Effects of Open-Label, Adjunctive Ganaxolone on Persistent Depression Despite Adequate Antidepressant Treatment in Postmenopausal Women: A Pilot Study

Abstract

Objective:

The neuroactive steroid metabolite of progesterone, allopregnanolone, is a positive allosteric modulator of GABA_A receptors and a putative treatment for mood disorders. We performed a pilot study to determine whether an oral allopregnanolone analog (ganaxolone) may be effective adjunctive therapy for persistent depression despite adequate antidepressant treatment in postmenopausal women.

Methods:

Ten post-menopausal women (age 62.8±6.3 years, range 53-69) with persistent depression despite adequate antidepressant treatment [current DSM-IV major depressive episode per the Structured Clinical Interview for DSM-IV (SCID), Montgomery-Asberg Depression Rating Scale (MADRS) ≥16, and treated with an adequately dosed antidepressant for ≥6 weeks] were studied from 12/2016 to 4/2018. Open-label ganaxolone (225 mg BID, increased to 450 mg BID if tolerated) was administered for 8 weeks, followed by a 2-week taper.

Results:

Mean total MADRS score (primary endpoint) decreased by 8 weeks [24.4±1.6 (SEM) to 12.8±2.9, p=0.015] and persisted over the two-week taper (p=0.019); 44% of subjects experienced response (score decrease ≥50%) and remission (final score <10), which persisted in 100% and 50% of subjects at 10 weeks, respectively. Secondary endpoints showed significant improvement, including the Inventory of Depressive Symptomatology-Self-Report (IDS-SR; p=0.003), MADRS Reduced Sleep subscale (p<0.001), Symptoms of Depression Questionnaire (SDQ) total score (p=0.012), and SDQ subscales for disruptions in sleep quality (p=0.003) and changes in appetite and weight (p=0.009) over 8 weeks. No significant effects were observed on quality-of-life or sexual function. All subjects experienced sleepiness and fatigue; 60% experienced dizziness.

Conclusion:

In this open-label, uncontrolled pilot study, adjunctive ganaxolone appears to exert antidepressant effects but produces sedation with twice-daily dosing. Ganaxolone may also improve sleep, which may be useful in patients with depression and insomnia.

INTRODUCTION

Resistance to selective serotonin reuptake inhibitor (SSRI) and serotonin norepinephrine reuptake inhibitor (SNRI) treatment occurs in about 50% to 70% of patients with Major Depressive Disorder (MDD), a condition associated with significant morbidity¹ and affecting women at higher rates than men.² Few well-tolerated, effective augmentation therapies are available for such patients.³ Therefore, new therapeutic strategies for persistent depression despite adequate antidepressant treatment are needed.

Ganaxolone (Marinus Pharmaceuticals), an allopregnanolone analog with similar biological activity,⁴ is a candidate for such a therapy. Allopregnanolone is a neuroactive steroid metabolite of progesterone and positive allosteric modulator of GABA_A inhibitory brain receptors, acting with 10-times the potency than benzodiazepines at these receptors.^5–7 Progesterone is converted to allopregnanolone by the enzymes 5α-reductase and 3α-hydroxysteroid dehydrogenase (3α-HSD). The addition of a methyl group in the 3β position of ganaxolone prevents back conversion to progesterone and extends its effect. Additionally, ganaxolone can be administered orally in the outpatient setting, while allopregnanolone itself must be administered by continuous intravenous infusion.

The rationale for studying the effects of a neuroactive steroid analog in women with persistent depression despite adequate antidepressant treatment has its basis in cross-sectional and observational studies. Our group has demonstrated that serum levels of allopregnanolone, but not its precursor progesterone, are inversely associated with depression symptom severity in a group of female volunteers with a wide range of weight (from anorexia nervosa to obesity) without MDD.⁸ Other groups have demonstrated inverse associations between allopregnanolone cerebrospinal fluid (CSF) and serum levels and severity of depression symptoms in patients with MDD, as well as lower allopregnanolone levels in depressed subjects compared with non-depressed controls.^9,10,11 Observational prospective studies also support a role for allopregnanolone in depression, and in particular, the data suggest that an increase in allopregnanolone levels (both CSF and serum) occur in patients after administration of SSRIs and other antidepressants.^9,11–13 The rationale for studying the effects of ganaxolone specifically in postmenopausal women is that progesterone, the precursor of allopregnanolone, is very low in this demographic; thus postmenopausal women likely are in a state of relative deficiency of this neuroactive steroid.

