
and risks to the baby (and siblings) from lack of attention
or negative interaction with the mother. The results of
our study support further development of brexanolone
for the treatment of post-partum depression.
Contributors
All authors contributed to the design of the study and writing of the
manuscript. SK, HC, and SM-B designed the study. SK and SM-B wrote
the first draft. CNE, KMD, RR, and DR were investigators in the study.
SR provided statistical analysis. All authors vouch for the accuracy and
completeness of the data, data analyses, and the fidelity of this report to
the study protocol.
Declaration of Interests
SK, HC, HG-B, AS, JD, EH, and JJ and are employees of Sage
Therapeutics, Inc, with stock or stock options, or both. SK has a patent
pending (SAGE-547 for neuropsychiatric conditions). SR, CNE, and DR
and are consultants for Sage Therapeutics, Inc. RA is an employee of
Sage Therapeutics, Inc. CNE, KMD, and SM-B report that their
institutions are receiving grants for the conduct of the clinical trial from
Sage Therapeutics, Inc. CNE is a consultant for Asarina Pharma. DR will
receive stock options for being on the clinical advisory board of Sage
Therapeutics. Inc. SP is a founder, board member, and shareholder of
Sage Therapeutics, Inc. RR declares no competing interests.
Acknowledgments
This study was supported by Sage Therapeutics, Inc. We thank
Jeffrey R Skaar and Anna K Talaga at Boston Strategic Partners
(supported by Sage Therapeutics, Inc) and Paul Miller at Sage
Therapeutics, Inc, for editorial support.
References
1 Gavin NI, Gaynes BN, Lohr KN, Meltzer-Brody S, Gartlehner G,
Swinson T. Perinatal depression: a systematic review of prevalence
and incidence. Obstet Gynecol 2005; 106: 1071–83.
2 Howard LM, Molyneaux E, Dennis CL, Rochat T, Stein A, Milgrom J.
Non-psychotic mental disorders in the perinatal period. Lancet 2014;
384: 1775–88.
3 Fellmeth G, Fazel M, Plugge E. Migration and perinatal mental
health in women from low- and middle-income countries:
a systematic review and meta-analysis. BJOG 2017; 124: 742–52.
4 Meltzer-Brody S, Maegbaek ML, Medland SE, Miller WC,
Sullivan P, Munk-Olsen T. Obstetrical, pregnancy and
socio-economic predictors for new-onset severe postpartum
psychiatric disorders in primiparous women. Psychol Med 2017:
47: 1427–41.
5 Kim JJ, La Porte LM, Saleh MP, et al. Suicide risk among perinatal
women who report thoughts of self-harm on depression screens.
Obstet Gynecol 2015; 125: 885–93.
6 Postpartum Depression: Action Towards Causes and Treatment
(PACT) Consortium. Heterogeneity of postpartum depression:
a latent class analysis. Lancet Psychiatry 2015; 2: 59–67.
7 WHO. International statistical classification of diseases and related
health problems, 10th revision (ICD-10). Geneva: World Health
Organization, 1992.
8 American Psychiatric Association. Diagnostic and statistical manual
of mental disorders, 5th edn (DSM-5). Arlington, VA: American
Psychiatric Publishing, 2013.
www.thelancet.com Published online June 12, 2017 http://dx.doi.org/10.1016/S0140-6736(17)31264-3 9
Articles
9 Johannsen BM, Larsen JT, Laursen TM, Bergink V, Meltzer-Brody S,
Munk-Olsen T. All-cause mortality in women with severe postpartum
psychiatric disorders. Am J Psychiatry 2016; 173: 635–42.
10 Khalifeh H, Hunt IM, Appleby L, Howard LM. Suicide in perinatal
and non-perinatal women in contact with psychiatric services: 15 year
findings from a UK national inquiry. Lancet Psychiatry 2016;
3: 233–42.
11 Noorlander Y, Bergink V, van Den Berg MP. Perceived and observed
mother-child interaction at time of hospitalization and release in
postpartum depression and psychosis. Arch Womens Ment Health
2008; 11: 49–56.
12 Field T. Postpartum depression effects on early interactions,
parenting, and safety practices: a review. Infant Behav Dev 2010;
33: 1–6.
13 Epperson CN, Jatlow PI, Czarkowski K, Anderson GM.
Maternal fluoxetine treatment in the postpartum period: effects on
platelet serotonin and plasma drug levels in breastfeeding
mother-infant pairs. Pediatrics 2003; 112: e425.