Finally, recent data demonstrating the positive effects of allopregnanolone administration in postpartum depression provide further evidence of the antidepressant potential of allopregnanolone.^14,15 In a double-blind, randomized, placebo-controlled trial in 21 women with severe postpartum depression,¹⁴ those who received IV brexanolone experienced a very substantial reduction in depression symptom severity compared to the placebo group.¹⁴ This finding was subsequently confirmed by two larger phase 3 randomized, double-blinded trials of the same formulation of allopregnanolone.¹⁶ Short-term studies of SAGE-217, an oral positive allosteric modulator of GABA_A receptors and a neuroactivesteroid as well, suggest potential efficacy in the treatment of moderate-to-severe major depressive disorder.^17,18

We hypothesized that oral ganaxolone would be an effective augmentation therapy for postmenopausal women with persistent depression, and we report here the results of a pilot study investigating this hypothesis. We also investigated the effects of ganaxolone on symptoms associated with depression and/or its treatment, including sexual function, fatigue and sleep disturbance. Finally, we hypothesized that subjects with lower allopregnanolone levels at baseline would demonstrate greater improvement with ganaxolone treatment.

METHODS

Participants

The protocol was approved by the Partners Human Research Committee and conducted from 12/216 to 4/2018. Written informed consent was obtained from all participants prior to any procedures being performed. Inclusion criteria were: female, age 50–75, postmenopausal, MDD with a current major depressive episode by DSM-IV criteria by the Structured Clinical Interview for DSM-IV (SCID),¹⁹ Montgomery-Asberg Depression Rating Scale (MADRS)²⁰ score ≥16, and currently treated with an antidepressant taken at an adequate dose for at least 6 weeks. The MGH Antidepressant Treatment Response Questionnaire (ATRQ)^21,22 was used to confirm source documentation of adequate antidepressant treatment dose and duration to fulfill trial inclusion criteria.²¹ Exclusion criteria were: serious suicide or homicide risk, current or history of psychotic features, or substance abuse disorder active within the last 6 months.

Design

The study schema is presented in Table 1.

Table 1.

Ganaxolone dosing plan and schedule of assessments.

	Baseline	Week 1	Weeks 2, 4, 6	Weeks 8, 10	Week 22a
Ganaxolone dosing plan	Start 225 mg BID	Dose Increase 450 mg BID	Continue 450 mg BID	Drug Taper	3 Months Post Treatment
Depression symptom severity
MADRS (primary endpoint)	x		x	x
CGI-S	x		x	x
IDS-SR	x		x	x	x
SDQ	x		x	x	x
Fatigue/sleepiness
BFI	x		x	x	x
FSS	x		x	x
ESS	x		x	x
Quality of life (SF-36)	x		x	x
Sexual function (DISF)	x		x	x
Cognitive function (CPFQ)	x		x	x
Suicidality (CHRT)	x	x	x	x

^aVisit occurred remotely; only self-administered questionnaires were performed. Abbreviations: MADRS, Montgomery-Asberg Depression Rating Scale; CGI-S, Clinical Global Impressions; IDS-SR, Inventory of Depressive Symptomatology-Self-Report; SDQ, Symptoms of Depression Questionnaire; SF-36, 36-Item Short Form Health Survey; BFI, Brief Fatigue Inventory; FSS, Fatigue Severity Scale; ESS, Epw orth Sleepiness Scale; DISF, Derogatis Interview for Sexual Function; CPFQ, Cognitive and Physical Functioning Questionnaire; CHRT, Concise Health Risk Tracking scale.

This was an open-label pilot study in which 10 post-menopausal women with persistent depression despite adequate antidepressant treatment were administered ganaxolone over an 8-week treatment period followed by a 2-week taper. Psychiatric measures were assessed at baseline, at weeks 2, 4, 6 and 8 of ganaxolone treatment, and immediately after the drug taper at 10 weeks. Self-rated questionnaires only were administered remotely at 22 weeks, which was 3 months after discontinuation of the study drug.