14 Bloch M, Schmidt PJ, Danaceau M, Murphy J, Nieman L,
Rubinow DR. Effects of gonadal steroids in women with a history of
postpartum depression. Am J Psychiatry 2000; 157: 924–30.
15 Epperson CN, Gueorguieva R, Czarkowski KA, et al.
Preliminary evidence of reduced occipital GABA concentrations in
puerperal women: a 1H-MRS study. Psychopharmacology (Berl)
2006; 186: 425–33.
16 Schiller CE, Meltzer-Brody S, Rubinow DR. The role of reproductive
hormones in postpartum depression. CNS Spectr 2015; 20: 48–59.
17 Paul SM, Purdy RH. Neuroactive steroids. FASEB J 1992; 6: 2311–22.
18 Majewska MD, Harrison NL, Schwartz RD, Barker JL, Paul SM.
Steroid hormone metabolites are barbiturate-like modulators of the
GABA receptor. Science 1986; 232: 1004–07.
19 Deligiannidis KM, Kroll-Desrosiers AR, Mo S, et al.
Peripartum neuroactive steroid and γ-aminobutyric acid profiles in
women at-risk for postpartum depression. Psychoneuroendocrinology
2016; 70: 98–107.
20 Zorumski CF, Paul SM, Izumi Y, Covey DF, Mennerick S.
Neurosteroids, stress and depression: potential therapeutic
opportunities. Neurosci Biobehav Rev 2013; 37: 109–22.
21 Belelli D, Harrison NL, Maguire J, Macdonald RL, Walker MC,
Cope DW. Extrasynaptic GABAA receptors: form, pharmacology,
and function. J Neurosci 2009; 29: 12757–63.
22 Luisi S, Petraglia F, Benedetto C, et al. Serum allopregnanolone levels
in pregnant women: changes during pregnancy, at delivery, and in
hypertensive patients. J Clin Endocrinol Metab 2000; 85: 2429–33.
23 Nappi RE, Petraglia F, Luisi S, Polatti F, Farina C, Genazzani AR.
Serum allopregnanolone in women with postpartum “blues”.
Obstet Gynecol 2001; 97: 77–80.
24 Maguire J, Mody I. GABA(A)R plasticity during pregnancy:
relevance to postpartum depression. Neuron 2008; 59: 207–13.
25 Paoletti AM, Romagnino S, Contu R, et al. Observational study on
the stability of the psychological status during normal pregnancy
and increased blood levels of neuroactive steroids with GABA-A
receptor agonist activity. Psychoneuroendocrinology 2006; 31: 485–92.
26 Deligiannidis KM, Sikoglu EM, Shaffer SA, et al.
GABAergic neuroactive steroids and resting-state functional
connectivity in postpartum depression: a preliminary study.
J Psychiatr Res 2013; 47: 816–28.
27 Harris B, Lovett L, Smith J, Read G, Walker R, Newcombe R.
Cardiff puerperal mood and hormone study. III. Postnatal
depression at 5 to 6 weeks postpartum, and its hormonal correlates
across the peripartum period. Br J Psychiatry 1996; 168: 739–44.
28 Milgrom J, Gemmil AW, Bilszta JL, et al. Antenatal risk factors for
postnatal depression: a large prospective study. J Affect Disord 2008;
108: 147–57.
29 Serati M, Redaelli M, Buoli M, Altamura AC. Perinatal major
depression biomarkers: a systematic review. J Affect Disord 2016;
193: 391–404.
30 First MB, Spitzer RL, Gibbon M, Williams JBW. Structured clinical
interview for DSM-IV axis I disorders, clinician version (SCID-CV).
Washington, DC: American Psychiatric Press, 1996.
31 Zimmerman M, Martinez JH, Young D, Chelminski I, Dalrymple K.
Severity classification on the Hamilton Depression Rating Scale.
J Affect Disord 2013; 150: 384–88.
 Articles
32 Kanes SJ, Colquhoun H, Doherty J, et al. Open-label, proof-of-
concept study of brexanolone in the treatment of severe postpartum
depression. Hum Psychopharmacol Clin Exp 2017; 32: e2576.
33 Bech P, Allerup P, Gram LF, et al. The Hamilton depression scale.
Evaluation of objectivity using logistic models. Acta Psychiatr Scand
1981; 63: 290–99.
34 Montgomery SA, Asberg M. A new depression scale designed to be
sensitive to change. Br J Psychiatry 1979; 134: 382–89.
35 Guy W. ECDEU assessment manual for psychopharmacology:
Rockville, MD: US Department of Health, Education, and Welfare,
Public Health Service, Alcohol, Drug Abuse, and Mental Health
Administration, National Institute of Mental Health,
Psychopharmacology Research Branch, Division of Extramural
Research Programs, 1976.
36 Spitzer RL, Kroenke K, Williams JB, Lowe B. A brief measure for
assessing generalized anxiety disorder: the GAD-7. Arch Intern Med
2006; 166: 1092–97.
37 Cox JL, Holden JM, Sagovsky R. Detection of postnatal depression.
Development of the 10-item Edinburgh Postnatal Depression Scale.
Br J Psychiatry 1987; 150: 782–86.
38 Spitzer RL, Kroenke K, Williams JB. Validation and utility of a
self-report version of PRIME-MD: the PHQ primary care study.
Primary Care Evaluation of Mental Disorders. Patient Health
Questionnaire. JAMA 1999; 282: 1737–44.
39 Barkin JL, Wisner KL, Wisniewski SR. The psychometric properties
of the Barkin Index of Maternal Functioning.
J Obstet Gynecol Neonatal Nurs 2014; 43: 792–802.
40 Posner KBG, Stanley B, Brent DA, et al. The Columbia-Suicide
Severity Rating Scale: initial validity and internal consistency
findings from three multisite studies with adolescents and adults.
Am J Psychiatry 2011; 168: 1266–77.
41 Hoddes E, Zarcone V, Smythe H, Phillips R, Dement WC.
Quantification of sleepiness: a new approach. Psychophysiology 1973;
10: 431–36.
42 Yonkers KA, Wisner KL, Stewart DE, et al. The management of
depression during pregnancy: a report from the American
Psychiatric Association and the American College of Obstetricians
and Gynecologists. Obstet Gynecol 2009; 114: 703–13.
43 Hantsoo L, Ward-O’Brien D, Czarkowski KA, Gueorguieva R,
Price LH, Epperson CN. A randomized, placebo-controlled,
double-blind trial of sertraline for postpartum depression.
Psychopharmacology (Berl) 2014; 231: 939–48.
10 www.thelancet.com Published online June 12, 2017 http://dx.doi.org/10.1016/S0140-6736(17)31264-3
44 De Crescenzo F, Perelli F, Armando M, Vicari S. Selective serotonin
reuptake inhibitors (SSRIs) for post-partum depression (PPD):
a systematic review of randomized clinical trials. J Affect Disord 2014;
152–54: 39–44.
45 Hendrick V, Altshuler L, Strouse T, Grosser S. Postpartum and
nonpostpartum depression: differences in presentation and response
to pharmacologic treatment. Depress Anxiety 2000; 11: 66–72.
46 Epperson CN, Anderson GM, McDougle CJ. Sertraline and
breast-feeding. N Engl J Med 1997; 336: 1189–90.
47 Cox EQ, Sowa NA, Meltzer-Brody SE, Gaynes BN. The perinatal
depression treatment cascade: baby steps toward improving
outcomes. J Clin Psychiatry 2016; 77: 1189–200.
48 Fournier JC, DeRubeis RJ, Hollon SD, et al. Antidepressant drug
effects and depression severity: a patient-level meta-analysis. JAMA
2010; 303: 47–53.
49 Bech P, Cialdella P, Haugh MC, et al. Meta-analysis of randomised
controlled trials of fluoxetine v. placebo and tricyclic antidepressants
in the short-term treatment of major depression. Br J Psychiatry
2000; 176: 421–28.
50 Murrough JW, Iosifescu DV, Chang LC, et al. Antidepressant efficacy
of ketamine in treatment-resistant major depression: a two-site
randomized controlled trial. Am J Psychiatry 2013; 170: 1134–42.
51 Zarate C, Singh JB, Carlson PJ, et al. A randomized trial of an
N-methyl-D-aspartate antagonist in treatment-resistant major
depression. Arch Gen Psychiatry 2006; 63: 856–64.
52 Concas A, Mostallino MC, Porcu P, et al. Role of brain
allopregnanolone in the plasticity of γ-aminobutyric acid type A
receptor in rat brain during pregnancy and after delivery.
Proc Natl Acad Sci USA 1998; 95: 13 284–89.
53 Klak J, Hill M, Pařízek A, et al. Pregnanolone isomers, pregnenolone
and their polar conjugates around parturition. Physiol Res 2003;
52: 211–21.
54 van Calker D, Zobel I, Dykierek P, et al. Time course of response to
antidepressants: predictive value of early improvement and effect of
additional psychotherapy. J Affect Disord 2009; 114: 243–53.