Psychiatric Measures

The study primary endpoint was the change in MADRS over the eight week treatment period.^23,24 The following measures were also used as secondary measures of depression symptom severity: Inventory of Depressive Symptomatology-Self-Report (IDS-SR)^25–27 and Clinical Global Impressions (CGI-S).²⁸ The Symptoms of Depression Questionnaire (SDQ)²⁹ was designed to more fully capture the heterogeneity of symptom presentations of depressive disorders and includes five factors focusing on the following dimensions: 1) lassitude, mood, and cognitive functioning; 2) anxiety, agitation, irritability and anger; 3) suicidal ideation; 4) disruptions in sleep quality; 5) changes in appetite and weight. Quality of life was measured using the 36-Item Short Form Health Survey (SF-36), a self-administered, validated, and widely used quality-of-life questionnaire.³⁰ Effects of drug on fatigue severity were measured using the Brief Fatigue Inventory (BFI)³¹ and Fatigue Severity Scale (FSS),³² and sleepiness was measured using the Epworth Sleepiness Scale (ESS).³³ Sexual function was assessed with the Derogatis Interview for Sexual Function – Female Version (DISF) total score.^34–36 The Cognitive and Physical Functioning Questionnaire (CPFQ), designed to assess cognitive function in mood and anxiety disorders, with a focus on executive dysfunction, was also administered.³⁷ The presence of suicidal thoughts was assessed using the Concise Health Risk Tracking scale (CHRT).³⁸

Laboratory Measures

Allopregnanolone, allopregnanolone-sulfate, pregnenolone and pregnenolone-sulfate were quantified at baseline and 8 weeks in human K2EDTA plasma using Turbo Ionspray LC-MS/MS (Keystone Bioanalytical Inc.) with the respective deuterated neurosteroids as internal standards. The lower limit of quantification was 25 pg/mL for allopregnanolone and pregnenolone, 2.5 ng/mL for pregnenolone sulfate and 250 pg/mL for allopregnanolone-sulfate.

Statistical Analysis

Paired T-tests were performed between baseline (pre-treatment) and 8-week time points for all endpoints, as pre-specified, and two-tailed p values are reported. If the two-tailed p value was < 0.05, then paired T-tests were also performed between baseline and 10 weeks and baseline and 3 months to assess whether the treatment effects were durable during the 2-week taper (8-10 weeks) and at the 3-month follow-up assessment (week 22), respectively. Neuroactive steroid variables were log transformed prior to analysis and Pearson correlation coefficients were reported.

RESULTS

Pre-treatment Clinical Characteristics

The mean age of the participants was 62.8 ± 6.3 (SD) years (range 53–69), and the mean BMI 26.2 ± 4.0 kg/m² (range 21.6 – 33.8). The mean time since menopause was 12.4 ± 7.5 years (range 3 – 30 years). Three participants (30%) reported hot flashes over the past month, only one of whom reported hot flashes that interfered with sleep. Two participants reported intermittent use of vaginal estrogen, but no participant was taking systemic estrogen. Six participants were on an SSRI only (60%), three were on an SNRI only (30%) and one was on an SSRI and bupropion. Duration of antidepressant use was 7 weeks for 2 participants and at least 10 weeks for 8 participants, with four of the latter group reporting a stable dose for 1 or more years. The mean MADRS Total Score was 24.4 ± 5.1 (range 18–34), indicating moderate depression.³⁹

Psychiatric Measures

All psychiatric measures are reported in Table 2.

Table 2.

Depression symptom severity.^a,b

	Treatment Period		Drug Taper	3 Months Post Treatment	Baseline vs Week 8 P-value
	Baseline	Week 8	Week 10c	Week 22c,d
Clinician-administered questionnaires
Total MADRS score	24.4 ± 1.6	12.8 ± 2.9*	14.4 ± 2.5*		p =0.015
Apparent sadness	2.4 ± 0.2	1.7 ± 0.4	1.4 ± 0.3		p =0.11
Reported sadness	2.9 ± 0.2	2.2 ± 0.5	1.8 ± 0.3		p =0.35
Inner tension	3.0 ± 0.4	1.9 ± 0.6	1.9 ± 0.5		p =0.11
Reduced sleep	3.1 ± 0.3	0.7 ± 0.4*	1.3 ± 0.4*		p <0.001
Reduced appetite	0.7 ± 0.4	0.1 ± 0.1	0.1 ± 0.1		p =0.17
Concentration difficulties	3.0 ± 0.4	1.9 ± 0.6†	2.2 ± 0.6		p =0.067
Lassitude	3.0 ± 0.4	1.3 ± 0.4*	2.0 ± 0.4†		p =0.024
Inability to feel	2.9 ± 0.5	0.9 ± 0.3*	1.7 ± 0.3		p =0.028
Pessimistic thoughts	2.6 ± 0.3	1.9 ± 0.5	1.6 ± 0.5		p =0.30
Suicidal thoughts	0.8 ± 0.3	0.2 ± 0.2	0.4 ± 0.3		p =0.31
CGI-S score	3.8 ± 0.1	2.4 ± 0.4*	2.9 ± 0.4†		p =0.004
Self-rated questionnaires
IDS-SR Score	29.7 ± 2.1	17.1 ± 2.9*	19.5 ± 2.2*	15.7 ± 3.1*	p =0.003
Total SDQ Score	136.4 ± 5.8	108.5 ± 8.0*	109.5 ± 7.4*	104.7 ± 6.8*	p =0.012
Factor 1 (lassitude, mood and cognitive functioning)	55.4 ± 3.3	44.6 ± 4.3†	44.6 ± 4.6	38.7 ± 4.6	p =0.064
Factor 2 (anxiety, agitation, irritability and anger)	41.2 ± 1.3	31.1 ± 3.0*	31.0 ± 1.9*	32.0 ± 1.8*	p =0.012
Factor 3 (suicidal ideation)	16.1 ± 0.8	13.0 ± 1.1*	13.0 ± 1.0*	12.3 ± 1.1*	p =0.041
Factor 4 (disruptions in sleep quality)	9.5 ± 0.9	5.1 ± 0.6*	6.7 ± 0.6*	6.7 ± 0.7†	p =0.003
Factor 5 (changes in appetite and weight)	9.5 ± 0.4	8.3 ± 0.2*	8.7 ± 0.2	9.0 ± 0.5	p =0.009

^*indicates p <0.05 for change from baseline.

^†Indicates trend (0.05<p < 0.1) for change from baseline.

^aFor all questionnaires, higher values indicate worse symptom severity.

^bAll data are reported as mean ± SEM.

^cPaired T-tests were only performed between baseline and the 10- and 22-w eek follow -up time points if a significant difference was found during the treatment period (baseline and 8 weeks).

^dClinician-administered questionnaires were not performed at week 22. Abbreviations: MADRS, Montgomery-Asberg Depression Rating Scale; CGI-S, Clinical Global Impressions; IDS-SR, Inventory of Depressive Symptomatology-Self-Report; SDQ, Symptoms of Depression Questionnaire.

Primary Endpoint

The primary endpoint was change in total MADRS score over the 8-week main-study period. Nine out of ten subjects completed the full 8-week treatment period. Forty-four percent of subjects (4/9) experienced both response to treatment (>50% reduction from baseline in total MADRS) and remission (final MADRS score of <10). Thus, every subject who responded also achieved remission. Responders/remitters did not differ with respect to use of SSRIs vs SNRIs, menopausal stage, presence of hot flashes, or current use of vaginal estrogen. The mean total MADRS score (primary endpoint) decreased significantly between pre-treatment (week 0) and 8 weeks [24.4 ± 1.6 (SEM) to 12.8 ± 2.9, p=0.015] (Figure 1A). MADRS subscales demonstrating significant reductions between pre-treatment and 8 weeks were Reduced Sleep (p<0.001) (Figure 1B), Lassitude (p=0.024) and Inability to Feel (p=0.028). There was no significant worsening of any subscales.

Figure 1.

A) There was a reduction in depression symptom severity, as measured by mean Total MADRS score, that remained reduced through a 2-week taper period. B) There was an improvement in sleep quality as demonstrated by a reduction in the mean MADRS Reduced Sleep Subscale score, that remained reduced through a 2-week taper period. *Indicates significant change compared to baseline with p-values noted. Error bars indicate SEM.

Secondary Endpoints

There was no significant worsening of any secondary endpoint scale or subscale.

Depression Symptom Severity:

There were significant improvements in the IDS-SR (Supplementary Figure), CGI-S and total SDQ (Figure 2A) scores. There were also significant improvements in 4 of the 5 SDQ subscales: Factor 2 (anxiety, agitation, irritability and anger), Factor 3 (suicidal ideation), Factor 4 (disruptions in sleep quality), and Factor 5 (changes in appetite and weight) (Figure 2C–F). There was a trend toward a significant improvement in Factor 1 (lassitude, mood and cognitive and social functioning) (Figure 2B).

Figure 2.

There were improvements in the SDQ Total Score (Panel A) and multiple SDQ Factors (Panels B-F) during the treatment period (Baseline-Week 8), some of which showed durable improvement after a 2-week drug taper (Week 10) and at three months off of drug treatment (Week 22). *Indicates significant difference and ^†indicates trend in difference compared to baseline value. Error bars indicate SEM.

Quality of Life:

There was a trend toward improvement with ganaxolone therapy over 8 weeks in the SF-36 Mental Health Component Summary Scale with no notable improvement in the Physical Component Summary Scale.

Fatigue, Sexual Function, and Cognitive Function:

There were no significant changes in any measures of fatigue (ESS, FSS and the BFI), sexual function (DISF) or self-reported cognitive function (CPFQ).

Post-Treatment Data

The effect of ganaxolone on depression symptom severity, as measured by the total MADRS score (primary endpoint), was retained at 10 weeks, after a 2-week drug taper (Figure 1A). All subjects who experienced a response at 8 weeks had retention of that response through 10 weeks. Half of subjects with remission at 8 weeks had retention of that remission at 10 weeks.

Significant improvement in the MADRS Reduced Sleep Subscale was durable through the 2-week taper (Figure 2B), while significant positive effects on the other two MADRS subscales, including Lassitude and Inability to Feel, were maintained as a trend through the 2-week taper.

The significant effect on depression symptom severity as measured by the secondary endpoint measure, IDS-SR, was also maintained after the 2-week drug taper (Supplementary Figure). The response persisted 3 months after study drug discontinuation. The significant positive effects on the total SDQ score, SDQ Factor 2 (Anxiety, Agitation, Irritability and Anger), SDQ Factor 3 (Suicidal Ideation) and SDQ Factor 4 (Disruptions in Sleep Quality) were also maintained off drug at 10 weeks (Figure 2A, 2C–E). Of these, the total SDQ score, SDQ Factor 2 (Anxiety, Agitation, Irritability and Anger) and SDQ Factor 3 (Suicidal Ideation) continued to showed improvement three months after drug discontinuation with a trend toward continued improvement in SDQ Factor 4 (Disruptions in Sleep Quality) (Figure 2A, 2C–E).

Neuroactive Steroid Levels

Baseline allopregnanolone levels were below the limit of quantification (<25 pg/mL) in the majority of subjects (6/10). Allopregnanolone-sulfate, pregnenolone, and pregnenolone-sulfate were quantifiable in all subjects (10/10). Allopregnanolone-sulfate levels (R=0.82, p=0.004) (Figure 3) were positively associated with greater depression severity by MADRS at baseline. Pregnenolone and pregnenolone-sulfate levels were not associated with baseline MADRS scores.

Figure 3.

Allopregnanolone-sulfate levels (R=0.82, p=0.004) were positively associated with greater depression severity by total MADRS at baseline.

Neither baseline levels or change in levels of allopregnanolone-sulfate, pregnenolone, or pregnenolone-sulfate over the 8-week treatment period predicted response to ganaxolone treatment. There were no differences in mean baseline neuroactive steroid levels between the responder and remitters (n=4) versus those with no response or remission of depressive symptoms. Allopregnanolone levels were undetectable in equal frequency across groups regardless of response (undetectable in n=3/4 responder/remitters and n=3/5 non-responder/non-remitters).

Drop-outs and Adverse Events

Nine out of ten study subjects completed the 10-week protocol. All ten reported adverse events that were possibly or probably related adverse events, as follows. All subjects experienced sleepiness and fatigue, and six of ten subjects experienced dizziness. One subject was discontinued after experiencing somnolence and dizziness after one dose of ganaxolone (225 mg). Four of the ten participants were able to tolerate dose increases (450 BID), three subjects tolerated the initial dose (225 mg BID) but were unable to tolerate attempted dose increases (450 BID), and two subjects did not tolerate the initial 225 mg BID dose well enough to have their doses increased. There was no change in CHRT score, an assessment of suicidal thoughts and behaviors. Mean transaminases (AST and ALT) did not increase over the 8-week study.

Discussion

This pilot study is the first to explore the effects of oral ganaxolone, an allopregnanolone analog and neuroactive steroid, for persistent depression despite adequate antidepressant treatment in postmenopausal women. Our data suggest that ganaxolone exerts antidepressant effects in a subset of such patients. Although all participants reported daytime sleepiness at the current dosing schedule, we also observed significant improvements in sleep during the treatment period, which was likely related. In addition, our data suggest possible positive effects on anxiety, agitation, irritability and anger and suicidal ideation with ganaxolone treatment, as well as persistence of treatment effects beyond the period of drug administration.

Our remission rate of 44% (4/9 subjects) appears greater than that of the STAR*D Augmentation Trial that demonstrated a remission rate of approximately 30% (by Hamilton Depression Rating Scale <7) in citalopram-treated subjects who were augmented using sustained release bupropion or buspirone.⁴⁰ There are sparse data on the treatment of persistent depression despite adequate antidepressant treatment specifically in postmenopausal women. Previous studies in peri- and postmenopausal women with untreated depression report wide-ranging remission rates of 38–88% with conventional antidepressants.^41–43 However, we have not identified any other studies of augmentation therapy in postmenopausal women with depression resistant to standard therapies. Therefore, it is not possible to determine how our remission rate compares with that of other adjuvant therapies specifically in postmenopausal women.

The results of two recent studies of SAGE-217, another positive GABA_A allosteric oral synthetic neurosteroid, have recently been reported in treatment of patients with major depressive disorder.^17,18 The medication was shown to be effective in a randomized, placebo-controlled trial in 89 men and women with moderate-to-severe depression (MADRS≥32 and HAM-D≥22) for 14 days reported recently in the New England Journal of Medicine.¹⁷ A second, larger study (n=323) of SAGE-217 did not demonstrate a greater reduction in depression symptom severity compared to placebo at 15 days.¹⁸ However, these results, which were negatively affected by a high placebo response, became significant in post-hoc analyses when subjects in the SAGE-217 group with no measurable drug concentrations were excluded or when only including patients with higher depression symptom severity at baseline. These studies differed from ours in that they tested a different medication--though with a similar mechanism of action--and they did not specifically target our study population of postmenopausal women with moderately severe persistent depression (MADRS≥16) despite adequate antidepressant treatment. Nevertheless, these studies highlight the importance of the non-specific, placebo-like effects in interpreting depression study results, particularly in patients experiencing less severe depression severity, who may be more susceptible to these effects.^44,45 Placebo-controlled studies of ganaxalone are therefore warranted to confirm the findings we report here.

The signal detected that ganaxolone may improve sleep is particularly important, given that 75% of patients with depression struggle with insomnia.⁴⁶ The durability of antidepressant effects over the 2-week taper period suggests that administering the entire dose before bedtime might provide assistance with sleep, followed by sustained daytime antidepressant effects without the significant somnolence observed with morning dosing. Thus, our study suggests that perhaps lower doses with fewer side effects and evening dosing may be particularly efficacious, taking advantage of these moderately sedating effects for improved sleep, though further studies would be necessary to determine if this is the case. This is consistent with the results from the Amaryllis study, which compared oral ganaxolone at a low dose (675 mg at dinner) and high dose (initiated at 675 mg at dinner and bedtime for two days followed by a transition to 1125 mg once daily at dinner) in women with postpartum depression. The higher dose regimen was well tolerated and showed a HAM-D-17 reduction that was greater than the low dose regimen. Thus, these data support use of the once daily 1125 mg oral evening dose in future clinical studies.⁴⁷

Our findings also suggested that improvements in depression, anxiety, agitation, irritability and anger, suicidal ideation, sleep quality and overall disease severity may be sustainable with drug taper and even three months after drug discontinuation. Part 1 of the Magnolia study (Marinus) showed a similar durable response at one month after IV ganaxolone treatment (48-hour infusion followed by 12-hour taper) in women with postpartum depression. Part 2 of the Magnolia study showed an immediate response to combined IV and oral ganaxolone treatment (6-hour infusion followed by one oral dose) in the same population but failed to demonstrate a durable response at 1-month post-treatment. Thus, these data raise the possibility of sustained antidepressant effects with ganaxolone administration, but further studies are needed to confirm this finding and determine the mechanism of response durability.

Our data demonstrate that postmenopausal women with persistent depression despite adequate antidepressant treatment have largely undetectable to low levels of allopregnanolone, consistent with our original rationale for investigating allopregnanolone analog administration in such women. However, we also found that allopregnanolone-sulfate, which was detectable in all subjects, was positively associated with depression severity by MADRS prior to ganaxolone treatment. The precise relationship of allopregnanolone-sulfate to allopregnanolone levels is not entirely understood, and we are limited in drawing conclusions regarding this relationship between allopregnanolone and allopregnanolone sulfate because of the number of undetectable allopregnanolone levels in our cohort. Of note, these data should be interpreted in the context of the fact that all study participants were receiving antidepressant medications upon entry to the study, and prior data suggest that antidepressants may increase allopregnanolone levels.^9,11,12 Despite these limitations, these data provide a plausible mechanistic role for the direct association between allopregnanolone-sulfate levels and depression severity in this cohort and further raises the scientific question of whether neuroactive steroid dysregulation could be a mechanism underlying persistent depression despite adequate antidepressant treatment.

Limitations of the current study include its uncontrolled design. As such, the signal observed in this study provides preliminary data for future more definitive studies in this area. In addition, we were unable to assess neuroactive steroid levels in non-depressed controls, nor were we able to reliably measure additional neuroactive steroids in the pathway to help identify possible enzymatic regulatory steps. Finally, we do not have data regarding prior history of hormone-related depression, which may be useful in this population. Further studies are warranted to determine whether neuroactive steroids play a direct role in the pathophysiology of persistent depression despite adequate antidepressant treatment.

In summary, our pilot data suggest that oral ganaxolone, an analog of allopregnanolone, may be a promising pharmacotherapy for postmenopausal women with persistent depression despite adequate antidepressant treatment and raise the question of whether neuroactive steroid dysregulation may contribute to depressive symptoms in this population. Furthermore, definitive studies in these areas are warranted, and the observed positive effects on sleep and the potential for sustained treatment effects also merit further study. Randomized, placebo-controlled studies are necessary to rule out placebo effects for both mood and sleep. Given the sedation experienced by a majority of participants and promising durability of antidepressant effects over the two-week period, bedtime dosing only should be considered for future studies. Finally, should rigorous studies confirm an antidepressant effect, it will be important to identify subsets of women who respond – for example, women with neuroactive steroid dysregulation – and to identify specific mechanisms of action, including regional brain targets.

Supplementary Material

Supplementary Figure

Supplementary Figure. Total IDS-SR score for all subjects at each time point. Each unique symbol represents a single subject.

Clinical Points.

• The neuroactive steroid allopregnanolone is a positive modulator of GABA_A receptors and a putative treatment for mood disorders. However, it has not been studied as augmentation therapy in postmenopausal women with persistent depression despite adequate antidepressant treatment.

• Oral allopregnanolone as augmentation therapy to traditional antidepressants led to improvement in mood in this open-label pilot study of postmenopausal women with persistent depression despite adequate antidepressant treatment.

• Additional randomized, double-blind, placebo-controlled studies are warranted to assess the antidepressant effects of oral allopregnanolone augmentation therapy in this population.